May 13, 2008
Nuclear Plant Costs Double To Quadruple Previous Estimates

Rebecca Smith of the Wall Street Journal reports on the thinking of big new nuclear power plant buyers. Nuclear power is seen to cost double to quadruple previous estimates.

A new generation of nuclear power plants is on the drawing boards in the U.S., but the projected cost is causing some sticker shock: $5 billion to $12 billion a plant, double to quadruple earlier rough estimates.

What became of all the efforts to develop newer lower cost designs? Are these cost increases due to the Asian demand for commodities driving up the cost of iron ore, concrete, and other construction materials?

The latest projections follow months of tough negotiations between utility companies and key suppliers, and suggest efforts to control costs are proving elusive. Estimates released in recent weeks by experienced nuclear operators -- NRG Energy Inc., Progress Energy Inc., Exelon Corp., Southern Co. and FPL Group Inc. -- "have blown by our highest estimate" of costs computed just eight months ago, said Jim Hempstead, a senior credit officer at Moody's Investors Service credit-rating agency in New York.

Oil costs a lot. Then coal and natural gas go up in price in response. Optimists think we still have nuclear as another substitute. But its costs have skyrocketed as well. The inflationary pressures seem inescapable. Photovoltaics might be our only hope for cheap future energy. It doesn't do well as a baseline energy source. But PV makers are coming up with innovations that lower costs.

The Congressional Budget Office still hopes for eventual cheaper nukes.

The Congressional Budget Office just finished a rosy-glasses report on nuclear economics. Even while acknowledging that historical costs for nuclear plants always doubled or tripled their initial estimates, the CBO took heart from promises made by manufacturers of next-generation reactors and a single on-time and on-budget project in Japan to project cheaper nuclear construction costs in the future.

Whether nuclear costs can come down depends on why they are high in the first place. Does anyone know where these cost increases come from? Can innovations in speed of construction yield big cost savings?

Back 8 months ago Moody's was already pretty pessimistic on nuclear costs. But they've become even more pessimistic.

While utilities reportedly have priced the cost of a kilowatt of nuclear power at $3,000 to $4,000, Moody's Investors Services said in October that a more realistic price would be $5,000 to $6,000. That puts the cost of a 1,500-megawatt nuclear plant at about $9 billion, according to reports.

The chairman of a German electric power plant operator puts the cost of new nukes at almost $10 billion.

Wulf Bernotat, chairman and chief executive of E.ON, the German energy giant that owns Powergen, has told The Times that the cost per plant could be as high as €6 billion (£4.8 billion) - nearly double the Government's latest £2.8 billion estimate.

Still, getting your own national nuclear power plant has become all the rage.

VIENNA -- At least 40 developing countries from the Persian Gulf region to Latin America have recently approached U.N. officials here to signal interest in starting nuclear power programs, a trend that concerned proliferation experts say could provide the building blocks of nuclear arsenals in some of those nations.

Share |      Randall Parker, 2008 May 13 10:09 PM  Energy Nuclear

Engineer-Poet said at May 13, 2008 11:29 PM:

The big winner from this appears to be wind.  At under $2000/kW and 18 months from go-ahead to production, it's the option with both the least cost per kWh and least risk.

Pete said at May 14, 2008 3:46 AM:

uh - oh do I have to tell an engineer that wind turbines' output over time is only between 25 and 30% of their rated power? Compare this to nukes with over 80% and your quoted figure gets multiplied by 3....

Or, does your figure imply that you can get wind turbines which only cost 650 bucks per rated kW?


allan said at May 14, 2008 6:04 AM:

If the commodities for building nuclear is going up then isn't the commodities for all going up? Coal plants use iron, steel, and concrete too. I don't think the price increases will be limited to nuclear production.

