Is solar really four times the cost of nuclear? No – but…

Nobody seems to have noticed, in amongst all the hoopla about shale gas and whatnot, that the UK government has just announced the proposed ‘strike prices’ to be paid for electricity generated by large-scale renewables from next year until 2019.

These numbers are especially important because one of the most thorny aspects of the energy debate is around how much the different options might cost. The fallback argument for anti-nuclear campaigners, for instance, is that nuclear power is far, far too expensive to deliver the low-carbon power we need on any kind of realistic scale. And certainly the dramatic cost overruns seen at Flamanville and Olkiluoto do give serious cause for concern about the potential cost escalations of new nuclear, in Europe at least.

However, many have made similar claims against solar power on the basis of cost – George Monbiot and others have critiqued the feed-in-tariffs currently paid on solar PV on the basis that the money could be better spent elsewhere. I hope yesterday’s numbers can bring some useful balance into the debate, because they represent the electricity prices the UK government is actually prepared to guarantee in order to incentivise renewables investment, rather than being yet another set of self-justificatory estimates from someone with an ideological or financial axe to grind.

Without further ado, here they are (from Investing in Britain’s Future, p.30):


Just to be clear what we’re talking about, ‘strike prices’ are guaranteed prices for 15 years for each renewable technology. Under the proposed Contracts for Difference system currently passing through Parliament as part of the Energy Bill, if the wholesale price of electricity is higher than the strike price in future, the generator will pay the difference back to the taxpayer; if it is lower they receive a subsidy to top it up. Either way this fixes the price for a decade and a half, allowing investments to be made on the basis of a known revenue stream for wind, solar, biomass or whatever. It’s pretty much the same as a feed-in-tariff in reality.

Here’s the important point about the new figures: nuclear is likely to be highly competitive with all the renewables, and may still be the cheapest option. Current negotiations around the ‘strike price’ to be paid for nuclear-generated electricity from Hinkley Point C are understood to be converging on a price in the £90-100 range – my guess is that the final deal will see the UK Government paying just under £95 per megawatt-hour for nuclear electricity under the new system (I’d put money on this – but not much!). This means that nuclear will cost about the same as onshore wind, and may even be slightly cheaper, as onshore wind has a strike price of £100 until 2017, after which it falls to £95.

We should know within a little as a month or two what the strike price will actually be for the nuclear electricity proposed to be generated at Hinkley C. EDF has already spent £1 billion preparing the site and conducting other preparatory work: I now think it is highly unlikely the whole thing will fall through, as yesterday’s announcement on loan guarantees for new nuclear development also suggests. (Important aside: The strike price for nuclear will be for a decade or more longer than the 15 years for renewables, reflecting the 60-plus lifetime of the proposed reactors as opposed to the 25 or so years average lifetimes of wind turbines and solar panels.)

Yesterday’s news also makes it very likely that nuclear will be cheaper than solar PV in the UK at least until the end of this decade. Nuclear also has higher value to the grid as carbon-free baseload power, as compared to intermittent renewables – yes, I know this is another huge debate, but no one can deny that solar will make a minimal contribution during the cold winter evenings when the UK sees its electricity demand typically reach an annual peak. However, the price difference is not huge: nuclear advocates cannot claim that solar is triple or quadruple the price of nuclear on the basis of these UK numbers; it is only about 30% more expensive, and is getting cheaper all the time. By 2019 it will by only 15% more costly, and after that – who knows?

Both onshore wind and solar PV are limited in terms of scale: you can appreciate how much from these one-page backgrounders produced for the DECC 2050 calculator (here’s the onshore wind one, and here’s the solar one). This means that the only very large-scale challenger to nuclear is offshore wind, which can conceivably be deployed at the scale of hundreds of gigawatts in the very conducive environment of the UK’s continental shelf. Offshore wind is considerably more expensive than nuclear, however, at around £150 per megawatt-hour. On this basis new nuclear cannot be ruled out in terms of the UK’s low-carbon energy supply on the basis of cost – and this I think is an important piece of context that is largely missing from the current debate.

