Mitigating Spontaneous Solar Module Glass Breakage and Catastrophic Hail Losses

RETC, a member of the VDE Group, conducted a subject matter expert interview with Jon Previtali —VDE Americas' product manager for natural catastrophe advisory services—as part of its research for the 2024 PV Module Index Report. Here, we share this in-depth PVMI interview, exploring spontaneous glass breakage, catastrophic hail losses, and a proposed new hail resiliency test standard.

RETC: Is spontaneous glass breakage something the engineering community is tracking?

JP: We have heard reports of a couple of incidents of spontaneous glass breakage from credible sources. While it’s unclear why it’s happening, we have some theories. One possibility is that the length of the rails that connect the tracker torque tubes to the modules is too short to provide enough support for the design wind load, which allows the modules to bend in the wind and causes the glass to crack. Another theory is that glass is breaking due to manufacturing defects, which could be inclusions, particles of pollutants in the glass, or other manufacturing process problems. Glass breakage could also be a construction issue, such as module clamps being torqued down with too much force during installation.

RETC: How can project stakeholders prevent spontaneous glass breakage?

JP: To address this risk, you can ask the engineer of record and independent engineer to check the length of the rails. We often see letters from PV module manufacturers confirming that rails are long enough, but without providing any supporting calculations. Since this issue isn’t covered under the PV module warranty, it’s important to get an independent third-party engineering review. To prevent manufacturing process issues, you can use a combination of accelerated lifetime testing and factory production oversight. For example, dynamic mechanical load tests done on a randomly selected group of modules can help identify issues before modules are produced and shipped. Also, production oversight inspections can monitor the glass manufacturing and handling process or verify whether testing has taken place for glass delivered to the module factory.

RETC: What is your experience with catastrophic glass breakage due to severe hail? 

JP: We are seeing more glass breakage due to hail because of the expansion of projects into hail-prone regions like Texas and the Midwest. The risk is exacerbated by larger, bifacial modules that use 2.0-mm [0.08-in.] glass for both front and back sides. The benefit of 2.0-mm double-glass bifacial modules is that they are not subject to the Section 201 tariff, and their larger format provides economies of scale cost reductions. The downside is that 2.0-mm glass is too thin to be fully tempered in most mass-production glass manufacturing lines. Because 2.0-mm glass is heat strengthened or semi-tempered only, it’s softer and more vulnerable to hail damage as compared to fully tempered glass, which is what we have with 3.2-mm [0.13-in.] front glass products. Unfortunately, modules of all kinds are being deployed in hail-prone regions without sufficient hail monitoring and stow protocols in place.

RETC: Is it possible to prevent catastrophic losses due to hail?

JP: Buying modules manufactured with fully tempered front glass will increase first costs but decrease financial exposure due to hail risk over the investment hold period. Module glass that is 3.2 mm or thicker is fully tempered and roughly twice as resilient to hail impacts as thinner heat-strengthened glass. We are aware of manufacturers evaluating different hail-hardened module design approaches, such as using thicker 4.0-mm [0.16-in.] front glass or using much thinner chemically treated front glass layers, like Gorilla Glass. While using hail-hardened modules is the best solution to the problem, all is not lost if you cannot. You can significantly reduce the risk of hail damage by using trackers that can put the modules into a steep stow position prior to a hailstorm. For material hail risk mitigation, we are talking about trackers that can stow at or beyond a 50° tilt angle under high wind conditions. This hail stow strategy significantly reduces the risk of glass breakage by reducing the glass area exposed to hail impacts and converting direct blows with a high kinetic impact energy into glancing blows with a far lower impact energy. While our recommended hail stow scenario is to tilt the trackers at a maximum tilt angle away from the wind, it’s very difficult—impossible even, using available technology—to know which direction the wind will be coming from when the hail falls. When there is a hail alert, we recommend that you put the modules into the closest maximum tilt position to avoid going through the flat position, which is most susceptible to hail damage. One of the ways that we support customers is by providing a hail stow protocol document that can be attached to an operating agreement.

RETC: Do you think the industry needs better ways to characterize impact resiliency?

JP: Currently, module hail resiliency testing follows the IEC 61215 standard, which requires 11 freezer ice ball impacts at different locations often on just one or two PV modules. Testing to this pass-fail standard does not provide enough data to accurately characterize module hail resiliency. It also doesn’t do a good job of picking up possible glass defects that would show up in a larger sample size of modules. To reliably characterize module hail resiliency, we need more than 11 data points, and we need to test modules to failure so that we know the probability of failure under different hailstone size conditions. To solve this problem, VDE Americas is working with RETC, our sister company, on a new hail impact characterization test that will use many more ballistic impacts and kinetic energy levels—representing a range of hailstone sizes and wind conditions—on approximately 30 randomly selected PV modules. Unlike IEC 61215, this new testing protocol will be a test to failure, meaning it will really show us what impacts modules are able to withstand. Importantly, this new test is designed to be not much more expensive than the typical IEC 61215 testing.

RETC: How are you able to control costs while generating more data?

JP: One issue with the IEC 61215 hail test is that it is difficult to make and maintain freezer ice balls that meet the testing requirements. You need special water and highly controlled freezer conditions, and you can’t store the freezer ice balls for more than a few days because they will sublimate, losing density and size. This complexity adds cost. To reduce cost and complexity, we will use synthetic ice balls instead of freezer ice balls. The synthetic materials we are evaluating have densities similar to those of freezer ice balls but are more consistent as a laboratory testing tool. So, by using a synthetic material as an ice proxy, we can develop an information-rich, cost-effective, and low-uncertainty test. Initially, the new test will be available only through RETC. However, once we have validated and defined this new hail impact characterization test protocol, our plan is to have VDE standards developers in Germany submit a proposed testing standard for IEC review.