Load Representation for Dynamic Performance Analysis

8/12/2019 Load Representation for Dynamic Performance Analysis

1/11

472 IEEETransactions on Power Systems, Vol. 8, No.2, May 1993LOAD REPRESENTATION FOR DYNAMIC PERFORMANCE ANALYSIS

IEEE Task Force onLoad Representation for Dynamic Performance

A b s t r a c t - This paper summarizes the state of the rt ofrepresentation of power system loads for qynamic performanceanalysis purposes. It includes definition of terminology, discussionof the mportance of load modeling. important considerations fordifferenttypesof loads and differenttypesof analyses. Typicalloadmodel data and methods for acquiring data arc reviewed. A list ofreccnt refmnccs is included.Key Words - Load Modeling, Power System Modeling, PowerSystem Dynamic Peaformance.

1. INTRODUCTIONThe purpose of this paper is to review the current state of the rtof power system load representation for dynamic performanceanalysis. The power system engineer bases decisions concerningsystemrcinforceanentsand/= system pafarmance in large part on heresultsofpower flow and stabilitysimulntionstudies. Representation

inadquacies that cause under- or over-building of the system ordegradation of reliability could prove to be costly. In performingpower system analysis, models must be developed for all w e n tsystem components, including generating stations, transmissionanddistribution equipment, and load devices. Much attention has beengiven to models for generation and transmissioddistributionqu ip me w The npresentationof he loadshas received less attentionand continuesto bean area of greater uncmainty. Many studies haveshown that load representation can have a significant impact onanalysis results. Therefore, efforts directed at improving loadmodelingarcof major importance.One objective of thispaper s to summtuizc basic load modelingconcepts and provide guidelines for different t y p e s of dynamicperformance analysis. The accurate modeling of loads continues tobea difficulttaskdue to several factors, ncluding:a. Large number of diverse load components,b. Ownershipand ocation of load devices in customer facilities

not directly Bcccssibleto the electric utility,c. Changing load composition with time of day and week,seasons, weather, and throughtime,d. Lackof precise nformation on the composition of the oad.

Utilities often collect information on load composition for loadforecasting purposes. However, this information is sometimes notreadily accessibleor in usable form for system analysis purposes.Recent advames in techniques for measuring load characteristics andfor detemining \them from composition data can contribute o moreaccurate load representation. A second objective of this paper is toreview these echniques, identify data sources and encouragemoreextensive load data acquisition.This paper is divided into ten sections. Basic load modelingconcepts and definitions are given in Section 2. The need forimproved load rcpresentation s discussed in Section 3. Section 4describes current industry practice,based on a survey conducted bythe task force. Section 5 is a discussion of issues to be considered inselecting load models. Section 6 resents guidelines for applyingload models to specific ypes of studies. The data for load models isdiscussed in Section 7, including both measurement-based andcomponent-based methods of obtaining data. Typical data is alsogiven. Techniques for validating load representationaxe discussed inSection 8. Conclusions regarding the state-of-the-art and furtherRBtD needsam presented in Section 9. A list of recent references isgiven in Section 10.2. BASIC LOAD MODELING CONCEPTS

This section provides basic definitions and concepts related toload modeling.L O A D - The term load can have several meanings in powersystem engineering. including:

a. A device. connected to a power system, that consumespower,b. The total power (active and/or reactive) consumed by alldevices co ~ e ~ t e do a power system,c. A portion of the system that is not explicitly represented in a

system model. but rather is treated as if it were a singlepower-consuming device connected to a bus in the systemmodel.d. The power output of a generator or generating plant.

e. Uncertainties regarding the characteristics of many loadcomponents, p&cul~ly for M e voltage or frequencyvariations.when themeaning is not clear from the context, the t e rn , loaddevice, system load, *bus load, and generator or plant load,respectively, may be used to clarify the intent.Definition c is the one that is of main concern in the presentpaper. As illustrated in Figure 1,load in this context includes, notonly the co~ectedoad devices, but some T all of the following:p r e , t i o nwas coordinated by weW.ice (Chairman) withcontribuoons from H-D. Chiang, H.K. Clark, C. C o n d a , D.C.Lee, J.C.Hsu, S. Ihara. C.A. King, C.J. Lin, Y.Mansour,K Srinivasan, C.W. Taylor, and E.Vaahedi.

92 Wn 126-3 PWRS A paper recommended and approved - Secondary distribution feedersby the IEEE Power System Engineering Committee of - Shuntcapacitorsthe IEEE Power Engineering Society for presentationat the IEEE/PES 1992 Winter Meeting, New York, NewYork, January 26 - 30, 1992. Kanuscript submitted - Customer Wiring, transformers, and capacitorsAugust 30, 1991; made available for printingDecember 31, 1991.

- Substation step-down transformers- Subtransmission feeders- primiuy distributionfeeders- Distribution transformers- VoltageregulatorsAccurate load representation thereforequires accounting for theeffects of these elements. Exactly what is included in the loaddepends on what is and is not represented in the system model.Studies of bulk power transmission systems often omit much of thesubtransmissionaswell as the distribution system.

0885-8950/93 03.00 992 IEEE


2/11

473

SubstationTransformer

Figure I BusLoad including feeder s, transformers, and shuntcapacitors,as well as load devices.In describing the composition of the load, the following terms,illustrated in Figure2, are recommended:

LOAD COMPONENT - A load component is the aggregateequivalent of al l devices of a specific or similar type, e.g., waterheater, mom air conditioner, fluorescent lighting.LOAD CLASS - A load class is a category of load, such as,residential, commercial, or industrial. For load modeling purposes,it is useful to group loads into several classes, where each class hassimilar load composition and load characteristics.LOAD COMPOSITION - The fractional composition of the loadby load components. This term may be applied to the bus load or toa specific oad class.LOAD CLASS MIX - The fractional composition of the bus loadby load classes.LOAD CHARACTERISTIC - A set of parameters, such aspower factor, variation of P with V, etc., that characterize thebehavior of a specified load. This term may be applied to a specificload device, a load component, a load class, or the total bus load.

The following terminology is commonlyused in describingdifferent ypes of load models:LOAD MODEL - A load model is a mathematical representation ofthe relationship between a bus voltage (magnitude and frequency)and the power (active and reactive) or current flowing into the busload. The term load model may refer to the equations themselves

BusLoad

4 PQ

ItLoad ClassMix Composition

Industrial

Commercial

Residential

IILoad

Classes

Resistive Heat0Room Air Cond.

LghtingUWater Heater-0

LoadComponents

or the equations plus specific values for the parameters (e.g.,coefficients, exponents) of the equations. Depending on thecomputational implementation of these equations in a specificprogram, the load power or current may not be calculated explicitly,but it is useful to think of the model in these terms.STATIC LOAD MODEL - A model that expresses the active andreactive powers at any instant of time as functions of the bus voltagemagnitude and frequency at the same instant. Static load models areused both for essentially static load components, e.g., resistive andlighting load, and as an approximation for dynamic load components,e.g., motor-driven loads.DYNAMIC LOAD MODEL - A model that expresses the activeand reactive powers at any instant of time as functions of the voltagemagnitude and frequency at past instants of time and, usually,including the present instant. Difference ordifferentid equations canbe used to represent such models.CONSTANT IMPEDANCE LOAD MODEL - A static loadmodel where the power varies directly with the square of the voltagemagnitude. It may also be called a constant admittance load model.CONSTANT CURRENT LOAD MODEL - A static load modelwhere the power varies directly with the voltage magnitude.CONSTANT POWER LOAD MODEL - A static load modelwhere the power does not vary with changes in voltage magnitude.It may also be called constant MVA load model. Because constantMV A devices, such asmotors and electronic devices, do not maintainthis characteristic below some voltage (typically 80 toW ), anyload models provide for changing constant M V A (and other staticmodels) to constant impedance or tripping the load below a specifiedvoltage.POLYNOMIAL LOAD MODEL - A static load model thatrepresents the power relationship to voltage magnitude as apolynomial equation, usually in the following form:

p.f.

1.

.a2

1.

Component CharacteristicsdP dQ dP dQdV dV df

2. 0. 0. 0.

.5 2.5 0.6 -2.8

1.54 0. 0. 0.

2. 0. 0. 0.

J

MotorParameters

etc.

Figure 2 . Terminology or C omponent-BasedLoad Modeling.


3/11

474The parameters of this model are the coefficients (ai to ag) and thepower factor of the load. This model is sometimes referred to as theZIP odel, since it consists of the sum of constant impedance (Z),constant current I),nd constant power P) terms. If this, or other,models areused for representing a specific load device, Vo should bethe rated voltage of the device, and Po and Qo should be the powerconsumed at rated voltage. However, when using these models forrepresenting a bus load, Vo, o. and Qo are normally taken as thevalues at the initial system operating condition for the study.Polonomialsof voltage deviation from rated (AV) are also sometimesused [24].EXPONENTIAL LOAD MODEL - A static load model thatrepresents the power relationship to voltage as an exponentialequation, usually in the following form:

Two or more terms with different exponents are sometimes includediw ac h equation. The parameters of this model are the exponents, npand nq, and the power factor of the load. Note that by setting theseexponents to 0, 1, or 2, the load can be represented by constantpower, constant current, or constant impedance models, respectively.Other exponents can be used to represent the aggregate effect ofdifferent types of load components. Exponents greater than 2 or lessthan 0may be appropriatefor some types of loads.FREQUENCY-DEPENDENT LOAD MODEL - A static loadThis is usuallyrepresented by multiplying either a polynomial or exponential loadmodel by a factor of the following form:model that includes frequency dependence.

where f is the frequency of the bus voltage, fois the rated frequency,and af is the frequency sensitivity parameter of the model.Bus Frequency - The frequency of the bus voltage is not aninherent variable in fundamental-frequencynetwork analysis and isnot used i n many dynamic performance analysis programs.However, it can be computed by taking the numerical derivative ofthe bus voltage angle. This frequency is required, not only in thestatic frequency-dependent oad model, but also in some other loadmodels, such as a dynamic induction motor model. The busfrequency is sometimes approximated by using an average systemfrequency, computed from a weighted average of synchronousmachine speeds. This approximation is wrong because it will notproduce the w m t mpact on damping of oscillations.EPRI LOADSYN STATIC LOAD MODEL - The static modelused in theEPRILOADSYN program [121:

The active power is represented by two aggregate components,one frequency-dependent and the other not. The first term of thereactive power equation represents the reactive consumption of the allof the load components. The second termapproximates the effect ofthe net reactive consumption of feeders and transformers minus shuntcapacitance, included in the bus load, to give the specified initialreactive flow (Qo) at the bus.

EPRI ETMSP STATIC LOAD MODEL - The general modelused in the EPRI Extended Transient Mid-term Stability Program[23]: P=P(C1)+P( CMVA)+P(CZ)+P( V. f )Q=Q(CI)+Q(CMVA)+ Q(cz)+Q(~,~)

where:P(C1) = Po (ICONCP/100) (V/Vo)QCI) = Qo (ICONCQ/l00) (V/Vo)P(CMVA) = Po (ICONPP/100)Q(CMVA) = Qo (ICONPQ/100)P(CZ) = Po (ICONZP/100) (V/Vof

+ K P2 loo) (V/Vo)a2 (1+KPf2(f- ~))]

ICONZP = 1 - CONCP + ICONPP + KP1+ KP2ICONZQ = 100- O N C Q + ICONPQ + KQ1+ KQ2INDUCTION MOT OR - For modeling of induction motors, moststability programs include a dynamic model based on the equivalentcircuit shown in Figure3. Other features available in some programsare additional rotor circuits, saturation, low voltage tripping, andvariable rotor resistance.

RI L1 L2

Figure 3 Induction motor equ ivalent circuit, where s is the motorslip, R2 and L2 a re rotor val ues, R1 and L1 are armaturevalu es, and L,,, is the mag netizing inductance.It is important to note that the slip used in this model is thefrequency of the bus voltage minus the motor speed. Someprograms incorrectly use either average system frequency or 1.0 inplace of the bus frequency. As with the frequency-dependent oad,such approximations will incorrectly represent damping effects.Several levels of detail, based on this equivalent circuit, may beavailable, including:

1. A dynamic model including hemechanical dynamics but notthe flux dynamics,2. Addition of the rotor flux dynamics,3. Addition of the stator flux dynamics.Stator flux dynamics are normally neglected in stability analysisand the rotor flux dynamics may sometimes beneglected, particularlyfor long-term dynamic analysis. Low voltage tripping is an

important feature for voltage stability analysis and other studiesinvolving sustained low voltage.CLOAD MODEL - A load model used in the PTI PSS E stabilityprogram including dynamic models of aggregations of large andsmall motors, non-linear model of discharge lighting, transformersaturation effects, constant MVA, shunt capacitors, and other staticload characteristics and a series impedance and tap ratio to representthe effect of intervening subtransmission and distribution elements.


4/11

475machines (near decelerating machines), there will be an optimisticimpact on the results. On the other hand, a constant MVA load modelwould have the opposite impact on the results since it would hold theload power at a higher value during the depressed voltage. It istherefore difficult to select a model that is guaranteed to beconservative for all parts of the system and for various disturbances.Also, static load models have proved to be inadequate in casesinvolving large deviations of voltage and/or frequency, especially inislanding situations [15]. For example, if a disturbance results inislanding some part of a system with a deficiency in generationaccompanied by considerable drop in voltage, a constant impedanceload model may indicate a frequency rise rather than drop.Moreover, while constant MVA is considered by many as aconservative representation for induction motor loads this may bem e only for the active part of the load. The reactive power part,however, is in fact expected to increase as the voltage decreasesbeyond a certain value. In these cases, there is no substitute forrealistic dynamic load models.Small-Signal Stability Damping Studies

Inter-area modes of oscillation, involving a number of generatorswidely distributed over the power system, often result in significantvariations in voltage and local frequency. In such cases, the loadvoltage and frequency characteristics may have a significant effect onthe damping of the oscillations. A recent study [16] for the westernNorth America power system showed that using a constantimpedance load representation in small signal analysis tended tooverestimate the damping by about 25 as compared with a moreaccurate load representation.Voltage Stability Analysis

