Weighted efficiency measurement of PV inverters: introducing...

5
JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS Vol. 15, No. 5- 6, May – June 2013, p. 550 - 554 Weighted efficiency measurement of PV inverters: introducing η İZMİR İLKER ONGUN * , ENGIN ÖZDEMIR a Ege University Ege Meslek Yuksekokulu, 35100 Bornova İzmir, Turkey a Kocaeli University, Faculty of Technology, Department of Energy Systems Engineering, 41380 Umuttepe Kocaeli. Conversion efficiency of DC/AC inverters depends on some parameters and fluctuates over the input power of the inverter. Since the PV inverters operate under a fluctuating input power supplied by the PV modules, conversion efficiency must be measured against the weights of the probable power ranges which represent the various irradiation values. This approach of having different weights for different irradiation ranges resulted in two basic weighted conversion efficiency models of η EURO and η CEC . These two models consider the irradiation distribution over the whole annual sunny time and prioritize the ranges with various weight factors. Since the irradiation profiles vary around the planet, inverter efficiencies must be evaluated against local irradiation profiles to get more precise annual energy yield estimation. This paper presents η İZMİR , a weighted conversion efficiency evaluation model, derived from the İzmir irradiation profile. This model has been developed in a way that it should be simple and accurate so it has been matched with other models for its estimation capabilities. The results are discussed here and suggestions being made. (Received December 18, 2012; accepted June 12, 2013) Keywords: PV inverters, conversion efficiency, weighted efficiency, η EURO , η CEC , η İZMİR 1. Introduction Penetration of grid connected photovoltaic power systems (PVPS) is rapidly increasing for two decades. As they are being an embedded part of the electric networks, their electricity generations have started to be studied more intensively. The incentives given by the governments for those systems due to the paradigm change in the energy field has triggered a further acceleration in those studies. Electricity generation of a PV power system depends on the solar irradiation received by the PV modules and the efficiency of the system. The efficiency of a PVPS on the other hand, is a multifold concept covering conversion efficiency of the PV modules along with the conversion efficiency, MPPT performance and some other properties of PV inverter used. PV inverters are evaluated with their overall efficiency. Overall efficiency is described as the ratio of the energy delivered by the PV inverter at the AC terminals to the energy provided by the PV array [1]. The two efficiencies involved in the inverters are conversion efficiency (η conv ) and MPPT efficiency (η MPPT ) described as, (1) and (2) respectively, where, ݐሻ· ݐis instantaneous value of the delivered power at the AC terminal, ݐሻ· ݐis instantaneous value of the power drawn by the inverter, p ሺtሻ · dt is instantaneous value of the MPP power provided by the PV array (or PV simulator). Thus, the overall efficiency including both efficiencies becomes; (3) Besides the straightforward mathematics here, especially with the grid connected power plants, annual yield estimation needs further effects to be studied accurately for realistic revenue projections. One these would be the input power fluctuations resulting from climatic conditions, since the value of PV array power explicitly affects both conversion and MPP tracking efficiencies of the PV inverter. The first weighted efficiency calculation concerning the effect of irradiation profile on the inverter efficiency has been introduced with north-western Germany climate data (Trier) in 1991 [2], [3]. The formula given in a footnote of a magazine article then became a well-known comparison tool among PV inverters Although the weather data used for calculating the weighting factors do not represent whole Europe – especially the South – the formula know is now known as the “European Efficiency – [4]. European Efficiency formula is given as; ሺ4ሻ 1 1 2 2 3 3 4 4 5 5 6 6 EURO EU MPP EU MPP EU MPP EU MPP EU MPP EU MPP a a a a a a η η η η η η η = + + + + + 0 0 () () M M T AC conv T DC p t dt p t dt η = 0 0 () () M M T DC MPPT T MPP p t dt p t dt η = 0 0 () () M M T AC T MPP p t dt t conv MPPT p t dt η η η = =

Transcript of Weighted efficiency measurement of PV inverters: introducing...

Page 1: Weighted efficiency measurement of PV inverters: introducing …laboratuar.kocaeli.edu.tr/gucelektronik/sci/... · 2013. 11. 14. · JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS

JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS Vol. 15, No. 5- 6, May – June 2013, p. 550 - 554

Weighted efficiency measurement of PV inverters: introducing ηİZMİR

İLKER ONGUN*, ENGIN ÖZDEMIRa Ege University Ege Meslek Yuksekokulu, 35100 Bornova İzmir, Turkey a Kocaeli University, Faculty of Technology, Department of Energy Systems Engineering, 41380 Umuttepe Kocaeli.

Conversion efficiency of DC/AC inverters depends on some parameters and fluctuates over the input power of the inverter. Since the PV inverters operate under a fluctuating input power supplied by the PV modules, conversion efficiency must be measured against the weights of the probable power ranges which represent the various irradiation values. This approach of having different weights for different irradiation ranges resulted in two basic weighted conversion efficiency models of ηEURO and ηCEC. These two models consider the irradiation distribution over the whole annual sunny time and prioritize the ranges with various weight factors. Since the irradiation profiles vary around the planet, inverter efficiencies must be evaluated against local irradiation profiles to get more precise annual energy yield estimation. This paper presents ηİZMİR, a weighted conversion efficiency evaluation model, derived from the İzmir irradiation profile. This model has been developed in a way that it should be simple and accurate so it has been matched with other models for its estimation capabilities. The results are discussed here and suggestions being made. (Received December 18, 2012; accepted June 12, 2013) Keywords: PV inverters, conversion efficiency, weighted efficiency, ηEURO, ηCEC, ηİZMİR

1. Introduction Penetration of grid connected photovoltaic power

systems (PVPS) is rapidly increasing for two decades. As they are being an embedded part of the electric networks, their electricity generations have started to be studied more intensively. The incentives given by the governments for those systems due to the paradigm change in the energy field has triggered a further acceleration in those studies.

Electricity generation of a PV power system depends on the solar irradiation received by the PV modules and the efficiency of the system. The efficiency of a PVPS on the other hand, is a multifold concept covering conversion efficiency of the PV modules along with the conversion efficiency, MPPT performance and some other properties of PV inverter used.

PV inverters are evaluated with their overall efficiency. Overall efficiency is described as the ratio of the energy delivered by the PV inverter at the AC terminals to the energy provided by the PV array [1]. The two efficiencies involved in the inverters are conversion efficiency (ηconv) and MPPT efficiency (ηMPPT) described as,

(1) and

(2) respectively, where,

· is instantaneous value of the delivered power at the AC terminal,

· is instantaneous value of the power drawn by the inverter, p t · dt is instantaneous value of the MPP power provided by the PV array (or PV simulator). Thus, the overall efficiency including both efficiencies becomes;

(3)

Besides the straightforward mathematics here,

especially with the grid connected power plants, annual yield estimation needs further effects to be studied accurately for realistic revenue projections. One these would be the input power fluctuations resulting from climatic conditions, since the value of PV array power explicitly affects both conversion and MPP tracking efficiencies of the PV inverter.

The first weighted efficiency calculation concerning the effect of irradiation profile on the inverter efficiency has been introduced with north-western Germany climate data (Trier) in 1991 [2], [3]. The formula given in a footnote of a magazine article then became a well-known comparison tool among PV inverters Although the weather data used for calculating the weighting factors do not represent whole Europe – especially the South – the formula know is now known as the “European Efficiency – [4]. European Efficiency formula is given as;

4 1 1 2 2 3 3

4 4 5 5 6 6

EURO EU MPP EU MPP EU MPP

EU MPP EU MPP EU MPP

a a aa a a

η η η ηη η η

= ⋅ + ⋅ + ⋅

+ ⋅ + ⋅ + ⋅

0

0

( )

( )

M

M

T

AC

conv T

DC

p t dt

p t dtη

⋅=

0

0

( )

( )

M

M

T

DC

MPPT T

MPP

p t dt

p t dtη

⋅=

0

0

( )

( )

M

M

T

AC

T

MPP

p t dtt conv MPPT

p t dtη η η

⋅= ⋅ =

Page 2: Weighted efficiency measurement of PV inverters: introducing …laboratuar.kocaeli.edu.tr/gucelektronik/sci/... · 2013. 11. 14. · JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS

wMt

hfWicC

eeaii

ts

elrapmIIBims

where isMPPT efficienthese values ar

The variahighly emphasfor the geograpWith the incnstallations ha

comparison toCommission u

Weig

Pa

PM

Weig

Pa

PM

The quickevaluates highemphasizes machieve more nverter, than nverter.

Evaluationthat, local clstrategies affec

2. ETA İz Since the

effective on ocations coul

represent Turkalmost in the process, the measurementsIn some assInstitute’s meBasic measuren DMİ Men

minute resolushown in Fig 1

W

s the weightinncy at partialre listed in Taations in thesized in this fphical locatio

creasing peneave been mad

ool has been inutilizing the Sa

ghting Factor

artial MPP power

MPP, PVS/PDC,r

ghting Factor

artial MPP power

MPP, PVS/PDC,r

k comparisonher irradiation

more on the loprecise annua maximum

n of weightelimatic condicts inverters’

zmir

effect of climefficiency,

ld be done. Hkey (lat. E42ºmiddle (lat. EState Meteo

s database wasessments, Eeasurement daement set is tanemen Obserution. Annual1.

Weighted effici

ng factor and l MPP power

able:1. e lower irradformula, makins with highe

etration of Pde in southern ntroduced by acramento clim

Table

0,03

0,05

Table

0,04

0,1

n of two tablen region in mower part. Boual yield esti

m conversion

ed efficiency itions (latitudenergy efficie

matic conditiofurther resea

Here, İzmir haº-E36º), for it E38º30’). Duriorology Direas used as primEge Universiata base has aken during 2rvation Statiol distribution

iency measurem

is the sr. The indices

diation levelsing it less suiter solar irradiaVPSs, and mlocations, anoCalifornia Enmatic data [5]

1. Weighted eff

0,06

0,1

2: Weighted eff

0,05

0,2

es show that Cmore detail and

th can be useimation for aefficiency of

approach shde) and tracencies [6].

ns is known tarch for varas been chose

geographicaling the evalua

ectorate’s climmary data souiy Solar En

also been u2009 through 2on no. 1778

of these dat

ment of PV inve 

static s for

s are table

ation. more other nergy ].

ηCECwas

who

fficiency formula

0,13

0,2

fficiency formula

0,12

0,3

CEC d EE ed to a PV f that

hows cking

to be rious en to lly is ation mate urce.

nergy used. 2012 9 at ta is

accotempmod

andcurr

wheTPV

OCV

SCI

erters: introduci

Using same C formula cons suggested as

ose coefficient

a coefficients fo

0,1

0,3

a coefficients fo

0,21

0,50

Fig. 1. 2009 irrM

Energy calcuording to EN

mperature depdule has been

irradiance anrent of a cSi m

ere,

1

4

CEC CEC

CEC

aa

ηη

= ⋅

+ ⋅

(, 1OC STCV β= ⋅ +

,SC STCSTC

GIG

= ⋅

ing ηİZMİR

structure witnsidering the h;

ts are shown i

or ηEuro.

0,4

0,5

or ηCEC.

1 0,5

0 0,7

radiation and tMenemen observ

ulations have 50530. In th

endent open calculated as,

nd temperatumodule has bee

is PV modu

1 2

4 5

MPP CEC

MPP CEC

aa

ηη

+ ⋅

+

( ))PV STCT Tβ ⋅ − ⋅

(1 ( PVC

Tα⋅ + ⋅ −

th the Europehigher irradia

in Table 2.

8 0,

5 1

3 0,0

5 1

temperature datvation station.

been made fohis document,

circuit volt,

ure dependent en calculated

ule temperatur

2

5

MPP CEC

MPP CE

aa

ηη⋅ +

⋅ +

ln 1G

G CC

⎛ ⎛ ⎞+ ⋅⎜ ⎜ ⎟⎜ ⎝ ⎠⎝

))STCT−

55

ean Efficiencytion condition

(5

,2

1

05

1

ta from DMİ

or cSi moduleirradiance an

tage of a cS

6

t closed circuas,

7

re ºC;

3 3

6 6

C MPP

EC MPP

ηη⋅

V RC C G⎞

− ⋅ ⎟⎟⎠

51

y, ns

5)

es nd Si

6

uit

7

Page 3: Weighted efficiency measurement of PV inverters: introducing …laboratuar.kocaeli.edu.tr/gucelektronik/sci/... · 2013. 11. 14. · JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS

552 İlker Ongun, Engin Özdemir  

TSTC is standard test condition temperature 25 C°; G is solar irradiation W/m2; α is temperature coefficient of current 0,25%/ ºC; β is temperature coefficient of voltage 0,4%/ ºC; and technology depending correction factors CR, is 2,514 · 10 W/m2 CV, is 8,593 · 10 CG is 1,088 · 10 m2/W.

