IMPROVED STABILITY OF CIGS-BASED THIN-FILM PV MODULES
Katsumi Kushiya, Satoru Kuriyagawa, Kenichi Tazawa, Tadashi Okazawa and Masayuki Tsunoda Showa Shell Sekiyu K.K., CIGS Project Team
(Atsugi) 123-1, Shimo-kawairi, Atsugi, Kanagawa, Japan
ABSTRACT The object of this study is to bring useful and effective data to establish the measurement procedures suitable to the CIGS-based thin-film PV technology. Stable performance in the outdoor exposure test is warranted if CIGS-based thin-film PV modules are prepared by employing the moduling (or packaging) technology for crystalline-Si PV modules. To verify our assumption that significantly reduced degradation or even no degradation may be observed by applying a light soaking during the Damp heat test, the modified Damp heat test (85 °C temperature and 85 % relative humidity for 1000 hours) applying a continuous irradiation of 150 to 200 W/m2 is performed. It is confirmed that this assumption is valid for the CIGSbased thin-film PV modules tested in the open-circuit condition due to the enhancement of FF. INTRODUCTION
EXPERIMENTS Applying the moduling (or packaging) technology suitable to crystalline-Si PV modules, our baseline CIGS-based thin-film PV module has been fabricated in a glass-glass sandwich structure as shown in Fig. 1. In this structure, a white tempered glass as a cover glass was put on a CIGS-based thin-film circuit (or submodule) by cross-linking reaction of ethylene vinyl acetate (EVA) as glue. Experiments reported in this contribution were performed using this structure of CIGS-based thin-film PV module.
Cover glass (Semi-tempered glass, 3.2 mm thick.) Aluminum Frame Ethylene Vinyl Acetate (EVA) as glue Substrate (Soda-lime glass, 1.8 mm thick.) CIGS-based Thin-film Circuit (3 ?m thick.) EVA/Back Sheet
In the PV2030 Roadmap issued by NEDO , Cu(InGa)Se2 (CIGS)-based thin-film PV technology is expected to achieve the same efficiency level as the crystalline-Si PV modules by the year of 2030. However, in the actual status of CIGS-based thin-film PV, the commercialization was the slowest among the thin-film PV technologies. The year of 2005 would be remembered as a CIGS year because three companies, such as Showa Shell Sekiyu K.K., Würth Solar GmbH and Honda Motor Co., Ltd., took a step forward this direction with 20 MW/a, 15 MW/a and 27.5 MW/a, respectively. As the result, the worldwide production volume of CIGS-based thin-film PV modules will be over 70 MW/a in 2007. However, this level of production would be recognized as a pilot production compared to the current production level of other PV technologies. At this stage, the following items will be completed: 1) to evaluate mass production capability and expandability of the baseline production technology developed at the R&D stage, 2) to establish the appropriate production engineering, and 3) to develop the advanced technologies for further cost down and more profitable (or competitive) performance. These issues are needed to move toward a full-scale production. Another important issue is product reliability for a long product life. In this contribution, unique behavior of CIGS-based thin-film PV modules on a light soaking would be discussed. This would be helpful to establish appropriate characterization or measurement procedures to this “new-comer” PV technology.
Fig. 1 Structure of our current baseline CIGS-based thin-film PV module
RESULTS AND DISCUSSION 1. Outdoor exposure test As shown in Fig. 2, the module efficiency was improved, on average, +0.2 % from the circuit efficiency by applying an outdoor sun soaking for 3 to 4 hours in a fine day after making a module. It was verified that a light soaking after moduling was necessary to bring out the hidden performance by our current moduling procedure. In the outdoor exposure test performed at the test site shown in Fig, 3 with our baseline CIGS-based thinfilm PV modules, stable performance has been demonstrated for over 1100 days as shown in Fig. 4. In this test, the same modules were regularly taken out from the test site and measured under the standard test condition (STC, 25 °C, AM1.5 and 1000 W/m2) using a class-A steady-state solar simulator with a single-crystal Si reference solar cell to adjust the light intensity.
