What does it cost to build a nuclear plant? What could it cost?
March 2nd, 2008
.... when the capital costs of nuclear energy is examined, it becomes apparent that current policies are far from generous to any potential plant constructors. In fact, the way things are currently set up, the costs due to regulatory reasons are enormous.
Note: This post applies to the United States, because obviously the laws and regulations will vary from country to country, and therefore it’s necessary to limit things to one jurisdiction. However, the basic principle that administration cost more than materials and that the actual construction costs are not necessarily very high applies in to almost every nation.
Over half of the cost of nuclear power plant construction is directly related to the cost of licensing, approval and other bureaucratic expenses. A recent proposal for plant construction by NuStar is expected to cost 520 million dollars for licensing. In other words, if everything goes according to plan, the company will have to spend half a billion dollars before they even break ground on the new plant. Although the license fees for any given power plant are only a few million dollars, the process requires numerous studies, surveys, hearings and can take many years. Each and every plant must be certified and each reactor on the site must go through the process, even if the plan is to build multiple identical reactors. Public hearings are conducted and petitions are accepted against the planned plant. Injunctions, hearings, contesting of studies and other such measures can add months or years.
In order to determine the minimum cost of a nuclear power plant, it is necessary to separate the regulatory and governmental costs and determine simply what the physical cost is for the materials needed and the labor to construct the plant. This is not as easy as it sounds, since many of the components of the nuclear power plant have
The estimates are derived from reported costs of construction of general purpose structures and items and the contracts which have been issued for major components of thermal power plants or related systems to various companies. The estimate are based on real world costs as they are now, although some might argue that these costs could be lower if the mass-production approach were applied to other components of a power plant, as might be the case in a large nuclear energy initiative.
One of the costs which is the most difficult to pin down is the cost of the reactor and reactor systems, since the regulatory costs and other administrative fees are built in. In this case, the cost represents integral boiling water or pressurized water reactors in the Generation III+ family. Newer reactor designs, such as the Pebble Bed or the Molten Salt Reactor could be considerably less costly (or more costly) to construct, but as such reactors have not yet been deployed commercially, this is more difficult to be certain of. The estimates are based primarily on information from Westinghouse relating to the construction and installation costs of reactors as well as some of the cost and feasibility studies done for the IRIS reactor.
Some have suggested that a modular reactor system could be built for under a few million dollars. In theory, this might be possible, and some of the smaller experimental reactors such as the Aircraft Reactor Experiment and Project Pluto were able to construct working reactors for, by modern standards, extremely low costs. A nuclear reactor is not necessarily as complicated a piece of equipment as one might think. Most of the design considerations are related to control systems and other incidental factors. The actual reactor, however, is basically just a big pressure vessel. Depending on the size of the vessel, it may or may not require specialized heavy industrial processes to fabricate. Modern power reactors can take up to two or three years from order to delivery, however mass production and modular fabrication has been demonstrated on comperable industrial equipment. Nuclear energy concepts which rely on multiple reactors of smaller size or do not need a single high pressure vessel, such as the CANDU reactor avoid this limitation.
Physical Cost Breakdown:
Based on these broad estimates, it appears that the non-nuclear aspects of a thermal power plant as well as the necessary security and administrative infrastructure will be of a cost of approximately a quarter billion US dollars. It may very well be less than this, as the estimates are generous. It may also be somewhat more due to other expenses. This cost does not include the infrastructure – the running of transmission lines to the plant site. Because nuclear power plants do not require constant fueling, pipelines or other fuel delivery systems are not necessary.
These costs, however, could be substantially less if the nuclear power plant is replacing an existing fossil fuel power plant. Electrical switching equipment, transformers and transmission lines can be reused in the new plant. Additionally, generators could be refurbished and reused for some or all of the power generation. Turbine-generator sets can theoretically be reused, however most fossil fuel plants run at a lower pressure than nuclear plants and therefore reusing some or all turbines may or may not be possible, depending on the circumstances.
