Power market developments

How much?

20 November 2007



For some utilities, the capital costs of a new nuclear power plant are prohibitive.


Just before the release of the US Energy Information Administration’s (EIA’s) Annual Energy Outlook 2005 (AEO 2005) the then senior vice president of nuclear generation and chief nuclear officer at the Nuclear Energy Institute (NEI), Marvin Fertel, told the Senate Energy and Natural Resources Committee that the assumptions made on new nuclear plant construction were erroneous. The EIA had assumed overnight capital costs of $1928/kWe, which Fertel claimed were “unrealistically high, and inflated.”

The EIA, Fertel said, “assumed that new nuclear plants would experience the same delays, lengthy construction periods and high costs experienced by some of the plants built in the 1980s and 1990s.” These assumptions were unrealistic owing to advances in construction techniques and new simplified, standardised plant designs. More realistic overnight capital cost estimates of new nuclear were of the order of $1400-1500/kWe for the first-of-a-kind and $300 less for the nth-of-a-kind, he claimed.

The EIA’s estimate, however – which has not changed significantly since AEO 2005 – was actually optimistic when compared to a 2003 report by Massachusetts Institute of Technology (MIT), titled The Future of Nuclear Power, where the base case overnight capital cost for nuclear was $2000/kWe (in 2002 dollars).

There are many other figures available, including the June 2007 report by The Keystone Center, titled Nuclear Power Joint Fact-Finding. This study, which was funded by several nuclear plant operators as well as other interested parties including General Electric and NEI, estimates overnight costs of $2950/kWe (in 2007 dollars). With interest, this figure translates to between $3600/kWe and $4000/kWe.

Interestingly, when Nuclear Power Joint Fact-Finding was released, the nuclear industry press chose to either focus on other aspects – in particular the ‘finding’ that nuclear is a viable option for dealing with climate change – or ignore the report altogether. Considering the number of organisations involved in the nuclear industry that backed the report, this low level of coverage is anomalous, and suggests a certain amount of discomfort with the findings.

However, prohibitively high though it may at first appear to be, even the figure for new build costs in The Keystone Center report is considered too low by some observers.

Independent energy consultant and former director of power planning and forecasting at Seattle City Light Jim Harding said he thought the lower figure of the report’s range, ie $3600/kWe, is no longer believable and the upper limit of $4000/kWe “is probably low.”

A paper by Harding, published by the Nonproliferation Policy Education Center in June 2007 and titled Economics of New Nuclear Power and Proliferation Risks in a Carbon Constrained World, states: “Final construction costs in real 2007 dollars range from $4300-4550/kWe. This is not far from a May 2007 estimate of $4000/kWe from Standard & Poor’s, but it is also probably too narrow, given construction time and real escalation uncertainties.”

Even more recently, a ‘special comment’ report released by Moody’s Investors Service on 10 October this year, titled New Nuclear Generation in the United States: Keeping Options Open vs Addressing An Inevitable Necessity, estimated the all-in costs of a nuclear plant to be between $5000 and $6000/kWe. The report did however provide a note of caution, stating: “While we acknowledge that our estimate is only marginally better than a guess; it is a more conservative estimate than current market estimates.”

Explaining the shortcomings of cost estimates in more detail, the report states: “All-in fact-based assessments require some basis for an overnight capital cost estimate, and the shortcomings of simply asserting that capital costs could be ‘significantly higher than $3500/kWe’ should be supported by some analysis. That said, Moody’s cannot confirm (and all of our research supports our conclusion) definitive estimates for new nuclear costs at this time. Moody’s can assert with confidence that there is considerable uncertainty with respect to the capital cost of new nuclear and coal-fired generating technologies, and that companies may decide not to proceed with financing and construction unless and until they have satisfied themselves (and, where necessary, their boards and regulators) that the investment is justified and that the plant can produce electricity and recover costs at a price that will not be overly burdensome to consumers.”

According to Moody’s, companies that build new nuclear plants will see marked increases in their business and operating risks because of the size and complexity of these projects, the extended time they take to build, and their uncertain final cost and cost recoveries. The report’s author, Moody’s vice president, senior credit officer Jim Hempstead, commented: “To the extent that a company develops a financing plan that overly relies on debt financing, which has an effect to reduce the consolidated key financial credit ratios, regardless of the regulatory support associated with current cost recovery mechanisms, there is a reasonably high likelihood that credit ratings will also decline.”

At least in the USA, this is probably the weakest link: granted that the industry has to address potential skills shortages, decommissioning costs, long-term waste management concerns, supply chain constraints, licensing and regulatory uncertainties, amongst many other issues – but faced with a lower credit rating, there aren’t many company boards that would give the go-ahead to a new nuclear plant.

