This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Seeking Consensus on the Internalized Costs of Nuclear

What is meant by “internalized costs”?

Internalized costs are the costs which can be accurately accounted for in our current systems. In energy production, these costs typically consist of capital costs, financing costs, operation and maintenance costs, and exploration costs. Some energy options incur these costs in various stages such as extraction, transportation and refinement. Profits and taxes are excluded wherever possible in order to isolate the pure cost of production.

Nuclear energy is the first energy option covered in this series where up-front capital costs dominate the overall cost picture. This requires a more precise estimation of effects such as financing costs, cost increases during construction and capital utilization rates (capacity factor). An example of the influences of these factors on the ongoing UK case is shown below. Thus, we have to take a very brief look at these cost elements before moving on.

Financing costs

Financing costs have a very large influence on the economics of capital-intensive technologies. For example, the LCOE of a technology where all costs are in the form of an up-front capital investment with an expected lifetime of 50 years will increase by a factor of 5 if the discount rate is increased from 0% to 10%.

Financing costs (discount rate) for energy infrastructure typically consist of four main elements: the time-value of money (~0-6% to keep the financial system stable based on regional GDP growth rates), a risk premium (~1-5% to compensate for losses due to accidents or underperformance), operating costs for the financial sector (~1-2% estimated from mortgage rates relative to central bank rates), and some clean profits to investors. This study will exclude investor profits, but it should be noted that a technology will not be deployed if investors do not see a clear profit potential.

Cost escalation under construction

Nuclear power plants typically have long construction times. These years involve some financing costs even before the plant generates any revenue. In addition, nuclear suffers from a “cost escalation curse” as illustrated below for the rather extreme US case. This implies that, by the time that the plant is actually fully constructed, real costs are often significantly higher than original estimates.

Capital utilization

Given a certain discount rate for future electricity, it is obviously best to get as much value as early as possible out of a capital-intensive generating technology. The capital cost component of the LCOE is inversely proportional to the capacity factor, implying that halving the capacity factor would almost double the LCOE of a capital-intensive technology like nuclear. It is therefore understandable that the global average capacity factor of nuclear plants is generally quite high (below), affording little freedom for load following.

Internalized costs of nuclear

The costs of nuclear reactors vary greatly from case to case. Overnight capital costs of plants reviewed by the IEA ranged from $1556/kW in Korea to $5858/kW in the Czech Republic. Nuclear energy is currently exclusively used for electricity, but estimates for transport and heating will also be given below for comparison to other energy options.


The LCOE of nuclear power is given below as a function of the two most important variables: plant capital costs and financing costs. Other assumptions include a $10/MWh fuel cycle cost, O&M costs expressed as a linear function of capital costs based on data in the aforementioned IEA report (also roughly $10/MWh on average), a plant lifetime of 50 years and a 90% capacity factor. A high capital utilization rate is selected for fair comparison to other capital-intensive technologies enjoying priority dispatch. The Excel datasheet used to create this figure can be accessed here.

In addition, the effect of capital cost escalation under construction is illustrated below for several compounded escalation rates and construction times. The illustration is based on overnight costs of $4000/kW.


The cost of using nuclear-generated electricity for heat is given below.


In order to facilitate an apples-to-apples comparison for transport costs from electricity sources, an estimate of power-to-liquids costs will be given here. Even though this article targets current internalized costs, estimated costs of mature (beyond 2020) power-to-H2 processes will be used. Assumptions include the following: $1500/kW capital costs, $60/kW/yr O&M costs, 15 year lifetime, 5% cost of capital and 70% conversion efficiency. A 50% synfuel production plant capacity factor will be assumed in a baseload-dominated system where synfuels will primarily be produced during low-demand periods at night and over weekends.

Numbers are based on PEM electrolysis in this report and Fischer-Tropsch syngas conversion numbers from this book. These costs are quite optimistic and show that the resulting synfuel costs are increasingly dominated by energy costs at an electricity cost beyond $40/MWh.


If you have a number that differs significantly from the estimates given above, please add it in the comments section below together with an explanation and a reference. 

Schalk Cloete's picture

Thank Schalk for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.


Schalk Cloete's picture
Schalk Cloete on May 5, 2014 3:27 pm GMT

DATA: Capital costs: $2500/kW in case of nuclear predominantly being expanded in the developing world.

This is based on estimations for Chinese costs by the IEA.


Schalk Cloete's picture
Schalk Cloete on May 5, 2014 3:33 pm GMT

DATA: Capital costs: $6000/kW in case nuclear revives in the developed world where capital intensive investments like nuclear make a lot more sense due to low economic growth.

Investment costs given by the IEA are in this range for developed nations. In the case of the US, the EIA estimates overnight costs around $5500/kW, but this probably excludes cost escalations under construction. 

Schalk Cloete's picture
Schalk Cloete on May 5, 2014 3:41 pm GMT

DATA: LCOE: $47/MWh in the case of a developing world buildout. 

This number assumes an 8% discount rate, primarily because the time-value of money is a very important factor in the rapidly growing developing world. In other words, one unit of nuclear electricity generated 30 years from now will be valued much lower than at present because people will be a lot richer by then.