Therefore, our theoretical nuclear power plant can clearly be built for well under one billion US Dollars (£660m)and quite possibly under half a billion US dollars (£330m). It may even be as low as a quarter of a billion US dollars, if it were to replace an existing fossil fuel plant.
Considering that nuclear power generation has a considerably lower fuel cost than other methods, even with the once through fuel cycle, and a substantially lower cost when advanced fuel cycles are employed, the return on such an investment would be enormous. Some of the advanced fuel cycles that make use of on-site reprocessing or even continuous reprocessing and can make use of alternative fuel cycles like thorium could cut the already low cost of fuel for reactors much further. And maybe, just maybe, could it someday be “too cheap to meter?” Perhaps we’ll find out.
Now, when can we get started on this?
Note: This post applies to the United States, because obviously the laws and regulations will vary from country to country, and therefore it’s necessary to limit things to one jurisdiction. However, the basic principle that administration cost more than materials and that the actual construction costs are not necessarily very high applies in to almost every nation.
Current Nuclear Plant Construction Cost:
The current theoretical overnight cost of constructing a nuclear power plant is about 2 to 2.5 billion dollars for a plant with two conventional reactors and generating about two gigawatts – a nominally sized plant. This compares favorably with fossil fuel plant. Westinghouse has estimated the cost of four power plants, each containing two AP1000 reactors and generating more than 2 gigawatts each to be about 8 billion US dollars. General Electric has stated that their new ESBWR design could reduce costs to below $1000 (£660 m) per kilowatt of installed capacity....Over half of the cost of nuclear power plant construction is directly related to the cost of licensing, approval and other bureaucratic expenses. A recent proposal for plant construction by NuStar is expected to cost 520 million dollars for licensing. In other words, if everything goes according to plan, the company will have to spend half a billion dollars before they even break ground on the new plant. Although the license fees for any given power plant are only a few million dollars, the process requires numerous studies, surveys, hearings and can take many years. Each and every plant must be certified and each reactor on the site must go through the process, even if the plan is to build multiple identical reactors. Public hearings are conducted and petitions are accepted against the planned plant. Injunctions, hearings, contesting of studies and other such measures can add months or years.
Each plant is independently designed and each design is studied and approved as if it were the first of its kind, even if it is identical to others. Additionally, the reactor design must be licensed before it can be used in any plant. GE has been working on the certification process for the ESBWR for nearly ten years and is just getting to the final steps of the design certification process. Despite not selling a single reactor, they have already invested $400 million in the licensing process alone. The French firm Areva has stated that it will cost at least a quarter billion dollars for the design work needed to get one of their reactors approved in the United States, despite being approved in the European Union which has equally comprehensive safety standards. They also estimate the cost of designing a nuclear power plant at half a billion dollars. FOR EACH PLANT!
If this same procedure were applied to aircraft, then every Boeing 747 to roll off the assembly would have to go through the extensive flight testing and design certification that the original prototype went through in 1969, even if it was just an updated version of an existing aircraft. The process would also need to be repeated each time an existing aircraft received an engine upgrade or had the seats reupholstered.
By comparison to fossil fuel power plants:
Fossil fuel plants are not that much cheaper than nuclear plants, even despite the massive regulatory expenses in nuclear power plants......
So what is the actual cost a nuclear power plant can be built for?In order to determine the minimum cost of a nuclear power plant, it is necessary to separate the regulatory and governmental costs and determine simply what the physical cost is for the materials needed and the labor to construct the plant. This is not as easy as it sounds, since many of the components of the nuclear power plant have
The estimates are derived from reported costs of construction of general purpose structures and items and the contracts which have been issued for major components of thermal power plants or related systems to various companies. The estimate are based on real world costs as they are now, although some might argue that these costs could be lower if the mass-production approach were applied to other components of a power plant, as might be the case in a large nuclear energy initiative.