Voltage stability analysis is a class of studies that does not lenditself to simple static load modeling. Reference [ I O ] shows the extentof load model details that the Swedish State Power Board had to useto duplicate the field records of the 1983 blackout. Initial effortsusing simple static load models failed to explain the voltage collapsescenario. It was then recognized that the load characteristics at lowvoltages (below 0.8 pu) will not follow the characteristics that aretraditionally used in stability studies. Detailed characteristics ofinduction motors, lighting, refrigerators, and air conditioners areexplained in the reference.Voltage stability studies done at Ontario Hydro for the OttawaArea reemphasized the need for accurate dynamic load modeling. Ina study of the impact of the loss of one of the transmission linesfeeding the area, a substantial difference was found between theresults obtained with static and dynamic models. The former led toentirely the wrong conclusion [15].While it is often assumed that the nearly constant MVAcharacteristics of induction motors will produce the most severeresults, B.C. Hydro's experience i n the 1979 voltage collapseincident of the northern part of the British Columbia system showedthe fallacy in generalizing this conclusion. The constant impedancecharacteristics of arc furnaces near the sending end of a radial systemresulted in increasing transfer to the receiving end due to the loadreduction with voltage near the sending end [22].Summary

The foregoing discussion indicates that there is no general rule asto which load model is reasonable to use. It is this task force'sopinion that there is no substitute for spending the effort to identifythe actual load characteristics of a given system and to use modelsappropriate to the type of study being performed. A considerableimprovement in modeling fidelity can be achieved by fully utilizingexisting models including dynamic models. However, furtherdevelopment, as described in Section 9, is also needed.

4. CURRENT INDUSTRY PRACTICEA survey regarding current industry practices and research wastaken by the task force. The survey was sent late in 1988 toapproximately 85 industry representatives, mostly in North America,

LOAD POWER ORDER - The value of the bus load (P and/or Qat rated voltage and frequency. In most studies, this value is heldinvariant throughout a simulation, so that the only variations in loadP and Q are due to changes in system voltage and frequency.However, load power order can be varied in some studies, e.g., toanalyze voltage stability during a load ramp or to study the effect ofimpact loads such as arc furnaces. This term has been coinedrecently to clarify the distinction between this quantity and the powerdemand, which is the power consumed at the actual voltage andfrequency.3 NEED FOR IMPROVED LOAD REPRESENTATION

Electric utility analysts and their managementrequire evidence ofthe benefits of improved load representation in order to justify theeffort and expense of collecting and processing load data and,perhaps, modifying computer program load models. The benefits ofimproved load representation fall into the following categories:A. If present load representation produces overly-pessimistic results:

1. In planning studies, the benefits of improved modeling willbe in deferring or avoiding the expense of systemmodifications and equipment additions,2 . In operating studies, the benefits will be in increasing powertransfer limits, with resulting economic benefits.1 . In planning studies, the benefit of improved modelingwll be

in avoiding system inadequacies that may result in costlyoperating limitations,2 . In operating studies, the benefit may be in preventing systememergencies resulting from overly-optimistic operating limits.

In addition, failure to represent loads in sufficient detail may produceresults that miss significant phenomena.It is difficult to quantify the benefits of improved loadrepresentation. Several studies, reported in the literature, havedemonstrated the impact that different load models can have on theresults of different types of studies. In some cases, the impact can besignificant. A recent voltage stability study for the Pacific NorthwestSystem showed that when the heating load of the residential part ofthe winter peak load was represented accurately (close to constantimpedance for first-swing stability studies), a considerably highertransfer limit was achieved. This may result in a major saving incapital expenditure. On the other hand, an optimistic load modelcould result in pushing systems beyond their actual limits and makingthem vulnerable to major collapses. For example, the Tokyo systemcollapse of 1987 was partly due to underestimating the characteristicsof the reactive power consumption of air-conditioning loads.A common philosophy, in the absence of accurate data on loadcharacteristics, is to assume what is believed to be a pessimisticrepresentation, in order to provide some safety margin in the systemdesigns or operating limits. This has been shown to be a dangerousapproach because it is not always possible to select representationsthat will be pessimistic for all parts of the system or all testconditions. Some of the reasons for this are as follows:First-Swing Transient Stability Studies

System voltages are normally depressed during the first angularswing following a fault. The power consumed by the loads duringthis period will affect the generation-load power imbalance andthereby affect the magnitude of the angular excursion and the first-swing stability of the system.Consider a case where the loads have a constant currentcharacteristic (power consumption varies directly with the magnitudeof the voltage) [12]. If a constant impedance load model were used,the power consumption would vary with the square of the voltageand, therefore, would be lower than the actual load during thedepressed voltage period. For loads near accelerating machines thiswill give pessimistic results, since the generation-load imbalance willbe increased. However, for loads remote from the accelerating

B . If present load representation produces overly-optimistic results:


5/11

476with a 40 response rate. A majority of those who responded werenot using special load modeling software. Of those who were, themajority wereusing theWRI LOADSYN package.For pre- and post-contingency power flow analysis, constantMVA load models were the overwhelming choice. However, severalutilitiesdid r ponusing voltagedependent load models for this typeof analysis.For first-swing and small-signal stability analysis. constantcurrent for active power load and constant admittancefor reactiveload was the dominant choice, though it was by no meansoverwhelming. Other load models reported in use included constanta@nittance for both active and reactive load, static polynomialmodels, and combinations of the above. At least two utilitiesreported the use of dynamic inductionmotormodels for certain ypes

The survey results indicated that approximately50 of thosewho responded were dissatisfied with their present load models andwere pursuing work to improve them. This work included gatheringmore detailed end-use data, installing digital and/or transientrecorders at selected locations within their systems, and conductingfield tests to develop new models or validate existing representations.Several of the industry representatives who responded to thesurveyraised concerns with regard to data-gathering difficulties. Itwas perceived to requ rea p a t deal of data-gatheringeffort to utilithe morecomplex load models. It was suggested that theLOADSYNprogramprovide generic data for a wider range of industries similarto that already supplied for steel mills and aluminum plants. Inaddition, it was suggested that more research work be done in theareaof data acquisition.

5. LOAD MODELING CONSIDERATIONS

of stability analysis.

A number of modeling considerations have been discussed inSection 3 to indicate the importance of good load modeling. T h i ssectionsummslrizes these and other important considerations or thefollowing types of system dynamic pexformance studies:first-swing,srnall-signal- damping,- synchronizingpower,load-generation imbalance,induction motor tability,cold-load pickup,voltage stability, anddynamic 9vervoltages.

Whether or not a particular oad component shouldbemodeled indetail i s a study depends on how much the component responseaffects the voltage and frequency excursions typical of that type ofstudy. Each of the above studies poses unique load modelingrequirements as outlined in the brief introductory paragraphs thatfollow.First-Swing -- First-swing problems exhibit large and rapid voltageexcursions during the initiating fault and slower voltage excursionsduring the first power-angle swing, which lasts one second or less.Load response to these voltages is important, as discussed in Section3. There is also a brief frequency excursion during the power-angleswing, so frequency characteristics of loads close to accelerating ordecelerating genera tm can also be important.Small-Signal Stability Damping -- In studies of the damping ofpower-angle oscillations, typically in the range 0.1 to 1.2Hz oadresponse to sinusoidal variations in voltage and frequency isimportant. Fkquency variations are greatest near machines with thegreatest participation in the particular mode of oscillation. Voltagevariations end tobe greatestat intermediate points between opposingmachines or groups of machines.Synchronizing Power Margin -- Studies of synchronizing powermargins may requireevaluation of periods from one or two minutes

to 20 minutes or more after a disturbance, thus making long-termload characteristics mportant. Long-term load voltage characteristicsand tap-changing ransformers(LTCs) ear the electrical center of asystemwill be imprtant when angles BCIDSS a system appraach90.Generation-Load Unbalance -- In studies of generation-loadunbalance, response of load components to frequency decay rates inthe range 0.1 to about 4 Hz per second is important. For lowfrequency decay rates, loads follow their long-term voltage andfrequency characteristics, while inertia and, sometimes, electricaltime constants of motors can come into play at higher frequencydecay rates.Induction Motor Stability -- In studies of induction motor stability,the important question is whether motors will reaccelerate or stallfollowing fault clearing. Motor and shaft load inertia, contactor hold-in characteristics, and motor electrical parameters such as the rotorcircuit time constant can be as important as the stiffness of thesystem.Cold Load Pickup -- In cold load pickup studies, almost all customerload device characteristics are important. For instance, thermostatsand people may add load devices to the system while the feeder isopen, while protective devices may disconnectsome motors. Motorstarting current and inertia affect the current in the feeder in the firstseconds after it is reclosed. Distribution substation LTCs are veryimportant, and if not modeled explicitly, mustbemade a part of theload model.Voltage Stabiliry -- Voltage stability is usually a longer-term problemmuch like that of maintaining synchronizing power margins to avoidinstability, and the load modeling requirements are similar.Characteristicsof loads under abnormally low voltage conditions andthe action of LTCs and other voltage control devices must beaccurately modeled.Dynamic Overvoltages -- Dynamic overvoltages are usuallyassociated with sudden loss of line or network loading as can occurupon bypassingor blocking of HVDC converters, load rejection onradial generation, and instability and breakup in higher voltagenetworks and networks in which shunt capacitors or cables areheavily used. A dynamic overvoltage study may address the designof an static var system (SVC) to limit overvoltages or the risk ofgenerator self-excitation. Though not much load may be in theproblem area or left on the network when the overvoltages occur,even a small quantity of load can have a significant effect. Henceload response to voltage, and in particular, saturation characteristicsof load devices, will be important. Since harmonics will affect theovervoltage, fundamental frequency analysis is not always adequatefor such studies.Load CharacterizationLoads can becategorized as follows:

loads that exhibit fast dynamic electrical and mechanicalcharacteristics the primary examplesare themechanical andelectrical time constants of induction motors. Adjustablespeed drives are another example, though they respond soquickly to voltage and frequency that they exhibit theirlonger-term characteristic hroughout most time simulations.loads whose response to voltage excursions exhibitsignificant discontinuities; examples include:

discharge lighting (constant current active part, voltage to4th powa for reactive part, extinguish at about 80 ofratedvoltage),adjustablespeed drives that shut down on low voltage ashigh as 90% of rated voltage),motor contactors that drop open during faults and voltageswings, removing motor load from the system, andmotor overload protection that removes stal led motorsfrom the systemafter about 10 seconds


6/11

477loads whose response to voltage excursions does not exhibitsignificant discontinuities or time lags; examples include:- very small motors,- incandescent lighting,- uncontrolled resistive loadsloads with slow dynamic characteristics such as:- loads controlled by thermostats and manually-controlledloads that are initially constant resistance, but change toconstant power over a 10 to 20 minute period after achange in voltage.

Specific load device characteristics include:Motors -- A motors inertia and rotor flux time constant are keydeterminants of the motors active and reactive power response to thevoltage and frequency swings that characteristically follow faults.Reaccelerating motors near accelerated generators tend to slowgenerators and improve stability. Reaccelerating motors in weakareas pull voltage down, reduce synchronizing power and thusdegrade stability. Motor starter contactors may drop open at 6575in the case of 2300 and 4000 V motors and 5565% in the case ofmotors at 460V and below. Motors may drop during the fault, orduring the voltage excursion that follows the fault. Near sending-endgenerators, this will degrade stability as the lost load power adds totransfers. Near the receiving end or in weak areas, the loss of motorload may improve stability. Motor active and reactive power leadvoltage and frequency during sinusoidal voltage and frequencyvariations, usually degrading damping. Motors consume about 70%of the electrical energy in the U.S., and are thus clearly an importantload element.Air Conditioners -- As load devices, air conditioners fall into themotor category, but many air conditioners are not served bycontactors and thus do not drop from the system during faults andvoltage excursions. Because of low inertia, air conditioners slowconsiderably during faults, and thus impose large, low power factorstarting currents on the network after the fault is cleared. If thenetwork is not strong enough to reaccelerate all of the air conditionerssimultaneously, voltage will be depressed and the air conditionerspeeds will continue to decay. The constant-torque mechanical loadimposed by the loaded compressor and the low starting torque designof the motor (the motor is chosen only to start an unloadedcompressor) contribute to the tendepcy to stall. Because the voltageis low, overload protection i s slow to operate; taking typically 10 to12 seconds to remove the air conditioners from the system. Voltagebegins recovering as the air conditioners are removed, but all air.conditioners must be tripped by overload protection even if voltagerecovers fully because at low speed motor torque will be below thatof the loaded compressor. Air conditioning can be as much as 50of the summer load in some areas.Discharge Lighting -- Discharge lighting includes mercury vapor,sodium vapor, fluorescent, and similar types widely used in industryand for parking lots and street lighting. This type of load mayrepresent up to 20 of system load in commercial areas. This loadextinguishes at about 80 voltage, and thus may drop to zero duringfaults and when voltages swing down during a power-angle swing.A momentary drop in load near the electrical center of a system(where the voltage excursion is the largest, see Figure4)will usuallyimprove the first-swing. On the other hand, lighting near sendingend generators that extinguishes and takes one to two seconds torestart after the fault, can increase generator acceleration and thusdegrade stability.

At voltages above the extinguishing point, conventional dischargelighting is typically constant current active part and voltage to the 4thpower reactive part. This reactive characteristic can help stabilizevoltage even where voltage does not reach the extinguish point. Thenewer electronic ballasts, high efficiency lamps, and light-controlleddimming fixtures are closer to a constant MVA load, but theirextinguish points need to be explored.

2 With Discharge Lighting ModeledWith Discharge Lighting Not Modeled

Figure 4 . When voltages swing below 80 following fault clearing,discha rge lighting w ill extinguish, reduce system load, andstabilize voltage between 70 and 80 . This will improvestability as shown in thisfigure.

Incandescent Lighting -- Incandescent lamps are often assumed tohave a constant resistance characteristic. However, due to the largetemperature swing that occurs in the filament when voltage changes,the filament resistance changes with voltage. The result is activepower that varies with about the 1.55 power of voltage; abouthalfway between constant current and constant resistance.There is a thermal time constant associated with incandescentlamps, but it is short compared with other time constants of concernin stability analysis. Only in lamps above about 200W is the thermaltime constant large enough to be of concern.Thermostat-controlled Loads -- Many loads, such as space heaters,molding and packaging machines, soldering machines, water heatersand the like, are very close to a constant resistance characteristic -- inthe short-term. Such loads operate in a temperature range where theeffect of changes in power input do not change temperature enoughto measurably affect resistance. However, beginning just secondsafter a drop in voltage, the reduction in heat output from such deviceswill be sensed by thermostats and the on part of the thermostatcycle will be extended. Thermostats that are in the off period oftheir cycle when the voltage change occurs will not respond to thevoltage drop until they enter the on portion of their cycle.Figure 5shows the temperature profile of a thermostat-controlledload with 100% and 90 voltage. As would be expected,temperature rises more slowly than normal during the onportion ofthe cycle when voltage is low, but temperature drops at the same rateduring the off portion of the cycle. However, the averagetemperature, and thus the average power input, are the same in bothcases. Because of the longer on period, more thermostats are in theon mode at any given instant when voltage is low, and the total loadis thus the same as it is at normal voltage. The result is thatthermostat-controlled loads are, effectively, constant power loads ona long-term basis. Since cycle periods vary 5 to 30 minutes), theload will completely restore only after the off period of the longestthermostat period.There are possible exceptions. Where voltage is so low thatsome thermostats can not be satisfied, the load will not be fullyrestored. Similarly, on a very cold day, thermostats on electric spaceheating may already be on at or close to 100% of the time, so thatfull restoration does not occur following a drop in voltage.


7/11

478

Ea

50 On-Time at 100 VoltageOn OnOHn

55.6 On-Time at 9OV0VoltageOn Ono- m

LAverage Temperature

7 2 LTime

Figure5 When voltage is low, thermostats on resistance typelo remain closed longer, effectively restoring the loadto a constant power characteristic in the steady-state.

There are many industrial, commercial, and residential loaddevices that have this characteristic. This component may make up20 to 40 of the system load.In first-swing studies and damping studies of less than about 5minutes, thermostat-controlled loads should be given a constantresistance characteristic. For long-term studies of several minutes ormore duration, a dynamic characteristic, which changes fromconstant resistance to constant power with a suitable time constant,should be used.Manually Controlled oads -- Manually controlled loads are quitesimilar to thermostat-controlled oads. For instance, a pot of waterwill be left on an elecmc range until boiling is clearly visible. Ifvoltage is low, the waterwll remain on the electric range for a longerthan normal period. Many people doing the same thing results inmore pots of water on ranges at any given instant when voltage islow. This component of load makes the transition from constantresistance o constant power in several minutes.Electronic Devices Regulated Power Su pplies) -- Many electronicdevices (e.g., computers, microwave ovens, TVs) include switch-mode and other regulated power supplies. These power supplies willprovide a constant dc output for input voltage down to about 90 ofnormal. Hence they are effectively constant power down to thisvoltage. Below this voltage some of these loads will cease tofunction, though some will continue to draw power. The powerinput thus may drop when voltage falls below about 90 of ratedvoltage.Adjustable Speed Drives - - Adjustable speed drives are likecomputers and electronic equipment, except that shutdown willtypically occur when voltage drops below the lowest tolerance, about90 of normal. Response to voltage is fast relative to rotor angleoscillations, so such loads are effectively constant power even infirst-swing and small-signal stability studies, so long as voltage isabove the point at which shutdown occurs.LTCs -- Though not a load component, LTCs on distributionsubstation transformers are not explicitly modeled in most powerflow data bases, and thus, by default, become part of the load. AsLTCs restore subtransmission and distribution voltages after adisturbance, they return the power flowing to voltage-sensitivedevices to pre-disturbance levels. The increasing load can bring onvoltage instability, deplete synchronizingpower margins, or degradedamping. LTCs typically begin moving about one minute after a dropin voltage, and complete the restoration of voltage within another oneto two minutes.System Structure ConsiderationsThe system structure s less important than the type of ppblem beinganalyzed. For instance, voltage stability problems can occur in anytype of system, and load response to low voltage will be importantregardless of the system structure.

Systems with long transmission distances or weakinterconnectionsare most susceptible to synchronizing power anddamping problems. Voltage response of loads will be importantwhen synchronizing power margins are being considered. Insystems with damping problems, frequency excursions will be largenear machines that have the greatest participation in the particularmcde of oscillation. Voltage excursionswill be large at intermediatepoints between opposing machines or groups of machines.Therefore, the response of loads to both voltage and frequency willThe first swing will be less sensit ive to load characteristics intightly coupled systems. In these systems a relatively severe fault isrequired to threaten stability, and the synchronizing powerwll belarge and voltage excursions will be modest and short-lived. Forthese reasons, load characteristics will play only a minor role in thestability of tightly coupled systems.In weakly coupled systems, voltage and frequency excursionswill be large and Synchronizing power will be modest. Loads thuswill tend to play a significant role in first-swing behavior in suchsystems.It is not always obvious when loqd characteristicsare importantin stability studies. A straight-forward procedure to assesssensitivity o load characteristics s to make comparative simulationswith a range of load models. For instance, one may compare aconstant impedance (active and reactive) load model with a constantcurrentor constant power load model.If a comparison of algebraic load models shows significant loadeffects, a more thorough evaluation using dynamic models for

motors may be warranted. Some portion of the load, typically50to 70%, can be replaced with motor models. The motor modelshould be based on typical induction motor data and have a loadtorque proportional to speed (average of pump and conveyercharacteristics). It is not necessary to include dynamic motors inevery bus load, so long as the desired portion of total load in an areais representedasmotors.Load Composition

Load Characteristics can vary significantlywith time of day, dayof week, season, and weather. It may be prudent in some studies toassume the worst-case load model for the problem under study.However, determining he worst-case load model may not be a trivialtask. It may be necessary to build a range of load models and testeach.In warm climates, summer load may include large amounts of airconditioning load. Winter loads may consist of electric heating inareas of low energy cost, or heat pumps in other areas. Where fossilfuels are the primary source of winter space heating, therewll be acomponent of motor load driving air handlers and fuel oil pumps.Both air conditioning and heating loads are seasonal but re alsovery sensitive to weather. Cold, windy weather will increase heatingload and hot, humid weather will increase airconditioning load.Weekday load is likely to be dominated by industrial loads. Upto 95 of the load in some industrial plants is motor load. Someprocess plants work around the clock and through weekends.Commercialbuilding loads consist argely of air conditioning anddischarge lighting. This component of load is largest duringweekday business hours.Light industry often includes significant elecmc heating devices(packaging, soldering, etc.). As noted earlier, this load is constantresistance in the short term but is continually returned to a constantpower level over a period of minutes.Agricultural load consists largely of motors driving pumps,especially in growing seasons. Synchronous as well as inductionmotors may be used for this application.Modeling Alternatives

Most present programs support one or more of the static loadmodels de sc ri w in Section 2, and dynamic induction motor models.Available dynamic motpr models include those with rotor inertia androtor flux transients represented and those with rotor inertiarepresented and electrical characteristics based on the motorequivalent circuit. The later type neglects rotor flux transients and

be lpXtant.