The 8% figure can be roughly broken down into 5% time-value of money (GDP growth rate), 2% risk premium and 1% finance industry costs. Risks are viewed to be reasonably small in the developing world since electricity demand will continue to rapidly expand and and the chance of reactors being shut down due to a black swan event in another country is quite low. 

Schalk Cloete's picture
Schalk Cloete on May 5, 2014 3:45 pm GMT

DATA: LCOE: $77/MWh in case of a revival in the developed world. 

This estimate is for a 6% cost of capital. The time-value of money in the developed world is very low at present (the base interest rate is essentially zero), but risks surrounding investments in nuclear power remain large (e.g. the German reaction after Fukushima). Thus, the breakdown is roughly 1% time value of money, 4% risk and 1% financial industry overheads. 

Nathan Wilson's picture
Nathan Wilson on May 7, 2014 3:21 am GMT

DATA: should include decommissioning and waste disposal cost.

It is worth noting that the decommissioning cost is normally internalized for nuclear.  As listed here the costs are on the order of 10% of the construction cost.  Using 6% for a discount factor, and a 50 years plant lifetime and 10 year delay, that implies each dollar of decommissioning cost is only worth 2.4¢ of upfront costs.  Hence the decommissioning effectively adds 0.24% of to the LCOE. 

In the US, waste disposal is also internalized at a 0.1¢/kWh charge paid to the Federal government.

Charles Weber's picture
Charles Weber on May 9, 2014 7:32 pm GMT

       Everyone is stressing excessive use of fossil fuels as causing a green house affect on climate. However, that is the least of our problems. Sucking our petroleum reserves dry, even our oil shale, will have disastrous consequences in the future on our economy (USA) and our security, especially military security. We should use foreign fuel as much as possible and atomic power short term. Long term could be a disaster. Losing the oil for so trivial a purpose as generating heat (therefore electricity, etc.) is bad enough. But at least the carbon is still on the surface of the Earth, expensive to make oil and chemicals out of, but not impossible. Uranium once burned is gone forever. If an important use arises then, our progeny will be cursing us. If that use were to be the only practical way to prevent a huge meteorite from destroying the Earth, they will be cursing us with their dyeing breath.

                                                   Sincerely, Charles Weber

Nathan Wilson's picture
Nathan Wilson on May 11, 2014 4:36 pm GMT

The report from the World Nuclear Association is that India is still building it’s domestically designed and built 700 MW PHWR for $1700/kW (it’s a heavy water moderated reactor which runs on natural uranium).  They have a new design for a 300MW AHWR derivative which will utilize more passive-safety features, but the first unit has not yet been built, and no cost has been announced.

They are building a few Russian LWRs, but these are apparently valued mainly to boost the nuclear build rate, and of course to boost uranium imports (the Russian reactors come with Russian fuel supply agreements).

The Indian AHWR will supposedly become the dominant reactor in India, and will be exported.  It seems like a good size for India’s 200 GW-peak grid (with its 6.3% annual growth = 13 GW).

Schalk Cloete's picture
Schalk Cloete on May 11, 2014 10:30 pm GMT

Impressive. Let’s hope that the ultimate costs at commissioning are close to this value and that the cost escalation curse is proactively prevented.

Now that China is trying to shift to a more consumer-based economy, India will probably gradually take over the role as the world manufacturing centre. Indian GDP is less than half that of China and it will soon have a greater population, implying that there is lots of cheap labour. If India can bring lots of cheap energy into the mix, the economy can develop quite quickly. Nuclear should play a significant role although it is currently still 40 times smaller than coal and this ratio has been steady since the turn of the century. 

Nathan Wilson's picture
Nathan Wilson on May 12, 2014 4:22 am GMT

Sucking our petroleum reserves dry, even our oil shale, will have disastrous consequences in the future on our economy

I’m not worried.  Petroleum-derived fuels are best for transportation, but they are only slightly better (for non-aviation uses) than the 2nd tier alternatives:  natural gas, methanol, and ammonia.  Natural gas we have lots of (assuming a price point of $4/gallon-of-gasoline-equiv, at which point the power plants won’t compete for access).  Methanol we can make from our abundant coal resources for about equal to today’s gasoline price.  Ammonia can be made from any primary energy source (including solar, wind, and nuclear if the price goes high enough).

Uranium once burned is gone forever.

Mostly wrong, and completely irrelevant.  Most of today’s reactors use enriched uranium (in which the U235 isotope is a larger fraction than in natural uranium), which is required for light water reactors.  In the enrichment process, 90% of the uranium comes out as “depleted uranium” (which has too little U235 to be of use for most reactors), which is stored as “enrichment tails” for future use.  Of the 10% of the uranium which is processed into fuel, 95% of that is unburned, and comes out as part of the “spent nuclear fuel”.

So our nuclear fuel cycle only “burns” 0.5% of the uranium ore which is dug up, the rest is stored for future use.  Fast breeder reactors can use all of natural uranium or the leftover enrichment tails, or the remaining uranium and plutonium from spent nuclear fuel.  The US has enough enrichment tails in storage to power the nation for something like 400 years, should we choose to deploy fast breeders.

