Nuclear capital costs, Three Mile Island and Chernobyl

Lovering, Yip and Nordhaus (Science Direct April 2016) reviewed construction cost data for 349 reactors in the US, France, Canada, West Germany, Japan, India, and South Korea, encompassing 58% of all reactors built globally, and concluded that there is no inherent cost escalation trend associated with nuclear technology. There is however a vast variation in construction costs from one country to another. Some countries like the USA, Canada, Japan and W Germany responded to the Three Mile Island accident by imposing regulations that pushed construction costs through the roof while France, S Korea and India did not. S Korea and India are still able to deliver nuclear power stations for $2 billion / GW ($2010) installed capacity which remains a small fraction of the capital cost of solar PV.

About half way down the list of articles in Blowout Week 113 was an abstract from a just-published paper entitled Historical construction costs of global nuclear power reactors , authored by Lovering, Yip and Nordhaus (hereafter LYN). It contains some interesting data which are worth summarizing in a post.

LYN reviewed cost data for “349 reactors in the US, France, Canada, West Germany, Japan, India, and South Korea, encompassing 58% of all reactors built globally”, and concluded that “there is no inherent cost escalation trend associated with nuclear technology”. Their results, however, allow us to deduce a little more than that, and here we will review them, starting with LYP’s Figure 12, reproduced below as Figure 1:

Figure 1: Overnight construction costs of global nuclear reactors

It plots overnight construction cost in 2010 US dollars against the date of construction start for all 349 reactors in the seven countries The most prominent feature is the cluster of blue points that extends skywards after about 1970. These are the US reactors that had the misfortune of being under construction at the time of the Three Mile Island accident in 1979. What happened in other countries is a little harder to see and we will look into it shortly, but first, what is overnight construction cost? I quote from LYN:

The Overnight Construction Cost (OCC) includes the costs of the direct engineering, procurement, and construction (EPC) services that the vendors and the architect-engineer team are contracted to provide, as well as the indirect owner’s costs, which include land, site preparation, project management, training, contingencies, and commissioning costs. For heavy-water reactors, the OCC includes the cost of the initial heavy-water inventory. The OCC includes back-fit costs but excludes retrofitting or capital expenditures after first operation and the cost of the initial fuel core. The OCC represents the single largest component of the total levelized cost of generating electricity with nuclear power, typically accounting for roughly 55%. In this study, we focus exclusively on OCC because the other lifecycle costs – approximately 15% for Interest During Construction (IDC), 15% for O&M and decommissioning provision, and 15% for fuel and provisions for used fuel – are more predictable and have had far less variation over time and country.

I don’t know whether this is the only way of doing it but at least LYP seem to have done their homework.

Figure 1, however, shows some apparently large differences between the countries which are nevertheless hard to pick out. Figure 2 shows overnight nuclear costs by country put together by overlaying heavy black squares over the dots shown in Figure 1 to make these trends more visible (the US is not included because the trend is already clearly visible in Figure 1):

Figure 2: Overnight costs by country

The three countries in the first row – Canada, Japan and west Germany – show significant cost escalation with time, with Canada showing an abrupt increase after 1980 and Japan and Germany showing abrupt increases in the mid-1970s. In the second row France and India show cost escalation over the same period but lower and much more stable costs overall. South Korea shows costs decreasing slightly since 1980 but has no pre-1980 data.

There are all kinds of country- and design-specific factors that will have influenced these costs, some of which are discussed by LYP, and a detailed recapitulation of them is beyond the scope of this post. Nevertheless it seems that something happened to increase costs for most reactors that began construction after about 1975, and the obvious culprit was the 1979 Three Mile Island accident, about which the IEA had this to say:

The psychological effect on the population in the neighbourhood, and eventually throughout the Western world, was immense. So was the damage to the plant itself and to the reputation of the nuclear power industry.

The effects of Three Mile Island were, however, immense only in some western countries. The US and some other countries went into regulatory tailspins that effectively stopped new nuclear development in its tracks, but France continued much as if nothing had happened, as did India. And South Korea, which began its nuclear program in 1980, was clearly totally unmoved.

The next nuclear accident occurred at Chernobyl in 1986, What was its impact? Figure 2 suggests that It might have contributed to the higher costs of the four reactors that started construction in Canada in the mid-1980s, but I’m not sure about that. The data for the US and Japan are scattered to the point where it’s difficult to say whether Chernobyl did anything or not, and Germany’s nuclear construction program was pretty much over by then. The impacts on India and South Korea were, however, negligible.

Chernobyl did have an impact on nuclear construction in France, although not a major one. As shown in the first graphic in Figure3 it increased construction lead times but didn’t increase costs. Contrast this with the second graphic, which shows the reaction of the US to Three Mile Island:

Figure 3: Overnight costs and construction duration before, during and after Three Mile Island accident, France and USA

Two important nuclear countries that are not included in the LYN analysis are Russia and China. In an attempt to fill the gap Figure 4 shows installed nuclear capacity growth in these countries with the X-scale shifted 5 years to the left to simulate a constant five-year construction lead time (data from the World Nuclear Association). Three Mile Island passed unnoticed in Russia, but new construction came almost to a halt after Chernobyl, remained depressed after the collapse of the Soviet Union in the early 1990s and is only slowly beginning to recover. China, on the other hand – well, China is China. Any annual growth rate of less than 10% is regarded as a result of failed economic policies.

Figure 4. Growth of installed nuclear capacity in Russia and China. To make the plot as comparable to other plots as possible the years shown are approximate “dates of construction start” estimated by subtracting five years from the year in which the plant went into operation.

And to round things off here’s a brief summary of overnight nuclear capital costs in Russia and China in comparison with other countries, again from the World Nuclear Association:

Nuclear overnight capital costs in OECD ranged from US$ 1556/kW for APR-1400 in South Korea through $3009 for ABWR in Japan, $3382/kW for Gen III+ in USA, $3860 for EPR at Flamanville in France to $5863/kW for EPR in Switzerland, with world median $4100/kW. Belgium, Netherlands, Czech Rep and Hungary were all over $5000/kW. In China overnight costs were $1748/kW for CPR-1000 and $2302/kW for AP1000, and in Russia $2933/kW for VVER-1150. EPRI (USA) gave $2970/kW for APWR or ABWR, Eurelectric gave $4724/kW for EPR. OECD black coal plants were costed at $807-2719/kW, those with carbon capture and compression (tabulated as CCS, but the cost not including storage) at $3223-5811/kW, brown coal $1802-3485, gas plants $635-1747/kW and onshore wind capacity $1821-3716/kW. (Overnight costs were defined here as EPC, owner’s costs and contingency, but excluding interest during construction.)

Finally, LYN’s Figure 13, which compares overnight nuclear costs with overnight solar costs as a function of total world installed capacity, is reproduced as Figure 5 below. Solar costs show a rapid decrease with increasing installed capacity but show signs of flattening out. Nuclear costs, on the other hand, show no clear trend with time, which is why LYN conclude that “there is no inherent cost escalation trend associated with nuclear technology”. The most interesting feature, however, is that solar still has no clear capital cost advantage over nuclear, which with nuclear capacity factors five or six times higher than solar capacity factors should mean that a levelized cost comparison will be a no-contest;

Figure 5: Overnight nuclear and solar costs as a function of global installed capacity

So what’s the bottom line? Basically that nuclear power is expensive only if a country chooses to make it so.

This entry was posted in Energy and tagged , , . Bookmark the permalink.

179 Responses to Nuclear capital costs, Three Mile Island and Chernobyl

  1. Euan Mearns says:

    This leaves a burning question about Hinkley C. Cost = £18 billion = $25 billion. Capacity = 3.2 GW = $7.8 billion / GW. EDF and the government really need to get a grip of this situation and fast. I blame Ed Davey.

    • Alex says:

      Interestingly, Ed Davey is in today’s Guardian saying Hinkley is a good deal and Osborne would have settled for a higher price. I don’t quite buy that.

      The real problem was that the EPR was (and still is) the only reactor approved by the ONR. If the ABWR and AP1000 had been approved, the Government might have been able to negotiate closer to £70/MWh, reflecting an over night cost of about $6000/KW.

      • Leo Smith says:

        CANDU took one look at the hoops to be jumped through and walked away…

        • Peter Lang says:

          In that case no electricity is safe enough because all other ways of generating electricity are less safe.

          You should be arguing for: no cars, trains, plains, bicycles, hospitals, manufacturing, chemicals or anything else because all are much less safe than nuclear power, even Gen II nuclear power.

          Do some objective research and seek professional help to get over your irrational paranoia.

          • nikopol92 says:

            Dis you read m’y tweet ? I mean that EPR is SAFER than AP1000 So Thé privé is not only important point as I read jour post 😉

          • Alex says:

            It’s generally reckoned the AP1000 is safer than the EPR, with a higher level of passive safety. That said, they’re both safer than just about any alternative.

            What you are highlighting is that the EPR is more attuned to French regulations than the AP1000. That is not surprising at all.

          • Peter Lang says:

            I responded to your comment which said “AP1000 is not safe enough for Europe…”

            Your comment is irrelevant because nuclear power is the safest way to generate electricity – any nuclear! Nuclear is 66% safer than wind and 5x safer than rooftop solar PV.

            only nuclear would survive if all technologies had to insure for the fatalities they cause. To understand this let’s estimate how much would society need to subsidise nuclear, or penalize other electricity generators, to equalize the compensation costs so all technologies pay for the fatalities they cause? Viewed another way, how much would we need to subsidise nuclear to reward the comparatively higher safety of nuclear power?

            A rough calculation suggests we should subsidise nuclear by $140/MWh to substitute for coal-fired generation and $37/MWh to substitute for gas fired generation in the USA (it’s different in each country). In that case, consumers should be paid around $50/MWh to consume nuclear generated electricity – “nuclear too cheap to meter” would be correct, except it would have to be metered to pay the subsidies to the consumers.


            That gives you a ball park quantification of the enormous safety advantage of nuclear over any other electricity generation technology.

          • Mark Pawelek says:

            Living without technology is more dangerous. Longer lifespans correlate well with wealth, which aligns with high energy use.

          • Peter Lang says:


    • climanrecon says:

      Sir Ed was on radio 4 “Today” this morning, defending the project and its price, apparently it would have cost even more if it had been left to those Tories. For Ed “low-carbon” justifies any price: “who knows what the cost-of-carbon” will be in 50 years, he says.

      For Lib Dems fossil fuels will inevitably be expensive in the future, either from scarcity, of from carbon taxes, or both. That was the basis for giving wind a license to print money.

      • Mark Pawelek says:

        That would be the same Ed Davey who believes wind power will certainly beat fossil fuels on price by 2020 [ according to Dieter Helm ]. Neither Ed nor the rest of the Lib Dems are famous for being pro-nuke. The whole party was anti-nuclear power until their 2013 Con: If I didn’t know better, I might assume the astronomic cost of Hinkley C is a conspiracy to make the green claim – “nuclear power is too expensive” – true. I don’t think that. It’s just another example of our politicos having no idea.

    • Mark Pawelek says:

      Three Moorside AP1000 reactors are supposed to cost £10bn : US $4312/kW, 56% the cost of The AREVA EPRs.

      • Alex says:

        Mark, that would be good, but do you have a source for that figure?

        As shown in the USA, it is possible, if you really try hard, to build AP1000s for $7,500/KW.

        • Mark Pawelek says:

          You want a source from me but you give no source for your counter claim: AP1000 reactors will cost $7500/kW! I find $7500/kW for the AP1000 very dubious. Google “Moorside AP1000” in google news. £10bn for 3 × AP1000 reactors is the consortium’s claim.

          I’m not a big fan PWR technology but the AP1000 family look close to the best it gets in terms simplicity, and making sure meltdowns are next to impossible. It’s a very lean design compared to the AREVA EPR. It should cost a lot less to build. WNA say “A contrast between the 1188 MWe Westinghouse reactor at Sizewell B in the UK and the modern AP1000 of similar-power illustrates the evolution from 1970-80 types. First, the AP1000 footprint is very much smaller – about one-quarter the size, secondly the concrete and steel requirements are lower by a factor of five*, and thirdly it has modular construction.”

          For comparison sake. In UAE, KEPCO charge US $5bn to build an APR1400, so 3 × APR1400 = £10.55bn. WNA claim for the APR1400 “Projected cost at the end of 2009 was US$ 2300 per kilowatt, with 48-month construction time.” KEPCO claimed $2300/kW 7 years ago. With a superior design, Westinghouse should be able to do it for less than twice the price.

    • Mark Pawelek says:

      Edf said they wouldn’t pour any Hinkley C concrete until 2019. The AP1000 will gain it’s ONR UK GDA within a year. Being smaller, less complex, with fewer components, and far less steel/concrete, the AP1000 will certainly take less time to build than EPRs. At this rate, Moorside will be on the grid before Hinkley C.

    • Mark Pawelek says:

      Change the UK regulatory framework to encourage advanced reactors such as molten salt reactors (MSR). An MSR is intrinsically safe because it is unpressurized, and, after an accident, will shut down without human intervention,

      Hinkley C EPR reactors : US $7770 / kW
      French 1970s/’80s nukes : < US $2000 / kW
      ThorCon MSR US $1000 / kW

      We could, in theory, build current technology (PWR, or BWR) nuclear reactors for US $2000 / kW in Britain. According to US nuclear startup ThorCon: ideally thorium MSRs can be built for as low as $1000 / kW, An MSR is more efficient because it runs a a much higher temperature (about 650 to 700C). It is intrinsically safe. In the event of an accident, it is designed to safely shutdown without human intervention. It faces no realistic disaster scenarios. Fuel burnup will be much higher, so the amount of waste much lower. The thorium fuel cycle will make far less problematic transuranics, so it is much cleaner. It is proven – a reactor similar to this ran for thousands of hours 45 years ago at Oak Ridge, USA. There are no technology hurdles stopping this; only regulatory, and institutional hurdles.

  2. Hugh Sharman says:

    Roger, thanks for this magesterial review! Now! How to get those dunderheads at Treasury and DECC to reform their pernicious regulations?

  3. Euan Mearns says:

    And a note on the CAPEX of solar and nuclear. Nuclear can be expected to run at 90% capacity while global solar perhaps 15% capacity. The capital costs must be adjusted for this which makes solar 90/15 = 6 times more expensive than nuclear to install. We can also add amortisation. A new nuclear station may last for 60 years while solar PV for 20 years. So we are now at 18 times more expensive for solar PV. And then there are all the ancillary costs of back up and load balancing. It is no wonder OECD economies are going down the tube.

    • K Periasamy says:

      This is something most of the people fail to appreciate. They blindly go by cost / MW of Installed capacity !

    • Nuclear gets permission for 40 years which can be extended to 60 years after another big chunk of capital was spend for renovation. So initial capital cost needs to be repaid over 40 years.
      On the other hand solar is projected to last at least 20 years with minimal maintenance costs. Currently there is no commercial PV plant with more than 10 years of exploatation record so we are here in realm of prognosis but with observed rate of capacity loss below 0,7% a year one can expect PV plant to generate reasonable amounts of power for over 50 years. It seems more probable that future replacement of PV modules will be performed to modernise the plant rather than to restore its capacity.

      • robertok06 says:

        “Nuclear gets permission for 40 years which can be extended to 60 years after another big chunk of capital was spend for renovation. ”

        Not true. A typical reactor upgrade aftere 30-40 years consists in new, more efficient turbines, or eventually replacement of the steam generators (for PWRs)… a “few” hundred million euro/dollars… in fact the main issue in that case is to do the work fast, as the loss of income during the reactor stop is what hurts the most.

        There’s plenty of examples one could give in support of what I’ve written.

        • K Periasamy says:

          Yes, very much true.
          The main systems / equipment like Civil structures, Reactor Vessel, Electrical & Control systems, Utilities, Safety systems, etc are not replaced. Hence, the refurbishment / up gradation does not cost much.

        • guber says:

          After 40 years you need completely new control systems, sindce for the old ones you do not get any spare parts any more. Which need to pass all safety and security issues, and must be adopted to the completely outdated inferfaces existing only in this old power station, but nowhere else any more in Industry. these parts are expensive, not tons of cheap concrete or steel.

        • Alex says:

          The French upgrade program has been estimated by a French court – probably a high end estimate – as costing €100 billion. This includes refurbishment and uprating of all the plants, and maintenance over the final 20 years.

          That comes out at about €2 billion per GW. That sounds like a big number to beat the nuclear industry with, but it’s for about 1000 GW-years, and comes out to about 1.2 cents per KWh.

          It illustrates that closing a nuclear reactor early, before it’s end of life, is financial and environmental vandalism. Only a country with a bonkers energy policy would consider it.

        • Mark Pawelek says:

          Nuclear gets permission for 40 years

          Untrue. The Hinkley CfD is for 35 years. It’s exceptional, because the capital cost the EPR design is exceptionally high. Too high. Other reactors, soon to be approved with a UK GDA, such as AP1000, and ABWR, will probably get CfDs closer to 25/30 years.

          Solar power is intermittent. It will need support from fossil plant. Yet who will build such fossil plant in today’s deformed pseudo-market where all the incentives go to non-carbon electricity? No one. UK government have been courting new natural gas for 18 months with no offers from the private sector. At least Hinkley is a full solution delivering firm power rather than a partial one like intermittents.

      • Peter Lang says:

        On the other hand solar is projected to last at least 20 years with minimal maintenance costs.


        The expected life time of solar PV in a fleet is the probable life of the installation. This is somewhere around 12 years for residential, roof top solar because people refurbish houses and other disruptions. If they put solar back it will often be a new system, not reinstall the old system

        Solar PV O&M is not low. It requires regular cleaning to maintain capacity factor, inverters fail and corrosion causes failures. Many owners do not even realise their solar panels are not working.

        Furthermore, fleet average capacity factor is below the optimal because shading increases over time (new buildings and trees cause shading), and many stop working and are not repaired.

    • willem post says:


      Exactly right.

      Wind and solar energy are short-lived cripples, cannot stand on their own.

      To equate nuclear, fossil, hydro and bio energy with wind and solar energy is blasphemy.

      Russia’s Rosatom has about $110 billion in order backlog for nuclear goods and services at end 2015, including at least 34 reactors (in construction, contracted, ordered).

      Russia’s turnkey, NPP average price to foreign customers is about $5000/kW.

      Russia usually provides attractive financing as well.

      China has many reactors planned and under construction, and will soon become a competitor of Russia.

      The nuclear goods and services orders of the US, EU and Japan are minor compared to Russia’s and soon China’s.

      • willem post says:

        Addition to comment:

        If nuclear is expensive, how about CSP in the US southwest?

        US Southwest: The Crescent Dunes CSP plant, tower-type, is located in the US southwest. Capacity: 110 MW, 10-h storage is required for continuous operation. Estimated production: 500,000 MWh/y of steady, dispatchable, 24/7/365 energy. CF = 500,000/(8,760 x 110) = 52%; a more likely CF would be 45 to 50 percent. Capital cost: $1.6 billion, or 14,545/kW, a very high cost. A quick way to calculate MINIMUM energy cost over 30 years = $1,600,000,000/(500,000 MWh x 30 years) = 10.7 c/kWh.

        If O&M, insurance, taxes, replacements, etc., and financing and paying interest on the investment over 30 years are included, the likely energy cost would be about 18 – 20 c/kWh, less with subsidies, cash grants, etc. Remember, all of this is STANDARD, WELL-DEVELOPED technology, i.e., no cost-reducing break-throughs can be expected.

        • Willem: The claim “10-h storage is required for continuous operation” might be better expressed as “10-h storage is required for continuous diurnal operation”. With this much storage the plant would indeed be capable of smoothing out day-night fluctuations, but it wouldn’t be remotely capable of smoothing out seasonal variations to the point where it delivers “steady, dispatchable, 24/7/365 energy” in the way a nuclear plant does. .

          • guber says:

            What use would this seasonal smoothing have in a desert climate, where demand is high when there is a lot of sun (air conditioning), and low when there is less sun?

          • singletonengineer says:

            @ willem post, March 9, 2016 at 2:07 pm plus Guber and others.
            1. Overnight smoothing is useless unless backed by conventional plant to cover events lasting more than 10 hours… such as winter time, when nights last 14 hours or more in the US mid-west.
            2. “10 hours smoothing”, but at what percentage of nameplate rated output? 10? 20? How is this adequate?
            2. Claims along the lines of “existing, proven technology, nothing more to worry about” ignore the simple fact that there is no existing 10-hour storage plant in service anywhere on the globe and certainly not at anything close to 100% nameplate output.
            3. I have asked for but never received an independently reviewed report of any solar project or any wind project, anywhere on the planet, with or without thermal or battery backup, that is cost-competitive without either preferential market access conditions or subsidies. Until that day arrives, no claim that unreliables are cost competetive under any scenario is just that – a claim.

          • willem post says:


            CSP with 10-h storage would provide steady (voltage, frequency, phase) energy, and it is dispatchable, a major improvement over PV solar and wind.

            During most of the daytime hours, energy would be stored in excess of what is needed to run the plant at a high percent of rated output.

            After the sun goes down, the plant would be run at 60% of rated output, or less, to ensure there is enough thermal energy left over for the next early morning.

            Hopefully, the sun will shine and the cycle is repeated. If not, nuclear has to take up the slack, assuming fossil is on the way out.

            The US southwest, with large, flat, uninhabited areas, could have thousands of such CSP plants.

            See Part II of this article.


          • Graeme No.3 says:

            willem post @ March 10, 2016 at 2:13 am

            CSP with 10 hours storage at 9 times the cost of coal fired is just the answer?
            10 hours storage means at least 14 hours sunlight, more because the energy from the sun low on the horizon isn’t enough, so 16/17 hours sunlight – at low latitudes?

            It seems many plants of this description use a gas turbine to generate early morning demand, with the waste heat warming the molten liquid. And have you costed the water needed for frequent washing of those mirrors?

    • robertok06 says:


      “A new nuclear station may last for 60 years while solar PV for 20 years. So we are now at 18 times more expensive for solar PV.”

      This point, I think, needs to be detailed a bit.
      The LCOE of PV versus nuclear is often touted has being close to one another (without, of course, considering the mandatory surplus production and cost of any storage if one wants to compare apples with apples, baseload generation (nuclear) with intermittent PV).

      Now, wit hout bothering to use the complete formula for the calculation of the LCOE, going on NREL’s LCOE calculator page and filling in the blank fields with proper values…

      … 90% CF for nuclear vs 15% for PV (10% in Germany or UK)

      …. 4000 Euro/kW for nuclear and 1500 Euro/kWp for PV… putting the same cost of financing (6%/year?)… zero cost for fuel for PV and 0.5 for nuclear, 10800 forhear rate of nuclear (PV zero, obviously), and IMPORTANT!… 20 years for the period of calculation, one gets:

      (using the median values of this for other costs)

      Fixed O&M: 91; Variable O&M: 0.006; Fuel cost: 0.5;

      LCOE kWh nuclear: 6.7 c$

      Same for PV: Fixed O&M: 30; Variable O&M: 0;

      LCOE kWh PV: 12.2 c$

      So…roughly speaking, without taking into account storage requirements, PV is 80% more expensive than nuclear.

      But, what is the LCOE? By definition, it is the cost that can be associated at each kWh in order for the capital cost and the interests to be re-paid in the time period chosen… with no actual money made by the owner of the system.

      So… the kWh of nuclear costs this amount provided that the capital cost be paid in full in 20 years…. but we all know that reactors last much longer, 40 minimum, 60 easy, new designs could go to 80 and 100 (tested new alloy material for RPV in Russia recently, withstands the high rate of displacements per atom generated by the high neutron fluence).
      PV,on the other hand, is generally “incentivized” for 15 or 20 years (at much higher costs), and guaranteed by manufacturers for 25-30… nobody knows what modern modules/panels will actually do in 30 years!… there are several studies/papers showing that they can die prematurely with some frequency.

      The only real advantage of PV over nuclear is that one can actually install and start production of several GWp per year in a typical EU country, while nuclear needs 5-10 years for construction, so the LCOE as calculated above does not capture the cost of financing during the long construction… but that’s mainly a political problem, dictated only in small part (IMO) by technology.


      • Euan Mearns says:

        Thanks Roberto for fleshing this out and for the great link. It would be interesting to take this to the next stage and look at cost of converting each technology to dispatchable supply. Cost of pumped hydro in the case of nuclear. Cost a gazillion power walls in the case of solar.

      • gweberbv says:


        your costs for PV are slightly too high (probably a factor of 2 – with a high probability for an additional drop by maybe 30% within the next 5 years). A tiny installation with less than 10 kWp capacity installed on a rooftop in Germany comes in now for roughly 1300 to 1500 Euros/kWp! For utility scale non-rooftop installations you are now well below 1000 Euros/kWp (even with German labour costs).
        However, a 1-to-1 comparison is not the right approach anyway. Solar PV obviously can act only as a part of an electricity system, while nuclear can do it alone (if necessary).

      • Willem post says:


        If a business owns a NPP, it would issue long term bonds at about 5% for 40 years for 70 percent and put up own money of 30 percent. The business typically likes to earn about 10% on its investment. Both sources of funds need to be amortized over 40 years.

        The NREL allows only one percentage.

      • willem post says:

        “But, what is the LCOE? By definition, it is the cost that can be associated at each kWh in order for the capital cost and the interests to be re-paid in the time period chosen… with no actual money made by the owner of the system.”

        LCOE can be anything one defines it to be.

        The simplest LCOE = Turnkey plant overnight capital cost/lifetime energy production.

        And it get more and more complicated from there on.


        US Southwest: The Crescent Dunes CSP plant, tower-type, is located in the US southwest. Capacity: 110 MW, 10-h storage is required for continuous operation. Estimated production: 500,000 MWh/y of steady (voltage, frequency, phase-angle), dispatchable energy. CF = 500,000/(8,760 x 110) = 52%; a more likely CF would be 45 to 50 percent. Capital cost: $1.6 billion, or 14,545/kW, a very high cost. A quick way to calculate MINIMUM energy cost over 30 years = $1,600,000,000/(500,000 MWh x 30 years) = 10.7 c/kWh.

        If O&M, insurance, taxes, replacements, etc., and financing and paying interest on borrowed money, and owner’s return on investment, over 30 years are included, the likely energy cost would be about 16 – 18 c/kWh, less with subsidies, cash grants, depreciation tax benefits, etc. Remember, all of this is STANDARD, WELL-DEVELOPED technology, i.e., no cost-reducing break-throughs can be expected.

      • Leo Smith says:

        Unfortunately with intermittent renewables you must add in either the cost of some alternative generation or some storage, to get the same equivalent performance.

        I guesstimated that at around 2p/kWh for UK gas..

        • willem post says:


          Plus cost of grid upgrades.

          In Texas about $7 billion of transmission was required to get Panhandle wind energy from west Texas to population centers in east Texas; it shows up as a surcharge on customer electric bills.

      • Token says:

        For Roberto: “We believe these risks,
        combined with the higher capital costs and longer construction schedules of nuclear plants as
        compared to other generation facilities, will make lenders unwilling at present to extend longterm
        credit.” (Citigroup, Credit Suisse, Goldman Sachs, Lehman Brothers, Merrill Lynch, and Morgan Stanley, Comments in response to Notice of Proposed Rulemaking on Loan Guarantees for Projects that Employ Innovative Technologies (RIN 1901-AB21), 72 Federal Register 27471 (May 16, 2007), July 2, 2007)
        This was before Fukushima, and already then they did not want to give credits to nuclear projects.

        • robertok06 says:

          Of course banks are not willing to finance nuclear… they are interested in making money, not producing large amounts of electricity 24h/24, 365 dd/year… they prefer the happy free meal of “incentives” for intermittent renewables, how could one blame them?… if I were an investor instead of a physicist I’d do the same choice.

          Fact is, all this crazy rush to install intermittent renewables has started with the pretext of “saving the planet” from the poisonous killer gas CO2 (I adhere, here, to the green dogma as per greenpiss leaflets)… and we all know that nuclear is the only scalable at will technology option which is 100% CO2-free at production… hydro cannot be scaled much more than it already is… and the silly panels and turbines will always need support from baseload power stations… if the latter are not nuclear they forcefully will be gas/coal/lignite/oil… there’s no other choice possible!

          So, what’s your position in this, token? Are you for saving the planet?… making money?… something else?

          Let’s hear.

  4. Nicola Terry says:

    I totally agree that nuclear power is often made unnecessarily expensive by regulation. However I think your comparison with solar power omits important differences – like that solar is low risk and can be installed in weeks, sometimes even without planning permission while nuclear power takes 5 years and umpteen controversial consultations. On purely cost grounds, solar is rarely the best choice but that is not the only factor.

    • Euan Mearns says:

      Nicola, I agree that the large upfront capital cost of nuclear is a problem. It is a problem that government should recognise and put in place procedures to overcome this. For example Bonds that are underwritten by the Government. But at the same time they need to get the cost way down.

      Hinkley C is like a 20 bedroom palace built out of granite. You need a big mortgage to pay for it, but it will last your whole life and even the life of your children. Solar PV on the other hand is more like a tent.

      • Rainer says:

        I live in a tent like this since more then 25 years.
        Not one cent cost to hold it in shape. No one hour time to hold it in shape.
        Producing between 60% and 90% of the yearly demand.
        Even did not clean it at all. And the trees are growing an bringing shadow!
        I love my house you call tent Roger.

        • robertok06 says:


          Reading what you’ve written at first sight reminds me the usual green propaganda… so please provide some details of your wonderful tent’s cover… the PV system you are talking about.
          I am highly skeptical of your data.. in particular

          “producing between 60% and 90% of the yearly demand”

          …since data show that for a sunny country like Italy the average domestic PV system (3-5 kWp) allows on average a self-consumption of 25-30%… so either millions of italians are dumb and you are a lot smarter than them or you have been carried a bit in your statement.

          Alternatively your mean “60% to 90% of the yearly demand”… your demand, but not necessarily consumed by you…so in that case you should justify the exorbitant costs of the kWh that the company/utility is obliged to buy from you.

          Please, provide data: country of installation, size of the system, type of the modules, any storage device… failing to do that I’ll put your messate in the “wishful thinking” folder.


        • Euan Mearns says:

          You are welcome to try this in Scotland.

        • Alex says:

          Your tent may be lovely, but it no use on a cold winter evening.

  5. Gaznotprom says:

    It’s a crazy crazy world.
    Oh Nuclear works – let’s make it so expensive so it doesn’t get built!
    Oh Solar is so expensive and doesn’t work, lets subsidise that!

  6. Gaznotprom says:

    Thanks for that Roger, great insight!

  7. Greg Kaan says:

    I had read the paper but did not think to factor in those incidents. Great analysis, Roger.

    Nicola, to deploy solar PV on the scale of a 2GW reactor, you would encounter planning permission and consultation road blocks as well. So far, utility PV deployment has been approved almost by stealth but the environmental implications are becoming increasingly clear and with that has come increased opposition to newer proposed installations.

  8. guber says:

    Nobody in germany pays 2000$/kWp for solar in residetal projects. See e.g.,
    utility scale Solar in germany is at 1000$/kWp or below, otherwise you don’t survive in the market.
    It is like with nuclear – it depends how many regulation costs you put on the systems.

    • Euan Mearns says:

      Yeh, but your load factors are also below 10%

      • guber says:

        Well in south germany load factor is about 0,12-0,15 , which is not good, but enough. But the important point are the costs. Germany is not a especially low cost country with cheap labour costs.
        Other countries – nearly all – have higher capacity factors, south of spain around 0,25, an chile up to 0,35.
        Also Solar needs no water. In many coutries supplying cooling water is a severe problem. That’s why power stations like monrupole or medupi have dry cooling – very expensice, and eating up efficiency.
        A capacity factor of 0,9 for nuclear is at the optimistic end – only whith a very good station, deep within the baseload region, no competing renewables around, this can be reached. It can be much lower:, also Thiange and Doel would love to reach a capacitiy factor of 0,5 again.

        Combined with fuel costs solar does not have, and higher project cost risks resulting in higher interest rates for nuclear, the ceiling for nuclear on the economic side and without costs for spent fuel and decomissioning is somwhere at 3* solar costs, (If solar remains somewhere in the region of maximum emand incl. load shiifting, and nuclear deep within baseload region). With the german utility scale prices this would be somewhere at 3000$/kWp for nuclear
        By the way, with a little maintenance solar lasts much longer than 20 years. Solar power installations since the 1970’s very rarely had to be closed down due to wearing out. And my personal experiece – the series of solar power supplies for traffic surveilance systems I once tendered in the early 1990’s are all still in operation, just needed a replace of battery+inverter once. So all generators run now for 25 years mantenance free, failure free.
        If these systems are typical, statistic says that the solar modules will outlast nuclear power stations.

        • willem post says:


          “Well in south germany load factor is about 0,12-0,15”

          What happens in Germany is what matters.

          On THIS site one must quote proper facts.


          • guber says:

            Yes, but numbers are without the rising share of in-house consumption – usually around 30% of produced power in residental installations. ( And In the earlier years the newly installed systems made up a large share of the installed capacity, but produced only a part of the year, thus havin load factors of 0,05 and below in the first calendar year.) So below 0,1 is incorrect, and the mentioned high capacity factors are for _south_ germany, not whole germany. In north germany capacity factors are as low as in UK.
            Which is why residential solar is very high in the montanious regions in the south, where capacity factor also due to the thinner athmosphere is reaching the 0,15, and significant lower in the north with the lower solar production.
            But for a worldwide view on things this is irrelevant – most people in the world live in the “solar belt” of the world, starting somewhere in southern france down to south africa, where production does not vary so much beween seasons. And where demand is high when production is high in the average of days (air conditioning).
            UK is about (roughly) the only industrialised country which is so far north and can not compensate wind power with Hydro. So it is not a “typical” country in the world.

          • Token says:

            Willem 34,930,000MWh /38,233MW*8,760h=0,104. Without in-house consumption, and with assumption that all new capacity installed in 2014 was installed at 1.1.2014 0:00 o’clock.

          • robertok06 says:

            “Yes, but numbers are without the rising share of in-house consumption – usually around 30% of produced power in residental installations.”

            No. The production figures include the self-consumption, it is the production metered.
            See data on Fraunhofer’s web site (for previous years, now they simply have a short document made)… average capacity factor of PV in Germany is slightly above 10%, and insolation in the sunniest part of Germany (if Bavaria can be called Germany… 🙂 ) is a bit lower than northern Italy, for obvious reasons…. maybe 12% in a good year.

            Also, 30% of self consumption is, I think, too high… note that self consumption during winter is virtually zero (as for 4 full months the average german PV produces at 5-7% CF)… and in summer there’s a limit to the amount one can consume… lights in the house are less used, daytime is longer, etc…

        • robertok06 says:

          “Well in south germany load factor is about 0,12-0,15 , ”

          15% no way. Italy’s average is 14.5% (data for 2014,entire year), and the biggest installations are in the southern part of the country. I give you a 12% maximum.

          Using the NREL LCOE calculator, even with 1200 $/kWp (approx 1000 Euro/kWp) the reduced (12%) capacity factor with respect to the 15% value I’ve considered in my other message makes the LCOE cost equal to 12.8 c$… practically no change.

          The capacity factor is what dictates, first and foremost, the LCOE cost… that’s why nuclear is ALWAYS among the cheapest forms of electricity production.

        • robertok06 says:

          “A capacity factor of 0,9 for nuclear is at the optimistic end”

          Strange… ’cause all of the 99 reactors in operation in the USA, and they are all seasoned units, have generated power in 2015 at close to 90%.
          Sweden’s reactors and Finland’s as well are at 92%… all pieces of junk with 30 years of operation, a couple of them are replicas of Fukushima’s BWRs.

          Most of the 8% time lost not operating is time lost because of bureucracy, paperwork and the like.
          During the long and extremely cold spell of 2 winters ago in the USA, the reactors on the middle/eastern part of the US have worked at an average capacity factor ABOVE 100%, because the unseasonably low temperatures have helped increase the Carnot efficiency of the cycle.

          These are all known facts, not my opinions, let me be clear.

          • guber says:

            Yes and worl average is 0,71. to tell the real numbers not pick the raisins.

          • robertok06 says:


            “Yes and worl average is 0,71. to tell the real numbers not pick the raisins.”

            Source, please? Let’s see who/what you quote…

            I am asking this because this one KNOWLEDGEABLE source says a much different thing:

            “Considering 400 power reactors over 150 MWe for which data are available: over 1980 to 2000 world median capacity factor increased from 68% to 86%, and since then it has maintained around 85%. ”


            Nice try, though, guber!… but you should know by now that pro-nuclear people around here are not gullible monkeys living on trees, believing anything they read.


            P.S.: note that I know exactly why you cited 71%… but that’s cheating!…

          • Peter Lang says:


            Yes and world average is 0,71. to tell the real numbers not pick the raisins.

            Your number is meaningless if your average includes load following nuclear or plants that are not fully dispatched, – e.g. because wind and solar are “must take” so nuclear is not dispatched.

            OCGT may have a capacity factor of 1% to 10% and is built because it serves an important role.

            I suggest you should follow your own advice and “not pick raisins”.

        • Alex says:

          10% is the assumed capacity factor for southern Germany. We get 11% with panels due south. You might get 12% up in the Schwarzwald or Algau, above the fog. But none of that is when the electricity is needed most.

          A modern nuclear power plant can achieve over 90%.

          • Leo Smith says:

            CANDU achieved 95%

            Uk old reactors before EDF took em over were poor at less than 60% IIRC, but a program of proper maintenance pushed them up into the 80%’s.

            The data should be there on Gridwatch

          • Peter Lang says:

            Lat time I looked, Wolsung 1 (a CANDU) had the world’s best life time capacity factor.

        • GeoffM says:

          Guber, you say solar needs no water just after mentioning how sunny Spain and Chile are, but you’re wrong. I read about a study in a desert environment which found that if solar panels aren’t cleaned, they lose 50% of their rating each 18 months. I have spent over a year of my life in deserts and believe me, dust is a huge problem, especially after a dust storm. Where is the water going to come from in the context that many desert regions have water shortages?

    • @ guber

      Agee stat put last years installed costs circa 1200 €/MW

  9. Rainer says:

    In nuclesr technology the building costs are not the problem.
    Why you allways forget the cost of thousands of years garding the dunping places?
    Not name all costs is an really old tricky thing to manupulate public.
    I am disapointed.

    • Greg Kaan says:

      Hello Ranier

      I was wondering what studies you have read on the issue of nuclear waste disposal/storage?

      Here is a recent article that was written by a fervent environmentalist in my country on this subject. It was written in the context of a proposal for a nuclear waste storage facility but is relevant in general on the hazards of nuclear waste


      • oldfossil says:


        The physical maintenance of the site won’t cost much. Mountains are low maintenance. The cost of staffing and maintaining a guarded perimeter won’t be very expensive either. This is not Fort Knox we’re talking about.

        • Rainer says:

          Exact that is the problem:
          I was travelling australia and really was shocked about the not existing filters in carbon plants!!! It is not the 1950ith. Just ripping of the complete continent to make a good just in time live. It is a shame.
          All nuclear fans do not calculate really costs. Same with emissions of other form of power plants. This part is a calculation of cost per MW. I still insist:
          If it is not calculated all the cost it is not the work to keep away the children and grandchildren save for blackout. It is just putting the cost and risk to the children and grandchildren. It is just a form to lie to yourself how good you are. You are not good, you are irresponsible.
          And the worst: you know better and not act in the way of your knowledge. Just propaganda not to say the texas word bu*****t.

      • Graeme No.3 says:

        Greg Kaan:
        I followed your link..very thought provoking except among some of the commentators who emulate old time Vikings and go berserk.

    • Wm Watt says:

      I wonder if spent fuel rods could be dumped down abandoned oil wells. I understand they need to be undisturbed for 30,000 years, not leak into underground water flows, and be out of the reach of bad people like terrorists. There might be old oil wells which don’t have underground water flows, and I can’t imagine bad persons fishing them up drill holes. Looks like a cheap solution to me.

      • Euan Mearns says:

        Well fluids normally flow through the reservoir part of an oil well. But above the reservoir is normally thick impermeable beds like mud rock, anhydrite or salt. So you simply fill in the bottom of your well with cement. Drop in your containers of waste, fill the well with cement and go home. In the North Sea, this could be over 10,000 ft below the sea bed.

        Its simple, and probabaly completely safe for ever. But the greens don’t want this problem to be solved.

      • Alex says:

        Yes and there’s been quite a bit of work recently on borehole disposal. However putting fuel rods into boreholes would be a complete waste of energy. They need to be recylced and the fuel reused. The remaining fission products can be put into any old borehole. No guards required.

      • Leo Smith says:

        You can dump treated waste almost anywhere. The real issue is that apart from some pretty well known stuff that is either reprocessed or is small in quantity, and can be vitrified, most nuclear waste is less dangerous than a council land fill.

        The problem of nuclear power is not waste disposal, it is public perception.

        As certain posts above amply display.

        The whole world is made of nuclear waste, and its powered by a giant highly dangerous nuclear reactor that kills 3000 people a year in the UK alone, from radiation exposure…

        The issue people have with man made nuclear waste, isn’t that its radioactive, its that its man made.

        Otherwise there would be calls to cover Dartmoor and Aberdeen in 20 ft of concrete and put an armed guard round them (might work for Aberdeen, at that ;-))

    • Peter Lang says:

      The back end of the nuclear fuel cycle coast about 1% of the cost of electricity.

  10. david says:

    funny article to read as the CFO of EDF Thomas Piquemnal just resigned!

  11. paolo pulicani says:

    Put it as you like but the reality is this:

    EDF forced by the French State to take a majority share in AREVA. The capital increase (5 billion €) will be backed by French State “in compliance with European regulations.” (I rather say: “in spite of European regulation”).

    Sources said chief financial officer Thomas Piquemal had resigned over concerns the project would put too much stress on EDF’s balance sheet

    • robertok06 says:

      Hello Paolo Pulicani, notorius anti-nuclear propogandist…

      “Put it as you like but the reality is this:

      EDF forced by the French State to take a majority share in AREVA. The capital increase (5 billion €) will be backed by French State “in compliance with European regulations.” (I rather say: “in spite of European regulation”).”

      Reality is that AREVA’s losses are worth a couple of billion Euros, while the technology that you like so much, photovoltaics, pulls out of the pockets of italian electricity users 6,7 billion Euro/year for TWENTY years, and all this to produce an insignificant fraction of the electricity that AREVA’s products generate.

      AREVA enriches uranium, among other things, for all of France’s 58 reactors, to the tune of 417 TWh in 2015… even if the 2 extra billion Euros were added to the cost of such a humongous amount of electricity that would be close to a detail…. 2/417… I let you do the math… while Italy’s PV generates only 24 billion kWh… and if you divide 6.7 by 24 you get close to THIRTY Eurocents per kWh… and this only during the day, never at night, 1/4 in winter of what it generates during summer.

      You can put it as you like it, man, but THE reality is this.


      • paolo pulicani says:

        Oh no dear Roberto,
        I’m not a “propagandist” of anything. And I don’t prefer PV or nuclear or hydro or wathever.
        I tend to look at this matter in terms of economic viability (both short and long term) and not as a “battle for civilization”.
        What I’m pointing out is that, in a epoch of high “turbulence” – and we are one of such epochs – long term planning (five, ten years) simply doesn’t work.
        If you wish, I favor small and quick solution (be they nuclear, PV, hydro or wathever) , which I consider better suited to the current time.

        • robertok06 says:

          “If you wish, I favor small and quick solution (be they nuclear, PV, hydro or wathever) , which I consider better suited to the current time.”

          … and which solution would that be for France, for instance (since you have mentioned EdF)?

          Let’s see.

          • robertok06 says:

            See?… the anti-nuclear troll has disappeared… cannot answer a simple question.

          • paolo pulicani says:

            Oh, I don’t have a solution (otherwise I’ll be hired with a BIG salary by French Government, or Chinese Government for that matter).
            What I’m pointing out is that the proposed “big” solutions (big plants) aren’t solutions, because it takes time and big money to complete these kind of plants and, in the meantime, the “boundary conditions” are likely to be completely different and not predictable. I think that Messier Thomas Piquemal had the same kind of risk aversion (not to bet the house in a single big project). The French Government thinks otherwise and I suppose they’ll decide to go ahead: good luck with this.

          • paolo pulicani says:

            ..and French Cour des Comptes agrees with Thomas Piquemal:

            “Even though the [Hinkley Point] deal has not been finalised, the complexity of the deal and especially the way it could impact the responsibility of EDF suffice to raise serious questions”.

            The auditor said the focus on nuclear in EDF’s foreign investments was a way to preserve its skill base at home, which it said was acceptable, as long as this was not more a defensive than an offensive stance.


  12. Syndroma says:

    BN-800 fast breeder reactor cost was 146 billion roubles. $2-4 billion depending on exchange rate.
    Another 9 billion roubles for MOX fuel assembly line.

    • Euan Mearns says:

      And the power rating?

      • Mark Pawelek says:

        The BN-800 has a design rating of 880 MW, yet the wikipedia entry says “The plant started producing electricity December 10, 2015, with a reduced power of 235 MW”. It uses ceramic fuel, so does not take advantage of the Integral Fast Reactor, IFR, pyroprocessing and cast alloy fuel cycle innovations. Wow. Pyroprocessing is over 20 year old technology, which to the best of my knowledge, no one has tried to use. That despite Argonne National Laboratory IFR researcher Roger Blomquist claiming the cost of pyroprocessing will be about one seventh the cost of PUREX (reprocessing). Regulators really do have the nuclear industry pussy-whipped eh?

        • Syndroma says:

          The plant obviously was not started at full power from day one. As of Feb 25, the output was 555 MW or 67% of design power. The next step is 85%, then a planned maintenance in the summer and 100% by the autumn.

          Right now BN-800 is loaded with a so-called “hybrid” core: both plutonium and highly enriched uranium oxide fuel assemblies. It’s a stopgap solution because commercial MOX facility was not ready in time. Actually, there was a nightmare hydrodynamic problem due to the different sodium flows through the different assemblies. More than a hundred assemblies had to be removed from the reactor and physically modified in an inert atmosphere.

          But the MOX facility is operational now and starting from the next reloading BN-800 will run with a MOX core. Plutonium for the fuel is taken from the stockpiles built with PUREX reprocessing. It’s a stopgap solution too, because pyroprocessing facility which is being built adjacent to the MOX facility, will become operational in 2018. It’s a small proof-of-concept facility, designed to demonstrate that the fuel reprocessing can be made without large volumes of low-level waste.

          Since BN-800 is a commercial plant, it requires an uninterrupted supply of fuel. That’s why fuel technologies with low technical risks were chosen. As for the innovative fuel, nitride fuel assemblies are being tested in BN-600 right now. Nitride fuel is interesting due to its high density, which is good for the breeding ratio. A considerable effort is being made to prove the commercial viability of the nitride fuel.

          And of course the minor actinides from the reprocessed fuel are expected to be burned in the BNs. It will solve the long-term problem of waste storage. Only decay products will have to be buried. And that means in 300 years their radioactivity will drop below the activity of natural uranium ore.

  13. I am confused. How can there be such a difference between utility in the US and residential solar in Germany with utility scale on the wrong, unexpected side?

    Also you can get some costs for solar (total) from the recent AGEE stat renewable electricity report. Costs in 2014 were 1228 €/MW but in 2008 it was up around 4000.

  14. Wm Watt says:

    Looking at the cost increase of longer term construction sites in the scatterplots reminds one of the changes imposed during construction on Canadian projects by vacillating politicians spooked by anti-nuclear groups. They seem to be no more disciplined than homeowners going during renovation projects with the same economic results. Increasing costs during past construction has been a big deterrent to new projects here.

    BTW I like the graphical display of data on this site. Numbers are important for keeping human speculation in perspective and graphs are an excellent way of presenting large amounts of data. Keep up the good work.

    • sod says:

      “BTW I like the graphical display of data on this site. Numbers are important for keeping human speculation in perspective and graphs are an excellent way of presenting large amounts of data. Keep up the good work.”

      I totally agree with this. The data presentation and also the hypothesis testing is excellent.
      It is also done to a level, that makes it hard to argue against the main points of the articles (The accidents most likely had an effect on costs of later projects)

      I do of course disagree with the majority of the interpretation of these results.

      For example i think Korea is a bad example, as it had a huge scandal about the costs of the reactors.

      I would also argue, that it is obviously democracies, that have a problem with nuclear after accident, and that is not by pure choice, when prices go up.

      So i would expect a similar development in other countries soon, it has already started in India:

      And Fukushima also had an effect on China:

      I was actually pretty shocked, when i read this Financial times article about EDF and Hinkley today:

      And the costs do not include abandoned projects, current extreme price developments (EPR fiascos) and future risks (an accident in China and India should be the next turning point).

      • robertok06 says:


        your message is full of untrue statements, or superseeded decisions by govenments.

        In particular you say that…

        “And Fukushima also had an effect on China:

        … which is old stuff, China simply blocked the authorization of new projects, especially those inland which would use river water for cooling, but has recently increased its goal for new installations… they will start tens of new reactors in a matter of few years, and have just re-authorized the construction of some reactors inland.

        Talking about protests in India… sure, there have been protests by local fishermen… few hundred people, often backed by “imported” protesters organized by GreenPiss.
        Some time ago I read an article stating that “thousands of protesters in Tokyo vent their opposition to nuclear power”… thousands (3000 at most) in a city of 20 million???? Excuse me???

        Try again.

        • ribinka says:

          Roberto, I hope somebody starts moderating your contributions as promised in the rules for the blog.

          • Euan Mearns says:

            Ribinka, Roberto is a physicist at the premier physics research institute in Europe. I tend to find his comments rooted in fact and science. Why would I moderate this?

            You have gone straight to comment moderation. Your response to this comment should begin with an account of your qualifications, who you are, what is your CV? Followed by a detailed account of why Roberto’s submissions need to be moderated.

          • robertok06 says:

            I know that I often loose my temper when these subjects come up… but as a physicist I can’t stand logic and data to be bent to ideology, misinformation and/or plain stupidity.
            Cheers, and please say something on the subject, make a contribution, rather than objecting to my commenting style.

          • Peter Lang says:

            Robertok06’s comments are some of the most informative on this web site. I hope he will not be discouraged by trolls.

      • robertok06 says:


        ” (an accident in China and India should be the next turning point)”

        I know that you anti-nuclear guys really hope for new accidents, but even if it happened I must tell you that tomorrow it will be the 5th anniversary of Fukushima’s accident, and the tally is still at ZERO deaths…while the tally for dead people in the land of Energiewende goes up by abot 10-15 every single DAY… courtesy of about 0.6-0.8 TWh/day of electricity from coal/lignite.

        In case you don’t believe my pro-nuclear words… read this, written by an acclaimed environmentalist:

        “Using historical production data, we calculate that global nuclear power has prevented an average of 1.84 million air pollution-related deaths and 64 gigatonnes of CO2-equivalent (GtCO2-eq) greenhouse gas (GHG) emissions that would have resulted from fossil fuel burning.

        On the basis of global projection data that take into account the effects of the Fukushima accident, we find that nuclear power could additionally prevent an average of 420,000-7.04 million deaths ”

        I don’t know what you think, but for me saving millions of lives of people around the world is a sufficient reason to opt for nuclear.

        Cheers,and have a nice reading.

  15. Mark Pawelek says:

    “$3860 for EPR at Flamanville in France”
    — Ha ha. At £18bn for Hinkley C, the rate is US $ 7770/kW. In comparison, the Moorside AP1000 reactors (£10bn for 3) will be US $4312/kW. Interested readers may also want to read what Bernard Cohen wrote about the effect of the NRC on US nuclear power in his 1990 book: “The nuclear energy option”. Chapter 9 – Costs – what went wrong?, it’s here: The other point about the NRC – I’m told they’ve licensed only 4 new US reactors since they began – about one per decade.

  16. guber says:

    @Singletonengineer – if there is a storage for 10 hours (maimum power) it is likely to last 14 hours with reduced power, too. usually damnd at night is much lower than during the ay. And the grin epands daylight for power production with every 1000-1500km east-west extension by one hour.

  17. Mark Pawelek says:

    Why is nuclear power so expensive? One reason : it’s stuck in the past by rules and red tape. I don’t just mean the nuclear regulators. Technologies can’t advance because anti-proliferators all but outlaw them.

    1) Lithium-7 is a good example. Li-7 is ideal for making molten salt reactors, which as a prerequisite for thorium reactors. Yet lithium is naturally a mixture of two isotopes: Li-6 and Li-7, in a ratio 1:3. Putting natural lithium in a reactor would result in the Li-6 making masses of tritium. It would waste a lot of neutrons too. The isotopes must be separated. US dept of defence are not kean on you doing that. Lithium-6 is considered an A-bomb proliferation risk because it can be used to make tritium (which is a H-bomb proliferation risk).

    2) Likewise the Integral Fast Reactor was killed on very dubious anti-proliferation grounds. We’ve seen no commercial development of pyroprocessing (although it 20+ years old), and PhD nuclear power researchers reckon it should cost one seventh PUREX.

    3) A third example is regulators slowing advanced nuclear power down to a snail pace with their rules and regulations over what materials can be used to make a reactor. The best materials can’t be used in molten salt reactors because they would take 15 or more years to test and certify with the US regulatory agency. Nuke startups like Terrestrial Energy and ThorCon must use nuke-certified stainless steel instead, if they want to see their designs in action before they die.

    • Mark Pawelek says:

      Let me add:
      4) The “cradle to grave” fuel cycle imposed on all Western countries by anti-proliferator rules. Companies like AREVA and Westinghouse make nuclear fuel and help ensure it can’t be diverted for A- or dirty bomb use. It’s a nice earner for these companies. “Cradle to grave” means the responsible people are always in control of the fuel material, even after use. Another cosy agreement the mainstream nuclear industry made with government to cripple nuclear power.

    • Mark Pawelek says:

      Apologies. Lithium is naturally a mixture of two isotopes: Li-7 and Li-6, in a ratio 12 : 1. I got it confused with another element I wrote about recently.

      I blogged this:

    • RDG says:

      Mark, why don’t you just summon the courage to admit that those designs (MSRs, IFR, EPR, etc) suck instead of blaming it all on the regulators. The days of govt providing funds to stupid ideas are over.

  18. I come a bit late to this discussion and see there is some bickering about wind and solar capacity factors, with numbers being quoted from 10-30% or so, amid much guesswork, and especially German numbers being thrown around that make no sense to me. For a few years up till early 2013 I collected the data from the German BundesMinisterium fuer Wirtschaft und Energie, and loaded them in a spreadsheet. The data covers hydro, wind, solar, six types of bio, and geothermal, for the three uses heating, electricity generation and transport. I wanted to paste here the wind and solar energy vs installed capacity and capacity factors for 1990-2012 for electricity production only, but this txt-based comment space does not allow a spreadsheet block (T36:AC67) to be pasted without making a glorious mess. So I quote the result:
    – CF wind for 1990-2012 is 16.4% avg, st dev = 2.4% i.o.w. 1 sd is from 14.0 – 18.8%
    – CF solar for 2000-2012 is 7.7% avg, st dev = 1.5% i.o.w. 1 sd is from 6.2 – 9.2%.
    These are based on the government-provided numbers here: /

    This should stop the guesswork. (Not stated is whether the BM also captures personal use ).
    There is lots more to learn there, such as the miserable contribution to the country’s electricity production of all the renewables except hydro. The numbers I see quoted in the press and journals for that contribution always exaggerate the actual numbers calculated by the spreadsheet (up to 4x higher) . Germans not only lie to others (which is ok) but to themselves (which is stupid). The overall picture is that Germany is on a hiding to nowhere.

    • Euan Mearns says:

      BM reports in the UK does not capture solar at all. Wind is split into two categories. Large wind farms are connected to the HV grid and are reported by BM / Gridwatch. Small windfarms are connected to the LV grid and are seen by BM as negative demand. But National grid report unmetered wind and solar separately. And the Renewable Energy Foundation also has a good data base.

      I had a look at UK load factors here:

      • gweberbv says:


        please excuse my naive question, but how you get paid for production when it is not metered? Are the unmetered renewables installations in UK paid for the sheer fact of their existence?

    • Roger Andrews says:

      This should stop the guesswork.

      Not exactly, because government statistics are themselves based on guesswork as to how much power was generated by unmetered solar installations and what the capacity of these systems might have been. I wrote a post on this a couple of years ago

      which concluded that government stats were not to be trusted and the only way of obtaining realistic estimates of solar capacity factors was by analyzing output data from individual operating solar systems. So I downloaded operating data for over 400 systems worldwide from the sunny portal site and analyzed them. Results by country are summarized in this plot:

      And in this table.

      Germany with a capacity factor of 12.2 +/- 1.2%, fits in well with the other countries on the plot. Germany with a 7.7% +/-1.5% capacity factor does not.

      • Willem Post says:


        Those are larger installations, likely in good locations, facing solar south.

        See my above CF calcs. and URL, which covers ALL installations.

        • The installations include rooftop panels, ground-mounted panels and a few ground-mounted single axis and dual axis tracking arrays, but they’re dominantly small to medium-sized rooftop installations (down to 1kW) that often don’t point in the optimum direction. A representative sample, I think.

          Your above calcs based on “official” German stats:


          Are too low for the reasons I discussed in my post.

    • gweberbv says:


      the numbers for photovoltaic installations (and also the other energy sources) are given for the end of each year. If you simply divide by the production by this number, you implicitly assume that the change in installed capacity during the year is negligible. But of course, it was not. I now took the production data from the years since 2010 and normalized them to the mean value of the installed capacitiy for this year and the last year. The results are as follows: 2010 -> 9.5%, 2011 -> 10.3%, 2012 -> 10.3%, 2013 -> 10.2%, 2014 -> 11.0%.

      • @ gweberbv

        yes, I am aware of the difference between End of Year capacity and avg of previous and current capacities or Middle of Year case. My spreadsheet has a switch to either use EOY or MOY . The numbers I quoted (20160310 12:41) were the EOY case. For MOY the solar CFs go up by approx 1%, the wind CFs go up approx 2.5%.
        The averages for MOY become:
        wind 1990-2012 CF avg = 18.6, st dev 2.9
        solar 2000-2012 CF avg = 9.6, st dev 1.7

        When I quoted numbers I forgot the switch and gave the numbers showing on opening the spreadsheet. Anyone interested in receiving the spreadsheet may ask Euan for my e-address .

        What makes a bigger difference is the comment by Roger Andrews (20160310 16:45) who points to the difference between (German) government stats I used and the actual situation that includes unmetered solar installations, which according to his CF vs latitude analysis brings German solar up to 12.2 ± 1.2 , which is 2.4% more than my MOY value 9.6 ± 1.7. It is still apples and oranges because my numbers are 13-year time averages of one country’s solar park, while Roger averages over a selection of systems. Be that as it may.
        There are no unmetered wind installations (except for tiny units like 3kW on a farm roof) so my wind numbers stand. Except, once again, that wind installations quote their output to the grid, never their house load which they take from the grid to operate a dozen supporting systems on the wind tower, and which is considerable. I saw US figures quoted as 10-12% of output, but lost the reference. Wind operators presumably hide the house load numbers because they are dreadful.

        I want to remove a wide-spread confusion about the use of capacity factor and load factor. The utility industry definitions are:

        CF is MWh over name-plate MW x 8766h (or whatever period ).
        LF is MWh over max power delivered that year in MW x 8766h.

        The difference can be important. The UK wind park of 7232 MW (MOY value) in 2012 never produced 100%, the best was 4105 MW or 56.7% of nameplate power, for one half hour on 25 Sep 2012 at 20:00. The CF for the 8-month period Apr-Oct was 15.8%. This is likely to be quoted as LF = 28.2%, which sounds so much better. The data came from NETA, which after some analysis showed the UK wind park delivered below 10% of nameplate for 43% of the time, less than 20% for 68% of time, less than 40% for 96% of time, and less than 56.7% for 100% of time.

        I suspect the wind industry quotes LFs only, not CFs. Can someone confirm ?

        And are the wind and solar CF percentages in the EIA page presented by T2M really CFs or the nicer LFs ?

        • Peter Lang says:


          Thank you for your comment. Regarding this bit:

          What makes a bigger difference is the comment by Roger Andrews (20160310 16:45) who points to the difference between (German) government stats I used and the actual situation that includes unmetered solar installations, which according to his CF vs latitude analysis brings German solar up to 12.2 ± 1.2 , which is 2.4% more than my MOY value 9.6 ± 1.7.

          Do these analyses estimate output based on insolation and capacity oriented at optimum orientation? Or do they have an allowance for the fact that most installations are not oriented at the optimal angel (many are far off optimum) and that many have shading for part of the day? If the latter, what is the basis for the assumptions used?

          Also, what allowance si made for the PV systems that are not operating (for whatever reason)? What statistics are available on the number and total capacity of the PV systems that are not working?

        • Peter Lang says:


          I want to remove a wide-spread confusion about the use of capacity factor and load factor. The utility industry definitions are:

          CF is MWh over name-plate MW x 8766h (or whatever period).
          LF is MWh over max power delivered that year in MW x 8766h.

          The difference can be important.

          US Transmission planning expert, Gene Preston, just confirmed for me the meaning of Capacity Factor and Load Factor in US (Australia uses these terms in the same way as the US):

          capacity factor is a source of power and load factor is a load or consumption of power. Its the energy divided by the max possible energy. Or it could be the average output or load divided by the maximum power out or consumed. They are not the same thing, but similar idea. Don’t use load factor to describe generation and don’t use capacity factor to describe load. [Geert’s] definitions below are ok. Name plate is generation and delivered is load.

          [My bold added]

        • Greg Kaan says:

          wind installations quote their output to the grid, never their house load which they take from the grid to operate a dozen supporting systems on the wind tower, and which is considerable. I saw US figures quoted as 10-12% of output, but lost the reference. Wind operators presumably hide the house load numbers because they are dreadful

          Geert, here is an article from 2011 by Willem Post (who often comments here) which analyses the perfomance of some US wind farms and small installations plus the consequences of wind intermittency on the Colorado and Texas grids.

          Additionally, there is a analysis of the performance of a single turbine at the University of Maine which includes the parasitic power drawn during a on windless day summer and winter day. The winter figure of 5.6kW was surprisingly high.

          Hopefully, Willem can provide a link to some more recent information.

          I also found the following article analysing the performance of a Vestas V82 turbine at the University of Minnesota which found that zero wind power consumption could peak around 50kW

          I looked at one of the production record files linked from the article to confirm this and if anything, 50kW peak appears to be an underestimate. The average consumption for this turbine during windless periods does not seem reliable, though, when you compare these to the maximum wind speed recorded in these periods..

        • Thinkstoomuch says:

          From the EIA Monthly report for January 2016 page 213.

          or Page 19 of 29(much smaller file) in:

          Average Capacity Factors
          This section describes the methodology for calculating capacity factors by fuel and technology type for
          operating electric power plants. Capacity factor is a measure (expressed as a percent) of how often an
          electric generator operates over a specific period of time, using a ratio of the actual output to the
          maximum possible output over that time period.
          The capacity factor calculation only includes operating electric generators in the Electric Power Sector
          (sectors 1, 2 and 3) using the net generation reported on the Form EIA-923 and the net summer capacity
          reported on the Form EIA-860. The capacity factor for a particular fuel/technology type is given by:

          (missing image of the equation)

          Where x represents generators of that fuel/technology combination and m represents the period of
          time (month or year). Generation and capacity are specific to a generator, and the generator is
          categorized by its primary fuel type as reported on the EIA-860. All generation from that generator is
          included, regardless of other fuels consumed. Available time is also specific to the generator in order to
          account for differing online and retirement dates. Therefore, these published capacity factors will differ
          from a simple calculation using annual generation and capacity totals from the appropriate tables in this
          End Quote

          Missing image CF=(generation /( capacity x time) {more or less}

          Also most here are probably aware but I was not is there is a difference between net and gross capacity. I believe EIA uses net. But wiki and such seem to use gross. Which matters somewhat for example Solana Generating Station has a gross rating of 280 MW but a net of 250 MW.

          Delivered 718,843 MWH in 2015 and 603,567 in 2014. Rated Gross would be 2,452,800 MWH and net of 2,190,000 MWH.

          Monthly output:

          I have not been looking at the wind stuff yet. There is boatload of stuff in there. I am not very good at sorting except in a slow tedious fashion..

          Thank you all for the thought materials.

          Have fun,

  19. robertok06 says:

    I don’t know you, but I’m getting tired of reading over and over again this nonsense BS about the capacity factor of PV in some countries… so I think it would be useful to make reference to something like this:

    Next time I hear someone mentioning Germany’s incredible electricity production, I’ll just ask the writer to look at the map and compare the colours. Note, for reference, the colour of Italy, and the one for the sunniest part of Germany, in the south. No comparison, right?
    Italy average CF is 14.5% in 2014 (last year for which data are available).

    And next time I hear some BS about the incredible potential of Scottish PV… I call a doctor. 🙂


    • Willem Post says:


      Please also show a wind map with 80 and 100 meter masts.

      Germany has lots of money to make fool of itself regarding solar.

      The German mass media have been conducting a campaign of organized lying regarding wind and solar.

      There is no way other EU countries will follow that folly, PLUS they do not have the money.

      Regarding wind, only offshore in the Baltic are winds good enough to justify the $4500 – $5000/kW cost, and on windy days much of that energy is absorbed by Norway’s hydro plants, which merely reduces water through its turbines.

  20. Thinkstoomuch says:

    With all the talk about Capacity Factors I thought I would share this link for the US.

    Just happened to stumble across it. Today.

    It is for utilities scale and is a fairly recent addition so numbers for wind only calculated back to January 2013 and PV back to August 2014. All on one chart with Nuclear.

    Make of it what you will.

    Have fun,

    • Peter Lang says:

      I find the capacity factors for solar PV hard to believe. They must all be located in deserts at low latitudes.

      • Greg Kaan says:

        The solar PV capacity factors also do not correlate well with the net output from this following page

        I tried dividing the monthly 2015 net output figures for PV against the corresponding capacity factor to give a resulting capacity for that period. If the installed capacity was fixed, the result should be constant while if there was capacity added each month, then the result should increase.

        The resulting capacities I get GENERALLY increase as the year progresses but there are dips. This occurs when using either the Utility Scale PV outputs or the Estimated Total PV output.

        I haven’t heard of any large retirements of PV farms in the USA that would account for this variation in capacity

  21. gweberbv says:

    Leo Smith said it right: The problem of nuclear power … is public perception.

    If public opinion demands that under no circumstances not the tiniest amount of radioactive material will ever be released by nuclear power plants, fuel processing facilities and waste disposals you need one safety net after the other, tripple checking of every screw and so on. And if it turns out that anything does not exactly match the standards, the construction or operation of the plant can be stopped anytime.

  22. disdaniel says:

    I think nuclear advocates need to include CF of Japan nuclear plants the past 5 years to their calculations.
    Also what is the LCOE of nuclear than stops producing entirely partway through the “lifecycle”?
    Nuclear power has now had three major “accidents”, I don’t think the industry will survive a fourth–anywhere for any reason.

    • Euan Mearns says:

      Could we have some stats on fatalities in wind, solar and nuclear industries normalised for power generated. Tks

      • disdaniel says:

        Are you talking about the death of birds, fish, turtles or people?
        Do you include the entire supply chain from dust to dust? or just the design or just the construction or just the operation, or just the decommissioning time-frames?
        And how exactly do these stats help you calculate Japan’s nuclear capacity factor between 3/11/11 and 3/11/16? or the cost to Japan of 40ish expensive (and now thoroughly unproductive) nuclear energy sinks?

        • Peter Lang says:


          Nuclear power has now had three major “accidents”, I don’t think the industry will survive a fourth–anywhere for any reason.

          You can get a quick answer to Euan Mearns’ question here:

          Fatalites from:
          nuclear = 0.09/TWh
          wind = 0.15/TWh
          solar = 0.44/TWh

          Wind generation is 67% more deadly than nuclear and solar 5 times more deadly. These are based on full life cycle analysis.

        • robertok06 says:


          Nuclear power has now had three major “accidents”, I don’t think the industry will survive a fourth–anywhere for any reason.”

          The problem is that some notorious anti-nuclear organizations and think-groups are traying to decrease the survival even of nuclear WITHOUT accidents!… just look at France’s reactors: some German and Swiss “green” groups want absolutely to shut down Fessenheim’s 2 reactors, on the pretense that they are dangerous… while in reality they have been upgraded very recently, bringing up to the most modern fleet status their control system, they have even added a couple extra meters of reinforced concrete UNDER the older one, to take care of a possible core melt and collect the corium.
          Another Swiss “green” group with the city of Geneva standing behind it want also to shut down the entire power station of Bugey, along the Rhone river, with its 4 reactors, again, citing its age… on the pretense that it could pollute the water of the Rhone in case of accident… please note that Geneva is UPSTREAM of Bugey!… so there’s absolutely zero chance of the radiation moving upstream… radiation is not migrating salmons coming back to hatch.

          Other example is the 2 Belgian reactors of Tihange and Doel, which had been found full of “cracks” on their reactor pressure vessel… after a couple of years of forced stop for tests (hundreds of tests with state of the art equipment) they have been cleared by the belgian nuclear safety authority which has been listening to the advice of an international panel of experts, Germans, French, even Oak Ridge National Lab in the USA… and yet the greens ask for their immediate closure.

          This is what happens when blind ideology takes the place of rational thinking based on known physics/technology.

          • sod says:

            ” just look at France’s reactors: some German and Swiss “green” groups want absolutely to shut down Fessenheim’s 2 reactors, on the pretense that they are dangerous… while in reality they have been upgraded very recently, bringing up to the most modern fleet status their control system, they have even added a couple extra meters of reinforced concrete UNDER the older one, to take care of a possible core melt and collect the corium.”

            I am sorry, i am really holding myself back on this topic, but these claims are simply false.

            The thickness of the concrete basemat has been increased by a meagre 50cm.


            The original concrete base was not build to catch a core, so these are really only rudimentary improvements. (to 2 meters, when in the real world you want something like 7 m (Fukushima), all of which being dedicated to catch a core!)


            It is also not “green groups” calling to finally shut down this old and dangerous plant. For example Basel Town and Basel “countryside” are requesting the shut down, and in both “parliaments” the greens are a rather unimportant group.



            Basically everybody in range of the plant wants it shut down, apart from the people of Fessenheim who depend a lot on income from the plant.

          • disdaniel says:

            “The problem is that some notorious anti-nuclear organizations and think-groups are traying to decrease the survival even of nuclear WITHOUT accidents!”

            You mean without additional major accidents, I assume…

            “so there’s absolutely zero chance of the radiation moving upstream…”

            I am sure this starement will surprise people living (or rather evacuated and now unable to live) upstream, but downwind, of Fukushima.

      • gweberbv says:


        fatalities are next to irrelevant.

        What really counts to people is the price they can ask for their house once they have to admit that radiation levels in their garden are a few times higher after this completely harmless accident in the NPP next door. This market price is not related to an abstract fatality rate. It is determined by public perception of ‘radioactive pollution’.

      • David McCrindle says:

        Don’t forget hydro power. Dam failures are also dangerous. Remember that the failure of the Fujinama dam (albeit irrigation rather than hydropower) following the Fukushima earthquake caused more direct fatalities than the nuclear stuff did. (My reference is wikipedia so may be suspect).

    • Stuart Brown says:

      One should welcome potential converts into the church… Fukushima Daiichi Unit 1 was commissioned on 26th March 1971, the last, Unit 6 in October 79. Between 2002 and 2005, some of the reactors were shut down due to the Tepco data falsification scandal. Up to 2009 the average CF for the plants including that shutdown was just under 63%, which is pretty poor for a nuclear plant. Even assuming that no power was produced in 2010 and 2011 (and of course every year since) because the numbers aren’t on Wikipedia where I’m looking the average CF to date is over 47%. Contrast and compare with the CF for wind or solar.

      I will have to leave LCOE calculations to cleverer people than me, but just note that even Fukushima Daini could be restarted if the will was there to do so.

  23. Pingback: The Myth Of Expensive Nuclear Power

  24. Peter Lang says:

    Nuclear power learning rates: policy implications
    by Peter Lang

    A revolution could be achieved with nuclear power if we remove the factors that caused the large cost increases during and since the 1970’s, i.e. return to the learning rates demonstrated before 1970.

    Main Points:

    • Learning rate is the rate costs reduce per doubling of capacity. Until about 1970 learning rates for nuclear power were 23% in the US and 27% to 35% in the other countries studied, except India.

    • Around 1970, learning rates reversed and become negative (-94% in the US, -82% in Germany, -23% to -56% in the other countries, except South Korea); clearly something caused the reversal of learning rates for nuclear power around 1970.

    • If the positive learning rates from 1953 to 1970 had continued, nuclear power would cost less than 1/10th of current cost.

    • If nuclear deployment had continued at 30 GW per year from 1980, nuclear would cost much less than 1/10th of what it does now; furthermore the additional nuclear generation would have substituted for 85,000 TWh of mostly coal-generated electricity, thereby avoiding 85 Gt CO2 emissions and 5 million fatalities.

    • In 2015, assuming nuclear replaced coal, the additional nuclear generation would have replaced half of coal generation, thus avoided half of the CO2 emissions and 300,000 future fatalities. If the accelerating rate of deployment from 1960 to 1976 had continued, nuclear would have replaced all baseload coal and gas generation before 2015.

    • High learning rates were achieved in the past and could be achieved again with appropriate policies.

    Continue reading:

  25. Franktoo says:

    Construction has started for new reactors in Georgia and approved for Hinkley Point C. These are newer reactors presumed to be safer in case of a loss of coolant – the mostly likely type to be built (in the developed world), especially after Fukushima. Both are expected to cost at least $6B/MW, not the $2B/MW shown in Figure 12. Since these points are missing from the graph, IMO you are looking at “cherry-picked” data missing the high overnight construction cost relevant to the US and the EU.

    • Peter Lang says:


      Both are expected to cost at least $6B/MW, not the $2B/MW shown in Figure 12.

      I have several questions and comments about your comment.

      1. How did you identify which reactors are which on Figure 12? Do you have the download of the data?

      2. I think you mean $B/GW, not per $B/MW. The y axis on figure 12 is $/kW which is the conventional way to state capital costs for electricity generators.

      3. Only reactors that are completed are included in Figure 12 (except for three S. Korean reactors which are near completion). All other points are for completed reactors. The last US reactor included is Harris -1, which commendec construction in 1978 and completed in 1987 (three Mile Island accident delayed it). I asked Lovering:

      Why are the reactors in Korea that commenced construction up to and during 2013 included, but the new reactors being built in the US are not included even though they commenced construction in 2013, and France Flamanville-3 started construction in 2007 and is not included? What criteria did you use for including or excluding reactors under construction in the past decade or so?

      Lovering replied:

      We only included reactors that have *completed* construction, because otherwise we wouldn’t have total costs for them that’s true of the five reactors under construction in the US and several in Korea and India. We basically included everything for which there is available data.

  26. Pingback: Recent Energy And Environmental News – March 28th 2016 | PA Pundits - International

Comments are closed.