8/11

479thus may not always give sufficiently accurate results in first-swingand damping studies. The effective rotor circuit time constant oftypical motors is in the range .05 to .15 seconds.Some researchers have investigated the relationship betweenvoltage and active and reactive load power, and a useful body ofinformation on load-voltage characteristics exists. However, fewinvestigations of the response of composite loads to frequencyvariations have been done. New models or improvements in existingmodels may be indicated by such tests.Development of special models may be warranted for largeunique loads such as aluminum smelters, chlorine plants, pumpingplants, and steel mills.An efficient composite load model may be useful. Presently allload components can be modeled individually by detailed dynamicload models, but these models greatly increase simulation times. Anefficient load model might accept directly the output of the EPRILOADSYN program and provide the appropriate dynamiccharacteristics for the study underway. Running time would bereduced, for instance, by neglecting frequency effects in voltagestability studies that do not involve significant frequency variation.

6. LOAD MODEL APPLICATION GUIDELINESSimple static models may be adequate where parametric casesshow such models give the same study results as more detaileddynamic models. Comparison of simple models and detailed modelsusing typical data in both cases may be taken as a first step in thecomparison. Where detailed load models are shown to be necessary,the typical data shouldbe replaced by data based on best estimates ofactual load device characteristicsor test results.The threshold of significance for differences in study resultscaused by load model changes will depend on the engineers point ofview. His point of view may be strongly influenced by his studyobjective and the engineering and computer time that is available toachieve it. Major factors will include the time required to gather theinitial data for detailed models, the annual update effort, the datahandling effort during a study (e.g., when changing from summer towinter conditions), and the power flow and simulation programrunning times. A planner may be less worried about accuracy on theone hand (especially if hes looking 10years into the future), but alsohave the time to prepare and use detailed models. Operating studiesmay need high precision to set actual day to day transfer limits, butbe constrained by the computer time available and the power flow orstability program running time (which will be longer with detailedload models). The engineer investigating a major disturbance willprobably have the greatest need for accuracy, and may also have thetime to achieve it. In each case, the engineer must evaluate hisrequirements and resources, and choose an appropriate compromise.Use load models that accurately reflect specific load types whenlarge special loads, such as chlorine plants, aluminum reductionplants, smelters, and electric furnaces, are in a part of the networkthat will experience significant voltage or frequency excursions.Power plant auxiliaries, including adjustable speed drives shouldbe modeled in detail in transient stability studies, especially in areaswhere faults will be placed. Because of the frequency variations thatoccur near generating plants during oscillations, plant auxiliaryinduction motors may be as important asmotors in industrial plants.Saturation effects in transformers, motors, and customer loadsare important in simulations involving transients connected with lossof load or system breakup and consequent high voltages.To ensure accuracy, stability studies should employ gooddynamic load models. The load model should include the effects ofmotor inertia, motor rotor flux transients, contactor drop-out,discharge lighting discontinuities, and similar phenomena. Tailored

load models should be used for large special loads such as aluminumreduction plants. The effects of LTCs on load magnitude in theminutes after a disturbance should be taken into account in dampingand voltage stability studies. Synchronizing power and voltagestability studies should take into account the longer-termcharacteristics of loads. The effects of thermostats and LTCs andmanual intervention by the customer are examples.

7 DATA FOR LOAD MODELINGThere are two basic approaches to obtaining data on compositeload characteristics. One is to directly measure the voltage andfrequency sensitivity of load P and Q at representative substationsand feeders. The other is to build up a composite load model fromknowledge of the mix of load classes served by a substation, thecomposition of each class and typical characteristics of each loadcomponent. Each of these approaches has advantages anddisadvantages, and each is described in more detail below. In

general both should be used, as they are complementary, in order tounderstand and predict load characteristics under varying conditions.Measurement-Based Data

Data for load modeling can be obtained by installing measurementand data acquisition devices at points where bus loads are to berepresented. These devices must measure voltage and frequencyvariations and the corresponding variation in active and reactivepower, either in response to intentional disturbances or to naturally-occuring events. These systems can also perform the function ofdisturbance data acquisition.Several utilities have been developing systems to gather such data[2,13,21]. A new generation of such devices is being developedunder the name of disturbance monitors [19,20,25].One such system developed at IREQ has been operating at aHydro-Quebec substation since May, 1989 [25]. It consists of asubstation resident unit, which monitors voltages and currentscontinuously and a central computer which can control and retrievedata from several substation resident units. Optionally, when thereare no disturbances, the mean and covariance of the voltage phasor,current phasor and frequency deviation over several minutes arecomputed and stored on a continuous basis. These may be used forsmall signal modeling purposes.A microprocessor-based transient data recording system has beendeveloped at Taipower company to investigate actual load behaviorsduring system disturbances [21]. A t the present time, nine sets ofself-acting monitoring system are installed at the level of primarysubstations and distribution substations of the Taipower system. Thelocations were selected at substations which supply powerpredominantly to residential loads, commercial loads, or industrialloads. When a system disturbance occurs, small or large, therecorders at each substation trigger automatically to record the three-phase voltages and currents onto a local computer. A digitalsampling technique is used to record voltages and currents at thesubstation. The Discrete Fourier Transform technique is used totransform the recorded data into a phasor form. The voltage andcurrent phasors at each substation are then used to compute both thereal and reactive power at the substation. Since June 1988, severaldisturbances caused by either short-circuit faults or generator tripaccidents have been detected by the load transient recorder system.The data are being used to identify parameters for both static anddynamic load models.The parameters of a load model are estimated by fitting these datato the assumed model. The determination of static load modelparameters from staged step changes in voltage are straightforward.Sophisticated estimation techniques may be needed to obtain morecomplicated models and to obtain models from data obtained duringnaturally-occuring disturbances. The data from such disturbances arebeginning to be available. These data are expected to provide newinformation on the load characteristics.Measurement-based techniques have the obvious advantage ofobtaining data directly from the actual system. However, there areseveral disadvantages, including:Application of data gathered at one substation to load modelsfor other substations may only be possible if the loads arevery similar,Determination of characteristics over a wide range of voltageand frequency may be impractical,Accounting for variation of load characteristics due to daily,seasonal, weather, and end-use changes requires on-goingmeasurements under these varying conditions.

-

--


9/11

480One way of overcoming the fiist disadvantage is to use theapproach reported by Ontario Hydro in Reference [9].Measurements were made on feeders supplying specific classes ofloads, e.g., residential,commercial,industrial, etc., toobtain typicalcharacteristic data for each load class. These characteristics werethen combined according to the mix of diferent oad classes presentat each bus. Measured static characteristics or different load classesatOntario Hydro appear in Reference191. These tests were repeatedfor other system conditions to enhance the accuracy of the LoadClassCharacteristics and the derived load models. The Load ClassMix data requi redfor this approach is the same as requi redfor thecomponent-based approach d scussedbelow.

Component-Based DataThe purpose of the the component-based approach is to developload models by aggregating models of the individual componentsforming the load. Component characteristics, e.g., for airconditioners, luasce nt lights, etc., canbe determined by theoreticalanalysis and laboratory measurementsand,once determined, used byall utilities. Much of t h i s datahas been determined and documentedby EPRI projects [1,3,12], lthough it needs to be updated for newand redesigned load devices.Since t would be difficult to determine directly the number ofeach type of load component on each feeder and substation in asystem, a simpler approach was devised in EPRI project RP849-7[12,14]. This approach, illustrated by Figure 2, requires thefollowing information:1. Typical compositions (fractions of load consumed by each

typeof load component)of each of several load classes, e.g.,residential,commercial, industrial,2. M i x of load classes at each bus.The EPRI LOADSYN program [14] rovides an automatedmeans of combining these data toobtain loads models of varyingcomplexity in standard forms for commonly-used power systemanalysis programs.The composition data will vary with seasonal, daily, and otherfactors. The EPRI Load Modeling Reference Manual [12] ncludestypical data on load compositions for several load classes,geographical regions, seasons, and heating fuels. This data,obtained from census and load research data, can be used as a firstestimate by individual utilities. Procedures for gathering morespecific load composition data are described in the manual.The load mix data,which describe the percentage contribution ofeach load class to the total load, vary from bus to bus in the system.The load mi x data are also dependent on time and weather, based onconsumption patterns and the underlying composition of each class.These functionalities of theLoad Mix Data make it difficult to dentifywith great certainty. It remains an area that deserves attention byutilities to set up standard methods for derivation and formats forcommunicating it. Development of models, such as those used forload forecasting, that would relate weather, time of day, day ofweek, and other factors to the relative consumptions of different loadcomponents n each load class, shouldbe considered.One of the m w e s available for identifying the load mix data isload consumption information recorded by utilities for otherpurposes. Utilities use data recordeTs o measure the power flows atmany points in their systems. romthese measurements hey derivethe consumption foreach load or a group of loads. This nformation,which is updated every few minutes. is kept on computers files andtherefore can easily be used for other purposes.A method describedin Reference [9] ses the customer billingdata to identify the load mix data. The billing data contains energyconsumptionfor each quarter of the year for each of several regionsand types of loads. This nfamation is beingused by Ontario Hydroto obtain seasonal load mix data for use with the LOADSYN

P r o g r a m .The composition-based approach to load modeling has theadvantage of ot requiringfield measurements and of being adaptableto different systems and conditions. Its main disadvantage is inrequiring the gathering of load class mix. and perhaps loadcomposition, data, which are not normally used by power system

analysts. Further industry attention to devising standard forms andmethods for collecting this data would ease this problem. Severalutilities are actively pursuing this data collection.Typical Data

Typical data haspreviously been published in a number of papersand reports. The most extensive compilations are found inConcordia and Ihara [6] nd he reportson he FPRIW849 projects[1,3,121.Table I shows a set of static load characteristics for various loadclasses, geographical regions of North America, seasons, andprincipal heating fuel. This data was obtained from the EPRILOADSYN program using the default load compositions andcomponent characteristics distributed with that program. Theparametersin the table are for a model of the following form:

Q =Q~ v/G Kqv [l + Kqf f - ~ ]Table ITypical Load Voltage and Frequency Parameters

RESIDENTIAL:Elec. HeatingNortheastSummerWinterNorth-CentralSummerWintersummerWinterSummerWinterNon-Elec. HeatingNortheastsummerWinterNorth-CentralSummerWinterSummerWinterSummerWinter

southWest

southWest

D.f.

.90.99

.90.99

.87.97

.92.99

.91.93

.91.96

.89.97

.94.97COMMERCI LElec. Heating

Non-elm. HeatingSummer .85Winter .90Summer .87Winter .90

INDUSTRIAL .85PRIMARYALUMLNUM .90STEELMILL .83POWER PLANTAUX. 30AGRICULTURALPUMPS .8

rn1.21.71.11.7.91.51.31.71.21.61.31.51.11.61.41.5

.5.6

.7. 8

.11.8.6.11.4

2.7 .7 -2.32.6 1.0 -1.72.6 . 8 -2.32.6 1.0 -1.72.4 .9 -2.12.5 .9 -1.82.7 . 8 -2.22.5 1.0 -1.52.8 .7 -2.33.1 .7 -1.92.8 .7 -2.23.0 . 8 -1.72.5 .9 -2.02.9 . 8 -1.62.9 .7 -2.12.8 .9 -1.3

2.5 1.2 -1.62.5 1.5 -1.12.5 1.3 -1.92.4 1.7 -0.9.6 2.6 1.62.0 1.5 .61.6 2.9 1.81.4 5.6 4.22.2 -0.3 .6

In the absence of any further information on the loadcomposition, the most commonly accepted static load model is torepresent active power as constant current and reactive power asconstant impedance. The constant current active power represents amix of resistive and motor (nearly const. MVA) devices. Thisrepresentationm a y be shifted more towards constant impedance orconstant MVA, if the load is known to be more resistive or moremotor-driven, respectively. This representation for an average loadis reasonably well supported by thedata n Table I.


10/11

48 1Investigation of possible existing sources of data on systemload characteristics, such as billing data, load researchstudies, and system disturbance monitors.Development of a plan for acquisition of improved load data,either through measurements or load midcompositionanalysis.

8. VALIDATION OF LOAD REPRESENTATIONThough the sophistication of load modeling in dynamicsimulations has improved greatly over the past several years, manyutilities still face the difficult question of deciding which load modelsto employ to best represent their system load performance.Connected with this is the need to validate the models that are chosenand the data obtained by one of the methods discussed above. Twomain approaches have been considered for load model validation.One is to make measurements directly at load substations in the same

manner as described in Section 7. The other approach, based onsystem-wide measurements, is described below.To address the question of load model validation, the EmpireState Electric Energy Research Corporation (ESEERCO) in 1986initiated a research project (EP86-7) with the goal of evaluating thesensitivity of New York Power Pool (NYPP) dynamic simulations toload modeling and validating the load models that are used in NYPPstudies. The technique employed to validate the load model involvesthe comparisonof simulated responses of important system variablesto recorded data for system disturbances.A digital acquisition system (DAS) is being developed forinstallation at ten critical stations within the NYPP system to providethe measured data. Measured responses to severe disturbances willbe analyzed and compared with simulations using load models ofvarying complexity. The comparison is based on fourier-derivedmodes of oscillation and damping factors; t h u s exact timesynchronization s not a requirement of the DAS units. The datacaptured by the DAS must be sufficient to compare with a 10-15second simulation.The DAS is designed to perform as a remote, unattended devicewhich is capable of communicating with a master processing uni t(MPU). Data transfer occurs via standard telephone lines. From theMPU, the engineer may examine and evaluate recorded data as wellas modify the various operational parameters such as the channeltrigger settings, scale factors, amount of pre-trigger data stored etc.The DAS itself is provided with 14 input channels to store stationvoltage and frequency as well as MW and MVar line flows. Asampling rate of 15 Hz is employed and storage capability is suchthat approximately 50 events of 30 seconds duration may be stored.A disturbance index file is maintained which contains the date andtime of the event as well as the channel which initiated the trigger.The DAS provides the user with a choice of two trigger algorithms, afixed value trigger and a moving range trigger whose limits will varyslowly with changing system conditions. The trigger settingsmay beeasily modified from the MPU. One challenge in analyzing themeasured results is to account for system changes that may resultfrom the disturbance, such as removal of portions of the system,including loads, by relay action.The advantage of the system-wide approach to model validationis that, if successful, t validates not only the load models but all ofthe models used in the analysis procedure. However, ifdiscrepancies are found between measured and computed responses,it may be difficult to determine the source of the problem. I t istherefore recommended that model validation include both system-wide comparisons and measurement on individual loads and othersystem equipment.

9 CONCLUSIONSThis paper has reviewed the state of the art of load representationfor dynamic performance analysis. It is concluded that considerableprogress has been made in recent years in understanding loadcharacteristics and, particularly, in methods for determiningimproved load model data.It is recommended that serious attention be given by powersystem analysts to the load models and data used in their studies.Steps should include:1. Familiarization with recent literature, including referenceslisted below.2 . Selection of most realistic models for near-term use, based onavailable data.

3.

4.

There are a number of areas where further research anddevelopment are needed, including:Models and data for load behavior at low voltages and overlong time periods, including self-mpping mechanisms forloads under low voltage conditions,Improved simplified modeling of induction motorcharacteristics, ncluding saturation, frequency dependency,and low voltage stall,mp, and recovery behavior,Updated models and component characteristics or new andredesigned end-use devices,Standardized procedures for collecting load class mix datawithin utility organizations,Procedures and formats for exchange of load model data,including between elecmc power retailers and wholesalers.Procedures for modeling the impact of weather, daily,seasonal, and other factors on load composition,Improved understanding and modeling of the effect ofsubtransmission and distribution network elements(transformers, feeders, capacitors,etc.) on the aggregate loadrepresentation, ncluding modeling of LTC transformers andof transformer saturation where important,Methods and equipment for measurements of loadcharacteristics,Enhancements n the modeling options available in dynamicperformance analysis computer programs, includingdischarge lighting, low-voltage motor behavior, and, forlong-term dynamics, LTC transformer and dynamic(thermostatic) oads.

10. REFERENCESThe following list of references is limited to papers publishedsince 1979. Reference 6 contains an extensive list of older referenceswhich should also be consulted. Reference [17] is a recent CIGREreport which includes a bibliography with 35 entries.

M.S. Chen, DeterminingLoad Characteristics or TransientPerformance, EPRI Report EL-840, Project RP849-3, 1979.K. Srinivasan, C. Nguyen, Y. Robichaud, A. St. Jacques,G.Rodgers, Load Response Coefficients Monitoring System:Theory and Field Experience, IEEE Trans. PAS-100, pp.T. Gentile, S. Ihara, A. Murdoch, N. Simons, DeterminingLoad Charactexistics for Transient Performance, EPRI ReportEL-850, Project RP849-1, 1981.S . Ihara, F.C. Schweppe, Physically Based Modeling ofCold Load Pickup, IEEE Trans. PAS-100, pp. 4142-50,1981.S Ihara, N.W. Simons, G.L. Paulsen, Bismarck LoadBehavior During Field Tests, IEEE Trans. PAS-100, pp.C. Concordia, S . Ihara, Load Representation in PowerSystem Stability Studies, IEEE Trans., PAS-101, pp. 969-977, 1982.T. Frantz, T. Gentile, S. Ihara, N. Simons, M. Waldron,Load Behavior Observed in LILCO and RG&E Systems,IEEE Trans. PAS-103, pp. 819-831, 1984.T. Ohyama, A. Watanabe, K. Nishimura, S . Tsuruta,Voltage Dependence of Composite oad in Power Systems,IEEE Trans. PAS-104, pp. 3064-73, Nov. 1985.E. Vaahedi, M.A. El-Kady, J.A. Libaque-Esaine, V.F.Caravalho, Load Models for Large-scale Stability Studies

3818-27, Aug. 1981.

4540-5, 1981.


11/11

482

r121t131

r141

c171Wl

from End-User Consumption, IEEE Trans. PWRS-2, pp.K. Walve, Modelling of Power System Components atSevere Disturbances,CIGREPaper 38-18,1986.F. Nozari, M.D. Kankam, W.W. Price, Aggregation ofInduction Motors for Transient Stability Load Modeling,IEEETrans. PWRS-2, pp. 1096-1103, 1987.W.W. Price, K.A. Wirgau, A . Murdoch, F. Nozari, LoadModeling for Power Flow and Transient Stability Studies,EPRI Report EL-5003, Project 849-7, 1987.T. Dovan, T.S. Dillon, C.S. Berger, K.E. Forward, AMicrocomputer Based On-Line Identification Approach toPower System Dynamic Load Modeling, IEEE Trans.W.W. Price, K.A. Wirgau, A. Murdoch, J.V. Mitsche, E.Vaahedi, M.A. El-Kady, Load Modeling for Power Flow andTransient Stability Computer Studies, IEEE Trans. PWRS-3,E. Vaahedi, H.M. Zein El-Din, W.W. Price, Dynamic LoadModeling in Large Scale Stability Studies, IEEE Trans.Y. Mansour, Application of Eigenanalysis to the WesternNorth American Power System, IEEE Pub. 9OTHO292-3,Eigenanalysis and Frequency Domain Methods for SystemDynamic Performance, 1990.CIGRE Task Force 38.02.05, Load Modeling andDynamics, Electra, May 1990.R. Ueda, S . Takata, Effects of Induction Machine Load onPower System, IEEE Trans. PAS-100, pp. 2555-62, May1981.IEEE Task Force Report, Instrumentation for MonitoringPower System Dynamic Performance, IEEE Trans. PWRS-2, pp. 145-152, February 1987.IEEE Task Force Report, Power System DisturbanceMonitoring: Utility Experiences, IEEE Trans. PWRS-3, pp.134-148, Feb. 1988.C.Y. Chiou, C.H. Huang, A.S. Liu, Y.T. Chen, T.H. Li andC.J. Lin, Development and Application of Microprocessor-based Load Transient Data Recording Systems, IASTEDInternational Conference, Power High Technique 1991,Taiwan, March 1991.IEEE Task Force Report, Voltage Stability of PowerSystems: Concepts, Analytical Tools , and IndustryExperience, IEEE Pub. 90THO358-2-PWR, 1990.EPRI, Extended Transient-Midterm Stability ProgramPackage FACTS Version Users Manual, EPRI Project1208-9, September 1990.EPRI, The Effects of Reduced Voltage on the Operation andEfficiency of Electric Systems, EPRI Report EL-3591, Vol.1 June 1984.K. Srinivasan, R. Jutras, Power System DisturbanceMonitoring System,Proc. IEEE T&D Conference,pp. 261-5, Dallas, September 1991.

864-72, NOV. 987.

PWRS-2, pp. 529-36, Aug. 1987.

pp. 180-7, 1988.PWRS-3, pp. 1039-45, 1988.

...

Documents

Load Representation for Dynamic Performance Analysis