Note also that subsidies for wave and tidal are very high indeed – at least three times the cost of onshore wind and nuclear. This reflects the slow development of these sectors, and their obvious need for heavy state funding support for the forseeable future. The pot of money isn’t limitless though: under the ‘Levy Control Framework’ it is capped at between £3.3 billion in 2014/15 rising to £7.6 billion in 2020/21.

Pretty much all these billions of pounds of state support will be for renewables: new nuclear isn’t expected to start generating until 2020 at the absolute earliest. So the good news for renewables enthusiasts is that the UK will see major investments in new wind, solar and other renewable capacity over the next decade, before any new nuclear comes online.

That is something I think we can all celebrate. The UK faces a major low-carbon energy supply crunch, as does the rest of the world, and renewables have an absolutely crucial part to play. Let’s get building.


  1. Robert Wilson


    Very sensible post.

    As you imply the real question is the relative costs of low carbon power sources that we can scale. For obvious reasons solar cannot in the UK, the sun goes and there is more or less none of it in winter. Solar can’t get much past 10% no matter how cheap it is, so it’s a side issue for UK climate mitigation.

    That basically leaves wind and nuclear. I agree with you that onshore wind either cannot, or is extremely difficult to scale. Low power density is the problem. As David MacKay pointed out covering half of the UK in wind turbines would provide 100% of its energy (roughly). That should end the idea that onshore wind is a solution in a densely populated country with a historic nimby streak.

    So, we’re down to nuclear and offshore wind (or CCS) in the long run, i.e. post 2020. Nothing else can get you a decarbonised electricity grid. And the costs of offfshore wind quoted above should make it clear to anyone (though it won’t) that ruling out nuclear will make decarbonisation incredibly expensive or impossible, unless we see quite astonishing reductions in the costs of offshore wind.

    These figures also make clear that the UK would be unwise to sign up to a 2030 renewables target. The 2020s seem to just see a straight choice between nuclear, offshore wind and gas (or CCS if somehow it works.) Unless offshore wind comes down in price a lot then any RE target will be folly.

    1. Peter Simmons

      I suggest you read up on new developments in solar; particularly materials research. Did you think it would all stop with glass-encased panels built in the 1990s? Your assessment of the UK’s solar potential is lamentably behind time, you may have noticed when the sun shines it’s very hot? That’s because the ozone layer is the thinnest it’s ever been, so more radiation is getting through. Add to that climate changes which are causing temperature sensitive species to move north [olants and animals] and the capacity of new PV materials to capture useful amounts of energy even on overcast days, and your dismissal of PV is shown as simplistic at best – Britain cold and wet, Spain hot and dry.

      ‘covering half of the UK in wind turbines would provide 100% of its energy ‘ there we have the classic one solution reasoning that bedevils this issue. Why would anyone want to supply 100% from only onshore wind? Straw man arguments still go round and round. With offshore already well established [albeit by foreign companies since too many in the UK have opposed renewables for so long our renewables industry is behind other European countries] and will grow. Onshore wind has a useful part to play also, as well as standalone wind turbines powering farms or small communities and larger clusters in suitable sites. There are plenty of other sustainable technologies and river energy harvesting hasn’t even begun to be considered, yet we have thousands of rivers which could be suitable; water mills as well as windmills were common before fossil fuels were discovered as the big idea to provide all energy.
      As for nuclear, same argument against still applies, which is conveniently ignored by Linas and other apologists for nuclear – the costs of disposal of waste. Offshore is so much cheaper than nuclear when this is factored in – that’s if one can put a price on a process not yet unvented for safe disposal – the rest of his argument, such as it is, is fiddling with figures and isn’t an argument but a stats exercise. And we all know how that can be manipulated to ‘prove’ anything.
      Another aspect of nuclear conveniently ignored by the brave new nuclear campaigners is cooling; the need for copious amounts of water means they are all on the coast, and, as Mark should know since writing 6 Degrees, the coasts are all under threat from sea level rise; currently 3.6mm annually ignoring storm surges and high tides. The nuclear installations are all at sea level, Sizewell which I know well certainly is. And the micro-brains of government talk glibly of replacing them. Would that be in situ or a few miles inland on higher ground in anticipation?
      The case for nuclear is always put as a technological one, yet basic logic is ignored and problems glossed over. Now we are told the government will underwrite costs for a foreign builder [French] to build these shining examples of fifties technology, and when they say they’ve lost money, we will pick up the tab, along with disposal costs likely to be many billions.
      A solar roof campaign in the UK could contribute so much more to the grid, yet the very fact that we haven’t done much with solar is used as a reason for ignoring it, stats again of solar’s contribution being miniscule, as a reason to dismiss it is twisted reasoning and totally unscientific.
      And we mustn’t forget Fukushima and the ongoing pollution that is understated by proponents who hope the radioactivity will quickly dissipate into the atmosphere and ocean and become our old ‘background radiation’. They can’t eat the food grown anywhwerre near it, and fish caught off the coast are dangerous to eat, but hey, small beer for big nuclear, which, seeing the loss of approval in Japan, and the decision by Germany to move away totally from it, they have to find room to expand elsewhere to continue making big profits.
      With all the other add-ins like insulation, energy waste reduction, storage, and small-scale local solutions, there’s still no argument for nuclear, despite the vociferous campaign by the lobby, of which Lynas is a part.
      Is it just men who want one big solution to everything? Are you impressed with size?

