Posted by: bmeverett | January 16, 2013

The state of play on solar energy

As promised last week, here is an update on the state of play of residential solar power in the United States. As with wind, we’ll start by agreeing on a set of rules for the analysis, summarized as follows:

First, we focus on cost, not jobs created or carbon emissions reduced.
Second, we look at the full cost.
Third, we exclude all the special taxes credits and subsidies.
Fourth, we use proper capital costs.
Fifth, we use the proper comparison basis.
Sixth and finally, we quantify externalities.

Consistent with these rules, my assumptions for solar energy economics are as follows:

a) The cost for residential solar panels is $2.40 per Watt, delivered to the house. With the 30% federal tax credit, the homeowner would pay only $1.68. As noted in my last post, however, this federal tax credit does not reduce the cost of the solar installation, but simply shifts it from the homeowner to the taxpayer. Someone still has to pay it, so we’ll stick with the true cost of $2.40. I cannot find any meaningful way of adjusting this cost for the many indirect subsidies, such as the federal grants to solar panel manufacturers. The cost structure of Chinese solar panel suppliers, who have recently dominated the US market, is even more opaque, so let’s take the $2.40/W as our basis.

This cost has come down quite dramatically over the last few years, but the reason for the decline is significant. The federal government decided some years ago to finance the construction of numerous solar panel manufacturing facilities, generally in the districts of powerful politicians. The idea was partly to cater to the green energy crowd, but mainly to boost investment and employment in politically important areas. When Chinese companies entered the US market in a big way a few years ago, much of the US manufacturing capacity became redundant. Many US suppliers were stuck with large inventories of solar panels that could only be sold at distressed prices. It remains to be seen whether these prices are sustainable. In any case, the current low cost of solar panels is not an indication of technological progress, but rather of market glut and low Chinese manufacturing costs.

b) The Balance of System (BOS) costs, which include the inverter, which converts DC power to AC, plus the wiring, switches and installation labor, total about 150% of the cost of the solar panels themselves. With $2.40/W panels, the total installed system cost would be about $6 per W. This is consistent with the average cost reported in by the State of California for recent residential solar installations (data at

c) Maintenance for residential solar systems is not high, and consists mainly of keeping the panels clean. The best estimates I can find are around 3¢ per installed Watt per year, so an average 4 kW system would require about $120 per year of maintenance. It’s worth noting that in the real world, people do not keep their solar panels very clean, just as they don’t keep their tires fully inflated. Dirty solar panels produce less power, but we’ll ignore that problem here.

d) The critical issue for solar systems is load factor, i.e., the amount of power a solar panel can actually produce over the course of a year. A natural gas combined cycle power plant has the capacity to produce more or less the same amount of power regardless of its location and construction, but solar output depends to a great extent on where the system is located and how it’s oriented to the sun. In order to avoid biasing the calculation, let’s pick Los Angeles, which is a pretty sunny area. Ideally, a rooftop solar panel in the northern hemisphere should be oriented directly south and tilted at roughly the same angle as the latitude of its location, in this case 34º. (Please note that a mechanically rotating system can be more efficient, but is much more expensive.) According to the DOE’s National Renewable Energy Laboratory, a perfectly aligned 4 kW rooftop solar unit in Los Angeles would generate 5,879 kilowatt-hours (kWh) of electricity every year. If the solar panels were generating power all the time, they would produce 35,040 kWh per year (24 hrs X 365 days X 4 kW), so the load factor for our ideal solar installation would be 16.8%. This calculation is fine if you are building a brand new house. The cheapest way of installing solar panels is to lay them on the roof, using the roof for support. Most existing houses, however, do not have a south-facing roof. Installing solar panels at an angle to the roofline requires a separate support structure and is both unsightly and expensive.

To be fair, let’s assume that, on average, all solar installations have the correct tilt angle, but the average orientation is 30° away from true south. In this case, the panels would generate 5,608 kWh or a load factor of only 16%. In addition, the rooftops of many houses are partially blocked by taller buildings, trees and other structures which reduce the available sunlight. Let’s assume that, on average, 5% of the sunlight is blocked. On balance, therefore, our residential solar panels in Los Angeles would generate 5,326 kWh per year or a load factor of 15.2%. Let’s use this number in our calculations.

e) The intermittency debit is a bit more complex for solar than for wind. Wind power comes mostly at night, when it competes with low-cost baseload power plants which are idled by the wind turbines. Solar, however, comes during the day and sometimes during peak hours when electricity is expensive. Let’s assume that the net intermittency debit is zero.

f) The proper basis for residential solar cost is the retail price of power to homeowners, rather than the generating cost to the utility. In Los Angeles, the average residential price is 14.4¢/kWh. Residential solar has no transmission debit.

g) Finally, let’s assume that the homeowner finances the entire solar installation with a home-equity loan. About 30% of homeowners are currently “under water” and have no equity, but let’s ignore this problem. According to Wells Fargo Bank, an LA homeowner with an average credit rating and $100,000 of home equity could obtain a 10 year variable rate home equity loan at 8% interest.

These assumptions produce a generation cost for solar power of 44¢ per kWh – over three times as expensive as retail power in California at 14.4¢/kWh. The entire argument for “grid parity” is based on a massive set of subsidies that do not reflect actual costs. As noted above, homeowners get a 30% federal tax credit, and Los Angeles customers currently get an additional rebate of $1.05/W for residential solar installations. These two programs alone would reduce the homeowner’s perceived cost to 27¢/kWh without reducing the true cost to the society.

Los Angeles is currently developing a “feed-in tariff” program that would pay residential customers 17¢/kWh for solar power generated by their homes but made available to the grid. If a solar installation sold, for example, 25% of its output back to the grid rather than consuming those kWh in the home, the breakeven cost of solar energy would be reduced another 1¢/kWh. These tariffs, however, are also subsidies, since other consumers must pay higher costs. Other states have property tax relief, low-cost financing, special incentives for low-income homeowners and other subsidies. If we add enough of these subsidies, we can make solar energy look inexpensive to the consumer, even though it would still be very expensive to the society.

What about externalities? If we start with the full social cost of 44¢/kWh, we would need a carbon tax of over $1,000 per metric tonne to make the cost of residential solar equal to the cost of grid power. To put such a tax in perspective, the US currently emits about 6 billion metric tonnes of CO2 per year. A $1,000 tax would be a burden of $6 trillion a year on our GDP, which is currently about $15 trillion.

We really should stop hiding and obscuring the true costs of solar energy just to make it look good. As with wind, the real issue is whether subsidizing solar will actually force down its cost. There’s no reason to believe that this is the case. Take helicopters, for example. The cheapest two-person helicopter currently on the market costs about $200,000 plus maintenance costs of about $13 per hour of flying time. Fuel efficiency is about 10 miles per gallon – about the same as a Hummer. If the government gave massive subsidies to, say, 100,000 helicopters per year, would we expect that helicopters would quickly become cost-competitive with automobiles for private transportation? I very much doubt it. The high cost of helicopters is inherent in their physics, not just the result of poor technology or low manufacturing scale. The same is true for solar power. We may ultimate find an inexpensive way of harnessing the sun’s energy, but solar photovoltaics are probably not it. In the meantime, let’s at least be honest about the status and cost of this technology.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: