reference for chapter 3

8/15/2019 Reference for Chapter 3

http://slidepdf.com/reader/full/reference-for-chapter-3 1/897

Siemens Solar Basic PV Technology Course Fundamentals – Introduction

Copyright © 1998 Siemens Solar Industries

1-1

Chapter OneIntroduction

Welcome to the World ofPhotovoltaicsThank you for your interest in photovoltaic technology and system design, andwelcome to the world of solar electricity! Siemens Solar is proud to be recognizedas the world-wide leader in solar electric power generation, and we wish to supportyou in your efforts to apply photovoltaic technology to solve electrical power

problems, whether you are an individual designing your own system, or a systemdesigner working on large applications.

Purpose of This ManualThis course material is designed to be a self-teaching and technical resource forpeople interested in learning about how to design stand-alone photovoltaic powersystems. Although the focus is on professionals that will be designing systems forclients, individuals interested in designing their own systems will also benefit fromthis program. This program is used by professionals from a wide variety of fieldsincluding industrial technology, business, finance, military, construction, real estate,and education. The topics are presented in small steps, with exercises throughout.

The main desired outcome is an ability to successfully deal with the many aspects ofphotovoltaic system design – not only the computational skills needed to arrive atthe array and battery size, but also the judgmental skills needed to see when andwhere photovoltaics can be a viable solution to power needs, and the integratingskills required to specify appropriate electrical and mechanical components from avariety of manufacturers.

For those people preparing to attend the Comprehensive Seminar offered by

Siemens Solar, please answer all the exercises as you proceed through the text.Take the time to think about the questions, and how the information applies toproblems you may face. Upon completion of the text, send the answers to theexercises and the seminar application form to Siemens Solar to become eligible toattend the Comprehensive Seminar.





1-2

How to Proceed Through TheseMaterials

1) Browse:

Browse through the workbooks to become familiar with the contents and the style ofpresentation.

2) Sequence:

Choose to begin with topics of interest, or proceed sequentially through all thechapters, beginning with the introductory/survey chapters, and continuing with thetechnical material.

3) View Video:

Begin work in a new chapter by watching the videotaped presentation with theworkbook open. Follow along and examine the graphics as they are discussed.

4) Read Workbook:

Periodically the video presentation asks you to “Stop the tape and refer to yourworkbook”. Take time to read the pages up to the pause, and review in greaterdetail the key ideas discussed in the video. Underline or highlight important phrasesto help reinforce your learning.

5) Do Exercises:

If the section has exercises, first scan all the questions to get a sense of the scopeof detail needed to answer. Then proceed carefully to work on each problem.(Completion of all the exercises is a pre-requisite to attending the ComprehensiveSeminars). If an exercise is difficult re-read the workbook section or review thevideo.





1-3

Caution

Products specified for use in this manual might not conform to the National ElectricCode and may not conform to local requirements if the system is installed on abuilding, movable structure or vehicle. Before assembly and installation of any

photovoltaic power system, you should consult with local authorities so that you maybe assured installation will safely conform to all local building code requirements. Apermit may be required.

Also, consult local codes before using products or installation procedures outlined inthis manual. Codes could possibly cover applicable inspections, permits forelectrical wiring, wire size, interconnections, grounding, enclosures, conduits, over-current protection, receptacles, load restrictions, disconnects and appliances.Failure to follow applicable codes constitutes misuse of the products.

Do not attempt installation before reviewing all applicable instructions.

The technical information and suggestions for installation, operation, use andmaintenance made herein are based on Siemens Solar Industries knowledge andexperience and are believed to be reliable, but such information and suggestions donot constitute a warranty, expressed or implied.

Since the conditions or methods of installation, operation, use and maintenance ofthe equipment described in this manual are beyond Siemens Solar Industriescontrol, Siemens Solar Industries does not assume responsibility and expresslydisclaims liability for loss, damage or expense arising out of or in any way connectedwith such installation, operation, use or maintenance.





1-4

(End of Chapter)





1-5

CHAPTER ONE

INTRODUCTION 1-1

Welcome to the World of Photovoltaics 1-1

Purpose of This Manual 1-1

How to Proceed Through These Materials 1-2

Caution 1-3



Siemens Solar Basic PV Technology Course Fundamentals - Basics of Photovoltaics

Copyright ©1998 Siemens Solar Industries

2-1

Chapter TwoThe Basics of Photovoltaics

“Photovoltaic” refers to the creation of voltage from light, and is often abbreviated as just “PV." A more common term for photovoltaic cells is “solar cells," although thecells work with any kind of light and not just sunlight.

A solar cell is a converter – it changes energy of light into electrical energy. A celldoes not store any energy, so when the source of light (typically the sun) is removed,there is no electrical current from the cell. If electricity is needed during the night,some form of electrical storage (typically a battery) must be included in the circuit.

In this chapter we will present some of the most fundamental concepts of energyand power that are the basis for understanding photovoltaic power systems. We willalso discuss some of the common terms used in photovoltaic technology, andpresent the prime benefits of using solar electricity for your power requirements.





2-2

What Are Solar CellsThere are many materials that can be used to make solar cells, but the mostcommon is the element silicon. (This is not to be confused with “silicone," asynthetic polymer.) Silicon is the second most abundant element in the Earth’scrust, next to oxygen, and silicon and oxygen together make quartz or common

sand. It is therefore very abundant, as well as non-toxic and safe. This is the samesilicon that is used to make computer chips, and some of the processing stepsinvolved in making solar cells are similar to the steps in making computer devices.However, solar cells are much larger than typical individual computer circuits, andthey must be much less expensive! A typical solar cell used for terrestrial (Earth-based) applications is 3-6 inches in diameter and costs only a few dollars, whereas atiny computer circuit device might be only a tenth of an inch in length and width andcost tens or hundreds of dollars.

The conversion process occurs instantly whenever there is light falling on thesurface of a cell. And the output of a cell is proportional to the input light: the more

light, the greater the electrical output. The cell does not use up any internal “fuel” toproduce output. The sunlight acts as the “fuel” for the conversion process. And that“fuel” is delivered free everywhere in the world. The solar resource is more uniformlydistributed over the Earth’s surface than other renewable sources of energy lightwind or hydro. These resources are plentiful in certain specific climates andgeographic locations, but may depend on exact details of land contour andelevation.





2-3

What Are Solar Cells?

• Thin wafers of silicon – Similar to computer chips

– But much bigger and much cheaper!

• Silicon is abundant (sand) – Non-toxic, safe

• Light carries energy into cell

• Cells convert sunlight energyinto electric current- they donot store energy

• Sunlight is the “fuel”





2-4

Key Benefits of Solar Electricity

Energy Independence

There are many other ways to generate electricity, but what are the specific benefitsof solar generated electricity from photovoltaic cells? One of the most attractivebenefits is that you will have “energy independence," the ability to create your ownelectrical power, independent of fossil fuel supplies or utility connections.

Fuel Is Already Delivered

In a sense you do need sunlight as the “fuel”, but that is already delivered for free allover the planet’s surface. Other conventional generation methods require access to

a site for fuel deliveries. This may limit the choice of suitable sites so as to haveroad access, and even then access may be prevented due to poor road conditions,or vehicle problems. The cost of delivering fuel to remote locations can besubstantial. For example, it has been estimated that it requires one unit of dieselfuel to deliver one unit of diesel fuel to remote villages along the Amazon River inBrazil. In other words, the cost of the fuel is doubled!

Minimal Maintenance

Solar electric systems typically require very minimal maintenance because there areso few moving parts. Contrast this with a diesel-powered system or even otherrenewable sources such a wind generators or hydro generators, which often havecostly repairs or regular maintenance of moving parts. Very complex photovoltaicsystems do have more parts and may require some maintenance. But when lookingat small power requirements, such as for home lighting or remotetelecommunications systems, only occasional battery maintenance is required. Itwould be a mistake to say that photovoltaic systems require NO maintenance, butthe absolute amount of time and money required for photovoltaic systems is quitelow.





2-5

Benefits of Solar Electricity

Energy independence

“Fuel” is already delivered free everywhere

Minimal maintenance

Maximum reliability

Generate the energy you need where you need it

Reduce vulnerability to power loss

Systems are easily expanded

Maximum ReliabilityThis is perhaps the primary advantage of photovoltaics when compared to any otherform of electrical power generation. Because there are typically few or no movingparts and the complexity of the systems can be kept low, the ultimate reliability ofphotovoltaic power systems in the real world is quite high. The photovoltaicgenerator typically is not affected by environmental effects such as lightning strikes,high winds or blowing sand, humidity and heat, or snow and ice. The key to reliabilityis quality and simplicity. If high quality components are used with the solid-statesolar generators, and if the component count and complexity of the system designare kept to a minimum, the chance of any failure occurring is remarkably low.





2-6

Generate Where Needed

You can think differently about designing power systems for your loads, and notalways have to consider a central generator large enough for all your currentdemands. You can distribute the generation of power to various sites, such as ateach classroom, or each house, rather than always having to install a large

generator and string power lines to individual users that might be separated by greatdistances.

Telecommunications systems designers can look to photovoltaics as a way toperhaps be more selective with the locations of their repeaters. For example, anengineer might be considering covering a certain area with repeaters, and think thathe is forced to choose sites that are easily access, so that diesel fuel can bedelivered and maintenance can be performed. But the sites may not be the optimumfor coverage of the area. Instead, by choosing to use photovoltaic power for hissites, he may now consider more remote, inaccessible sites, and may actually beable to install fewer total repeaters but end up giving the same coverage that the

more numerous accessible sites would give.

Because photovoltaic generators can be as small as a few watts, you can trulyconsider installing just the amount of power that you need at each site. This flexibilityis not available from other forms of generation.

Reduced Vulnerability

Because you can avoid stringing long power lines for many miles or kilometers fromsome central generation source, many of the problems with utility power losses canbe avoided. Ice storms or vehicle accidents can cause power lines to go down,perhaps tens or hundreds of miles from where the power is actually needed. With areliable photovoltaic power system at your site you could still have power, whileothers around you have none.

And if you have chosen to distribute the generation of power to various load sites atyour location, you can insure even more reliability and less vulnerability to each loadsite. For example, if separate homes have their own lighting system, and one useroverdischarges their batteries, or damages their system, the other users will beunaffected.

This also applies to deliberate vandalism or terrorism. For example, having electricalgeneration for lighting and security distributed to each building would make loss ofsecurity for the whole site more difficult or practically impossible.





2-7

Easily Expanded

Photovoltaic power generators are modular by design. More power can be added toan existing array easily. Old modules can be added to new ones without any penalty(if the voltages are properly matched). Just enough power can be purchased andinstalled today to meet your current needs, and as demand grows more modules

can be added in later years. This also means that financially it is easy to start with aminimal power system today, and then add to the power as your budget allows later.





2-8

How Solar Cells Work

People often say that solar cells work by “magic” because there is nothing moving,the result is instantaneous, and no fuel is apparently needed! The basic process bywhich solar cells convert sunlight into electricity can seem “magical”, but actually is

simple. In a later chapter on “Photovoltaic Physics” more details will be given. Fornow, we can however give a simple explanation.

Internal Field and Electron Flow

Most typical solar cells are made of the element silicon. (For cells made of othermaterials, the explanation is still basically accurate). When light shines on a solarcell the energy of the light actually penetrates into the solar cell, and on a randombasis “knocks” negatively charged electrons loose from their silicon atoms. To

understand this we can think of light as being made of billions of energy particlescalled “photons”. (These are not the positively charged “protons” located at thenucleus of atoms). The incoming photons act much like billiard balls, only they aremade of pure energy! When they collide with an atom the whole atom is energized,and an electron is ejected or ionized from the atom.

The freed electron now has extra potential energy, and this is what we call “voltage”or electrical “pressure”. The freed electron has energy that could be used to chargea battery or operate an electric motor for example. But the problem is how to get thefreed electron out of the solar cell. This is accomplished by creating an internalelectro-static field near the front surface of the cell during manufacturing. Other

materials besides the basic silicon are “grown” into the silicon crystal structure. Theycreate an electrical imbalance that results in a one-way electrical “broom” that“sweeps” the freed electrons out of the solar cell and pushes them on to the nextcell, or on to the load.

As billions of photons flow into a cell that is exposed to light, billions of electrons areknocked loose and gain extra energy. They flow though the internal electro-staticfield, and out of the cell or module. This flow of electrical charges with extrapotential energy or voltage is what we call “electrical current”.





2-9

How Solar Cells ChangeSunlight Into Electricity

• Light knocks loose electronsfrom silicon atoms

• Freed electrons have extraenergy, or “voltage”

• Internal electric field pusheselectrons to front of cell

• Electric current flows on toother cells or to the load

• Cells never “run out” ofelectrons

P/N junction

e-

photon

e-

h+internalfield





2-10

Water Analogy

It is often helpful to give an analogy to water flowing. Imagine a water pumpconnected to a circuit of pipes that are already full of water. The pipe circuit alsoincludes some sort of “load” like a water wheel, and all the water returns back to the

pump through the pipes, so that no water is ever lost due to evaporation orsplashing. In the analogy the pump is the solar cell, the pipes are the wiresconnecting the cell to an electrical load and back to the cell, and the water in thepipes is like the electrons already in the wires.

When sunlight (or any light for that matter) shines on the cell it delivers the “fuel” thatis needed by the cell needs to operate. Electrons are freed and set in motion. Theinternal electro-static field pushes the freed electrons out of the cell and into thewire. The analogy would be as if the pump was turned on. The pump begins topush water into the pipe. But the pipe is already full of water, so water flows almostsimultaneously throughout the whole system.

The water flows on to the “load” like a water wheel, where its pressure and flow allowuseful work to be done. All of the water is then captured and flows again thoughpipes back to the pump. The pump continues to push new water to the load throughthe pipes.

In the real case of the solar cell the electrons freed by the incoming sunlight photonsflow out of the cell and on to the electrical load. They give up their extra potentialenergy or “voltage” there and allow useful work to be done. The electrons thencontinue to flow to the back of the solar cell, where they become available onceagain to be knocked loose and flow on to the load.

The electrical circuit is closed, just like the pipe system was closed, so no electronsare ever “used up” in the process. The solar cell never “runs out” of electrons. Itonly needs continuous input of “fuel” in the form of light energy to keep running.





2-11

Cells Into Modules

Because typical silicon solar cells produce only about 1/2 volt we need to connectcells together to give more useful voltages. When electrical generators areconnected together in “series”, or positive to negative, the voltage of each generator

adds up.

Usually 30-36 solar cells are connected together in a circuit to give a final voltage ofabout 15-17 volts, which is enough to charge a 12-volt battery. Charging batteries isthe primary use for photovoltaic modules, so most are designed around doing that

job.

Connect Cells To MakeModules

• One silicon solar cellproduces .5 volt

• 36 cells connected togetherhave enough voltage tocharge 12 volt batteries andrun pumps and motors

• Module is the basic building

block of systems• Can connect modules

together to get even morepower

But manufacturers could produce different module designs to better match other

loads, for example high voltage motors for water pumps or utility connected systemsthat often operate at hundreds of volts.

If the voltage or current from one module is not enough to power the load, thenmodules can also be connected together, just as the cells were. Manufacturesusually build modules with convenient junction boxes that allow interconnecting inseries or parallel.





2-12

Terms and Definitions

There are some basic terms that we have been using that should be carefullydefined, so we all speak the same language. Often people not familiar with thetechnology may use the terms incorrectly. Specifically, we want to be clear on thedifference between “modules” and “panels”.

• The CELL is the basic building block of a manufacturer of solar modules.

The fundamental physics of the materials used determines the voltage of a cell, andthe size determines the current. Usually manufacturers settle on one or two basicsizes and designs for their cells and then proceed to make millions of them. Theycan be packaged and exported to other module manufacturing facilities, and builtinto specialized products such as lanterns, radios and garden lights. Think of thecell as the smallest unit to work with. But it is fragile, and its voltage is low (typically1/2 volt). For use in the real world it must be protected and connected to other cells

to give useful voltage.

• The MODULE is really the basic building block for real-world remote powersystems.

It is a collection of cells interconnected by usually flat wire, and includesencapsulation to protect the cells and interconnecting wires from corrosion andimpact. It usually includes a frame to allow easy mounting and a junction box toallow wiring to other modules or to the battery and loads. The number of cellsconnected in series determines the final voltage of the module. Usually this is 30-36

silicon cells to give a voltage suitable for charging 12-volt batteries, but somemodules are made to deliver higher voltages for use in utility power systems. Aphotovoltaic power system can be a simple as one module connected to a batter ora motor. If more current or voltage is needed then modules must be connectedtogether.

• A PANEL is a collection of modules physically and electrically grouped togetheron a structure.

This would be the building block for larger power systems. Usually the modules are

wired together on the panel to give the final system voltage (for example 12, 24, 48volts or higher) and the panels are wired together or individually through field junction boxes and then on to the system controls and batteries. Individual panelscan be disassembled or maintained while the other panels are operational.





2-13

• An ARRAY is the full collection of all solar photovoltaic generators.

Sometimes an array is so large that it is grouped into SUB-ARRAYS, for easierinstallation and power management. An array can be as small as one module (for asimple home lighting system) or as large as 100,000 modules or more for very largeutility connected systems!

Terms Used

CELL -- basic building block in factory

MODULE -- smallest unit that can do

real-world work; building blockin the field

PANEL -- physically connected modules on astructure

ARRAY -- all solar generators in one

installation





2-14

Basic Concepts of Energy andPower

We should take some time to make sure that the meanings of some commonly usedterms are quite clear. All future discussions in this course will depend onunderstanding the differences between energy and power, and between voltage andcurrent. Once again, using a water analogy can help in understanding these terms.

Voltage

We can draw an analogy between the term “voltage” and water pressure. Imaginewater held behind a large dam where there is tremendous pressure but no forwardmovement of the water. An electrical system with high voltage potential means thatthe electrons in the components and wires have stored “pressure” and is capable ofdoing work if released to flow, just as the water could do work if released from thedam.

The unit of measure used for electrical voltage is the “volt”.

It is important to realize that voltage does not flow. It is a measure of the differencein force or pressure between two points in an electrical circuit.

CurrentThe analogy to electrical “current” is the rate of flow of water. Electrons actually flowpast any point in an electrical circuit, just as water actually flows past when releasedfrom a dam (or any other source of pressure).

The unit of measure for electrical current is the “ampere” or just “amp”. The numberof electrons flowing is enormous, amounting to over 1,000,000,000,000,000,000electrons per second in just one amp!

And don’t refer to current flow as “amps per second”. It is simply “amps” -- the unit

already includes the notion of number of electrons per second.





2-16

Energy

If power is the rate at which work is done, then energy is the AMOUNT of work thatis done during a specific period of time. The rate of doing work (power) must bemultiplied by the time to give the amount of work.

Amount = Rate X Time

Energy = Power X Time

= Current X Voltage X Time

The unit of measure of electrical energy is watt-hour or kilowatt-hour (kWh).

Energy is so fundamental that it is hard to define. It is the fundamental “stuff” of theuniverse. Scientists and engineers have learned how to describe the conversion ofenergy and how it transforms from one form (heat for example) into another (light forexample). But no one can very clearly define exactly WHAT IT IS. What we doknow is how to use it to do “work”.

So ENERGY is what you have to work with, while POWER is the rate at which youconvert or use it.





2-17

Battery and Module Work Together

The importance of understanding the difference between energy and power can nowbe presented in the context of a photovoltaic system. Most photovoltaic systemsuse batteries to store the energy converted by the solar modules during a day intochemical energy for use during the night or on stormy days. The battery acts as areservoir of ENERGY and mediates between the POWER that might be available atany moment from the solar modules and the POWER that the loads might want todraw at that instant. If the loads need more power than the modules can produce,then the battery discharges a bit to supply the difference. During the night, forexample, the modules produce no power so the battery must discharge to supply all

the power needed by the loads. During a day, if the loads do not require all thepower available from the modules, then the extra power goes into recharging thebattery.

Battery and Module WorkTogether

DaytimeCharging

UseAnytime





2-18

Balance the “Energy Diet”

The amount of time that the loads draw their power determines the total amount ofENERGY that they draw in a typical 24-hour day. The modules must, on theaverage, replace that ENERGY during the few hours of sunlight that are available.

So the loads can operate anytime, day or night, on cloudy or clear days. Themodules replace the used energy only during daylight hours.

Another way to look at this is: The loads can only use up the energy that themodules can produce.

The power of the loads can exceed the power of the modules, but the energy usedby the loads cannot exceed the energy produced by the modules. If it did, over timethe battery would try to fill in the deficit and would quickly become fully discharged.

For short times the loads can draw more power from the battery than the modulesare producing at that moment. At night the loads can draw power even though themodules are producing no power at all.

What must be kept in balance is the ENERGY produced and consumed. If theaverage daily load energy consumed is balanced with the average daily moduleenergy produced, then you have a proper solution to your energy problem!

During this course, you will learn how to predict the energy demand of the loads anddetermine how many solar modules are needed to meet that energy demand.





2-19

(End of Chapter)





2-20

CHAPTER TWO

THE BASICS OF PHOTOVOLTAICS 2-1

What Are Solar Cells 2-2

Key Benefits of Solar Electricity 2-4Energy Independence 2-4Fuel Is Already Delivered 2-4Minimal Maintenance 2-4Maximum Reliability 2-5Generate Where Needed 2-6Reduced Vulnerability 2-6Easily Expanded 2-7

How Solar Cells Work 2-8Internal Field and Electron Flow 2-8Water Analogy 2-10Cells Into Modules 2-11Terms and Definitions 2-12

Basic Concepts of Energy and Power 2-14Voltage 2-14Current 2-14Power 2-15Energy 2-16Battery and Module Work Together 2-17Balance the “Energy Diet” 2-18



Siemens Solar Basic PV Technology Course Fundamentals – Market OverviewCopyright ©1998 Siemens Solar Industries

3-1

Chapter ThreeMarket Overview

In this section, we will examine some of the key players in the worldwidephotovoltaic market, including the major manufacturers and the commontechnologies used today. We will look at some trends in application and marketgrowth that will give you a sense of the excitement we feel for the future. The mainpurpose of this section is to give you an orientation to the current and future statusof this power generation technology and how Siemens Solar is positioned to stay theworld leader.




3-2

Major ManufacturersThe number of manufacturers of solar photovoltaic cells and modules peaked duringthe late 1970’s and early 1980’s at almost 30, but has decreased during the 1980’sand 90’s to about a dozen major contributors. The cumulative shipments of the topdozen manufacturers during the last fifteen years are shown below. There is

manufacturing occurring in all regions of the world, including India, Europe, Japan aswell as the US.

H e l i o s

S h

a r p

C E L

S o l e

c

E u r

o s o l a

r e

P h

o t o

w a t t

A S

E B P

K y o c e r a

S

o l a

r e x

S i e

m e n

s

0

2 0

4 0

6 0

8 0

1 0 0

1 2 0

1 4 0

M

e g a w a t t s

S h i p p e d

Cumulative Shipments1980-1996

Source: Strategies Unlimited

In 1996 Siemens Solar celebrated a milestone for the world photovoltaic industry bybecoming the first company to ship a cumulative total of 100 Megawatts of solarcells and modules. The other major manufacturers are Solarex (U.S.), Kyocera(Japan), and BP Solar (U.K.).




3-3

The overall worldwide market for photovoltaic power is shared by the majormanufacturers. The market share of the key manufacturers is shown below for1995. Again Siemens Solar has the largest single portion of the worldwide market,with about 23% of the total.

Market Share

1996

(by shipments)


Siemens Solar 21%

Solarex 11%

Kyocera 11%

BP Solar 9%

Eurosolare 3% ASE 4%

Astropower 3%Sharp 3%

Other 35%

89.5 MW




3-4

Market and Application Forecasts

Market Growth Projections

How big is the photovoltaic market and how fast is it growing? These questions are

of interest both to people actively participating in the business and to financial andgovernmental institutions looking to photovoltaic technology as one answer to theirapplication problems.

One prediction of how the market for photovoltaic power will grow in the next fewdecades has been developed by Strategies Unlimited, a consulting firm that hasfollowed the growth of photovoltaics worldwide for many years. The assumptionsbehind each of these scenarios are discussed a little later in this section.

The “usual” growth projection shows 20-22% annual growth through the year 2010,and anticipates no technical breakthroughs and no increase in government

incentives. Even at this normal pace, the market size is projected to be about 1800megawatts of annual production worldwide. This would mean an increase of 4000%from 1990 to 2010!

Long Term Market Forcast1990-2010

• Business As Usualgrowth~22% to 2000~20% to 2010

• Accelerated growth~29% to 2000~22% to 2010

Source:Strategies Unlimited Report PM-39 March 1993

0

500

1000

1500

2000

2500

3000

3500

1990 1995 2000 2005 2010

Accelerated As Usual

ShipmentsMWp

42.7

3200

1800




3-6

Application Growth Forecasts

We turn our attention from general market forecasts to specific application areas tolook at how photovoltaic power is used around the world. The application groups ofphotovoltaic power can be divided into five broad groups:

• Remote Industrial: This has been the major application area for 30 years, includingtelecommunications, cathodic protection, telemetry, navigational systems and otherunmanned installations in harsh remote sites. The load demands are well knownand the requirements for reliable power are the highest.

• Remote Habitation / Consumer Power: This segment includes applications thatare typically occupied, such as cabins, homes, villages, clinics, schools, farms, aswell as individually powered lights and small appliances. The load demands in thissegment are not as well defined, and are more flexible.

• Grid Connected: These systems are typically multi-kilowatt or megawatt scalesystems that are directly connected to an existing power grid network. Electricpower is generated only during daylight hours, and is either consumed at the site ofgeneration (as on commercial buildings) or is fed into the general utility grid systemand consumed as a part of the normal power system. Small 4-10 kilowatt rooftopsystems can be located on top of individual homes, while larger 30-100 kilowattsystems can be associated with commercial or industrial buildings to offset theirdaytime lighting or air-conditioning loads. Large 100-500 kilowatt systems can beinstalled along utility feeder lines close to their full capacity to improve power qualityand postpone rewiring or installing new larger transformers.

• Government Demonstration: This market segment has always been small, and

constitutes projects funded by government or military organizations. These aretypically funded early in a government’s learning curve about photovoltaics, toprovide information on installed costs and reliability.

• Consumer Indoor: These products use photovoltaic cells to provide the smallamount of power needed for small electronic devices such as watches andcalculators.




3-7

The latest estimate for the way modules are currently used worldwide is shown onthe next page. The remote habitation application area is about 50% of the totalworldwide market, with the traditional remote industrial applications making up thenext largest market segment.

Market Segments

Remote

Habitation

Remote

Industrial

Grid

Connected

ConsumerOther

Forcast 1997


50%

27%

10%

8%5%




3-8

Projections have been made to the year 2010 estimating the changes that mightoccur in the relative size of different application groups. Once again, they present a“business as usual” forecast and an “accelerated” forecast. The two scenarios aresimilar in most respects. The remote industrial application group, the traditional“bread and butter” of the industry, will continue to be a major segment, but willcontinue to yield to remote habitation and grid connected systems. The majordifference in the two scenarios is the timing and the extent of the growth of the grid

connected application group.

Source: Strategies Unlimited Report PM-39 March 1993

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 1995 2000 2005 2010

Consumer Indoor

GovernmentDemonstration

Grid Connected

Remote Habitation

Consumer Power

Remote Industrial

Long Term Forcast (business as usual)

Major Application Groups




3-9

Long Term Forcast (accelerated)

Major Application Groups

Source: Strategies Unlimited Report PM-39 March 1993

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990 1995 2000 2005 2010

Consumer Indoor

Government

Demonstration

Grid Connected

Remote HabitationConsumer Power

Remote Industrial

The future will have photovoltaics used closer to where we live and work, as the cost

of delivered power comes down to compete with generators and utility power.

Remote habitation already constitutes the largest market segment. In developedcountries, this would constitute power for remote homes, cabins, and farms. Fordeveloping regions, these applications would be home lighting, schools, clinics,farms and village power.

The most important trend to understand is the impending growth of the gridconnected market. As the cost of photovoltaic power continues to decrease, and astraditionally generated utility power costs continue to increase, the marketacceptance of photovoltaic power will accelerate. In these scenarios, the market

share due to grid connected systems will be near or greater than 50% by 2010. Thiswill indicate the maturation of the photovoltaic market, as what has been traditionallyconsidered a “remote” power solution becomes economic for the vast urban market.




3-12

Ribbon Silicon

The slicing process in the previous two methods is wasteful, often converting 40-50% of the material into dust. This is because the wafers are only approximately0.015" thick, and the saw blade is about this thickness as well.

One method of producing wafers avoids most of this waste by growing a thin ribbonfrom the melted silicon. The ribbon is either pulled sideways off the top of the melt,or pulled up through a die.

Very fast growth rates are possible, but the speed results in polycrystallinestructures. If the pulling process is done very carefully, near single crystal structureis possible. The ribbon thickness is approximately 0.010"-0.015", so no furthersawing is necessary. The ribbon is simply scribed and broken to producerectangular wafers.

However, the surface of the wafer is not typically flat and often bulges. Waviness in

the surface makes further manufacturing steps and interconnection difficult.

Efficiencies similar to polycrystalline silicon are typical.

Thin Films

All of the previous methods produce a single cell as the basic building block. Thismeans that many cells must be connected together to produce a module of usefulvoltage, because each cell produces approximately 0.5 volts. The interconnectingand subsequent lamination steps are costly.

In the past 10 years great progress has been made in manufacturing solar modulesby depositing extremely thin films of semiconductors onto glass or metal substrates.This process has many advantages over the "traditional" methods mentioned abovethat produce individual cells.

The semiconductor layers are only a few hundred atoms thick, so expensive materialcosts are reduced. The entire module is made as a unit, so interconnectingmachinery and processing are eliminated. And the cell size can be modified easily,so it becomes easy and cost effective to make modules of different power output fordifferent applications or products. A unique characteristic of some thin film solardevices is that the light that does not interact to knock loose electrons can passthrough the device because the layers are so thin. This means semi-transparentfilms are possible. Car sunroofs, boat hatches, and building glass, for example, canbe made that produce useful electric power and also serve as windows!

Several approaches are being pursued around the world to develop thin film solarcells and modules. Some of these are described next.




3-13

Thin Film Silicon:Hydrogen (TFS:H)

Thin films of silicon:hydrogen alloy are the predominant technology mass producedtoday, and are found in solar calculators, watches and PV modules.

The atomic structure of the thin film is not totally ordered, like the structure of singlecrystal silicon. The probability is very high that an electron freed by light will

recombine with a hole before it can get very far. The light energy would just beturned into low-grade waste heat. That is why the doped layers are placed on eitherside of a large intrinsic, or undoped, layer. Instead of waiting for electrons knockedloose by photons to wander into a region of electric field, the electric field spansalmost the entire thickness of the semiconductor. Electrons that are knocked looseby incoming light are pushed immediately by the field, and have a very good chanceof being swept out of the cell before they recombine with a hole. The sandwiching ofan intrinsic layer of silicon:hydrogen alloy between P-type and N-type layers is calleda P/I/N structure.

Currently the efficiency of mass produced large area thin film silicon:hydrogen alloy

modules is about 3-6%, lower than that of the "traditional" technology modules. Butdevices of 12% efficiency have been reported.

Copper Indium Diselenide (CIS)

Another combination of semiconductor materials that shows great promise is copperindium diselenide (CuInSe2 , often shortened to just CIS). When combined with athin layer of another semiconductor, usually cadmium and zinc sulfide (CdZn)S, thedifference between the two materials forms a p-type/n-type junction. With both thesingle crystal and the thin film silicon cells discussed before, one material (silicon)

was treated with small amounts of impurities (dopants) to create a P/N or P/I/N junction. This is called a "homojunction" or "single junction" device. In the case ofCIS, the junction is created by placing two different materials in contact with eachother, creating a "heterojunction" device.

The response of CIS to light extends from the middle of the visible range far into thenear-infrared region of the solar spectrum, and allows for more of the available lightto be used by the cells compared to single or polycrystalline or amorphous silicondevices. Efficiencies of greater than 12% have been achieved on small devices, andover 10% on one square foot large area modules.




3-14

Tandem TFS/CIS

An exciting prospect for high efficiency and low cost lies in combining thin filmdevices of different spectral responses together in one module. One plan for such a"tandem" module involves placing a module of thin film silicon:hydrogen alloy with atransparent back conductor in front of a module made from CIS. The front thin filmsilicon:hydrogen module absorbs the short wavelength blue, green and yellow light,

and lets the orange, red, and infrared light pass through to the CIS module.

Research devices achieving over 15.5% have been made using this method, and itis anticipated that 18-20% is achievable. The combination of higher efficiencies andlower costs than any traditional design makes this a very exciting near-termtechnology.

Concentrator Cells

Covering large areas with solar cells and intercepting what light falls on the surfaceis called a "flat plate" method. This is the common approach taken bymanufacturers of single and polycrystalline silicon and thin film devices. Anotherapproach is to focus the light onto a small area and have a specially made cell at thefocus. This seeks to reduce overall cost by reducing the amount of cell materialrequired and substituting inexpensive lens or mirror materials.

Concentrator cells must be specially made to handle the large currents produced.From 10-500 times the normal sun intensity can be produced on the cell surface,resulting in an equally large amount of current compared to what would be producedby flat plate converters.

To keep the focus on the small cell, the entire module assembly must veryaccurately track the sun. Since tracking is essential, concentrator devices areappropriate only where clear skies are predominant.

Recent advances in concentrator cells have been achieved using a "point contact"structure. Thousands of p/n junctions are made on the back of a single silicon chip.Thousands of interconnections must be made to make a working cell, but some ofthe losses that occur in single junction devices are avoided and efficiencies ofgreater than 30% have been reported. The transfer from laboratory to large-scalecommercial production has yet to be achieved.




3-15

Market Share By Cell TechnologyThe latest figures for 1995 for shipments worldwide separated by cell technologyshow that single crystal silicon cells and modules dominate the photovoltaic market.Approximately 57% of all shipments of photovoltaic power generators were made upof single crystal or CZ technology, most of this was produced by Siemens Solar.

The next largest component of the photovoltaic market is polycrystalline technologywith about 25%. And thin film technology, primarily amorphous silicon cells forconsumer products such as solar calculators and watches, constituted a little morethan 13% of the worldwide shipments.

Single Crystal Dominates

World Shipments• Worldwide Shipments By Technology: 1996

T h i n F i lm s

1 3 %

R i b b o n

3 %

P o l y c r y s t a l l i n e

S i

2 7 %

O t h e r

3 %

S i n g l e C r y s t a l

S i

5 4 %

48.3 MW 11.7 MW

24 MW89.5 MW

Source: The Solar Letter, February 14, 1997

CdTe,

Concentrators

The total worldwide production of photovoltaic power generators in 1995 was about81 megawatts. This is large for the photovoltaic manufacturers, but is quite smallwhen compared to the generating capacity of conventional power plants. One largepower plant can produce 800-2000 megawatts, continuously day and night. Thisshows how far the photovoltaic market has to grow to become a major contributor tothe world’s energy needs.




3-16

It is interesting to look at how the primary technologies (single crystal, polycrystalline,and thin film silicon) have evolved in the marketplace in recent years. Below isshown the relative market share of each technology in 1991 and 1995. The actualproduction level of thin film silicon has decreased by 20%, and the production ofpolysilicon has decreased slightly also, while the production level of single crystalsilicon has increased by 137%.

Worldwide Growth Rates OfPrimary Technologies

1991-1996

0

10

20

30

40

50

60

70

80

90

1991 1996

W o r l d P r o d u c t i o n M W p

Thin F ilm Po ly CZ

54 MWp

84 MWp

+145%

+15%

-15%

Source: Solar Letter, 2/17/97

CZ

Poly

Thin Film




3-18

In general, single crystal silicon devices are the most efficient, with polycrystallinesilicon cells slightly lower in efficiency, and thin film devices of amorphous silicon thelowest in efficiency. Data that has been gathered over a period of seven years fromactual outdoor testing of these different technologies is shown. This data wasgathered as a part of a US government/utility program called Photovoltaics for UtilityScale Applications (PVUSA), where multi-kilowatt arrays of various technologieshave been installed and are being monitored for performance in real world

conditions.

Single Crystal Silicon EfficiencyIs Superior

• PVUSA testing various systems since 1989

• Actual field performance

0

2

4

6

8

10

12

J a n - 8 9

J u l

J a n - 9 0

J u l

J a n - 9 1

J u l

J a n - 9 2

J u l

J a n - 9 3

J u l

J a n - 9 4

J u l

J a n - 9 5

J u l

J a n - 9 6

D C E f f i c i e n c y %

Single Crystal

Polycrystalline

Amorphous

Source:PVUSA ProgressReport March 1996

The PVUSA data shows clearly how the different technologies compare in efficiency.The average efficiency of the single crystal silicon array is about 10%, while theefficiency of the polycrystalline modules is about 8%. The thin film amorphoussilicon array averages only about 3% efficient. This comparison shows the normalseasonal variations in efficiency due to ambient heat in the summer months for thesingle and polycrystalline modules. One interesting feature of thin film amorphoussilicon devices is that they are less affected by temperature variations, as shown bythe nearly flat efficiency data throughout the hot and cold seasons.




3-19

Advantages and Weaknesses ofCurrent TechnologiesEvery technology available today has some positive features as well as some

negatives. It is useful to put the different technologies side-by-side and make somegeneral statements about how they compare. In the future, the weaknesses of anyone technology may be overcome, and we may see new manufacturing techniquesthat allow for currently weak technologies to become more competitive in themarketplace. But for now, we seek to identify some very general characteristics ofeach technology and give you a highly simplified overview of how they compare.

Cell Technology Comparison Chart

Cell TypeBest CellEfficiency

Module AreaEfficiency

Advantages Weaknesses

Silicon 22.7 % 12-15 %Well understood;

Receivingrenewed attention

Indirect band gaplimits efficiency;

How thin?

CdTe/CdS 15.8 % 6-8 %Low cost;

High depositionrates possible

Cd liability;Needs moredevelopment

AmorphousSilicon

13.2 % 4-9 % Low costLooses power

over time;Low efficiency

CuInSe2 16.9 % 10% 23% potential;Low cost

Manufacturing yieldsare low; Needs more

development

Single JunctionConcentrator

28.7 % NAHybrid PV / thermal;

central powergeneration

Lacks productioneconomy of scale;

Complex BOS

MultijunctionConcentrator

35 % NAHybrid PV/thermal

Space

Lacks productioneconomy of scale;

Complex BOS




3-20

The bulk silicon (both single crystal and polycrystalline) and thin film amorphoussilicon technologies have been described earlier. The cadmium telluride(CdTe/CdS) material is another heterojunction thin film technology that shows muchpromise and is being pursued by a variety of manufacturers. The multijunctionconcentrators are more exotic devices that utilize different cell technologiesdeposited on top of one another, to capture different regions of the solar spectrum.These offer the highest potential efficiencies, but are by far the most complex and

expensive. Their best application would be where space is an expensive premium,such as in large central power generation plants or in orbiting space facilities.

The primary trade-offs that are evident from this comparison are potentially highdevice efficiency vs. reduced efficiency for real modules, and potential low costs vs.process stability and manufacturing yield issues. There is not one clear choice forthe “best” technology, as all of them have potential improvements and difficulties toovercome.

The “module area” efficiencies mentioned in this comparison are general values foractual production volume power modules, and are contrasted with the “best cell

efficiency” values that are often reported in press releases and research papers. Itis very important to draw a distinction between the peak efficiency of a singleresearch device and the consistently attainable efficiency of a high volumeproduction module. A single research device may have a high efficiency, and showthe potential that production devices can attain, but actual full production modulesincorporate interconnect space, frame space, gridline coverage and other physicalfeatures that decrease the efficiency of the overall device.

Another concept that is highlighted in this table comparison of technologies is thatconcentrator technologies have not achieved mass production status yet. Theyhave the potential for the highest efficiencies of all the technologies presented, but

involve extensive balance-of-system (BOS) equipment, such as specially designedsun-position tracking frameworks, concentrating lenses, computer controls, andspecial heat sinks. Their ability to compete with flat-plate technologies will dependon the cost-effectiveness of the entire package needed to create useable power,and not just on the efficiency of the small individual photovoltaic cells.




3-21

(End of Chapter)



Siemens Solar Basic PV Technology Course 4-1 Fundamentals – Skills for System Designers

Copyright ©19978Siemens Solar Industries

Chapter FourGeneral Skills of System

DesignersMany people feel that photovoltaic system design involves using a computer orcalculator and solving some equations that lead to one unique solution. The truth isthat designing solar power systems involves a great deal of judgment before andafter any calculations are made, and any recommendation by a computer programmust be tempered by experience and a keen sense of unpredictable human andenvironmental factors. Skilled system designers are critical of their own

assumptions and question their results to see if all contingencies have beenconsidered.

In this section, we will discuss some of the approaches that should be followed bysystem designers to gathering the initial information, dealing with objections anduncertainties, and thoroughly transferring the final system to the user. Theseinclude:

• Considering Tradeoffs • Anticipating and “Second-Guessing”

• Identifying Potential Applications • Handling Customer Objections • Gathering Information • Designing the System • Installation • Transferring the System to the Client

Each of these skills will be discussed next.





Consider TradeoffsMost of the time, system design involves some degree of compromise betweencompeting and desirable qualities. There are many choices that have to be made:type and size of equipment, location, amount of backup or redundancy, degree ofprotection, level of safety, amount of complexity, initial costs or costs over time, and

so on. Factors that may influence the choices might include the client’s budget, theremoteness of the site, how critical the loads are, the sophistication of the actualusers, and future growth possibilities.

System Design InvolvesTradeoffs

EfficiencySimplicity

andReliability

InitialCost

LifetimeCost

CentralizedGeneration

DistributedGeneration

• Choices based on budget, remoteness,how critical is load

Discussed next is some fundamental tradeoffs that must be made constantly when

designing photovoltaic systems. Both of the extremes of the tradeoffs are desirable,so the dilemma facing the designer is how far to go towards one side or the other.





Efficiency vs. Simplicity and Reliability

People designing technology based systems often tend to seek maximum efficiency,and consider any additional equipment or expense to be justifiable if it results inmeasurable efficiency increases. But this tendency must be tempered by real worldconcerns for field failures and general unpredictability in weather and load demands.

The primary characteristic of photovoltaic generators is simplicity--no moving parts!And from this comes an extremely high reliability. Any move away from thissimplicity and associated reliability must be scrutinized carefully. Most users wouldprefer a slightly inefficient system that continues to perform year after year over asophisticated system that has failed due to a component failure or unanticipatedglitch. It is the user’s perspective that must be considered first, and not thedesigner’s desire to show analytically under “standard conditions” a high theoreticalefficiency.

Initial Cost vs. Lifetime CostFor some clients the most important factor they face is keeping the initial cost of asystem down. It doesn’t matter to them that certain components, such as batteries,are not the highest quality types and will not last more than a few years. They canafford to buy replacements when needed a few years down the line.

For other clients, the possibility of reducing maintenance costs is extremelyattractive. For example, industrial customers designing critical and very remotesystems understand the high cost of diesel refueling or helicopter time to replacebatteries. These costs over time can be shown to easily justify the increased cost ofchoosing the highest quality components from the start.

Centralized vs. Distributed Generation

Photovoltaic technology works well as a distributed source of power, where eachload or home or user can have their own array. However, there may be situationswhere a centralized power generation approach is desirable.

A centralized approach may be appropriate when for example a large and complex

hybrid diesel-PV system is needed for a village or commercial site. Trainedpersonnel can be assigned to operate and maintain the equipment. Responsibilityfor proper operation falls on someone other than the final users. But this can alsomean that the users do not appreciate the necessity for conservation in their energyusage. They may abuse the system, add unnecessary or unauthorized loads, and ingeneral cause problems for the others in the centralized system.





On the other hand, if a distributed system is designed, then each site or user mustbe responsible for their own energy generation and usage. If they abuse or overusetheir system, their system may fail or shut itself down. But this will not affect theother users, and the abuser will quickly learn to be more conscious of their energyusage. Also the geography of the overall site may favor a distributed approach. Forexample, Navaho Indians in the central region of the US have their individual homeslocated perhaps a kilometer or so from the nearest neighbor. Stringing power lines

to each home would prove uneconomical compared to installing separate and stand-alone power systems for each home.

Anticipate and Second-GuessEven when you have seemingly reliable information for your design calculations, youshould think about what could go wrong and try to incorporate equipment that canhelp prevent abuse of the system and premature failures. Cultural differences orclimatic conditions that are different from the designer’s own should be understood.

The final users may not have any appreciation of how to maintain high technologyequipment or they may think that the system can take care of itself, or they may beafraid of it--all factors that can lead to misuse and failure.

Often photovoltaic power systems are being installed where no electrical power wasavailable before. Once users become accustomed to using it, they may use muchmore that the system designer was told they would. Lights and appliances may beleft on for more hours than planned, or extra lights and loads may be added,overstressing inverters and wiring as well as draining the batteries.

Simple choices like lights with timers and “hard wiring” appliances to a central power

distribution center can help to keep the system working within the limits it wasdesigned for. Feedback to the users from low voltage alarms can help educatethem about their overuse – so they stop using non-essential loads before thebatteries are over discharged.

There is not a formula to apply in the system design process that will assure that allcontingencies have been considered. It is up to the designer to stop and think--haveI put myself in the place of the intended user and into the intended climate, andthought about what could possibly happen? The more thorough the second-guessing, the more robust and reliable the final system design.





System Design InvolvesSecond-Guessing

• Anticipate what could gowrong

• Incorporate equipment toprevent improper operation

• Question your initial loadand location information





Identify Potential ApplicationsThe process of system design cannot begin until good cost-effective applicationshave been identified. Several key qualities that indicate a good opportunity forphotovoltaic power are given.

Certainly remoteness and any requirement for high reliability clearly point toconsidering photovoltaics. And relying on photovoltaic systems instead of utility linesor generators can reduce vulnerability to failure due to extremely bad weather, suchas ice or windstorms or blowing sand. And having distributed photovoltaic power foreach critical load or site means that if one system is damaged, the others remainoperational.

If there are critical loads at a site, then more than one source of power is desirable.Photovoltaics can be combined with conventional sources to add reliability anddiversity and insure that critical systems stay on line no matter what may happen.

A unique feature of solar photovoltaic generators is that they can generate smallamounts of power and output at relatively low voltages. Many small loads, forexample a remote telemetry system or an electric fence charger, need only a fewwatts of power, while generators cannot be found below a few kilowatts ofgenerating capacity, and utility line extensions for such small loads would surely betoo costly.

The low voltage of the modules can come in handy when compared to the extremelyhigh voltages of transmission lines. For example, aircraft warning lighting atop highvoltage transmission towers could tap off the very power lines they support forpower, but the cost and weight of the transformer needed would be prohibitive. But

lightweight solar modules and batteries can be mounted on the tower itself andprovide 12 or 24 volt power to lights without the need for transformers.

Certainly if there is a concern about noise or air pollution or environmentalsensitivity, these are flags that a photovoltaic solution may be appropriate. Andclients may have general social concerns about using renewable sources of power,or about being independent of utility networks.

The low voltage of typical solar modules makes them safe to handle and install. Ofcourse once many modules are connected together in series, potentially dangerouspower is present. But most small photovoltaic systems involve only 12 or 24 volt

power, which is safer than the common 110 or higher voltages from utility power.





Identify Potential Applications

•Modularity

•Safety

•High Insolation

•Independence

•Social Concerns

•Environment

•Portability

•Remoteness

•Reliability

•Vulnerability

•Critical Loads

•Low Voltage

•Small Power

•Noise

•Pollution

There may be a desire for a modular approach to power generation, where somepower is needed now and more will be needed in the future. Instead of buying and

installing an oversized generator anticipating long term future power requirements,you can install only the power that is needed today, and add modules at any time inthe future when budgets or load demands increase. Entire village power systemscan grow this way, starting small and growing as the community grows.

Individual modules are lightweight and can be easily transported if necessary.Modules have been used for mobile radios, camera battery recharging, lighting fortemporary camps, as well as transported on trucks or by hand from one water pumpto another.

And certainly if there is abundant solar insolation, photovoltaics should be

considered. Why not use a resource that is plentiful, and delivered for free!





Handle Customer ObjectionsAfter potential applications have been identified, designers may face the obstacle ofdealing with objections or questions about the technical aspects of photovoltaicpower systems. Users may not understand how the modules will survive in harshenvironments, or work with other components to give reliable power day or night.

Some typical objections are listed.

Handle Objections

• Needs Maintenance

• User Education

• Breakable

• Theft

• Safety Concerns

• Building Codes

• Need the Sun

• Initial Cost

• Produces DC• Area Limitations

Certainly a common objection is the relatively high initial cost of a photovoltaic powersystem compared to a generator only system. This must be handled by presentingthe true costs of operating a generator over time, including regular and unscheduled

maintenance, parts, labor, fuel (including delivery costs) and other “invisible” costs.The problem of high initial cost can also be addressed by developing creativefinancing arrangements, so that users pay for their energy over time just as they dofor conventional energy.





Another key objection is that photovoltaic devices make DC electricity, while mostcommon appliances and other loads are designed to operate from AC power. Thiscan be dealt with by either proposing the use of DC appliances, or by including asmall inverter that converts DC to AC. The extra cost of the inverter can becompared to the usually higher prices of DC appliances, and a choice can be madeas to which way to go.





Gather Load InformationOnce any objections have been overcome, it is time to begin gathering critical dataabout the situation. There are three areas of information that are important: theapplication; the climate; and the client or user. Application information must be asdetailed and quantitative as possible. Otherwise slightly wrong information at the

beginning of the sizing process will result in completely incorrect recommendations.

Gather Information About theApplication

• Load requirements

• Load profile

• Surges

• Power quality

• DC or AC

• Critical loads

• Ease of access to site

The individual loads must be well understood, including their current and voltagerequirements, and how many hours each day they will be on. In addition, the profileof the load demand over a typical week or for each month must be specified. It will

be important to know if the load is greater in the summer or winter. This informationwill help determine the proper tilt angle of the modules so that they intercept themost solar radiation when the loads need it. Also if the load is used only a few dayseach week, as in the case of a weekend cabin, can affect the balancing of modulesand battery storage and final system cost.





Certainly if AC loads are to be included, their requirements for power quality andsurge currents will affect the choice and sizing of a DC-to-AC inverter. Theinefficiency of the inverter must be added into the calculations, and the input voltageof the inverter chosen will determine the voltage of the solar array and battery bank.

Seasonal Load Profiles

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0

10

20

30

40

50

60

70 Summer peak profile

Winter peak profile

If there are any critical or essential loads this will influence the choice of sizing of thearray and battery to handle unexpected bad weather. And the ease or difficulty ofreaching the site may influence the reserve time built into the battery, as well as thesetting of any low voltage alarm signals to prompt maintenance actions.





Gather Climate InformationAccurate location and weather data is of course critical to proper system sizing.Data on solar radiation or insolation must be obtained, and not just for a season orfor one year, but accumulated over many years. Specific data for a site taken over ashort period may not reflect the normal weather that can be expected. The usual

format for solar insolation data is in units of kilowatt-hours/square meter (kwh/m2), orin Langleys (calorie/cm

2), or sometimes in megajoules/m

2 or btu-hours/ft

2.

Measurements of simply “hours of sunlight” or “number of bright days” is notquantitative enough and cannot be used. Common sources of solar insolation dataare airport authorities, agricultural stations, government agencies, and universities.Siemens Solar has compiled a database of solar insolation for over 2000 sitesaround the world. This is used by our computer sizing programs to accuratelypredict module output.

Gather Information About theClimate

• Latitude, longitude

• Insolation

• Temperature

• Variability of weather• Harshness





Temperature ranges as well must be well know, as modules loose voltage potentialand batteries loose life expectancy with higher temperatures. Temperatures will alsoinfluence the choice of the module design, as 36 cells module circuits are typicallyused for the hottest climates, while 33 cells in series are adequate for moremoderate climates.

Local knowledge about the variability or harshness of the weather is important to

gather. Statistical data on solar insolation is usually based on averaged data, soparticular variation patterns from day to day are lost. The severity and duration oftypical seasonal storms would be very helpful in deciding how much reserve capacityto build into the battery bank, and whether backup generator capacity should beconsidered.





Gather Client InformationA third critical source of information that directly affects the choices you make indesigning a photovoltaic system is the client or user of the system. The informationgathered here may not be as mathematically precise or as easily quantifiable as theprevious application and climate information, but it is quite important.

The client’s budget may be a useful starting point for your design process. You mayoften have to begin with the budget and work backward to see how much solarpower they can afford at the start. You can always encourage them to add moresolar power as their budget allows.

The reliability of the information that they give to you should be examined. Is theload information that they have given accurate, or is it just a guess? Are theyrepeating what others have said, or are they sources of original data? Sometimesclients are seeking replacements for fuel-powered generators, and they may makethe simple mistake of thinking that the load is equal to the power rating of the

generator, without looking at detail at the actual loads. If a large generator wasinstalled for surge power or because it happened to be on hand or for whateverreason, then any calculations made based on that capability will be far too high forthe actual site.

The technical skill level of the users will have a definite influence on your systemdesign choices. The sophistication of controls and feedback, and the accessibility tothe system will depend greatly on your assessment of the user’s abilities andunderstanding. If the user is fairly uneducated or unskilled with electricity, then youwill want to make the feedback simple and the access quite limited. This may alsoinfluence your choice of battery type, where you would probably choose a

maintenance-free battery. Whereas, if the user is a sophisticated engineer, you willprobably want to allow greater access to the system controls, and may feel free touse a battery that requires proper and regular maintenance.





Gather Information About theClient

• Budget

• Reliability of their data

• Level of technical skill

• Future growth possibilities

• Aesthetics

You should keep in mind the probability of near-term future growth of demand, and

design your system to accommodate that growth. If the load is well known andfixed, with no possibility of increase, then you can design your system to fit that loadvalue exactly. But if there is the real chance of load expansion, for example morechannels of radio being utilized, or more hours of lighting needed or more users in avillage, then expansion and increased power flows should be planned. Spaceshould be left in the module mounting structures, larger wires should be used,regulators and inverters should be oversized, and so on. This will help to reduce theneed to purchase more components in the near future.

And aesthetics may have an influence on your design choices. Whether the userwants the array to be visible or not will influence your location options and may affectyour wire lengths and sizes for example. Instead of mounting the array on theground at the proper tilt angle, the user may want the array to be mounted on a roofwith a tilt or azimuth angle that is not optimum. This will definitely affect your arraysizing calculations.





Design the SystemIt seems that we would never get to the point of finally calculating the array andbattery size! But not we are ready. It takes all this previous data gathering to be in aposition to make the choices needed during system design. Design is not justplugging in some numbers. It involves balancing calculations with judgments, and

selecting equipment based on the information about the climate, the load and theclient.

Design the System

• Array sizing• Battery bank sizing

• Wiring

• Safety components

• User feedback

• DC or AC or Hybrid

• Mounting

• Accomodate future growth

The number of modules in series and parallel for the array can now be calculated,as well as the capacity and voltage of the battery bank. All of the system wiring

must be chosen, with attention to safety equipment such as circuit breakers, fuses,disconnects and grounding. Electrical code considerations may dictate the use ofequipment that adds to the overall system complexity and cost, but also adds tosafety.

At this time you should consider the hardware needed to give proper user feedback.Simple dials or fancy digital meters can be chosen, and even remote transmission ofsystem parameters can be designed.





Fundamental decisions need to be made here about whether the system is all DC,or all AC, or some combination of the two. You have to decide whether the systemwill be a pure stand-alone photovoltaic system or a hybrid design with a fuelgenerator or wind or hydro assist.

The array mounting design must be made, perhaps including empty space for futuremodule additions. Components such as inverters charge regulators, and wiring

should be chosen to allow maximum growth, so that costly upgrades or changes canbe avoided.





InstallationYou can perform all the calculations in the world, but if the equipment is not installedcorrectly, your system could still fail. Bad communications between the designersand the installers can lead to wasted time, money and faulty system performance.

As much of the system should be pre-assembled as possible and shipped to the siteready for installation. This allows you to test the components as they work togetherand make repairs or adjustments as needed at the home shop (where tools andspare equipment is readily available). Charge controls, system disconnects,inverters and feedback components can often be pre-mounted to a strongbackboard, transported to the site, and installed quickly.

Installation

• Pre-assemble as much as possible

• Site selection

• Safe practices

• Codes and inspections





The installation site should be visited before system design is completed so thatspecial circumstances can be taken into account. Perhaps there are trees or othersources of shading at the site. This might require clever mounting solutions orlonger and larger wire than anticipated. Foundation design might depend on thetype of soil, the wind loading, the growth of vegetation and so on.

Safe practices should be followed at all times during the installation phase. All

appropriate local and regional electrical and building codes should be followed.

Each component and the overall system should be fully tested before the installationis considered complete. Abnormal operations should be corrected, and all systemvoltage drops should be carefully measured to insure that all connections are proper.





Transfer the System to the ClientThe system design process is still not complete until the system has been properlytransferred to the client or user. Often this step is left out with the assumption thatthe user understands what to do. This can be disastrous. Users might notunderstand the importance of no shading on the modules or of regular battery

maintenance. The system can still eventually fail, even if all care has been taken indesigning and installing the components, if the users don’t properly maintain theirsystem.

Transfer System to Client

• Simple language or pictorial

• Troubleshooting Guide

• Service/Maintenance Plan

• Get user sign-off





A Troubleshooting Guide should be created for the user, showing “symptoms,possible causes, corrective actions”. A Service and Maintenance Manual should beon hand, showing the user with simple pictures what to check and when, and howthey should perform the service.

It would be a good idea to get the user to sign-off on these transfer documents. Thisadds to the perceived seriousness of the transfer process, and forces them to be

aware of the content of the documents.

Examples of Transfer Documents

By way of example we have included two samples of documents that serve totransfer understanding and responsibility for proper system operation tounsophisticated users.

Presented on the next page is a reduction of a large color poster created for a rural

electrification program in South America. It explains in simple words and drawingsthe proper operation and maintenance of a home lighting system.

And on the following pages is an owner’s manual created for small homeelectrification systems installed on the Navajo Indian Reservation in the UnitedStates.

Please look at these documents and think about how you would best structure yourtransfer documents so that they would be useful to the people involved. Differencesin education and sophistication will require vastly different documentation. Somedocumentation for rural systems users may involve only pictures or cartoons, with

very little text, while complex engineered systems may require detailedspecifications. The user should be the starting point for designing transferdocumentation.





The Homeowner’s Manual presented on the following pages was prepared for theowners of simple 12-volt DC home electrification systems installed on the NavajoIndian Reservation in the United States.





(End of Chapter)





CHAPTER FOUR

GENERAL SKILLS OF SYSTEM DESIGNERS 4-1

Consider Tradeoffs 4-2Efficiency vs. Simplicity and Reliability 4-3Initial Cost vs. Lifetime Cost 4-3Centralized vs. Distributed Generation 4-3

Anticipate and Second-Guess 4-4

Identify Potential Applications 4-6

Handle Customer Objections 4-8

Gather Load Information 4-10

Gather Climate Information 4-12

Gather Client Information 4-14

Design the System 4-16

Installation 4-18

Transfer the System to the Client 4-20Examples of Transfer Documents 4-21



Siemens Solar Basic PV Technology Course 5-1 Fundamentals–- Applications & Typical Systems


Chapter FiveApplications and Typical

SystemsIn this chapter, we look at some of the most popular and economical types ofapplications for solar photovoltaic power systems. There are many new applicationsthat are being developed, but we can discuss here those that have already proven tobe cost effective and reliable.

We can also look briefly at some common system configurations, to become familiar

with some of the equipment that is used in systems. Most all of the equipmentmentioned (such as inverters and charge controls) will be discussed in more detail inlater chapters.





Flexibility of Application

One of the blessings of photovoltaic technology is that there is a tremendousflexibility in the kinds of designs and applications that solar power canaccommodate. There is a wide variation in the size of system, the complexity, and

the dependence or independence on natural cycles. There is not one “best” way todesign a photovoltaic system.

Application Flexibility

1 watt

Near existingutility power

Daytime only

Purely PV

Centralizedgeneration

Megawatts

Remote

Any weather, anytime

Hybrid system

Distributedgeneration





Power Flexibility

The inherent modularity of solar photovoltaic systems means that there istremendous flexibility in the amount of power you can install. Wind generators beginat about 500 watts of generating capacity, and fuel generators begin at a fewthousand watts. But you can install as little as a few watts of photovoltaic generating

capacity if that is all that is needed, for example to charge an electric fence energizeror operate a small telemetry unit.

Hundreds or thousands of watts of power can be installed for homes, farms,industrial or commercial uses, schools and clinics.

And hundreds of thousands or millions of watts of generating capacity can also beinstalled economically as well. Large-scale utility systems serving thousands ofhomes or businesses have been installed.

Near Utility Power or Remote

The historical market for photovoltaics has been remote applications where the costof extending utility power lines was prohibitively expensive. Even when the distancefor an extension is not very far and the cost is moderate, if the power needed issmall (for example to run a single water pump on a farm) it is still more economicalto install photovoltaic at the site instead of extending the line.

But you should not think of photovoltaics as being appropriate only for remoteapplications. The costs involved in utility line hookup even when the distances arevery short can still be higher than the costs of installing a stand-alone photovoltaicpower system. Urban street lighting, bus shelters, advertising lighting and othercommercial uses of power might involve relatively short hookups, but thetransformers and equipment needed, the civil work to trench for power lines or installpower poles, and all the associated costs of utility hookup can make even a shortconnection costly.

A perfect example of the cost effectiveness of photovoltaics in an urban electrifiedsetting is emergency roadside telephones. Thousands of these systems have beeninstalled along highways, where regular utility power is only a few feet away. But thecosts involved in civil work and equipment make installing photovoltaics theeconomic choice.





Daytime Only or Anytime

Just because the photovoltaic modules use the sun this does not mean that theloads need to operate during daylight. And it doesn’t mean that you can’t run yourloads during stormy weather. By using batteries for nighttime operation, andincluding backup generators such as diesel, wind or hydro, you can design a system

that will operate in the harshest of environments.

But photovoltaics do offer the unique opportunity to avoid the cost and complexity ofthese components if you indeed need only power during daylight! Water pumping isa perfect example of how solar photovoltaics can match a load requirement withminimal cost and complexity. A DC pump can be connected directly to a solar array,and will automatically operate whenever there is sufficient sun. During the summerwhen there is greater need for water, the system pumps more water automaticallybecause there is more sunlight! Attic or ceiling fans are other application areaswhere direct coupling to solar modules results in low cost highly reliable operationduring daytime only.

Photovoltaics Only or HybridGeneration

Systems can be designed to operate with only photovoltaics as the source of power,to maximize reliability and minimize maintenance and complexity. Or backupgenerators can be included in the system design to provide power day and night andduring seasons of low solar radiation. A diesel, gasoline or propane generator cancharge the battery bank and operate AC loads as well. Wind generators or hydrogenerators can also be connected to a system to provide power.

By combining the reliability and simplicity of photovoltaic modules with the 24-houravailability of standby generators, you can get the best of both worlds.





Centralized or DecentralizedGeneration

Photovoltaics are inherently a decentralized generation technology. The power thatis needed can be installed where it is needed. Each house, schoolroom,

commercial site or transmitter can be outfitted with its own independent system. Ifone system fails or needs maintenance, or if the user exceeds the energy available,the other systems continue to operate without interruption.

But decentralized systems require vigilance by their owners. If the skills orknowledge of how to operate or maintain the system were not available, thenperhaps a centralized approach would be best.

A centralized photovoltaic power system could have all the modules, batteries,inverter capacity and controls needed for an entire village. Trained personnel couldmaintain the equipment, and would be responsible for proper operation. The users

of the energy are not as directly connected to the source of their energy, but perhapsthis approach is best for their situation.





Outdoor Lighting ApplicationsOutdoor lighting applications are often decentralized and relatively small, providing aperfect match to photovoltaic power advantages. Each lighting unit can have its ownarray and battery and controls for maximum reliability and flexibility in site location,and minimum vulnerability to power loss. Often systems can be designed to operate

from only one or two modules.

Roadside Flashers

Highway Safety Signs





Telecommunications ApplicationsTelecommunications has always been an important market for photovoltaics. Theequipment used in many telecom applications is operated on DC power, making thematch to DC generated solar power simple and economical. Reliability in severeclimates is perhaps the most important attribute of photovoltaics for this market.

Downtime is terribly expensive and may be life threatening, so reliability is critical.And the modularity and flexibility of photovoltaic system design means that modulescan be used to power even the smallest telemetry station as well as very largemicrowave repeaters.

Microwave, TV orRadio Repeaters

Telemetry Stations





Radio and Telephones





Navigation ApplicationsSimilar to the telecom market, aides to navigation require the highest reliability.Signals or lights must operate under all conditions, in all seasons. They are typicallylocated in remote sites, and traditional power sources such as utility line extensionsor diesel generators would be quite costly. Each navigational aid can have its own

photovoltaic power supply, making vulnerability to overall power loss extremely low.And this market is a perfect example of how a relatively small amount of power canproduce a very large benefit.

Railroad Signals

Buoys





Airport Approach Systems

Offshore Oil Platforms





Cathodic Protection Applications

Cathodic protection involves the electrical prevention of rusting. Metals in contactwith the earth or moisture will naturally corrode. But this process is an electro-chemical reaction, and can be reduced or practically stopped by applying a reverse

current to the natural rusting reaction. The electro-chemical reaction involves DCcurrent flow, so photovoltaic power is a perfect solution. Small systems can beplaced along remote stretches of buried pipelines, for example, and systematicallyand automatically prevent corrosion with minimal maintenance requirements.

Well Heads

Oil and Gas Pipelines





Water Pumping Applications

Productivity in farming can be drastically improved with greater access to watersupplies. Often photovoltaic power systems are used to replace windmills that have

just broken down too often, or cannot lift the water due to lowering water tables.

Instead of installing a diesel powered pump system in remote reaches of a farm, adirect-coupled solar system can be installed. Such a water pumping system wouldrequire no batteries, no need for fuel deliveries or engine maintenance, wouldautomatically produce more water in the summer when it is needed more, and wouldgreatly improve the productivity of the land. Simple water systems can also beinstalled in remote villages or homes. Instead of drinking polluted surface watervillagers can drink fresh, clean ground water, thereby reducing disease relatedproblems and costs.

Livestock Watering

Village Drinking Supply

Irrigation





Remote Habitation Applications

One of the most exciting markets for photovoltaics is for home lighting andappliances. Single homes or whole villages can generate their own electric powerwithout the need for sophisticated maintenance or regular fuel deliveries. Small

clinics can operate vaccine refrigerators, sterilization equipment, emergency radiosand other critical loads. Typically they are located in very remote locations, andmoney is not available for costly generator maintenance or fuel. Vacation homes orcabins, or even regular residences can have all the comforts of modern life from aphotovoltaic power system, including lights, appliances, radio, TV, VCR, microwaveoven, computers, power tools and refrigeration.

Rural Home Lighting Systems





HomeElectrification

Schools

Hospitalsand Clinics





Power for Mobile Applications

Solar modules can be used in mobile applications for battery charging and operationof appliances and equipment. For example mobile recreational vehicles (RV’s) canhave a few modules mounted permanently on the roof, to supply quiet DC power for

lights, radio, TV, appliances, coolers and other small appliances. Boats can usemodules to keep starting batteries fully charged, and also for full operation whilerunning, such as lights, radio, radar and sonar, and other equipment.

And modules can be used for electric vehicle charging stations. Commuters withelectric vehicles can park under the shade of the solar array, and have their vehiclebattery recharged while shopping or during work hours.

Electric Vehicle Charging Station

Recreational Boats / Vehicles





Commercial Farm DSM Individual Home RooftopFacility System

Commercial Building with Integrated Module Facade





Reduces Daytime Consumption at theCustomer’s Site

An example is shown below of the performance of a DSM system using

photovoltaics to reduce the daytime demand of a commercial office building in theSacramento, California area. Solar modules generate approximately 15 kW of peakpower during the middle of the day, and act to reduce the overall demand of thecommercial building on the utility system. For large customers, utilities might havetime-of-day pricing, where they would pay a higher price for electricity during peakdemand periods. The added cost of the photovoltaic equipment could be comparedto the reduced demand for exceptionally costly utility electricity.

PV Output Reduces Peak Demand

0

20

40

60

80

100

120

0 2 4 6 8 10 12 14 16 18 20 22 24

Hour of Day

L o a d ( k W )

PG&E R&D Building, June 21, 1990

HVAC (minus PV)

PV Output

Lighting

Computing





6 MW Grid Support Facility

500 KW Grid Support Facility

200 KW Grid Support Facility





Correct Power Factor

Another problem that utilities face at the end of long feeder lines is that the “powerfactor” of their AC electricity can be poor. The power factor is a measure of how wellthe voltage and current waveforms are synchronized with each other. A power

factor of 1.0 would mean that the current and voltage hit their peaks at exactly thesame moment, and maximum real power is transmitted. Over long transmissionlines there is a lot of capacitance, and this tends to push the voltage and currentwaveforms out of “sync”. Also large motors at various commercial sites can alsopush the waveforms apart. The more out of “sync” the current and voltage are, theless real power is being delivered to the customer, even though the utility mustgenerate the real power in the first place.

Photovoltaic power systems putting their power through properly adjusted DC-to-ACinverters could “push” the waveforms closer together, and result in improved powerfactors and better efficiency.

Postpone Costly TransformerUpgrades

Utility feeders have large multi-megawatt transformers that reduce the voltage of theelectric power from the very high values used for long transmissions down to moremanageable values for local distribution. The transformers are installed with morepower capacity than is needed, but with community growth they may reach near theirfull capacity within a few years. At that time the utility might face the need to replacethe old transformer with a larger one. This can cost millions of dollars.

By installing a large utility interconnected photovoltaic power station downline fromthe transformer, the required power can be delivered to the customers, and thecostly upgrade can be postponed for years. The cost benefit to the utility isimmediate and substantial, and shows the true cost effectiveness of photovoltaics.





Delivered Cost Of Utility Power Is CloseTo Photovoltaic Cost

Various utilities have recently examined the true DELIVERED cost of their electric

power, as opposed to just looking at the cost of generating the power, and havefound that the dollar value of the electricity at the end of the wire can be substantiallyhigher than the “busbar” cost at the generator. This is extremely important becauseeconomic analyses of conventional means of power generation (nuclear, coal, gas)often look only at the generation costs but do not include all of the costs associatedwith transmission and distribution (“T & D”). This is especially unfair whencomparing the costs of photovoltaic generated power, because photovoltaics isinherently a distributed form of generation, and power can be generated close to thecustomer thereby avoiding many of the T&D losses and costs of conventionalsources. This cost advantage must be looked at to fairly judge the costeffectiveness of using photovoltaics in utility scale applications.

The results of two economic analyses are shown, one conducted by the Pacific Gasand Electric (PG&E) utility for a 500 kW grid support system and the other by a utilityin Western Australia. The generation costs of utility electricity are low compared tothe relative costs of photovoltaics, but when all the T&D costs have been added, thefinal effective cost of electricity at the customer’s site is almost doubled!

The cost components used for the analysis are defined below:

• Energy: The value of fuel savings

• Capacity: The value of deferring new power plants

• Min. Load Savings: Avoid power plant operation at inefficient minimum loads• QF Savings: Reduced avoided cost payments to qualified facilities

• Loss Savings: Locating generation at the load reduces resistance losses

• Voltage Support: Reducing losses in a high impedance network

• Substation: Value of avoiding or delaying substation upgrades

• Reliability: Value to customers of avoiding outages

• Transmission: Value of avoiding transmission line upgrades

• Environment: Value of avoiding emissions

The real cost of large-scale photovoltaic generated power compares favorably to

these utility costs TODAY. There is little need for utilities to wait until the cost ofphotovoltaics comes down – it is cost effective for them today!





Economic Analysis of Grid

Support Systems

0

100

200

300

400

500

600

700

800

PG&E SECWA A n n u a l L e v e

l i z e d V a l u e ( $ / k W - y r )

Environment

Transmission

Reliability

Substation

Voltage Support

Loss Savings

QF Savings

Min.Load

Capacity

Energy

Figure 5-4 Two utility analyses show real cost of delivered utility power





Typical System Configurations

Most photovoltaic power systems are custom designed for their particular situation,but we can generalize somewhat to show the common approaches to solving power

supply problems. In the following pages we present simple diagrams of the typicalconfigurations of equipment that are used to serve the applications that we havediscussed previously.

We can group system types into five broad categories:

• Simple Single-Module DC Systems

• Large DC Systems

• AC Power Systems

• PV-Generator Hybrid Systems

• Utility Connected Systems





Simple Single-Module DC Systems

Simple DC SystemConfiguration

DC Loads

• Rural Electric Lighting Systems: A single module is connected to a single lowcost battery through a simple charge regulator. The regulator has terminals forconnection to a fluorescent light. The regulator should have a low voltage alarmor relay that turns off the light if the battery voltage becomes too low. The userwould have to wait until the module charged the battery back to an acceptable

intermediate voltage before they could turn on the light again. The battery couldbe a commonly available automobile battery. It wouldn’t be expected to have along life, but its low cost would allow the initial system price to be acceptable.Replacing the battery every year or so could be an acceptable price for clean,reliable, safe lighting.





• Street Lighting Systems: Single or multiple modules can be connected througha charge regulator to a deep cycle battery, which powers a streetlight. These canbe used for roadside illumination or for parking lighting or pathway lighting. Anelectronic dusk-to-dawn timer is included in the system for automatic operation ofan efficient fluorescent or low-pressure sodium lamp. The light is switched onautomatically at dusk, and either operates for a pre-set length of time or untildawn. Rugged yet simple charge control and timer circuitry are demanded, for

reliable operation in severe environmental conditions. Sealed batteries arepreferred so that minimal maintenance and no corrosive gassing occur near thecontrols. Yet these types of batteries may have shortened life in the outdoors(hot) conditions expected for this application.

• Portable Lanterns: Rugged portable lighting systems are a most affordableoption for low-income rural areas. Lanterns may have the solar module built intothe lantern housing, or more commonly have the module separate, connected bya removable cord. Sealed maintenance free lead acid batteries are commonlyused, and nickel-cadmium batteries can also be used. Typical arrangements usea 5-7 watt fluorescent lamp, and a 5-10 watt solar module. Usage time is

typically 4 hours.





Simple DC Water Delivery Systems

Simple DC SystemConfiguration

• Personal water delivery systems: One or two modules connected directly to apositive displacement DC water pump can deliver hundreds of gallons frommoderate depths. Typically a waterproof submersible motor and pump islowered down a well or borehole into the water. Modules can be directlyconnected to the pump, or a maximum power tracking device can be connectedbetween the modules and the pump to maximize the electrical match andimprove output. During summer months when more water is typically needed fordrinking or irrigation and livestock, greater solar insolation means that more wateris pumped automatically.

• Surface mounted pumps: An inexpensive solution to surface mounted water

pumping is to remove an AC motor and replace it with a DC motor that can bedirectly connected to solar modules. An inexpensive jet pump conversion can bemade to surface mounted pumps to allow pumping from greater depths, althoughthe total output will be low due to the internal return flow of water to operate the

jet.

• Floating pump: A floating pump is appropriate for canals, lakes, ponds andeven open wells.





Large DC Systems

Large DC System Configuration

DC

Loads

• Multiple module arrays: Larger systems can be designed by adding moremodules and batteries. A larger charge controller would be needed to handle theincreased current flow from the array. Again, some DC loads could beconnected directly to the regulator.

• Load distribution box: If the load circuits were too numerous, a DC circuitbreaker distribution box could be used to allow connection of multiple loadcircuits.

• Multiple charge controllers: If the array is larger than the current capacity ofone single charge controller, multiple charge controllers can be connected inparallel. Slightly different charge voltages can be set to allow the battery togradually reach full charge.





AC Power Systems

AC and DC System Configuration

DC

Loads

AC

Loads

Inverter

• Inverter operates AC loads: AC appliances can be powered by adding an AC-to-DC inverter. This complex component switches electronically the DC voltageand current from the battery to produce alternating current (AC) and voltage,commonly used by household and office equipment. Such a power system couldthen operate any load that might be needed, including computers, fax machines,radio, TV, VCR and CD players, refrigerators and freezers, power tools andkitchen and bathroom appliances. DC loads could still be included in the systemas well.

Inverter





Photovoltaic-Generator HybridSystems

Hybrid System Configuration

AC

Loads

• Back-up generator: A common choice is a fuel powered motor generator, eithergasoline, propane, or diesel. By combining the reliability and quiet operation ofphotovoltaic modules with the assured availability of generator power during anyseason, you can assure power availability upon demand.

• Transfer switch: Generator AC output power can be passed directly on to ACloads. A transfer switch is needed to prevent generator power from feedingbackwards into the inverter. The transfer switch could be a fast acting electronic

design, or a simple manual switch that the user operates when needed.

• Rectifier charges battery: The generator can also pass power through arectifier that changes the AC back to DC current and voltage. This DC power ispassed to the battery bank, to recharge the batteries after a long period of belowaverage weather. A smaller battery bank can be installed compared to stand-alone systems because system autonomy is now provided by the generator.





Hybrid System Configuration

ACLoads

• DC charging only: A hybrid system can be designed to have the generator actonly as a battery charger. No AC output from the generator is used to run loads.Instead all the AC power for the loads is output only from the inverter. Thegenerator is turned on, either automatically by the charge control system or

manually by the user, whenever the battery voltage gets too low during badweather.

• Avoid transfer switch: Operating all loads from the inverter means there is notransfer switch “glitches” that could harm electronic equipment.

• Operates generator at full power: The generator is sized so that it can operateat its full rated output to charge the battery. When the batteries are sufficientlyrecharged, the generator is turned off, and the finishing charge is supplied by thesolar modules. Operating at full output means maximum fuel efficiency and longlife.

• Battery life extended: The generator is turned whenever needed to keep the

battery from staying discharged for more than a few days. Battery life ismaximized due to minimal sulfation.





Utility Interactive Systems

Utility Connected Configuration

Instead of having a system independent of the utility grid a system can be designedto work with the grid. A specially designed utility-interactive inverter is needed, andmany models are available worldwide. The basic arrangement of a utility-interconnected system is quite simple.

• The solar array is connected to the inverter, as are the utility lines. The output isconnected to the normal distribution box for the house or business.

• During the day power generated by the array is fed into the inverter and changedinto pure sinusoidal AC power that is synchronized with the grid frequency.

• If that power is needed in the home it is passed on. If the load demand is lessthan what the array is producing, the excess is fed into the utility grid system, andenergy is credited to the home.

• If more power is needed than the array can produce at a particular moment thenpower is drawn from the utility to add to the array power.

• Typically there are no batteries in utility-interconnected systems, so at night allthe power needed flows from the utility.





Recently, new designs of bi-directional inverters have been created that allow utility-interconnected systems to have battery backup as well. AC power from the utilitypasses through the normal meter and into a standard distribution center.

Synchronized sinewave power from the inverter is connected to this distributioncenter as well. Inverter power is also available from a second AC output connectionand can be sent to a dedicated distribution box for critical loads.

Bi-Directional Utility Configuration

InverterBatteryRegulator

Meter

Critical Loads

(Operate frombattery)

Non-Critical Loads .(Operate onlyfrom utilitypower)

• The inverter continuously shares between AC utility power and DC battery power.

• AC power from the utility can be rectified into DC power and used to recharge thebattery bank during bad weather.

• Battery power is drawn upon instantly if there is a utility power failure or

brownout, much like an uninteruptable power supply (UPS) system.• The non-critical load distribution center looses power when there is utility failure,

but the critical center continues to draw power from the inverter and battery.

Bi-directional utility interconnected system offer both independent power duringutility failures and utility backup during normal utility operation.





(End of Chapter)





CHAPTER FIVE

APPLICATIONS AND TYPICAL SYSTEMS 5-1Flexibility of Application 5-2

Power Flexibility 5-3Near Utility Power or Remote 5-3Daytime Only or Anytime 5-4Photovoltaics Only or Hybrid Generation 5-4Centralized or Decentralized Generation 5-5

Outdoor Lighting Applications 5-6

Roadside Flashers 5-6Highway Safety Signs 5-6Bus Shelters 5-7Garden Lights 5-7Portable Lanterns 5-7

Telecommunications Applications 5-8Microwave, TV or Radio Repeaters 5-8Telemetry Stations 5-8Radio and Telephones 5-9

Navigation Applications 5-10Railroad Signals 5-10Buoys 5-10Airport Approach Systems 5-11Offshore Oil Platforms 5-11

Cathodic Protection Applications 5-12Oil and Gas Pipelines 5-12

Water Pumping Applications 5-13Irrigation 5-13Livestock Watering 5-13Village Drinking Supply 5-13





Remote Habitation Applications 5-14Rural Home Lighting Systems 5-14Home 5-15Electrification 5-15Schools 5-15Hospitals and Clinics 5-15

Power for Mobile Applications 5-16Electric Vehicle Charging Station 5-16Boats / Vehicles 5-16Recreational 5-16

Utility Interconnected Applications forDemand Side Management (DSM) 5-17

Decreased Need for Costly Peaking Generation 5-17Same Effect as Installing More Efficient Loads 5-17

Reduces Daytime Consumption at the Customer’s Site 5-19

Utility Applications for Grid Support & Bulk Power Generation 5-20Improve Voltage at the End of Long Lines 5-20Correct Power Factor 5-22Postpone Costly Transformer Upgrades 5-22Delivered Cost Of Utility Power Is Close To Photovoltaic Cost 5-23

Typical System Configurations 5-25Simple Single-Module DC Systems 5-26

Simple DC Water Delivery Systems 5-28Large DC Systems 5-29AC Power Systems 5-30Photovoltaic-Generator Hybrid Systems 5-31Utility Interactive Systems 5-34



Siemens Solar Basic PV Technology Course Fundamentals – Photovoltaic Physics


6-1

Chapter SixThe Physics of

Solar CellsThe process of direct conversion of light into electricity seems almost magic. In thischapter, you will learn about the basics of how this process occurs, without the useof intimidating formulas or confusing jargon. The purpose of this explanation is tomake you more comfortable with the fundamentals of how silicon solar cells work.Other semiconductor materials can be used to make PV devices, but describing asilicon-based device will illuminate the general principles common to all.





6-3

When single silicon atoms are combined, they connect together to form a solid.Neighboring atoms share outer electrons, forming “bonds”. These bonds whereelectrons are shared between atoms is what holds all matter together. Crystallinesilicon consists of orderly bonding of each silicon atom with 4 neighboring siliconatoms. Such a highly ordered structure of atoms is also called a crystal lattice.

A model of crystalline silicon is shown below. Each atom is represented by a large

ball, and the interconnecting “rods” represent the shared electron “bonds”. Eachsilicon atom is connected or bonded to four other silicon atoms, forming a beautifulrepeating lattice structure. The lattice is the same arrangement as found in adiamond, except that silicon atoms are present instead of the carbon found in adiamond.





6-4

At the atomic level, light acts as a flux of discrete particles called photons. Photonscarry momentum and energy but are electrically neutral. When semiconductormaterial is illuminated by light photons of light actually penetrate into the material,traversing deep into the solid.

Photons with enough energy that collide with electrons can dislodge them from theirbond. The photon disappears and its energy is transferred to the electron, which

becomes free to wander throughout the semiconductor material as a conductionelectron, carrying a negative charge and usable energy.

It is at the moment of releasing the electron that sunlight energy has been convertedinto electrical energy.

e- e-e- e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e- e- e- e- e- e- e- e- e- e-

e-

e-

e-

e-

e- e-

e-

photon

e- e-

Holes Created by Photon





6-5

Whenever an electron is knocked loose it leaves behind a bond that is missing anelectron. Such an incomplete bond is called a “hole”.

A nearby electron can jump from its bond into the hole and fill it, but this leaves ahole where the electron came from. In this way the hole moves in the material. Butwherever there is a hole, an electron is missing, resulting in a localized net positiveelectrical imbalance there. The hole appears to be a positive charge moving in the

solid, although it is really an absence of an electron moving about. Overall the netelectrical charge of the material is neutral.





6-6

P-Type Dopants

In the absence of any external electric field electrons freed and energized byphotons will wander for a short time and then recombine with a wandering hole.The energy originally transferred to the electron from the photon is simply lost to the

semiconductor lattice as heat. The key to producing usable output current is tosweep the freed electrons out of the material before they recombine with a hole.

When there is just silicon atoms in the lattice, the material is called “intrinsic” or pure.One way to alter the electrical properties of silicon is to introduce elements into theintrinsic semiconductor that contribute excess holes and electrons. Materials thatsignificantly alter the properties of semiconductors are called “dopants” and theprocess of placing them into the semiconductor is called “doping”.

Boron P-Type Doping

e- e-e- e-

e-

e-e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e- e- e- e- e-

e-

e-

e- e-

e-

e-B

Defecit ofone electronleaves "hole"

e-

Nearby electronscan move in and"fill" hole.





6-7

One dopant used with silicon is boron, which has only three outer electrons, one lessbonding electron than silicon. Each boron atom can only bond with threeneighboring silicon atoms, leaving one bond half-complete. A nearby electron froma silicon atom can vibrate into the hole site next to the boron, “filling” the hole. But itleaves behind an absence of an electron, a hole, where it originally came from.Another nearby electron can vibrate into that hole, but leaves behind a hole where itcame from. So there exists in the semiconductor structure a wandering absence of

an electron.

Wherever this absence of electron (hole) is there is one too few electrons to balancethe positive charges of the nucleus of that particular atom. This results in a netpositive charge at that atomic site. As the hole site moves around, so does the netpositive charge. Since the moving charges in this material have a positive charge,the semiconductor material is called positive-type or P-type.

The concentration of boron is quite low, usually around one boron atom to every10,000,000 silicon atoms.

It is important to understand that the overall net charge in the semiconductor isneutral, but if you look at small regions, there will be net negative charges at theboron atom sites and net positive charges moving around with the holes. The boronatoms will have a negative charge because a neighboring electron has fallen into theincomplete bond and brings with it a net negative charge. The local negative chargeat the boron atom is balanced by the local positive charge near the hole, this beingnow a silicon atom that has temporarily lost one of its outer electrons.





6-8

N-Type Dopants

Another dopant used with silicon is phosphorous. Each atom of phosphorous hasone more outer bonding electron than silicon. The fifth electron breaks away fromthe phosphorous atom easily because there is not bond to hold it. It becomes a

freely moving negatively charged particle. Since the moving charges in this type ofsemiconductor carry a negative charge, this type of doped semiconductor is callednegative-type or N-type.

The concentration of phosphorous atoms is again quite low, but typically greaterthan the boron concentration, usually around one phosphorous atom for every 1000silicon atoms.

Phosphorous N-Type Doping

e- e-e- e-

e-

e-e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e- e- e- e- e-

e-

e-

e- e-

e-

e-Pe-e-

Extra electroncan wanderin crystal

e-

Once again, the net charge of the entire semiconductor is neutral. But in small localregions you can find fixed net positive charges where phosphorous atoms arepermanently missing their fifth outer electron. And there will be moving negativecharges that are those fifth outer electrons, freely wandering throughout the lattice.





6-9

Creating An Internal Electric Field

Regions of p-type boron-doped silicon and n-type phosphorous-doped silicon arecreated adjacent to one another. Some free electrons in the n-type region crossover and fall into the holes in the p-type region where they remain permanently.

As this cross over process continues, every boron site that contributed a holebecomes permanently negatively charged, and every phosphorous site that gave upan electron becomes permanently positively charged. Two equivalent but oppositelycharged regions grow on either side of the p-type/n-type interface, creating anelectric field.

The electric field is oriented to push electrons in one direction, toward the n-typeregion, and any holes are swept toward the p-type region. Any freely movingcharges that enter the zone of influence of the field are immediately swept out of thatzone, so that zone is called the depletion region. A name given to this type of

electric interface is a p/n junction.

e- e-e- e-

e-

e-e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e- e-

e-

e-

e-

e-

e- e-e- e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e-

e- e- e- e- e- e- e- e- e-

e-

e-

e-

e-

e- e-

e- e-

e-

e-B P

e-

Permanent internal electric field and lines of force

N-type silicon regionP-type silicon region

Creating Electric Field





6-10

The strength of the internal electric field is quite strong. The distance across theregion of the field is only about one micron (1/10,000 of a centimeter, or 1/25,000inch), but the field potential is about one volt. This means that if the field were toextend across one inch it would be about 25,000 volts! This strong field isequivalent to an “electronic broom” that can “sweep” freed electrons out of the celland create the one-way flow of electrons that can be called electric current.





6-11

Internal Electric Field SweepsElectrons Out of Cell

It is this internal electric field that sweeps electrons out of a cell. When lightpenetrates into the semiconductor material, knocking free electrons and giving thempotential energy, the freed electrons wander until they are pushed by the electricfield across the P/N junction. They are forced out of the cell, and are available foruseful work.

Internal Field Sweeps Electrons

Out of Solar Cell

Electronsreturn fromexternal

fill holes.

h+

e-

photon

e-

circuit to Freed electron

Lines of force ofpermanent internalelectric field

P/N Junction region

Newly created hole

Electronsflow on tonext cellor out ofmoduleand intoexternalcircuit.

In a module, a number of cells are connected together in series. The electrons flowfrom one cell into conductors that carry them to the next cell. In the next cell theyare once again struck by photons, given more potential energy, and swept out of thecell. Finally the electrons leave the last cell in the module and flow to the load.





6-12

For every electron that leaves a cell, there is another that is returning from the loadto replace it. The wire that is used to make the circuit from the module to the batteryor load and back to the module contains electrons, so as soon as an electron leavesthe last cell in a module and enters the wire, an electron at the other end of the wiremoves into the first cell in the module.

So the PV device cannot “run down” like a battery, nor can it “run out of electrons”.

It produces output (electrical energy) in response to input “fuel” (light energy). A PVcell cannot store electrical energy, it can only convert light energy into electricalenergy.





6-13

Current Is Regenerated In EachCellThe current that is generated in a cell does not simply flow on through all the other

cells in series with it and eventually on to a load. The electrons that are liberated inone cell flow through the internal p/n junction and on to the second cell, but they fallinto holes there and stop. They must be energized by photons again in that secondcell in order to move out of that cell and on to yet another, and so on. This is why itis so important for all cells in a solar module to be equally illuminated by light, withno shadows or dark regions. The current flow from one cell must be regenerated ineach and every cell in series with it in order to have the electric current flow properly.We will discuss the severe negative effect of shading on cell and module output laterin the program.





6-14

Voltage Increases As CurrentPasses Through Series CellsIn each cell electrons gain about one volt of potential energy (or “voltage”) when they

are energized and ionized by the photon. In passing through the p/n junction, theylose about one half volt through collisions and accelerations, so they are left withabout one half volt of potential as they leave the front of a cell and move on to asecond cell. They enter into the back of the second cell, and “fall” into holes in thesilicon structure. This flood of electrons with about one half a volt of potential energyraises the entire potential of the second cell. The voltage potential of electrons inthe second cell is about one half volt higher than it was in the first cell.

When electrons in this second cell are again struck by photons, they add theapproximate one volt of energy from the photon to their new higher “base” energy ofone half volt, so now they have about 1½ volts of potential. But again in passing

through the p/n junction of this second cell they lose about one half volt again, sothey are left with a net voltage gain of about one volt when leaving this second cell.They carry this one-volt on to the third cell, where they again “fall” into holes, andawait yet more photons to energize them again, and the cycle repeats for all thecells in series.

Manufacturers connect enough cells in series to produce a final voltage that isuseful. Typically modules are used to charge 12-volt batteries, which actually needabout 14-15 volts to be fully charged. So 30-36 cells are usually connected togetherin series to make a single module that can produce about 15-18 volts.

If higher voltages are needed, for example to charge higher voltage battery banks oroperate high voltage motors or interface with the high voltages of utility power, thenmodules are connected together in series. Some photovoltaic systems that areconnected to inverters that produce AC power to feed into the standard utility powergrid operate at voltages of 450 volts! This would mean that over 1000 cells areconnected together in series! Remember though that each cell contributes onlyabout one half volt, whether it is the first cell in the series or the last.





6-15

Exercises





6-16





6-17

Spectral Response

Different sources of light may appear equal in brightness to the human eye but willcontain different amounts and intensities of colors. For example, fluorescent light istypically stronger in the blue than incandescent lights. Also throughout a day there

is a difference in the spectral content of morning, noon, and evening sunlight, as canbe observed by looking at the sky.

“Light” is just a narrow range of all electromagnetic radiation that is emitted by thesun. Radiation is a moving electric-magnetic field, and the field vibrates regularly ata very rapid pace. The distance between the peaks of the vibrations of radiation iscalled the “wavelength” of that radiation. “Light” is radiation between approximately400 nanometers or nm (violet color) and 800 nm (red color).

200 400 600 800 1000 1200 1400

1.2

1

0.8

0.6

0.4

0.2

PowerDensity

( mw/cm2/nm)

VisibleIRUV Light

Wavelength ( nanometers, nm )

Standardized SolarSpectral Distribution





6-18

The spectral response is a measurement of the “response” (measured by generatedcurrent) when a PV device is exposed to a spectrum or range of light. A 100%response would mean for 100 photons of a certain wavelength that are absorbed,100 electrons would be freed and swept out for use.

Spectral Response

0.00

0.20

0.40

0.60

0.80

1.00

0 0.2 0.4 0.6 0.8 1 1.2

Wavelength (

m)

Q

u

a

nt

u

m

E

f

f

i

c

i

e

n

c

y

CZ

CIS

TFS:H

Light of the same color or wavelength will produce different amounts of current indifferent semiconductor devices. The spectral response of a typical CZ silicon cellbegins about 350 nm, peaks around 800 nm and falls off rapidly beyond 1100 nm.The response range spans the entire visible spectrum and reaches into the nearinfrared. The spectral response of a typical cell made from thin filmsilicon:hydrogen, on the other hand, is narrower, beginning around 300 nm, peakingaround 500 nm, and falling off beyond 700 nm. And a third type of PV devicematerial, copper indium diselenide or CIS, has yet another spectral response range,beginning just short of the other two but extending further into the invisible infraredregion.

Because a cell’s response to light depends on the wavelength of that light, justknowing the total energy of the light is not enough information to predict cell output.Two sources of light can have the same total energy density but one source couldhave its energy in the form of a few high energy blue photons, while another could





6-19

contain many low energy red or infrared photons. Also, two light sources canappear to be similar in brightness to the human eye, but one may emit a great dealof extra radiation beyond the visible range where our eye will not notice but to whichthe PV device will respond. A standard “typical outdoor spectrum” has been definedas the spectrum from the sun that filters through 1.5 thickness of our atmosphere,and is referred to as “Air Mass 1.5”. This serves as a common reference spectrumfor comparing device output.

Exercises





6-20

(End of Chapter)





6-21

CHAPTER SIX

THE PHYSICS OF SOLAR CELLS 6-1Holes and Conduction Electrons 6-2

P-Type Dopants 6-6

N-Type Dopants 6-8

Creating An Internal Electric Field 6-9

Internal Electric Field Sweeps Electrons Out of Cell 6-11

Current Is Regenerated In Each Cell 6-13

Voltage Increases As Current Passes Through Series Cells 6-14

Spectral Response 6-17



Siemens Solar Basic Photovoltaic Technology 6-1 Photovoltaic Physics

Chapter 6 – Answers The Physics of Solar Cells

The photovoltaic effect describes the property of certain materials to produce electronswhen light falls on (and is absorbed by) the material's surface.

Electron, positive (+)

b. An electron around a silicon atom

c. Get the electrons out before they recombine with a hole

N, negative, electron

P, positive, "hole" (absence of an electron)

Negative, positive

N, phosphorous

b. Convert energy



Siemens Solar Basic PV Technology Course 7-1 Components – Module Manufacturing & TestingCopyright © 1998 Siemens Solar Industries

Chapter Seven

Module Manufacturingand Testing

The manufacturing process used by Siemens Solar involves many proprietary elementsand is quite involved. But it is important to discuss the manufacturing process so thatyou can appreciate the technical features and benefits of the modules, and can haveconfidence in the durability and longevity of the products.

The manufacturing process can be presented as three segments or phases: crystal andwafer fabrication; cell manufacturing; and module assembly. Each of these processesis discussed next.

Manufacturing Flow

Melt and grow

into single crystal

Crop, slab andsaw into wafers

Etch surface

Diffuse withphosphorus

Coat with anti-

ref lect ive coating

Print contacts

Electrical test

Solder cel lstogether

Laminate

Attach frame,

j- bo x, di od es

Final electricaltest

Change quar tzinto pure si l icon

Crystal andWafer Fabrication

Cell Manufacturing Module Manufacturing




Wafer Manufacturing

Raw Materials

The process of making a solar module ultimately begins by mining the raw material.This usually consists of high quality silica or quartz (SiO2) from mines around the world.Silicon (Si) is an abundant element on the earth, and can be found in common sand,but it is more economical today to begin the process with silica that has been somewhatnaturally purified. In large furnaces, the silica is reduced to silicon and then purifieduntil it becomes 99.9999% pure! Examples of these purified rocks of silicon are shownbelow.




Crystal Growing

Chunks of purified silicon are carefully loaded into a crystal growing furnace, along witha small amount of the element boron. The furnace is sealed and the chunks are heatedto greater than 1400 deg.C. (over 2500 deg.F.) until they melt (Figure 7-4). A small

“seed” crystal of pure silicon about the size of a pencil is lowered inside the furnace untilit touches the liquid. The cooler seed acts as a template and atoms of silicon andboron freeze onto it, and the seed begins to grow (Figure 7-5). By carefully controllingtemperatures and the speed of growth, the diameter of the growing crystal is increasedto about 5" and then held steady by computerized controls. The crystal is then grown inlength until almost all of the original silicon material is exhausted.

This process of growing a single crystal structure from a small seed crystal is called the

Czochralski method, usually abbreviated as just CZ.




Cropping

After removal from the crystal growing furnace, the ingot top and bottom (“crown” and“tail”) are cropped off and recycled (Figure 7-6). The ingot is cut into short sections foreasier handling. The sides of the section are cropped along its length to make a

squared-off block. The “M” line of cells is cut almost square for the greatest packingdensity in a module. The “Pro” line of cells is cut less, leaving more of a rounded shape(Figure 7-7).




“M” and “Pro” Series Cells

• M series cell pack

together more densly

• Pro series cell requires

less overall processing




Slicing Wafers

The section is mounted into a “wire saw” to cut individual wafers. The wire saw involvesa single long thin wire wrapped many hundreds of times around four rotating drums.The ingot section is placed in the center of the four drums, and is slowly pushed up

through the web of the wires. Hundreds of thin silicon wafers are cut all at the sametime this way.

Surface Etching

The wafers are then etched to remove some surface damage caused by the wire sawabrasion, and to create a surface that helps absorb light (Figure 7-9). The final wafersurface is made of millions of tiny 4-sided pyramids, following the pure crystal structurein the original seed. These pyramids reflect light amongst themselves and allow morelight to be absorbed (Figure 7-8).




Cell Manufacturing

Diffusion

Wafers are next loaded into diffusion tubes which are long tubes of silica glass,surrounded by large resistive heaters, and that will contain a hot gas of phosphorous. Acomputer controls the entire process, slowly moving wafers into the chamber, allowinggases to flow for precise amounts of time, and then removing the wafers (Figure 7-10).This process is conducted in “clean room” conditions similar to the semiconductorindustry. Clean room conditions are important in optimizing the performance of thecells.

Inside the tube atoms of phosphorous penetrate or “diffuse” into the wafers a shortdistance, only 12 millionths of an inch or 0.3 microns, and embed themselves in thecrystal structure. This embedding of select impurities into a semiconductor like siliconis called “doping”.

The differences between phosphorous-doped silicon and the base material of boron-doped silicon set up a permanent electric field in each wafer. This field sweeps outelectrons that are knocked loose by light, and makes the wafer into an active “solar cell”when light shines on it.




Anti-Reflective Layer Deposition

The diffused wafers are coated with thin films of metal oxide to effectively reduce theamount of light reflected from the surface, and therefore increase the amount of lightabsorbed and converted into electricity. This “AR” coating turns the cell surface dark

blue. This is because the cells are most effective with green, yellow, and red light, anddo not use blue light very well. So the design of the AR coating is optimized for themost effective light, and the blue light is allowed to be partially reflected, so the cellslook blue.

Screen Printing of Gridlines

The cells are loaded onto a series of automated high-speed screen-printers that deposita pattern of metal paste onto the front and back (Figure 7-11). The pattern of thin linesand contact pads is designed to efficiently collect electrons that reach the cell surface,

without blocking out too much sunlight.




The metallic gridlines are the final structure that makes up a complete cell.

Structure of CZ Cell

N-type layer

P-type layer

300 nm

250000 nm

MetallicGridlines

MetallicGridlines

Anti-reflectiveCoating

150 nm

Electric Fieldat the p/n junction

Cell Operation

e- Thin N-type Layer

Thick P-type Layer

P/N Junction

region of

electric field




Every cell is tested under simulated noonday sunlight to determine its electrical output,and cells of similar output are sorted and grouped together (Figure 7-14).




Module Manufacturing

Interconnecting Cells

Cells are connected together in series to add voltage. This is done by an automatedsoldering machine that connects the back of one cell to the front of the next with tworibbons of tin-coated copper ribbon (Figure 7-15). The cells are soldered to the ribbonsat multiple points to allow for expansion and contraction of the copper against thesilicon. Two ribbons are used to add reliability to this critical interconnection.

Lamination

Strings of 10, 11, or 12 cells are soldered together to create final groups of 30, 33, or 36cells, and then are laminated into a sandwich of support materials designed to insurelong operating life. The sandwich consists of tempered glass, then one layer of EVA(ethylene vinyl acetate), the cell circuit, another layer of EVA, and finally a back cover ofmultiple polymer sheets that prevent moisture penetration. The interconnected cellcircuit is therefore “floating” in a sea of plastic. The sandwich is first sealed in vacuumlaminators that remove moisture and air and soften the plastics, and is then thermallyset in large ovens.

Final Module Assembly Steps

An edge gasket is placed along the perimeter of the laminate to create a soft cushionfor the final metal frames. A strong frame is attached to the edges of the laminate forstrength.

In modules which have three strings of cells (such as the SM55, SM50-H and SM46),separate positive and negative junction covers are attached at each end. Each box issecured to the laminate with strong adhesive, and contains a bypass diode, whichdecreases the negative effects of any shading of cells that might occur.

In module designs with four strings of cells (such as the SP75), a single junction box isattached at one end of the module. Dual bypass diodes can be installed in the single

junction box, with each diode protecting one half of the module circuit.




Every completed module is tested again under simulated noonday sunlight to insure itmeets minimum power output specifications. The output is recorded on computer forquality control and customer assurance. The modules also are tested for safe highvoltage operation by putting 3000 volts between the circuit and the metal frame. Allmodules are also visually inspected for blemishes or flaws before being packaged 4 to

a box for shipment.

Cells Connected In Circuit

e-

CZ cell Interconnectingcopper ribbons




Exercises




Qualification Testing ofModule Designs

PV power systems must be reliable even in the harshest environments and operate formany years to be economically justified. Siemens Solar currently guarantees moduleoutput for 10 years, and life cycle cost analyses are done based on module life of 20years or greater. To be confident to make such long-range projections, modulemanufacturers subject their modules to intense testing to “qualify” their electrical andphysical design. These tests are not normally done on every production modulebecause they are “destructive tests” and they can cause some discoloration ordegradation to the tested modules. These tests described below are “qualificationtests” which means that they are used to “qualify” an existing or new module design asa truly robust, long-life design, suitable for installation in any climate for many years.These tests are typically done on any new module design or when changes are madeto existing designs. Module manufacturers may have their Quality AssuranceDepartment perform some of these tests on an ongoing basis on samples of regularproduction modules, to continually check that manufacturing processes are producingquality products.




CEC Tests (JPL Block V Tests)

The first work done on creating standards for testing terrestrial solar modules was done

by the Jet Propulsion Laboratory (JPL) of the California Institute of Technology for theU.S. Department of Energy back in the late 1970’s. JPL was asked to develop a set oftests to demonstrate module durability and reliability. Over a period of a few years, thetests evolved with modules being obtained from U.S. manufactures in “block buys”, sothe tests became know at the “Block I Test”, then the “Block II Test”, and so on. Thelast version of the JPL test design was Block V, and these are detailed in JPLDocument No. 5101-162, “Block V Solar Cell Module Design and Test Specification forResidential Applications - 1981”. The basic test design developed by JPL has beenmodified by various organizations around the world.

The Commission of European Communities (CEC) and the International EuropeanCommunity (IEC) have developed updated qualification tests based on the initial workdone by the Jet Propulsion Laboratory. The latest version of the CEC test is CEC 503,which supersedes CEC 502 and 501. To “pass” modules may not exhibit greater than5% loss of electrical power in any test.




CEC Tests Summary:

UV Exposure:Modules are exposed to 15 hours continuous of 1000 watts/m

2 of UV light, to

accelerate any degradation of the encapsulant plastic.

Damp Heat:Modules subjected to 1000 hours continuous of 85% relative humidity at 85

oC. This

simulates severely humid tropical environments and will reveal any weaknesses tomoisture penetration.

Thermal Cycling:Modules are heated to 85

oC then cooled to -40

oC for 200 cycles. This is a greater

temperature swing than any real module on the earth would experience. This tests foreffects of expansion and contraction of contacts and lamination.

Humidity-Freezing:Modules are thermal cycled from 85

oC to -40

oC in 85% relative humidity for 10 cycles.

This tests for ingress of moisture that could result in corrosion and deterioration of thelaminating plastics.

Mechanical (Wind) Loading:Modules are mounted in a frame and subjected to 50-pound/square foot (psf) or 244kg/m

2, equivalent to 125 mph (200 km/h) winds, on alternate front and back for 10,000

cycles. This is to simulate severe wind conditions and to check for loosening ofcontacts and possible cell breakage.

Twisting:Modules are fixed at three corners and the forth is lifted approximately 1" (3 cm), toagain simulate wind conditions and torquing, to check for cell breakage and loss ofelectrical contact.

Hail Stone Impact:Ice balls are shot directly at the module glass front surface from a cannon usingcompressed air. Various impact points are chosen, including above cell edges, thecenter of the glass, and above interconnects. Siemens Solar modules withstand 1 1/2"hailstones traveling at 63 mph in this test (4cm diameter at 100 km/h). In a real

hailstorm, many stones would impact the glass at angles, and would glance off.

High Electrical Potential:3000 volts DC is applied between the cell circuit and the metal frame to test for anycurrent leakage. A maximum of 50 microamps is allowed to flow at room temperature.




UL Approval of Siemens Solar Modules

Siemens Solar modules have qualified for the Underwriter’s Laboratory (UL) listing(Figure 7-16). This listing assures consumers as well as inspectors, installers, lenders

and other professionals that the module design is safe. A wide variety of tests wereperformed to insure that the modules are safe for consumers. Some are presentedbelow:

Construction Integrity:Insulating materials, current carrying parts, internal wiring wireways, assembly andinstallation factors, connecting means, bonding, material compatibility, spacing, wiringcompartments, corrosion resistance, sharp edges, accessibility, fire resistance andencapsulation.

Performance Tests:Temperature, voltage and current measurement, leakage current, strain relief, dielectricvoltage withstand, inverse current overload, installation/maintenance, impact, fire,exposure to water spray, accelerated aging of gaskets and seals, temperature cycling,humidity cycling, metallic coating thickness, hot spot endurance, arcing, mechanicalloading, junction cover crush resistance.

Underwriter's Lab Tests

• Consumer and Inspectorconfidence in UL label

• Construction Integrity – materials used, safe design, fire

resistance

• Performance Tests – impact, inverse current, water,

crushing, gaskets

Siemens Solar IndustriesCamarillo, CA 93011

MODEL M55PHOTOVOLTAIC MODULEAT 1000 W/M SOLARIRRADIANCEAND 25 C CELL TEMPERATURE

2o

MAX. POWER SHORT CKT. RATED 53 WATTS 3.35 A 3.05 A

MAX. SYST. OPEN CKT. V. OPEN CKT. RATED 600 VOLTS 21.7 V 17.4 V

FIRE RATING SERIES FUSE CLASS C 5 A

FIELD WIRING BYPASS DIODECOPPERONLY, 14 AWG MIN. INSTALLATION GUIDE

INSULATED FOR75 C MIN. 233-701500-20 MADE IN U.S.A.

30B9 LISTED




U.S. Coast Guard Tests

The US Coast Guard has developed its own rigorous qualification test program detailed

in Specification 401 (III), “Solar Photovoltaic Arrays for 12 volt DC Marine Aids toNavigation”. Some are variations of the JPL type of tests, while some, like the “PIT”test, are unique to the USCG. Special modules (MAR 10, 20, and 35 modules) aremanufactured by Siemens Solar to meet these extremely tough requirements:

Pressure Immersion Temperature Cycling (PIT):Modules are immersed in salt water and alternately pressurized from 0-5 psig (0-3500kg/m2). The water temperature is cycled from 40-50

oC. to 3-9

oC. for 500 cycles. This

is a severe test of a module’s ability to handle operating in harsh marine environments,where navigational buoys might be tossed by large waves.

Shock:Modules are lifted at one edge to a height of 4" (10 cm) and dropped.

Twisting:This twist test is similar to the CEC test where one corner is lifted approximately 1”(3cm).

Steel Ball Impact:This is a more severe test than the ice ball impact test, where a 2.36 oz. 1" diametersteel ball (67 gm 2.5 cm diameter) is dropped onto the module glass front from a height

of approximately 3 feet (1 meter).

Termination Robustness:The module is suspended by its electrical output cable.

Temperature Shock:Modules are cycled 3 times between 71

oC. and -57

oC. with a transition time of less than

5 minutes. This is more severe than the gradual temperature transitions of the thermalcycling described earlier, and would simulate modules mounted on offshore buoysbeing tipped into freezing arctic waters.

Vibration:Modules are subjected to varying frequency of vibrations from 5 to 200 Hz for 84minutes along each of the three axes.




Specific Features and Benefits ofSiemens Solar Module Technology

The entire goal of the manufacturing engineering done at Siemens Solar is to produce amodule that can last in the real world. The preceding discussion of modulemanufacturing and testing cannot be of value unless it is connected to benefits to thefinal user of the modules. Some of the ways that the technical features of the modulescan result in direct user benefit are listed below.

Single Crystal (Czochralski Method) Silicon Cells

The extremely quick output response to even dim light means high efficiency duringdawn and dusk and in overcast environments, and superior performance during themost critical time of the year. This superior performance leads to lower cost per usableamp-hour delivered to loads.

Textured Cell Surface

Tiny pyramids etched on the cell surface scatter all wavelengths of light, and increaseabsorption and reduce reflection inexpensively across the full solar spectrum. Thisfeature is unique to single crystal lattice materials, as polycrystalline surfaces do nothave uniform lattice orientation and therefore cannot achieve uniform pyramidal

reflection enhancement.

Anti-reflective Coating on Cells

Multiple layers of thin metal oxides cover the front surface of each cell, to moregradually change the index of refraction from glass to silicon, thereby decreasingreflection and increasing light absorption and efficiency. The Siemens Solar TOPStm

process is our proprietary method for optimizing and matching both a texturized surfaceand an antireflective coating, which produces a superior absorbing surface.

Emissive Back Surface

A final back layer of highly emissive white Tedlar minimizes module and celltemperature, and therefore maximizes cell output.




Choice of Single or Separate Junction Boxes

The SM-series of modules have separated positive and negative junction boxes, whichmakes wiring easy and reduces the possibilities of wiring errors in the field.

The SP-line modules (and SM55-J module) have a large single junction box thataccepts either conduit or single conductor cable pushed through a waterproofelastomer seal. The large J-box contains dual bypass diodes that are attached withscrews for easy replacement, as well as allowance for a screw-in blocking diode. Allstainless steel screw terminals and two isolated floating terminals insure easy andreliable connections.

6 or 12 volt Configuration

The new large single J-box allows for field adjustable operation at either 6- or 12-volts.A wiring diagram is molded into the terminal cover for permanent reference.

White Background Around Cells

The space between cells is filled with an extremely bright white background. Light thatdoes not hit a cell is scattered by this specular surface and is internally reflected withinthe glass until it reaches a cell edge. Thus light that might not have been used isconverted into current. This is more efficient than other module designs with a blue orless reflective surface.




Cell Crystal Planes Oriented at 45o Angle to Cell Edges

The single crystal structure is oriented so that the orthogonal crystal planes are notparallel to the cell edges, but are angles at 45

o. Any small cracks will probably

propagate under the dual interconnect ribbons and not cut off an edge section from

contributing power. This reduces the chances of local hot spots and increases thereliability of output over a long module life. Polycrystalline cells cannot be oriented inthis way.

External Grounding Screw

Positive grounding of module frame is easy and increases chances of survival inlightning strikes.

Dual Cell Interconnect Ribbons

Two solder coated copper ribbons fully span both the top and bottom of each cell. Thisinsures cell interconnection even when a cell is cracked or when glass is shattered.

Cells Soldered at Multiple Points

Interconnect ribbons are not solidly soldered to the silicon but have multiple contactpoints. This reduces tension from the difference in thermal expansion of silicon andcopper, and increases the life expectancy of the module.

Tempered Water-White Low Iron Glass

Maximum protection and maximum transmission of light are possible. The glass doesnot appear green when viewed on edge, due to extremely low amounts of iron thatwould scatter and absorb light.

Glass Roughness (Stipple) Faces Inside

The rough surface from the tempering is placed on the inside, so that a smooth surfaceis presented to the environment. This allows for the natural cycles of wind and rain toefficiently clean the module, virtually eliminating the necessity of cleaning.

Junction Box Lid Attached to Junction Box

A strong connecting strand keeps the lids with the modules, preventing loss duringinstallation or maintenance.




Dual Bypass Diode Protection

Some manufactures allow for only one bypass diode to be installed in their junction box.All Siemens Solar power modules have two bypass diodes installed at the factory. Thenumber of series cells “removed” from the circuit in the case of local shading is not the

full 36 cells in one module, but is limited to 24 cells in the SM-series and only 18 cells inthe SP-series of modules. This reduces the heating of any shaded cells, and preservesarray output power even better during periods of local cell shading.

High Dielectric Polymer

Current leakage from the cells to the metal frame is limited well below 50 microampseven at 3000 volts, for added safety in high voltage arrays.

Back Contact Pattern is a Grid

An open grid of silver lines gives equivalent conductivity to a solid back contact, butreduces cell cost and allows for unused long wavelength light (infrared) to pass through,reducing module temperature and maximizing voltage.

UL Approval

The Underwriter's Laboratory approval gives extra assurance to developers, installers,lending institutions, and customers that the technology is proven and safe.




Modules Are Extremely Reliable

Perhaps the most essential benefit of using photovoltaic power is high reliability. The

technical features described earlier combined with high overall manufacturing qualityresult in an extremely reliable product, as our module return experience illustrates:

Module Return History

• Solar modules are extremely reliable!

• During 1983-1992 181,611 SM50-H sold

• Total installed time 637,558 years

• Total returns 385 modules

+ Failure rates: .069 failures/ million hours.0006 failures/ year

An analysis was done in 1994 for our popular SM50-H module. The total returns due tomodule failure were accumulated for a ten year period, and compared to the total hoursof operation worldwide for all the SM50-H modules shipped and installed during thattime. With only 385 modules returned for failure during the ten years, and a totalestimated operating time of over 637,000 hours for all the 181,000 modules shipped,the failure rate calculates to be only .069 failures in one million hours of operation! Orabout one failure in 14 million hours!

It must be emphasized that a complete power solution usually involves more than justmodules. Batteries, charge regulators and control systems, fuses and circuit breakers,diodes, and of course load devices are involved. All of these other components havefailure rates higher than solar modules, and care must be taken when designing andinstalling photovoltaic power systems to minimize their effect on overall reliability.




Exercises




World Class Module Specifications

The previous discussion illustrates the extent that Siemens Solar seeks to achieve the

highest quality products for the world market. We seek to set the standard for solarmodule performance and quality. Many of you will be involved in the process of writingor reviewing proposals and tenders for solar equipment. We would hope that youwould apply the highest standards to module manufacturing, testing and performancemeasurements.

Below are listed the various standards of performance and testing that Siemens Solarmodules have met. They should be the standard for consideration to insure a worldclass project or system!

Include In Tender Specifications...

• Performance tested in accordance with IEC904 and ASTM E1036

• Pass environmental tests required by JPL5101-161 (Block V)

• Meet UL 1703 requirements and listed underUL file No. E79555

• Qualification certificate issued by CEC JointResearch Center, Ispra Establishmentregistration No. PV-MT-503-83/95 according torequirements of CEC specification No. 503




(END OF CHAPTER)




CHAPTER SEVEN

MODULE MANUFACTURING AND TESTING 7-1

Wafer Manufacturing 7-2

Cell Manufacturing 7-7

Module Manufacturing 7-11

Qualification Testing of Module Designs 7-14

CEC Tests (JPL Block V Tests) 7-15UL Approval of Siemens Solar Modules 7-17U.S. Coast Guard Tests 7-18

Specific Features & Benefits of Siemens SolarModule Technology 7-19

Modules Are Extremely Reliable 7-23

World Class Module Specifications 7-25



Siemens Solar Basic Photovoltaic Technology 7-1 Module Manufacturing and Testing

Chapter 7 – Answers Module Manufacturing & Testing

The incorrect steps are:f. 20% of all modules made are tested for output power (100% of the modules are

tested for output power) j. Boron is added to each wafer before printing (boron is added when the silicon is

melted to make crystals)

The proper order for the steps is:l - d - b - m - g - i - a - e - c - k - h

Condition Test Another test perhaps

A. Modules in a moist tropical climate h g, aB. Modules bobbing on buoy in ocean e c, f, g, hC. Modules on alpine mountain top g d, gD. Modules in large 450 VDC array a d, gE. Modules in thunderstorm area b c, d

Note: other combinations are possible.



Siemens Solar Basic PV Technology Course Components – Output CurvesCopyright © 1998 Siemens Solar Industries

8-1

Chapter EightOutput Curves

The single most important technical aspect of photovoltaic cells and modules is thecurrent-voltage output curve. Understanding the curve allows a system designer toanticipate how a module will be influenced by the environment, and how a particularload will interact with the module. The output curve also shows how the moduleoutput is "bounded" and therefore safer than battery or generator in the case of shortcircuits. As you read this chapter sketch the curve often, label points on the curve,and sketch how the curve might be influenced by the environment. This will help inunderstanding sizing and wiring concepts presented later in this book.




8-4

Module Output Curves

Most photovoltaic module manufacturers produce a variety of modules with differentvoltage and current characteristics to suit different markets and customers. As anexample, the standard module I-V Curves for modules made by Siemens Solar are

shown below.

SM Modules 1000 w/m 2 and 25 deg.C.

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

0.00 5.00 10.00 15.00 20.00 25.00

Voltage (volts)

C u r r e n t ( a m p

s )

SM6 SM1 SM20

SM55SM50-H

SM46

It is the number of solar cells connected in SERIES that determines the maximumvoltage of a module. Each silicon solar cell can produce a maximum of about 0.6volts at full sunlight. So 36 cells in series will result in a maximum voltage of about21 volts. Some modules are made with 33 cells for a maximum of about 19 volts.And modules made with only 30 series cells will produce a maximum voltage of only18 volts.

And it is the SIZE (area) of a single cell that determines the maximum currentoutput. The Siemens Solar modules are made from two basic sizes of cell. The

square cell used in the SM-series of modules is about 104 mm in across, andproduces a maximum current of about 3.4 amps. The larger cell used in the SPseries of modules is about 125 mm across and produces about 4.8 amps maximum.




8-5

To create a module that produces less current than the standard cell, cells are cutinto half or quarter, to give 1/2 or 1/4 the standard current. This is the case for theSiemens Solar SM20, SM10, and SM5 (with 1/8 cell size). And the Siemens SolarSP36 and SP18 are made of 1/2 and 1/4 cells cut from the larger SP75 cells.

SP Modules 1000 w/m 2 and 25 deg.C.

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

0.00 5.00 10.00 15.00 20.00 25.00

Voltage (volts)

C u r r e n t ( a m p s )

SP75

SP36

SP18




8-6

Exercises

Manufacturerand model

Isc Voc Imp Vmp Pmax # SeriesCells

Cell Size




8-7

Effect of Environment onModule Output Potential

The IV curve is really just a "snapshot" view of the potential output of a photovoltaicdevice under static environmental conditions. If the environmental conditions arechanged, the output potential of the device changes. The three main environmentalconditions that we will examine are:

• Light Intensity or Irradiance

• Cell Temperature

• Light Spectral Content

Effect of Light Intensity (Irradiance)When the intensity, or irradiance level, of light changes the number of photonsentering the PV device changes, and the number of electrons released changes. Sothe direct result of a change in light intensity is a change in output current at allvoltages.

Effect of Light Intensity (Irradiance)

Lower irradiancereduces current

Voc drops

slowly withlowerirradiance




8-8

The Isc is DIRECTLY PROPORTIONAL to the light intensity, and the Voc variesmore slowly in a logarithmic relationship. In other words, the ratio of the Isc to thelight intensity or irradiance will be the same. If the light intensity is halved, the Iscwill drop to half.

Isc1 = Isc2 Light intensity1 Light intensity2

Example: A module has a rated Isc of 3.4 amps at 1000 w/m2. We can calculate

the Isc at another light level by taking the proportion of the new lightlevel to the standard level of 1000. If the actual light intensity on amodule were 850 w/m

2, then the actual Isc would be given by the

proportion:


3.4 amps = Isc2 1000 w/m

2850 w/m

2

Isc2 = 3.4 X ( 850 / 1000 )

= 2.9 amps

Standard Condition for Irradiance

A standard condition of light intensity has been established so that PV device outputcan be compared. The internationally accepted standard is 1000 watts/m

2 (or 100

mw/cm2). This is called "one sun" or "peak irradiance".

Standard Condition for Irradiance = 1000 watts/m2




8-9

Circumstances For Exceeding the“Standard”

Most locations will measure less than this at noon, but many will measure more than

this. Sometimes white clouds will reflect more light onto the module surfaceproducing irradiance values of 1400 w/m2 or higher.

Measurements at high altitudes may be greater than the standard 1000 w/m2

because less air is between the sun and the surface, so more light gets through.Sometimes ground reflectance from white sand or rocks or reflective building orwater surfaces will also raise the total intensity of light on a surface to greater thanthe standard 1000 w/m

2.

Clear Day

1000 w/m

Cloudy Day

1000 w/m




8-10

Effect of Temperature

As the cell temperature rises, the main effect is to reduce the voltage available atmost currents. There is a slight rise in current at very low voltages. The change involtage is DIRECTLY PROPORTIONAL to the rise in temperature. It is important to

note that these factors refer to cell temperature, not just ambient or air temperature.The relationship between air temperature and cell temperature depends on lightintensity, and is discussed later in this chapter.

Effect of Temperature

Higher temperature

reduces voltage

Isc rises slightly as temperature goes up

Manufacturers of modules anticipate the lost of voltage in real world hot conditions,and compensate by building modules with enough cells in series so that even when

very hot, the module has enough voltage to charge batteries or operate the loaddevice.

The general formula for determining the change in voltage with temperature is givenon the next page. The effect of temperature on Voc for one cell is multiplied by theamount of temperature change and the number of cells in series. This will give thevoltage change for the whole module. Subtracting this voltage change from thevoltage at one temperature gives the voltage at another temperature.




8-11

Example Cell Temperature Coefficients:

Siemens Solar125 mm cell (140 cm

2)

Siemens Solar103 mm cell (104 cm

2)

Parameter Changeper

oC

% Changeper

oC

Changeper

oC

% Changeper

oC

Voc -2.15 mV -0.36 % -2.15 mV -0.36 %

Vmp -2.18 mV -0.44% -2.19 mV -0.45 %

Isc 2.06 mA -0.04% 1.20 mA -0.04 %

Imp -4.37 mA -0.10% -3.23 mA -0.10 %

Pmax -9.53 mW -0.45% -7.08 mW -0.47 %

Example Module Temperature Coefficients:

Parameter SM55 SP75 SM50-H30(33 cells)

Voc -0.077 volt /oC -0.077 volt /

oC -0.071 volt /

oC

Vmp -0.079 volt /oC -0.078 volt /

oC -0.072 volt /

oC

Isc 1.20 mA /o

C 2.06 mA /o

C 1.20 mA /o

C

Imp -3.22 mA /oC -4.42 mA /

oC -3.25 mA /

oC

Pmax -0.255 watt /oC -0.345 watt /

oC -0.234 watt /

oC




8-12

Calculating Voltage Changes

Voltage Change = Voltage Loss Factor X Temperature Change

and

Voltage2 = Voltage1 - Voltage Change

Example: A module has 36 cells in series for a Vmp of 17.3 volts at 25oC. In real

world conditions, the cells will easily heat up to 50oC. Using the factors

presented, and multiplying by 36 cells and 50 - 25 = 25oC.

temperature difference, we get

Vmp Change = Factor X Temperature Change

= -0.079 v/ oC X 25

oC

= -1.98 volts

Vmp2 = 17.3 volts Vmp (at 25oC) - 1.98 volts

= 15.3 volts Vmp (at 50oC)

which is still enough to fully charge a typical "12 volt" battery thatactually needs up to 15 volts to reach full charge.




8-13

Calculating Power Changes

The overall reduction in maximum power (Pmax) is a general factor to use toestimate the effect of temperature on output power, not just current or voltage. Thepower reduction factor for solar modules as indicated in the table presentedpreviously is generally about -0.45% to -0.5%, and this includes changes in current,

voltage, and the curve shape. In other words, a cell or module will loose a little lessthan 1/2% of its peak power with every degree in temperature rise.

Typical Maximum Power Reduction Factor = - 0.45 % /oC

The formula for determining the effect of temperature on overall device power isgiven below. Use the overall reduction in Pmax factor given above and multiply bythe amount of temperature change. This will give the percentage of power change.

% Power Change = Pmax Reduction Factor X Temperature Change

Example: A module is rated at 55 watts Pmax at 25oC. If it is operating outdoors

and heats up to 50 oC., then the Pmax will be reduced.

% Power Change = -0.45% / oC X (50 oC - 25 oC)

= -.45% /oC X 25

oC

= -11.25% change

Pmax2 = Pmax1 X (1 - % Power Change) 100

= 55 watts Pmax (at 25 oC) X ( 1 - 11.25 )

100 = 48.8 watts Pmax (at 50

oC)




8-14

Standard Condition for Cell Temperature

Since PV device output is affected by cell temperature, a standard cell temperatureof 25 oC has been accepted internationally. Stating the output of modules and cellswith reference to this common temperature allows for proper comparisons to be

made.

Standard Condition for Temperature = 25 oC

Temperature Effect Depends on the

Number of Series CellsThe factors given previously help to calculate the change in Voc (and Vmp), but theydon’t really help to understand the effect temperature has on module output into abattery. Battery charging is the most common use for photovoltaic modules, and it isthe battery voltage that determines where on the IV curve the module operates. Anaverage battery voltage of 13.5 volts can be used to illustrate the point.

The number of series cells determines the final module voltage and the voltage ofthe “knee”. The difference between a 33-cell module and a 36-cell module becomes

critical when we look at the effect of temperature. The IV curve is reduced in voltageat higher temperatures. This reduction causes the “knee” of the curve to moveinward, and since the battery is holding the module operating point to around 13.5volts, the operating point actually moves downward along the “knee”. The amount ofdecrease in current depends on how many cells in series.




8-15

~13.5 volts average battery voltage

SM5536 cells

0.00

1.00

2.00

3.00

4.00

0.00 5.00 10.00 15.00 20.00 25.00Voltage

C u r r e n t ( a m p s )

25 deg. C.

47 deg. C.

65 deg. C.

SM50-H33 cells

0.00

1.00

2.00

3.00

4.00

0.00 5.00 10.00 15.00 20.00 25.00

Voltage

C

u r r e n t ( a m p s )

25 deg.C.

47 deg. C.

65 deg C.

We can see in the figures that the 36-cell module (Siemens SM55 in this example)are less affected by high operating temperatures than the 33-cell module (SiemensSM50-H). The 36-cell module drops from 3.3 amps at 25

oC to only 3.0 amps at 65

oC, while the 33-cell module drops from 3.25 amps at 25

oC to less than 2.5 amps at

65oC. However the 33-cell module drops to only 3.0 amps at a more moderate 47

oC. Thus we see that a 36-cell module is designed to operate well in the hottest of

climates, while a 33-cell module performs well in more moderate temperatures.




8-16

Exercises

2

2

2

2

2

2




8-17

Spectral Content of Light

The spectral distribution of light refers to how much of the light's energy is deliveredat each different wavelength or color. The spectral distribution of the sun's lightabove the atmosphere is nearly that of a "black body" or perfect radiator at 6000

oC.

The greatest output from the sun is in the visible range of wavelengths (roughly fromviolet at 380 nm to red at 750 nm), but there is substantial energy output in shortwavelength ultra-violet (UV) and in long wavelength infrared (IR) and radio waves.

As the light passes through our atmosphere certain wavelengths are scattered andabsorbed more than others by air, moisture and aerosols. For example, the largedips in the spectral distribution graph shown are due primarily to absorption by watervapor. The proportion of reds, greens, blues, etc., will change depending on thethickness of atmosphere that sunlight has to penetrate. That thickness varies hourlyas the sun rises in the sky. The thickness of atmosphere light must penetrate iscalled the Air Mass (AM). If the sun were directly overhead, light would pass through

1 thickness of atmosphere, or 1 AM to reach the earth's surface. In the afternoon ormorning, the sun is at a lower angle, and the distance light must pass through mightbe 2 or more times the noon time thickness (i.e.. 2 AM or more).

Effect of Light Spectrum

Sun at noon

Earth

Atmosphere

Sun at

mid-morning

1.5 Air Masses

( AM 1.5 )

One Air Mass .

(AM 1 )




8-18

A standard set of atmospheric conditions has been established for the photovoltaicsindustry by NREL, the US National Renewable Energy Laboratory (formerly SERI,Solar Energy Research Institute) in Golden, Colorado, for the U.S. Department ofEnergy. This standard is listed as ASTM #E892, and among other factors is aspectral distribution that is equivalent to light passing through 1.5 thickness ofatmosphere, or “Air Mass 1.5” (AM 1.5).

Solar Spectral Distribution = ASTM #E892 or “Air Mass 1.5”

By standardizing the amount of atmosphere under which measurements of cell andmodule output are made, we also fix the proportion of red, green, blue, etc. light, orthe spectral distribution. This standard is necessary because photovoltaic cells

respond differently to different wavelengths of radiation. (Recall the discussion ofspectral response in the chapter on Photovoltaic Physics). If a CZ cell wereexposed to blue light and red light of the same total energy, the current producedwould be higher for the red light. The standard spectral distribution of AM 1.5 givesa common reference condition so modules from different manufactures and ofdifferent construction can be fairly compared.




8-19

Summary of Standard Conditions

Because the electrical output of a photovoltaic device is so affected by theenvironment, we must standardize the conditions under which they are measuredand compared. Once those conditions are set, the output potential of the device is

absolutely determined.

The three standard conditions that are used to specify the environmental conditionsfor photovoltaic devices are summarized below.

Light Intensity or Irradiance = 1000 watts/m2

Cell Temperature = 25oC

Solar Spectral Distribution = ASTM #E892 or“Air Mass 1.5”




8-20

Exercises




8-21

Other Rating Conditions

The relationship of module output as rated under Standard Operating Conditions(SOC) and the actual output that a user would be able to measure under realoutdoor conditions is very poor. Typically the irradiance level will be lower and thecell temperature will be higher than SOC. Other test conditions have been defined

over the years to try to account for this discrepancy. Some of these sets ofconditions are described next.

Other Rating Conditions

• Nominal Operating Cell Temperature (NOCT)

– Irradiance 800 w/m2

– Ambient Temperature 20 oC

– Wind Speed 1.0 m/s maximum

– Module at Open Circuit

• Standard Operating Conditions (SOC) – Irradiance 1000 w/m2

– Cell Temperature NOCT

• Nominal Operating Conditions (NOC)

– Irradiance 800 w/m2

– Cell Temperature NOCT

Nominal Operating Cell Temperature

The NOCT is not a condition but rather an empirically determined value for a specificmodule design and construction. The NOCT is intended to represent the typical real

cell operating temperature for a module. The actual temperature will depend on thethermal characteristics of the cell and the module packaging. For example, glass-on-glass modules will typically have a higher NOCT than white background modulesdue to the greater thermal mass of the two layers of glass. Blue backgroundmodules may have a slightly higher NOCT than white background modules due toabsorption. Generally a module manufacturer will seek to have as low an NOCT aspossible.




8-23

Defining Module Output EfficiencyOne common measure of the quality of a solar cell or module is the efficiency. Butthere is not just one way to define “efficiency” for solar devices. This should bemade clear, because statements are often made comparing manufacturer’sefficiencies, and it can be deceiving if you are not clear on exactly what “efficiency”

is being quoted.

In general, efficiency is defined as the ratio of output from a device compared to theinput to the device.

Efficiency = OutputInput

Total Area Efficiency

This definition involves the ratio of maximum electrical power output compared to thetotal light power incident on the ENTIRE device. The device area includes moduleframe, interconnects and gridlines on the surface of the cells. This is the "real world"efficiency of a module. This value indicates what you really have to work with.

Total Area Efficiency = PmaxTotal Device Area X Input Light Power

Example: Calculate the total area efficiency of the following module:

Pmax = 55 watts Length = 1.293 mWidth = 0.330 m

Total Area Efficiency = 55 watts1.293m X 0.33 m X 1000 w/m

2

= 0.129 or 12.9%




8-24

From the example, we see that the overall efficiency of the example module is12.9% under full sun conditions. In other words, of the total solar radiation that fallson the total module are (including all the inactive areas like frames, inter-cell spaces,and gridlines) 12.9% of that energy is output as electrical power. This may seemlike a low figure, with about 87% of the incident power not being converted, butremember that the fuel for the electrical power is free sunlight. This efficiency ranksamongst the highest module efficiencies in the world, and is the direct result of the

single crystal technology used in the Siemens Solar cells.

Aperture Area Efficiency

Another way to calculate efficiency is called “aperture area” efficiency. This refers tothe module efficiency calculated without any frame dimensions. This would includeall area inside of the frame, including any gridlines or cell interconnects. Thismodified way to calculate efficiency removes the effect of a large frame, which reallydoesn’t reflect the quality of the cell efficiency anyway.

Active Area Efficiency

This definition involves calculating efficiency based only on the area of the devicethat is exposed to light or “active” semiconductor. Light incident on shaded areaslike interconnect wires, cell gridlines, and frame area is not included. It is alwayshigher than the total area efficiency will be for a completed device withinterconnected cells. Active area efficiency is often announced by research groupsor by manufacturers and confused with total module area efficiency. It should beunderstood to refer to fundamental physics of a single cell only, and not to finalmanufactured modules where cells will be covered with gridlines and interconnectwires.

Active Area Efficiency = PmaxActive Area Only X Input Light Power




8-25

Exercises

Module Manufacturer and Model Total Area Efficiency (%)




8-26

Combining Cells and CurvesSolar cells can be combined in series and parallel to increase voltage and current.When cells are connected in series, the current flow is the same through each celland the resulting voltage is the sum of the voltages of each cell. When cells areconnected in parallel, the voltage across each cell is the same and the currents add

to produce a final current.

The exact output curve for a combination of cells can be created by adding thecurves for all the single cells. For series connections, the current flowing througheach of the cells is the same. So at any current level, the voltage of one cell isadded to the voltage of the next. The output curves are added horizontally orvoltage-wise, as shown. For parallel connections, the voltage across each cell is thesame and the currents flow together. So at every voltage value, the current of eachcell is added. The curves are added vertically or current-wise, as shown.

Combining Cells To Make A Module

• Series connected cellsincrease voltage potential

+=

+ =

+ =

+ =

• Parallel connected cellsincrease current potential




8-27

A solar module is a collection of cells connected in series and sometimes in parallelto produce a basic building block with enough voltage to do useful work. The mostcommon load is a "12 volt" battery, which I place in quotes because it actually needsapproximately 14-15 volts to be fully charged. So most modules are made ofenough cells in series to produce at least 14.5 volts at maximum power, to be able toeffectively charge the batteries. To achieve this voltage 30-36 single crystal siliconare needed in series. A composite curve for a 36-cell CZ module is shown. The IV

curve for a single cell is added along the voltage axis 36 times. The final IV curvefor the entire module reaches out to a Voc of more than 21 volts.

Cells Combine To Make OutputCurve For Module

3.4 amps

...36 cells in series...

21 volts0.6 voltseach cell

If the main application for the modules were not 12-volt battery charging, then someother number of cells might be more appropriate. For example, if the ultimateapplication was to be direct connection to utility power at voltages of 120 or 240volts, then there would be no reason to keep the voltage of the modules around 12volts. Modules with a Vmp of 40 or 60 volts could be designed. The ultimateapplication voltage determines the number of cells needed in series. The mostcommon market for large power photovoltaic modules is still 12 volt battery charging,so that is why most manufacturers produce modules with 30-36 CZ cells in series.




8-28

If more voltage or current than one module can produce is needed, modules canalso be connected in series and parallel to achieve practically any final voltage andcurrent. Systems producing 600 volts DC and hundreds of amps have beeninstalled, and successfully operated for years.

Combining Modules To MakeAn Array

6.8 amps

84 volts

For example, in the drawing above there are 4 modules connected in series to makea nominal 48-volt string, and two strings are connected in parallel to increase currentoutput. The composite curve for the entire 8-module array can be constructed byeither adding voltage-wise first and then adding current-wise, or visa versa.




8-29

If we add voltage first, the 4 modules in series are added voltage-wise to reach outto beyond 80 volts at Voc, and then the 2 parallel strings are added current-wise toshow that the current adds.

If we use the values calculated previously for a 36-cell module, the resulting curvefor the entire array has values given below:

New parameters for array of four series X two parallel modules

21.7 volts Voc (module) X 4 series modules = 86.8 volts Voc (array)

17.3 volts Vmp (module) X 4 series modules = 69.2 volts Vmp (array)

3.4 amps Isc (module) X 2 parallel modules = 6.8 amps Isc (array)

3.05 amps Imp (module) X 2 parallel modules = 6.1 amps Imp (array)




8-30

Exercise

a. b. c. d.

0

2

4

6

8

10

12

14

16

0 20 40 60

0

1

2

34

5

6

7

8

9

10

0 20 40 60 80

0

2

4

6

8

10

12

14

16

0 20 40 600

1

2

34

5

6

7

8

9

10

0 20 40 60 80

0

2

4

6

8

1 0

1 2

1 4

1 6

0 1 0 2 0 30 4 0 5 0

s ing le ce l l

(1 )

(2 )

(3 )

(4 )




8-31

Determining Output Over Time

The IV curve, adjusted for environmental conditions, is only one half of theinformation required to determine what a module will output. The other criticalinformation is the LOAD operating characteristics. It is the load that determines

where on the IV curve the module or array will operate, not the other way around.Different types of loads will interact with the PV device in different ways, and theoutput of a module under the same environmental conditions will vary depending onthe load.

Three factors determine output potential: Cell Temperature

Irradiance

Spectral Distribution

Forth factor determines actual output: Load Interaction

An IV curve gives only a snapshot of potential output of a PV cell or module or arrayunder static conditions of light, spectrum and temperature at a given instant. Amodule in the real world will experience changing conditions throughout a day. A

more relevant measure of module output comes from measuring or predicting thevarying output over a typical day. And the output is determined by the load.

The three basic types of loads that photovoltaic devices interact with are:

• Maximum Power Tracking Device

• Battery

• DC Motor

How these different types of loads interact with a solar module and determine theactual output are discussed next.




8-32

Cell Temperature Is Affected By Irradiance

The cell temperature, and therefore the voltage available from the PV device, isrelated to light level or irradiance. The greater the irradiance, the more energy isbrought to the module and the greater the internal heating. The exact relationship

between irradiance and cell temperature depends on the module construction. Thethickness of glass, the emissivity of the back layer, the darkness of the cells -- all thishas an impact on the final cell temperature. The relationship between light level andcell temperature for Siemens Solar modules is shown as an example. The rise incell temperature above the ambient air temperature is linear with irradiance.

Cell Temperature Affected by

Irradiance

0

10

20

30

40

0 200 400 600 800 1000 1200

Irradiance (Watts/m2)

T e m p e r a t u r e

( D e g C )




8-33

Knowing the air temperature and the irradiance level, the cell temperature can bedetermined, as given below.

Cell Temperature = Ambient Temperature + Temperature Rise

Example: A 36-cell module is rated at 25oC. (the standard temperature discussed

earlier). If in a real situation, the insolation level is 900 watts/m2, the

temperature rise is given by the graph as approximately 25oC. above

ambient. If the air is at 35oC., the cell temperature will be

Cell Temperature = 35oC (ambient air) + 25

oC (temperature rise)

= 60oC.

Once the actual cell temperature has been calculated, it can be used to predict howthe IV curve voltage potential is reduced. All computer programs used for arraysizing use some sort of adjustment for cell temperature to better predict actualmodule performance in the real world.

It is important to see that the irradiance and ambient temperature work together topredict the actual cell temperature in a module. For example, we just calculated thatif the ambient air temperature were 35

oC. and the irradiance was 900 w/m

2, then

the cell temperature would be 60oC. But what if the same module was located in a

cold alpine climate with an ambient temperature of only 5oC. The same irradiance

level of 900 w/m2 would result in the same temperature rise value of 25 oC., but thefinal cell temperature would be less.

Example: The same solar module in a cold alpine climate with only 5oC ambient

in 900 w/m2

Cell temperature = 5oC (ambient air) + 25

oC (temperature rise)

= 30oC

The cell temperature is much lower, and therefore the module voltage would be

much higher in the alpine climate than in the hot climate. So just knowing theirradiance on the module is not enough to predict module performance. Theambient air temperature must be known as well.




8-34

Exercise

2

2

2

2




8-35

Maximum Output Potential During a Day

By predicting how the air temperature and irradiance vary during a day, theinstantaneous output potential of a cell or module, as represented by the IV curvecan be accurately predicted for any time of the day. By continuously varying the

environmental conditions on the module, a series of IV curves can be calculated.Since power is a rate of delivery of energy, adding the maximum power possiblefrom a module (Pmax) over the day gives the maximum total energy that can bedrawn from a module during a day.

This might actually occur for modules connected to a utility interconnected inverterfor example. Usually utility interactive inverters have Maximum Power Tracking(MPT) circuitry that continuously adjusts the operating voltage of an array so that themaximum power is output at any instant. Such MPT devices are also popular fordirect coupled water pumping systems. The solar array is connected to the MPTdevice, and its maximum power is extracted and converted into power at the

optimum voltage for the pump at any instant.

The variation of Pmax for a 36-cell 55-watt module is shown over a typical summerday. The area under the curve represents the total maximum power that onemodule could actually deliver to a load. In this case, that is 253 watt-hours (Wh), orapproximately 1/4 kilowatt-hour (1/4 kWh).

The previous discussion has talked about IV curves and how they represent thepotential output at any instant under fixed environmental conditions. But the outputadded up over time is the energy that could go into a load during a day. The totalenergy available to a load is the value we have been seeking. If we know the total

energy needed by a load in a typical day, we can now determine how many moduleswould be needed to supply all that energy. For example, using the above figure of1/4 kWh from one module, if the daily load was 5 kWh/day for a remote application,then approximately 20 modules would be needed to produce all that energy. This isthe principle behind the array sizing that will be presented in the chapter on SystemSizing.




8-37

Exercise

Hourof Day

WattsFeb

WattsJune

4 0.815 2.046 3.917 9.288 0.48 16.279 5.60 22.27

10 10.79 26.3111 14.09 28.6512 15.18 29.361 14.00 28.512 10.67 26.063 5.51 21.964 0.47 15.985 9.086 3.87 2.08 .8




8-38

Graph the watts against time to show the output over a typical day.

Watts

35

30

25

20

15

10

5

0

4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8

Time of Day




8-39

Interaction with a Battery LoadThe output value given previously was from a module operating at its maximumpower point all day long. This can be achieved by using some sort of electronicmaximum power tracking circuitry, that will be described later in chapter on EnergyEnhancement Through Tracking. However, the most common load for a PV module

or array is a battery. A battery has a voltage, depending on its state of charge, age,temperature, rate of charge or discharge and other factors. This voltage determinesthe voltage of everything connected to the battery, including the module or array.Even if the system has motors, inverters, and other devices connected to thebattery, or even other loads connected directly to the array, the battery will set theoperating voltage for every component in the system.

It is sometimes thought that a module with a maximum voltage potential of 20 voltswill push a 12-volt battery to 20 volts and harm the battery. But it is the battery thatfixes the operating voltage of the module, not the other way around.

A detailed discussion of battery characteristics and performance is presented later inthe chapter on Battery Technology. For now, we only need to realize that as abattery is charged its voltage rises, and as it is discharged the voltage falls. Thevoltage of a battery varies within a fairly narrow range, usually between 11-15 voltsfor a "12 volt" battery, with average operation around 13-14 volts.

This is why modules are designed with up to 36 solar cells in series, giving a Vmp at25

oC of more than 17 volts. At first this may seem too high for a 12-volt battery. But

a module in a hot climate will loose approximately 2 volts due to heat, bringing theVmp down to approximately 15 volts. A "12 volt" battery actually needs about 14.5volts to reach full charge.

So it is the battery voltage that determines the operating point on the IV curve, andtherefore the charging current into the battery. As shown in the figure, the module orarray is affected by temperature and light level throughout a typical day. Since thebattery holds the module operating point within a narrow range of voltage around 13-14 volts, the resulting output will be seen as varying current into the battery. Theoperating voltage will rise slowly over a day as the battery is charged. But comparedto the wide changes in the module curve, the voltage of the battery changes onlyslightly.

As shown in the figure, the total energy into a battery is slightly less than the

maximum possible energy that could have been delivered. This is because thebattery held the operating point of the module at a lower voltage than the Vmp of themodule. Module manufacturers try to create modules that are good matches involtage to the ultimate load, in this case a battery. However, because ofenvironmental uncertainties and variations, the match of module Vmp and loadvoltage cannot always be perfect. This is a case where the match is close, but notperfect.




8-40

Output into Battery

0

5

10

15

20

25

30

35

40

4 6 8 10 12 14 16 18 20

Time of Day

Power (watts)

Maximum power possible

Power into battery

0 2 4 6 8 10 12 14 16 18 20

Voltage (volts)

Current (amps)

2.8

2

1

0

(B)

(C)

8:00 a.m.

3:00 p.m.

12:00 noon

(A)

Changes Over a Day (M55 in Phoenix)

247 wh




8-41

Exercise

Hourof Day

WattsApril

WattsJuly

6 1.37 7.4 8.18 20.3 18.99 32.2 28.0

10 39.0 32.1

11 41.7 33.112 41.6 32.71 40.1 31.42 37.2 29.33 31.0 25.74 19.9 17.75 7.1 7.66 1.0




8-42

Graph the watts against time to show the output over a typical day

Watts

45

40

35

30

25

20

15

10

5

0

4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8

Time of Day




8-43

Interaction with a DC Motor

A DC motor also has an "I-V curve", usually called a load curve. A typical curveramps up in current with increasing applied voltage. When overlaid on a module IVcurve, this defines a precise intersection point with the I-V curve. This intersection

point sets the operating current and voltage for the module or array. So the motordetermines the voltage and current output of the module or array.

A direct coupled motor system has the module connected directly to a DC motor andpump without any controls or battery. It has the advantage of the fewestcomponents and thus high reliability. But the system suffers in overall efficiency dueto mismatch each morning and evening. A direct coupled motor/PV array system isusually designed so that has the motor operates the modules near their point ofmaximum power during the middle of the day (see point B in the figure). But whenthe module current output is lower in the morning and evening, there will be a verypoor match, as shown in the figure. The motor operates the module at far lower

voltage than the module's Vmp (points A and C). So the motor may operate, but it isgetting less than the maximum possible power the module or array could deliver.

A maximum power tracking device can be placed between the array and the motorto force the array to operate at its maximum power point all day long. This would beone way to increase the efficiency of the overall system. Maximum power trackersare discussed more in the chapter on Energy Enhancement Through Tracking.




8-47

In real outdoor conditions, the light level would probably be lower than the standardlevel of 1000 w/m

2, and the cell temperature would probably be higher than the

standard temperature of 25oC. A 36 cell 36-watt module at outdoor conditions of for

example 800 watts/m2and 50

oC cell temperature has Vmp = 17 volts and Imp = 2.1

amps (approximate). Dividing these values will give a resistance that will operatethe module near its maximum power point in typical field conditions:

Resistance needed to operate at Pmax = Vmp(35 Pmax, typical outdoors) Imp

= 17 volts 2.1 amps

= 8.1ohms

So connecting an 8.1-ohm resistor to a 36-watt module in the field will operate themodule very near its maximum power point, to allow checking of output other than

just Voc and Isc.

0 4 8 12 16 20

Voltage (volts)

Current (amps)

3

2

1

0

Operating at Maximum Power

with Resistive Load

(Vmp, Imp)

R = Vmp/Imp




8-48

To calculate the size of resistor needed for an array of modules, multiply the aboveresistance value by the number of modules in series, and divide by the number ofmodules in parallel. The final resistance value will operate the array near itsmaximum power point.

Resistance array = Resistance module X # Modules in Series# Modules in Parallel

Example: You wish to do field diagnostics on a small array of four 36-wattmodules in series by two in parallel. What value of resistance wouldyou need to operate the array near its maximum power point in actualfield conditions?

Resistance array = 10 ohms x 4 in series

2 in parallel

= 20 ohms

The example above shows that to operate an array of four 36-watt modules in seriesby two in parallel, you would need a resistor of approximately 20 ohms. This isshown graphically below.

20-ohm resistor operates array near Pmax




8-49

Exercise:




8-50

(End of Chapter)




8-51

CHAPTER EIGHT

OUTPUT CURVES 8-1Basic Current-Voltage Curve (The “I-V Curve”) 8-2

Output Curve Terminology 8-3

Module Output Curves 8-4

Effect of Environment on Module Output Potential 8-7Effect of Light Intensity (Irradiance) 8-7Standard Condition for Irradiance 8-8

Circumstances For Exceeding the “Standard” 8-9Effect of Temperature 8-10Standard Condition for Cell Temperature 8-14Temperature Effect Depends on the Number of Series Cells 8-14Spectral Content of Light 8-17

Summary of Standard Conditions 8-19Other Rating Conditions 8-21

Defining Module Output Efficiency 8-23

Total Area Efficiency 8-23Aperture Area Efficiency 8-24Active Area Efficiency 8-24

Combining Cells and Curves 8-26

Determining Output Over Time 8-31Cell Temperature Is Affected By Irradiance 8-32Maximum Output Potential During a Day 8-35

Interaction with a Battery Load 8-39

Interaction with a DC Motor 8-43

Interaction with a Resistive Load 8-45



Siemens Solar Basic Photovoltaic Technology 8-1 Output Curves

Chapter 8 – Answers Output Curves

Isc(Amps)

Voc(Volts)

Imp(Amps)

Vmp(Volts)

Pmax(Watts)

SM55 3.45 21.7 3.15 17.4 55

SM50-H 3.35 19.8 3.15 15.9 50

SM46 3.35 18.0 3.15 14.6 46

SM20 1.6 18.0 1.38 14.5 20

SM10 0.71 19.9 0.61 16.3 10

SM6 0.42 19.5 0.39 15.0 6

SP75 (12V) 4.8 21.7 4.4 17.0 75

SP36 (12V) 2.4 21.7 2.1 17.0 36

SP18 (12V) 1.2 21.7 1.1 17.0 18

c. 0.8

d. 0.9

b. No

b., c.





2.0 amps = Isc2

1000 w/m2

850 w/m2

Isc2 = 2.0 X ( 850 / 1000 )

= 1.7 amps

If the two days are the same temperature, then we can calculate the Isc at standardirradiance for each module.

Module 1:

2.5 amps = Isc2 745 w/m

21000 w/m

2

Isc2 = 2.5 X ( 1000 / 745 )

= 3.36 amps for Module 1 at 1000 w/m2

Module 2:

2.1 amps = Isc2 650 w/m

21000 w/m

2

Isc2 = 2.1 X ( 1000 / 650 )

= 3.23 amps Module 2 at 1000 w/m2

Under standard irradiance, Module 1 has an Isc of 3.36 and Module 2 has an Isc of3.23. Therefore the correct answer is:

a. Module 1

Note: If the temperatures were significantly different on the two days that themeasurements were taken, we could not answer the question without knowing theactual temperatures.




Using the cell temperature coefficients for either the SP cell or the SM cell, we see thatthe change in Voc per

oC is

-2.15 mV/ o

C or -0.36%/ oC

We calculate the voltage loss factors for 33 cells as follows:

Voltage loss factor = -2.15 mV/ o

C X 33 cells

= -0.00215 V/ o

C X 33 cells

= -0.071 V/ o

C

Voc Change = Factor X Temperature Change

= -0.071 V/ o

C X [55oC - 25

oC]

= -0.071 X 30

= -2.13 Volts

Voc2 = 19.9 Volts Voc (at 25oC) - 2.13 Volts

= 17.77 Volts Voc (at 55oC)

Using the percentage loss figures:

Voltage loss factor = -0.36 %/

o

C

Voc Change = Factor X Temperature Change

= -0.36 %/ oC X [55

oC - 25

oC]

= -0.36 X 30

= -10.8 %

Voc2 = 19.9 Voc (at 25oC) X (1 - % Voc Change)

100

= 19.9 Volts X ( 1 - 10.8 )100

= 19.9 X (1-0.108)

= 19.9 X (0.892)

= 17.75 Voc (at 55oC)




The work below shows the calculations and results for the SM10, SM50, SM55, SP75,and SR100 modules. The method for other modules will be the same.

The datasheet for the SM10 module gives the following information:


Total Area Efficiency = 10 watts0.360 m X 0.330 m X 1000 w/m

2

= 0.084 or 8.4%

For the SM50 module:



2

= 0.118 or 11.8%

For the SM55 module,



2

= 0.129 or 12.9%

For the SP75 module:



2

= 0.119 or 11.9%




For the SR100 module:



2

= 0.112 or 11.2%

The data sheet for the SM55 module gives the following information:

Voc: 21.7 Volts Isc: 3.45 Amps

Vmp: 17.4 Volts Imp: 3.15 Amps

With 8 modules in series, the voltages will be multiplied by a factor of 8:Voc = 21.7 X 8 = 173.6 VoltsVmp = 17.4 X 8 = 139.2 Volts

Four strings of modules in parallel will increase the currents by a factor of 4:Isc = 3.45 X 4 = 13.8 AmpsImp = 3.15 X 4 = 12.6 Amps

The correct representation is sketch c.

a. Most cells in series _(4)_ b. Fewest cells in series _(1)_ c. Next fewest cells in series _(2)_ d. 2 strings of cells in parallel _(3)_ e. Most cells in parallel _(1)_ f. Same Imp as another module (2) or (4)

Reading the graph of cell temperature rise vs. irradiance, we estimate that at 800 watts/ m

2 the cell temperature rise will be approximately 22 °C. Then the cell temperature will

be: Cell Temperature = 32 °C (ambient air) + 22 °C (rise)

= 52 °C




Looking at the graph of cell Temperature Rise vs. Irradiance, we determine that anirradiance of 700 watts/m

2would result in a temperature rise of about 20

oC. If the

ambient temperature is 35oC, then the cell temperature is 35

oC + 20

oC = 55

oC. The

temperature coefficient stated in the chapter for Vmp is -0.072 V /oC. We calculate the

change in Vmp as:

Vmp change = Factor X Temperature Change

= -0.072 V/ oC X [55

oC - 25

oC]

= - 2.16 Volts

Vmp (at 55oC) = 15.9 - 2.16 Volts = 13.74 Volts

By summing the each hour's energy, we get the output for each month:

Output in February = 76.8 WhOutput in June = 247.1 Wh

0

5

10

15

20

25

30

35

4 6 8 10 12 2 4 6 8

Time of Day

Watts

Feb

Jun




The module literature gives the following values for the SP75 module:Vmp = 17.0 volts Imp = 4.4 amps

The resistance that will operate a single SP75 module at the maximum power point

under standard conditions is then:

Resistance = Vmp = 17.0 = 3.86 OhmsImp 4.4

Therefore, connecting a 3.86-ohm resistor to the 75 watt module in the field will operatethe module very near its maximum power point. To operate the whole array at themaximum power point, we calculate the resistance as:

Resistance (array) = 3.86 ohms x 6 in series4 in parallel

= 5.79 ohms

Note that under real field conditions, the values of Vmp and Imp may be slightlydifferent. This is because the irradiance may be less than 1000 watts/m

2 and the cell

temperature will likely be hotter than 25 °C.



Siemens Solar Basic PV Technology Course Components – Load EstimationCopyright © 1998 Siemens Solar Industries

9-1

Chapter NineLoad Estimation

The loads cannot be taken as given while all other calculations are carefullyscrutinized. The loads influence every aspect of system design, and must be asefficient and reliable as possible. As you read, think about what influences loadefficiency, and how you might have to argue with a client to replace their existing orproposed load with a more efficient one.

The entire system design is based on the size of the load. If the information isinaccurate the initial costs will be too high or the array and battery could be too smalland the system will eventually fail. It is, therefore, essential that time be taken to

look carefully at the load requirements and the expected usage pattern.

Using literature values for load consumption is common, but it is more accurate tohave the load demand measured to be sure. Often nominal numbers are presentedin literature, and a particular piece of equipment may require more or less powerthan stated.

If an existing application is being retrofitted with a PV power system it is veryimportant to not just look at the old generator capacity and try to recreate it with PV.Very often an oversized diesel generator was installed, perhaps because that sizewas used elsewhere or to allow for future growth. A PV power system can be

designed to accurately match the current load requirement without limiting the abilityto expand in the future to meet greater demand.

The load profile throughout the year must be accurately determined. Any seasonalvariation might influence the choice of tilt angle or battery size for autonomy. The"duty cycle" or hours of operation for intermittent loads must be estimated carefully.In the case of telecommunications equipment, not only the hours of transmitting butalso the hours of standby or quiescent operation need to be included in the loadcalculations.

Improving load efficiency is the quickest way to reduce PV power system cost. Amore efficient load device may even be slightly more expensive than an existing orconventional load device. But if the power consumption of the efficient load issignificantly lower than the other, the cost savings in the array and battery may offsetthe higher load cost in many cases, and result in a total system cost (modules +battery + loads) that is lower than if inefficient loads are used.




9-2

Using oversized diesel generators may have allowed load device efficiency to beneglected in the past. But when a small load is operated by a large capacitygenerator, fuel efficiency is low so fuel is wasted. Simply retrofitting a new PV powersystem to an existing inefficient load is a mistake if a more efficient load device canbe installed.

Load Sizing FormsWe have developed a form to use for calculating the average daily load. Both ACand DC loads can be added. The Daily Load Sizing Form has you add up all theloads to estimate the largest daily load demand on your batteries. The Week-Averaged Load Sizing Form has you average the load demands over a 7-day period.This allows for slightly smaller array sizing if your loads vary during the week. Thebattery should be sized to meet the largest daily load demand, while the array canbe sized to meet a more averaged load value. The sizing forms are presented next.




9-3

Daily Load Sizing Form

System Description :

DC Loads Qty. Amps Hours/day Daily Demand (Ah)

: X X =

: X X =

: X X =

: X X =

: X X =

: X X =

: X X =

: X X =

DC Loads (Ah) =

AC Loads Qty. Watts Hours/day Daily Demand (Wh)

: X X =

: X X =

: X X =

: X X =

: X X =

: X X =

: X X =

: X X =

AC Sub-Total (Wh) =

Continuous Watts =Surge Est. =

Inverter Choice:

[ ] ÷ [ ] ÷ [ ] + [ ] = ____________ AC Sub-Total Efficiency Input Voltage DC Loads Daily Load (Ah)




9-4

Week-Averaged Load Sizing Form

System Description:

DC Loads Qty. Amps Hours Days/wk Weekly Demand (Ah)

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

Weekly DC Loads (Ah) =

AC Loads Qty. Watts Hours Days/wk Weekly Demand (Wh)

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

: X ______ X ______ X =

Weekly AC Sub-Total (Wh) =

Continuous Watts = ______

Surge Est. = ______ Inverter Choice:

[ ] ÷ [ ] ÷ [ ] + [ ] =AC Sub-Total Efficiency Input Voltage DC Loads Total Weekly

Demand (Ah)

[ ] ÷ 7 days =Total Weekly Week-AveragedDemand (Ah) Daily Load (Ah)




9-5

Daily Load Demand

You calculate the total daily load demand by multiplying each load demand times thetime that the load operates in a typical 24-hour period. DC loads are estimated byusing amp-hours, while AC loads are estimated by using watt-hours. This is

because the DC loads can draw their current directly from the battery, while the ACloads must draw their power from a DC-to-AC inverter that will change the voltagefrom the (usually 12, 24 or 48 volts) DC level to the (usually 110 or 220 volts) AClevel. The input and output voltage of the inverter must be included in thecalculations separately.

DC Load Demand = DC Load Current (amps) X Hours of Operation

AC Load Demand = AC Load Power (watts) X Hours of Operation

If DC loads are given in watts instead of amps, then you can easily convert bydividing by the nominal operating voltage. This is usually 12 volts, and sometimes24 volts.

DC Load Current (amps) = DC Load Power (watts)

Nominal DC Voltage

Example: A typical 12 volt 40-watt fluorescent light has a current given by

DC Load Current = 40 watts12 volts

= 3.3 amps

Example: It is very common for microwave repeater equipment to operate at 24volts nominal. If the operating power of a repeater is 200 watts (at 24volts) the load current is given by

DC Load Current = 200 watts24 volts

= 8.3 amps




9-7

Loads May Vary Seasonally

Most loads will not be operated the same amount of time every day and everymonth. There will be a seasonal variation to the load demand, and this will influencethe choice of tilt angle for the array as well as the array and battery size. An

example of two seasonal profiles is shown below, one peaking in the summermonths, and the other peaking in the winter.

Seasonal Load Profiles


0

10

20

30

40

50

60

70 Summer peak profile

Winter peak profile

An example of the variation in insolation throughout a year is shown on the nextpage. The insolation on a flat surface varies greatly with the least during the winterand the greatest during the summer. The best system design has the array tilted toan angle so that the profile of insolation throughout the year matches the loadprofile. In this way, the fewest modules are needed to meet the load demand.

If the load demand is small in winter and large in summer, as with air conditioningloads or water pumping for irrigation, then tilting the array near a flat angle will givean insolation profile that best matches the load requirements.




9-8

New Delhi Insolation Profiles

Latitude 29 deg. N

0

100

200

300

400

500

600

J a n

F e b

M a r

A p r

M a y

J u n

J u l

A u g

S e p

O c t

N o v

D e c

I n s o l a t i o n (

L a n g

l e y s )

0 deg. 15 deg. 30 deg. 45 deg. 60 deg.

If the load is relatively constant every month of the year, as might be the case for anavigational aid or a constantly transmitting repeater, then a different angle is better.Tilting the array up increases the insolation intercepted during the winter months andsacrifices some during the summer months, with a resulting profile that is moreconstant throughout the year to better match the requirements of a constant load.

If a tilt is chosen that does not match the insolation profile to the load profile, thenthe array will have to be quite large to produce enough output when insolation is lowand may produce excess power that is just wasted when insolation is high. Bychoosing a tilt angle for the array that gives the best match, array size is minimized,and a reliable fixed angle mounting structure can be designed for the system.Computer models or repetitive hand calculations can predict the best angle to give amatch between load demand and insolation.




9-10

For practice at adding AC loads, let’s estimate the demand for a remote school witha variety of AC appliances, lights and other loads.

Example: A small school in a remote area wants to remove their generator anduse only photovoltaics to power all their loads. The school has eight

40-watt fluorescent lights, small lights in the bathrooms, fourcomputers, an overhead projector, and a small microwave oven forheating lunches, and a refrigerator.

AC Loads Qty. Watts Hours WeeklyDemand (Wh)

Lights : 8 X 40 X 8 = 2,560PL lights : 2 X 11 X 2 = 44

Computer : 2 X 200 X 4 = 1,600

Projector : 1 X 300 X 3 = 900

Microwave : 1 X 800 X 2 = 1,600Refrigerator : 1 X 200 X 12 = 2,400

AC Sub-Total (Wh) = 9,104




9-11

Exercise




9-12

Choosing An Inverter

Before you can complete the sizing forms for AC loads, you must choose aninverter. This choice will determine the average inverter efficiency and the nominalDC voltage of the array and battery, both of which are needed to finish the

calculations indicated at the bottom of the sizing forms.

Meeting Continuous Power Demand

There are two parameters that will help choose an adequate inverter to handle theAC load demand. First the inverter must be able to operate all the AC loads thatmight be on at one time continuously. This value is calculated by adding up thewatts of the AC loads and filling in the “Continuous Watts” line in the sizing form. Besure to multiply by the quantity to get the correct total load requirement.

Estimate of Continuos Watts: Add all AC loads that might be on at same time

Meeting Surge Power Demand

The second parameter that the inverter must meet is the surge requirement.

Inductive loads, such as motors (example: washing machines, water pumps),compressors (example: refrigerators), and even certain types of fluorescent lightballasts. In an inductive load, there are coils of wire that must be “loaded” withenergy. During this short period of loading, the current that can flow can be 4-6times the continuous running current! So any inverter chosen to operate such loadsmust be able to supply this surge of current, for fractions of a second or sometimesfor seconds.

To estimate the total surge power that the inverter might have to deliver, you mightadd up all the surge powers of all the inductive loads and add this to the continuouspower demand of the non-inductive loads. But this approach is much too severe.

The probability of two or more inductive loads turning on at exactly the same time isquite low. And the result of such an approach would perhaps be the choice of anunnecessarily large and expensive inverter.




9-13

A more practical approach to estimating total surge power is to identify the load withthe largest surge power and add this value to the continuos power demand of all theother AC loads that might be on at the moment that the large load is turned on.Thus inverter would have to operate all the loads (except the one with the largestsurge) and be able to turn on the load with the largest surge as well.

Estimate of Surge Watts: Add the surge of the largest loadto the power of all the other AC loadsthat might be on at the same time

If you are given no information on the surge demand of a particular inductive load, itis safe to assume that six times the continuous current will be drawn as a surge tostart the load.

An example of a residential refrigerator load profile is shown. As time passes(moving from right to left) the compressor turns on and then off, trying to keep therefrigerator box cool. A refrigerator’s efficiency is measured by not just thecontinuous power level, but also by how many hours in a typical day the compressoractually operates. Each time it turns on, there is a spike or surge in the current thatis drawn. The continuous running current of this particular refrigerator is about 2amps, but the surge current needed is about 12 amps, a factor of 6 times greater!

Refrigerator Load Profile -- Continuous and Surge Levels




9-14

Estimating Continuous and Surge Powerfor Complex Systems

Sometimes it is not practical to add up the continuous and surge requirements ofeach load. This is particularly true when dealing with residential systems, where

there might be a large number of AC appliances and loads, but their time usage isunknown. It would be improbable that you or the user could accurately predictexactly how much time each and every load would be used, and exactly what loadswould be operating at the same time. For large AC systems it is better to estimatethe continuous and surge power demands by using statistical averages.

To estimate the continuous demand of a complex system, an Energy ConsumptionChart such as the one shown below is quite useful. These are often prepared bylocal utilities or energy departments. The annual energy consumption of variouscommon loads is indicated. This number is based on a wide variety of users, andrepresents only a statistical average. You would add up the annual energy

consumption of the loads, and then divide by 365 days/year to get the estimateddaily load.

Example: A home is being built in a remote area, and is to be solar powered.The owner has made a list of the appliances that will be used in the house.Annual energy values have been taken from the chart.

Refrigerator/Freezer 1135 kWh/yearMicrowave oven 190Toaster 39Dishwasher 363Coffee Maker 106

Washing machine 103Room air conditioner 1350Bathroom lights 60Bedroom lights 200Dining room lights 144Hall lights 108Kitchen light 100Stereo 108Television 320Various personal care items 35

Total Annual Energy Requirement 4361 kWh/year

Estimated Daily Load = 4361 kWh/year365 days/year

= 12 kWh/day

The system designer and the client could look at this list and see what items contribute mostto the energy burden, and perhaps discuss alternative or more efficient loads to reduce theoverall energy and therefore the solar array and battery size as well.




9-15

Source: Photovoltaic User Guide, California Energy Commission, July 1984




9-16

Using another statistical tool presented next, you can estimate the continuous andsurge power that might be required from an inverter. A chart has been preparedshowing how much continuous power and surge power might be needed to serve arange of daily loads.

Estimating Surge

0 4 8 12 16 20 24

24

22

20

18

16

14

12

10

8

6

4

2

0

surge

range

continuousrange

Average Daily Load (kWh)

Estimated PowerRequirement (kW)

Read up from the estimated daily load requirement to find the approximatecontinuous and surge demand. For example, if the total daily load was about 20kWh/day, you would read up from 20 to find that the continuous power output of theinverter may be between 4-7 kW, and that the surge demand may be between 10-20kW. There is quite a range allowed, and this method is not precise. But it is a quickway to estimate values that otherwise could be tedious to calculate and which mightbe ultimately inaccurate anyway.

Example: Using the daily load demand of 12 kWh/day already estimated for theremote home, the inverter chosen should be able to supply

Continuous Watts: 2-4 kW

Surge Watts: 6-12 kW




9-17

Reading Inverter Literature

With continuous and surge load power values in hand, you can now select aparticular inverter for your system and determine the average efficiency and theinput DC voltage that it requires. These pieces of information will then be used tocomplete the load sizing calculations.

Use actual manufacturer’s literature to determine the pertinent factors and values.First locate the continuous power output capability of the various models, and thentry to determine the maximum surge capability as well. If these meet or exceed yourcalculated values, then this particular model is suitable for your load demands.(Other technical features such as output waveform, or voltage and frequencyregulation, will also guide you to choose one inverter over another. These aspects ofinverter technology are discussed in the chapter on Inverter Technology).

Once you have chosen an inverter, you can read from the literature the averageefficiency and the input DC voltage. Sometimes an inverter may be available in

more than one input DC voltage. The higher the DC voltage, the smaller the DCcurrent that is needed for the same amount of power, and therefore the smaller thewires, fuses, circuit breakers and the lower the resistive losses in these componentsas well.

If you are only operating AC loads in the entire system, then you are free to chooseany input DC voltage that you wish. To minimize costs and resistive losses, youmay want to choose the higher voltage, such as 24 or 48 volts. (Some large inverterseven come with 110 volts input).

However if you are planning on having both DC and AC loads operating from the

same photovoltaic power system, then you are more constrained in your inverter DCinput voltage. Most DC loads operate at 12 volts. Some fluorescent light ballasts,refrigerators, and pumps operate at 24 volts, as do many telecommunicationsdevices. If you want to operate 12 or 24 volt loads you will have to choose aninverter model with that same input DC voltage. (Sometimes a DC-DC convertercan be used to allow multiple DC voltages in the same system. For example, a 48-volt inverter and battery may be installed to operate AC loads, and a DC-DCconverter can be installed as well to convert from 48 to 12 volts for some small DCappliances).

Exercise




9-18

Calculate Ampere-Hours For AC Loads

With average efficiency and voltage values from manufacturer’s literature, you cancomplete the sizing forms for AC loads. The sizing forms show that to calculate theDaily Load (Ah), you must take the AC Subtotal (Wh) and divide by the inverterefficiency and input DC voltage.

AC Load (Ah) = AC Subtotal (Wh)Efficiency X Voltage

Why do we divide by the efficiency? We want to determine the amount of DC amp-hours that the solar array must produce. What we have been calculating so far inthe Load Sizing Forms is the amount of AC watt-hours that the loads need. Since

the inverter is not 100% efficient, more energy must be put into the inverter than isoutput to the AC loads. How much more is determined by the efficiency.

Inverter Efficiency

InverterAC power outDC power in

power out = Efficiencypower in

power in = power out Efficiency




9-19

Efficiency is defined as the ratio of output power or energy divided by the inputpower or energy. Therefore the energy or power in is given by the energy or powerout divided by the efficiency.

So dividing the AC Subtotal (Wh) by the efficiency CONVERTS the watt-hours out tothe AC loads into the watt-hours needed to be put into the inverter on the DC side.

DC Watt-Hours Into Inverter = AC Subtotal (Wh) Efficiency

Then to complete the calculation, divide the DC watt-hours into the inverter by theDC voltage to get the DC amp-hours into the inverter. This is the Ah load demandthat we need to deliver from the batteries and array.

DC Amp-Hours Into Inverter = DC Watt-Hours Into InverterDC Input Voltage of Inverter

If the DC voltage of the inverter and any DC loads are the same, then both can beoperated from the same battery bank and the same array. The Ah values for the DCloads and AC loads can be added to give the grand total for the whole system. If thevoltages are different, then either a DC-DC converter must be used, or separatesystems can be designed for each.




9-20

Exercise:




9-21

Daily Load vs. Occasional Load

If a load is the same every day, then the solar array should be sized to meet thatload demand. But if a load is used heavily some days of a week and less on otherdays, then the solar array can be sized smaller than the heaviest demand and larger

than the smaller demand, and on the average during the week meet the loaddemand. The Week-Averaged Load Sizing Form is intended to more economicallysize a solar array to operate a load that is not the same every day.

A good example of how this approach is applied is a weekend cabin. The loads areused heavily on the Saturday and Sunday (2-day/week) and are not even operatedat all Monday-Friday (five days/week). The battery should be sized to meet theheavy weekend demands, but the array has all five weekdays to “catch up” on whatit could not produce during the weekend.

If the load varies during the week, then the array can be sized based on the average

demand spread through the seven days of the week, using the Week-AveragedLoad Sizing Form. At the end of the week, the small array has still fully rechargedthe battery bank. The battery is always sized using the Daily Load Sizing Form

If Load Varies During Week

Array Output Spread Through Week

1

0

10

20

30

40

50

60

70

80

90

100

Battery Net

Array Load Battery Net

At End OfDay




9-22

If the load is the same every day, then the array and battery are sized to meet thatconstant daily demand, using the Daily Load Sizing Form.

If Load Is Same Every Day

Array Matches Load Each Day

1

0

10

20 30

40

50

60

70

80

90

100

Battery Net

Array Load BatteryNet At End

Of Day

To see the impact of approaching load estimation from an averaging method, on thenext pages we redo the examples for the small remote cabin and school, this timeestimating what the “averaged” load would be. This might lead to a substantiallysmaller load value and therefore a smaller array.




9-23

Example: The small remote cabin that was previously used (page 11) will only beoccupied on the weekends. Recalculate the Ah demand using the Week-Averaged Load Sizing Form. Compare the result to the daily usage valueoriginally calculated.

DC Loads Qty. Amps Hours Days/wk WeeklyDemand

(Wh)

Lights : 2 X 3.3 X 4 X 2 = 52.8PL lights : 3 X 0.92 X 5 X 2 = 27.6

TV : 1 X 3.3 X 4 X 2 = 26.4Fan : 1 X 2 X 8 X 2 = 32

Weekly DC Loads (Ah) = 138.8

[ 138.8 ] ÷ 7 days = 19.8Total Weekly Week-Averaged

Demand (Ah) Daily Load (Ah)

On the next page an example should be worked out for the remote school discussedearlier. Recalculating the loads using the Week-Average Load Sizing method, as ifthey were spread out over all seven days of the week, can be done to see if the loadis substantially smaller taking this approach.

Array and battery sizing calculations will be discussed in the chapter System Sizing.But you can see that by averaging the load demand over the entire week that theenergy demand from the array is reduced. The difference between the values for

daily loading and week averaging is greater for the remote cabin than for the school.This is because the cabin changes from a regular daily loading to only a two day perweek loading. The change for the school design from daily loading to having theweekends “off” is not so great.

If the load distribution during a typical week is fairly even, then use the Daily LoadSizing Form value for both array and battery sizing. Use the Week-Averaged LoadSizing Form value for array sizing only if the distribution during a 7-day week is reallyskewed to a few days. Then you might see some economics in sizing the array forthe average. In all cases, used the Daily Load Sizing Form value for battery sizing.The battery has to be sized to handle the actual heavy daily loads.




9-24

Example: The school also previously used (page 12) is to be occupied only duringthe week (Monday-Friday). Recalculate the Ah load demand if it isspread out over the full seven days of the week using the Week-Average Load Sizing Form. Note that the refrigerator is not turned offand must be kept operating during all seven days. How does this valuecompare to the previously calculated load demand?

AC Loads Qty. Watts Hours Days/wk WeeklyDemand

(Wh)

Lights : 8 X 40 X 8 X 5 = 12,800PL lights : 2 X 11 X 2 X 5 = 222Computer : 2 X 200 X 4 X 5 = 8,000Projector : 1 X 300 X 3 X 5 = 4,500

Microwave : 1 X 800 X 2 X 5 = 8,000Refrigera

tor: 1 X 200 X 12 X 7 = 16,800

Weekly AC Sub-Total (Wh) = 50,320

[ 50,320 ] ÷ 7 days = 7,188Total Weekly Week-AveragedDemand (Wh) Daily Load (Wh)




9-25

Exercise




9-27

Fluorescent

In fluorescent lamps mercury vapor is excited inside a tube by an alternating current,and light is emitted by the gas. The light strikes the inner coating of the tube, whichis some type of fluorescing material, producing a soft glow. The light produced fallswell within the visible range, so the efficiency is good.

Traditionally fluorescent bulbs have been long cylinders ranging from a few watts to40 watts. New "parallel length" or PL lamps are now available that are much morecompact and can often times replace incandescent bulbs in standard light fixtures.The PL lamps fold back the long cylinder to make a compact "H" shape. Double "H"or "quad" PL lamps are also available. PL type lamps are available in 5, 7, 9, 13, 18,24 and 36-watt models. To operate from DC power a ballast is needed to producehigh frequency AC current. These are discussed further in this section.

New types of traditionally shaped fluorescent lamps are available that produce morelumens/watt, better color, and longer life (up to 24,000 hours). These lamps are asmaller tube diameter (1 5/16") and are called "T-10" tubes.

High Intensity Discharge (HID) Lamps: Mercury Vapor Metal Halide Low Pressure Sodium High Pressure Sodium

These lamps are ideal for outdoor applications like security lighting where lamps willbe operated for extended periods of time.

The mercury vapor lamp produces a harsh bluish-white light and is the least efficientof the HID lamps. They have a low initial cost, but higher operating costs than otherHID choices.

The low-pressure sodium lamps produce a harsh yellow-gold color and are suitableonly for general outdoor lighting. They are the most efficient of the light sources, butthe harshness of the light makes their applications limited.

High-pressure sodium lamps are the most efficient of the white light sources. Theyproduce a soft amber color and are used widely for street lighting and other exteriorlighting applications.

Metal halide lamps have good color rendering capabilities and are widely used forsporting event lighting.




9-29

Light Efficacy

Efficacy (lumens/watt)

100

50

0

Incandescent

Tungsten Halogen

Mercury Vapor

Fluorescent

High Pressre Sodium Low Pressure

Sodium




9-30

The life of the lamp must also be a consideration in system design. It is encouragingthat the more efficient or efficace lamps are also the more long life products as well.Fluorescents can last 8,000-12,000 hours (some even up to 24,000 hours)compared to typical incandescents that last only about 500-1500 hours. The entirefamily of high intensity discharge lamps (HID) typically lasts about 15,000-25,000hours.

Lamp Life

L i f e

( h o u r s

0

5000

10000

15000

20000

25000

Incandescent Tungst enHalogen

Mercury Vapor Fluorescent High PressureSodium

Low PressureSodium

Hours of Life

It is critical that customers and users of photovoltaic generated power understandthe value of more efficient lights. The least efficient incandescent light is perceivedas being cheaper because the lamp bulb cost is so low, whereas a compactfluorescent may cost 10 times as much. But with the fluorescent lasting 6-8 times aslong and consuming 4-6 times less, the cost over time of the incandescent can be

shown to be much higher than the cost of the “more expensive” fluorescent choice.




9-31

Fluorescent Ballasts

Incandescents need only to have adequate voltage and current to operate. They willglow immediately. All the other forms of lamps need some form of ballast to give aninitial pulse to start the lamp, or to convert the power to an alternating current to

keep the gas glowing.

Traditional ballasts have been made using a "core and coil" technology. Theseoperate using the available 60 Hz utility power, and are not particularly efficient.There is a slightly noticeable flicker to the light, and sometimes a delay or pulsingwhen starting lamps. Recently available improved high frequency electronic ballasts(20,000 Hz output) help to reduce these problems. They are almost twice asefficient, turn on the lamps instantly, operate the lamps with no flicker or electronicinterference, and some even offer a degree of dimming to fluorescent lamps.

The overall efficiency of the lamp must have the efficiency of the ballast included to

determine final power requirements. The small electronic ballasts for PL lampsconsume approximately 2 watts, while the traditional ballasts use almost 4 watts.

The cost of the ballast must also enter into the cost effectiveness equation. A smallPL lamp may cost $5 and the ballast may add $15-25 to that. When compared to a$1-3 incandescent lamp, this looks bad. But the PL is expected to last 10 times aslong as the incandescent and consumes 4-5 times less power to give the samebrightness. Over the life of the PL lamp and ballast, the total cost of theincandescent choice becomes much greater.

The ballasts typically do not operate more than one PL lamp however. Each PL

lamp for example must have its own ballast, so usage in track lighting is difficult. Anoption is to use small 12-volt Quartz-Halogen bulbs, although they are not asefficient.

Ballast and fluorescent lamp design may have to be modified to allow starting in coldclimates. The ballast may not be able to pulse strongly or long enough to start acold lamp. Special pre-heaters can be installed in the bulb to allow for starting coldlamps.




9-32

Color Quality

One measurement that relates to color quality is the color temperature of the light.At high temperatures, matter emits a spectrum of light, and the spectrum changeswith higher temperature. The surface of the sun is approximately 5900 deg. C. or

6200 deg.Kelvin, another temperature scale similar to the Celsius (C) scale butbeginning at "absolute zero" degrees. Our sunlight spectrum is very close to thespectrum emitted by a body at that temperature. The hotter the temperature of aradiating source of light, the higher the percentage of high-energy blue light isemitted. The colder the temperature, the higher the percentage of low energy redlight is present in the spectrum. So the spectrum emitted from lamps can becompared to the spectrum that would be emitted by a perfect radiator. A colortemperature is given to a lamp to give a measure of how "red" or "blue" the lampspectrum appears. The small energy efficient PL lamps are usually available withcolor temperature of 2700 deg.K, but are also available at 3000, 3500, and 4100deg.K.

Another measure of the quality of light is the color rendition index (CRI). This is afigure of merit that compares how well lamp light matches the balance of colors indaylight. Two lamps may be compared only if they share similar color temperatures.

Incandescents can give fairly good color rendition. Fluorescents have traditionallybeen more blue than incandescents, but now "warm" and "full spectrum"fluorescents are available. The mercury and sodium lamps do not produce a broadgradually changing spectrum of wavelengths and so their color is quite harsh.

Using Outdoor LightOne consideration to make when determining lighting needs is how much naturalsunlight can be brought into the space. The cost of the light is free, and the colorrendition is excellent. Glass is not as good an insulator as wall materials, andadjustments may have to be made to the heating plans for a living space. Butpassive solar heating and cooling can be anticipated and often controlled. There arerooftop mechanisms that track the sun and reflect light down into a living orworkspace efficiently throughout a day. These could be operated using PVmodules, adding no burden to the power requirements of the space.




9-33

Exercise

(“PL”)




9-34

Efficient Refrigeration Loads

A second major residential load that warrants scrutiny is refrigeration. People areoften unaware of how much the compressor is running in their refrigerator or of howmuch power it is consuming. Efficiency improvements here can result in

tremendous savings in array and battery costs.

Types of Refrigerators

Refrigeration units operate from two basic sources of power, electricity or liquid fuel.Electric units can be DC or AC or switchable between the two. Fuel types can usekerosene or propane.

There are units that automatically switch from DC to AC to gas power, most

commonly used on recreational vehicles (RV's). These units are typically veryinefficient and use a great deal of power (up to 400 watts) on DC power. They aretherefore not a good choice for a remote PV powered site.

It is not uncommon for remote sites using PV for lighting; communications and toolsto use a gas powered refrigeration unit. Gas may already be on site for cooking, andavoiding module and battery costs for refrigeration keeps the overall system morecost effective.

However there are many manufacturers of DC powered refrigeration units. Thesevary from small portable units of only 0.4 cubic feet (10 liters) capacity, to full sized

17 cubic feet (480 liters) capacity units for a home or village.

Designs for Efficiency

It is especially important for PV powered refrigeration units to be designed forefficient operation. The walls should be thick and well insulated. And the storagechamber should be no larger than required for the expected usage. For example, asmall unit for vaccine storage does not need to be as large as a unit intended forstoring food or milk.

The seals around the doors should seal tightly. A double seal helps to insure noleaking.

Top loading designs are best at keeping the cold in whenever the lid is opened. If afront door style is used, efforts should be made to keep the cold from escaping. Thiscan be done by hanging plastic baffles with slits, to allow access to foods butprevent gross movement of air. Or sections can be enclosed with their own doors,so that access to one area does not expose another area to loss of cold.




9-35

The condenser coils should be located to allow maximum escape of heat withoutheating the walls of the unit. Many common AC units are inefficient because theylocate the condenser coils underneath the refrigerator, forcing all the escaping heatto pass up and around the refrigerator walls! Coils can be located on the top of theunit, or ideally they can be located away from the unit, perhaps outside where theambient temperature may be lower than inside the home.

There are available on the market super efficient refrigerator designs. A particularmodel of PV powered refrigerator, manufactured by SunFrost, needs only five 50-watt solar modules to operate a 17 cubic foot capacity unit, even at 90o F ambient airtemperature. The unit alone costs more than double what a conventional ACpowered design costs, but the total cost of the modules, batteries, and inverterrequired for the inefficient conventional AC unit far surpass the total cost of theefficient DC unit plus its 5 modules and small battery bank. The SunFrost productalso comes ready for AC power and is only slightly less efficient. It can beincorporated into a conventional all-AC residential system design if so desired.




9-36

Exercise




9-37

Question Your Load Information

All of your accurate and precise calculations for load sizing will not overcomeincorrect initial data! You should be vigilant and always on the lookout for incorrectload data. Incorrect assumptions, inaccurate estimations, unrealistic forecasts,

unexpected increases--all of these and more can be lurking behind your nice lookingcalculations. You should at the very least ask yourself some of the followingquestions when dealing with load estimation.

Is Future Growth Anticipated?

Telecom channels can be added, and hours of operation can increase. Populationscan increase, bringing more demand. Enjoyment of the quiet and reliable nature ofphotovoltaic power can lead to expansion.

Think ahead when doing your system design. This also applies to componentselection as well. Choose charge regulators, wires, inverters and other componentsthat are adequately oversized to accommodate some future growth without majorchanging out of hardware.

Is The Load Profile Well Known orIs It A Guess?

Some clients may not know an existing site’s real energy or power consumption, andmay guess for purposes of your calculations. Remember that most users ofelectricity are aware of their POWER requirements but not of their ENERGYrequirements. They may confuse the two, and assume that if an existing generatoror source of power is present, that that must be the actual power used by the load.For example, if a 10 kW diesel generator is on site for whatever historical reason,this does not necessarily mean that the load demand is 10kW.

Has seasonal variation been taken into account? Lighting demand, for example,may increase in the winter months, while water demand may increase in the

summer.




9-39

(End of Chapter)



Siemens Solar Basic Photovoltaic Technology 9-1 Load Estimation

Chapter 9 – Answers Load Estimation

We first calculate the current for each load type:

Fluorescent lights = 40 Watts = 3.3 amps12 Volts

Ceiling Fans = 20 Watts = 1.7 amps12 Volts

Vaccine refrigerator = 60 Watts = 5.0 amps12 Volts

This information is used to complete the Daily Load Sizing Form:

System Description: Remote Clinic - Daily Loads

Hours DailyDC Loads Qty. Amps /day Demand (Ah)Fluorescent Lights : 6 X 3.3 X 6 = 118.8

Ceiling Fans : 3 X 1.7 X 12 = 61.2

Vaccine Refrigerator : 1 X 5 X 10 = 50.0

DC Loads (Ah) 230.0




System Description: Remote Clinic - Weekly Loads

WeeklyDC Loads Qty. Amps Hours Days/W

kDemand (Ah)

Weekday Loads- Fluorescent Lights : 6 X 3.3 X 8 X 5 = 792.0

Ceiling Fans : 3 X 1.7 X 8 X 5 = 204.0

Weekend Loads- Fluorescent Lights : 6 X 3.3 X 4 X 2 = 158.4

Ceiling Fans : 3 X 1.7 X 4 X 2 = 40.8

Constant Loads - Vaccine Refrig : 1 X 5 X 10 X 7 = 350.0

Weekly DC Loads (Ah) 1545.2

0 ÷ ÷ + 1545.2 = 1545.2

AC Sub- Effic. Input DC Total Weeklytotal Voltage Loads Demand (Ah)

1545.2 ÷ 7 = 220.7

Weekly Days Avg LoadAh/day (Ah/day)




To clarify the power requirements, look at the amount of time that the loads are intransmit and standby. The combined time on and off must equal 24 hrs (1 day).

Radio transmitter #1 Transmit 12 hrs / dayStandby 12 hrs / day

Radio transmitter #2 Transmit 8 hrs / dayStandby 16 hrs /day

System Description: Telecom site with two transmitters

Hours DailyDC Loads Qty. Amps /day Demand (Ah)Trans. #1 – Transmit : 1 X 8 X 12 = 96.0

Trans. #1 – Standby : 1 X 0.5 X 12 = 6.0Trans. #2 – Transmit : 1 X 5 X 8 = 40.0

Trans. #2 – Standby : 1 X 0.3 X 16 = 4.8

DC Loads (Ah) 146.8

During the winter months, the stated transmit times are reduced:

Radio transmitter #1 Transmit 8 hrs / dayStandby 16 hrs / day

Radio transmitter #2 Transmit 5 hrs / dayStandby 19 hrs /day

System Description: Telecom site with two transmitters

Hours DailyDC Loads Qty. Amps /day Demand (Ah)Transm. #1 – Transmit : 1 X 8 X 8 = 64.0

Transm. #1 – Standby : 1 X 0.5 X 16 = 8.0Transm. #2 – Transmit : 1 X 5 X 5 = 25.0

Transm. #2 – Standby : 1 X 0.3 X 19 = 5.7

DC Loads (Ah) 102.7




Refer to the manufacturer's literature.

The refrigerator will be the largest surge load. We will first take all the other AC loadwatts, then add six times the refrigerator load. As a worst case, we assume that all ofthe loads can be on at the same time.

Continuous AC LoadsLoad Qty Watts TotalLights 8 40 320PL lights 2 11 22Computer 2 200 400Projector 1 300 300Microwave 1 800 800

Total continuous load 1842

The refrigerator is 200 Watts continuous. At 6 times the rating, this gives a surgepower of 1200 Watts for the refrigerator.

The total surge requirement is 1842 + 1200 = 3042 Watts.

Select an inverter with a surge capacity of at least 3042 Watts.

Results will vary depending on student input.




We first calculate the load size during the week:

System Description: Telecom site - load during the week

Hours DailyDC Loads Qty. Amps /day Demand (Ah)Trans #1 - Transmit : 1 X 10 X 24 = 240.0

Trans #2 - Transmit : 1 X 15 X 8 = 120.0

Trans #2 - Standby : 1 X 0.8 X 16 = 12.8

DC Loads (Ah) 372.8

Hours DailyAC Loads Qty. Watts /day Demand (Wh)

Computers - On : 2 X 300 X 8 = 4,800Computers - Off : 2 X 75 X 16 = 2,400

Lights : 4 X 40 X 8 = 1,280

AC Loads (Wh) 8,480

Continuous Watts = 760Surge Est. = 1560

8,480 ÷ 90% ÷ 24 + 372.8 = 765.4

AC Sub-total Effic Input DCLoads Daily(Ah/Day)

Note: We calculated the Continuous Watt load to be two computers (on) plus all fourlights. So, 2 X 300 + 4 X 40 = 760. The Surge Estimate was calculated by taking6 times the power for the lights and adding it to the computer loads: 2 X 300 + 6 X4 40 = 1560.




We next calculate the average weekly load:

System Description: Telecomm Site - Weekly Average Loads

WeeklyDC Loads Qty. Amps Hours Days/W

kDemand (Ah)

Repeater #1 : 1 X 10 X 24 X 7 = 1680.0

Repeater #2: Transmit (M-F) : 1 X 15 X 8 X 5 = 600.0

Standby (M-F) : 1 X 0.8 X 16 X 5 = 64.0

Transmit(Weekend)

: 1 X 15 X 4 X 2 = 120.0

Standby(Weekend)

: 1 X 0.8 X 20 X 2 = 32.0

Weekly DC Loads (Ah) 2496.0

WeeklyAC Loads Qty. Watts Hours Days/W

kDemand (Wh)

Computer Use:

On - (M-F) : 2 X 300 X 8 X 5 = 24,000

Off - (M-F) : 2 X 75 X 16 X 5 = 12,000

Off - Weekend : 2 X 75 X 24 X 2 = 7,200

Fluorescent Lights : 4 X 40 X 8 X 5 = 6,400

Weekly AC Loads (Wh) 49,600

Continuous Watts = 760

Surge Est. = 1560

49,600 ÷ 0.85 ÷ 24 + 2496.0 = 4927.4

ACLoad

Effic Input DCLoad

Weekly Ah

4927.4 ÷ 7 = 703.9

WeeklyAh

Days Avg Load(Ah/day)

The Continous Watt load and Surge Estimates are the same as above. Note that theload averaged out over the week is slightly less than the load from Monday to Friday,704 Ah/day vs. 765 Ah/day. This is because the weekend loads are lower, whichreduces the average.




When sizing a battery for this system, we should use the higher load (765 Ah/day) forMonday through Friday. This is because the higher load during the week represents asignificant portion of the amount of energy storage. In other words, a 5-day batteryneeds to be sized for the worst 5 days of load. We might size the array based on theweekly load, however, since over a 7-day period, we only need to replace the averageload each day (704 Ah/day).

We first consider the costs of the incandescent bulbs. Over the total time period of8000 hours, the total amount of electricity used by the bulbs is:

8000 hours X 75 Watts = 600,000 Watt-hrs = 600 kilowatt-hours

The cost of this energy is:

600 kilowatt-hours X $0.10 / kWh = $60.

The lifetime of a single incandescent bulb is 1000 hours, so we will need to purchase 8total bulbs for the 8000 hour period. The cost of the bulbs and the energy is:

8 X $0.50 + $60 = $64.

Now we calculate the same costs for the compact fluorescent. The electricity used is:

8000 hours X 15 Watts = 120,000 Watt-hrs = 120 kilowatt-hours

120 kilowatt-hours X $0.10 / kWh = $12.

One compact fluorescent should last for 8000 hours, so we only need a single bulb.The cost of the bulbs and energy is then:

$15 + $12 = $27.

We see that over the lifetime of the compact fluorescent, the total cost is less than halfof the using incandescent bulbs.




We first calculate the costs of using the conventional refrigerator. The energy used inone day by the conventional refrigerator is:

220 Watts X 12 hrs / day = 2640 Watt-hrs

The number of modules required to produce this energy is:

2640 Watt-hrs / 160 Wh per day = 16.5, rounded up to 17 modules.

17 modules X $320 = $5440

The energy required for the super-efficient refrigerator is found the same way:

60 Watts X 13 hrs / day = 780 Watt-hrs

And we need to buy a number of modules equal to:

780 Watt-hrs / 160 Wh per day = 4.9, rounded up to 5 modules.

5 modules X $320 = $1600

We can now compare the total costs:Conventional Refrigerator

Super-efficient Refrigerator

Refrigerator $0 (already purchased) $2500Modules $5440 $1600

Total $5440 $4100

So even with the purchase of a new refrigerator, the cost savings in the energy result ina lower overall cost.



Siemens Solar Basic PV Technology Course Components – Battery TechnologyCopyright © 1998 Siemens Solar Industries

10-1

Chapter TenBattery Technology

In stand-alone photovoltaic power systems, the most important and most poorlyunderstood component next to the PV modules is the battery. There is no oneperfect type of battery for all remote photovoltaic power systems. There are manyfactors that influence the choice and performance of a battery in a photovoltaicsystem. And battery technology is changing, with new constructions andperformance levels available to choose from. The following discussion is intended togive an overview of battery technology and to equip a photovoltaic system designerwith the questions needed to determine which battery is best suited for a particularapplication.




10-2

Purpose of Batteries in Photovoltaic Systems

A storage battery is an electrochemical cell that stores energy in chemical bonds .When a battery is connected in a circuit to dc electrical load, such as anincandescent light or resistor, the chemical energy within a battery is converted toelectrical energy, and there is a flow of current through the circuit. An understandingof storage battery design, terminology and performance characteristics is essentialin the design of stand-alone PV systems.

In stand-alone photovoltaic systems the electrical energy produced by the PV arraycannot always be used when it is produced. Because the demand for energy doesnot always coincide with its production electrical storage batteries are commonlyused in PV systems. The three primary functions of a storage battery in a PV

system are:

• Energy Storage Capacity and Autonomy to store electrical energy when it isproduced by the PV array and to supply energy to electrical loads as needed oron demand.

• Voltage and Current Stabilization to supply power to electrical loads at stablevoltages and currents, by suppressing or 'smoothing out' transients that mayoccur in PV systems.

• Supply Surge Currents to supply surge or high peak operating currents toelectrical loads or appliances.




10-3

Energy Storage Capacity and Autonomy

Because the production of energy from a photovoltaic array may not always coincidewith the energy demand for electrical loads, batteries are required in most stand-alone PV systems. This is perhaps the most important function of batteries in stand-

alone PV power systems – to allow the loads to operate when the PV array by itselfcannot supply enough power. Energy storage is required if electrical loads arerequired to operate at nighttime or during extended periods of cloudy or overcastweather.

In the system sizing process the PV array is generally sized to satisfy the averagedaily load demand during the period with the lowest insolation to electrical load ratio(usually during winter months), to ensure that sufficient energy is available at alltimes of the year. The battery storage capacity is generally sized to meet theaverage daily electrical load for a specified number of days without input from the PVarray, or for a specified autonomy period.

Autonomy or the ‘days of storage’ are often referred to when speaking about thebattery storage capacity of a stand-alone PV system. A stand-alone PV system isdescribed as having "autonomy" if sufficient battery storage capacity is available tooperate the electrical loads directly from the battery, without any energy input fromthe PV array. The greater the design autonomy period the larger the batterycapacity required for a given load demand. For common, less critical PVapplications, autonomy periods are typically designed for between two and six days.For critical applications involving essential loads or public safety or where weatherpatterns dictate, autonomy periods may be greater than ten days. Longer autonomyperiods result in a lower average daily depth of discharge, and lower the probability

that the allowable depth of discharge, or minimum load voltage is reached.




10-4

Voltage and Current Stabilization

Another purpose for batteries in stand-alone PV power systems is to stabilize or“level out” the potential wide variations in voltage and current that may occur in a PVelectrical system. As discussed previously in this manual a PV array can

theoretically operate at an infinite number of operating points between the short-circuit current and open-circuit voltage. When electrical loads are directly connectedto a PV array, the load impedance dictates the operational voltage of the PV array,which may not be optimal to operate the load at its prescribed conditions, or topermit full utilization of the maximum power available from the PV array. Batteriesare used to allow the loads to operate within a prescribed voltage and current range,as well as to ensure that the PV array is operated near its maximum power voltage.By acting as a buffer between the PV array and loads, a battery can also stabilizethe voltage and current supply to electrical loads in which the load powerrequirement oscillates or varies with respect to time.




10-5

Supply Surge Currents

Another important function of batteries in stand-alone PV power systems is to supplysurge or peak currents required by electrical loads that are higher than normalsteady-state operating currents. Since PV devices are inherently current-limited in

their output by short-circuit current and irradiance, a PV array by itself may not beable to supply enough current to meet the surge requirements of some electricalloads. While the PV array may be large enough to supply the total energy neededby a load over a day, it may not be large enough to meet a momentary powerdemand by a load at any particular time. A battery is capable of delivering highcurrents, and can supply large currents to the loads for short periods, while beingcharged by the array at lower currents over the course of a day.

Electric motors are common loads that often require large surge currents for starting.For example in refrigerators, compressors, power tools and other motor loads, thesurge current may be 5 to 10 times the normal running or operating current level. A

battery can supply hundreds of amperes for short periods, thereby meeting thesemomentary surge requirements.




10-6

Battery Design and Construction

Battery manufacturing is an intensive, heavy industrial process involving the use ofhazardous and toxic materials. Batteries are generally mass produced, combiningseveral sequential and parallel processes to construct a complete battery unit. After

production, initial charge and discharge cycles are conducted on batteries beforethey are shipped to distributors and consumers.

Manufacturers have variations in the details of their battery construction, but somecommon construction features can be described for most all batteries. Someimportant components of battery construction are described below.

Cell

The cell is the basic electrochemical unit in a battery consisting of a set of positive and negative plates divided by separators, immersed in an electrolyte solution andenclosed in a case . In a typical lead-acid battery, each cell has a nominal voltage ofabout 2.1 volts, so there are 6 series cells in a nominal 12-volt battery. Figure 10-1shows a diagram of a basic battery cell.

Active Material

The active materials in a battery are the raw composition materials that form the

positive and negative plates, and are reactants in the electrochemical cell . Theamount of active material in a battery is proportional to the capacity a battery candeliver. In lead-acid batteries, the active materials are lead dioxide (PbO2) in thepositive plates and metallic sponge lead (Pb) in the negative plates, which react witha sulfuric acid (H2SO4) solution during battery operation.




10-7

Batter Cell Com osition

Active material

Grid Grid

Separator

Electrolyte

Case

Active material

Electrical load

Negative platePositive plate

ElectrolyteThe electrolyte is a conducting medium that allows the flow of current through ionictransfer or the transfer of electrons between the plates in a battery. In a lead-acidbattery the electrolyte is a diluted sulfuric acid solution, either in liquid (flooded) form,gelled or absorbed in glass mats. In flooded nickel-cadmium cells, the electrolyte isan alkaline solution of potassium hydroxide and water. In most flooded batterytypes, periodic water additions are required to replenish the electrolyte lost throughgassing. When adding water to batteries it is very important to use distilled or de-mineralized water, as even the impurities in normal tap water can poison the batteryand result in premature failure.




10-8

Grid

In a lead-acid battery the grid is typically a lead alloy framework that supports theactive material on a battery plate , and which also conducts current. Alloying

elements such as antimony and calcium are often used to strengthen the lead grids,and have characteristic effects on battery performance such as cycle performanceand gassing . Some grids are made by expanding a thin lead alloy sheet into a flatplate web. Others are made of long spines of lead with the active material platedaround them forming tubes, or what are referred to as tubular plates .

Plate

A plate is a basic battery component, consisting of a grid and active material,sometimes called an electrode . There are generally a number of positive andnegative plates in each battery cell , typically connected in parallel at a bus bar orinter-cell connector at the top of the plates. A pasted plate is manufactured byapplying a mixture of lead oxide , sulfuric acid , fibers and water on to the grid .

The thickness of the grid and plate affect the deep cycle performance of a battery.In automotive starting or SLI type batteries many thin plates are used per cell. Thisresults in maximum surface area for delivering high currents, but not much thicknessand mechanical durability for deep and prolonged discharges. Thick plates are usedfor deep cycling applications such as for forklifts, golf carts and other electricvehicles. The thick plates permit deep discharges over long periods, whilemaintaining good adhesion of the active material to the grid, resulting in longer life.

Separator

A separator is a porous, insulating divider between the positive and negative plates in a battery, used to keep the plates from coming into electrical contact and short-circuiting, and which also allows the flow of electrolyte and ions between the positiveand negative plates. Separators are made from microporous rubber, plastic orglass-wool mats. In some cases, the separators may be like an envelope, enclosingthe entire plate and preventing shed materials from creating short circuits at the

bottom of the plates.




10-10

Battery Types and ClassificationsMany types and classifications of batteries are manufactured today, each withspecific design and performance characteristics suited for particular applications.Each battery type or design has its individual strengths and weaknesses. In PVsystems, lead-acid batteries are most common due to their wide availability in many

sizes, low cost and well understood performance characteristics. In a few critical,low temperature applications nickel-cadmium cells are used, but their high initial costlimits their use in most PV systems. There is no “perfect battery” and it is the task ofthe PV system designer to decide which battery type is most appropriate for eachapplication.

In general, electrical storage batteries can be divided into to major categories,primary and secondary batteries.

Primary BatteriesPrimary batteries can store and deliver electrical energy, but cannot be recharged .Typical carbon-zinc and lithium batteries commonly used in consumer electronicdevices are primary batteries. Primary batteries are not used in PV systemsbecause they can not be recharged.

Secondary Batteries

A secondary battery can store and deliver electrical energy, and can also be recharged by passing a current through it in an opposite direction to the dischargecurrent. Common lead-acid batteries used in automobiles and PV systems aresecondary batteries. The table lists common secondary battery types and theircharacteristics that are important to PV system designers. A detailed discussion ofeach battery type follows.




10-11

Secondary Battery Types and Characteristics

Battery Type Cost Deep CyclePerformance

Maintenance

Flooded Lead-Acid

Lead-Antimony low good high

Lead-Calcium Open Vent low poor medium

Lead-Calcium Sealed Vent low poor low

Lead Antimony/Calcium Hybrid medium good medium

Captive Electrolyte Lead-Acid

Gelled medium fair low

Absorbed Glass Mat medium fair low

Nickel-Cadmium

Sintered-Plate high good none

Pocket-Plate high good medium




10-12

Lead-Acid Battery Classifications5.3.3.1 Lead-Acid BatteriesMany types of lead-acid batteries are used in PV systems, each having specific

design and performance characteristics. While there are many variations in thedesign and performance of lead-acid cells they are often classified in terms of one ofthe following three categories.

SLI Batteries

Starting, lighting and ignition (SLI) batteries are types of lead-acid batteries designedprimarily for shallow cycle service, most often used to power automobile starters.These batteries have a number of thin positive and negative plates per cell designedto increase the total plate active surface area. The large number of plates per cellallows the battery to deliver high discharge currents for short periods. While theyare not designed for long life under deep cycle service, SLI batteries are sometimesused for PV systems in developing countries where they are the only types of batterylocally manufactured. Although not recommended for most PV applications, SLIbatteries may provide up to two years of useful service in small stand-alone PVsystems where the average daily depth of discharge is limited to 10-20%, and themaximum allowable depth of discharge is limited to 40-60%.

Motive Power or Traction Batteries

Motive power or traction batteries are a type of lead acid battery designed for deepdischarge cycle service, typically used in electrically operated vehicles andequipment such as golf carts, fork lifts and floor sweepers. These batteries have afewer number of plates per cell than SLI batteries, however the plates are muchthicker and constructed more durably. High content lead-antimony grids areprimarily used in motive power batteries to enhance deep cycle performance.Traction or motive power batteries are very popular for use in PV systems due totheir deep cycle capability, long life and durability of design.

Stationary Batteries




10-15

These are typically flooded batteries, with capacity ratings of over 200 ampere-hours. A common design for this battery type uses lead-calcium tubular positive electrodes and pasted lead-antimony negative plates. This design combines theadvantages of both lead-calcium and lead-antimony design, including good deepcycle performance, low water loss and long life. Stratification and sulfation can alsobe a problem with these batteries, and must be treated accordingly. These batteries

are sometimes used in PV systems with larger capacity and deep cyclerequirements. A common hybrid battery using tubular plates is the Exide Solarbattery line manufactured in the United States.




