This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


The Cost of Wind Energy: Part II

Charles Barton's picture
Nuclear Green

I am a retired counselor. My father was a nuclear scientist and I have had a life long interest in and fascination with his work.

  • Member since 2018
  • 654 items added with 95,648 views
  • May 28, 2013

wind energy costs

In Part I, we explored past and present wind costs and noted rapid inflation. This was the first step in an exploration of David Roberts’ claim that renewable electricity was less expensive than nuclear generated electricity, but there are many other factors that I did not touch on or barely mentioned in Part I that require further exploration.

Nuclear reactors typically generate ninety percent of their faceplate electrical capacity. That ninety percent is called “capacity factor”.  Reactors are typically taken off line for maintenance at periods when electrical demand is not at its peak, thus reactors are almost always available when consumers demand electricity.

In contrast, wind generators typically produce electricity at a capacity factor of about thirty percent. To equal the gross electrical output of a nuclear reactor, three wind generators producing equivalent nameplate capacity would be required, but it is more complex than that. If those three wind generators produce constant electricity ninety percent of the time, then the cost of wind would simply be three times the cost of one wind generator. This cost itself would take the cost of wind into the same range as the cost of nuclear power or higher, but there are more costs associated with wind. In the first place, wind does not blow at constant speeds even over a large area. More wind generators are required to compensate for periods of slow wind, but there are also periods of very slow wind or no wind at all. During periods of slow wind, more wind generating capacity is required to keep electrical output from wind installations up.

Mark Z. Jacobson claims that by spreading wind facilities over Texas, Oklahoma,  Kansas, and New Mexico and linking them with high voltage power lines something like a reliable power source can be accomplished eighty percent of the time. Five wind generating facilities with the same nameplate capacity as one nuclear power plant would be spread over the four states and linked by high voltage power lines. Even then it would fall short of goals twenty percent of the time. Jacobson does not tell us how much his scheme would cost, but it would be a pretty penny. Day time winds drop in speed as temperatures soar in Southwestern states such as Texas. As temperatures soar, the demand for air conditioning swells as well, thus the generating output of Jacobson’s wind system is poorly matched to Texas electrical demand in the summertime. Some backup must be found to Jacobson’s already expensive wind system.

In 2007, when I was arguing with Roberts, I pointed out the problem of wind fluctuation and the need for backup. One of Roberts’ readers responded that the wind system could simply be connected to the grid and fluctuations could be covered as they already are on the grid. Fluctuations on the grid are covered by so called spinning reserve. That is, power plants that are kept running without covering load. If a power plant is suddenly taken off line, or if consumer demand rises quickly, spinning reserve is brought on line and begins to supply electrical energy, but spinning reserve requires fossil fuel backup. If our goal is to have one hundred present replacement of fossil fuels as the energy source for the grid, we will have to eliminate fossil fuels from our backup mix.

Furthermore, studies of wind penetration of the grid suggests that wind displaces natural gas at low levels of penetration and only begins to displace coal when wind penetration rises above twenty percent of electrical demand. When wind penetration arises above twenty percent, the cost of electricity begins to rise as well. For relatively small wind penetration levels, wind simply supplements other electrical generating systems. For example, in the Pacific Northwest, wind is matched to electricity generated by water driven turbines along the Columbia River. Wind, when it is blowing, is a useful tool in managing the Columbia River electrical generation system. Unfortunately, the wind stops blowing sometimes. Which means water pools created by Columbia River dams will have to be drawn down in order to meet electrical demand. Sometimes this wind failure lasts for a week or more. If the wind failed in other parts of the country where there is less hydroelectric generation capacity, relying on conventional grid resources would mean relying on fossil fuel generated electricity.

Since wind tends to displace natural gas fired generators first, it means very limited effect on grid CO2 output. As wind penetration rises, the cost of electricity rises as well. As wind penetration rises, the challenge of locating good wind generation facility sites becomes more and more difficult as the best sites are used first. Eventually, adding new sites means adding very little real world generation capacity. Adding new wind powered electrical generation facilities becomes more and more expensive per unit of output. Thus, continued use of the current grid system to backup wind does not offer a satisfactory and inexpensive means of shutting down the emission of greenhouse gases.

If the conventional grid offers no solution to the problem of wind in a post carbon world, are there alternative backup systems that can solve this problem? Several technologies have been proposed as offering means to backup wind. These include pump storage, compressed air storage, and batteries. Pump storage involves pumping water to the top of a mountain and storing it in a reservoir. As electrical demand rises, the water can be released back down the mountain to run through an electrical turbine at the mountain base. The water can be transferred between two pools, one at the base of the mountain and the other at the top of the mountain, however water evaporates from the pool therefore new water has to be added to the bottom pool. A huge amount of water would be required to provide backup electrical generating capacity to wind in the United States.

Water is not a land efficient energy source. The Tennessee Valley Authority (TVA) has dammed virtually all of the rivers that flow through the Tennessee Valley. They allow their water to flow through turbines to generate electricity. These dams produce together about five percent of the electricity generated by TVA. In order to backup wind generation virtually every mountain top in Tennessee would have to leveled and turned into a lake. This would not entirely please conservationists and environmentalists. In addition, the waters of Tennessee’s rivers are committed to a variety of uses including navigation, recreation, wildlife preservation, and household water. The pump storage approach would draw water from all of these commitments and utilize it to generate electricity. Because water evaporates from lake surfaces, the amount of water that the system discharges would be significantly less than the amount of water that currently flows through the river. If enough reservoirs were built, evaporation would greatly diminish the flow of water from the Tennessee River so that by the time the river reaches it’s mouth, very little water would be released into the Ohio River. Thus, pump storage does not offer a suitable backup for wind generated electricity.

Compressed air storage is a second backup scheme proposed by wind advocates. In a compressed air storage system, air is drawn into an underground chamber under pressure. When the wind is blowing, but consumers do not want the electricity generated, then the electricity is used to pump air into a storage chamber. At the time that wind is not blowing, the air is released through turbines which then power generators. There is a major problem with the compressed air storage approach. Compressed air pumped into underground chambers heats up. As the air comes into contact with the walls of the chamber, some of the heat is released into the walls of the chamber and from the walls of the chamber into the earth. When the air is discharged, it expands and as it expands, its temperature drops. Humidity in the air freezes as the air chills. As the air blows through turbines. ice particles are blown along with the air. The turbines are struck by the ice particles and are damaged by them. Think of the compressed air system as a heat pump which chills the air to be discharged. The loss of heat in the stored air is an inefficiency that cost us forty percent of the electrical energy used to pump the air into the underground chamber. In order to increase the amount of energy into the exiting air and melt the ice particles, natural gas is burned in the air stream. This does increase the generating power of the system, but also leads us back to the problem of CO2 discharge. Thus, compressed air storage is expensive, inefficient, and not an entirely useful decarbonation tool.

Finally, wind advocates note batteries as the third backup technology, but current battery technology would be too expensive and otherwise unsatisfactory for a wind backup technology even when significant advances in battery technology are factored in. The battery backup picture does not look promising.

Highly efficient batteries are expensive, while inexpensive batteries are not efficient. For example, lead acid batteries i.e. batteries used in cars are heavy, that is, they use lots of material, but they hold a relatively small charge especially when their size is considered. It is certainly conceivable that the efficiency of lead acid batteries can be increased in the future, but even if they are ten times more efficient they still would be heavy and require a considerable amount of material. Lead batteries also do not have long useful lives and must be replaced every few years. Lead batteries even if made ten times more efficient would not be satisfactory power sources for automobiles or trucks.

High temperature batteries may weigh less and have longer lives, but like lead batteries, they may not be satisfactory energy sources. It remains to be seen whether high temperature batteries can be made efficient enough to serve as backup to wind generated electricity, but I am not going to put my money on it yet. At any rate, high temperature batteries are probably going to be quite expensive compared to nuclear sources.

Although lithium batteries are useful for small mobile devices, it is doubtful that they would be equally useful for large scale backup of wind generated electricity because of their cost. Lithium batteries are relatively lightweight, but improving their efficiency is proving challenging.

Are there any technologies that I have not mentioned that could backup wind generators? Some time ago, on “Nuclear Green” I offered a brief study on the use of Molten Salt Reactors as backup for wind. Molten Salt Reactors would seem to offer a possible route to solving all of the problems associated with wind backup, but they offer a problem as wind backup, namely that Molten Salt technology would not simply function as a wind backup, but as a wind replacement as well. Therefore, if you start building large numbers of Molten Salt Reactors there would be no need for wind generators which are not very useful to begin with.

Charles Barton's picture
Thank Charles for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member
Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Kevon Martis's picture
Kevon Martis on May 28, 2013

Absent practical storage, it is very difficult to compare the cost of wind generation directly to dispatchable generation, as Dr. Paul Joskow at MIT has noted: 

“Economic evaluations of alternative electric generating technologies typically rely on comparisons between their expected life-cycle production costs per unit of electricity supplied. The standard lifecycle cost metric utilized is the “levelized cost” per MWh supplied. This paper demonstrates that this metric is inappropriate for comparing intermittent generating technologies like wind and solar with dispatchable generating technologies like nuclear, gas combined cycle, and coal. Levelized cost comparisons are a misleading metric for comparing intermittent and dispatchable generating technologies because they fail to take into account differences in the production profiles of intermittent and dispatchable generating technologies and the associated large variations in the market value of the electricity they supply. Levelized cost comparisons overvalue intermittent generating technologies compared to dispatchable base load generating technologies. These comparisons also typically overvalue wind generating technologies compared to solar generating technologies. Integrating differences in production profiles, the associated variations in the market value of the electricity at the times it is supplied, and the expected life-cycle costs associated with different generating technologies is necessary to provide meaningful economic comparisons between them. This market-based framework also has implications for the appropriate design of procurement auctions created to implement renewable energy procurement mandates, the efficient structure of production tax credits for renewable energy, incentives for and the evaluation of electricity storage technologies and the evaluation of the additional costs of integrating intermittent generation into electric power networks.”

To make a meaningful LCOE comp between wind, one must add the cost of the shadow generation necessary to make wind dispatchable, which Taylor and Tanton have done here:

I am sure neither of these studies is news to regular readers here, so I offer them as a service to new readers.

Kevon Martis




I K's picture
I K on May 28, 2013

An unlikely global grid gets rid of storage problems anything else is an impossible fantasy.

One alternative I came up with was to feed concentrated solar via many fibre optics into a rock formation or a massive lump of iron. once it is at a hot temp use it as your heat source to boil water to generate electricity at around 38 percent efficency (considerably higher than PV directly)

Not only do you go from highly variable solar to potentially base load output but you could even use it for load following

John Miller's picture
John Miller on May 28, 2013

Charles, the bottom line is that for power grids to supply uninterruptable, on-demand and reliable power to its customers, essentially 100% backup natural gas intermediate/peaking power is most often required.  Another problematic issue is that state-of-art NGCC generator’s efficiency varies with rate/design capacity.  When the wind conditions allow increased wind power generation, the backup-peaking NGCC power plants generation capacity is normally reduced to less than optimal design-max. efficiency rates.  This normally increases the level of natural gas fuel consumption and associated carbon emissions per KWH.


As you state, it’s a myth that any intermediate wind (or solar) power can displace baseload coal.  Until industrial scale power storage becomes a reality, only natural gas, nuclear, hydropower/pumped storage and geothermal power are currently able and available to displace significant baseload coal.

Nathan Wilson's picture
Nathan Wilson on May 31, 2013

The other major shortcoming of wind with fossil backup is that ordinarily the fossil backup produces the majority of the combined output.

As discussed, there is no storage technology on the horizon that can cost effectively remedy this.  However, a large dispatchable load, such as fuel synthesis, could (H2 or NH3).

I did some calculations based on Jacobson’s data for smoothing with19 wind farms, and got an interesting result.  If the instantaneous power output of the farms is split so that all power up to a certain limit goes to the grid and the remainder goes to dispatchable load, and if that limit is adjusted so that on average equal parts go to each, then the fraction that goes to the grid actually has an impressive 90% capacity factor (with a nameplate output of 24% of the total) and the dispatchable load sees a capacity factor of 32%.  

Using the latest EIA estimate for the cost of wind, we have  $0.0866/kWh, and a capital cost of $2438/kW.  I assume a cost of $1/W and efficiency of 60% for fuel synthesis; this gives a fuel cost of $0.0866/kWh*(1000+2438)/2438/.60= $0.204/kWh, with 34.4 kWh/gge (gallon of gas energy equivalent), that’s $7.02/gge (fuel from wind power, before transportation and retail markup).

So this is not cost effective in the US at today’s wind power prices.  However, it could be if the grid power sales are used to subsidize the dispatchable load power sales (e.g. if grid power was sold for around $0.13/kWh, then the fuel plant could buy the remaining power for $0.04/kWh the synthetic fuel could be around $3.5/gge).

It is a little interesting; it is another way of showing that the variability of wind power incurs a cost to match demand, but it beats the cost of power smoothing with batteries.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »