guidelines for modeling power electronics in electric power engineering applications

8/4/2019 Guidelines for Modeling Power Electronics in Electric Power Engineering Applications

1/10

IEEE Transactions on Power Delivery, Vol. 12, No. 1, January 1997 505

GUIDELINES FOR MODELING POWER ELECTRONICSIN ELECTRIC POWER ENGINEERING APPLICATIONS

A.M.Gole, Albert Keri, C. Nwankpa, E.W.Gunther, H.W . Dommel, I.Hassan, J.R.Marti, J. A.MartinezK.G.Fehrle, L.Tang (Chairman), M.F. McGranaghan, O.B.Nayak, P.F.Ribeiro, R.Iravani, R. asseter

Power Electronics Modeling Task Force & Digital Simulation Working GroupABSTRACT -This paper presents a summary of guidelinesfor modeling power electronics in various powerengineering applications. This document is designed foruse by power engineers who need to simulate powerelectronic devices and sub-systems with digital computerprograms. The guideline emphasizes the basic issues thatare critical for successfully modeling power electronicsdevices and the interface between power electronics and theutility or industrial system. The modeling considerationsaddressed in this guideline are generic for all powerelectronics modeling independent of the computational tool.However, for the purposes of illustration, the simulationexamples presented are based on the EMTP or EMTP typeof programs. The procedures used to implement powerelectronics models in these examples are valuable for usingother digital simulation tools.1. INTRODUCTIONAs a consequence of the advances in power electronicstechnologies over the last two decades, power electronicsapplications have quickly spread to all voltage levels, fromEHV transmission to low voltage circuits in end userfacilities.Commonly observed power electronics applications includeHVDC terminals, various static var compensation (SVC)systems, high power ac to dc converter for dc arc furnaces,static phase shifter, isolation switch, load transfer switch,converterlinverter based drive technologies, active lineconditioning, energy storage and instantaneous backuppower systems, renewable energy integration, andnumerous others covered under subjects of Flexible ACTransmission Systems (FACTS) and Custom PowerSystems (CPS). The need for power electronics modelingand simulation is driven by both existing and newapplications.

96 SM 439-0 PWR D A paper recommended and approved by the IEEETransmission and Distribution Committee of the IEEE PowerEngineering Society for presentation at the 1996 IEEWPES SummerMeeting, July 28 - August 1, 1996, Denver, Colorado. Manuscriptsubmitted January 3, 1996; made available for printing May 7, 1996.

A comprehensive review of various benefits, technicalissues, capabilities of digital time-domain simulationpackages, and existing difficulties associated with digitaltime-domain simulation of power electronic switches andcircuits is given in [11.Objectives of the simulation include:

verify an application designpredict the performance of a systemidentify potential problemsevaluate possible problem solutions

The simulation is specially important for a conceptvalidation and design iteration during a new productdevelopment.Power electronics applications are relatively new to manypower system engineers. This guideline provides generalprocedures to help these engineers to make their ownsimulation cases as needed. The theories of powerelectronics are not in discussion. Attention is focused onthe simulation of the interaction between the powerelectronic sub-system and the connected power system.Thus, a model for a power electronic switching device canbe greatly simplified. More details about the devicerepresentation is presented in the later section of this paper.In the last section of the paper, the references related tovarious power electronics simulations are listed.2. GENERAL CONSIDERATIONS OF POWERELECTRONICS MODELING2.1 Types of ProblemsPower electronics modeling can be divided into two basiccategories, depending on study objectives. The firstcategory covers all steady state evaluations. The focus is onthe power system response to the harmonics injected from apower electronics sub-system. Examples of this typeinclude a study of the steady-state harmonic propagation ina transmission and distribution system, harmonic frequencyresonance, system voltage and current distortion, filteringdesign calculation and performance evaluation, telephoneinterference analysis and system losses associated withharmonics. In this type of study, the harmonic currentinjection can often be assumed independent of the voltagevariations at a point of common coupling (PCC), an

0885-8977/97/$10.000 996IEEE


2/10

506electrically dividing point between the utility system and thecustomer circuit. Therefore, the power electronics sub-system can be greatly reduced to a shunt circuit equivalent.The second type of power electronics modeling covers amuch more extensive and complex range of practicalproblems. In many applications, operation of a powerelectronics sub-system depends closely upon the operationstate of the connected system. To evaluate the dynamic andtransient performance of a systenl with power electronicsinterfaces, the monitoring and control loops of the system,including detailed signal processing and power electronicsdevice firing need to be modeled. Examples of this type ofapplications can be a SV C system, a SuperconductingMagnetic Energy Storage (SMES), active powerconditioning system and various adjustable speed drivesapplications. When modeling these applications, variationsof system parameters need to be used to derive the powerelectronic controls so the output of the power electronicssub-system changes accordingly as demanded. Because inthese cases, the power electronics sub-system directlyaffects overall system operation, a separate treatment of thesupply system and the power electronics sub-system isunacceptable.2.2 Frequency Dom ain and Time Dom ain SimulationAs far as the solution methods are concerned, there are twobasic approaches: the frequency domain solution and thetime domain solution. Digital computers can only simulatecircuit phenomena at discrete frequencies or at discreteintervals of time (step size Af or At). This leads totruncation errors in all digital simulations.Compared with the time domain calculation, a frequencydomain simulation is more robust because a circuit solutionis found at each individual frequency and truncation errorsare not accumulated. The programs using this solutionmethod often treat the non-linearity of a system as knowncurrent sources. For a harmonic evaluation, the frequencydomain solution usually requires less computation timecompared with a time domain solution. However, mostavailable frequency domain solution programs havedifficulties in handling the system dynamics, controlinterfaces and fast transients. The time domain solution isbased on the integration over a discrete time interval. Thenumerical methods applied in different programs can useeither iterative techniques or direct solution methods. Thesolution stability and accuracy achieved are closely relatedto the time step size selection. Because truncation errorscan accumulate from step to step, the solution may divergefrom the true solution if an improper time step is selected.The time domain simulation has great advantages over afrequency domain simulation in handling the systemdynamics, power electronic interface and transients.

2.3 Tools for Power Electronics Application SimulationAccording to their main functions, commonly used tools forpower electronics related simulations can be classified intothree groups.1. Power electronics simulation tools2. General transients simulation tools or EMTP type of

programs3 . General harmonics simulation tools or frequencydomain simulation tools

The tools of the first group may offer the best productivityif a detailed application topology and complicated operationcontrols need to be simulated and if the main interest of thestudy is the power electronics sub-system. These toolsoften have difficulties when an extensive utility networkneeds to be included in the simulation or when the maininterest of the study is overall system dynamics andtransients. In such cases, the EMTP type programs areusually more suitable tools. The currently availableversions of the EMTP type programs may be less efficientfor detailed power electronics modeling. However, theseprograms are very attractive for an application orientedpower electronics simulation because these programs offertremendous capability and flexibility in characterizingvarious types of power system components, includingpower electronics switches with reasonably simplifiedcharacteristics. As graphic interfacing features aregradually incorporated in these programs, a level ofdifficulty in using these tools will decrease.3. MODELING GUIDELINES3.1 Representation of Semiconductor Switching DevicesFor a power level application, the commonly usedswitching devices are power diode, thyristor, Gate Turn-Off thyristor (GTO) and Insulated Gate Bipolar Transistor(IGBT). Except for the diode that is a two-terminal,uncontrollable device, the others are three terminalcontrollable devices.Thyristors are commonly used in applications where onlyturn-on control is required. The device can turn-off withthe load commutation or by a forced commutation whenrequired. GTOs have found an increased number ofapplications because of their gate turn-off and high powercapabilities. The drawbacks of a GTO is its switchingfrequency limitation, around lOOOHz, and its highpercentage gate turn-off current requirement, 1/7 to 1/5 of aload rating current. IGBT is another alternative that hasbecome popular in industrial applications in recent years.The device has gate voltage control type turn-on and turn-off capability and the device can be switched at frequenciesup to several tens of kHz in practical applications. At thepresent time, IGBT applications are basically limited by its


3/10

507

Characteristic

I_

power rating. As of today, 3 MW IGBT based drives arecommercially available.

CharacteristicIgl IgO

For most power electronic simulations addressed in thisdocument, a simplified device characteristic is acceptable.Therefore, instead of trying to represent the switchingcharacteristic of a diode as shown on the left in Figure 1, asimplified characteristic or an idealized characteristic,shown by solid and dotted lines respectively on the right,can be used.

1Real Device 1haracteristic CharacteristicSimulated DeviceI Cathodea node I

Figures 1 . Actual and Simplijied Switching Characteristicof Diode DeviceRepresenting the reverse recovery characteristic, leakagecurrent, and forward voltage drop of a diode is often notnecessary for an application study. In commonly usedsimulation programs, the simplified diode model may beavailable as a built-in device or it can be realized with avoltage controlled switch. If the diode model does notallow to specify on-state and off-state resistances, they caneither be neglected with minimal loss of accuracy or can bemodeled by inserting a series resistance for on-state loss anda parallel resistance for off-state loss.

