July 16, 2006
Future Of Nuclear Power Surveyed
Writing for the New York Times Magazine Jon Gertner has written an excellent article surveying the state of the nuclear power industry and signs that new nuclear power plant construction will commence in less than 10 years. For anyone seriously interested in energy policy I urge you to read this long article in full.
Thanks partly to large government incentives and to market forces that have pushed the price of other electric plant fuels (especially natural gas) to historic heights, the prospect of starting a new nuclear reactor in this country for the first time in 30 years has become increasingly likely. By early summer a dozen utilities around the country had informed the U.S. Nuclear Regulatory Commission, which oversees all civilian nuclear activity in this country, that they were interested in building 18 new facilities, nearly all of which would be sited next to existing nuclear reactors.
The electric power industry is taking nuclear power very seriously.
The sooner and higher the carbon taxes come the more attractive nuclear power will become. But nuclear power plants take several years from beginning of planning to first power production. So the electric power industry must make multi-billion dollar guesses about the state of emissions regulations in future decades.
Moreover, what makes the choice of fuels such a knotty problem is that something that is cheap now, like coal, may not be so cheap in 10 years. This isn’t because we’re running out; we probably have at least a century’s worth of coal reserves in the United States alone. But if the government were to impose a tax or a cap on carbon emissions, something that almost everyone I spoke with in the energy industry believes is inevitable, or if new laws mandate that coal plants must adopt more expensive technologies to burn the coal cleaner — or to “sequester” the carbon-dioxide byproducts underground — the financial equation will change: a kilowatt-hour generated by coal suddenly becomes more expensive. There are other contingencies at play, too: fuels, like natural gas, could experience a supply interruption that leads to enormous price spikes. As for the hope that wind and solar power will generate large amounts of clean, affordable electricity in the near future? I encountered great skepticism inside and outside the utility companies. “Maybe in 40 years,” Paul Joskow, of M.I.T., told me.
Looking out over decades the electric power industry also has to guess about the rate of technological advances in wind, photovoltaics, and other non-fossil fuels based alternatives for generating electric power. Nuclear power plants do not pay back their capital costs for decades. So the cost of competing electric power sources 20, 30, and 40 years hence have to figure into decisions about whether to start building nuclear power plants today.
If carbon taxes become a major cost then that might drive the cost of coal electric well above nuclear. But another risk that nuclear faces is the potential for innovations that lower the cost of carbon extraction when burning coal. So even if evidence of global warming from carbon dioxide becomes very strong that's not a guarantee that nuclear will become the lowest cost electric power source.
Some see construction of new electric plants as avoidable by use of technologies that greatly improve energy efficiency.
There is a counterargument to building large new power plants. One view — voiced most forcefully, perhaps, by Amory Lovins, a physicist who runs Rocky Mountain Institute, which advises corporations and utilities on energy efficiency — is that we don’t need to increase our electrical supply. We need to decrease demand by rewarding utilities for getting customers to reduce electricity use by, say, updating their appliances, furnaces and lighting. Lovins, a longtime critic of nuclear power, contends that it remains financially uncompetitive and that the 30-year absence of new plants is proof that the market has rejected nuclear power as a viable technology. When we spoke about whether utilities need to build more big generating plants in this country, he told me no — not now, not in 15 years, not even after that. “I think if you do,” he remarked, “your shareholders and ratepayers will be asking awkward questions that you would really rather not want to answer.” Yet the concern, even among Lovins’s admirers, is that if he is mistaken — that is, if either his estimates on efficiencies can’t accommodate population and industrial growth, or because what is possible in principle for energy efficiency is not possible in the real world — then the utilities will require an alternative plan. And that would entail more supply, likely meaning more big base-load plants (whether they rely on uranium, gas or coal) as well as large investments in renewable sources like wind and solar power.
My view: Only a big rise in the cost of electricity will substantially reduce per capita electric usage. Rising living standards will make electric power more affordable. People will find more ways to use electricity if they can afford it. They'll get bigger televisions, faster computers, run air conditioners to a lower temperature, and so on. Sure, technological advances will improve energy efficiency. But when energy efficiency rises part of the response is to do more of whatever is now more efficient to do. For example, make cars more fuel efficient and people will drive more miles and get bigger cars. Also, other technological advances will raise incomes and so people will buy more gadgets that use more power. This is especially the case in the industrializing countries, most notably China. So I do not see conservation as a solution. Increases in energy efficiency can raise living standards. But it is unlikely they will stop the increase in demand for energy.
Westinghouse with their AP1000 design and other nuclear reactor designers claim they've gotten their costs down far enough to be competitive. But read the full article for reasons behind the uncertainty about their cost estimates.
But the appeal of the AP1000 remains doubtful, even as 11 utilities, including the Southern Company, have expressed interest in the design. Westinghouse maintained to me that the cost will ultimately be somewhere between $1.4 billion and $1.9 billion. “We’re negotiating contracts,” Dan Lipman, who runs the new-power-plant division at Westinghouse, told me over lunch at the company cafeteria. “We’re well beyond the should-we-do-nuclear phase. It’s now a matter of, How should we do it?” So I asked Lipman what it would mean to actually cut a deal with a utility for a new plant, the first in 30 years. Would it happen a year from now? Two years? “If your definition of a deal is, when do you first start getting money, then that could happen very soon,” he said. “I look for that this year, with big money committed after licensing by the N.R.C.” From his continuing negotiations, Lipman said, it’s clear that his customers are interested in “off-ramps”: clauses in the contracts that allow them to bow out if they hit an unexpected financial or construction snag.
The industry has a number of advantages that it did not have during the last wave of nuclear reactor construction. First off, computers can track design changes, automate communications, manage order tracking and parts inventory, and otherwise manage the design and construction process. Computers have made large construction projects more manageable. Also, the industry is going to use standard designs this time around. So each new plant won't have a large assortment of unique problems to work out. The industry has even formed a consortium for constructing the first reactors that use the new designs. This consortium will allow a great deal of sharing of regulatory forms and knowledge about costs and technological problems encountered during the construction process.
The second cushion is the creation of an industry consortium, called NuStart, to test the licensing process. NuStart is filing several applications for nuclear plants, on behalf of its members, with the Nuclear Regulatory Commission. These applications — for the Grand Gulf plant in Mississippi and the Bellefonte site in Alabama — have preceded all others and may end up being built first. One goal of NuStart is to prove to Wall Street that utilities can get a license in a timely manner. Another goal is to establish a way for the industry to pool risk and information. If NuStart’s construction-and-operating applications for its two sites are approved, in other words, any utility in the consortium (including Entergy, Exelon and Southern Company) can copy huge parts of the approved application for its own use, thus saving time and money.
A lot is going to hinge on the costs of building the initial reactors that test out the regulatory process and the new designs. We will find out from the costs and schedules of those reactors how far the nuclear power industry has progressed toward making nuclear power competitive. The big wild card for nuclear power is global warming. If the global warming threat starts looking serious enough to justify large carbon taxes then I expect a huge shift toward nuclear for new electric power plants.
Again, read the full article if you are seriously interested in the energy debate.
The sad thing about this article is that it focused exclusively on American power plant providers as if they were the only game in town or even the leaders in the industry. They are not. The French are the real leaders in nuclear power plant construction these days. While the boys in Pittburg have been making their pretty designs, the French have been building plants. Westinghouse is expecting the NRC to give full approval to its AP1000 design. The French are actually building a plant in Finland using their competitive EPR design.
Don't think that the French are distant players in the American market. Through their American subsidiary companies they are the real leaders in nuclear reactor services and a major supplier of fuel to American plants. They have also recently made major investments in the specialised heavy steel fabrication needed for reactor vessel construction.
The fact is that over the last 30 years American companies have been shrinking their business in nuclear power because the market here has contracted. And the French have been buying those assets up as they were sold. They are also recruiting and hiring nuclear engineers. Special plans are being put into place to utilize the skills of the current generation of engineers as they now approach retirement. When the time comes for ramping up the American companies will be playing catch-up ball.
The excitement over new nuclear reactor designs starting up during the next decade may yet need be altered by wildcard fusion advances that use magnetohydrodynamic or electrodynamic conversion of plasma to electrical power. The lower waste heat from such systems would result in far smaller reactors hence low capitalization compared to thermal nuclear fusion.