JMG3Y said at May 14, 2008 7:00 AM:

Is this primarily the monetization of increasingly "scarce" (decreasing EROEI) embodied energy contained in the commodities? Yes, there are other factors, such as the declining concentration of good source material (e.g., iron ore deposits) and the bottlenecks of regional processing and transport capacity but are these secondary? If so, it seems to me that projecting and then incurring future costs contains a very nasty lagged feedback component that is very hard to get ahead of, making contracting for projects with a significant build time a very risky business at the moment. For an example from the food side, the lag is at least a year. Fertilizer that was $400 per unit the last planting season is over $1000 per unit this season but products containing this grain crop won't hit shelves for another year. Unlike the stock market, supply and demand responses of sellers and buyers in the interlocking commodity supply chains are separated by a cascade of lagged feedback loops that aren't well captured by current economic models, either implicit models in people's heads or explicit ones in computers.

Brian Wang said at May 14, 2008 9:36 AM:

“Although the PCCI has been on an upward trend since 2000, a surge that began in 2005 has pushed costs up 76 percent in the past three years,” according to Scott. “The latest increases have been driven by continued high activity levels globally, especially for nuclear plants, with continued tightness in the equipment and engineering markets, as well as historically high levels for raw materials.” Excluding nuclear plants, costs have risen 79 percent since 2000, she noted

All power costs are up

From the Sept 2007 Batelle, Edison foundation report. Nuclear power costs have been staying more stable than other kinds of energy. Nuclear is the lowest line, which means prices moved the least.

China contracted for $5.3 billion for four AP1000 in 2007. Construction started. The contract was for $1,130/kw

four AP1000 in the USA, contract in 2008 $13.7 billion, $2927/kw

The overnight costs quoted in the FPL are from $6.7 billion, or $2,444/kW, to $9.8 billion, or $3,582/kW.

In March 2008 Progress Energy published estimates for building two new Westinghouse AP1000 units on a greenfield site in Florida. If built within 18 months of each other, overnight capital cost for the first would be $5144 per kilowatt and the second $3376/kW. The costs include land, licence application, initial core load, cooling towers, owner's costs, insurance and taxes, escalation and contingencies. This would appear to be a wider scope for overnight capital cost than usual. Interest adds about one third to the combined figure - $3.2 billion, and infrastructure - notably 320 km of transmission lines - about another $3 billion. The units are expected on line in 2016 and 2017 and are expected to save customers some $930 million per year relative to natural gas-fired generation.

Randall Parker said at May 14, 2008 10:17 AM:


I wonder the same thing about the energy component of nuclear costs. Supposedly nukes have a very high EROEI (energy return on energy invested). If those published reports about nuclear EROEI are true you would not expect rising oil prices to boost the costs of a nuclear plant so much.


I see wind as having less potential for cost reduction. Yes, in the short term it is the winner. But if only one solar photovoltaic company comes in with a big price breakthrough then solar will become the low cost leader.

Brian Wang,

Yes, nuclear's cost problems are less bad than natural gas's cost problems. Natural gas is going to get much worse as people shift to using natural gas as a substitute for oil. In dollars per million BTUs natural gas seems to stay at around half oil's cost. I'm seriously wondering when natural gas-powered cars are going to start making sense. You can special order them from Honda for many of their models btw.

Randall Parker said at May 14, 2008 10:20 AM:


Nuclear power plants are the most capital intensive as compared to natural gas and coal plants with the same rated capacity. So run-ups in commodity prices impact nuclear power most heavily.

Also, nuclear power plants take the longest time to build. This increases costs because those years where partially built plants sit there getting built have big interest rate costs. You spend the first billion dollars or euro on a power plant you pay interest on it for a few years until the rest of the plant gets built and it finally starts generating revenue to cover its interest costs.

jp straley said at May 14, 2008 10:46 AM:

Doesnt anyone think that building new types of nukes, namely thorium/U or pebble bed reactors, whose thermal efficiency (esp. in the case of pebbles and Molten Salt rxors) are much higher than BWRs and their ilk, should not be the case to compare? So much of the cost of BWR is safety, and with intrinsically more safe designs, some economy to obtain the same level of safety is possible. Liekwise with Th/U, burnup is better, disposal is simpler, and the nuclear ash is not a feedstock for bomb-making. Valuable benefits, eh?

Of course, the research to provide a good engineering basis must be done now--some on the Molten Salt already done thirty years ago---still relevant, or course.