But in any case, the Earth’s uranium and thorium resources are in fact so enormous as to be inexhaustible.  A ton of ordinary/average rock has an order of magnitude more energy in the form of nuclear fuel than coal has when burned in a coal plant.  And much of Earth’s uranium occurs in concentrated deposits.

Joris van Dorp's picture
Joris van Dorp on May 28, 2014 11:17 am GMT

Thankfully, it looks to me like the German developments in nuclear have increased nuclear project financial risk only temporarily. After the German anti-nuke policy is understood to be a mistake – which I think will happen rather soon judging from the mounting problems of the Energywende and the resulting political discussion – the risk premium for nuclear projects in the future may well fall below what is was before Fukushima. Presumably, governments the world over are going to be jarred awake when Germany eventually makes an about-face on their nuclear phaseout policy, potentially granting future nuclear investments solid political backing as opposed to the risky, fickle backing they have now?

Joris van Dorp's picture
Joris van Dorp on May 28, 2014 3:09 pm GMT

I worry every time I see such estimates for new nuclear.

I worry that overreliance on the (currently true) $6000 figure is obscuring the rather likely expectation that building hundreds of identical nuclear plants as opposed to the current handful is going to result in an average cost which is far lower than $6000.

Is there any analysis which attempts to quantify the cost-reduction effect (‘learning effect’) of building hundreds of AP1000’s or EPR’s? I’ve not seen any. Until I do, I am going to assume that building hundreds of EPR’s in for example Europe will yield an average cost of less than $3000/kW. I’ll even assume that the 100th EPR or AP1000 will likely cost less than $2000/kW.

Anyone disagree?

Bas Gresnigt's picture
Bas Gresnigt on May 28, 2014 4:30 pm GMT

Your current $6000 figure is old.
Taking into account the delays, subsidies etc. that figure is now ~$9000,=.

In 2013, the US Energy Information Administration estimated the levelized cost of electricity from new nuclear power plants to be $108/MWh.

As that is twice the whole sale market price in competitive markets (~3 times the German price at the Leipzig exchange), no one will want to invest.
Only governments that have the guts to spill the money of their tax-payers, such as UK, Georgia.

Bas Gresnigt's picture
Bas Gresnigt on May 28, 2014 6:03 pm GMT

The cost estimates for new nuclear in UK may be in line with the UK report, but if you look at the data for the planned EPR at Hinkley those are unrealistic low. The report is probably biased in order to support government policy. The Hinkley data:

1. Government guarantees £92.50/MWh (based on 2012 pounds, inflation corrected) for all produced electricity during 35years. With 2.5%/year inflation that implies a guaranteed price of:
 – £121/MWh (=$200/MWh) in 2023 at the start of the new NPP
 – £184/MWh (=$300/MWh) in 2040 halfway the subsidy period.

2. Government delivers loan guarantees for £10billion of loans. So the risk premium of ~4%/a is carried by the tax payer and not the operator (EDF). 
A subsidy worth another £18/MWh (90% load factor).

3. Government takes all “excessive” decommission and waste costs as well as accident costs…
You can estimate those between £10/MWh and £100/MWh depending whether you are optimistic or pestimistic.

So the real costs of the new NPP in UK are at least:
 – £149/MWh (=$246/MWh) in 2023 at the start of the new NPP
 – £21284/MWh (=$350/MWh) in 2040 halfway the subsidy period.

Compare that with the decreasing costs of:
– onshore wind; German FiT €88/MWh (=$120/Mwh) during 15years.
– PV-solar; German FiT €90/MWh (=$122/MWh) during 20years
no risk guatantees, etc. After those periods whole sale prices that are now at $50/MW level, while futures (e.g. at the market at Leipzig) show a long term trend to towards even lower prices.

Bob Meinetz's picture
Bob Meinetz on May 28, 2014 7:34 pm GMT

Bas, I’m not sure why you link to the Wikipedia entry for EIA – it offers no validation for your figures.

In fact, neither does the EIA website. The 2014 EIA cost estimate for advanced nuclear coming online in 2019 is $96.1/MWh, 11% lower than your figure.

The estimated levelized cost for 2019 PV solar is $130/MWh – 26% more expensive than nuclear. Of course, that doesn’t include fossil fuels which must be burned as backup.

Germany once “had the guts” to waste its taxpayers money, but seems to be losing its nerve with Vice Chancellor Sigmar Gabriel recently admitting that its celebrated renewables Energiewende is on the verge of failure.

Bas Gresnigt's picture
Bas Gresnigt on May 28, 2014 8:25 pm GMT

German utilities proposed last week to transfer all NPP’s (incl.the €30billion in the decommissioning fund) for free into a government foundation that would do the job.
Government reject.

My rough estimation is that the construction costs of those NPP’s were roughly €30billion…
Apparently utilities and government estimate that the decommissioning costs are much higher…
So you should multiply your decommissioning cost estimation with at least a factor 10 (not 10% but 100% of construction costs), probably with a factor 20 or more…

So if the decommissioning fund contains only 10% of construction costs, the tax-payer will have to step in again and subsidize ~95% of those huge costs.

Just remember that the costs to clean up the nuclear waste pile at Sellafield are estimated to be around £100billion (to be carried by the UK tax-payer)!


