Manufacturers hold the details of their processes close to the chest, but you can use their sale price for a back-of-the-envelope upper bound. Searching Google for “photovoltaics”, the first price I found was $2.38/W (in one of the ads). Assuming the manufacturer only breaks even, and that entire price was spent on energy used for manufacturing the panels and the materials used to make the panels, then at 5 cents/kWh (which according to http://www.neo.ne.gov/statshtml/115.htm is a good price), making the panel couldn’t have used more than 47kWh/W. If that much energy was used, and the solar panel operated at rated capacity for 8 hours/day, then it would take 16 years to produce the energy used to make it. By comparison, the warranty on the same panel guarantees at least 80% capacity for 25 years.
However, this is a very loose upper bound, in that I assumed that the entire purchase price was spent on electricity. This is probably off by an order of magnitude, since the cost of solar cells is dominated by labor, R&D, factory equipment, and profit, not by energy use.
I appreciated this calculation. Although as you show it’s unlikely that solar panels represent a net loss in energy, it’s still kind of off-putting that, given my local power rate of 5 cents/kWh, I’d have to wait 16 years to make back my money if I buy solar panels to reduce the power I take from the grid. Of course, this results mainly from the government subsidy of residential power, a bizarre policy completely at odds with the same government’s exhortations to be “power smart”.
Manufacturers hold the details of their processes close to the chest, but you can use their sale price for a back-of-the-envelope upper bound. Searching Google for “photovoltaics”, the first price I found was $2.38/W (in one of the ads). Assuming the manufacturer only breaks even, and that entire price was spent on energy used for manufacturing the panels and the materials used to make the panels, then at 5 cents/kWh (which according to http://www.neo.ne.gov/statshtml/115.htm is a good price), making the panel couldn’t have used more than 47kWh/W. If that much energy was used, and the solar panel operated at rated capacity for 8 hours/day, then it would take 16 years to produce the energy used to make it. By comparison, the warranty on the same panel guarantees at least 80% capacity for 25 years.
However, this is a very loose upper bound, in that I assumed that the entire purchase price was spent on electricity. This is probably off by an order of magnitude, since the cost of solar cells is dominated by labor, R&D, factory equipment, and profit, not by energy use.
I appreciated this calculation. Although as you show it’s unlikely that solar panels represent a net loss in energy, it’s still kind of off-putting that, given my local power rate of 5 cents/kWh, I’d have to wait 16 years to make back my money if I buy solar panels to reduce the power I take from the grid. Of course, this results mainly from the government subsidy of residential power, a bizarre policy completely at odds with the same government’s exhortations to be “power smart”.