One of the costs which is the most difficult to pin down is the cost of the reactor and reactor systems, since the regulatory costs and other administrative fees are built in. In this case, the cost represents integral boiling water or pressurized water reactors in the Generation III+ family. Newer reactor designs, such as the Pebble Bed or the Molten Salt Reactor could be considerably less costly (or more costly) to construct, but as such reactors have not yet been deployed commercially, this is more difficult to be certain of. The estimates are based primarily on information from Westinghouse relating to the construction and installation costs of reactors as well as some of the cost and feasibility studies done for the IRIS reactor.Some have suggested that a modular reactor system could be built for under a few million dollars. In theory, this might be possible, and some of the smaller experimental reactors such as the Aircraft Reactor Experiment and Project Pluto were able to construct working reactors for, by modern standards, extremely low costs. A nuclear reactor is not necessarily as complicated a piece of equipment as one might think. Most of the design considerations are related to control systems and other incidental factors. The actual reactor, however, is basically just a big pressure vessel. Depending on the size of the vessel, it may or may not require specialized heavy industrial processes to fabricate. Modern power reactors can take up to two or three years from order to delivery, however mass production and modular fabrication has been demonstrated on comperable industrial equipment. Nuclear energy concepts which rely on multiple reactors of smaller size or do not need a single high pressure vessel, such as the CANDU reactor avoid this limitation.
Physical Cost Breakdown:
Non-Power Related:Land Acquisition and Clearing: 0~5 million USDAdministrative office building: 20 million USDFixtures and other incidental: 2 million USDRoads and parking: 500,000 USDOther Misc: 500,000 USD~25 million dollarsSecurity:Perimeter security (fence, gate, systems): 2 million USDGuardhouse, other security: 2 million USDOn-site emergency services: 4 million USDFour One Megawatt diesel generators: 250,000 USDSix 125 kilowatt diesel generators: 200,000 USDUninterruptible Power systems: 150,000 USDControl Room Systems and Redundancy: 1 million USD~10 million dollarsPower Generating:Steam Turbine Generator Sets: 160 million USD
Piping, cooling, regulation: 30 million USDTurbine building: 10 million UDSMisc support and service equipment: 5 million USD
Transformers and switching: 15 million USD~220 Million dollars
Based on these broad estimates, it appears that the non-nuclear aspects of a thermal power plant as well as the necessary security and administrative infrastructure will be of a cost of approximately a quarter billion US dollars. It may very well be less than this, as the estimates are generous. It may also be somewhat more due to other expenses. This cost does not include the infrastructure – the running of transmission lines to the plant site. Because nuclear power plants do not require constant fueling, pipelines or other fuel delivery systems are not necessary.
Nuclear Related:
Containment Structures: 40 million USD
Fueling and Spent Fuel Handling: 20 million USD
Reactor and Reactor subsystems: 150-200 million USD
Therefore, the physical cost of constructing a nuclear power plant, using existing systems and generation III+ reactor technology could be reasonably estimated at as low as half a billion US Dollars for a two gigawatt capacity plant. In reality, however, there will always be some administrative costs. This only represents the "overnight cost" of the construction. In other words, it does not include interest on loans or bonds used to build the plant, which will vary depending on how much of the capital is borrowed. There will also be a need for insurance, taxes on items and some regulatory and licensing fees. It hardly seems unreasonable that the necessary inspections and design considerations could be accomplished for under one hundred million dollars, especially considering that this would be much higher than the approval costs of nearly any non-nuclear installation of equal or greater environmental hazard.
Therefore, our theoretical nuclear power plant can clearly be built for well under one billion US Dollars (£660m)and quite possibly under half a billion US dollars (£330m). It may even be as low as a quarter of a billion US dollars, if it were to replace an existing fossil fuel plant. Considering that nuclear power generation has a considerably lower fuel cost than other methods, even with the once through fuel cycle, and a substantially lower cost when advanced fuel cycles are employed, the return on such an investment would be enormous. Some of the advanced fuel cycles that make use of on-site reprocessing or even continuous reprocessing and can make use of alternative fuel cycles like thorium could cut the already low cost of fuel for reactors much further. And maybe, just maybe, could it someday be “too cheap to meter?” Perhaps we’ll find out.
Now, when can we get started on this?