10-16

Captive Electrolyte Lead-Acid Batteries

Captive electrolyte batteries are another type of lead-acid battery. As the nameimplies, the electrolyte is immobilized in some manner and the battery is sealedunder normal operating conditions. Under excessive overcharge, the normally

sealed vents open under gas pressure. Often captive electrolyte batteries arereferred to as valve regulated lead acid (VRLA) batteries, noting the pressureregulating mechanisms on the cell vents. Electrolyte cannot be replenished in thesebattery designs, therefore they are intolerant of excessive overcharge.

Captive electrolyte lead-acid batteries are popular for PV applications because theyare spill proof and easily transported, and they require no water additions makingthem ideal for remote applications were maintenance is infrequent or unavailable.However, a common failure mode for these batteries in PV systems is excessiveovercharge and loss of electrolyte, which is accelerated in warm climates. For thisreason, it is essential that the battery charge controller regulation set points are

adjusted properly to prevent overcharging.

This battery technology is very sensitive to charging methods, regulation voltage andtemperature extremes. Optimal charge regulation voltages for captive electrolytebatteries vary between designs, so it is necessary to follow manufacturersrecommendations when available. When no information is available, the chargeregulation voltage should be limited to no more than 14.2 volts at 25

oC for nominal

12-volt batteries. The recommended charging algorithm is constant-voltage , withtemperature compensation of the regulation voltage required to prevent overcharge.

A benefit of captive or immobilized electrolyte designs is that they are less

susceptible to freezing compared to flooded batteries. Typically, lead-calcium gridsare used in captive electrolyte batteries to minimize gassing, however some designsuse lead-antimony/calcium hybrid grids to gain some of the favorable advantages oflead-antimony batteries.

In the United States, about one half of the small remote PV systems being installeduse captive electrolyte, or sealed batteries. The two most common captiveelectrolyte batteries are the gelled electrolyte and absorbed glass mat designs.




10-17

Gelled Batteries

Initially designed for electronic instruments and consumer devices, gelled lead-acidbatteries typically use lead-calcium grids. The electrolyte is 'gelled' by the addition ofsilicon dioxide to the electrolyte, which is then added to the battery in a warm liquidform and gels as it cools. Gelled batteries use an internal recombinant process to

limit gas escape from the battery, reducing water loss. Cracks and voids developwithin the gelled electrolyte during the first few cycles, providing paths for gastransport between the positive and negative plates, facilitating the recombinantprocess.

Some gelled batteries have a small amount of phosphoric acid added to theelectrolyte to improve the deep discharge cycle performance of the battery. Thephosphoric acid is similar to the common commercial corrosion inhibitors and metalpreservers, and minimizes grid oxidation at low states of charge. Gelled batteriesrepresent over 90% of the captive electrolyte batteries used in small PV systems inthe United States.

Absorbed Glass Mat (AGM) Batteries

Another sealed, or valve regulated lead-acid battery, the electrolyte in an AGMbattery is absorbed in glass mats that are sandwiched in layers between the plates.However, the electrolyte is not gelled. Similar in other respects to gelled batteries,AGM batteries are also intolerant to overcharge and high operating temperatures.Recommended charge regulation methods stated above for gelled batteries alsoapply to AGMs.

A key feature of AGM batteries is the phenomenon of internal gas recombination.As a charging lead-acid battery nears full state of charge, hydrogen and oxygengasses are produced by the reactions at the negative and positive plates,respectively. In a flooded battery these gasses escape from the battery through thevents, thus requiring periodic water additions. In an AGM battery the excellent iontransport properties of the liquid electrolyte held suspended in the glass mats, theoxygen molecules can migrate from the positive plate and recombine with the slowlyevolving hydrogen at the negative plate and form water again. Under conditions ofcontrolled charging the pressure relief vents in AGM batteries are designed toremain closed, preventing the release of any gasses and water loss.

While these batteries are successfully used in PV systems, unfavorableperformance of AGM batteries in early PV applications has limited their use to lessthan 10% of the captive electrolyte batteries currently used in small PV systems inthe United States.




10-18

Lead-Acid Battery Chemistry

Now that the basic components of a battery have been described, the overallelectrochemical operation of a battery can be discussed. Referring to Figure 10-1,the basic lead-acid battery cell consists of sets positive and negative plates, divided

by separators, and immersed in a case with an electrolyte solution. In a fullycharged lead-acid cell the positive plates are lead dioxide (PbO2), the negativeplates are sponge lead (Pb), and the electrolyte is a diluted sulfuric acid solution.When a battery is connected to an electrical load, current flows from the battery asthe active materials are converted to lead sulfate (PbSO4).

Lead-Acid Cell Reaction

The following equations show the electrochemical reactions for the lead-acid cell.

During battery discharge the directions of the reactions listed goes from left to right.During battery charging, the direction of the reactions is reversed, and the reactionsgo from right to left. Note that the elements as well as charge are balanced on bothsides of each equation.

Lead-Acid Cell Electrochemical Reactions

At the positive plate or electrode:

PbO H e Pb H O2

2

24 2 2+ + ⇔ ++ − +

Pb SO PbSO2

4

2

4

+ −+ ⇔

At the negative plate or electrode:

Pb Pb e⇔ ++ −2

2

Pb SO PbSO2

4

2

4

+ −+ ⇔

Overall lead-acid cell reaction:

PbO Pb H SO PbSO H O2 2 4 4 22 2 2+ + ⇔ +




10-19

Some consequences of these reactions are interesting and important. As thebattery is discharged, the active materials PbO2 and Pb in the positive and negativeplates, respectively, combine with the sulfuric acid solution to form PbSO4 andwater. Note that in a fully discharged battery the active materials in both the positiveand negative plates are converted to PbSO4, while the sulfuric acid solution isconverted to water. This dilution of the electrolyte has important consequences interms of the electrolyte specific gravity and freezing point that will be discussed later.

Formation

Forming is the process of initial battery charging during manufacture. Formation of alead-acid battery changes the lead oxide (PbO) on the positive plate grids to lead dioxide (PbO2), and to metallic sponge lead (Pb) on the negative plates. The extentto which a battery has been formed during manufacture dictates the need foradditional cycles in the field to achieve rated capacity .

Stratification

Stratification is a condition that can occur in flooded lead-acid batteries in which theconcentration or specific gravity of the electrolyte increases from the bottom to top ofa cell. Stratification is generally the result of undercharging, or not providing enoughovercharge to gas and agitate the electrolyte during finish charging. Prolongedstratification can result in the bottom of the plates being consumed, while the upperportions remaining in relatively good shape, reducing battery life and capacity. Tallstationary cells, typically of large capacity, are particularly prone to stratificationwhen charged at low rates. Periodic equalization charges thoroughly mix theelectrolyte and can prevent stratification problems.




10-20

Specific Gravity

Specific gravity is defined as the ratio of the density of a solution to the density ofwater, typically measured with a hydrometer . By definition, water has a specificgravity of one. In a lead-acid battery, the electrolyte is a diluted solution of sulfuric

acid and water. In a fully charged battery, the electrolyte is approximately 36%sulfuric acid by weight, or 25% by volume, with the remainder water. The specificgravity of the electrolyte is related to the battery state of charge , depending on thedesign electrolyte concentration and temperature .

In a fully charged flooded lead-acid battery, the specific gravity of the electrolyte istypically in the range of 1.250 to 1.280 at a temperature of 27

oC, meaning that the

density of the electrolyte is between 1.25 and 1.28 times that of pure water. Whenthe battery is discharged, the hydrogen (H

+) and sulfate (SO4

2-) ions from the sulfuric

acid solution combine with the active materials in the positive and negative plates toform lead sulfate (PbSO4), decreasing the specific gravity of the electrolyte. As the

battery is discharged to greater depths, the sulfuric acid solution becomes diluteduntil there are no ions left in solution. At this point the battery is fully discharged,and the electrolyte is essentially water with a specific gravity of one.

Concentrated sulfuric acid has a very low freezing point (less than -50oC) while

water has a much higher freezing point of 0oC. This has important implications in

that the freezing point of the electrolyte in a lead-acid battery varies with theconcentration or specific gravity of the electrolyte. As the battery becomesdischarged, the specific gravity decreases resulting in a higher freezing point for theelectrolyte.

Lead-acid batteries used in PV systems may be susceptible to freezing in someapplications, particularly during cold winters when the batteries may not be fullycharged during below average insolation periods. The PV system designer mustcarefully consider the temperature extremes of the application along with theanticipated battery state of charge during the winter months to ensure that lead-acidbatteries are not subjected to freezing. Table 10-2 shows the properties andfreezing points for sulfuric acid solutions.




10-21

Properties of Sulfuric Acid Solutions

Specific Gravity H2SO4 (Wt%) H2SO4 (Vol%) Freezing Point(oC)

1.000 0.0 0.0 0

1.050 7.3 4.2 -3.31.100 14.3 8.5 -7.81.150 20.9 13.0 -151.200 27.2 17.1 -271.250 33.4 22.6 -521.300 39.1 27.6 -71

Adjustments to Specific Gravity

In very cold or tropical climates the specific gravity of the sulfuric acid solution inlead-acid batteries is often adjusted from the typical range of 1.250 to 1.280. Intropical climates where freezing temperatures do not occur, the electrolyte specificgravity may be reduced to between 1.210 and 1.230 in some battery designs. Thislower concentration electrolyte will lessen the degradation of the separators andgrids and prolong the battery’s useful service life. However, the lower specificgravity decreases the storage capacity and high discharge rate performance of thebattery. Generally, these factors are offset by the fact that the battery is generallyoperating at higher than normal temperatures in tropical climates.

In very cold climates, the specific gravity of the electrolyte may be increased above

the typical range of 1.250 to 1.280 to values between 1.290 and 1.300. Byincreasing the electrolyte concentration, the electrochemical activity in the battery isaccelerated, improving the low temperature capacity and lowers the potential forbattery freezing. However, these higher specific gravities generally reduce theuseful service life of a battery.

While the specific gravity can also be used to estimate the state of charge of a lead-acid battery, low or inconsistent specific gravity reading between series connectedcells in a battery may indicate sulfation, stratification, or lack of equalization betweencells. In some cases a cell with low specific gravity may indicate a cell failure orinternal short-circuit within the battery. Measurement of specific gravity can be a

valuable aid in the routine maintenance and diagnostics of battery problems instand-alone PV systems.




10-22

Sulfation

Sulfation is a normal process that occurs in lead-acid batteries resulting fromprolonged operation at partial states of charge. Even batteries that are frequentlyfully charged suffer from the effects of sulfation as the battery ages. The sulfation

process involves the growth of lead sulfate crystals on the positive plate, decreasingthe active area and capacity of the cell. During normal battery discharge, the activematerials of the plates are converted to lead sulfate. The deeper the discharge, thegreater the amount of active material that is converted to lead sulfate. Duringrecharge the lead sulfate is converted back into lead dioxide and sponge lead on thepositive and negative plates, respectively. If the battery is recharged soon afterbeing discharged the lead sulfate converts easily back into the active materials.

However, if a lead-acid battery is left at less than full state of charge for prolongedperiods (days or weeks), the lead sulfate crystallizes on the plate and inhibits theconversion back to the active materials during recharge. The crystals essentially

“lock away” active material and prevent it from reforming into lead and lead dioxide,effectively reducing the capacity of the battery. If the lead sulfate crystals grow toolarge they can cause physical damage to the plates. Sulfation also leads to higherinternal resistance within the battery, making it more difficult to recharge.

Sulfation is a common problem experienced with lead-acid batteries in many PVapplications. As the PV array is sized to meet the load under average conditions,the battery must sometimes be used to supply reserve energy during periods ofexcessive load usage or below average insolation. As a consequence, batteries inmost PV systems normally operate for some length of time over the course of a yearat partial states of charge, resulting in some degree of sulfation. The longer the

period and greater the depth of discharge, the greater the extent of sulfation.

To minimize sulfation of lead acid batteries in photovoltaic systems the PV array isgenerally designed to recharge the battery on the average daily conditions during theworst insolation month of the year. By sizing for the worst month’s weather, the PVarray has the best chance of minimizing the seasonal battery depth of discharge. Inhybrid systems using a backup source such as a generator or wind turbine, thebackup source can be effectively used to keep the batteries fully charged even if thePV array can not. In general, proper battery and array sizing, as well as periodicequalization charges can minimize the onset of sulfation.




10-23

Nickel-Cadmium Batteries

Nickel-cadmium (Ni-Cad) batteries are secondary , or rechargeable batteries andhave several advantages over lead-acid batteries that make them attractive for usein stand-alone PV systems. These advantages include long life , low maintenance ,

survivability from excessive discharges, excellent low temperature capacity retention , and non-critical voltage regulation requirements. The main disadvantagesof nickel-cadmium batteries are their high cost and limited availability compared tolead-acid designs.

A typical nickel-cadmium cell consists of positive electrodes made from nickel- hydroxide (NiO(OH))and negative electrodes made from cadmium (Cd) andimmersed in an alkaline potassium hydroxide (KOH) electrolyte solution. When anickel-cadmium cell is discharged, the nickel hydroxide changes form (Ni(OH)2) andthe cadmium becomes cadmium hydroxide (Cd(OH)2). The concentration of theelectrolyte does not change during the reaction so the freezing point stays very low.

Sintered Plate Ni-Cads

Sintered plate nickel cadmium batteries are commonly used in electrical testequipment and consumer electronic devices. The batteries are designed by heatprocessing the active materials and rolling them into metallic case. The electrolytein sintered plate nickel-cadmium batteries is immobilized, preventing leakage,allowing any orientation for installation. The main disadvantage of sintered platedesigns is the so called 'memory effect', in which a battery that is repeatedly

discharged to only a percentage of its rated capacity will eventually 'memorize' thiscycle pattern, and will limit further discharge resulting in loss of capacity. In somecases the 'memory effect' can be erased by conducting special charge anddischarge cycles, regaining some of its initial rated capacity.

Pocket Plate Ni-Cads

Large nickel cadmium batteries used in remote telecommunications systems andother commercial applications are typically of a flooded design, called flooded

pocket plate . Similar to flooded lead-acid designs, these batteries require periodicwater additions, however, the electrolyte is an alkaline solution of potassiumhydroxide, instead of a sulfuric acid solution. These batteries can withstand deepdischarges and temperature extremes much better than lead-acid batteries, and theydo not experience the 'memory effect' associated with sintered plate Ni-Cads. Themain disadvantage of pocket plate nickel cadmium batteries is their high initial cost,however their long lifetimes can result in the lowest life cycle cost battery for somePV applications.




10-24

Nickel-Cadmium Battery Chemistry

Following are the electrochemical reactions for the flooded nickel-cadmium cell:

Nickel-Cadmium Cell Reactions


2 2 2 2 22 2 NiO OH H O e Ni OH OH ( ) ( )+ + ⇔ +

− −


Cd OH Cd OH e+ ⇔ +− −

2 22( )

Overall nickel cadmium cell reaction:

Cd NiO OH H O Cd OH Ni OH + + ⇔ +2 2 22 2 2( ) ( ) ( )

Notice these reactions are reversible and that the elements and charge are balancedon both sides of the equations. The discharge reactions occur from left to right,while the charge reactions are reversed.

The nominal voltage for a nickel-cadmium cell is 1.2 volts, compared to about 2.1

volts for a lead-acid cell, requiring 10 nickel-cadmium cells to be configured in seriesfor a nominal 12-volt battery. The voltage of a nickel-cadmium cell remains relativelystable until the cell is almost completely discharged, where the voltage drops offdramatically. Nickel-cadmium batteries can accept charge rates as high as C/1, andare tolerant of continuous overcharge up to a C/15 rate. Nickel-cadmium batteriesare commonly subdivided in to two primary types; sintered plate and pocket plate .




10-25

Exercises




10-26




10-27

Battery Strengths and Weaknesses

Each battery type has design and performance features suited for particularapplications. Again, no one type of battery is ideal for PV system applications. Thedesigner must consider the advantages and disadvantages of different batteries with

respect to the requirements of a particular application. Some of the considerationsinclude lifetime, deep cycle performance, tolerance to high temperatures andovercharge, maintenance and many others. The table below summarizes some ofthe key advantages and disadvantages of the different battery types discussed in thepreceding section.

Battery Type Advantages DisadvantagesFlooded Lead-Acid

Lead-Antimony low cost, wide availability,good deep cycle and hightemperature performance,

can replenish electrolyte

high water loss andmaintenance

Lead-Calcium Open Vent low cost, wide availability,low water loss, canreplenish electrolyte

poor deep cycle performance,intolerant to hightemperatures and overcharge

Lead-Calcium Sealed Vent low cost, wide availability,low water loss

poor deep cycle performance,intolerant to hightemperatures andovercharge, can not replenishelectrolyte

Lead Antimony/ Calcium Hybrid

medium cost, low waterloss

limited availability, potentialfor stratification

Captive Electrolyte Lead-Acid Gelled medium cost, little or no

maintenance, lesssusceptible to freezing,install in any orientation

fair deep cycle performance,intolerant to overcharge andhigh temperatures, limitedavailability

Absorbed Glass Mat medium cost, little or nomaintenance, lesssusceptible to freezing,install in any orientation

fair deep cycle performance,intolerant to overcharge andhigh temperatures, limitedavailability

Nickel-Cadmium

Sealed Sintered-Plate wide availability, excellentlow and high temperatureperformance,

maintenance free

only available in lowcapacities, high cost, sufferfrom ‘memory’ effect

Flooded Pocket-Plate excellent deep cycle andlow and high temperatureperformance, tolerance toovercharge

limited availability, high cost,water additions required




10-28

Battery Cycling

One of the main parameters that distinguishes battery types is their ability to “cycle”.A battery cycle refers to the process of charging and discharging a battery. Batterydischarge is the process that occurs when a battery delivers current, quantified by

the discharge current or rate . Charging is the process when a battery receives oraccepts current, quantified by the charge current or rate . A discharge followed by arecharge is considered one cycle.

The discharge can be very small or shallow, or it can be very severe or deep. A 100percent depth of discharge cycle provides a measure of the total battery capacity ata given discharge rate . All batteries can be cycled, but the question is how deeplyand how many times before a permanent loss of capacity occurs. Batteries used inphotovoltaic applications will definitely be subjected to cycling on a daily basis, andperhaps deeply cycled occasionally.

We can think of a battery as being “full of charge”, even though it is actually full ofchemicals that hold potential energy. We use the capacity of a battery bank in aphotovoltaic power system to operate the loads during each night, and duringperiods of heavy load use or below average insolation. If a series of below averageweather days occurs in a row, then the battery is not fully recharged at the end ofeach daily cycle, and the capacity and state of charge of the battery reduces daily.

Battery “Cycling”

• One discharge andrecharge is one “cycle”

• Battery li fe depends onhow deep and howman y times the batteryis cycled

• Batteries designed forshallow cyc ling willwork, but for short time

• Solar systems needdeep cycling batteries

Battery “capacity”

one day’s use

leave unused

available

for use

+ -




10-29

Battery Discharging

When a battery is discharged the chemical reaction on the plates proceeds inwardtoward the grid. The deeper the discharge, the deeper the chemical reaction occurs.In lead-acid batteries the lead-sulfate molecules that are formed are larger than the

lead or lead-oxide molecules, and the bonding of the active material to the plates isgradually weakened due to grid growth. Figure 10-3 illustrates the dischargeprocess.

Discharge Process

• Discharge reactionproceeds inward towardgrid

• 100% dischargeweakens adhesion

• Increased resistanceproduces heat

• Degradation accelerates

grid

active material

coated onto grid




10-30

Charge/Discharge Rates Expressed asTime

Battery manufacturers often refer to rates of charge or discharge not in amperes but

by the time it would take to completely charge or discharge a battery at a specificcurrent. The rate of charge or discharge of a battery is expressed as a ratio of thenominal battery capacity to the charge or discharge time period in hours. Forexample, a 100 Ah battery is discharging at the rate of 2 amps. The time tocompletely discharge a fully charged battery at this rate would be the capacitydivided by the current, or 100 Ah / 2 amps = 50 hours. So we would say that thebattery is discharging at the "50 hour rate" or at "C/50".

Rate in Amperes =Capacity in Amp - Hours

Time in Hours

C

T=

This notation is helpful because it allows us to talk about relative rates of batterycharge and discharge, without referring to the exact size of a battery. For example,most manufacturers recommend charging their batteries no faster than the C/5 rateto limit gassing and overcharge. This means 20 amps for a 100 Ah battery, and 100amps for a 500 Ah battery. Moderate charge rates are around C/20 or C/30, whiletrickle charging at C/100 will hardly produce any gassing at all in most batteries.

This notation is used for discussing discharge rates as well. For example, batteriesused for UPS systems generally have their capacity measured at the C/5 or C/2 hourrate, because in the event of a utility power failure, the UPS system is expected tooperate for only 2 or 5 hours. Batteries used in electric forklift operations will havetheir capacity rated at the C/8 hour rate, because it is anticipated that they will bedischarged during a typical 8-hour shift.

Batteries used in typical PV systems experience very low rates of charge anddischarge compared to these common industrial applications. For example, themaximum charge rates from the PV array to battery are commonly about C/40, andtypical discharge rates supplied to the loads may be as low as C/100 to C/200.




10-31

Depth of Discharge (DOD)

The depth of discharge (DOD) of a battery is defined as the percentage of capacitythat has been withdrawn from a battery compared to the total fully charged capacity.By definition, the depth of discharge and state of charge of a battery add to 100

percent. The two common qualifiers for depth of discharge in PV systems are theallowable or maximum DOD and the average daily DOD and are described asfollows:

Allowable Depth of Discharge

The maximum percentage of full-rated capacity that can be withdrawn from a batteryis known as its allowable depth of discharge. The allowable DOD is the maximumdischarge limit for a battery, generally dictated by the cut off voltage and discharge rate . In stand-alone PV systems the low voltage load disconnect (LVD) set point of

the battery charge controller dictates the allowable DOD limit at a given dischargerate. Furthermore, the allowable DOD is generally a seasonal deficit, resulting fromlow insolation, low temperatures and/or excessive load usage. Depending on thetype of battery used in a PV system, the design allowable depth of discharge may beas high as 80% for deep cycle, motive power batteries, to as low as 15-25% if SLIbatteries are used. The allowable DOD is related to the autonomy , in terms of thecapacity required to operate the system loads for a given number of days withoutenergy from the PV array. A system design with a lower allowable DOD will result ina shorter autonomy period.

Average Daily Depth of Discharge

The average daily depth of discharge is the percentage of the full-rated capacity thatis withdrawn from a battery with the average daily load profile. If the load variesseasonally, for example in a PV lighting system, the average daily DOD will begreater in the winter months due to the longer nightly load operation period. For PVsystems with a constant daily load, the average daily DOD is generally greater in thewinter due to lower battery temperature and lower rated capacity. Depending on therated capacity and the average daily load energy, the average daily DOD may varybetween only a few percent in systems designed with a lot of autonomy, or as highas 50 percent for marginally sized battery systems. The average daily DOD is

inversely related to autonomy; meaning that systems designed for longer autonomyperiods (more capacity) have a lower average daily DOD.




10-32

Autonomy Determines Capacity and Depthof Discharge

The number of days of reserve or autonomy is the main factor that determines the

size of the battery and therefore the magnitude of the daily battery depth ofdischarge. The greater the number of days of autonomy sized into the battery, thelarger the total capacity and therefore the smaller the percentage used each day fora typical daily cycle.

The relationship between days of autonomy and depth of discharge is shown below,for both deep cycling batteries (maximum allowable DOD = 80%) and for shallowcycling batteries (maximum allowable DOD = 50%).

Autonomy (Reserve)Determines Daily Discharge

1 3 5 7 9 11 13 15 17 19

Number of Days of Autonomy

Depth of

80

70

60

50

40

30

20

10

0

50% Maximum D.O.D.

80% Maximum D.O.D.

Discharge (%)




10-33

Example: A photovoltaic power system is designed with deep cycling batteries togive 5 days of autonomous operation to a final discharge depth of80%. Reading up from 5 days on the graph to the “80% MaximumDOD” line, the typical daily depth of discharge would be only 15%.

If the battery were a shallow cycling type and could only be 50%discharged at the end of 5 days, the average daily depth of dischargewould be only 10%.

Since most remote PV power systems are designed with at least 4 days ofautonomy, batteries in remote photovoltaic systems will be shallow cycled on anaverage daily basis, whether deep or shallow cycling type batteries are used.However, we recommend using deep cycling batteries in remote photovoltaic powersystems, even though they will typically be shallow cycled, perhaps 15-10% daily or

even less. By using deep cycling type batteries, the system can withstand theexpected seasonal drops in capacity due to below average insolation. Shallowcycling batteries can be used as well, but shorter life should be expected.




10-34

Exercises




10-35

State of Charge (SOC)

The state of charge (SOC) is defined as the amount of energy in a battery,expressed as a percentage of the energy stored in a fully charged battery.Discharging a battery results in a decrease in state of charge, while charging results

in an increase in state of charge. A battery that has had three quarters of itscapacity removed, or been discharged 75%, is said to be at 25% state of charge.The figure below illustrates the seasonal variation in battery state of charge anddepth of discharge for a typical PV system.




10-36

Temperature Limits Discharge Depth

As discussed earlier, if the internal temperature of a battery reaches the freezingpoint of the electrolyte, the electrolyte can freeze and expand, causing irreversibledamage to the battery. In a fully charged lead-acid battery, the electrolyte is

approximately 35% by weight and the freezing point is quite low (approximately -60oC). As a lead-acid battery is discharged, the becomes diluted, so the concentration

of acid decreases and the concentration of water increases as the freezing pointapproaches the freezing point of water, 0

oC. The figure below shows the

relationship between the maximum allowable depth of discharge and correspondingminimum temperature to avoid freezing in a typical flooded lead-acid battery. Noticethat this is the similar to the information presented in Table 10-2 on the discussion ofspecific gravity.

Temperature Limits MaximumDischarge Depth

-60 -40 -20 0

Lowest Battery Temperature (deg.C)

Maximum D.O.D. (%)

80

60

40

20

0




10-37

The maximum depth of discharge must not be allowed to exceed the point thatwould allow freezing. The photovoltaic power system must be designed with enoughbattery capacity so that after the maximum expected days of autonomous operation,the depth of discharge does not exceed the danger point. If the coldest expectedbattery temperatures are not known, the coldest 24-hour average temperature forthe location may be used.

You can see that if the coldest temperature does not go below -8o

C, then there is noeffect on the allowable maximum depth of discharge to prevent freezing. This factoris only applicable in very cold climates.

Example: At a remote telecommunications site on a mountain top, the coldest24-hour temperature is approximately -20

oC. Read up from -20

oC in

Figure 10-6 to a value of approximately 50%.

This means that the maximum allowable depth of discharge would beapproximately 50%, even if "deep cycling" type batteries were used

and the manufacturer had indicated that they can be designed formaximum discharges of 80%.

The example system should be designed so that at maximum discharge, thebatteries would be only 50% discharged (at the end of the autonomy period). Ifallowed to discharge more during this coldest time of the year, they might be indanger of freezing.




10-38

Exercises




10-39

Shallow and Deep Cycle Batteries

Batteries are sometimes classified as deep or shallow cycling, but there are onlygeneral classifications and are not always truly indicative of actual batteryperformance and cell design. Shallow cycle batteries generally a type of SLI or

automotive starting battery. These batteries are made with lead-calcium grids andshould not be cycled greater than 15% on a daily (or nightly) basis. They should notbe allowed to discharge more than 50% under any circumstances, because theybecome very difficult to recharge from beyond that depth of discharge. This isespecially true for low cost automobile batteries that might be used for simple ruralelectrification systems. These types of batteries are designed to provide largecurrents for short periods, and are not designed to sustain deep discharges. Theywill give about 500-1000 cycles to 15% depth of discharge before loosing too muchcapacity and requiring replacement.

Shallow Cycle Batteries

500 - 1000 cycles to 15% maximum daily discharge

50% maximum allowable depth of discharge

Deep cycle batteries are often traction of motive power types and can handle greaterdischarges for longer times than typical shallow cycle batteries. Even so they should

not be cycled to 100% discharge. Most manufacturers recommend that deep cyclingbatteries be discharged to no more than 80% of the rated capacity, to preventreactions from occurring close to the grid. The expansion from lead and lead dioxidemolecules into large lead sulfate molecules can severely weaken the bonding of theactive materials to the grid, causing increased internal resistance and heating.Typically deep cycle batteries can deliver 1500-1800 cycles to 80% depth ofdischarge before needing replacement, and will deliver 3000-4000 cycles ifdischarged more moderately to 25% depth of discharge or less.

Deep Cycle Batteries

3000 - 4000 cycles to 25% depth of discharge

500 - 1800 cycles to 80% maximum allowable depth of discharge




10-40

The difference in cycle life between deep and shallow cycle batteries is shown in thefigure below. The general trend for both types of batteries is that a greater depth ofdischarge yields a fewer number of cycles. All batteries deliver fewer cycles atgreater depths of discharge.

Depth of Discharge Affects Cycle Life

10 30 50 70

Depth of Discharge (%)

5000

4000

3000

2000

1000

0

Deep Cycling Batteries

Shallow Cycling Batteries

Number of Cycles to 20% Loss of Capacity

What determines when a battery should be replaced? Most all batterymanufacturers recommend that when a battery has lost 20% of its capacity it hasreached the end of its useful life, and should be replaced. The cycle graphpresented above is based on the number of cycles until 20% capacity has been lost,or in other words until only 80% of the original capacity is left. However, this chartshould only be used as a rough guideline for battery cycle life. The specific batterytype and design, temperature, rates, charging methods, maintenance and otherfactors all have an effect on battery cycle life.

Example: If a shallow cycling battery is discharged to approximately 15% each cycle,you could expect about 500 cycles before it would need replacing. Thiswould correspond to about 1-2 years of operation if the cycling were once aday. If it were discharged to about 50% each cycle, you could expect a few

hundred cycles, corresponding to less than a year.

If a deep cycling battery were discharged to 80% each cycle, you couldexpect about 1500-1800 cycles, corresponding to about 4-5 years ofoperation if there were one cycle each day. Cycling to only 20% each cyclecould give about 4000 cycles, or about 10 years of operation.




10-41

Battery Capacity

Capacity is a measure of a battery's ability to store or deliver electrical energy,commonly expressed in units of ampere-hours .

Ampere-Hour Definition

The ampere-hour (Ah) is the common unit of measure for a battery's electricalstorage capacity , obtained by integrating the discharge or charge current in amperesover a specific time period. An ampere-hour is equal to the transfer of one ampere over one hour, equal to 3600 coulombs of charge. For example, a battery thatdelivers 5 amps for 20 hours is said to have delivered 100 ampere-hours.

Factors Affecting Battery Capacity

Capacity is generally specified at a specific discharge rate , or over a certain timeperiod. The capacity of a battery depends on several design factors including; thequantity of active material , the number, design and physical dimensions of theplates , and the electrolyte specific gravity . Operational factors affecting capacityinclude; the discharge rate , depth of discharge , cut-off voltage , temperature, ageand cycle history of the battery. Sometimes a battery's energy storage capacity isexpressed in kilowatt-hours (kWh), which can be approximated by multiplying therated capacity in ampere-hours by the nominal battery voltage and dividing the

product by 1000. For example, a nominal 12 volt, 100 ampere-hour battery has anenergy storage capacity of (12 x 100)/1000 = 1.2 kilowatt-hours.




10-42

Effects of Temperature on Capacity

Cold temperatures decrease the total capacity available from a battery. Duringdischarge, the electrolyte does not penetrate as deeply into the active material onthe plates, and the cut-off voltage is reached sooner at cold temperatures.

PV system designers must be acutely aware of the effects of temperature on batterycapacity. Battery manufacturers generally rate capacity at a temperature of 25

oC.

If the required battery size in a PV system is calculated based on the expectedcapacity at 25

oC, the battery may be too small to provide the storage necessary to

achieve the design autonomy period during cold temperatures. As a result thebattery could be severely discharged and the system loads may not be satisfied.Additional battery capacity must be installed in stand-alone PV power systems tocompensate for the expected reduction in capacity at low temperatures.

Standard Temperature for Rated Battery Capacity = 25o

C

Battery Capacity Must Be Derated For Low Temperature Operation

Conversely, a battery operated at temperatures greater than 25oC will deliver more

than the rated capacity. However, under no circumstances should a battery beheated or operated at elevated temperatures to increase the available capacity.Higher than rated operating temperatures significantly reduce battery life. Mostbattery manufacturers recommend their batteries be operated in temperature rangesof between 20 and 30

oC.

How is this information used in the sizing and design of batteries in PV systems?We apply this data during our battery sizing calculations to insure that even whenthe batteries are at their coldest, the system has the capacity it requires. In otherwords, when we calculate the amount of capacity we need to give the days ofautonomy desired, the calculated value must be increased by an appropriate factor ifthe battery is operated below 25

oC.




10-43

Example: Assume that 1000 ampere-hours of battery capacity is required tooperate a specific load for the required days of autonomy. Alsoassume that the average discharge rate is C/120 or about 8 amps.

If the battery temperature could get as low as -20oC during operation,

1000 Ah of capacity is still required at these lower temperatures. Since

most battery manufacturers rate capacity at 25o

C, the rated capacitymust be derated to the available capacity at -20

oC. How much

capacity is required at 25oC to ensure that 1000 Ah are available at -

20 o

C at the C/50 rate?

To determine the increase in capacity required, we refer to the figurebelow which shows the percent of standard rated capacity (at 25

oC) as

a function of the discharge rate and temperature. At -20 oC at the

C/120 rate, the capacity of a typical lead-acid battery is about 82% ofthe rated capacity at 25

oC. To calculate the increased capacity

required, simply divide the 1000 Ah required at 25oC by 0.82 to give

1000 / 0.82 = 1220 Ah. Therefore, by installing about 1220 Ah ofbattery capacity (rated at 25

oC), we still have 1000 Ah available when

the battery is at -20oC.

30

40

50

60

70

80

90

100

110

120

-30 -20 -10 0 10 20 30 40

C/500 C/120

C/50 C/5

C/0.5

empera ure an sc arge a e

Effects on Lead-Acid Battery Capacity

Battery Operating Temperature -oC

P e

r c e n t o f R a t e d C a p a c i t y




10-44

Discharge Rate Affects Capacity

When a battery is discharged, the lead sulfate and water reaction products get in theway of further reaction. At high discharge rates, the reactions are confined to thelayers of the active material that are in immediate contact with the free electrolyte,

limiting the cell capacity. This is because there is insufficient time for the electrolyteto diffuse into the pores of the plates, and the sulfate molecules forming at thesurface clogs the pores, preventing full use of all the active material. This effect isamplified as the discharge rate increases. The result is that final cut-off voltage isreached sooner and less total capacity is usable at faster rates of discharge. Sobattery capacity is not a fixed value, but depends on how fast the battery isdischarged. Discharging a battery slowly delivers more capacity from the battery,while discharging it quickly delivers less total capacity.

This effect is not permanent. For example, if a battery is discharged at a fast rate, itonly delivers a fraction of its rated capacity. If it is then fully recharged and it is

discharged at a slower rate, then more capacity will be available. However, a fewcycles may need to be performed on the battery to achieve stable capacities at thenew rates.

The rate of discharge must therefore be included in any statement of batterycapacity. Rates are not usually stated in amperes, but in hours for full discharge to aspecified cut-off voltage. Most manufacturers use the ten-hour rate for their nominalratings (some use eight hours or twenty hours). For example, if a battery capacity ispresented by the manufacturer as "180 Ah at the 10 hour rate", this means that if thebattery were discharged for ten hours at a constant C/10 rate (180/10 = 18 amps),then 180 Ah would have been discharged from it to the cut off voltage. Batteries

used for industrial or motive power applications often have capacity measured at the6 or 8 hour rate, because the time of a normal working shift is that long.

The terminology of the rate of discharge in hours, for example the “10 hour rate” ,tells us how fast the battery is being discharged. It does not necessarily mean that abattery will in fact actually be discharged for ten hours and drop in voltage to the finalcut-off voltage. It means only that it is being discharged at a rate that would fullydischarge it in that time. In practice, it may only be discharged for a few hours.

Capacity ratings for longer periods are appropriate for photovoltaic power systemsbecause the battery is usually designed to discharge steadily over many days of bad

weather. For example, discharging a battery bank for a remote home designed withfive days autonomy would take 120 hours (5 days x 24 hours/day). And a batterydesigned for a critical remote telecom repeater with 14 days of autonomy would takeover 300 hours (14 days X 24 hours/day) to completely discharge.




10-45

Typical Discharge Times

Industrial, motive applications 10 hours

Photovoltaic applications 100-300 hours

The time we need to specify is the time to TOTAL discharge (100% DOD). But wedo not allow batteries to fully discharge every cycle. In fact, most manufacturersrecommend that shallow cycling type batteries only be discharged to a maximum of50% of their full capacity, and that deep cycling batteries be discharged to amaximum of 80% of their capacity. This is to prevent weakening of the bondbetween the active materials and the grids.

Maximum Recommended Depth of Discharge for Lead Acid Batteries

Shallow cycling types 50%

Deep cycling types 80%

Thus we need to take into account the maximum allowable depth of discharge incalculating the time it would take to fully discharge a battery. The time (hours) todischarge to the maximum DOD limit is simply the number of days of autonomyreserve times the number of hours each day that the loads operate.

Time to discharge to max. DOD = Days of Reserve X Load Operating Time

For continuous loads, such a microwave repeaters or navigational systems, theoperating time might be 24 hours/day. For simple home lighting systems, the load (asingle fluorescent light for example) might be specified as 4 hours/day. For systemswith a variety of intermittent loads, a “weighted average” for the load operating timecan be calculated by summing each load (amps DC, or watts AC) multiplied times itsoperating time and dividing this by the sum of the load values alone. This “weights”the average load operating time for a collection of different loads based on the sizeof the load and its operating time.




10-46

Load Operating Time:

• Continuous Loads: use 24 hours

• Single Load Systems: use load operating time

• Multiple Load Systems: use weighted average load operating time

Weighted Average Load Operating Time =load time

loads

×∑∑

The formula for the time to total discharge (not just to max. DOD as describedabove) is time to discharge to the maximum DOD divided by the maximum alloweddischarge.

Time to fully discharge = Days of Reserve X Load Operating TimeMaximum Depth of Discharge (%)

Example: The small remote cabin used in the Load Estimation Chapter has avariety of loads. The weighted average load operating time would begiven by dividing the sum of the load X time by the sum of the loadsonly:

Weighted Average = 69.4 AhLoad Time (2 x 3.3) + (3 x .92) + (1 x 3.3) + (1 x 2)

= 4.7 hours

The system will be designed with 4 days of reserve autonomy in thebattery bank. If low cost shallow cycling batteries are to be used, thelimit to maximum DOD is 50%. The battery discharge rate, given intime to total discharge would be given by:

Time to full discharge = 4 days X 4.7 hours/day0.5

= 37.6 hours




10-47

Example: A small remote telemetry system is to be designed with 5 days ofreserve battery capacity. Deep cycling industrial batteries are to beused that can discharge to 80%. The load is a continuous (24-hour)load. The average discharge rate, expressed in hours, is given by

Time to full discharge = 5 days X 24 hours/day

0.80

= 150 hours

In the example a battery is designed to give five days of reserve or autonomy. Wechoose a deep cycling type battery, so the maximum depth of discharge (DOD) canbe up to 80%, or in other words down to 20% state of charge. As the battery isdischarged over time the state of charge decreases. By the end of 5 days or 120hours, we would have discharged the battery to 80% depth of discharge (20% stateof charge) if there were no input from the PV array.

But the figure shows that the time to fully discharge the battery is actually 150 hours,or about 6.25 days. So we are really discharging this particular battery at the “150hour rate” or the C/150 rate, and not the C/120 rate. We would not actuallydischarge this battery over 150 hours to 100% depth of discharge or to 0% state ofcharge, but this is the time that it would take if we did. Although it is this number thatis used to determine the available capacity from the battery with this autonomyperiod.




10-48

0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 00

2 0

4 0

6 0

8 0

1 0 0

1 2 0

Hour Rate Indicates Time toTotal Discharge

Stateof

Charge(%)

Time Duration of Discharge (hours)

Limit of dischargerecommended by manufacturer

# Days X 24 hours

Time to total discharge: #Days X 24 hrs

Max % Discharge

5 days (120 hours) reserve to 80% DOD gives 150 hours discharge time

Literature Example Showing Capacity at Different Rates of

Discharge

A portion of a typical battery capacity specification sheet is shown on the next pageto demonstrate the dependence of the rate of discharge on battery capacity.

As an example, battery type 3-35A27 listed at the bottom of the sheet will deliver468 Ah if discharged in 8 hours, but will deliver 610 Ah if discharged more slowlytaking 48 hours, and will deliver 624 Ah if discharged over 100 hours. Note that thecut-off voltage of 1.75 volts per cell (VPC) is included in the specification. Thiswould mean that when a 12-volt battery, made of two of the 3-35A27 batterymodules connected in series, has reached 10.5 volts (6 x 1.75) the manufacturer

considers the battery to be fully discharged.

This manufacturer has listed a "Nom AH Cap" or nominal ampere-hour capacity foreach battery. The numbers shown are slightly higher than those found in the “8 Hr”column, so they reflect capacity obtained by discharging a bit more slowly than the 8hour rate. Most likely they are using a common battery industry standard 10-hour




10-49

rate. So the manufacturer’s nominal capacity for the 3-35A27 is 470 Ah at the 10hour rate, but the capacity obtained at the slower 100 hour rate would be 624 Ah. Ifa range of rates is given, use the formula presented previously for calculating thetime for full discharge to determine the proper rate to use. If the exact dischargetime is not given by the manufacturer just use the time that is closest or less than theactual time to conservatively size the battery.

Example: In our example of the remote telemetry system presented above, therate of discharge, stated in hours, was given as 150 hours. Themanufacturer’s literature presented here goes only to the 100-hourrate. So use that column to select the capacity available from thebattery. The capacity at the 150 hour rate would be slightly greater,but not that much different.

If 600 Ah at 12 volts were required for the telecom project, you couldselect two 6-35A15 batteries and connect them in parallel, to get 2 x336 Ah = 672 Ah at the 100 hour rate.

Or you could select the 3-35A27 that has 624 Ah at the 100 hour rate,but it is only 6 volts. So you would need to use two of them andconnect them in series to get 12 volts.




10-50

Type Volts NominalAH

Capacity

Horizontal Stacking Dimensions AH Capacities of Single Cell

to 1.75 vpc average

at 25 deg.C.

Width Height Depth Cell Type 8 h 24h

48h

100h

in mm in mm in mm

6-35A05 12 75 17.23 438 8.60 218 14.2 361 35A05 72 86 94 96

6-35A07 12 110 21.67 550 35A07 108 129 141 144

6-35A09 12 145 26.17 665 35A09 144 172 188 192

6-35A11 12 180 30.67 779 35A11 180 215 235 240

6-35A13 12 215 35.17 893 35A13 216 258 282 288

6-35A15 12 250 39.67 1008 35A15 252 301 329 336

3-35A17 6 290 24.52 623 35A17 288 344 376 3843-35A19 6 325 26.77 680 35A19 324 387 423 432

3-35A21 6 360 29.02 737 35A21 360 430 470 480

3-35A23 6 395 31.27 794 35A23 396 473 517 528

3-35A25 6 430 33.52 851 35A25 432 516 563 576

3-35A27 6 470 35.77 909 35A27 468 559 610 624




10-51

Cut-Off Voltage Affects Capacity

The cut off voltage is the lowest voltage that a battery system is allowed to reach inoperation, and defines the battery capacity at a specific discharge rate .Manufacturers often rate capacity to a specific cut off, or end of discharge voltage at

a defined discharge rate. For lead-acid batteries the cut off voltage used to ratecapacity is generally 1.75 volts per cell, or 10.5 volts for a nominal 12-volt battery.The cut off voltage for nickel-cadmium cells is typically 1.0 volt. If the same cut offvoltage is specified for different discharge rates, the capacity will generally be higherat the lower rate. Note that the cut off voltage defined by a battery manufacturermost often represents a fully discharged battery. Typically, batteries used in PVsystems are never allowed to reach this low of a cut off voltage, and in practice aregenerally limited to no more than an 80% depth of discharge as determined by thelow voltage load disconnect point of the battery charge controller.

Battery Voltage Over Time

1.5

1.61.71.81.9

22.12.2

0 20 40 60 80 100

Time (hours)

V

o l t a g e ( v o l t s )

C/100 C/10




10-52

Effects of Self Discharge Rate on Capacity

In open-circuit mode without any load or charging a battery undergoes a naturalreduction in state of charge, due to internal mechanisms and losses within a battery.Different battery types have different self-discharge rates, the most significant factor

being the active materials and grid alloying elements used in the design. Highertemperatures result in higher discharge rates, particularly for lead-antimony designs.The figure below shows the effects of battery cell temperature on the self-dischargerate for lead-acid batteries with lead-calcium and lead-antimony grids.

0

5

10

15

20

-50 -25 0 25 50 75

Lead-Antimony Grid (end of life)

Lead-Antimony Grid (new)

Lead-Calcium Grid (typical)

Lead-Acid Battery Self Discharge Rate

Battery Operating Temperature (oC)

S e l f D i s c h a r g e R a t e

( % o f

r a t e d c a p a c i t y p e r w e e

k )




10-53

Exercises




10-54

Voltage Changes During Chargingand Discharging

We will now begin to examine some of the many factors that affect batteryperformance. We begin with the most basic of characteristics--battery voltage.Nominal values for battery voltage were given previously, namely 2.0 volts for a leadacid battery cell and 1.2 volts for a nickel-cadmium cell. But these values are onlynominal, and the actual voltage of a cell or a battery will depend on several factors.

The fact that batteries operate over a range of voltages has direct impact on at leastthree other components of a typical photovoltaic power system. The DC loads andthe inverter must be designed to accept the wide voltage range typical inphotovoltaic systems. If there is a problem with the extent of the range, a DC powersupply or voltage regulator must be included in the system, to keep the voltage

range with the limits of the inverter or DC loads. And the current or chargeregulators commonly used in photovoltaic systems must be designed to anticipatethe wide battery voltage range.




10-55

Battery Voltage and State of Charge areRelated

A fully charged battery is said to be at 100% state of charge (SOC). As the battery

is discharged, the chemical balances change, and it moves away from 100% SOC tosome partial SOC. The change in SOC is reflected in both the composition anddensity of the electrolyte and in the voltage of the battery.

If the electrolyte could be accessed the state of charge could be determined bymeasuring the specific gravity as discussed earlier. In a fully charged lead-acid cell,the specific gravity is typically about 1.265, while a completely discharged cell wouldhave a specific gravity close to water, or 1.000.

While specific gravity can be used to approximate the battery state of charge inremote PV systems, it is seldom recorded except during periodic maintenance

checks. However, the open-circuit battery voltage (without a charge or dischargeload, battery at rest for some time) may be used to estimate the state of charge.The figure below illustrates that there is a fairly linear relationship between the SOCof a lead-acid battery and the open-circuit voltage.

For example, if a battery were 50% discharged, or at a SOC of 50%, we couldmeasure a specific gravity of approximately 1.17, or a battery open-circuit voltage ofabout 12.0 volts. So in this case, reading 12.0 volts on a "12 volt" battery meansthat the battery is about 50% discharged! Precise digital meters are needed toaccurately determine the voltage and SOC of a battery.




10-56

It is essential to know the immediate history of the battery prior to measurement ofthe battery voltage, since a charge or discharge immediately preceding themeasurement may significantly alter the accuracy of the measurement.

Lead-Acid Battery Voltage andSpecific Gravity as a Function of

State of Charge

Voltage(volts)

13

12

11

100

1.300

1.200

1.100

1.000

SpecificGravity

SpecificGravity

Open Circuit Voltage

State of Charge (%)

80 60 40 20100

Note that there is little variation in specific gravity with state of charge for nickel-cadmium cells and voltage trends are different than lead-acid batteries.




10-57

Battery Voltage Depends on Rate ofCharge or Discharge

Discussing battery open-circuit voltage refers to a battery that is at rest, neither

being charged nor discharged. When a battery is being charged, the voltage istemporarily elevated above its open-circuit voltage. And conversely when underdischarge, the voltage is temporarily depressed below the open-circuit voltage. Thisis primarily due to the internal resistance of the battery. Other reasons are due tothe variation in electrolyte concentration at the immediate vicinity of the pores of theplates, which is a result of the diffusion of electrolyte from an area of higherconcentration (at the bulk of electrolyte) to an area of lower concentration (near theplates) while the battery is under discharge. During the charging process theelectrolyte concentration is higher at the plates when compared to the bulk of theelectrolyte. So not only is voltage affected by the state of charge of the battery, butan added factor is the rate of discharge or charge.

The voltage measured at the battery terminals when the battery is being charged ordischarged is called the terminal voltage. As the rate of charge or discharge isincreased, the deviation from the pure open circuit voltage is increased. Forexample, when charging, the voltage measured at the 20 hour rate of charge (C/20)is always higher than for the 50 hour or 100 hour rate at a given battery SOC.

It is difficult therefore to determine battery state of charge simply by measuringvoltage, because in remote PV power systems the battery may be charging ordischarging. The voltage measured (terminal voltage) may be elevated ordepressed with respect to the accurate measure of SOC, the open circuit voltage.

One practical application of this occurs with charge regulator design in photovoltaicsystems. A typical regulator will allow full array current to flow into the battery bankuntil the voltage begins to rise steeply, usually around 14.5 volts for a 12 volt battery.In Figure 10-13, this corresponds to about 80% SOC. At this steep rise area, thevoltage of the battery could continue to go up, but no progress toward 100% SOCwould occur. This is the region of "gassing", where there is enough voltage to breakup water molecules into hydrogen and oxygen gas. A typical regulator design willreduce the array current at this point. For example, the regulator might reduce thecurrent down from the C/20 rate to the C/50 rate. At the lower rate, the voltagedrops below 14.5, and begins to rise more slowly as the battery approaches 100%SOC. The regulator will continue to decrease the array current as the voltage risesagain, and in this way very gently approach 100% SOC. Referring again to Figure10-13, the battery depicted will be very near 100% SOC at the C/100 rate and still bebelow 14 volts.




10-58

In addition to estimating the state of charge from the battery terminal voltages atdifferent rates of charge and discharge, these curves can also be used to determinethe voltage set point for the charge controller low voltage load disconnect set pointor alarm settings. For example, if it is desired to limit the maximum battery depth ofdischarge to 80% at the 40 hour rate, the corresponding load disconnect voltageshould be about 11.5 volts.

Battery Charging andDischarging

0 20 40 60 80 20 40 60 80 100

Voltage (volts)

17

16

15

14

13

12

11

10

Open Circui t Vol tage

40 hour rate

5 hour rat

20 hour rate

50 hour rate

100 hour rate

State of Charge (%) Depth of Discharge (%)

cut-off voltage




10-59

Battery Charging

Methods and procedures for battery charging vary considerably. In a stand-alonePV system the ways in which a battery is charged are generally much different fromthe charging methods battery manufacturers use to rate battery performance. The

various methods and considerations for battery charging in PV systems arediscussed in the next chapter on battery charge controllers.

Battery manufacturers often refer to three modes of battery charging: normal or bulk charge, finishing or float charge and equalizing charge .

• Bulk or Normal Charge: Bulk or normal charging is the initial portion of acharging cycle, performed at any charge rate which does not cause thecell voltage to exceed the gassing voltage. Bulk charging generally occursup to between 80 and 90% state of charge.

• Float or Finishing Charge: Once a battery is nearly fully charged, mostof the active material in the battery has been converted to its original form,and voltage and or current regulation are generally required to limit theamount over overcharge supplied to the battery. Finish charging is usuallyconducted at low to medium charge rates.

• Equalizing Charge: An equalizing or refreshing charge is usedperiodically to maintain consistency among individual cells. An equalizingcharge generally consists of a current-limited charge to higher voltagelimits than set for the finishing or float charge. For batteries deepdischarged on a daily basis, an equalizing charge is recommended everyone or two weeks. For batteries less severely discharged, equalizing mayonly be required every one or two months. An equalizing charge istypically maintained until the cell voltages and specific gravities remainconsistent for a few hours.

Battery Gassing and Overcharge

Gassing occurs in a battery during charging when the battery is nearly fully charged.At this point, essentially all of the active materials have been converted to their fully

charged composition and the cell voltage rises sharply. The gas products are eitherrecombined internal to the cell as in sealed or valve regulated batteries, or releasedthrough the cell vents in flooded batteries.




10-60

In general, the overcharge or gassing reaction in batteries is irreversible, resulting inwater loss. However in sealed lead-acid cells, an internal recombinant processpermits the reforming of water from the hydrogen and oxygen gasses generatedunder normal charging conditions, allowing the battery to be sealed and requiring noelectrolyte maintenance. All gassing reactions consume a portion of the chargecurrent that cannot be delivered on the subsequent discharge, thereby reducing thebattery charging efficiency.

In flooded lead-acid and nickel-cadmium batteries, gassing results in the formationof hydrogen at the negative plate and oxygen at the positive plate, requiring periodicwater additions to replenish the electrolyte. The following electrochemical reactionsshow the overcharge process in typical lead-acid and nickel-cadmium cells.

Lead-Acid Cell Overcharge Reaction


2 2 2 H e H + −+ ⇒


H O e O H 2

1

2 22 2− ⇒ +− +

Overall lead-acid cell overcharge reaction:

H O H O2 2

12 2

⇒ +

Nickel-Cadmium Cell Overcharge Reaction


4 4 2 42 2

H O e H OH + ⇒ +− −


4 2 42 2OH H O O e− −

⇒ + +

Overall nickel-cadmium cell overcharge reaction:

2 22 2 2

H O H O⇒ +

Flooded Batteries Require Some Gassing




10-61

Some degree of gassing is required to agitate and prevent stratification of theelectrolyte in flooded batteries. When a flooded lead-acid battery is charged, heavysulfuric acid forms on the plates, and falls to the bottom of the battery. Over time theelectrolyte stratifies, developing greater acid concentrations at the bottom of thebattery than at the top. If left unmixed the reaction process would be different from

the bottom to the top of the plates, greater corrosion would occur, and batteryperformance would be poor. By gently gassing flooded batteries, the electrolyte ismixed preventing electrolyte stratification. However, excessive gassing andovercharge dislodges active materials from the grids, reducing the battery life.Excessive gassing may also lead to higher temperatures, which acceleratescorrosion of the grids and shortens battery life.

Captive Electrolyte Batteries Should Avoid Gassing

Gassing control is especially important for captive electrolyte or sealed batteries.

These are not flooded, and electrolyte cannot be replaced if allowed to escape dueto overcharging. These batteries do not need for their electrolyte to be mixed, as inflooded batteries. For these types of batteries, the charging process should becontrolled more carefully to avoid gassing.




10-62

Charge Regulation Voltage AffectsGassing

The charge regulation voltage, or the maximum voltage that a charge controller

allows a battery to reach in operation plays an important part in battery gassing.Charge controllers or regulators are used in photovoltaic power systems to allowhigh rates of charging up to the gassing point, and then limit or disconnect the PVcurrent to prevent overcharge. The highest voltage that batteries are allowed toreach determines in part how much gassing occurs. To limit gassing and electrolyteloss to acceptable levels proper selection of the charge controller voltage regulationset point is critical in PV systems. If too low of a regulation voltage is used, thebattery will be undercharged. If too high of a regulation voltage is used, the batterywill be overcharged. Both under and overcharging will result in premature batteryfailure and loss of load in stand-alone PV systems. In general, sealed “maintenancefree” valve-regulated batteries (using lead-calcium grids) should have lower charge

regulation voltage set points than flooded deep cycling batteries (using lead-antimony grids).

Other Factors Affecting Battery Gassing

The onset of gassing in a lead-acid cell is not only determined by the cell voltage,but by the temperature as well. As temperatures increase, the correspondinggassing voltage decreases for a particular battery. Regardless of the charge rate,the gassing voltage is the same, however gassing begins at a lower battery state ofcharge at higher charge rates. The grid design, whether lead-antimony or lead-

calcium also affects gassing. Battery manufacturers should be consulted todetermine the gassing voltages for specific designs. The figure on the followingpage shows the relationships between cell voltage, state of charge, charge rate andtemperature for a typical lead-acid cell with lead-antimony grids.




10-63

2.0

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.82.9

3.0

0 20 40 60 80 100

Lead-Acid Battery Charging Voltage as a

Function of State of Charge

Battery State of Charge (%)

C

e l l V o l t a g e ( v o l t s )

Lead-Antimony GridsCharge Rate

C/20

C/5

C/2.5

Gassing Voltage at 27oC

Gassing Voltage at 0


By examining the figure, one can see that at 27oC and at a charge rate of C/20, the

gassing voltage of about 2.35 volts per cell is reached at about 90% state of charge.At a charge rate of C/5 at 27

oC, the gassing voltage is reached at about 75% state

of charge. At a battery temperature of 0o

C the gassing voltage increases to about2.5 volt per cell, or 15 volts for a nominal 12-volt battery. The effects of temperatureon the gassing voltage is the reason the charge regulation voltage is sometimestemperature compensated - to fully charge batteries in cold weather and to limitovercharge during warm weather. This type of information is needed to properlyselect battery charge controller voltage regulation set points in order to limit theamount of gassing for a specific battery design and operational conditions.

Some recommended ranges for charge regulation voltages (at 25oC) for different

battery types used in PV systems are presented in Table 10-4 below. These valuesare typical of voltage regulation set points for battery charge controllers used insmall PV systems. These recommendations are meant to be only general in nature,and specific battery manufacturers should be consulted for their suggested values.




10-64

Battery Type

ChargeRegulation

Voltage at 25oC

FloodedLead-

Antimony

FloodedLead-

Calcium

Sealed, ValveRegulatedLead-Acid

FloodedPocket Plate

Nickel-

Cadmium

Per nominal 12volt battery 14.4 - 14.8 14.0 - 14.4 14.0 - 14.4 14.5 - 15.0

Per Cell 2.40 - 2.47 2.33 - 2.40 2.33 - 2.40 1.45 - 1.50

The charge regulation voltage ranges presented in Table 10-4 are much higher thanthe typical charge regulation values often presented in manufacturer’s literature.This is because battery manufacturers often speak of regulation voltage in terms ofthe float voltage , or the voltage limit suggested for when batteries are float charged for extended periods (for example, in non-interruptible power supply (UPS) systems).In these and many other commercial battery applications, batteries can be “trickle” orfloat charged for extended period, requiring a voltage low enough to limit gassing.Typical float voltages are between 13.5 and 13.8 volts for a nominal 12-volt battery,or between 2.25 and 2.30 volts for a single cell.

In a PV system however, the battery must be recharged within a limited time (usuallyduring sunlight hours), requiring that the regulation voltage be much higher than themanufacturer’s float voltage to ensure that the battery is fully recharged. If chargeregulation voltages in a typical PV system were set at the manufacturer’srecommended float voltage, the batteries would never be fully charged.




10-65

Exercises




10-66

Balancing Life Cycle and Initial Costs

First costs often determine which battery is selected by a designer; however,recurring costs, including battery maintenance and replacement are burdened by thesystem owner/operator. The sum of the initial cost and amortized recurring costs

are what determine a battery's life cycle cost. In many cases, a battery with highinitial costs may have the lowest life cycle costs, all factors being considered. This issometimes the case in the use of higher cost nickel-cadmium cells used for criticalapplications.

When deciding between the lower cost of shallow cycling batteries, or the higherinitial cost of deeper cycling batteries, photovoltaic system designers are faced withthe common decision of “pay now or pay later”. There is not one perfect batterychoice for all users and applications, and “paying now” may be best in one situationwhile “paying later” may be more appropriate in another.

Battery Trade-Off “Pay Now or Pay Later”

• Pay less initially for starting type batteries, but get short life (1-2 years)

• Pay more initially for deep discharge batteries, and get longer life (4-8years)

• Starting types may be more available, but may not be the best “value” inthe long run

For low-income rural villagers, the high cost of a “better” longer life deep cyclingbattery may be prohibitive. But they may be able to afford a lower cost, more readilyavailable and replaceable shallow cycling, starting type battery. In a year or two,they can afford to buy a replacement. For them, the more affordable choice may beto not pay now for the “best” (longest life) but to pay later (for replacements).

On the other hand, a remote telecom site in a difficult or dangerous location wouldbe best designed with the longest life battery possible. The cost of poorperformance would be much greater than the small savings of a cheaper battery.The budget would probably be present for purchasing the “better” longer life batteryinitially. The cost savings in reduced maintenance and replacement costs over timewould be of primary consideration in this case.




10-67

Factors Affecting Battery Life

Lower than expected battery life is a common problem in small stand-alone PVapplications. The battery bank is typically the second most expensive component inmost photovoltaic power systems, next to the PV array. Over a twenty to thirty year

life of a PV system, the replacement and maintenance costs for batteries may be thehighest life cycle component in a PV system. For these reasons, PV systemdesigners and users should have a good understanding of the issues affectingbattery life.

Battery lifetime is dependent upon a number of cell design and operational factors,including the components and materials of battery construction, temperature,frequency and depth of discharges, average state of charge and charging methods.As long as a battery is not overcharged, overdischarged or operated at excessivetemperatures, the lifetime of a battery is proportional to its average state of charge.A typical flooded lead-acid battery that is maintained above 90% state of charge will

provide two to three times more full charge/discharge cycles than a battery allowedto reach 50% state of charge before recharging. This suggests limiting themaximum allowable and average daily DOD to prolong battery life. The actual pointthat determines the end of life for a battery is arbitrary, but most often it is assumedto be when the battery will no longer deliver 80% of it’s full rated capacity. It isimportant that the PV system designer understand the consequences of capacityloss as a battery ages, as this may affect the reliability of system operation.

Lifetime can be expressed in terms of cycles or years , depending on the particulartype of battery and its intended application. Exact quantification of battery life isdifficult due to the number of variables involved, and generally requires battery test

results under similar operating conditions. Often, battery manufacturers do not ratebattery performance under the conditions of charge and discharge experienced inPV systems. This makes the accurate estimation of battery life in PV systems verydifficult, if not impossible. However, there are well known operating practices andprocedures that tend to maximize battery life. A discussion of these considerationsfollows.

What Maximizes Battery Life?

There are several general guidelines that can lead to maximizing the useful life of

batteries in PV systems. Each of these is reviewed next.

• Cool operating temperatures

• Shallow depth of discharge

• Prevention of overcharging and discharging

• Proper and frequent maintenance and water additions

• Full recharges after discharging as soon as possible




10-68

Temperature Affects Battery Life

Temperature has important effects on battery life that are important to PV systemdesigners. The main enemy to battery life is operation at high temperatures. Ingeneral, as the temperature increases by 10 oC, the rate of an electrochemical

reaction doubles, resulting in the recommendations from battery manufacturers thatbattery life decreases by a factor of two for every 10 oC increase in average

operating temperature. Higher operating temperatures accelerate corrosion of thepositive plate grids, result in greater gassing and electrolyte loss, and shortenedbattery life. Lower operating temperatures generally increase battery life, howeverthe capacity is reduced significantly, particularly for lead-acid batteries. Wheresevere temperature variations from room temperatures exist, battery are sometimelocated in an insulated or other temperature regulating enclosure to minimize batterytemperature swings.

10

100

1000

5 10 15 20 25 30 35 40 45

Lead-Antimony Grids

Lead-Calcium Grids

Nickel-Cadmium

Effects of Temperature on Battery Life

Battery Operating Temperature (oC)

B a t t e

r y L i f e

( % l

i f e

a t 2 5 o C )




10-69

As sown in the figure, the grid alloying elements play an important part in battery life.The lifetime of batteries with lead-calcium grids is more affected by temperature thanlead-antimony grids or nickel-cadmium batteries.

The importance of temperature can be seen by looking at an example of amanufacturer’s warranty statement. A portion of the warranty for GNB Absolyte IIP

batteries is presented on the next page. The portion of interest to us now ishighlighted below.

The wording of the warranty statement has been condensed to focus attention onthe temperature factor.

“GNB also warrants ...that GNB ABSOLYTE IIP batteries used in photovoltaicservice will furnish... 80% of the specified capacity...for “m” years...provided that thefollowing conditions are satisfied.”

“.....The average daily ambient temperature in the area of use in any year isestablished:

not to exceed 25 degrees C “m” equal to 10 yearsnot to exceed 30 degrees C “m” equal to 7 yearsnot to exceed 35 degrees C “m” equal to 5 years”

Notice that the warranty period decreases substantially as the average operating

temperature of the battery increases. Notice that with a temperature change from25 to 35oC the warranty period decreases by 50%, as per the general rule stated

previously.




10-70




10-71

Effects of Discharge Depth on Battery Life

The battery depth of discharge also affects battery life. If the discharge depth is keptshallow (less than 25%), then the adhesion of the active materials to the gridsremains strong, prolonging the life and performance of the battery. In PV systems,

the average daily depth of discharge is generally low because the battery capacity issized to provide a large number of days of reserve or autonomy. Typically, thebattery bank in PV system is designed to provide 5 to 10 days of reserve operationduring the lowest insolation to load ratio period, so the average daily depth ofdischarge is generally only 10 to 20% of the total installed capacity. The more daysof reserve or autonomy that are designed in to the battery, the shallower the averagedaily depth of discharge is, and the longer the battery life. Of course, the greatercapacity installed to give more reserve days results in higher initial costs. Onceagain there is a tradeoff of initial cost against long life.

10

100

1000

10000

0 20 40 60 80 100

Motive Power Battery (deep cycle)

Automotive (SLI) Battery (shallow cycle)

Effects of Depth of Discharge onLead-Acid Battery Cycle Life

Battery Depth of Discharge (%)

B a t t e r y L i f e

( C y c l e s )




10-72

Overcharging Shortens Battery Life

Overcharging leads to excessive gassing and loss of electrolyte. The gas bubblesgenerated can actually scrub the outside of the plates as they rise, and canaccelerate the erosion of active material from the plates, reducing battery life.

Excessive heating during overcharging accelerates the natural corrosion process.Corrosion is the electrochemical activity resulting from the immersion of twodissimilar metals in an electrolyte, or the direct contact of two dissimilar metals,causing one material to undergo oxidation, or lose electrons, and the causing theother material to undergo reduction, or gain electrons. Corrosion of the gridssupporting the active material in a battery is an ongoing process, and may ultimatelydictate the battery's useful lifetime.

Overcharging is prevented in PV systems by using a charge controller or regulatorbetween the PV array and batteries that limits battery gassing, the maximum voltage

and state of charge.

Maintenance Affects Battery Life

Any component of a system will last longer if it is frequently maintained, and thiscertainly applies to batteries. Keeping the electrolyte level “topped up” means theacid concentration is kept constant and at the correct value, and that the plates arealways fully immersed in liquid. If the electrolyte level is allowed to drop below thetop of the plates, corrosion and loss of capacity and life are accelerated greatly.Keep the electrical terminals clean and free of corrosion, and make sure the batterycables are firmly connected to the terminals. Loose or corroded terminalsconnections can lead to voltage drops and “hot spot” heating, and batteries canexperience unequal charging. With unequal voltage potentials across batteries,ones with good connections will be overcharged while ones with bad connections willbe undercharged. Both of these conditions will lead to loss of battery life over time.

Fully Recharging Affects Battery Life

If lead-acid batteries are left at partial state of charge for weeks, sulfation will set in

and capacity can be permanently lost. Sulfate crystals can grow and warp or shortout the plates. To the degree that systems can be designed to fully recharge thebatteries, the battery life will be extended.




10-73

Battery Efficiency

Battery efficiency is dependent on the type of battery, charging methods, rates ofcharge and discharge, depth of discharge and temperature. In general, theefficiency of a battery is much greater at lower states of charge than when the

battery is nearly fully charged. The total or round-trip battery energy efficiency iscomposed of two types of efficiencies. There is a voltage or voltaic efficiency, and acharge or coulombic efficiency. They measure different characteristics of batteries,and are often confused.

Battery Voltage (Voltaic) Efficiency

The voltaic efficiency of a battery is determined by the charge and discharge ratesand the battery temperature. The voltaic efficiency is expressed as the ratio of the

battery voltage under discharge to the voltage under charge. High rates and lowtemperatures act to decrease battery voltaic efficiency.

Battery Voltaic Efficiency =Voltage During Discharge

Voltage During Charge

Recall how battery voltage increases during charging and decreases duringdischarging as compared to the open-circuit voltage. Battery voltage in a PV systemmay vary considerably, depending on state of charge and rate of charge ordischarge. An overall average voltaic efficiency can be calculated by assuming thaton the average, a battery is charged at about 14 volts and discharged at about 12volts at the given charge and discharge rates in the system. This approximationyields an overall voltaic efficiency of about 85%.

Approximate Voltaic Efficiency =

12 volts

14 volts 0.85 85%= =




10-74

The charge and discharge curves presented below illustrate battery voltaicefficiency. During charging the voltage is elevated above the open circuit voltage,and during discharging the voltage is depressed below it. Since the rate of chargeor discharge determines the battery voltage at a certain state of charge, the voltaicefficiency is higher at lower rates of charge and discharge.

10.0

11.0

12.0

13.0

14.0

15.0

16.0

0 20 40 60 80 100

Battery Voltaic Efficiency


B a t t e r y V o l t a g e

Charge Rates

C/20C/5

Open-Circuit Voltage

Discharge RatesC/5

C/20




10-75

Battery Charge (Coulombic) Efficiency

The coulombic or charge efficiency of a battery is defined as the ratio of the ampere-hours withdrawn from a battery during discharge to the ampere-hours provided torecharge the battery. Due to electron losses from gassing and other internal

mechanisms, a battery can not deliver as many amp-hours as it takes to charge it atthe same rate. At low states of charge, a battery accepts current readily, there islittle gassing and the coulombic efficiency is high. As a battery nears full state ofcharge, gassing and internal heating tend to reduce the coulombic efficiency.

Battery Coulombic Efficiency =Discharge Amp - Hours Output

Charge Amp- Hours Input

Typical industrial applications using motive power or traction batteries must fullyrecharge deeply discharged batteries in a limited time, usually less than 8 to 10hours. To accomplish this, high charge rates (C/5 to C/10) are required resulting inlow coulombic efficiencies. In PV systems, the coulombic efficiency of batteries isgenerally very good due to the relatively low charge rates used. Typical charge ratesin PV systems are often C/20 or lower, because the amount of battery storagerequired for autonomy is relatively large with respect to the PV array chargingcurrents. While an accurate estimation of coulombic efficiency is difficult todetermine, 90% is a typical value for most batteries used in small stand-alone PVsystems.

Using Coulombic Efficiency in Sizing Calculations

The coulombic efficiency is used to calculate the PV array size required in aphotovoltaic system. In the load estimation chapter, we showed how to calculate anaverage daily load demand in units of ampere-hours. The PV array must not only besized to meet the load amp-hour demand, but must also overcome the batterycoulombic inefficiency. To determine the total amp-hours that the PV array mustproduce, we divide the total load demand by the estimated battery coulombicefficiency.

Amp- Hours PV Array Must Produce =Daily Load Demand in Amp- Hours

Battery Coulombic Efficiency




10-76

Example: The load demand for a remote home is estimated to be 200 Ah/day at12 volts. Using an average battery coulombic efficiency of 90%, thetotal Ah that the array must produce is given by:

Ah From Array = 200 Ah Daily Load Demand 0.90

= 222 Ah

Battery Round-trip or EnergyEfficiency

The energy or round-trip efficiency of a battery is defined as the product of thecoulombic and voltaic efficiencies. This efficiency defines the ratio of the energywithdrawn from a battery during discharge to the amount of energy supplied to bringa battery back to full state of charge.

Battery Energy = Voltaic Efficiency X Coulombic EfficiencyEfficiency

= .85 X .95

= .80 or 80%

The round-trip battery energy efficiency is often used when discussing batteries.However it is the coulombic efficiency that is most commonly used in photovoltaicsystem sizing methods using the amp-hour approach. In cases where energy orwatt-hour approach is used in system sizing, the energy or round-trip batteryefficiency is sometimes used to account for the voltage as well as the chargeefficiency of the battery, especially for quantifying losses due to gassing. For thesizing calculations presented in the chapter on System Sizing, the only battery

efficiency that needs to be considered will be the charge or coulombic efficiency,which is usually about 90% for most batteries.




10-77

Exercise




10-78

Battery Selection Criteria

The selection of a battery for use in PV systems involves many decisions and tradeoffs, and depends on many factors. While no specific battery is appropriate for allPV applications, common sense and a careful review of the battery literature with

respect to the particular application needs will help the designer narrow the choice.Some decisions on battery selection may be easy to arrive at, such as physicalproperties, while other decisions will be much more difficult and may involve makingtradeoffs between desirable and undesirable battery features. With the properapplication of this knowledge, designers should be able to differentiate amongbattery types and gain some application experience with batteries they are familiarwith. The list below summarizes some of the key considerations in battery selection.

Battery Selection Criteria

• Nominal system voltage• Charge regulation requirements

• Required capacity or autonomy

• Ampere-hour capacity at discharge rate

• Daily and maximum depth of discharge

• Self-discharge rate

• Gassing characteristics

• Efficiency

• Temperature effects

• Size, weight and structural needs

• Susceptibility to freezing• Susceptibility to sulfation, stratification

• Electrolyte type, concentration

• Auxiliary equipment

• Maintenance requirements

• Terminal configurations

• Battery life (cycles/years)

• Availability and shipping requirements

• Cost and warranty




10-79

Desirable Battery Construction Features

Selection of a durable battery design is as equally important as the proper sizing andtreatment in improving the longevity of batteries in PV systems. A few of theimportant battery design characteristics promoting deep-cycle, long-life performanceare listed as follows:

• Low-antimony, thick plates for reducing water loss and providing adequatemechanical strength for long cycle life.

• Separator envelopes around the positive plates, and appropriate plate edgeprotection to minimize internal short-circuit potential.

• Large electrolyte volume below plates to allow for shed materials to accumulatewithout causing short circuits.

• Sufficient electrolyte volume above plates to minimize frequency of water

additions.

• Transparent battery containers to allow for visual inspections.

Size and Weight

Due to their large size and weight, the physical characteristics of batteries oftenplace restrictions on battery selection. A typical 12 volt, 100 ampere-hour floodedlead-acid battery weights anywhere between 20 and 40 kilograms, depending on theweight of active material, grid and interconnect design and amount of electrolyte.

Special lifting devices may be required for large capacity batteries.

For remote PV applications where the transportation of heavy equipment iscumbersome or otherwise difficult, the designer may choose to use several smallerbatteries or cells rather than a few larger batteries. The physical properties ofbatteries also require consideration of the strength and dimensions of the enclosureor area in which they are to be installed.

Availability

Due to the weight and hazardous nature of batteries, shipping costs are high, andresults in most batteries produced at regional plants being used to supply to localmarkets. In developing regions, the optimal battery for a given application may notbe locally available. In these cases, the best battery available should be used, withspecial attention paid to properly sizing and treating the battery for maximum life.




10-80

Battery Auxiliary Equipment

Battery auxiliary equipment includes any systems or other hardware necessary tosafely and effectively operate a battery system. Some of the more important batteryauxiliary systems and equipment are discussed below.

Ventilation

Batteries often produce toxic and explosive mixtures of gasses, namely hydrogen,and adequate ventilation of the battery enclosure is required. In most cases,passive ventilation techniques such as vents or ducts may be sufficient. In somecases, fans may be required to provide mechanical ventilation. Required air changerates are based on maintaining minimum levels of hazardous gasses in theenclosure. Under no circumstances should batteries be kept in an unventilated areaor located in an area frequented by personnel.

Catalytic Recombination Caps

A substitute for standard vented caps on lead-antimony batteries, catalyticrecombination caps (CRC’s) primary function is to reduce the electrolyte loss fromthe battery. CRC’s contain particles of an element such as platinum or palladium,which surfaces adsorb the hydrogen generated from the battery during finishing andovercharge. The hydrogen is then recombined with oxygen in the CRC to formwater, which drains back into the battery. During this recombination process, heat isreleased from the CRC’s as the combination of hydrogen and oxygen to form water

is an exothermic process. This means that temperature increases in CRC’s can beused to detect the onset of gassing in the battery. If CRC’s are found to be atsignificantly different temperatures during recharge (meaning some cells are gassingand others are not), an equalization charge may be required. The use of CRC’s onopen-vent, flooded lead-antimony batteries has proven to reduce electrolyte loss byas much as 50% in subtropical climates.

Battery Monitoring Systems

Monitoring and instrumentation for battery systems can range from simple analog

meters to more sophistication data acquisition systems. Lower level monitoring ofbattery systems might include voltage and current meters or battery state of chargeindicators, while higher level monitoring may include automated recording of voltage,current, temperature, specific gravity and water levels. For small stand-alone PVsystems, monitoring of the battery condition is generally done only occasionallyduring routine maintenance checks, or by simple meters or indicators on the batterycharge controller.




10-82

The performance of the Cool CellTM

depends on the ambient temperature and theextent and type of cloud cover. In areas with exceedingly high ambienttemperatures, this enclosure can provide a significant reduction in batterytemperatures compared to a standard box or insulated enclosure. Cloud cover playsan important effect on the performance of the Cool Cell

TMas this affects the

radiation heat transfer from the lid. Clear skies generally have an effective

temperature (for purposes of radiation calculations) of below -40o

C, however cloudyskies have a much higher effective temperature. Since the radiation heat transfer isproportional to the effective temperature raised to the fourth power, overcast skiescan reduce the cooling effectiveness of this type of battery enclosure.




10-83

The relative performance of the Cool Cell is illustrated below. Four conditions arecompared: (a) the widely varying case of an uninsulated steel box; (b) the ambientair; (c) a concrete vault with an insulated lid; and (d) the moderated temperatureswings of the passively cooled enclosure.




10-84

Battery Safety ConsiderationsDue to the hazardous materials and chemicals involved, and the amount of electricalenergy which they store, batteries are potentially dangerous and must be handledand used with caution. Typical batteries used in stand-alone PV systems can deliver

up to several thousand amps under short-circuit conditions, requiring specialprecautions. Depending on the size and location of a battery installation certainsafety precautions are required.

IMPORTANT:

Exercise extreme caution and follow recommended practices whenworking with batteries!

Handling Electrolyte

The caustic sulfuric acid solution contained in lead-acid batteries can destroyclothing and burn the skin. For these reasons protective clothing such as apronsand face shields should be worn by personnel working with batteries. To neutralizesulfuric acid spills or splashes on clothing, the spill should be rinsed immediately

with a solution of baking soda or household ammonia and water. For nickel-cadmium batteries, the potassium hydroxide electrolyte can be neutralized with avinegar and water solution. If electrolyte is accidentally splashed in the eyes, theeyes should be forced open and flooded with cool clean water for fifteen minutes. Ifacid electrolyte is taken internally, drink large quantities of water or milk, followed bymilk of magnesia, beaten eggs or vegetable oil. Call a physician immediately.

If it is required that the electrolyte solution be prepared from concentrated acid andwater, the acid should be poured slowly into the water while mixing. The watershould never be poured into the acid. Appropriate non-metallic funnels andcontainers should be used when mixing and transferring electrolyte solutions.




10-86

Selected ReferencesInstitute of Electrical and Electronics Engineers, "IEEE Recommended Practice for Installation and

Operation of Lead-Acid Batteries for Photovoltaic (PV) Systems", ANSI/IEEE Std. 937-1987, NewYork, NY, 1987.

Institute of Electrical and Electronics Engineers, "IEEE Recommended Practice for Installation andOperation of Nickel-Cadmium Batteries for Photovoltaic (PV) Systems", ANSI/IEEE Std. 1145-1990, New York, NY, 1990.

Stand-Alone Photovoltaic Systems - A Handbook of Recommended Design Practices, Sandia NationalLaboratories, SAND87-7023, revised November 1991.

Naval Facilities Engineering Command, Maintenance and Operation of Photovoltaic Power Systems,NAVFAC MO-405.1, December 1989.

Exide Management and Technology Company, Handbook of Secondary Storage Batteries and ChargeRegulators in Photovoltaic Systems - Final Report, for Sandia National Laboratories,SAND81-7135, August 1981.

Bechtel National, Inc., Handbook for Battery Energy Storage in Photovoltaic Power Systems, Final Report,SAND80-7022, February 1980.

S. Harrington and J. Dunlop, "Battery Charge Controller Characteristics in Photovoltaic Systems",Proceedings of the 7th Annual Battery Conference on Advances and Applications, Long Beach,California, January 21, 1992.

H.A. Kiehne, “Battery Technology Handbook”, Marcel Dekker, Inc., 1989.

G.W. Vinal, “Storage Batteries”, John Wiley & Sons, Inc., Fourth Edition, 1954.

D. Linden, “Handbook of Batteries and Fuel Cells”, McGraw Hill, Inc., 1984.




10-87

(End of Chapter)




10-88

CHAPTER TEN

BATTERY TECHNOLOGY 10-1

Purpose of Batteries in Photovoltaic 10-2

Systems 10-2Energy Storage Capacity and Autonomy 10-3Voltage and Current Stabilization 10-4Supply Surge Currents 10-5

Battery Design and Construction 10-6Cell 10-6

Active Material 10-6Electrolyte 10-7Grid 10-8Plate 10-8Separator 10-8Element 10-9Terminal Posts 10-9Cell Vents 10-9Case 10-9

Battery Types and Classifications 10-10Primary Batteries 10-10Secondary Batteries 10-10

Lead-Acid Battery Classifications 5.3.3.1 10-12

Lead-Acid Batteries 10-12SLI Batteries 10-12Motive Power or Traction Batteries 10-12Stationary Batteries 10-12

Types of Lead-Acid Batteries 10-13Lead-Antimony Batteries 10-13Lead-Calcium Batteries 10-14Lead-Antimony/Lead-Calcium Hybrid 10-14Captive Electrolyte Lead-Acid Batteries 10-16




10-89

Lead-Acid Battery Chemistry 10-18Lead-Acid Cell Reaction 10-18Formation 10-19Stratification 10-19

Specific Gravity 10-20Sulfation 10-22

Nickel-Cadmium Batteries 10-23Sintered Plate Ni-Cads 10-23Pocket Plate Ni-Cads 10-23Nickel-Cadmium Battery Chemistry 10-24

Battery Strengths and Weaknesses 10-27

Battery Cycling 10-28Battery Discharging 10-29Charge/Discharge Rates Expressed as Time 10-30Depth of Discharge (DOD) 10-31Autonomy Determines Capacity and Depth of Discharge 10-32State of Charge (SOC) 10-35Temperature Limits Discharge Depth 10-36Shallow and Deep Cycle Batteries 10-39

Battery Capacity 10-41

Ampere-Hour Definition 10-41Factors Affecting Battery Capacity 10-41Effects of Temperature on Capacity 10-42Discharge Rate Affects Capacity 10-44Cut-Off Voltage Affects Capacity 10-51Effects of Self Discharge Rate on Capacity 10-52

Voltage Changes During Charging and Discharging 10-54Battery Voltage and State of Charge are Related 10-55Battery Voltage Depends on Rate of Charge or Discharge 10-57

Battery Charging 10-59Battery Gassing and Overcharge 10-59Charge Regulation Voltage Affects Gassing 10-62

Balancing Life Cycle and Initial Costs 10-66




10-90

Factors Affecting Battery Life 10-67What Maximizes Battery Life? 10-67Temperature Affects Battery Life 10-68Effects of Discharge Depth on Battery Life 10-71

Overcharging Shortens Battery Life 10-72Maintenance Affects Battery Life 10-72Fully Recharging Affects Battery Life 10-72

Battery Efficiency 10-73Battery Voltage (Voltaic) Efficiency 10-73Battery Charge (Coulombic) Efficiency 10-75Battery Round-Trip or Energy Efficiency 10-76

Battery Selection Criteria 10-78

Battery Auxiliary Equipment 10-80Ventilation 10-80Catalytic Recombination Caps 10-80Battery Monitoring Systems 10-80Enclosures 10-81

Battery Safety Considerations 10-84Handling Electrolyte 10-84Personnel Protection 10-85

Dangers of Explosion 10-85Battery Disposal and Recycling 10-85

Selected References 10-86



Siemens Solar Basic Photovoltaic Technology 10-1 Battery Technology

Chapter 10 – Answers Battery Technology

a. Positive Plate

PbO2 + 4H+ + 2e

- ⇔ Pb

2+ + 2H2O

Element / Ions Qty on Left Qty on RightLead (Pb) 1 1Oxygen (O) 2 2Hydrogen (H) 4 4Ion Charges +4 -2 = +2 +2

Pb2+

+ SO42-

⇔ PbSO4

Element / Ions Qty on Left Qty on RightLead (Pb) 1 1Oxygen (O) 4 4Sulfur (S) 1 1Ion Charges +2 -2 = 0 0

b. Negative Plate

Pb ⇔ Pb2+

+ 2e-

Element / Ions Qty on Left Qty on RightLead (Pb) 1 1Ion Charges 0 +2 -2 = 0

Pb2+

+ SO42-

⇔ PbSO4

Element / Ions Qty on Left Qty on RightLead (Pb) 1 1Oxygen (O) 4 4Sulfur (S) 1 1Ion Charges +2 -2 = 0 0

c. Overall lead - acid cell reaction

PbO2 + Pb + 2H2SO4 ⇔ 2PbSO4 + 2H2O

Element / Ions Qty on Left Qty on RightLead (Pb) 2 2Oxygen (O) 10 10Sulfur (S) 2 2Hydrogen (H) 4 4




For a 24-volt (nominal) battery, we need

24 volts / 2.1 volts per cell = 11.4 = 12 lead acid cells

24 volts / 1.2 volts per cell = 20 nickel cadmium cells

Note: in most cases a lead-acid cell is assumed to have a nominal voltage of 2.0 volts.

We assume the nominal voltage of a lead-acid cell is 2 volts. Then 24 cells will have atotal nominal voltage of 24 X 2 Volts = 48 Volts.

b. Lower than

b. Left at partial state of charge for a long period


b. Less

a. A smaller

Referring to the chart of Temperature Limits for Maximum Depth of Discharge, we see

that at a temperature of -25 °C, the maximum recommended depth of discharge isapproximately 45%.




Referring to the chart of Temperature Limits for Maximum Depth of Discharge, we see

that at a temperature of -40 °C, the maximum recommended depth of discharge isapproximately 25%.

At first, we might be tempted to answer that we need a battery that is four times aslarge. However, this would only be true if we assumed that we could use 100% of theoriginal battery. Since the maximum recommended depth of discharge is 80% atnormal temperature, the battery for the mountain application needs to be larger by afactor of

80% = 3.2 times larger25%

An example helps make this clearer. Assume we need 800 Ahr of storage. A 1000 Ahrbattery designed for use at normal temperatures can provide 800 Ahr storage to 80%.

Now, if we can only discharge the battery to 25% DOD, the battery must provide thesame 800 Ahr capacity. Using the ratio of 3.2 that we calculated above, we choose a3,200 Ahr battery. Notice that 25% discharge = 25% X 3200 = 800 Ahrs as is required.

Since low temperatures can have a significant impact on the required battery size, wewould be very interested in any methods that might keep the battery from getting cold.Some possible techniques to do this might include:

Keeping the batteries in a heated area (e.g. an equipment shelter)Locating the batteries in a buried vaultProviding insulation to minimize the temperature changes


a. 0oC and C/120? 92%

b. 0oC and C/50? 85%

c. -10oC and C/120? 90%

d. 25oC and C/500? 100%

e. 10oC and C/120? 95%

These values are taken from the chart showing Temperature and Discharge effects onLead Acid Capacity.




Referring to the chart, we see that at -15 °C and C/500, the battery will have about 90%

of its standard rated capacity (at 25 °C). The increased capacity required is

1500 Ahr = 1667 Ahr at 25 °C0.90

At 5 °C and C/50, the battery will have about 87% of its standard rated capacity (at 25

°C). The increased capacity required is

200 Ahr = 230 Ahr at 25 °C0.87


At 12 volts, the current required by the loads is:

Lights 20 watts / 12 volts = 1.7 ampsRefrigerator 100 watts / 12 volts = 8.3 amps

System Description: 12 Volt Rural Electrification System

Hours DailyDC Loads Qty. Amps /day Demand (Ah)Lights : 2 X 1.7 X 6 = 20.4

Refrigerator : 1 X 8.3 X 12 = 100.0

DC Loads (Ah) 120.4

Ah From Array = 120.4 Ah Daily Load Demand

0.90 coulombic efficiency

= 133.8 Ah



Siemens Solar Basic PV Technology Course 11-1 Components –Inverter TechnologyCopyright © 1998 Siemens Solar Industries

Chapter Eleven

Inverter TechnologyThe photovoltaic array and battery produce DC current and voltage. If the loads requireAC power an inverter can be used to convert from DC to AC. Commonly availableinverters can output in 1- or 3-phase, 50 or 60 hertz, and 117 or 220 volts and canrange in continuous output power from a few hundred watts to thousands of kilowatts.Large utility scale inverters are made to output at 480 volts AC or higher and havecapacities exceeding 1000 kilowatts.

In this section we will examine some of the technical features and characteristics ofinverter technology commonly used in photovoltaic systems. The three most importantcharacteristics of inverters that we will discuss will be:

• Output power and surge power

• Output efficiency

• Output waveform




Inverter Power Rating

Power Is A Function Of Time

An inverter's capability to output power depends on its ability to dissipate the heatgenerated within the inverter. An inverter's rated output power is generally the powerlevel it can maintain for a long period. An inverter can put out greater power than itsnominal rating before it overheats and shuts off, but only for a short period of time.Some inverters can output a great deal more power than their nominal rating for a fewminutes, allowing very large loads to operate, for example large power tools and orpump motors.

However, there is no standard period of time accepted by all manufacturers for inverteroutput rating. There is a trend for manufacturers to rate their inverters on a“continuous” basis indicating the power the inverter could output for many hours. Thisis a safe, conservative rating method, although most inverters would not be operating attheir full rated output level for 24 hours a day. In most real situations the output isrequired for a few hours or minutes.

Some manufacturers rate their inverter output for 30 minutes and some even rate thembased on 15-minute output. The problem with this situation is that you as a systemdesigner have the burden of discovering the exact method of output rating so that youcan make a fair comparison.

Instead of reading a single power number it is best to find a chart or graph thatdescribes the output against different times, as shown below.

An example of an inverter with a fairly generous power vs. time curve is shown in thefigure. The literature mentions that the inverter is “rated at 2000 watts output”, butnotice that this is for 30 minutes. For periods longer than 30 minutes, the power levelwill be less. Continuous output would be more like 1700 watts.




Surge Power

Notice that the output power of the inverter in the figure above can exceed the nominal

2000 watts for a few minutes. Often this is referred to as the “surge capability” of theinverter, although strictly surge power may be defined as the power that the inverter canoutput for less than a second to start large inductive (motor) loads. Recall from ourdiscussions in the chapter on Load Estimation that many inductive loads (motors,compressors, refrigerators, washing machines, and large water pumps) can draw 5-6times their nominal running power when they start!

For example, a remote home might have non-inductive loads of 1000 watts and a3/4-hp water pump motor. Normally the water pump might draw about 700 watts, sothe total continuous power needed from the inverter would be about 1700 watts. Butwhen the pump turns on it draws about 3500 watts. If it turned on when the other 1000watts of load were operating, the inverter would have to output about 4500 watts (1000non-inductive + 3500 inductive surge) for a few moments to start the motor. Theinverter example used above in the figure could output about 6000 watts for a fewseconds, so this would be adequate. An inverter with a lower surge capacity might notbe able to start the pump.




Power vs Time

0 20 40 60

Time (minutes)

Output Power (watts)

6000

5000

4000

3000

2000

1000

Power of “2000” watt inverter

The surge power of an inverter is again not a single fixed number, but varies with theamount of time. A good inverter can typically output 200-300% of its nominal power fora few seconds. Some low cost inverters can only output 110% of their output for ashort time. These would have limited surge capacity and would not be useful forsystems where inductive loads would operate.




Temperature Affects Power Rating

Another source of confusion in inverter rating is the temperature of the rating. The

output can be maintained so long as the inverter design adequately dissipates theinternal heat generated. But this ability will depend on the ambient temperature. If theair temperature is hotter, the inverter design will not be able to passively dissipate asmuch heat, and the power rating will be lower.

Most US manufacturers rate their output at 25o

C. Some others (often internationalmanufacturers) will rate at 20

oC. And some manufacturers take the extreme case of

rating their inverter output at 40o

C. This must be looked at to fairly compare productcapability.

The use of a high temperature such as 40oC is conservative and safe. An inverter

rated at 40oC will output more power for a longer period of time at lower temperatures.

But it may not be difficult for the electronics of an inverter to see 35-40o C conditionsinside the inverter case, so this rating is not unreasonable.




Inverter Efficiency

An inverter consumes some power itself. This adds to the total load that the

photovoltaic array must operate and increases array size and initial cost. The invertershould be as efficient as possible, certainly above 90% over most of its normaloperating range. Many moderately priced inverters can achieve 94% efficiency. Theefficiency may depend on the nature of the load. Pure resistive loads may operate theinverter at a higher efficiency than inductive loads that absorb the power differently.

Efficiency Depends On Load

Inverter efficiency is not a single fixed value, but varies depending on the amount of

power being generated. A curve of efficiency vs. output power is the best way toexamine how the inverter will perform in a variety of situations. It is difficult orimpossible to predetermine exactly what will be the power demand on the inverter atany moment, so guesses must be made. You must estimate at what power levels yoursystem will operate and determine average efficiency from a curve such as the onepresented below.

Older technology using silicon controlled rectifiers (SCR) were not as efficient as morecurrent designs using new power transistor technology. Efficiencies can now reachabove 90% over much of the range of operation. Notice how the efficiency curve risesquickly even for small loads, and then gradually drops as power continues to increase.

This shows the effect of the internal power losses increasing with higher currents.

For example, reading from the “new transistor” efficiency curve, you can see that theefficiency at full rated load is about 85% while the efficiency at only 10% of full load isabout 93%. This is good news compared to older technology. Because it means thatyou can choose a large inverter to handle the heavy loads, you might have in a complexresidential system and yet not sacrifice on efficiency when operating only a few smallappliances or lights in the evening for example. With older technology you would pay asteep efficiency penalty when running loads much smaller than the full rated capacity ofthe inverter.




Inverter Efficiency

0 20 40 60 80 100

100

90

80

60

50

40

30

20

10

0

Efficiency

Percent of Full Load (%)

70

old SCR type

new Transistor type




Inductive vs. Resistive Loads

Resistive loads are the simplest for inverters to operate. Another measurement would

use inductive loads, such as motors, and the efficiency curve would probably be slightlydifferent. Inductive loads tend to push some power back into the inverter, and somemanufacturer’s designs may handle the inductive loads more efficiency that others.Measuring the AC output to inductive loads becomes more complex because thecurrent and voltage will not be in phase as they are with pure resistive loads. You mustuse a power meter on the output, or measure the current and voltage and also thephase difference between them. This is the “power factor” and it can reduce the realpower delivered to an inductive load 10-30% or more. This is why most manufacturersmeasure their inverter output power and efficiency into pure resistive loads. Thismethod however does not give the total picture of what efficiency you would actually getin a complex residential or industrial system with many large inductive loads.

Test Set-up For Measuring Efficiency

It is fairly easy to create your own inverter efficiency curve by plotting power into avariable load against the power consumed by the inverter. A circuit of incandescentlights or heaters can serve as a variable pure resistive load. Measure the current andvoltage output to the loads, and compare to the current and voltage drawn from thebattery bank by the inverter. Operate loads from about 10% of the full rated output towell beyond 100% of the rating, to create an efficiency curve similar to that shown.

battery

(+)

(-)

Inverter

12 v DC 120 v AC(+)

(-)

variableload(ex: lights) .

shunt #1

shunt #2




Exercise




Inverter Output Waveform

The waveform of the inverter output can be an important factor in matching inverter to

load. The waveform describes the way the current and voltage vary over time. Thereare three general classes of waveform: squarewave; modified squarewave or modifiedsinewave; and sinewave.

Squarewave

Generally the most inexpensive inverters are "squarewave" types. The input DC poweris chopped and boosted in voltage, with little filtering or modulation of the output. Theresulting output contains many unwanted harmonics, or waves of various frequencies

that fight against each other. The measure of this unwanted waveform content, thetotal harmonic distortion, can be 40% or greater. Typically squarewave inverters cannotsurge significantly, perhaps only 10-20 % above the maximum continuous power. Theefficiency can be as low as 50-60%. And they have little output voltage regulation.However these inverters may be useful for small inductive loads or resistive loads.

Squarewave Inverters




Modified Squarewave or“Modified Sinewave”

Another type of inverter output is available that produces a "modified squarewave"output. Sometimes these are referred to as “modified sinewave” in manufacturer’sliterature, although the form more closely resembles a squarewave with a delay throughzero. The peak voltage of the waveform is only about 150 volts, while utility sinewavepower peaks at about 175 volts. The total harmonic distortion is significantly less than asquarewave, perhaps less than 20%. Also they can surge 300-400% above thecontinuous power, and have good voltage regulation. Efficiencies of greater than 90%are common. These types of inverters are quite popular in remote homes, and canoperate a wide variety of common appliances and electronic loads including homecomputers and stereos. However there are reported problems with operating somespecific loads. Some models of laser printers malfunction or are damaged; microwaveovens may not heat as much and may have to be operated longer; and small electronicclocks may run twice as fast. You should check with the manufacturer to see if anyincompatibility with your planned loads have been reported.

Modified Squarewave Inverters




Sinewave

Finally there are types of inverters that output a nearly pure "sinewave. These types of

inverters may involve extensive and carefully tuned filtering or digital synthesis. In thepast these types of inverters tended to have efficiencies and surge capabilities slightlybelow the modified sinewave types, but recently models have been introduced withefficiencies above 90% and costs below some modified squarewave models. Ingeneral, it is always best to choose a sinewave inverter and supply a sinewave outputto your AC loads. This is what they were designed for, and little or no problems willoccur with respect to harmonic distortions or inadequate peak voltages. Some classesof electronic loads, such as telecommunications or delicate instrumentation may requirea sinewave power waveform.

Sinewave Inverters




Exercise




Inverter Circuit Topologies

The topology of inverter design refers to the electrical approach to the conversion

process and the electrical principals applied to designing the control and power circuits.

One fundamental topology difference is where and how the inverter circuit changesfrom DC to AC and from the input voltage to the output voltage. Basically there are twochoices

• First convert to the final output frequency of 60/50Hz AC and then transform voltage

• Upconvert the voltage first and then convert to 60/50 Hz AC for final output.

Final Frequency First, Then TransformVoltage

In the first topology a 60/50 Hz switching circuit is used to “chop” the input DC voltageat the desired final output frequency. The input voltage (typically 12 or 24 volts)remains the voltage of the output of this first stage of the inverter circuitry. Then thislow voltage AC waveform is put on one side of a large transformer. The ratio of thenumber of windings on either side of the transformer determines the change in voltagethrough the transformer. For example, to transform 12 volts AC to 120 volts AC, theoutput side (120 volts) of the transformer would need to have ten times the number of

windings of the input side (12 volts). The output of this stage of the inverter would thenbe a desired frequency and would be at the desired voltage. The transformer isbetween the switching transistors and the AC output side of the inverter. This may beseen as an advantage as the transformer can absorb high energy spikes coming IN tothe inverter (from lightning, load spikes, etc.) and act to protect the more delicatetransistor switching components.




First Transform Voltage, Then FinalFrequency

In the second topology, very high frequency switching of the order of 25,000 Hz isperformed on the input DC voltage. The output at this stage is still 12 or 24 volts butthe frequency is very high. This waveform is then applied to one side of a transformerwith the winding ratio needed to output at 165 volts. This voltage is the peak-to-peakvoltage of nominal 120-volt power. This high frequency high voltage waveform is thenrectified, or returned to DC power. It is still at 165 volts. The power is now switched atthe desired final output frequency (60/50 Hz), to give the final output waveform at thedesired voltage and frequency. The advantage of the high frequency switching is thatthe size of the transformer, and the associated weight and cost, is less than thatneeded for 50/60 Hz switching. However the transformer is now no longer between theAC output and the final switching transistors, and they are more vulnerable to highenergy spikes and noise coming into the inverter.

Pulse Width Modulation

An approach to creating a sinewave output is to apply Pulse Width Modulation (PWM)switching to the power. Pulses are first made very short, then gradually longer until inthe middle of the wave the pulse is the longest. Then the pulses gradually get shorter.In this way the time-average of the energy of the pulses forms a sinusoidal voltagecurve. This PWM switching can be done either before or after the main transformer, asdescribed above.

Multiple Transformers

Another newer topology (used in the sinewave utility interactive Trace Engineeringinverters for example) involves not one but three transformers. One transformer ratio is3:1, another is 9:1 and another is 27:1. These are switched on and off in an additive orsubtractive manner to create a variable ratio configuration producing many small steps(around 50 for the Trace model, depending on battery voltage). This allows for astepped approximation to a sinewave output waveform.




Utility Interactive Inverters

There is a class of inverters used in photovoltaic systems that interact with the utility

grid. They draw power from the grid when the solar array cannot supply all the loads,and feed power into the grid when excess power is available from the array. In order tosynchronize with the utility power waveform, their output must be sinewave also. Untilrecently they have been completely dependent on utility power to operate and have notallowed battery interaction. But recent developments have produced models that are“bi-directional” and that can feed utility power into batteries or can instantaneously drawfrom batteries if there is a utility power failure.

These types of inverters are primarily used in situations where the grid is reliable, asdifferent from traditional markets for remote photovoltaic power systems where gridpower is non-existent or unreliable. Power from an array can be fed directly into a

home or business and displace power that would be needed to run daytime loads, suchas air-conditioning or lighting. This approach is especially cost effective where utilitieshave established time-of-day rates. Typically they charge higher rates during the peakof the day, right when photovoltaic systems would be displacing the most power andsaving the most. Then at night when power would be draw from the utility, the rates arelower and the costs to the user are lower.

Systems that are utility interactive are simpler than typical stand-alone power systems.They do not involve batteries or charge regulators and are therefore lower in total costper watt.

Some models of utility interactive inverters operate at nominal 45 volts so they wouldneed strings of only three modules in series, while others operate at up to +/- 240 voltsand require strings of 32 modules in series.




Bi-Directional Utility Interactive Inverters

Until recently utility interactive inverters have been completely dependent on utilitypower to operate and have not allowed battery interaction. But new developments have

produced models that are “bi-directional” and that can feed utility power into batteries orcan instantaneously draw from batteries if there is a utility power failure. For example,Trace Engineering in the US has introduced a 4000-watt bi-directional inverter thatgives new possibilities to the utility interactive market. Their inverter operates from 24or 48 volts DC input and outputs a synthesized sinewave output waveform suitable forthe utility grid. Their design draws power instantaneously from either the grid or thebattery and solar array to power AC loads, and can pass power back to the utility or thebattery as needed. If there is a utility power failure, power is seamlessly drawn from thebattery and there is no transfer switch “glitch”.

A critical requirement of utility interactive inverters is to stop supplying power into theutility grid when the grid is shut off. This is an important safety requirement, becausewhen utility linemen shut off a section of the grid for maintenance or repairs, they do notwant any sources of power to be feeding into those lines! Any photovoltaic powersystem connected to those lines must also be shut off immediately. Otherwise a small“island” of dangerous power would exist on the line. In the past concerns over“islanding” would lead utilities to require an external switch that the linemen couldmanually switch off outside of any utility connected residence. Now, however, utilityinteractive inverters use a variety of methods to detect that grid power is absent andstop inverting automatically. These methods include (a) voltage over or under detect;(b) frequency over or under detect; and (c) waveform “steps” detection.

The market for utility interactive photovoltaic systems is still in its infancy with only a fewthousand systems worldwide. But new developments in inverter technology are makingthat market more cost effective and exciting.




Exercise




Summary of Inverter Features

Below is presented a summary table listing various features of inverters, and discussing

some of the most important issues related to photovoltaic system design.

Feature Explanation Considerations in PV SystemDesign

OutputVoltage andCurrent

Inverters manufactured in theUS are 120 or 240 vac 60 Hzoutput to match utility output.

Almost all US inverter makersalso make "export" models with220 or 240 vac 50 Hz output.Some inverters are "cascadable"such that their power outputscan be combined.

WaveformType

See previous discussion in thismanual. There are three types;sine wave, modified sine wave,and square wave

Sine waves are best forapplications requiring low noise;e.g. audio/video, computers, andcommunications. Modified sinewaves are best for situationsrequiring high efficiency. Squarewaves are cheap and best suitedfor resistive loads.

PV Input

Voltage Limits

Inverters are designed for a

specific DC input voltage (i.e.12, 24, etc.). For the specifiedinput voltage, there will be upperand lower limits. For a typical12-volt inverter, it might rangefrom 11 to 16 volts.

If one is using alkaline batteries

(Nicads), this figure may be ofimportant as alkaline batteriesmay reach 16 VDC while underrecharge.




No LoadConsumption

This is the amount of power thatthe inverter consumes whenoperating, but not powering anyloads.

Many inverters have sleepcircuits, which puts the inverterto "sleep" in a mode where nopower is consumed.

TotalHarmonicDistortion

THD is a measure of howperfectly an inverter's waveformmatches the ideal sine wave.Sine wave inverters have a THDof 1 - 5%. Modified sine wavehave THD of 10-30% and squarewave of a THD > 30%. As areference, in the US, commercialutility power has a THD ofaround 3 - 8 %

The lower the THD, the lessamount of noise and radiofrequency interference.

Output Power This refers to the "continuous"output power for the inverter

A few inverter manufacturersmay still publish output powerthat is not continuous, butderated over time.

Surge Power The surge capability of aninverter allows it to have astarting wattage of 2 to 4 timesits rated output wattage. Thetype of motor that you are

running can require a surge of 3times (Brush Type) to 7 times(split phase). The bigger theinverter surge wattage, the largerthe electric motor you can run.

The new 25,000 Hz"transformerless" inverters havelower surge capabilitiescompared to the traditional 60Hz inverters. This means that

you may need more invertercapacity in order to run motors ifyou are using 25,000 Hzinverters.




RMS Voltageand PeakVoltage

To be close to emulating gridpower, inverter RMS voltageshould to be within 10% of thestandard utility RMS voltage and

peak voltage should be within15% of the standard utility peakvoltage.

Low RMS and peak voltage willcause poor operation of motorsand appliances that havetransformers (TVs, stereos, etc.).

RMS voltage and peak voltagesthat are too high will damageappliances

Power Factor Ideally an inverter should beable to deal with all powerfactors, from -1 to 1, where +1represents a pure resistive loadand -1 representing a pureinductive load.

The inverter industry has beensuccessful in producing productsto accommodate a wide range ofpower factors. Most reactiveloads (appliances) in homeshave power factors greater than0.5.

ProtectiveCircuitry

Inverters are now designed witha broad range of protection (e.g.fault, overload, over-temperature, and high/lowvoltage).

Good design will have thisprotection.

BatteryCharger

Some manufactures build in thecircuitry for battery charging intothe inverter housing. Since thetransformer already exists, the

inverter drives the transformerbackwards as a battery charger.

Three-stage charging ispreferred; Stage One: BulkCharge with maximum chargeamps to maximum voltage.

Stage Two: ConstantVoltage/Absorption, wherecurrent tapers off at rate requiredto hold voltage constant. StageThree: Float Voltage, whichholds full charge withoutgassing. In addition anequalization stage should beincluded which allows chargingto a voltage high enough to gasbatteries in order to remove

sulfation.




UtilityTransferRelay

Many inverters with batterychargers are capable oftransferring their load to theutility or an engine/generator.

If the transfer is performed withelectro-mechanical relays, thenthe time of transfer mayunsuitable for continuous

operation of computers andother electronic loads.

UL Listing This tells the user that theinverter is not a fire hazard.

Important feature to have inmeeting NEC code forresidential systems.

Grounding NEC requires that the greenequipment grounding conductorand the white neutral conductorbe grounded on the AC side.

Current should not flow in thegrounding wire. The equipmentgrounding conductor should besized for the DC input.

Proper grounding of the inverteris important to meet code.Many non-UL listed inverters donot have a provision for the

equipment ground connection.Some inverters ground theirchassis to the negativeconductor that will not pass ULstandards.

Efficiency To date efficiency has beenmeasured using resistive loads,which has little meaningbecause most loads being runare primarily reactive like motors

and appliances.

This is considered by many ameaningless figure until invertersare tested for efficiency withreactive loads.




Further Discussion of SelectedInverter Characteristics

When choosing an inverter for a particular system, the factors of surge capacity,efficiency and waveform are perhaps the most fundamental. The inverter must be ableto surge to start all the expected inductive loads. It must have enough continuousoutput power to handle all the expected loads that might be on at one time. And thewaveform must be appropriate for all the electronic equipment. There are some otherdesign parameters that should be considered as well. These are described below.

DC Input Voltage

The input DC voltage tends to be a function of the size of the inverter. As the powerthrough the inverter increases, more current flows and there is greater internal heating.Manufacturers tend to increase the DC voltage for larger inverters to keep the currentsat a manageable level, around 100 amps or so. So small inverters from 100-2500 wattscontinuous tend to be available in 12 volts, while inverters in the 2000-3500 watt rangeare made to operate at 24 volts, and models from about 2400-5000 watts are designedwith 48 volt input. Inverters with output greater than 5 kW can have input DC voltagesof 120 to 240 volts and even higher.

If all the loads in a system will be AC then the choice of input voltage is fairly free to

make. The higher the DC voltage the lower the currents into the inverter. This meansthat smaller wire can be used, as well as smaller and less expensive disconnectswitches or circuit breakers and fuses. The DC voltage of the inverter will set thevoltage of the battery bank and the array as well.

Do not make the common mistake of thinking that a 24-volt inverter will need twice asmany modules as a 12-volt inverter. The 24-volt choice will use just as many totalmodules, but they will be configured differently. Instead of all the modules beingconnected in parallel for 12 volts, half as many will be connected in parallel but eachmodule will be connected to another in series. The total will be the same. Andcomparing a 48-volt choice to a 24-volt choice would be similar. Strings of four

modules in series would be used to develop the 48 volts instead of strings of two inseries for 24 volts, but there would be half as many strings in parallel compared to a 24-volt case. If DC and AC loads are to be operated in the same system, then the choiceof input voltage is pretty well restricted to 12, or perhaps 24 volts in some cases. Mostall DC loads are designed to operate at 12 volts. Some ballasts for fluorescent lightsand some refrigerator models come in a 24 volt version.




Voltage Regulation

At high power levels, the inverter draws larger currents from the battery. This causesthe battery voltage to fall. The inverter should be able to compensate for this voltage

drop and maintain output AC voltage fairly well. This information may not be presentedwith the usual efficiency curves and is quite important. If output voltage drops toomuch, then loads may not operate correctly.

Serviceability

The inverter design should allow easy servicing in the field, or allow for cards to beswapped and exchanged to minimize down time.

Adjustable Threshold

Most inverters have some threshold of load power requirement before they actually"turn on" and commutate to produce AC power. This may be a fixed power level, ormay be adjustable by the user. If the load threshold is higher than some small loads ina house, for example a small electronic clock, the inverter may not sense the load and itwill not operate alone. Some other loads must be on so that the total power is greaterthan the threshold value. Some inverter models offer an adjustable threshold level.However, the lower the threshold level – the slightly lower the efficiency of the inverter.

Paralleling For Greater Power

Some inverters can be connected together in parallel and synchronized to increase themaximum continuous power output. This allows for a small inverter to be installedinitially, and then more power installed later if loads increase or if budget allows.




(End Of Chapter)




CHAPTER ELEVEN

INVERTER TECHNOLOGY 11-1

Inverter Power Rating 11-2 Power Is A Function Of Time 11-2 Surge Power 11-3 Temperature Affects Power Rating 11-5

Inverter Efficiency 11-7 Efficiency Depends On Load 11-7 Inductive vs. Resistive Loads 11-9

Test Set-up For Measuring Efficiency 11-9

Inverter Output Waveform 11-11 Squarewave 11-11 Modified Squarewave or “Modified Sinewave” 11-12 Sinewave 11-13

Inverter Circuit Topologies 11-15 Final Frequency First, Then Transform Voltage 11-15 First Transform Voltage, Then Final Frequency 11-16

Pulse Width Modulation 11-16 Multiple Transformers 11-16

Utility Interactive Inverters 11-17 Bi-Directional Utility Interactive Inverters 11-18

Summary of Inverter Features 11-20

Further Discussion of Selected Inverter Characteristics 11-24 DC Input Voltage 11-24

Voltage Regulation 11-25 Serviceability 11-25 Adjustable Threshold 11-25 Paralleling For Greater Power 11-25



Siemens Solar Basic Photovoltaic Technology 11-1 Inverter Technology

Chapter 11 – Answers Inverter Technology



The average efficiency of the transistor based inverter is about 92%. The average

efficiency of the SCR based inverter is about 75%.


Compare waveforms to those shown in the text.




Siemens Solar Basic PV Technology Course Components – Charge Regulators/ControlsCopyright © 1998 Siemens Solar Industries

12-1

Chapter TwelveCharge Regulators and

System ControlsPhotovoltaic modules are highly reliable and virtually maintenance-free; but,modules alone do not solve a customer’s power problem. Other components aregenerally required to properly control, distribute and store the energy produced bythe array to power an electrical load.

In stand-alone PV systems the array is usually connected to batteries that store the

array energy and supply power to the electrical loads on demand. In most cases,when batteries are included in a PV system, a charge regulator or controller isrequired to protect the batteries from being overcharged by the array andoverdischarged by the system loads. In some designs the charge regulator orsystem controller may also provide status information to the user/operator on systemperformance and battery state of charge.

In this chapter we will discuss the purpose of battery charge regulators and controlsin PV systems, their important features, and how different types operate to controlthe flow of energy in a remote PV power system.




12-4

Prevent Battery Overdischarge

During periods of below average insolation and/or during periods of excessiveelectrical load usage the energy produced by the PV array may not be sufficientenough to keep the battery fully recharged. When a battery is deeply discharged thereaction in the battery occurs close to the grids, and weakens the bond between the

active materials and the grids. When a battery is excessively discharged repeatedlyloss of capacity and life will eventually occur. To protect batteries fromoverdischarge most charge regulators include an optional feature to disconnect thesystem loads once the battery reaches a low voltage or low state of chargecondition.

In some cases the electrical loads in a PV system must have sufficiently highenough voltage to operate. If batteries are too deeply discharged the voltage fallsbelow the operating range of the loads, and the loads may operate improperly or notat all. This is another important reason to limit battery overdischarge in PV systems.

Overdischarge protection in charge regulators is usually accomplished by open-circuiting the connection between the battery and electrical load when the batteryreaches a pre-set or adjustable low voltage load disconnect (LVD) set point . Mostcharge regulators also have an indicator light or audible alarm to alert the systemuser/operator to the load disconnect condition. Once the battery is recharged to acertain level, the loads are again reconnected to a battery.

Non-critical system loads are generally always protected from overdischarging thebattery by connection to the low voltage load disconnect circuitry of the chargecontroller. If the battery voltage falls to a low but safe level, a relay can open anddisconnect the load, preventing further battery discharge. Critical loads can be

connected directly to the battery, so that they are not automatically disconnected bythe charge regulator. However, the danger exists that these critical loads mightoverdischarge the battery. An alarm or other method of user feedback should beincluded to give information on the battery status if critical loads are connecteddirectly to the battery.




12-5

Provide Load Control Functions

In some cases regulators may have optional features that allow regulation or controlof the PV system electrical load. Load control in PV lighting system regulators is apopular feature. This control can take place at sunset or sunrise as sensed by aphotosensor or the array current or voltage output. In other cases the controller may

have a timing function to cycle the load operation for a specified period or at acertain time of day. Experience has shown that these load control functions mayrequire adjustment and proper specification for the array type and site conditionssuch as temperature and background lighting. While these features add to the costand complexity of the controller they can also greatly simply the use and operation ofthe PV system.

Where load voltage fluctuations may be detrimental to certain loads, load voltage regulation may be desirable to limit the voltages at which the loads operate. A fewadvanced PV system controllers are designed to perform this function.

Provide Status Information To Users

Many charge regulators used in PV systems can provide status information on theoperation of the system and condition of the battery. These optional features ofregulators can allow users to intelligently manage their use of energy and gain abetter understanding of how the system operates to fully utilize it’s potential.

Battery voltage and/or state of charge is an essential piece of information that canbe indicated by charge regulators. This can be incorporated as a “gas gauge” with asimple dial voltmeter showing green, yellow or red regions corresponding to differentbattery voltage ranges. Or a digital readout of exact battery voltage can be providedfor more sophisticated users. While random battery voltage readings are interesting,the user must know how to interpret this information. For example, they must knowat what voltage levels they should begin to curtail their energy usage, to preventbattery overdischarge.

Many regulators also indicate with a lamp or ammeter whether the array is chargingthe battery, or if the array or load is disconnected. Load currents or even the netbattery amp-hours can also be displayed on some regulator/controller designs.Knowledge of the array and load currents gives users a sense of how much energythey are producing with the array and consuming with electrical loads. This mayallow them to plan their load usage to better correspond with the energy availabilityof their system during low insolation periods.




12-6

Interface and Control Backup EnergySources

In the case of a hybrid PV power systems using one or more backup energy sources

in addition to the array more advanced system control centers may be designed tointerface these alternate sources with the PV electrical system. One example wouldbe a controller that activates a backup generator at low battery state of charge. Thecontrol center would start the generator at a pre-set low battery voltage, and turn itoff when the battery is recharged or reaches a higher voltage limit.

The control of backup energy sources can also be performed by other componentsin the PV system. For example, some stand-alone inverters used in PV systems willstart a backup generator or divert the loads to utility power when the battery reacheslow state of charge. Once the battery has been recharged to a pre-set level, theloads are again connected to and powered by the battery bank.

Divert PV Energy To Auxiliary Load

Batteries in photovoltaic systems are often fully recharged by the middle of the dayduring the summer. Normally, the charge regulator disconnects the array to preventbattery overcharging, wasting valuable array energy. To utilize this excess energy,some controllers allow the diversion of array energy to power an auxiliary load oncethe primary battery bank is fully charged. In most cases the regulator will bebypassed, directly-coupling the PV array to the auxiliary load. The auxiliary load istypically a backup battery, a DC water pump, a resistive element in a water heater,or a fan or some other simple motor load that can be operated from the array withoutvoltage regulation. In this way all the energy that the array can produce is beingutilized for some purpose.

When the battery voltage falls, and array power is once again needed to run theregular loads, the auxiliary load is then disconnected. So the auxiliary load must bean optional, non-critical load that can be operated whenever excess array energy isavailable.




12-7

Serves As Wiring Center

In most cases the charge regulator or system controller serves as the terminationand connection point between the conductors leading to the various components ina PV system. For example, the charge regulator in a small residential lightingsystem commonly has the PV array, battery and load all connected to the regulator

terminals. A fuse or circuit breaker for array and battery protection can also beincluded in the regulator design.

Larger PV systems generally have overcurrent protection and disconnect devicesincluded as part of the system control center. With these devices included in thesystem controller, the conductors leading to the PV array, battery and loads areconnected at a centralized point in the system. The control center may also be theprinciple grounding point in the system and include surge suppression devices. Withthis centralized configuration for the system connection and controls, the installation,operation and maintenance of the system is greatly simplified.




12-8

Terminology

While the specific regulation method or algorithm may vary among different chargeregulators, all have basic functions and characteristics. Charge regulatormanufacturer's data generally provides information about these functions and their

specifications.

Nominal System Voltage

The nominal system voltage is the voltage at which the battery and charge regulatoroperate in a PV system. Most charge regulators are designed for operation at aspecific nominal system voltage, while some may allow operation at multiplevoltages, for example with 12 or 24-volt systems.

As mentioned other places in this manual, the selection of the nominal systemvoltage has important ramifications on many aspects of design and equipmentselection. For systems with higher load power demands a higher nominal systemvoltage is generally used to lower the peak operating currents, reducing the size andratings of conductors, overcurrent and disconnect devices and the charge regulator.However, most equipment for small stand-alone PV systems is widely availability in12 or 24-volt dc models, and this may dictate the designer’s selection of the nominalsystem voltage.

Nominal Load and PV Array CurrentCharge regulators are rated for their ability to handle certain maximum and nominalcurrents for the PV array and load. Often, surge conditions may exist from the arrayand loads, and the regulator must be tolerant of these conditions. A furtherdiscussion of selecting and sizing charge regulators is presented later in thischapter.




12-9

Charge Regulator Set Points

The battery voltage levels at which a charge regulator performs control or switchingfunctions are called the regulator set points. Four basic control set points aredefined for most charge regulators that have battery overcharge and overdischargeprotection features. The voltage regulation (VR) and the array reconnect voltage

(ARV) refer to the voltage set points at which the array is connected anddisconnected from the battery. The low voltage load disconnect (LVD) and loadreconnect voltage (LRV) refer to the voltage set points at which the load isdisconnected from the battery to prevent overdischarge. Figure 12-1 shows thebasic regulator set points on a simplified diagram plotting battery voltage versus timefor a charge and discharge cycle. A detailed discussion of each charge controllerset point follows.


Time

B a t

t e r y V o l t a g e

Charging Discharging

Low Voltage Load Disconnect (LVD)

Voltage Regulation (VR)

Voltage Regulation Hysteresis (VRH)

Low Voltage Disconnect Hysteresis (LVDH)

Load Reconnect Voltage (LRV)

Array Reconnect Voltage (ARV)




12-10

Voltage Regulation (VR) Set Point

The voltage regulation (VR) set point is one of the key specifications for chargeregulators. The voltage regulation set point is defined as the maximum voltage thatthe charge regulator allows the battery to reach, limiting the overcharge of thebattery. Once the controller senses that the battery reaches the voltage regulationset point, the controller will either discontinue battery charging or begin to regulate

(limit) the amount of current delivered to the battery. In some regulator designs, dualregulation set points may be used. For example, a higher regulation voltage may beused for the first charge cycle of the day to provide a little battery overcharge,gassing and equalization, while a lower regulation voltage is used on subsequentcycles through the remainder of the day to effectively ‘float charge’ the battery.

Proper selection of the voltage regulation set point may depend on many factors,including the specific battery chemistry and design, sizes of the load and array withrespect to the battery, operating temperatures, and electrolyte loss considerations.For flooded batteries the regulation voltage should be selected at a point that allowsthe battery to achieve a minimal level of gassing. However, gassing should be

avoided for sealed, valve-regulated lead-acid (VRLA) batteries. Temperaturecompensation of the voltage regulation set point is often incorporated in chargecontroller design, and is highly recommended for VRLA batteries and if batterytemperatures exceed ± 5

oC from normal ambient temperatures (25

oC). A

discussion on voltage regulation set point selection and temperature compensationis contained later in this chapter.

An important point to note about the voltage regulation set point is that the valuesrequired for optimal battery performance in stand-alone PV systems are generallymuch higher than the regulation or ‘float voltages’ recommended by batterymanufacturers. This is because in a PV system, the battery must be recharged

within a limited time period (during sunlight hours), while battery manufacturersgenerally allow for much longer recharge times when determining their optimalregulation voltage limits. By using a higher regulation voltage in PV systems thebattery can be recharged in a shorter time period, however some degree overovercharge and gassing will occur. The designer is faced selecting the optimalvoltage regulation set point that maintains the highest possible battery state ofcharge without causing significant overcharge.




12-11

High Voltage Alarm

In some regulator designs a high voltage alarm is included to alert the systemuser/operator of a dangerously high battery voltage condition. In the event of acontroller failure, or failure of a backup source regulator, the high voltage alarmsends an audible or visible signal to the operator if the battery voltage exceeds apre-set level. The settings for the high voltage alarm should be slightly higher than

the maximum expected voltage regulation set point of the charge regulator (includingany increases for temperature compensation). For most applications typical highvoltage alarm settings range upward from about 14.8 volts.

Array Reconnect Voltage (ARV) Set Point

In interrupting (on-off) type regulators, once the array current is disconnected at thevoltage regulation set point, the battery voltage will begin to decrease. The rate atwhich the battery voltage decreases depends on many factors, including the chargerate prior to disconnect, and the discharge rate dictated by the electrical load. If the

charge and discharge rates are high, the battery voltage will decrease at a greaterrate than if these rates are lower. When the battery voltage decreases to apredefined voltage, the array is again reconnected to the battery to resume charging.This voltage at which the array is reconnected is defined as the array reconnect voltage (ARV) set point .

If the array were to remain disconnected for the rest of day after the regulationvoltage was initially reached, the battery would not be fully recharged. By allowingthe array to reconnect after the battery voltage reduces to a set value, the arraycurrent will ‘cycle’ into the battery in an on-off manner, disconnecting at theregulation voltage set point, and reconnecting at the array reconnect voltage set

point. In this way, the battery will be brought up to a higher state of charge by‘pulsing’ the array current into the battery.

It is important to note that for some regulator designs, namely constant-voltage andpulse-width-modulated (PWM) types, there is no clearly distinguishable differencebetween the VR and ARV set points. In these designs the array current is notregulated in a simple on-off or interrupting fashion, but is only limited as the batteryvoltage is held at a relatively constant value through the remainder of the day. Adiscussion on these types of regulators is included later in this chapter.




12-12


The voltage span or difference between the voltage regulation set point and thearray reconnect voltage is often called the voltage regulation hysteresis (VRH). TheVRH is a major factor that determines the effectiveness of battery recharging forinterrupting (on-off) type regulators. If the hysteresis is to great, the array currentremains disconnected for long periods, effectively lowering the array energy

utilization and making it very difficult to fully recharge the battery. If the regulationhysteresis is too small, the array will cycle on and off rapidly, perhaps damagingregulators which use electro-mechanical switching elements. The designer mustcarefully determine the hysteresis values based on the system charge and dischargerates and the charging requirements of the particular battery.

Most interrupting (on-off) type regulators have hysteresis values between 0.4 and1.4 volts for nominal 12-volt systems. For example, for a regulator with a voltageregulation set point of 14.5 volts and a regulation hysteresis of 1.0 volt, the arrayreconnect voltage would be 13.5 volts. In general, a smaller regulation hysteresis isrequired for PV systems that do not have a daytime load. This is because daytime

loads will pull the battery voltage down past the reconnect voltage and initiate arraycharging during the day. If only nighttime loads are operated, then the amount oftime that on-off type regulators spend with the array connected will be less than ifloads were operating throughout the day.

Low Voltage Load Disconnect (LVD) Set Point

Overdischarging the battery can make it susceptible to freezing and shorten itsoperating life. If battery voltage drops too low, due to prolonged bad weather forexample, certain non-essential loads can be disconnected from the battery to

prevent further discharge. This can be done using a low voltage load disconnect (LVD) device connected between the battery and non-essential loads. The LVD iseither a relay or a solid-state switch that interrupts the current from the battery to theload, and is included as part of most regulator designs. In some cases the lowvoltage load disconnect unit may be a separate unit from the main charge regulator.

In regulators or controls incorporating a load disconnect feature the low voltage load disconnect (LVD) set point is the voltage at which the load is disconnected from thebattery to prevent overdischarge. The LVD set point defines the actual allowable maximum depth-of-discharge and available capacity of the battery operating in a PV system. The available capacity must be carefully estimated in the PV system design

and sizing process using the actual depth of discharge dictated by the LVD set point.

In more sophisticated designs a hierarchy of load importance can be established,and the more critical loads can be shed at progressively lower battery voltages.Very critical loads can remain connected directly to the battery so their operation isnot interrupted.




12-13

The proper LVD set point will maintain a healthy battery while providing themaximum battery capacity and load availability. To determine the proper loaddisconnect voltage, the designer must consider the rate at which the battery isdischarged. Because the battery voltage is affected by the rate of discharge, a lowerload disconnect voltage set point is needed for high discharge rates to achieve thesame depth of discharge limit. In general, the low discharge rates in most smallstand-alone PV systems do not have a significant effect on the battery voltage.

Typical LVD values used are between 11.0 and 11.5 volts, which corresponds toabout 75-90% depth of discharge for most nominal 12 volt lead-acid batteries atdischarge rates lower than C/30.

A word of caution is in order when selecting the low voltage load disconnect setpoint. Battery manufacturers rate discharge capacity to a specified cut-off voltagethat corresponds to 100% depth of discharge for the battery. For lead-acid batteries,this cut-off voltage is typically 10.5 volts for a nominal 12-volt battery (1.75 volts percell). In PV systems, we never want to allow a battery to be completely discharged,as this will shorten its service life. In general, the low voltage load disconnect setpoint in PV systems is selected to discharge the battery to no greater than 75-80%

depth of discharge.

In cases where starting (SLI) batteries are used or it is otherwise desired to limit thebattery depth of discharge to prevent freezing or prolong cycle life, a higher LVD setpoint may be desired. To protect the battery from freezing, the LVD set point maybe temperature compensated in some cases to increase the load disconnect voltageautomatically with decreasing battery temperature.

To properly specify the LVD set point in PV systems the designer must know howthe battery voltage is affected at different states of charge and discharge rates. In afew designs, current compensation may be included in the LVD circuitry to lower the

LVD set point with increasing discharge rates to effectively keep a consistent depthof discharge limit at which the LVD occurs.




12-14

Low Voltage Alarm

An optional regulator feature is a low voltage alarm (LVA) designed to alert the userof a low battery state of charge condition. If the charge regulator were to fail, or thearray damaged, or during long periods of below average insolation, the battery stateof charge might drop below safe levels. A low-voltage alarm can be included in aremote power system to monitor battery voltage and send an audible, visual or

electronic alarm if the battery voltage drops below some predetermined safe range.Typical low voltage alarm settings for nominal 12-volt batteries are about 11.8-11.5volts, corresponding to about 60-70 % depth of discharge.

One advanced feature of some controllers is “current compensation” of the LVAvoltage. When high load currents are flowing the battery voltage will be temporarilylowered due to the internal voltage drop in the battery. This temporary low voltageshould not activate the LVA. It should only come on when a low battery voltage issustained for a long period of time. Current compensation circuitry monitors the loadcurrent and lowers the LVA set point correspondingly, so that nuisance alarms areavoided.

Load Reconnect Voltage (LRV) Set Point

The battery voltage at which a regulator allows the load to be reconnected to thebattery is called the load reconnect voltage (‘LRV). After the regulator disconnectsthe load from the battery at the LVD set point, the battery voltage rises to its open-circuit voltage. When the array or a backup source provides additional charge thebattery voltage rises even more. At some point the controller senses that the batteryvoltage and state of charge are high enough to reconnect the load, called the load reconnect voltage set point .

The selection of the load reconnect set point should be high enough to ensure thatthe battery has been somewhat recharged, while not to high as to sacrifice loadavailability by allowing the loads to be disconnected too long. Many controllerdesigns effectively ‘lock out’ loads until the next day or when the controller sensesthat the array is again recharging the battery. Typically LVD set points used in smallPV systems are between 12.5 and 13.0 volts for most nominal 12-volt lead-acidbatteries. If the LRV set point is selected too low, the load may be reconnectedbefore the battery has been charged, possibly cycling the load on and off, keepingthe battery at low state of charge and shortening it’s lifetime.

As in the selection of the other regulator set points, the designer must consider thecharge rates for the loads and array and how these rates affect battery voltage atdifferent states of charge.




12-15

Low Voltage Load Disconnect Hysteresis (LVDH)

The voltage span or difference between the LVD set point and the load reconnectvoltage is called the low voltage disconnect hysteresis (LVDH). If the LVDH is toosmall the load may cycle on and off rapidly at low battery state-of-charge (SOC),possibly damaging the load or controller, and extending the time it takes to fullycharge the battery. If the LVDH is too large, the load may remain off for extended

periods until the array fully recharges the battery. With a large LVDH, battery healthmay be improved due to reduced battery cycling but with a reduction in loadavailability. The proper LVDH selection for a given system will depend on loadavailability requirements, battery chemistry and size, and the PV and load currents.




12-16

Other Functions Associated WithCharge Regulation

Backup Energy Source ControlAnother optional feature of some charge regulators is the control of auxiliary orbackup energy sources. Sometimes operating in conjunction with low voltage alarm(LVA) circuits, the controller switches power to start a backup generator andconnects it to the battery when the state of charge and voltage reach a low level. Anaudible alarm can be used for an occupied site to alert the user that it is time tomanually turn on the generator. Or the control unit can automatically switch on agenerator with electric start. Another control can be set to a higher voltage to turnthe generator off as the battery voltage reaches a predetermined higher state ofcharge.

Equalizing Charge Capability

For some battery types, particularly for tall flooded lead-antimony types, periodicequalization charges are required to maintain optimal battery performance. Undermost circumstances, this requires user intervention in system operation, bybypassing the normal charge regulation circuit for the duration of the equalizationcharge. To perform equalization charging automatically, some new regulators allowtheir VR set point to be overridden periodically, either when the battery voltage has

dropped particularly low, or at regular time intervals (e.g. 15-20 days). Thefrequency and extent to which equalization charges occur can be programmed intothe controller microprocessor circuitry.

Set Point Adjustability

While some charge controllers use fixed resistors to pre-set the controller set pointsmany have the ability to adjust or change the regulation and load disconnect setpoints. Some also have provisions to adjust the hysteresis values as well.

Adjustments are typically made with potentiometers (single or multi-turn), DIP (dualin-line package) switches, or circuit board jumpers. Although the system designer orinstaller may need access to properly set the controller for the type of battery andsystem configuration, user/operator access to regulator adjustments should bediscouraged.




12-17

Load Voltage Regulation

For some PV applications with critical loads such as telecommunications and

telemetry equipment, the load must operate at a specific voltage or within a

prescribed voltage range. While an external dc-dc converter between the battery

bank and load sometimes accomplishes this task, some advanced charge regulator

types can provide load voltage regulation. A DC-DC converter may be incorporatedin the controller circuitry, or a simple voltage regulator circuit may be used if the load

voltage requirements are less than the lowest expected battery voltage.

Regulation/Control Element Design

The regulation or switching control element in charge regulators can be either a

solid-state device or an electro-mechanical relay. Simple interrupting or on-off

regulators may use relays, however in most cases MOSFETs or power transistorsare used rather than relays because they have lower power requirements, are

smaller, and can operate for many more cycles. The regulator switching elements

must be properly rated for the application. The steady-state dc current ratings for

the regulator element should be at least 125% of the maximum PV array short-circuit

current. The peak or surge DC current ratings for the element should be at least

150% of maximum expected PV array short-circuit current.

The switching element that controls the low voltage load disconnect circuit is

generally an electro-mechanical relay, as the number of cycles this switching

element experiences is very limited compared to the regulation element. Generally,relay contacts should be plated, or hermetically sealed and rated to handle at least

125% of the rated dc load current. Peak or surge dc current ratings for the load

switching element should also be able to handle expected high current surges from

the system loads.




12-18

Operational Limits

The environmental and mechanical design limits of charge regulators are an

important consideration for most PV applications. Generally, PV systems and

components are installed in remote areas, in unconditioned spaces, and subject to

the extremes of the weather. For these reasons most regulators have minimum and

maximum ratings for ambient temperature, battery temperature and relativehumidity. Where extreme environmental conditions exist the designer should

consider these specifications when selecting charge regulators.

The packaging and physical characteristics are another important characteristic of

charge regulators. In general, the regulator circuitry should be sealed from the

environment, either by conformably coating or ‘potting’ the circuitry. A rigid case

should protect the regulator or the controller may be installed in a weatherproof

enclosure.

Terminations used to connect wiring to the charge regulator should be corrosionresistant, and be large and sturdy enough to accept the conductor sizes that may be

used in the system.

Surge Protection and Grounding

Most regulators include some type of surge suppression devices on the array and

load circuits. Commonly, these devices are metal-oxide varistors (MOVs), which are

connected between the positive and negative terminals, and from this terminal toground. Under normal operating conditions MOVs are high impedance devices.

However under surge conditions, MOVs shunt energy to ground, bypassing sensitive

electrical circuits of the controller. While these devices do not always protect the

controller circuitry from harmful surges, they are strongly recommended for

regulators used in lightning prone areas and in dc/ac systems using inverters.




12-19

Service Disconnects and OvercurrentProtection

Other features that are sometimes included system control centers are overcurrent

protection and disconnect devices. The requirements for overcurrent protection anddisconnect requirements are discussed in detail elsewhere in this manual. Ingeneral, disconnects and overcurrent protection are needed between the PV arraysource circuits and regulators, between the battery and regulators, and perhaps onload circuits. When these components are included as part of the system controlcenter, the installation, operation and maintenance of the system are simplified.




12-20

Standard Configurations OfCharge Regulation and ControlSystems

There are a number of ways that battery charge regulation and system controls areconfigured photovoltaic systems. Depending on the type and size of the system, thecost and availability of hardware, and designer choice, certain regulation and controlconfigurations may be more appropriate than others. Some of the more commonconfigurations of charge regulation and system control are described below. Theelectronic design of charge regulators is presented in the next section.

Simple Series Path ConfigurationThe most common configuration is to install the charge regulator/controller in seriesbetween the array and the battery. When the battery reaches the charge regulationvoltage set point of the controller, the regulator mechanism may either open-circuitor short-circuit the array, or limit the current to lower than the peak array current.

The following figure shows two typical series regulator configurations used in PVsystems. The top diagram shows a simple design where the charge regulator is onlyused to protect the battery from overcharge, but does not protect it fromoverdischarge. In this configuration, the regulator is placed in series between the

PV array and battery, while the load is connected directly to the battery.

The bottom diagram in the figure shows a series configuration with batteryoverdischarge protection. The charge regulator used in this design includes a lowvoltage load disconnect circuit, and the load is connected to the charge regulator atthe load terminals rather than directly to the battery. Charge regulators of this typeare common and are designed to handle PV array currents up to 30 amps. Wherethe load currents exceed the ratings of these controllers an external relay is oftendriven by the charge regulator to control the load.




12-21

Auxiliary Load Path Configuration

Photovoltaic systems are usually sized so that the array produces enough energy tofully operate the loads during the winter months, in order to keep the battery fullycharged. When the solar insolation is high and the load usage is low the chargeregulator disconnects the array from the battery to prevent overcharge, wastingvaluable energy. Where a backup or non-essential load can be utilized the excessarray power can be diverted directly to the backup load or through the battery to thebackup load.

Simple Series Configurations

PV Array Battery LoadCharge

Regulator

Without Overdischarge Protection

With Overdischarge Protection

PV Array Load

Battery

ChargeRegulator




12-22

The figure below shows a basic configuration for auxiliary load connection in PVsystems with similar battery overdischarge protection as described for the seriespath configuration. In this case when the battery reaches full state of charge, thecharge regulator diverts the array power to an auxiliary load while the primary loadcontinues to operate from the battery. In most cases these regulators connect thearray directly to the auxiliary load, and bypass the circuit to the battery. In this

design it should be noted that the array power and voltage delivered to the load willnot be regulated, so the load must be compatible with the array output. Electricalloads that can be directly connected to the array output include auxiliary batteries,DC motor appliances such a pumps or fans, or resistive elements used for water orspace heating. Whenever excess energy is available from the array, these types ofloads can absorb it. When the primary battery voltage drops and array power isonce again needed to recharge it the auxiliary load’s power is cut off.

During the summer months charge regulation limits the array energy utilization by asmuch as 50% in most small stand-alone PV applications. The use of an auxiliaryload can make full use of all the array energy throughout the year that would

otherwise be wasted. Array energy diversion to auxiliary loads is usually used inremote homes or cabins, where the primary load usage is intermittent or infrequent,and where a non-essential load can be utilized. Generally, commercial andindustrial PV applications have little need for the common types of auxiliary loads.

Auxiliary Load Configuration

PV Array BatteryChargeRegulator


AuxiliaryLoad

PrimaryLoad

Auxiliary load is OPTIONALto utilize excess array energy




12-23

Parallel Path Configuration

A less common approach to battery charge regulation is a parallel pathconfiguration. In this configuration the auxiliary load is no longer an option - it isnecessary and required to protect the battery from overcharging.

The figure below shows a diagram of a parallel path configuration withoverdischarge protection. The charge regulator is not in series between the arrayand battery, but is placed in parallel with the battery. The current from the array isfed directly to the battery at all times. When the battery reaches full state of charge,the regulator begins to pass excess array power to the auxiliary load. If the batteryis fully charged then all the power from the array flows across the battery terminalsand to the auxiliary load. The primary load will continue to operate.

The concern with this configuration is that the auxiliary load must always be available and must be able to fully utilize the maximum expected array power. If not,failure of the load or battery overcharge may occur. The advantage of this type of

configuration is that multiple sources of power can be connected to the battery.Sources such a wind generators that produce large variations in current coulddamage some regulators configured in series. But in this approach, the large batterybank acts as a buffer, absorbing all the wind generator energy, and the regulatormust only gradually divert the energy away to the auxiliary load as the batterybecomes charged.

Parallel Path Configuration

PV Array Battery ChargeRegulator


AuxiliaryLoad

PrimaryLoad

Auxiliary load is MANDATORYto regulate battery overcharge




12-24

Sub-Array Switching Configuration

For larger arrays, the commonly available charge regulators used for small stand-alone PV applications may alone not be rated to handle the higher array currents.Charge regulation for arrays larger than 500 to 1000 peak watts is generally

accomplished by grouping the array into parallel sub-groups or sub-arrays, and byregulating each of the sub-arrays independently. The common operational mode forsub-array switching controllers is that each sub-array regulator is set to a slightlydifferent cut-off voltage. As the battery voltage rises, first one sub-array disconnectsat the lowest regulation voltage, and then the next is disconnected, graduallyreducing the charge current into the battery as the voltage increases to the highestregulation voltage set point.

Sub-Array Switching Configuration

Sub-Array #1

Battery

Optional LVD

Sub-Array #2

Sub-Array #3

Charge

Regulator

Charge

Regulator

Charge

Regulator

Load

Sub-Array #3

One sub-array may be connected directly to thebattery for finishing charge




12-25

For this configuration, a common design approach is that several regulators areused in parallel, each with its own sub-array connected. Another approach is onethat a master controller operates relays switching each sub-array, each set to aslightly different voltage point. For example, a 12 volt array could have one sub-array disconnected at 13.8 volts, a second at 14.0 volts, a third at 14.2 volts, and soon. In this way the charge current is reduced in steps, and the battery gradually

approaches full charge. This approach allows for the use of simple on-off typecontrols and yet results in the same gentle approach to full charge as complex multi-stage and constant-voltage regulators.

The switching elements used in sub-array switching control designs can either beelectro-mechanical relays or long-life mercury displacement relays. By separatingthe control mechanism (relay) from the charge control circuitry greater flexibility indesign can be enjoyed, and very large array currents can be regulated with simplesingle stage on-off type control elements (relays).

In systems with many sub-arrays one small sub-array may be left connected directly

to the battery (with appropriate fusing and disconnect switch for wire protection andcode compliance). This final sub-array current provides a gentle finishing charge atthe end of the day. The general requirement is that this one sub array contributesonly low charge rates (for example C/90 or less) to the battery to limit the batteryvoltage.

The number of parallel sub-arrays determines the number of regulators or relaysrequired. To determine the number of regulators or sub-array relays required, dividethe maximum possible array current by the current capacity of one regulator or relay.The maximum array current is the short circuit current of the entire array increasedby a safety factor for possible cloud and ground reflectance, usually a factor of 130%

is used.

Number of Sub-Arrays = Number of Parallel Modules X Isc X 1.3and Regulators Current Capacity of One Regulator




12-26

Example: A large microwave repeater system requires 4 kW of solar array at 24volts. If 35-watt modules are used, this means that a total of 114modules are required.

4000 watts ÷ 35 watts/ module = 114 modules

Since the system requires 24 volts, this means that two modules arewired in series and 57 strings of two-in-series are wired in parallel. Sothe number of parallel modules is 57.

114 modules total = 2 in series X 57 in parallel

A typical size charge regulator is rated at 30 amps (some are larger,some smaller). Using the Isc of 2.15 amps for an 35 watt module, thenumber of regulators and sub-arrays is given by

Number of Regulators = 57 parallel X 2.15 amps X 1.30

30 amps/regulator

= 5.3 regulators, rounded up to 6




12-27

Exercise




12-28

Electronic Designs for ChargeRegulation.

The discussion in the preceding section show the different ways battery chargeregulation can be configured in photovoltaic systems. There are also a number ofvariations in the function and electronic design of charge regulators.

Two basic methods exist for controlling or regulating the charging of a battery from aPV module or array - shunt and series regulation. While both of these methods areeffectively used, each method may incorporate a number of variations that alter theirbasic performance and applicability. Simple designs interrupt or disconnect thearray from the battery at regulation, while more sophisticated designs limit thecurrent to the battery in a linear manner that maintains a high battery voltage.

The algorithm or control strategy of a battery charge regulator determines theeffectiveness of battery charging and PV array utilization, and ultimately the ability ofthe system to meet the electrical load demands. Most importantly, the controlleralgorithm defines the way in which PV array power is applied to the battery in thesystem. In general, interrupting on-off type controllers require a higher regulation setpoint to bring batteries up to full state of charge than controllers that limit the arraycurrent in a gradual manner.

Some of the more common design approaches for charge regulators are describedin this section. Typical daily charging profiles for a few of the common types ofregulators used in small PV lighting systems are presented in the next section.




12-29

Shunt Regulator Designs

Since photovoltaic cells are current-limited by design (unlike batteries) PV modules

and arrays can be short-circuited without any harm. The ability to short-circuit

modules or an array is the basis of operation for shunt regulators.

The figure shows an electrical design of a typical shunt type regulator. The shunt

controller regulates the charging of a battery from the PV array by short-circuiting the

array internal to the regulator. All shunt regulators must have a blocking diode in

series between the battery and the shunt element to prevent the battery from short-

circuiting when the array is regulating. Because there is some voltage drop between

the array and regulator and due to wiring and resistance of the shunt element, the

array is never entirely short-circuited, resulting in some power dissipation within the

controller. For this reason, most shunt controllers require a heat sink to dissipate

power, and are generally limited to use in PV systems with array currents less than

20 amps.

The regulation element in shunt regulators is typically a power transistor or MOSFETdepending on the specific design. There are three general variations of the shuntregulator design. The first is a simple interrupting, or on-off type regulator design.The second type limits the array current in a gradual manner, by increasing theresistance of the shunt element as the battery reaches full state of charge. A thirdapproach is to pulse the current through the shunt element and vary the pulse timeor width depending on the battery state of charge (Pulse Width Modulation or PWM).These variations of the shunt regulator are discussed next.




12-30

Basic Sh unt R egula tor D es ign

P VArray

+

D CLoad

Block ing Diode Load Sw i tch ing Element

LV DContro l

Regulat ionContro l

Shunt E lementBattery

Shunt-Interrupting Design

The shunt-interrupting regulator completely disconnects the array current in aninterrupting or on-off fashion when the battery reaches the voltage regulation set

point. When the battery decreases to the array reconnect voltage, the regulatorconnects the array to resume charging the battery. This cycling between theregulation voltage and array reconnect voltage is why these regulators are oftencalled ‘on-off’ or ‘pulsing’ controllers. These are NOT to be confused with a truepulsing or PWM type control, discussed later. Shunt-interrupting regulators arewidely available and are low cost, however they are generally limited to use insystems with array currents less than 20 amps due to heat dissipation requirements.In general, on-off shunt regulators consume less power than series type regulatorsthat use relays (discussed later), so they are best suited for small systems whereeven minor parasitic losses become a significant part of the system load.

Shunt-interrupting charge regulators can be used on all battery types, however theway in which they apply power to the battery may not be optimal for all battery

designs. In general, constant-voltage, PWM or linear regulator designs are

recommended by manufacturers of gelled and AGM lead-acid batteries. However,

shunt-interrupting regulators are simple, low cost and perform well in most small

stand-alone PV systems.




12-32

The PWM design allows greater control over exactly how a battery approaches full

charge, and the high speed switching elements generate little heat. PWM type

regulators can be used with all battery types, however the controlled manner in

which power is applied to the battery makes them preferential for use with sealed

VRLA types batteries over on-off type controls. To limit overcharge and gassing, the

voltage regulation set points for PWM and constant-voltage controllers are generally

specified lower than those for on-off type regulators. For example, a PWM regulator

operating with a nominal 12 volt flooded lead-antimony battery might use a VR set

point of 14.4 to 14.6 volts at 25o

C, while an on-off regulator used with the same

battery might require a VR set point of between 14.6 and 14.8 volts to fully recharge

the battery on a typical day.

Pulse-W idth-M odula ted (PW M )

R egu lator D esign

Battery a t Low Sta te of Charge

Battery at High State of Charge

Unti l the battery is ful ly charg ed,the current pulses to the b atteryare wide.

Onc e the battery becom es ful ly

charged, the current pulses tothe battery becom e narrower.




12-33

Series Regulator Designs

As the name implies, this type of regulator works in series between the array andbattery, rather than in parallel as for the shunt regulator. There are severalvariations to the series type regulator, all of which use some type of control orregulation element in series between the array and the battery. While this type of

controller is used in small PV systems, it is the most practical choice for largersystems due to the current limitations of shunt controllers. The figure shows anelectrical design of a typical series type regulator.

In a series regulator design array current is controlled in one of three ways.

• A relay or solid-state switch can open the circuit between the array and thebattery to discontinue charging.

• Array current is pulsed rapidly, and the pulse time or width is varied to keep thebattery voltage constant (PWM type).

• A control circuit limits the current in a series-linear manner to hold the battery

voltage at a high value.

Because the series regulator open-circuits rather than short-circuits the array as inshunt-controllers, no blocking diode is needed to prevent the battery from short-circuiting when the controller regulates.

Basic Series Regulator Design

PVArray

+

DCLoad

Load Switching Element

LV DControl

RegulationControl

Battery

Series Element




12-34

Series-Interrupting 1-Step Design

The most simple series regulator is the series-interrupting type, involving a one-stepcontrol, turning the array charging current either on or off. The charge regulatorconstantly monitors battery voltage, and disconnects or open-circuits the array inseries once the battery reaches the regulation voltage set point. After a pre-setperiod of time, or when battery voltage drops to the array reconnect voltage set

point, the array and battery are reconnected, and the cycle repeats. As the batterybecomes more fully charged, the time for the battery voltage to reach the regulationvoltage becomes shorter each cycle, so the amount of array current passed throughto the battery becomes less each time. In this way, full charge is approachedgradually in small steps or pulses, almost identical to the shunt-interrupting typecontroller discussed earlier. The principle difference between the two is whether theregulation is through shunting or series interrupting.

As mentioned with the shunt-interrupting type regulator, the 1-step series-interrupting type designs are best suited for use with flooded batteries and NOT forsealed VRLA types, due to the fact that gassing results from the relatively long “on

pulses”.

Series-Interrupting, 2-Step, Constant-Current Design

This type of controller is similar to the 1-step series-interrupting type, however whenthe voltage regulation set point is reached, instead of totally interrupting the arraycurrent, a limited constant current remains applied to the battery. This ‘tricklecharging’ continues either for a pre-set period of time, or until the voltage drops tothe array reconnect voltage due to load demand. Then full array current is onceagain allowed to flow, and the cycle repeats. Full charge is approached in acontinuous fashion, instead of smaller steps as described above for the on-off typeregulators. Some two-stage controls increase array current immediately as a loadpulls down battery voltage. Others keep the current at the small trickle charge leveluntil the battery voltage has been pulled down below some intermediate value(usually 12.5-12.8 volts) before they allow full array current to resume.

Series-Interrupting, 2-Step, Dual Set Point Design

This type of regulator operates similar to the series-interrupting type, however there

are two distinct voltage regulation set points. During the first charge cycle of theday, the controller uses a higher regulation voltage to provide some equalization

charge to the battery. This may be applied on a preset frequency (e.g. every 15-20

days) or may be programmed to occur after the battery has been discharged to a

low state of charge.




12-35

Once the array is disconnected from the battery at the higher regulation set point,

the voltage drops to the array reconnect voltage and the array is again connected to

the battery. However, on the second and subsequent cycles of the day, a lower

regulation voltage set point is used to limit battery overcharge and gassing.

This type of regulation strategy can be effective at maintaining high battery state of

charge while minimizing battery gassing and water loss for flooded lead-acid types.

The designer must make sure that the dual regulation set points are properly

adjusted for the battery type used. For example, typical set point values (at 25oC)

for this type of controller used with a flooded lead-antimony battery might be 15.0 to

15.3 volts for the higher regulation voltage, and between 14.2 and 14.4 volts for the

lower regulation voltage.

Series-Linear, Constant-Voltage Design

In a series-linear, constant-voltage regulator design the regulator maintains the

battery voltage at the voltage regulation set point. The series regulation element

acts like a variable resistor controlled by the regulator battery voltage sensing circuit

of the regulator. The series element dissipates the balance of the power that is not

used to charge the battery, and generally requires heat sinking. The current is

inherently controlled by the series element and the voltage drop across it.

Series-linear, constant-voltage regulators can be used on all types of batteries.

Because they apply power to the battery in a controlled manner, they are generally

more effective at fully charging batteries than on-off type regulators. These designs,

along with PWM types are recommended over on-off type regulators for sealed

VRLA type batteries, because they gradually approach, but do not exceed, thegassing voltage.

Series-Interrupting, Pulse Width Modulated (PWM) Design

This algorithm uses a semiconductor switching element between the array andbattery that is switched on/off at a high frequency with a variable duty cycle tomaintain the battery at or very close to the voltage regulation set point. Shunt-typePWM designs (with the switching element across the array and battery) were

discussed earlier. Power dissipation within the series-interrupting PWM regulator isconsiderably low, similar to the series-linear, constant-voltage algorithm.

As with all series type designs, there is no need for a blocking diode to preventbattery shorting, unlike the shunt-type PWM designs.




12-36

Exercise




12-37

Daily Operational Profiles forCharge Regulators

The following sections present typical daily operational profiles for a few of thedifferent types of battery charge regulators commonly used in small stand-alone PVsystems. These daily profiles show how the different charge controller algorithmsregulate the current and voltage from the PV array to protect the battery fromovercharge.

About the Charge Regulator DailyProfiles

The data presented in the graphs were measured during tests on operational PVlighting systems at the Florida Solar Energy Center (FSEC) in February 1993.Several identical systems were monitored, with the exception that each system useda different battery charge controller. The data presented here are for a selected‘clear day’ with no cloud cover, clearly showing the charge controller regulationeffects.

To properly understand the data presented in the graphs, it is helpful to know howthey were measured. The measured parameters included among others the solarirradiance (Sun), battery voltage (Vbat) and current (Ibat), and PV array voltage

(Vpv) and current (Ipv). The designations in parenthesis are used in the legend keyfor the daily profiles.

Each parameter was sampled every 10 seconds and averaged over a six-minuteperiod and recorded for a total of 240 data points daily. In addition, the minimumand maximum of the battery voltage samples were recorded every six minutes.These minimum and maximum voltages (based on 10 second samples) are key tounderstanding how a battery charge controller operates.

In each of the following figures showing charge controller daily performance, thereare two graphs. The top graph shows the battery and PV array voltage versus time

for the ‘clear day’. Note that for clarity, the battery voltage is plotted on the left y-axis, while the PV array voltage is plotted with respect to the right y-axis on adifferent scale. The bottom graph shows the battery and PV array currents over theday, as well as the solar irradiance. Note that the currents are plotted on the left y-axis, and the irradiance is plotted on the right y-axis.




12-38

The sizing of the battery, PV array and load profile in the test systems wasconfigured to typify commercially available PV lighting systems. The differentcharge controllers were selected from those commonly used in these type and sizesof systems. The following table lists the nominal specifications for the FSEC testsystems.

Nominal System Specifications

• Location: Florida Solar Energy Center (FSEC)

• Design Insolation: 5 kWh/m2-day

• PV Array: Nominal 100 watt Pmp, 6 amps Imp

• Battery: Flooded Lead Antimony, 12 volt, 100 Ah @ 20 hr rate

• Load: Nominal 3 amps, 8 hours nightly, 24 amp-hours per day

A final word of caution when examining the following daily operational profiles for thedifferent charge controllers. Since these were test systems designed to investigatenot only the behavior of the different controllers, but the effects the regulation setpoints had on maintaining battery state of charge, the set points were not alwaysoptimized for the specific system design. In some cases this was intentional while inother cases was the result of the controller operating characteristics. The main pointto emphasize however is that the daily profiles presented here show how the chargecontrollers typically operate in PV systems.




12-40

The hysteresis is an important specification for on-off controllers, and must beselected properly to achieve good array energy utilization and proper batteryrecharging.

Towards the end of the sunlight hours (1600-1700 hours) the PV array currentoutput reduces to a low enough value, in this case about 2.5 amps, whereinregulation is not required to limit the battery voltage below the regulation set point of

the controller. Once the sun sets (about 1800 hours) the battery voltage begins agradual decrease to it’s open-circuit voltage.

Notice how the open-circuit voltage at the end of the day is higher than in themorning, indicating that indeed the battery was charged and is now at a higher stateof charge. At about 2030 hours, the 3-amp load is reconnected and the batteryvoltage begins to steadily decrease in transition to the next day.




12-41

Shunt-Interrupting Charge Regulator

Clear Day Operational Profile

11

12

13

14

15

4 8 12 16 20 24

Time of Day (EST)

Battery Voltage (V)

0

5

10

15

20

25

PV Array Voltage (V)

Vbat, avg

Vbat, max

Vbat, min

Vpv, avg

-5

-2.5

0

2.5

5

7.5

10

4 8 12 16 20 24

Time of Day (EST)

Battery & PV Array Current (A)

0

200

400

600

800

1000

Irradiance (W/m2)

Ibat, avg

Sun

Ipv, avg




12-42

1-Step Series-Interrupting Regulator Daily Profile

A 24-hour daily profile for a small stand-alone PV lighting system operating with a 1-

step series-interrupting (on-off) type battery charge regulator is shown in Figure 12-10. The night and early morning activity is similar to the first system describedearlier.

Until about Noon-time (1200 hours) the PV array current and the battery voltageincrease steadily with increasing insolation as the battery is being recharged. Notethat during this period the battery charge controller is not regulating and the PV arraycurrent is approximately the same as the battery current. However, the minimumbattery voltage shows values slightly lower than the average and maximum batteryvoltages during the morning charging period. This is a particular characteristic of thecharge controller in this test system, by which the array is periodically disconnected

from the battery to sense nighttime conditions.

At about Noon (1200 hours) the battery voltage reaches the regulation voltage of thebattery charge controller (about 14.1 volts), and the controller begins to regulate thePV array current. When this occurs the battery current decreases to the jaggedcharacteristic of the 1-step interrupting (on-off) algorithm. The series characteristiccan be seen by the fact that once regulation begins, the average PV array currentdecreases (time averaging between battery charging current and zero), while theaverage PV array voltage approaches the array open-circuit voltage (time averagingbetween battery charging voltage and Voc). When this controller open-circuits thearray during regulation, the result is zero PV current while operating the array at theopen-circuit voltage point.

With the onset of regulation the minimum and maximum battery voltages aredistinguished from the 6-minute average voltage, and show the approximatecontroller set points. After regulation, the maximum battery voltage is about 14.1volts. This maximum battery voltage corresponds to the voltage regulation set pointfor the battery charge controller. The minimum battery voltage is between 13.2 and13.4 volts, corresponding to the voltage at which the charge controller reconnectsthe array to the battery to resume charging.

After sunset (~1800 hours) the battery voltage begins a gradual decrease to it’sopen-circuit voltage, which is again higher than it was in the morning, indicatingbattery recharging has occurred. At about 2030 hours, the 3-amp load isreconnected and the battery voltage begins to steadily decrease in transition to thenext day. In comparison with the shunt-interrupting controller discussed previously,the regulation set point for this 1-step series-interrupting controller was considerablylower, resulting in a lower battery state of charge. This is indicated by the lowerbattery voltage just prior to the load being disconnected in the early morning.




12-43

Series-Interrupting Charge Regulator

Clear Day Profile in PV Lighting System

11

12

13

14

15

4 8 12 16 20 24

Time of Day (EST)

Battery Voltage (V)

0

5

10

15

20

25PV Array Voltage (V)

Vbat, avg

Vbat, max

Vbat, min

Vpv, avg

-5

-2.5

0

2.5

5

7.5

10

4 8 12 16 20 24

Time of Day (EST)


0

200

400

600

800

1000

Irradiance (W/m2)

Ibat, avg

Sun

Ipv, avg




12-44

Modified Series Interrupting 2-StepConstant Current Charge RegulatorDaily Profile

A 24-hour daily profile for a small stand-alone PV lighting system operating with amodified series interrupting 2-step constant current type battery charge regulator isshown in Figure 12-11. The morning profiles are again similar to the previouspatterns.

At about Noon (1200 hours) the battery voltage reaches the regulation voltage setpoint for the battery charge controller (about 14.9 volts), and the controller begins toregulate the PV array current. In contrast to the series- and shunt-interruptingcontrollers discussed previously, the battery current is not entirely disconnected fromthe battery, but only limited to a lower value. When this occurs the battery currentdecreases to below 2 amps, and remains in a current-limited mode through theremainder of the day. The series characteristic is shown by the fact that onceregulation begins, the average PV array current also decreases while the averagePV array voltage approaches the open-circuit array voltage. This controllerregulates the array in a series-linear manner, by increasing the resistance betweenthe PV array and battery. The resistance is held at such a value that a limitedamount of current is allowed to flow from the PV array to battery after initialregulation.

With the onset of regulation, the minimum and maximum battery voltages areindistinguishable from the six-minute average voltage, indicating that the controller isNOT an on-off interrupting type design. After the initial battery regulation at 14.9volts, the voltage after regulation remains at about 14.1 volts through the remainderof the day.

Once the sun sets (~ 1800 hours), the battery voltage begins a gradual decrease toit’s open-circuit voltage. At about 2030 hours, the 3-amp load is again reconnectedand the battery voltage begins to steadily decrease as the battery is discharged.




12-45

Modified Series Charge Regulator


11

12

13

14

15

4 8 12 16 20 24

Time of Day (EST)

Battery Voltage (V)

0

5

10

15

20


Vbat, avg

Vbat, max

Vbat, min

Vpv, avg

-5

-2.5

0

2.5

5

7.5

10

4 8 12 16 20 24

Time of Day (EST)


0

200

400

600

800

1000

Irradiance (W/m2)

Ibat, avg

Sun

Ipv, avg




12-46

Daily Profile for Constant-VoltageSeries Charge Regulator

A 24-hour daily profile for a small stand-alone PV lighting system operating with a

constant-voltage series type battery charge controller is shown in Figure 12-12.Again the morning profiles are similar to previous cases.

At about Noon (1200 hours) the battery voltage reaches the regulation voltage setpoint for the battery charge controller (about 14.5 volts), and the controller begins toregulate the PV array current. When this occurs the battery current graduallydecreases to about 1 amp by the end of the day. The series characteristic of thiscontroller is shown by the fact that once regulation begins, the average PV arraycurrent also decreases while the average PV array voltage approaches the open-circuit array voltage. In principle this controller regulates the array in a series-linearmanner, by increasing the resistance between the PV array and battery through

semiconductor devices such as MOSFETs. The resistance is held at such a valuethat limits amount of current that is allowed to flow from the PV array to battery afterinitial regulation, while holding the array voltage at a constant value corresponding tohe controller’s regulation voltage.

With the onset of regulation the minimum and maximum battery voltages areindistinguishable from the six-minute average voltage, indicating that the controller innot an on-off interrupting type design. After the initial regulation at 14.5 volts, thevoltage after regulation remains at this level through the remainder of the day.

Moving toward sunset (about 1800 hours), the array current is no longer highenough to maintain the battery at the regulation voltage, and the battery voltagebegins a gradual decrease to it’s open-circuit voltage. At about 2030 hours, the 3amp load is again reconnected and the battery voltage begins to steadily decreaseuntil the next day when charging resumes.




12-47

Constant-Voltage Series Charge Regulator


11

12

13

14

15

4 8 12 16 20 24

Time of Day (EST)

Battery Voltage (V)

0

5

10

15

20


Vbat, avg

Vbat, min

Vbat, max

Vpv, avg

Vbat, avg

Vbat, max

Vbat, min

Vpv, avg

-5

-2.5

0

2.5

5

7.5

10

4 8 12 16 20 24

Time of Day (EST)


0

200

400

600

800

1000

Irradiance (W/m2)

Ibat, avg

Sun

Ipv, avg




12-48

PWM Series Charge RegulatorDaily Profile

A 24-hour daily profile for a small stand-alone PV lighting system operating with apulse-width-modulated (PWM) series type battery charge regulator is shown in

Figure 12-13.

At about Noon (1200 hours) the battery voltage reaches the regulation voltage setpoint for the battery charge controller (about 14.5 volts), and the controller begins toregulate the PV array current. When this occurs the battery current decreases in a

jagged manner, and remains in a current-limited mode through the remainder of theday. The series characteristic can be seen by the fact that once regulation begins,the average PV array current also decreases while the average PV array voltageapproaches the open-circuit array voltage.

In the PWM design an oscillating signal operating at a frequency of several hundred

Hertz is used to regulate the array current. When the controller is not regulating, thefull array current is applied to the battery. When the regulation voltage is reached,the current pulses are gradually reduced to hold the battery voltage at the regulationset point.

In effect, the PWM design operates similar to the constant-voltage controller, withthe exception that there is a small hysteresis between the minimum and maximumbattery voltage after regulation. The PWM is essentially a high switching speed on-off type or interrupting type controller that does not allow the battery voltage to dropsignificantly during regulation.

Once the sun sets (about 1800 hours), the battery voltage begins a gradualdecrease to it’s open-circuit voltage. At about 2030 hours, the 3-amp load isreconnected and the battery voltage begins to steadily decrease in transition to thenext day.




12-49

Pulse-Width-Modulated Series Charge Regulator


11

12

13

14

15

4 8 12 16 20 24

Time of Day (EST)

Battery Voltage (V)

0

5

10

15

20


Vbat, avg

Vbat, max

Vbat, min

Vpv, avg

-5

-2.5

0

2.5

5

7.5

10

4 8 12 16 20 24

Time of Day (EST)


0

200

400

600

800

1000

Irradiance (W/m2)

Ibat, avg

Sun

Ipv, avg




12-50

Exercises




12-51

Voltage Regulation Set PointSelection

In stand-alone PV systems the ways in which a battery is charged are generallymuch different from the charging methods battery manufacturers recommend. Abattery in a PV system must be fully recharged during the few daylight hours, amuch shorter time period than typically used for recharging standby batteries in UPSapplications (days or weeks) or industrial cycle applications (18-24 hours). For thisreason, the voltage regulation set point must be set higher than the usual “float”voltage given by battery manufacturers, to permit full utilization of the array current,but not to high as to excessively overcharge and gas the battery.

The optimal voltage regulation set point allows a maximum amount of Ah to bereplaced into a battery with minimum water loss and gassing. This is usually arrived

at empirically, and is different for different types of batteries and even differentmanufacturers of the same type of battery.

The set points are probably more important than the particular type of regulatordesign. A relatively simple charge regulator design with proper set points may workbetter than a sophisticated controller which is not set properly for the type of batteryor the environment.

Battery Gassing is Key To Voltage

Regulation Set Point SelectionThe onset of gassing in a lead-acid cell is determined by cell voltage, temperature,and charge rate.

• Under standard conditions, at a certain voltage (usually ~ 2.35 volts), thechemical potential is high enough to separate water into hydrogen and oxygen,causing water loss through gassing.

• At higher battery temperatures, the internal chemical reactions occur faster andthe corresponding gassing voltage decreases. This is the primary reason whytemperature compensation is used for the voltage regulation set point.

• At higher charge rates the internal voltage drop in the battery elevates thevoltage and brings on gassing sooner than would occur for a slower charge rate.




12-52

Some degree of gassing and overcharge is required for flooded lead-acid batteries,but can be harmful to sealed VRLA captive electrolyte type batteries for which lostelectrolyte can not be replaced. In general, sealed “maintenance free” valve-regulated batteries (using lead-calcium grids) require lower charge regulation voltageset points than flooded deep cycling batteries (using lead-antimony grids). Batterymanufacturers should be consulted to determine the gassing voltages for specific

designs.

The figure shows the relationships between cell voltage, state of charge, charge rateand temperature for a typical lead-acid cell with lead-antimony grids. At 27

oC and at

a charge rate of C/20, the gassing voltage of about 2.35 volts per cell is reached atabout 90% state of charge (point “A”). At a charge rate five times faster (C/5) at 27

o

C, the gassing voltage is reached at only 75% state of charge (point “B”). At a lowbattery temperature of 0

oC gassing does not occur until about 2.5 volt per cell, or 15

volts for a nominal 12-volt battery.

2.0

2.1

2.2

2.3

2.42.5

2.6

2.7

2.8

2.9

3.0

0 20 40 60 80 100

Lead-Acid Battery Charging Voltage as a

Function of State of Charge


C e l l V o l t a g e

( v o l t s )

Lead-Antimony GridsCharge Rate

C/20

C/5

C/2.5


Gassing Voltage at 0


“B”“A”




12-53

Suggested Voltage Regulation SetPoints

Some recommended ranges for charge regulation voltages at 25o

C for different

battery types used in PV systems are presented in the table below. These valuesare typical of voltage regulation set points for battery charge controllers used insmall PV systems. These recommendations are meant to be only general in nature,and specific battery and regulator manufacturers should be consulted for theirsuggested values.

Regulator Parameters Battery Parameters

RegulatorDesign Type

ChargeRegulationVoltage at

25o

C

FloodedLead-

Antimony

FloodedLead-

Calcium

Sealed,Valve

Regulated

Lead-Acid

FloodedPocket Plate

Nickel-

Cadmium

On-Off,Interrupting

Per 12 voltbattery

14.6 - 14.8 14.2 - 14.4 14.2 - 14.4 14.5 - 15.0

Per Cell 2.44 - 2.47 2.37 - 2.40 2.37 - 2.40 1.45 - 1.50

Constant-Voltage,

PWM, Linear

Per 12 voltbattery

14.4 - 14.6 14.0 - 14.2 14.0 - 14.2 14.5 - 15.0

Per Cell 2.40 - 2.44 2.33 - 2.37 2.33 - 2.37 1.45 - 1.50

The charge regulation voltage ranges presented in the table are much higher thanthe typical values presented in manufacturer’s literature. This is because batterymanufacturers often speak of regulation voltage in terms of the float voltage , forwhen batteries are float charged for extended periods (for example, in non-interruptible power supply (UPS) systems). In these and many other commercialbattery applications, batteries can be “trickle” or float charged for an extendedperiod, requiring a voltage low enough to limit gassing. Typical float voltages arebetween 13.5 and 13.8 volts for a nominal 12-volt battery, or between 2.25 and 2.30

volts for a single cell.

In a PV system however, the battery must be recharged within a limited time (usuallyduring sunlight hours), requiring that the regulation voltage be much higher than themanufacturer’s float voltage to ensure that the battery gets as much charge aspossible in a short period of time. If the charge regulation voltage in a typical PVsystem were set at the manufacturer’s recommended float voltage, the batterieswould never be fully charged.




12-54

Temperature Compensation

As discussed previously, the electrochemical reaction and gassing in a battery ishighly dependent on temperature. Lower battery temperatures slow down thereaction, reduce capacity and raise the voltage required for gassing. Conversely,

higher temperatures accelerate the reaction, increase grid corrosion, and lower thegassing voltage.

For these reasons temperature compensation (TC) of the VR set point isrecommended in PV systems where battery temperature might vary more than ±5

oC

from 25oC. Temperature compensation is also required for all type of sealed VRLA

captive electrolyte batteries where any gassing means permanent loss of electrolyteand life. By using TC a battery can be fully charged during cold weather, andprotected from overcharge during hot weather.

A widely accepted value of temperature compensation for lead-acid batteries is -5

mV/

o

C /cell. For a nominal 12 volt battery, this amounts to -30 mV /

o

C. Wherebattery temperatures vary by as much as 30o

C, temperature compensation mayresult in the regulation set point varying by as much as 1.0 volt in a 12 volt system.It is important to notice that the TC coefficient is negative, meaning that increases intemperature require a reduction in the charge regulation voltage. If batterytemperatures are lower than the design condition, the regulation voltage is increasedto allow the battery to reach a moderate gassing level and fully recharge.Conversely, the regulation set point is reduced if battery temperatures are greaterthan design conditions.

Battery temperatures may be sensed with an external probe connected to the

regulator, or approximated with an on-board sensor in regulator circuitry.

Typical Temperature Compensation Coefficient

-.005 volts/o C for one lead-acid cell

-.030 volts/o

C for a 12-volt lead-acid battery




12-56

Example: A 12-volt battery has a nominal final charge voltage of 14.5 volts atstandard temperature (25

oC).

If the battery operated in the winter at about 15oC., the final charge

voltage should not be 14.5 volts but should be increased to almost14.8 volts.

Voltage Change = factor X temperature change

= -.030 volts/ o

C X (15o

C - 25o

C)

= +.3 volts

Final voltage = 14.5 + .3 = 14.8 volts

And conversely, if the battery is in a hot climate, and the battery

temperature is 35o

C., the final charge voltage should be reduced toabout 14.2 volts.

With temperature compensation built into the regulator, batteries in cold climates willreceive the full charging they require, and batteries in hot climates will not beovercharged and gassed excessively. It is often considered an option, but should beincorporated into every system. It should always be included when sealed batteriesare used, as they are especially sensitive to temperature effects. The battery is tooimportant and expensive a component to not include such a feature.




12-57

Exercises




12-58

Oversizing Charge Regulators

The regulator must not only be able to handle typical or rated voltages and currents,but must also be sized to handle expected peak or surge conditions from the PVarray or required by the electrical loads that may be connected to the regulator. It is

extremely important that the regulator be adequately oversized for the intendedapplication. If an undersized regulator is used and fails during operation, the costsof service and replacement will be higher than what would have been spent on aregulator that was initially oversized for the application.

Typically, we would expect that a PV module or array produces no more than itsrated maximum power current at 1000 W/m

2 irradiance and 25

oC module

temperature. However, due to possible reflections from clouds, water or snow, thesunlight levels on the array may be “enhanced” up to 1.3 times the nominal 1000W/m

2 value used to rate PV module performance. The result is that peak array

current could be 1.3 times the nominal peak rated value if reflection conditions exist.

For this reason the peak array current ratings for charge regulators should be sizedfor at least 130% of the nominal peak maximum power current ratings for themodule or array.

The figure below shows the irradiance profile for a typical clear day and for a cloudyday with periods of enhanced irradiance due to cloud-reflectance. Note that duringthe cloudy day irradiance levels peaked above 1300 w/m

2!




12-59

The size of a regulator is determined by multiplying the peak rated current from anarray times this “enhancement” safety factor. The total current from an array isgiven by the number of modules or strings in parallel, multiplied by the modulecurrent. To be conservative, the short-circuit current (Isc) is generally used insteadof the maximum power current (Imp). In this way, shunt type regulators that operatethe array at short-circuit current conditions are covered safely.

Regulator Size (amps) = # Modules or Strings in Parallel X Isc X 1.3

Example: An array of 10 parallel 35-watt modules is needed for a 12-volt remotehome system.

Total array current = 10 modules X 2.15 amps Isc

= 21.5 amps array Isc

Regulator size = 21.5 amps X 1.3 safety factor

= 28 amps

In the example, the array would produce 21 amps of current, but the regulatorshould be sized to control up to 28 amps of current. A single regulator rated at 30

amps would be adequate, or two regulators rated at 15 or 20 amps could be used inparallel. If two controllers were chosen the array would be split into two sub-arraysof five modules each, and the output from the two regulators would be combined intoone battery. The two regulators could be set to slightly different final chargevoltages, and a two-step charging system would be the result.

Consult with regulator manufacturers to determine if they have already built a safetyfactor into their rating value. Oversizing the regulator by 130% may not benecessary if you are sure that the regulator design can handle any possible highcurrents for short periods.




12-60

Exercises




12-61

Operating Without a ChargeRegulator

In most cases a charge regulator is an essential requirement in stand-alone PVsystems. However there are special circumstances where a charge regulator maynot be needed in small systems with well-defined loads. Beacons and aids tonavigation are a popular PV application, which operate without charge regulation.By eliminating the need for the sensitive electronic charge regulator, the design issimplified, at lower cost and with improved reliability.

The system design requirements and conditions for operating without a chargeregulator must be well understood because the system is operating without anyovercharge and overdischarge protection for the batteries. There are two caseswhere battery charge regulation may not be required:

• when a low voltage “self-regulating module” is used in the proper climate

• when the battery is very large compared to the array.

Each of these cases is discussed next.




12-62

Using Low-Voltage “Self-Regulating”Modules

The use of “low-voltage” or “self-regulating” PV modules is one approach used to

operate without battery charge regulation. This does not mean that the moduleshave an electronic charge regulator built-in, but rather it refers to the low voltagedesign of the PV modules. When a low voltage module, battery and load areproperly configured, the design is called a “self-regulating system”.

Typical silicon power modules used to charge nominal 12-volt batteries usually have36 solar cells connected in series to produce and open-circuit voltage of greater than21 volts and a maximum power voltage of about 17 volts. Why do we generally usemodules with a maximum power voltage of 17 volts when we are only charging a 12-volt battery to maybe 14.5 volts? Because voltage drops in wiring, disconnectsovercurrent devices and controls, as well as higher array operating temperatures

tend to reduce the array voltage measured at the battery terminals in most systems.By using a standard 36 cell PV module we are assured of operating to the left of the“knee” on the array I-V curve, allowing the array to deliver it’s rated maximum powercurrent. Even when the array is operating at high temperature, the maximum powervoltage is still high enough to charge the battery. If the array were operated to theright of the I-V curve “knee”, the peak array current would be reduced, possiblyresulting in the system not being able to meet the load demands.

In the case of using “self-regulating” modules without battery charge regulation wewant to take advantage of the fact that the array current falls off sharply as thevoltage increases above the maximum power point. In a “self-regulating” low voltagePV module, there are generally only 28-30 silicon cells connected in series, resultingin an open-circuit voltage of about 18 volts and a maximum power voltage of about15 volts at 25

oC. Under typical operating temperatures, the "knee" of the IV curve

falls within the range of typical battery voltages. As a battery becomes chargedduring a typical day, its voltage rises and results in the array operating voltageincreasing towards the maximum power point or “knee” of the IV curve. In addition,the module temperature increases, resulting in a reduction of the maximum powervoltage. At some point, the battery voltage high enough that the operating point onthe IV curve is to the right of the “knee”. In this region of the IV curve, the currentreduces sharply with any further increases in voltage, effectively reducing the chargecurrent and overcharge to the battery.




12-63

Self-Regulation UsingLow-Voltage Module

0 2 4 6 8 10 12 14 16 18 20

Voltage (volts)

Current (amps)

2

1

0

M55

M65

Phoenix in July at 3:00 p.m.

The figure shows a comparison of operating points in the afternoon in a hot climatebetween a 36-cell (M55) module and a "self-regulating" 30-cell (M65) module. Asthe battery voltage rises there is a more dramatic reduction in current from the 30-cell module. In the afternoon, in this example, the battery voltage has risen to about14.4 volts, and the current from the 30-cell module is almost one third that from the36-cell module. The battery is reaching full charge, and the current should bereduced. The reduction in charging current that would be accomplished by a chargeregulator for the 36-cell module is performed automatically with the low-voltage 30-cell module.

“Self-regulating” modules are popularly used in the recreational vehicle (RV) market.A single 30-cell module can be safely matched to a single large 120-150 Ah batterywithout needing a regulator. The current from the module in the afternoon is lowenough so that the battery is not overcharging.




12-64

However, there are conditions that must be met, even with RV’s, for a “self-regulating” design to work properly. If more than one module were connected toonly one 120-150 Ah battery, or if the battery were too small (less than 120 Ah), thenovercharging could occur each day, and a small charge regulator would be neededin the system. Other factors must also be considered before using a “self regulatingmodule”. These are discussed below.

Special Considerations For Using Self-Regulating Module

Using a "self-regulating module" does not automatically assure that a photovoltaicpower system will be a self-regulating system. For self-regulation and no batteryovercharge to occur, the following three conditions must be met:

1. The load must be used daily. If not, then the module will continue toovercharge a fully charged battery. Every day the battery will receive excessivecharge, even if the module is forced to operate beyond the "knee" at current

levels lower than its Imp. If the load is used daily, then the amp-hours producedby the module are removed from the battery, and this energy can be safelyreplaced the next day without overcharging the battery. So for a system to be"self-regulating", the load must be consistent and predictable. This eliminatesapplications where only occasional load use occurs, such as vacation cabins orRV’s that are left unused for weeks or months. In these cases, a chargeregulator should be included in the system to protect the battery.

2. The climate cannot be too cold. If the module stays very cool, the "knee" ofthe IV curve will not move down in voltage enough, and the expected drop off incurrent will not occur, even if the battery voltage rises as expected. Often "self-

regulating modules" are used in arctic climates for lighting for remote cabins forexample, because they are the smallest and therefore least expensive of thepower modules, but they are combined with a charge regulator or voltagedropping diodes to prevent battery overcharge.

3. The climate cannot be too hot. If the module heats up too much, then the dropoff in current will be too extreme, and the battery may never be properlyrecharged. The battery will sulfate, and the loads will not be able to operate.

In summary, a “self-regulating system” design can simplify system design byeliminating the need for a charge regulator, however these type of designs are onlyappropriate for a very narrow range of applications and conditions. In most allstand-alone PV system designs, a charge regulator is required.




12-65

Using a Large Battery or Small Array

A charge regulator may not be needed if the charge rates delivered by the array tothe battery are small enough to prevent the battery voltage from exceeding thegassing voltage limit when the battery is fully charged and the full array current is

applied. In certain applications, a long autonomy period may be used, resulting in alarge amount of battery storage capacity. In these cases, the charge rates from thearray may be very low, and can be accepted by the battery at any time withoutovercharging. These situations are common in critical application requiring largebattery storage, such as telecommunications repeaters in alpine conditions orremote navigational aides. It might also be the case when a very small load andarray are combined with a large battery, as in remote telemetry systems.

In general a charging rate of C/100 or less is considered low enough to be toleratedfor long periods even when the battery is fully charged. This means that even duringthe peak of the day, the array is charging the battery bank at the 100-hour rate or

slower, equivalent to the typical trickle charge rate that a regulator would produceanyway.

To determine if a regulator is needed calculate the peak rate of charge expectedfrom the array and see if this rate is near or slower than the “100 hour” rate. Thenumber of modules in parallel is multiplied by the module peak current (Imp) to givethe array peak current. Dividing the battery capacity by this current gives the hoursto fully charge the battery.

Peak Charge Rate (hours) = Installed Battery Capacity (Ah)Number of Parallel Modules X Imp




12-67

The preceding discussion illustrates that the conditions and circumstancesthat allow photovoltaic systems to be designed without charge regulators arequite special. In most applications, a charge regulator is required.

Exercises




12-69

Selected References

Stand-Alone Photovoltaic Systems - A Handbook of Recommended Design Practices, Sandia NationalLaboratories, SAND87-7023, revised November 1991.

Naval Facilities Engineering Command, Maintenance and Operation of Photovoltaic Power Systems,NAVFAC MO-405.1, December 1989.

Exide Management and Technology Company, Handbook of Secondary Storage Batteries and ChargeRegulators in Photovoltaic Systems - Final Report, for Sandia National Laboratories,SAND81-7135, August 1981.

Bechtel National, Inc., Handbook for Battery Energy Storage in Photovoltaic Power Systems, Final Report,SAND80-7022, February 1980.

S. Harrington and J. Dunlop, "Battery Charge Controller Characteristics in Photovoltaic Systems",Proceedings of the 7th Annual Battery Conference on Advances and Applications, Long Beach,California, January 21, 1992.




12-70

(End of Chapter)




12-71

CHAPTER TWELVE

CHARGE REGULATORS AND SYSTEM CONTROLS 12-1

Purpose of Charge Regulators and System Controls 12-2Prevent Battery Overcharge 12-3Prevent Battery Overdischarge 12-4Provide Load Control Functions 12-5Provide Status Information To Users 12-5Interface and Control Backup Energy Sources 12-6Divert PV Energy To Auxiliary Load 12-6Serves As Wiring Center 12-7

Terminology 12-8

Nominal System Voltage 12-8Nominal Load and PV Array Current 12-8Charge Regulator Set Points 12-9

Other Functions Associated With Charge Regulation 12-16Backup Energy Source Control 12-16Equalizing Charge Capability 12-16Set Point Adjustability 12-16Load Voltage Regulation 12-17

Regulation/Control Element Design 12-17Operational Limits 12-18Surge Protection and Grounding 12-18Service Disconnects and Overcurrent Protection 12-19

Standard Configurations Of Charge Regulation andControl Systems 12-20

Simple Series Path Configuration 12-20Auxiliary Load Path Configuration 12-21Parallel Path Configuration 12-23

Sub-Array Switching Configuration 12-24

Electronic Designs for Charge Regulation 12-28Shunt Regulator Designs 12-29Series Regulator Designs 12-33




12-72

Daily Operational Profiles for Charge Regulators 12-37About the Charge Regulator Daily Profiles 12-37Shunt-Interrupting Charge Regulator 12-39Daily Profile 12-39

1-Step Series-Interrupting Regulator 12-42Daily Profile 12-42Modified Series Interrupting 2-Step Constant CurrentCharge Regulator Daily Profile 12-44Daily Profile for Constant-Voltage Series Charge Regulator 12-46PWM Series Charge Regulator Daily Profile 12-48

Voltage Regulation Set Point Selection 12-51Battery Gassing - Key to Voltage Regulation Set Point Selection 12-51Suggested Voltage Regulation Set Points 12-53

Temperature Compensation 12-54

Oversizing Charge Regulators 12-58

Operating Without a Charge Regulator 12-61Using Low-Voltage “Self-Regulating” Modules 12-62Using a Large Battery or Small Array 12-65

Selecting Charge Regulators 12-68



Siemens Solar Basic Photovoltaic Technology 12-1 Charge Regulators and System Controls

Chapter 12 – Answers Charge Regulators & System Controls

According to the data sheet, an SP75 module has a peak power of 75 Watts and an Iscof 4.8 amps (when configured at 12 Volts).

Since the array is 750 Watts total, there must be 750 ÷ 75 = 10 modules.

With a 12-volt system, the array will have all 10 modules in parallel. The maximumamount of current that the array can produce is 10 times the Isc of one module. Wecan therefore calculate the number of regulators required as

Number of Regulators = 10 parallel X 4.8 amps X 1.30 (Safety factor)

30 amps/regulator

= 2.08 regulators, rounded up to 3

The array of 1200 Watts is split into two sub-arrays. Therefore, each sub-array is 600Watts. Using 75-Watt modules, the number of modules in one array is

600 watts ÷ 75 watts/ module = 8 modules

At 12 volts, the array is configured as 8 modules in parallel. We use 4.8 amps as theIsc for the SP75 modules. Since the sub-array is supposed to have a single regulator itmust handle the full amount of current for the 8 modules (including the safety factor).

Size of Regulator = 8 parallel X 4.8 amps X 1.30 (Safety factor)

= 49.92 amps

This is greater than 30 amps, so we can not use a 30-amp regulator. However, we canuse a 50-amp regulator since the maximum expected current is slightly less than thisvalue. The correct answer is

d. 50 amps.





The shaded area shows indicating greater total charge being put into the battery overthe period of the day.

-5

-2.5

0

2.5

5

7.5

10

4 8 12 16 20 24

Time of Day (EST)


0

200

400

600

800

1000

Irradiance (W/m2)

Modifiedseries

Constantvoltage

According to the Voltage Regulation Setpoint Table, the proper ending voltage for asealed VRLA battery is 2.33 - 2.37 volts when using a PWM controller. We will use theaverage of this range, 2.35 volts, as our starting point. For a nominal 12-volt system,the ending voltage will be equal to 6 times this (since there are 6 two-volt cells).

6 X 2.35 volts = 14.1 volts.

So, under normal circumstances (25 °C), we would stop charging when the batteryreached 14.1 volts. However, this needs to be reduced because of the hightemperatures.

Voltage Change = factor X temperature change

= -0.030 volts/ °C X (40 °C - 25° C)

= -0.45 volts

Final voltage = 14.1 - 0.45 volts = 13.65 Volts





At 12 volts, all four modules will be in parallel. The maximum array current will be fourtimes the Isc of the module.

Total array current = 4 modules X 4.8 amps Isc

= 19.2 amps

Regulator size = 19.2 amps X 1.3 safety factor

= 24.96 amps

The necessary size is greater than the smallest regulator size available (20 amps). We

need to use a regulator that is greater than 24.96 amps, so the correct answer is

b. 25-amp regulator

The shunt regulator is rated for 25 amps. However, this needs to include the 1.30safety factor. The amount of array current that is allowed is:

Array Current = 25 amps

1.30

= 19.23 amps

Using an Isc of 2.4 for the SP36 module, the maximum number of modules is

Maximum Modules = 19.23 = 8.012.4

We round down and can use a maximum of 8 modules. The correct answer is:

b. 8 modules




According to the module literature, a 6-Watt module has a Imp of 0.39 amps. The peakcharge rate is determined as

Peak charge rate = Battery capacityArray peak amps

So,Battery capacity = Peak charge rate X Array peak amps

= 100 hours X 0.39 amps

= 39 Ah

Remember, if the battery size is greater than 39 Ah, the charge rate will be slowerthan a C/100 peak charge rate. The decision to use a charge regulator would dependon the available types of regulators, the pattern of load use and the environment. Since

the application is located in the desert (hot conditions) this might not be a good situationto use a self-regulating module.

The required battery capacity is given as 250 Ah battery capacity. The peak chargerate is:

Peak charge rate = 250 Ah2.5 amps

= 100 hours peak charge rate

This is just at the edge of the recommended peak charge rate for self-regulated system.A regulator is probably not necessary from a charging standpoint. However, thedecision whether or not to use a regulator requires good judgement of the advantagesand disadvantages, as well as attention to overall system needs.



Siemens Solar Basic PV Technology Course Components – TrackingCopyright © 1998 Siemens Solar Industries

13-1

Chapter ThirteenEnergy Enhancement

Through TrackingTechnology

There are two methods for enhancing the output of a photovoltaic power system thatare commonly employed by system designers. Both of these approaches improvethe match between the potential output of solar modules and the load. Both

methods involve the word “tracking”, although one is electronic and the other ismechanical. In this chapter we will discuss these methods for getting more out ofyou solar generators.

The first method involves physically tracking the sun as it moves throughout the dayand the seasons. The second method involves electronically monitoring themaximum power point of an array and transferring this maximum power to the loads.




13-2

Tilting The Array To Match TheLoad Profile

The most basic approach to “tracking the sun position” is to mount the array at afixed angle tilted up from horizontal so that the module surface faces the sun as itpasses through the sky each day. The mounting hardware can be ground mountedor have the modules mounted at the top of a pole.

The sun path is high in the summer and low in the winter. The best system designhas the array tilted to an angle so that the variation or profile of insolation throughoutthe year matches the load profile.

An example of the variation in insolation throughout a year is shown. The insolationon a flat surface varies greatly, with the least during the winter and the greatest

during the summer. If the load demand is small in winter and large in summer, aswith air conditioning loads or water pumping for irrigation, then tilting the array near aflat angle will give an insolation profile that best matches those load requirements.

If however, the load is relatively constant every month of the year, as might be thecase for a navigational aid or a constantly transmitting repeater, then a differentangle is better. Tilting the array up increases the insolation intercepted during thewinter months and sacrifices some during the summer months, with a resultingprofile that is more constant throughout the year. This more constant insolationprofile better matches the profile of a constant load.

Notice that as the tilt angle increases from horizontal, that a “double peak” profileemerges. The insolation profile never gets perfectly flat. The spring and fall havemore insolation than the winter or the summer. This double-peak profile is typical forlatitudes above the Tropic of Cancer or Capricorn. Nearer to the equator, the sun’spath actually passes overhead twice during the year. During the spring and fall, thepath will pass nearly straight overhead. During the winter and summer, the path willactually be further down from straight above.

If a tilt angle is chosen that does not match the insolation profile to the load profile,then the array will have to be quite large to produce enough output when insolationis low and will produce wasted energy when insolation is high. By choosing a tiltangle for the array that gives the best match, array size is minimized. Computermodels or repetitive hand calculations can predict the best angle to give a matchbetween load demand and insolation.




13-3

New Delhi Insolation Profiles

Latitude 29 deg. N

0

100

200

300

400

500

600

J a n

F e b

M a r

A p r

M a y

J u n

J u l

A u g

S e p

O c t

N o v

D e c

I n s o l a t i o n ( L a n g

l e y s )


Trivandrum Insolation Profiles

Latitude 8 deg. N

0

100

200

300

400

500

600

J a n

F e b

M a r

A p r

M a y

J u n

J u l

A u g

S e p

O c t

N o v

D e c

I n s o l a t i o n

( L a n g l e y s )





13-4

Exercise

Insolation Profile Data (Langleys) Trivandrum Latitude: 8 deg. N

-15 deg. 0 deg. 15 deg. 30 deg. 45 deg.

Jan 404 461 519 546 540Feb 452 494 533 540 516

Mar 495 515 528 509 464Apr 474 471 461 424 366May 454 436 414 367 305Jun 382 362 342 302 252Jul 388 371 352 313 262Aug 428 419 406 370 317Sep 433 441 444 421 378Oct 389 413 436 433 408Nov 359 399 442 458 447Dec 366 422 481 511 509

Insolation Profile Data (Langleys) Leh Latitude: 34 deg. N

0 deg. 15 deg. 30 deg. 45 deg. 60 deg. 75 deg.

Jan 207 268 316 343 349 334Feb 262 317 355 369 363 336Mar 332 375 394 388 364 323Apr 438 464 461 434 388 327May 476 484 462 419 361 293

Jun 503 502 471 420 356 284Jul 494 497 470 423 361 290Aug 457 476 465 431 380 314Sep 422 467 481 467 430 373Oct 340 404 445 456 442 404Nov 262 336 393 423 427 406Dec 209 278 335 369 381 369




13-5

Trivandrum Insolation Profiles

600

500

400

300

200

100

Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec

Leh Insolation Profiles

600

500

400

300

200

100

Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec




13-6

Sun Position Tracking

Although fixed angle mounting structures are simple and worry free, modules do notget good solar exposure in the morning and evening. By mounting the modules on atracking structure, gains in total daily output of power of 30% or greater can be

achieved because the modules are facing the sun directly during the morning andevening hours. Sun position trackers can move along one axis to capture most ofthe sun’s energy daily, or can also adjust along a second axis for the altitude of thesun and enhance seasonal performance.

The effect of sun position tracking is shown below for Leh at 34-deg. N latitude. Themodule is tilted to 50 degrees, to give the best output during the winter. The dailyenhancement is shown for the summer and for the winter, as well as theperformance over a typical year.

During the middle of the day, the single axis tracking gives the same output as the

fixed array, because they are at the same angle at that time. But in the morning andevening, notice that the single axis tracking gives more output. The array is tilted toface the morning or afternoon sun directly, while the fixed array is still facing dueSouth in the Northern Hemisphere (and due North in the Southern Hemisphere).The increased area under the curve for tracking indicates the increased energyproduced by the modules.

The winter performance in the same location is shown below. During the winter, thesun is making a low path with only a few hours of exposure. The sun is out in frontof the fixed array during most of the arc of the day, so tracking and fixed receiveabout the same amount of exposure. The enhancement during winter is lower than

in summer because there is less sun path to track!

Tracking adjustments by employing a second axis adds some output, depending onthe month and the latitude. The sun makes the same path in the sky during springand fall. It deviates up about 23.5 degrees in the summer and deviates down about23.5 degrees in the winter, for a total angle change of about 47 degrees summer towinter. Adjusting for this seasonal change adds mostly during the summer, whenthe path in the sky is large and there are many hours of exposure.




13-7

One SL-35 in Leh, June

0

0.5

1

1.5

2

5 7 9 11 13 15 17 19

Hour

A m p s

Fixed Single Axis Double A xis

One SL-35 in Leh, Janua ry

0

0.5

1

1.5

2

5 7 9 11 13 15 17 19

Hour

A m p s

Fixed Single A x is Double A x is

One SL-35 in Leh

02

468

101214

1618

Jan Mar May Jul Sep Nov

A h / d

a y

Fixed Single A xis Double Ax is




13-8

The enhancement of sun position tracking will be different depending on the latitude.The performance curves of a module in Minicoy are shown below. The latitude ofthe site is only 8 deg. north of the equator. The enhancement from tracking is moreuniform throughout the year, compared to the heavier gain only in the summer forLeh at 34-deg. latitude. Winter and summer daily enhancements are about thesame, because the sun’s path is always a large arc in the sky.

One SL-35 in Minicoy, June

0

0.5

1

1.5

2

5 7 9 11 13 15 17 19

Hour

A m p s

Fixed Single Axis Double Axis

One SL-35 in Minicoy

024

68

101214

1618


A h / d a y

Fixed Single Axis Double Axis




13-9

Exercise




13-10

Solar Powered (“Passive”) Sun Tracker

The typical commercial solar powered tracker design involves two tubes of Freon onthe east and west sides of the array. Each tube is partially shaded by a metal sheetor cover, as shown in Figure 13-7. As the sun moves, one tube becomes more

exposed to the sun than the other. The Freon expands and either pushes a pistonor transfers oil to the other side, which causes the structure to move to follow thesun.

Trackers can follow the sun along only one axis or can have dual axis tracking forcomplete seasonal compensation. Most large utility scale photovoltaic systemshave the modules mounted on trackers, to maximize module output and minimizeacreage costs.

At very high latitudes, the tilt angle of the structure is great, and the difference in

weight between the two sides becomes less. Therefore, it may be more difficult forsolar powered trackers to adequately perform at high latitudes.

Freon Tubes

Shadow Masks




13-11

Electrically Powered (“Active”) SunTracker

Electrically powered trackers use motors to move the structure holding the modules.

One design uses photocells at the base of a tall vertical block. As the sun moves,the shadow from the block moves across the photocells. Logic in the trackercontroller tells the motors to move the structure until the photocells “see” the sameamount of irradiance. This keeps the block, and the modules, pointed straight at thesun all day long. The energy needed to operate the tracker is drawn from themodules themselves, but is quite small, about 1/2 watt during daylight hours.

Two designs for active trackers are shown below. One design uses a “tilt-and-roll”design where linear motors push the array around the azimuth and tilt the array upor down to accommodate altitude tracking. A second approach uses a “vertical axis”tracking design where a gear drive mechanism rotates the array about a vertical axis

while a linear motor tilts the array for altitude adjustments. Both designs use thesame form of electronic control unit mounted on the upper edge of the array.




13-12

Trackers can be designed to hold from 4-24 modules or even more. Wind loading oflarge tracker structures should be a concern. But most designs can handle winds upto 40 mph or greater.

Because the tracking is driven by electronics, it can operate equally well in alllatitudes. There is no greater difficulty tracking at high latitudes as might occur withsolar powered trackers.

Exercise




13-13

Electronic Power Tracking

A second method for enhancing the output of a solar array into a load involves theuse of an electronic maximum power point tracking circuit (MPT) between themodules and the load. Such a device can be used for battery charging systems, but

has its greatest benefit when used with direct coupled water pumping systems.

Power Tracking With DC Motors

We have discussed how the load determines the operating voltage for an array. DCmotors (and batteries to some extent) will probably operate an array away from itsmaximum power voltage (Vmp) throughout a typical day. This means that themodule-load matching is not optimized and that system efficiency is not maximized.

A maximum power tracking device (MPT) operating between the array and load willforce the array to operate at its maximum power voltage at all times. The circuitrychops the DC from the array at high frequency, and rectifies it back to DC but at adifferent voltage and current, keeping the total POWER (current X voltage) thesame. It translates array power into power at the optimum voltage for the load. Sothe array and the load operate at different voltages and currents.

Pmax (array) = Power (load)

Imp (array) X Vmp (array) = I (load) X V (load)

An IV curve is shown with the point of maximum power indicated. If this power istranslated to a lower voltage, the current must rise to keep the value of powerconstant. Thus a MPT effectively changes the shape of an IV curve from having aflat region where current is almost constant into a descending geometric curve ofconstant power. Thus the current available at voltages less than the Vmp of themodule actually increases. Don’t think that energy or power has somehow beenmagically created. The total power available stays the same. It is just the current

and voltage values that have been changed. Their product, the power, stays thesame.

This substantially increases efficiency for direct coupled DC motors and waterpumps and can even improve battery charging efficiency under certaincircumstances. The effect for a DC motor is shown below. During the middle of theday, the module and motor should be matched in voltage by design. But in the early




13-14

morning and late afternoon, or during overcast conditions, the module wouldnormally supply very little current directly to the DC motor. The motor would operatethe module at point A, well away from its maximum possible power output at point C.By using the MPT, the motor now interacts with the power curve at point B,substantially increasing the current and the voltage. The water pump begins to pumpearlier in the day, and continues to pump later in the afternoon than would bepossible without the MPT.

M axim um Pow er Tracking

0 4 8 12 16 20 24

Voltage (volts)

Current (amps)

4

3.5

3

2.5

2

1.5

1

0.5

0

curve of constant power

maximum power point

of module (Vmp X Imp)

Op tim ized O peration W ithDC Motors

With MPT

Without MPT

A

C B

module motor

module motor

0 4 8 12 16 20 24Voltage (volts)

Current (amps)

4

3.5

3

2.5

2

1.5

1

0.5

0

motor curve

Phoenix, Arizona, 8:00 a.m.

A

B

C

M55 m odule curve




13-15

Power Tracking With Batteries

The enhancement of MPT for battery charging is less dramatic. This is becausemodules are designed to charge typical 12-volt batteries, and the number of solarcells in series has been chosen deliberately to match the voltage requirements for

charging batteries. In very cold conditions however, there may be a slight mismatchof the battery operating voltage and the module Vmp. The cold module temperaturewill keep the voltage potential of the IV curve out beyond the typical charging voltageof the battery.

An inherent efficiency of about 95% has been assigned to the MPT device. One ofthe issues of using MPT with batteries is that the enhancement must be greater thanthe internal efficiency of the device itself, or there will be no net gain at all.

An example of battery charging with MPT is shown below for a hot location, Raipur,during the coolest and hottest months. A 36-cell SL-35 module is used. Even in the

coolest month in Raipur, September, the gain from MPT is so small that the inherentinefficiency of the MPT actually penalizes the output into the battery, compared todirectly connecting the module to the battery. During the hottest month, May, theinput into the battery is even more reduced. The module output voltage potential isalready reduced due to high temperatures, and the MPT only subtracts itsinefficiency burden from the output.

The annual performance is shown below as well. In this hot climate, MPT does notincrease input into the battery at all during the year. The annual input without MPTis approximately 57.3 kwh/year from one module, while the input with MPT at 95%efficient is only 56.0 kwh/year.




13-16

One S L-35 in Raipur in Se ptem ber

0

5

10

15

20

25

30

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Hour

W

a t t s

MPT @ 95% ef f ic iency Battery

One SL-35 in Raipur in May

0

5

10

15

20

25

30

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Hour

W a t t s





13-17

One SL-35 in Raipur

0

50

100

150

200


W

h / d a y





13-18

Another example is shown, this time in Leh. The temperatures here will be lowerthan in the case of Raipur, and the difference in benefit of MPT with battery chargingwill show more clearly. The enhancement is positive throughout the year, higher inthe cool month of January and less in the hot month of August. But here again, therelative heat in all months makes the gain small. The annual total charge put into abattery directly is 50.3 kwh/year while the input through a MPT is 53.6 kwh/year.

One Sl-35 in Leh in January

0

5

10

15

20

25

5 7 9 11 13 15 17 19

Hour

W a t t s

MPT @95% eff iciency Battery

One Sl-35 in Leh in August

0

5

10

15

20

25

5 7 9 11 13 15 17 19

Hour

W a t t s

MPT @ 95% eff iciency Battery




13-19

The annual performance profile shows that there is a greater contribution from MPTin a cold climate throughout the year, again more so in the winter than in the warmersummer.

One SL-35 in Leh

0

50

100

150

200


W h / d a y

MPT @ 95% efficiency Battery

The negative trade-off that must be considered is the cost of the device, and the

increased risk of failure. Also the MPT must overcome its own energy inefficiency(usually about 5%) to prove economical.




13-20

Summary and Comparison ofPower Tracking and Sun PositionTracking

By way of summary, we should compare the benefits of the two types of trackingthat have been described in this chapter. Both sun position tracking and maximumpower tracking enhance the transfer of power from a solar array to a load. They doso in different ways, and are optimum in different environmental conditions.

Simply put, a sun position tracker maximizes the input of power into the system inthe first place, while an MPT device maximizes the output of the power to the load(but does not affect or increase the power input to the modules).

A sun tracker would be beneficial in a climate of clear sky and not extremely highwinds. If the sky is overcast often, then the tracker will not yield any benefit. Itwouldn’t matter where the module was facing

An MPT tracker would be of most benefit in the overcast situation. The moduleswould capture what solar energy they could, and the MPT would help transfer thatenergy on to the loads.

Both types of trackers can enhance output into batteries, although MPT matching isless dramatic than sun position tracking. But both systems would be of greatestbenefit when coupled with DC motors, as in water pumping. There are thousands of

working systems worldwide that use sun position tracking or MPT to enhance waterpumping.

Exercise




13-21

(End of Chapter)




13-22

CHAPTER THIRTEEN

ENERGY ENHANCEMENT THROUGH TRACKING

TECHNOLOGY 13-1

Tilting The Array To Match The Load Profile 13-2

Sun Position Tracking 13-6Solar Powered (“Passive”) Sun Tracker 13-10Electrically Powered (“Active”) Sun Tracker 13-11

Electronic Power Tracking 13-13Power Tracking With DC Motors 13-13

Power Tracking With Batteries 13-15

Summary and Comparison of Power Tracking andSun Position Tracking 13-20



Siemens Solar Basic Photovoltaic Technology 13-1 Energy Enhancement Through Tracking

Chapter 13 – Answers Energy Enhancement through Tracking Technology

Location: Trivadrum

Trivandrum

0

100

200300

400

500

600


-15

0

1530

45

The total annual insolation is

-15 Deg 0 Deg 15 Deg 30 Deg 45 Deg

Annual 152,740 158,169 162,810 157,789 144,693




Location: Leh

Leh

0

100

200

300

400

500600


0

15

30

45

60

75

The total annual insolation is

0 Deg 15 Deg 30 Deg 45 Deg 60 Deg 75 Deg

Annual 134,051 148,188 153,617 150,351 139,972 123,245

Location: Leh, Latitude 34 N

MonthFixed Axis

Ah/dayDouble Axis

Ah/dayPercentIncrease

Largest June 9 16 77%Smallest Jan. 8 10 25%

Location: Minicoy, Latitude 8 N

MonthFixed Axis

Ah/dayDouble Axis

Ah/dayPercentIncrease

Largest June 8 13 62%Smallest Jan. 13 17 31%

The differences in enhancement are due to the different site latitudes.



Siemens Solar Basic PV Technology Course Components– Diodes in PV SystemsCopyright © 1998 Siemens Solar Industries

14-1

Chapter FourteenDiodes in Photovoltaic

SystemsThis chapter discusses the function of a very specific piece of equipment used inphotovoltaic systems, the solid state “diode”. This device performs two differentfunctions, depending on where it is installed, and those functions are often confusedand misunderstood. The purpose of this chapter is to clarify the differences inperformance and to make clear why diodes are used in photovoltaic systems.

A diode is a solid state device made of P-type and N-type silicon. In fact a siliconsolar cell is a diode. Solar cells use the internal electrostatic field of their P/N junction to prevent electrons from flowing back into the cell after they have beenknocked loose by light. Diodes use their internal field to allow electric current to flowone way in a circuit and prevent it from flowing back.

When diodes are installed in series with a string of modules they perform a blockingfunction, preventing backflow down the module string. When diodes are installed inparallel with modules they perform a bypass function allowing current to pass arounda shaded area of a module. These two functions will be explained next.




14-2

Bypass Diodes For Shading

Shading Causes Mismatched Currents

When part of a photovoltaic module is shaded the shaded cells will not be able toproduce as much current as the unshaded cells. Since all the cells are connected inseries the same amount of current must flow through every cell. The unshaded cellswill force the shaded ones to pass more current than their new Isc. The only waythe shaded cells can operate at a current higher than their Isc is to operate in aregion of negative voltage, that is to cause a net voltage loss to the system. Thecurrent times this negative voltage gives the negative power produced by the shadedcells. In other words, the shaded cells will dissipate power as heat and cause "hotspots". And the shaded cells will drag down the overall IV Curve of the group ofcells.

Mismatched Cells in Series

unshaded cell 25% shaded cell series combination

+ =




14-3

Effect Depends On How Shaded

A module curve is shown with varying degrees of shading. Even with only one cellout of 36 shaded by 50%, there is a significant loss of power in the voltage range ofbattery charging. One cell completely shaded is even worse, but note that the

module is not "turned off" by one completely shaded cell.

For a module with three cells shaded the impact is of course worse still. But noticethat the effect of 25% shading on three cells is not as bad as 75% shading on onecell, the same total area of shading. Having the shading spread over many cells isnot as severe as having all the shading located in one or a few cells.

Effect of Shading on OutputPotential

One cell shaded

-10 0 10 20

3

2

1

25%

50%

75%

100%

Voltage

Current

-10 0 10 20

3

2

1

25%

50%

75%

100%

Voltage

Current

Three cells shaded

same amount of shading




14-4

How Bypass Diode Reduces Loss

An example of a module partially shaded will illustrate how bypass diodes can helpto reduce the looses associated with shading. Let’s look at two cases, one with amodule with three cells 50% shaded, and another with three cells 100 % shaded.

The exact amount of loss of power will depend on the amount of shading and onwhere the shaded module operates on its new reduced IV curve. Where a shadedmodule operates depends on the operating current of the other modulesconnected to it.

In the example in the figure, the shaded module is connected in series with manyother modules. The other unshaded modules try to operate at a their normal peakcurrent, creating a high current level. If the new reduced Isc of the shaded moduleremains above the operating current of the other series modules in the group, thenthe shaded module will be able to operate in the forward part of its curve (at point Ain the figure), and will still be able to contribute voltage to the series string. The

voltage out of the module or cell string will be higher than going into the module, asis the normal case.




14-5

Shaded Module In Series String

AB

Shaded module can still contribute voltageShaded module removesvoltage from system

5 10 15 20-5-10

3 cells 50% shaded

(-)

(+)

Current

Voltage

(+)

(-)

CurrentVoltage

3 cells 100% shaded

Level of current in array is

determined by other modules

Isc for unshaded cellsIsc for shaded

cells is below

the Isc for

unshaded cells.




14-7

Shaded

ModuleWithBypassDiode

-10 0 10 20

(+)

(-)

(-)

(+)

Current at

passes through

module

Current above

passes through diode

Severely shaded module is

limited to operating at only

-0.7 volts by the bypass

diode.

No current

passes

backwards

through bypassdiodeAll current

passes through

module and

on to loads.




14-8

Placement Of Bypass Diode In ModuleCell Circuit

The simplest way to include bypass diodes in a module cell circuit is to connect the

negative of the diode to the negative starting point of a module, and the positive ofthe diode to the positive end of the module. When there is severe shading, then thebypass diode will bypass the entire module circuit.

To preserve more of the module’s voltage, a diode can be placed around a portionof the cells in the circuit. Two methods can be used due to different junction boxconfigurations. For modules with three strings of cells, as shown below, with thecircuit beginning at one end of the module and ending at the opposite end, onebypass diode is installed in each of the two junction boxes at opposite ends of themodule. If the cell circuit is stretched out into a long string, you can see more clearlyhow each diode is connected around 24 cells, and how the two diodes overlap in cell

coverage.

In the case of severe shading, current can flow through a group of 12 cells and thenthrough the bypass diode. Thus only 24 cells or 2/3 of the module circuit isbypassed.

Dual Bypass Diode PlacementOverlapping Design




14-11

For this example, assume that three cells in one module become 75% shaded. Wefirst show the result without bypass diodes, and then on the next page show theeffect with bypass diodes in place.

The entire array output potential is dragged down. The battery as alwaysdetermines the operating point for the array, and now results in an array current ofabout 1.3 amps. The new Isc for the shaded module is only 1.1 amps, so the

shaded module is forced to operate at a negative voltage of -8 volts. The negativevoltage times the current gives the net power dissipated by the shaded module:

-8 volts X 1.3 amps = -10.4 watts

So about 10 watts of power is lost in shaded cells, in the form of heat. Rememberthat an M55 module produces only 53 watts peak at standard conditions, and about45 watts in field conditions when it is heated. So this potential power loss is quitehigh.

Three Shaded CellsReduces Entire Array Output

-20 0 20 40 60 80 100

3

2

1

Array current isdragged downby shaded cells

Shadedmoduleis forcedto operateat anegativevoltage

Voltage (volts)

Battery determines voltage

for entire array.




14-12

(For simplicity, we show the benefit of one single bypass diode around the entiremodule). With a bypass diode installed in parallel with the shaded module, theresult is quite different. There are two regions to examine – one below the Isc of theshaded module and the other above it.

Looking at current levels below the new Isc of the shaded module, everything is leftas is. The bypass diode can do nothing here.

Looking above the new Isc of the shaded module however, the shaded modulecannot operate beyond approximately -0.7 volts, as this is the maximum negativevoltage allowed by the diode. So the output potential of the array is not draggeddown as severely (-0.7 volts instead of -8 volts!). For current levels above the newIsc of the shaded module, the shaded module is completely bypassed, so thecomposite curve for the entire array is that of only four unshaded modules instead offive.

The battery voltage now operates the entire array at a current of almost 2.0 amps,nearly back to the full-unshaded potential (instead of only 1.3 amps as before with

no diode).

The shaded module operates at a net power loss of only:

-0.7 volt X 2.0 amps = -1.4 watts

Thus the total dissipated power in shaded cells and bypass diode is only about-1.4 watts instead of -10 watts as before. This can be broken down into the powerdropped in the shaded cells and the power loss through the diode. The modulepasses slightly more than its Isc of 1.1 amps at -0.7 volts so it is dissipating only asmall amount of power itself:

-0.7 volts X 1.1 amps = -0.77 watts

And the current going through the diode is just the difference between the shadedmodule current the array current, given as 2.0 amps - 1.1 amps = 0.9 amps. At avoltage of -0.7 volts across the diode, this means that the power lost in the diode isgiven by:

-0.7 volts X 0.9 amps = -0.63 watts

In summary, we have reduced the power lost from 10.4 watts to only 1.4 watts, with

0.77 watts dissipated in the cells and 0.63 watts dissipated across the diode. Thetwo main benefits of bypass diodes are quite evident: (1) array output is preserved;and (2) local heating is minimized.




14-13

Bypass Diode PreservesArray Output

-20 0 20 40 60 80 100

3

2

1

Voltagelossthroughshaded

module islimited to-0.7 volts

Above this point, array outputis missing one module, butis no longer dragged downby shaded cells.




14-14

Bypass Diodes In High Voltage Arrays

Recall from the discussion in the chapter on Wiring, there are two general ways towire systems of voltage higher than 12 volts. The modules can be wired in parallelfirst to create current, and then have the parallel wired groups connected in series to

get voltage. Or modules can be wired in series strings first to get voltage, and thenthe strings can be wired in parallel to get current. The recommended way is thelatter, to wire in series groups. The discussion below illustrates another reason whythis is the preferred method.

Parallel Wired Groups

If modules are wired in parallel groups there is a need for a large external diodearound each parallel group, as shown below.

Bypass Diodes ForParallel Wired Groups




14-15

If modules in a parallel group are shaded, then the group cannot produce the currentof the other groups in series with it. The unshaded modules in the group cancompensate by trying to produce more current. They do this by operating at lowervoltage, which moves their operating point up in current along their IV curve.

If the shading is bad enough, and the current compensation needs to be very large,then the entire group will continue to compensate by operating further and further

down in voltage. The group can actually go past zero volts and operate in the regionof negative voltage to try to get the current level up to that of the other unshadedgroups in series with it. If this happens, all the bypass diodes in the separatemodules will begin to pass current. Now the current is not just that of one module,but could be the current of the entire array. The small diodes for each module wouldbe overloaded and could fail.

The solution is to install an external diode large enough to handle the current of theentire array. If a shaded group compensates by going into negative voltage, thenthe current from the other groups bypasses the entire group through the large diode,and no damage occurs. This can involve extra wiring and cost.

Current Compensation ofParallel Wired Modules

+Voltage-Voltage

current level of other modules




14-16

Series Wired Groups

In the case of modules wired in series groups the bypass diodes installed in eachmodule will be sufficient. No extra diode protection is needed. This is because thecurrent that can flow through a bypass diode is just one module’s worth and nomore. The current flowing in a string is just the current of one module. If a module

becomes shaded, the current flows through the bypass diode and on to the othermodules in that string.

Bypass Diodes ForSeries Wired Strings




14-17

Blocking Diodes Prevent Leakage

Diodes placed in series with cells or modules can perform another function – that ofblocking reverse leakage current backward through the modules. There are twosituations where blocking diodes can help prevent this phenomenon.

• Blocking reverse flow of current from battery through module at night.

In battery charging systems the module potential drops to zero at night, and thebattery could discharge all night backward through the module. This would notbe harmful to the module, but would result in loss of precious energy from thebattery bank. Diodes placed in the circuit between the module and the batterycan block any nighttime leakage flow.

• Blocking reverse flow down damaged modules from parallel modules during day.

Blocking diodes placed at the head of separate series wired strings in highvoltage systems can perform yet another function during daylight conditions. Ifone string becomes severely shaded, or if there is a short circuit in one of themodules, the blocking diode prevents the other strings from loosing currentbackwards down the shaded or damaged string. The shaded or damaged stringis "isolated" from the others, and more current is sent on the load. In thisconfiguration, the blocking diodes are sometimes called "isolation diodes".

Each of these situations will be discussed next.




14-18

Preventing Nighttime Leakage

The IV curves presented on the next page show how the module is operated duringthe day and night by the battery. At night the IV curve passes through zero volts andamps, and slowly slopes downward from there. The battery maintains its voltage

even at night, so the module is operated at about 12 volts. Since the entire curve isbelow zero amps, the current passing through the module is "negative", or going thewrong way. The amount of current leaked from the battery depends on the shape ofthe IV curve at night. The fewer number of cells in series in the module, or thepoorer the curve shape, the greater the leakage current. Leakage currents for somemodules are only about 50 mA, so the power dissipated through the cells would beonly 12 volts X 0.050 amps = 0.6 watts. The power is spread evenly through all thecells, so the power dissipated by any one cell would be only about 0.6 watts / 36cells = 0.017 watts or 17 milliwatts.

Curve Shape DeterminesNighttime Leakage

(+) Current

(-) Current

Daytime flow

of current

(+) (+)

( - )( - )

(+) (+)

( - ) ( - )

Leakage flow ofcurrent at night




14-19

This can be prevented by installing a diode in series between the array and battery.The diode allows current to flow into the battery, but "blocks" reverse flow at night.So in this configuration, the diode is called a "blocking" diode. There is a smallpenalty to pay in the form of a 0.6-1.0 volt drop across the diode.

12 Volt System

+

-

PVModule Battery Load

Blocking Diode




14-21

Preventing Daytime Leakage

The preferred method of wiring higher voltage arrays is to first connect modules inseries strings and then connect them in parallel. Each series string should have itsown blocking diode. Do not think that if there are four strings, as shown, that the

current from any string must flow through all four diodes. Current flows from eachstring through only one diode and combines with the current from the other strings,so there is no increase in voltage penalty.

Blocking diodes placed at the head of each parallel string can perform anotherfunction during daylight operation. If one string becomes severely shaded, or if thereis a short circuit in one of the modules, the blocking diode prevents the other stringsfrom loosing current backwards down the shaded or damaged string. The shaded ordamaged string is "isolated" from the others, and more current is sent on the load.In this configuration, the blocking diodes are sometimes called "isolation diodes".

Series-Wired Strings




14-22

Typically in this configuration all the blocking or isolation diodes are groupedtogether in a field combiner box located at each structure. All the positive wires fromthe series strings are fed separately into the box, and the diodes are mounted onsome type of wiring block. The separate outputs from the diodes are combined intoa large diameter wire, and all the current from the strings flows on to the controlsand battery. The negative wires are combined into a large diameter wire also.

Series-Wired Strings WithCombiner Box

Field Combiner Box

Dual J-boxstyle modules

Single J-boxstyle modules

Other functions that can be performed at the field combiner box include a disconnectswitch or circuit breaker and lightning protection devices. These could beconveniently built in to the box and would allow further protection of the system fromshort circuits and lightning strikes.




14-23

An example of a commercial field combiner box is shown below. This unit ismanufactured by Solar Electric Specialties in the US and includes blocking diodes,separate pull fuses for each series string, a DC rated circuit breaker and a siliconoxide varistor (SOC) lightning protection unit. The negative block combines thenegative wires from the strings and a ground wire from the lightning protection unit.The block is then connected to a ground rod driven into the earth at the array. Ifthere is a lightning strike at the array, the dangerous current spike will be shunted to

ground at the array instead of passing through the wires all the way back to thesystem controls.




14-24

Current and Voltage Rating ofDiodes

To insure the diodes are not stressed, the current rating of blocking and bypassdiodes should be twice the Isc that might flow through them, and the reverse voltagerating should be twice the entire array Voc.

Current Rating of Diodes Double Isc

Voltage Rating of Diodes Double Voc

Most power modules produce current in the range of 3-4 amperes, so this meansthat the current rating of bypass diodes should be at least 6-8 amps, a relativelysmall value. Since the manufacturer does not know what the voltage of the finalarray will be, the voltage rating of bypass diodes should be the highest reasonablecommercial value, usually 600 volts.

In the case of a single blocking diode for a 12-volt parallel string, the current ratingmust be the Isc for the entire array. For example, if 8 35-watt modules wereconnected in parallel for a 12-volt system, the single blocking diode needed shouldhave a current rating of at least

2 X 8 parallel modules X 2.15 amps (Isc) = 34.4 amps.

The voltage rating of the blocking or isolation diodes should be twice the full arrayopen circuit voltage (Voc). Do not use just the nominal system voltage. For thesimple 12-volt array with 8 modules in parallel, the number of series modules is onlyone, so the voltage rating should be

2 X 1 series module X 20 volts (Voc) = 40 volts.

For a 48-volt nominal array, with 4 modules in series, the voltage should be at least

2 X 4 series modules X 20 volts (Voc) = 160 volts.




14-25

Exercise




14-26

(End of Chapter)




14-27

CHAPTER FOURTEEN

DIODES IN PHOTOVOLTAIC SYSTEMS 14-1

Bypass Diodes For Shading 14-2Shading Causes Mismatched Currents 14-2Effect Depends On How Shaded 14-3How Bypass Diode Reduces Loss 14-4Placement Of Bypass Diode In Module Cell Circuit 14-8An Example Of Bypass Diode Reducing Losses 14-10Bypass Diodes In High Voltage Arrays 14-14

Blocking Diodes Prevent Leakage 14-17

Preventing Nighttime Leakage 14-18Preventing Daytime Leakage 14-21

Current and Voltage Rating of Diodes 14-24



Siemens Solar Basic Photovoltaic Technology 14-1 Diodes

Chapter 14 – Answers Diodes in Photovoltaic Systems

First Array:

Bypass

Bypass

Bypass

Bypass

Blocking

A diode should be rated for twice the expected Isc and twice the full array Voc. Thebypass diodes are connected to 4 modules in parallel, so their rating should be:

Current Rating 8 X Isc Voltage Rating 8 X Voc

The blocking diode must handle all 4 parallel strings, so its rating should be:Current Rating 8 X Isc Voltage Rating 8 X Voc



Siemens Solar Basic Photovoltaic Technology 14-2 Diodes

Second Array:

Blocking

Bypass

Bypass

Bypass

Bypass

Blocking

The bypass diodes are now connected to a single module, so the rating should be:Current Rating 2 X Isc Voltage Rating 8 X Voc

Each blocking diode is connected to a single parallel string, so each will be rated at:Current Rating 2 X Isc Voltage Rating 8 X Voc



Siemens Solar Basic PV Technology Course 15-1 System Design – System SizingCopyright © 1998 Siemens Solar Industries

Chapter Fifteen

System SizingThe purpose of photovoltaic system sizing is to calculate the number of solarmodules and batteries needed to reliably operate the load throughout a typical year.This involves balancing the often-opposing goals of maximum reliability andminimum cost.

This section presents simple methods for calculating array and battery size forstandalone systems. While reading this chapter keep in mind that not all choices in

system sizing are based on calculations. There are decisions that require judgmenton the part of the designer and the user. The mechanics of the calculations arequite simple. It is the judgment of the designer about the efficiency andappropriateness of the loads that makes a system well designed and cost effective.




Basic Principles

• The solar array is sized to replace the load on a daily basis based on averageweather conditions. The average is made up of below average days and aboveaverage days so the array and battery must work together.

• The array is NOT sized based on how quickly it can recharge the array after afew days of below average weather. This would result in a large array most ofwhich is not needed or used during most of the year.

• The battery has the job of supplying energy to the loads when the insolation isbelow average. The array will replenish the battery during subsequent days ofabove average insolation.

If there are concerns about quickly recovering from a storm or prolonged period ofbelow average weather then the designer should be considering a hybrid systemdesign. The backup generator (usually a diesel, gas or propane generator) would beused to bring the batteries back to near full charge every few days during the winteror periods of prolonged bad weather. The generator might not even be neededduring the summer months.

The sizing calculations presented in this chapter are based on designing astandalone photovoltaic, non-hybrid and system. Hybrids are discussed in thechapter on Hybrid Systems




Size Array For Worst Season

When first thinking about array sizing for standalone systems one could begin bysizing so that the array output would be equal to the mathematical average of theload during the entire year. In that way, on the average over the year, the array

would put back into the system as much energy as the load used. But this meansthat for half the months of the year, every year, the battery would be partiallydischarged. During such long periods of sustained partial charge sulfation of thebattery plates would occur. In standalone systems there is no generator available“on demand” to recharge the batteries during below average weather. Batteryperformance and life would be very poor, and the overall cost of the system overtime would be very high.

The proper approach to array sizing is to calculate the array needed during the worstseason of the year. This means that one would expect the battery to be fullyrecharged even during the worst season and certainly during all the rest of the year.

This approach will reduce the sulfation that might occur on the battery plates andlead to long system operating life and low maintenance costs over time.

If the worst season of the year is very far below the average for the year then sizingfor the worst month will give an array that is quite oversized for the rest of the year,resulting in high initial costs. In that case it would make sense to consider a hybridsystem with a backup generator. But for small systems where a generator would betoo expensive, or in remote situations where the operation and maintenance of agenerator would be a burden, designing a standalone photovoltaic system will be themost cost-effective approach.




Battery Sizing

Days of Autonomy or Reserve

The battery bank is sized to operate the loads during a long sequence of belowaverage insolation days. We can think of the battery as being “full of charge”.During a below average day the array cannot supply all the amp-hours of chargeneeded to replace what the load draws from the battery. So the battery ends upbeing discharged at the end of the day. If the next day is again below average thenthe battery again discharges some to operate the loads. This process can go on foronly so long before the battery is discharged to a point that is may becomedamaged. The system designer must build into the battery capacity enoughequivalent days of charge to operate the loads autonomously, meaning without anyinput of energy from the solar array. We refer to these equivalent days of reserve as

“days of autonomy”.

The usual rough values that are used in most system sizing calculations are about 3-5 days for non-critical applications and 7-14 days or even more for more criticalapplications. Non-critical situations usually involve occupied systems, where theusers can adjust their load demand a little to accommodate the bad weather.Critical situations might involve commercial or governmental systems forcommunications or navigation, or important health needs such as hospitals andclinics. And very remote systems must have a large capacity in their battery bank toallow for the time it would take a maintenance crew to arrive at the site.

Rough Guide to Days of Reserve or Autonomy

Non-critical Applications 3-5 days Typical 4 days

Critical Applications 7-14 days Typical 10 days




Basic Battery Sizing Calculation

The basic formula for calculating battery size is presented below. It involvesmultiplying the number of days of reserve or autonomy times the amount neededdaily for the load. This gives the first approximation to the size of the battery

capacity.

But we cannot allow all of the battery capacity to be discharged during the “days ofautonomy”. Manufacturers recommend that only 80% of even deep cyclingbatteries, and only about 50% of shallow cycling batteries, be discharged. So wemust divide by the maximum percentage usable to give the amount of capacity toinstall.

Maximum Percentage Usable

Deep Cycling Battery Type Up to 80%

Shallow Cycling Battery Type Up to 50%

The basic version of the formula is given below for calculating the battery capacitythat must be installed.

Battery Capacity = Number of Days of Autonomy X Daily LoadMaximum % Usable

Each battery has a nominal voltage, depending on how many single cells areconnected by the manufacturer. Some large cells are only two volts; some smallerunits are six or twelve volts. The formula for the number of batteries to connect inseries to give the voltage for the loads is simply the nominal system voltage dividedby the nominal battery voltage.

Number of Series Batteries = Load Nominal VoltageBattery Nominal Voltage




To illustrate these basic equations let’s use the remote school system that was sizedon Page 11 in the Load Estimation chapter as an example.

Example: The AC load demand for the school summed to 9104 Wh/day. If we use theTrace 4024 sinewave inverter, and assume that the efficiency is about 90%and the input voltage is 24 volts, then the DC load will be 421 Ah/day (9104

Wh ÷ 0.9 ÷ 24 V = 421 Ah).

We shall consider this a non-critical system because the users can be a bitflexible in their usage depending on the weather, so we will design for 5 daysof autonomy. We will assume that deep cycling industrial batteries are used.The maximum % usable will therefore be 80%.

Battery Capacity = 5 days X 421 Ah/day = 21050.8 0.8

= 2631 Ah of capacity

If they were to use Trojan L16 deep cycling flooded batteries, rated at 350 Ahand 6 volts then they would need about eight in parallel and four in series:

(2631 Ah ÷ 350 Ah = 7.5, round up to 8) and (24 volts ÷ 6 volts = 4).

As another example, look again at the small remote cabin system sized on Page 10of the Load Estimation chapter. Assume that it will be used as only a weekendcabin and that cost is a problem so low cost batteries will be used. Recall that it is apure DC system with no AC loads.

Example: The remote weekend cabin load was calculated to be about 70 Ah/day at 12volts. If low cost starting batteries are to be used, then we can use only 50%of the capacity for days of reserve. If the cabin is to be used only for theweekends, then only 2 days of reserve need to be built into the batterycapacity.

Battery Capacity = 2 days X 70 Ah/day = 1400.5 0 .5

= 280 Ah of capacity

If they were to use Delco 2000 shallow cycling batteries, rated at 105 Ah at

12 volts each, they would need three batteries in parallel (280

÷105 = 2.7)




Exercise




Correcting For Cold Temperature

The basic formula presented must be modified to properly account for someperformance factors that affect battery capacity and depth of discharge. The firstfactor that must be taken into account is the fact that batteries loose capacity whenthey get cold. Recall from the discussion in the Battery Technology chapter that a

lead acid battery can deliver only a little more than 90% of its rated capacity at 0

o

Cand only about 80% at -20oC.

This reduction in available capacity with cold temperatures means that MOREcapacity must be installed than you would calculate at standard temperatures of25

oC. Install more than you need at 25

oC so that when the battery gets cold, it will

still have the capacity that you need.

We do this by dividing by the Capacity Correction Factor given in manufacturer’sliterature. A typical reduction curve is shown below. Estimate the averagedischarge rate using the method given on page 46 of the Battery Technology

chapter. The average rate of discharge for photovoltaic systems is given by theformula:

Average Rate of Discharge (hours) = # Days of Autonomy X Load Operating TimeMaximum % Usable

where Load Operating Time can be estimated as follows:



• Multiple Load Systems: Weighted Average Load Operating Time =load time

loads

×∑∑

30

40

50

60

70

80

90

100

110

120

-30 -20 -10 0 10 20 30 40 50

C/500 C/120

C/50 C/5

C/0.5

emperature an sc arge ate

Effects on Lead-Acid Battery Capacity

Battery Operating Temperature - oC P e r c e n t o f R a t e

d C a p a c i t y




Cold Temperature Data

Data indicating the lowest 24-hour average temperature are extracted from theSiemens Solar database for various sites and presented in the Appendix section. Itis assumed that the battery is a large thermal mass and will not quickly cool off tothe coldest temperature in the night, but rather will slowly move between the heat of

the day and the cold of the night.




Specifying Battery Capacity At ProperDischarge Rate

Another factor that needs to be considered when sizing batteries is that battery

capacity increases with slower discharge rates. Typically manufacturer’s list theircapacity at a standard rate of 8 or 10 hours of discharge. But in photovoltaicsystems, it is typical for systems to have discharge rates of 100-200 hours or slower,due to all the extra capacity built into the battery for autonomy.

Most manufacturers give the capacity of their batteries at different rates ofdischarge. You need to be able to specify the capacity needed for your systemat the proper discharge rate. Average rate of discharge was discussed in theBattery Technology chapter, and is presented again here.

The average rate of discharge for photovoltaic systems is given by the formula:

Average Rate of Discharge (hours) = # Days of Autonomy X Load Operating TimeMaximum % Usable

The average load operating time is either 24 hours for a continuous load, or theactual load operating time if a single load or the weighted average time if severalintermittent loads are present.

Load Operating Time:



• Multiple Load Systems: use weighted average load operating time

Weighted Average Load Operating Time =load time

loads

×∑∑

Some examples of industrial battery manufacturer’s literature are shown on the nextpage.




C&D Battery Company:Cell Type AH Capacity @ 25 deg. C. Overall Dimensions Weight

1.75vpc

to 1.9 vpc Length Width Height

8 h 72 h 120h 240h 480h in. mm in. mm in. mm lbs. kgs.

3DCPSA- 3 31 44 46 48 49 5.28 134 7.38 187 10.31 262 27.7 12.6

3DCPSA- 5 62 72 73 74 75 5.28 134 7.38 187 10.31 262 33.6 15.2

3DCPSA- 7 94 133 139 145 150 9.47 241 7.38 187 10.31 262 54.3 24.6

DCPSA-11 156 222 230 240 249 6.38 162 7.38 187 10.75 273 37.0 16.8

DCPSA-13 188 267 277 290 300 6.38 162 7.38 187 10.75 273 39.0 17.7

DCPSA-15 219 295 300 306 310 6.38 162 7.38 187 10.75 273 41.0 18.6

3KCPSA- 5 225 315 324 333 340 8.53 217 10.44 265 18.25 464 131.0 59.4

KCPSA- 7 337 384 389 394 400 3.62 92 10.44 265 18.25 464 58.0 26.3

KCPSA- 9 450 506 513 520 524 4.62 117 10.44 265 18.25 464 79.0 35.8

KCPSA-11 562 618 628 638 644 5.59 142 10.44 265 18.25 464 96.0 43.5

KCPSA-13 675 743 755 766 773 6.59 167 10.44 265 8.25 464 113.0 51.3

KCPSA-15 787 1011 1031 1047 1060 8.53 217 10.44 265 18.25 464 139.0 63.0

Varta Battery Inc.:Type Capacity (Ah) to 1.85 vpc final voltage

5 hour 10 hour 24 hour 48 hour 72 hour 120 hour 240 hour

Vb2306 270 300 360 408 432 450 468

Vb2308 360 400 480 544 576 600 624

Vb2310 450 500 600 680 720 750 780

Vb2312 540 600 720 816 864 900 936Vb2407 630 700 840 952 1008 1050 1092

Vb2408 720 800 960 1088 1152 1200 1248

Vb2409 810 900 1080 1224 1296 1350 1404

Vb2410 900 1000 1200 1360 1440 1500 1560

Vb2412 1080 1200 1440 1632 1728 1800 1872

Vb2414 1260 1400 1680 1904 2016 2100 2184

Vb2416 1440 1600 1920 2176 2304 2400 2496

Vb2418 1620 1800 2160 2448 2592 2700 2808

Vb2420 1800 2000 2400 2720 2880 3000 3120




Example: A remote telecom site will use deep cycling industrial batteries (maximumDOD of 80%). The system will have 10 days of autonomy designed into thebattery. Two loads are present:

Load 1: Microwave Repeater 10 amps 24 hours/dayLoad 2: Radio 7 amps 12 hours/day

Weighted AverageLoad Operating Time = 10 amps X 24 hours + 7 amps X 12 hours

10 amps + 7 amps

= 19 hours

Average Rate of Discharge = 10 days X 19 hours/day 0.8 max. discharge

= 238 hour rate

The nearest rate used by either manufacturer shown on the previous page isthe 240 hour rate, so use values from this column for this system. If yourcalculated average discharge rate is between rates used by a manufacturer,use the next faster rate (shorter time) to be conservative.

---------------------------Compare the capacity available at the 240-hour rate to the nominal rates ofthe manufacturers:

C&D Literature: use as an example the KCPSA-15 battery. Capacity at theslow 240 hour photovoltaic rate is 1047 Ah, and capacity at their “standard” 8

hour rate is 787 Ah. So a typical C&D battery will deliver about

1047 Ah @ 240 hr / 787 Ah @ 8 hr = 0.33 or 33% more capacity at the slowphotovoltaic rate compared to their standard rate.

Varta literature: use as an example the Vb2410 battery. Capacity at the slow240 hour photovoltaic rate is about 1560 Ah, and capacity at their “standard”10 hour rate is 1000 Ah. So a typical Varta battery will deliver 1560 Ah @240 hr / 1000 Ah @ 10 hr = 0.56 or 56% more capacity at the slowphotovoltaic rate compared to their standard rate.

Rule of Thumb: You will have about 30% more capacity available from a battery attypically slow photovoltaic discharge rates (usually C/100 to C/300) compared tomanufacturer’s standard or nominal rates (usually C/8 or C/10). This is only a rough“rule of thumb” to use when no other capacity vs. rate information is available.




Limiting Maximum Depth of Discharge ToPrevent Freezing

A final factor that must be included in battery sizing is the adjustment to the

maximum allowed depth of discharge to prevent freezing. As a battery dischargesthe electrolyte turns into water, and the freezing point rises toward the freezing pointof pure water at 0

oC. In cold climates, if the battery is allowed to discharge too

much, the electrolyte might freeze, damaging the battery.

Even if a deep cycling industrial battery is used in a system, the maximum depth ofdischarge might have to be limited to less than the usual 80%. The adjustment canbe read from a chart as given below.

Correct for Limited DischargeDue to Possible Freezing

-60 -40 -20 0

Lowest Battery Temperature (deg.C)

Maximum D.O.D. (%)

80

60

40

20

0

Using the lowest 24-hour average battery temperature read up to find the maximumallowed depth of discharge. This concern applies only for temperatures below minus8

oC. For example, if the battery would get to -20

oC, the maximum allowed depth of

discharge should only be about 53%, not the usual 80%.




Complete Battery Sizing Calculation

Putting all these correcting factors together, we develop a final formula forcalculating the capacity of a photovoltaic battery bank.

Battery Capacity = Number of Days of Autonomy X Daily Load(@ specified rate) Max % Usable X Temperature Derating Factor

(A) The Maximum Percent Usable is either the standard value of 50% for shallowcycling batteries, 80% for deep cycling batteries, or a reduced value based onfreezing concerns. The designer can also reduce this value simply as a way to buildin more life to the system. For example, a designer might use a shallow cycling

battery, but size based on using only 30% of the capacity and not the usual 50%.This might not impact the cost of the system too much, and would mean the batterylife would be extended.

(B) The Temperature Derating Factor is included to make sure that more capacity isinstalled at 25

oC so that when the battery gets cold and loses some capacity, there

will still be the required capacity present.

(C) The specified rate insures that you are taking into account the change ofcapacity with slow discharge rate. Use manufacturer’s literature to choose a batterythat will give you the capacity you need at the average discharge rate of your

system.

An Array/Battery Sizing Form is presented next. Use this form for your batterycalculations. All the factors are indicated. An example is worked for your reference.




Example: Calculate the capacity needed for the remote school example again, this timewith more information. The local meteorologist reports that it gets to an

average of -20o

C in the winter at the site. This will mean that wecompensate for both loss of capacity with cold temperature, and limit themaximum depth of discharge to prevent freezing.

(A) Using the freezing phenomenon chart in Figure 15-2, we determine thatthe maximum allowable depth of discharge is about 50% or 0.5 , even thoughwe are using “deep cycling” type batteries that usually can be discharged to80%.

(B) Calculate average rate of discharge. The weighted average loadoperating time for the loads on p. 9-11 is given by:

Weighted AverageLoad Operating Time = 9104 Wh/day

(8X40)+(2x11)+(2x200)+ 300+800+200 watts

= 4.5 hours

The number of days of autonomy was given as 5 and the maximumdischarge is limited to 50% due to freezing, so the average rate is given by

Avg. Rate of Discharge = 5 days X 4.5 hours/day = 45 hr rate0.5 max. discharge

(C) Use Figure 15-1 to determine the Temperature Derating Factor. Use the50-hour curve (closest to our 45-hour rate), and determine that theTemperature Derating Factor at -20

oC is 70% or 0.70.

(D) Calculate the capacity. We will use Varta batteries for this example.

Battery Capacity = 5 days X 421 Ah/day load0.5 X 0.70

= 6014 Ah @ 45 hour rate

Looking at the Varta capacity Table 15-2, use the 48 hour column (closest toour 45 hour rate) to select a particular model to use. We could choose 4 ofthe Vb2412 units, as each gives about 1/4 of the capacity we need at thatrate.

Since they are 2-volt cells, we would need strings of 12 cells connected inseries to give the 24 volts for the inverter.




Array/Battery Sizing Form

System Description: Remote School

Max. Daily Load = 421 Ah Number of Reserve Days = 5

System Voltage = 24 V Maximum Battery % Usage = 0.5

Average Discharge = Number of Days X Load Operating Time = 45Rate Maximum Battery % Usage (hours)

Coldest Avg Temperature. = -20oC Temperature Derating = 0.85

Battery Capacity = Number of Days X Max. Daily Load

(@ 45 hour rate) Maximum % Usage X Temp. Derating

= [ 5 ] X [ 421 ] = 6014 Ah[ 0.5 ] X [ 0.70 ]

Chosen Battery : Varta Vb 2412

Capacity = 1632 @ 48 hour rate Voltage = 2-volt cell

Parallel Batteries = Battery Capacity = [ 6014 ] = 3.7 round to4

Chosen Battery Cap. [ 1632 ]

Series Batteries = System Voltage = [ 24 ] = 12Chosen Battery V. [ 2 ]

-----------------------------------------------------------------------------------------------------------------Daily Load = ________ Module : _______ Module Output = ______

Parallel Modules = Daily Load .Module X Derating X ChargeOutput Factor Efficiency

= [ ] = __________ [ ] X [ ] X [ ]

Series Modules = System Voltage = [________] = __________ 12 [ ]




Exercise




Battery Cell Size

When you calculate the capacity of your battery bank, you will have to decide howmany real battery units you will put together in parallel to give you your capacity. Doyou choose just one huge cell, or a couple in parallel, or three, or ten?

For example, if you calculated that you needed 1000 Ah of capacity, would youchoose a single 1000 Ah cell, or two 500 Ah cells, or four 250 Ah cells?

In general, you want to keep the number of parallel connected batteries to aminimum. This will reduce the possibility of the batteries getting out of equalizationwith each other. Some parallel connected batteries could become more or lesscharged than others.

Rule of Thumb: As a general guideline, it is recommended that you connect no

more than four batteries in parallel. Use this as a general rule of thumb.

If you find that your choice of battery means you have to connect 20 batteries inparallel, for example, you should look at choosing a larger capacity unit so you needfewer in parallel!

Many system designers prefer to design their system with two parallel strings. In thisway, if there is a problem with a battery cell, part of the battery bank can be removedand worked on, while the system still has nominal voltage and can operate.Different designers will have different preferences.

Exercise




Array Sizing

In the previous discussion we showed how battery sizing was based on operatingthe loads during a period of below average insolation days, or days of autonomy.Now we look at how to size the array based on replacing the daily load on the

average every day.

Basic Array Sizing Calculation

The basic method to calculate the array size is to divide the average daily Ah load bythe number of Ah that one module will produce in a day. This will give the number ofmodules needed in parallel to produce the current for the load. The number ofmodules needed in series is given by dividing the nominal system voltage by thenominal voltage of one module.

Number Parallel Modules= Daily Load Demand (Ah)Module Daily Output (Ah)

Number Series Modules = Nominal System Voltage (volts)Nominal Module Voltage (volts)




Modifications To Array Sizing Formula

Derate Module Output By 10%

In the real world, modules loose some of their output due to environmental factor.

Dirt, dust and gradual aging can reduce the module output. The standard practice isto derate the calculated output of a module by 10% to account for theseunpredictable factors. Think of this as a normal engineering safety factor, requiredin the case of photovoltaic system design because we are relying on the weather.Variations are to be expected. So we build in a margin of safety to insure that thesystem works year after year.

Increase Load By 10% For Battery Coulombic Efficiency

During the charging process, flooded lead acid batteries will gas. This means that

some of the charge pushed through the battery by the modules is not turning intousable capacity but is escaping the system. Thus a little extra current must begenerated to overcome this loss. This is the coulombic efficiency of the battery.Usual practice allows for about 5-10 % loss due to gassing. So we must increasethe array size by about 10% over the load demand to account for this inefficiency inthe battery.




Complete Array Sizing Calculation

We must therefore modify the simple array sizing formula to account for thesefactors.

(A) Divide the Daily Load by the Battery Coulombic Efficiency. This effectivelyincreases the daily load and gives the true load that the array must replace.

(B) Multiply the Module Daily Output by the Derating Factor. This reduces theexpected output from a module due to environmental and aging losses, and gives amore conservative estimate of what can be expected from a module in the realworld.

Together these factors give the complete array sizing formula.

Number Parallel Modules = Daily Load (Ah)Coulombic X [ Module X Derating ] Efficiency Output Factor

Number Series Modules = Nominal System Voltage (volts)Nominal Module Voltage (volts)




Seasonal Variations Require Monthly Calculations

If the load profile is not constant it is wise to perform this calculation for the differentseasons or for each month. There is usually more output available in the summer,spring and fall than in the winter. But the load may be greater in the summer aswell. The expected module output for each month or season should be divided into

the load for that month or season to calculate the number of modules needed foreach month. The required size would be the largest number needed during theyear.

For example, you may calculate that you need nine modules in the winter, but onlyneed seven modules in the summer. You would have to install the larger number tomeet your load demand throughout the year.




Choose A Module With The Proper NumberOf Cells In Series

The daily output of a module will depend greatly on the number of cells in series. All

cells and modules loose voltage with increasing temperature. This is a naturalphenomenon well understood by manufacturers. Modules are designed to operatein different temperature climates for this reason.

36-Cell Modules For Hot Temperatures or MPT Applications

Manufactures usually produce a standard line of modules with 36 cells in series.Modules with 36 cells in series are designed to operate near their Imp even in thehottest of climates. The presence of 36 cells in series means that the Vmp atstandard temperature of 25

oC is about 17 volts, much more than needed to charge

a 12-volt battery. When these modules are placed in hot climates, they will looseabout two volts due to heat, and will have their Vmp at about 15 volts. This isenough to charge all types of batteries, even in the hottest of climates. Modules with36 cells therefore should be considered for all hot climates. They are also themodule of choice for systems with MPT, such as direct coupled water pumpingsystems and utility interactive systems. All the potential power of the 36 cells will beextracted by the MPT device, so no voltage is wasted.

33-Cell Modules For Moderate Temperatures

Solar module manufacturers may create modules with different number of cells in

series, so reduce module cost and to match different climates. The modules with 33cells in series are designed to operate in moderate climates. The presence of 33cells in series means that the Vmp at standard temperature of 25

oC is about 16

volts, slightly more than needed to charge a 12-volt battery. When the module heatsup to about 40-45

oC it will loose about one volt and will have its Vmp at about 15

volts. This is enough to successfully recharge any battery. But if the module is in avery hot climate, then it will loose more voltage. If it heats up to 50

oC or more, then

the voltage will drop to about 14 volts or less, and some current output reduction willoccur. This is not damaging to the module at all, but simply means that less thanoptimal current will be generated. Thus 33 cell modules are best suited formoderate climates and are not recommended for India.




Choosing A Module

• 36 cells: Vmp @ 25oC = 17.4 volts@ 50oC = 15.3 volts

– hot climates

– direct couple to DC motors

– utility interconnection with MPT

• 33 cells: Vmp @ 25oC = 15.9 volts@ 50oC = 14.0 volts

– moderate climates – general battery charging




“Peak Hour” Method Of Estimating ModuleDaily Output

You can estimate module daily output by using weather data. This method involves

converting the actual measured insolation on a tilted surface into the equivalentnumber of hours of “peak hours” of standard full sun irradiance at 1000 w/m2.

Multiplying the number of “peak hours” times the module “peak output” (Impmeasured at 1000 w/m

2) gives an estimate for the number of Ah/day from a module.

Module Output Is Peak Hours X Peak Power

If the insolation data on a tilted surface is in units of kwh/m2 then the conversion to

peak hours is easy. For example, if the average insolation for a month is 6.6kwh/m

2, this can be rewritten as 6.6 hours X 1 kw/m

2. But recall that 1 kwh/m

2 is just

1000 watts/m

2

, and is just the standard irradiance used by manufacturers to ratemodule output. It is as if the module was exposed to the standard peak irradiancelevel of 1000 w/m

2 for 6.6 hours. This of course did not happen, but it makes for a

simple calculation. The standard irradiance level of 1000 w/m2 is important because

module output at that condition is given in all manufacturer’s literature, so you cansimply use manufacturer's literature values. To calculate the Ah/day from a moduleusing peak hours, multiply the peak hours times the module Imp.

Estimating Module OutputUsing “Peak-Hour” Method

• Pretend that all real insolation was delivered at peakirradiance level of 1000 W/m2

ex: 6.6 kwh/m2 = 1kw/m2 X 6.6 hours

= peak X peakirradiance hours

• Module Daily Output = Imp X peak hours(for 35 watt module)

= 2.0 amps X 6.6 hours

= 13.2 Ah




For example, the average daily irradiance profile is shown for a surface tilted to 20o

for Madras. The total insolation falling on that surface in March is 6.6 kw/m2. This

can be rewritten as 6.6 hours X 1 kw/m2. In the figure we show the equivalent

irradiance profile with the sun suddenly “turning on” to the peak irradiance level of1000 w/m

2 and staying at that level for 6.6 hours. Module output would be given by

multiplying this value for equivalent peak hours times the module Imp. For a typical35 watt module with Imp of 2.0 amps, this would be 6.6 X 2 amps = 13.2 Ah/day.

For a typical 75-watt module with Imp of 4.4 amps, this would be 6.6 X 4.4 amps =29 Ah/day.

If we look at the average daily irradiance profile for Madras in July, we see that thetotal insolation that fell on a surface tilted to 20

o would be only 4.5 kw/m

2. This

would equate to 4.5 peak hours of irradiance.

Calculate array size based on the month with the lowest insolation for a load profilethat is constant throughout the year, or for each month and compare to the loadrequired for that month and choose the largest number of modules needed in anymonth.

Peak Hours

Madras, Surface Tilted to 20oS

0

200

400

600

800

1000

1200

4 6 8 10 12 14 16 18 20

Hour of Day

I r r a d i a n c e ( w / m 2 )

March July

4.5 kwh/m2

6.6 kwh/m2




If insolation data is reported in units other than kwh/m2, such as Langleys

(calorie/cm2), megajoules/m

2, or Btu/ft

2, then multiply by the conversion factor

presented below to change into kwh/m2 and therefore peak hours.

Conversion Factors For Peak Hours

Unit of Insolation Multiply by this factor to convert to peak hours...

kwh / m2 / day 1.0

Langley / day 0.01162

MJ / m2 / day 0.2777

Btu / ft2 / day 0.003155




Weaknesses in the Peak Hour Method

• There are some simplifications in the peak hour approach that should beunderstood. For example, the effect of temperature on module output isneglected in the peak hour approach. As a means to compensate the moduleImp is used in calculations. Since battery loads usually operate the module at

voltages slightly lower than the Vmp and therefore at currents slightly higher thanthe Imp, using Imp for module output is being conservative. The effect oftemperature on module output is greater for modules with fewer cells in series.So the peak hour method is more accurate for 36 cell modules, and lessaccurate for 33 cell modules, especially in hot climates. Predictions in coldclimates will be more accurate for all modules.

• In the peak hour method, the total measured solar insolation is translated intopeak hours of operation. Actually, at the beginning and ending of each day,there will be some time when the irradiance will be too low and the module orarray voltage will not be sufficient to charge a battery. This error is usually quite

small, but is more pronounced for modules with fewer cells in series. Predictionsof Ah/day/module will be more accurate with this method for 36 cell modules thanfor 33 cell modules.

• The peak hour method assumes that module output is completely linear withirradiance. It assumes that all modules will convert solar irradiance into electricalpower the same. But this is not the case. For example, high efficiency singlecrystal solar cells can convert at low light levels more efficiently than some othertechnologies. So this peak hour approach of multiplying hours times ratedcurrent can overestimate the output of certain technologies.

In general however, the peak hour approach is a useful method for quicklyapproximating module output given local measured insolation data on a tiltedsurface.




Exercise




LehLatitude: 34

oN

Best Tilt: 50o

S


4.0 4.3 4.4 4.9 4.7 4.6 4.7 4.8 5.3 5.3 5.0 4.4

MadrasLatitude: 13

oN

Best Tilt: 15oS


5.9 6.5 6.6 6.4 5.8 5.0 4.5 4.9 5.3 5.4 5.2 5.4




Two examples of how to use the data are presented next. The first example is theremote school, and the second is the remote cabin.

Example: The remote school 24-volt DC load demand was determined to be 421Ah/day. Use the insolation data given for Leh. We would use the 50

o tilt to

produce the flattest insolation profile for the year to match the expected

constant load for the school used throughout the year.

The lowest insolation during the year at 50o tilt angle is 4.0 kwh/m

2 during

January. This translates into 4.0 peak hours. Calculating the output from atypical 75 watt module:

Module Output = 4.0 peak hours X 4.4 amps Imp

= 17.6 Ah/day

To calculate the number of modules needed for this load, we assume a

conservative battery coulombic efficiency of 90%, and apply the usual 10%derating of module output:

Number of Parallel Modules = 421 Ah/day daily load0.9 X [ 17.6 Ah/day X 0.9 ]

= 29.5 , round up to 30 modules

Number of Series Modules = 24 volt system voltage12 volt module voltage (nominal)

= 2 in series

Total number of modules = 2 series X 30 parallel = 60 modules

The work for this example is worked out on the next page, using the lower portion ofthe Array/Battery Sizing Form.





System Description: Remote School (location: Leh worst month: Jan)

Max. Daily Load = _____ Number of Reserve Days = _____

System Voltage = _____ Maximum Battery % Usage = _____

Average Discharge = Number of Days X Load Operating Time = _____ Rate Maximum Battery % Usage (hours)

Coldest Avg Temperature. = _____ Temperature Derating = _____


(@ ___ hour rate) Maximum % Usage X Temp. Derating

= [ _____ ] X [ _____ ] = _____ Ah

[ ] X [ ]

Chosen Battery : ____________________________

Capacity = _____ @ _____ hour rate Voltage = _____

Parallel Batteries = Battery Capacity = [ _____ ] = _____ Chosen Battery Cap. [ ]

Series Batteries = System Voltage = [ _____ ] = _____ Chosen Battery V. [ ]

-----------------------------------------------------------------------------------------------------------------

“Week-Averaged” = 421 Module : 75 watt Module Output = 17.6 Ah/day Daily Load


= [ 421 ] = 29.5 -> 30

[ 17.6 ] X [ 0.9 ] X [0.9 ]

Series Modules = System Voltage = [24] = 2 12 12





System Description: Weekend Cabin (Madras worst month: July)




Coldest Avg. Temperature. = _____ Temperature Derating = _____



= [ _____ ] X [ _____ ] = _____ Ah[ ] X [ ]

Chosen Battery : _____



Series Batteries = System Voltage = [ _____ ] = _____

Chosen Battery V. [ ]-----------------------------------------------------------------------------------------------------------------

“Week-Averaged” = 19.8 Module : 35 watt Module Output = 9.0 Ah/day Daily Load


= [ 19.8 ] = 2.7 -> 3[9.0 ] X [ 0.9 ] X [0.9 ]

Series Modules = System Voltage = [_ 12 _] = 1 12 12




Siemens Solar Module Output Tables

Siemens Solar has developed a variety of computer sizing programs that assist indesigning photovoltaic power systems for a wide variety of applications. Theadvantage of using the output of a computer program is that all the important

variables, such as the effect of temperature on module performance, is taken intoaccount. Greater precision is possible when accounting for variables such asbattery efficiency and sun angle. The results of these computer calculations aremade available to you in the Appendix section to allow for more precise calculationsthan is possible with peak hour estimations.

The Siemens Solar Module Output Tables contain module output valuescalculated from these programs for over a hundred sites in the form ofAh/day/module for each month of a typical year. Use these values in the completearray sizing formula to calculate the number of modules needed.

Assumptions

For each location, the data includes the latitude and longitude, the Siemens Solar“map code” identifier number, and then twelve output values, one for each month.The values are calculated by the program PVSYSTEM, and are based on thefollowing assumptions:

(1) Constant load every month. This simplifies the automatic calculation processand is applicable to a wide variety of situations.

(2) Best tilt angle for constant load. The computer chooses the array tilt angle thatbest matches a constant load. This tilt angle is presented at the end of eachlocation’s data.

(3) Module operating voltage of 14.3 volts. This is based on an assumed averagebattery voltage of about 13.7 volts plus about .6 volts added for system losses

(4) Ground reflectance of 0.20 from “dry grass”.

Multiply SM55 Data By 1.44 To Create SP75 Data

There is a table of data for the 36-cell SM55 and for the 33-cell SM50-H. The valuesfor the SM55 can be used for the SP75 as well. Simply multiply the SM55 value by1.44 to account for the larger cell used in the SP75.




The first page of the North American Region data table is presented to illustrate theinformation presented. There are five regions of data: North America; Latin andSouth America; Europe; Asia; and Africa. The most appropriate regional data foryou has been included.

NORTH AMERICANREGIONModule = SM55

Ah/day/moduleLocation Map Lat Long : JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Best

Code : Tilt

:

UNITED STATES :

:

Alabama :

BIRMINGHAM 867 34 87 : 11.38 13.12 14.14 14.75 13.98 13.39 13.14 14.20 14.64 15.48 13.41 11.34 55

MOBILE 868 31 88 : 12.73 14.37 14.82 14.63 13.55 12.63 12.10 13.09 13.98 15.86 14.14 12.36 55

MONTGOMERY 224 32 86 : 11.84 13.49 14.39 15.00 14.02 13.49 13.13 14.14 14.50 15.82 14.01 12.12 55

Alaska :

ADAK 900 52 177 : 5.49 8.03 9.48 10.03 9.26 8.54 8.39 8.27 8.74 8.70 6.96 4.90 70

ANNETTE 901 55 132 : 4.67 7.46 10.07 11.49 11.55 10.33 10.67 10.41 9.88 7.10 5.21 3.34 75

BARROW 902 71 157 : 0.00 2.24 11.21 15.56 12.09 15.13 15.22 10.36 6.47 2.79 0.03 0.00 55

BETHEL 136 61 162 : 3.45 8.09 12.27 13.00 11.67 10.80 9.61 8.28 9.20 7.50 4.10 1.62 80

BETTLES 904 67 152 : 0.02 5.56 12.58 15.49 15.30 14.47 12.89 11.22 10.62 6.34 1.28 0.00 80

BIG DELTA 905 64 146 : 1.61 7.40 13.34 14.65 14.32 13.35 12.84 12.46 11.46 7.56 3.29 0.00 80

FAIRBANKS 158 65 148 : 0.79 6.79 12.88 14.22 13.89 13.26 12.37 11.30 10.61 6.79 2.66 0.00 80

GULKANA 907 62 145 : 2.67 7.78 13.41 14.83 13.37 12.82 12.47 12.25 11.31 8.78 3.83 0.72 80

HOMER 908 60 152 : 4.17 8.07 12.21 13.26 12.54 12.26 11.85 10.96 10.41 8.99 5.51 2.18 80

JUNEAU 909 58 135 : 3.24 5.89 9.04 11.02 10.55 10.42 9.85 9.10 7.93 5.53 3.72 1.63 75

KING SALMON 910 59 157 : 4.95 9.04 12.60 12.45 11.51 10.62 10.04 9.27 9.91 9.59 6.33 3.46 80

KODIAK 911 58 152 : 4.57 7.96 12.24 12.88 11.20 11.21 10.83 10.95 10.30 9.75 5.93 3.26 75

KOTZEBUE 912 67 163 : 0.01 5.25 12.14 15.32 15.58 15.23 13.35 11.34 10.36 6.53 0.71 0.00 75

MATANUSKA 185 62 149 : 3.62 9.12 16.20 14.72 13.17 12.30 11.50 11.05 9.92 7.75 4.76 1.75 80

MCGRATH 913 63 156 : 2.04 7.15 12.31 13.52 12.41 11.67 10.68 9.76 9.72 6.77 3.41 0.33 80

NOME 914 65 165 : 0.62 6.66 11.58 13.97 13.52 13.22 11.21 9.73 9.77 7.11 1.99 0.00 80

SUMMIT 915 63 149 : 2.02 7.07 12.62 14.34 13.81 12.08 11.01 10.11 9.98 7.86 3.72 0.20 80

YAKUTAT 916 60 140 : 2.94 5.74 9.66 11.35 10.50 10.03 9.39 8.81 8.10 6.46 3.56 1.37 75

Arizona :

PHOENIX 197 33 112 : 17.62 19.61 20.31 20.48 19.07 17.59 16.73 18.01 19.81 20.32 18.76 17.05 55

PRESCOTT 871 35 112 : 17.95 19.60 20.84 21.31 20.72 19.95 17.58 18.21 20.72 20.84 19.14 17.31 50

TUCSON 220 32 111 : 18.38 20.12 21.00 21.20 20.05 18.72 17.02 18.14 19.98 20.53 19.14 17.60 50

WINSLOW 873 35 111 : 17.46 19.62 21.00 21.47 20.51 19.64 17.81 18.60 20.53 20.55 18.93 16.76 50

YUMA 874 33 115 : 18.56 20.43 21.69 21.59 20.37 19.16 17.43 18.81 20.35 20.74 19.42 17.92 50

Arkansas :

FORT SMITH 875 35 94 : 12.76 14.02 14.34 13.87 13.73 13.67 14.00 14.87 14.98 15.62 13.93 12.45 60

LITTLE ROCK 180 35 92 : 12.28 13.88 14.20 13.71 13.72 13.64 13.71 14.64 15.02 15.83 13.60 12.04 60

California :

ALPINE 692 33 117 : 13.55 17.49 15.57 16.90 14.40 16.62 17.05 17.65 19.18 19.56 19.03 16.76 55

ARCATA 881 41 124 : 10.07 12.19 13.11 14.06 13.40 13.03 12.59 12.93 14.12 13.07 10.66 9.55 65

ARROWHEAD 693 34 117 : 13.08 16.73 16.96 18.50 18.38 18.69 18.64 19.69 21.39 21.18 19.76 17.06 55

BAKERSFIELD 882 35 119 : 13.28 15.79 17.86 18.03 17.46 17.09 17.26 18.64 19.84 19.34 15.77 12.40 60

BLYTHE 695 34 115 : 17.02 19.47 20.59 20.40 19.01 18.19 17.20 18.69 19.70 20.45 19.71 17.83 50

BUTLER VALLEY RANCH 1115 41 124 : 9.59 8.90 10.91 13.02 12.72 13.51 14.98 15.89 16.43 11.57 9.17 8.71 65

CARLSBAD 696 33 117 : 13.34 17.29 17.90 16.85 14.55 14.59 14.86 15.59 16.58 15.67 17.84 15.92 55

CARRISA PLAIN 1223 35 120 : 16.38 16.23 15.93 16.45 16.89 15.95 16.17 17.29 17.58 18.96 14.60 14.04 60

CHULA VISTA 697 33 117 : 14.65 18.11 17.63 17.36 14.86 14.56 15.37 16.02 17.66 17.43 18.00 16.02 55

DAGGETT 884 35 117 : 17.08 18.53 19.78 19.28 17.72 16.88 16.53 18.04 19.60 19.86 18.27 16.63 60

DAVIS 152 39 122 : 11.62 13.20 16.69 17.51 17.18 16.82 17.50 18.64 19.56 17.98 14.57 11.18 60




Please find the Data Tables appropriate for your region in the Appendix section.




Two examples of how to use the data are presented next. The first example is theremote school, and the second is the remote cabin.




Example: The remote school DC load demand was determined to be 421Ah/day. Assume the location is near Yuma, Arizona. Looking at thedata for Yuma, we find that the lowest output during the year for anSM55 module is 17.43 Ah/day during July. Assume a conservativebattery coulombic efficiency of 90%. Apply the usual 10% derating ofmodule output.

The number of SM55 modules needed for this applications is given by

Number of Parallel Modules = 421 Ah/day daily load.9 X [ 17.43 Ah/day X .9 ]

= 29.8round up to30 modules in parallel

Number of Series Modules = 24 volt system voltage12 volt module voltage (nom.)

= 2 in series


The work for this example is worked out on the next page, using the lower portion ofthe Array/Battery Sizing Form.





System Description Remote School







= [ _____ ] X [ _____ ] = _____ Ah

[ ] X [ ]

Chosen Battery : _____________




-----------------------------------------------------------------------------------------------------------------“Week-Averaged” = 421 Module : SM55 Module Output = 17.43 Ah/day Daily Load


= [ 421 ] = 29.8 -> 30[17.43 ] X [ 0.9 ] X [0.9 ]

Series Modules = System Voltage = [_ 24 _] = 2 12 [ 12 ]




Example: The weekend cabin system was determined to have a daily loaddemand of 69.7 Ah/day. But if the system is really only going to be forweekend use, then the array can be sized based on the “WeekAveraged Load Demand” . The array can be sized to replace the 2-day weekend demand over the entire 7-day week. Recalculating the

69.4 Ah/day daily load spread over the week: 69.4 Ah/day X 2 days / 7 days = 19.8 Ah/day “week averaged”. Now we size the array toreplace this amount on a daily basis.

Assume the location is in the mountains near Arrowhead, California.Looking at the data for Arrowhead, we find that the lowest output for anSM55 module is 13.08 Ah/day during January. We have decided touse SP75 modules in this application. So the output from a SP75would be adjusted from this value for an SM55 by multiplying by thefactor of 1.44.

SP75 output in January = SM55 output X 1.44= 13.08 Ah/dayX 1.44= 18.84 Ah/day

Assume the same conservative battery coulombic efficiency of 90%,and the usual 10% module output derating factor.

The number of SP75 modules needed for this applications is given by:

Number of Parallel Modules = 19.8 Ah/day “avg” daily load.9 X [ 18.84 Ah/day X .9 ]

= 1.3 modulesround up to2 modules in parallel

Number of Series Modules = 12 volt system voltage12 volt module voltage (nom)

= 1 in series


The work for this example is again worked out on the next page, using the lowerportion of the Array/Battery Sizing Form.





System Description: Weekend Cabin







= [ _____ ] X [ _____ ] = _____ Ah

[ ] X [ ]

Chosen Battery : _____________




-----------------------------------------------------------------------------------------------------------------“Week-Averaged” = 19.8 Module : SP75 Module Output = 18.84 Ah/day

Daily Load


= [ 19.8 ] = 1.3 -> 2[ 18.84 ] X [ 0.9 ] X [0.9 ]

Series Modules = System Voltage = [ 12 ] = 1 12 [12 ]




Exercise




Checking Battery Sizing

Checking Depth of Discharge

There are a couple of checks that you can perform on your array and batterycalculations to learn more about how the system might perform. The first is tocalculate the average daily depth of discharge for the installed battery. This checkcan be interesting because you may have chosen to install a deep cycling battery,but if the number of days of autonomy is large, the actual daily depth of dischargecan be quite shallow.

Daily Depth ofDischarge = Daily Load (Ah)

Total Installed Capacity

= Daily Load (Ah)Number Parallel Batteries X Battery Capacity

Example: The remote school problem worked on pages 16 and 32 has beendesigned with 3175 Ah of installed capacity. The daily load was 421Ah at 24 volts. A check on the actual average depth of dischargeshows:

Daily DOD = 421 Ah daily load3175 Ah battery capacity

= .13 or about 13% daily




Checking Maximum Rate of Charge

Another check on your calculations is the maximum rate of charge of the batterybank by the installed array. The battery should not be charged too fast during peakirradiance periods or damage could occur. Battery manufacturers will tell you themaximum rate of charge for specific models of batteries. If no specifics are

available, refer to this chart of typical maximum allowable charge rates for differenttypes of batteries.

Typical Maximum Charge Rates

Pure lead grids, stationary battery C/25Lead calcium grids, shallow cycling battery C/15Lead antimony grids, deep cycling battery C/5

The formula for checking the maximum charge rate of your installed system is givenbelow. You simply divide the total peak current that can come from the installedparallel modules into the total installed battery capacity.

Maximum RateOf Charge = Total Installed Capacity (Ah)

Total Array Peak Current (Amps)

= Number Parallel Batteries X Battery Capacity

Number Parallel Modules X Imp

Example: The remote school problem worked on pages 16 and 32 has been designedwith 30 75-watt modules in parallel by 2 modules in series for 24 voltoperation. There will be 3175 Ah of installed capacity. A check on themaximum rate of charge shows:

Max. Rate = 3175 Ah battery capacity30 parallel modules X 4.4 amps Imp (75 watt module)

= 24 hours

This is slower than the typical maximum of five hours for deep cyclingbatteries, so the array is not too big for the battery.




Calculating Tilted Insolation

The insolation data used must be incident on a tilted surface. Raw weather data isusually measured on a flat (horizontal) surface, and it will not represent the amountof insolation available when the modules are mounted on a tilted surface. Data must

either be measured on a tilted surface, or some mathematical work must be done tothe raw horizontal data to predict insolation on a tilted surface.

General Concepts in Translating From Horizontal to TiltedInsolation

For many regions of the world, the only available solar radiation data is in the form ofmonthly averages for daily global horizontal insolation. Photovoltaic systemdesigners require insolation data for tilted surfaces to properly size systems. For thisreason, methods are needed to translate actual measured horizontal data to

reasonably accurate values for tilted surfaces. Presented next is a number ofempirical and geometric relationships required for this estimation.

The total solar insolation or radiation intercepted on a tilted surface is made up ofthree contributors: (1) direct beam radiation coming from the actual disk of the sun;(2) diffuse whole-sky radiation coming from the entire canopy of the sky; and (3)reflected radiation bouncing from the ground or nearby surfaces.

Total Insolation on a Tilted Surface = Direct (beam) Insolation

+ Diffuse (whole-sky) Insolation+ Reflected Insolation

When measurements of total (global) insolation on a horizontal surface arerecorded, and then averaged, these components are lost, and all that remains is atotal “bucket” of insolation with no information about how much was direct and howmuch was diffuse.

We need to separate these two components from the total, because they contribute

to the total insolation on a TILTED surface in different ways. The diffuse componentis just a function of how much of the total canopy of the sky the tilted surface “sees”.The direct component needs to be added up as the angle of the sun changes duringthe day, and is a function of the exact path of the sun on that day at that latitude.The reflected component just depends on the tilt angle of the surface as itaccumulates the same fraction of both the direct and diffuse.




Global Horizontal Insolation

Horizontal

terrestrialsurface

Sun

Directbeamradiation

Diffuse whole-sky radiation

Tilted Insolation

Tiltedterrestrialsurface

Sun

Directbeam

Diffuse whole-sky radiation

Reflected

radiation




Summary of Steps

Summary of steps to calculate insolation on a fixed tilted surface using monthlyaverage insolation data measured on a horizontal surface:

Step Procedure Variable

1 Establish the desired tilt angle of the modules and thelatitude of the site.

s, L

2 Read from literature the measured average terrestrialinsolation on a horizontal surface for each month.

H

3 Determine the declination and sunset hour angle for arepresentative day for each month. These are geometricvariables that depend only on day of year and latitude.

δ, hs

4 Determine the extraterrestrial horizontal insolation. Thisis the radiation outside the atmosphere, a variable duethe non-circular orbit of the Earth around the Sun.

H o

5 Calculate the clearness index, the ratio of actualterrestrial to extraterrestrial insolation. This indicateshow much of the theoretical radiation actually got throughto the surface.

K t

6 Determine the ratio of the diffuse to total insolation. Thiscomes from a model of the sky, and separates thediffuse or indirect radiation from the total. (The remainderwould be the beam or direct radiation coming from thedisk of the sun).

H diffuse

7 Determine the sunset angle for a tilted surface (ifdifferent from the angle for a horizontal surface).

hs

'

8 Determine the beam radiation tilt factor. This is ageometric factor that adjusts for the surface areapresented at a tilt angle compared to horizontal.

Rb

9 Select the ground cover and the associated groundreflectivity factor.

ρ g

10 Calculate the final total insolation on a tilted surface byadding up the contribution for direct, diffuse, andreflected insolation.

H tilted




Equations for Conversion from Horizontal to Tilted Insolation

First, the declination, δ, must be calculated for a representative day for each monthof the year:

Where n = julian day number for the representative day in each month.

The following table gives the recommended average day for each month of theyear:

Date Julian Day number

January 17 17February 16 47

March 16 75April 15 105May 15 135June 11 162July 17 198August 16 228September 15 258October 15 288November 14 318December 10 344

Source: S.A. Klien, “Calculation of Monthly Average Insolation on Tilted Surfaces”,Solar Energy, vol. 19, p.325, Permagon Press Ltd., 1977.

Next, the sunset hour angle, h s , is determined for the representative day for eachmonth:

( )δ = +

2345

360

365 284. cos n Eqn. 2-1, p.12 (Hsieh)

( )h Ls = −−cos tan tan

1 δ Eqn. 2-10, p.15 (Hsieh)

Where L = site latitude.




The monthly average daily solar insolation on an extraterrestrial horizontal surface,

H o , is then calculated. The over bar designates a monthly average value.

H H n

L h h

Lo sc ss= +

⋅ +

241 0 033

360

365

2

360π δ

π δ . cos cos cos sin sin sin

Eqn. 3-38, p.44 (Hsieh)

Where H sc = solar constant = 1353 W/m 2

Next, the monthly average clearness index , K T , is defined by the ratio of themonthly average daily insolation on a terrestrial horizontal surface to the radiation onan extraterrestrial horizontal surface:

K H

H

T

o

= Eqn. 3-42, p.52 (Hsieh)

Then, the diffuse to total radiation ratio , H H diffuse , is given by Collares-Pereira &

Rabl:

( )[ ] ( ) H

H h h K

diffuse

s s T = + − − + − −0 775 0 00653 90 0 505 0 00455 90 115 103. . ( ) . . cos

Eqn. 3-51, p.61 (Hsieh )

The sunset hour angle , h ' s , on tilted surfaces facing the equator is determined by:

( )[ ]{ }′ = − −−h h L ss smin , cos tan tan

1 δ Eqn. 3-55, p.62 (Hsieh)

Where s = surface tilt angle.




The monthly average beam radiation tilt factor , R B , is in general a complicated

function of atmospheric transmittance. However, Liu and Jordan have estimated R B

as the ratio of extraterrestrial radiation on the tilted surface to that on a horizontal

surface for the month. For surfaces facing the equator, R B is estimated by:

( ) R

L s h h L s

L h h L

B

s s

s s

=− ′ + ′ −

+

cos cos sin sin( ) sin

cos cos sin sin sin

δ π δ

δ π

δ

180

180

Eqn. 3-53, p.61 (Hsieh)

The monthly mean total radiation tilt factor , R , is calculated by:

R H

H

H

H R

H

H

s stilted diffuse B

diffuse

g= = −

+

+

+

−

1

1

2

1

2

cos cosρ Eqn. 3-52, p.61 (Hsieh )

Where ρ g = ground reflectivity, generally use 0.2 as default value. A table of ground

reflectivity factors is presented below for various common surfaces:

Ground Condition Ground Reflectivity factor

ρ g

Earth roads .04

Water surfaces from above .07

Coniferous forest (winter) .07Weathered blacktop .10

Gravel, bituminous roof .13

Soils (clay, loam) .14

Dry grass .20

Crushed rock .20

Weathered concrete .22

Green grass .26

Forest in autumn, ripe field .26

Dark building surfaces .27

Dead leaves .30

Light building surfaces .60Snow .75




Finally, the monthly average daily total solar radiation , H t , on tilted surfaces isapproximated by:

This final equation can be re-written in another form to more clearly show the threecomponents adding up to give the total insolation:

We can rewrite H H diffuse− as the direct component H direct . Then the

equation can look like the following:

Total Insolation on tilted surface = direct + diffuse component + reflected

H H R H s

Hr s

tilted direct b diffuse g= + +

+

−

1

2

1

2

cos cos

H R H tilted = ⋅ Eqn. 3-52, p.61 (Hsieh )




Tracking Surfaces

For tracking surfaces, the monthly average insolation must be estimated using long-term hourly radiation data. First, a more comprehensive form of the equation for the

monthly average beam radiation tilt factor, R B , must be used to account for the daily

variation in surface tilt (s) and azimuth angles (ψ ) for the tracking surface. An

equation for R B for surfaces of any orientation has been developed by Klien:

[ ] R A B C D E L h h L B s s= − + + + ÷ +

2

180cos cos sin sin sinδ

π δ Eqn. 3-56, p.62 (Hsieh)

Where,

( ) ( ) A s L h hs r = ′ − ′cos sin sinδ π

180

( ) ( ) B L s h hs r = ′ − ′sin cos sin cosδ ϕ π

180

( )( )C L s h hs r = ′ − ′cos cos cos sin sinδ

( )( ) D L s h hs r = ′ − ′cos cos sin sin sin sinδ ϕ

( )( ) E s h hs r = ′ − ′cos sin sin cos cosδ ϕ

′hr = sunrise hour angle on tilted surface, deg.

Charts used for rapid graphical solutions of R B are presented in Hsieh.




The hourly average radiation on a tilted surface can be approximated from monthlyaverage daily insolation by the following analytical expression suggested byCollares-Pereira and Rabl:

( ) ( )r a b h

h h

h h h

s

s s s= +

−

−

π

π 24 2 360cos

cos cos

sin cos Eqn. 3-44, p.58 (Hsieh)

Where r = ratio of hourly total radiation to the daily total radiation

h = hour angle in degrees at mid point of hour

a hs= + −0 409 0 5016 60. . sin( )

( )b hs= − −0 6609 0 4767 60. . sin

The hourly totals for insolation on the tracking surface must then be integrated overthe entire day, over the range of surface tilt and azimuth angles to determine thedaily total radiation on tilted surfaces.

In practice, insolation enhancement factors are generally applied to the insolation onequator facing latitude tilt surfaces to approximate the insolation on trackingsurfaces. For single-axis tracking, insolation enhancements of between 15 and 20percent are typical. For two-axis trackers, enhancement values of between 25 and30 percent are common. In general, passive trackers and areas with higher diffuseto total radiation ratios will have lower insolation enhancements due to tracking.

References

Hsieh, J.S., Solar Energy Engineering, Prentice-Hall, 1986.




(End of Chapter)




CHAPTER FIFTEEN

System Sizing 15-1

Basic Principles 15-2

Size Array For Worst Season 15-3

Battery Sizing 15-4Days of Autonomy or Reserve 15-4Basic Battery Sizing Calculation 15-5Correcting For Cold Temperature 15-8Cold Temperature Data 15-10

Specifying Battery Capacity At Proper Discharge Rate 15-11Limiting Maximum Depth of Discharge To Prevent Freezing 15-14Complete Battery Sizing Calculation 15-15Battery Cell Size 15-19

Array Sizing 15-20Basic Array Sizing Calculation 15-20Modifications To Array Sizing Formula 15-21Complete Array Sizing Calculation 15-22Choose A Module With The Proper Number Of Cells In Series 15-24“Peak Hour” Method Of Estimating Module Daily Output 15-26Siemens Solar Module Output Tables 15-36Checking Battery Sizing 15-45Calculating Tilted Insolation 15-47



Siemens Solar Basic Photovoltaic Technology 15-1 System Sizing

Chapter 15 – Answers System Sizing

A photovoltaic system needs to be sized for the worst case to insure that it will operatereliably year round. If the system were only sized to meet the average requirements,then the system would be undersized for half the time. This could result in a lack ofsufficient energy to support the load during the periods of poor weather. Even if thesystem could support the load in these conditions, the battery would remain at a lowstate of charge for long periods of time. This would damage the battery and reduce itslifetime. Sizing for the worst case results in a PV system that will operate reliablythroughout the year and extend the lifetime of the battery.

The load demand calculated in Problem 9-1 is 230 Ah/day. Since low cost batteries willbe used, the maximum % useable capacity will be 50%.

Battery capacity = 5 days X 230 Ah/day0.50

= 2,300 Ah

Since the system is 12 volts, we will need a 12-volt battery of this capacity or sufficientsmaller voltage cells in series.

The load demand calculated in Problem 9-3 is 146.8 Ah/day. Using deep cycleindustrial batteries, we have 80% useable capacity.

Battery capacity = 14 days X 146.8 Ah/day0.80

= 2,569 Ah

The overall battery bank voltage must be configured at 24 volts. This could be two 12-volt batteries in series, four 6-volt batteries, or twelve 2-volt cells.




Lead-acid cells should not be completely discharged, even for high quality, deep cyclebatteries. Discharging lead-acid batteries beyond the 80% DOD limit can damage theplates, cause the plates to break apart, or reverse the cell polarity. Any of these couldsignificantly reduce the battery lifetime.

The only capacity information we have for the batteries is the "rated capacity" of 80 Ah.We do not know how this varies with rate. To get an assumption of the capacity at areasonable discharge rate, we will use the "rule of thumb" indicated in the text. We willmultiply the rated capacity by 130%. So the capacity will be:

80 Ah X 130% = 80 X 1.30 = 104 Ah

The load demand is 30 Watts / 12 volts = 2.5 amps. The total daily Ah load is 2.5 amps

X 12 hours = 30 Ah/day. The battery autonomy is specified to be 5 days. Low costbatteries will result in a maximum % useable capacity of 50%. The load operating timeis 12.

The average rate of discharge is

Avg Rate of Discharge = 5 days X 12 hours50% useable capacity

= 120 hour rate

The coldest 24-hour temperature is 0 °C, which results in a temperature derating of92% (using the C/120 curve from Figure 15-1). There is no limit on the depth ofdischarge due to freezing concerns, so the maximum % useable capacity is 50%.

Battery Capacity = Number of Days of Autonomy X Daily Load(@ specified rate) Max % Usable X Temperature Derating Factor

Battery Capacity = 5 days X 30 Ah = 326 Ah(@ 77 hour rate) .50 X .92

The number of parallel batteries is:

Parallel Batteries = 326 Ah104 Ah per battery

= 3.13 batteries rounded to 4 batteries.




Since this is a 12 V system the number of batteries in series is 1. Note that we required4 batteries in parallel for a very small load. A larger load would have required morethan 4 batteries in parallel, but this is not good practice. This illustrates one of thelimitations of using automotive batteries for solar applications.

We calculated the load demand in Problem 9-3 to be 146.8 Ah/day. The number ofdays of autonomy is given as 14. The deep cycling batteries will allow a maximumcharge of 80%. However, we have an additional limitation due to concerns aboutfreezing, which limits the maximum % capacity to 75% (refer to Figure 15-2).

For the weighted average load operating time, we should only take the peak loads thatwill be on at the same time (the transmit loads for both transmitter #1 and #2):

Weighted operating time = 8 X 12 + 5 X 8

8 + 5

= 136 / 13 = 10.5 hours

The average discharge rate will be:

Average Discharge rate = Days of Autonomy X Operating TimeMaximum Discharge %

= 14 X 10.5 = 196 hours0.75

At - 30 °C, we need to derate the battery capacity 88% for temperature (using theC/120 curve of Figure 15-1). The battery capacity is then:

Battery Capacity = 14 days X 146.8 /day@ C/196 .75 Max discharge X .89 temperature factor

= 3079 Ahrs at the 196 hour rate

Looking at Table 15-1, we see columns for 120 hours and 240 hours. To beconservative, we will use the 120-hour column (since this gives the smaller rating for the

same cell). There are no cells with a capacity of 3114 Ah, so we will need to usemultiple cells in parallel. The capacity of KCPSA-15 cells is 1031. Three of these cellsin parallel give 3093 Ah, slightly more than we need. To get a 24-volt string of 2-voltcells, we need 12 cells in series. Multiplying by 3 cells in parallel, we will need 36 cellstotal.




Our requirement from Problem 15-6 is for 3079 Ah at a discharge rate of C/120. Thereare a number of possible configurations:

Cells inParalle

l

CellModel

CellCapacity Total Ah

2 Vb2412 1800 36003 Vb2407 1050 31504 Vb2312 900 3600

Although 5 strings of the Vb2310 (750 Ah) would also give sufficient capacity, we rulethis out because it exceeds the Rule of Thumb for using no more than 4 parallel strings.Since the Vb2407 is the smallest size that meets our requirements, it would probably bethe best choice. However, we would need to look at overall price, cell size and weight,battery layout and similar details.

Location: Leh

The month with the lowest insolation is January, with 4 kWh/m2 /day. To convert to

peak hours, we multiply by the conversion factor (in this case just 1).

4 kWh/m2 /day X 1 = 4 peak sun hours

Looking on the data sheet, we find the following information:

Module Pmax ImpSP36 36 2.1 AmpsSP75 75 4.4 Amps

The output for each module is the Imp multiplied by the peak sun hours

SP36: 2.1 amps X 4 peak sun hrs = 8.4 Ah/day

SP75: 4.4 amps X 4 peak sun hrs = 17.6 Ah/day




Location: Madras

The month with the lowest insolation is July, with 4.5 kWh/m2 /day. The peak sun hours

are:

4.5 kWh/m2 /day X 1 = 4.5 peak sun hours

Again, we take the Imp and multiply by the peak sun hours:

SP36: 2.1 amps X 4.5 peak sun hrs = 9.5 Ah/day

SP75: 4.4 amps X 4.5 peak sun hrs = 19.8 Ah/day

The load demand in Problem 9-1 is 230 Ah/day. We assume a nominal 12-volt system.Looking at the data for Montgomery, Alabama, we find that the lowest output during the

year for an SM55 module is 11.84 Ah / day in January. We use the same deratingfactors in the example: a battery coulombic efficiency of 90% and a module outputderating of 10%.

Number of Parallel Modules = 230 Ah/day daily load0.9 X 11.84 Ah/day X (1 - 0.10)

= 230 = 23.98 modules9.59

= 24 modules (rounded up)

Number of Series Modules = 12 volt system voltage12 volt nominal module

= 1 in series

Total number of modules = 1 series X 24 parallel = 24 modules.




The load demand in Problem 9-3 is 146.8 Ah/day. According to the data for Yuma,Arizona, the lowest output during the year for an SM55 module is 17.43 Ah / day in July.Since we are using the SP75 module, we adjust the output by a factor of 1.44.

SP75 output = SM55 Output X 1.44

= 17.43 Ah /day X 1.44

= 25.1 Ah/day

Assume the same conservative battery coulombic efficiency of 90% and a moduleoutput derating of 10%.

Number of Parallel Modules = 146.8 Ah/day daily load0.9 X 25.1 Ah/day X (1 - 0.10)

= 146.8 = 7.22 modules20.33

= 8 modules (rounded up)

Number of Series Modules = 24 volt system voltage12 volt nominal module

= 2 in series

Total number of modules = 2 series X 8 parallel = 16 modules.



Siemens Solar Basic PV Technology Course 16-1 System Design - System WiringCopyright © 1998 Siemens Solar Industries

Chapter Sixteen

System WiringAll the precise calculations in the world can be sabotaged by improper wiring of thecomponents. Poor choice of wire size can restrict battery charging and eventuallycause system failure even with an adequately sized array. Imagine someone workingon the electrical system. Anticipate problems or dangers and how you could eliminatethe dangers with safety disconnect switches and fuses. Thinking about safety andwiring requires that you consider what can possibly go wrong – not assuming thateverything will be safe and that people will not do stupid things. Design the system as ifyou were living with it and consider how you would want yourself and your familyprotected from the danger of fire and system failure.

Electrical wiring is a specialized field and this chapter cannot possibly cover all theissues completely. In the United States the National Electric Code (NEC) covers allissues of wiring in depth and specifically covers solar photovoltaic systems in Article690. A professional electrician or electrical contractor should be involved in the wiringof photovoltaic power systems.

The entire photovoltaic system should be diagrammed before costing or installation.This should include the wiring of modules in the array or sub-arrays, controllers, AC andDC load centers, batteries, inverter if present, grounding and circuit protection. Thewiring process is the time to consider safety features such as fuses, circuit breakers,ground fault interrupters, and grounding rods and wires. Also efficiency and costconcerns will affect many specific choices such as the sizes and types of wires to use.

In this chapter we will look at three aspects of photovoltaic system wiring: (1) properwire sizing and insulation; (2) array wiring in series and parallel groups; and (3) safetyequipment wiring.

Proper wiring and the design of safe, user-friendly photovoltaic systems are themost overlooked aspects of PV system design. Adherence to the NationalElectric Code and safe practice will result in reduced hazards associated withelectrical installations. Careful design of the wiring subsystem will result inefficient and reliable PV systems that are safely and easily serviced.




The National Electric Code andthe UL ListingThe National Electric Code (NEC), first developed in 1897, is a code of standards forthe design of safe electrical power systems in the US. Article 690 of the Codeaddresses safety standards (such as overcurrent protection) for the installation of PVsystems. Most local electrical inspectors in the US use the NEC as the code of practice(sometimes with slight modifications) in their jurisdiction. The NEC states thatadherence to the recommendations made will reduce the hazards associated withelectrical installations.

The NEC suggests and many inspectors require that an approved testing laboratorytest electrical equipment used. The most common of these national testingorganizations is the Underwriters Laboratories or UL. Most building inspectors expectto see UL on electrical products used in electrical systems. This presents a problem forthe PV industry as many products have low production rates that do not justify the costsfor the testing. As a result it is still difficult to design a PV system that has used all ULlisted components. Proper documentation on the use of the non-UL listed equipmentwill assist in gaining the approval of the local inspector. Another helpful source ofinformation on the NEC as it applies to PV is “Photovoltaic Power Systems and theNational Electric Code, Suggested Practices”. This publication is available from theSandia National Laboratories, US Dept. of Energy, (505) 844-3698.

An important concept to understand is that the code is not designed to create the mostcost effective or efficient systems. Its focus is safety. There may be a tension betweena system designer’s desire to make a system low cost and the code requirements thatlead to added components or larger size choices. Safety, reliability, efficiency and costeffectiveness must all be considered by the designer and by the user.

Exercise




Proper Wire Type and InsulationWire can be solid or stranded. Often AC house wiring is done with solid wire. Inphotovoltaic systems, however, the DC currents are higher than typical for AC

situations, and therefore the wire size must be larger. Solid wire becomes stiff anddifficult to work with in large sizes and photovoltaic installers often use stranded wire.

Wire can be made of aluminum or copper. Aluminum wire can be considered for verylarge wire runs, needing 4/0 for example because it costs less than equivalent copperwire. For most wiring applications, however, copper wire should be used.

Most wire is insulated with a type of plastic called thermoplastic. Thermoplastic isdesigned to withstand considerable heat. When current flows through a wire itsresistance causes power to be dissipated. If the current and/or resistance is high thepower dissipated can be significant, causing the wire to get warm or even hot. For this

reason the insulation must be able to withstand heat. While most wire is made ofthermoplastic, rubber insulation is also available.

Thermoplastic and rubber insulation are designated by a letter code that indicates thetype of environment that is most appropriate for its use. The most common type is T ,which means thermoplastic insulation for use in dry areas where temperature will not

exceed 60 °C. Type TW insulation should be used for wet (or dry) areas. Somecommon types of wire and their U.S. designations are listed in the table.

Letter Designation Type of Insulation For Use in:

T Thermoplastic Dry Areas where temp will not exceed 60 °C

TW Thermoplastic Wet Wet (or Dry) Areas where temp will not exceed 60°C

THW Thermoplastic, Hightemperature, Wet

Wet (or Dry) Areas where temp will not exceed 75°C

THHW Thermoplastic, High,High temperature, Wet


THHN Thermoplastic High,High temperature,Nylon jacket

Dry Areas where temp will not exceed 90°C.

R Rubber insulation Dry Areas where temp will not exceed 60°C

RHW Rubber, Hightemperature, Wet


RHH Rubber, High, Hightemperature

Dry Areas where temp will not exceed 90°C

RHW-2 Rubber, High, Hightemperature, Wet

Wet or Dry locations where temp will not exceed

90°C.




Color Coding of Wire

The color-coding of wire makes wiring easier and is used to designate its function. Italso minimizes the possibility that incorrect connections will be made. For AC house

wiring in the US white or gray is always used for neutral or main system grounds (asopposed to equipment ground that is commonly green or bare wire). The hot wires canbe black, red, blue or yellow. Black is the most common; but, in cables with two hotwires, black and red are used.

In PV systems the NEC specifies that in a DC circuit the system grounded conductor bewhite. There is no convention designating the color of the ungrounded conductor buttypically red or black are used. Green or green with yellow-stripped is used for theequipment ground.

Wire AC(Below 600 Volts) DC (Below 600 Volts)

Neutral or Ground White or Gray White

Hot (or high side) Black, Red, Blue, Yellow No NEC Requirement, buttypically black or red.

Equipment Ground orGrounding

Bare, Green, or Green withYellow Stripe

Bare, Green, or Green withYellow Stripe

Conductor Types

A cable is a connection of two or more wires combined in a common outer covering. Itis pre-assembled with wires of various sizes and insulation covers. All of the wires areprotected with an outer sheath made of plastic, rubber, or metal. Most of the wire youwill be working with comes in cable form, which is easier to work with than multipleindividual wires.

The most commonly used cable for electrical wiring of AC systems is nonmetallicsheathed cable or type NM. This is commonly referred to as Romex, which is used inhouse and building wiring in dry locations. Non-metallic sheathed cable is available indifferent versions for a variety of other applications including areas of moisture (NMC)and wet areas buried in the ground (UF ) commonly referred to as underground feedercable.




Wires and cables used in PV systems are typically designed for outdoor situations withmoisture and sunlight resistance or are placed in conduit to protect against weather.The following cable types are the most commonly used in PV systems.

Wire/Cable Type Design Use Applications inPhotovoltaic

Systems

Design Issues forPhotovoltaic

SystemsUSE or USE-2(Underground ServiceEntrance)

or

SE (Service Entrance)

When made to UL

Standard, it has a 90°Crating and is sunlightresistant. Moistureproof, resistant to acidsand chemicals.

Best cable for moduleinterconnects

The “-2” designationmeans that theinsulation is rated for

90°C. wet or dry, and isespecially suited for thehot temperatures thatmodule wiring wouldendure.

TC (Tray Cable) Used for moduleinterconnects

TC is sunlight resistantand generally marked as

such.UF (UndergroundFeeder)

Designed for use in wetareas especially outsidewhere it may beexposed to weather,also may be burredunderground.

For use in array wiring,and between arraywiring and the PVdisconnect device

Not recommendedwithin batteryenclosures, Shouldspecify sunlight resistantcoating if used outside.

THHN (Dry Locations)

and

THWN (Wet Locations)

Must be used inconduit

Used between the arraywiring and the PVdisconnect device

Must be used in conduitwith insulation rated at

90 °C. In larger sizes, isdifficult to bend.Conduit sizing dependsupon number and size

of conductorsUSE/RHH/RHW Multi-stranded flexible

UL listed cable availablein large sizes

Used for battery /Inverter wiring

Welding Cable Multi-stranded flexiblecable available in largesizes

Used for batteryinterconnects and forbattery/inverter wiring

Not UL Listed

NM (“Romex”) Interior house wiring, drylocations

Limited. Used inconduit to connect arrayto PV disconnect deviceand for short DC wiringin the house.

Conduit sizing dependsupon number and sizeof conductors




Consider Both Ampacity andVoltage Drop in Wire Sizing

There are two major considerations in sizing wire for a photovoltaic system:

• Ampacity

• Voltage drop

Ampacity is the current carrying capacity of the wire, or its ability to carry current safelywithout overheating. Ampacity is based upon the wire size or cross-sectional area, thetype of materials its constructed of (e.g., copper, thermoplastic insulation), its length,and its temperature. For these different conditions, the National Electric Code hasdetermined a maximum ampacity for each wire type (Romex, USE, etc.) for varioustemperatures and lengths. This is the same standard for which PV systems need to bedesigned.

The second consideration in wiring is voltage drop. Unlike higher voltage AC systemswhere a small voltage drop will not have measurable impact on the efficiency anddeliverability of energy, PV systems are very sensitive to voltage changes. Thesevoltage drops must be considered in the design of a photovoltaic system.




Wire Sizing Based on AmpacityAmpacities for common types' wire used in photovoltaic systems are presented in thetables on the next page. Wire sizes are given in both the American Wire Gauge (AWG)

and the metric method of cross sectional sizing in square millimeters. The value in theregular font is for wires in conduit while the italic numbers are the ampacity values ofsingle conductors in free air. These values represent the maximum current that the

conductor can safely carry continuously while the ambient air temperature is at 30 °C.If the ambient temperature is higher then the maximum current must be reduced. Thereduction for temperature is discussed later in this section.

We can identify general areas in photovoltaic system designs where wires are sizedbased on ampacity: (1) short connections of DC wiring made indoors from batteries toswitches, controls and other components; (2) AC wiring from the inverter to adistribution panel and on to the AC loads; (3) battery-to-inverter wiring; and (4)

conductors carrying array current.

Short DC Wiring From Battery

Wires to connect switches, controls and other components are put within conduit. Theruns are usually so short that voltage drop is not a problem and the wires are sizedbased on ampacity. Usual safety practice is to oversize the wire by 125% above thecontinuous current that the wire might handle. (This does not include wire carryingcurrent flow from the array into the battery. There is an additional safety factor that is

applied. This is explained later.)

AC Load Wire Sizing

For AC load wiring usual electrical practices apply. Circuits are designed to carrytypically 15 or 20 amps and usually AWG 14 or 12 wire is sufficient. Wires are usuallycarried within conduit. Again, the usual safety factor for ampacity recommended by theNEC is 125% of the maximum current that might flow continuously.




Wire Sizing (AWG)

Wire Size(AWG)

14

12

10

8

6

4

2

1

0

00

000

0000

Ampacity (amps) UF USE,RHW, THHN,THHW

THHW,THWN RHW-2,USE-2

60o

75o

90o

20 25

25 30

30 40

40 60

55 80

70 105

95 140

110 165

125 195

145 225

165 260

195 300

20 30

25 35

35 50

50 70

65 95

85 125

115 170

130 195

150 230

175 265

200 310

230 360

25 35

30 40

40 55

55 80

75 105

95 140

130 190

150 220

170 260

195 300

225 350

260 405

Regular Font = NEC 1993, Table 310-16. Ampacities of Insulated Conductors Rated 0-2000 Volts.Not More Than Three Conductors in Raceway, Cable, or Earth (Directly Buried), Ambient Temp. of 30 oC.

Italic Font = NEC 1993, Table 310-17. Ampacities of Single Insulated Conductors Rated 0-2000 Volts in Free Air Based on Ambient Temp. of 30 oC.

Wire Sizing (metric)

Wire ampacity rated at ambient temperature of 30 deg.C., maximum con ductor temperature of 60

deg.C., and assuming cable touches surfaces of walls, etc. Voltage loss factor rated at 20 deg.C.

Wire Size(m m 2)

2.5

4

6

10

16

25

35

50

70

95120

150

Ampacity(amps)

32

42

54

73

98

12 9

15 8

19 8

24 5

29 234 4

39 1




Inverter Wire Sizing

For the case of battery-to-inverter wiring the DC currents can be very high and care

must be taken to choose a wire adequate for the job. Inverters used for largeresidential or commercial applications are often rated to produce 2000-4000 wattscontinuously. If the input voltage to the inverter is 12 volts this will mean DC currents of180-400 amps. Even if the inverter input voltage is 24 volts the current can be 100-200amps. Very large wires are needed to handle such currents.

The distance between the battery and the inverter should be kept as short as possibleto minimize voltage drop at such high currents. Usually no more than about six feet ortwo meters is recommended as the maximum length for inverter cables.

The current that should be used for inverter wire sizing should correspond to the fullload continuous output of the inverter, not just the load demand being sized for today.A system designer may often choose to install a larger inverter than is needed, toaccommodate future growth. The wire size installed, however, should be designedaround the full continuous demand of the inverter. This will allow future growth in thepower to the loads without requiring costly re-wiring from small wires perhaps neededtoday to larger wires needed in the future.

The formula for calculating the current that should be used as the basis for invertercable wiring is presented next. The continuous power rating of the inverter is divided bythe full load efficiency. This gives the input current needed to overcome internalinverter inefficiency. This value is divided by the lowest DC input voltage to the inverter(usually about 11 volts for 12 volt systems and about 22 volts for 24 volt systems) notby just the nominal voltage (12, 24 or 48 volts). This is to account for the fact that at alower battery voltage the inverter will draw higher currents to meet load powerdemands.

The NEC safety margin of 125% should be applied on top of this real current value forpurposes of choosing wire with adequate ampacity. In other words the calculatedcurrent might actually be flowing for extended periods of time, and the wire chosenshould have an ampacity of at least 125% more than this value.




Ampacity of = Maximum Continuous Input Current X 1.25Inverter Wire

Maximum = Continuous Inverter Output Rating (watts)Continuous Full Load Efficiency X Lowest Input VoltageInput Current

Nominal Voltage Lowest Input Voltage12 1124 2248 44

Example: The total load demand for a residential system comes to about2.5 kW continuous. The system designer chooses to install theTrace 4048 sinewave inverter, with 48-volt input and rated forcontinuous output at 4000 watts. Efficiency at 4000 watts is about87%.

Max. Inverter Input Current= 4000 watts.87 X 44 volts

= 104 amps

Applying the 125% safety factor, the wire chosen should have anampacity of at least:

Ampacity value = 104 X 125% = 104 X 1.25

= 130 amps

So in this example, a recommendation might be to use type USE/RHH/RHW cable withan ampacity rating of at least 130 amps.




Conductors Carrying Array Current

Wires carrying current from the modules or array are to be sized based on their short

circuit current (Isc), rather than just on their maximum power current (Imp). The usualNEC code practice is to choose wire with ampacity rating of at least 125% higher thanthe array Isc.

Modules listed to Underwriters Laboratory Standards have an additional factor applied.Module Isc is to be increased by 125% before any other calculations are considered, toaccount for possible environmental enhancements of output such as bright clouds orreflective ground cover. This factor is to apply even before the NEC safety factor of125% is applied.

In other words, apply two factors to wires carrying array current:

• Derate the ampacity of the wire by 20% (apply as 125% factor) to insure that thewire is not operated continuously at more than 80% of its rated value.

• Derate the expected continuous output current of a solar module by 125% toaccount for expected continuous enhancements to output.

The combination of the two factors means that a factor of 156% should be applied tothe published Isc value of a module for ampacity sizing purposes.

Array wiring ampacity safety factor (UL and NEC) = 156%

This factor should be applied to all short lengths of wire interconnecting array power toswitches, controls and other equipment.




Adjustment To Ampacity For Temperature

The ampacity of a conductor is based on the ability of the insulation to adequately

dissipate the heat generated by the current flow through the copper. Most ampacityratings are based on the ambient temperature around the wire being 30 °C.

If the ambient temperature is greater than the rated value of 30 °C. then the insulationwill not be able to dissipate heat as well. The ampacity rating of the conductor shouldbe decreased to reflect the increased ambient temperature.

The temperature derating factors recommended by the NEC for various conductors arepresented below. These must be applied against the standard values presented earlierin Tables 16-4 and 16-5. The same factors apply for conductors in conduit or in freeair.

Ampacity Temperature Correction Factors

AmbientTemperature (

oC)

Standard Temperature Rating

60oC 75

oC 90

oC

26-30 1.00 1.00 1.0031-35 .91 .94 .9636-40 .82 .88 .91

41-45 .71 .82 .8746-50 .58 .75 .8251-55 .41 .67 .7656-60 --- .58 .7161-70 --- .33 .5871-80 --- --- .41




Example: SM55 modules are mounted on a fixed structure in the desert. Theback of the module is exposed to free air. Ambient temperaturescan get up to 45

oC. Type USE-2 (rated for 90

oC) AWG #10 is

used to interconnect the modules.

The temperature derating for 90oC conductors at 45

oC is .87.

The standard ampacity of USE-2 wire AWG #10 in free airis 55 amps. [Refer to the italic numbers in Table 16-4].

The combined derating for the ampacity is given by:

Ampacity = 55 amps rating X .87 ÷ 1.25 NEC ÷ 1.25 UL

= 30 amps

Example: SP75 modules are mounted on a roof. The conductors are inconduit, and the back of the module is mounted very close to theroof surface where temperatures can get up to 75

oC. Type USE-2

(rated for 90oC) AWG #8 is used to interconnect the modules.

The temperature derating for 90oC conductors at 75

oC is .41.

The standard ampacity of USE-2 wire AWG#8 in conduitis 55 amps. [Refer to the regular font numbers in Table 16-4].

The combined derating for the ampacity is given by:

Ampacity = 55 amps rating X .41÷ 1.25 NEC

÷ 1.25 UL

= 15 amps

Temperature can lead to a substantial reduction in the ampacity of conductors,especially in photovoltaic systems where wires interconnecting modules and connectingan array to the main system components are usually exposed to high ambientconditions.

Array wiring from the array in the field is usually sized even larger than this factor would

dictate. The array in the field is typically a long distance from all the other equipment inthe power system. The long distance of this run of wire requires a different approach towire sizing – one based on voltage drop not on ampacity. This is discussed later in thischapter.




Exercise




Wire Sizing Based on Voltage Drop

The second factor in wire sizing is voltage drop. This is often a small consideration in

common utility connected high voltage residential and commercial systems butbecomes critical in photovoltaic system design. Sizing based on voltage drop isespecially important for the wire run between the array and the battery. This distance isusually quite large and voltage drop can be a big problem.

All conductors have some small resistance that causes a loss of voltage depending onthe diameter and length of the wire. The primary effect of small wire between aphotovoltaic array and a battery bank is to reduce the amount of current that flows intothe battery. This is because the resistance of the wire causes the voltage potential ofthe array to be reduced at all current levels. More voltage is loss at higher arraycurrents. This results in a "dragging in" of the IV curve of the array, as shown below.

Effect of Voltage Dropon Module Output

-4 0 4 8 12 16Voltage (volts)

Current

(amps)

Range ofbatteryvoltage

no wire loss12

10

8

6

4

2

AWG #4

AWG #10

AWG #14

Phoenix50 feet to battery

4 parallel SM50-H modules




The batteries operate in a narrow range of voltage close to the "knee" of the array IVcurve if the “knee” is brought in due to small wire. The operating point on the curvethen drops down the IV curve and less current flows to the battery. The net effect is tooperate the array at a lower current level. The figure shows that if there was novoltage loss in the wire and the batteries were operating at about 13.5 volts, then the

current from the array would be about 11 amps. With the losses expected using AWG#10 wire the battery would get only about 8 amps. By using common residential wiresize of AWG #14 the current would be only about 6 amps, close to half of what isexpected. The current would not be what the designer anticipated and over time thebattery would never be fully recharged and the system would fail.

Size For 2% Maximum Loss

The National Electric Code recommends that there be no more than 5% voltage drop inany circuit between a source of power (typically a main service entrance in a building)and the furthest load. In stand-alone photovoltaic systems we can apply this rule tomean there should be no more than 5% voltage loss from the modules to the DC loadsor into the inverter for AC systems.

An even more stringent recommendation is to not have more than a 2% voltage drop inany feeder circuit or any branch circuit. A feeder circuit is any circuit from a source ofpower (for example, a photovoltaic array or a battery bank) to a distribution centerwhere circuit protection is available (for example, a central wiring center or DC loadcenter). A branch circuit is any circuit from a distribution center to a load (for example,from a DC load center to the loads). Some photovoltaic system designers use the 2%factor from the modules to the battery and then from the battery to the loads topreserve voltage and improve performance. This results in larger wires and slightlygreater costs but preserves more of the expensive energy generated in the first place.




Voltage Drop Factors

The voltage drop in a wire is calculated using the Voltage Loss Factors listed on the

next page (ohms/foot or volts/ampere/foot). The current and the total length of wire(going to and returning from) are multiplied by the Factor to give the volts lost.

Voltage Drop in Wire = Current x Total Wire Length x Voltage Factor

The voltage loss in high voltage AC circuits is usually quite small whereas the losses inlow voltage DC systems can be quite high. This is not because the current is DC or AC,

but rather because the voltages in AC systems are usually higher (typically 120 or 240volts) while the voltages in DC systems are usually low (typically 12 or 24 volts). Someexamples will illustrate this.

Example: A 1000 watt load (ex: heater, microwave oven, hair dryer) is located 25feet from a load center. The total length of wire for the circuit is therefore50 feet. The amps of current to the load would be given by

Current = Power ÷ Volts = 1000 watts ÷ 120 volts

= 8.3 amps

According to the ampacity charts, AWG #14 wire would be more thansufficient to handle 8 amps.

If the wire used was standard household AWG #14 (voltage loss factor of.002823 volts/amp/foot), the voltage loss in the circuit would be:

Voltage Drop = 1000 watts X 50 feet X 0.002823 v/a/ft(#14 wire) 120 volts(120 volts)

= 1.18 volts

The recommended maximum of 2% voltage loss for a 120-volt circuit would be a dropof 2.4 volts. The 1.18-volt drop in this circuit is less than 2.4, so in this case the wirechosen based on ampacity is also acceptable in terms of voltage loss.




Voltage Loss Factors (AWG)

Voltage Loss Factor(volt/amp/foot)

.002823

.001775

.001117

.0007023

.0004416

.0002778

.0001747

.0001385

.0001099

.0000871

.0000691

.0000548

Wire Size(AWG)

14

12

10

8

6

4

2

1

0

00000

0000

Voltage Loss Factors (Metric)

Wire Size (mm2) Voltage Loss Factor

(volt/amp/meter)

2.5 0.00848644 0.0052881

6 0.0034434

10 0.0021112

16 0.0013206

25 0.0008371

35 0.0005614

50 0.0004183

70 0.0003151

95 0.0002406

120 0.0001907

150 0.0001518




If the same load were operating at nominal 12 volts, however, the current would bemuch higher to produce the same power and the wire size needed would be muchgreater.

Example: The same 1000 watt load is operating at 12 volts nominal, and islocated 25 feet from a DC load center. The total length of wire forthe circuit is therefore 50 feet. The amps of current to the loadwould be given by

Current = Power ÷ Volts = 1000 watts ÷ 12volts

= 83.3 amps (ten times higher than before)

According to the ampacity charts, the AWG #14 wire used in the

previous example would be insufficient to handle this high current.

We would need at least AWG #4 wire if we used 75 °C ratedconductor.




Another example illustrates that using wire with enough ampacity does not necessarilygive the acceptable voltage drop of 2%.

Example: An array of four SP75 modules is wired in parallel at 12 volts. The

array is located 30 feet from the load center and other componentsof the power system.

The nominal Isc from the array will be given by

Array Isc = 4 X 4.8 amps Isc = 19.2 amps

The NEC and UL safety factors apply as this is module current.Conductor ampacity must be overrated by 125% and current mustbe overrated by 125%. The current that should be used forampacity is given by

Array Current for Ampacity = 19.2 X 1.25 X 1.25

= 30 amps

Assume the conductors are in buried conduit with ambient of about

30 °C. so no temperature derating will apply at this time. If we use

THHN (90 °C rated) conductor, then AWG #12 wire would appearadequate to handle the current.

But now we look at the voltage drop that would occur using

# 12 wire. The total length of conductor is 2 x 30 feet = 60 feet.Using the nominal array Isc value of 19.2 and the voltage lossfactor for #12 wire we find that the voltage drop would be

Voltage Drop = 19.2 amps X 60 feet X .001775 v/a/ft

= 2.05 volts

The nominal system voltage is only 12 volts. This voltage drop of 2.05 volts is more

than 17% of that nominal voltage! The acceptable voltage drop for a 12-volt system isonly 2% of 12, or only .24 volts.




This illustrates that choosing wire based on ampacity for the long run from the array intothe main system components is unacceptable. We must size that run of wire based onvoltage drop. The wire chosen will be much larger than dictated by ampacity alone. Tosize for acceptable voltage drop, we must use the equation on Page 14 in reverse.Insert the wire length, the array current, and the acceptable voltage drop and solve for

the voltage loss factor. An example will illustrate.

Example: Using the above array again, the array current is 19.2 amps, thetotal wire length is 60 feet. The acceptable voltage drop is 2% of12 volts or .02 X 12 = .24 volts.

Voltage Loss Factor = Acceptable voltage dropCurrent X Total Length

= .24 volts

19.2 amps X 60 feet

= .000208

Referring to the Voltage Loss Factor chart of Table 16-7, the factor for #4 wire(.0002778 volt/amp/foot) is not small enough, so go to the next larger wire. This meansthat #2 wire would be the size needed to keep the voltage drop less than 2%. Comparethis to the #12 that was acceptable with regard to ampacity only.




Voltage Drop Tables

A set of tables has been created to allow easy wire sizing based on voltage drop.

• Tables have been created for nominal 12, 24, 36, 48 and 120-volt systems.

• Distances indicated are for the one-way length of the conductor. The total round-triplength of conductor has been taken into account in the calculations.

• American Wire Gauge (AWG) tables indicate lengths in feet and metric (mm2) tables

indicate in meters.

Procedure for using the tables:

1. Determine the current that might be flowing in the wire ( use Isc for array wire

calculations);2. Read across the row to find the one-way distance (if between values go to the next

larger number);3. Read up to find the acceptable wire size for 2% voltage drop.

If your current is greater than indicated in the tables then use the formula on Page 17and the Voltage Loss Factors on Page 18 to determine the acceptable wire size. Whenusing the formula be sure to use the total length of the conductor, not just the one-waydistance that is used in the tables.

(Ampacity limits have not been indicated in these charts. Be sure to check that theampacity of the wire is not exceeded especially in the case of high current values).




Voltage Drop Table -- 12 Volt Nominal

Maximum one-way distance for 2% voltage loss in 12 volt systems

Wire Size (AWG) [distances in feet]14 12 10 8 6 4 2 1 0 00 000 0000

Amps1 43 68 107 171 272 432 687 866 1092 1377 1737 21902 21 34 54 85 136 216 343 433 546 689 868 10953 14 23 36 57 91 144 229 289 364 459 579 7304 11 17 27 43 68 108 172 217 273 344 434 5485 9 14 21 34 54 86 137 173 218 275 347 4386 7 11 18 28 45 72 114 144 182 230 289 3657 6 10 15 24 39 62 98 124 156 197 248 3138 5 8 13 21 34 54 86 108 136 172 217 2749 5 8 12 19 30 48 76 96 121 153 193 24310 4 7 11 17 27 43 69 87 109 138 174 21915 3 5 7 11 18 29 46 58 73 92 116 14620 2 3 5 9 14 22 34 43 55 69 87 11025 2 3 4 7 11 17 27 35 44 55 69 8830 1 2 4 6 9 14 23 29 36 46 58 7335 1 2 3 5 8 12 20 25 31 39 50 6340 1 2 3 4 7 11 17 22 27 34 43 5545 1 2 2 4 6 10 15 19 24 31 39 4950 1 1 2 3 5 9 14 17 22 28 35 4460 1 1 2 3 5 7 11 14 18 23 29 3770 1 1 2 2 4 6 10 12 16 20 25 3180 1 1 1 2 3 5 9 11 14 17 22 2790 0 1 1 2 3 5 8 10 12 15 19 24

100 0 1 1 2 3 4 7 9 11 14 17 22

Wire Size (mm2) [distances in meters]

2.5 4 6 10 16 25 35 50 70 95 120 150Amps

1 15 24 35 63 99 154 217 311 441 583 745 9302 7 12 18 31 50 77 108 155 221 291 373 4653 5 8 12 21 33 51 72 104 147 194 248 3104 4 6 9 16 25 38 54 78 110 146 186 2335 3 5 7 13 20 31 43 62 88 117 149 1866 2 4 6 10 17 26 36 52 74 97 124 1557 2 3 5 9 14 22 31 44 63 83 106 1338 2 3 4 8 12 19 27 39 55 73 93 1169 2 3 4 7 11 17 24 35 49 65 83 10310 1 2 4 6 10 15 22 31 44 58 75 9315 1 2 2 4 7 10 14 21 29 39 50 6220 1 1 2 3 5 8 11 16 22 29 37 4725 1 1 1 3 4 6 9 12 18 23 30 3730 0 1 1 2 3 5 7 10 15 19 25 3135 0 1 1 2 3 4 6 9 13 17 21 2740 0 1 1 2 2 4 5 8 11 15 19 23

45 0 1 1 1 2 3 5 7 10 13 17 2150 0 0 1 1 2 3 4 6 9 12 15 1960 0 0 1 1 2 3 4 5 7 10 12 1670 0 0 1 1 1 2 3 4 6 8 11 1380 0 0 0 1 1 2 3 4 6 7 9 1290 0 0 0 1 1 2 2 3 5 6 8 10

100 0 0 0 1 1 2 2 3 4 6 7 9







(amps)1 85 135 215 342 543 864 1374 1733 2184 2755 3474 4380

2 43 68 107 171 272 432 687 866 1092 1377 1737 21903 28 45 72 114 181 288 458 578 728 918 1158 14604 21 34 54 85 136 216 343 433 546 689 868 10955 17 27 43 68 109 173 275 347 437 551 695 876

6 14 23 36 57 91 144 229 289 364 459 579 7307 12 19 31 49 78 123 196 248 312 394 496 6268 11 17 27 43 68 108 172 217 273 344 434 5489 9 15 24 38 60 96 153 193 243 306 386 487

10 9 14 21 34 54 86 137 173 218 275 347 43815 6 9 14 23 36 58 92 116 146 184 232 29220 4 7 11 17 27 43 69 87 109 138 174 21925 3 5 9 14 22 35 55 69 87 110 139 175

30 3 5 7 11 18 29 46 58 73 92 116 14635 2 4 6 10 16 25 39 50 62 79 99 125

40 2 3 5 9 14 22 34 43 55 69 87 11045 2 3 5 8 12 19 31 39 49 61 77 97

50 2 3 4 7 11 17 27 35 44 55 69 8860 1 2 4 6 9 14 23 29 36 46 58 7370 1 2 3 5 8 12 20 25 31 39 50 6380 1 2 3 4 7 11 17 22 27 34 43 55

90 1 2 2 4 6 10 15 19 24 31 39 49100 1 1 2 3 5 9 14 17 22 28 35 44


2.5 4 6 10 16 25 35 50 70 95 120 150Amps

1 29 47 71 126 198 308 433 622 882 1165 1491 18602 15 24 35 63 99 154 217 311 441 583 745 9303 10 16 24 42 66 103 144 207 294 388 497 620

4 7 12 18 31 50 77 108 155 221 291 373 4655 6 9 14 25 40 62 87 124 176 233 298 3726 5 8 12 21 33 51 72 104 147 194 248 3107 4 7 10 18 28 44 62 89 126 166 213 2668 4 6 9 16 25 38 54 78 110 146 186 2339 3 5 8 14 22 34 48 69 98 129 166 20710 3 5 7 13 20 31 43 62 88 117 149 18615 2 3 5 8 13 21 29 41 59 78 99 12420 1 2 4 6 10 15 22 31 44 58 75 9325 1 2 3 5 8 12 17 25 35 47 60 7430 1 2 2 4 7 10 14 21 29 39 50 6235 - 1 2 4 6 9 12 18 25 33 43 5340 - 1 2 3 5 8 11 16 22 29 37 4745 - - 2 3 4 7 10 14 20 26 33 4150 - - 1 3 4 6 9 12 18 23 30 3760 - - 1 2 3 5 7 10 15 19 25 31

70 - - 1 2 3 4 6 9 13 17 21 2780 - - 1 2 2 4 5 8 11 15 19 2390 - - 1 1 2 3 5 7 10 13 17 21

100 - - 1 1 2 3 4 6 9 12 15 19







Amps1 128 203 322 513 815 1296 2061 2599 3276 4132 5211 65712 64 101 161 256 408 648 1030 1300 1638 2066 2605 32853 43 68 107 171 272 432 687 866 1092 1377 1737 21904 32 51 81 128 204 324 515 650 819 1033 1303 16435 26 41 64 103 163 259 412 520 655 826 1042 13146 21 34 54 85 136 216 343 433 546 689 868 10957 18 29 46 73 116 185 294 371 468 590 744 9398 16 25 40 64 102 162 258 325 409 517 651 8219 14 23 36 57 91 144 229 289 364 459 579 73010 13 20 32 51 82 130 206 260 328 413 521 65715 9 14 21 34 54 86 137 173 218 275 347 43820 6 10 16 26 41 65 103 130 164 207 261 32925 5 8 13 21 33 52 82 104 131 165 208 26330 4 7 11 17 27 43 69 87 109 138 174 21935 4 6 9 15 23 37 59 74 94 118 149 18840 3 5 8 13 20 32 52 65 82 103 130 16445 3 5 7 11 18 29 46 58 73 92 116 14650 3 4 6 10 16 26 41 52 66 83 104 13160 2 3 5 9 14 22 34 43 55 69 87 11070 2 3 5 7 12 19 29 37 47 59 74 9480 2 3 4 6 10 16 26 32 41 52 65 8290 1 2 4 6 9 14 23 29 36 46 58 73

100 1 2 3 5 8 13 21 26 33 41 52 66


2.5 4 6 10 16 25 35 50 70 95 120 150Amps

1 44 71 106 188 298 462 650 933 1324 1748 2236 27912 22 35 53 94 149 231 325 466 662 874 1118 13953 15 24 35 63 99 154 217 311 441 583 745 9304 11 18 27 47 74 115 162 233 331 437 559 6985 9 14 21 38 60 92 130 187 265 350 447 5586 7 12 18 31 50 77 108 155 221 291 373 4657 6 10 15 27 43 66 93 133 189 250 319 3998 5 9 13 24 37 58 81 117 165 218 280 3499 5 8 12 21 33 51 72 104 147 194 248 31010 4 7 11 19 30 46 65 93 132 175 224 27915 3 5 7 13 20 31 43 62 88 117 149 18620 2 4 5 9 15 23 32 47 66 87 112 14025 2 3 4 8 12 18 26 37 53 70 89 11230 1 2 4 6 10 15 22 31 44 58 75 9335 1 2 3 5 9 13 19 27 38 50 64 8040 1 2 3 5 7 12 16 23 33 44 56 70

45 1 2 2 4 7 10 14 21 29 39 50 6250 1 1 2 4 6 9 13 19 26 35 45 5660 1 1 2 3 5 8 11 16 22 29 37 4770 1 1 2 3 4 7 9 13 19 25 32 4080 1 1 1 2 4 6 8 12 17 22 28 3590 0 1 1 2 3 5 7 10 15 19 25 31

100 0 1 1 2 3 5 6 9 13 17 22 28







Amps1 170 270 430 683 1087 1728 2748 3466 4368 5510 6947 87612 85 135 215 342 543 864 1374 1733 2184 2755 3474 43803 57 90 143 228 362 576 916 1155 1456 1837 2316 29204 43 68 107 171 272 432 687 866 1092 1377 1737 21905 34 54 86 137 217 346 550 693 874 1102 1389 17526 28 45 72 114 181 288 458 578 728 918 1158 14607 24 39 61 98 155 247 393 495 624 787 992 12528 21 34 54 85 136 216 343 433 546 689 868 10959 19 30 48 76 121 192 305 385 485 612 772 97310 17 27 43 68 109 173 275 347 437 551 695 87615 11 18 29 46 72 115 183 231 291 367 463 58420 9 14 21 34 54 86 137 173 218 275 347 43825 7 11 17 27 43 69 110 139 175 220 278 35030 6 9 14 23 36 58 92 116 146 184 232 29235 5 8 12 20 31 49 79 99 125 157 198 25040 4 7 11 17 27 43 69 87 109 138 174 21945 4 6 10 15 24 38 61 77 97 122 154 19550 3 5 9 14 22 35 55 69 87 110 139 17560 3 5 7 11 18 29 46 58 73 92 116 14670 2 4 6 10 16 25 39 50 62 79 99 12580 2 3 5 9 14 22 34 43 55 69 87 11090 2 3 5 8 12 19 31 39 49 61 77 97

100 2 3 4 7 11 17 27 35 44 55 69 88


2.5 4 6 10 16 25 35 50 70 95 120 150Amps

1 58 94 142 251 397 615 866 1244 1765 2330 2981 37212 29 47 71 126 198 308 433 622 882 1165 1491 18603 19 31 47 84 132 205 289 415 588 777 994 12404 15 24 35 63 99 154 217 311 441 583 745 9305 12 19 28 50 79 123 173 249 353 466 596 7446 10 16 24 42 66 103 144 207 294 388 497 6207 8 13 20 36 57 88 124 178 252 333 426 5328 7 12 18 31 50 77 108 155 221 291 373 4659 6 10 16 28 44 68 96 138 196 259 331 41310 6 9 14 25 40 62 87 124 176 233 298 37215 4 6 9 17 26 41 58 83 118 155 199 24820 3 5 7 13 20 31 43 62 88 117 149 18625 2 4 6 10 16 25 35 50 71 93 119 14930 2 3 5 8 13 21 29 41 59 78 99 12435 2 3 4 7 11 18 25 36 50 67 85 10640 1 2 4 6 10 15 22 31 44 58 75 93

45 1 2 3 6 9 14 19 28 39 52 66 8350 1 2 3 5 8 12 17 25 35 47 60 7460 1 2 2 4 7 10 14 21 29 39 50 6270 1 1 2 4 6 9 12 18 25 33 43 5380 1 1 2 3 5 8 11 16 22 29 37 4790 1 1 2 3 4 7 10 14 20 26 33 41

100 1 1 1 3 4 6 9 12 18 23 30 37







Amps1 425 676 1074 1709 2717 4320 6869 8664 10919 13774 17369 219022 213 338 537 854 1359 2160 3434 4332 5460 6887 8684 109513 142 225 358 570 906 1440 2290 2888 3640 4591 5790 73014 106 169 269 427 679 1080 1717 2166 2730 3444 4342 54755 85 135 215 342 543 864 1374 1733 2184 2755 3474 43806 71 113 179 285 453 720 1145 1444 1820 2296 2895 36507 61 97 153 244 388 617 981 1238 1560 1968 2481 31298 53 85 134 214 340 540 859 1083 1365 1722 2171 27389 47 75 119 190 302 480 763 963 1213 1530 1930 243410 43 68 107 171 272 432 687 866 1092 1377 1737 219015 28 45 72 114 181 288 458 578 728 918 1158 146020 21 34 54 85 136 216 343 433 546 689 868 109525 17 27 43 68 109 173 275 347 437 551 695 87630 14 23 36 57 91 144 229 289 364 459 579 73035 12 19 31 49 78 123 196 248 312 394 496 62640 11 17 27 43 68 108 172 217 273 344 434 54845 9 15 24 38 60 96 153 193 243 306 386 48750 9 14 21 34 54 86 137 173 218 275 347 43860 7 11 18 28 45 72 114 144 182 230 289 36570 6 10 15 24 39 62 98 124 156 197 248 31380 5 8 13 21 34 54 86 108 136 172 217 27490 5 8 12 19 30 48 76 96 121 153 193 243

100 4 7 11 17 27 43 69 87 109 138 174 219


2.5 4 6 10 16 25 35 50 70 95 120 150Amps

1 146 236 354 628 992 1538 2166 3109 4412 5825 7453 93022 73 118 177 314 496 769 1083 1554 2206 2913 3727 46513 49 79 118 209 331 513 722 1036 1471 1942 2484 31014 37 59 88 157 248 385 542 777 1103 1456 1863 23265 29 47 71 126 198 308 433 622 882 1165 1491 18606 24 39 59 105 165 256 361 518 735 971 1242 15507 21 34 51 90 142 220 309 444 630 832 1065 13298 18 29 44 79 124 192 271 389 551 728 932 11639 16 26 39 70 110 171 241 345 490 647 828 103410 15 24 35 63 99 154 217 311 441 583 745 93015 10 16 24 42 66 103 144 207 294 388 497 62020 7 12 18 31 50 77 108 155 221 291 373 46525 6 9 14 25 40 62 87 124 176 233 298 37230 5 8 12 21 33 51 72 104 147 194 248 31035 4 7 10 18 28 44 62 89 126 166 213 26640 4 6 9 16 25 38 54 78 110 146 186 233

45 3 5 8 14 22 34 48 69 98 129 166 20750 3 5 7 13 20 31 43 62 88 117 149 18660 2 4 6 10 17 26 36 52 74 97 124 15570 2 3 5 9 14 22 31 44 63 83 106 13380 2 3 4 8 12 19 27 39 55 73 93 11690 2 3 4 7 11 17 24 35 49 65 83 103

100 1 2 4 6 10 15 22 31 44 58 75 93




Exercise




Series/Parallel Array Wiring

We now look at how to arrange modules into an array. Modules can be connected in

parallel, plus-to-plus, to add current. Or they can be connected in series, plus-to-minus,to add voltage. And arrays can be built using both wiring methods.

We recommend drawing array circuits in two steps, first for clarity as a theoreticalschematic, and then for actual details as they would be installed.

Schematic Drawing

We first draw the arrangement of modules in a schematic form with boxes representingmodules placed above one another for series connections, and placed side by side forparallel connections. This is very clear visually. We can think of higher voltage as

“higher stacks” of series connected modules and increased current coming frommodules “pouring or flowing together” side-by-side in parallel.

Installed Drawing

Then we show the modules as they would be installed in actual field arrangement.There are generally two junction box arrangements used in solar modules: separatepositive and negative boxes on opposite ends of a module; or single junction boxescontaining both positive and negative terminals.

For the single junction box arrangement all the junction boxes are lined up on the same

side, so all installed drawings look similar except for the details of where the wiresconnect.

For the separate J-box arrangement installed 12-volt systems look similar to theirschematic drawing, but series wired systems of higher voltage (typically 24 and 48volts) involve “flip-flopping” the modules so that the plus and minus connections areclose together.

We demonstrate recommended wiring arrangements for both single junction box anddual J-box arrangements for 12, 24 and 48-volt systems next. For example, the SM55has separate positive and negative terminal junction boxes, while the SP75 module has

a single J-box.




12-Volt Array Wiring

The schematic arrangement of a 12-volt array involves placing the modules side by

side, with all the positive terminals connected in parallel, and all the negative terminalsalso connected in parallel. The current from each module adds to give the total arraycurrent while the voltage of the entire array is the same as the voltage of one module.

12 Volt ArraySchematic




The actual installed wiring of a 12-volt array using dual J-box modules would alsohave all the positive connections along one side and all the negative ones along theother.

12 Vo lt ArrayAs Installed

Dual J-box

The installed wiring of a 12-volt array for single J-box modules would also have all thepositives connected together in parallel, and all the negatives connected together.

12 Volt ArrayAs Installed

Single J-box




Higher Voltage Array

When we move on to systems with voltages greater than 12 volts there are two

approaches that can be taken. We can connect the modules in parallel groups first, toproduce the current that is needed, and then connect these groups in series to give therequired voltage. This is shown on the left in the figure below. In actual field conditionshowever, it may be necessary to remove modules from an operating array for repair orother maintenance work. If the array is wired as shown on the left, in parallel groups,then removing a group leaves the rest of the array without enough voltage to properlyoperate. The whole array must be shut down.

The second method involves connecting modules in series groups first to produce therequired voltage, and then connecting similar series strings in parallel to produce therequired current. This is the recommended approach. If a series group is removedfrom the circuit, the rest of the strings still operate at the nominal system voltage, andthe array continues to charge the batteries or operate the load.

Parallel Groups orSeries Groups

Buss bar

Buss bar





For a 24-volt array we need strings of two modules wired in series. Any number of such

series strings can then be connected in parallel to give the required current.

A schematic view of a 24-volt array is shown below. Strings of two modules are“stacked” vertically, indicating “higher” voltage. Current flows “up” through the lowermodule, gaining nominal 12 volts. The current from the lower module in each stringthen flows on “up” through the second module in the string where it gains another 12volts. The current out of each “top” module combines to flow on to the battery or load.The total array current returns from the battery or load to the “bottom” of the array.There the current splits into equal parts and each string draws one-module’s-worth ofcurrent.

2 4 V o l t A r ra y W ir in gS c h e m a t i c

+

-+

-+

-+

-

+

-+

-+

-+

-




The installed wiring for 24 volts using modules with separate J-boxes is shown below.The two-module strings would be “flip-flopped” alternating positive and negative J-boxes. The intermediate wires for each series string, connecting the positive of the“bottom” module to the negative of the “top” module, are shown to the right of the array.The positive outputs from the two-module strings would all combine, and the negative

inputs would also combine. Typical ground mount structures hold up to eight modules,so four 24-volt strings could be mounted on one structure.

2 4 V o l t A r ra yA s I n s t a l le d

D u a l J -b o x

The installed wiring for single J-box modules is shown again with all the J-boxes alignedalong one side. The intermediate connection within each string, from the negative of

one module to the positive of the second module, is shown.

24 V o l t A r rayA s Insta l led

S ing le J -bo x





In the case of 48-volt systems, we need four modules connected in series. Strings of

four modules can be combined in parallel to give the final array current.

48 Volt ArraySchematic

+-

+-

+-

+-

+-

+-

+-

+-




Installed field wiring for a 48-volt system with dual J-box modules would again have themodules flipped. Typical ground mount structures hold up to eight modules, so two48-volt strings could fit on one structure.


Dual J-box

Installed wiring for single J-box modules would also have groups of four modulesconnected in series.


Single J-box




Exercise




“zero volts”

Load

12

3

4

5

6

7

8 9 10

11




Wiring of Safety Equipment

Battery Bank Wiring Limit the Number of Parallel Batteries

When connecting batteries we can increase the total capacity by connecting batteries inparallel. This connects plates of the same type of material (lead or lead-dioxide) andeffectively makes a bigger battery at the same voltage. It is recommended that notmore than four parallel connections be made in a bank. Parallel strings of batteries aremore susceptible to shorter battery life due to uneven charging of different strings. Thegreater the number of parallel connections the greater the chance for advanced agingof the battery. If more than four parallel strings are needed this would indicate that

perhaps a battery cell with too small of capacity is being used. Choose a largercapacity cell so that fewer parallel strings are needed.

Battery Bank Wiring

(-)

(+)

+

-

(+)

(-)

+

-

+

-

+

-

+

-

+

-

+

-

+

-

(+)

(-)

+

-

+

-

(+)

+

-

+

-

(-)

A B C D

Max 6-8 in parallel

Connect at opposite corners




Series and Parallel Connections

To operate at a higher voltage connect batteries in series. For example, the secondbattery bank shown (B) could be four 12-volt batteries connected in series to give a finalvoltage of 48 volts nominal.

If both large capacity and high voltage are needed then connect in series and parallel.For example, the battery bank on the far right (D) has four batteries connected in seriesfor 48 volts and then three strings connected in parallel to give more capacity at 48volts.

Tap Opposite Corners

If series and parallel connections are made it is good practice to tap off the battery atopposite corners, as shown. This helps to equalize the path of the current through the

batteries and helps to insure equal charging and discharging.

Use Solid Buss Bars

Another good practice is to avoid individual wires connecting parallel strings. Insteaduse a large copper bar drilled with holes for the parallel connected terminals. The solidcopper bar gives excellent conductivity and reduces the possibility of corrosion or looseconnections at the terminals.




Overcurrent Protection Devices

Whenever there are voltage sources such as batteries, photovoltaic systems, or

generators, there is the potential for dangerously large currents. A battery bank in ashort circuit condition can discharge hundreds or thousands of amps causing damageto equipment and even physical harm or death. Care must be taken to protectequipment from such high current conditions. As a result NEC requires that everyungrounded conductor must be protected by an overcurrent device, which breaks thecurrent if too much current flows. This is done by using fuses or circuit breakers in thecircuits. A circuit breaker is a switch that opens a circuit if excessive current flows. Afuse consists of a wire or metal strap that will melt (open) if excessive current flows.

Most fuses and circuit breakers have been developed for AC use and as a result areunrated for DC use. Those that are rated for DC current are in many cases undersizedfor the necessary currents when short circuits occur. As a result it is very important touse fuses and circuit breakers rated for their intended operation when used in DCcircuits. Since most PV systems may have transients surges (lightning and motorstarting), inverse-time circuit breakers (the standard type) or time-delay fuses should beused in most cases. Presented below are some important considerations forovercurrent protection, as specified by the US National Electric Code.

Overcurrent Protection

• Each source circuit must be protected by an overcurrent device on the ungroundedconductor.

• Use DC rated circuit breakers and fuses in order to meet NEC Code.

• Batteries must have “current limiting fuses” for overcurrent protection.

• All PV equipment (e.g. controllers, inverters, etc.) must have safety disconnects todisconnect it from all sources of power.

Specifying Overcurrent Protection

• Overcurrent devices must be able to handle the rated current that will open circuit.

• All fuses and circuit breakers should be in enclosures.

• Circuit breakers or fuses should not be paralleled in the same circuit.• Protection ratings of the overcurrent protection device must never exceed the

current ratings of the wire.

• Grounded conductors should never be fused or switched.




Types of Fuses and Circuit Breakers

There are a number of fuses and circuit breakers that have been UL listed for DCoperation that are commonly used in PV systems.

Manufacturer Type Voltage(volts)

ContinuousRating(Amps)

AIR(Amps)

Comments

Heinemann Circuit Breakers 65-250 15-700 5,000-25,000 These are not currentlimiting and hence can notbe used to protect otherfuses and circuit breakers.Can be used between thebattery and inverter.

Square D Circuit Breakers 48 15 - 70 5,000 These are not currentlimiting.

BussmannGouldLittlefuse

Class T Fuse -Fast ActingCurrent Limiting

125 - 300 1 - 1200 20,000 It is a fast acting currentlimiting fuse but seems toproduce good resultswhere time delay fuses arespecified.

BussmannGouldLittlefuse

RK-5 -Time DelayCurrent Limiting

125-600 0.1 -600 20,000 Most common fuse usedfor protecting from batteryshort circuit. Has timedelay and current limitingfeature with high AIR.

Continuous Ratings Of Overcurrent Devices

Fuses and circuit breakers have two current ratings:

• Continuous

• Maximum interrupt level or AIR (Amperes Interrupting Rating).

The continuous rating is the current that the device will allow to pass continuously. Forexample, a 100-amp fuse will allow 100 amps to pass continuously. The fuse will blowwhen the current exceeds this continuous level. How long the device lets the currentpass before it breaks the circuit depends on the exact design of the device and its time

delay. Many devices will allow 125% of the continuous rating pass without breaking.Currents exceeding 200% may blow the device in a few minutes. Levels exceeding500% may blow the device in a few seconds and so on. The exact performance of afuse or circuit breaker is shown in a graph of current vs. time. An example of a timegraph for a circuit breaker and a fuses is shown next.




Fuse Example

Presented below is the current vs. time characteristics of FLNR series time-delaycurrent-limiting fuses, DC rated, Class RK5. The “time-delay” models are reallydesigned to allow surge currents to flow through to loads such as motors and

compressors. They will have a slower response time curve than “fast acting” fusesdesigned for sensitive electronic equipment. The continuous current range is shownfrom 15 amps to 600 amps. For each value of fuse the amount of current that it willpass is shown compared to the time before the fuse will interrupt the current.

Example: The 200-amp rated fuse curve is shown (as the dark line). Thecurve indicates that the fuse will pass 600 amps for about 60seconds, and 1000 amps for about 10 seconds and so on. It willallow 2000 amps for only about 0.2 seconds, and 4000 amps foronly about 0.02 seconds. So at higher currents, the time to

interrupt is quite short.




Circuit Breaker Example

The current vs. time curves presented below are for Heinemann Series GJ circuitbreakers. Three ranges are shown with the top range indicating standard time delay,the middle range representing medium time delay, and the lower range illustrating short

time delay. The current values are indicated as percent of model continuous rating.The current range for GJ type breakers is from 100 amps to 250 amps. DC interruptratings range from 10,000 amps DC for the 125 VDC version to as high as 25,000amps for the 65 VDC designation.

Example: Examine the top “standard time delay” range. If we were to specifya 200 amp rated breaker, then 500% corresponds to 1000 amps.The curve indicates that the breaker will allow 500% (about 1000amps for a 200 amp rated breaker) to flow for between 2-10seconds. 2000 amps would correspond to 1000% for our 200 amp

breaker, and the current allowed at 1000% is between 0.01-0.4seconds.




Interrupt Rating of Overcurrent Devices

The second current rating is the interrupt rating known as AIR, Amperes InterruptingRating. This is a measure of how high above the continuous rating the device can

safely break current flow without being damaged by short circuit conditions. Forexample, a common AC breaker for homes might have a continuous rating of 60 ampsand an interrupt rating of 10,000 amps of AC current.

Batteries Are The Major Concern

Because photovoltaic modules are inherently current limited high short-circuit currentsfrom the solar array are not a problem. However, a single 220 amp-hour 6-volt deepdischarge battery may produce short-circuit currents as high as 8000 amps for afraction of a second. Such high currents can generate excessive thermal and magnetic

force that can cause an underrated device to burn apart. As a result the interrupt ratingis very important when designing protection for battery circuits.

Derate AC Ratings By 90%

Overcurrent protection devices such as fuses and circuit breakers, that are designed tointerrupt currents in AC systems expect the current to fall to zero every half-cycle, orabout every 8 milliseconds. The arc that is created by large currents can be easilyextinguished at that point. But with DC circuits the current does not go to zero.Overcurrent devices with interrupt ratings for AC operation must be derated

usually by at least 90% for DC operation! For example, this means that a fuse orcircuit breaker rated to interrupt 10,000 amps AC should be expected to safely interruptonly 1000 amps of DC current.

Current-Limiting Devices

Batteries can produce thousands of DC amps under short circuit conditions. A 100-Ahbattery could release up to 2000 amps of short circuit current, while a 1000-Ah cellmight be able to produce up to 20,000 amps! This level of current can overwhelm anddamage downstream overcurrent protection. This downstream overcurrent equipment

needs protection too! Specially designed fuses, called current-limiting fuses, caninterrupt large currents. These are generally classified as RK5 or RK1 (DC rated ClassJ and T fuses can be used as well).




A typical high interrupt rating for a current limiting fuse is 200,000 amps of AC current.Under DC conditions this is derated to 20,000 amps. So a current limiting fuse shouldbe connected to each battery string up to about 600-1000 Ah capacity. In this waydownstream overcurrent equipment is protected from excessively high currents. Thismeans that designers should plan to break very large battery banks into units of

maximum 600-1000 Ah. Single banks of greater than 1000 Ah will need speciallydesigned overcurrent protection systems.

Current Limiting Fuses

(-)

(+)




Exercise




Safety Disconnects

Safety disconnects or switches are placed into power systems to allow equipment to besafely installed and maintained. The NEC requires that all source circuits or voltage

sources in photovoltaic systems such as photovoltaic arrays, battery banks, andgenerators, must be able to be isolated in case there is a problem. Circuit breakers andfused disconnect switches provide a means to disconnect safely, as well as providingfor overcurrent protection. The code limits the total number of switches or circuitbreakers in a PV system to six. They are further required to be grouped and markedand readily accessible to be pulled.

All safety disconnects in photovoltaic systems must be DC rated for voltage as well ascurrent. This may involve specifying a more expensive or heavy duty version of an ACrated model. DC rated switches are marked to indicate added protection againstarcing.

Typically there are four locations where disconnect devices are needed in photovoltaicsystems. They are: (1) between the array and the charge regulator; (2) between theregulator and the battery; (3) between the battery and any DC loads or load center; and(4) between the battery and the inverter. Disconnect means may be built into smallcharge regulators or system controllers. For example, the load terminals on chargeregulators may have circuit breaker protection built -in.

Safety Disconnects

Modules

Regulator

BatteryDC Loads

Inverter & AC Loads




Safety Disconnect Design Using Circuit Breakers

There are many ways to design safety disconnect protection into a photovoltaic powersystem. One approach that is quite flexible and compact is to use DC rated circuit

breakers and pull-fuses. These components can be sized for different currents and yethoused in the same box for convenience of service and ease of installation. Allcomponents must be DC rated to be able to break the arcs associated with DCcurrents. A general schematic of a possible approach is shown below.

Circuit Breaker MethodOf Safety Disconnect Design

system groundarray ground

Reg-

Inv-

DC Load-

Array-Battery-

Array+Battery+

Reg Array+

Reg Battery+

Inv+

DC Load +

A

B

C

D

E

The charging circuit passes current from the Array + through circuit breaker A and on tothe charge regulator (terminal designated as Reg Array +). Current returns back fromthe regulator (from terminal designated as Reg Battery +) and through circuit breakerB, and then on to the Battery +. The negative current from the battery enters from theBattery -, and passes through the central negative wiring block E, where it continues on

to the Array -. This completes the charging circuit.




The load circuit is adjacent and uses common components. The current begins at theBattery +, flows through a jumper to circuit breaker C and then on to the Inverter +terminal. The current can also flow to one of possibly many DC load breakers D whereit then flows on to the DC Load + terminal of various DC loads. Current would returnfrom the Inverter -, or from the DC Load -, and connect at the central negative wiring

block E, where it would then flow back to the Battery -. This would complete the loadcircuit.

The flexibility of this method is that the charging circuit breakers and the load breakersdo not have to be of the same size. For example, the array might need only a 30-ampbreaker while the inverter might need a 200-amp breaker and the DC loads might use a100-amp breaker.

The dashed lines represent the equipment grounding wires. All equipment grounds arecollected at one central point, the central negative wiring block E. This includes thegrounding wire from the array ground. This grounding wire must come all the way in

from the metal mounting structure and the module frames in the field. All negativeelectrical conductors also connect at E. Thus this point becomes the one central pointin the circuit where electrical and equipment grounds are bonded. A main systemground wire then connects from this central connection point to a ground rod thatmakes the firm ground to earth.




Example of Circuit Breaker Method:

“Power centers” are now commercially available that have all necessary means forarray, battery, and load circuit protection built-in. An example of the schematic wiring ofa Powercenter from Ananda Power Technologies Inc. is shown. The system uses large

current limiting fuses in a dual pull-out design as the disconnecting means between thearray and battery (top fuse in Pull Out), and between the battery and inverter and DCload breakers (bottom fuse in Pull Out). This is a compact and low cost method forincluding both overcurrent protection (with current limiting strength) and disconnectingmeans in one space-saving unit.

The Array Mercury Contactors are controlled by charge regulator circuit boards. Therelays open the array charging circuit to prevent battery overcharge, and stay open atnight to prevent leakage from the battery through the array.




Safety Disconnect Design Using Fused Disconnect Switches

Another approach to safety disconnect design is to use fused switches. These arelarge industrial grade knife switches with integrated fuse holders. The fuse acts as the

overcurrent protection and the knife switch acts as the disconnect means. Again thesemust be DC rated heavy-duty models to handle the array, battery and load DC currentsof photovoltaic systems.

The difficulty with this approach is that switches of different current levels may not beable to fit in the same box. Separate disconnect boxes may be needed for the arraycircuit and the load circuit, as shown above.

The charging circuit passes current from the Array + through a fused switch and on tothe charge regulator (terminal designated as Reg Array +). Current returns back fromthe regulator (from terminal designated as Reg Battery +) and through another fused

switch, and then on to the Battery +. The negative current from the battery enters fromthe Battery -, and connects to a negative wiring block A, where it continues on to theArray -. This completes the charging circuit.

A separate box services the load circuit. The current begins at the Battery +, flows intoa fused switch and then on to the Inverter + terminal. The current can also flowthrough a short jumper wire to an adjacent fused switch where it then flows on to theDC Load + terminal of the DC load center. Current would return from the Inverter -, orfrom the DC Load -, and connect to another negative wiring block B, where it wouldthen flow back to the Battery -. This would complete the load circuit.

All the ground wires from the equipment connect in to one of the negative wiring blocks,shown at block A in this drawing. A final system ground wire connects from the blockA to a ground rod driven into the earth.




Fused Switch Method

array ground

system

ground

inverterground

DC

loadground

Array-

Battery -

Array+

Battery+

Reg Array+

Reg Battery+

Reg-

Inv+

DC Load+

Inv- DC Load-

regulatorground

AB




Example of Fused Switch Method:

An example of a 24-volt AC and DC photovoltaic system using the fused disconnectswitch approach is shown on the next page. The charging currents from parallel 24-voltstrings flow into a field junction box at the base of each sub-array. Each string flows

through a blocking diode and combines into a larger wire. From the field junction boxesthe current from the sub-arrays travels the long distance to the central equipment of thesystem.

The currents are wired into separate fused switches and then on to separate chargeregulators. The currents from the regulators combine at positive block A and negativeblock B. The positive current flows on to a fused switch and then on to the positiveterminals of the battery bank. Thus the charge regulators are protected from both thearray and the battery.

The battery is shown with three parallel strings each with a current limiting fuse to

protect the downline fuses and disconnects from possible battery short circuit currents.The negative wire runs from the opposite corner of the battery bank and connects to thenegative block B. Current flows back through the charge controllers and back to thesub-arrays.

There are two load circuits. One circuit has current flowing from the battery positiveterminal through a fused switch and then on to the DC distribution box. The other loadcircuit has current flowing through a larger fused switch and then on to the positiveterminal of the inverter. The inverter AC output is connected to the AC distribution box.Negative conductors from the two load types returns to the negative block B and thenon to the negative of the battery bank.




24-Volt AC/DC system diagram illustrating fused switch approach

Controller #1 Controller #2

Inverter

(+)

(-)

ACloadcenter

DCloadcenter

(-)

(+)

+ +- -




Exercise




Grounding Details

• System and equipment grounding conductors share the same electrode.

• There should be only one single point where the system ground and the equipmentground are bonded together.

• DC and AC subsystem should share the same electrode.• Equipment grounding wire should have the same ampacity as the overcurrent

device protecting the particular circuit.

• For lightning protection at the array use system of grounding electrodes bondedtogether.

Grounding Electrode (Rod)

• Locate as close to system grounded conductor as possible.

• Connect the grounding conductor to the system on the battery side of the charge

controller at a point to the system.• Grounding electrode conductor needs to be as large as the largest conductor in the

system is and be at least No. 8 AWG.

• NEC requires that the grounding electrode be 5/8” diameter with at least 8 feetdriven into permanently moist soil.

If the solar array will be a great distance away from the central ground rod for thesystem a separate ground rod for the array structure can be driven close to the array.However there should be a bonding wire that connects this remote ground rod to themain system ground rod. This bonding wire keeps both ground rods at the same

voltage potential. If the two rods were not connected a nearby lightning strike wouldresult in a momentary surge of current into the earth that would cause large differencesin ground potential due to the resistance of the earth. The two ground rods would notboth be a "zero" potential and large induced currents will be created in the groundsystem. By having the wire connecting the two rods the resistance of the earth isovercome, and the two rods stay at the same potential and no induced current result.




System and Equipment

Grounding

Electrical equipment

Array disconnect

Equipmentground System

groundingelectrode

Array framegroundingelectrode

Bonding wire

System ground




Disconnect Arrangement forUngrounded and Grounded Systems

The U.S. NEC code stipulates that for systems above 50 volts (Voc for photovoltaictype systems), one DC conductor shall be grounded. This means that systemdesigners can choose an ungrounded configuration for 12 and 24-volt nominal systems.For ungrounded systems disconnects are required on both the positive and negativeconductors leading to and from the battery. For grounded systems disconnect andovercurrent protection are only required on the ungrounded conductor. The figurebelow shows the overcurrent and disconnect configurations for grounded andungrounded PV systems.

Battery

Batter Overcurrrent and Disconnect Re uirements

from chargingcircuits

to load

circuits

current limiting

disconnect

Grounded Systems Ungrounded Systems

Battery

fuses

switches

from chargingcircuits

to load

circuits




Lightning and Surge Protection

Solar modules mounted on the tops of buildings or high can attract lightning strikes.

Besides grounding equipment such as the array frames, system designers must provideappropriate means to deal with lightning induced surges coming into the systemconductors. Metal oxide varistors (MOV) and silicon oxide surge arrestors (SOV) areused as surge suppression devices on PV systems. Both have a breakdown voltage atwhich they dissipate energy to ground. SOV’s are preferred due to the fact that they donot draw current when they are off and if they fail, they fail in the open circuit condition.They are rated for surge currents of up to 100,000 amps at 300 volts or higher. SOV’sare commonly installed either at the array in the array field junction box or as part of thesystem controller or power center.

A simple schematic of the concept and an actual drawing of a field junction box withSOV surge protection are shown. In the schematic notice that both the positive andnegative conductors from the array are connected through SOV’s to ground. Normallythe SOV would be an open circuit and no connection would be made to ground. But incase of a large voltage surge the SOV would short circuit and the dangerous voltagespike would be shunted to ground at the array, thereby protecting the downlineequipment.

Simple Surge Protection

Diagram

From array frame

Bonding wire

to system electrode

Array grounding electrode

From array To controlsand battery

SOV's




An example of a field junction box with surge protection is presented below. Allowancefor up to eight separate array positive wires is made. Each positive connection is madeto a fuse and blocking diode and then combined into one positive buss bar. Currentflow through a field disconnect circuit breaker and then on to the positive outputterminal block. The array negative(s) would connect to the negative terminal block.

The SOV surge arrestor has connections to both the positive buss bar and to thenegative terminal block. It is also connected to the ground block. This is where thehigh lightning induced surge current would flow.

The array grounding electrode would be connected to the ground block as would thebonding wire to the system grounding electrode.




Fully Worked Example

To help illustrate the application of the rules and concepts presented in this chapter a

complete wiring exercise is presented below.

Example: Remote Cabin, DC-Only System

Array Size: Six 12 volt 75-watt SP75 ModulesBatteries: 700 amp hours at 12 volts (4 Trojan L-16)Load: 75-Watt Peak DC

Description: The modules are mounted on a rack on a hill behind the house, 50feet from the batteries and control center. The modules are wired

in parallel. Non-metallic conduit is used to run the cables from themodule rack to the control panel. A disconnect and control panelis mounted on the back porch. The batteries are in an insulatedbox under the porch within a few feet of the control panel. All theloads are DC with a peak combined power of 75 watts at 12 volts,primarily due to a pressure pump on the gravity-fed water supply.The battery bank consists of four Trojan L-16 (350-amp-hour,6-volt) deep-cycle batteries wired in series and parallel.

Module Interconnects:

Determine Short Circuit Current

The first step in sizing the wiring is to determine the short circuitcurrent of the array and then adjust for the NEC and UL safetyfactors and temperature.

The array short circuit current can be determined by multiplying thenumber of SP75 modules (six modules) by their rated short circuitcurrent (Isc = 4.8 amps):

Array Isc = number of parallel modules X rated Isc= 6 X 4.8 amps= 28.8 amps




Module Interconnects (continued):

Adjust for UL / NEC and Temperature

To be conservative, use the highest temperature rated insulation,

90° C. Assume that the back of the modules will get up to 65° C.The temperature derating for 90° C conductors at 65° C is 0.58.

Applying the NEC oversizing factor of 125% and the UL oversizingfactor of 125% and the temperature derating of 0.58 to determinethe required ampacity rating at standard conditions:

Ampacity

Required = 28.8 amps ÷ 0.58 X 1.25 NEC X 1.25 UL

= 77.6 amps

Type USE-2 (rated for 90°C) will be used to interconnect themodules. The standard ampacity of USE -2 # 8 AWG in free air is80 amps. This is sufficient to meet the above requirement.

Wire From Field Junction Box to Control Panel:

Because wire runs are typically a relatively long distance betweenthe module field junction box and control panel, voltage dropbecomes the main concern. However, it is good design to first

look at ampacity before calculating voltage drop.

Check Ampacity for the Chosen Wire

Cable in conduit exposed to 40° C temperatures is chosen for the

module to control panel runs. If RHW-2 cable with 90° C rating isselected, then the derating factor is 0.91.

Ampacity

Required = 28.8 amps ÷ 0.91 X 1.25 NEC X 1.25 UL

= 49.5 amps

This is safely covered by the chosen wire size of #8 AWG, which

has a rated ampacity in conduit of 55 amps at 30° C.




Wire From Field Junction Box to Control Panel (continued):

Size Wire for Voltage Drop

Using the 2% maximum voltage drop rule, we can refer to

Table 16-2 (Voltage Drop Table--12 Volt Nominal) for the 50 feetrun to determine wire size. The voltage drop should be sized forthe array current at peak power, not the short circuit condition.

The array current at peak power can be determined by multiplyingthe number of modules (six modules) by their Imp (Imp = 4.4amps):

Array Imp = number of panels x rated Imp

= 6 X 4.4 amps

= 26.4 amps

Using the Voltage Drop Table we see that for 25 Amps, 1/0 AWGis required for a 2% drop with a maximum one-way distance of 55feet. So we must use 1/0 size wire for this part of the circuit.

Connecting 1/0 AWG to the #8 AWG from the modules will requirea power distribution block in the field junction box.

Wiring in the Control Center and to the Batteries:

These wires are short (typically a few feet) and hence voltage dropwill not be a concern. However, they should be sufficient size tohandle the ampacity requirements. Cable type THHN size # 6AWG is used in the control center. We already have determined(see above) that #8 AWG THHN has sufficient ampacity. The 1/0cable from the negative DC input is used to the point where thegrounding electrode conductor is attached in order to meet code.

The same #6 AWG is used to connect from the control center tothe batteries. If this distance is greater than a few feet, one shouldconsider fusing right at the battery bank and increasing the size ofthe conductors from the battery to the control center.




Grounding Conductors:

Equipment grounding conductor sizing requires a number ofcalculations and tables. (See Section 250 in NEC). #6 AWG is asufficient size to handle the grounding for this system. The

grounding electrode conductor should be # 1/0 AWG, as it isrequired to be as large as the largest conductor in the system is.

Overcurrent Protection and Disconnect:

Each source circuit must be protected by overcurrent devices. Acurrent-limiting fuse in a fused disconnect switch is required bycode to protect the load and the array circuit from the batteries.

Array Circuit

The design current of the array circuit fuse must be greater thanthe array short circuit current:

28.8 amps X 1.25 NEC X 1.25 UL = 44 amps

A 50 amp RK-5 type fuse is chosen to meet this requirement.

Each component must be able to be disconnected from all sourcesof power. The charge control is between the array and the battery.Hence a 50-amp single-pole circuit breaker (cheaper than a fuseddisconnect switch) is specified to isolate the controller from the

array.

Load Circuit

To determine the protection on the load circuit, the maximum loadpower of 75 watts at 12 volts is converted to

75 ÷ 12 = 6.25 amps. The protection rating in this circuit must onlymeet the standard NEC overrating factor:

6.25 amps X 1.25 NEC = 7.8 amps.

A 30-amp fuse will meet this requirement easily.

Both fuses should have an amperes interrupting rating (AIR) ofaround 20,000 amps sufficient to withstand the short circuitcurrents from the battery under fault conditions.




Voltage Ratings:

Disconnects, fuses, and circuit breakers must have a voltage ratingof at least 1.25 times the system open circuit voltage. Standardmodule Voc is 22 volts.

22 volts X 1.25 NEC = 27.5 volts.




Wire Sizing Worksheet

An example of a worksheet for collecting information about the wiring design for a

photovoltaic system is presented on the next page.

Charge controller capacity can be calculated, including the environmental safety factorof 1.3 for array current enhancement from reflections.

Wiring between equipment can be listed and sized based on voltage drop or ampacityconsiderations.

Size inverter wire and most small connections based on ampacity. Remember tooversize any conductors of array current by the NEC and UL safety factors. Applytemperature corrections as appropriate.

Size the run from array to central wiring, and any long runs to DC loads using voltagedrop considerations. Size based on only 2% voltage drop.




Controller and Wire Sizing Worksheet

Charge Control :

Total Charge Controller Capacity = # Parallel Modules X Isc X 1.3

= [ ] X [ ] x 1.3 = ______________ amps

Charge Controller Model : __________________________________________________________________________

Charge Controller Capacity : ________________ amps Number of Controllers = ________________ and Sub-arrays

Sub-Array :

Number of Parallel Modules = __________________ Nominal Sub-Array Current = ________________ amps

Wiring :

Wire Segments Endpoints Voltage % loss Current One-way Wire Wire Ampacity FuseDistance Gauge Type or CB

Rating

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______

________________ to ________________: _______ _______ _______ _______ _______ _______ _______ _______




(End of Chapter)




CHAPTER SIXTEEN

SYSTEM WIRING 16-1

The National Electric Code and the UL Listing 16-2

Proper Wire Type and Insulation 16-3 Color Coding of Wire 16-4 Conductor Types 16-4

Consider Both Ampacity and Voltage Drop in Wire Sizing 16-6

Wire Sizing Based on Ampacity 16-7 Short DC Wiring From Battery 16-7 AC Load Wire Sizing 16-7 Inverter Wire Sizing 16-9 Conductors Carrying Array Current 16-11 Adjustment To Ampacity For Temperature 16-12

Wire Sizing Based on Voltage Drop 16-15 Size For 2% Maximum Loss 16-16 Voltage Drop Factors 16-17

Voltage Drop Tables 16-22 Voltage Drop Table -- 12 Volt Nominal 16-23 Voltage Drop Table -- 24 Volt Nominal 16-24 Voltage Drop Table -- 36 Volt Nominal 16-25 Voltage Drop Table -- 48 Volt Nominal 16-26 Voltage Drop Table -- 120 Volt Nominal 16-27

Series/Parallel Array Wiring 16-29 12 Volt Array Wiring 16-30

Higher Voltage Array 16-32 24 Volt Array Wiring 16-33 48 Volt Array Wiring 16-35




Wiring of Safety Equipment 16-39 Battery Bank Wiring 16-39 Overcurrent Protection Devices 16-41

Safety Disconnects 16-48 System And Equipment Grounding For Photovoltaic Systems 16-57 Disconnect Arrangement for Ungrounded & Grounded Systems 16-60 Lightning and Surge Protection 16-61

Fully Worked Example 16-63

Wire Sizing Worksheet 16-68



Siemens Solar Basic Photovoltaic Technology 16-1 Wiring

Chapter 16 – Answers System Wiring

Refer to the National Electric Code or the Photovoltaic Power Systems and the National Electric Code, Suggested Practices.

The inverter has a continuous output of 2400 watts. A nominal voltage of 24 voltsmeans that the input voltage could be as low as 22 volts (see table in the text). Givenan efficiency of 85%, we calculate the maximum input current as:

Max Inverter Input Current = 2400 Watts.85 X 22 volts

= 128.3 amps

Using the 125% safety factor from the NEC, our wire needs to have an ampacity of atleast:

Ampacity = 128.3 X 125% = 128.3 X 1.25

= 160.4 amps

Per Table 16-4, a conductor made of 14 AWG wire and type THHN (75 °C) insulation is

rated for not more than 20 amps. At an ambient temperature of 40 °C, this value needsto be derated by 0.88 as shown in Table 16-6. We also need to divide by the NECsafety factor:

Derated Ampacity = Ampacity at 25 °C X Temperature Correction FactorNEC Safety Factor

= 20 X 0.88 = 14.1 amps1.25 (NEC)




The current flow required by the loads is equal to:

Load Current = Qty X Watts = 4 X 40Volts 12

= 13.3 amps

Since the ampacity under these conditions is greater than the load current, theconductor can be used to carry the load current. The answer is then "Yes."

The temperature derate for 90 °C conductors at 40 °C is .91. Using the array Isc, wecalculate the minimum ampacity of the wire as:

Minimum Ampacity = 19.2 amps X 1.25 NEC X 1.25 UL

0.91 Temperature Factor

= 32.9 amps

With a 75 °C conductor, the temperature derate for a 40 °C ambient is .88. Theminimum ampacity of the wire is:

Minimum Ampacity = 19.2 amps X 1.25 NEC X 1.25 UL0.88 Temperature Factor

= 34 amps

We first evaluate the voltage drop only, using a current of 6 amps and a total distanceof 20 X 2 = 40 feet. Based on a 12-volt system, the acceptable voltage drop is 2% of12 volts or .02 X 12 = .24 volts.


= .24 volts

6 amps X 40 feet

= 0.001

Referring to Table 16-7, we see that the first conductor size with a voltage loss factorless than this number is 8 AWG.




Next looking at the ampacity, we calculate the minimum required ampacity as:

Minimum Ampacity = 6 amps X 1.25 NEC = 7.5 amps

Using THHN (90 °C) in conduit, we find using Table 16-4 that 14 AWG has an ampacityof 25 amps, easily accommodating this current. This is a case where the wire size willbe determined by the voltage drop rather than just the ampacity.

We will first evaluate the voltage drop. The current is 2 X 40 Watts ÷ 24 volts = 3.3amps. The total distance is 35 X 2 = 70 feet. Based on a 24-volt system theacceptable voltage drop is 2% of 24 volts or .02 X 24 = .48 volts.


= .48 volts3.3 amps X 70 feet

= 0.0021

Per Table 16-7, we see that the first conductor size with a voltage loss factor less thanthis number is 12 AWG.

For the minimum ampacity required, we calculate it as follows (there are notemperature derates):

Minimum Ampacity = 3.3 amps X 1.25 NEC = 4.1 amps

Referring to the values of THHN (90 °C) in conduit in Table 16-4, we see that 14 AWGhas an ampacity of 25 amps, easily accommodating this current. Again, the voltagedrop in this case will determine the necessary conductor size.




We will first evaluate the voltage drop for the array. Using 5 modules in parallel the Iscfor the array will be 5 X 3.45 = 17.25 amps. The total distance is 50 X 2 = 100 feet.Based on a 24-volt system the acceptable voltage drop is 2% of 24 volts or .02 X 24 =.48 volts.


= .48 volts17.25 amps X 100 feet

= 0.00028

Per Table 16-7, we see that the first conductor size with a voltage loss factor less thanthis number is 4 AWG.

We next look at the minimum ampacity required. Remember, for an array circuit, weneed to use both the NEC and UL safety factors:

Minimum Ampacity = 17.25 amps X 1.25 NEC X 1.25 UL = 26.9 amps

Referring to the values of THHN (90 °C) in conduit in Table 16-4, we see that 12 AWGhas an ampacity of 30 amps.

PointCurrentFlowing Voltage

1. 12 amps 0 volts2. 3 amps 0 volts3. 6 amps 0 volts4. 3 amps 0 volts5. 3 amps 12 volts6. 3 amps 24 volts7. 3 amps 36 volts8. 3 amps 48 volts

9. 3 amps 48 volts10. 9 amps 48 volts11. 12 amps 48 volts




Referring to Figure 16-14, we first identify the curve for a 60A fuse. This is the thirdcurve from the top. We next find the time 0.1 seconds across the top and move downto where this intersects the 60A fuse curve. Reading across horizontally, we find this isabout 900A of overcurrent.

An overcurrent of 400 amps for a 100-amp breaker represents a 400% load. Referringto Figure 16-15, we find the 400 percent load on the bottom and then move up tostandard time delay range (top range shown). This indicates a time between 4-11seconds to interrupt.

If we were to use a short time delay breaker (bottom range), the time is .03 and .2seconds to interrupt.

Again looking at Figure 16-15, we find the 300% load at the bottom and move up untilwe reach the 5 second mark. This occurs within the medium time delay range.

Refer to the appropriate diagram.

Refer to the appropriate diagram.



Siemens Solar Basic PV Technology Course 17-1 System Design – Water Delivery SystemsCopyright © 1998 Siemens Solar Industries

Chapter Seventeen

Water Delivery SystemsWater pumping is a fast growing and wonderful application for photovoltaic powersystems. There is a natural match between the availability of sunlight and the need forwater. Most photovoltaic pumping systems do not use batteries thus avoiding a costlyand high maintenance component and increasing the reliability greatly. In ruralundeveloped areas there is a critical need for fresh underground water to help preventdiseases spread by using surface waters. A low maintenance photovoltaic poweredwater system can bring health and prosperity to remote villages, without the burdens ofpaying for maintenance, spare parts or fuel. In this chapter we discuss issuesassociated with photovoltaic powered water delivery systems including performancecharacteristics of different types of pumps and sizing concerns.

The focus of this discussion is on direct coupled solar powered water pumpingsystems. These systems output water in proportion to available solar insolation. Theypump more water during summer months, typically when more water is needed, andavoid the need for batteries and regulators. Such direct coupled systems may not beas “efficient” as battery coupled designs in the sense that they cannot pump during anyweather or at night. But the advantages of reliability, simplicity and low cost makethese types of systems very attractive for a wide range of applications and users aroundthe world.




Why Solar Water Pumping

Advantages of Solar Water PumpingPhotovoltaics offer many advantages to traditional water pumping technologies such asa diesel engine generator or a windmill. A properly specified PV pumping system willrequire little or no maintenance, has no fuel requirements, and will have a long lifeexpectancy particularly for the solar panels. The major disadvantage of PV is the initialcost.

Advantages Disadvantages

Solar

Pumping

• No fuel required

• Little maintenance• Environmentally Benign

• Panel life is 20-30 years

• Cost effective for smallpower demands

• High initial capital cost

• Uses unfamiliar technology• Parts may be hard to obtain

DieselEngineGenerator

Low initial capital cost

• Familiar Technology• Requires regular

maintenance and fuelbrought to site

• Requires dependableoperator and servicesupport

• Environmentally harmful

• Expensive whenconsidering life cycle cost




Unique Characteristics of Solar WaterPumping

PV water pumping technology (typically DC) has been developed around a variablepower source, the sun, as compared to traditional pumping (AC) which relies on aconstant power supply such as the grid or an engine generator. As a result PV waterpumping systems have the following unique characteristics.

1. Premium on Efficiency: Because sunlight is expensive to convert to electricity,there is a high premium to use efficient motors and properly sized pump heads inorder to keep the power requirements and hence the initial cost down for the PVsystem.

2. Use of DC Motors: DC motors are commonly used with PV for pumping to avoidthe loss of efficiency and complexity when converting DC power to AC. DC motorswork well at varying voltage and speed which is ideal for operating from a constantlychanging power source such as solar energy.

3. Time of Delivery: Direct coupled PV pumps delivery water only when the sun isout. This may require some type of water storage in order to satisfy the need whenthe sun is not out.

4. Range of Effectiveness: PV pumps are only available for a set range of conditionsof head and daily pumped volume. The upper limits for a PV pump are 200 metersof head and 400 m3 of daily pumped volume.

5. Portability: PV pump systems can be made to be portable (trailer mounted forexample). This can offer many advantages. For example, in the western USAlivestock is often moved seasonally to different pastures. Instead of havingpermanent pumps one PV pump can be moved to different well sites.




Markets and Applications

The number of solar pumping installations is estimated to be over 10,000 worldwide in

1990. Cumulative pump sales is believed to be over 5000 per year (1994) increasing ata rate of 30 to 40% per annum in the developing world. PV pumping systems are mosteconomical for small power demands typically less than 1000 Wp. They can vary fromone or two modules (about 80 Wp) for livestock watering or household use to severalkilowatts for village water pumping. This range is well matched to pumping demandscommonly found in the developing world.

The reasons for increasing pump sales are: (1) falling real prices of PV pump systemsdue to decreasing cost of modules and pumps; (2) more reliable pump technology (thefirst pumping systems on the market in the early 80’s had a high failure rate); and (3)increasing awareness and acceptance by users such as NGOs (non-government

organizations) and governments in the developing world.

There are four main markets for photovoltaic powered water pumping systems with agrowing variety of applications serving people around the world.

Village Water Supply

Solar pumping is best suited for villages because of the small quantities of waterinvolved and the high value of delivering clean water for drinking. Hand pumps might

already be used, or villagers may be simply gathering surface water and carrying it backto their village. But surface water is easily contaminated and disease control becomesa big problem. Sub-surface water pumping using photovoltaics brings both a reliableand a clean water supply to people in remote areas.

Solar pumping becomes cost competitive with diesel for village water supply whenaverage daily water requirements are less than a volume-head product of 800 m

4 (e.g.,

40 m3 /day through a 20 m head) and where average solar irradiation is greater than 2.8

kWh/m2 /day. (Source: World Bank Technical Paper #168).

If explicit water purification is needed, then solar power systems can operate them as

well. Systems that operate on the basis of reverse osmosis, UV irradiance, or chlorinecan all be operated directly by photovoltaic power systems.




Irrigation

Water can be pumped from nearby streams or canals or lakes to irrigate fields. Or itcan be draw up from even hundreds of feet. Photovoltaic systems can be permanently

mounted in the fields, be portable so they can serve multiple farmers, or be moved todifferent fields as needed.

Drip irrigation is more efficient than flood irrigation and is especially suited for hotclimates. However, it involves more equipment and is more expensive. Photovoltaicpower systems can operate the pumps as well as the pressurizing systems to passwater through filters that are often needed for drip systems.

In general solar water pumping for irrigation is not competitive when compared to dieselpumping. An irrigation system usually requires a large amount of power for a shortseasonal period, whereas solar water pumping has its greatest advantages with smallscale year-round use.

However, there are niches for PV pumping with such applications as micro irrigation forvegetable gardens and plots of less than 1 hectare. As a general rule solar pumpingsystems for irrigation are most cost effective compared to diesel pumps where the peakdaily water requirements (volume-head product) are less than about 250 m

4 and where

insolation is greater than 4 kWh/m2 /day.

Consider solar water pumping if the following conditions are met:

Volume-Head

Product (m4)

Average Daily

Solar Insolation(kWh/m2 /day)

Village Water Supply Less than 800 Greater than 2.8

Irrigation Less than 250 Greater than 4.0

Source: Solar Pumping: World Bank Technical Paper #168




Livestock Watering

Livestock watering using photovoltaics has been a high value application in developedcountries. The cost of maintaining utility lines to pumps in remote areas is often greater

than the total revenue for the power to run the pump. As a result there is a movementin the US utility industry to offer photovoltaic pumps as a substitute for extending gridpower to such small and infrequently used loads.

Systems can be installed in remote ranch areas to provide a reliable water supply withlittle maintenance and noise. Water can be pumped day and night with the use ofbatteries. Or systems can be designed to work without batteries and accumulate thewater pumped during daylight in storage tanks, which then can deliver water day ornight.

Interesting applications such as pond aeration are also possible. Modules operatepumps that bubble water into ponds to mix up temperature gradients and prevent orpostpone freezing.

Residential Needs

Homes have a variety of pumping needs including submersible pumps for wells,pumping water to pressure lines, and water for irrigation and landscape maintenance.A variety of specialized DC pumps for use with photovoltaics have been developed forthis purpose.

PV pumping for homes is typically a part of a larger residential solar electric powersystem. In many cases standard AC pumps are used because they use the AC poweravailable from the inverter in the system.




Typical Usage Requirements

The following are some suggestions for water requirements in developing and

developed regions of the world. They are meant only as a rough guideline to help givea sense of what might be required for different applications.

Typical Human Water Requirements in Developing Areas (per person per day)

Region Gallons (Liters)

Africa 4 - 10 15 - 35Southeast Asia 8 - 20 30 - 70Western Pacific 8 - 25 30 - 95

Eastern Mediterranean 10 - 22 40 - 85Europe(Algeria, Morocco,Turkey)

5 - 17 20 - 65

Latin America, Caribbean 20 - 50 70 - 190

World Average 10 - 24 35 - 90

Typical Animal Water Requirements (per animal per day)

Animal Gallons Liters

Horse 12 45Cow (dry) 15 57Cow (milking) 40 150Beef Cattle 12 45Hog 4 15Sheep 2 10Turkeys(per 100) 13 50Chickens(per 100) 4 15

Hens(per 100) 8 30

Rule of thumb 10 gpd per 1000 lbs. 37 lpd per 450 kg.




Breakdown of Daily Use (per day)

Use Gallons Liters

Developing Areas

Survival 0.6 2Public Hydrant 11 40Single Home Faucet 11-16 40-60Multiple Connectionsw/bath, toilet, sink

45-65 80-240

Industrialized Areas

Personal (includingcooking)

10 38

Shower 25-60 95-225Bath 30-35 115-135Sink 1-2 4-8Washing Machine 30-50 115-190Dish Washer 10-20 38-75




Battery Powered Or Direct Couple To Array

There are basically two types of PV powered water delivery systems: (1) pump andmotor directly connected to an array (or using a maximum power tracking circuit); and(2) motor connected to battery bank charged by a solar array.

The advantage of direct coupled water delivery systems is the underlying simplicity ofthe design. There is more water pumped during summer periods of high insolation,which is usually when more water is required, especially for livestock watering andirrigation. During periods of low insolation there is less water pumped, but there is lessevaporation as well. The system has few components to fail and the initial cost can bekept low. As photovoltaic and maximum power tracking technologies have becomemore efficient and reliable over time the trend worldwide favors installing direct coupledpumping systems over systems using batteries.

Battery Power orDirect Coupled to Array?


DirectCoupledto Array

•Simplicity•Reliability•Low maintenance•Low initial cost•Match to season

•Low efficiency if noMPT installed•Requires storage tanksfor 24-hour availability

BatteryPowered

•Predictable supply•Higher efficiency•Supply starting surgecurrent

•Maintenance•Complexity•High costs over time•Charge control failure




One of the disadvantages of direct coupled systems is the inefficiency of the matchbetween a DC motor and a PV array during periods of low irradiance. The array can bedesigned to have the motor operating at the array Vmp during the peak hours of theday, as discussed previously in the chapter on Output Curves. This means that eachmodule operates near 15 volts instead of being held back to around 12 for battery

loads. But during the morning and evening hours or days of low insolation the array willoperate the motor far below its optimum voltage and away from the array Vmp. Thearray may be producing power, but it is not being efficiently passed to the load. Theusual solution for this problem is to include a maximum power tracking device betweenthe array and DC motor. It operates the array at its maximum power voltage andconverts the power into the most useful voltage and current for the motor. Of course itadds somewhat to the initial cost, has some inherent inefficiency (usually 95%), andadds some complexity back to the system. But batteries and charge controllers are stillavoided. And the overall efficiency of batteries (about 80% overall) is still lower thanthe efficiency passed through a typical maximum power tracker.

Another disadvantage of direct coupled systems is the need for some form of waterstorage if water is required at night or during low insolation days. This meanseffectively higher initial costs, and perhaps some maintenance costs associated withmaintaining the tank and its support structure. But it can be argued that it is moreeffective to store water than to store electricity in batteries.

The advantage of battery coupled systems is just the opposite. All the array energy ispassed to the batteries, and can be extracted during day or night. The motor can beoperated at its optimum voltage all the time, not just the peak hours of the day, andeven on overcast days, thus prolonging motor life and improving system efficiency. Andanother key feature of battery coupled systems is the ability to supply a high surgecurrent to start difficult motors or counter a high torque portion of a pumping cycle.

The key disadvantage of battery coupled systems is that batteries are involved at all. Ifthey are treated well they can last for many trouble-free years. However they areadversely affected by high temperatures, will require cleaning of the terminals andwater level maintenance, age with time, and can sulfate, corrode, and fail. Some formof charge control is required to prevent overcharge, and a motor controller is necessaryto tell the motor when to operate and when to shut off. This all adds up to a morecomplex system, with an associated higher probability of some sort of failure.

Which type of system is best for a particular situation depends on comparisons of initialcost, costs over time, reliability, efficiency and availability.




Terminology

The most critical pieces of load information for doing water system sizing are the total

vertical lift, or “head”, and the volume of water per day required. Other details will helpto fine tune the design, but these two pieces of information are critical.

Flow: The rate at which water is delivered by the pump, usually measured ingallons/minute or liters/second.

Volume: The total amount of water needed daily. Usually given in gallons/day or cubicmeters/day.

Suction Head: Vertical distance from surface of water to center of pump when pump islocated above water. There is no suction head for a submerged pump.

Discharge Head: Vertical distance from center of pump to surface of storage tankwater or point of free discharge.

Static Head: Vertical distance from surface of water to surface of storage tank water orpoint of free discharge.

Static Head = Suction Head + Discharge Head

Pressure Head: If a final discharge pressure is desired, the pump must be able tosupply the flow with the needed energy. The Pressure Head is the equivalent feet ofhead the pump needs to be able to pump to supply this final pressure.

Pressure Head (feet) = Pressure (psi) x 2.31

Friction Head: There is a loss of energy as water moves through a pipe. The smallerthe pipe diameter and the faster the flow, the greater the loss. The pump must be ableto supply enough energy to overcome these losses.




The Friction Head is the equivalent distance the pump must be able to push water tohave enough energy to overcome these losses. Tables of friction loss coefficients fordifferent size pipes, different types of pipe, and for fittings, are presented in theAppendix. Factors are available for fittings, elbows, as well as for different pipe types(steel, plastic, copper) and for different pipe diameters.

Draw Down: When the pump draws water from the well, the level of the surface maydrop depending on the ability of the surrounding earth to replenish the well. The drawdown is the distance from the surface of the water when it is being pumped to thesurface level when it is static. This can amount to many tens of feet, and depends onrate of pumping. The higher the rate, the greater the draw down.

Pumping “Head” Components

Hf friction head

Hp pressure head

(2.3 X psi.)

Hd discharge head

Hs suction head

Hdr drawdown

surfacemountedpump/motor

Total Dynamic Head: The final total head the pump must be able to deliver at thedesired rate of pumping, including all the previous heads.

Total Dynamic Head = Suction Head + Discharge Head +Pressure Head + Friction Head + Draw Down




Calculating Total Head

The purely vertical displacements of suction head and discharge head and drawdown

are relatively simple to determine by measurement. The other adjustments to totalhead, such as friction head and pressure head must be calculated.

Friction Head Calculation

Water flowing through straight length of pipe experiences friction, and this reduces theeffective output of a pump. The amount of loss depends on the diameter of the pipeand the flow rate through the pipe. It is as if the pump has to overcome more verticalfeet of lift, even though the water may be moving horizontally. The friction losses havebeen translated into effective vertical feet of lift or head in various tables presented

next. The losses are presented as equivalent vertical feet per 100 feet of actualpipe length.

A component of friction loss is the turbulence and friction caused by fittings, elbow joints and other connections. Their energy reducing effect has also been converted intoequivalent vertical feet of head in various tables presented next. The losses arepresented as equivalent vertical feet.

Friction head is computed in two steps:

(1) First the equivalent feet of head due to fittings is calculated. Then this length isadded to the actual length of pipe, to give the total equivalent length of pipe.

(2) The friction loss equivalent feet of head is then computed for this total length, andadded to the true vertical displacements of suction and discharge head and drawdownto give the final total head of the system.




Size of Fitting, inches 1/2” 3/4” 1” 1 1/4” 1 1/2” 2”

90o Ell 1.5 2.0 2.7 3.5 4.3 5.5

40o Ell 0.8 1.0 1.3 1.7 2.0 2.5

Long Sweep Ell 1.0 1.4 1.7 2.3 2.7 3.5

Close Return Bend 3.6 5.0 6.0 8.3 10.0 13.0

Tee--Straight Run 1 2 2 3 3 4

Tee--Side Inlet / Outlet 3.3 4.5 5.7 7.6 9.0 12.0

Glove Valve Open 17.0 22.0 27.0 36.0 43.0 55.0

Angle Valve Open 8.4 12.0 15.0 18.0 22.0 28.0

Gate Valve--Open 0.4 0.5 0.6 0.8 1.0 1.2

Check Valve (Swing) 4 5 7 9 11 13

Check Valve (Spring) 4 6 8 12 14 19

GPM 3/8” 1/2” 3/4” 1” 1 1/4” 1 1/2” 2”

1 4.25 1.38 .356 .11 2 15.13 4.83 1.21 .38 .10

3 31.97 9.96 2.51 .77 .21 .10 4 54.97 17.07 4.21 1.30 .35 .16 5 84.41 25.76 6.33 1.92 .51 .24 6 36.34 8.83 2.69 .71 .33 .10 8 63.71 15.18 4.58 1.19 .55 .1710 97.52 25.98 6.88 1.78 .83 .2515 49.68 14.63 3.75 1.74 .5220 86.94 25.07 6.39 2.94 .8625 38.41 9.71 4.44 1.2930 13.62 6.26 1.81

35 18.17 8.37 2.4240 23.55 10.70 3.1145 29.44 13.46 3.8450 16.45 4.67

(Equivalent vertical feet/100 feet of length)




GPM 3/8” 1/2” 3/4” 1” 1 1/4” 1 1/2” 2”

1 4.30 1.86 .26 2 15.00 4.78 1.21 .38

3 31.80 10.00 2.50 .77 4 54.90 17.10 4.21 1.30 .34 5 83.50 25.80 6.32 1.93 .51 .24 6 36.50 8.87 2.68 .70 .33 .10 7 48.70 11.80 3.56 .93 .44 .13 8 62.70 15.00 4.54 1.18 .56 .17 9 18.80 5.65 1.46 .69 .2110 23.00 6.86 1.77 .83 .2512 32.60 9.62 2.48 1.16 .3415 49.70 14.70 3.74 1.75 .5220 86.10 25.10 6.34 2.94 .8725 38.60 9.65 4.48 1.3030 54.60 13.60 6.26 1.8235 73.40 18.20 8.37 2.4240 95.00 23.50 10.79 3.1045 30.70 13.45 3.85


GPM 3/8” 1/2” 3/4” 1” 1 1/4” 1 1/2” 2”

1 6.2 1.8 .39 2 19.6 6.0 1.2 5 30 5.8 1.6 7 53 11.0 3.2 2.210 19.6 5.3 3.915 37.0 9.9 6.2 2.118 55.4 16.1 6.9 3.220 18.5 10.4 3.925 27.1 14.3 5.3 1.530 39.3 18.7 7.6 2.1

35 48.5 25.4 10.2 2.840 30.0 13.2 3.545 39.3 16.2 4.250 19.4 5.1





Example: A water pumping system is proposed to supply water to a storage tanknear a village. The piping used throughout is to be 1-inch diameter and the flow ratewill be about 10 gpm. It will draw from a nearby stream, so assume no drawdown.

The system uses a surface mounted centrifugal pump, so water is drawn up to the

pump and discharged above the pump. The total vertical lift will be 150 feet, from thetop of the stream to the top of the water in the tank (5 feet suction to pump, 145 feetdischarge from pump to water in tank).

There will be the following components:370 feet of 1 inch plastic pipe; a check valve (spring type); three 90 deg. elbow joints;two 45 deg. elbow joints/

Calculating the equivalent feet for the fittings (using Table 17-6):

90 deg. elbows: 3 X 2.7 feet = 8.1 feet

45 deg. elbows: 2 X 1.3 feet = 2.6 feetspring check valve: 8 feet

Equivalent length for fittings: 18.7 feet

Length for friction calculation: 370 feet + 18.7 feet

= 388.7 feet

Friction Head (using Table 17-7):

388.7 feet X 6.88 feet/100 feet 100

= 26.7 feet

So the total dynamic head will be

suction head + discharge head + friction head

5 + 145 + 26.7 = 176.7 feet




Types of Motors For PV Pumping SystemsThere are three types of motor types used in photovoltaic powered pumping systems:(1) brushed (Permanent Magnet) DC motors; (2) brushless (Permanent Magnet) DCmotors; and (3) AC motors.

The most important distinction is between AC and DC motors. DC motors can becoupled directly to the array producing the most efficient and simplest systems. ACmotors require an inverter, thereby increasing complexity and adding inefficiency.However, AC motors are more available, cheaper and more familiar to pump peoplearound the world. AC motors are typically used in larger photovoltaic pumping systemswhile DC motors are best used in smaller systems.

The DC pumps used for PV applications are generally of the permanent magnet type.In a conventional DC motor a magnetic field is produced electromagnetically by the fieldwindings. In a permanent magnet motor, the magnet produces the magnetic field.Thus no power is consumed in the windings leading to higher efficiencies, which isalways the preference in PV systems.

No one type of motor dominates the marketplace, as each one has certain advantagesand disadvantages, as presented in the next table.




Motor Type Advantages Disadvantages Special Features

Brushed DC

The traditional DCmotor in whichbrushes conductcurrent into thespinning portion ofthe motor.Permanent magnetsproduce a magneticfield inside the motorshell

• Simplest and

most efficientmotor for usewith PV systems

• No complexcontrol circuitry.Motors startwithout a greatcurrent surge andwill run slowly butnot overheat withreduced voltage.

• Brushes need

replacingperiodically(Typicalreplacementinterval, 2000 to4000 hr. or 2years)

• Requires

maximum powertracker foroptimumperformance

• Only available insmall motor sizes

• Increasingcurrent (byparallelingmodules)increases torque,Increasingvoltage (by seriesmodules)increases speed

Brushless DC

High technologymotor using acomplex electronicsystem to preciselyalternate the currentcausing the rotor tospin

• Efficient• No maintenance

required

• Electroniccomputation addsextra expenses,complexity,increasing risk offailure

• In most cases, oilcooled, can’t besubmerged asdeep as watercooled AC

• Growing trendamong PV pumpmanufacturers touse brushlessDC, primarily forcentrifugal typesubmersibles

AC Motors

Inductive motors

• No brushes to

replace• Can use existing

AC motor/pumptechnology, whichis cheaper andavailableworldwide. Canhandler largerpumpingrequirements.

• Requires an

inverter tochange DC intoAC adding costand complexity

• Less efficientthan DC units

• Prone tooverheating ifcurrent is notsufficient to startmotor, or ifvoltage is too low

• Can be single or

three phase• Inverters are

designed to varyfrequency tomaximize powerto the motor inresponse tochanging lightlevels




Types of Pumps For PhotovoltaicPumping Systems

There are two general classes of pumps: (1) centrifugal; and (2) positive displacement.Both centrifugal and positive displacement pumps can be further classified into thosewith motors that are (a) surface mounted and those, which are (b), placed into the water(“submersible”).

Centrifugal pumps have blades or impellers that rotate at high speed, creating apressure and forcing water to flow. Positive displacement pumps move a volume ofwater through a distance by means of a plunger or cavity and then displace that waterwith another quantity of water behind it, and so on. Pumps in both classes haveadvantages and disadvantages, which are discussed next.

Pump Types

Centrifugal

Surface

Straight orSelf-Priming

Jet Pump

VerticalTurbine

Centrifugal

Submersible

Submersible

PositiveDisplacement

Surface

HelicalCavity

Jack Pump

PositiveDisplacement

Submersible

Diaphragm




Centrifugal Pumps - Surface

In general centrifugal pumps are good for moving large volumes of water. The highspeed of the impellers makes moving gritty water a problem though. The performance

of a centrifugal pump connected directly to a photovoltaic array is very sensitive to thepeak irradiance. As the irradiance goes down the current from the array goes downand the motor rotates more slowly. A change in the speed will be directly proportionalto a change in the flow rate, but will have a geometric effect on head capability. That isa 20% reduction in irradiance will result in 80% of original current and 80% of originalflow rate capability, but will result in (.80) x (.80) = .64 or 64% of original head capability.Small changes in irradiance will result in large changes in the output of the pump, andmay lead to the pump not meeting the head requirement and effectively "shutting off"until higher irradiance returns.

Straight CentrifugalInstalled above the water, it has a single impeller. Suction head is limited byatmospheric pressure to approximately 20 feet, but can push above pump (dischargehead) hundreds of feet. Must be "primed" (filled with water) before each start-up.Commonly used for surface water pumping, from streams, lakes, or to move wateralong the land through pipelines.




Self-Priming Centrifugal

Has a chamber above impeller which holds water after shut-down so the pump is"primed" before each start-up.

Jet Pump

Centrifugal with a portion of the flow returning to a venturi on the input side. Thisincreases suction head to 150 feet but with a reduction in net flow. The venturi can beplaced just in front of the impeller chamber, or all the way down at the input of thesuction pipe. Low cost solution for relatively low flow rates and heads. Low efficiencyhowever.

Vertical Turbine

Series of impellers in one long narrow cylindrical casing that is submerged below water

level, connected to the motor on the surface by a long drive shaft. Allows for deeppumping at high rates by using huge motors above ground. Heads limited by shaftlength. Efficiency is reduced due to twisting, friction, vibration, and weight of shaft andbearings. Typically used for large-scale irrigation with large AC or diesel motors.




Centrifugal Pumps - Submersible

Submersible: Similar to vertical turbine but has waterproof DC or AC motor connected

directly to pump, submerged below water level. Pump and motor assembly issuspended by the pipe that carries water to surface. No drive shaft means can pumpfrom great depths (1000 feet).




Positive Displacement Pumps - Surface

In general positive displacement or volumetric pumps are good for meeting large headrequirements with small or moderate volumes. The rates are not as high as for

centrifugals, but the use of total available solar energy is generally better. Smallchanges in the irradiance on an array will slow the motor but will not reduce its ability tomeet the head requirement as much as for centrifugals. This is because it is stillpushing one small unit of water after the next, just a bit slower. So once a positivedisplacement type pump has pumped enough to meet the head requirement, it willcontinue pumping all day long. Generally more of the solar energy over the day will bedirectly translated into volume of water pumped.

Helical Cavity: Helical rod rotates inside a cavity of slightly different pitch, creating avertically moving cavity. Submerged under water and connected to motor aboveground by long drive shaft. Motor power is lost in friction, weight, vibration and twistingof drive shaft and bearings. Requires high torque, may necessitate batteries ormaximum power tracker to supply surge to overcome tremendous friction. Can movevery gritty water, whereas high-speed centrifugals would be eroded quickly. Suitable forlow flow rates and moderate depths.




Jack Pump: Motor pulls on oscillating jack above ground which pulls on a long driveshaft connected to a plunger with flapper valve below water level. Each cycle of the

jack moves a volume of water into the drop pipe, and displaces water previouslypushed into the pipe. Low flow rates achieved but great depths possible. Flapper valveneeds periodic replacement. Above ground mounting means can use wide variety of

either AC or DC motors.

A variation of the jack pump design has motor and small jack mechanism packaged innarrow cylindrical casing like submersible. Pump and motor can be submersed belowwater level.




Positive Displacement Pumps –Submersible

Diaphragm Pump: The motor rotates an uneven cam, which opens and closes valves.They produce low to medium flow rates and operate to medium to high heads (fewhundred feet). They can be built as a surface mounted design or in a waterproofsubmersible design. Many commercially available small diaphragm type pumps forpersonal water supply operate at 24 volts needing only 2-4 modules.




Pump Use Advantage Disadvantage

SurfacePositiveDisplacement- Rotary orMono pumps(HelicalCavity)

• Medium to high-head, medium-flow,Mono or Moynopumps

• Head: < 150 m

• Very robust

• Efficient over widerange of head exceptfor under 20 meters

• Simple construction• Self priming

• No back-flow valverequired

• Sand or very hardwater can causepremature degradationof rubber stators

• Requires gearing• Can overload motors if

downstream valvesare closed

• Installation is difficult

• Requires batteries orpower-conditioning tosupply high startingtorque (current)

SurfacePositive-Displacement

-ReciprocationPiston(Jack orNoddingDonkey

• High-head, low-flow,down hole pistonand cylinder drivenby sucker rod fromsurface

• Head: < 150 m

• Efficient over a widerange of motorspeed

• Can pump fromgreat depths

• Simple design, easyto repair

• Maintenance requiresperiodic replacementof leathers andcylinder, must removepiston and cylinderfrom well to serviceseals

• More expensive thancentrifugals of samesize

• Cannot tolerate sandor sediment

• Efficiency decreasesas piston seals wear

• Requires batteries orpower-conditioning to

supply high startingtorque (current)

SubmersiblePositiveDisplacement(Diaphragm)

• Medium head, lowflow applications

• Head: < 80m

• Few moving parts

• High efficiency overthe range of head

• Low internal friction• Tolerant of sand and

other particles

• Serviceable withhand tools

• Small size makesthem easy to removefrom wells

• Can be used inportableapplications.

• Low capacity

• Not appropriate fordeep wells

• Diaphragms must bereplace every 1-2years

• Most designs usebrushed DC motorrequiring regular brushreplacement

• Limited to heads

below 100 meters• Not all models can be

rebuiltAdapted from “Pump Selection, Water and Sanitation for Health Project”, US AID p. 42




1

10

100

1 10 100

Daily Volume (m3/day)

T o t a l H e a d ( m )

Optimal Operation Ranges forDifferent Types of Pumps

Positive Displacement

Submersible

Surface Suction

CentrifugalHandpump




Pump Type Comparison

Centrifugal

High speed impellers

Large volumes

Moderate head

Loss of flow rate with higher head

Low irradiance reduces ability toachieve head

Possible grit abrasion

Positive Displacement

Volumetric movement

Lower volumes

High head

Flow rate less affected by head

Low irradiance has little affect onachieving head

Unaffected by grit




Exercise

Model Type Head Flow Rate Power Needed




Primary RequirementsFor Choosing A Pump

The first requirement that any pump must satisfy is the total head. If the head cannotbe reached then no water is pumped at all. If possible the total dynamic head shouldbe know including friction losses in the pipes and any drawdown that will occur. Oftenall this information is not available. In that case certainly at least the suction anddischarge head must be known.

The second requirement is that the pump be able to deliver the required volume ofwater in a typical day to the specified head. Direct coupled solar pumping systems willoperate only during daylight, and the output will vary with the daily insolation. A simpleapproach to choosing a pump begins with estimating the amount of “peak sun hours”available at the site, and dividing this into the total daily volume needed. This gives therate of pumping during “peak sun hours” that the pump must perform. Choose a pumpthat can operate at the required head and that can output at this rate. The powerneeded by the pump to operate at this rate and this head will then determine the size ofthe solar array.

(1) Required Total Dynamic Head (including friction and drawdown)

(2) Required Pump Flow Rate = Daily Volume Required Peak Sun Hours

Factors for converting insolation values into “peak sun hours”:

Unit of Daily InsolationMultiply measured insolation by this

factor to convert to peak hours...

kwh / m2

1.0

Langley (cal/cm2) 0.01162

MJ / m2

0.2777

Btu / ft2

0.003155




Supplementary ConsiderationsFor Choosing A Pump

Site, Well, and Water Considerations

The diameter of the borehole must be known because it can restrict your choices forpump models. Typically submersible pumps are available in 4 inch or 6 inch diametersor in larger sizes. A system designer who chooses a large pump on the basis ofmeeting the volume demand must also confirm that the pump will indeed fit in the wellin the field!

The flow capacity of the well should be known so that the pump does not “over pump”the well. If the rate of pumping far excess the recovery rate of the well from thesurrounding ground then the water level will continually drop during pumping, and thepump could be exposed to air and pump in dry conditions. This might harm the pump.If more volume is needed than can be delivered by a direct coupled pumping systemoperating quickly during sunlight hours, then perhaps a battery coupled system shouldbe considered. In such a design a slower pump rate could be sustained 24 hours a dayperhaps staying under the natural recovery rate of the well.

Abrasive particles in the water are another consideration in choosing a pump type.Very gritty water can cause high-speed impellers in centrifugal types to erode quickly.

Positive displacement type pumps should be considered because they are less affectedby grit. Or perhaps manufacturers of centrifugal pumps could supply an appropriatefilter to reduce the grit effect. Such filters will probably reduce output performance tosome degree.




Electrical Considerations

One electrical consideration is whether the system should be direct coupled to modules

or battery coupled. The usual trade-offs are simplicity, reliability, low maintenance, andlower cost for a direct coupled system against 24-hour availability of water and highermotor operating efficiency with batteries.

Another choice is whether a fixed array mount should be used or a sun position tracker.Fixed structures are less expensive and can withstand extremely high winds. Buttrackers can increase water output by 50-60% during summer months. The reduction inthe number of modules needed could equal or exceed the cost of the tracker.

If a pump model is chosen that uses an AC motor then an inverter will be required.Most common commercial inverters require a battery bank for stable and narrowvoltage window. (Grundfos by way of contrast makes a special inverter designed tooperate directly connected to the solar array).

If a pump with a DC motor is chosen then the consideration is whether to use a brush-type commutated motor or a brushless electronically commutated design. Brushes willhave to be serviced periodically while electronic commutation circuitry could increasethe complexity and decrease the reliability of the overall system. Details of the motorand control design must be considered.




Pump Output Curves

Just as solar modules have output curves so do water pumps. Solar module IV curves

compare current against voltage given conditions of light irradiance and temperature.Water pump curves compare flow rate against head given conditions of motor power tothe pump.

If a pump is paired with different motors or the motor is operated at different voltages,the output of the pump will be different. In AC systems often manufacturers present apump performance diagram when it is matched to different horsepower motors. Pumpsto be operated by DC motors (typical for solar pumps) often present performancecurves for the same motor operated at different DC voltages.

Pump curves show the flow rate possible from a pump at a given head. Typically, head

is listed along the vertical axis (in meters or feet), and flow rate is listed along thehorizontal axis (in gallons per minute or liters per second).

To read a typical head-flow curve begin with the head requirement and read across tothe curve and then down to find the flow rate that would be produced at that head withthat power. As an example, performance curves for the A.Y. McDonald “Solar Sub” DCwater pumps are presented on the next pages, showing performance at variousvoltages.

Derate Array Power To 80%When looking at photovoltaic coupled pump curves it is important to realize that thepower available from a solar array in field conditions will be less than the standard ratedpower of the array, due to heat and dust. Typically a derating of 80% is applied to thestandard peak power of the array to give the expected actual power available to thepump in field conditions. Modules will operate typically at about 50-60

oC. so their

power is derated by about 12-15% (recall 1/2% power loss peroC above 25

oC.).

Added to this is expected loss due to dirt and dust in agricultural settings of about 5-10%, giving a total derating of about 20%. In other words power values indicated insolar coupled water pumping literature should be increased by about 20% to give the

size of array at standard conditions needed for the performance indicated.




Example: An A.Y. McDonald model 211012DK pump is being used to pumpwater 100 feet above the surface of a well. If the pump is coupledto a nominal 45 volt array (three modules in series operating nearVmp of about 15 volts), read across from 100 feet to the 48 volt

curve (assumes each module operating at 16 volts) and then downto find a rate of 350 gallons per hour (1.3 m3 /hr). If a nominal 60

volt array were used (four modules in series), read across from 100feet to the 66 volt curve (assumes each module at 16.5 volts) andthen down to find a rate of about 630 gph (2.4 m

3 /hr).

A.Y. McDonald

211012DK

0

50

100

150

200

250

300

350400

450

500

0 100 200 300 400 500 600 700

Flow (gph)

H e a d ( f e e t )

90 V, 1190 W

75 V, 880 W

66 V, 680 W

48 V, 350 W

34 V, 170 W

Power neededat pump




A.Y.McDonald

211009DP

0

50

100

150

200

250

300

350

0 200 400 600 800 1000 1200 1400 1600 1800Flow (gph)

H e a d ( f e e t )

90 V, 1190 W

75 V, 880 W

66 V, 680 W

48 V, 350 W

34 V, 160 W

Power neededat pump

The 211009DP model of pump has similar shaped curves but different values. Thismodel of pump cannot achieve the same total head, but can pump more volume.

Centrifugal pump curves always curve slowly downward, as seen in the A.Y. McDonaldcurves. At lower head the same pump/motor will deliver much higher flow. Orconversely, as head is increased the flow rate decreases drastically. This effect ofdecreasing flow with increasing head is partially because at higher head. More of thespinning impeller's energy is used to just hold up the column of water above it. Lessenergy is available for moving the column. So flow rate goes down as head goes up.




Positive displacement type pumps show little decrease in flow rate with increasing headcompared to centrifugal pumps. This is because the column of water is held up by amechanical design of the pump (a flapper valve, diaphragm, or progressing screw, forexample) so little more energy is needed to hold up a large vertical column of waterthan a small one. But positive displacement types tend to be able to pump less volume

than centrifugals types.

An interesting shape of curve occurs for a jack-type positive displacement pump. Athigh head the nature of the curve is similar to positive displacement pumps. That is,there is little effect of increasing or decreasing head on flow rate. But at low heads the

jack-type pump seems to be able to increase the flow more than the progressing cavitytypes. The curves for the Solarjack SJA series of jack type pumps are shown as anexample. The power values given are power required at the pump.

Pump Jack Output Curves

0

100

200

300

400

500

600

700

800

900

0 2.5 5 7.5 10 12.5 15

Flow (gpm)

H e a d ( f e e t

470 w atts 660 w atts




Pump EfficiencyManufacturers may offer a pump model that can deliver to a required head at differentflow rates depending on the voltage or array power delivered. After a designer has

selected various pumps that can deliver to the specified head they must choose thebest pump among the choices. Perhaps the primary factor affecting this choice is theefficiency of the pump at that particular head. Choosing a pump that operates at ahigher efficiency will mean that fewer modules will be required, so initial cost is lower.

The efficiency of a pump (including the motor) has a maximum at some value of headand flow and is less anywhere away from that point. The efficiency is determined bydividing the output power of the flow by the input electrical power to the motor.

Efficiency of Pump/Motor = Output power of flow

Input electrical power from array

Ideal efficiencies of between 25% and 60% are possible although real worldperformances will probably be limited to 40-50%. Added losses are due to heating inthe motor and friction and twisting losses in shaft type systems.

The formula for calculating efficiency divides the theoretical power needed to movewater by the actual power specified for the pump by the manufacturer.

Efficiency of Pump/Motor = Head x Flow x Fc

Electrical Power

Fc = conversion factor for units ( from MJ to kWh)

2.7 if units are: meters and m3 per hour

163.5 meters and m3 per minute

0.003 feet and gallons per hour0.189 feet and gallons per minute

You may have determined that a number of manufacturer’s pumps will reach yourrequired head and also pump enough water. Use this formula to calculate theefficiency at the required head for each of the pumps. Compare the efficiencies andthe flow rates or volumes, and choose the pump that has the highest efficiency, unlessother factors override this choice. You will then be delivering water to your headrequirement with the fewest modules per volume.




Example: Examine the 211012DK pump from A.Y. McDonald, and calculatethe efficiency for various voltage and power conditions. Pickpoints all along the curves so as to see where the efficiencyreaches a maximum.

First examine efficiency along the 48 volt curve:

Efficiency @ 150 ft. = 150 ft X 160 gph X 0.003 factor350 watts

= .206 or about 21% efficient


= .30 or 30% efficient



Next examine efficiency along the 66 volt curve:












The 211009DP pump cannot achieve as high of head as the DKpump (320 feet maximum head Vs 490 feet), but it can pump atgreater flow rates.

Examine this pump at various points along its 66 volt curve andcompare to the DK pump:







So the DP choice is more efficient than the DK pump, and both pumps show thecharacteristic peaking of efficiency in the midrange of the pump head range.

The are both about the same efficiency at 150 feet, showing that different pumps maybe similar in efficiency at particular operating points, but that over their range one modelis more efficient than another.

It is this type of analysis that needs to be done to decide which pump is the best choicefor a particular head and flow rate. Of course other must also be weighed, and pumpefficiency may be overridden if these other factors indicate a more cost-effectivesolution.




Exercise




Considerations and Calculations:Designing a Solar Pumping System

Choosing the right solar pumping system for a project is not an easy task. It requires abackground in photovoltaics and water pumping and experience working withphotovoltaic pumps in the field. Up to a few years ago there were only a few PV pumpson the market representing a narrow range of head and daily volume.

Today there are more diverse PV pumps on the market representing a broader range ofpump types and applications. Photovoltaic pumping systems are able to pump up to adepth of 200 meters and can provide up to 400 m

3 per day under optimum conditions.

The following steps are a guide to getting the right pump chosen for your specificapplication.




Step Action

1 Visit the site; determine the rough layout of the system and thelikely water source.

2 Obtain information on the well and water resources (i.e. welldiameters, depths, yields, drawdowns, seasonal river levels, andmonthly groundwater information). Consult with local authoritiesand private welldriller who have knowledge of the area. (See“Supplementary Considerations for Choosing a Pump”).

3 Estimate the likely daily ‘demand’ for water for the applicationsfor each month. Be sure to consider water supply needs in thefuture in for example a growing village population.

4 Gather monthly data on the solar insolation or “peak hours”(convert from kWh/m

2 /day, MJ/m

2, Langleys, etc.)

5 Calculate an estimate for peak flow rate, based on the average

daily insolation available.6 Determine the total dynamic head, including drawdown and

friction losses if possible.7 Choose a pump type (Positive Displacement, Submersible,

Floating or Surface Suction) that best fits the head, volumerequirements. (Recall the discussion previously on “OptimumOperation Ranges for Different Types of Pumps”).

8 Choose motor/pump systems that will deliver to your requiredhead, and that can meet or exceed your flow requirement. (Usemanufacturer’s literature and help from your pump supplier).

9 Calculate the efficiency of the pump(s) at your required head.

Choose the pump with the highest efficiency that can deliver yourrequired flow rate.

10 Estimate the array size for each month, using the Array Sizeformula presented next for this calculation, based on pumpefficiency and available peak hours of insolation.

11 Choose the largest array size as calculated above.12 Compare manufacturers “best case” performance with your array

estimation. Adjust system size as necessary. Be conservative.13 Reevaluate your decision. If you haven’t used the pump before,

get feedback from other users. Make sure that it fits otherconsiderations (i.e. computability with the water quality, ease of

maintenance, availability of support, etc.)14 Complete the system design. (See “Component Design

Considerations.”15 Procure and install system. (See “System Installation”).




Solar Pumping Array Sizing Formulas

Array Size

Use the following formula to calculate the size of solar array need to pump a givenamount of water to a given head. (This is a slightly different version of the efficiencyformula presented earlier).

Array Size (Wp) = H x V X FcP.H. x Fm x Ft x eff.

where:

H = total pumping head in meters or feetV = desired daily water volume in m

3 /day or gallons/day

Fc = conversion factor for units ( from MJ to kWh)2.7 if units are: meters and m

3 per hour

0.003 if units are: feet and gallons per hourP.H. = equivalent peak hours of daily insolation

(global solar radiation in the array plane (kWh/m2 /day))

Fm = array/load matching factor0.95 for pumps with maximum power trackers0.9 for centrifugal

0.8 for othersFt = temperature derating factor for array power loss due to heat

0.8 for warm climates0.9 for cool

eff. = pump/motor efficiency at required head (wire to water)

Forming Array into Series and Parallel Modules:

# Series Modules = Pump nominal voltage

15 volts/module

# Parallel Modules = Array Wp# Series Modules X Module Wp




Peak Flow Rate

Use the following formula for estimating the peak flow rate that the pump must deliverat the specified head to meet the daily volume requirement. (This was presented in theearlier discussion of “Primary Requirements for Choosing a Pump”).

Flow Rate = Daily Volume Requirement = VPeak Sun Hours P.H.

Factors for converting insolation values into “peak sun hours”:

Pump/Motor Efficiency

Use the following formula for estimating the efficiency of the pump at the required head.

Efficiency = H x Flow Rate x FcPump Electrical Power Requirement

Reality Check : Manufacturer literature usually quotes the efficiency as measuredunder ideal conditions, as high as 50 to 60%. Field measurements of pump efficiencyare rarely ever above 40%. A pump/motor (wire to water) efficiency of 30 to 40% is arealistic estimate based on field testing. Surface centrifugal are less efficient hence werecommend using a 30% efficiency, centrifugal submersibles are in mid-range (35%efficiency) and positive displacement (surface and submersibles) have the highest

efficiency (40%). Higher head may drive the efficiency of centrifugals even lower.

Unit of Daily Insolation P.H. = Insolation times this factor...

kWh / m2

1.0

Langley (cal/cm2) 0.01162MJ / m

20.2777

Btu / ft2

0.003155




Example: A water pumping system is proposed for a remote rural community.The projected population for the town in five year is 300 people.

Step (2) We are given the following information:

Water is in a well, and needs to be pumped a long way throughexisting piping to a elevated storage tank for the village.

Vertical distance, static water to tank inlet pipe 30 mDrawdown estimate 7 mTotal length of pipe 500 mDiameter of pipe 1.5 inch

PVC pipe

Step (3) If we assume typical consumption of 55 ltrs/person/day, then the

requirement will be 300 people X 55 l/day = 16,500 liters/day, or16.5 m3 /day, or about 4340 gallons/day.

Step (4) Local weather data is provided. The best month is June with 9.1kwh/m

2, and the lowest month is January with 5.7 kwh/m

2. Use the

lowest month to be conservative. Convert to 5.7 peak hours.

Step (5) Required peak flow rate = Volume / Peak Hours

= 16.5 m3 / 5.7 peak hours

= 2.89 m

3

/hour or 764 gph

Step (6) Total vertical head = drawdown + discharge= 7 m + 30 m= 37 m

Friction head: assume at least one 90° elbow, whichadds 4.3 feet or 1.3 m to the totalequivalent length of pipe.

Equivalent length = 500 m + 1.3 m

= 501.3 m

Use friction loss factor based on 800 gphand 1.5 inch pipe, given as 1.74 feet/100 ftor 1.74 m/ 100 m.




Friction Head = (501.3 m / 100) X 1.74= 8.7 m friction head

Total Dynamic Head = total vertical + friction heads= 37 m + 8.7 m

= 45.7 mRound to 46 m.

Step (7) A DC centrifugal submersible with maximum power tracking isproposed for the project.

Step (8) An AY McDonald 211009DP pump (with built-in maximum powertracking) is selected to be used for the project. Our total dynamichead of 46 m converts to about 150 feet. Reading across from 150feet, the 75-volt configuration yields about 800 gph (or 3 m

3 /hr),

which exceeds our requirement of 764 gph. So we will work with

this curve. The manufacturer indicates that the pump requires 880watts in this configuration.

Step (9) Efficiency = 150 feet X 800 gph X 0.003 as Fc880 watts pump power

= 0.41 or 41%

Step (10) Array Wp = 150 feet X 4340 gpd X 0.003 Fc .5.7 hours X 0.95 Fm X 0.8 Ft X 0.41

= 1100 watts

We choose to use a 75 watt module for this project.

# Series Modules = 75 nominal volts15 volts/module

= 5 modules in series

# Parallel Modules = 1100 watts5 series modules X 75 Wp

= 2.9 modulesRound up to 3 modules.

Array design: 5 series X 3 parallel = 15 total




Component Design Considerations

A typical pump installation (direct coupled) includes the following major components:

Array, Pump Controller, pump cable, pipe, grounding wire and safety rope.

Array

An important choice is whether a fixed array mount should be used or a sun positiontracker. Fixed structures are less expensive and can withstand extremely high winds.But trackers can increase water output by 30 to 40% during summer months. Theamount gained depends upon your insolation levels, the time year, and your latitude.Using a tracker could reduce in the number of modules needed, which could then equalor exceed the cost of the tracker. Trackers, though generally very reliable, do addcomplexity to the system and may need to be serviced overtime.

Pump Controllers or Power Conditioning Units for Pumps

Most PV pump manufacturers include now with their system some type of impedancematching device so that the systems will operate at optimum power (max. power point),matching the electrical characteristics of the load and the array. In these conditionsboth the motor and the array can function close to their maximum efficiency over arange of conditions and light levels. For instance, high current may be produced so thatthe motor/pump will start in low solar irradiance levels. If the pump does not come witha maximum power tracker or linear current booster it is advisable to consider oneparticularly when using positive displacement pumps, which have to work against the

full pumping head on start-up.