      And anyone labouring under the misaprehension that nuclear is a carbon-free sustainable energy technology should consider the costs [both in money and carbon] of: mining; crushing; processing; shipping a fossil fuel such as uranium. The heat nuclear generates worldwide which will only increase is added to the atmosphere, as is the carbon, but no carbon is removed as with biomass, and it’s not clean and free as with wind. Cost of decommissioning and removing a nuclear plant against removing all wind farms and returning the land to its natural state?

      Concord, another old technology, was retired years ago, nuclear power is from the same era and should also be retired. We have much better, more subtle and clever, technology now than smashing atoms together to make steam. Bit stone age innit?

    2. Clyde Davies

      OK, Peter, let’s (politely) examine some of the points you make. You comment that “Your assessment of the UK’s solar potential is lamentably behind time, you may have noticed when the sun shines it’s very hot? That’s because the ozone layer is the thinnest it’s ever been, so more radiation is getting through.”

      Actually, no. The ozone layer is getting thicker and has recovered substantially since its low point. It’ll take five decades, but the increase in (UV) radiation being admitted by a holey ozone layer has to be balanced against the ever growing influence of aerosols in the atmosphere, which have caused global dimming of about 5%.

      Photovoltaic solar cells don’t make use of UV light: most of their sensitivity is in the red-orange region of the spectrum (which is why they look blue: the blue light is reflected preferentially). There happen to be more red photons in light than at any other visible wavelength.

      The reason we haven’t done very much with domestic solar, I expect, is because this idiotic Coalition decided that take up was getting too successful and so decided to halve the feed-in tarriff, making it less of a clear cut decision to install it.

      You then go on to say: “As for nuclear, same argument against still applies, which is conveniently ignored by Linas and other apologists for nuclear – the costs of disposal of waste.”. Well, you’d be surprised at how little waste a nuclear reactor now generates: read Mark’s book ‘The God Species’. Moreover, part of the UK’s nuclear plans are to build some fast reactors as well as the more conventional variety. Fast reactors burn *everything*, including waste, and can even make their own fuel. Reprocessing for fast reactors does not involve an enrichment step so there’s no danger of proliferation of weapons-grade material.

      Fukushima is another moot point. The number of people who have died from Fukushima’s disaster is ZERO. Zip. Diddley squat. The ongoing risk is also small, as Wikipedia states “In 2013, two years after the incident, the World Health Organization indicated that the residents of the area were exposed to so little radiation that it probably won’t be detectable. They indicated that for those infants in the most affected areas, the lifetime cancer risk would increase by about 1%.” That 1% is relative, not absolute risk. In fact, if you add up all the deaths and sickness per megawatt hour from coal or oil generation, it probably outstrips nuclear tenfold.

      No energy generation is totally carbon-free, but it stands to reason that if you are going make substantial impacts in our CO2 burden, the best place to start is stopping burning the bloody stuff. And then to make up the shortfall with a power source that is available 24/7, when the sun don’t shine, and the wind don’t blow.

  2. Clyde Davies

    Excellent article. If cost were the pivotal factor in this argument we’d all still be burning coal and oil, so perhaps the anti-nuclear lobby had best not rely on that tactic. Looking at the summer we’ve been having, solar doesn’t seem like a particularly profitable bet either.

