How much would a 100% nuclear energy system cost?

Joris van Dorp, MSc.
The Kernel
Published in
15 min readFeb 28, 2018


In discussions about the future of energy, there has been for decades a lively debate on the question of what role nuclear energy should play, if any, going forward. While this debate is fascinating I want to focus here on one aspect that I feel is sometimes misunderstood: the economics of running entire societies on nuclear energy alone, for 100% of energy supply.

One of the most powerful nuclear power plants in the world, the 6400 MW Bruce generating station in Canada serves the electricity demand of 4 million Canadian households at 12000 kWh each.

I will show that 100% nuclear energy would be a cost effective option, contrary to the alternative extreme of 100% renewable energy which is not.

I will first motivate why I find this hypothetical relevant, and then I will provide simple calculations and conclusions supporting the case, followed by concluding remarks about secondary aspects of the topic.

The relevance of a hypothetical 100% nuclear energy system

Centrists in the energy discussion (in a laudable attempt to achieve compromise between the dueling factions of pronuclearism and antinuclearism) sometimes try to calm the debate by stating that “all energy options will be needed to solve the climate/energy challenge”. This statement comes in other forms, such as: “no one energy option can solve the challenge”, “There is no silver bullet for solving climate/energy” or “We cannot afford to exclude options just because we don’t like them”.

While I sympathize with such statements — if only because they are evidently true given the urgency today of reducing co2 emissions — I am concerned about a misunderstanding that such statements could create about the potential role of nuclear energy. It could obscure that nuclear can solve the challenge without any help from solar, wind, biomass, geothermal, hydro, tidal energy, energy storage, demand response, long distance transmission, smart grids and all the other technologies often noted as being necessary for solving the challenge.

How much would a 100% nuclear energy system cost, compared to what we are used to paying for energy today? Is it true what the centrists seem to say, that 100% nuclear energy is just as unfeasible as 100% renewable energy would be, or is the magnanimous notion that “we need all available technologies” to solve the climate/energy challenge fundamentally incorrect, because nuclear could actually do it all as has recently been argued by these scientists?

The cost of 100% nuclear energy, assumptions

To determine the cost of 100% nuclear energy, I will make the following assumptions intended to simplify the analysis. (Readers may skip forward to the next section, Results)

  • I will use total United States of America hourly consumption of electricity for 2017 (data: EIA) as the reference energy demand to be met.
  • I will assume all other types of energy demand (district heat, process heat and fossil fuels that cannot readily be replaced by electricity) will be provided by some combination of electricity, hydrogen synfuels produced by electrolysis, or heat from nuclear boilers and heaters. However, I will leave these energy demands aside under the assumption that - despite adding a substantial amount of electricity demand to the grid - they will not do so in a way that would make non-nuclear energy sources more attractive than nuclear energy sources. In other words: they will likely increase electricity demand very significantly, but not in a way that affects the discussion here about the relative economics of 100% nuclear.
  • I will assume there are only two energy options available to meeting demand: natural gas and nuclear power.
  • I will assume all plants are built to replace capacity at existing locations, so no expansion of the transmission system or greenfield energy development is required.
  • I will use EIA outlook 2018 figures for the cost components of new generation, operation and fuel supply.
  • I will ignore reserve generating capacity, losses in transmission and distribution, and all other practical detail that has limited significance here.
  • I will apply a simple financial model to determine annual costs and hence cost per kWh for the grid, consisting of three parts: Capital costs, Rent costs, and Operating & Fuel costs, defined in turn:
  • Capital cost is the initial capital cost divided by the plant lifetime (40 years for a nuclear plant, 20 years for a gas combustion turbine.) Annualised, it reflects payments made to pay back the initial investment.
  • Rent cost is the return on capital for the capital provider. It’s a fixed percentage of the initial capital paid each year to investors (for example pension funds and stock holders) in addition to the annual capital cost repayment. The rent is also known as the Weighted Average Cost of Capital (WACC) typically consisting of various types of debt and equity, and which has a further division into nominal (after tax) and real (before tax). In order to keep things simple I will use a simple rent (% of investment paid out per year throughout the life of plant) for the WACC, so ignoring tax impacts, debt/equity ratio’s and so on.
  • Operating & Fuel cost is the total cost of keeping the plant running and supplied with fuel: salaries, supplies, maintenance, fuel and other ongoing payments that are not rent and not payments to pay back the initial investment.


