How Much Wind?

Yesterday the Department of Energy released a major study on the potential of wind power in the US, suggesting that by 2030 it could supply 20% of our electricity needs, at little incremental investment over and above what would be necessary anyway, to keep up with the growth of demand. This is an encouraging result for those who see renewable energy as a vital component of any effort to reduce greenhouse gas emissions and improve our energy security. While I can't possibly do justice to a 200-page report in a brief posting based on a quick skim-through, several interesting assumptions and observations leaped out at me. While they don't necessarily undermine the headline results, they serve to emphasize that the report is a finding of feasibility, not a forecast.

To an engineer, the wealth of data contained in a document such as this is irresistible. My first instinct was to dig out my calculator and start comparing the numbers to each other, and to current actual data. The report states that the quantity of wind power capacity required to supply 20% of US electricity demand in 2030 is 305,000 MW, or 305 GW. That reflects an 18-fold expansion of 2007 year-end wind capacity, for a compound annual growth rate of 14% over the next 22 years. Compared to an average growth rate of 36% over the last three years, that figure seems modest, but then it will get progressively harder to sustain double-digit growth as the installed base grows larger. But that's not the most interesting figure. That honor goes to the report's assumption of a 50% improvement in the capacity factor for wind--representing the ratio of actual output to nameplate capacity--from roughly 30% today to 45% by 2030.

Although I eventually found the support for that assumption in Chapter 2 of the study, I backed into it by comparing the 1,200 terawatt-hours (1.2 trillion kWh) of generation required to cover 20% of US demand in 2030 to the 305 GW of wind capacity. Although the study describes how this improvement could be achieved through a combination of advanced turbine technology, increased turbine size, and aggregation of the output of many turbines across a wide area, it goes beyond the capabilities of current turbines and what I have read of the experience of high-penetration European wind operations, such as Denmark. The reason this is important is that, if this significant improvement fails to materialize, then either a much higher level of wind installations will be required to supply 20% of US electricity in 2030, or the 305 GW proposed here would only deliver 800 terawatt-hours per year, or about 20% of 2006 demand.

Another interesting aspect of that 1,200 terawatt-hr figure is that it highlights the report's assumption of a 50% expansion in US electricity demand over the next 22 years. That works out to 1.8% per year, which seems reasonable based on past experience, but might not be compatible with efforts to reduce US greenhouse gas emissions significantly in the same period, unless it also assumes that a significant portion of new electricity demand will come from displacing petroleum from transportation via electric vehicles. The only mention of such substitution that I found in a full-document search was of plug-in hybrid cars as a potential outlet for off-peak wind generation. If 20% market penetration represents a practical upper bound, and electricity growth were only 1.1%/year, then maximum wind capacity in 2030 would be 258 GW, instead of 305 GW, still 15 times today's level.

The report also includes a set of wind power supply curves, showing the levelized cost of tapping all potential US onshore and offshore wind resources, ranging from 6 cents per kWh to over 14 cents. If I've interpreted it correctly, the incremental cost of the last land-based portion of that 305 GW in 2030 would be about 8 cents/kWh, while the offshore component would run between 10 and 11 cents, in current dollars and excluding the Production Tax Credit (PTC) that is currently up for renewal. This doesn't quite live up to the suggestion of some wind power advocates that it will be cheaper than coal, unless the coal power in question is from advanced generating plants with carbon capture and storage, or paying a substantial cost to emit CO2. At a minimum, the supply curve suggests that wind power will continue to need assistance on the order of the current 2 cent per kWh PTC, to penetrate beyond a small number of sites with the best economics.

The idea that wind power might someday supply a fifth of US electricity demand won't startle anyone who believes that the experience of Denmark could be translated to a much larger country. However, rather than viewing the DOE's "20% Wind Energy by 2030" study as confirmation of this notion, it should be recognized as a detailed scenario describing what would be necessary to reach that threshold. It spells out in considerable detail the improvements in wind turbine technology, transmission capacity, and load management that would be required. Much hard work remains to be done, to turn feasibility into practical possibility.