    Offshore wind seems like the easiest option for now. But we’ll need nuclear to provide backup all the same.

  3. Bill

    Actually not completely correct, Mark. If you are discussing the conventional mega muclear projects then yes. But there are nuclear power systems that are smaller, much mroe economical and completely practical. For example, Hyperion Power (recently renamed to Gen4 Energy) which has developed a small scall nuclear generator that is affordable and uses a non critical mass source.
    Solar & wind simply are not practical for large scale purposes and are completely useless in the higher latitudes.

  4. William Hughes-Games

    I wasn’t certain from your article if the cost of decomissioning a nuclear plant and the cost of storing and ultimately disposing of nuclear waste was included in the price for nuclear power. Leaving that aside, though, small solar is already economically competitive. No ridiculous FIT’s such as Germany has are needed. We simply need true NET metering and a costing system which is fair to both the small generator and the power company.

  5. Jeff Walther

    Wind and solar do cost three times as much as nuclear — five times or more for solar, once you factor in the cost of transmission lines and backup capacity. Who pays those costs? Is it the folks setting up the wind towers, or will it be like in Texas, where the wind turbine owners sell their electricity for about twice the cost of other sources, and then the taxpayer spends $10 billion to put in long distance transmission lines to bring their electricity to market.

    Britain may be in a slightly different situation, being more geographically compact.

    And back up requirements matter as well. With wind and solar you must buy the capital infrastructure at least twice. And that must be paid for somehow, or one does not have reliable power. Hence, Germany starting to realize that it will have to subsidize peak gas and coal plants just to keep them in existence for when needed.

    Your analysis does not come close to charting the hidden costs which are never mentioned in just getting wind and solar onto the grid in a manner which grid reliability can survive.

    But as you say, let’s ignore theoretical numbers which are vulnerable to ideology. Everywhere that has subscribed to wind has seen huge increases in their electric bill, far beyond what could be accounted for by a simple 20% increase in a small percentage of their electricity source.

    Here in Austin, TX the way it works is that energy customers can “subscribe” to wind. They pay the “cost” of wind electricity and in theory all their electricity comes from wind.

    And all of our rates went up 20% last year after no increases for 18 years, at a time when natural gas is at record low prices. Why?

    Well, the wind subscribers pay the “wind rate”, which is a little higher. But when the wind isn’t blowing, the city must purchase electricty on the spot market to compensate. So, rather than subscribing to more affordable base-load power, the city is purchasing a much higher percentage of spot market electricity to compensate for the unreliable wind. With the increased spot market purchases, just to allow wind on the grid, all of our rates have gone up, whether we subscribed to wind or not.

    That’s on top of the $10 billion coming out of our taxes to pay for the transmission lines.

    The transmission line cost alone would have been enough to build nuclear reactor capacity which would out-produce all the wind in West Texas. That’s before you factor in the cost of the wind turbines and the back up capacity.

    If all the money wasted on wind had been spent on nuclear electricity generation, Texas would see a vastly larger reduction in CO2 emissions. And isn’t that what it is all about?

    1. Bas

      Nuclear gets huge liability subsidies.
      The two major one’s concern:

      – the costs of major accidents; all costs, above a ridiculous low amount, have to be paid by the citizen / tax payer. And chance for that is ~once in 6000 reactor years as history shows, while the costs of such disaster (e.g. a jumbo crash) in UK may easily be >1000bilionGBP!
      That imply a subsidy of ~20 GBP/MWh.
      – the costs for the radio-active wast storage. Nuclear plants only have to make reservations for a limited tine (100years?), after that the tax payer has to pay. And these costs are also big as Sellafield shows.
      That implies also a subsidy of ~10 GBP/MWh

      If you take these subsidies away, no nuclear power plant can compete against renewable solar & wind incl. (pumped) storage in the near future.
      Just look how the costs of solar and wind went down and realize this is only the beginning.

      Mark Lynas is wrong regarding the numbers of Wind turbines as he uses the older 2.5MW turbines. In a few year 8MW turbines will be the standard and these will also have more production (produce with lower winds, etc).
      An EU study showed that 20MW turbines are feasible with present technology. So in the end the estimates about the huge numbers of wind turbines are probably a factor 10 to high.

      Furthermore, solar costs per watt went down with 8%/a during the last 30years and all signs are that that will continue if not accelerate.
      So with these strike prices solar farms will become extremely profitable in UK…

  6. Christopher Willis

    Curious, these don’t seem to be cost estimates, but price limits. Won’t these limits just tell us later if the cost of the source is accurate? With that said, the costs aren’t that different from our own EIA’s estimates of LCOE. Nuclear has a lot of potential for prices to come down as well, as some have mentioned with some gen4 projects. Will those be ready for the primetime when it counts remains to be seen, but it can be done if the will is behind it.

    Additionally, other nuclear projects around the world have seen much greater success than the Areva project. We even canceled their fuel reprocessing stuff here in the states for cost overruns. The problem might be as easily fixed as not buying Areva stuff until they shape up their cost performance. With several Korean and Chinese projects finishing in 5 years and billions less than Areva systems in Europe, perhaps nuclear isn’t as out of shape cost wise as we have come to expect here in the west. It makes sense that energy hungry Asia would come out swinging as nuclear bulls, they need non-coal energy in a bad way.

    1. Keith Pickering

      Using EIA’s estimates of LCOE to compare nuclear to renewables is a very difficult prospect, as EIA themselves note, because of the difference in value between dispatchable and non-dispatchable electricity. Beyond that, EIA (and everyone else, as far as I can tell) ignores one of nuclear’s big advantages over renewables, which is the long lifetime of the powerplant. NPP’s are licenced for 40 years in the US, while wind and solar installations are engineered for a 20 year lifespan. So you’d end up buying a windfarm about twice over during the life of the NPP.

  7. Thomas

    Thanks for this article. Yes, in order to produce our electricity in the future, we will need all of them: Solar, Wind, Hydropower, Nuclear, and more efficient use of fossil fuels. Splitting pennies among these vs. the pounds we are releasing currently makes no sense. Unfortunately, a market-based economy is by design unable to provide for long-term solutions. What may appear as the cheapest solution right now may turn out to be an expensive.

    Btw.: What about the 80% of energy that is not electricity? Transport and heating are now mostly reliant on fossil fuels. Wind power is not an option here. Neither Nuclear nor Solar will be able to step in quickly.

    I am sure that in 30 years, or 50, or 70, what is now oil and gas will be covered with a new energy carrier based on renewable fuels. Hydrogen has been proposed for this some 30 years ago..

    1. Peter Vaughan

      Thank heavens for someone who talks sense. Of course if we think that rising carbon emissions is the biggest problem facing the world then we will need to develope all the alternatives. Thus we will need energy provided by wind, solar, nuclear and hydro (including wave/ tidal). I suspect the precise mix will depend on the locality. I.e if one lives in regions with plenty of land to spare (not too sure where these are apart from desert regions) Then wind and solar may be more attractive than nuclear. In a small over crowded Island like Britain it is hard to see how we can meet our future enegy requirements in a low carbon way without having a significant contribution from nuclear. Eventually electric cars and improved electrical heating systems should reduce the amount of oil required for transport and heating. But this will depend on plentiful supply of non carbon energy sources. A plentiful supply of energy will also enable the world to combat the growing shortage of water by increasing the number of desalination plants. It is very hard to see how this can happen unless we include nuclear with the outher low carbon energy sources.

  8. PCAH

    First, nuclear is not low carbon. Existing AGR reactors routinely discharge carbon dioxide into the atmosphere. Second, EDF want taxpayer subsidies for 40 years, not 15 years like renewables. Third, nuclear waste and decommissioning costs are not included in the strike price and will all be met by the taxpayer; that’s £100billion just for legacy waste – and counting. Fourth, EDF and DECC have announced that Hinkley C would not begin operating before 2030 at the earliest. EDF could get on shore wind and solar farms operational on all their UK nuclear sites within two years, connecting into the existing transmission grid which must also be linked into the European wide grid.
    Fourth, the Severn Barrage is a large infrastructure project which will produce twice the amount of electricity as Hinkley C for double the operational lifetime. Investors are ready to go. What is the energy minister waiting for?
    Fifth, none of the electricity from wind, wave, tidal and solar power will have any ‘health detriment’. Even the government’s justification consultation admitted ‘health detriment’ was a cost of nuclear power. ‘Health detriment’ for victim communities around Hinkley Point has left whole families wiped out from premature deaths and heritable genetic mutations. The costs to the NHS have been and continue to be enormous.
    If the lights go out it will be because the government have failed to follow Germany’s example and get on with producing electricity from renewable sources. In any case, those of us struggling to survive the radiation would rather spend time in darkness than continue to pay the cost of nuclear with our lives and the lives of our children.

