Sunday, March 15, 2009

Rooftop solar

Many people are attracted to the idea of placing solar panels on the roofs of houses to generate electricity. This post by Matt Yglesias is typical. However the idea makes no economic sense.

Yglesias claims there are no economies to scale when generating electricity with solar panels. This is obviously false. Imagine millions of solar panels lined up in rows in desert areas of Arizona (or Nevada or California). Compare this to a few solar panels on the roofs of millions of houses. Which arrangement will be easier and cheaper to build and maintain? The answer is obvious. Furthermore the desert panels will generate more electricity as few houses get as much sun as the southwest desert.

So given that utilities are not generating electricity from solar panel farms in the desert how do you create the illusion that generating electricity from less efficient roof top installations is sensible. The answer is massive subsidies. Some of these subsidies like tax credits and low interest loans are obvious. Another is less so.

It arises from the way the residential electricity consumption is currently billed. Typically there is a small fixed charge and a per kwh charge based on usage which accounts for the bulk of the bill. However in reality most of the cost of providing electrical service to a typical residence is fixed consisting of the costs of constructing and maintaining the power generation and distribution system. In most cases the marginal cost of actually generating the kwh is not the bulk of the cost of service. The effect of the difference between the actual costs and how electrical service is billed is that large users tend to subsidize small users.

So what does this mean for rooftop solar panels. It means the reduction in your bill from generating some of your own power overstates the savings to the utility. The difference will have to be made up by the other customers which represents a subsidy to you (since you are not disconnecting from the utility entirely and still expect utility power to be available whenever you want it). The subsidy is even greater if you are allowed to sell excess power to the utility at the retail rate (run your meter backwards). By pricing rooftop electricity at the retail rate it may appear to be a better deal than electricity from solar farms in the desert priced at the wholesale rate but this is an illusion. It would in fact be cheaper and thus better for society if it was generated in bulk in large installations.

Installing roof top solar panels is a symbolic way of demonstrating concern for the environment. Too bad they don't actually make any sense.


  1. My residential electric bill in San Diego does not work in the idealized way that James describes (i.e., a fixed charge plus a simple per kWh charge). In fact it includes a complicated and opaque list of charges, fees, and rebates that make it impossible to predict the total payment from any given number of kilowatt hours billed. But it's clear that the cost per kWh actually increases steeply with usage, presumably reflecting some societal desire to subsidize starving widows and penalize electricity hogs, regardless of the actual cost of providing service. For example, in February we used 461 kWh and were charged $63.33, corresponding to a rate of 13.7 cents per kWh. For comparison, in January we used 553 kWh and were charged $92.17. The additional 92 kWh cost us $28.84 extra, for a marginal rate of 31.3 cents per kWh.

    This does not detract from James' main point that it surely must be more efficient to operate solar panels in bulk rather than distributed among individual homes and that rooftop installations must be subsidized in order to be competitive. But there already are large subsidies and penalties involved in individual household bills, and it's not obvious to me that this one is larger than any of the others. How electricity should be rationally priced is complicated in California, which has considerable hydropower that is very cheap but now insufficient to serve a swollen population and expensive long-term contracts that were signed during a crisis provoked by an ill-conceived experiment in deregulation.

    Two final points: There is some cost in building utility lines to bring power from the desert to where it is needed and some loss of power during transmission, both of which don't apply for rooftop installations. Finally, from an aesthetic point of view, some of us would rather see solar panels on rooftops than spread over vast areas of currently pristine desert.

  2. I agree with you, Peter, on the problems with solarizing the desert due to the fragile ecosystem. Take for example the cryptobiotic crusts in desert regions such as at Arches National Monument. The crust is composed of bacteria, fungi and algae and can take 500 years to form. It protects the desert soils from erosion. All this can be wiped out in a moment by heavy work boots, not to mention RV's.