flow ehandbook

8/22/2019 Flow eHandbook

1/12

Fl Habk

Mii FlStart u


2/123

tabl f CsKp mi pblms fm bbbli p 4Entrained gas can aect Coriolis meters but you can take steps to optimize perormance.

usa h Aaci f Mams 9Magnetic owmeters provide accuracy and can be used in a variety o applications and environments

tak a diff Lk a Cifal Pmps 13An unconventional assessment can provide insights or eective control.

tPrevious Page Next Pageu
http://www.usa.siemens.com/


3/12 4

Kp mi pblms fm bbbli pEntrained gas can aect Coriolis meters but you can take steps to optimize perormance.

and environments

By Tim Patten, Micro Motion, Inc.

CorIoLIS MeterS have long been used verysuccessully on single-phase luids. However,

liquids that contain bubbles (air or gas) cause

dynamic changes to a Coriolis meter that are not

present in a single-phase luid and that lead to

measurement errors.

A Coriolis meter operates by driv ing one or

two tubes at a resonant, or natural, requency. In

the meter, the electronics (or transmitter) send a

drive signal to the sensor that tracks the requen-

cy o the tube and maintains the proper vibration

amplitude. Driving on the resonant requency is

important because it enables luid density mea-

surement and minimizes power requirements.All modern Coriol is meters are intrinsical ly

sae (IS), which limits the amount o power that

is allowed to drive the sensor. Bubbles moving

around in the l iquid tremendously increase uid

damping, which results in power requirements

that ar exceed what IS restrict ions permit. So, the

tube amplitude signicantly decreases. Tis condi-

tion is sometimes called stall, although the tubes

usually do continue to vibrate to some extent.

When the tube amplitude decreases, the

signal-to-noise ratio also alls, making it a chal-

lenge to extract the mass ow signal rom the

Figure 1. When uid velocity cant overcome buoyancy, bubbles get caught in

inlet leg.



4/12 5

relatively high level o noise. Older analog signal-

processing techniques are highly sensitive to

entrained air because signa l amplitude is low and

noise is high; no algorithm is available to enhance

the measurement signal, thereby improving the

signal-to-noise ratio. In contrast, digital signal

processing (DSP) can eectively lter the noise

to yield a good stable measurement so long as the

tube is vibrating, even at reduced amplitudes.Note: Even with DSP, when gas is present in a liq-

uid stream the meter can only provide total-product

density (including the gas), not liquid-only density.

tHe IMPACt oF FLuId dynAMICS

Coriolis meters are not sensitive to ow prole

and other disturbances that aect other metering

technologies. For instance, since the undamental

measurement o delta comes rom the relative

values o each o two tubes in bent-tube designs,

swirl upstream o the meter doesnt impact the mea-

surement because it doesnt matter how much ow

goes through one tube or the other. Accuracy is not

degraded even when one tube is completely plugged.

However, when gas is present in a liquid, the

ow prole can become a concern. Although the

undamental measurement is unaected (that is,

the relative delta ), the tubes can become imbal-

anced due to the large density dierence between

them (air in one, liquid in the other, or instance).

An imba lance can cause meter zero errors; there-ore, measuring low ow rates can be problematic.

An equally signicant problem occurs at rates

too low to sweep bubbles out o the tubes. I the

uid velocity is less than approximately 0.6 m/s,

air will hang up in tube regions where the ow

is against gravity (Figure 1). Bubbles get caught in

the inlet tube leg because uid velocity is not great

enough to push the bubbles down and out against

gravity orcing the bubbles up. Tis issue is present

in any bent-tube meter design because at some loca-

tion in the tube the uid velocity is ghting gravity.

Te solution is to keep ow rate high enough

Figure 2. Even at low ow rate, measurements or 10,000-cP toothpaste with 2-5% void raction are within specifcation.



5/12 6

such that uid velocity can purge the sensor o

air. A rate o 20% o meter nominal ow (1 m/s in

the ow tube) or higher is adequate to completely

purge the meter o bubbles and give good peror-

mance. In a U-shaped meter, mounting the sensor

in a vertica l pipe run with ow going up helps to

keep the bubbles moving through the meter.

tHe roLe oF FLuId ProPertIeS

Pressure, luid temperature and viscosity all

impact how a Coriolis meter deals with varying

levels o entrained air.

As pressure increases or decrea ses , the appar-ent void raction changes, o course. his means,

or instance, i two meters are piped in series, the

downstream meter is at a distinct disadvantage

because the pressure is lower and thereore the

void raction is higher.

emperature plays a minor role, in that it a-

ects viscosity and surace tension. It also impacts

void raction to a small degree (higher tempera-

ture results in higher void raction).

