Programm Photovoltaik Ausgabe 2009 ... - Bundesamt für Energie BFE

1. Traceable performance measurement of PV devices (SP1)
1.2 ISAAC SP1 Activities (2008)
1.2.1 Round Robin tests (WP1.1)
Two measurement round robins (RR) were planned within the first half of the project (2006/2007) and
two within the second half (2008/2009). Each time a representative set of crystalline and thin film
modules is tested. The first round had the goal to document the current measurement differences for
commercial PV modules (crystalline silicon, thin-film and multi-junctions) within 8 of the major European
test laboratories (TÜV, ARSENAL, ISAAC-TISO, ISE, JRC-ESTI, CIEMAT, ECN, CREST),
whereas the second round has the goal to identify the progress in measurement procedures and test
equipment introduced by all laboratories within the intermediate optimisation phase. The results of the
first c-Si RR have been already described within the last annual report (2007). This report will concentrate
on the summary of the first thin film RR results and a short description of the second RR phase
started in 2008.
Results of the first thin film Round Robin
A detailed description of the thin film RR approach and all test samples can be found in the ‘ISAAC
activity report 2007’. The complete results of this RR have been published and presented at the last
European PV conference in Valencia [1].
This report summarises the main results of the first thin film round robin as well as the lessons learned
by the ISAAC institute. Figure 1 shows the comparison of the maximum power measurements among
the laboratories for the single thin film technologies. The data are always referred to the average of
each test sample. The spread for reported Pmax values measured under standard test conditions (STC)
lies in the range of ±7%, when excluding some outliers which have been explained by the absence of
any pre-conditioning procedure in case of CIGS and an insufficient initial stabilisation of one of the a-
Si/a-Si modules (all marked by red circles). The best comparability was obtained for single junction a-
Si modules, whereas the largest spread occurred for a-Si triple-junction and CIGS modules. No conclusions
could be drawn for the originally included CdTe technology due to the breakage of the test
samples at an early stage.
Figure 1: Discrepancies of PMAX data at STC as measured by 7 test institutes (represented by different
symbols) divided by type of test sample.
The much higher spread in Pmax compared to the ±2% of the first c-Si RR (see ISAAC activity report
2007) can be mainly explained by a higher uncertainty caused by the current mismatch of the layers
within multi-junction devices, the use of non-optimal reference devices and the consequential need of
spectral mismatch corrections and last but not least the short and long term pre-conditioning effects
influencing many of the thin film technologies. Even if all samples were exposed to a prior long-term
light exposure in order to stabilise the electrical output and to fix it to an initial reference point, most
PERFORMANCE, G. Friesen, SUPSI, ISAAC-TISO
210/290
2/8