The EIA provides hourly data on total US electricity consumption. The following chart presents that data.

Total 2017 hourly US electricity demand data, downloaded from the US EIA website

Apparently, maximum electricity demand was in the summer, at just over 700 GW (gigawatts, or billions of watts). Total annual consumption was 3943 TWh (terawatthours, or trillions of watt-hours). The average demand was 450 GW. Demand fluctuates on a daily and seasonal cycle.

To supply electricity using only nuclear and natural gas a total generating capacity of about 700 GW is needed. For example, 400 GW of nuclear and 300 GW of natural gas would yield the following plant utilization curve:

Example dispatch model result for nuclear and natural gas generation if 400 GW of nuclear is installed

The chart shows that in this simple model, nuclear (orange) is being used almost continuously, dipping toward 300000 MW (which is 300GW) only whenever demand falls below 400 GW. Natural gas (grey) is covering the remainder of the demand. Demand is thus met each hour of the year by either nuclear alone, or nuclear and natural gas working together.

To investigate the economics of nuclear power, I have run the simple model multiple times to calculate the cost and CO2 performance of different combinations of nuclear and natural gas across a range from having no nuclear at all to having 700 GW of nuclear, with natural gas serving the remaining demand as necessary. As more nuclear is added to the system the capacity factors of nuclear and natural gas change, visualized in the following chart.

Capacity factors of nuclear and natural gas when meeting US 2017 hourly electricity demand, and when nuclear is dispatched before natural gas. The horizontal axis of the chart represents the amount of nuclear power capacity installed on the grid, ranging from none on the left, to 700 GW on the right (enough to meet all US electricity demand in 2017).

For the capital cost of nuclear power, I have considered two alternative capital cost cases.

The first is based on recent EIA figures for new nuclear plants in the USA, which yields $5946/kW. This figure is historically high and indicative of a distressed industry that has been struggling in a society that doesn’t value it and doesn’t care about it.

The second case is informed by the work of Lovering et. al. which considers historical global nuclear industry in various countries, revealing a much lower average cost for nuclear of $2273/kW in today’s dollars.

This cost can be regarded as representative of a settled nuclear industry that is mature and able to focus on producing electricity, without having to constantly deal with costly shocks due to order cancellations, extraordinary policy uncertainty or uncompromising public/political resistance. This cost figure happens to be close to the cost target of $2500/kW that China has for the export of its indigenous Hualong One technology, a cost target which it has already achieved at home, thanks to its deliberate long-term policy of developing a vibrant nuclear industry. I will use the current China capital cost as the second capital cost case.

I consider the China cost figure as more relevant than the EIA figure for any country which chooses, like China, to embark on a nuclear new build program of some magnitude. When building dozens or hundreds of reactors, nuclear industry (and society) will be able to deliver, as it has in the past.

I have also considered different levels of rent, to investigate the impact of rent on the cost of electricity. If have considered 0% rent (interest free loan) through 8% rent (typical for commercial investors) in increments of 2%.

Running the models for the various financial assumptions leads to the following results, shown in the chart below, which I will discuss.

Economic model results for 0 to 700 GW of nuclear, EIA or Chinese nuclear capital cost, and WACCs ranging from 0% to 8%. The meaning of various lines in the chart is described below.

The left vertical axis represents the cost of electricity in $/kWh (please excuse the € symbols in the chart which should be $'s), being the total annual cost of running the collection of nuclear and/or natural gas plants divided by the total amount of electricity supplied. The sloping colored lines represent the cost of electricity for each of five different rents, ranging from 8% rent to 0% rent and assuming current EIA capital cost estimates of new nuclear power in the US. The dashed colored lines represent Chinese nuclear capital cost.

The right vertical axis represents CO2 intensity in grams of CO2 per kWh. The dark blue line sloping down and to the right shows that as more nuclear power is used, the lower the average CO2 intensity of the grid becomes. The intensity falls linearly as nuclear is added, until about 400 GW of nuclear is installed. Above 400 GW, as nuclear capacity factor starts falling, each additional nuclear GW is reducing average grid emissions by a relatively smaller amount. The cost per kWh starts curving upwards after 400 GW of nuclear has been added, because with each additional nuclear GW, cost is rising faster than the amount of electricity supplied.

The difference in terms of CO2 emissions between the zero nuclear and 700 GW nuclear cases is about 1,7 billion tons of CO2. Assuming a Social Cost of Carbon of $40 per ton of CO2, this equates to a CO2 cost savings of almost 68 billion dollars per year, equal to 1,7 cents per kWh saved in the 700 GW nuclear case. These savings have not been included in the results and are mentioned only for reference.

