Should you spring clean your solar panels? Google ran the data to find out:
We have two different sets of solar panels on our campus — completely flat ones installed on carports, and rooftop ones that are tilted.Since the carport solar panels have no tilt, rain doesn’t do a good job of rinsing off the dirt they collect. (Also, our carports are situated across from a sand field, which doesn’t help the situation.) We cleaned these panels for the first time after they had been in operation for 15 months, and their energy output doubled overnight. When we cleaned them again eight months later, their output instantly increased by 36 percent. In fact, we found that cleaning these panels is the #1 way to maximize the energy they produce. As a result, we’ve added the carport solar panels to our spring cleaning checklist.
The rooftop solar panels are a different story. Our data indicates that rain does a sufficient job of cleaning the tilted solar panels. Some dirt does accumulate in the corners, but the resulting reduction in energy output is fairly small — and cleaning tilted panels does not significantly increase their energy production. So for now, we’ll let Mother Nature take care of cleaning our rooftop panels.
I love their emphasis on data, but I’m pretty sure I could have provided that conclusion by eye-balling the two set-ups.
The financial analysis:
Our analysis now predicts that Google’s system will pay for itself in about six and a half years, which is even better than we initially expected.
I wouldn’t recommend using payback period for most financial decisions, but it is easy for a mass audience to understand. If we expect the solar panels to deliver in perpetuity, that would imply a 15 percent rate of return. If they degrade at 1 percent per year, as implied by the slides, that’s still impressive. In fact, even if the solar panels only deliver for 10 years, that’s still a nine-percent return with, apparently, very little risk. This implies that solar should be much more popular with bottom-line financial types. Hmm…