Bas Gresnigt's picture
Bas Gresnigt on May 28, 2014 9:23 pm GMT

Sorry. My mistake. I copied the whole sentence incl. the link from this wikipedia page:
(it is in the last paragraph under the heading “cost per KWh”)
Thanks for your correction.

I thought we discussed the exceptional high cost estimate for PV solar by EIA already.
Just compare:
– the recent deal by the Austin (USA) utility; 150MW of PV solar for $50/MWh for the next 20years!
– the German solar Feed-in-Tariff for bigger PV-solar installations (<10MW*): €90/MWh = ~$120/MWh.
And realize that:
– those Feed-in-Tariffs go down with ~1% per month, long term trend -8%/year.
So in 2019 those will be €60/MWh (=$80/MWh).
– Germany is in general far more north than USA (more at the latitude of Canada), hence far less sunlight. So the USA cost price ($/MWh) should be cheaper assuming US installers at that time have similar efficiency as German installers.

So I think that the EIA estimate is far to high. It should be ~$86/MWh in 2019 assuming US installer are then still ~20% less efficient than German installers.

*) installations >10MW have no guaranteed FiT in Germany, so they have to arrange their own commercial sales.

Nathan Wilson's picture
Nathan Wilson on May 29, 2014 6:40 am GMT


The purpose of the inflation adjustment in the Hinkley electric contract is to keep the electricity price the same (when measured in 2013 dollars), so Bas’s use of the inflation “correction” actually serves to misrepresent the future affordability of energy from the plant.

Assuming the loan guarantee is free (unlike the US government loan guarantee for Vogtle, which is sold to the investors at a fair market price), removing a risk premium would reduce the cost of energy.  But if the risk is due to policy uncertainty (which is likely the major risk), then the cost that is eliminated is not intrinsic to nuclear power (but rather is a cost associated with inconsistent policy), so it not fair to add this to the cost of nuclear.

In the US, each plant must a have decomissioning fund (funded by electricity sales) which is adequate to decomission the reactors.  The appropriate size of this fund is determined by regulators.  Because the decommissioning funds builds interest over so many decades, the impact on energy cost is very small.  The worst that is likely to happen if the fund is too small at plant closing, is that the decommissioning must be delayed while interest accumulates in the fund.  

The waste disposal fee is also determined by government regulators (which has historically been only $1/MWh in the US).  The biggest risk here is political (since the political process can be used to impose unnecessary costs on any industry), so again it is inappropriate to assign these costs to nuclear, since it is all political whim (i.e. our peer nations like China will not have policy risk here).

The Hinkely Point C reactor is effectively a first of a kind unit in the UK (3rd for all of Europe, but much work will be done by UK suppliers).  It is unlikely that future reactors will cost this much.  The learning effect is universal, and will surely be apparent for nuclear if deployments ramp up.  

The latest (as of May 2014) EIA data for the US predicts of nuclear cost of $96.1/MWh, which I believe is their forecast assuming little nuclear industry growth.

Nathan Wilson's picture
Nathan Wilson on May 29, 2014 4:59 am GMT

Again, I will point you to a list of reactors here which have already been decommissioned, at an affordable cost, so we know it is possible.  Of course we also know that Germany has chosen to use the nuclear industry as a political whipping boy and has economically slapped them around with special taxes and fees, and will likely continue to do so.  So I think it is completely fair to pass the economic risk of polical harassment on to the harassers.

Bas Gresnigt's picture
Bas Gresnigt on May 29, 2014 12:48 pm GMT

You really amaze me again.
How can you think that?? Considering that:

– during the last years polls show more support for the Energiewende than ever before. Unprecedented high levels of ~90%. So a German politician said that it would be political suicide to propose any postponement of the closure of the NPP’s.

– last autumn elections whiped out the only party that supported postponement of the closure of the NPP’s (FDP), out of parliament.

– EON wants to close its Grafenrheinfeld NPP in spring next year, 7 months earlier than agreed with government.

– German utilities offered all NPP’s to a government owned fund for free incl. the €30billion in the decommision fund.
So they consider their NPP’s now to be a burden.


Bas Gresnigt's picture
Bas Gresnigt on May 29, 2014 1:58 pm GMT

Nathan,”…purpose of the inflation adjustment in the Hinkley electric contract is to keep the electricity price the same..

The issue is that the cost prices of wind and solar electricity go down during next decades as they did in past decades. Wind with ~3%/a, solar with ~8%/a. Other renewables, such as geothermal have similar trends.

So in 2023 German FiT for:
– solar will be ~€42/MWh going down further towards <€30/MWh in 2030
– wind will be ~€67/MWh going down towards ~€40/MWh (when ~20MW turbines are the standard).

As the guaranteed price for Hinkley goes up from €121 in 2023 towards €144/Wh in 2030, that implies that with normal whole sale markets UK government has to subsidise 50% in 2023 growing towards >70% in 2030, and more thereafter. 
Probably paid by the rate payers via a nuclear levy on their electricity bill.

While at that time the Energiewende levy in Germany will be declining towards very low levels.

UK government tries to hamper this development by allowing only offshore wind and mimimizing solar.
So UK rate payers will pay anyway far more than German rate payers after ~2030.

You can see this development already comoing in the whole sale price developments.
Those of the UK are much higher than on the continent (so our connection to UK is mainly utilized to export electricity which we buy in Germany).

Nathan Wilson's picture
Nathan Wilson on May 30, 2014 3:36 am GMT

UK government tries to hamper this development [falling electricity prices due to cheap wind and solar]…”

Why would the government do that?  I don’t believe it is because they have are stupid or love wasting money.  The best explanation I can come up with is that they simply do not agree with your projections of cheap wind and solar at high penetration.  Perhaps they have noticed that winter sunshine is inadequate that far north, and run calculations that show that smoothing wind power fluctuations with advanced batteries is extraordinarily expensive.  Maybe the political support for on-shore wind power is falling in the UK due to aesthetics.

Schalk Cloete's picture
Schalk Cloete on May 30, 2014 11:24 am GMT

As requested in the article, please give only one number as your best estimate for the internalized costs of new nuclear. You can separate developed and developing world costs as I did below if you want. Otherwise I cannot include your opinion in the calculation of the TEC consensus on nuclear energy costs. 

Also please note that wind/solar power is off topic for this article. Some new articles especially for these energy sources will come out soon where everyone can weigh in on the topic of internalized costs of wind and solar power. 

Schalk Cloete's picture
Schalk Cloete on May 30, 2014 11:42 am GMT

As wind/solar is off-topic in this article, I suggest that we continue this discussion below another article I wrote where a very interesting paper calculating the lowest cost electricity mix for Northwest Europe was reviewed. That paper showed an optimum long-term share (assuming the capacity mix is perfectly optimized to accommodate intermittent renewables) of 20% onshore wind and 2% solar assuming 30% further cost reductions for wind and 60% further cost reductions for solar. For the existing capacity mix (sunk investments), the optimial wind share is 7% assuming a further 30% cost reduction. These values assume a €20/ton CO2 price. As the CO2 price increases to very high levels, the optimal share of intermittent renewables decreases as wind+fossil backup becomes more expensive than dispatchable nuclear. 

Schalk Cloete's picture
Schalk Cloete on May 30, 2014 11:51 am GMT

Future internalized technology costs will be the next criterion tackled by the Seeking Consensus column. Each criterion will take roughly 6 months to complete, implying that we will get to the future costs of nuclear in about 6 months’ time. Then everyone can weigh in with their cost projections for future decades. 

Schalk Cloete's picture
Schalk Cloete on May 30, 2014 11:53 am GMT

Let’s try to keep the discussion focussed on nuclear power for this article. Wind/solar will get their turn soon. 

Schalk Cloete's picture
Schalk Cloete on May 30, 2014 11:57 am GMT

As pointed out in a comment above, future technology costs will be the next assessment criterion covered in this column. The future costs of nuclear will be up for discussion roughly 6 months from now. There we can discuss at length about the opposing views of historical increasing cost trends vs. possible significant cost reductions due to new technology and a structured developing world rollout. 

Bas Gresnigt's picture
Bas Gresnigt on May 31, 2014 12:56 pm GMT

New nuclear in W-Europe.

Data: Capital
The new (EPR; 1600MW) reactors in W-Europe will cost at least €8billion.
Capacity factor of ~75% *). So av. production: 1200MW.
Capital costs: €6,667/KW or $9,067/KW **).

Those reactors will operate in an environment with a substantial solar+wind share (in Finland the CF may be higher as little sun+wind). The variable costs of Solar+Wind are below $1/Mwh. With such low prices even NPP’s bring their production down to the minimum as German experience shows.
That implies lower capacity factors mainly due to economics.

Of course costs in e.g. China are, just as PV-panels and Wind turbines are cheaper in China, mainly thanks to low wages.
Due to those difference you can only compare within a region. There are even big costs differences between USA and Germany regarding PV-installations. Those in Germany are much (~30-50%) cheaper while using the same China produced PV-panels.

Data: Cost / Kwh
1. Operating costs.
The capacity of the German Grafenrheinfeld NPP (totally written off; no capital costs), is near the EPR: 1345MW. The owner, E.ON. anounced that it will close the reactor prematurely (spring 2015) due to its bad financial results.
So the costs of only operating are more than the average whole sale price in Germany.
The German year-ahead whole-sale price is ~€37/MWh = $50/MWh.
So similar amount for the operating costs of the new EPR.

2. Capital costs, depreciation, decommissioning
With the increasing share of renewable and continued price fall of those, it is extremely risky to assume a longer life than the 35years that UK government guarantees for Hinkley.
Especially since even a fully depreciated 1300MW reactor cannot compete in Germany while the share of wind+solar will be only ~17% in 2015.

That implies: depreciation: 2.86%. The EPR will produce 10.512GWh/year.
So depreciation: €22/MWh = $30/MWh.

6% interest, incl. an optimistic risk premium (check history and bleak prospectives) delivers:
€23/MWh = $31/Mwh

Decommissioning: €5/MWh = $7/MWh. Rather low considering the size of the German decommissioning fund of €30billion, which is considered as not enough.

Assume the major waste storage costs to be paid by tax-payers. Same as now in UK (Sellafield costs ~£100billion), France and Germany. A difficult to quantify subsidy. I once did a try and arrived at $50/Mwh.

These optimistic low estimates add-up to: €87/MWh or $118/Mwh.

Why optimistic?
Even Grafenrheinfeld does not reach 35years economic viability (E.ON), while no wind & solar in 1975 (start construction) or even 1982 (start operation).
The price of those came down with a factor 10-100 while predictions are that this will continue for next decades (especially solar).
Wind & solar now already compete at/below this estimated cost-price (look at the Austin deal).
The 35years assume the EPR will still be competitive in ~2060 (10years building period) with these prices. Seems unrealistic to me. I estimate that there is <10% chance that NPP’s will still operate after 2040 in a competitive environment. Even UK government / tax payers are then probably so tired of  paying the huge subsidies to Hinkley, that they will seek a way to stop it. 

I assumed that tax-payers will continue to take the costs of final waste handling / storage.

This price is substantial lower than UK government guaranteed to EDF for Hinkley: 
€114/MWh = $155/MWh (2012 GBP’s, inflation corrected, first correction in 2013, etc).
While UK also delivers many other substantial subsidies as stated in my previous post. 

Bas Gresnigt's picture
Bas Gresnigt on May 31, 2014 1:03 pm GMT


Those other technologies MSR, etc) were already known in the sixties. At that time nobody chose to develop those further, so you may assume that general estimations were that those would be not-cost effective   / could not compete against PWR / LWR.

(Fast) breeders were developed, but those turned into very expensive adventurous failures in all western countries; France (Phenix and SuperPhenix), Japan, Germany.

So I think your estimation is rather optimistic.


Nathan Wilson's picture
Nathan Wilson on May 31, 2014 8:18 pm GMT

Say Schalk,

I think it would be very helpful if we had a discussion on energy storage before proceeding with wind and solar.  It is very important to draw a distinction between renewable+storage and renewables+coal, both for cost and environmental impact.

Nathan Wilson's picture
Nathan Wilson on Jun 1, 2014 12:31 am GMT

REBUTTAL:  The linked article on the Germany Grafenrheinfel nuclear power plant, didn’t say the plant was being closed due to “bad financial results“, but rather due to onerous taxes: 

If E.ON want to operate the power plant to the designated maturity date, approximately 80 million euros fuel tax would have been incurred.”

Given that only 7 months were left in its licensing period, that amounts to €15/MWh just for taxes (assuming it can achieve 80% capacity factor in spite of solar and wind getting priority grid access)!

So it is similar to the American reactors at San Onofre and Vermont Yankee, which were effectively bullied out of business by the government.  These market-distorting government policies can gain public support and votes in the short term, but in the long term, the truth will show up as higher energy costs, weaker economies, and less relevance of western nations in the future world economy.

Schalk Cloete's picture
Schalk Cloete on Jun 1, 2014 5:15 am GMT

Thanks for the detailed post, Bas. This one will definitely go into the stats. 

Incidentally, I get $118/MWh LCOE under the assumptions of $9067/kW capital costs, 80% CF and 6% discount rate. I can therefore confirm your calculations.

It really is amazing how much nuclear power costs vary across the world. The Indian reactor that Nathan referred to below is more than 5 times cheaper to buid than the European example you cited. I therefore agree that the future of nuclear in the developed world is quite bleak if costs are really as high as your sources suggest. Capital-intensive technology with low running costs is normally a good idea in stagnant economies where the time-value of money is very low, but capital costs approaching $10000/kW might be pushing it a little too far.

Schalk Cloete's picture
Schalk Cloete on Jun 1, 2014 5:25 am GMT

I agree that it could be helpful to have a special post about the internalized costs of various types of energy storage mechanisms. It could lead to some interesting discussions and can be a handy reference in future posts. I’ll set it up. Thanks for the suggestion. 

Bas Gresnigt's picture
Bas Gresnigt on Jun 1, 2014 1:13 pm GMT


I copied the statement about financial results from the Wikipedia page to which the first link in my post refers. The concerned chapter states at the end: “The reason given was the lack of profitability”. 

Note that this NPP was part of a study that showed significant (20%) increased frequency of cancers in children living in the surroundings of the NPP’s in Bayern. Not clear how many died.

This is in line with studies that showed significant changed sex-ratio in new borns for people living in the surroundings of NPP’s in Germany, Switserland and France. That changed sex-ratio indicates enhanced levels of DNA damage in offspring, which is in general rather detrimental.

Not sure why utilities opposed vehemently installing a good continuous radiation measurement network (also measuring fast sub-atomic particles) around their NPP’s and radio-active waste storages, as that may have cleared them from the accusation being the cause of these increased health damages for next generations.

Bas Gresnigt's picture
Bas Gresnigt on Jun 1, 2014 1:50 pm GMT

Renewable+coal (or lignite) is not an option that any country targets as far as I know.

Certainly not Germany! Also shown by the decrease of electricity produced using coal+lignite (check the stromerzeugungs tabel) since the Energiewende started in 2001, and
the big increase in renewable; from 7% towards 24%.
Despite the closing of the majority of their nuclear fleet (10 NPP’s since 2003; the other 9 will be closed before 2022).

Closing NPP’s has priority as those are considered to be far more damaging, especial for offspring and next generarions.

Joris van Dorp's picture
Joris van Dorp on Jun 2, 2014 8:52 am GMT

“- during the last years polls show more support for the Energiewende than ever before. Unprecedented high levels of ~90%. So a German politician said that it would be political suicide to propose any postponement of the closure of the NPP’s.”

90% may support the principle of the energywende, but support for more windturbines and powerlines, and higher energy costs is far less in Germany. Therefore, this statistic merely illustrates what happens when the public intelligence has been utterly destroyed by unmitigated anti-nuclear propaganda.

“- last autumn elections whiped out the only party that supported postponement of the closure of the NPP’s (FDP), out of parliament.”

See my previous comment.

“- EON wants to close its Grafenrheinfeld NPP in spring next year, 7 months earlier than agreed with government.”

The reason they want to close it is because the government has imposed unreasonable costs on exploiting and decommissioning nucleaer power plants, by special taxes and spurious conditions. Additionally, the lack of reasonable nuclear policy in Germany exposes the owner of a nuclear power plant to the manifest risk of being randomly and viciously attacked by unfounded accusations and slander from the powerful anti-nuclear lobby groups, causing costoverruns and delays in almost anything concerning nuclear power including decommissioning. Operating and decommissioning a nuclear power plant under such antagonistic and unscientific societal conditions is obviously far more risky than originally anticipated when Germany was one of the leading countries in terms of nuclear knowhow and development. So again, your soundbite has absolutely nothing to do with the businesscase of operating a nuclear power plant, especially one which has already been amortised.

“- German utilities offered all NPP’s to a government owned fund for free incl. the €30billion in the decommision fund. 
So they consider their NPP’s now to be a burden.”

In a country which is racked by the most grotesque anti-nuclear propaganda seen in any country for decades, and the resulting unprecedented state-mandated sabotage of the nuclear business case through special taxes and unreasonable operating constraint – which is what Germany has sadly become – of course nuclear power plants are eventually seen to be a burden by the utilities which own them. Again, this has nothing to do with the fact that exploiting a nuclear power plant throughout its life cycle is extremely profitable and cost-effective for society under normal circumstances.

Nathan Wilson's picture
Nathan Wilson on Jun 3, 2014 4:46 am GMT

I’m not familiar with that study, but the claims that excess cancers have been detected as a result of routine operation of a Germany nuclear plant is quite unlikely to be true, given that normal reactors release orders of magnitude less reactivity than the Fukushima accident, and the Fukushima accident is predicted to have no measurable health impacts (see this United Nations committee report which predicts no detectable health impact from Fukushima radiation, or thisone from the WHO).

If there is any topic on which junk science has been published by people with an agenda, it’s anti-nuclearism.  Just because some self-proclaimed experts says it, does not make it true.  Scientists use the peer review and referee processes for a reason.  

Bas Gresnigt's picture
Bas Gresnigt on Jun 3, 2014 10:01 am GMT

You didn’t read your second link or chose to ignore the health damage stated in the second yellow box in your link:

4%-7% bigger risk to get cancer for chilren, 70% more risk for thyroid cancer.

Your link states further: “low risk” and “no observable health effects in wider population“.
That also implies health effects. Only numbers are difficult as Japanese government hinders good research (it is better to prevent bad results), just as Ukraine, Belarus and Russian governments did.

Those health risks are very difficult to observe as extra low level radiation generates extra cancers, etc:

– after a latency of decades (20 – 60 years as shown by the RERF/LSS and medical studies;
Similar as low level smoking, asbestos, etc.

– Fukushima radiation of 10mSv (3-6times natural yearly background) delivers ~1% more cancers and other serious deseases, creating ~0.6% more premature deaths. Those are of course extremely difficult to detect as: 

– those extra cancers do not differ from other cancers, so detection that the extra Fukushima radiation is the cause is difficult and will be refuted by pro-nuclear (often without substantial ground), or ignored.
Even rock-solid studies that showed the 60% rise of Down syndrome, malformations, stillbirth, etc. in newborn after Chernobyl only in districts that got some fall-out (0.5mSv/a) in Germany (>1000miles away) are ignored.

– that 1% may seem low but with >100,000 people affected it concerns >1000 people.

Furthermore the WHO chose to ignore genetic effects (showed after Chernobyl) that affect quality of next generations, and ignored even the health effects on fetuses.
Fetuses are very vulnerable as also shown medical studies (so in hospitals no medical radiation of the womb when the woman is pregnant), due to their fast rate of cell division (during cell division the DNA is single stranded and cannot be repaired).

But compared to the WHO of 2006 that simply followed IAEA (prime target; more nuclear & NPP’s) in its ridiculous statements as a slave (due to the 1959 agreement), this WHO report shows the organisation makes significant progress.

The studies I refer are published in peer reviewed scientific journals.

Engineer- Poet's picture
Engineer- Poet on Jun 3, 2014 3:06 pm GMT

There are people living, happily and apparently in good health (for their age), in the Chernobyl exclusion zone.  Look up “Chernobyl babushkas” if this is news to you.

Aside from a few hot spots, the radiation levels in the Fukushima evacuation zone are half or less the levels in areas which have been inhabited for centuries with no evidence of harm to humans.  The beaches at Guarapari (Brazil) and Kerala (India), and the springs in Ramsar (Iran), expose people to as much as 260 mSv/yr.  There is no significant radiation health threat remaining in Fukushima, outside the bounds of the plant itself.  There is only nuclear paranoia.  Sadly, the Japanese have this in abundance, even turning it into a profitable media franchise.

Nathan Wilson's picture
Nathan Wilson on Jun 4, 2014 4:17 am GMT

Looking back at the press report on the WHO article, it still appears to me that they are saying the Fukushima health impacts will be too low to measure (for the small group of most exposed people, the lifetime cancer risk grows to 30.4%, whereas the baseline in Japan is 29.0%, a difference too small to measure in a small group).  Regarding other medical effects of radiation, the WHO said the effects of the accident are not expected to cause an increase in the incidence of miscarriages, stillbirths and other physical and mental conditions that can affect babies born after the accident.”

But in calling this a large effect, you forgot to account for the health impacts of the coal that was not burned, and also, that the impacts of the Fukushima accident represent nearly the total impacts of all of the 375 GWatts of Gen II nuclear power that has served the world for decades.  See NextBigFuture-Deaths_per_TWatt-Hour, which finds the nuclear is the safest.

And since no one is building Gen II reactors anymore, for comparison to other new-build energy sources, we should use the Gen III safety forecasts, which are a couple of orders of magnitude better.  The net impact of nuclear is even better when we consider that it is normally used by US utilities as an alternative to coal (whereas renewables most often reduce natural gas consumption).  

And of course if we are to make a fair comparison, the low-carbon alternative to nuclear is really a 50/50 blend (a little better or worse depending on your optimism) of renewables and flexible generation (i.e. fossil fuel).   By this metric, variable renewables are nearly as dirty and dangerous as fossil fuel.

Nathan Wilson's picture
Nathan Wilson on Jun 4, 2014 3:20 am GMT


The Fukushima accident produced the vast majority of harm from the entire world’s 375 GWatt nuclear fleet, for two decades.  What would be the external costs of replacing all of that energy with fossil fuels?  Keep in mind that fossil fuel kills over 10,000 Americans each year, and degrades the health of many, many others.  So any plausible mix of proven renewables and fossil fuel would also loose.

Also, were it not for the baseless fear and mis-information spread by anti-nuclear activists, the Fukushima evacuations could have impacted a lot less people for a much shorter time period.  

Joris van Dorp's picture
Joris van Dorp on Jun 4, 2014 9:29 am GMT

Even rock-solid studies that showed the 60% rise of Down syndrome, malformations, stillbirth, etc. in newborn after Chernobyl only in districts that got some fall-out (0.5mSv/a) in Germany (>1000miles away) are ignored.”

It saddens me that you continue – yet again – to refer to this study – which has been debunked in the scientific literature. It is NOT rock solid. It does NOT show the horrible effects you mention. It does NOT support your arguments.

What it does do is show yet again how agenda-driven your antinuclear propaganda is, Bas Gresnigt. It’s disgraceful. Again, I sincerely feel that I must protest. I again urge TheEnergyCollective to formulate and enforce basic rules of civil discussion, including a rule that prohibits the willfull promotion of grievous nonsense when it has been pointed out multiple times. Please consider this. We cannot have people like mr. Gresnigt sabotaging these important discussion by his unscientific agenda-driven propaganda. Thank you.

Readers who are interested in the actual status of the typical ‘rock solid studies’ that Bas Gresnigt is perpetually trying to push into serious discussions may find some answers here. It is a comment on the ridiculous ‘study’ posted by mr Gresnigt which purports to show impacts on children from chernobyl fallout. It demonstrates what Bas should have told us, namely that the authors of this ‘study’ are known for data-mining in support of their (and Bas’s) crusade against nuclear power at any cost.

Abstract: “The recent claim made in this journal that nuclear bomb tests and the Chernobyl disaster caused distortions in the secondary sex ratio is shown to be a likely artifact of data mining, misused statistics, and misreading of the evidence. In particular, the concept of statistical “significance” and its limitations do not seem to be fully understood, and important confounding factors have not been accounted for.”

Please excuse me, but I am a person who gets sick to his stomach at the sight of mr Gresnigt continual anti-nuclear propaganda, which costs millions of people including many women and children their lives, and gravely threatens humanity and our common future. I simply cannot remain silent on this crucial issue. Thank you.


Bas Gresnigt's picture
Bas Gresnigt on Jun 4, 2014 10:41 am GMT

You state: “..this study – which has been debunked in the scientific literature.”
But no link/reference??

Joris van Dorp's picture
Joris van Dorp on Jun 5, 2014 11:33 am GMT

What are you talking about. The reference is right there in my comment:

Abstract: “The recent claim made in this journal that nuclear bomb tests and the Chernobyl disaster caused distortions in the secondary sex ratio is shown to be a likely artifact of data mining, misused statistics, and misreading of the evidence. In particular, the concept of statistical “significance” and its limitations do not seem to be fully understood, and important confounding factors have not been accounted for.”

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »