文档库

最新最全的文档下载
当前位置:文档库 > A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV Dispersed Generator:A Power Quality Analysis

Within the IEEE519

Alejandro R.Oliva and Juan Carlos Balda,Senior Member,IEEE

Abstract—The use of environmentally clean photovoltaic(PV) dispersed generation will become more widespread in the future due to anticipated cost reductions in PV technology.This paper summarizes the results of a power quality(PQ)study performed on a PV generator in order to estimate the effects that inverter-in-terfaced PV dispersed generation might have upon the quality of electric power.Different interpretations of the harmonic distortion limits set in the IEEE519-1992standard are performed together with a comparison with the BC Hydro’s harmonic current limits. This paper also includes a statistical analysis of all measurements recorded with the help of two PQ monitors,an evaluation of the results from a connection/disconnection test,and harmonic simu-lation results.

Index Terms—Dispersed generation,harmonic distortion,IEEE standards,IEEE519-1992,power quality,photovoltaic power gen-eration.

I.I NTRODUCTION

T WO100-kW photovoltaic generating systems(called here the Solar Park)were installed by Central and South West (CSW)in the Ft.Davis,TX area(West Texas)as part of the com-pany program on renewable energy sources.Each photovoltaic (PV)system is interfaced to the distribution system through a pulse-width-modulated(PWM)inverter.The associated injec-tion of harmonic currents into the distribution system may cause malfunction of harmonic-sensitive equipment if the injection of harmonic currents is allowed to reach excessive levels. Therefore,the main objective of our work was to establish if the200-kW Solar Park might cause any degradation of the quality of the electric power supplied by West Texas Utilities (WTU),the affiliate of CSW in the Ft.Davis area.This paper illustrates the application of the IEEE519-1992Standard and the BC Hydro harmonic current limits to the PQ measurements recorded along the distribution feeder associated with the Solar Park.In addition,a statistical analysis of all measurements,re-sults from a connection/disconnection test,and harmonic simu-lation results are presented to complete the PQ study.

II.S TUDY M ETHODOLOGY

In order to accomplish the objectives set forth in this project, it was necessary to[1]

Manuscript received June5,2000.This work was supported in part by EPRI and CSW under Research Contract RP3797-04.

A.R.Oliva is with the Department of Electrical Engineering,Uni-versidad Nacional del Sur,Bahía Blanca,BA8000,Argentina,(e-mail: Aoliva@http://www.wendangku.net/doc/358a88d1b14e852458fb575b.html.ar).

J.C.Balda is with the Department of Electrical Engineering,University of Arkansas,Fayetteville,AR72701USA(e-mail:jcb@http://www.wendangku.net/doc/358a88d1b14e852458fb575b.html).

Digital Object Identifier10.1109/TPWRD.2003.809685a)Model the Ft.Davis distribution system[2];

b)Select the buses to be monitored:a worst-case analysis was

performed to identify those buses with the largest voltage total harmonic distortions

A PV dispersed generator_ a power quality analysis within the IEEE 519

(THDs)

1/h rule[2],

A PV dispersed generator_ a power quality analysis within the IEEE 519

H three-phase ac

choke,a

A PV dispersed generator_ a power quality analysis within the IEEE 519

(ungrounded)480-12470-V500-kV A step-up

three-phase transformer bank(which serves both PV systems).

The impedances of the isolation and step-up transformers are

2.83%and2.5%of their respective bases.

The PWM inverter is an Omnion?series3200converter spe-

cially designed for PV power conversion applications[8].

Table I displays the most important components of a typ-

ical current waveform at the Solar Park.The PWM switching

in combination with the ac choke and the at the low-voltage side of the

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

500-kV A step-up transformer http://www.wendangku.net/doc/358a88d1b14e852458fb575b.htmling the recorded

data,it was established that the average peak apparent power at

the Solar Park is approximately140kV A,yielding a peak av-

erage demand

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

OLIV A AND BALDA:A PV DISPERSED GENERATOR:A POWER QUALITY ANALYSIS WITHIN THE IEEE 519527

TABLE II

IEEE-519H ARMONIC C URRENT D ISTORTION L IMITS IN P ERCENTAGE FOR THE S OLAR P ARK C ONSIDERED AS D ISPERSED G ENERATION (AND [ISC =IL]>50

)

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

TABLE III

IEEE-519H ARMONIC C URRENT D ISTORTION L IMITS IN P ERCENTAGE FOR THE S OLAR P ARK C ONSIDERED AS U TILITY -D ISTRIBUTION

G ENERATING E

QUIPMENT

A PV dispersed generator_ a power quality analysis within the IEEE 519

TABLE IV

A PV dispersed generator_ a power quality analysis within the IEEE 519

IEEE-519H ARMONIC C URRENT D ISTORTION L IMITS IN P ERCENTAGE FOR THE

S OLAR P ARK C ONSIDERED AS A C

USTOMER

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

in Tables II–IV [2].Fig.1illustrates a scatter plot of all current total demand distortions (TDDs)where the data points were plotted versus the hour of the day to correlate with the time dependency.The figure also shows in a solid line,an average curve.An analysis of the current TDD reveals that the peak and the peak of the average TDD values are approximately 8%and 6%,respectively,satisfying the IEEE-51912%TDD limit for a customer.However,the average current TDD measured at the Solar Park was more than the 3.75%and 5%IEEE-519limits recommended for dispersed generation and generation equipment,respectively.

According to IEEE 519,the steady-state harmonic limits can be exceeded by 50%for short periods of time (up to one hour per day or approximately 4%of the time).This limitation is consistent with having the harmonic limits not exceeded 95%of the time (i.e.,the 95%probability point).Fig.2shows the probability distribution of the TDD corresponding to the same data in Fig.1.It can be noted that the 5%limit is not exceeded 72%of the time while 98%of the time,the TDD does not ex-ceed 7%.Furthermore,one can mention that the 7.5%limit (i.e.,

1.5

A PV dispersed generator_ a power quality analysis within the IEEE 519

(12470/480));we have

multiplied the 20-A limit set by BC Hydro for the 69-kV supply system by the transformer ratio.

Fig.4displays the average harmonic current injection versus the generated apparent power and its standard deviation (abso-lute values);this information describes the harmonic

character-

Fig.1.Scatter plots of the current TDDs at the Solar Park corresponding to the period from June 16to November 2,

1995.

Fig.2.Probability distribution of the TDD measured at the Solar

Park.

Fig.3.Probability distribution of the THC measured at the Solar Park.

istics of these PWM inverters for PV systems and can be really meaningful in future work.

When the Solar Park is considered as either dispersed gen-eration or generating equipment,almost every single harmonic component is above the IEEE-519limits shown in Tables II and III,respectively [2].

An analysis of the current harmonic components when the Solar Park generated 142kV A showed that odd harmonics were within the customer IEEE-519limits but even harmonics between the 18th and 48th (except for 34th)exceeded their

528IEEE TRANSACTIONS ON POWER DELIVERY ,VOL.18,NO.2,APRIL

2003

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

Fig.4.Harmonic current injection versus apparent power.

TABLE V

C URRENT H ARMONIC

D ISTORTIONS AT TH

E S OLAR P ARK FOR THE

E VEN -H ARMONIC C OMPONENTS B ETWEEN THE 18TH AND 48TH C

OMPONENTS

A PV dispersed generator_ a power quality analysis within the IEEE 519

A PV dispersed generator_ a power quality analysis within the IEEE 519

IEEE-519limits.Table V illustrates the even harmonics for this range.

In summary,the Solar Park considered,as a customer would be within the IEEE-519TDD limits,but out of range if consid-ered as both dispersed generation and generating equipment.

VI.V OLTAGE W A VEFORM PQ C HARACTERISTICS

The voltage PQ characteristics were analyzed using all data recorded at the three monitored sites during the complete moni-toring period.An important figure of merit is the statistical

dis-

Fig.5.Cumulative histogram of the voltage THD at the three monitored

sites.

Fig.6.V oltage THD versus time of the day at the Observatory.

tribution of the voltage THD,as shown in Fig.5.The voltage THDs at the substation and the Solar Park were within the 5%IEEE-519limit,while the voltage THD at the Observatory was often at the 5%limit.

Most maximum values of the voltage THD recorded at the Observatory occurred during the day.However,Fig.6shows that only a small increase in the average THD value occurred when the Solar Park was generating.It must be noted that most generating periods were normally coincident with the times when most customer loads would be operating.Furthermore,a “background”level for the voltage THD when the Solar Park was not generating is observed in Fig.6due to other harmonics

OLIV A AND BALDA:A PV DISPERSED GENERATOR:A POWER QUALITY ANALYSIS WITHIN THE IEEE519

A PV dispersed generator_ a power quality analysis within the IEEE 519

529

Fig.7.Average individual voltage harmonic distortions at the Ft.Davis substation,the Solar Park,and the Observatory.

sources.Fig.7illustrates the most important harmonic compo-

nents of the voltage THD at the Ft.Davis substation,the Solar

Park,and the Observatory.