Note that switching from fossil fuels to nuclear fuels also has indirect benefits in the form of avoided air pollution. These benefits appear to be substantial, but have not been quantified here because they would equally count for a 100% renewable energy system, assuming such a renewable energy system would not rely on burning too much biomass of course.


Using EIA nuclear cost assumptions without any learning effects, and 8% rent, the cost of electricity generation in the USA would reach up to 14 cents per kWh if the USA built 700 GW of new nuclear power to achieve a zerocarbon electricity supply. However, if the cost of installing nuclear power would fall to the more typical level described by Lovering et al (and achieved today by China) then the cost of 700 GW of nuclear would be less than 8 cents per kWh. If the required rent for nuclear would be reduced from 8% to 4%, then the cost of 100% nuclear electricity would be under 10 cents in the EIA capital cost case, and under 6 cents in the Chinese case.

Considering the above, it certainly appears economically feasible to supply 100% of US electricity with nuclear power. There is a rise in cost when supplying more than 60% of electricity with nuclear, but not a sharp upturn as is the case when increasing the supply of solar and wind power. This ability to supply 100% of energy at modest additional cost results from nuclear’s complete lack of reliance on exponentially rising additional investments in storage and transmission to allow it to achieve 100% market share.

If the USA embarks on a national project to install 700 GW of nuclear power, if it succeeds in allowing nuclear industry to reduce costs back to historical trends, and if the rent for new nuclear would fall to a more modest 4%, then the United states could achieve a 100% nuclear powered (and zero carbon) electricity system costing less than 6 cents/kWh. This would rival the cost of new natural gas.

Furthermore, no investment in smart-grids, long distance transmission or energy storage would be needed. There would also be no need for greenfield energy development of any kind, since the nuclear plants could simply be installed on sites currently occupied by fossil fuel burning plants that will be retired sooner or later anyway if co2 emissions are to be eliminated. And comparing the zero nuclear and 700 GW nuclear cases shows that the 700 GW nuclear case reduces co2 emissions by 1,7 billion tons annually, which comes down to a savings of 1,7 cents per kWh in terms of co2 emissions.

Finally, nuclear allows individual countries, states or even individual cities to become zero-carbon, since there is no need in principle to exchange electricity over long distances in an attempt to smooth demand and supply from intermittent renewable energy sources like solar and wind. Every community can decarbonize as much as it wants to, without depending on developments on the other side of the continent.

Clearly then, there appears to be no inherent economic or technical impediment to providing 100% of electricity for the US with nuclear power at a cost similar to a conventional fossil fueled electricity system, if there is a deliberate and sustained national pursuit of such a goal.

Thank you for reading this article. I gladly welcome any comment and criticism that can correct, improve or add to what has been presented. I can be contacted on twitter @energyjvd

An excel sheet with the EIA electricity demand data, the simple dispatch model, the financial parameters and the charts is available here.

Concluding remarks

I want to make clear that I have no ties whatever with nuclear industry, never have, and likely never will. I am an engineer, husband and father who is concerned about the public debate and direction of energy policy in many countries, policy which — I truly believe — is not up to the task and will cause significant and unnecessary harm to people and nature. I have learned a lot by reading the scholarly work and opinions of others far more qualified than I to grapple with the wicked problem of climate/energy and I urge all those who want to help solve climate/energy to do the same. The internet provides easy access to quality information that can satisfy the curiosity of all, whether you have a technical background or not. There are plenty of bona fide experts who maintain a social media presence and who will often be happy to respond to any question and doubt you may have.

There is much more to say about the topic dealt with in this article, and interested readers may find some of those things touched upon in the remarks below.

Remarks on renewable energy

Solar and wind power have not been considered in this article, but can and will play a useful role in the energy system of the future, although their role is inherently limited by their weather dependence, if little else. It looks like solar and wind are today cheap enough to allow them to work economically as a fuel saving technology with natural gas. And if nuclear costs stay as high as they are today, it even looks as though a combination of storage, wind, solar, demand response and nuclear may be an optimal mix for a zero carbon energy system. However, this does not detract from the fact that nuclear power as a single technological concept is evidently sufficient to allow achieving a low-cost zero-carbon energy system, with no help needed at all from any wind power, solar power or anything else, which is the only thing this article was intended for.

Hydropower, geothermal and biomass energy constitute special renewable energy sources which will have a role to play even if the US or other countries choose to solve the climate/energy problem entirely with nuclear energy.

Hydropower in some cases provides essential water management and irrigation services as well as electricity, in which case it would not be beneficial to decommission such hydro plants in order to return rivers to nature. However, hydro projects which exact an environmental toll and have no particular water management purpose can be shut down and replaced by nuclear energy.

Geothermal energy is a potentially viable resource which however requires more development and will remain dependent on suitable geological conditions.

Biomass should not be burned for energy, but will prove important as a low-cost, sustainable and reasonably abundant replacement source of (fossil) carbon for many different industrial and chemical applications. Nuclear power can be used cost-effectively to replace all current applications of biomass burning except niche applications in locations where biomass burning does not exact a significant environmental or economic toll.

Remarks on nuclear technology and external costs

The impact on nuclear economics and applications of non-traditional (yet proven) nuclear technologies such as breeder reactors and high temperature reactors has not been considered in this article, though these technologies will likely prove useful and necessary in a future where nuclear energy is applied confidently and broadly to the solution of global energy challenges. Particularly nuclear breeder reactors, such as uranium/plutonium FBR’s and thorium MSRs, which can use 100% of nuclear fuel instead of only 1%, will be essential in the medium term in order to render nuclear fuel supply inexhaustible. High temperature reactors and heat-only reactors will be useful for decarbonizing industry, hydrogen fabrication and providing district heat, or in effect to decarbonize entire cities.

Nuclear waste, nuclear accidents and (to a much lesser degree) nuclear proliferation risk are issues that are tied to civilian nuclear energy and as such add to its societal cost. However, the cost of nuclear waste and accident risk is already internalized in the cost of nuclear power, via obligatory insurance and nuclear waste fees. Nuclear waste and nuclear accidents consequences moreover are cost centers which are strongly affected by public perception. While nuclear waste management is cheap per unit of energy provided (it’s about ten times cheaper than adding a carbon capture and storage system to a fossil fueled power plant, for example), there is no limit to how high the cost of nuclear accidents (like Fukushima) can escalate if latent extraordinary public fear and distrust lead to draconian and harmful interventions such as evacuating entire regions, or shutting down whole fleets of nuclear plants and replacing their output with vast quantities of costly imported fossil fuels.

Proliferation risk is a unique risk which is ultimately part and parcel of humanity’s arrival in the atomic age. Proliferation risk will never go away and will require ongoing vigilance, institutional attention and international diplomacy. The choice to use more or less civilian nuclear energy does little or nothing to affect proliferation risk. Countries seeking weapons can and will build them without the use of civilian nuclear energy technology, and terrorists wanting to deploy a radioactive “dirty bomb” will obtain radioactive material far more easily from hospitals (nuclear medicine and imaging) or agriculture (sterilization of perishables) or industry (imaging and fault detection), than from inside a civilian power reactor or nuclear waste repository.

Nuclear energy is sometimes presented as not being zero carbon. This is incorrect. Nuclear energy is at least as zero carbon as solar or wind power.

Remarks on public participation

Nuclear power cannot readily provide value to a society in which it is slandered and demonized incessantly through the use of scary half truths and misinformation promoted by prominent organisations that claim to be concerned about people and the environment. As concerned citizens, the most important thing we can do individually to stimulate nuclear energy and help solve the climate/energy challenge is to simply and publicly not be set against civilian nuclear power. Enabling a nuclear powered future doesn’t depend on structural subsidies or even a price on carbon (though that would help speed things along). It only requires the kind of confidence and policy stability that led to the rapid and successful construction 40 years ago of the first wave of nuclear power plants still operating today around the world. It requires that we ask for and obtain accurate and objective information that helps us understand the issues, as opposed to remaining ignorantly stuck in a frightful Cold War mindset that has little if any relevance in the 21st century.

There will always be people and groups with a political, ideological or financial interest to deliberately undermine our confidence by stressing (and fabricating) disadvantages of civilian nuclear energy, but it is up to us whether we decide to sheepishly go along with their scary, misleading narratives or instead join actual climate and energy scientists who urge us to reconsider nuclear energy while there is still time to avoid the worst impacts of global warming and (ultimately) impending fossil fuel depletion.



Joris van Dorp, MSc.
The Kernel

Engineer consulting on energy in the built environment