Viscosity is a very important uid parameter

because it directly inuences the propensity o

the uid to hold up air (or gas). In a low-viscosity

liquid such as water, air bubbles coalesce rom nely

distributed small bubbles into large ones that collect

at high points in the line. In contrast, i the bubbles

stay nely distributed, as happens in high-viscosity

liquids, they will be purged rom the meter easily

and not collect and metering will be accurate.

Figure 2 shows results or toothpaste with a viscosity

o 10,000 cP and entrained air level between 2 and

5%. Rates are quite low or a 2-in. meter (


6/127

be measured because o the previously described

separation issues. However, noise rejection im-

provement with new DSP techniques will allow

the minimum rate to be pushed lower.

A signicant problem with any two-phase

ow (water/air, dog ood with solids suspended in

water, oil/gas, etc.) is at zero ow. When the ow

is stopped, the multiple phases separate by gravity,

prompting an imbalance in the tube. Tis imbal-ance causes an apparent meter zero change. Work

on signal processing improvements to address this

problem is currently a signicant area o research.

A SPeCIAL CASe

Empty-ull-empty batching can pose a related measure-

ment issue. Such batching is most common to avoid

cross-contamination o products when lling large

tanks such as rail cars or trucks. Tereore, the loading

line is purged with air or other inert gas between loads,

leaving the meter empty beore and ater the batch.

Generally, this application is not too difcult

because the batches tend to be long (greater than

one minute). Any transient meter behavior at the

beginning and end o the batch is sma ll compared

to the whole batch, so errors a re washed out.

However, when batches are short (less than one

minute), the transient errors can account or a signi-

cant raction o the total error. Air may be entrained

or a brie period, but the main issue is the time ittakes to ll the meter with uid. For instance, an ap-

plication running at 3 m/s will take about 0.1 s to ll

i the tube length is 0.3 m; an application at 0.3 m/s

will take a ull second simply to ll the meter. Experi-

ence has shown that i the meter ll-time is less than

0.1 sec., good batching perormance can be achieved,

regardless o the meters tube geometry.

tIM PAtten is director o measurement technology or

Micro Motion, Inc., Boulder, Colo. E-mail him at Tim.Pat-

[email protected].

mailto:[email protected]:[email protected]://www.emersonprocess.com/mailto:[email protected]:[email protected]:[email protected]


7/12

usa h Aaci f MamsMagnetic owmeters provide accuracy and can be used in a variety o applications

and environments

By David W. Spitzer

MAgnetIC FLowMeterS are among themost versatile o lowmeter technologies. hese

meters measure liquid velocity, rom which the

volumetric low rate is inerred. he measure-

ment is linear with liquid velocity and exhibits a

relatively large turndown. In addition, the range

o accurate low measurement is relatively large

and easy to change a ter installation.

Straight-run requirements are relatively short,

so magnetic lowmeter technology can be applied

where l imited straight run is available. In addi-

tion, the technology has no Reynolds number

constraints, so it can be used or liquids with high

or varying viscosity. However, liquid electricalconductivity constraints must be satisied or

these lowmeters to unction.

he only wetted parts o the lowmeters are

the liner and electrodes, both o which can be

made rom materials that can withstand cor-

rosion. his makes the lowmeters suitable or

use in chemical plants where corrosion may be

a concern. wo-wire magnetic lowmeters are

available that do not require power wiring. hese

can replace an existing lowmeter using the exist-

ing conduit or wiring with little or no electrical

rework.

PrInCIPLe oF oPerAtIon

Magnetic owmeters use Faradays Law o electro-

magnetic induction to determine the velocity o a

liquid owing through a pipe. Following Faradays

Law, ow o a conductive liquid through a magnetic

eld will generate a voltage signal. Tis signal is sensed

by electrodes located on the ow tube walls. When

the coils are located externally, a non-conductive liner

is installed inside the pipe to electrically isolate the

electrodes and prevent the signal rom being shorted.

For similar reasons, non-conductive materials are used

to isolate the electrodes or internal coil designs.

Te uid itsel is the conductor that will move

through the magnetic eld and generate a voltagesignal at the electrodes. When the uid moves aster,

it generates more voltage. Faradays Law states that the

voltage generated is proportional to the movement o

the owing liquid. Te transmitter processes the volt-

age signal to determine liquid ow.

SeLeCtIon FACtorS

Many actors must be considered when selecting

a owmeter, including the ambient conditions to

which the owmeter primary and transmitter will be

exposed. For the most part, the ambient temperature

rating o the owmeter primary is higher than that

9 Next PageutPrevious Page


8/12 10

o the transmitter and does not limit applicability.

