Assessing PV Module Reliability
by Mike Lucini
PV modules in the field are exposed to a host of trying environmental conditions. Thermal cycling, humidity, UV exposure, corrosion, hail, wind, and snow all come to mind. For a solar installation to be successful, it is critical that PV modules perform to expectations over the entire 25+ year service lifetime. To protect stakeholders in the event of failure, most manufacturers offer 10+ year product warranties, and 20-30 year power warranties. Some take the extra step of offering warranty re-insurance.
The assurances provided by warranties are nice, but it’s best if the modules never have an issue in the first place. No installer or system owner wants to deal with the headache of replacing modules, especially when product lines are continuously evolving, and some manufacturers are falling by the wayside. To that second point, for larger systems it may be wise to store a few spare modules long-term, so that a few bad apples don’t spoil the bunch.
Individual systems are important, but it’s more critical that reliability be addressed across the industry. How often do modules fail, and for what reasons? What exactly constitutes a failure? Like most mass-produced goods, PV module failure rates follow the classic ‘bathtub’ curve. It is characterized by a relatively high rate of “infant mortality”, a fairly constant rate of random failures, and accelerating failures towards end of life.
PV manufacturing has an inherent statistical distribution, so manufacturers perform in-line testing on each module to sort and sell them by power class. This presents an opportunity for rigorous quality assurance. Since each module’s performance and workmanship is tested by flash and electroluminescence testing (which reveals cracks and other defects), they can ensure that failing or obviously soon-to-fail modules never leave the factory. A module can be considered failing if it does not produce at least the power guaranteed by its warranty, since this is the minimum assumed performance.
All PV modules used in the United States must receive UL 1703 safety and IEC 61215 performance certifications. These certifications merely qualify a module design, and do not address long-term performance. To ensure long-term reliability, module designs should be tested well-beyond those benchmarks, and then manufacturing processes be adjusted to repeat those designs within acceptable tolerances. This should be a continuous, iterative process, where real sample batches are chosen at random from manufacturing lines and markets.
Many third-party qualification firms have begun such comparative testing programs. For example, Fraunhofer ISE and CSE published the first wave of results from its Photovoltaic Module Durability Initiative this past June. Since short-term tests are used to assess long-term reliability, the programs are accelerated to expose modules to a series or combination of extreme conditions. A module could undergo damp heat and UV exposure together.
Some of the more prominent failure modes that these endeavors may be looking for are listed below. These failure modes are not necessarily mutually exclusive. They can occur in tandem, accelerate the development of each other, and so on.
Potential-induced degradation: PID was discovered from observing modules in the field. PID occurs when system voltages cause ions to drift across different module materials. When the chemistry of the cell is changed, electrical performance is adversely affected. For a time, there was intense focus on determining exactly which system voltage and grounding configurations could best prevent PID. Empirically, it was found that sodium in glass was a major culprit. Manufacturers have adapted to incorporate sodium-free glass and other module materials that are not conducive to PID. Many modules can now be verified as “PID-free” to varying degrees by third-party testing.
Since PID is so sensitive to the materials stack, it is important that any change in material, even an equivalent material from a different supplier, be thoroughly vetted by a manufacturer before being introduced into their product lines. Cutting corners with inferior module materials, or switching materials without proper qualification, could cause PID down the road.
UV degradation: High-energy UV light can affect changes in any of the module materials. As with PID, any change in the chemistry and physics of the module can adversely affect electrical performance. UV degradation is known to cause browning of the backsheet, which may cause delamination.
Hot Spots: Hot spots are exactly what their name implies. Typically, a poor electrical interconnection between cells creates a point of high series resistance. Energy densities at these points can become extremely high.
Delamination: Delamination is a separation of the module materials stack. It can happen at any interface between the glass, encapsulant, cell, and backsheet. It can be the final catastrophic expression of a number of underlying causes. These include PID, hot spots, moisture intrusion, and wear-out from thermal cycling.
Diode failures: PV modules contain diodes in the junction box to minimize the effects of hotspots and somewhat mitigate partial shading. However, the diodes themselves present another possible point of failure. It is important that they be able to withstand thermal and humidity cycling, especially in high-temperature environments.
Most parties agree that accelerated testing can provide a valuable, quantitative assessment of module reliability. However, there is disagreement as to what extent that should go. Some groups are pushing for new standards in PV certification testing and/or manufacturing. Others believe that stakeholders should simply be smart about their purchasing decisions. To some, it is worth a small premium to select a product whose manufacturer has a long history in PV, with many GW of installed capacity and a reputation for quality. Looking at product recall history, or warranty claim rate, can go a long way towards deciphering the good from the bad.
D. Brearly. “Industry perspectives on c-Si PV module reliability and the rise of comparative testing.” SolarPro. Volume 6, Issue 6. October/November 2013.
D. Meakin, C. Schmid, G. Kinsey, C. Ferrara, S. Stecklum. “Fraunhofer PV Durability Initiative for solar modules.” Photovoltaics International. June 2013.
“What makes a PV module truly bankable? The important of quality materials for long-term reliability and performance.” Webinar presented by Suniva, featuring Dr. Sarah Kurtz of NREL. July 18, 2013.