7/28/2019 10.1.1.196.6809 (1).pdf

1/4

IMPACT OF LARGE PHOTOVOLTAIC PENETRATION ON THE QUALITY OF SUPPLYA CASE STUDY AT A PHOTOVOLTAIC NOISE BARRIER IN AUSTRIA

B. Bletterie; M. HeidenreichArsenal research, Business Unit Renewable Energy Technologies

Faradaygasse 3, A-1030 Vienna, AustriaPhone: +43(0)50 550-6355, Fax: +43(0)50 550-6390

E-mail: benoit.bletterie@arsenal.ac.at

ABSTRACT: With the steadily increasing penetration of distributed generation (DG) in electric power systems,concerns have been raised that reliable and safe electricity supply might be jeopardized by a further deployment of decentralized power plants. In this context, Power Quality (PQ) and more generally the Quality of Supply (QoS) arekey issues which have recently gained increased attention. QoS has been the subject of many studies during the lastdecade: customers as well as network operators are well aware that secure and efficient operation of power networksand customers equipment is intimately connected with a high QoS level.A consortium of research centers, universities, utilities and manufacturers is currently working on this topic withinthe framework of the European project DGFACTS. In this context, a power quality measurement campaign has beenlaunched with the aim to assess the impacts of various DG technologies such as PV plants, wind farms and small

hydro power plants.This paper gives an overview of the results obtained from PQ measurements performed at a 101 kW photovoltaicnoise barrier in Gleisdorf, Austria. Harmonics, flicker, unbalance and RMS events are characterized and discussed.Furthermore, a short discussion on the role played by the decoupling protection is provided.Keywords: Distributed generation, Power quality, Grid-connected PV-systems

1 INTRODUCTION

As one of the most essential bases for contemporarylife, electricity supply and, in particular, the quality of supply has gained increasing attention in the past years.

Nowadays referred to as a product with a definedquality, electricity is rather a unique product because of

its intangible and transient nature. With the liberalizationof electricity markets and the steadily increasing

penetration of Distributed Generation from RenewableEnergy Sources (RES) driven by favorableenvironmental policies and regulatory frameworks inEuropean countries, concerns have been raised regardingthe impacts on the quality of supply. Various studies[1,2,3] led to the conclusion that compatibility limitshave, in some cases, already been exceeded. Closeattention must be paid to ensure a high level of thequality of supply in order to guarantee the successful andsecure integration of distributed generation (DG) into theelectrical power system.

In this context, arsenal research is currently

investigating this issue in the framework of the European project DGFACTS (Improvement of the quality of supply in Distributed Generation networks through theintegrated application of power electronic techniques),

part of the European Research Project Cluster "Integration of RES + DG". The objective of this projectis to solve the series of power quality problems arisingfrom the integration of distributed generation into electricdistribution networks by developing new devices basedon the experience of the FACTS (Flexible AlternativeCurrent Transmission Systems). Among other contributions, arsenal research has initiated a power quality measurement campaign for photovoltaicinstallations in Austria.

The 101 kW photovoltaic noise barrier situated inGleisdorf (Austria), alongside the A2 motorway betweenVienna and Graz, was selected for power qualitymeasurements because of its relatively high capacitycompared to the strength of the network and because

many customers are connected at the same PCC. In theframework of the measurement campaign, the assessmentof the power quality at this plant was conducted duringMay and June 2003. Voltage variations, voltage dips andswells, flicker, harmonics, unbalance, etc. were measuredand analyzed according to the standard EN 50160 [4].Furthermore the network configuration was investigated

in order to conduct a comprehensive analysis of themeasurements.

2 THE EUROPEAN PROJECT DGFACTS

The 3-year project which started in January 2003 iscarried out by a consortium of 12 partners from 7European countries. The goal of this project is to developmodular systems (the so-called DGFACTS) aimed atimproving the quality of supply in order to allow a higher degree of RES+DG penetration into the current andevolving distribution systems. Fig. 1 illustrates theapplication of the DGFACTS devices in distribution

networks.

