Editor's Note: Scientific American's George Musser will be chronicling his experiences installing solar panels in Solar at Home (formerly 60-Second Solar). Read his introduction here and see all posts here.

Someone commenting on one of my recent posts posed an interesting question:

I wonder how much the albedo change of your roof offsets gains from electricity, much of the suns short wave energy must be being converted to heat therefore enhancing greenhouse   (as well as producing some electricity) can you prove you are actually energy balance positive???

In other words, solar panels both help the climate (by reducing the need to burn fossil fuels) and hurt it (by absorbing sunlight and warming the surface). Does the benefit really outweigh the cost? Nathan Myhrvold also raised this issue in a New York Times blog last week.

Silicon solar cells convert about 1/6 of incident sunlight into electricity and dissipate most of the remaining 5/6 as heat. So in terms of their direct climate effect, they have an albedo, or reflectivity, of 1/6. This is comparable to the albedo of standard asphalt shingles, so for most people, installing solar panels doesn’t have a net heating effect. But I had just restored my 1868-vintage tin roof and painted it white, giving it an albedo of about 5/6. So my panels do warm the planet.

The increase in heat absorption is a one-off climate forcing, equivalent to adding a certain amount of carbon dioxide to the atmosphere. By comparison, fossil power generation is an ever-increasing climate forcing, since each unit of energy entails the emission of more carbon dioxide. So there is some breakeven point beyond which the solar panels help the climate.

Let’s do a quick back-of-envelope calculation. My system is rated at 3 kW of DC power, so it dissipates on the order of 104 W of heat. Its area is about 10 square meters, about one part in 1014 of the Earth’s surface, so the system amounts to a global climate forcing of about 10-10 W per square meter. To convert this to a carbon-dioxide equivalent, note that the cumulative industrial emissions of 200 billion tonnes of carbon dioxide have produced a forcing of 1.5 W per square meter, or about 10-11 W per square meter per tonne.

Therefore by installing panels, I have had the same effect as releasing 10 tonnes of carbon dioxide. Burning fossil fuel emits roughly 1 tonne of carbon per MW-hr of electricity, so we need to produce 10 MW-hr to offset the albedo decrease. This will take about three years.

This is very rough and I’d be interested in other back-of-envelope calculations -- please let me know if I made a mistake. (Not that I have to implore readers of a blog to point out mistakes.)

I asked Gavin Schmidt of NASA GISS and he did the calculation a slightly different way. He imagines that it would take a photovoltaic array about 200 kilometers square to power the entire U.S. and estimates the albedo forcing to be about 0.01 W per square meter, equivalent to about 1 ppm of atmospheric carbon dioxide, which the U.S. adds to the atmosphere in about three years.

So the albedo effect is not negligible. It does need to be taken into account when calculating the payback period of solar energy. But it is hardly a deal-breaker.

Dark solar panels on George’s white roof