For simplification reasons, a 1 000 W PV array with Uoc,STC=100 V and Isc,STC=10 A has been considered.

The irradiation data has first been evaluated for annual energy distribution against irradiation classes. Results are presented in Fig:2. A quick inspection of the graph reveals

that, one third of annual energy yield would be harvested at and below 500 W/m2 irradiation levels. The other one thirds would be harvested between 500-750 W/m2 and above 750 W/m2 irradiation classes respectively.

This clearly shows that yield estimations made based on European efficiency wouldn’t be valid for İzmir irradiation since it assumes 79% of annual yield would be harvested at and below 500 W/m2 irradiation levels (referring Table:1).

CEC efficiency on the other hand, shows a closer match with İzmir irradiation profile at lower levels since it assumes 42% energy yield for that range (referring Table:1). However, this model still don’t show a proper match for medium and high irradiation levels for it assumes 95% of annual yield would be harvested below 750 W/m2 irradiation levels, which is not the case with İzmir data.

Fig. 2. 2009 irradiation profile evaluation for annual energy yield distribution.

The reason for European efficiency formula is not grasping İzmir energy yield can be its being based on hourly irradiation averages [4] instead of a high resolution irradiation measurement data set.

The failure of CEC efficiency may be resulted from its lack of considering temperature effect significant enough.

Further inspection of İzmir irradiation data and the energy yield calculations gives the weights in the Table:3 for 10% irradiation classes, values in the PMPP/PSTC row is representing the midpoints of these bins.

Table 3. Weighted efficiency formula coefficients for ηEuro.

PMPP/PSTC

(%) 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

IZM (ALL) (%) 1,31 2,83 4,02 4,90 5,72 6,80 8,24 9,12 9,49 11,45 14,41 13,35 7,11 1,03 0,18 0,03 0,01

Cumulative energy yield 1,31 4,14 8,16 13,06 18,78 25,58 33,82 42,94 52,43 63,88 78,29 91,64 98,78 98,78 99,96 99,99 100

0,00

1,00

2,00

3,00

4,00

5,00

6,00

7,00

8,00

9,00

10,00

0,00

0,00

0,01

0,10

1,00

10,00

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

850

900

950

1000

1050

1100

1150

1200

1250

1300

Energy (%

)

Irradiation Time (%

)

Irradiation Classes (W/m2)

Annual Energy Distribution by Irradiation Classes (2009 ‐Menemen)

Energy %

Page 4: Weighted efficiency measurement of PV inverters: introducing …laboratuar.kocaeli.edu.tr/gucelektronik/sci/... · 2013. 11. 14. · JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS

cPm1

dSw

manTİa

a

bnmtb

One of th

calculations hPMPP/PSTC ratimodule tempe1 000 Wp cS

The valuedistribution ofSacramento. Iwould have hi

Since themuch of the yand CEC effnecessary for Thus, a reducİzmir efficiencand their respe

Table 5: Re

PMPP/PST

IZM (5) IZM (4)

These val

and

These sug

been tested agnumber of commeasures at ththese evaluatibelow.

(5)IZMη =

(4)IZMη =

W

he most impohere is, thereio, due to terature. This i PV array c

PM

EURCEIZMIZM

es in the tablef a PVPS in İzInverter efficiigher weights e lower generyield, the resofficiencies at r hot climateced number ocy formula. Fective weights

educed number efficien

TC 10% 302,3% 13,4% 12

ues in the tabl

ggested evalugainst Europemmercially avhe AIT Inveron graphs are

10%

70%

0,02

0,28

η

η

= ⋅ +

+ ⋅ +

10%

70%

0,04

0,53

η

η

= ⋅ +

+ ⋅

Weighted effici

ortant result e is no instathe diminishimeans that th

could produce

Table 4: We

MPP/PSTC 5RO 3C

M (EURO) M (CEC)

e reveal that, zmir is differeiency values fin İzmir cond

ration classes lution shown lower parts

s with high of weightings

Five and four s are shown in

of evaluation cncy formula.

0% 50% ,1% 21,8% 2% 21%

le yields,

uation formulean and CEC vailable PV invrter Test Labe shown Fig.

30%

90%

0,13

0,35

η

η

+ ⋅ +

+ ⋅

30%0,12 η+ ⋅ +

iency measurem

derived fromance above ng effect ofhere is no w

e 1 000 W po

eights for İzmir

5% 10% 153% 6%

4% 1% 6%

2%

energy generaent than in Trifor higher po

ditions. do not repreboth in Euro does not ssolar irradia

s are proposeevaluation cla

n Table 5.

classes for İzmir

70% 90%27,6 35,2%53% -

las have togeefficiencies f

verters’ efficiboratory. Thre

3 through Fi

50%0,22 η+ ⋅

50%0,21 η+ ⋅

ment of PV inve 

m the 90%

f the way a

ower

throman

EuroTab

data calculated

5% 20% 2513% 5%

10% 7%

ation er or wers

esent pean seem ation. ed as asses

r

8

9

ether for a ency

ee of ig. 5

erters: introduci

oughout the ynners.

A comparisoopean and C

ble 4.

d for Euro and

5% 30% 5010% 412% 216% 212% 1

Fig. 3. Comp

Fig. 4. Compa

ing ηİZMİR

ear in İzmir,

on of İzmir CEC efficien

CEC classes.

0% 75% 108% 1% 53%9% 9% 26%

arison for a wid

rison for a rela

unless it is c

data based cncies are pre

00% 20% 5%

39% 34%

de voltage rang

atively low volta

55

ooled by som

alculations foesented in th

ge inverter.

age inverter.

53

me

or he

Page 5: Weighted efficiency measurement of PV inverters: introducing …laboratuar.kocaeli.edu.tr/gucelektronik/sci/... · 2013. 11. 14. · JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS

5

s

ciTd

ensTfd

crac

554

Fig. 5. Com 3. Concl The rem

summarized bThere is

climatic condinstalled in İz

This brings in data set to ach

Although evaluation of Pnot suitable fsince it is baThis inadequformula whicdifferent weig

Here withcould be madreliability. Besalso simpler consider.

mparison for a r

usion

markable conelow: a distinctive

itions, MPP pzmir would no

a need for furhieve a more re

it provides PV inverters, for yield esti

ased on a noruacy has beech makes usehting pattern.

h the İzmir effde with localsides the increthan the oth

relatively high v

nclusions of

e indication power generatot exceed 85%rther researcheliable result. a good basisEuropean eff

imations in srthern climateen fixed by e of the sam

ficiency formulized data seteased reliabilithers with 33

İlker

voltage inverter

this study

that, due toted by a PV a% of STC po

h using a multi

s for performficiency formusouthern locate data evalua

CEC efficime approach

ula, the evaluat in an enhaty, new formu

3% less term

r Ongun, Engin 

r.

are

o the array ower. iyear

mance ula is tions

ation. ency with

ation anced ula is

ms to

Inst“DiEU

doctConInveInst

[1] T

[2] R

[3] A

[4] B

[5] F

[6] V

____*Co

n Özdemir

Acknowledg The authors

titute of Techstributed EneProject no. 22The content

torate thesis nversion Efficerters”, studititute, İzmir, T

References

TS EN 50530 connected phR. Hotopp; „PStromerzeugu Auflage, RWAuf den SpureS. 62 – 65 B. Berger et.a the Basis of O CEC EfficienFinal Guidebo8th ed., DecemEnergy CommV. Salas et.al.and EuropeanWorld Congrepp 1590-1594

____________rresponding a

gements

of this papehnology suppoergy Resource28449”. ts of this p“Determinati

ciency for Tied in Ege

Turkey.

s

0, 2010, Overahotovoltaic InvPrivate Photovungsanlagen im

WE Energie AGen von „Euro-

al., “Are We BOutdated Defincy?”, ook, Emergingmber 2006; pumission; CEC-, “Analysis B

n Efficiency”, ess 2007 (Vol4

____________author: ilker.on

er thank to Aorting this reses Research I

paper are takion of Turke

Total Energy University

all efficiency overters. voltaik- m NetzparalleG, Essen, 199-Eta“, Photon

Benchmarkinginitions of the

g Renewablesublished by the-300-2006-00

Between EnergProceedings o

l. I – Vol. V),

__ [email protected]

AIT – Austriasearch througInfrastructure

ken from they’s Weighte

Yield of PVSolar Energ

of grid

elbetrieb“, 2. 1 , June 2004,

g Inverters on e European and

s Program, e California 1-ED8F-CMF

gy Efficiency of ISES 2009,

du.tr

an gh s,

he ed V

gy

d

F