I ICathode& Anode

Figure 2. An Actual and Simpl$ed SwitchingCharacteristic of thyristor DeviceIn Figure 2, switching characteristics of an actual, asimplified and an idealized thyristor are compared. Torepresent the simplified thyristor device, the turn-on control

is added on the simplified diode model. If the control isapplied continuously, this switch simulates the diode whichallows unidirectional current flow when the switch isforward biased. Delaying the gate pulse allows control overthe turn-on instant of the forward biased switch.For numerical simulations, if the gating circuit powerrequirement is excluded from the study, there is very littledifference between modeling a GTO, IGBT or any otherthree-terminal, controllable, unidirectional current flowingdevice. The device can all be represented by a simplifiedswitch with gate turn-on and turn-off controls. Thedifferent switching characteristics can be realized byapplying different firing controls.

G a t e" ' O w d e

-N- mof z L J - O(1) 12 ) (31

Figure 3. Alternativesfor Three-terminal ControllableSwitching Device RepresentationIn many actual power electronics applications, in order toprovide a continuous current flowing path for an inductiveload, a reversal diode (free wheeling diode) is used inparallel with a controllable switching device to form thebasic power electronics switching unit. However, severalalternatives are available for implementing this basicswitching unit in digital simulations.Considering only the terminal equivalent at the connectionpoint of the application, the simplest model that can be usedto present one leg of the bridge rectifier or inverter can beconstructed with one simplified bi-directional currentflowing switch with gate controls as in Figure 3 (1). Thismodel uses the least number of devices for a given powerelectronics topology. The problem with this bi-directionalcurrent flowing switch representation is that it fails tocorrectly represent an operation state during the idlingperiod. This can be better explained with an example,Figure 4, of a simplified inverter scheme used in a UPS.

D CBattery A CSupply

I I I

Figure 4. Simplified Inverter Scheme in a U P S System


4/10

50 8

When all the controlled switching devices are blocked, thedc battery is being charged through the rectifier bridgeconsisting of six reversal diodes. However, if the ideal bi-directional switches are used, while the UPS is idling, thebattery will not be charged by the ac supply.When the idling mode needs to be more realisticallyrepresented, the simple bi-directional device model can beimproved with a paralleled reversal diode as shown inFigure 3 (2). When not only terminal effects but alsoindividual device current carrying conditions are of interest,a further improvement on the model can be made by addinganother diode as shown in Figure 3 (3).It should be noted that for some programs, multiple switchconnections on the same circuit node can lead to a singularsystem admittance matrix. When using a program withsuch restriction, a small series resistor or inductor can beused to create intermediate nodes and avoid the problem.3.2 Representation of Power Electronics (PE) SystemIf every individual power electronic switching device isrepresented, a system model containing power electronicapplications can easily reach a complication level that ishard to be implemented. For an example, a HVDCterminal contains tens to hundreds series and parallelthyristor devices in one converter leg for high voltage andMVA ratings. Obviously, if one wants to represent eachindividual thyristor device in this HVDC system model, onewill quickly end up with a huge model.Fortunately, (except for some failure mode analyses), forthe purposes of most application simulations it is notnecessary to represent all individual devices. What need tobe simulated usually is the terminal characteristics of apower electronic sub-system and how it interfaces with theconnected system. Thus, the following procedures can beused to reduce the modeling complexity:

Using one or a few equivalent devices to representseries and parallel combination of a group of devicesRepresent power electronic loads with similarcharacteristics by an equivalentUse the simplest device model which is appropriate forthe applicationRepresent a power electronic sub-system by equivalentsource injection whenever it is acceptableRepresent only the front end of the drive system whenthe major concern is utility interfacingInclude the system dynamic and controls only whennecessaryUse modular approach for large scale modeldevelopment

In this guideline, detailed considerations related toapplying these system reductions are presented. Someimportant issues addressed are:0 Harmonic cancellation when multiple loads are

Existing system distortion.0 Appropriate source topology for PE sub-system

represented by their lumped equivalent

representationSystem unbalanceEffects of a dc link or the inverter side connection onthe front end interface with the power supply systemCurrent or voltage sharing among the parallel or seriesswitching devices

0

0

o Switching loss prediction3.3 Representation of the Power SystemSimilar to the situation in a power electronics sub-system, apower supply system can easily extend to a large electricaland geographic radius and become too complicated tomodel. Therefore, the power system needs to be simplified.The proper level of system reduction depends on the studyobjectives.If the purpose is to characterize the harmonics generated bya particular type of power electronics application, thepower system model can be significantly reduced. When apre-existing voltage distortion level at a power electronicsinterfacing bus is low, the rest of the power system can besatisfactorily represented by one or a set of first orderequivalents connected to the bus at a higher system voltagelevel. For an example, if the power electronics applicationinterfaces with the system at the low voltage bus of a step-down transformer, the equivalent of the system can beplaced on the high voltage bus of the transformer. When apre-existing voltage distortion level is greater than 2 % , oneneeds an adequate harmonic source to properly represent thebackground distortion.If the objective is to evaluate effects of the powerelectronics on a connected utility system, the model shallbe extended to cover all sensitive loads (i.e. rotatingmachines and all other major power electronics) within aconcerned electrical radius. Special attention is needed ifan unbalanced system condition is involved.Extensive power system model is required for a harmonicpropagation and resonant study. The main systemcomponents and dominant topology need to be kept in thepower system model. Capacitor and filter banks, allnonlinear passive circuit components, and all otherharmonic injection sources should be represented.Frequency dependent characteristics of the systemcomponents might need to be considered.


5/10

5093.4 Representation of System ControlsThe system control is one of the most important aspect of apower electronics simulation. As illustrated in this paper, aswitching device is greatly simplified. The properswitching performance of a device is realized viaappropriate gate controls. Modeling of power electroniccontrols consists of three steps:1. Monitoring and sampling2.3. Device gating signal generation.Most simulation tools provide some means to implementsystem controls. In some later developed programs, thecontrol block diagram and flow-chart structures aresupported for modeling different levels of system controls.Using these tools, a user can define the specified controls ina simulated system with great flexibility. Some key issuesensuring a correct control modeling is briefly mentionedbelow. These issues are more thoroughly treated in theguideline with iIlustration examples.

Signal processing and control reference derivation

with all signal filters should be carefully implementedin the control model in order to accurately predictcontrol response. This is particularly important whenthe objective of the simulation is to verify controldesign and to evaluate the response of a powerelectronic application to primary system dynamics.All power electronic devices have their limit inswitching frequency. When a load commutation or astandard PW M type scheme is simulated, the highestswitching frequency in the simulation is controlled bythe system frequency or by a carrier frequency. Evenconsidering a variable carrier frequency, the number ofswitching per fundamental frequency cycle is knownand the highest switching frequency can be made undera physical limit of the simulated device. However, ifthe device firing is determined by a simple comparisonbetween the system control reference and the systemoutput, a device switching may take place in simulationwhenever a comparison difference is detected.Therefore, the switching frequency becomes highly

For a time domain simulation, the highest resolutionfor a signal sampling is determined by a selected timestep. In general, this presents no problem for analogcontrol. However, for digital control simulation, if theselected time step is too large and if the simulatedsampling resolution is significantly different from the

dependent on the time step size, and the averageswitching frequency becomes unpredictable. Whenusing this type of firing logic, user should always takeextra measures, such as introducing a hysteretic loop,to ensure that the modeled device is working under itsphysical switching capability.

3.5 Snubber Treatment in PE ModelingThe simulation programs using trapezoidal integration

real system sampling resolution, significant errors canbe introduced and even lead to instability.For a time domain thedoes not reflect the time method are inherently prone to spuriousoscillations (alsologic response bowns chatter) in capacitive and inductive circuits whentime.reasonable time delay to match with the limitations of current injection and switching.the control hardware and software.When modeling a control response, it is important to

remember to introduce a subjected to sudden changes such as step change in voltage,some EMTP typeprograms take special measures to detect and remove these

to thescillations in which it is not aunderstand the program time between simulation engineer. Otherwise, Some of the measuresthe primary system and the interface' For an listed later in this section may have to be implemented toexample, the control model may introduce one (could obtain correctbe more than one) time step delay because of structureand This may no t The finite nature of the simulation time step of the EM"

type programs also poses another problem for powerause problems in some simulations. However, if theerror accumulate over the time, it can eventually result snubber circuits across fast acting power electronicin the divergence' The problem be switches. Note that in some situations the snubberR and Ccorrected in most cases by reducing the size of the time values of the system may or may not work instep or avoiding the possible accumulation mechanism simulationsusing In this case, the and

values of the snubbers needed for stable simulation isn the control model.

method Of the program'logic makes this time caused electronic circuit simulation which necessitates use of