A breakthrough announced recently by a founder of the US's nuclear fusion program, Robert Bussard (yes, the Bussard whose interstellar fusion ramjet has become known among scifi fans), may be the kind of thing that comes out of left field. He provided some details of his recent fusion breakthrough at:
The money quote:
"...a fusion rate of about 1E9 /sec. This works our to be about 100,000 x higher than the data of Hirsch/Farnsworth at similar well depth and drive conditions."
Indeed, he claims the physics model is now well-enough validated that the next rational step is to follow the scaling law (fusion output = scale^7) to an energy break-even device costing approximately $200M buildable as a demonstrator within the next 5 years.
Well, there was another money quote concerning the loss of funding from the Navy for this project:
"It was not a cutoff of OUR funding, but the entire Navy Energy Program was cut to zero in FY 2006, and we were a part of this cut. The funds were clearly needed for the more important War in Iraq."
Erratum: "thermal nuclear fusion" -> "thermal nuclear fission"
I suppose I wouldn't expect a big emphasis on renewables in an article on nuclear, but I was disappointed by the very brief look at wind & solar.
The one fact he gave was out of date: wind is about 1% of US kwhrs, not less than .5%.
The expert he quoted has a lot of experience with energy analysis, but most of it is either very old or related to nuclear. I see no sign that Joskow has ever seriously looked at renewables, and especially not lately. That's a big mistake, with things changing so quickly.
If you look at Amory Lovins' initial article in Foreign Affairs back around 1975, you will find a chart which maps the Hard Path and the projections of energy usage under that regime and the Soft Path and the much lower projections under that concept. These days, Amory shows that chart and then updates it. We are marginally above the projected energy usage under the Soft Path regime. We followed the Soft Path whether we liked it or not.
I would suggest that, in terms of projected electricity usage, Amory Lovins has a track record which means he is probably more likely to be correct than those who proclaim that only ever-increasing energy usage is the wave of the future.
You can read Amory's recent thinking on nuclear power at http://www.rmi.org. He believes that centralized power is going to be replaced by decentralized power, that nuclear will be outproduced and underpriced by what he calls "a herd of mice," cheap, decentralized power in a variety of forms - renewables, cogeneration, efficiency....
Concerning Areva, Kendall is right on several points. Keep in mind that Areva's American subsidiaries are really well-known American companies that have been acquired and merged over the years with the French company, whose name is only about five or six years old. So, what is now called Areva in the US once had names like Babcock and Wilcox and Exxon Nuclear. The result is a combination of American know-how in nuclear combined with the what has been learned by the Europeans (most importantly, France and Germany) over the last several decades in programs -- at least, in the case of France -- that are still ongoing and strong, unlike the industry in the US, which has languished since the late 70's. As Kendall points out, of all of the new designs that are being considered in the US, only Areva's EPR is currently being built anywhere in the world.
It is funny, but to call Areva in the US French and, by the same token, to call Westinghouse American is to ignore that, until recently, Westinghouse was owned by the British. Now, they're owned by the Japanese. Why not call them Toshiba?
Concerning fusion, that is very fascinating, but aside from specialists and sci-fi fans, it isn't particularly important. Even assuming that a major breakthrough is eminent, the amount of time required before this technology can be applied on a scale that could make even a minor impact is quite long. When you consider how long it requires for a nuclear power plant (a technology with which we have over 50 years of experience) to be designed, reviewed, certified, sited, and approved, any new technology, however revolutionary, will certainly require more time. All of this means that it will be too late to affect the current events surrounding (conventional fission) nuclear power -- either nuclear will be well into a comeback by then or it will be a dead technology.
Concerning renewables, I thought that the article gave a very realistic coverage of technologies such as wind and solar. As John Holdren states, "we'll need them all." As for how much they will be used, that depends on how economical they are (and how long the production tax credits last).
"realistic coverage of technologies such as wind and solar"???
Here's what he said:
"Wind and solar power make up less than one-half of 1 percent of what we use on a typical day. In part because the wind doesn’t always blow and the sun doesn’t always shine (and in part because wind turbines and solar cells are expensive to build) neither technology is yet good enough to generate large, reliable quantities of inexpensive electricity, or what utility companies call “base load” power. "
This is the kind of thing that I would expect from a nuclear advocate, not an evenhanded analysis. The .5% is out of date: it's 1%, and doubling every 2 years. "the sun doesn’t always shine" is very misleading: the fact is that solar production correlates better with consumption than nuclear. Don't forget that current consumption curves have been adjusted for the convenience of utilities: a lot of commercial/industrial electrical use is at night due to peak demand charges.
"wind turbines and solar cells are expensive to build". No, wind is not expensive. In good areas it can be $.03 per kwhr. I would argue that if you include all costs that wind is now cheaper than coal or nuclear.
"base load". Again, don't forget that what utilities refer to as base load isn't really a natural baseload: much of nighttime demand is only there due to peak demand charges.
Utilities are notoriously conservative. Wind and solar require innovation, and they see that as risky. I would argue that it's not: the innovations required require engineering and management, not new technology. Some utilities get that, which is why wind and solar are indeed growing so fast: doubling every 2 years. That rate of growth could get wind to 15% of demand in 8-10 years, with solar only 6-8 years behind.
“Maybe in 40 years,” Paul Joskow, of M.I.T., told me.
40 years is ridiculous. He clearly is out of touch, or biased.
"For example, make cars more fuel efficient and people will drive more miles and get bigger cars."
Doubtless this is true - but there is a limit as to how far people want to drive, and how big they want their cars to be. I'm at my limit for how far I'm willing to commute - many days I think I'm past the limit. And if all cars got the same fuel economy, I still don't think I'd want anything bigger than a midsize - too much of a pain to parallel park. But it might have a bit more horsepower. :)
On the other hand, your statement doesn't take capital cost into consideration - make cars more fuel efficient and they'll probably be more expensive (at least assuming they're not smaller or weaker) - this would discourage people from going to bigger cars as well.
Regarding peak and base load:
1) It is my understanding that the base demand has grown faster than the peak demand in recent years. Reason? Massive computer server farms and the internet. They run 24x7.
2) We need dynamic pricing. This would work against natural gas and for nuclear, solar, and wind.
3) Dynamic pricing would accelerate the development of technologies and organizational changes that would allow demand to shift to when prices are lower.
4) Better battery technology for cars will (and by "will" I mean I really do expect electric cars to take off) probably cause demand to grow more when electric prices are lower - though only if we move to dynamic pricing.
5) Solar does not work so well during short winter days. Keep in mind that many predictions about solar's future competitiveness often use SoCal as the example. Er, how about Minnesota or Massachusettes or Seattle?
6) Wind does not work so well in the US Southeast which has the weakest winds in the United States. But advances in transmission line technology would reduce that problem. Not sure when we will get those advances though.
As for technology forecasting: Someone will eventually be right when they predict solar's imminent competitiveness. But when?
1) Cars will become longer lasting. They've been doing that for decades. That lowers cost.
2) Computers that do the driving for you will eventually make long road trips easier to do.
3) Increased buying power translates into people buying bigger cars and SUVs.
4) Increased buying power also translates into increased demand for air travel which consumes even more fossil fuels.
5) Increased buying power also translates into building larger houses and using more energy to heat and cool them.
The US economy has become steadily more efficient in terms of energy used per $ of GDP. I've seen graphs on this and it is impressive. But the total GDP has risen faster than the ratio of energy to $ of GDP. I expect that trend to continue unless energy costs go much higher. But how can energy costs go much higher? Coal-to-Liquid puts a long term ceiling on oil costs. The high price of natural gas just shifts electric generation toward nuclear and coal.
I don't know where Nick is getting his numbers, but less than a 0.5% contribution by wind and solar combined agrees with the values that the EIA has for 2004 (the latest year for which they have posted data on their website). Go add the numbers for electricity net generation yourself.
I calculate that in 2004 wind and solar combined generated 0.37% of the total US electricity produced (net generation) that year. The electricity that they produced is only 0.40% of the total electricity consumed in 2004 (i.e., electricity actually sold to customers); however, this assumes that all of the electricity produced by wind and solar was sold to customers, which may not be the case.
Brian, 2004 is badly out of date. Wind capacity will have more than doubled from the end of 2004 to the end of 2006, from 6.7 MW to 14.5 MW. For 2005 capacity see http://www.awea.org/pubs/factsheets/Wind_Energy_An_Untapped_Resource.pdf
Note that the 24,800 gwhrs given is an average for the whole year, and some installations weren't completed until the end of the year, but they still had a 31% capacity factor. Now, add to the 9.149 GW for 2005 the 5.391 GW being installed in 2006 - see http://www.nei.org/documents/Energy%20Markets%20Report.pdf
That gives you 14.54 GW at the end of 2006, which would produce .97% of the US's 450 GW average at 30% capacity factor.
"As for technology forecasting: Someone will eventually be right when they predict solar's imminent competitiveness. But when?"
Definately not Joskow and his 40 years.
You can see solar costs declining 50% per decade since the 60s. Based on what I've read, the next ten years will likely see even better than a 50% cost reduction.
We followed the Soft Path whether we liked it or not.
The US has offshored a great deal of energy-intensive basic industry. We're still causing that energy to be consumed (in the sense of buying products whose materials embody that energy expenditure), but it isn't showing up as domestic consumption.
Nick, okay, so the figures given for wind (solar has not changed much) in the article are a year and a half old. I think, however, that your predictions for 2006 (which isn't over yet) are a bit optimistic. If you look at the historical record posted by the EIA for the five-year period 2000-2004, capacity factors for wind never reached 27%. For a couple of the years, it was around 20%. What makes you think that wind's capacity factor will hit 31% in 2006?
Anyhow, these are minor details. Your main point is taken, and this is encouraging news. I agree with John Holdren when it comes to renewables: we'll need them all.
What I still don't understand, however, is what your beef is with the article. Whether it is 0.5% or 1%, wind's contribution is still relatively small. Wind turbines don't work when the wind doesn't blow. The fact that talk of wind's capacity factor starts at 31% as optimistic and only goes down is evidence of that. When your capacity is as small as wind's was several years ago, it is easy to double capacity in two years. I don't think that it is realistic, however, to believe that this rate of growth will go on indefinitely.
One of the big factors driving wind is state-level mandates for achieving percentages of electric generated from renewable sources. California has such regulatory requirements and I've posted about this in the past.
Whether those mandates will remain in effect depends in part on what they will do to total electric costs as the percentages increase. For example, from the comments of a previous post:
The California state government requires public utilities in California to increase their use of renewable sources.
SAN FRANCISCO, April 27 /PRNewswire-FirstCall/ -- Pacific Gas and Electric Company announced it has signed up an additional 158 megawatts (MW) of clean, renewable wind energy resources to help meet its customers' future electricity needs. All together, generation from these new renewable energy resources will add enough generation to supply nearly 120,000 PG and E customers.
The 2005 RPS solicitation will be PG and E's third competitive solicitation for renewable energy since 2002. Since then, it has entered into contracts for 368 MW of renewable power from wind, geothermal, biomass, and hydro resources, including the three agreements being submitted today to the CPUC.
California's RPS Program requires each utility to increase its procurement of eligible renewable generating resources by 1% of load per year to achieve a 20% renewables goal. The RPS Program was passed by the Legislature and is managed by California's Public Utilities Commission and Energy Commission.
So we are going to see a huge growth in wind power in California. But what is the real cost of these deals?
What percentage of total increase in wind capacity is due to state regulatory mandates? How much of it is due to subsidies?
California has to achieve 20% of electric from renewables by 2017. So that is going to drive a lot of wind capacity increase.
Check out United States wind maps and notice how the Southeast is wind poor. That is from 1987. Later state-level wind maps are more accurate and detailed. I can not at the moment find the later national level wind map that I've previously encountered. Would appreciate a link if anyone can find it.
It was about time that the Times did a piece on nuclear power. They are always luke warm at best. The piece was balanced, but no mention was made of newer type reactors which hold promise for hydrogen, proliferation resistance, minimizing radioactive waste, lower cost and greater safety. The reporter focused on industry. Industry can't risk the cutting edge. Only the State can do the necessary R & D. nomination in 2008.
"If you look at the historical record posted by the EIA for the five-year period 2000-2004, capacity factors for wind never reached 27%. For a couple of the years, it was around 20%. What makes you think that wind's capacity factor will hit 31% in 2006?"
Sigh. The EIA does a very bad job of presenting renewables info. There's a lot of weekly and monthly oil & gas info - why are renewables so out of date?
Anyway, Brian, have you tried charting the data? You'll notice that capacity factors trend upward strongly. You'll also notice an odd pattern of up and down. Well, the kwhr data is an average from the whole year, and the capacity figure is from the end of the year, so the capacity figure is always understated. The extent of the error depends on growth: in the big years, when the credit was in effect, the error is large. As the base grows, the error trends downward. 29 to 31% seems to be a reasonable range: the AWEA figures for 2004 and 2005 come out to 29% and 31%, respectively. I tend to use 30% as a general figure.
Randall, California gets a big portion of it's power from natural gas, perhaps 50%. As wind is much cheaper than nat gas, and the gap will widen, wind will save CA a lot of $.
Brian, wind will contine to get cheaper, with larger turbines, manufacturing economy of scale and other cost reductions - for instance, the blades are a big cost component, and a lot of work is going into reducing their labor cost (the overwhelming majority of the cost of wind installations). GE just opened a plant in China, which will provide them with pretty cheap labor.
Wind at about 5% per kwhr is already cheaper then nat gas, and is pretty close to coal (cheaper if you include even a portion of external costs like pollution, GW, occupational health, etc). Wind will be very easy to expand, as there are a lot of farmers just dying to get a wind farm or expand the one they already have, and manufacturing of wind turbines is a pretty straightforward thing to expand. Not overnight, which is why wind developers are currently limited by the turbines they can lay their hands on, but pretty straightforward.
Wind is neat, but it isn't baseload like and it isn't peaking. Its whenever it wants. It would mix well with a hydroelectric plant that can store head during times of plenty and release it during times of want...
Nuclear is necissary, while wind is convenient. I myself favor a ban on all municipal coal power plant construction because we have nuclear avaliable for providing baseload power. This keeps the coal free for nice stuff like conversion to diesel fuel.
Wind will also mix well with the storage provided by electric personal transportation.
"Nuclear power plants do not pay back their capital costs for decades. So the cost of competing electric power sources 20, 30, and 40 years hence have to figure into decisions about whether to start building nuclear power plants today."
Well that is optimistic because US Army Corp of Engineers in report titled "Energy Trends and Their Implications for U.S. Army Installations" puts TOTAL global reserves of uranium at 33 to 43 years at current comsumption - nevermind building more stations to consume the resource.
There's also a whole body of growing evidence out there now with CO2 emmissions that as the rich ore bodies are used up and lower grade ores are used you get to a point where you may as well burn the fossil fuels directly to generate electricity to stay ahead. If in doubt have a read of this research from two scientists who worked in the industry http://www.stormsmith.nl/
Here's a taste of their findings;
"The production of electricity by nuclear reactors, as long as rich uranium ores are still available, leads to considerably less CO2-emission than does the use of fossil fuels for the purpose. In the course of time, as the rich ores become exhausted and poorer and poorer ores are perforce used, continuing use of nuclear reactors for electricity generation will finally result in the production of more CO2 than if fossil fuels were to be burned directly. In this revision we have given, in Chapter 2, a complete, up-to-date overview of all of the known or presumed uranium ore bodies, and the amount of net energy that burning them would deliver. In our original website we had estimated that using these resources to exhaustion would provide only three years of electrical energy at the present world rate of electricity use. In this revision, this estimate has been raised to about four years, partly because the data on the reserves is much more complete than it was a few years ago.
These are the salient results of the research reported on this website.The technical background of these is presented in an introduction and five chapters which can o be downloaded for study from the table of contents below.
The remarkable conclusion of this research might prompt one to ask how it was possible that an entire energy industry was built up when in fact, using all available resources, it could only provide such a small amount of electrical power. There are two reasons that may explain this remarkable fact, one arising from unrealistic, and easy to refute, assumptions concerning the (energetic) yield of nuclear power and the other on an (up to the present) unjustified technological optimism...."
To my mind the nuclear industry selling us a lemon with the global warming argument.
Wind turbines average the wind. Which is always blowing somewhere.
Given turbines spread out over a lage enough area (about 50 mile radius) wind can provide about 20% of rated peak capacity as base load.