Inverters

There are a variety of PV pump manufacturers who have developed PV pumpingsystems using inverters to drive single or three phase AC motors in direct coupledsystems. These inverters differ from those used for battery systems in that they notonly convert DC to AC but include circuitry that involves some form of impedancematching as well. Typically these are variable frequency inverters that control thefrequency and hence voltage of an AC pump to best match the output of a PV system.At this stage they have proved very reliable and efficient, though they add additionalcost to the overall systems as compared to a direct coupled DC motor/pump.

If a standard AC off-the-shelf pump is specified, then it is necessary to have a standardcommercial inverter with a battery bank. Though less efficient, because of their fixedvoltage operation, designer can optimize the motor/pump subsystem for maximumefficiency.




Water Level Sensors, Float Switches, and Wire

External water level sensors are used in wells with uncertain or marginal capacitieswhen the pump/controller does not provide dry running protection. If the outputcapacity of the well is close to the maximum pumping capacity, sensors should be

installed to prevent damage to the pump. Some pump manufacturers, however, claimthat their pumps are not damaged by running dry. Float switches keep water storagetanks from overflowing.

Wire used to connect the pump to the array must be sized properly to minimize linelosses. This is especially true for low voltage systems (12 and 24 Volts) where a smallvoltage drop can cause significant power losses. Wire type should be selected for theapplication (i.e. water submersion, sunlight, et al). The pump cable should be carefullyspliced and sealed when attached to the pump leads to avoid corrosion.

Disconnects, Grounding, Fuses

Pumping systems, like other PV systems, need to be safe. In the US this meanspumping systems must be designed to meet National Electric Code. The majorrequirements for PV pumping systems are:

1. Proper grounding of conductors and equipment2. Disconnects on the array and all equipment3. Fuses or circuit breakers on all ungrounded conductors4. Properly sized and insulated wires.




Water Pumping System Installation

Each pump type (e.g. surface centrifugal, pump jack, diaphragm) can require a wide

range of different tools, equipment, etc. to install. As a general rule surface pumps areeasier to install and require few tools. Setting pumps down deep wells (50 meters ormore) may require specialized pump setting equipment only available from the localwell driller. The majority of PV pumping systems failures is cause by controller or pumpproblems. Proper installation will minimize the occurrence of these problems.

The most common installation problems are the following:

Varying Water Levels: The water level in a well may vary seasonally as well as hourly.Most pumps will fail if run dry. Pumps must be mounted to keep the inlet below thewater level at all times. If the replenishment rate of a well is less than the maximum

possible pump rate then a level switch should be included to keep the pump fromrunning dry.

Protect the Pump Inlet: Sand is a primary cause of pump failure. If a pump islocated where dirt and sand may be pulled into the pump, a sand screen should beused. Consult the manufacturers for recommended methods.

Ground the Equipment: Water pumps attract lightning because of the excellentground they provide. Avoid locating arrays on high spots. Consider erecting lightningrods on high ground around the pump to attract lightning away from the pump. Groundall equipment. Do not ground to the pipe because the ground would be interrupted

during maintenance.

Avoid Long Pipe Runs: Friction can significantly increase the head and thus the sizeof the PV array.

Protect the Control Equipment: All electronic control equipment should be housed inwater-resistant boxes. All wire should be approve for outdoor use or installed inconduit. Any cables submersed in water should be appropriate for those applications.

Protect the Well and Array: The solar array, well and pump (surface mounted) needsto be protected from people and animals.

Protect the Pump from Freezing: Surface pumps and pipes need to be automaticallydrained in order to prevent against freezing. A frozen pipe can cause a pump to burnout in winter conditions.




Matching Array To Pump Direct Coupled Arrays

It is not enough to connect an array with enough power to a water pump to achievepeak performance. The array must be configured so that the peak power voltage andcurrent are closely matched to the motor load curve during the day. Presented below isa comparison of the performance of a surface centrifugal pump connected to the samesize (power) array but configured in two ways: 3 in series by 10 in parallel (high voltage,low current); and 2 in series by 15 in parallel (lower voltage, higher current). This willillustrate the effect of array / pump mismatch. Later we present these two arrayconfigurations connected to the same pump using maximum power tracking and howthis helps eliminate the problem.

This simulation uses solar data for Madras, India during the month of March. The array

is made of thirty 35-watt modules (peak array power of 1050 Wp) with the array tilted to15 deg. facing south. The surface centrifugal used is the A.Y. McDonald 820309DS1.

Each simulation shows the array IV curve at 8:00 am, 9:00 am and Noon, and the totaldaily water pumped is indicated. The pump/motor load curve is also indicated to showhow the pump motor operates the array.

In the first configuration, 3 series X 10 parallel, the array IV curve intersects thepump/motor load curve at about 23 volts at 8:00 and about 31 volts at 9:00 and only atabout 32 volts at Noon. The maximum power points of the array are indicated bycrosses and are all at about 40-50 volts during these time periods. Therefore, we see

that this configuration of array is not being operated near any of its maximum powerpoints by the pump. The total output from the pump is 6836 gallons per day in Marchwith this configuration of the 1050 Wp array.

In the second configuration, 2 series X 15 parallel, the array IV curve intersects thepump load curve at 30 volts at 8:00, about 33 volts at 9:00, and about 34 volts at Noon.These voltages correspond more closely to the indicated maximum power points of thearray during these times and the pump is able to deliver 9322 gallons per day – anincrease of 36%!

The second configuration of array, with more modules in parallel for more current, is a

better match to this particular pump/motor. At Noon the 3 X 10 and the 2 X 15 arraysare delivering about the same power to the pump (660 and 673 watts respectively). Butearly in the morning (and symmetrically in the late afternoon as well) the 2 X 15 array isdelivering more current to the pump than the 3 X 10 array (12 amps vs. 9 amps). The2x15 array is operating the pump up its load curve higher and sooner than the otherconfiguration.




Direct Coupled to 3 Series X 10 Parallel Array

0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Voltage

C u r r e n t

DirectCouple

6836 gpd

8am 210 watts +

+9am 429 watts

noon 660 watts

+

Direct Coupled to 2 Series X 15 Parallel Array

0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Voltage

C u r r e n t

Direct

Couple

9322 gpd

8am 356 watts +

+9am 589 watts

Noon673 watts+




Maximum Power Tracking

The introduction of maximum power tracking (MPT) to this water pumping system

removes the mismatch problem. The peak power available from the 1050 Wp array istranslated into current and voltage that best operates the pump, independent ofwhether the array is wired at 3 series X 10 parallel or 2 series X 15 parallel.

Both configurations are able to deliver 348 watts at 8:00, 561 watts at 9:00, and 760watts at Noon. Both configurations allow the pump to produce 10,055 gallons per dayin March. Notice that this volume is greater than the amount delivered by the bettermatched direct coupled system presented above. That system configured as 2 series X15 parallel delivered 9322 gpd. The MPT is able to better match even thatconfiguration to the pump and deliver in this case about 8% more water in March. (Themore mismatched configuration of 3 series X 10 parallel is increased by 47% usingMPT!)

This illustrates the desirability of having MPT in solar water pumping systems. Theconfiguration of the array is not such a critical factor with MPT. In fact, the array can beconfigured for maximum voltage, and therefore minimum current and voltage lossesand smaller wire size and subsequent lower cost.

This is one major reason why solar pump manufacturers typically supply a MPT unitwith their pumps, or even build the MPT circuitry into the motor. The solar waterpumping market is one area where a bit of added complexity is worth the price.





3 Series X 10 Parallel Array

0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Voltage

C u r r e n t

MPT

10,055 gpd

8am 348 watts

9am 561 watts

noon

760 watts



0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Voltage

C u r r e n t

MPT10,055 gpd

8am 348 watts

9am 561 watts

noon760 watts




Sun Position Tracking

We can also look at the effect of tracking the array to face the sun and examine itseffect on pump output. Again both array configurations are shown directly coupled to

the centrifugal pump – with with no MPT. (If we examined the combined effect of MPTand sun position tracking we would again see that the final output of the twoconfigurations would be equal. The MPT would translate the available power from botharrays, each at 1050 Wp, to the pump equally well.)

Both systems show significant improvement in volume output with sun position tracking.The more mismatched configuration (3 X 10) shows a 54% increase, from 6836 to10,534 gpd, and the better matched system (2 X 15) changes from 9322 to 13,745 gpdfor a 47% increase. These increases are slightly different due to the way the array IVcurve intersects with the pump load curve.

The 3 X 10 configuration profits a bit more from sun position tracking because all theeffect of tracking translates into more current into the pump. This system is gettingmore power at 8:00 with sun position tracking than it had even after 9:00 as a fixedangle system.

The 2 X 15 configuration also profits from sun position tracking but the gain is lessbecause the pump is operating the array beyond its maximum power voltage. The gainfrom 8:00 to Noon is small.

The decision of whether to use sun position tracking or MPT with solar water pumpingsystems may depend on a number of factors, including the remoteness and harshnessof the location, the clarity of the sky (sun position tracking won’t show a large gain withovercast skies), and the cost of the two methods. Reliability and simplicity should stillbe the guiding factors in deciding how best to enhance the output of your solar array.




Direct Coupled and Sun Tracking


0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Voltage

C u r r e n t

TrackingDirect

Couple

10,534 gpd

8am 507 watts

++

9am 600 watts

noon 660 watts

+

Direct Coupled and Sun Tracking


0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70

Voltage

C u r r e n t

Tracking

Direct

Couple

13,745 gpd

8am 667 watts

++

9am 700 watts

Noon

673 watts+




(End of Chapter)




CHAPTER SEVENTEEN

WATER DELIVERY SYSTEMS 17-1

Why Solar Water Pumping 17-2Advantages of Solar Water Pumping 17-2Unique Characteristics of Solar Water Pumping 17-3

Markets and Applications 17-4Village Water Supply 17-4Irrigation 17-5Livestock Watering 17-6Residential Needs 17-6

Typical Usage Requirements 17-7

Battery Powered Or Direct Couple To Array 17-9

Terminology 17-11Calculating Total Head 17-13

Types of Motors For PV Pumping Systems 17-17

Types of Pumps For Photovoltaic Pumping Systems 17-19

Centrifugal Pumps - Surface 17-20Centrifugal Pumps - Submersible 17-22Positive Displacement Pumps - Surface 17-23Positive Displacement Pumps – Submersible 17-25

Primary Requirements For Choosing A Pump 17-31

Supplementary Considerations For Choosing A Pump 17-33Site, Well, and Water Considerations 17-33Electrical Considerations 17-34

Pump Output Curves 17-35Derate Array Power To 80% 17-35




Pump Efficiency 17-39

Considerations and Calculations:Designing a Solar Pumping System 17-43

Solar Pumping Array Sizing Formulas 17-45

Component Design Considerations 17-50

Water Pumping System Installation 17-52

Matching Array To Pump 17-53Direct Coupled Arrays 17-53Maximum Power Tracking 17-55Sun Position Tracking 17-57



Siemens Solar Basic Photovoltaic Technology 17-2 Water Delivery Systems

Referring to the information for the A.Y. McDonald pump in Figure 17-12, we find thecurve for 680 Watts. At a head of 300 feet, the output is about 50 gph. This is equal to

50 gph ÷ 60 = .83 gallons per minute (gpm). We can calculate the efficiency as:

Efficiency of Pump Motor = Head X Flow X FcElectrical Power

= 300 X .83 gpm X 0.189680 Watts

= .069 = 6.9%

Looking at figure 17-14 for the Solarjack SJA pump, we look at the curve for 660 Wattsat 300 feet of head. The output is about 6.25 gpm. This is equal to 6.25 gpm X 60 =375 gph. The efficiency is then:

Efficiency of Pump Motor = 300 X 6.25 gpm X 0.189660 Watts

= .537 = 53.7%

There is a sizable difference in efficiency. The McDonald pump is a centrifugal typeand does not have as wide of range of efficiency. The Solarjack pump is operatingnear its peak efficiency.

According to the data sheet for the Shurflo 9300, the flow rate at 100 feet is 103 gallons

per hour. This is equal to 103 ÷ 60 = 1.72 gallons per minute. The motor current at100 feet is given as 2.6 amps. Using a system voltage of 30 volts, the power is

Power = 2.6 amps X 30 volts = 78 Watts.

The efficiency is:

Efficiency of Pump Motor = Head X Flow X FcElectrical Power

= 100 X 1.72 gpm X 0.18978 Watts

= .417 = 41.7%




The estimated insolation is 6.5 peak sun hours. The farmer's required flow rate is then:

Required Flow Rate= 4000 gallons/day = 615 gallons/hr6.5 peak sun hrs

a. Using the McDonald DK Pump:

We look at Figure 17-12. At 110 feet, it looks like the output is just about 615gallons/hour using the 66V curve. However, to be conservative, we will use the 75Vcurve. Using the 75V curve, the output is about 700 gallons/hour, and the pumprequires 880 Watts.

Efficiency = 110 feet X 700 gph X 0.003 as Fc880 watts pump power

= 0.26 or 26%

Array Wp = 110 feet X 4000 gpd X 0.003 Fc6.5 hours X 0.95 Fm X 0.8 Ft X 0.26

= 1028 Watts required

We choose to use a 75-watt module for this project.


= 5 modules in series


= 2.74 modules rounded to 3 modules

Total array: 5 series X 3 parallel = 15 total

The actual flow rate will be 700 gph X 6.5 peak hours = 4550 gallons/day. The outputper module is:

Output per module = 4550 gallons = 303 gallons/module15 modules




b. Using the McDonald DP Pump:

We look at Figure 17-13, we see that at 110 feet of head, we get about 850gallons/hour using the 680-watt curve (66 V). This is sufficient for our dailyrequirements.

Efficiency = 110 feet X 850 gph X 0.003 as Fc680 watts pump power

= 0.41 or 41%

Array Wp = 110 feet X 4000 gpd X 0.003 Fc6.5 hours X 0.95 Fm X 0.8 Ft X 0.41

= 652 Watts required

Again, using a 75 Watt module-


= 4.4 modules in series rounded to 5 modules


= 1.73 modules rounded to 2 modules

Total array: 5 series X 2 parallel = 10 total

The actual flow rate will be 850 gph X 6.5 peak hours = 5525 gallons/day. The outputper module is:

Output per module = 5525 gallons = 553 gallons/module10 modules

Clearly, the DP pump is much better suited for this application. At the required head / flow combination, the efficiency of this pump is much higher than the DK pump. Thisshows the importance of selecting the correct pump for the specific project

requirements.



Siemens Solar Basic PV Technology Course 18-1 Hybrid SystemsCopyright © 1998 Siemens Solar Industries

Chapter Eighteen Hybrid Systems

Multiple Energy Sources - FlexibleApproaches

Hybrid power systems include a combination of energy sources. The energysources might be solar arrays, wind turbines, diesel generators, biomass plants, orothers. Together, these sources provide the total energy required by the load.These training materials consider mainly photovoltaic - diesel hybrids, although othertypes of systems are briefly discussed.

The general purpose of the hybrid system is to get the best benefits of eachtechnology while reducing some of the disadvantages. A pure stand-alone PVsystem, for example, requires very little maintenance but depends on the amount ofsun light for all its energy. On the other hand, a diesel generator can provide adependable amount of energy whenever it is required, but it requires fuel andfrequent maintenance. A hybrid system using both a PV array and diesel enginecan provide energy regardless of changing weather patterns with less maintenanceand fuel use than a dedicated prime-power engine. This illustrates how the featuresand benefits of each energy source are blended in a hybrid system.

Hybrid systems can take many forms, and there is no single design that is best forevery situation. Instead, the hybrid designer must consider the differenttechnologies and components available and then choose a combination that best fitthe particular requirements. Good hybrid design consists of finding a balancebetween many factors.




Fundamentals of Hybrid PowerSystems

On-Demand Power Generation

One of the key distinctions of most hybrid power systems is the ability to produce“on-demand” power generation. On-demand means that the system itself candirectly control the production of energy. In a simple stand-alone photovoltaicsystem, there is no on-demand power source. The only power source is the sun,and energy is only produced when the sun is shining. This energy is stored in abattery, but if the battery runs low on energy, then the system must wait until the sunrises and begins to produce additional energy. If a diesel engine and batterycharger are connected to this system, the engine and charger represent the on-demand source. If more energy is needed, then a control system starts the engineand produces additional energy.

Examples of variable energy sources• Solar• Wind• Small scale hydro-power

Examples of on-demand sources• Utility (if reliable)• Diesel generators• Gasoline generators• Thermoelectric generators• Biomass plants

Most hybrid systems include an on-demand power source. This provides muchbetter control of energy production in the system and allows the system greaterflexibility to deal with changes in either the load or in the variable energy resource.




Better Use of Renewable EnergyProduction

Stand-alone systems using renewable energy are often designed around a worst-

case situation. Because the renewable energy is variable, the system must bedesigned to accommodate the period of minimum energy production (the worst-casedesign period).

Example: The table below shows horizontal insolation for Denver, Colorado. Italso shows the results of tilting the array (at 65°) to optimize the worstcase month, as well as tilting the array (at 35°) to optimize the annualaverage. The system designed for the worst case month will be basedon 4.77 sun-hours each month, while the system designed for theannual average will be based on the average of 5.73 sun-hours.

Month HorizontalInsolation Tilt to OptimizeWorst Case Tilt to OptimizeAnnualAverage

Jan 2.65 5.30 4.69Feb 3.55 5.65 5.36Mar 4.82 5.55 5.93Apr 5.92 5.18 6.23May 6.73 4.85 6.41Jun 7.41 4.85 6.75Jul 7.16 4.85 6.62Aug 6.44 5.07 6.42

Sep 5.44 5.47 6.19Oct 4.10 5.51 5.54Nov 2.78 4.87 4.47Dec 2.31 4.77 4.18

Average 4.94 5.16 5.73

Since stand-alone systems are designed with the worst case in mind, they areoversized for the rest of the time. This results in extra installed capacity. Theadditional energy in the peak periods can not be used and is wasted. In effect, thewhole system performance suffers from the worst month. If the difference is severebetween the worst case and the other months, this can result in wasted energyequal to 100% or more of the design load.

Continuing the example from above, the system designed for the worstcase month has a worst case month (Dec.) of 4.77 sun-hours. Thebest month is 5.65 sun-hours (February). So during February, thesystem produces 5.65 / 4.77 = 118% of the energy necessary. If allthe energy over 4.77 sun-hours is excess, then an additional 18% ofthe load energy is wasted.




A properly designed hybrid system can make better use of the renewable energy.This is because the hybrid system can be designed around the average case, notthe worst case period. In general, this results in a smaller renewable energycomponent and less wasted power. It is important to remember that the renewableenergy source in a hybrid system is actively supported by another power source.

Exercise

°

°

MonthHorizontalInsolation

Tilt to OptimizeWorst Case

Tilt to OptimizeAnnualAverage

Jan 6.50 4.85 6.20

Feb 5.71 4.94 5.91Mar 4.49 4.43 4.83Apr 3.47 4.19 4.13May 2.43 3.55 3.21Jun 1.84 2.84 2.50Jul 2.10 3.22 2.84Aug 2.85 3.83 3.59Sep 3.71 3.98 4.14Oct 4.92 4.43 5.07Nov 6.08 4.76 5.91

Dec 6.34 4.63 5.99Average 4.20 4.14 4.53

°

b)




Load Matching

Another advantage of hybrid power systems is a better match between powerproduction and the load. This can be as simple as being able to turn on a powersource for peak loads. A carpentry shop may use a small amount of power during

normal operation, but need more power to run a large lumber saw. It makes senseto provide the baseline amount of smaller power for most of the day and then run anon-demand source specifically to power the saw. Similarly, a remote village mayhave a large peak demand for lighting and entertainment in the evening hours.

Step Load

0 4 8 12 16 20 24

Time

L o a d S i z

e ( k W )

Sometimes the size of the load indicates that a hybrid might be useful. Largeapplications require large current flows and higher voltages. These can be costly toprovide using only solar, for example.

For village or facility power systems, there is often a collection of different types ofload equipment. Different types of equipment may be suited to different sources ofpower. Rather than try to supply all the loads with a single power source, a hybridmay match different resources to each individual load. PV, for example, is very wellsuited for water pumping for humans and agriculture. The source of power (the sun)tends to be most available during hot, dry periods. Other loads may be better suitedfor a diesel generator. Equipment that has high start-up surges for example may run

well with a generator. For example, an ice-making unit, which has a largecompressor, has high starting currents that can be supplied by a generator. 3-phaseloads are also typically run from a generator. While such loads can be run fromphotovoltaics and appropriate power-conditioning units, it may be simpler to providean engine for a specific load.




Applications of Hybrid Systems

Hybrid systems are generally used for larger loads than pure stand-alone systems.Generators represent a compact source of high power output. Since the generatoroffsets some of the photovoltaic power, the same size array can serve a larger load

in a hybrid system. For smaller systems, the cost of generators and additionalcontrols tends to outweigh any real economic advantage to a hybrid system.

Because of the inherent flexibility in a hybrid approach, hybrid systems are also usedwhen there is variability in the load. Any time there is a base load with a limitedduration of higher loads, there may be a good match for a hybrid power system. Ifthe system must serve a mix of loads, with different characteristics, it may also be agood application for a hybrid system.

Typical Size Ranges – Hybrid systems can be cost effective in the range of 10 kWhto 220 kWh. Note that some of the other advantages of hybrid system may result in

their use outside of this range. Conversely, if the load is within this range, it doesnot necessarily imply that a hybrid system is the correct solution. Other factors suchas alternative sources of power, availability of technical support, renewableresources, and customer preference must be taken into account. If the utility grid isreliable and easily accessible, for instance, there may be very little justification for ahybrid based only on the load size.

Telecomm Power – Certain types of telecommunications equipment can require ahigh level of power (1-4 kW or more). While diesel generators power manytelecommunications loads, this becomes a problem in very remote areas.Maintenance at such sites is very difficult and expensive. Some sites may only be

accessible by helicopter, making it impractical to bring in large amounts of fuel andoil. While some of these sites become stand-alone systems, for others a hybridsystem is a good compromise between initial cost and on-going maintenance.

Village Power – Small rural communities may be candidates for hybrid powersystems. In some areas of the world, these villages are difficult to electrify bytraditional grid-extension. The size of the loads and the variability caused by manyhouseholds can justify a hybrid system. In addition, the reduced maintenancerequirements in comparison to a diesel-only system are a benefit when skilledtechnicians may not be available.

Facility Power – Some sites are dedicated facilities for a specific use. This couldbe for commercial use such as an ice making plant for fishing, for governmental usesuch as clinic or a recreational use such as a hotel. These types of facilities oftenhave higher load requirements than a simple stand-alone system can provide. Inaddition, there can be sufficient load diversity to make use of different powersources. There may also be additional reasons to consider a hybrid power systemover a diesel. A hotel promoting the ecological beauty of its setting, for example,would not want to have a diesel running 24 hours per day.




Advantages of PV/Diesel Hybrid Systems

The purpose of designing a hybrid system is to utilize the best advantages of thevarious technologies. Hybrid systems have several advantages over either stand-alone PV systems or prime power diesel systems:

Better Use of Renewable Energy – Compared to a stand-alone renewable energysystem, a hybrid can make better use of the renewable energy received. Sincehybrids can be designed for the average case, instead of the worst case, they canmore evenly use the energy through the year.

Higher System Availability – In a normal stand-alone system, there is always somechance that there will not be enough energy to support the design load. This isbecause the incoming energy follows a statistical pattern, and there is always asmall chance of a long period of very poor sun. The ability of a stand-alone systemto support its load under these variable conditions is called availability. It is

important to recognize that availability is different from reliability. Typicalcommercial PV systems have an availability of approximately 99.7% or higher.However, to get higher availability even more PV must be added. A hybrid systemwith an on-demand power source avoids this problem, since the on-demand sourcecan cover the small amount of time of poor weather.

Lower Diesel Maintenance than Prime Power – Maintenance on a diesel engineis directly related to the amount of run time. Therefore an engine in a hybrid systemthat runs half the hours compared to a prime power engine will need only half themaintenance. In addition, poor engine loading can also result in maintenanceproblems. Prime power diesels are often loaded at some fraction of their full rated

capacity. Ideally, an engine should be loaded between 85-95%. However, becauseof peak loads, available equipment or uncertainty about loads, some enginesoperate at a much lower loading, 75%, 50% or even less. This results in a numberof problems, including coking, glazing, slobbering, which can result in highermaintenance costs.

Higher Fuel Efficiency – In addition to the maintenance problems presented by lowloading, running a diesel at significantly less than full load also results in poor fuelefficiency. While a fully loaded engine uses 100% of the rated fuel consumption, anengine that is 10% loaded may still require as much as 50% of the full consumption.Therefore, the amount of energy produced per gallon or liter of fuel will be much

lower. In a hybrid system, the loading of the generator can be controlled for higherfuel efficiency.

Reduced Capital Costs – Standalone PV systems require very little maintenance inthe field, but they have a higher initial cost. Prime power diesels on the other hand,may be very cheap initially, but cost more in terms of continued fuel andmaintenance costs through the life of the system. A hybrid system may cost lessthan the equivalent stand-alone system and use less fuel than a simple diesel-onlysystem.




Better Flexibility for Loads – Hybrid systems offer great flexibility in their ability tomeet changing loads. Since the generator can provide additional energy by simplyrunning longer, it is easier for a hybrid to accept an increase in the loads. Standalone PV systems can not handle a significant increase in the load above the designload, and the only way to allow for growth is to over-size the system for the initialload. Similarly, a prime power diesel can not easily accept load growth and musteither sized for the actual load or run poorly at low load levels initially.




Disadvantages of PV/Diesel HybridSystems

Like any other technology, hybrid systems have disadvantages. To properly specify

and design hybrid power systems, these must be understood as well. Compared tostand-alone PV systems and/or prime power diesel systems, hybrids can have thefollowing disadvantages:

Higher Control Complexity – The consequence of using multiple power systemstogether and utilizing more sophisticated operation is increased control complexity.In addition to being able to control and monitor each separate power sub-system,the controls must also allow for the interactions between sub-systems. However, theintroduction of microprocessor controls increases both the functionality and reliabilityof the control system.

Greater Initial Engineering – Hybrid systems are relatively more complex thaneither stand-alone systems or prime power diesels to design and engineer. Theinitial engineering work to create and manufacture a hybrid requires a broadknowledge of various components, as well as the ability to perform systemintegration. It can often be difficult to determine what solutions are acceptable andwhich will cause further problems. Experienced firms with the right capabilities canaddress the many tradeoffs in hybrid design to assure that the system is designedcorrectly.

More Frequent Maintenance than Stand-Alone Systems – Although hybridsystems can require significantly less maintenance than a prime power system,

hybrid system do require more maintenance than a stand-alone PV system. Thismaintenance is essential for the operation of the hybrid. Since batteries are cycledharder, cell watering and check-ups are very important. The addition of a diesel alsoincreases the number of maintenance tasks as well as the expertise required for thetechnician.

Higher Technical Expertise Needed for Troubleshooting – Problems with hybridoperation can require higher technical competence to identify and solve. This maybe a problem for locations where there is a shortage of skilled technicians. It canalso present difficulties if the end-user of the system does not have access to thoseskills (in a remote village for example). This means that the repair and service of thefinal system should be addressed before installation.

Increased Pollution and Noise – A photovoltaic power system uses clean, silentpower -- the sun. Introducing a diesel into such a system also brings unwantedpollution and noise. Diesel engines produce carbon dioxide (CO2) and otherundesirable emissions. The noise of the engine operating may disturb nearbypeople or bring unwanted attention to the system. Compared to a prime powerdiesel, however, both noise and pollution can be much reduced, and the system canoperate more efficiently.




Comparisons of Various System Types

HybridSystem

Stand-AlonePV

CycleCharge

Prime Power(Diesel only) Utility

Initial Cost High High Moderate Low Low - High

RecurringCosts

Low-Moderate

Low Moderate High Low

Maintenance Moderate Low Moderate High LowFuel Use Low-

ModerateN/A Moderate High N/A

LoadFlexibility

High Low Moderate Low-High High

SystemComplexity

High Low Moderate Low Low

Use ofRenewables

High Moderate-High

N/A N/A N/A

Pollution Moderate Low High High LowNoise Low-

ModerateLow Moderate-

HighHigh Low




Hybrid Configuration and Operation

Power Sources

As described previously, the key characteristic of a hybrid system is that it includes amix of power sources. This combination of power sources gives the system moreflexibility than if it relied on only a single source. Furthermore, using more than onesource may make the system less expensive than if it had to rely on a scarcerenewable resource during part of the year. This section describes some of thevarious power sources available in hybrid systems.

Photovoltaic Arrays

PV is well suited for use in remote areas, since it is extremely reliable and requires

very little maintenance. Many telecommunications loads use PV as the sole powersource, and hybrid systems are a logical extension of this application of PV. Thedifficulties with photovoltaics are the initial cost and the requirement for unobstructedland to place the array. Nevertheless, the advantages of PV make it one of themore common renewable energy sources used for hybrid systems.

PV Arrays


• Very lowmaintenance

• High reliability• Long product life

• High initial cost• Requires land with

no shading• Seasonal resource

Engine Generators

Engine generators represent the single most common on-demand power source.They are very common, compact in size and inexpensive to purchase. Theirdisadvantages are that they require on-going maintenance and fuel, and producenoise and pollution.

Engine GeneratorsAdvantages Disadvantages

• Widespread use andsupport

• Small size• Low initial cost

• Require on-goingmaintenance andfuel

• Produce noise andpollution




Wind Turbine Generators

Wind Turbine Generators (WTG’s) convert wind energy into electrical energy via arotating electric generator. WTG’s are commercially available in sizes ranging from50 watts to many hundreds of kilowatts. The larger units (> 100 kW) are typicallyused in utility-scale wind farms. They are not suitable for use in remote powersystems, although there are beginning to be exceptions to this. The medium size

units (5 kW to 50 kW) are often suitable for remote power use, and the small ones(50W - 5 kW) are typically configured specifically for charging batteries.

The WTG may be connected to the hybrid system on either the AC side or on theDC side. If the unit is connected on the AC bus, it feeds AC directly into the grid,which can cause dramatic fluctuations in voltage and frequency. Unless a verysophisticated control system is used, the wind contribution in this type of system islimited to a small percentage of the overall energy.

Alternately, the AC power delivered out of the WTG can be converted to DC andused to charge batteries. The batteries exert a dramatic stabilizing effect on system

performance, although, they too can be overwhelmed with drastic fluctuations. In asystem with multiple generating sources, it is important to coordinate chargingbetween the various sources. For example, if a WTG is charging the batteries attheir maximum rate during a windy period, and the sun comes out bright and strong,the net effect could be to overheat the batteries due to the excessive current.Similar coordination is needed during diesel charging—the generator cannot bestopped and started continually just to compensate for fluctuating wind speeds.

There are a number of specific points to note when considering WTG hybrids:

First, wind energy is much more sensitive to wind speed than solar energy is to

irradiance. In fact, WTG output theoretically varies as the cube of the wind speed(Power ∝ V3). This has benefits and disadvantages. The disadvantage is that if youhad estimated the wind speed to be somewhat higher than it really is, the net energyoutput of the WTG would be dramatically less than predicted. On the other hand, ina good wind climate, the output of the turbine is so large that other generatingsources cannot even compete economically. In general, if the average wind speedis greater than 4 meters per second, a WTG will be less expensive in a hybridsystem than a PV array, even in a good solar climate. However, because of thewide variability of wind speed on a daily, seasonal and annual basis, it may stillmake sense to have both solar and wind generators in a hybrid system.

Second, WTG’s are much more sensitive to location than PV arrays. It is not tooinaccurate to assume that the insolation at a particular site will be relatively similar tothat actually recorded for nearby sites (airports, city centers, etc.). However, thewind speed is very dependent on specific and highly localized factors, such as thelocation of nearby hills and trees. The wind speed is also dependent on the heightof the turbine, so an anemometer at 3 meters will not accurately represent the




energy available to a turbine on a 20-meter tower. The best advice is to monitor thewind speed at a site for at least one year before making final decisions on WTGinstallation. There are a number of excellent wind monitoring kits which can helpyou perform this task.

Third, because of the extra energy farther off the ground, it pays to install as tall atower as possible. Care must be taken not to shade the PV array with this tower.

Summary of Energy Considerations with Wind Turbine Generators1. Wind Turbine output depends very heavily on wind speed2. Wind speeds can vary significantly over small areas3. Height increases the amount of energy available

In summary, wind turbines can form an excellent generating source in a hybridsystem. However, since the output is so sensitive to the wind speed, and since thewind speed is so sensitive to the particular climate, it pays to make a carefulresource assessment before committing large amounts of investment in windturbines.

Wind Turbine Generators


• Does not require anyfuel

• Potentially lowestcost source ofrenewable power

• Requires additionalsite work to install

• Requires additionalmaintenance

• Usually requiresextended period ofdata monitoring

Thermo-Electric Generators

Thermo-electric generators (TEG’s) operate based on the principle that electricitycan be produced from the temperature difference between dissimilar metals. A fuel,usually natural gas, is burned in the presence of a pile or stack of dissimilar metals,producing a continuous DC current and voltage. Commercially available units havebeen widely used in the gas pipeline industry, where there is an abundant supply ofcheap gas for fuel.

The basic principles of TEG's result in a number of specific advantages and

disadvantages. TEG's have no moving parts and there is no mechanical wear ascompared to an engine generator. TEG's can be started after a long shutdown,which makes them suited for seasonal use. The thermoelectric affect works betterand is more efficient in colder weather. A TEG does, however, require fuel and fuelstorage. The efficiency is very low compared to other types of generators, typically4% or less. The advantage of working well in cold climates becomes adisadvantage in a hot climate. Finally, typical commercial units are only in the rangeof 10 - 1,000 Watt output, limiting their practical use to smaller systems.




Micro-Hydro


• Requires no fuel• No pollution• May provide

economic power

• Depends on specificsites

• Requires additionalcivil works at site

• Typically requiressome resourcemonitoring

• Maintenanceconcerns

Utility / Grid

Although systems with a utility supply are sometimes not considered to be truehybrids, there are times when the overall system behavior resembles a typical hybridsystem. These can be instances when there is utility power present, but it is not veryreliable. In such circumstances the utility is treated more like a variable energysource rather than an on-demand power source.

Exercises




Bus Configurations

One of the basic distinctions in hybrid design is the bus configuration. Broadlyspeaking, this is the general architecture of the system. Hybrids are usuallycategorized as DC bus systems or AC bus systems.

DC Bus Hybrids

In a DC bus system, a generator / rectifier combination charges the batteries. Thecenter of the system is considered to be the common DC “bus” where the battery,rectifier and PV array are all connected together. This is the most common hybridsystem configuration. This system was developed from the idea of “cycle charged”diesel generators, and it is ideal for telecom hybrid systems, since the loads aretypically DC. Note that a DC bus system can supply AC loads by adding a separateinverter to provide AC power.

The PV array charges the battery based on the available solar energy. The DEG is

normally only started when the batteries reach a low voltage state and need to berecharged. The DEG is connected to a dedicated battery charger that supplies DCcurrent. Sometimes there may be an emergency transfer switch so that thegenerator can power an AC load directly. Under normal conditions, however, theengine only runs to charge the battery and to provide the additional energy neededfor restore the battery to full charge.

Pictorial Representations of DC bus configurations

DEG

Rectifier

BatteryBank

PVArray

DCLoads

(+) (-)

AC

DC




DEG

Rectifier

BatteryBank

PVArray

DCLoads

ACLoads

Inverter

TransferSwitch

(+) (-)

AC

DC

DC

AC

AC Bus Hybrids

AC bus hybrids utilize the generator to power some or all of the loads. The batteriesare charged using a bi-directional power conditioning unit that serves as bothinverter and rectifier. The focus of the system is considered to be the common AC“bus” where the generator, inverter and AC Loads are all connected together. Sincethe generator takes a much more active role in the system, these are alsosometimes called interactive hybrids. An AC bus hybrid may have some DC loads,but it is more common to have only AC loads.

The PV array still supplies energy to the battery depending on the solar insolation.

The generator however, can now charge the batteries and supply power to the load.The bi-directional inverter can take power from the battery bank to supply AC power,or can draw AC power from the generator to charge the batteries. Some bi-directional units are capable of operating in synchronized or “parallel” mode whereboth the inverter and generator can supply the load together. This allows for greatflexibility in handling peak loads.




DEG

Inverter/ Rectifier

BatteryBank

PVArray

DCLoads

ACLoads

(+) (-)

AC

DC

Exercises




Hybrid Operation

Daily Operation

A hybrid system will go through shorter and more intense cycles than a stand-alone

system. While a stand-alone system may remain at almost a full state of charge forlong periods of time, most hybrid systems will cycle through the battery every coupleof days. Rapid changes in the battery result caused by a sequence of array andrectifier charging.

The figure below illustrates a sample hybrid system's performance over 4 days. ThePV array and rectifier both contribute about half the total energy in the system. Thebattery bank provides about 1.5 days of storage. This results in a system with rapidcycles and heavy utilization of the battery. Three lines are shown, one depicting theaverage battery state of charge, one line showing the energy provided by the PVarray, and one line showing the rectifier contribution. For this example, the load is

assumed to be constant.

D a y 1 D a y 2 D a y 3 D a y 4

B a t t e r y

P V A r r a y

R e c t i f e r

The figure shows the battery initially at nearly full charge. During the first day, thebattery state of charge falls as the load draws power. The PV array replaces some,but not all of the energy used by the load. On the Day 2, the load again draws downthe battery, with some of the loss being replaced by the PV. At the end of thesecond day the battery is discharged sufficiently to bring the rectifier on line. Therectifier provides a large amount of current in a short interval of time, bringing thebattery back nearly to full charge. As the rectifier shuts down additional PV energy




finishes the charging. At the end of the third day, the load keeps drawing powerfrom the battery. On the fourth day, the load continues to draw down the batterywith some replacement from the PV array. This is very similar to Day 1 and we cansee that the system will continue to run in cycles with the battery eventually reachinga low point, causing the generator to start.

Performance Factors

The pattern of charge and discharge cycles shown above is typical of hybridsystems. Obviously, the cycles may not be identical each day because of severalchanging factors:

1. The load may change from day to day.2. The battery may be at a different state of charge at the start of the day.3. The weather may change, resulting in more input from the PV array.4. The generator may be programmed to run differently, such as providing an

equalize cycle.

The annual performance of a hybrid system is typically measured by the following:• Total Energy Production• Contribution (%) by each source• Number of battery cycles• Generator run hours• Number of generator starts• Generator fuel consumption

Each of these factors is determined by the size of the various system components.Some of the relationships may not be obvious at first. However, if you stop to thinkabout what is actually happening within the system, then the explanation should be

apparent. Below are some general guidelines for most hybrid systems. Note thatcomputer programs must be used for detailed performance results.

Battery Size – The size of the battery affects how many cycles the battery goesthrough. The smaller the battery the more frequently the battery will be chargedand discharged.

Generator / Rectifier Size – The sizes of the generator and rectifier mainlydetermine how long the generator will run for each charge cycle. They do not affecthow much of the total energy comes from the generator. So, to keep the annual runhours low choose a relatively large engine.

Array Size – The size of the array does affect how much of the annual energy issupplied by PV. Obviously the bigger the array the more energy it will produce.

Inverter Size – In a DC bus system the size of the inverter does not affect theannual performance directly. In an AC bus system, the size of a bi-directionalinverter affects the generator run-time similar to the rectifier size.




Different Roles of Generators

Although generators are frequently included in a hybrid system, there are severaldifferent ways that a generator can be used. The operation of the generator can besimplified into the following types:

Prime Power – The generator runs whenever the load requires power. Some primepower applications have an engine running continuously throughout theentire year.

Back-up Power – Another power source is the primary means of supplying the load.The generator runs only rarely, and it is only intended to handleexcessive demand.

Cycle Charge – A pure cycle charge system consists of only a generator, batterycharger and battery bank. The generator provides all the power, but itruns only when the battery requires recharging.

Emergency Power – The generator is not intended to run under normalcircumstances. In some systems the emergency generator is activatedby manual operator control.

In actuality there can be a range of energy contributions and run-times, so thathybrid systems will tend to fit somewhere in between these types. While eachapproach may be suited to a particular set of circumstances, it is important toconfirm the user’s expectation for how the system will operate.

Prime Power

Back-UpEmergency

CycleCharge




Exercises




Battery

The battery bank is a key element in a hybrid system. The battery is literally at thecenter of the system, and the quality of the battery can influence the entire systemperformance. A well-designed system makes good use of the battery storage. A

poorly designed system suffers when the battery does not perform. The batterybank also represents a significant cost in the system and is worthy of extra attention.

Choose a Deep Cycle Battery

Because the battery in a hybrid system is used much more heavilythan in a standalone PV system, it is important to select the properbattery. A heavy duty, cycling type battery is most appropriate, withtypical specifications requiring a cycle life of 1500 cycles to 80%DOD. This will allow the battery to reach 7-10 years of life, evenwith cycles every 2-3 days. Forklift batteries are appropriate, even

though some high antimony types may only have five-year designlives. Higher quality “tubular plate” low antimony telecom batteriescan last up to twice as long, but can also be significantly moreexpensive.

Batteries designed for float applications and UPS operation are definitely notappropriate, and neither are SLI (auto) batteries. “Marine” type cycling batteries area marginal choice, while wheel chair and golf-cart batteries may be used in a carefuldesign. Care should be taken with sealed (gelled or absorbed electrolyte) VRLAtype batteries, since most of these were developed for low cycling applications.

They are also much less forgiving of the variable rates of charge and dischargeexperienced in a hybrid system.

ComponentSelection

Choose a deep cycle battery

• Don't use SLI or float type of

batteries

• Take care with VRLAbatteries




Engine / Generator

The engine generator is a critical component of a true hybrid system, since itsupplies a significant part of the energy—it is not just a “backup.” There are a widerange of choices for construction, fuel and accessories that must be considered.

Prime Power vs. Stand-By

Because the engine in a hybrid system may be running for manyhundreds or thousands of hours per year, it should be specified as aprime-duty engine rather than a standby-duty engine. Prime powerengines are rated for continuous duty environments. Stand-byengines are only expected to run on an infrequent basis. Using astand-by engine in a hybrid application may result in excessive wearand early failure. Note that some manufacturers will rate a specific

engine for both prime duty and stand-by, with the prime duty ratingbeing lower than the stand-by rating. In a standard hybrid system,the engine should last the life of the system without replacement.

Air-Cooled vs. Water-Cooled

Generators have two types of cooling systems: air-cooled and watercooled designs. Air-cooled engines have the advantage of simplicity

and ruggedness. Water-cooled engines are smaller and quieter for a given poweroutput. Air-cooled engines require proper ventilation for both intake and exhaust

airflow. Water-cooled engines require the availability of clean water and propertechnician training. Because reliability is typically more important than size andnoise, air cooled engines are preferred for remote power systems, at least in thesmall sizes.

Choice of Fuel

The choice of fuel for the generator is of primary importance. Generators usingdiesel fuel are known as diesel electric generators (DEG’s) and are very commonaround the world. Diesel fuel is commonly available (although the quality can varysignificantly). Service and trained diesel technicians are also comparatively easy tofind, making diesel generators a good choice for many types of systems. Dieselgensets are often rated for 20,000 to 50,000 hours of operation. In cold weather(approaching 32 °F or 0 °C) diesels can have starting problems although these canbe addressed through the addition of glowplugs and appropriate grades of fuel.

ComponentSelection

Use an engine rated for prime power

Be sure the engine is designed for remote starts

• Use air-cooled

engines for small systems

• Choose fuel type based on application

• Add proper accessories for the system




Diesel Fuel


• Readily Available• Difficulty with cold

weather starting• Fuel can be stored

for several months

• More expensive thangasoline engines insmall sizes

• Noise and pollution

Gasoline/petrol generators are very lightweight and compact for the amount ofpower they produce, making them ideal for automotive applications and light duty“contractor” generators. They also tend to be the least expensive generatorsavailable. Their two major limitations are fuel storage and design life.Gasoline/petrol is much more difficult to store in large quantities because it is morevolatile (meaning it burns easier) than other fuel options. This makes it prone toexplosions if not stored properly, or if there is a leak in the system. In addition these

engines are typically rated to provide only about 1,000-2,000 hours of engine life(compared to 20,000-50,000 hours for diesel gensets) making them suited only forsmall backup systems such as cabin power systems.

Gasoline (Petrol)


• Readily Available• Usually technicians

are available• Engines are small

and inexpensive

• Hazardous to storelarge quantities

• Fuel can not bestored for longperiods

• Engines typically notrated for long life

Another common fuel choice for engines in commercial hybrid systems is liquefiedpetroleum gas (LPG). These are sometimes modified DEG’s. While LPG is lessefficient in terms of fuel consumption compared to gasoline, it is much cleaner andrequires less engine maintenance interval, thus justifying its use in many situations.Certain environmentally sensitive locations such as National Parks will only allow

LPG generators because of concerns regarding pollution. LPG is volatile andrequires specialized fuel feed systems.




Liquefied Petroleum Gas


• Very clean burning• Less engine

maintenance• Low pollution

• Not always easilyavailable

• Specializedtechnicians

• Requires special fuelsystems for very coldclimates

Natural gas is occasionally used for generators, particularly on a pipeline or otherlocation where there is essentially a “free” supply. Similar to LPG, natural gas canbe very clean burning. The reliability of natural gas engines can also be quite good.However, unless there is a supply of fuel nearby, this fuel is very seldom used forremote hybrid applications.

Natural GasAdvantages Disadvantages

• Very clean burning• Less engine

maintenance• Low pollution

• Typically availableonly in specificapplications

• Specializedtechnicians

Remote Starting

It is very important to ensure the engine selected has remote start capability (unlessan operator will be on the site at all times). The engine start circuit must bespecifically designed to allow remote starting, otherwise it will not operate properly ina hybrid system. An engine for a hybrid system must also include an automaticchoke that will automatically adjust the air mixture between starting and normaloperation.

Even if the engine can be remotely started, it is also important to determine the typeof signal(s) needed to start and stop the engine. The simplest system is a two-wire

start system in which there is a single pair of wires. When the wires are connectedthe engine will start and continue to run. When the wires are disconnected, theengine will stop. Other types of start systems take 3 or more wires and haveseparate signals to start the ignition and to operate the engine. These type ofsystems are more complicated, since the control system must now determine howlong to close the starter circuit, whether to allow repeated crank attempts, etc.




Whether or not the engine has a two-wire start system, the important factors are thatthe engine is designed for remote starting and that the type of start system is closelycoordinated with the system controls.

Generator Accessories

Engine-generators in hybrid systems have special duties, compared to a traditional

prime diesel or standby diesel engine. Hybrid engines work in a demandingenvironment, with frequent starts, extended hours of operation and typicallyinfrequent maintenance. For these reasons, a number of the following accessoriesmay be justified.

Long run lube systems

Because maintenance intervals are such a large cost factor in hybrid systems, itmakes sense to reduce maintenance as much as possible. The best way of doingthis is to reduce the frequency of the most common maintenance—changing the

engine oil. Typical standby diesel engines have an “oil change interval” of only 100hours, since nobody expects them to be run very much. In continuous (prime)operation this represents just over four days of use, meaning that the oil would needto be changed almost twice per week! Even in a hybrid system with 1000 hours ofdiesel operation per year a crew would need to visit the site every month.

Diesel engines designed for prime service typically have oil change intervals rangingfrom 250 hour to 500 hours, and they often have available options to extend the oilchange interval to 1500 hours or more. With standard engines the maintenanceinterval for a hybrid is thus approximately every 3 months with the potential to reducethat to semi-annual or even annual visits.

Heavy duty start systems

A remote site diesel engine is only useful if it starts when it is needed. For thisreason alone, it is usually worth the investment in some sort of enhanced startersystem such as a 24V heavy-duty starter. In addition if the start reliability can beincreased a smaller battery can be used, and the system can have significantsavings. It also pays to use a multiple-attempt, crank-limited start controller toprevent damage to the engine during start attempts.

Solar assisted start batteries

One of the primary causes for start failures is a weak battery. In a hybrid system,where the engine may not be called on to start for a month or more, the starterbatteries may self-discharge to the point where they won’t support a start whenrequested (especially in hot seasons, when the diesel is used least). A small PVmodule dedicated to charging the diesel start battery is often a worthwhile




investment. The system should be checked, however, to ensure that it does notovercharge the batteries.

Heavy duty air filters

Keeping the engine air clean is just common sense. If you keep the engine clean, it

will last longer. Dry type filters are preferable to oil bath filters in remoteenvironments.

Dual fuel filters and/or fuel-water separators

Water and other contaminants in the fuel are one of the primary causes of enginefailure at remote sites. A heavy-duty fuel filter, including a fuel-water separator(supplied by RAYCOR and others) is a wise investment.

Reduced vibration mounts

Since the engine is a hybrid system is often installed near expensive batteries andelectronic equipment, it makes sense to specify shock absorbent mounting of theengine. This is a relatively cheap option that will pay for itself in subtle ways.

“Critical zone” muffler system

Engines will typically ship with an “industrial grade” muffler, which is intended for aloud, industrial setting. “Residential grade” mufflers are available which can reduce

the noise to about 95 dBA. If noise is a particular issue (for a village system, forexample), then a “critical grade” (hospital zone) muffler is the solution. Thesemufflers reduce the noise to about 85 dBA. No one will even notice that the diesel isoperating!




Control Systems

Hybrid systems require significantly more control than eitherstandalone PV or prime diesel power systems. The control systemis a central part of a hybrid system and must be integrated with the

overall operation and control. There are several choices in selectingthe proper control system, including control functions, datamonitoring, and types of control systems. A good system designerneeds to understand the range of choices available.

Hybrid Control Functions

In addition to charge regulation and load management, the controllermust provide control of the generator and rectifier, as well as monitoring theadditional currents, voltages and equipment associated with a hybrid system.

Typical functions provided by the control system include the following items.

Control Functions

Starting EngineA hybrid control system needs to be able to start the generator under theappropriate circumstances. The type of relays and control points requireddepend on the start circuit of the engine. In addition, the controller needs theproper data to determine when the engine should be started, which mayinclude battery voltage, current and amp-hour readings, time of day or otherparameters.

Stopping EngineThe engine must be stopped at the conclusion of a cycle.

Regulating ArrayThe power from the array into the battery must be regulated. The relays usedto disconnect the array must be sized for the maximum array current(including edge of cloud effects).

High Voltage DisconnectA high voltage disconnect protects the battery from over charging bydisconnecting the battery from all charging sources. The relays must be sizedaccording the maximum output for each source.

Low Voltage Load Disconnect

A low voltage load disconnect protects the battery from over discharge. Therelays used to disconnect the load must be sized for the maximum loadcurrent.

Exercise EngineEngines that sit unused for a long period of time may experience problemsstarting. To avoid this, hybrid systems sometimes include an engine exercisefunction, which periodically starts and runs the engine for a short time.

ComponentSelection

Identify all necessary control functions

• Design the control systems as part of the total system




Equalize BatteryIn addition to normal battery charge cycles, the system may include anequalize cycle to reduce any voltage differences between battery cells.

System Alarms

Engine Alarms

One of the most important functions of a hybrid system controller is to monitorthe engine. The control system must be able to recognize potentialdangerous conditions and shut the engine down before physical damageoccurs. There are certain specific indications that should be monitored as aminimum. While the control system does not need to have a separate alarmsignal for each of the following situations, it should be able to respond to anyof them.

- Low oil pressure- Engine over-temperature- Engine over-speed- Engine over-crank (engine has failed to start)

Battery VoltageAn essential feature of a controller is to protect the battery bank from over-charge and over-discharge. The actual control functions for high voltagedisconnect and load voltage disconnect are explained above. In addition, it isoften useful to sound an alarm so that the user is aware that the voltage is outof normal limits.

Data Monitoring

System VoltagesSystem voltages can provide a quick check on the state of the system. A

number of readings through the course of a day also depict arrayperformance, load use and approximate battery state of charge.

- Array Voltage- Load Voltage- Battery Voltage

CurrentThe current in each of the main circuits is also an important variable. Fromthe current, one can calculate array output, load demand and chargingefficiency.

- Array Charging Current- Load Current

- Battery Current (Charge and Discharge)

TemperatureTemperature information is used to adjust the battery charging algorithm andmay also be important to monitor for the load equipment.

- Battery Cell Temperature- Load Enclosure Temperature- Ambient Temperature




Miscellaneous ParametersDepending on the system, there may be additional items that should berecorded. For example:

- Fuel level- Wind speed- Site insolation

Types of Control Systems

There are many different classes of control systems available. This sectiondiscusses the three major classes, starting with the simplest.

Manual Controls

In this control scheme an operator is always available. When the battery voltagefalls to a certain point the operator turns the diesel generator on and supervises

charging of the battery. When the battery is fully charged the operator shuts theengine off manually.

While this is the simplest technique in terms of control hardware it is also very proneto errors in judgment or attention. It can also be very costly both in terms of operatorlabor and in consequences of a simple lapse in attention. This type of labor is oftenclassified as “long hours, low effort” and poses a special challenge to maintainingskilled labor.

Separate Subsystem Controls

In this type of system each subsystem has its own controls. For instance the PVarray has its own controller while the generator is controlled via a separate voltagesense circuit. If a wind turbine is present its control is completely separate from theother controllers and likewise for the inverter.

Obviously this technique only works for fairly simple DC bus systems and does notallow tight system integration which holds the potential for future cost reductions.

System Level Controls

In this type of control system a single central controller or a network of distributedcontrollers provides coordinated control of all system functions. This type of controlsystem is usually implemented with microprocessors and can include long-term datalogging and remote monitoring capabilities. Some modern inverters include some orall of the control functions listed below. The hybrid system design must decide theproper approach based on all the parts of the system.




Battery Chargers / Rectifiers

The battery charger is the interface between the AC generator and the DC energystorage system (battery), so it is a critical, but often overlooked component of ahybrid system. Proper selection of a battery charger is essential to good hybrid

system design. Three phase chargers are preferred over single-phase unitsbecause of reduced AC current, increased efficiency, and less DC ripple. However,for smaller systems only single-phase chargers may be available.

Types of Chargers

Most battery chargers are meant to be operated off of utility-gradeAC “mains” power and special attention is needed for use with dieselgenerators. The process of charging batteries has twocomponents—changing the AC to DC (rectification) and control of

the charging current. There are a number of different types ofchargers commercially available, each of which has advantages anddisadvantages. These are discussed below.

Diode Bridges (half bridge, full bridge and three phase bridge)

These units are the most basic type of battery chargers and tend to be very cheap.They offer very little control of the end of charge. Single-phase half bridge chargershave very significant ripple even when filtered heavily. Single phase, full bridgechargers have less than 5% ripple and are very efficient; however, they have no

inherent control.

SCR Chargers

This class of chargers is based on silicon-controlled-rectifiers (SCRs). In essencethese are diodes that can be “turned off” thus offering a method of regulation. SCRrectifiers tend to be inexpensive and moderately efficient but also have a reputationfor being extremely “noisy” in electrical terms. This affects both the AC input circuitand the DC output. The high frequency spikes caused by switching the SCRs offtend to cause problems with the alternators and voltage control circuits used with

engine generators. Most manufacturers require 20-40% oversizing to ensure properoperation.

ComponentSelection

• Choose the right

combination of factors:

- Cost

- Efficiency - Temperature

rating - Power

Quality




Hybrid Bridge / SCR Chargers

This class of rectifier uses a diode bridge to provide the lower portion of the chargepotential (e.g. 48V in a 48V system) with SCRs regulating the last few volts (e.g.,the 48-56V range in a 48V charger). This improves efficiency and reduces the noiseproblem significantly; however, it adds significant cost to the charger since twopower stages and two control circuits are required. They are available in an

extremely wide range of voltages and currents.

Ferroresonant chargers

These chargers produce very clean power using ferroresonant transformers anddiode bridges. Their drawback is size, weight and low efficiency, but they areavailable in a wide range of voltages (up to 260 V) and currents (up to 800 amps).

Saturated Amplifier (Mag-Amp) Chargers

These chargers use a specialty circuit that takes advantage of specialized behaviorof certain types of amplifier circuits. The 3-phase versions have moderateperformance, low cost, and moderate weight, and they tend to be extremely simpleand therefore rugged. (They are commonly used in mines and other harshenvironments.) They are available in a wide range of voltages (up to 260 V) andcurrents (up to 800 amps). The single-phase versions are horrible (70% efficiencyand power factors as low as 0.55).

PWM (high frequency) chargers

The latest technology makes use of high frequency switching circuits similar to thoseused in some PV controllers and sine wave inverters. These units are lightweight,efficient, and highly controllable. They tend to be more expensive than other optionsand are usually only available in smaller sizes (50 amps and less, typically 48Vmax).

Bi-directional inverters as chargers

Some inverters operate in fully bi-directional mode; that is, they can operate either

as DC to AC converters (inverters) or AC to DC converters (rectifiers). They can beefficient and cost effective, since the same circuit supplies two different functions.

Ferroresonant, Saturated Amplifier and PWM rectifiers are most commonlyused in hybrid systems although other types find uses as well. Check to ensure thatthe selected charger is compatible with other components in the system especiallyvoltage regulators on engines.




Efficiency and Power Factor

In general one should choose a battery charger to be as high as efficiency aspossible. Typical specifications for chargers are:

• Three phase charger 85-92% efficient 0.8 to 0.9 power factor

• Single phase chargers 75-85% efficient 0.7 to 0.85 power factor

Power Quality and Other Considerations

Power quality for a battery charger affects two areas – the DC current into thebattery and the AC power used by the charger. The quality of the DC output willaffect any DC loads on the system. The amount of disturbance is typically rated asan amount of ripple. Specify ripple specifications for the DC leg (2% ripple absolutemaximum, typically less). AC power quality will affect the generating source (andmay affect the AC loads on an AC bus system). Check the AC noise specificationsalong with power factor.

In some models of battery chargers acoustic noise can also be a problem. Find outwhat the unit produces under worst case situations.

Finally confirm that the rectifier is designed for a rugged environment (e.g. widetemperature variations, high humidity, dust or marine atmosphere) not an air-conditioned computer room.




Inverters

Inverter technology can be critical to hybrids that are specified to supply AC power.There are three main classes of DC-AC converters—square wave, modified squarewave, and sine wave. The technology follows the same basic classes as batterychargers discussed above. The best choice is a single or 3-phase, transistorized

PWM, sine wave inverter provided the cost is appropriate.

Power Quality

Inverter power quality is very important to AC loads. The most basicevaluation of power quality is the waveform. The quality of power isthe worst with a square wave and best with a sine wave. Somedevices may not operate with square wave power or may experienceoverheating.

Wave Form Power Quality

The line below indicates the extremes of power quality. Squarewave is the worst and pure sine wave is the best. However, costs

usually increase with power quality, too.

Square Wave Modified Square Wave Sine Wave

(Worst Quality) (Best Quality)

Voltage and Frequency Tolerances

The next general evaluation of power quality is the voltage and frequency tolerancesof the output power. Voltage and frequency tolerances are usually specified aspercentages of the nominal value.

Total Harmonic Distortion

The most specific measure of wave quality is the total harmonic distortion. As

explained in the section on AC Power, total harmonic distortion measures how farthe waveform deviates from a pure sine wave. Modern sine wave inverters canachieve a total harmonic distortion of less than 5%.

DC Charge Current

If the inverter is bi-directional, then the DC charging current power quality shouldalso be examined. Refer to the section above on battery chargers for thisdiscussion.

ComponentSelection

• Choose the right combination of factors:

- Cost - Efficiency - Temperature

rating - Power

Quality




Structures and Enclosures

Hybrid components require some method of mounting and protection from the directeffects of the weather. This may be accomplished by placing each group ofcomponents in a separate enclosure or there may be one shelter that contains all

the components. Generally smaller systems use the first approach while largerhybrids will have a building or even several buildings dedicated to the system.Prevention of vandalism is another reason for providing the proper enclosures forthe hybrid system.

Often it is desirable to separate the batteries from any electronicequipment such as inverters, rectifiers or load equipment. Floodedlead acid batteries produce hydrogen and trace amounts of acidvapor when under heavy charge. The acid vapor can acceleratecorrosion in sensitive electronics. For this reason the batteriesshould be placed in a separate room or enclosure.

A system enclosure both solves certain problems and createsadditional ones. An equipment shelter may protect the componentsfrom very low outside temperatures and from wind-blown dust anddebris. However, the temperature and ventilation within the shelter

must also be considered. Diesel engines produce very large amounts of heat andso do power electronics to a lesser extent. Proper ventilation must provide cool,clean air for these types of equipment.

Fuel storage represents an additional concern in a hybrid system. Diesel enginesoften use a smaller day-tank together with a main fuel tank. The National Electrical

Code limits the amount of diesel fuel that can be stored inside a building to 660gallons (2500 liters). Any outside fuel storage should be secured with a lock, sincediesel fuel is a valuable commodity, especially in remote areas!

Any structure or components exposed to the ambient conditions should beengineered for the proper corrosion resistance. Sand, wind, rain, sun and salt-spraycan rapidly work to produce rust and corrosion in all but the most durable materials.Stainless steel, quality hot-dipped galvanized steel and aluminum are the preferredchoices.

ComponentSelection

• Choose structures / enclosures based on:

- Weather

- Battery location

- Corrosion - Vandalism

protection




PV / Diesel Hybrid Sizing and Design

The flexibility of hybrid power systems is matched by the possible variations insystems. There are many different approaches to designing a hybrid system. It isnot important whether the system follows a particular design approach, but whether

the system functions properly and supports its design loads.

The hybrid designer must determine several parameters to size a system:

• Design Loads and Load Profiles • Type of Bus Configuration • Battery Bus Voltage • Battery Sizing • Generator and Battery Charger Sizing • Generator Operation • PV and other Renewable Sizing

Some system-level design is necessary to make the proper choices during the basicsizing. It is very important to keep in mind the overall system operation during thisprocess. There are many interactions between components in a hybrid system andthe designer must ensure that all the overlapping requirements are met.

This section addresses some of the fundamentals of system sizing and designspecifically for PV / Diesel hybrids. Although this particular combination of powersources is given emphasis many of the design principles are valid for other types ofsystems.

Trade Offs in Hybrid System Design

Customization

Simplicity Standardization

Flexibility

Design Limits

Initial Cost

Operating Cost




Design Loads and Load Profiles

Just as defining the design load is a crucial step in stand-alone PV system design,the overall load for a hybrid system is equally important. For AC loads, additionalinformation about the frequency, number of phases and power factor is necessary.

Questions to ask to determine system loads:

1. What are the different types of loads to be used with the system?

2. For all loads (both AC and DC): What is the voltage? What are the voltage limits (max and min)? What is the power (or current)? Does the power or current change according to voltage? When does the load operate (duty cycle)?

What is the peak power drawn by the load?

3. For each AC load: What is the frequency? Is the load single phase or 3 phase? What is the power factor?

Not only the size of the load is important, but also the hourly load profile. Thesystem must be able to accommodate any peaks. The load profile might alsosuggest certain control strategies. For example, in an AC bus design load peaksshould be coordinated with the generator since this improves system efficiency.

Example: DC Load Calculation

For a certain telecommunications application, the load is given as:225 Watts at 24 Volts for 14 hours per day100 Watts at 24 Volts for 10 hours per day

So225 W x 14 hours = 3,150 watt-hrs/day100 W x 10 hours = 1,000 watt-hrs/day

Total load = 4,150 watt-hrs.4150 / 24 hrs = 173 W continuous. 225 W peak

DesignGuidelines

Get complete,accurate load information.

• For DC and AC loads:

- voltage - power - duty cycle - peak power

• For AC loads: - frequency - 1 ph or 3 ph - power factor




Example: AC Load Calculation

In a certain rural community, the community leaders would like to allowthe following equipment per household. There are 40 households totalin the community:

2 x 50 watt fluorescent lights (12 hours per day), power factor = 0.61 x 25 watt radio (6 hrs per day), power factor = 1.0

For one household2 x 50 Watts x 12 hours = 1200 Watt hrs/day1 x 25 Watts x 6 hours = 150 Watt hrs/dayTotal load = 1,350 Watt hrs /day. 1350 / 24 hrs = 56.3 Watts

continuousPeak power = 2 x 50 Watts + 25 Watts = 125 Watts

The apparent power is:2 x 50 Watts / 0.6 power factor = 167 Volt-Amperes (VA)1 x 25 Watts / 1.0 power factor = 25 VATotal apparent power = 192 VA

For 40 householdsTotal load = 40 x 1,350 Watts = 54 kWhrs/day.Total real power = 40 x125 Watts = 5000 WattsTotal apparent power = 40 x 192 VA = 7680 VATotal power factor = 5000 W / 7680 VA = 0.65




Bus Configurations

Determining the Correct Configuration

The choice of whether to use a DC bus or AC bus is determined by the loads and

the overall system operation. If all the loads are DC power, then a DC bus makesthe most sense. If the majority of the loads are AC then an AC bus system mightmake sense. AC bus hybrids are particular useful when the generator is required tosupport part of the load.

In general an AC bus design requires more sophisticated controlsand is more complicated to operate. However, they can also bemore efficient, since energy from the generator can go directly to theloads. In the DC bus system the generator energy must go throughan AC/DC converter (rectifier) and then back out to a DC/ACconverter (inverter) to reach the load. Depending on the efficiency

of these components this can add significant losses. For example, ifthe rectifier efficiency is 80% and the inverter efficiency is 85%, thenany energy going through the pair of components is only at a 0.8 x.85 = .68 = 68% efficiency.

The hybrid designer must be careful making such comparisons.These efficiency gains for the AC bus system only apply to theenergy that is supplied directly by the generator. In addition an ACbus system may need to use single-phase alternators and rectifiers

based on the load requirements. Since the generator and rectifier pair for a DC bussystem can be independent of the load some efficiency can be gained by using

three phase alternators and rectifiers, which are more efficient than single-phaseunits. This can result in the net efficiency of a small, single phase AC bus hybridbeing less than an equivalent DC bus hybrid using a 3-phase generator and batterycharger.

DesignGuidelines

• For a DC-only system, use a DC bus.

• For an AC- only system,use an AC bus.

• For a mixed system,evaluate both options.




Bus Voltage Selection

The battery bus voltage in a hybrid system can make an important difference onboth the cost and the efficiency of the system. In general the battery voltage shouldbe as high as possible within the constraints of equipment availability and localsafety codes. This is because the higher battery voltage allows lower operating

currents resulting in lower losses and higher overall efficiency. (Remember thatwattage losses are proportional to the square of the current.) Since the cost ofwires, fuses, circuit breakers and other auxiliary equipment is also a function ofcurrent capacity a higher voltage will result in savings for those components as well.

Comparison of Losses,Assuming 120W Load with 0.5V Voltage Drop

Voltage Current Loss % of Total48 V 2.5 A 1.25 W 1%24 V 5 A 2.5 W 2%12 V 10 A 5.0W 4%

For DC bus systems the DC operating voltage of the load equipmentoften determines the bus voltage. This can become a problem withlarge 12-volt loads and the designer sometimes needs to considerthe economics of a DC/DC converter to step the voltage down fromsay 48 volts at the battery to 12 volts at the equipment. Similarly if

there are mixed load voltages usually the largest load determinesthe bus voltage since this reduces the size of the rest of the DC/DCconverters.

For systems with AC loads the battery voltage is determined by theinput DC voltage of the inverter. In general all but the smallestsystems should use at least 48-volt equipment while larger villagesystems (> 25 kWh / day) should use 120-volt or 240-volt batteries.The largest PV-diesel hybrids in commercial use today use 480 Vbatteries (which keep the maximum operating voltage below the 600VDC limit). The following example shows a comparison between the

use of 12V and 48V batteries for a hybrid system with a 4 kW inverter:

Design

Guidelines

• Choose a higher voltage

to reduce the current.

• DC systems: Match the voltage of the largest load

• AC systems: Match the input voltage of the inverter




Example: Effect of Bus Voltage on Currents

A hybrid system uses a 2.5 kW array of 53-watt modules, a 12 kWrectifier (90% efficient), and an 8 kW inverter (90% efficient).

At 12 V, the following conditions apply:

Parallel Modules = 48 (48 module array @ 1S x 48P)Array Current = 48 modules X 3.5 A (Isc) = 168 ARectifier Current = 12,000 W X 0.90 ÷ 12 V = 900 AInverter Current = 8000 W ÷ 0.90 ÷ 10 V (min) = 890 A


Parallel Modules = 12 (48 module array @ 4S x 12P)Array Current = 12 modules X 3.5 A (Isc) = 42 A

Rectifier Current = 12,000 W X 0.90 ÷ 48 V = 225 AInverter Current = 8000 W ÷ 0.90 ÷ 40 V (min) = 222 A


Parallel Modules = 5 (45 module array @ 9S x 5P)Array Current = 5 X 3.5 A (Isc) = 18 ARectifier Current = 12,000 W X 0.90 ÷ 120 V = 90 AInverter Current = 8000 W ÷ 0.90 ÷ 100 V (min) = 89 A

A 100A disconnect for the 120V system may be less than 10% of theprice of the disconnect for the 12V system. Other components willhave cost savings as well.




Battery Sizing

Stand-alone photovoltaic systems will often provide 5-7 days or more of battery-backup for the load. For critical applications or in locations where the solar insolation ispoor this may increase to 10-15 days or more.

The autonomy of the battery bank is the number of days that the battery can supplythe load under no charging conditions. Thus a battery bank that can support theload for 5 days without charging is said to have 5 days of autonomy.

For a hybrid system the battery bank is usually much smaller, often2-3 days of autonomy. The reason that the battery can be smaller ina hybrid system is because there is an on-demand power source.When the battery gets low the system can put a generator in actionand replenish the battery as needed. In a stand-alone system thebattery functions as a reserve of energy. This reserve should bekept as full as possible to guard against periods of poor weather.

The battery in a hybrid system is serving a slightly different purpose.The battery allows the system to control the use of each powersource. By providing storage the battery allows the generator to runat optimum times. It also provides a means of utilizing the energygathered by the array. In advanced hybrid control strategies thereare even times when the system deliberately seeks to leave thebattery at a low state of charge.

The smaller the battery capacity, the more frequent the cycles. Thisresults in a shorter battery life. A good hybrid design strikes abalance between a smaller economic battery and reasonable life. A

common error about hybrid performance is that a small batteryincreases the amount of diesel energy contributed to the system.

On a first order this in not true. A smaller battery results in more frequent dieseloperation and more generator starts, but the total amount of energy produced isabout the same. By using high-cycling batteries and a reliable start system a hybridcan use a smaller battery and potentially save significant money.

Since the size of the battery does determine the maximum charge current thebattery capacity does have an effect on the size of the battery charger. This in turnaffects the maximum size of the generator because the generator should not be toomuch larger than the battery charger. Forcing a very small engine size will increase

the total hours of operation, which might not be desirable.

DesignGuidelines

• A larger battery reduces the number of engine starts and battery cycles

• Start with a 2- 3 battery

• Check the number of cycles for the battery

• Verify the maximum charge rate is

less than C/5.




Example: Effect of Battery Size on Cycle Life

A 2-day battery will result in approximately 180 cycles per year. Whatis the estimated life of the battery if it is rated for 1200 deep cycledischarges at 25 °C?

A 2-day battery will result in approximately 180 cycles per year. 1200 / 180 = 6.7 years at 25 °C. Note if the battery temperature is greaterthan 25 °C for significant periods of time, the life will drop sharply. A6.7 year cycle life is reasonable for a hybrid system, since the batteriesare not likely to last much longer.




General Engine Derating Factors:

Temperature-0.36% of full power for every oC above rated temperature

Elevation

-3.5% of full power for every 300 m above rated altitudeHumidity

Per manufacturer’s recommendations, up to -6% of full power

Example: Calculation of Engine Derating due to Elevation

A 10 kW diesel generator has a nominal rating of 10,000 watts up to700 meters altitude. If the generator is installed at 3000 meters, itsoutput will be reduced.

Derating = -3.5% X (3000m actual - 700m rated ) 300 m= -0.035 X 2300m 300m= -0.27 or -27 % derating

Output Power @ 3000 meters = 10,000 watts X (1-0.27)= 10,000 watts X 0.73= 7300 watts

So the “10,000” watt generator could really only produce about 7,300watts at that altitude. If 10 kW were really required, then a largergenerator would have to be installed.

The example above shows that the rated power must be larger than then requiredpower to allow for the derating. You would divide the required power by [1 - thederating factor] to get the sea level rating of the inverter you would need.

Power @ rated conditions = Power REQUIRED at installed conditions[1 - Derating Factor]




Example: Calculating Rated Engine Size to Give Required Size

Following the above example, in to actually have 10 kW output available at 3000meters, you would need to install a 14 kW rated model:

10,000 Watts= 13,700 watts, rounded to 14 kW (1- 0.27)

Generator Energy Contribution

The more energy the generator provides the smaller the PV array can be. Thisreduces the initial cost of the system but also increases the fuel and maintenancecosts of the engine.

There are several simplified approaches to determining the amount of energycontributed by the generator and the total number of run-hours. These are useful fordetermining a rough estimate of the generator performance. In order to make

accurate predictions about generator performance more detailed computer-basedsimulations are used.

The best approach to energy contribution is to determine the desired number of runhours and then estimate the amount of energy produced in the system. As abovethis method requires careful consideration of the efficiencies and losses. Thiscalculation must take into account the efficiencies and losses through the system.

A second approach is to just determine a percentage of the total energy. Forexample a designer might want the DEG to provide 50% of the annual energyrequirements. This condition then determines how many hours the engine runs.

However this method is arbitrary and can result in systems that are not welloptimized.

Note: It is very important to keep in mind that engine run-hour estimates in thesetypes of calculations are rough estimates only. Because the total number of hourscan be small any small change in the actual values can result in a very largepercentage change.

Example: Consequences of Small Changes in Diesel Calculations

Suppose the duty cycle for an engine is estimated to be 10%. If the

actual duty cycle is 12%, the total hours of operation change from 876per year to 1051 or an additional 175 hours. This may not seem like alarge amount. However, it reflects a 20% additional increase, whichwill be reflected in fuel consumption, maintenance requirements, andso on. If the diesel fuel tank were sized with only an extra 10%capacity, the engine would run out of fuel!

The above example illustrates the importance to treating engine calculationscarefully and allowing for ample design margins.




Exercises




PV Array and Other Renewables

The photovoltaic array in a hybrid system provides energy without the need for fuelor frequent maintenance. PV energy therefore helps reduce the operating costs andrequired maintenance of the system. The proper tilt angle and the correct amount ofPV required for a hybrid system are slightly different than for a stand-alone PV

system.

Array Tilt Angle

Stand-alone photovoltaic systems need to maximize the insolationfrom the short winter sun to minimize battery size and system cost.This usually means tilting the array surface 10-20 degrees morethan the latitude angle to face the low winter sun.

But in the case of a hybrid system there is no longer the need tomaximize any particular seasonal performance from the solar array.

Array output can be optimized for the greatest output over theentire year letting the diesel generator fill in during any particularseason as necessary.

Set the Array Tilt Angle Equal to Latitude for Maximum Annual Output

The angle that results in maximum annual output from a solarmodule or array is usually the latitude angle of the site. At this anglethe sun travels exactly perpendicular to the surface of the modules

during the spring and fall solstice, and the winter and summer sun paths are equallyspaced away from perpendicular.

Example: Best Tilt Angle for Annual Output

An example will illustrate the maximum energy production that resultsfrom the array being tilted to the latitude. We can look at a singleSM55 module operating in Denver, Colorado at latitude 40° N:

Array TiltAngle AnnualWh / year Lowest MonthWh /day (Month) Highest MonthWh /day (Month)20° 93,849 158 (Dec) 329 (Jun)30° 97,031 186 (Dec) 314 (Jun)40° 97,506 207 (Dec) 293 (Sep)50° 95,454 220 (Dec) 278 (Mar)60° 90,959 228 (Dec) 272 (Oct)

DesignGuidelines

• A larger array results in more energy from PV and reduces the amount of energy from

other sources

• Tilt the array to get the best annual output.

• Size the PV array between 25% and 75% contribution.




As the array surface tilt angle increases from 20° to 60°, the totalannual energy output increases and then decreases, with themaximum occurring somewhere in between. This maximum output is97,506 Wh/year at 40° tilt, the same angle as the latitude.

Note that at 40°, the worst-case output (Dec.) is only 207 Wh/day,

while at 60° the worst-case output rises to 228 Wh/day. For a stand-alone system, you would choose the 60° tilt angle to minimize the sizeof the array and the discharge of the battery bank. However, with ahybrid system, you do not have to rely on the solar array to rechargethe battery during bad weather--you have the generator.

This illustrates that to maximize the output of a solar array, and therebyminimize the energy required from a diesel generator as backup, thebest angle to choose is the latitude of the site.

Annual Energy Contribution

The effect of adding more renewables to a system is to reduce the amount ofgenerator run time. At the same time the renewables also increase the initial cost ofthe system. The general range PV contribution is between 25% and 75% of the totalannual load.

A large part of the economic advantage of a hybrid system comes from moreeffective capture of solar energy. For this reason it is usually advisable to size thesolar array so that no energy is wasted even in the highest solar month. Since the

battery is smaller than a comparable stand-alone system a hybrid with a very largerenewable energy fraction may not be able to store all the renewable energyproduced.

Because of the variability of solar insolation from day to day the practical limit isabout 90% PV contribution in the best month in order to avoid regulating the arrayunnecessarily. This will typically limit the annual effective solar contribution to 70-80%. If the energy is wasted then the system efficiency drops and may result in aneconomically poor design.

In some design situations there are specific goals for energy contributions during

certain seasons. For this reason it may be acceptable to design the array so thatsome energy is regulated during summer months.




Checklist for Hybrid Sizing

Site and Project Information

System Description: Site:

Site Latitude (deg): Site Elevation (m):

Average Temperature: Avg. Insolation: (tilted):

DC Loads

Description Quantity Watts Hours/dayDaily Demand(DC Whrs)

Total

AC Loads

Description Quantity Watts P.F.Volt-Amperes Hours/day

Daily Demand(AC Whrs)

Total

= + Total AC Whrs Inverter Effic. AC Input (Whrs) DC Whrs




Sizing Checklist, Continued.

DC

Volts

DC

Amps

AC

Volts kW kVA Hz

Phase

1 or 3ph Other CharacRectifier

FuelType

TotalDerating

Generator

ContinuousRating

SurgeRating

Inverter

EstimatedPeak

EstimatedSurge

TotalLoad

Capacity(kWh) Type

NPa

Battery

QuantityTiltAngle

NPa

PV Array

OtherRenewable



Siemens Solar Basic PV Technology Course System Design – Hybrid SystemsCopyright © 1998 Siemens Solar Industries

18-54

Simplified Sizing/Design Example

The following example illustrates some of the principles discussed in this section.

Proposed ApplicationA customer has telecommunications equipment that needs to operate a remote site.The customer has had poor experience with diesel engines in the past and is

concerned about the amount of maintenance required at this distant location.

The technical load is 30 A continuous at 48 V DC. There is a small air conditioningunit 600 watts at 220 V AC, 50 Hz that is only used when technicians are on site.Further discussions with the customer determine that the technicians typically bringa portable generator with them to the site for test equipment and power tools.

The proposed site is located at an elevation of 1500 m. The average sitetemperature is 30 °C during the summer season. The climate is dry so humidity isnot an issue. The average solar insolation provides 5.5 kWh/day.

LoadsThe telecommunications load is 30A at 48VDC.

30 A x 48V = 1440 W = 34.6 kWh/day = 12,629 kWh /year

Based on the description of the intended use for the air conditioner it isrecommended that the portable generator power this load. A separate power hook-up can be made to allow the generator to be connected to the air conditioner and toallow technicians to power other equipment. This approach greatly simplifies thedesign of the system and eliminates the need for an inverter.

Hybrid ConfigurationBecause of the separation of the technical DC loads from the air conditioner a DCbus is the most suitable.

Bus VoltageSince the only loads are 48V DC this is the most appropriate choice for the busvoltage.

BatteryA battery capacity with 3 days autonomy to a maximum 80% depth-of-dischargeyields:

34.6 kWh/day load = 34.6 X 3 / 0.8 = 129.75 kW

129.75 kWh / 48V = 2703 Ah per string




18-55

Suppose we have the following battery cells to choose from:

Capacity 2000 Ah 2500 Ah 3000 Ah 4000 Ah

The 3000 Ah capacity cells are the most appropriate, since they are the smallestsize that provides at least 3 days of autonomy. The maximum charge rate of C/5 for

this size battery is approximately:

3000 / 5 = 600 Amps.

Battery Charger and GeneratorSuppose we have the following rectifiers to choose from at 48V (Efficiency = 0.80):

Amps 100 150 200 400 800Input (kW) 6 9 12 24 48Phase 1 ph 1 ph 3 ph 3 ph 3 ph

The 400A rectifier is a little smaller than the maximum allowed charge rate, but willstill charge the batteries at an appropriate rate (about C/8).

We next determine the size of the generator. Suppose that the following generatorsare available and the ratings are given for 900 m and 25 °C:

Capacity 25 kW 27.5 kW 30 kW 40 kWFuel Diesel Diesel Diesel DieselPhase 3 ph 3 ph 3 ph 3 ph

First we calculate the range to provide a 75-90% loading on the engine.

Max 24 / 0.75 = 32 kWMin 24 / 0.90 = 26.7 kW

We need somewhere between 26.7 and 32 kW installed output from the generator.We must now allow for any deratings due to the elevation and temperature.

Temperature -.36% / °C x (30 - 25) = -1.8%Elevation -3.5% / 300 x (1500 - 900 ) = -7.0%Total derating 1.8 + 7.0 = 8.8%

The rated range must be:Max 32 kW / ( 1- 0.088) = 32 / .912 = 35.1 kWMin 26.7 kW = 26.7 / .912 = 29.3 kW

The 30 kW engine would be a good match.




18-56

Generator ContributionThe customer has stated a preference to reduce the amount of generator run time.However, given the size of the load it will not be economical to power the loadsexclusively from PV. We might select 500 hrs as the maximum desired generatorrun time in one year (typically 1 maintenance interval per year).

The rectifier output is 400A x 48 V = 19.2 kW. Assuming an additional 10% system

losses (0.90 multiplier) and overall battery efficiency of 80% we can calculate theenergy output of the rectifier / generator combination as:

500 x 19.2 x 0.90 x 0.80 = 6,912 kWh/year

PV ArrayThe PV array needs to make up the remaining portion of the annual load = 12,629 -6,912= 5,717kWh /year. The average insolation is 5.5 peak sun hours and weassume the following multipliers for array output:

High temperature = 0.85Dirt = 0.9Battery Eff = 0.8

The size of the array is calculated as:

Array Size (kW) = Annual Energy (kWh/year)Peak Sun Hrs x 365 days x 0.85 x 0.9 x 0.8

= 57175.5 x 365 x 0.85 x 0.9 x 0.8

= 4.65 kWp = 4650 Wp

The array voltage is already determined by the voltage of the bus to be 48 V. If weselect SM50 modules with a rated power of 50 Watts, then the array needs toconsist of:

4650 / 50 = 93 modules, rounded up to the next multiple of 4 = 96 modules.




18-57

“Rules of Thumb” for Sizing

With so much flexibility it is difficult to do any hybrid sizing by hand. It really requiresa computer program to balance the many factors and allow quick modification to seethe effect. However, it is possible to give some very general “rules of thumb” to getyou started with some hybrid designs. Keep in mind that these guidelines are verygeneral and rough. Further refinement is always possible.

Start with the average daily load for the system in kWh/day. Then some simpleratios can be established for the other components of the system such as the size ofthe generator, rectifier, battery capacity and inverter. And a rough relationship canbe given between hours of generator run and total generator energy in the year.From this information array size can be estimated given some weather information.The factors below are separated for a DC bus system with DC loads only and an ACbus system with AC loads only.

Units DC Bus Hybrid AC Bus Hybrid

Daily Load Demand kWh/day

= Y = Z

Inverter Power kW (Not applicable) = 0.3 Z

Battery Capacity kWh = 4.3 Y = 5 Z

Rectifier Rating kW = 0.48 Y = 0.56 Z

Generator Rating kW = 0.67 Y = 0.78 Z

Annual GeneratorEnergy

kWh/yr = Gen Run Hrs X 0.4 Y = Gen Run Hrs X 0.4 Z

Annual Total Energy kWh/yr = 365 Y = 365 Z

Annual PV Energy kWh/yr = Annual Total - AnnualGenerator Energy

= Annual Total - AnnualGenerator Energy

PV Array PmaxRating

kW = Annual PV EnergyPeak Hrs X 365 X 0.6

= Annual PV EnergyPeak Hrs X 365 X 0.5

(Note: The derivation of each factor is given at end of the section)

The following example illustrates using these guidelines to do a rough hybrid systemsizing.




18-58

Example: Hybrid Sizing Using “Rules of Thumb” Method

A remote village will have a hybrid photovoltaic diesel power systeminstalled. The typical daily energy demand for the village has beenestimated to be about 24 kWh/day. Since the loads are AC, an ACbus is most appropriate.

Inverter Power = 0.3 x 24 = 7.2 kW

Perhaps this would be rounded up to the next commercially availablecapacity, probably 10 kW. Typically at this size, the input DC voltagewould be 48 or 120 VDC.

Battery Capacity = 5 x 24 = 120 kWh

The Ah of capacity will depend on the input DC voltage of the inverter.For example, if the input voltage were 48 volts, the capacity would be

2500 Ah. At 120 VDC the capacity would be 1000 Ah.

Rectifier Rating = 0.56 x 24 = 13.4 kWGenerator Power = 0.78 x 24 = 18.7 kW

Again, these might be rounded up to the next commonly available size.

Assume that the recommended interval between maintenance is about500 hours, and that we do not want to have more than onemaintenance trip per year.

Annual Generator Energy = 500 hours x 0.4 x 24= 4800 kWh per year

Total Annual Energy Demand = 24 kWh/day x 365days/year

= 8760 kWh

Annual Energy From PV = 8760 - 4800= 3960 kWh

Assume that the region receives about 5 “peak hours” of insolationdaily.

Array Power Required = 39605 X 365 X 0.5

= 4.3 kW

So we find that a 4300 watt array would provide the energy that thegenerator did not supply during its 500 hours of annual operation.




18-59

Derivation of Hybrid Sizing “Rules ofThumb”

All Systems

1. AssumedParameters:

Battery bank autonomy = 3 daysMax battery depth of discharge = 70%

Max battery rate of charge = C/10

Rectifier efficiency = 80%

Generator derating = 90%

General system efficiency = 90% (Due to wiring losses,etc.)

Battery total energy efficiency = 80%

PV Array temperature factor = 90%

PV array dirt factor = 90%

Inverter efficiency = 85% (AC Bus systems only)

DC Bus Systems AC Bus Systems

2. Assume the totaldaily load is givenas:

Y = daily DC load (kWh) Z = daily AC load from inverter(kWh)

3. The daily amountof DC energydrawn from thebattery is:

= Y = Z / inverter efficiency

= 1.2 * Z

4. The total energy

available in thebattery is:

= 3 * Y = 3 days * Z / inverter efficiency

= 3.6 * Z

5. The total batterycapacity is:

= 3 * Y / Max. DOD

= 4.3 * Y

= 3 * Z / (inverter eff * max.DOD)

= 5 * Z

6. The rectifierrating is based onthe size neededto deliver themaximum chargerate to the battery:

= C/10 of the battery, increasedby system losses.

= C/10 * Battery Capacity

.9

= 4.3 * Y / 10 .43 * Y.9 .9

= .48 * Y

= C/10 of the battery, increasedby system losses.

= C/10 * Battery Capacity

.9

= 5 * Z / 10 .5 * Z.9 .9

= .56 * Z




18-60

6. The generatorrating is based onthe size neededto power therectifier:

= rectifier rating increased bythe rectifier efficiency andgenerator derating

= .48 * Y

.9 * .8

= 0.67 * Y

= rectifier rating increased bythe rectifier efficiency andgenerator derating

= .56 * Y

.9 * .8

= 0.78 * Y

7. The annualamount of energyproduced by thegenerator inkWh/year is givenby:

= Gen Run Hrs * final DC loadoutput

= Gen Run Hours * Gen Rating* rectifier eff * gen derate *system losses

= Gen Run Hours * .66Y * .8 *.9 * .9

= Gen Run Hours * 0.4*Y

= Gen Run Hrs * final AC loadoutput

= Gen Run Hours * Gen Rating* rectifier eff * gen derate *system losses * inverter eff

= Gen Run Hours * .78Z * .8 *.9 * .9 * .85

= Gen Run Hours * 0.4*Z

8. The inverterrating is assumedto be:

(Not applicable) = 0.3 * Z (Typical ratio of loadto continuous rating)

9. The annualamount of energyproduced by thePV array inkWh/year is givenby:

= Total Annual Energy - Energyproduced by generator

= 365 * Y - Gen Run Hours *0.4*Y

= Total Annual Energy - Energyproduced by generator

= 365 * Z - Gen Run Hours *0.4*Z

10. The PV energyis related to the

array rating asshown by:

So the PV arrayrating is given by::

Array Energy = PV ArraykW * peak hours * 365

days/year * deratings

where deratings = heat factor* dirt * battery eff

= .85 * .9 * .8 = 0.6

= Annual PV Energy

peak hours * 365 * 0.6

Array Energy = PV ArraykW * peak hours * 365

days/year * deratings

where deratings = heat factor* dirt * inverter eff * batteryeff

= .85 * .9 * .85 * .8 =0.5

= Annual PV Energy

peak hours * 365 * 0.5




18-61

Hybrid System Maintenance

Since a hybrid system is a “cross-breed” between two standalone power systems itsmaintenance requirements will reflect each of those systems. However, carefuldesign of the hybrid system will preserve much of the “maintenance free”

characteristics of a PV system with only a limited amount of diesel maintenance.

PV System Maintenance

The maintenance of a PV array in a hybrid system is no different than standardmaintenance in a regular PV system. A good check of the system will include:

1. Checks for damaged modules.2. Checks for loose electrical connections.

3. Cleaning the array surface if unusually dirty.4. Checks for any loose mechanical connections.

Engine Generator Maintenance

The maintenance of a diesel engine generator (DEG) is proportional to the number of hours of operation.

This is a very important point in a hybrid system. Even though the energycontribution of the diesel segment may be 50% or more the duty cycle is typicallymuch less than this often 10% or less. Because maintenance is related to thenumber of hours rather than the total energy delivered this offers a tremendousadvantage to a hybrid system.

Consider the following maintenance schedule which is typical of a small, air-cooled,prime diesel engine:

Maintenance Task Interval

Oil Change 250 hrsDecoke 1,500 hrs

Overhaul 6,000 hrs




18-62

An operating manual lists the following Three Levels of Diesel Maintenancerequirements:

Frequency Task(s) To Perform Qualifications Cost Estimate250 hours • Oil and filter change.

• Inspect air and fuel filters, fuelsystem, starter battery, and

system electrical connections.

Can beperformed by atrained technician

at the site.

$25 to $100plus travel ifrequired.

1,500hours

Decarbonization (top endoverhaul): In addition to oilchange tasks-• Replace air and fuel filters• Remove the cylinder head and

clean valves, valve seats,injector nozzles, etc.

• Replace all top end gaskets.

Can only beperformed by atrained enginemechanic, sitework is possible.

$250-500 plustravel for atypical twocylinder aircooled engine.

6,000hours

Full overhaul. In addition to taskslisted above• Perform full engine overall.• Clean all combustion and

inlet/outlet chambers.• Replace all engine gaskets.• Replace camshaft / crankshaft

bearings, valves, valve springs,injectors, fuel pumps, pistons,piston rings, starter battery,start solenoid and enginestarter, fuel solenoid, fuel

pipes, and other parts asnecessary.

Requires acompetentmechanic.Typicallyperformed in arepair shop whichmeans the enginemust betransported to theshop, and areplacementengine providedto the site.

$1,500-3,000,plustransportationand loaner ofreplacementengine.

Note that many standby engines have only a 100-hour oil change interval that, ifrunning in a prime power application, would necessitate nearly 90 oil changes peryear -- almost twice per week! Engines made for prime power often have optionsthat can extend the primary maintenance interval to 500, 1000 or even 1500 hoursdramatically reducing the maintenance requirements.

Using the hybrid sizing example in the previous section the generator runs

approximately 500 hours per year vs. 8760 hours for a continuous prime diesel. Thefollowing table combines the maintenance schedule with the operation of theengines to calculate the total number of maintenance visits required:




18-63

MaintenanceTask

Interval(Hours)

Hybrid(500 hours)

Prime Power(8760 hours)

Oil Change 250 2 per yr 35 per yrDecoke 1500 1 every 3 yrs 6 per yr

Overhaul 6000 1 every 12 yrs 3 every 2 yrs

In addition the higher loading of the hybrid engine will keep the engine running moreefficiently reducing some of the common problems with diesel generators -- cloggedfuel injectors, heavily carbonized valves, slobbering and glazing. This means thatthe intervals for extended maintenance items such as top-end overhauls and fulloverhauls might be extended even further than for a variable load prime diesel.




18-64

Battery Maintenance

The second most maintenance intensive item in a hybrid system is the battery.Unlike batteries in PV systems that are treated quite “gently” hybrid system batteriesare exposed to frequent deep discharges (typically two per week or more) andvariable rate charges (often at high rates). This last point is very important --

charging must be coordinated between the various generators.

For example a system may include a diesel engine, a couple of wind turbines and asubstantial size PV array. The engine/battery charger is designed for a high raterecharge (C/6 or so). If the battery charger is on and the wind suddenly picks up orthe sun comes out the charge rate may exceed the battery manufacturer’srecommendations causing damage to the battery and a potentially dangeroussituation with excessive gassing and hydrogen evolution. If VRLA batteries are usedthis point is even more critical.

In general the batteries should be checked at regular intervals. Unless there are

serious access constraints to the site the batteries should be examined at leastevery 3 months. They should be filled with water, equalized (if the controller doesnot have an automatic equalization function) and checked for tight connections,evidence of corrosion, etc. Once a year a complete record should be made ofindividual cell voltages and any suspect cells tested and replaced if necessary.

Since most hybrid systems use flooded cycling batteries they often use increasingamounts of water as they age. This should be taken into consideration whenplanning maintenance visits. In a village type system a local villager can be trainedto check the water level at weekly intervals, keep records and report to themaintenance crew if there are any “thirsty” cells. In telecom systems this is not

possible; however, careful monitoring of separate battery strings might show if one isdrawing excessive amounts of current during charge -- a sure sign of a problem cell.

In general the use of a high quality controller and battery charger will help ensurelower maintenance requirements and longer life of any hybrid system battery.




18-65

Controls Maintenance

The amount of maintenance on the system controls is usually very small andconsists primarily of checking that all the wiring connections are tight and that thereare no problems with the fuses or surge protection.

Some systems with on-site data storage may use this time to download the data orto replace storage modules for transportation back to a central site.

It is also a good idea to use the maintenance interval as a chance to review thesystem operation. Does the system operate as intended? Is there any unusualbehavior that requires attention. Reviewing the performance data can be a way toidentify potential problems before they become serious. The system set points andcontrol strategies can sometimes be adjusted to accommodate new changes in theload demand or different conditions than were originally expected.




18-66

Applications and Case Studies

Telecommunications

Telecommunications applications are often candidates for hybrid system for severalreasons. Many telecommunications loads are relatively large and a hybrid systemhelps offset some of the initial cost. Companies in the telecommunications industryoften have experience with diesel engines and already have trained technicians whocan perform the necessary maintenance. Also telecommunications equipment oftenis located in places where there may be very limited insolation during a portion of theyear or where there is not sufficient area for a full size array.

Telecommunications loads that might be power by a hybrid include:

• Microwave repeaters

• Fiber optics relays• Rural telephony• Cellular phone sites• Television transposers

Telecommunications power sources demand very high reliability. Some systems willinclude a second engine generator as a backup. This is based on the idea that asystem should have duplicates any essential component or system. Sometimes thisis referred to as "1 + 1 redundancy." Another approach used to assure performanceis to over-size some of the components. A telecommunications system may includea much larger battery or have a very high PV array contribution. These features

allow the system to operate for a longer period of time after a diesel failure, hopefullygiving a repair crew sufficient time to reach the site and make repairs.

A system designed for telecommunications must also give thought to protecting theload equipment. Unlike a village power system where all the loads are physicallyseparate from the power system a telecomm hybrid usually needs to include somespace for the load equipment. This should be removed from the generators andbatteries. In addition there may be heating and cooling requirements for the loadequipment. The system designer must include the power necessary to heat or coolthe enclosure in addition to the load equipment draw.




18-67

Village Power

Many rural communities in the developing world do not have electricity. A smallamount of electricity can make a dramatic change in the lifestyle of these people byproviding lighting, refrigeration and access to television and radio programs. Hybridsystems can be used for rural electrification projects. Of those communities that are

already electrified many are powered by a diesel generator and already havedistribution lines. In communities without power a hybrid offers the ability to supplypower while reducing fuel costs, pollution and noise.

Village load profiles are often well suited for a hybrid system particularly if there is asignificant lighting component. Since the lights are primarily used during the eveninghours the load level may be several times the daytime value. This results in veryinefficient prime power diesel systems, but the same load variation can be handledby a properly designed hybrid power system. The presence of a generator to supplypower on-demand is also a cost-effective means to power large commercial loads(e.g. an ice-making facility) that would be very costly for a stand-alone system.

Hypothetical Village Load

4.8 kW Peak / 50 kWHr/Day

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00

Time of Day

L o a d ,

k W

There are many social aspects to bringing power into a village. How will the

lifestyles of the villagers change? What new uses for the power will they find? Thesize of the load can often change according to special occasions or holidays.Sometimes the total load in the village grows dramatically after the system has beeninstalled.




18-68

Facility Power

Another application of hybrid systems is for facility power. This could be anyisolated site with a collection of diverse loads. Some example of facility applicationsinclude the following:

• Hotels• Government offices• Medical clinics• Military facilities• Recreation Sites (Marinas, Campgrounds, etc.)

While the characteristics of each system can vary significantly these applicationsrequire fairly large amounts of power and typically include AC loads. In addition theload demand is rarely constant and may be composed of a number of variableloads. It is important to carefully define the load requirements of the system andensure that the hybrid will have sufficient capacity to meet peak loads.

Since many facility power applications are existing sites the hybrid power system isoften intended to supplement an existing diesel-based power system. In suchcircumstances the primary customer concern is to reduce the amount of dieseloperation. This is in effort to reduce the amount spent on fuel and maintenance.Pollution and noise may also be significant secondary considerations especially forecologically sensitive sites.

One feature of a facility power system is that someone is usually designated toprovide maintenance for the system. This can be an advantage over village powerprojects where on-going maintenance can be difficult to support.




18-69

Design Examples

Pulau Pemanggil, Three Islands Project, Malaysia

The Pulau Pemanggil system is located in Malaysia and is designed to provide

power for a small community located on an island near the coast. The original loadrequirements were calculated as follows:

ITEM Avg.Wh/Dayper unit

No. ofUnits

DailyEnergyWh/Day

AnnualEnergy(kWh)

Domestic House 276 20 5520 2014.80Domestic House + Shop 380 3 1140 416.10Mosque 834 1 834 304.41Community Hall 464 1 464 169.36Police Post 1048 1 1048 382.52School 3970 1 3970 1449.05Clinic (excluding vaccine storagesystem)

740 1 740 270.10

Power Plant House 904 1 904 329.96Total Average Wh/Day 14620 5336.30Add 10% Safety Factor (to accountfor cable losses, load variations,etc).

1462 533.63

Total Average Energy required. 16082 5869.93

Array 7.2 kW peakBattery 124 kWh @ 120 VoltsGenerator 26 kW diesel generatorInverter 10 kW, 240 VAC, 50 Hz, 1 phaseDesign Load 16.1 kWh/day @ 4.5 Sun hours

The system was originally designed as a PV system with the generator primarilyused for backup. The PV array was to supply most of the loads. After the systemwas installed the loads grew significantly to the point where the generator supplies asubstantial portion of the loads.

PV Array – The photovoltaic array uses high-efficiency, single crystal modules.There are three ground mounted arrays.

Controls – The system uses one set of controls for the PV array and another set ofcontrols for the generator. In addition there is a data acquisition system that recordsperformance data for the system.

Battery – The system uses a single string of sealed batteries.




18-70

Block Diagram - Pulau Pemanggil

7.1 kWPV Array

124 kWhBattery

11.3 kWRectifier

240 VACLoads

26.4 kWGenerator

10 kWInverter

For additional information on this system contact Siemens Showa Solar, Singapore.




18-71

SunWize™ Power Station

The SunWize™ Power Station is a series of pre-packaged hybrid designs that aredesigned to server telecommunications or village power applications. The systemdesign offers an excellent example of combining all the important factors of a hybridsystem – performance, cost, reliability and system integration. In addition thesystem is designed as a pre-assembled unit that can be tested in the factory. This

significantly improves the overall quality of the product.

There are several versions of the Power Station. The main components andfeatures for a PSG1500 system are shown below. It is important to recognize thatratings and performance depend on the specifics of the installation.

Array 1500 W peakBattery 24 kWh @ 24 VoltsGenerator 4.8 kWInverter/Rectifier 4 kW, 120 VAC, 60 Hz, 1 phase / 24 VDCDesign Load 6.5 kWh/day AC loads @ 4.5 Sun hours

PV Array – The photovoltaic array uses high-efficiency, single crystal modules.These modules provide higher amp-hour per watt output.

Array Structure – A self-supporting structure provides a framework to mount all thecomponents. This design uses the weight of the system to ballast the system. Thearray structure is adjustable from 30-60 degrees tilt.

Inverter/Battery Charger – The system uses a bi-directional inverter/batterycharger. This unit provides both the AC output as well as allowing current from thegenerator to charge the battery.

Batteries – The battery bank is composed of flooded 2V cells that are contained in adedicated enclosure. The battery enclosure is ventilated to prevent theaccumulation of hydrogen.

Generator – This system uses an air-cooled propane generator.

Controls – The system is completely automatic with the inverter controlling thegenerator and DC regulator controlling the solar array functions. The controls anddistribution are located in a weather resistant enclosure.




18-72

Block Diagram - SunWize Power Station

1.5 kWPV Array

24 kWhBattery

4 kWInverter/ Rectifier

120 VACLoads

4.8 kWGenerator

For additional details refer to the brochure located in the Appendix. The SunWize™Power Station is manufactured by SunWize Technologies, Inc. located in Kingston,NY, USA.




18-73

McPherson Peak System

The McPherson Peak system* is an example of a large telecommunication hybrid.This system powers cellular repeater equipment and was designed and installed bySolar Electric Specialties. The system is located near Santa Barbara, California. Asa telecommunications power system the hybrid system is designed for high reliabilityand low maintenance:

Array 11 kW peakBattery 200 kWh @ 24 VoltsGenerator 4.8 kW 24 VDC Output, Dual Units.Inverter 1.5 kW, 120 VAC, 60 Hz, 1 phase / 24 VDCDesign Load 70 kWh/day @ 4 Sun hours

PV Array – The photovoltaic array uses single crystal modules. About 75% of thearray is mounted on the shelter – the rest is ground mounted.

Generators – The engine generators are propane fueled. There are two DC

generators for the site. One generator is designated as the primary engine and theother serves as backup (1 + 1 redundancy).

Note that the array size is very large while the generator is relatively small for thegiven load. This system was designed to operate at 80-90% PV contribution butalso to allow for significant load expansion. This way the load may increase withoutany need to add additional hardware.

System Controller – The system controller at this site allows for remote monitoringof key system values including fuel levels. This allows the customer to access thesite via a telephone modem and determine the status of the system.

Inverter – There is a small inverter to provide power to 120 VAC loads within theshelter. Most of the equipment loads operate directly off the 24-volt power bus.

Batteries – The battery bank is composed of sealed batteries. Because thegenerator provides a relatively small amount of charge current the risks ofovercharging are well managed.

Shelter/Enclosure – The load equipment is contained in a 24' concrete shelter.The shelter also provides a mounting location for some of the PV array. Coolingfans ensure that the shelter temperature does not exceed the design limit.

* Note: There are actually two hybrid systems located on McPherson peak both ofserving similar loads. The system in this example is slightly larger than the othersystem.




18-74

Block Diagram - McPherson Peak System

70 kWPV Array

200 kWhBattery

1.5 kW

Inverter

24 VDCLoads

4.8 kW DCGenerator

120 VACLoads

4.8 kW DC

Generator

For additional information on this system contact Solar Electric Specialties in Willits,California, USA.




18-76

(End of Chapter)




18-77

CHAPTER EIGHTEEN

HYBRID SYSTEMS 18-1

Multiple Energy Sources - Flexible Approaches 18-1

Fundamentals of Hybrid Power Systems 18-2On-Demand Power Generation 18-2Better Use of Renewable Energy Production 18-3Load Matching 18-5Applications of Hybrid Systems 18-6Advantages of PV/Diesel Hybrid Systems 18-7Disadvantages of PV/Diesel Hybrid Systems 18-9Comparisons of Various System Types 18-10

Hybrid Configuration and Operation 18-11Power Sources 18-11Bus Configurations 18-16Hybrid Operation 18-19

Component Selection 18-23PV Array 18-23Battery 18-24Engine / Generator 18-25Control Systems 18-30Battery Chargers / Rectifiers 18-33

Inverters 18-36Structures and Enclosures 18-37

PV / Diesel Hybrid Sizing and Design 18-38Design Loads and Load Profiles 18-39Bus Configurations 18-41Derivation of Hybrid Sizing “Rules of Thumb” 18-59

Hybrid System Maintenance 18-61PV System Maintenance 18-61Engine Generator Maintenance 18-61Battery Maintenance 18-64

Controls Maintenance 18-65

Applications and Case Studies 18-66Telecommunications 18-66Village Power 18-67Facility Power 18-68Design Examples 18-69



Siemens Solar Basic Photovoltaic Technology 18-1 Hybrid Systems

Chapter 18 – Answers Hybrid Systems

Looking at the table of insolation data, the worst case for the system tilted at 55° is 2.84

peak sun hours and occurs in June. In comparison, the average number of sun hoursfor the hybrid system tilted at 25 is 4.53. Therefore, the first advantage to using thehybrid is that we can design around an insolation value of 4.53 sun hours, rather than2.84.

The highest insolation for the system tilted at 55° is 4.94 sun hours, occurring inFebruary. A stand-alone system designed for the worst case of 2.84 will lose much ofthis energy. The amount of excess energy is

Maximum Insolation = 4.94 = 1.74

Worst Case 2.84

This implies that 74% of the required energy is wasted in February.

a. PV might be preferred to wind or micro-hydro in an area where construction isdifficult. Both wind turbines and micro-hydro turbines can require significantconstruction. Very little civil works are required for a small PV array. Other issues thatmight make wind or micro-hydro less desirable are: difficulty in obtaining permits, localwind patterns, water access, undesirability of obstructing water flow, etc.

b. Where there is a steady, significant wind flow, a WTG might be preferable to PV.Under the proper conditions, wind can provide a very low cost energy source.Additional factors that might make wind more favorable might be: extreme seasonalsolar variations (e.g. the arctic), limited available ground area, etc.

c. A thermoelectric generator might prove the best solution if there is a cheap source ofenergy close by. This is common is gas field applications where the TEG can useplentiful natural gas. Another common application is in extreme cold climates. Not onlydoes the TEG operate better under such conditions, it can also provide heat to keepequipment from freezing.

Note that other answers are possible.




1. DC Bus (DC Loads Only) – This system configuration makes the most sense forlarge DC loads. The generator provides power to the rectifier, which then charges thebattery. It is the simplest to design and control, and it does not require a sophisticatedinverter. Good applications for this type of system include telecommunications

systems.

2. DC Bus (DC and AC loads) – A DC bus hybrid with mixed loads has some of thesame control simplicity as a basic DC bus system. AC loads are connected to adedicated inverter that draws from the battery. This works best if the AC loads aregenerally small and do not have much variation. This system again works well fortelecommunications, perhaps for a system including an equipment shelter with lights orother small AC loads.

3. AC Bus (AC Loads) – This type of system has the most flexibility in supplyingvariable AC loads. The generator is connected in parallel with the output of the inverter.

This allows generator power to flow directly to the loads and is very efficient. Theinverter must be more sophisticated and the control system needs to include morefeatures. This type of system works well for a village with many AC loads or a facilitypower system, e.g. a small eco-tourist hotel.

An AC bus hybrid can still supply DC loads directly from the battery bank. The bi-directional inverter keeps the battery charged, supplying any DC loads that may beconnected. Note that for larger hybrid system, the battery bank voltage may be too highfor most normal DC loads.

Choose any three of the following factors:

Battery SizeThe larger the size of the battery bank, the fewer cycles the generator will make.A large battery bank will also reduce the number of cycles on the battery, therebyextending its life.

Generator/Rectifier SizeThe larger the generator / rectifier combination, the shorter the time necessary torecharge the battery bank. This does not change the total amount of energyprovided by the diesel; it just reduces the hours of generator operation.




We now need to take into account deratings for both temperature and altitude. Thealtitude derating was calculated in the previous problem to be 35%. The temperaturederating is:

Temperature Derating = -0.36% X (35 °C - 25 °C rated)

= .0036 X 10 = .036 = 3.6% derating

The final output will be the rated power multiplied by both derating factors:

Output Power = 50,000 Watts X (1 - .35) X (1 - .036)

= 50,000 X .65 X .964

= 31,330 Watts or 31.3 kW @ 4000 m, 35 °C

We first calculate the derating factor for the elevation.

Derating = -3.5% X (1900 m actual - 700 m rated)300m

= -0.035 X (1200)300

= .14 or 14%

The necessary rated power is then calculated by:

Power @ rated conditions = Power REQUIRED[ 1 - Derating Factor]

= 20,000 Watts[ 1 - .14 ]

= 23,260 Watts or 23.3 kW @ rated conditions

Therefore, in order to have 20 kW available at 1900m, the engine must have a rating ofat least 23.3 kW at the rated altitude of 700m.




The sizes of the various installed components (array, battery, rectifier and generator)may be different than the sizes resulting from the "rules of thumb." There are a numberof reasons for these kinds of differences, including:

• Customer requirements, including higher availability

• Availability of equipment sizes

• Different design goals

• Different operational strategies

• Cost parameters

It is important to remember that the rules of thumb are a rough guide. The hybridsystem designer still needs to evaluate the sizing and selection of components to meethis/her specific needs.



Siemens Solar Basic PV Technology Course 19-1 System Design – Installation & TroubleshootingCopyright © 1998 Siemens Solar Industries

Chapter Nineteen

Installation andTroubleshooting

As you read this chapter, consider the benefits of planning and the costs of a mistakeonce you have begun the installation process. As we have seen throughout thiscourse, the calculations needed for a typical photovoltaic installation are simple. It isthe proper consideration of all the small details that distinguishes a professionalinstallation from an amateur one.

Also consider the cost of repair or maintenance visits to a remote site compared to thecost of initially planning the system for extra reliability. There is never a free lunch,even with a photovoltaic system. But you can pay a little at the beginning of the life of asystem, or a lot more over time, especially when it is difficult or costly to access the site.

The information presented in this section is intended as a survey of technical topics,and does not constitute a complete course covering all aspects of installing systems. Itis expected that the student will continue his or her own education, to insure that propersafety and design concerns are included in any system design, installation,maintenance and repair.

The technical information and suggestions for installation, operation, use andmaintenance made herein are based on Siemens Solar Industries' knowledge andexperience. They are believed to be reliable, but such information and suggestions donot constitute a warranty, expressed or implied.

Careful planning, knowledge of electricity, and safe practices will reward thesystem installer with a dependable, low maintenance system.




Safety Issues

The safety record for photovoltaic power systems has been excellent, and we in the

industry want to keep it that way. There is the danger of physical harm from workingwith equipment, as well as the danger of electrical shock from working with solarmodules, batteries, inverters, and engine generators.

National Electric Code

In the previous System Wiring chapter we discussed Section 690 of the NationalElectric Code in how it addresses photovoltaic power systems. Many issues dealingwith proper design and installation of photovoltaic systems are covered in that section,

and many other issues are referenced from that section to previously written sectionson wiring, grounding, installing, etc. You should become familiar with the Section 690,or refer any questions regarding proper electrical methods to a trained andknowledgeable electrician.

Adhering to the NEC helps you avoid the following:

• Shock hazards - dry and wet leakage.

• Wiring that will deteriorate.

• Inadequate or no over-current protection and disconnects.

• Non-standardized wiring methods.

• Inadequate grounding.

• Poor workmanship.




General Safety Practices

The following are general guidelines for working safely with photovoltaic systems.

•

Know the system you are working on.

• Know the test equipment and tools you are working with.

• Never work on a photovoltaic system alone.

• Review first aid procedures. Know how to get help.

• The best safety system is an alert mind, a skeptical mind, and a slow hand.

• Follow National Electrical Code and local codes.




Safe Practices With Batteries

Batteries deserve their own special attention here because they are potentiallydangerous if improperly handled, installed, or maintained. Dangerous chemicals and

high voltages and currents are potential hazards. Anyone who is working with batteriesmust first be familiar with safety procedures and system design. (See also theANSI/IEEE standard 937-1987, Recommended Practice for Installation andMaintenance of Lead-Acid Batteries for Photovoltaic Systems.)

The following suggestions for safety are presented to aid in properly and safelyhandling, installing, checking and replacing batteries in photovoltaic power systems. Allbatteries should be installed and handled in accordance with the manufacturer'sinstructions and local codes and regulations. Observance of proper techniques will helpto insure long battery life, low maintenance and safe system operation.

Personal Safety

• Remove any jewelry from around neck or from hands or wrists before workingaround batteries.

• Use non-metal hard hats to avoid possible electrical shock.

• Wear protective clothing while working with batteries. This includes acid-resistantgloves, apron, and eye protection.

• Have fresh water easily accessible in case acid splashes on skin or eyes.

• Have baking soda easily accessible in case of acid spills.




Conditions for Safety in Working with Batteries

• Keep open flames and sparks away from battery area.

• Discharge body static electricity by touching a grounded conductor before touching

battery terminal posts.

• Disconnect battery bank from any sources of charging or discharging before workingon batteries.

• Have exposed battery terminals covered by strong non-conducting caps or covers.

• Design battery area with adequate ventilation, and adequate protection from theenvironment.

• Lift batteries only in a manner approved by their manufacturer, and never by the

terminal posts.

• Use lifting aides, dollies and other moving aides to transport batteries. Do notsimply pick up batteries and carry over terrain. Plan the transportation process.

Tools for Working with Batteries

• Tools should be shorter than the length between terminal posts, to reduce thepossibility of dropping and causing a short circuit.

• Have non-working or grip end of metal tools covered with a strong non-conductiveplastic coating or tape.




Exercise:




Planning the Installation

Following is a general discussion of the importance of planning and preparedness forthe installation of PV systems.

The suggested process steps presented here are broken into two phases: the planningstage followed by the installation. Key to a successful installation is a site visit in theplanning stage to finalize the system design and to determine the layout of theequipment at the site. By visiting the site first and proper planning, the system designerwill avoid many costly mistakes allowing for a trouble-free installation.

The most important first step in the planning stage is to have proper documentation ofthe system design including electrical schematic if available along with information on

the major components: controls, inverter, etc. The designer should be familiar with theoverall system design, the major components and other considerations specific to theproject before visiting the site.

In planning a PV system installation the system designer should account for thefollowing general and specific considerations:




General Installation Considerations

• Be familiar with all relevant documentation on the system including schematic,information on the various components, system operation, etc.

• Make provisions to meet all permits and codes

• Make sure to follow proper installation and operation safety procedures

• Plan to consider aesthetics and architectural compatibility

Specific System Considerations

• System Sizing and Load OptimizationAll pertinent data should be collected and calculations of array and battery sizeshould be made. The client should be involved in deciding those issues thatinvolve judgment, such as autonomy period in battery sizing, criticality of theloads, load sensitivity to voltage ranges and waveform. All efforts should bemade to use efficient loads. The system sizing process may iterate many timeswith options examined to arrive with the best design.

• Consider shading and aesthetics in choosing the array mounting locations

• Plan for lightning protection: grounding, surge arrestors, etc.

•

Plan to locate batteries, controls, inverters, and electronic components such thatthey are protected and access is controlled (locked room).

• Plan to protect PV panels from animals, people, and falling objects. Considerfencing surrounding the array and structure.

• Consider the shortest possible wire runs, to minimize voltage drop/wire sizes, andwire protection, in the placement of equipment.




Site Visit and Analysis

If at all possible, the designer, installer and client should visit the proposed site for the

system. If no visit is possible, then the designer/installer should have extensivephotographs or maps available to become familiar with the terrain and anticipate anyproblems with actually installing the system. A visit will also allow for better planning onthe placement and sizing of equipment as well as the logistics of bringing equipment tothe site. Array mounting or location may be greatly influenced by the solar access atthe site, and this may affect later equipment considerations and sizing calculations.

Equipment for Site Analysis

When visiting the site it is recommended to take along the following tools.

• a solar siting device to determine shading for various times of the year

• a compass for azimuth

• a measuring tape and a notebook

Solar Siting Device

A solar siting device is a device that allows you to view the path of the sun throughout

the year and also any objects that might block the sun and cast shadows on the array.An example of a solar siting device is shown in the figure.

The dome allows viewing of the entire horizon without having to look directly at the sun.By looking down on the dome, the installer can see the outline of any trees, structures,or other objects that might obscure the sun’s path.

Specially designed cards slip under the dome and indicate the hourly position of thesun for each month of the year. The installer can move the siting device around at thesite to find a location where the array will not be shaded during peak hours of insolation,usually between 9:00 a.m. and 3:00 p.m. during the shortest day of the year.




The device can even be used to estimate quantitatively the amount of shading thatmight occur due to objects in the sun’s path. Values are assigned to each hourlysegment that indicate the percentage of total solar path contained in each hourlysegment. Since the sun is up for fewer hours during the winter, each hour of the sun’spath during those months will constitute a greater percentage of the whole than an hoursegment during the summer.

The installer can draw the horizon onto the card, including any objects that intercept thesun’s path. Each hourly segment that is obscured during a particular month can beadded up to give the total percentage of the sun’s path that is blocked. This can thenbe subtracted from the expected total solar insolation for a typical day during that monthto give the real insolation that would be available. But all effort should be made at theoutset to site the array away from objects that block the sun’s path.




Array Spacing From Objects

One particular aspect to examine on the site visit is the best location for the array.There may be a conflict between the desire to keep the array as close to the batteryand loads (to minimize voltage losses in the wires) and the best location for receiving

solar radiation. Structures, trees, chimney, fences, and other potential shading objectsshould be avoided. Also the prospect of vandalism should be considered, and mayimpact the location of the array.

Finding room for placement is usually not an issue for remote sites where ample spaceand options for locating may be available. For restricted rooftop mounting, or for largearrays where land cost is a factor, a detailed analysis of land cost vs. power loss maybe advisable.

Locate Array Away FromTall Objects

cr i t ical

an g le

s un

ar r ayh e i g h t

d i s t a n c e

The proper distancedepends on the latitude,the time, and the height ofthe nearest tall object

An array should be placed far enough away from an object so that there is no shadingbetween the hours of best insolation, usually from 9:00 a.m. to 3:00 p.m., on the day

with the longest shadows, December 21 in the Northern Hemisphere and June 21 in theSouthern Hemisphere. It is very helpful to have a solar site locator to determinethis (as discussed earlier). If you do not have access to a siting device, use thefollowing calculations to insure that the array will be located away from potentialshading.




Array Orientation and Azimuth

Normally, an array should face true south in the Northern Hemisphere, not just towardsthe magnetic south indicated by a compass (and true north in the SouthernHemisphere). The deviation of magnetic north from true north is called the magnetic

declination. Consult local magnetic declination maps to find the azimuth correction thatshould be applied to a compass reading for your particular site.

The declination map for the United States is shown below. The map shows east andwest declinations. For example, if your location is in the region of the map where the“east” declinations are shown, you should say “the north point of a compass pointsabout XXX degrees east of true north”. So true north is actually that many degreeswest of where the compass points to North.

Example: The northern “pan handle” of Texas falls between the 10o E and 12

oE

lines of the map. This means that the north point of a compass pointsabout 11

o EAST of true north. So true north would actually be about

11o WEST of where the needle points.




Array Considerations

When laying out the array, the system designer should consider the following:

Mechanical Considerations • Must support all loads including wind and snow• Should maintain strength over array lifetime

• Provide access for array maintenance

• Protect array from animals and vandalism(e.g. install fencing around array site)

• Provide ample space for air cooling

• Avoid the use of dissimilar metals

Electrical Considerations • Use appropriate wiring and protection

• Minimize electrical circuit losses such as voltagedrop

• Select equipment based on expected moisture,temperature, radiation, and salt environmentalconditions

• Consider safety, lightning protection, andgrounding

Aesthetics • Blend array with building lines and colors

• Tilt array at slope and orientation

• Avoid harsh contrasts and patterns

• Make mounting hardware inconspicuous




Equipment Location

The site visit is also an opportunity to plan the location of the control equipment, batterybank, and other wiring concerns. Ease of access to the equipment and safety should

be considered. Building new cabinets or structures, or modifications of existing ones,may be planned at this time.

The location of a ground rod or locating of a suitable grounding pipe should be planned.The quality of the soil may impact the type of grounding scheme used. A great distancebetween the array and the control and battery systems may cause the designer toconsider two ground rods, one at the array and another near the battery and controls.This would impact the grounding wiring plan.

Equipment location is very important. One should consider the following:

• Install controls, power conditioning equipment, instruments, and batteries in such a

way that access is controlled

• Plan to post a clearly written method for disconnecting power to equipment beforeservicing and for reconnecting

• Consider that the batteries will need to be vented and protected from sparks. Donot locate the controller or inverter over the batteries

• Do not locate an engine generator in the same room as the batteries if they arevented lead-acid type

• Plan for properly sized electrical disconnect switch or circuit breaker for all source

circuits and to isolate major components such as the controller and the array




Foundation Design

Purpose of Foundations

The purpose of a foundation is (a) to prevent the array from sinking; (b) to maintainproper array orientation with respect to the sun; and (c) to prevent the array from"uprooting" during severe winds. Many factors are involved when choosing an arrayfoundation: site accessibility, local topography, soil properties, local building codes, andavailability of labor.

Types of Foundations

The slab type foundation requires a large mass of concrete and relatively flat terrain.

The slab can be poured at the site or prefabricated slabs can be transported to the site.It is thus not well suited for remote applications where the cost of transporting cementto the site is prohibitive. Nor is it desirable in very rugged terrain because of the siteexcavation that would be required before placing the foundation.

The block type foundation is most ideally suited for rugged terrain and remote locationsbecause it is relatively light and transportable, and can be prefabricated wherevercement and equipment are available. Little excavation is needed and the blocks canusually be positioned fairly easily, minimizing alignment problems. Blocks must beassembled with reinforcing steel, and all cavities fully filled with concrete or mortar.

A variation of the block type foundation is to use sauna tubes (with concrete) or woodenpressure treated posts to make a beam foundation. In this type the tubes or posts areset in below the frost line. Beams of wood, steel or aluminum create a frame acrossthe tops of the posts and tubes. Bolted to this frame is the module mounting hardware,typically a standoff type. The beam foundation is a compromise between the slab andblock types. It is well suited for "rolling hill" terrain, and provides easy alignmentbetween adjacent arrays.

There is no foundation that is applicable to all situations all the time, for any suchfoundation would have to be grossly over designed and would be uneconomical in mostsituations. The foundations discussed represent the most common designs used for

PV power systems. Although variations in these foundations exist, one of them willgenerally be suitable for a particular application.

If a pedestal or tracker structure is to be used, consult the manufacturer for therecommended foundation. Often simply setting the vertical pole in a hole filled withconcrete will be enough.




Loading Values

The foundation weight must be determined as a part of the foundation design process.The weight depends upon expected array loading and soil type.

Array loading includes wind force pushing against or lifting the array, snow pack thatmight accumulate on the array, and the dead weight of the modules. The SiemensSolar Standard Support Structure has been designed to withstand 50 lb./ ft

2 loading

acting directly on the solar panels. It would take a 140-mph wind (225 km/h) to producethis force or one meter of snow. A foundation should be designed to support an arrayduring these "worst load" conditions.

R1 R4

R2 R3

Forces on Foundation from Standard Support Structure

The reaction forces to this loading are shown below, both for horizontal and verticalforces. The values shown are not the required foundation weights, but are the forces

that will be transmitted to the foundation during worst load conditions. For example, ifan 8-module structure at 45-degree tilt was subject to worst load conditions, the verticalforce transmitted to one front foot would be 75 lbs. and a rear foot would experience700 lbs. vertical force. The front foot would be pushed horizontally with 325 lbs. andthe rear foot would feel 300 lbs. push horizontally. To remain stationary, the foundationmust react with at least this much force in the opposite direction.




The reaction forces have both positive and negative values because the forces act inboth directions. The forces will be in one direction when the array is loaded from thefront, but will act in the opposite direction when the loading is on the back of the array.

Vertical forces are resisted by the sheer weight of the foundation. Resistance to

horizontal forces is related to soil density, cohesiveness, and soil aggregate and tofoundation weight and design. In general, an engineering soil analysis may need to beconducted on site, and the results analyzed to determine required foundation weightand design.

The values presented below are not weights, but the force that the foundation mustresist. The final weight and shape of the foundation to resist these forces are derivedfrom knowing the soil characteristics as well as the bulk weight of the foundationmembers. All values are for both directions with respect to the “R” vectors.

LOAD FORCES ON FOUNDATION

Tilt Angle R1 R2 R3 R4

4-Module Structure

15 175 425 525 25020 150 300 425 25025 150 200 350 25030 125 125 325 25035 100 50 300 25040 75 50 300 250

45 50 50 275 25050 50 50 275 25055 50 75 300 25060 50 100 275 27565 125 125 325 325

8-Module Structure

15 250 300 525 57520 125 175 125 67525 100 200 150 70030 50 250 175 70035 50 275 225 70040 50 300 250 70045 75 325 300 70050 150 350 325 70055 200 350 375 72560 275 350 425 72565 475 425 400 850




Array Mounting Configurations

There are a number of ways to mount photovoltaic panels. The two major types areground mounts and roof mounts. Ground mounts are usually done in rack mount or ona pedestal, as is the case for tracking. Roof mounts are either standoff or directly

integrated onto the roof.

Ground Mount - Rack • Most common and preferred mounting configuration

• Works well also on flat roofs

• Fixed azimuth orientation that can be adjusted foroptimum tilt

• New construction or retrofit

Ground Mount -Pedestal or Tracker

• Used for smaller arrays

• Allows flexible tilt and azimuth orientation

• Can be used on rough terrain

• Wind loading must be considered

Roof Mount - Standoff • Mount arrays > 8 cm above roof to allow air flow

• Minimize number of roof attachments

• Hinged panels provide safe accessibility

• Pre-assembled sub arrays can minimize installationtime

Roof Mount - Direct • Attach directly to roofing felt• Operates 15 to 30 degrees C hotter than equivalent

rack mount

• May work well with new thin film materials




Battery Location and CompartmentDesign

Batteries should be located in a dry location out of the elements. They should be on alevel surface that is structurally sound for the weight required. When batteries areracked they can weigh several tons over a small surface area (< 10 ft

2 or 1 m

2).

Batteries should not be in contact with cold surfaces such as concrete, and should alsobe located where access is controlled such as a lockable room or battery compartment.

The battery should be enclosed in a container, box or room that allows for: (a) properventilation of evolved gasses; (b) safe and easy access for maintenance andreplacement; (c) reduced exposure to environmental extremes and temperature swings;(d) restricted access to untrained or unauthorized persons; and (e) isolation frompotential hazards such as sparks from generators.

If the temperature of the site will get very cold, burying the battery compartment shouldbe considered. An insulated space below the frost line will keep the batteries at a morereasonable temperature throughout the year. The space should be heavily insulatedand made watertight.

The battery compartment should be vented to meet the relevant code requirements. Alarge battery bank may require fans, while a small battery bank may need only a venthole or tube. Filling the vent tube with loosely packed fiberglass or other suitablematerial is one way to allow escape of hydrogen gas without allowing gusts of wind fromoutside to cool the chamber. Another means of venting batteries is to drill a hole ineach battery cap and attach it to a plastic tube that is then vented to the outside through

a central receiving pipe. Recombination caps or “catalytic recombiners” will minimizebattery gassing and reduce the need for ventilation.

A battery box can be built using plywood and rigid insulation or other suitable materials.It should allow access for maintenance, and allow safe lifting or sliding for removal ofbatteries when necessary. The box should be closed with a lock or other means toprevent casual contact by unauthorized persons. It is good practice to place batteriesin a plastic vessel to contain any spills of the electrolyte.

In many small systems, batteries are placed outside in weatherproof or rainproof metalor fiberglass boxes (e.g. NEMA 3r, NEMA 4x). In these cases, it is important to

insulate the boxes to minimize temperature extremes that can reduce battery life.Venting may be necessary if the batteries are a flooded lead acid type. It is importantthat the batteries are isolated from “HOT” electrical components that might ignite thehydrogen gas.




System Component Pre-Wiring andTesting

For all but the very small and simple systems, it is a good practice to pre-wire as muchof the system as possible. Pre-wiring forces the installer to plan the installation to fit thelocation, and to have all pieces and tools required for a safe and correct assembly.Faulty equipment can be identified and replaced, mistakes can be avoided and designimprovements can be made.

Electrical Connections

All wire connections, switches, receptacles and breakers should be mounted inelectrical boxes. Following are the common ways of connecting wire.

Wire or Conduit to a Box • Use strain relief connectors with round wire suchas Type SO if the wire will be pulled.

• Use cable connector or Romex connector for non-metallic sheathed cable (NM).

• Use conduit couplings to connect lengths ofconduit.

Wire to Wire • Three acceptable wire-to-wire connections arecrimp, solder, or wire nuts. Crimping with solder ispreferred.

• All wire to wire connections must be mounted in anelectrical box.

• When crimping, use heaving duty industrial typecrimpers.

Wire to Terminal • For large wire or when multiple wires areconnected to one terminal, use screw lugs or bolt-on connectors.

• For small wire use ring terminals and spadeterminals (non-insulated are preferred).

• Spade terminals can be used in non-vibratingapplications, ring terminals for connections thatcan loosen.

• Crimping and soldering the ring or spade isrecommended particularly in areas of highhumidity.

• When crimping, use heaving duty industrial typecrimpers.




Control Center

A primary candidate for pre-wiring is all the control and main safety equipment. Thistypically includes the main disconnect switches for the array and battery, as well as thecharge controller(s) and any alarms and metering equipment, and AC and DCdistribution boxes or load centers. One of the great advantages of pre-wired and code-approved power centers such as the products made by Ananda Power Technologies isthat you do not have to pre-wire and test all the main control and safety equipment.

It is often desirable to pre-wire and actually pre-assemble these components onto aplywood board for transportation to the site as an assembled unit. Carefulmeasurements must have been taken at the site to insure the board will fit properly withthe required safe spacing from walls or other equipment.

Even if transporting to the site fully assembled is not possible, pre-assembly is a good

way to test the overall layout and design, and insures that all details are checked. Endsof wires can be wrapped with a number or letter code, and a legend made, so that inthe field the proper connections can be made quickly.

Locate the control center to avoid temperature extremes. Use insulated enclosures forcold weather, shaded enclosure for hot weather. Many small system controllers aremounted in rainproof boxes such as a NEMA 3r. This should be separate from anybattery compartment unless you are using maintenance free batteries that do not emitgasses.

Array

The array should be pre-assembled and wired, to again insure that all hardware isincluded. Module interconnect wires can be pre-cut and have soldered crimpconnections or wire stripped ready for installation into the module junction box. Thepath for the main wires from an array can be planned, and means for securing thesewires against wind can be planned. The method for grounding the array and mountingstructure should be worked out.




Battery

Planning for batteries involve how to safely handle and move them. Due to their largeweight, it may be necessary to use a mechanical lift to move them. In addition, it isimportant that there is sufficient width allowed to move batteries in and out without

damaging them.

It is best to have batteries delivered a few days before installation. The battery suppliershould charge them before delivery. Before taking them to the site, the installer shouldverify their state of charge and if necessary bring all batteries to an equal full charge.This is also the best time to screen batteries. Quality control can be a problem withbatteries in many Third World countries. By checking voltages and observing how theytake a charge, batteries can be screened for weak cells and poor performance beforeplacing them in the field.

Pre-wiring of the battery bank insures that proper lugs or clamps are obtained and pre-

soldered to heavy gauge wire. Also a method can be practiced for attaching anytemperature sensor to the battery cases, for charge controller temperaturecompensation.

Other Wiring

The full system of array, battery, controls, loads, and inverters should be planned. Wireand conduit can be obtained and cut, and interconnects and fittings acquired. Themounting and wiring of any inverter can be planned. An AC and/or DC load center ordistribution box can be obtained and pre-wired with appropriate circuit breakers or

fuses. All switches for loads can be examined for the best way to interconnect thewires. Using closed lugs on the switches will assure reliable operation even if theconnecting screws loosen slightly.




Planning The Logistics Of Installation

Besides the equipment to be installed, there are other elements that must be taken intoaccount for a successful installation. These primarily concern the logistics on how to

get there, how to get the equipment there, and the planning of what will get done, when,and by whom.

If the installation is to be done at some remote site where access to comforts andplentiful tools will not be possible (often the case for PV power systems!), it becomeseven more critical that all the logistics of the work be thought out ahead of time.

Transportation Plan

Appropriate transportation equipment must be obtained. The vehicles must be large

enough to carry all the equipment in a safe and secure manner. All possibilities forweather should be considered to be prepared for snow, wind storms, rain, and roughterrain or other problems.

There may be a need for small lifting and transporting equipment such as two-wheelhand trucks or dollies to move the batteries or inverters.

Human Needs Plan

Planning for human needs can include sleeping and housing arrangements, taking

adequate food and water or insuring that such will be easily available near the site.First aid equipment should be transported.




Tool Selection

Some tools for installing and maintaining photovoltaic power systems are listed below.Of course other equipment for building and construction will be useful. Careful planningof the installation process will help prevent loss of time due to missing tools and

accessories.

Generally Required or Advised

Digital Voltmeter, with at least 10-ampcurrent capability, spare batteries, external50-100 amp current shunt.

Clamp-on DC ammeter.

Compass and Solar Site Locator

Tilt Angle Indicator, or plumb line andprotractor

Tape Measure

Wrenches: Specific sizes, for all mountingbolts; Adjustable, for unexpected on-siteproblems; Vise-Grips, for variable and heavyduty

Screw Drivers: Flat Blade, in sizes for allmounting hardware; Phillips, in sizes for allmounting hardware; Small jewelry size, foradjusting controls

Wire Crimping, Stripping and Cutting Tool(s)

Electrical Tape

Miscellaneous for connections: Split bolts,wire nuts, lugs, solder-less connectors

Optional

Temperature probe for voltmeter

Light meter

Soldering Iron and solder

Knife

Hammer

Socket Set, English and Metric

Metal Hack Saw, Wood Saw

Chisel

AC Amp meter




Installation ProcessesAs opposed to the pre-installation processes at some comfortable location, the actualinstallation process involves working quickly and safely at the installation site to install

and commission the PV power system and loads.

Foundation Alignment

The most critical element of foundation installation is alignment. The points where themounting structure is to meet the foundation must be level and any mounting bolts mustbe spaced correctly. It is critical that careful measurements be made both for spacingand for flatness.

The orientation of the foundation must face true (solar) south (unless local climate orload factors require facing away). Care should be taken to adjust for magneticcorrections to compass readings.

Array Assembly Procedure

Once the mounting structure is in place, it is time to attach the panels and wire themtogether with interconnects. Make sure you have thought through your wiring planbefore starting and, if necessary, refer to a schematic. It is important at this stage to

remind oneself about safety. Large arrays can produce currents sufficient to shockand/or hurt you. If the array is exposed to sunlight while wiring, cover the arraywith blanket or other opaque covering to reduce the hazard of electrical shock.

Be careful to follow proper wiring techniques including using correct colors, fasteners,connections, etc. Mark wires as positive or negative if a color convention cannot befollowed. Leave proper documentation if necessary.

The following general procedure outlines the steps to install Siemens Solar modules ona Siemens Solar Standard Support Structure, consisting of two long "C" channels andtwo adjustable support legs with angle feet. Some of this procedure will also apply to

other mounting, such as on trackers. Consult the manufacturer of any mountingstructure for recommended procedures. If you are using the Siemens Solar StandardSupport Structure, follow the guide enclosed with each structure.




Exercise:

°

°

°




Step Array Installation Procedure:

1 Check that all tools and mounting hardware are available and inworking order. Check that foundation is properly installed andlocations for mounting feet (if applicable) are spaced correctly.

2 Remove modules from shipping boxes and inspect. Check electricaloutput of each module at Voc, Isc and using a 50 watt 6 ohm resistor toapproximate maximum power.

3 Place modules face down, side-by-side, on a clean flat non-abrasivesurface. The shipping boxes will work as a "table" if desired. Arrangethe modules with the junction boxes located so that series and parallelwiring will be efficient.

4 Place the two support structure "C" channels on top of the modules,parallel, with the "C" facing inward and the mounting holes facing downaligning with the mounting holes in the modules.

5 Attach the modules to the structure channels using a 1/4" stainless

steel bolt, a flat washer, a lock washer and a nut in each of four places.Fasten only finger tight at this point. Continue for all modules on thestructure.

NOTE: Up to 8 modules can be mounted on a 9-foot long structure.There should be about 1 foot of empty space left between the lowestmodule and the ground, to allow for debris to blow free and not build upagainst a module surface. Modules should be mounted flush to the topof the structure. Empty spaces can be filled with a cross-brace,available from Siemens Solar.

6 Attach all module interconnect wires. Use sunlight resistant type UFwire or equivalent. Remove about 5/8" of insulation (3/4" for AWG #8wire). Do not use wire lugs. Insert stripped wire through foam seal and

under terminal screw slider. Tighten screw securely using ascrewdriver. Attach large sized strain relief clip enclosed with modules.For small wires (AWG #14 and #12), use the second smaller strainrelief clip also.

7 Assemble the two rear support legs to the correct length to give theproper final tilt angle. Use four 3/8" bolts, washers and nuts to fullyassemble each support leg.

8 Attach four-foot angles to foundation.

9 Attach each rear support leg to a rear foot angle using 3/8" bolt, flatwasher, locking washer and nut.

10 Move module assembly into place with modules face up, and attachbottom of side channels to front foot angles.

11 Lift up assembly and bring ends of rear support legs up to meet theassembly. Attach each rear support leg to a side channel using a 3/8"bolt, washer, locking washer and nut.

12 Tighten all 1/4" nuts and bolts to 80 in-lbs., and all 3/8" nuts and boltsto 230 in-lbs.




Battery Bank Assembly

First and foremost, review battery safety procedures discussed at the beginning of thischapter.

Do not mount the batteries directly on a concrete floor. Mount onto wooden or othernon-conducting rails.

Make sure the batteries are fully charged, and that the electrolyte level is atmanufacturer's recommended level. Check all cell voltages and write down on a statussheet for later comparison.

Handle batteries with extreme care, and use tools carefully. The greatest danger willoccur if wires are hastily connected, or if tools are dropped onto the bare batteryterminals. All connections should be "walked through" a few times, perhaps withanother installer present to confirm, before actual wiring is done.

Make sure to place a sign at the batteries warning unauthorized personnel about thedangers.

Control Center

Assemble all controls, disconnect switches, alarms and meters and load centers as pre-planned.

Make sure there is the required safe space between any boxes and walls or pipes orother equipment.

Check that all connections are secure and clean, that all wiring is color coded ormarked for correct polarity and that all wire lugs are fastened tight to their wires.

Before hooking up the inverter, make sure the polarity is correct. Most inverters canbe damaged if hooked up with the wrong polarity.

Leave a schematic of the system wiring (secured to the control center if there isappropriate room) for later maintenance and troubleshooting. Label all main switches

so that an uninformed person with the operation guide can safely and easily disconnectthe system in an emergency.




Other Wiring

Connect all equipment with proper wire and conduit if necessary.

Make sure all wiring is secured by gently but firmly pulling on all connections. Checkthat insulation is not cut during installation work, or when wrapped around corners ornear edges of conduit.

Make sure that the color coding of the wire used is consistent and conforms to localcodes.

Commissioning

The full system should be tested and put into commission before the installers leave thesite.

All safety equipment and switches and circuit breakers should be cycled and tested.

Some controllers and inverters have a very specific hook-up sequence to bring themon-line. Be prepared to follow that sequence.

All hard-wired loads should be operated and all outlets should be tested.

Restricted equipment should be locked and the keys secured.

A system description with circuit diagrams and an operation/maintenance plan shouldbe presented to the client. The designer or installer should read through theoperation/maintenance plan with the client. Perhaps have them sign a copy assertingthat they have read the instructions. A copy of the operation/ maintenance plan shouldbe left with the client. All major components and switch gear should be labeled and beeasily accessible for service.




Maintenance and TroubleshootingA maintenance program should be developed and followed for every photovoltaic powersystem. The maintenance program should include consideration for structural,

electrical and mechanical components of the system. It should be clearly and simplywritten so that the system owner can easily maintain the system.

The maintenance requirements of a typical photovoltaic power system are quite simplecompared to those for a typical conventional power system such as a diesel generator.There are very few mechanical components to wear out, and little need for adjustmentsafter the system has been first commissioned. However, like anything placed in theenvironment for years, a system will benefit from occasional attention to corrosion andproper performance.

Maintenance includes procedures that are expected and normal for the system.

Troubleshooting covers procedures that should be followed if system performance isout of normal tolerances. All safety precautions previously mentioned should befollowed.

Batteries Are The Main Focus

The most important maintenance item for a solar electric system is taking care of thebatteries to maximize their service life. Care of the battery is primarily a matter ofmonitoring and maintaining their state of charge. Each battery type requires certain

maintenance. The most commonly used battery is flooded lead-acid and needs to bewatered often and equalized periodically.

A well-designed system will come with some type of voltage or state-of-charge meter.The system owner should be taught how to use the meter to find out the battery state-of-charge. Ideally batteries kept between 70% and 100% state-of-charge (SOC)indicates a healthy system and leads to long battery life. Consistently low SOC isindicative of excessive use of the PV system. Understanding the usage of the battery(excessive or light) will be very important in determining maintenance intervals, servicelife, and will help in diagnosing system problems in the future.




Sample Maintenance Schedule

A sample maintenance schedule is presented below to indicate typical frequencies ofmaintenance actions.

Weekly Maintenance • Observe battery state charge

Monthly Maintenance • Check and add water to the battery electrolyte

• Wipe electrolyte residue from top of batteries

• Inspect the charge controller for proper indicator lightsequence

• Equalize batteries if specific gravity difference betweenany adjacent cells is greater than 15 points(e.g. 1.250 vs. 1.265)

• Observe controller for proper indicator light sequence• Inspect array for broken panels. Wash array.

Annual Maintenance • Check array wiring for physical damage and windchafing

• Check array mounting hardware for tightness

• Inspect inverter. Remove dust or dirt.

• Inspect system wiring for poor connections. Look forsigns of excessive heating

• Inspect battery terminals for corrosion. Clean andapply anti-oxidant grease as necessary

• Inspect controller for proper operation

• Verify output from the array(Isc and Voc, and if possible Imp and Vmp)




Battery Maintenance

Physical Considerations

Observe water level in every battery cell and fill to lever indicated with distilled water.

Inspect all terminals for corrosion and loosened cable connections. Clean and tightenas necessary. After cleaning, add anti-oxidant to exposed wire and terminals.

Examine outer surface of batteries for film of liquid. This may indicate excessivegassing, bubbling electrolyte out of battery, it also may indicate a controller problem.

Confirm that batteries are not sitting directly on floor. Some elevation allows forcleaning and refilling water to flow completely away from battery cases, and prevents

bottom of battery from operating at lower temperature that top of battery.

Make sure the battery enclosure is well ventilated. Inspect any ventilation pathway orsystem for blockage.

Electrical Considerations - Determining Battery State of Charge(SOC)

There is some controversy as to the best way to determine battery state of charge. It isgenerally accepted among designers that the most accurate way is to use a hydrometer

to measure the specific gravity of the battery cells (if batteries are vented). This methodis cumbersome, can expose the installer to acid, and can inadvertently lead tocontamination of the cells. As a result, for ease and general accuracy, a voltagereading (while batteries are at rest, not being charged or discharged) is preferred bymany in the industry. Fortunately more and more controllers are coming withreasonably accurate voltmeters that make the appropriate adjustments for temperature,charge and discharge cycle. This is certainly the trend for the future.

To determine specific gravity, remove battery caps and measure specific gravity of allcells using an accurate hydrometer. Determine the battery state of charge by referringto manufacturer's literature for the mix of electrolyte specified. It is common in hot

climates to install batteries with a reduced concentration of acid (lower specific gravity)to reduce internal battery corrosion due to the high temperatures. The greatestaccuracy occurs when corrections for electrolyte temperature are included. Alwayswear gloves and eye protection when working with the acid electrolyte.




To determine voltage, measure the open circuit voltage of each battery cell, or of eachblock of cells if internally connected. The best time of day is late afternoon, after thearray has had the most time to charge the battery. Ideally, this should be done afterdisconnecting any array and load connections and waiting at least 15 minutes after anycharging or discharging has occurred. Sometimes it is helpful to first discharge the

batteries for a few minutes to remove any surface charge from the battery plates.Remember to compensate for temperature.

Use the manufacturer's chart of state of charge vs. Voltage (or the figure presented inthe chapter on Battery Technology) to determine the state of charge from the voltagereading. If the batteries must remain connected, then measure the charging ordischarging current. Use the manufacturer's charts for voltage at different currents toestimate the battery state of charge.

Presented below are typical values for state-of-charge, specific gravity and voltage, forlead acid batteries in temperate climates (25 deg C):

SOC Specific GravityBattery Voltage

12 volt 24 volt

100% 1.265 12.68 25.3590% 1.250 12.60 25.2080% 1.235 12.52 25.0570% 1.225 12.44 24.8860% 1.210 12.36 24.72

50% 1.190 12.28 24.5640% 1.175 12.20 24.4030% 1.160 12.10 24.2020% 1.145 12.00 24.0010% 1.130 11.85 23.700% 1.120 11.70 23.40




Array Maintenance

Physical Considerations

Replace any broken modules. If full replacement is not possible at time of visit check ifthe electrical requirements allow for wiring around the broken module, to allowcontinued array output without the broken module.

Wash module surface as needed, using soft cloth and water. Perform cleaning in earlymorning or wait until evening, to avoid thermal shock to glass. Perform washing onlywhen modules are not in direct sunlight, when the sun is positioned below the horizon.

Check module tilt angle using tilt indicator or level and protractor. Tilt angle should bewithin five degrees of that specified by sizing calculations.

Check module azimuth angle using compass. Compensate for local deviations of trueor solar North from magnetic North. Normally azimuth is true south in the NorthernHemisphere and true north in the Southern Hemisphere. However different angles maybe chosen depending on local weather conditions and load profile matching. Azimuthangle should be within 15 degrees of that specified by sizing calculations.

Confirm that no objects cast shadows on the array surface between 9 a.m. and 3 p.m.

Verify that all bolts are secure, and that the mounting structure is well attached to anyfoundation.

Examine all wiring connections for corrosion or looseness. Clean and tighten asnecessary.

Check that all junction boxes are covered and that any seals are in place.

Inspect modules for broken cells, de-lamination, and discoloration. These alone maynot result in any noticeable deterioration in electrical output, but may be useful if otherproblems are observed.




Electrical Considerations

It is very helpful to periodically verify the electrical output of the array.

Measure Short Circuit Current (Isc) using clamp-on meter or passing current directly

through meter. Isc is directly proportional to sunlight intensity. Use a reference cell orlight meter to independently check light level. The ratio of light level to 1000 w/m

2

should be the ratio of Isc to literature value.

Measure Open Circuit Voltage (Voc) using meter. Voc decreases with increasingtemperature. Use a reference cell or light meter to independently check light level anda temperature gauge to measure ambient (air) temperature. Use the graph of celltemperature above ambient temperature to estimate actual cell temperature. Calculatedrop in Voc for operation above 25 deg.C by using voltage factors presented in Chapter8. Compare calculated value to measured value.

Actual module current will probably not be at literature values, due to low irradiance andhigh temperatures. A rough guideline for module current into a battery is given below.If a module is connected to a 12-volt battery and operating under common outdoorconditions (irradiance about 800 w/m

2 and ambient air temperature about 30 deg.C),

then the following currents should be observed:

Typical Field Values

36 cell module (75 W)

nominal current 4.4 amps

3.5 amps

36 cell module (53 W)nominal current 3.1 amps

2.5 amps


2.4 amps


1.6 amps




Field Measurement Using Resistors

If possible, measure array voltage and current at some intermediate point between Iscand Voc (preferably near Vmp) by using some dummy load. Use 6-ohm (approximate)resistor for a single 40-50 watt 12-volt module. Multiply this resistance by the number

of modules in series and divide by the number in parallel to determine resistance to usefor an array. This resistance will operate the array near maximum power. Thismeasurement gives more information about the curve shape of the array.

Resistancearray = 6 ohms X # Modules in Series# Modules in Parallel

Res istive Load Sized To

Check Module

0 4 8 12 16 20

Voltage (volts)

Current(amps)

3

2

1

0

(Vmp, Imp)

R = Vmp/Imp = 6 ohms

R e s i s tiv e L o a d S i z e d T oC h e c k A r ra y

R = 6 o h m s X # S e r ie s

# Para l le l




Electronic Equipment Maintenance

Check that all controls, alarms, meters, etc. are securely attached.

Follow manufacturer's procedures to verify that all controls, etc. are functioning properlyand are calibrated. If possible, include verifying the voltage setting of the chargecontroller, to insure that batteries are not over- or under-charged.

Inspect all connections for corrosion and loosened wire connections. Clean and tightenas necessary.

Inspect relay contact points for pitting and corrosion. Replace if damaged.

Load Maintenance

Lubricate all moving parts of motors and loads as recommended by the manufacturer.

Check refrigerator door seals. Clean refrigerator coils.

Clean or replace any air filters.




Wiring and Safety EquipmentMaintenance

Check that all safety disconnects are functioning.

Examine all wiring connections for corrosion or looseness. Clean and tighten asnecessary.

If possible measure actual voltage drop along wires during normal operation.

Check that all wiring is secured, and inspect insulation for wear, especially at bends andwhere secured.

Confirm that all fuses are conducting.

If possible, safely introduce a short circuit path and check that circuit breakers arefunctioning.

Verify that all metal objects involved in the electrical system (boxes, raceways, conduit)are properly grounded, and that the grounding path is continuous all the way to theground rod or grounding object.

Optionally check potential between grounding rod and surrounding earth with properequipment.

Verify that properly sized wire has been installed. Determine the maximum current that

might flow for each circuit, and the total length of each circuit. Use charts (for examplein Chapter 8) to determine the proper size wire so that there is not more than 5%voltage drop from array to battery and from battery to load.




Troubleshooting

The following guide is presented for troubleshooting of generic photovoltaic powersystems. A more detailed and relevant guide can be created for a specific system with

input from the various component manufacturers.

Symptom Chart

Symptom Probable Cause

Array Wiring Controller Load

Low battery voltage A,B,C,D,E F,G,H I,J,K,M P

High battery voltage B I,L

Load inoperative A,B,C,D,E F,G,H I,J,N O

System erratic or noisy F,G,H I,M Q

Exercise:




Corrective Action Chart

Cause Corrective Action

Array

A Loose wires TightenB Improper wiring ReconfigureC Broken module ReplaceD Shading Remove obstruction, or relocate arrayE Improper

orientationReorient

Wiring

F Excessive voltage

drop

Use larger size wire

G High resistanceconnections

Clean and tighten

H Short circuit Repair

Controller

I Looseconnections atterminal block

Tighten

J High resistance atrelay contacts

Replace controller

K Inoperative Replace controllerL Relay fused shut Replace controllerM Electrical

interferenceInstall filter, or remove source

N Load sheddingoccurred

Wait for battery to recharge, check actualload usage against sizing calculations.

Load

O Failed Replace. Check for improper voltage.P Excessive current

demandCheck actual load usage against sizingcalculations, check load for failure.

Q Generatingelectrical noise

Filter the circuit




Field EvaluationAn example of useful field evaluation procedures is presented next. Field techniciansor engineers investigating the status or performance of an installed system can use

such worksheets. All parameters may not apply in all systems, but these provide youwith the basis for a thorough assessment of how a system was installed and is working.

The first field evaluation worksheet is for complete system evaluation. This can beused for commissioning a system or for checking a system after some time ofoperation. It is especially useful in the context of checking a system installed by someother organization than your own, either by government agencies checking the qualityof subsidized systems, or by a company called in to service a system installed by someother organization.

The second field worksheet is for checking the operation set points of a charge

regulator. This simple procedure is needed in the field because the system may notcycle through its extremes of operational values during a typical field visit within a shortperiod of time. The field technician or engineer must force the regulator to operate toits set point values to check for proper operation.




Solar Photovoltaic System Field Evaluation Worksheet

PRELIMINARY INFORMATION

Date: ____________________ Inspector Name__________________________ System Location: ___________________________________________________________ System Description: ___________________________________________________________

General Application (End Use): ______________________________________________________________________________ ______________________________________________________________________________

PHOTOVOLTAIC MODULES AND ARRAY

Manufacturer: _______________ Model #_______________

Array Configuration: __________ series X __________ parallelTotal Number of Modules: __________ (#)Array Mount Design/Type _____________________________ (ground/roof, rack, tracking)Array Tilt Angle __________ (deg) Array Azimuth _________ (deg)Seasonal or Daily Adjustable Tilt or Tracking (describe) _________________________________

Module/Array Specifications @ STC (1000 W/m 2 , 25

o C )

Module Array Open-Circuit Voltage (Voc) __________ (V) __________ (V)Short-Circuit Current (Isc) __________ (A) __________ (A)Maximum Power Voltage (Vmp) __________ (V) __________ (V)Maximum Power Current (Imp) __________ (A) __________ (A)

Maximum Power (Pmp) __________ (W) __________ (W)

BATTERY SUBSYSTEM

Battery Manufacturer: _______________ Model #: _______________ Battery Chemistry/Type _______________ (antimony, calcium, VRLA)Battery Bank Configuration _______________ series X _______________ parallelTotal Number of Batteries _______________ (#)

Battery Specifications at 25o C

Cell/Battery Bank

Capacity __________ (Ah) __________ (Ah) Rate __________ (C/#)Nominal Voltage __________ (V) __________ (V)

Open-Circuit Voltage at 100% SOC __________ (V) at 50% SOC __________ (V)Specific Gravity at 100% SOC __________ (#) at 50% SOC __________ (#)

Cycle Life ____________ (cycles) __________ (20% DOD) ____________ (cycles) __________ (50% DOD) ____________ (cycles) __________ (80% DOD)




ELECTRICAL LOADS

Device DC/AC Amps/Watts Hrs/Day Ah/Day 1. ______________ __________ __________ __________ __________ 2. ______________ __________ __________ __________ __________

3. ______________ __________ __________ __________ __________ 4. ______________ __________ __________ __________ __________ 5. ______________ __________ __________ __________ __________ 6. ______________ __________ __________ __________ __________ 7. ______________ __________ __________ __________ __________ 8. ______________ __________ __________ __________ __________

Seasonal Variation in Load? ___________________________________________________ Special Load Performance (water volume, illumination levels, etc.) _____________________

___________________________________________________________________________

Total Electrical Load

DC Loads: Peak Amps ________ (A) Avg. Daily ________ (Ah)AC Loads: Peak Watts ________ (W) Avg. Daily ________ (Wh) Avg. Daily ______ (Ah)Total Average Daily Load DC & AC ________ (Ah)

CHARGE CONTROLLER/VOLTAGE REGULATOR

Manufacturer ____________________ Model # _______________ Regulation Type ____________________ (series, shunt, PWM, constant-voltage)

Controller Specifications:

Nominal Voltage ________ (V)Maximum Rated Currents: PV Array ________ (A) Load ________ (A)

Array Regulation Set Points: VR: ________ (V) VRH: ________ (V)Array Regulation Switching Element: ________ (SS/EM) Leg: ________ (pos/neg)

Load Disconnect Set Points: LVD: ________ (V) LVDH: ________ (V)Load Disconnect Switching Element: ________ (SS, EM) Leg: ________ (pos/neg)

Temperature Compensation: __________ (mV/ oC/cell) Type: ________ (probe/board)

Other Controller Functions/Characteristics (meters, indicators, alarms, etc.) _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________




INVERTER/POWER CONDITIONER

Manufacturer: ____________________ Model #: _______________ Waveform ____________________ (sine, quasi-sine, square)Output Voltage Regulation __________ (%)Harmonic Distortion __________ (%)

Minimum Input/Operating Voltage __________ (V)LVD Function On Inverter DC Input __________ yes ________ no ________ (V)

Inverter Specifications (25o C)

Nominal DC Input Voltage __________ (V)Nominal AC Output Voltage __________ (V) Frequency _________ (Hz)Maximum Continuous Output Power __________ (W)Standby Power Requirement (zero load) __________ (W)Surge Capability: __________(W) ________ (minutes)

__________(W) ________ (minutes)Nominal Efficiency (resistive load) __________(%) at 25% Load

__________(%) at 50% Load __________(%) at 75% Load __________(%) at 100% Load

Other Inverter Features (battery charger, disconnects, over-current protection, etc.): ____________________________________________________________________________________ ____________________________________________________________________________________

OTHER SYSTEM SPECIFICATIONS AND INFORMATION

Over-current Protection and Disconnect Devices (list types, locations and ratings): __________________________________________________________________________________ __________________________________________________________________________________ Grounding (system grounding, equipment grounding, earth electrode):

_________________________________________________________________________________ _________________________________________________________________________________ Surge Protection Devices (list types, locations and ratings):

_________________________________________________________________________________ _________________________________________________________________________________ Blocking and Bypass Diodes (list types, locations and ratings):

__________________________________________________________________________________ __________________________________________________________________________________

Conductors and Wiring (list diameters {sizes}, types, lengths and color codes):PV Array Interconnects: _____________________________________________________________ PV Array J-Box to Controller: ________________________________________________________ Controller to DC Loads: _____________________________________________________________ Battery to Controller: _______________________________________________________________

Battery to Inverter: ________________________________________________________________ Inverter to AC Loads: _______________________________________________________________

Other Information: __________________________________________________________________________________ __________________________________________________________________________________




SYSTEM EVALUATION AND DATA MEASUREMENT RECORD

Photovoltaic Array

Sub Array ArrayOpen-Circuit Voltage (Voc) ________ (V) ________ (V)

short-circuit Current (Sic) ________ (A) ________ (A)Irradiance (sunlight) ________ (W/m

2)

Temperatures: Array ________ (OCR) Ambient ________ (

OCR)

Normalized Array Short-Circuit Current (1000 W/m2) ________ (A)

Within Specification? ________ yes ________ no

Array to Battery Charging Current ________ (A) Irradiance ________ (W/m2)

Battery Voltage ________ (V)Normalized Charging Current (1000 W/m

2) ________ (A)

Battery Subsystem

Battery Temperature ________ (oC)

Open-Circuit Cell Voltages: 1. ______ (V) 2. ______ (V) 3. ______ (V)4. ______ (V) 5. ______ (V) 6. ______ (V)7. ______ (V) 8. ______ (V) 9. ______ (V)10.______ (V) 11.______ (V) 12.______ (V)

Avg. ________ (V) Minimum ________ (V) Maximum ________ (V)

Specific Gravities: 1. ______ (#) 2. ______ (#) 3. ______ (#)4. ______ (#) 5. ______ (#) 6. ______ (#)7. ______ (#) 8. ______ (#) 9. ______ (#)10.______ (#) 11.______ (#) 12.______ (#)

Avg. ________ (V) Minimum ________ (V) Maximum ________ (V)

Need for Equalization (difference of more than 0.15 SG): ________________________ Battery State of Charge Estimate ________ (%)

Battery Load Test (C/1-C/0.3, 5% of capacity): ________ (A) ________ (min) ________ (Voc) ________ (Vend) ________ (Vrest, min)

Electrical Loads

DC Load Currents 1. ________ (A) 2. ________ (A) 3. ________ (A) Voltage ______ (V) 4. ________ (A) 5. ________ (A) 6. ________ (A)

AC Load Power 1. ________ (W) 2. ________ (W) 3. ________ (W)

Total DC Load Current ________ (A) Total AC Load Power ________ (W)All Operational?________ yes ________ no

Special Load Operational Performance (water volume/head, illumination levels, etc.) _________________________________________________________________________________




Charge Controller Operation

Voltage Drops Through Controller: PV to Battery ________ (V) at Current ________ (A)Battery to Load ________ (V) at Current ________ (A)

Set Point Tests: Controller Regulation ________ (VR) ________ (VRH)Load Disconnect ________ (LVD) ________ (LVDH)Within Specifications ________ yes ________ noQuiescent Current ________ (A)

Other Functions - Operational Performance: _______________________________________________ ____________________________________________________________________________________ ____________________________________________________________________________________

Inverter Operation

DC Voltage (V) DC Current (A) DC Power (W) AC Power (W) Efficiency (%) __________ __________ __________ __________ __________ __________ __________ __________ __________ __________ __________ __________ __________ __________ __________

Standby Power at Zero Load __________ (W)

System Operation

Steady-State Operation (no controller regulation, no changes in load or irradiance)

PV Array Current ________ (A)

DC Load Current ________ (A)Inverter DC Current ________ (A)Controller Quiescent Current ________ (A)Sum of all currents should add to zero _________ (check)

Other Noteworthy Comments and Data ____________________________________________________________________________________ ____________________________________________________________________________________ ____________________________________________________________________________________ ____________________________________________________________________________________

____________________________________________________________________________________ ____________________________________________________________________________________ ____________________________________________________________________________________ ___________________________________________________________________________________




(This page is blank)




In-Field Measurement of Charge Controller Set Points

Why a Procedure Is Needed to Measure Charge Controller Set Points In the Field

In general, it is not possible to determine charge controller set points in the field during system operationbecause:

• Often the system will not be in regulation.• More often, the system will not be in a load disconnect condition.

• The time need to observe these conditions in actual system operation may be a very long time.

For these reasons, a procedure is needed to force the regulator to operate to its set point values, to allowthe field personnel to witness first-hand whether the regulator is operating within acceptable limits.

Charge Controller Set Points - Review

•

Regulation Voltage (VR):

the maximum voltage the controller allows the battery to reach before thearray is disconnected.• Regulation Voltage Hysteresis (VRH): the voltage span between the VR and the voltage at which

the array is reconnected to the battery.• Low Voltage Disconnect (LVD): the battery voltage at which the controller disconnects the load,

defining the maximum DOD.• Low Voltage Disconnect Hysteresis (LVDH): the voltage between the LVD and the voltage at which

the load is reconnected.


Time

B a t t e r y V o l t a g e

Charging Discharging

Low Voltage Load Disconnect (LVD)

Voltage Regulation (VR)


Low Voltage Disconnect Hysteresis (LVDH)

Load Reconnect Voltage (LRV)

Array Reconnect Voltage (ARV)




Procedure and Methodology

The charge controller must be isolated from the system to perform set point tests.• Disconnect the PV array, battery and DC loads via switches or circuit breakers provided in the system.

• Remove all conductors from the terminals of the charge controller.

Connect a small battery or electrolytic capacitor to the charge controller battery terminals.• The battery voltage must be the same as the nominal charge controller voltage.

• If a battery is used, the capacity should be low, approximately 5-10 amp-hours (ensures that the battery can be cycled in ashort period of time)

• If a capacitor is used, the working voltage must be above the array open-circuit voltage. The capacitance should be between5,000 and 50,000 micro-farads.

Observe correct polarity!

Connect a PV module (or series string of modules) to the charge controller array terminals.• The array nominal voltage must be the same as the nominal charge controller voltage.

• The module (or series string) should have a low output current, but large enough to quickly charge the small battery if used.Approximately one to three amp modules should be sufficient.

Observe correct polarity!

Connect a resistive load to the charge controller load terminals.• The resistive load should be large enough to discharge the battery at a high rate, at least C/1 but not greater than C/0.25.

• Furthermore, the load must consume a lower current than the test PV array provides.

The charge controller should now be cycling between the voltage regulation and array reconnect setpoints.

Connect a voltmeter to the charge controller battery terminals.• The voltmeter must have the capability to record minimum and maximum voltages.

Wait for the controller to complete a full regulation cycle, and note the minimum and maximum voltagesrecorded on the meter.• These voltages will be the array regulation (VR) and array reconnect set points.

Next, reset the minimum and maximum voltage recording for the voltmeter.

Disconnect the test array from the charge controller array terminals (or shade the array).• Leave the load connected and adjust the load for a current higher than the PV array current.

Wait until a minimum voltage is recorded.• This will be the load disconnect voltage (LVD).

Reconnect (or uncover) the array and wait until a maximum voltage is recorded.• This will be the load reconnect voltage.




Charge Controller In-Field Test

Configuration

VTest PV Module

or Array Voltmeter

Charge ControllerUnder Test

VariableResistanceLoad

Small Battery orElectrolytic Capacitor




Special Considerations

• The relative sizes of the test battery (or capacitor), the PV array and test load must be correct toensure the operation of the controller between all set points.

• Disconnects and overcurrent devices should be used on the battery connection to the charge

controller during the tests.

• Special controller algorithms and timing circuits may require special considerations for these tests.

• A variable resistor may facilitate the performance of these tests.

• Repeat the test a few times to ensure that accurate data is measured. Average the final results fromall of the measurements.




[End of Chapter]




CHAPTER NINETEEN

INSTALLATION AND TROUBLESHOOTING 19-1

Safety Issues 19-2National Electric Code 19-2General Safety Practices 19-3Safe Practices With Batteries 19-4

Planning the Installation 19-7Site Visit and Analysis 19-9Foundation Design 19-16Battery Location and Compartment Design 19-21

System Component Pre-Wiring and Testing 19-22Planning The Logistics Of Installation 19-25

Installation Processes 19-27Foundation Alignment 19-27Array Assembly Procedure 19-27Battery Bank Assembly 19-30Control Center 19-30Other Wiring 19-31Commissioning 19-31

Maintenance and Troubleshooting 19-32Batteries Are The Main Focus 19-32Sample Maintenance Schedule 19-33Battery Maintenance 19-34Array Maintenance 19-36Electronic Equipment Maintenance 19-39Load Maintenance 19-39Wiring and Safety Equipment Maintenance 19-40

Troubleshooting 19-41Field Evaluation 19-43



Siemens Solar Basic PV Technology Course System Design – Economic AnalysisCopyright © 1998 Siemens Solar Industries

20-1

Chapter Twenty Economic Analysis

It is not uncommon to have a request to compare renewable generators (wind, PVhydro) with non-renewable generators (diesel engines, grid-connected systems). Itis very difficult to compare these technologies fairly since their cost structure isentirely different. For example, a diesel generator has a low initial cost while a PVsystem to supply the same energy requirement is significantly more expensive. Onthe other hand, the PV system, once installed, uses no fuel and has very lowmaintenance costs, while the diesel generator requires constant purchases of fueland maintenance on a regular basis. An approach to economic analysis that looksat the total life costs of these different systems is required. This is discussed next.

Why Life Cycle Analysis Is Important

Life cycle cost (LCC) analysis is a tool used to compare the ultimate delivered costsof technologies with different cost structures. Rather than comparing only the initial(capital) costs or operating costs, life cycle cost analysis seeks to calculate the costof delivering a service over the life of the project. This means adding the capital

costs and the operating costs over all the years of operation of the project, thendividing by the total kilowatt-hours (kWh) of electricity provided. This gives a value incost per kWh, no matter what technology is used to deliver the electricity.

The following sections give an overview of the factors and methods used to providea life cycle costs analysis for comparing renewable and nonrenewable energysystems, along with a more detailed explanation of "net present value analysis."




20-2

System CostsTo properly account for all the expenses in a power system that operates over manyyears, three different types of costs should be analyzed. These are: (1) initial capitalcosts of purchasing equipment and installation; (2) recurring costs that occur everyyear of operation such as fuel and maintenance; and (3) nonrecurring costs that may

occur on an irregular basis, such as equipment replacement or repairs. Each ofthese costs is discussed next.

Capital Costs

Generators

The cost of a generator which uses fossil fuel (gasoline / petrol, diesel, or LPG) asits fuel source varies depending on the construction of the engine. Typically,

gasoline is used for portable generators because of the high weight / power ratio ofthese engines. However, these engines are typically only designed for a lifetime of2,000 hours or less, and require near continuous maintenance. In addition, thestorage problems (flammability and "shelf life") associated with gasoline makediesel or LPG the fuel of choice for prime power generation.

Diesel generators are typically specified either as "standby" or "prime" duty cycleengines. The prime rating is usually less than 90% of the standby rating. Inaddition, prime diesels usually have longer maintenance intervals than "standby"engines, savings costs later in the life cycle process. Typical life of a dieselgenerator is 20,000 to 40,000 hours. Single generators are usually sized to meet

annual peak loads, which means that there are typically sized at 6-8 times theaverage load.

The cost of generators depends on the fuel, "duty cycle" and size. Gasolinegenerators are typically less expensive than diesel or LPG generators. Unitsdesigned for "prime" duty are usually significantly more expensive than those ratedfor "contractor" or "standby" duty. Finally, larger generators typically cost much lessthan small generators. AA typical example is that a generator which is four times aslarge will only cost two as much as a base case generator.

For specific examples, a recent "Grainger" catalog lists a gasoline powered 4 kVA

"contractor" generator at $1,089 (approx. $270 per kVA), while a "standby" dieselgenerator rated for 8 kW continuous sells for $7,500 (approx. $940 per kW ofgenerator capacity) and a 35 kW standby generator sells for approximately $15,000(approx. $430 per kW).




20-3

BOS costs such as control systems, fuel tanks, fuel containment structures,generator buildings, redundancy (dual engine systems) and storage / powerelectronics (batteries and rectifiers for telecom systems) add additional costs to agenerator system. Translated into terms of $/kW of load, the capital cost of dieselsystems is usually $10K-$30K per kW of load.

PV Modules

PV modules are typically sold in terms of $/W(p) (or "dollars per peak watt") for aspecific module size. Although there are some discounts for bulk purchases, thecost of a PV system over a wide range of power outputs is much more linear thanthe cost of a diesel generator.

A typical PV array will cost $6,000 to $8,000 per rated kilowatt ($36K-$48K per kWof load), plus array structure and BOS. The array that is required to meet a specificload is typically sized for worst case solar insolation, and is thus 6-8 times the ratingof the load. Thus a 6 kW(p) PV array might be required to meet a 1 kW load in atypical environment.

Batteries

PV systems usually have batteries equivalent to 5-10 days of storage, with thebatteries rated to 80% of design capacity. Thus, a 1 kW PV system would have120-240 kWh of useful capacity, equivalent to 150 to 300 kWh of rated capacity.Batteries usually cost $125-$250 per kWh of capacity, so the net cost of a battery fora 1 kW load PV system is $19K-75K.

Diesel based telecom systems usually have batteries rated to 3-8 hours of load, andso are significantly less expensive than PV batteries.

BOS

A PV system must include a support structure for the PV array, charge controller toregulate battery charging, and housing for the system battery. A diesel basedtelecom system usually includes housing for the telecom equipment, batteries, andbattery chargers and often for the diesel generators as well.

Line Extensions

Capital costs for line extensions are usually based on the distance that the load isaway from an established distribution transformer. Typical line extension costs for asmall system range from $1K to $5K per km, plus the actual connection charges ($1-2K). Stringent environmental requirements for remote sites can increase the cost ofextension by up to a factor of 10 times.




20-4

Recurring Costs

Recurring costs are those costs that are incurred every year of operation. For afossil-fuel generator, this includes fuel and engine maintenance costs, while for a PVsystem, it is typically limited to checking the batteries one to four times per year.Fuel costs are usually segregated from maintenance costs because of a difference

in inflation rates between the two categories. If the system is financed, the recurringcosts include loan payments.

Fuel

Annual fuel costs are dependent on the fuel consumption and the cost of fuel at thesite. Fuel consumption in a gas or diesel generator is a function of how highly theengine is loaded. For example, a diesel engine that is loaded at 25% of its ratedcapacity will still use 40% of its full load fuel consumption. Another way to expressthese numbers is as specific fuel consumption or kWh per liter of fuel. A fully loaded

diesel has a specific fuel consumption of approximately 3.1 kWh/L, meaning that itwill deliver 3.1 kWh for every liter of fuel consumed. At 25% load, specific fuelconsumption goes down to approximately 1.9 kWh per liter, while at 12% load theengine will only deliver approximately 1.2 kWh for every liter of fuel used. This is avery important factor in comparing renewable and nonrenewable technologies, sincemost prime diesel plants operate at an average load of between 12% and 25% ofrated capacity. Well-designed hybrid power systems, however, operate the engineat 75-95% of its rated load, making them much more fuel-efficient.

The cost of fuel at the site is the sum of the purchase price of fuel plus the deliverycosts. While delivery costs for some sites are negligible, the delivery costs at many

remote sites (mountaintop telecom installations, remote villages, and islandcommunities) can easily double or even triple the purchase price of fuel.

Maintenance

System maintenance is the other recurring (annual) cost. This category includesequipment maintenance, site maintenance, system supervision, etc.

For PV and wind systems, the generators typically require very little maintenance.However, if there are batteries in the systems, these will require inspection and

topping up approximately every 3-6 month.




20-5

Generators have much more complex maintenance requirements. A typicalexample is an air-cooled diesel generator used for prime power applications. Anoperating manual lists the following Three Levels of Diesel Maintenancerequirements:

Frequency Task To Perform Qualifications Cost Estimate250 hours Oil and filter change.

Also inspect air and fuel filters,fuel system, starter battery, andsystem electrical connections.

Can be

performed by atrained technicianat the site.

$25 to $100 plus

travel if required.

1500 hours Decarbonization (top endoverhaul). In addition to oilchange tasks, replace air andfuel filters. Remove thecylinder head and clean valves,valve seats, injector nozzles,etc. Replace all top end

gaskets.

Can only beperformed by atrained enginemechanic, sitework is possible.

$250-500 plustravel for a typicaltwo-cylinder air-cooled engine.

6000 hours Full overhaul. In addition totasks listed above, perform fullengine overall. Clean allcombustion and inlet/outletchambers. Replace all enginegaskets. Replace camshaft / crankshaft bearings, valves,valve springs, injectors, fuelpumps, pistons, piston rings,starter battery, start solenoid

and engine starter, fuelsolenoid, fuel pipes, and otherparts as necessary.

Requires acompetentmechanic.Typicallyperformed in arepair shop whichmeans the enginemust betransported to theshop, and a

replacementengine providedto the site.

$1,500-3,000,plustransportation andloaner ofreplacementengine.

For a prime power engine which is running continuously (8,760 hours per year), thesample maintenance schedule listed above would require 35 oil changes, 6decarbonizations, and 1.5 overhauls per year. Engines made for prime power oftenhave options that can extend the primary maintenance interval to 500, 1,000 or even1,500 hours, dramatically reducing the maintenance requirements. Note that manystandby engines have only a 100-hour oil change interval, which would necessitatenearly 90 oil changes per year – almost twice per week!

Loan Payments

If the system is financed through a bank or loan agency, the amortized loanpayments can be treated as a recurring cost rather than a capital cost.




20-7

Economic Factors

Life Cycle Term (“term”)

In performing a life cycle cost analysis, the primary economic factor is the "life cycle" of thesystem. The length of "term" of the analysis is chosen to be the service life of the longest-lived component. For example, a system which compared the costs of a diesel generatorand PV system over the course of a only a single year would ignore the fact that the PVsystem will continue to operate for many years without additional expenses.

In the case of PV system comparisons, the useful life of a PV module is on the order of 20-30 years, so this should be the "term" chosen for the analysis. Similarly, a comparisonbetween standard "prime" diesel generators and "cycle charged" systems would probably bebased on the life of the battery in the cycle charged system, or about 10 years.

Discount Rate (“DR”)

The next most important factor in NPV analysis is the "discount rate." This is the factor thatdescribes the changing value of money over time. It is basically equivalent to the amount ofmoney you could make with your capital if you chose to invest it in a bank or otherinvestment program rather than in a power system. The discount rate is applied to the priceof something in the future to “bring it back” to an equivalent value today.

Typical values for the discount rate at 7-15% when used in this type of analysis. In generalhigher discount rates make future expenses less important, and favor putting offexpenditures to the future. And in general, lower discount rates keep the impact of futureexpenses high, and favor capital intensive projects such as renewable energy systems.

Fuel & General Escalation (“FE”, “GE”)

Cost Escalation accounts for the fact that components and services traditionally get moreexpensive over time. For remote power systems, this applies to fuel, maintenance costsand replacement parts. Traditionally, fuel costs are escalated at a separate (and slightlyhigher) rate than other costs. This represents the volatility of the international energymarket. While it is very difficult to predict long term trends in fuel costs, it is important to

pick a value that has some real world representation for the specific situation beinganalyzed.

Typical annual escalation rates for remote power systems are 5-10% for fuel and 3-8% fornon-fuel expenses. Of course, local economic factors may cause large variations in thesenumbers.




20-8

Net Present Value

A mathematical method called "Net Present Value" (NPV) analysis takes intoaccount the fact that the value of money changes over time. This method seeksto calculate the costs in terms of "constant" currency. This is often expressed in"constant 1995 dollars" or some similar expression. The basis of NPV analysis is to

express a series of future costs in terms of present currency, taking into account thechanging value of money.

An item or event that has a cost in the future will probably cost more than the sameitem or event today due to general “price escalation”. However, money investedtoday to pay for that future cost would also increase in value, depending on how thatmoney today is invested. The net present value of an event or item occurring in thefuture is in a sense the equivalent amount of money you would need to invest todayto be equal to the cost of the event or item at that future date. The idea of NPV isbest expressed in an example:

Example: A full tank of fuel for a generator costs $100 today. It will need to be replacednext year. If the discount rate is 10%, then the amount of money neededtoday to pay for $100 of fuel next year will be given by

PV = 1 X $100 = $90.911+ 0.10

So $90.91 is the present value of the future $100 needed to pay for the fuel inone year. If the $90.91 was put aside this year and “invested” at the discountrate of 5%, then in one year we would have $100.

If there is price escalation happening as well, then the future cost of the tank

of fuel will be slightly higher than the price today. If the price escalation ofdiesel fuel is 5%, then the price in one year will be given by

FV = (1 + 0.05) X $100 = $105.00

So you would need to have $105.00 in one year to pay for the fuel.

The NPV of next year’s fuel will be the escalated price discounted back to thepresent. This involves combining the two formulas above.

NPV = (1 + 0.05) X $100 = $95.45(1+ 0.10)

So the net present value of the tank of fuel to be purchased next year is$95.45 in today’s money. Putting this much aside today will increase in valueat the discount rate of 10% and be enough to cover the escalated price of$105.00 that will be needed one year from now.




20-9

Exercise:




20-10

LCC Analysis Methodology

The basic life cycle cost analysis method is as follows:

1. Calculate the initial (capital) cost of the system;

2. Calculate the annual recurring fuel cost, then multiply by a factor which accountsfor the discount rate (DR), fuel inflation rate (FE) and life cycle term (term);

3. Calculate the annual recurring maintenance cost, then multiply by a factor whichaccounts for the discount rate, non-fuel general inflation rate (GE) and life cycleterm;

4. Determine the schedule for each non-recurring cost and multiply the cost by afactor to account for the discount rate and inflation rate in the year of occurrence;

5. Add these four costs together and divide by the total number of kWh produced todetermine the life cycle cost of energy.

Capital Costs

Capital costs are input into the formula with a factor of 1.0 since they are assumedto occur in "year 0," that is, before the start of system operation.




20-11

Recurring Fuel Costs

The formula for life cycle fuel costs is:

LCCFuelCost AnnFuelCost FE

DR FE

FE

DR

Term

= +

−

× −

++

* 1

1 1

1(Eqn 20.1)

whereAnnFuelCost is the annual fuel expenditure

FE represents Fuel Escalation

DR represents Discount Rate, and

Term is the life cycle term

(Note: If the discount rate is equal to the fuel escalation rate, the first term has a"division by zero" which will cause it to go to infinity. This can be remedied by usinga number with a very small difference, i.e. 0.06999 instead of 0.07 for a 7% discountrate. In any event, if the escalation factor is the same as the discount rate, then theFuel Factor (the term in the brackets) is simply equal to the life cycle term. So thetotal life cycle fuel cost would simply be the annual fuel expenditure times thenumber of years in the term. The same concern applies to the Maintenance Factordescribed in the next section)

Recurring MaintenanceThe formula for life cycle maintenance costs is:

LCCMaintCost AnnMaintCost GE

DR GE

GE

DR

Term

= +

− × − +

+

*

11

1

1(Eqn 20.2)

whereAnnFuelCost is the annual non-fuel expenditure

GE represents General Escalation


Term is the life cycle term




20-12

Non-Recurring Costs

The formula for nonrecurring costs is:

LCCReplCost ItemCost GE

DR

RY

= +

+

∑ * 1

1(Eqn 20.3)

whereItemCost is the nonrecurring expenditure in

present day costs

GE represents General Escalation


RY is the "Replacement Year"

Note that the costs for each non-recurring cost need to be calculated separately,then summed together to get the total nonrecurring costs.

Final Result - Life Cycle Energy Cost

The formula for the life cycle energy cost is:

LCC$ / kWh CapitalCosts LCCFuelCost LCCMaintCost LCCReplCost

Term kWh d

= + + +

* * / 365

(Eqn 20.4)

where

LCC$/kWh is the life cycle cost per kWh energy

LCCFuelCost is calculated above

LCCMainCost is calculated above

LCCReplCost is calculated above

Term is the life cycle cost term, and

kWh/d is the daily kWh output of the system




20-13

Exercise:




20-14

Sample AnalysesThis section presents analysis of four different power system options for a sampleapplication, a small village power supply. The costs associated with thesecalculations are estimated only -- they are not intended to be a formal analysis of theoptions presented. Rather, they are intended to show the method used and how a

comparison of various power solution options might be presented. A thoroughevaluation would require more detailed designs, actual system and operating costs,and a careful selection of economic factors.

The analyses will compare four typical energy solutions:

• Pure Standalone PV System• PV Diesel Hybrid System• Prime Diesel System• Utility Grid Extension




20-15

Sample Application Parameters

The sample application is a small rural village in Madras, India. The initial load isestimated to be 42 kWh AC per day, with a 15 minute peak load of approximately 8kW, with occasional surges up to 16 kW (recall the continuous and peak powergraph in the chapter on Load Estimation). The site is moderately remote (15 km

from the nearest power line), but accessible for the entire year by dirt road.

Site Factors Load Factors Economic Factors

Location Madras,India

Daily 42 kWh Life CycleTerm

20 years

Latitude 13 deg.N Annual 15,330kWh

DiscountRate

12%

Longitude 80 deg.E Total inLife Cycle

306 MWh FuelEscalation

10%

Distancefrom grid

15 km SurgeEstimate

20-40 kW GeneralEscalation

7%

Fuel Cost(delivered)

$0.40 /liter




20-16

PV System

A PV array to meet this load in this location would be approximately 16.5 kWp (4series X 55 parallel 75 Wp modules). A battery bank with 5 days of autonomy isrecommended with a total rated capacity of 196 kWh DC @ 10 hour rate (4080 Ah@ 10 hr @ 48 volts).

The major cost during the life of the system would be the replacement of the batterybank. It is estimated that the battery will need replacement after year 7 and againafter year 14. A small cost for battery maintenance trips during each year isincluded.

The resulting LCC analysis shows that the total life cycle (20 year) system cost,including initial capital and maintenance and battery replacement, comes to$168,486 NPV (discounted to constant dollars). The system produces 306,000 kWhin 20 years, for a LCC Cost/Energy ratio of $0.55/kWh.




20-17

Life Cycle Cost An alysis

System D escription Location Average Energy Requirement

Pure PV Madras, Ind ia 42 kW h/day

Initial Cap ital Costs

Item Q uantity Cost Extended Cost

Solar Array 220 $420.00 $92,400

Battery Bank ( kW h @ $/W h ) 196 $0.14 $27,440

System Contro ls 1 $1,000 $1,000

Inverter ( kW continuous @ $/watt) 10.0 $0.70 $7,000

Building 0 $0 $0

C ivil W orks (installation) 0 $0 $0

$0

$0

$0

$0

To ta l I nit ia l C a pi ta l $ 1 27 ,8 4 0

Recurring Costs

Fuel Annual H ours L iters/hour Cost/unit Annual Cost

-------------- -------------- -------------- -------------- --------------

Maintenance F re qu en cy /yea r C os t/E ve nt E xte nded Cos t

G enera l 1 .0 $350 $350

$0

$0

$0

$0Total Recurring Cost $350

Non-Recurring Items

Item Cost (today) Year NPV Factor NPV Cost

1 0.955 $0

2 0.913 $0

3 0.872 $0

4 0.833 $0

5 0.796 $0

6 0.760 $0

7 0.726 $0

Battery Repl + 10% labor $30,184 8 0.694 $20,946

9 0.663 $0

10 0.633 $0

11 0.605 $0

12 0.578 $0

13 0.552 $0

14 0.528 $0


16 0.482 $0

17 0.460 $0

18 0.440 $0

19 0.420 $0

20 0.401 $0

T ota l N o n- Re cu rr in g C o st $ 36 ,1 61

Economic Factors

Item Value

LCC Period (years) 20

Discount Rate 12% NPV Factor

Fuel Escalation (FE) 10% 16.64

General Escalation (FE) 7% 12.82

Cost Summary

Life Cycle Cost

Initial Capital $127,840 $127,840

All Non-Recurring Costs $36,161 Annual Cost NPV Factor $36,161

Annual Fuel -------------- 16.64 $0

Annual Maintenance $350 12.82 $4,485

Total LCC Cost (NPV) $168,486

Total Energy (kWh) 306,600

Cost/kWh (NPV) $0.55




20-18

The graph below illustrates that the array is sized large enough to operate all theloads even during the lowest insolation month (July), and that there is excess arrayenergy produced in every other month of the year. The array is deliberatelyoversized so that it can adequately operate the load in every month. The oversizingis evident by the fact that the array potential is larger than the load demandthroughout the year (except for July). The largest oversizing is evident in March,where the array could have produced almost 40% more energy than was needed by

the load.

No generator is included in the standalone system, so there are no generator startsin any year.

S ystem P e rforma nce

0

200

400

600

800

1000

1200

14001600

1800

2000

JAN FEB M AR APR M AY JU N JU L AU G SEP OCT N OV D EC

0

1

Ge n S ta rts M o nthly Lo a d Arra y (use d ) Arra y (p o te ntia l)




20-19

A cash flow analysis shows that there is a large initial outlay for all the equipment,and a battery replacement in year eleven. All other costs are small. The totalaccumulated cash flow (not discounted) comes to $278,333 over the life of thesystem.

Cash Flow Analysis - Pure PV System

$0

$20,000

$40,000

$60,000

$80,000

$100,000

$120,000

$140,000

1 2 3 4 5 6 7 8 9 10 1 1 12 1 3 14 15 16 17 1 8 1 9 20 21

Year

A n n u a l C o s t

$0

$50,000

$100,000

$150,000

$200,000

$250,000

$300,000

Accum ulated Costs $278,333




20-20

Prime Generator

A prime diesel generator power system would include two (redundant) generators,to extend life and insure reliable power availability. They will be sized at 20 kWeach, but could range from 20-40 kW, to handle the possible surge requirements ofa variable village load profile (recall the discussion of surge requirements for

complex large systems in the chapter on Loads). The lowest of the possible sizes ischosen here, to be most conservative in the cost comparison.

The generators would be alternated, so they would both age approximately thesame, each operating for 4380 hours per year. The manufacturer wouldrecommend replacement after 30,000 hours of this relatively continuous operation,which would occur after 7 years of system operation (30,000 hours / 4380 hourseach generator = 6.8 years), and then again after year 14 of the system life.

Fuel and maintenance costs would add up to a considerable expense during thesystem life. The fuel consumption rate is estimated to be 3.0 liters/hour as the

generator is expected to operate on the average at only about 25% load. Operatingfor 8760 hours/year at this rate consumes 26,400 liters/year at a cost of $10,775.Maintenance is expected for oil changes every 250 hours, decoking every 1,500hours and major overhaul every 6,000 hours (with full replacement needed after30,000 hours).

The resulting LCC analysis shows that the total life cycle (20 year) system cost,including initial capital and fuel, maintenance and generator replacement, comes to$354,816 (discounted to constant dollars). The system produces 175,200 kWh in 20years, for a LCC Cost/Energy ratio of $1.16/kWh.




20-21

Life Cycle Cost Analysis(Sugg ested ge nerato r size rang e and fuel rat es

System Description Location Average Energy Requirement to meet surge requirements

Prime Diesel Madras, India 42 kWh/day for variab le village loa d)

21 42 kW

Initial Capital Costs 3.12 5.96 liter/hour @25%

Item Quantity Cost Extended Cost

Diesel Gen: Power (kW) 20 2 $12,000 $24,000

Fuel System 1 $1,000 $1,000

Building 0 $10,000 $0

System Controls 1 $1,000 $1,000

Civil Works (installation) 0 $10,000 $0

$0

$0

$0

$0

$0

Total Init ial Capital $26,000

Recurring Costs Liters/hour

Fuel Annual Hours @25% Load Cost/unit Annual Cost

Diesel fuel (liters) 8760 3.00 $0.41 $10,775

3.0

Maintenance Period (hours) Frequency/year Cost/Event Extended Cost

Oil Change 250 35.0 $150 $5,256

Decoke 1500 5.8 $300 $1,752

Overhaul 6000 1.5 $1,500 $2,190

General 0.0 $300 $0

$0Total Recurr ing Cost $9,198

Non-Recurring Items


1 0.955 $0

2 0.913 $0

3 0.872 $0

4 0.833 $0

5 0.796 $0

6 0.760 $0

7 0.726 $0

DEG Replace +10% labor $26,400 8 0.694 $18,320

9 0.663 $0

10 0.633 $0

11 0.605 $0

12 0.578 $0

13 0.552 $0

14 0.528 $0

DEG Replace + 10%labor $26,400 15 0.504 $13,307

16 0.482 $0

17 0.460 $0

18 0.440 $0

19 0.420 $0

20 0.401 $0

Total Non-Recurring Cost $31,627

Economic Factors

Item Value





Cost Summary

Life Cycle Cost (NPV)Initial Capital $26,000 $26,000


Annual Fuel $10,775 16.64 $179,314

Annual Maintenance $9,198 12.82 $117,874







20-22

A cash flow analysis shows that there is an initial outlay for all the equipment, andthen generator replacement in year 8 and 15 of system life. The escalated cost ofgenerators after 7 and 14 years is shown. The fuel and maintenance costscontinually add up during system life, and also increase in cost as their valuesescalate over time. The total accumulated cash flow (not discounted) comes to$420,312 over the life of the system.

This analysis shows the sensitivity of long-term prime diesel system operation costto the discount rate value used. There are substantial costs occurring in far futureyears. The higher the discount rate, the more discounted the far future costs arereduced in the LCC type analysis (with all future costs being “brought back” to initialconstant dollars). So different discount rates can end up drastically affecting the lifecycle cost of such a system with much of its cost occurring in the far future.

Cash Flow Analysis - Prime Diesel

$0

$20,000

$40,000

$60,000

$80,000

$100,000

$120,000

$140,000

$160,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Year

A n n u a l C o s t

$0

$200,000

$400,000

$600,000

$800,000

$1,000,000

$1,200,000

$1,400,000

A c c u m u l a t e d C o s t

A cc um u la te d C os ts $1,226,510




20-23

PV-Diesel Hybrid

The hybrid system designed for this example included enough solar array to supply40% of the annual energy load, with the generator operating to provide theremaining energy throughout the year. The battery needed for such a system canbe quite small, as there is no need for many days of “autonomy” because the

generator is available “on demand”. A battery bank of 2 days of autonomy wasused, with the generator turning on and recharging the battery whenever it droppedto 75% depth of discharge (deep cycling batteries). This resulted in a total batterycapacity of only 101 kWh (C/10 rate), a reduction of 50% of the capacity needed forthe Pure PV system.

One generator is installed with a capacity of 13 kW, sized based on charging thebattery bank at its maximum allowable charge rate. The fuel consumption rate isestimated to be 4.8 liters/hour, as the generator is expected to operate at full loadwhen it is needed to recharge the battery. This is of course larger than the rate usedfor the prime diesel example, because the generator was only operating on the

average at about 1/4 of full load. The hours of operation at this full rate are nowestimated to be only about 1296 hours/year (instead of 8760 hours), only about 15%of the amount of hours for the prime diesel system example.

Operating for 1296 hours/year at this full power rate consumes 6220 liters/year at acost of $2,550. This is less than 25% of the cost for the prime diesel system.

Maintenance is estimated at the same periods as for the prime diesel system, withoil changes every 250 hours, decoking every 1,500 hours and major overhaul every6,000 hours. Maintenance costs are estimated to come to $1253/year, less than15% of the prime diesel costs.

At the end of the full system life term of 20 years, the diesel is expected to haveoperated on 25,920 hours total (20 years X 1296 hours/year), less than theestimated life of 30,000 hours. So the generator will not have to be replaced duringthe system life, compared to two replacements during the prime diesel system’s life.

The battery life is expected to be similar to the pure standalone system, due to themore intensive deep cycling of the batteries but less sulfation resulting from lesstime at partial state of charge. The generator is expected to start a total of about108 times per year, so the batteries will cycle down to 75% depth of discharge atleast that many times. This would mean about 756 such deep cycles in 7 years.Although some battery manufacturer literature states longer expected cycle life tothis depth of discharge, the accumulated loss of life over real time was estimated toresult in the need to replace the batteries after year 7 and again after year 14 of thesystem life.

The resulting LCC analysis shows that the total life cycle (20 year) system cost,including initial capital and fuel, maintenance and battery replacement, comes to$141,300 (discounted to constant dollars). The system produces 175,200 kWh in 20years, for a LCC Cost/Energy ratio of $0.46/kWh.




20-24

PV-Generator Hybrid System Sizing Report

Loc a tion Informa tionWhidb ee Island , WA.

Array Informa tion: Module Quantity 76

Mo dule Type SP75

Array Power 5.7 kW

Genera tor Info: Power: 13 kW

Fuel Consumption 4.80 liters/ hour

Output Dera ting Fac tor 100 %

Battery Informa tion: Days of Autonomy: 2 d a ys

Ba ttery Voltage Effic iency 88 %

Battery Charge Effic ienc y 90 %

Battery Energy Capac ity 132 kWh @ C/ 108

(ad just by 1.3 to C/ 10 ra te) 101 kWh @ C/ 10

Ba ttery Charge Cap ac ity 2745 Ah

Ba ttery Nomina l Voltage 48 Volts

Ba ttery Max. DOD 75 %

Inverter Informa tion: Effic ienc y: 85 %

Rec tifier Informa tion:Effic ienc y: 80 %

Output Ra ting : 275 amps

Da ily Monthly Array kWh Gen kWh # Gen Hours of Fuel

Loa d kWh Loa d kWh Sup ply Sup ply Sta rts Op era tion se d (liters)

JAN 42.0 1302.0 562.3 672.0 8 96.0 461

FEB 42.0 1176.0 553.5 672.0 8 96.0 461MAR 42.0 1302.0 627.2 672.0 8 96.0 461

APR 42.0 1260.0 579.1 672.0 8 96.0 461

MAY 42.0 1302.0 545.5 756.0 9 108.0 518

JUN 42.0 1260.0 453.5 756.0 9 108.0 518

JUL 42.0 1302.0 422.9 924.0 11 132.0 633

AUG 42.0 1302.0 461.4 840.0 10 120.0 576

SEP 42.0 1260.0 486.1 756.0 9 108.0 518

OCT 42.0 1302.0 509.5 840.0 10 120.0 576

NOV 42.0 1260.0 481.4 756.0 9 108.0 518

DEC 42.0 1302.0 509.5 756.0 9 108.0 518

ANNUAL Tota ls: 15330 6192 9072 108 1296 6220

PV/ Diesel Ra tio PV 40% enera tor 59%




20-25

Life Cycle Cost Analysis

System Description Location Average Energy Requirement

Hybrid System Madras, India 42 kWh/day

Initial Capital Costs


Diesel Gen Power (kW) 13 1 $6,000 $6,000

Fuel System 1 $1,000 $1,000

Building 0 $10,000 $0

Rectifier 1 $0 $0Solar Array 76 $420.00 $31,920

Battery Bank ( kWh @ $/Wh , 10 hr rate ) 101 $0.14 $14,190

System Controls 1 $1,000 $1,000

Inverter ( kW continuous @ $/W ) 10.0 $1.00 $10,000

Civil Works (installation) 0 $10,000 $0

$0

Total Initial Capital $64,110

Recurring Costs

Fuel Annual Hours Liters/hr Cost/liter Annual Cost

Diesel fuel 1296 4.80 $0.41 $2,550

Maintenance Period (hours) Frequency / year Cost/Event Extended Cost

Oil Change 250 5.2 $150 $777

Decoke 1500 0.9 $300 $259

Overhaul 6000 0.2 $1,000 $216

General 0.0 $300 $0$0

Total Recurring Cost $1,253

Non-Recurring Items


1 0.955 $0

2 0.913 $0

3 0.872 $0

4 0.833 $0

5 0.796 $0

6 0.760 $0

7 0.726 $0


9 0.663 $0

10 0.633 $0

11 0.605 $0

12 0.578 $013 0.552 $0

14 0.528 $0


16 0.482 $0

17 0.460 $0

18 0.440 $0

19 0.420 $0

20 0.401 $0

Total Non-Recurring Cost $18,700

Economic Factors

Item Value





Cost Summary

Life Cycle Cost



Annual Fuel $2,550 16.64 $42,438

Annual Maintenance $1,253 12.82 $16,052


Total Energy (kwh) 306,600

Cost / kWh (NPV) $0.46




20-26

The graph below shows the array and generator performance over a typical year.The array is sized to meet only 40% of the load, so the generator is operated everymonth, varying from 8 starts in January-April (about once every 3-4 days) to 11starts in July (about once every 2-3 days). All of the array potential output is utilized,in contrast to the pure standalone system where the array met the load in July andwas oversized all the rest of the year, peaking at nearly 40% oversize in March.

System Performance

0

200

400

600

800

1000

1200

1400

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

0

12

3

4

5

6

7

8

9

10

11

12

Gen Starts Monthly Load Array (used) Array (potential)




20-27

A cash flow analysis shows that there is an initial outlay for all the equipment, andthen battery replacement in year 8 and 15 of system life. The escalated cost of thebattery bank in those future years is shown. The fuel and maintenance costs arequite small compared to the prime diesel system. The total accumulated cash flow(not discounted) comes to $349,600 over the life of the system.

Cash F low Ana lys is - H ybr id Sys tem

$0

$10,000

$20,000

$30,000

$40,000

$50,000

$60,000

$70,000

1 2 3 4 5 6 7 8 9 10 1 1 1 2 1 3 1 4 15 1 6 1 7 1 8 1 9 2 0 2 1

Y e a r

A n n u a l C o s

$0

$50,000

$100,000

$150,000

$200,000

$250,000

$300,000

$350,000

$400,000

A

c c u m u l a t e d C o s

A c c u m u la t e d C o s t s $349,600




20-28

Grid Connect

A grid connection would entail the costs of extending the line, estimated for thisexample at $6,000 / km, and the cost of a step-down transformer (estimated atabout $7,000). There is no diesel fuel, but there is an electricity charge estimated tobe $0.10/kWh, and an annual maintenance cost for the line of about $30/km.

The resulting LCC analysis shows that the total life cycle (20 year) system cost,including initial capital and electricity and maintenance, comes to $128,279(discounted to constant dollars). The system produces 175,200 kWh in 20 years, fora LCC Cost/Energy ratio of $0.42/kWh.

Of course, this analysis does not account for any of the issues associated with gridpower in remote areas, such as power quality or outages. There can be realeconomic costs attached to low voltages and poor power factor, in that customersmight have to pay for power conditioning equipment at their end of the wire, or mayhave unacceptable operation from motors or electronic equipment. And power

brown outs or outages will have very measurable impact on productivity. Thisaspect of the grid extension alternative do not show up from this simple type ofanalysis, and should also be considered when comparing true costs of variouspower solutions.

The cash flow analysis shows the large initial cost of extending the grid and then thesmall-accumulated cost of the electricity and line maintenance over the life of thesystem.

Cash Flow Analysis - Utility Extension

$0

$20,000

$40,000

$60,000

$80,000

$100,000

$120,000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Year

A n n u a l C o s t

$0

$50,000

$100,000

$150,000

$200,000

$250,000

A c c u m u l a t e d C o s t

Accum ulated Costs $213,322




20-29

Life Cycle Cost Analysis

System Description Location Average Energy Requirement

Utility Extension Madras, India 42 kWh/d

Initial Capital Costs


Utility Line Extension (km @ $/km) 15 $6,000 $90,000

Transformer 1 $7,000 $7,000

$0

$0$0

$0

$0

$0

$0

$0

Total Initial Capital $97,000

Recurring Costs

Fuel Annual kWh Cost / kWh Annual Cost

Electricity 15,330 $0.10 $1,533

Maintenance Extension Length Frequency/year Cost / km Extended Cost

Line Maintenance 15 1 $30 $450

$0

$0

$0$0

Total Recurring Cost $450

Non-Recurring Items


1 0.955 $0

2 0.913 $0

3 0.872 $0

4 0.833 $0

5 0.796 $0

6 0.760 $0

7 0.726 $0

8 0.694 $0

9 0.663 $0

10 0.633 $0

11 0.605 $0

12 0.578 $013 0.552 $0

14 0.528 $0

15 0.504 $0

16 0.482 $0

17 0.460 $0

18 0.440 $0

19 0.420 $0

20 0.401 $0

Total Non-Recurring Cost $0

Economic Factors

Item Value





Cost Summary

Life Cycle Cost


All Non-Recurring Costs $0 Annual Cost NPV Factor $0

Annual Fuel $1,533 16.64 $25,512

Annual Maintenance $450 12.82 $5,767







20-30

Comparison of Results

The real value of a LCC analysis is not in the absolute number but in a comparisonof values from various alternative choices. Now we can compare the four energysolutions side-by-side, to see how their values and components relate.

Economic Factors HybridSystem Prime Diesel Pure PV UtilityExtensionInitial Capital $64,110 $26,000 $127,840 $97,000Annual Fuel (NPV) $42,438 $179,314 $0 $25,512Annual Maintenance (NPV) $16,052 $117,874 $4,485 $5,767All Non-Recurring Costs(NPV)

$18,700 $31,627 $36,161 $0

Total LCC Cost $141,300 $354,816 $168,486 $128,279Cost/kWh (NPV) $0.46 $1.16 $0.55 $0.42Accumulated Cash Flow $349,600 $1,226,510 $278,333 $213,322

One useful method of comparison is to graph the component costs. Each solution isgraphed below, with the initial capital, annual fuel and maintenance costs, and anymajor non-recurring expenses shown. This makes it easier to see which alternativehas the largest initial capital costs, and which has greater costs over time. Forexample, the diesel has the lowest initial costs of all, but accumulates the greatestoperating costs of all.

Life Cycle Cost Comparison (NPV)

$0

$50,000

$100,000

$150,000

$200,000

$250,000

$300,000

$350,000

$400,000

Hybrid

System

Prime Diesel Pure PV Utility

Extension

All Non-RecurringCosts (NPV)

AnnualMaintenance(NPV)

Annual Fuel(NPV)

Initial Capital




20-31

Notice that the benefits of the hybrid solution over the pure PV solution are not justin a lower LCC, but also in a dramatically lower initial cost.

Another interesting method of comparison is to look at the cash flows over time.Each alternative is graphed below against the same scale, to more easily see whentheir accumulated (non-discounted) expenditures cross one another. Each systemis assumed to be producing the same amount of total energy annually, so this

comparison shows when each system becomes “cost effective” compared to theothers.

For example, in the first years of the systems, the prime diesel system is producingenergy for the lowest accumulated cost. At about 5 years, the accumulated costs offuel and maintenance push the prime diesel system above the total accumulatedcost of the hybrid system. Therefore after about 5 years, the hybrid system is themost cost effective (total energy / total expenditure). And after about 7 years, eventhe pure standalone PV system, with its very high initial costs, is more cost effectivethan the prime diesel system.

Cash Flow Analysis

$0

$200,000

$400,000

$600,000

$800,000

$1,000,000

$1,200,000

$1,400,000

0 2 4 6 8 10 12 14 16 18 20

Year

A c c u m u l a t e d C a s h F l o w

Hybrid System Prime Diesel Pure PV Utility Extension




20-32

Parametric Analysis

In this analysis, various factors, or parameters, are varied while holding all otherfactors constant to determine the sensitivity of the final result to the initialassumptions. Typical factors to compare through parametric analysis include:

• Fuel Cost• Discount Rate• Inflation Rate• Maintenance Rate• Grid Extension Distance• Grid Electricity Cost• Battery Cost• Diesel Cost• PV / Diesel Ratio

Sensitivity to Fuel Cost

As an example, the initial analysis presented above used a fuel cost of $0.40 perliter. By varying this value only, the following parametric analysis table can bedeveloped:

Alternative Fuel Cost ($/liter) Sensitivity

0.20 0.40 1.00

StandalonePV

0.55 0.55 0.55 none

Prime Diesel 0.86 1.14 2.00 high

Hybrid 0.39 0.46 0.66 moderate

GridExtension

0.42 0.42 0.42 none

This analysis shows that PV and grid options are not at all sensitive to fuel cost(since they use no fuel directly), while hybrids are moderately sensitive and dieselsystems are highly sensitive. Therefore any assumptions about diesel fuel costhave a tremendous effect on the life cycle costing of the Prime Diesel alternative,and should be validated as much as possible for this type of analysis to have realmeaning.




20-33

Sensitivity to Discount Rate

Varying the discount rate shows another example of parameter sensitivity. Thisfactor should be of major concern when comparing various projects that havesignificant costs in the future.

Alternative Discount Rate Sensitivity

6% 12% 18%

StandalonePV

0.66 0.55 0.49 low

Prime Diesel 2.00 1.16 0.76 high

Hybrid 0.66 0.46 0.36 moderate

GridExtension

0.50 0.42 0.38 low

This analysis shows that the capital intensive projects (pure standalone PV and gridextension), where most of the expenses occur at the beginning of the project, arenot very sensitive to discount rate, while the project with the highest proportion ofrecurring expenses in the future (prime diesel) is highly sensitive. Once again, thispoints out that assumptions about discount rate must be carefully exposed and

validated for the conclusions of this type of analysis to have meaning. Theassumptions of risk and real financial conditions must be carefully considered,especially since we are considering alternatives that have very different future costs.A high discount rate for example will make a Prime Diesel alternative lookinexpensive compared to a PV solution, whereas a low discount rate will show aPrime Diesel system to be very expensive over time.



LCC Sensitivity

Another interesting analysis is to evaluate the total Life Cycle Cost of the hybridsystem for different PV/Diesel ratios. It is interesting to find the ratio of PV energy togenerator energy that gives the lowest life cycle cost. Other factors such as initial

cost may take precedence, but this type of sensitivity analysis helps to show “howmuch” of a hybrid the system should be to have the lowest cost over time