Different methods may be used to synchronize power primarily dependent on the time step and secondarily ongating with required 'Y stem system configuration (capacitors and inductors in thereferences. In many cases, a phase-locked-loop system), and the load current level. Programs us hg special(pLL) features such as variable time stem (verv short time stewe greatly simp1ified to reduce thesystem complexity. However, when the system

contains significant waveform distortions, eitherharmonics or transient disturbances, a practical PLLI \ .during switching) or interpolate switching (simulate the

switching very close to the required instant using linearinterpolation between time steps) do not require fictitious


6/10

510snubber circuits. For using a program that does not addressthese problems, one of the following measures or theircombinations can be taken to prevent numerical instabilityin the simulation:* Select a smaller time step* Use artificial snubber circuits

Introduce a small smoothing reactor for dc links* Introduce proper stray capacitances in the system modelProvide a parallel damping for lumped system

3.6 Simulation Errors and ControlErrors in a PE simulation can come through the followingchannels:*

control system simplificationtime step related truncation* program structure and solution method introducedinterfacing time delayincorrect system initial conditions

switching device approximation and system reductionadded circuit elements for numerical oscillation control

For application simulations, some errors resulting from thesystem simplification and measures of numerical oscillationcontrol are acceptable. The fourth and fifth items in the listcan be controlled by reducing the time step size. Arecommended time step size should not be greater than 1/5to 1/20 of the period of the highest concerned frequencycycle. For an example, for an IGBT inverter simulationwith 5000Hz PWM switching, a selected time step could be10 ps. However, if the objective of the simulation is to seethe detailed transient at the terminal of the induction motorwhich is fed by the inverter through a section of the cablewith an 1.0 ps travel time, an adequate time step should be0.2 ps or smaller.

* Configuration of modeling system, important systemcomponents and parametersAssumptions and specified accuracy requirementControl block diagram, detailed control functiondescription and complete listing of the controlconstants.

0 Table or figure summary of important simulationresults

The details of these examples are given in a separatedocument.4.1 Simulation of the Pulse Width Modulation (PWM)Voltage Source Inverter (VSI) Adjustable Speed DriveThe first example is a PWM-VSI ac drive simulation usingEMTP. The ac drive consisting of a three-phase diodebridge rectifier, capacitive dc link and three-phase PWMoutput inverter. The switching losses of the drive are asecondary order consideration in the analysis and the

*e Procedures of system simplification*

idealized switching characteristics are used.AC Supply DC Link InductionSystem Motor

IM

SixPulse PWMDiode Rectifier Inverter

Figure 5. Electrical Circuit Con figuration of anAdjustable Speed DriveErrors caused by incorrect system initial conditions can bereduced by just letting the simulation run for a period oftime to reach a corrected initial condition. This takes morecomputing time, but saves in model construction, especiallyif the program allowed to restart. There are some methodsdeveloped which help to accelerate the system into thecorrect initial condition quickly. The examples of usingsome of these methods can be found in the related referencelisted in this paper.4. SUMMARY OFDOCUMENTED EXAMPLESSummary of three simulation examples related with powerelectronics application are presented in this paper. Theexample can be either a real case study or exercise forillustration purposes. For each example, the followingimportant items are documented.* Objectives of simulation and major considerations

PWMVSI>VS480A-CNVRTA(Type 9)400, I I I

Figure 6. SimulatedAC Input Currentof A PWM-VSlAdjustable Speed D riveThe built-in diode models are used to construct the frontend rectifier. The same switching devices with addedouen/close controls are used to remesent outmt inverter


7/10

51 1IGBTs. The EMTP input data modules are used to buildthis example case. Both the output reference frequency andthe PWM carrier frequency are made to be controllable.Modeling of a signal processing and firing pulse generationis illustrated in this example. The motor load of the drive isrepresented by its R+ jX equivalent branch. The simulatedac input current, carrier and reference signal for the PWMcontrol, ac output voltage and current are presented inFigure 6 through Figure 8.