The future will not allow the same types of projects--we must be smarter.

JP Straley

Kent Gatewood said at May 14, 2008 11:17 AM:

Mr. Parker is there some reason trucks arn't running on natural gas now?

Thanks, Kent

Randall Parker said at May 14, 2008 11:56 AM:


Range is one big advantage of diesel. But UPS is going for both diesel hybrids and CNG vehicles in its fleet.

UPS (NYSE: UPS) announced it has ordered 200 hybrid electric vehicles (HEVs)--the largest commercial order of such trucks by any company--in addition to another 300 Compressed Natural Gas (CNG) vehicles for its U.S. delivery fleet.

With fuel costs starting to out pace labor costs for truck operation this creates incentives to spend more labor on more frequent fill-up times since CNG is cheaper.

Idling trucks is another way to cut energy costs.

With the cost of a fill-up for an 18-wheeler now well over $1,000, fuel surpasses labor as the biggest cost for some trucking firms. Big companies face falling profits. Small and independent truckers, some 20 percent of the industry, face devastation. Trucking firm failures jumped to "catastrophic levels" in the first quarter of the year, reports industry analyst Donald Broughton. He estimates that 42,000 trucks, or 2.1 percent of the nation's capacity, were idled in the first quarter—with nearly 1,000 companies going bankrupt.

The older and less fuel efficient trucks will get idled first.

Hybrid diesels only make sense for delivery trucks that do a lot of stop and go. So for UPS they make a lot of sense. But out on the highway there's not much hitting of the brakes. So regenerative braking doesn't help much.

Wolf-Dog said at May 14, 2008 12:36 PM:

The new types of nukes such as the Pebble Bed Reactor, Molten Salt Reactors, were sabotaged by various groups, (government funding discontinued). The current expensive estimates are understandably based on the expensive and very clumsy pressurized water reactors that require a very costly reactor container and a very advanced water pumping system.

But let me elaborate more about the other reactors and the pros and cons:

1) The Pebble Bed Reactor is very cheap and very fast to build due to its simplicity ( it can be built in less than a year and it would surely cost far less than $1 billion per Gigawatt), and its operation is also very cheap due to its low maintenance, so that it would generate electricity at 5 cents per kWh. But on the DOWNSIDE this kind of reactor is rather inefficient for burning uranium, and the resulting waste would accumulate at least as much as the pressurized water reactors, and the world is running out of uranium. Thus although the Pebble Bed reactors should definitely be developed and quickly built (thanks to their modular design that makes it possible to build these at the factory and ship the complete pieces to be mounted at the intended sites), it must be noted that in order to reprocess the nuclear waste from Pebble Bed Reactors, new kinds of waste transmutation reactors or reprocessing plants must be built, since the whole point of the Pebble Bed reactor was to build thousands of these reactors in the world, not hundreds, and then the nuclear waste would accumulate unless it is reprocessed. Also uranium would run out in the world if thousands of inefficient reactors are built, and Pebble Bed is not efficient for fuel conservation, even though its operating expenses make it possible to yield cheap electricity at 5 cents per kWh.

2) The Molten Salt reactors (and the related integral fast breeder reactors) have the advantage of being nearly 100 times more uranium fuel efficient because they would burn the long term nuclear waste as new fuel, and this means that they would breed new fuel from the nuclear waste. Since less than 1 % of uranium is the fissile version, this means that nearly 100 % the mined uranium would become fuel for these reactors, and the waste problem would also be solved. But on the downside, unlike the Pebble Bed reactor which is very simple to develop, this molten salt reactor is more complex, and a lot more money and at least another 5-10 years would be needed to commercialize this, even with new government money, although the feasibility is demonstrated.

Now let me mention HOW the new integral fast reactor was sabotaged:

A) After UC Berkeley demonstrated the feasibility of the fast breeder integral reactors (which burn their long term nuclear waste as their own fuel), the Bill Clinton government immediately discontinued the funding. The fear of the anti-nuclear groups was that this new progress, would cause a new wave of interest in nuclear power, even though this new reactor promised to clean up the nuclear waste that the antinuclear people were against.

