Solar cell junction temperature measurement of PV module

(1)

Solar cell junction temperature measurement of PV module

B.J. Huang

a,⇑

, P.E. Yang

a

, Y.P. Lin

a

, B.Y. Lin

a

, H.J. Chen

b

, R.C. Lai

b

, J.S. Cheng

b a_{New Energy Center, Department of Mechanical Engineering, National Taiwan University, Taipei, Taiwan}

b_{Photovoltaic Technology Center, Industrial Technology Research Institute, Hsinchu, Taiwan}

Received 22 July 2010; received in revised form 3 November 2010 Available online 18 December 2010

Communicated by: Associate Editor Manuel Romero-Alvarez

Abstract

The present study develops a simple non-destructive method to measure the solar cell junction temperature of PV module. The PV module was put in the environmental chamber with precise temperature control to keep the solar PV module as well as the cell junction in thermal equilibrium with the chamber. The open-circuit voltage of PV module Vocis then measured using a short pulse of solar irradi-ation provided by a solar simulator. Repeating the measurements at diﬀerent environment temperature (40–80°C) and solar irradiation S (200–1000 W/m2), the correlation between the open-circuit voltage Voc, the junction temperature Tj, and solar irradiation S is derived. The fundamental correlation of the PV module is utilized for on-site monitoring of solar cell junction temperature using the measured Vocand S at a short time instant with open circuit. The junction temperature Tjis then determined using the measured S and Vocthrough the fundamental correlation. The outdoor test results show that the junction temperature measured using the present method, Tjo, is more accurate. The maximum error using the average surface temperature Taveas the junction temperature is 4.8°C underestimation; while the maximum error using the present method is 1.3°C underestimation.

Keywords: Solar cell junction temperature; Solar cell measurement; Solar PV

1. Introduction

Solar cell junction temperature can seriously aﬀect the power output of solar PV module (King et al., 1997; Nord-mann and Clavadetscher, 2003). Hence, an accurate solar cell junction temperature measurement is important in order to correctly assess the temperature eﬀect of a solar PV module.

The solar cell junction temperature depends on the packaging of solar cells and the environmental factors such as ambient temperature and wind speed/direction. There are some methods to measure the solar cell junction tem-perature. Usually, the junction temperature Tjis taken as

the average value of the surface temperatures of the bottom side Tbotand the top side Ttop, i.e.

Tj¼ Tave ¼ ðTtopþ TbotÞ=2 ð1Þ

Due to the thermal resistance of the packaging materials and the interior temperature gradient, the junction temper-ature measured in this way may be inaccurate. Some inves-tigators use numerical method and a simple energy balance to predict the junction temperature (Mattei et al., 2006).

The other method as adopted in IEC904-5 Standard (IEC Standard 904-5, 1993) (IEC method) uses the pre-deter-mined Voc1at known S1and Tj1to determine the equivalent

cell temperature (ECT) from the measured S2and Voc2using

the average temperature coeﬃcient of Voc, CT, and the given

diode thermal voltage D for a given ns(number of series

con-nection of PV cells). The measurement is however tedious. In the present study, we attempt to develop a simple non-destructive method to measure the solar cell junction temperature which can be used in on-site monitoring of solar PV performance.

⇑ Corresponding author.

E-mail address:bjhuang@seed.net.tw(B.J. Huang).

www.elsevier.com/locate/solener

(2)

2. Measurement of solar cell junction temperature

It is known that the open-circuit voltage of solar cell Voc

decreases with increasing junction temperature Tj. Voc is

also aﬀected by solar irradiation S. That is, Voc follows

the function of Eq.(1).

Voc¼ f ðS; TjÞ ð2Þ

Measuring Vocand solar irradiation S allows us to

deter-mine Tj, if the basic relation, Eq. (2), is pre-determined

experimentally.

The relations of Vocvs. Tjat various solar irradiations S

can be determined in an environmental chamber which keeps the solar PV module in thermal equilibrium within a temperature-controlled chamber. In an on-site application, the solar cell junction temperature can be determined by a suddenly disconnection of the solar PV module for a short period of time in order for measuring the Voc. The junction

temperature Tj can then be converted from the measured

Vocand solar irradiation S using the pre-determined

rela-tion. The measurement is simple, non-destructive and can be very accurate.

