The phrase “garbage in, garbage out” explains that analysis based on flawed input assumptions will result in flawed results and conclusions. A recent textbook example of that comes from the American Coalition for Clean Coal Electricity (ACCE) and the special interest group Institute for Energy Research (IER). IER has been down this road before, and once again their analysis isn’t grounded in reality.
Wind is the cheapest source of new electric generating capacity
The fact is utilities and their regulators are buying renewable energy because it reduces electric bills for homes and businesses. Market prices for renewable energy purchase contracts confirm wind and solar energy are often the lowest cost option these days, as USA Today noted earlier this week. Wind power purchase prices averaged $18.91/megawatt hour (MWh) in 2017, while preliminary data indicate that solar prices averaged $22.30/MWh in 2018. The utility Xcel Energy has indicated its recently announced wind projects have a levelized price of $15-25/MWh. These purchase prices include the federal tax credit for renewable energy, which further reduces the cost of renewable energy to consumers.
However, even without incentives, in many parts of the country wind is still the cheapest source of new electric generation according to sources like investment firm Lazard Inc. For example, researchers from the University of Texas Austin’s Energy Institute created an interactive map showing the most affordable energy sources across the country. Check out all the green—that’s where wind is cheapest.
Renewable energy increasingly saves consumers money even without the federal incentives, which are currently being phased out. The following table uses real-world performance data to calculate the unsubsidized levelized cost for newly installed generation.
Notably, the preceding table uses actual capacity factor data for new power plants. As the IER authors correctly noted in their analysis, the actual levelized cost for new gas and new coal generation is dramatically higher than DOE estimates because DOE uses maximum theoretical output levels for fossil power plants instead of real-world performance data.
A mistake-filled report
Once renewable cost errors (discussed below) are corrected, IER’s own analysis shows new wind and solar are also competitive with existing fossil and nuclear generation. This confirms a wide variety of analysis, much of it conducted by coal power plant owners. Regardless, many existing coal and nuclear plants are also reaching the ends of their useful lives, and require hundreds of millions of dollars in repairs and upgrades if they are to continue operating.
True to the “garbage in, garbage out” maxim, the IER authors reach the wrong conclusions because their assumed costs for wind and solar energy are three to four times higher than all estimates by independent experts, as shown below.
How did the authors get the answer so wrong? Part of the problem is that they use old data for the productivity of wind and solar plants. The authors use the average productivity of the entire fleet of wind and solar plants that are currently operating, which misses that new plants are far more productive than those installed even a few years ago. This error adds $10.50/MWh to wind costs and $7.70/MWh to solar costs. In reality, solar capacity factors have increased 23 percent since 2010, while wind capacity factors are up 41 percent since 2009. Partially as a result, wind’s costs have fallen by 69 percent since 2009.
The report then compounds that error by inventing other cost adders. Most notably, the report incorrectly claims that adding renewable generation requires the addition of conventional generation. First of all, almost all U.S. regions already have excess power plant capacity. More importantly, adding wind or solar to the power system always provides positive capacity value to the power system to help meet electricity demand, and therefore reduces and never increases the need for conventional power plant capacity. Due to that error alone, the report incorrectly adds an “imposed cost” of $23.6 per MWh to wind generation and $21/MWh to solar generation.
In addition, the study lowballs estimates for the contribution of wind and solar to meeting peak capacity needs. The authors give solar credit for only 12.9 percent of its installed capacity, while many grid operators have shown solar’s contribution to be roughly five times higher. They similarly understate wind’s contribution by a factor of nearly three. For inexplicable reasons, the authors then add costly conventional capacity and generation to make up the difference, even though it is not needed.
IER’s convoluted “imposed cost” method is not necessary, as the Department of Energy’s Energy Information Administration (EIA) has already developed a method that accurately accounts for the different energy and capacity value of all energy sources, fully accounting for differences in their dispatchability and time of production. That method shows that the value of wind energy is only $4.60/MWh, or about 12 percent lower than that of more dispatchable gas combined cycle generation, while solar generation is actually $2/MWh more valuable than gas combined cycle generation.
The cost of reliably integrating large conventional power plants is actually larger than that of renewable resources, as data from Texas indicate. That means accurately accounting for integration costs and other costs imposed on the system would actually indicate a net benefit from increasing use of renewable energy. Large conventional power plants can fail instantly and without warning, and therefore require more expensive, faster-acting operating reserves than the slower-acting reserves used to accommodate gradual and predictable changes in wind and solar output. Regardless, in Texas both costs are well under $1/MWh, while the Midwest regional grid operator has noted that the impact of wind generation on its need for fast-acting reserves is “little to none.”
Large coal and nuclear power plants are also inflexible, in that they are unable to quickly change their level of output in response to fluctuations in electricity demand, which imposes a cost on the power system and other power plants. As a result, the Independent Market Monitor for the Midwest grid operator has found that coal plants account for the vast majority of the region’s generation scheduling error and associated operating costs. In contrast, most wind and solar plants are now fully dispatched by grid operators through electricity markets, and can respond hundreds of times faster than conventional power plants.
ACCE and IER’s study should serve as a warning sign for analysis based on faulty assumptions and dated information. If you put garbage in, you’ll get garbage out, and the end result won’t tell you much about the world that exists in reality.