Fig. 1: Project DGFACTS improvement of quality of supply for DG networks

7/28/2019 10.1.1.196.6809 (1).pdf

2/4

In order to provide the necessary inputs in terms of specifications and requirements for the development of the DGFACTS prototypes, the following preparativetasks were carried out:- Analysis of the existing and on-going normative and

legislative information regarding quality of supply (seereport of the DGFACTS project [5])- Power quality assessment at selected locations bymeans of in-situ measurements- Simulation of the impact of the interconnection of DG units to the distribution network regarding power quality

3 INSTALLATION SELECTED FOR THE CASESTUDY

The photovoltaic noise barrier Gleisdorf was selected because its capacity is relatively high compared to the

short-circuit power of the network at the point of common coupling (about 3%). The ratio between therated power of the generator and the networks short-circuit power gives an idea of the size of the generator with respect to the network strength, and provides arough gauge whether or not power quality problems may

be expected. The photovoltaic generator consists of about1 800 modules (1/3 of the installed power multi-crystalline and 2/3 amorphous silicon modules) whichare integrated into the 1 300 m length of the noise barrier.They are connected to 55 string inverters (single-phasetransformer-less type) which feed the electricity into thelow-voltage network. The annual yield of the installationas committed by the plant operator amounts to 86 000

kWh.

4 ASSESSMENT OF THE QUALITY OF SUPPLYBY MEANS OF IN-SITU MEASUREMENTS

The quality of supply was assessed over a period of more than 5 weeks using a power quality analyzer connected to a GSM modem for remote access to thedata. The measurements were analyzed on a weekly basisaccording to the standard EN 50160; the conclusionswere then presented in a report issued as a deliverable of the project. Fig. 2 gives a simple description of themeasurement system.

Fig. 2: Setup of the power quality measurements at the101 kW PV noise barrier

5 RESULTS OF THE POWER QUALITYASSESSMENT

In the following subsections, the results of theanalysis of the power quality parameters (RMS events,

harmonics, flicker and unbalance) are presented in detail.

5.1 RMS events (voltage events)RMS events were identified as the major source of

concern at this PV installation. At the beginning of themeasurement campaign, an extremely large number of voltage dips, voltage swells and short interruptions wererecorded at the installations bus bars: in total, more than1 000 RMS events were recorded during the first week.Fig. 3 shows the measured voltage profile for a sunnyday. The voltage (blue line) and current (red line) plotsare 10-minutes-minimum and 10-minutes-mean valuesrespectively.

Fig. 3: Generated current and terminal voltage for RMSevents during the peak generation (sunny day)

Fig. 3 clearly shows that the RMS events occurredduring times of generation peaks, which indicates acorrelation between the PQ problems and the PV plantoperation; an explanation is provided later.

The RMS events which occurred during the firstweek are classified in Table 1 according to their magnitude (residual voltage) and duration.

Table 1: RMS events UNIPEDE Table for phase L1,first week

7/28/2019 10.1.1.196.6809 (1).pdf

3/4

The information provided in Table 1 can be presented on the so-called magnitude-duration scatter plot (also called CBEMA diagram): see Fig. 4. Thisallows having an estimate of the events severity. Eachindividual point represents a RMS event whereas the two

envelope curves show typical sensitivity curves of ITequipment.

Fig. 4: CBEMA magnitude-duration scatter plot, phaseL1, first week

Most of the events which appear on this diagram arewithin the sensitivity zone which means that they can beclassified as severe: equipment subjected to suchdisturbances would usually trip.

After discussions with the network and plantoperator, the occurrence of these numerous events wasexplained by a too tight adjustment of the over-voltage

protection. The voltage variations/ fluctuations resultingfrom the current variations generated by the PV plantcombined with natural voltage variations have resultedin trips of the over-voltage protection. Although thevoltage variations were not significant, the plant wasrepeatedly disconnected and reconnected from thenetwork because of a too tight adjustment of thedecoupling protection.

After a change of this adjustment, the number of RMS events dropped back to reasonable numbers.

5.2 Harmonic distortionA careful analysis of the measurement data led to the

conclusion that the PV plant does not seem to have any

impact on the harmonic content of the line voltage. Thisconclusion could be derived from the fact that thedistortion level observed during nighttime, during sunnydays with large PV generation and cloudy days, iscomparable. The distribution of the THD values over thewhole monitoring period is depicted on Fig. 5. Table 2

provides the characteristics of this distribution.

Fig. 5: Voltage THD distribution (phase L1)

Table 2: Voltage THD (phase L1)10-minutes MIN value 1,3 %

10-minutes MAX value 4,3 %10 minutes MEAN value 2,7 %Standard deviation 0,4 %95% value of the 10-minutes values 3,5 %

The harmonic distortion level at the PCC iscomparable to levels which can be typically found ondistribution networks (with a predominance of the 5 th and7th harmonics).

5.3 Flicker Similarly, the operation of the PV plant did not have

any noticeable impact on the flicker level, except for those time intervals where RMS events occurred. Thisillustrates the importance to flag measurement intervalswhich contain RMS events; otherwise events would becounted twice. This issue is addressed in the standardIEC 61000-4-30 [6].

5.4 UnbalanceThe voltage unbalance (defined in [4] as the ratio

between negative and positive sequence) is rather small:the 95% value is about 0.3%. The fact that the PV power is not equally distributed over the three phases does nothave any noticeable influence on the voltage unbalance.Even though the current unbalance reaches levels of about 8% during the generation peak, the voltageunbalance level does not show any significant increase.

The results for harmonic distortion, flicker andvoltage unbalance are summarized in Table 3.

Table 3: Voltage harmonics, flicker and voltageunbalance summary(three phases, whole monitoring period)

95%-values EN 50160valuesVoltage distortionTHD U

4,8% 8%

Flicker Plt

0,6 1

Voltage unbalanceUU

0,3% 2%

7/28/2019 10.1.1.196.6809 (1).pdf

4/4

6 CONCLUSIONS

The following conclusions regarding the impact onthe quality of supply at this PV noise barrier can bedrawn from the measurement campaign:

- Harmonic distortion, flicker and voltage unbalancelevels are comparable to those which can be expected ontypical LV distribution networks without DG. Asignificant negative impact of the DG unit could not beobserved during the monitoring period.- However, some power quality problems wereidentified: numerous RMS events (voltage dips/swellsand short interruptions) occurred during a part of themonitoring period.- This abnormal occurrence of RMS events wasidentified to be the result of an interaction between thegenerator itself and the plants decoupling protection,caused by a too tight adjustment of the over-voltage

protection function.

The main conclusion derived from this study iscertainly that the decoupling protection, as a corecomponent of every DG installation, deserves particular attention. Too sensitive settings or inappropriate designcan not only lead to problems regarding the reliableoperation of the plant but furthermore also be a source of

power quality disturbances in the grid.In the further case studies carried out in the

framework of the project, spurious trips due to voltagedisturbances have been reported (e.g. at wind power andcogeneration sites). Such spurious trips can negativelyaffect the plant performance, components lifetime and

even the network operation at a larger scale (voltagestability).

Great attention has to be paid to this cross-cuttingissue.

7 REFERENCES

[1] Eurelectric, Power Quality in European ElectricitySupply Networks 1 st edition, Feb. 2002[2] Eurelectric, Power Quality in European ElectricitySupply Networks 2 nd edition, Nov. 2003[3] L. Berthet, D. Boudou, X. Mamo, P. Eyrolles, J.Martinon, EDF, State of play of the harmonic levels on

the French low-voltage networks, CIRED 2003[4] EN 50160:1999 Voltage characteristics of electricitysupplied by public distribution systems[5] B. Bletterie, M. Heidenreich, Evaluation of thequality of supply requirements specified by existingstandards, national legislation and relevant technicalreports inside and outside EU (available at www.http://dgfacts.labein.es/dgfacts)[6] IEC 61000-4-30 (2003:02): Power qualitymeasurement methods

8 ACKNOWLEDGEMENTS

The project DGFACTS is partially funded by theEuropean Commission, DG ResearchDuration: 01.01.2003-31.12.2005

Project no.ENK5-CT-2002-00658

The authors are solely responsible for this publication, itdoes not represent the opinion of the EuropeanCommunity and the European Community is notresponsible for any use that might be made of dataappearing therein.