Many primary and transmitter enclosures that are

rated or NEMA 4X or IP67 provide adequate pro-

tection against ambient humidity and precipitation

encountered in outdoor installations.

Operating conditions inside the pipe include

pressure, temperature and liquid conductivity. In

addition, the liquid can be corrosive or abrasive. Tese

conditions are typically addressed using appropriatemechanical design and material selection. Pressure

requirements are addressed by appropriate design o

the ow tube or the application. One supplier makes

a specially designed magnetic owmeter that can

withstand 1,500 to 2,000 bar (more than 20,000 psi).

Many primaries are available with polytetrauoro-

ethylene (PFE) or peruoroalkoxy (PFA) liners that

are rated to about 266 Degrees F and 356 Degrees F

(130 Degrees C and 180 Degrees C), respectively.

Less expensive liners rated to lower temperatures are

oten available to handle less demanding applications.

Appropriate electrode and liner material selection can

reduce the eects o corrosion and abrasion. ake care

when using ceramic liners because they can shatter

when temperature gradient constraints are exceeded.

Whereas the conductivity o the liquid in a typical

magnetic owmeter must be maintained above about 5

mixro-Siemens/cm (micro-S/cm), special low-conduc-

tivity designs are available that operate as low as about

0.01 micro-S/cm. Some owmeters require more than

50 micro-S/cm, however, they are low-cost units that

are oten applied to water or wastewater service where

this conductivity is usually not a constraint.Te amount o straight-run pipe required to

achieve the stated accuracy o the owmeter is a

reection o the quality o the design and the tight-

ness o the accuracy specication. In many applica-

tions, these owmeters will unction accurately

with about three nominal pipe diameters upstream

and two nominal pipe diameters downstream o

the electrode.

Magnetic owmeter operation requires good

electrical connections between the electrodes and

the liquid. Te quality o this connection can

degrade i an electrode becomes coated or corroded;

this can compromise AC owmeter accuracy by

shiting the zero, and may cause the owmeter to

ail to operate. Te advent o DC-pulse excitation

transmitters reduced much o the need to address

this issue. In addition, some manuacturers have

designed their transmitters to exhibit a relatively

high input impedance to help decrease the eects o

connection quality.

Magnetic owmeter coils can use and store sig-nicant amounts o energy relative to the amount o

energy needed to cause ignition. Most magnetic ow-

meter transmitters are designed to be non-incendive, so

normal transmitter operation will not cause ignition.

However, when installed in some hazardous locations,

ormal approval is required, and the transmitter must

be designed and installed to address the hazard.

A hazard may be present not only in the general

location o the primary and transmitter, but also inside

the pipe where the electrodes can provide a source o

ignition. o mitigate this hazard, the circuits o some

designs limit the energy available at the electrodes to an

amount less than that required or ignition.

Maintaining equipment is simplied when sel-

diagnostics are available to help the user. Te extent

and quality o the diagnostics and their ease o use

varies by manuacturer. Changing ranges is easier

and more accurately perormed in a digital manner.

Potentiometer adjustments and step switches are more

prone to problems.

otHer ConSIderAtIonS

Te market or magnetic owmeters is competitive,so prices are relatively low. Magnetic owmeters or

water and wastewater service can be economical due

to the economies o scale and the relatively low cost o

liners and electrodes or this service. However, applying

magnetic owmeters to corrosive or abrasive services

can signicantly increase the cost o the meters.

Magnetic owmeters or use in the chemical in-

dustry are typically more expensive than vortex shed-

ders. In some applications, the cost can rival that o

turbine owmeter or orice-plate owmeter systems.

Magnetic owmeters are typically more economical

than Coriolis mass owmeters.



9/12 11

FLowMeter PerForMAnCe

Te purpose o installing a owmeter system is to

accurately measure ow in a reliable manner. Issues

related to physical properties, process parameters,

electronic eatures and interconnections are oten

given much consideration. Relatively little empha-

sis, however, is given as to how well the owmeter

will perorm its intended purpose. Adding to the

conusion are the dierences in how perormance isexpressed and the incomplete nature o the avail-

able inormation. Nevertheless, the quality o ow

measurement should be a concern.

Te perormance o a owmeter is quantied by

its accuracy statements. Te reader must under-

stand not only which parameter is being described,

but also the manner in which the statement is

expressed. In ow measurement, parameters are

commonly described in terms o a percentage o

the actual ow rate, a percentage o the ull scale

o rate, or a percentage o the meter capacity. Tese

terms are mathematically related, so it is possible to

convert one to another (able 1). Note that when

compared on a common basis, such as percent o

rate, these statements describe signicantly dier-

ent perormance.