B) The Oil and Coal industries (which employ tens of millions of people) evidently did not like this new reactor, since it would take business away from these industries.

C) The Uranium Mining industry and the Uranium Fuel Manufacturing Industry did not like this new system because it would be 100 times more fuel efficient (requires only 1 % of the uranium fuel used in a similar pressurized water reactor with the same output), and also because its fueling would be a new system of molten salt, instead of the already standardized fuel pellets which are shipped from the fuel factory in a convenient way, and the uranium industry did not want to develop the molten salt type fuels since it is more work for them.

In other words, for once in their lives, the interests of the above-mentioned 3 mutually incompatible groups that hate each other, actually converged, and they all opposed the future research and development on advanced reactors.

K said at May 14, 2008 1:30 PM:

Nuclear construction is very risky. A big reason is the small volume; each plant is unique and you are always near the bottom on the learning curve.

Experience counts. In the US probably 95% of those involved in building our next nuclear plant will be working on their first nuclear plant. I live near Palo Verde which was finished about 1983; the senior people on that construction would be sixty to eighty years old today.

Also, politics dominates all nuclear decisions and directions can change with every election. Lenders, contractors, and utilities have to make provision for the uncertainties and it shows up as cost.

Who knows how much of these Cost Guesses - one can hardly dignify them as 'estimates' - is for risk and learning, but when numbers range from $2B to $10B the cost of materials is not likely to be the driver.

Brian mentioned an estimate for two identical Florida reactors that said the second would cost about 60% of the first. Maybe our big costs are not for the building but for the not building.

aa2 said at May 14, 2008 1:54 PM:

Basically America is like an unstable African nation whose policies can turn on a dime and leave capital investors stranded and where there is no consistency in the law. Whether its one of dozens of regulatory agencies who can block things for any reason they feel like, hundreds of courts where judges have unlimited power(a low level judge can stop a 5 billion dollar project on a whim), different politicians who can change the rules at any moment with elections going on all the time for many law making bodies.. it makes it all very risky. And that has to be built into the cost.

Long story short, imo America needs to concentrate power and control into a single strong willed person.

allan said at May 14, 2008 4:39 PM:

aa2 - I volunteer. Make me king for two years with unlimited powers and I'll have the changes going so strongly that it would near impossible to reverse them.

Engineer-Poet said at May 14, 2008 4:44 PM:
do I have to tell an engineer that wind turbines' output over time is only between 25 and 30% of their rated power?
Capacity factor in West Texas is closer to 40%.  At $1800/kW and 40%, wind would yield $4500 per average kW.  Current-generation CAES plants burn natural gas, but yield roughly 75-80% electricity out over gas in (compared to ~46% for the best simple-cycle gas peaking turbines).  This stretches the gas supply and makes e.g. biogas a competitive fuel.

I expect we'll need nuclear (up to maybe 40% from today's 19%), but the combination of wind+PHEV+CAES will fill a fairly large niche.

aa2:  The entire licensing game for nuclear changed several years ago.  Licensing is now one-stop combined construction and operating.  The industry has enough clout to block detrimental changes in the rules, so the tales of the "bad old days" aren't a picture of where we're going.  There will be much wailing and gnashing of teeth at Greenpeace, and that's about all they'll be able to do.

Fat Man said at May 14, 2008 9:46 PM:

Gee, if we, as a country, dither and stall long enough we won't have to build any new power plants because we will all be dead.

The only thing that is going to be more expensive than building new power plants is not building them.

I still object to comparing wind and solar plants with nuclear, without accounting for the fickleness of the wind, yeah it blows 40% of the time in West Texas*, but there will be periods of time of days in a row when it does blow at all, or the noted tendency of the sun to disappear at night. The cost of storage -- and lots of it -- in addition to generation, must be accounted for.

ep wrote: "At $1800/kW and 40%, wind would yield $4500 per average kW."

Since the highest figure Brian quoted above is $5144, wind is not a huge bargain.

*the rest of the time it sucks.