1-4244-0016-3/06/$20.00 ?2006 IEEE
Delta efficiency (= Module -Circuit) [%]
0.5 0.4 0.3 0.2 0.1
Average of 324 modules
As commonly recognized on the outdoor exposure test results [2,3], it has been verified that applying the moduling technology developed for crystalline-Si PV modules can warrant long-term durability to the CIGSbased thin-film PV modules. However, further R&D issues on the moduling would be a cost down approach keeping sufficient durability. 2. Modified Damp heat test Our current baseline CIGS-based thin-film PV modules have passed the test sequence defined in the IEC 61646, 1st edition, but partly considered the coming 2nd edition, which was carried out by the Japan Electrical Safety & Environment Technology Laboratories (JET). However, as shown in Fig. 5, it has been commonly observed in the Damp heat test in the IEC 61646 that the CIGS-based thin-film PV modules, which have been stored in the dark for 1000 hours at 85 °C temperature and 85 % relative humidity, show a temporary degradation after the test, but they can recover within a 95 % range of the initial performance by light soaking with a steady-state solar simulator or a metal-halide lamp to keep the constant irradiation or outdoor sun soaking [4,5].
0.0 39.0 -0.1 -0.2 -0.3
Module output [W] 41W 42W 43W
Fig. 2 Enhancement of the efficiency by outdoor sun soaking for 3 to 4 hours after moduling (324 modules, Average of +0.2%)
Fig. 3 Outdoor exposure test site at the Central R&D Lab. (ARL), Showa Shell Sekiyu K.K. (11kW output of 280 40W-CIGS mosaic modules (140 modules each x 2 segments), which has been monitored since February, 2003 Fig. 5 Typical result of Damp heat on CIGS-based thinfilm PV modules which were stored in the dark for 1000 hours at 85 °C temperature and 85 % relative humidity It is not clearly understood yet on the mechanism to explain why such light soaking by applying the constant irradiation for some duration is effective to recover from the (temporary) degradation after the Damp heat test. We assumed that CIGS-based thin-film PV modules would show significantly reduced degradation (or no degradation) if they were irradiated during the Damp heat test. To verify our assumption, a special apparatus for a modified Damp heat test, as shown in Figs. 6-1 and 2, was employed, which was consisted of two parts (i.e. a module set-up chamber for the Damp heat test and a light source of a class-B steady-state solar simulator). In this modified Damp heat test, two 30cmx120cm-sized module samples were stored in the module set-up chamber with a PYREX? glass window
Fig. 4 Outdoor exposure test result on the same seven CIGS-based thin-film PV modules regularly taken out from the 11-kW system shown in Fig. 3
and kept under the Damp heat test condition. A light 2 soaking of 100 to 300 W/m , which corresponds to a cloudy day, was irradiated continuously during the test period through this window and maintained by adjusting the Xe lamp intensity in a class-B steady-state solar simulator. The module temperature was monitored with an attached thermocouple (TC) at the rear side of the module and controlled at 85°C even when the above irradiation was applied. For this test, a thin transparent blue-colored plastic-sheet was coated on a full surface of a cover glass of the module to keep the constant irradiation by avoiding the formation of white deposits by the reaction of heat and water (humidity). 1) Module set-up chamber CIGS-based Thin-film PV Module
2) Light source of a class-B steady-state solar simulator
module temperature of 85 °C even when the irradiation 2 of 150 W/m was applied. As shown in Fig. 7, the results were that 1) the modules tested in the opencircuit condition had less degradation than the modules tested in the short-circuit condition, which was attributed mainly to less degradation of fill factor (FF), 2) the modules tested in the open-circuit condition showed a recovery within a 95 % range of initial output after light soaking for 2 hours which was performed under the STC, and 3) the modules tested in the short-circuit condition showed a degradation of almost 20 % drop from the initial output due to significantly depressed FF, which was in the same range as the normal (no irradiation) Damp heat test performed in the open-circuit condition.
Pyrex? glass window
Fig. 6-1 Line up of a special apparatus for the modified Damp heat test
Fig. 7 Result of the modified Damp heat test with the CIGS-based thin-film PV modules in the open-circuit condition keeping the irradiation intensity of 150 W/m2 (No light soaking was applied in the measurement under a steady-state solar simulator.) 1) Module set-up chamber (A module temperature of 85°C was measured at the rear side and a relative humidity was kept at 85 %.) These experimental results might be explained by either the previously proposed mechanism related to the Zn(OH)2 in the Zn(O,S,OH)x buffer deposited by the chemical bath deposition (CBD) technique  or the mechanism on carrier injection. Based on the former mechanism, depressed FF by heating at 85 °C was a result of hydration reactions of ZnO with free water molecules in the buffer, which led to increase in the amount of Zn(OH)2 in the buffer and to form a worse pn interface, while enhanced FF by light soaking was a result of dehydration reactions of Zn(OH)2 to form more ZnO. These two reactions would control the change of FF or interface quality. Based on the latter mechanism, it might be suggested that a photo-generated minority carrier (i.e. electron) by light soaking could compensate increased shallow defects by heating at 85 °C. In the open-circuit condition, this might happen inside the module. In the short circuit condition, recombination happened outside the module, so that no impact by light soaking was observed on the performance. Further investigation would be required to understand what happened in the Damp heat test as a specific test
2) Light source (A class-B steady-state solar simulator) Fig. 6-2 Special apparatus for the modified Damp heat test Experiments with the modules in the open-circuit or short-circuit condition were carried out maintaining the
condition. Another indication is that CIGS-based thinfilm PV modules should be stored in the open-circuit condition. In the experiment to increase the measurement frequency during the test, on-site (or in-situ) measurement was carried out after the following preparation: 1) correction of the irradiation intensity was done by using the formula incorporated in the monitoring system of a flash (or a long-pulse) solar simulator, 2) correction of the temperature was performed using the temperature coefficients measured by JET, 3) the irradiation intensity of 150 W/m2 was adjusted at the surface of the fully covered modules with a transparent blue plastic sheet through the PYREX? glass window by employing a single-crystal Si reference cell. As shown in Fig. 8, 1) FF was enhanced or did not degrade due to the light soaking and 2) significantly reduced degradation or almost no degradation was observed in the open-circuit condition and even at the end of 2000 hour test. Therefore, some 2 light intensity of over 200 W/m will be needed to cancel the temperature effect at 85 °C as the probably main cause of degradation.
tested in the open-circuit condition could be reduced 2 remarkably if a weak irradiation of at least 150 W/m was continued during the Damp heat test, further investigation on the light soaking effect to enhance the FF would be required. Although it was estimated that 2 irradiation intensity of over 200 W/m would be needed to enhance the FF at the module temperature of 85 °C, the intensity level of irradiation should be investigated to find the cross over point between the temperature and the irradiation intensity that is effective to prevent the (temporary) degradation. CONCLUSIONS The object of this study was to bring useful and effective data to establish the measurement procedures suitable to the CIGS-based thin-film PV technology. The following findings were verified in this contribution: 1) Light soaking after moduling was necessary to bring out the hidden performance in CIGS-based thin-film PV modules during the packaging process. 2) Stable performance of CIGS-based thin-film PV modules in the outdoor exposure test was warranted if the moduling technology for crystalline-Si PV modules was appropriately applied. 3) In the modified Damp heat test, in which module samples were stored applying a weak irradiation of 150 W/m2 continuously under the Damp heat test condition (i.e. module temperature of 85 °C and relative humidity of 85 %) for over 1000 hours, remarkably reduced degradation due to the enhancement of FF was observed on the modules tested in the open-circuit condition. This gave us some suggestions to consider the mechanism of light soaking on the CIGS-based thin-film PV modules. ACKNOWLEDGEMENT This study was consigned from New Energy and Industrial Technology Development Organization (NEDO). REFERENCES
Fig. 8 I-V parameters obtained by the on-site measurement as the function of Damp heat time in the modified Damp-heat test for the CIGS-based thin-film PV module tested in the open-circuit condition keeping 2 the irradiation intensity of 150 W/m The Damp-heat test condition in the IEC61646 is completely different from the actual outdoor application or the storage. Therefore, such test condition should be reconsidered in the future to fit or estimate the actual and realistic outdoor condition. Although it was verified that degradation of CIGS-based thin-film PV modules
 2030 (PV2030), June 2004) http://www.nedo.go.jp/informations/other/161005_1/gaiy ou_j.pdf#search='PV2030 Roadmap’  D. E. Tarrant and R. D. Wieting, Tech. Digest 14th “Sunshine Workshop” on Thin Film Solar Cells (2001), p. 80.  T.R. Strand, B.D. Kroposki, R. Hansen and D. Willett, Proc. 25th IEEE PVSC (1996), p. 965.  D. E. Tarrant, Solar Cells 30 (1991), p. 549.  K. Kushiya, Proc. WCPEC 3 (2003), 2PLC12.  K. Kushiya and O. Yamase, Jpn. J. Appl. Phys. Vol. 39 (2000), p.2577.