The harmonic distortion(HD%)at the Observatory for the

largest third,fifth,and seventh harmonics versus the time of the

day were plotted in Fig.8for all voltage measurements[2].The

third,fifth,and seventh HD%for the time of the peak voltage

THD(see Fig.7)are1.8%,3.2%,and0.7%,respectively.The

IEEE519states a maximum of3%for the individual voltage

harmonic distortions;thus,the fifth harmonic component was

slightly over that limit.Harmonics simulations with the Solar

Park injecting the measured average maximum demand current

(see Table I)gave voltage THDs under0.4%and an H5under

0.14%along the feeder(with the maximum value occurring at

the Observatory)[2].Harmonic injections from other customers

and/or the large transfer impedance close to the parallel reso-

nances at the fourth and seventh harmonics might have been the

cause for these slightly high harmonic voltages at the fifth har-

monic component.

A number of researchers have proposed different approaches

for the calculation of power in electric circuits with distortion.

Some of these methods are discussed in[9]which concludes

that the method in[10]“accurately describes the rating of power

compensation equipment that ought to be used in practice and is

a good resolution method to indicate the severity of distortion in

practical power systems.”The nonfundamental distortion power

is the distortion power to be compensated[10].Fig.9shows

the normalized nonfundamental distortion power for the Solar

Park;an average curve is also plotted.It can be noted that the

distortion power is low at average generating powers.

VII.C ONNECTION/D ISCONNECTION T EST

The inverters of the Solar Park were switched off on March

26,1996,at approximately6P.M.in order to record the THD

“background”level during daylight hours without the influence

of the Solar Park.On March27,1996,only one of the Solar

Park PV systems was switched on and off several times after12 P.M.to analyze the effects of the connection/disconnection of the Solar Park around the time of the greatest insolation.The other

PV system was disconnected during the test.

The H5measured at the substation was still high(an average

of1.25%)when the Solar Park was disconnected(i.e.,during the

first day).The H5then rose up to2%during the second

A PV dispersed generator_ a power quality analysis within the IEEE 519

night.Fig.8.Individual voltage harmonic distortions at the

A PV dispersed generator_ a power quality analysis within the IEEE 519

Observatory.

Fig.9.Normalized nonfundamental distortion power at the Solar Park. Harmonic simulations were performed assuming a back-ground voltage distortion at the substation bus,corresponding to a voltage THD of2.8%(with harmonic components as those measured when the Solar Park was disconnected).The simulation results showed a4%voltage THD and a4%H5 at the Observatory[2].These results support that the Solar Park did not have an important contribution to the recorded fifth-harmonic voltage at the Observatory since its harmonic

530IEEE TRANSACTIONS ON POWER DELIVERY,VOL.18,NO.2,APRIL2003

current injections were below the“customer”IEEE-519limits up to the18th harmonic.It is not within the scope of this paper to determine the source of this fifth harmonic not originating at the Solar Park.

VIII.D ISCUSSIONS

A complete analysis of all the recorded and simulation data yielded the following results[2]:

?the current injected by the Solar Park had a TDD below 12%(an IEEE519-1992limit that should be accomplished as a customer)for all recorded waveforms;

?however,the Solar Park harmonic current injection when considered as dispersed generation or generating equip-ment is above their respective IEEE-519limits;

?an analysis of the individual harmonic components re-vealed that only the even-harmonic current components between the18th and the48th(except for the34th)ex-ceeded the IEEE-519limits(when considered as a cus-tomer).Unfortunately,[5]does not provide the reasoning for the25%limits whose justification is beyond the scope of this paper.It is our opinion that this even-harmonic cur-rent injection should not cause any PQ problem due to the small energy associated with them;

?like the Solar Park,the Observatory also showed exces-sive values for the even-harmonic components of the phase currents for the same harmonic range;

?the5%limit for the voltage THD was often encountered at the Observatory;

?the individual3%fifth-harmonic voltage HD limit was exceeded at the substation and the Observatory.From an analysis of a connection/disconnection test of the Solar Park and simulation results,it was concluded that the Solar Park contribution to the H5harmonic distortion was minimal;

?an average“background”voltage THD of approximately 3%was noticed at the Observatory,but it could not be attributed to the Solar Park,because it was also registered during the night when the Solar Park did not generate;

?neither of the PQ monitors(i.e.,the one at the substation and that one at the Solar Park)recorded no voltage tran-sient during the connection/disconnection test due to the soft-start control of the PWM inverters[8];

?the THC at the Solar Park is below BC Hydro ampere limit.

In summary,the Solar Park should not produce any PQ problem in the Ft.Davis distribution system,regardless of the IEEE-519limits for dispersed generation or generating equipment.This statement is presently supported by the lack of customer complaints.Furthermore,a customer with an equivalent loading(the only difference would be the power flow)would be treated differently.Why is this required if both are interfaced to the distribution system through power electronics?It is obviously not a technical matter.The answer might be an economical consideration rather than a technical issue.

Nevertheless,it is true that a better inverter technology would produce a considerably lower TDD.Although the Solar Park in-verters have a relatively low switching frequency of6kHz,the low order harmonics should not be present unless they are asso-ciated with over modulation,asynchronous PWM,or blanking time.The TDD problem of the Solar Park could be probably cor-rected with an appropriate calibration of the PWM controller or further filtering.

Moreover,the authors’opinion is that the treatment of dis-persed generation in the IEEE519should be revised considering the future deregulation of the electric market in the U.S.

A CKNOWLEDGMENT

The authors appreciate the system data and technical support provided by Mr.S.Baker and Mr.B.Champion from WTU.

R EFERENCES

[1]“Electrotek Concepts,Power Quality Assessment Procedure EPRI Final

Rep.,”,CU-7529,1991.

[2]J.C.Balda and A.Oliva,“The Impact of Dispersed Generation Upon the

Quality of Electric Power:The Solar Park in the Ft.Davis Distribution System EPRI Final Rep.,”,TR-107725,1997.

[3]J.Arrillaga, D.Bradley,and P.Bodger,Power System Har-

monics.New Y ork:Wiley,1989.

[4]Electrotek Concepts,PQNode Application and System Software,1993.

[5]IEEE Standard519-1992IEEE Recommended Practices and Require-

ments for Harmonic Control in Electrical Power Systems,1992.

[6]W.Xu,Y.Mansour,C.Siggers,and M.B.Hughes,“Developing utility

harmonic regulation based on IEEE Std519-B.C.Hydro’s approach,”

IEEE Trans.Power Delivery,vol.10,pp.1423–1431,June1995.

[7]Electrotek Concepts HarmFlo+User’s Guide,1992.

[8]Omnion Power Engineering,Omnion Series3200Converter—User

Manual.

[9]J.H.C.Pretorius,J.D.Van Wyk,and P.H.Swart,“An evaluation

of some alternative methods of power resolution in a large industrial plant,”in Proc.8th Int.Conf.Harmonics and Quality of Power,Athens, Greece,Oct.14–16,1998,IEEE/PES and NTUA.0-7803-5105-3/98, pp.99–106.

[10]“Practical definitions for powers in systems with nonsinusoidal wave-

forms and unbalanced loads:A discussion,”IEEE Trans.Power De-livery,vol.11,pp.79–101,Jan.1996.

Alejandro R.Oliva received the B.S.degree in electrical engineering from the Universidad Nacional del Sur(UNS),Bahia Blanca,Argentina,in1987,and the M.S.E.E.degree from the University of Arkansas at Fayetteville(UAF),AR, USA,in1996.He is currently pursuing the Ph.D.degreee at the UAF. Currently,he is an Assistant Professor at the UNS.From1987to1988,he was with Hidronor S.A.,developing a data base for modeling large hydraulic plants.Since1988he has been with the UNS..His main research interests are power electronics and digital signal processor(DSP)-based control.

Juan Carlos Balda(S’77–M’90–SM’94)received the B.S.E.E.degree from the Universidad Nacional del Sur,Bahía Blanca,Argentina,in1979,and the Ph.D. degree from the University of Natal,Durban,South Africa,in1986. Currently,he is Associate Professor at the University of Arkansas,Fayet-teville,where he has been since1989.His main research interests are electric power distribution systems,electric power quality,motor drives and power elec-tronics,and digital control of electric machines.

He is a senior member of the honor society Eta Kappa Nu and is also a coun-selor of the IEEE Student branch at the University of Arkansas.