3. Experimental setup

Fig. 1shows the experimental apparatus setup to deter-mine Eq. (2) of a PV module. We used a solar simulator

(Burger PSS30) to provide artificial light incident upon the PV module. The open-circuit voltage of the PV module is measured at the time when the solar simulator flashes the light in a short period of time, 10 ms. This flash time is too short to cause temperature rise of solar cell during mea-surement. The environmental chamber is designed with an electric heater connected to a PID (proportional–inte-gral–derivative) controller to control the heating power so as to control the chamber at a fixed temperature.

The design of the environmental chamber is as shown in

Fig. 2which was made of aluminum. The chamber is ther-mally insulated and uses a glass cover to allow solar light to pass through only. The PV module is ﬁxed inside the cham-ber. Inside the chamber a fan was used for air circulation to create a uniform temperature over the PV module.

Table 1shows the characteristic of the solar PV module used in this experiment. The data ofTable 1are taken sep-arately from the standard test results of I–V curve of the solar PV module. During the heating process, the ature uniformity inside the chamber will aﬀect the temper-ature uniformity of the PV module and may lead to experimental errors. The ﬁrst test is to measure the temper-ature distribution of the PV module surfaces as well as the solar cell. Twenty-seven T-type thermocouples were mounted evenly on the top and bottom surfaces of PV module and inserted between the glasses and the solar cell, Nomenclature

Aoc intercept coeﬃcient of Voc, V

CT temperature coeﬃcient of Voc, V/°C

S solar radiation intensity incident upon the PV module, W m2

Tbot surface temperature on the back of solar PV

module,°C

Ttop surface temperature on the top of solar PV

mod-ule,°C

Tj solar cell junction temperature,°C

Tjd real solar cell junction temperature measured by

direct method,°C

Tjo solar cell junction temperature measured by the

present method,°C

Tave the average temperature of top and bottom

sur-faces of PV module,°C

Voc open-circuit voltage of solar PV module, V

Solar simulator

chamber

PID controller _{PV module}

(3)

for a specially-made PV module. Three thermocouples were installed in each location (total nine locations): top surface, cell surface (between the glasses), and bottom sur-face. The surface temperature non-uniformity of the PV module is less than 1°C. This experiment can also deter-mine the temperature response of solar PV module for reaching a steady state between measurements at diﬀerent operating conditions. The PV temperature uniformity test uses a specially-made PV module with thermocouples inserted onto the solar cell surface as well as outside sur-faces to check the design of the environmental chamber. After this temperature uniformity test, a real PV module is then put into this chamber for regular experiment.

The chamber temperature is ﬁrst set at 40°C with incre-ment 10°C for successive experiments. The PID controller regulates the chamber temperature as well as the PV mod-ule surface temperature. The test results indicate that the top surface temperature response of the PV module to the new setting of the chamber temperature is about 20 min to reach a steady state and the surface temperature uniformity of PV module is satisfactory, within ±1°C.

It is found that the cell temperature takes about 30 min to obtain thermal equilibrium in each temperature setting of the chamber. The test results indicate that any change of the operating condition during test requires at least 30 min to reach a thermal equilibrium.

4. Results and discussion

4.1. Determination of fundamental correlation of PV module After the PV module in thermal equilibrium with the chamber, which takes about 30 min for each temperature

setting, the open-circuit voltage of the PV module Vocis

measured when the solar simulator is turned on in a short period of time, 10 ms. This time period is too short to cause a signiﬁcant temperature rise of solar cell during measure-ment. Vocwas measured repeatedly every 10°C increment

from 40°C to 80 °C for diﬀerent S. The experimental result ofFig. 3shows that at ﬁxed S, Vocdecreases linearly with

increasing Tjwhich satisﬁes the relation of Eq.(2):

Voc ¼ Aoc CTTj ð3Þ

where Aocis the intercept coeﬃcient of Vocat Tj= 0°C and

CTis the temperature coeﬃcient of Voc. A relation between

Aoc and S, Eq. (4), can be derived from the determined

glass

insulation

fan PV module

Electric heater support

Solar radiation

Fig. 2. Environmental chamber design.

Table 1

Single-crystalline solar PV module characteristics.

Eﬃciency (%) 14.50–14.74

Maximum-power output (W) 2.29 Maximum-power current (A) 4.6 Short circuit current Isc(A) 4.377

Maximum-power voltage (V) 0.497 Open-circuit voltage Voc(V) 0.598 0.4 0.45 0.5 0.55 0.6 30 40 50 60 70 80 90 Voc (Volt) Junction temperature Tj(oC) S=1000 W/m2 800 W/m2 600 W/m2 400 W/m2 200 W/m2

Fig. 3. Variation of Vocwith Tj.

Table 2

Experimental determination of the relation of Voc.

Irradiation (W/m2₎

Linear relation Intercept coeﬃcient Aoc(V) Temperature coeﬃcient CT (V/°C) 1000 Voc=0.002160Tj+ 0.6526 0.6526 0.002160 R2_{= 0.9999} 800 Voc=0.002200Tj+ 0.6476 0.6476 0.002200 R2= 0.9993 600 Voc=0.002210Tj+ 0.6394 0.6394 0.002210 R2= 0.9994 400 Voc=0.002240Tj+ 0.6322 0.6322 0.002240 R2= 0.9998 200 Voc=0.002330Tj+ 0.6108 0.6108 0.002330 R2= 0.9998 Aoc= 0.0256 ln(S) + 0.4762 0.6 0.61 0.62 0.63 0.64 0.65 0.66 0 200 400 600 800 1000 1200 intercept Aoc (Volt) Solar irradiation S (W/m2₎ Fig. 4. Variation of Aocwith S.

(4)

parameters of Table 2. Fig. 4 shows the correlation be-tween Aocand S.

AocðSÞ ¼ 0:0256 ln S þ 0:4762 ð4Þ

Fig. 5shows the correlation between CTand S. A

rela-tion of CT with S, Eq (5), can be derived from the

deter-mined parameters ofTable 2.

CT ¼ 0:000188 lnðSÞ þ 0:003525 ð5Þ

Combining Eqs. (3)–(5), we can obtain the functional relation of Eq.(6).

VocðS; TjÞ ¼ ð0:0256 ln S þ 0:4762Þ

ð0:000188 ln S þ 0:003525ÞTj ð6Þ

Eq.(6) is the fundamental correlation of Voc, Tj, and S

which can be utilized in the determination of solar cell junction temperature in the ﬁeld operation of PV module. 4.2. On-site measurement of solar cell junction temperature Eq. (6) is the fundamental correlation of a PV module which can be utilized in on-site solar cell junction temper-ature monitoring if Vocand S are measured in a short time

instant at open circuit. The same PV module used in the present study was installed outdoor and run with a con-stant load (resistor) to simulate the performance of a solar PV power generation. In order to determine the PV junc-tion temperature Tj, a circuit was designed to disconnect

the load (open-circuit of PV module) for a very short per-iod of time, 10 ms in the present study, for measuring Voc.

The measuring circuit is shown in Fig. 6. Instantaneous junction temperature is then calculated from Eq. (6)using the measured solar irradiation S and Voc, which is denoted

Tjo. For monitoring purpose, Tjo was measured every

3 min. The energy loss due to the disconnection of power generation circuit during Tj measurement is only 10 ms/

180 s = 0.05%, which is negligible.

To verify the accuracy of the above measured results Tjo,

a thin T-type thermocouple was installed into the PV junc-tion during the module packaging in order to directly mea-sure the junction temperature Tjdand compare with Tjo.

CT = -0.000188 ln(S) + 0.003525 0.0022 0.00225 0.0023 0.00235 0.0024 0.00245 0.0025 0.00255 0.0026 0 200 400 600 800 1000 1200 Temperature Coefficient CT , V/ oC Solar irradiation S (W/m2₎ Fig. 5. Variation of CTwith S .

Fig. 6. Electronic circuit for measuring Voc.

30 40 50 60 70 0 200 400 600 800 1000 1200 10:00:05 10:24:13 10:48:15 11:12:18 11:36:18 12:00:11 12:24:15 12:48:24 13:12:24 13:34:03 13:57:45 Tem p erature (ºC) Irradiance (W/m 2) WindSpeed (m/s) Time (2008/09/19) Irradiation S Tjd(direct measurement) Tjo(monitored) Tave Wind Speed (x100)

Fig. 7. Outdoor test result of PV module. Table 3

Measurement errors of diﬀerent methods. Real junction temperature Tjd(°C) Tave_(°C) Error (°C) Max error 65.9 61.1 4.8 Min error 60.3 59.8 0.5 Real junction temperature Tjd(°C) Present method Tjo(°C) Error (°C) Max error 57.9 56.6 1.3 Min error 62.7 62.7 0.03

(5)

Fig. 7 shows the outdoor test results of the same PV module used in the experimental determination of the fun-damental correlation. The solar cell junction temperature Tjwas determined using the present method and the

aver-age of the surface temperatures at the bottom side Tbotand

the top side Ttop, Eq.(1). It is seen that the junction

temper-ature determined using the present method, Tjo, is more

accurate.Table 3shows the maximum and minimum mea-surement errors between Tjoand Tave. It is found that the

maximum error using the conventional Taveas the junction

temperature is 4.8°C underestimation; while the maximum error using the present method is 1.3°C underestimation. 5. Conclusions

The present study develops a simple non-destructive method to measure the solar cell junction temperature of PV module. Indoor experiment using a simple environmen-tal chamber and a solar simulator is required in order to determine the fundamental correlation of a PV module, i.e. Voc, Tjand S. The PV module has to be put in the

envi-ronmental chamber with precise temperature control to keep the solar PV module as well as the cell junction in thermal equilibrium with the chamber.

The junction temperature of the PV module in the chamber is not directly measured. But it is assumed to coin-cide with the chamber temperature itself, within ±1°C (accuracy of the temperature controller), if thermal equilib-rium has been reached. Many tests have shown that it will take about 30 min for the PV module to reach thermal equilibrium with the chamber. The measurement of the chamber temperature T, S and Vocare done after having

imposed the conditions that guarantee a good thermal equilibrium for the test PV module.

The open-circuit voltage of PV module Vocis then

mea-sured using a short pulse of solar irradiation provided by a solar simulator. Repeating the measurements at diﬀerent environment temperature (40–80°C) and solar irradiation S (200–1000 W/m2), the fundamental correlation between the open-circuit voltage Voc, the junction temperature Tj,

and solar irradiation S is derived.

The fundamental correlation of the PV module can be utilized for on-site monitoring of solar cell junction temper-ature if Vocand S are measured simultaneously at a short

time instant with open circuit. In order to measure Voc,

we designed a circuit to disconnect the load (open-circuit of PV module) for a very short period of time, 10 ms, for measuring Voc. The junction temperature Tjis then

deter-mined using the measured S and Voc through the

funda-mental correlation.

The outdoor test results of the same PV module used in the experimental determination of the fundamental correla-tion show that the junccorrela-tion temperature measured using the present method, Tjo, is more accurate. The maximum

error using the average surface temperature Tave as the

junction temperature is 4.8°C underestimation; while the maximum error using the present method is 1.3°C underestimation.

The present method of measuring the junction tempera-ture of PV module on-site can be accurate if the PV module used in the experimental determination of the fundamental correlation of the PV module is the same or is similar enough to the PV module used on-site. However, it is highly recommended that the sampling of PV modules used in field installation for the experimental determination of the fundamental correlation is a necessary step in order to obtain an accurate result in measuring the junction tem-perature on-site. The design of the new environmental chamber to fit the different sizes of PV modules as shown in Fig. 2is not difficult. The method of the experimental determination of the fundamental correlation of the PV module is exactly the same as described previously (Fig. 1). Acknowledgements

This publication is based in part on work supported by Photovoltaic Technology Center, ITRI, Taiwan, and Award No. KUK-C1-014-12, made by King Abdullah University of Science and Technology (KAUST), Saudi Arabia.

References

IEC, International Standard 904-5:1993 (ﬁrst ed.), Photovoltaic Devices – Part 5: Determination of the Equivalent Cell Temperature (ECT) of Photovoltaic (PV) Devices by the Open-Circuit Voltage Method, pp. 1–11.

King, D.L., Kratochvil, J.A., Boyson, W. E., 1997. Temperature coeﬃ-cients for PV modules and arrays: measurement methods, diﬃculties, and results. In: 26th PVSC. Anaheim, CA, USA, September 30– October 3, 1997. IEEE 0-7803-3767-0/97.

Mattei, M., Notton, G., Cristofari, C., Muselli, M., Poggi, P., 2006. Calculation of the polycrystalline PV module temperature using a simple method of energy balance. Renewable Energy 31, 553–567. Nordmann, T., Clavadetscher, L., 2003. Understanding temperature

eﬀects on PV system performance. In: 3rd World Conference on Photovoltaic Energy Conversion, Osaka, Japan, May 11–18, 2003.