Thermal Expansion Causes Solar Panel to Fail

Gordon Henderson sends this example of a solar panel that failed, in true monkey fashion, because it was inadequately designed for operation in sunny Scottsdale, Arizona!

“In the spring of 2008, I purchased two 15-watt solar panels. These solar panels are popular, inexpensive, and continue to be sold by numerous companies. I thought they presented a good value, and I began using them with a low-power solar project. My solar panels are in daily use in Scottsdale, Arizona, where the ambient summer temperatures exceed 110º F on many days. Therein lies a contributor to the problem, since the temperature of the solar panels is much higher than ambient, due to the solar-loading on their glass front surfaces.In July of 2008, when the temperatures were high, I noticed that one solar panel produced power in the morning, while ambient temperatures were lower, but stopped producing power in the afternoon, when ambient temperatures were higher. I originally wondered about a possible problem with the blocking diode that prevents reverse current, and I continued to monitor the symptoms. The symptoms continued on a daily basis. Soon, the second solar panel started to have similar behavior, so I had to act.

At this point I could have returned the solar panels to the manufacturer for warranty. But, had I sent them back to the manufacturer for warranty replacement, they would likely have been replaced with solar panels that would eventually demonstrate the same intermittent behavior due to the inherent problem. My failure analysis experience and engineering curiosity forced me to open the frames of each solar panel and have a look.

Upon opening the frames, I found all of the wiring potted with a substance that appeared similar to silicone RTV. First, I dug away the potting from around the diodes, checked the diodes, and proved that they were functioning properly. I was hesitant to dig the potting away from the positive and negative wires, because I suspected the wires were soldered to the solar panel via their thin, narrow, foil leads. I didn’t want to tear the delicate foil leads, especially if the leads were intact. Fortunately, I was able to prove which leads had the disconnect problem by pushing on the potting surrounding the offending lead to make or break the connection.

After very carefully digging away the potting, I could see that one foil lead was severed on each of the solar panels. These leads were straight with no stress relief to accommodate for movement due to thermal expansion within the potting. They were apparently pulled apart during thermal expansion of the potting at extremely high, solar-loaded temperatures. The broken leads would return to a position of slight contact at lower temperatures.

I was able to repair both leads by cleaning the foil and soldering additional foil leads between them and their wires. The new foil leads contained a generous stress relief. At this point, I suspected that the other two unbroken leads were marginal, and removed the potting to view them as well. As suspected, they had tearing of the foil or solder joint, although they were still functional. I repaired them in the same way.

During reassembly, I had to apply RTV potting to the wires to keep moisture out, but I avoided potting the foil lead areas. One year later, both solar panels are still performing normally.

The manufacturer should have used a generous stress relief or service loop for the delicate foil leads. This seems obvious under normal conditions. I would have avoided potting the leads altogether. If forced to use a potting substance, I would have picked one that would be more compatible and then tested it using the extreme hot and cold temperatures encountered by solar panels.

Did the manufacturer design and test these solar panels in a location with a cooler climate? Probably. I don’t believe the manufacturer accounted for the extreme solar-loaded temperatures encountered during Arizona summers. I wonder how many of these solar panels have been returned to the manufacturer for the same type of problem.

In the end, these solar panels were not a good value, but I still like them and use them. This problem was more than a nuisance, since the solar panels effectively lost total power on hot afternoons and provided poor performance when they were barely producing power in the cooler mornings. I have an additional four of the same solar panels that have only been used in cooler locations, and they are still functioning normally. I am hesitant to use them in a hot climate or install them in a permanent location. I have since bought additional low-power solar panels from another manufacturer, who uses a different and hopefully better method of attaching the leads. We’ll see.”