Old plant, new build

Back in May 2002, the Tennessee Valley Authority (TVA) board approved the restart of Browns Ferry 1, which had been offline since 1985 when TVA had decided to shut the plant down following problems of poor performance and management. Approving the restart project, a TVA statement read: “The unit 1 recovery project will cost from $1.7 billion to $1.8 billion and will take five years to complete.” The unit now has a capacity of 1155MWe – 105% of its originally licensed capacity – and work has been carried out that would allow the plant to be uprated to 1280MWe. Assuming permission to operate at the higher capacity level is obtained, the cost of the restart project was some $1400/kWe.

While this figure is comparable to Fertel’s estimate of what is “realistic” for a new plant, one should also bear in mind that a new nuclear plant would be expected to operate for 60 years or more. Browns Ferry 1, on the other hand, is currently planned to operate for a further 27 years (until 2034), having received a 20-year operating licence extension (to its original 40-year operating licence) last year.

More recently, this summer the TVA board approved the completion of Watts Bar 2, at an estimated cost of $2.49 billion. Translating to a figure approaching $2000/kWe (2007 dollars), this doesn’t even take into account the fact that the Westinghouse PWR is considered to be around 60% complete.

While these projects give some indication as to how much a utility is willing to spend on new nuclear capacity, the figures give no indication as to how highly TVA values the lower level of risk associated with these projects compared to so-called Generation III+ plant.

Lowering the risk was also a main consideration behind NRG Energy’s decision to choose the ABWR design in its combined construction and operating licence (COL) application, which was submitted to the US Nuclear Regulatory Commission (NRC) on 24 September 2007. NRG spokesman David Knox told Nuclear Engineering International (NEI): “We chose the ABWR technology because of the certainty that it provides: it’s already certified by the US NRC and with it having been built four times in Japan, the people that are building it know exactly how much concrete you need, how much steel you need, and how many man hours it will take for the construction.” Although it is not really possible to make an accurate estimate for projects that are still several years out into the future, Knox added: “With this one we actually have a lot more certainty there because we know exactly what the requirements are going to be.”

A fact sheet on NRG’s proposal to build two 1350MWe ABWRs at the South Texas Project (STP) states that construction of the units would “cost more than $6 billion.” Upwards of $2200/kWe then, but of the same order of magnitude as EIA estimates and the completion of Watts Bar 2. However, it is worth noting that when NRG announced its repowering programme, which included the proposed nuclear units at STP 3&4, the figure given in a 21 June 2006 press release was only $5.2 billion for the project.

Furthermore, this project is likely to be first in line for the new build ‘incentives’ available under the US Energy Policy Act of 2005 – and, as a result of the involvement of Japanese companies in the project, financial support from the Japanese government, in the form of export credits or political risk insurance, might also be available.

The utilities’ view

Two of the utility backers of The Keystone Center study were American Electric Power (AEP) and Florida Power & Light (FPL). Since its publication, heads of these utilities have publicly said they believe the cost of a new plant to be in line with the report’s figure. At the end of August, Tulsa World reported that Michael Morris, CEO of AEP, said AEP is not planning to build new nuclear plants due to likely delays, supply chain constraints, and that “realistic” costs would be around $4000/kWe.

The same issues were covered in detail by Lew Hay, chairman and CEO of FPL, in the Biannual General Meeting of the World Association of Nuclear Operators (Wano), held in Chicago, Illinois on September 23-25. Hay told the meeting: “Although suppliers keep quoting overnight costs of $2500 to $3500 per kilowatt, I believe the all-in costs are likely to be much higher – possibly twice as much once you factor in owners’ costs such as land, cooling towers, switchyard, etc, interest during construction and cost escalation due to inflation and cost overruns. And of course we have to have a contingency as well.”

He went on: “If our cost estimates are even close to being right, the cost of a two-unit plant will be on the order of magnitude of $13 to $14 billion. That’s bigger than the total market capitalisation of many companies in the US utility industry and 50% or more of the market capitalisation of all companies in our industry with the exception of Exelon.”

This, he said “is a huge bet for any CEO to take to his or her board.”

It is a huge bet, but not one that Hay dismissed out of hand, as Morris seems to have done. However, it is hard to imagine the FPL board not raising a few eyebrows at these figures, especially after the company’s recent acquisition of the two 525MWe units at Point Beach. On 1 October 2007, FPL Energy announced that it had completed the purchase of the Point Beach PWRs from We Energies for around $719 million for the reactors, plus some $205 million for fuel and other items. The units received 20-year operating lifetime extensions at the end of 2005, extending their expected closure dates to 2030 and 2033.

Returning to the ‘huge bet” of new nuclear, it is clear that UniStar Nuclear is certainly one organisation that is prepared to take on the odds.

Although the NRG Energy COL application discussed earlier is the first full one to be submitted to the NRC, earlier this year, on 13 July, UniStar Nuclear (a partnership between Constellation Energy and Areva, Inc), submitted the environmental report portion of its reference plant (Calvert Cliffs) COL application to the NRC. The documentation was published on the NRC website. Although later redacted, one of the documents stated that the project is expected to be financed with 80% debt and 20% equity, assuming the debt is completely covered by federal loan guarantees. Elsewhere, it noted that debt would be $5500 million, making the expected capital cost of a 1600MWe USEPR unit $6875 million, or around $4300/kWe.

Earlier this year Constellation provided an estimated, overnight cost, levelised over four units, for the USEPR of $2400/kWe in 2005 dollars. This figure correlates well with Hay’s assertion that all-in costs are likely to be of the order of twice the amount quoted for overnight costs. Constellation told NEI that it intends to provide a revised cost estimate early next year.

Proprietary information

It is not so easy to verify these figures since the industry is often averse to revealing such information. For example, the value of the contract to supply four AP1000 units at two sites in China would provide a good indication as to what to expect an AP1000 plant in the USA to cost. But this seems to be a closely guarded secret. On the other hand, at a press briefing at the beginning of September 2007, president and CEO of Westinghouse Steve Tritch was asked whether his company would say how much an AP1000 would cost. He replied that Westinghouse would give a price within the next couple of months, once it had done more work on the design. One of the mistakes of the past, he said, was to allow utilities to have free rein over the design and that nowadays the tendency was to standardise designs as much as possible. Once Westinghouse had got the AP1000 design to around 85%, he said, then it would be in a position to give a figure on the cost of a new unit.

Vendors are not known to overestimate plant costs, and it now seems that the widely-quoted €3 billion turnkey price for the EPR under construction at Finnish utility TVO’s Olkiluoto site will cost vendor Areva dearly. Nevertheless, the project has provided Areva with many lessons learned, giving the company a competitive edge in the US and UK new build markets. For the time being though, at least for earlier than nth-of-a-kind projects, the nuclear industry remains hard pressed to point to projects where it got the costs right.

Government incentives

Even if the industry is capable of achieving the construction times it claims are possible – and if projects are not held up for any other reason – it still looks as if the industry will need more help from the government.

One year ago, speaking at Wilmington Conferences’ European Nuclear Power Debate, held in London on 4-5 September 2006, Bill Coley, CEO of British Energy said: “I don’t believe subsidies are needed at all. I think, given some certainty about permitting, planning and licensing – that assures us that we can put standard designs in place – we can do so at a very competitive cost.” While he was referring specifically the UK industry, he added that he was also familiar with the US Energy Policy Act of 2005, and had provided input to the US government’s energy review. But this input did not include asking for subsidies, he said. “I was somewhat surprised when the Energy Policy Act was passed and I saw incentives were in it. I think what you’re seeing is that the nation – the USA – fundamentally believes that nuclear is critically important to the future energy supply and economy of the country and they want to do whatever they can to induce people to make that investment and to restart nuclear build in the USA. But you don’t have to have the subsidies and I would not lobby for those.”

One year on, in its response the UK nuclear consultation, British Energy said: “Nuclear is cost competitive without subsidy, provided the fossil fuel alternatives carry the cost of the carbon emissions associated with their use.”

Others, such as Electricité de France (EdF) subsidiary EDF Energy, took the same line. EDF Energy’s response stated: “Investors such as ourselves are prepared to invest in nuclear without government subsidy,” but called for a long-term guarantee on a minimum carbon price, which, the company claimed, “is not a subsidy for nuclear.”

The same is true in the USA, where more government support is available to new nuclear projects. Even with the incentives in the US Energy Policy Act, it seems unlikely that new nuclear plants will be ordered without a price on carbon emissions. As Lew Hay put it: “Based on our analysis, nuclear power will not be economic without a sustainable carbon price of roughly $30 per ton, or more.”


Related Articles
The American way
Halfway through the longest build in history
Leading America
New company for US ABWRs

FilesHistorical US construction costs
Recent construction cost experience

Recent construction cost experience Recent construction cost experience
Historical US construction costs Historical US construction costs
Calvert Cliffs Calvert Cliffs


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.