Other terminology may be used to express these

concepts. When this occurs, conrm exactly what

the other terms mean so they can be understood.

Perormance statements apply to a range o

ow or, stated dierently, between a minimum and

maximum ow velocity. It is important to identiy

the range in which the statement applies becauseperormance can be signicantly degraded or

undened when the owmeter operates outside o

this range.

Complicating the issue are some owmeters that

have dierent perormance statements or dier-

ent measurement ranges. For example, a owmeter

may have a reerence accuracy o 0.25% o rate

rom velocities o 1 to 10 m/s, and an absolute er-

ror o 0.0025 m/s rom 0.1 to 1 m/s. Perormance

is undened below 0.1 m/s. able 2 describes this

perormance using the above inormation. Note how

perormance degrades at low ows.

PerForMAnCe CLAIMS

For the most part, the claims made by suppliers

regarding magnetic owmeters are true statements,

even though they may seem extraordinary. Te

problem is that the statement may be incomplete,

and may not include certain acts and inormation

that clariy the statement. Sometimes claims are

simplied or convenience and easier understand-

ing. However, in many cases, urther investigationmay reveal other motives or doing so.

For example, consider a magnetic lowmeter

that has a reerence accuracy o 0.25% o rate and

a turndown o 1,000:1. he implication is that

the lowmeter can measure within 0.25% o rate

over a 1,000:1 range o low. aken individua lly,

both parts o the claim are likely true statements.

Yet when combined, t hey can be misle ading by

omission. Further investigation will show that

the reerence accuracy o 0.25% o rate applies

only within a range o low rates. Below the

minimum low rate o the range, the reerence

accuracy becomes a ixed absolute error. So as the

low rate decreases, the accuracy expressed as a

percentage o rate will increase.

Assuming that the reerence accuracy o 0.25%

o rate applies between 5% and 100% o meter

capacity, and that between 0.1% and 5% o meter

capacity, the reerence accuracy is xed at the abso-

lute error at 5% o meter capacity. able 3 calculates

reerence accuracy throughout the range o ows.

Tis illustrates that above 0.5 m/s, the reerence

accuracy is 0.25% o rate and that the turndown is10/0.01 or 1,000:1, both a s claimed. What is not

stated in the claim is that the reerence accuracy

degrades below 0.5 m/s and can approach 12.5%

o rate. Also not stated is that in actual installa-

tions, ows near meter capacity would rarely be

encountered, so the 1,000:1 turndown would rarely

be achieved. Assuming a more reasonable ull-scale

calibration range o 0 to 2 m/s, this owmeter

would achieve a 0.25% o rate reerence accuracy

rom 0.5 to 2 m/s, or a 4:1 turndown, and only a

200:1, or 2/0.01, turndown when the stated peror-

mance at low ow rates is included.



10/1212

In addition to high turndown, some suppliers

claim that their lowmeter operates at extremely

low low rates. Consider a claim to measure

velocity as low as 0.01 m/s. For a meter with a

capacity o 10 m/s, this corresponds to a 1,000:1

turndown. Although the lowmeter may operate

at this low rate, able 3 shows that it does so

with a reerence accurac y o 12.5% o rate.

Statements about magnetic owmeters otenclaim high reerence accuracy. What oten is not

stated is that it may apply over a range o higher

ows, and much o this range may not be encoun-

tered in actual operation. Furthermore, the reer-

ence accuracy a s a percentage o rate generally

degrades or is undened at lower ow rates (see

tables). When the ca librated ull-scale is low, and

the high reerence accuracy statements are limited

to a small range o high ow rates, the stated

reerence accuracy may not be achieved.

In general, reerence accuracy should be clear-

ly and completely stated or all ow rates prior to

perorming any analysis. Te range o applicabil-

ity o the high accuracy statement and the actual

operating ow range should be compared.

Magnetic owmeters are among the most

versatile o owmeter technologies. However, the

user should be aware o the manner in which their

application and operation are described in orderto ensure that the proper magnetic owmeter is

selected and installed.

dAvId w. SPItzer has more than 25 years o experience in

speciying, building, installing, commissioning and trouble-

shooting process-control instrumentation. Spitzer is a principal

in Spitzer and Boyes LLC, which oers consulting services or

the process industries in addition to product development,

marketing and distribution consulting or manuacturing and

automation companies.

http://www.burkert-usa.com/


11/12 13

tak a diff Lk a Cifal PmpsAn unconventional assessment can provide insights or eective control.

By Andrew Sloley, Contributing Editor

oFten, CrItICAL understanding o a systemcomes rom turning the common analysis on its

head. With centriugal pumps, this means thinking

that ow results rom back-pressure on the pump

discharge, not that pump discharge pressure varies

with ow rate.

Tis unconventional approach was crucial in

addressing a troublesome control system or the

overhead o a natural gas liquids plants debutanizer,

which goes to an accumulator downstream o a reux

pump (Figure 1). Varying the rate o liquid product

controls tower pressure. A dual-range controller

handles the reux drum pressure -- one valve lets in

uel gas to pressurize the system when pressure drops,

another vents the drum when pressure rises to too

high a level.

Tis rather odd system did not work. Te reux

pump cavitated all the time and pressure control was

erratic.

Te owner, which acquired the unit during a

company buy-out, lacked tower drawings, exchanger

inormation, pump curves, control valve inorma-

tion and historical operating data. Current operating

personnel had no experience with the unit and neverhad seen it work stably.

Lack o inormation doesnt justiy ignoring the

problem. So, lets examine this systems undamentals

and explore the most serious shortcomings.

Centriugal analysis starts by looking at two

things: the system curve and the pump curve. Te

system curve is the head loss required versus ow rate

through the system. Te pump curve is the dynamic

head generated by the centriugal pump. Te intersec-

tion o the system and pump curves denes the ow

rate the system will get.

We most commonly attain the required ow rate by

adjusting the system curve by adding an extra pressuredrop via a control valve. Alternatively, we can change

the pump curve using an adjustable speed drive.

In Figure 1, the reux control valve is a hand-op-

erated control valve (HCV). Te reux system doesnt

include an automated pressure drop. It essentially has

a xed system curve. Tis brings us to thinking about

the pump operation: pump ow stems rom back-

pressure on the pump.

Now, lets consider the tower pressure-control

system. PC3 adjusts the product ow out o the

system with the intent o changing the liquid level

in the condensers. Varying wetted condenser area

on the process side allows or pressure control. Tis

is a simple, ast-acting and eective system or total

condensation services.

Meanwhile, the product drum pressure-control

system maintains a constant destination pressure or

the net product rom the reux pump.

Te static head to the top o the tower ar exceeds

the pressure change between the tower and product

drum.

Te problem comes rom how the systems interact.

Te pressure control system requires level to existsomewhere up in the heat exchangers. Tink o the

exchangers and piping as a tall narrow vessel -- i more

liquid exits a vessel than goes in, the level drops, and

vice versa. Te pressure control system should perorm

similarly to a tight level control system.

Te system curve or the reux stream includes

two components: static head and system pressure

drop. Static head doesnt change with ow rate, but

system pressure drop does. I static head makes up

most o the system curve, the curve is relatively at.

Flat system curves create large ow rate changes rom

small pressure drop changes.



12/12

Te ow rate out o the vessel (heat exchangers

and piping) is very sensitive to changes in the HCV

position. Te HCVs purpose is to generate enough

pressure drop in the reux line so the unit can operate

in a sweet spot where the pressure control system will

work. In this case, we suspect the sweet spot is toosmall. Te back-pressure on the pump imposed by

the HCV usually is too low. At low back-pressure the

pump capacity exceeds the liquid rate. More liquid

is leaving the vessel (heat exchangers and piping)

than going in. Liquid level drops quickly. Finally, the

pump cavitates.

How can we address this problem?

One way is to try to put as much dynamic pres-

sure drop on the HCV as tolerable. Tis makes the

reux system curve steeper, which gives more stable

control. Tis is cheap and quick.

A second, and better, approach automates the

HCV. Control systems should transer a disturbance

rom where its important to where its not. Whats

important here is the ow rate out o the vessel

-- so we can maintain tight level control. We must

move the disturbance to something unimportant.

Many dierent congurations are possible. Tecloud in Figure 1 shows one o the simplest and

easiest options. A strap-on ultrasonic ow meter

along with a bolt-on actuator on the HCV enables

ully automated control o the overhead system. Te

disturbance now is in the valve pressure drop -- an

unimportant spot.

Tere are other ways to approach this problem.

But a undamental understanding o the system

comes rom looking at the pump backwards.

Andrew SLoLeyis a Chemical Processing Contributing

Editor. You can e-mail him at [email protected]

Debutanizer OverheaD

Figure 1. Lack o an automated pressure drop in the reux system made control difcult.
mailto:[email protected]:[email protected]

flow ehandbook

Documents