Optimal Planning of Energy Resources in a small energy system under the Smart Home Concept

Abstract: This paper investigates the optimal management of a renewable based energy supply design for a residential building equipped with smart communication and metering infrastructures, approaching the smart home (SH) concept. The best economically optimal model is selected in each case based on the net present cost (NPC) and the cost of energy (COE) criteria. Mutual effects of the inflation rate uncertainty, project life-time span and a comparative model to study the use of combined heat and power (CHP) technology on the objective function have been modelled. An economic index related to environmental issues and one related to the thermal load supply have been defined and analysed. Finally a conclusion has been done according to results achieved throughout the study. Keywords: Smart home (SH), Hybrid energy system (HES), Economic feasibility, Net present cost (NPC), Sensitivity analysis, Environmental issues

1. Introduction One of the key features of the smart grid is the

integration of renewable and storage energy resources at the consumption side [1-5]. From the smart grid’s point of view, buildings are small flexible consumers which can embed distributed energy resources (DERs) and exchange power when needed. Therefore, they are called prosumers in some references [6]. From the consumers’ point of view, using DERs help them have a more reliable and sometimes economical energy supply system.

Environmental issues accompanied by the economic aspects are the most effective factors for the electrification investors to choose the best option among the various types of power resources. In terms of environment, side costs due to the emission of pollutants from fossil fuel plants which vary from high medical costs incurred on the society to that of related to emission penalties, make the electricity energy more expensive than what it may seem at the first sight. In addition to all above, the increasing fuel price, the standby reserve cost and the start-up/shut-down cost of bulky generators deteriorate the condition and make it more expensive to rely only on conventional fossil fuel plants.

Smart home (SH) is a new born concept emerging as a result of a good combination of smart grid and DERs’ usage concepts. SH supports the idea of integrating DERs

in distribution networks at the nearest point to consumers retaining lots of benefits for residents [1].

Among essential infrastructures such as energy supply, communication, metering, etc. one of the most important elements is the hybrid energy system (HES) design. It is the main factor of determining the total cost of the smart home implementation. We will focus on HES in this paper and try to find the best possible configuration.

A number of optimization techniques for hybrid system design have been presented. Among them genetic algorithm (GA), particle swarm optimization (PSO) and simulated annealing (SA) have attracted much attention and been used vastly by researchers in the literature [7].

Various methods in addition to above given approaches such as linear programming, evolutionary algorithms, neural networks, simplex algorithm, dynamic programming, stochastic approach, iterative and probabilistic approaches, design space based approach ,etc. have been used by researchers to design hybrid systems optimally in terms of economy and environmental issues [7].

In this paper, we consider a residential building and assume that it is equipped with smart communication and metering infrastructures. Then we will define a HES design problem and analyse it using different case studies which include a base case study, a case study to analyse impacts of applying CHP technology, one for modelling the inflation rate uncertainty and finally one for studying project life-time impacts. Economic feasibility of such a design will be at the center of attention throughout the article via judgments based on NPC and COE indexes. However, a brief discussion is devoted to environmental issues that may be considered as an important criterion to design HES projects as more attentions are being paid to social health.

2. Problem Definition

2.1 Concepts Smart home supports the idea of integrating DERs in

distribution networks at the target consumption point

Mohammad Hassan Amirioun1*, and Ahad Kazemi2* *Center of Excellence for Power System Automation and Operation, Department of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran,1mhamirioun@elec.iust.ac.ir,2kazemi@iust.ac.ir

978-1-4673-5634-3/13/$31.00 ©2013 IEEE

while enabling both the distribution system operator (DSO) and the smart home central controller (SHCC) to send and receive information via communication infrastructures. All the appliances are equipped with a local controller (LC) that is in coordination with SHCC i.e. sending data and receiving commands. Power exchange between the building and the grid is controlled by the distribution system aggregator (DSA) which collects small power bids from different DERs’ holders and participates in the retail market with a large bid (Fig. 1) [8]. We assume that the needed communication and metering infrastructure has been provided and will focus on the energy supply system.

2.2 Purposes and Criteria

As declared before, the main purpose of the presented work is the selection of the most optimal configuration of HES for the considered building. Different cost types used in this work are listed as below: 1) Capital Cost (Ccap, $): The cost incurred by the launch of the project is called the capital cost. It mainly includes the fund needed to buy and install the components and provide essential infrastructures. Here, it is mainly related to the cost of resources’ provision and their accompanied equipments and also the preparation and installation cost. 2) Operation and maintenance cost (CO&M ,$/year ): It includes the cost of running and keeping the system ready to use at each hour. The overall system O&M cost is the sum of all the components O&M cost. 3) Replacement Cost (Crep , $/year): The system replacement cost is the overall cost of replacing a system component at the end of its lifetime. 4) Salvage cost (Csal , $/year): The value remaining in a component of the power system at the end of the project lifetime is called the salvage cost. In other words, it must take a negative value to show its concept. 5) Grid Sale cost (Csale ,$/year): The revenue earned by selling energy to the grid which clearly takes negative values to signify its concept. 6) Fuel cost (Cf): It is clearly the overall cost of fuel for the components needing fuel to work. Let TANN

nC be the annual total cost related to the nth

year, then to calculate the cost of the project in the present value, Equation (1) is used:

1 (1 )

TANNNn

TNPC nn

CCi

(1)

The uniformed annualized (UANN) cost of the system is calculated using Equation (2). In fact, the present value of the overall cost of the system is distributed evenly in the entire years of the project lifetime. The relationship between CNPC and CUANN is shown in Equation (2):

( , )UANN NPC i NC C CRF

Where N is the project lifetime and i is the annual real interest rate and is given in Equation (3):

100* / 1i r f f (3)

where r is the annual average interest rate and f is the annual average inflation rate. By defining the interest rate in this way, inflation is factored out of the economic analysis. CRF is the capital recovery factor and is calculated by Equation (4):

( , )

(1 )

(1 ) 1

N

i N Ni i

CRFi

Equation (5) depicts the relation between COE and the annualized cost. In fact, ignoring the loss power, COE is the pure electrical cost (difference between the annualized cost of the system and that of the boiler) per kWh of the useful electrical energy produced by HES, (sum of the annual load and the grid’ sale, here in $/kWh).

UANNANNboiler

ANN ANNload sale

COEC CE E

As a conclusion it is to be noted that NPC and COE will be used as the main indexes to evaluate the case studies in this paper.

2.3 Problem Formulation Five cost types explained above can be formulated

through Equations (6) to (10) and finally the problem objective function which is the total NPC cost of the project is written as Equation (11) and the constraints include Equations (12) and (13).

1

compN

capcomp

capcompC C

(6)

&&

1( )O M

comp

NcompnO M

compC tC

, year (7)

Where comp and Ncomp denote for a system’s componenet and the number of that componenet, respectively.

8760

1( ) ( )n sale

sale salet

C E t t

, year (8)

DSO DSA

SHCC

LC1 LCn ..............

Command Data Negotiation Fig. 1: Conceptual relation among different agents of a SH

Where Esale(t) is referred to the electricity energy sold to external grid during hour t of the year n and ( )sale t is the energy price at that hour.

n jrep rep

j

C C , year (9)

Where j denotes for any component replacement case during the nth

year.

n j remsal rep

j comp

RC CR

, year (10)

Where Rrem is the remaining age of the component when it fails to continue working and Rcomp is the nominal component age. So the objective function can be written as in Equation (11):

&1

({{ Nn

TNPC O Mn

ncap repMinimum C C C C

nsalC + ))(1 }n n

fnsale CC i

(11)

Subject to: min max( )j j jp p t p , t (12)

that depicts the power limitation of each element. Next equation applies the power balance constraint at each hour:

( ) ( ) ( ) ( )gen load netgrid lossP t P t P t P t , t (13)

This indicates the total power generated is the sum of load, net power exchanged with the grid and the loss power at each time. Loss is mainly due to converters, batteries and transmission cables which all have been ignored in this study. It should be noted that constraints are automatically applied by hybrid optimization model for electric renewable software (HOMER) and there is no need to insert them separately. It is a linear minimization problem and HOMER solves the problem using a black box code.

3. Simulations

3.1 Considered System 3.1.1 Load

The average electric and thermal load profile are illustrated in Figs. 2 and 3. It should be noted that the load consumption for all seasons and months has been considered the same and a near average has been applied. 3.1.2 Weather condition

The study area is located between 31°42ʹN and 51°42ʹE, urban part of Shahreza. The solar radiation profile of considered region is assumed to be like Fig. 4 for a one-year period according to NASA Surface Meteo-

rology and Solar Energy [9]. A de-rating factor of 90% reduces the PV production by 10% to take into account the varying effects of temperature and dust on the panels.

Fig. 5 shows the profile of wind speed of the region over a one-year period [9]. The annual average of wind speed is 6.7 m/s, providing a rather good potential for wind energy. 3.1.3 Resources

TABLE I shows detailed data on resources useable in our HES design. Input data on option costs, sizing and other parameters are presented in this table. 3.1.4 Energy Prices A time of use (TOU) Electricity purchase and selling back price has been applied and shown in Fig. 6 [10]. 3.1.5 Economic Parameters

Value of variable parameters will be declared in case studies later. In this section we assign fix value parameters.

Fig. 2: Electric load profile of the desired building

Fig. 3: Thermal load profile of the desired building

Fig. 4: Average solar radiation of the considered region

Fig. 5: Average wind speed profile of the considered region

TABLE I: Detailed Data on Resources Useable in HES Design

It should be noted that the reference value for the price

of 1$ of the U.S is assumed to be 12260 Rials [11]. The annual average interest rate; r is set to be 0.14 (14%) [10]. The diesel price is assumed to be 0.3$/Lit, considering the point that diesel is not discounted for private electricity generation [10]. 3.2 Case Studies

In this section we define and simulate each case study and then discuss the results. 3.2.1 Case 1

This case is set as a base case to check the variations due to changes in next cases. TABLE II shows the values used in this case. A heat recovery coefficient (HRC) has been defined and set value in this table to show whether a CHP technology is applied or not.

TABLE II: Applied Values for Parameters in Case 1 Parameter r( %) f (%) Life-time (yr) CHP HRC (%) Value 14 10 15 0

TABLE III summarizes the results of simulations for

this case. A seen, solar PV arrays do not qualify for the application due to their high capital cost in comparison with other resources. Only 30% of power production is supplied by renewable resources. It is expected that with higher inflation rates, the use of renewable resources especially solar PV will become more economical, because their operation cost is very small in comparison with their capital cost.

TABLE III: Summary of Results for Case 1 Component Value

NPC ($) 37795 COE ($/kWh) 0.082

Boiler Total Cost ($) 1853 CO2 Emissions (ton/yr) 20.22

Energy Sold to Grid (kWh/yr) 486 Grid Contribution (%) 70 Wind Contribution (%) 30 Solar Contribution (%) 0 Diesel Contribution (%) 0 Renewable Fraction (%) 30

3.2.2 Case 2

In this case we consider a microturbine equipped with CHP technology for diesel generator with a HRC equal to 60%. Results show that neither the first nor the second economical configuration has a CHP microturbine. This is mainly due to its high capital cost. However, the third economical option consists of a CHP microturbine accompanied by two wind turbines and the external grid. TABLE IV summarizes results obtained in this case. As can be seen there is still no place for solar PV in this case. Approximately half of the thermal load is supplied by CHP microturbine and the boiler cost has decreased. However, the overall cost of system has increased. This case is environmentally friendlier than case 1, because the TABLE IV: Summary of Results for Case 2

Component Value NPC ($) 39769

COE ($/kWh) 0.087 Boiler Total Cost ($) 1487

CO2 Emissions (ton/yr) 11.25 Energy Sold to Grid (kWh/yr) 3637

Grid Contribution (%) 37 Wind Contribution (%) 58 Solar Contribution (%) 0

CHP Electricity Contribution(%) 5 CHP Thermal Contribution (%) 46

Renewable Fraction (%) 58

Options Options on size and unit numbers Life Other information Capital

cost Replacement

cost O&M cost

Wind 3 kW, DC 0, 1, 2, 3 (turbines) 20yrs Hub height : 25m 7600

$/turbine 5000

$/turbine 100$/yr.

Solar 0, 1, 2, 4, 6 ,8 (kW),DC 20yrs De-rating factor: 90% 6000$/kW 4200$/kW 40$/yr. Diesel

CHP

0, 1, 2, 4, 6 ,8,10 (kW) 15000hrs Minimum load ratio: 30% 400$/kW 300$/kW 0.01$/kW/hr.

0, 1, 2, 4, 6 ,8,10 (kW) 45000hrs Minimum load ratio: 60% 750$/kW 600$/kW 0.5$/hr.

Battery 0,1, 2, 4, 6, 8, 10, 12,15 batteries, DC

10623 kWh Nominal capacity :1900Ah 100$/bat 70$/bat 5$/yr.

Converter 0,1,2,4,6,8 kW 20yrs Converter Efficiency: 90% Rectifier Efficiency: 85% 700$/kW 500$/kW 5$/yr.

Fig. 6: Grid purchase and selling back rates

Purchase rate

Selling back rate

grid contribution has dramatically decreased. A thermal-economic index is defined and analysed here to evaluate the extra cost of thermal supply by CHP microturbine. It can be formulated as Equation (14):

For the values achieved in case 1 and case 2, this factor is equal to 0.65$/kWh/yr; that is to supply part of the thermal load an extra cost equal to 0.65$ for a kWh of the thermal load should be paid annually. 3.2.3 Case 3

This case models the inflation uncertainty and investigates impacts of the inflation rate variation on NPC and the optimal combination of the system. We consider all the assumptions in case 1, except that a range of inflation from 0 to 25% has been applied instead of setting a fixed value for that.

As a rule of thumb when the inflation increases according to Equations (1) to (4) NPC will increase (Fig. 7). Actually analysing NPC is not reliable enough when comparing results of inflation variations. According to Fig. 8 COE decreases with the inflation increase. This conclusion is justifiable based on Equations (1) to (5). It means that the cost of energy supply for the desired building reduces when the inflation rate increases. In other words, it will become more economical to operate the system with higher values of the inflation rate. Fig. 9 illustrates the PV production variation versus that of the inflation rate. As shown, for inflation rates less than 20%, solar PV does not qualify for power production applications. However, as it goes over this value solar PV will become economical. This refers to the fact that the large part of solar cost includes the capital cost, therefore it will be preferred to those having large operation cost in high inflation rates. The same reasoning is dominant for the renewable fraction as shown in Fig. 10.

3.2.4 Case 4

This case models the project life-time uncertainty and studies impacts of its variation on NPC and the optimal combination of the system. All the assumptions in case 1 are applied here, except that a range of life-time from 10 to 25 years has been used instead of setting a fixed value for that.

Surely NPC will increase with the increase of the life-time span. However, based on the same reasoning presented for case 2 previously, we refer to COE variations. As Fig. 11 shows, a smooth drop occurs when the life-time increases. Comparing with Fig. 10, it is clearly perceived that COE drop is much smoother than that of case 2. Fig. 11 shows the renewable fraction variations. As shown, the maximum renewable fraction contribution is about 0.27 in the considered life-time span. Comparing with Fig. 9, it can be understood that the life-time span is less effective than the inflation rate to diminish COE. It is worth to mention that no PV production is preferred in the considered life-time span. It can be concluded that output variables are less sensitive to life-time span than that of the inflation rate. In other words, the inflation augment makes the hybrid energy system design more economical than that of the life-time span.

Fig. 7: NPC variation versus inflation increase

Fig. 8: PV production variation versus inflation increase

Fig. 9: Renewable fraction variation versus inflation increase

Fig. 10: COE variation versus inflation increase

Fig. 11: COE variation versus life-time span increase

In order to obtain a better sight of the topic and find out another contradistinction between the impacts of the inflation rate and the project life-time on the system efficiency in terms of energy supply, the net grid purchase energy for both cases 3 and 4 are shown in Fig. 13. The net grid purchased energy at an operation point is a good sign to measure the effectiveness of the design. The more the design is useful and effective, the less the net purchased energy is. As it shows, in both the cases the net purchased energy is decreased with the increase of the inflation rate or the project life-time. However, in case 3 it drops significantly with the inflation rate rise while in case 4 this drop is regular.

4. Conclusion An economic feasibility study was done for a hybrid

energy system designed for a residential building to investigate impacts of variable parameters on decision making factors especially economic standards. The entire study was conducted under the assumption that the primitive concepts of smart grid have been achieved and implemented in considered electric grid and the building under the study is equipped with smart communication and metering infrastructures. A base case was defined and then simulated using constant value assigned parameters and technical, economic and environmental aspects were analysed. As the second case, the effect of applying a CHP microturbine was investigated. A new index was defined and calculated in this case to obtain the extra cost per kWh of the thermal load supplied by the microturbine. Base on our results this index was equal to 0.65 $/kWh/yr. In case 3 the inflation rate uncertainty was studied based on its

fluctuations over the past years and the information obtained from credible economic reports and valid organizations. Results showed that the energy supply cost will become cheaper and the renewable contribution increases as the inflation rises. However, PV production remains zero till the inflation rate reaches 20% and increases with a positive slope. As the last case to study the expected life-time of the project, we applied a span of years instead of using a constant value for that. Simulations indicated a drop in energy supply cost, although it was much smoother than that caused by the inflation variation. The same behaviour was seen for the renewable fraction; i.e. the maximum contribution of renewable resources reaches 0.27 while it reached about 0.70 when the inflation rises. It could be concluded that among the effective variable parameters under the study in this paper, the inflation rate variation has the most impact on different system design outputs. Finally HOMER was found an excellent tool to study hybrid energy systems.

Refrences [1] V. Hamidi, K. S. Smith, and R.C. Wilson, “Smart grid technology review within the transmission and dtistribution sector ,” in Proc. 2010 Europe IEEE PES Innovative Smart Grid Technologies Conf., pp.1–8 . [2] N. Radhika, K. Sivalingam, and V. Anand, “Network architecture for smart grids “, in 2011 International Conf. on Computer, Communication and Electrical Technology (ICCCET), pp. 24-26 [3] Jinsoo Han, Haeryong Lee, and Kwang-Roh Park, “Remote-controllable and energy saving room architecture based on ZigBee communication,” IEEE Transactions on Consumer Electronics, Vol. 55, I. 1, pp. 264-268, 2009 [4] W.-H.E Liu, K. Liu, D. Pearson, “Consumer-centric smart grid,” in Proc. 2011 IEEE PES Innovative Smart Grid Technologies Conf. (ISGT), pp.1–6. [5] Smart Grid Conceptual Framework, IEEE Smart Grid Model, June 2011, [Online]. Available: http://smartgrid.ieee.org/smart-grid-framework-diagram. [6] S. Mariethoz, and M. Morari, “Modelling and hierarchical hybrid optimal control of prosumers for improved integration of renewable energy sources into the grid,” in Proc. 2012 American Control Conf. (ACC), pp.3114-3119. [7] O. Erdinc, and M. Uzunoglu, “Optimum design of hybrid renewable approaches,” Renewable and Sustainable Energy Reviews, Vol. 16, pp. 1412-1425, 2012 [8] A. R. Al-Ali, Ayman El-Hag, Mujib Bahadiri, Mustafa Harbaji, and Yousef Ali El Haj,“ Smart Home Renewable Energy Management System ,” Energy Procedia, Vol.12, pp.120-126, 2012 [9] http://eosweb.larc.nasa.gov/sse/ [10] www.tavanir.org.ir [11] http://www.cbi.ir/

Fig. 12: Renewable fraction variation versus life-time span

Fig. 13: Net grid purchase variation, a: versus inflation rate (%), b: versus project life-time (yr)