PWMVSI>TACS -DUMOS8(Type 9)I 5 , . . ,

- I S + : : . : I , : : ! i i : ! ' i i .40 43 50 55 EOTime (ms)

Figure 7. Simulated Ca rrier and Reference Signalsf o r APWM- VSI Adjustable Speed Drive Firing ControlPWMVWNVRTA-lNVRTE(1ype 8)800

Figure 8. Simulated AC Output Line-to-line Vo ltage of APWM-VSI Adjustable Speed Drive4.2 Simulation of Voltage Notching Caused by Operationof Current Source Inverter (CSI) ASDThe second example is based on a case study. The involvedsystem is illustrated by the one line diagram in Figure 9.The 25 kV distribution system is supplied through a 10MVA transformer from the 144 kV transmission system.The customer causing the voltage notching problems has a6000 hp induction motor supplied through a CSI adjustablespeed drive[2]. This drive is at a 4.16 kV bus suppliedthrough a 7.5 MVA transformer. Harmonic filters (5th,7th, 11th) are included to control the lower ordercharacteristic harmonics of the six pulse drive. The actualadjustable speed drive and motor load were represented to

reproduce the notching oscillations observed in themeasurements.Operation of the 6000 hp motor and drive resulted insignificant oscillations on the 25 kV supply system. Theseoscillations caused clocks to run fast at the customer withthe 6000 hp motor (clocks were fed separately from the 25kV system) and failure of surge capacitors on the 800 hpmotor at the customer located on the parallel feeder. Theobjective of the simulation is to identify the power qualityproblem associated with this 6000 HP, CSI ASD operation.The simulation was carried out using EMTP.The worst notching problems are associated with a firingangle at about 70% load. The simulated waveform for the25 kV bus voltage is shown in Figure 10. The oscillationsat each commutation point are in good agreement with themeasurement results.

Q ,hort Circuit Level74MVA44 kVI Lkm 266 MCM

10 km1IO

7.5 MVA& 4other4.16 kV loads

5.75%

S f S+rLdnic f ifilters 6000 haI 6.7.11Istherloads

I I1500 kVA 1500kVA4.7%

4.16 kV

Motor Load (160 kvar)(650 hp)surge capacltors 800 hp

Figure 9. One line diagram fo r the firs t example system.


8/10

512Simulated Voltage on 25 kV System - Base CoseJww , . 1I

lime (ms)

Figure 10.operating. Simulated 25 kV system voltage with drive

Figure 11 illustrates the voltage waveform at the 4.16 kVbus where the 800 hp motor surge capacitors causemagnification of the oscillations. The potential forproblems at this location is quite evident.

Voltaqe a i 4160 Voll Surge Copacilor - Ease Case 1

Figure 11 . Simulated waveform at surge capacitor location(4.16kV bus of cusfomer on parallelfeder)4.3 Simulationof HVDC Terminal an d Shunt TSCPTCRCompensationThe third example is an illustration case for modeling ofanHVDC system with shunt TSC/TCR compensation at theinverter bus [3]. The simulation example is made usingPSCADIEMTDC. The schematic system shown in Figure12is a modified version of the GIGRE Benchmark Modelfor HVDC Control Studies [4].

1000M w

Figure 12 . Study System

The inverter short circuit ratio has been reduced from itoriginal value of 2.5 to 1.5 to make the study moreinteresting. The dc link is a 1000MW, 500kV, 12 pulsemonopolar system. There are damped low and high passfilters at each converter terminal to reduce the distortiononthe ac bus. The control scheme for the HVDC systemconsists of a rectifier current controller with the gammacontroller.The SVC system is a -200/+300 MVAr, 12 pulse, TCRand TSC (two stage) combination connected to the inverterbus through a step up transformer. The SVC controls aredesigned to coordinate the control of TCR and TSC in sucha way that the combined susceptance of the SVC iscontinuous over its entire operating range. The basiccontrol mode is voltage control and has as a voltage droopbuilt into the controls. Several studies to evaluate therecovery to full power after a contingency were simulated[3]. The performances of several compensation optionswere compared. These options included fixed capacitors(FC), SVC, synchronous condenser (SC) and half-and-halfmix of the two (SVC+SC).

1.6 Ii

" 11 . 13m

Time (s )

Figure13 Inverter AC VoltageFollowing a Permanent DCBlockFigure 13 shows the typical results obtained from thesimulation for a permanent dc block. The fixedcompensator case does not control the overvoltage;however, all other options do. The SVC option is thefastest to respond followed by the SVC+SC option and


9/10

513M. Szechtman, T. Wess, C.V. Thio, First benchmarkmodel for HVDC control studies, Electra, No. 135,April 1991.L. Dube, H.W. Dommel, Simulation of controlsystem in an Electromagnetic Transient Program withTACS, IEEE Trans. on Power Industry and ComputerApplications, 1977EMTP Rule Book, EPRI/DCG Version 1.0.D.Goldsworthy, J.J. Vithayathil, EMTP model of anHVDC transmission system, Proceedings of the IEEEMontech 86 Conference on HVDC PowerTransmission, September 26-October 1, 1986, pp. 39-46L.X. Bui, S . Casoria, G. Morin, Modeling of digitalcontrols with EMTP, CEA Meeting, March 25-29,1989, Montreal, CanadaJ. Reeve and S.P. Chen, Versatile interactive digitalsimulator based on EMTP for ac/dc power systemtransient studies, IEEE Trans. on Power Apparatusand Systems, Vol. 103, No. 12, December 1984, pp.K.G. Fehrle, R. H. Lasseter, Simulation of controlsystems and application to HVDC converters, IEEETutorial Course 81 EHO173-PWR on DigitalSimulation of Electrical Transient Phenomena, 1981.L.X.Bui, G. Morin, J.Reeve, EMTP TACS-FORTRAN interface development for digital controlsmodeling, 91 SM 417-6 PWRSG. Morin, L. X. Bui, S. Casoria, J.Reeve, Modelingof the Hydro-Quebec - New England HVDC systemand digital controls with EMTP, IEEE Trans. onPower Delivery, Vol. 8, No. 2, April 1993, pp. 559-566.R. H. Lasseter and S. Y. Lee, Digital simulation ofstatic var system transients, IEEE Trans. on PowerApparatus and Systems, Vol. PAS-101, No. 10, pp.4171-4177, October 1982.A. M. Gole and V.k. Sood, A static compensatormodel for use with electronmagnetic ransientssimulation programs, IEEE Trans. on PowerDelivery, Vol. PWRSJ, No. 3, pp. 1398-1407, July1990A.N. Vasconcelos et. al. Detailed modeling of anactual static Var compensator for electromagnetictransients studies, IEEE Trans. on Power Systems,Vol. PWRS-7, no. 1, pp. 11-19, February 1992S.Y. Lee et al., Detailed modeling of static Varcompensators using the Electromagnetic TransientsProgram (EMTP), IEEE Trans. on Power Delivery,Vol. 7, no. 2, pp. 836-847, April 1992S. Lefebvre and L. Gerin-Lajoie, A staticcompensator model for the EMTP, IEEE PESMeeting, San Diego, July 28-August 1, 1991, Paper 91

3625-3633

SM 46 1-4 PWRS.

lastly the SC option. This simulation setup can be used toconduct almost any type of performance study including athyristor miss-fire in HVDC valve group or in the SVCitself.5. CONCLUSIONSThe appropriate characterization of the power electronics isvery important in power system simulations involvingpower electronics operations. In most these simulations,detailed representations of the power electronics are notnecessary. Depending on the objective of a study, theinvolved power electronics sub-system can be alwaysreduced to some extent with minimal loss of accuracy.Numbers of the digital computation tools are capable ofsimulating power electronics. However, in power systemsengineering community, the EMTP type of programs aremore commonly used. This results mainly from the greatcapabilities and flexibility of these programs in handlingconventional power system dynamics and electro-magnetictransients beside their capabilities of handling powerelectronics. With an adequate power electronics device andcircuit simplification, the EMTP type of programs arepowerful for modeling various types of power electronicsapplications. Because these programs are based on the timedomain solution method, the dynamic interaction betweenthe power electronics and the rest of the system can beeasily incorporated in simulation.The important considerations for simulating powerelectronics applications have been summarized in thisguidelines. Three modeling examples using the EMTPtype of programs were presented. The procedures used toimplement power electronics models in these examples arevaluable for using other digital simulation tools.6. REFERENCES AND BIBLIOGRAPHIC1.

2 .

3.

N. Mohan, W.P. Robbins, T.M. Underland, R.Nilsen, 0. MO, Simulation of power electronics andmotion control1 systems - an overview, Proceedings ofthe IEEE, Vol. 82, No. 8, August 1994, pp. 1287-1302.L.Tang, M.F. McGranaghan, R.A. Ferraro, S.Morganson, b. Hunt, Voltage notching interactioncaused by large adjustable speed drives on distributionsystems with low short circuit capacities, IEEE PESO.B.Nayak, A.M. Gole, D.G. Chapman, and J.B.Davies, Dynamic performance of static andsynchronous compensators at an HVDC inverter bus ina very weak ac system, IEEE Trans. on PowerSystems, Vol. 9, NO. 3, August 1994, pp. 1350-1358.

95 SM 388-9-PWRD.

~

4.

5.

6.7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.


10/10

51 418. L. Dube and I. Bonfanti, MODELS: A new

simulation tool in the EMTP , European Transactionson Electrical Power Engineering, Vol. 2, no. 1, pp.45-50, JanuarylFebruary 1992.

19. Leuven EMTP Center (ed.), ATP Rule Book, 1990.20. J.A. Martinez, Simulation of a microprocessor-

controlled SVC, 21th European EMTP Meeting, June5-7, 1992, Crete (Greece).21. H.W. Dommel, EMTP Reference-Manual (EMTPTheory Book), BPA, 1986.

22. J. A. Martinez, Simulation of power electronics usingthe EMTP, Part I: Power converters, A survey,UPEC94, September 14-16, 1994, Galway.

23. G.A. Capolino, H. Henao, ATP simulaiton for powerelectronics and AC drives, 15th European EMTPUsers Group Meeting, Paper 88R-027, October 17-18,1988, Leuven.

24. G.A. Capolino, H. Henao, Simulation of electricalmachine drives with EMTP, 18th European EMTPUsers Group Meeting, Paper M7, May 28-29, 1990,Marseilleartinez, G.A. Capolino, TACS andLS:Drive simulation languages in a generalpurpose program, Proc. MCED91, Marseille, July 1-2, 1991, pp. R1-R13.

26. G.A. Capolino, H. Henao, ATP advanced usage forelectrical drives, EMTP Summer Course, July 5-8,1993, Leuven.

27. H. Knudsen, Extended Parks transformation for 2 by3-phase synchronous machine and converter phasormodel with representation of harmonics, IEEE PESSummer Meeting, Paper 94 SM 350-9 EC, July 24-28,1994, San Francisco.AC electrical drives based on field-oriented controlmethod using a general purpose program, ProceedingsPCIM, pp. 350-364, 1986, Munchen

29. Z. Daboussi, N. Mohan, Digital simulation of field-oriented control of induction motor drives usingEMTP, IEEE Trans. on Energy Conversion, Vol. 3 ,pp. 667-673, September 1988.

30. L. Tang, Mark McGranaghan, Modeling an activepower line conditioner for compensation of switchingtransients, Proceedings of First InternationalConference on Power Systems Transients (IPST95),Lisbon (Portugal), pp. 403-408.

31. X . Z . Meng, J.G. J. Sloot, H. Rijanto, Modelling ofsemiconductor fuses in EMTP, Proceedings of FirstIntenational Conferenceon Power Systems Transients(IPST95), Lisbon (Portugal), pp. 481-486.Capolino, Advanced representation of powersemiconductors using the EMTP, Proceedings of First

28. M. Mazzucchelli, G. Sciutto, Digital simulation of

32. J.A. Martinez-Velasco,R. Abdo, G.A.

International Conferenceon Power Systems Transients(IPST95), Lisbon (Portugal), pp. 505-510.

33. Serge Lefebvre, Ricardo D. Range1,Modeling ofpower electronics devices in EMTP-TACS,Proceedings of First International Conferenceon PowerSystems Transients (IPST95), Lisbon (Portugal), pp.511-516.

34. G.A. Capolino, H. Henao, D. Leduc, V.T. NguyenPhuoc, CAD of field-oriented induction motor drivesusing a general purpose program, Proceedings PCIM,1989, Munchen.

35. E. Bassily, G.A. Capolino, H. Henao, Simulation ofdiscrete DC drive using sliding mode control, 22ndEruopean EMTP Users Group Meeting, Paper 92R-018, November 9-10, 1992, Leuven.

36. E. Bassily, G.A. Capolino, H. Henao, Simulation anddesign of brushless motor drive control with fuzzy piregulator, Proceedings of First European Conferenceon Power Systems Transients (EPST93), Lisbon(Portugal), pp 84-91, June 1993.

37. A.K. Khan, R. Dwyer, M . Mcgranaghan, L. Tang,Evaluation of Harmonic Impacts from CompactFluorescent Lights on Distribution Systems, IEEEPES 95 WM 105-7-PWRS.

38. J.A. Martinez, EMTP simulation of digitally-controlled static Var system for optimal loadcompensation, IEEE PES Summer Meeting, Paper 94SM 452-3 PWRD, July 24-28, 1994, San Francisco.

39. J.A. Martinez and G.A. Capolino, EMTP simulationof power electronics and drives using datamodularization, Proceedings First InternationalAegean Conferenceon Electrical Machines and PowerElectronics, Kucadasi (Turkey), May 27-29, 1992.Swe, 400 MW SMES power conditioning systemdevelopment and simulation, IEEE Trans. on PowerElectronics, Vol. 8, No. 3, July 1993, pp. 237-249.

40. Ibrahim D. Hassan, Richard M. Bucci, and Khin T.

guidelines for modeling power electronics in electric power engineering applications

Documents