Learning from the test

You’ve now tested another module. For the second time, a manufacturer whose module was graded “good” doesn’t want to publish the fact. Why’s that?

For some manufacturers, “good” simply isn’t good enough. Furthermore, even modules tested out as “good” can show weaknesses in individual categories. Some companies don’t want a relatively poor detail result to be printed, even if the test as a whole yields a good overall score. We always publish the detailed tables as well, showing up the strengths and weaknesses of a module, so some companies do decide against publication.

What critical points does the test reveal?

We sometimes see suboptimal results, for example, after the module has been aged in climate chambers. In our detailed assessment, some modules will perhaps get only one plus instead of three. If you look more deeply into the detailed results, you can see specific weaknesses in some modules, say a degradation of performance following the accelerated aging tests.

For the overall score, “satisfactory” results in one category can be compensated for by “excellent” results in another. In view of that, how meaningful is the overall score?

There are, of course, very many values that have to be combined into an overall score by means of specific weighting. But if there actually are serious defects in a given category and these are concealed when we combine individual assessments into an overall score, we run an extra evaluation. So the overall score is meaningful. It tells you clearly whether a module is excellent, good or satisfactory.

How did you set up the weighting of the various parameters?

The weighting was established by the advisory board. Industry representatives met several times to set it up. For example, they discussed how much value should be put on electrical safety, and how strict the performance and aging tests should be. These were then openly decided on. The basic approach was that a module is a complex component and, if you want to judge it meaningfully, you have to combine many individually weighted categories. This distinguishes PV+Test from other tests that take account of only a few categories when arriving at an assessment.

The test is intended to have a positive influence on the market. Can you see any results of this yet?

Yes, I do see some. First, there genuinely are many enquiries from interested manufacturers who say they can also maintain this level of quality and want to take part. Further tests are being conducted right now. Second, we have enquiries from manufacturers who have already taken part and wish to improve their modules. They want to pick up criticism we’ve revealed and perhaps have another go with a modified module. Some manufacturers want to see the many very detailed measurement results that we haven’t published. Additionally, they ask for information about test equipment – say, how we test modules for sharp edges. One manufacturer now wants to use the test in their own quality assurance procedures.

What unpublished data are of interest to manufacturers?

Electroluminescence imaging, for example, electrical insulation values under dry and wet conditions, and individually measured curves from performance testing. That would have been too much data for publication.

Module manufacturers can collect these data themselves, after all. What further help do the test data give them?

Measuring output isn’t trivial. You need a reference against a valid international norm. A great many parameters affect a measurement: temperature, the spectral nature of the light, and the homogeneity of the test area. It’s nothing unusual for module manufacturers’ own measurements to differ from TÜV Rheinland’s test results. Besides, the modules we test aren’t selected by the manufacturer but are bought anonymously. So we measure and analyze a random sample of the modules produced.
PV+Test results Since first publication in February (pv magazine 02/2011, p. 92), another module has scored “good” on the test, but its manufacturer doesn’t want it to be published (see interview). The PV+Test experts feel the tests have gone very well. Most modules scored “good” or “excellent” and corresponded with the state of the art. Only one module has so far tested as “poor.” Its manufacturer also declined publication.
TÜV Rheinland’s overall assessment is based on a large number of measurements, not all of which have been included in the table. Some can, however, be found in the individual reviews in the February issue. Each measurement is given a number of points between zero and ten, which are then separately weighted and added to give the overall result. The final scores are based on the total number of points, as follows: ? 90% of the maximum number of points: excellent; ? 80%: good; ? 70%: satisfactory; ? 50%: poor; < 50%: very poor. The sub-items are correspondingly assessed as: +++, ++, +, o, -.
The weighting for the sub-items towards the overall score was decided on in the industry’s advisory committee.
* Conergy: In the isolation measurement after temperature fluctuations, leakage current was detected at the plug of one of the two modules tested, in violation of the IEC standard. A repetition of the measurement was, however, successful. PV+Test is copyright protected. The table may only be disseminated with the consent of Solarpraxis AG and TÜV Rheinland.