The Fraunhofer USA Center for Sustainable Energy Systems (CSE) began pursuing its Photovoltaic Durability Initiative (PVDI), a set of panel tests that go beyond safety and performance, in 2010, along with Germany’s Fraunhofer Institute for Solar Energy Systems (ISE).

PVDI is “a protocol to provide quantitative comparative data for modules in various operating environments,” explained Fraunhofer senior technical staff member David Meakin, and “to identify the field performance capabilities necessary to reduce risk to investors.”

Durability, which Fraunhofer focuses on instead of bankability, depends on a lifetime assessment of a panel’s performance, Meakin said, and the data on that is inadequate. “The best we could do is a probability because we don’t have enough technical information -- nobody does --  to make lifetime predictions yet.”

Fraunhofer is a participant in the NREL International PV Module Quality Assurance Forum. Its nine working groups, composed of hundreds of scientists and engineers, hope “to generate the data necessary to start to make those predictions,” Meakin said. Ultimately, NREL hopes to produce “a complete quality control system for PV module manufacturing.”

Meakin has run five solar manufacturers’ modules through the Fraunhofer PVDI protocol and is currently running three more. The first public report is in preparation. Meakin hopes to eventually accumulate enough module data to assess “durability within a given environment.”

And, Meakin added, “as we learn about the testing standards that inform us about a module’s ability to survive in a given environment over time, we may change the protocol.”

Fraunhofer does not accept manufacturer-supplied modules. “The manufacturer contracts with us to test their modules, but we buy them on the open market from a supplier of our choice. This is to avoid cherry-picking. What we are looking at are typical modules that any buyer might get on the open market.”

The protocol tests include, Meakin said, potential induced degradation (PID, an electrical stress), damp heat (a moisture stress), UV exposure (a photonic stress), dynamic and static loading (a mechanical stress), thermal cycling (a thermal stress), and a long-term outdoor exposure test.

“We characterize the modules before we test them, we characterize the modules while they are being tested, and we characterize them after they have been tested,” Meakin explained. “The various test protocols are the stressors. What we are looking for is not only the absolute performance but how that performance changes over time with the stress.”

Though these tests are not very different from those being performed by the other labs described in this article series (Renewable Energy Testing Center, Intertek and PV Evolution Labs), under the heading of bankability, the Fraunhofer protocol -- like the others -- probably does not capture every detail a bank, developer or manufacturer might want to know about a module, Meakin admitted.

“But the stresses cause all the modules to degrade or change their performance to some degree, and they will be ranked on that basis.” The PVDI, Meakin said, is quantitative, not qualitative.

“There are two standards, UL 1703 and IEC 61215,” Meakin explained. “The UL testing is more geared toward safety. The IEC standard takes into account both safety and construction. They have minimal pass-fail criteria.” They are, at best, tests of “infant mortality,” Meakin said, not “long-term durability.”

In assessing reliability, he explained, there are three basic failure mechanisms at work: so-called infant mortality, constant (or random) failure, and wear-out. “Put the three together and you get the classic bathtub curve.”

 

The bathtub curve is a probability curve. “When you start out, you have a high risk of failure, known as infant mortality. That risk comes down very rapidly. For most of the life of anything, there is a low but constant failure rate. At the end of something’s life, because of wear-out, the probability of failure rapidly increases.”

The unanswered question about PV modules, Meakin said, is how long until the probability of failure starts increasing.

“Accelerated testing methods are part of it. But does a test for 1,000 hours relate to one year, ten years or 1,000 years? Depending on the material, it can be any one of these. We understand some of the acceleration coefficients, but not all of them. The other part is what we observe in modules that have been in the field for a long time. Sometimes we have to rely on accelerated testing because we don’t have twenty years to wait to see what happens.”

Part of what PVDI essentially does, Meakin said, is Highly Accelerated Lifetime Testing (HALT). “PVDI takes a series of modules, exposes them to outdoor stress and then ranks them according to that stress.” Meakin expects the process to render more acceleration coefficients.

Meakin wants to create environmentally specific test procedures so that a panel buyer in the desert can “choose modules ranked for top performance in high-irradiance, high-thermal-stress conditions and ignore performance in humidity. This ranking allows you to say these modules do well in this type of environment.”

It is too soon to know what a final international protocol will look like, Meakin acknowledged, “but I can guarantee you this kind of testing will go on. It is being called for, not by the module manufacturers, but by the downstream business entities -- installers, bankers, financiers, and plant operators, who understand their risk."

And, for manufacturers, the benefits of testing outweigh the cost, Meakin said. “A rule of thumb is that for every dollar it costs to catch a quality problem in the factory, it will cost you twenty times that if your customer catches it. There are all kinds of customer relationship costs and soft costs of allowing something to get out of the factory.”

***

This is the third in a series of articles on module testing that also includes Testing and Ranking Solar Module QualityWhat Do Solar Module Test Procedures Prove?, and A More Realistic Take on Solar Module Testing.