Fat Man said at May 14, 2008 9:55 PM:

RP wrote: "Idling trucks is another way to cut energy costs. ... The older and less fuel efficient trucks will get idled first. ... Hybrid diesels only make sense for delivery trucks that do a lot of stop and go. ..."

My first reaction was why don't they turn them off instead of letting them idle. Then I realized you meant complete removal from service.

Nonetheless, truck manufacturers, truck stop owners, and truckers need to get together on protocols to allow trucks to shut down their motors while they are parked overnight, and to obtain heat and electricity from land lines.

Randall Parker said at May 14, 2008 10:24 PM:

Fat Man,

Necessity is a mother. They'll get smarter in all sorts of ways because costs are so high.

The biggest unknown on future oil prices: The rate of demand destruction. So far demand destruction hasn't gotten as big as Asian economic growth. But maybe that could change? There's a lead time between high prices and people shifting en masse to lower energy lifestyles.

aa2 said at May 15, 2008 2:48 AM:

Engineer-Poet that is good news if the regulatory and legal system for nuclear has changed, but it seems companies are still skeptical as they are charging a lot more in the US. Although they are going for it, and before they weren't.

Something to consider with nuclear is this generation plants are supposed to last 60 years. And I presume will be able to last a lot longer then that with repairs and retrofits way down the road. Especially imo for those who buy the mass produced reactors. The Westinghouse AP1000 or Areva EPR's or GE reactors. I believe there will be many dozens of each built, so that is going to give a lot of expertise on how best to maintain them and probably how to uprate them.

So say the cost is an astronomical 4,500$ per kilowatt hour in the USA. So the 1600 MW Areva EPR would cost 7.2 billion to construct, and the 1117 MW Westinghouse AP1000 would cost 5 billion to construct. You know I just calculated the costs and at 4,500 a kilowatt hour nuclear would still deliver at 5.6 cents a kilowatt hour, the big thing is the low interest rates right now and those loan guaruntees, I assumed they would be able to borrow at 5%, and a 30 year loan.

The beauty is once that 30 year loan is gone, the cost drops to about 1.6 cents a kilowatt hour. And you have 30 years at that rate, then however long you can repair and retrofit.

aa2 said at May 15, 2008 2:51 AM:

Allan, thats what you would have to do is make the changes to organization and structure so that it would be very hard to reverse down the road. Sort of like 10 years of a dictator or someone with a mandate to reform all the dysfunctional areas then transition back to democracy.

Dystopianator said at May 15, 2008 4:31 AM:

South Africa and California were both once-great industrial powers who are dying from lack of energy. Poor planning. Stifling regulations. Stupid leaders. Anti-energy attitudes. Corruption. Laziness.

California is like a dysfunctional African nation. And the US as a whole is becoming more like California.

Engineer-Poet said at May 15, 2008 3:38 PM:
So say the cost is an astronomical 4,500$ per kilowatt hour in the USA. So the 1600 MW Areva EPR would cost 7.2 billion to construct, and the 1117 MW Westinghouse AP1000 would cost 5 billion to construct.
You mean $4500/kW.  At $4500/kWh, I doubt anyone would want to run more than a LED flashlight. ;)

Recall that $4500/kW is the projected price today; by the time the manufacturing date rolls around, the prices of important pieces could be much higher.  Shortages could stretch out construction and drive costs up further.  The times ahead will be volatile, and the shorter the construction time the less risk is involved.

I'm all for nuclear, but maybe we should test thorium-breeder fuel for PBMR's and revive molten-salt reactors as hedges against some of the things out there.

Jerry Martinson said at May 18, 2008 3:53 PM:

FatMan wrote: "Nonetheless, truck manufacturers, truck stop owners, and truckers need to get together on protocols to allow trucks to shut down their motors while they are parked overnight, and to obtain heat and electricity from land lines."

Many truck stops are starting to offer this service throughout the United States.

IdleAire is a company that makes such equipment:

I'd assume that the major trucking companies like Schneider and JB Hunt are going to roll this or similar technology out at their service plazas if they haven't already.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©