Some people in the PV industry say that solar modules are commodities. If they’re right, then solar modules are the only commodity in the world that comes with a 25-year warranty.

We’re not talking about toasters or televisions, where consumers don’t really care if the products don’t last longer than a decade. Billions of dollars are staked on solar modules meeting performance expectations for 20 years or more, in any and all environmental conditions. Although photovoltaic technology has generally proven to be extremely reliable -- crystalline silicon modules installed in the 1970s are still performing well today -- we’ve learned a lot over the last three decades that should inform both manufacturing operations and module purchase decisions.

In 1996, when my boss’s boss’s boss, CEO of Suntech Dr. Zhengrong Shi, was just my colleague, we were both hopeful that crystalline silicon-on-glass (CSG) thin-film modules would be the next big thing in solar. Dr. Shi was then the Deputy Research Director at our small thin-film startup in Sydney, Australia, called Pacific Solar (later renamed CSG Solar), a spin-off from a research lab at the University of New South Wales (UNSW). Although competitive CSG conversion efficiencies were merely a dream -- and still are -- it was our goal to bring a commercial product to market before the turn of the century. Consequently, we had to think seriously about module degradation and the vulnerabilities of deploying a new silicon substrate less than two microns thick (think human hair).

In the absence of significant field data for CSG module performance -- there was no such product in existence, and we couldn’t wait twenty years -- all we had to go on was internal product testing data. At the time, the major solar product standards bodies used a similar variety of product durability tests to simulate accelerated module degradation, including damp heat, thermal cycling, and humidity freeze. A common shortfall was that each test was always performed on a fresh module. The standards did not -- and still do not -- require the same module to undergo a variety of simulated, real-life environmental conditions. As a result, these standards are at best equivalent to three to five years of outdoor exposure and should be viewed as a baseline test protocol (Hoffman & Ross 1978, McMahon et al. 2000).

Consequently, we didn’t have a lot of faith that the standard testing requirements would provide insight into actual long-term module degradation patterns, especially when deployed in extreme climates. That’s a big problem if you’re responsible for a product for longer than you’re legally responsible for your own children. We believed our tests had to go above and beyond the required standards, leaving no doubt that the product would meet or exceed an aggressive 25-year power output warranty.

We developed an extended sequence of the IEC 61215 standard, which we called Combined Cycle testing, as a way to simulate the impacts of real-life, long-term environmental stressors on our modules, and on each of the individual components within. After Suntech acquired CSG, Suntech has continued implementation of this process to ensure the performance and reliability of the more than 25 million solar modules (over 6,000 megawatts) shipped to more than 80 countries and every environmental condition imaginable. We can now more effectively cross-reference our simulated test results against a range of real-world performance data, allowing for the development of better testing methodologies and better modules.

Here’s how our test works: one Combined Cycle consists of each of the standard exposures conducted in sequence, that is, 200 Temperature Cycles (from -45⁰ C to +85⁰ C), followed by 1000 hours Damp Heat (85% RH and 85⁰ C), then 10 Humidity Freeze cycles (-45⁰ C to +85⁰ C and 85% RH), each applied sequentially to the same modules. One sequence of such tests takes about 4 months to complete and is called one Combined Cycle. If the module survives, we do it again and again until the module’s power performance drops below 80% of its initial value (our warranty promise at 25 years). Linear interpolation is used to estimate when output drops to the 80% value, so we can make a fairly good estimate of the module’s durability performance in terms of number of Combined Cycles withstood (Green et al. 2004).

A summary of historical results for our Combined Cycle testing of Suntech and other commercially available PV modules is provided below. The commercial modules used are all from leading global manufacturers, as the few modules tested from third-tier manufacturers have been unable to withstand a single Combined Cycle.

Modules from all manufacturers lose some power due to current degradation. The primary cause of degradation for both crystalline silicon and thin-film modules is an increase in the module’s series resistance, as determined through lumped circuit modeling of the module under test. The causes of increased series resistance vary between modules, but factors such as delamination, cell cracking, tabbing connections or local hot-spots are often the culprits. Just like in real life.

Importantly, Combined Cycle testing alone cannot simulate all potential real-world stresses. Although better than the standard IEC tests, Combined Cycle does not include any UV or mechanical load testing, which may accelerate or identify new failure mechanisms. Therefore, in addition to Combined Cycle, Suntech participates in a number of other long-term accelerated test programs, including:

  • TÜV Rheinland Long Term Sequential Exposure test (TÜV 2010) – combines Combined Cycle and Double Length, whereby each module gets double exposure period of each of the three exposures plus some UV exposure.
  • Atlas 25plus (Zielnik 2010) – includes UV, salt spray, condensing humidity, temperature cycle, humidity freeze and outdoor exposure (in Arizona).

Suntech’s module testing facility in Wuxi, China was the first in China to be awarded the Underwriters Laboratories (UL) Witness Testing Data Program (WTDP) Certificate. In other words, with oversight from UL, Suntech can perform UL certifications in-house for their own products.
It’s no wonder these tests are far from ubiquitous, as they all require significant resources and time to conduct. A PV manufacturer really can’t properly test a new product in less than six months -- it's more likely to take a year. This has important implications for introducing new PV products to the market and it is imperative that accelerated lifetime testing is considered early in product development plans.

But it’s well worth it.

Reducing performance risk for our customers drives down the levelized cost of solar electricity generation and increases project returns. The basic test standards do not sufficiently separate the wheat from the chaff. Many lower quality modules -- modules that do meet IEC standards but which have not undergone Combined Cycle testing -- could underperform their 25-year power output warranties, and that’s not good enough in today’s highly-competitive market environment. Construction companies, integrators, and project developers who care about bankability can and should ask for more from their module suppliers than basic IEC compliance.

There’s no question that cost per watt will remain the baseline metric in analyzing the competitiveness of solar products. However, unlike commodity products, not all solar module watts are made equal. Even a 1 percent performance difference over 25 years can equate to hundreds of thousands of dollars in revenue, depending on the project size. We should ask for more lifetime energy from solar modules, rather than merely lower prices.

Stefan Jarnason leads Suntech Australia’s team of engineers in providing system design, energy simulation, and product development. Stefan has 15 years of experience in photovoltaic research, durability testing, manufacturing, product development and commercialization in Germany and Australia.