    1. quokka

      What an appalling collection of untruths.

      1. Nuclear power is low carbon. Furthermore regardless of the truth about AGRs, no more are to be built. The only technologies on the table for the UK are PWRs and BWRs with the possibility of a couple of sodium cooled fast reactors. The UK Parliamentary Office of Science and Technology emissions assessment may be found here:

      2. Operators of new nuclear power plants in the UK by law must meet the costs of spent fuel management and decommissioning. Claiming this is not “part of the strike price” is ridiculous.

      3. EDF and DECC have said Hinckly C not operating before 2030 ????? – provide a source for this claim.

      4. Severn Barrage is both expensive – look at the strike price – and environmentally destructive. The largest proposals would generate about 17 TWh/year. Two EPRs at Hinckly Point would be expected to generate about 26TWh/year.

      5. As for “whole families wiped out from premature deaths and heritable genetic mutations”???? Extraordinary claims need at least some credible evidence.

  9. Bryan

    Ooops ,you mention nuclear too receive 60% longer subsidies but claim £100 over 15 years and £100 over £35 years equal the same amount- No.

    15 x100 =1500 (OnshoreWind) Cheapest
    15 x 155 ==2325 (Offshore wind)
    35 x 100= 3500 – (Nuclear)
    15 x 305 = 4575 (New tech wave and tidal) Most expensive

    Which is higher?

    Which payment would you prefer to receive?
    which bill would you prefer to pay?

    Nuclear to receive 60% MORE ,Making nuclear 60% More expensive than onshore wind!.

    Jeff Walther: Why do you claim USA Nuclear needs no backup? Is USA nuclear failproof and perfect? – Unlikely.

    What happens when they are refuelled or need Maintance? Powercuts?

    Nuclear power needs no power lines then? mmm

    ALL generators need backup until storage becomes a bigger part of the grid ,here or in Texas.

    1. Andy Dawson

      “Why do you claim USA Nuclear needs no backup? Is USA nuclear failproof and perfect?”

      Perfect? No, but very much more reliable than wind or or other renewables.

      There’s a concept called “firm capacity” which is used in grid planing, to decide how much back up you need to carry. It’s basically the weighted average of production which can be called on at short notice from a station.

      A typical firm capacity for a nuclear station (or converntional plant) is around 90% of nominal capacity. For wind it’s around 15%.

      It’s fairly complext to model this into a reliability forecast, but to give you a simple example, say I need to reliably supply 5,000MW. If my nuclear units come in 1000MW units at 90% firm rating

      If I have 6 units I fail to supply if I have 2 stations out of action , the probability of which is (0.1^2) = 0.01.

      If I have 7 units to fail to supply needs 3 stations out, the probability is (0.1^3) = 0.001%

      If I have 6 units of wind, my chances of failing to supply is (0.85^2) = 72%

      If I have 7 units of wind, I’ve a mere 61% chance of failing to supply.

      You see there’s something of a difference? Oh, and note, that’s assuming that the availability of the various units are totally independent – not taking into account something like a large area of low wind.

      If I can respectfully suggests something? It’s not possible to discuss these issues credibly in a purely qualitative manner – it’s essentially a quantative discussion, if it’s going to be meaningful. Or, more crudely – it’s a numbers game.

    2. David

      Certainly a relevant consideration if the nuclear strike price contract turns out to be longer than the renewables one – but poorly represented here. You also have to factor into account the cost of electricity in the future. If the carbon levers work then we can expect this to go up. A long term contract like this cuts both way and forecasting is a mugs game. There is a every chance that EDF will end up subsidizing the UK electricity consumer, especially in the later part of the payment period.

      Whether or not that happens, it is just flat out wrong to multiply the strike price by the contract period to give an indication of the subsidy. For a meaningful assessment, you have to subtract the price paid to generators *at the time* from the strike price. This bit is really important since Hinkley can be expected to be generating during the winter demand peak, when demand (hence prices) are high and renewable generation may be quite low.

      The entire point of the CfDs is to provide long term cost stability to low carbon generators, encouraging their deployment. However in In the case of baseload nuclear this will flow through to consumers as well. Increasing levels of intermittent generation will have the opposite effect for consumers, driving spot prices in all directions depending on hour by hour production. The costs of intermittency are not in any way covered by the CfD, rather they are externalized via the capacity payments mechanism and by simply allowing network operators to charge more for balancing services.

      Upshot: nuclear will be a bargain

  10. David Walters

    “Existing AGR reactors routinely discharge carbon dioxide into the atmosphere”

    Really? Do tell? Care to put some numbers to that?

    1. John Logsdon


      They use CO2 as a coolant and it doesn’t leak or the pressure would be collapsing all the time and so would need replenishing at the same rate. So you would see fleets of tanker loads of CO2 lorries going to Heysham, Hartlepool etc.every day. I think not.

      The problem with AGRs is their enormous size (well Magnox was even bigger but they have all been closed). Approximate figures for power density: Magnox 0.7MW thermal per cubic metre, AGR 1MW thermal per cubic metre, PWR 100MW thermal per cubic metre. AGRs are more efficient at producing electricity because of their higher temperatures but otherwise PWRs trump the lot for being small as well as being able to load-follow, which AGRs most certainly can’t.

      And with their enormous size comes enormous construction costs and enormous waste disposal problems, particularly of intermediate waste.

      PWRs as in the EPR are still pretty large and suffer from being a large shock to the grid if they go off.

  11. kitemanSA

    Actually, the strike price is a lie in that it doesn’t take into account the cost of the power source needed to back up the unreliables. When those costs are included, the REAL cost of unreliables goes way up.

    1. Peter Simmons

      You clearly haven’t heard of storage. With electric cars becoming more plentiful they could provide quite a lot of storage; charged overnight when demand drops but the wind continues, used during the day when demand at peak. And the oldest storage technology is of course the reservoir, with turbines to extract the energy as needed, instantly. There are others. There is no cost for filling in the gaps as you suggest, it’s just a mix of different sources providing spare capacity. Rivers continue to flow, taking energy out of them isn’t rocket science.

  12. Andy Dawson

    I will – having worked on them.

    It’s in the order of 1-5 tonnes per week for a 660MWe unit.

    So, taking the upper range of that, to calculate g/kWh:

    5*10^6 / (660*10^3*24*7) = 0.04 g/kWh

    A typical number for CCGT is 490 g/kWh

    A typical number for coal is 950g/kWh.

    As to “Fourth, the Severn Barrage is a large infrastructure project which will produce twice the amount of electricity as Hinkley C”

    No it won’t – it’s got twice the installed capacity, but like most renewables operates with a low capacity factor – so average production would be about 1.7-1.9GW – about 60 percent of Hinkley C.

    On Mark’s original point – Mark, you’re mixing support level and actual cost. Investment in large solar has basically evaporated since the FiT levels were cut back to the numbers you quote, suggesting investors expect prices to be substantially above those levels.

    Similarly, although near-shore wind schemes are continuing, no-one’s committing for far-offshore wind farms – which would be necessary to deliver the large capacity required (see David McKay).

    Finally, although I’ve no “inside information”, I think your expectation on the iHinkley strike price may be too high. The loan guarantee basically fixes EDF’s cost of capital on the programme, by de-risking the project in terms of cash availability. This as a BIG leverage effect on the strike price. If my “back of a fag packet” estimates are right, it should drop it at least into the lower 80s GBP, and possibly into the 70s.

  13. Peter Simmons

    Mark has obviously scared himself with what he researched for 6 degrees, and rather than losing sleep worrying about his kids future, he consoles himself with the big comforter. Environmentalists, who’ve been involved in and learning about how the environment works are on a different level to a writer who found out some shocking facts about how climate change is going to impact. Full marks for doing the research, and writing the book, but you now have some strange bedfellows, from Nimby anti windists to full blown anti-green conspiracists. You may not have chosen them Mark, but there they are, quoting you, chortling to their simple selves that ‘another environmentalist says renewables are useless’ and it may seem like you’re getting support, but where is it coming from?

  14. J. M. Korhonen

    Thanks for this post, Mark. There are again some usual talking points expounded without much evidence to back them up, so here are some claims and findings from my own literature search.

    1. “Nuclear waste problem has not been solved.” Not true; you’re welcome to visit Onkalo in Finland to see just one practical example re: how to solve it. Sweden, too, has been working on a similar repository. Even better options than burial are likely to be available soon, sodium reactors being just one possibility.

    As far as this claim is amended to “but you can’t guarantee perfect safety for all eternity,” I have to concur – there are no human activities for which you can do that. However, when the risks are weighed and evaluated, I for one am happy with the level of safety provided: the most conclusive report so far concludes that the worst-worst case leak could possibly result to a dose rate increase for the single worst affected person in the region of 1 mSv per year. (Likely dose rates even from serious leaks are about 1/10th of a mSv.) Given that this is about the same rate increase one accepts on average when living in Finland, and about 1/5th of what employees and MPs in our esteemed House of Parliament get as a result of working 8 hours per day in a granite building, I refuse to be too concerned.

    These estimates come from Finnish Nuclear Safety Authority (STUK), and while the obvious counter is that they’re just in cahoots with the Big Bad Nuclear Industry, the inescapable physical fact is that the full Onkalo – containing all the high-level waste from four reactors over 60 years of operation – will at maximum have as much water-soluble, mobile and dangerous radioisotopes as approximately 27 self-luminescent EXIT signs. That’s because only radioisotopes relatively close to fuel pellet surface are even theoretically able to escape to groundwater, and while the pellets themselves contain more – a lot more – the time it takes for them to migrate along pellet grain boundaries combined with the relatively short half-life of dangerous products means that the maximum level is never exceeded. The only way escape could be faster is if the pellets themselves are physically crushed or dissolved, an unlikely enough occurrence for small ceramic items in completely backfilled tunnels 500 meters down the bedrock. (You tea-drinkers can easily estimate how fast ceramics dissolve in water by counting the number of teacups you’ve had to replace as a result of hot water corrosion. Although not a tea-drinker, I’d hazard to say the number would be reasonably low.) And even if this happens, the other mobility barriers have to fail in order for pretty much anything to reach the surface.

    It is notable that even Finnish Greenpeace, despite its known position on these matters and spending the majority of its effort in fighting nuclear (as opposed to, say, fighting coal), has not challenged these estimates, nor have any other campaigners I’m aware of. They merely claim that some material could eventually reach surface (which is true) while resolutely refusing to discuss the likely biological effects. (I might add that there are cities in Finland where at least a portion of the inhabitants are likely to receive 10 mSv extra per annum, and the average extra dose we get is that magical 1 mSv. If these levels cause excess disease, it has so far eluded detection, despite Finland having so comprehensive medical registries that our data is much sought after for various epidemiological studies.)

    By the way, the costs of disposal are more than covered by the special fee nuclear operators have to pay for each kWh sold.

    2. “Nuclear is not carbon-free energy source.” Of course not – currently no energy source is. However, rather extensive research on the subject, including IPCC assessments, concludes that the overall lifecycle greenhouse gas emissions per unit of energy produced from nuclear power – specifically including those things so many claim are missing from the numbers, such as mining, enrichment and disposal – are comparable to those from the best renewable, wind. Nuclear has also, according to IPCC figures, only some 1/3th of the carbon footprint of solar PV – and that’s before counting the carbon cost of storage or power grid extensions that renewables inevitably require. Furthermore, novel enrichment methods and the spread of fast reactors that don’t require enrichment are capable of slashing those already low figures even further.

    Similar results apply to EROEI studies, where nuclear is clearly the best non-fossil technology even when accounting for the full lifecycle. A recent study that specifically compares “apples and apples” is

    Weißbach, D., Ruprecht, G., Huke, A., Czerski, K., Gottlieb, S., & Hussein, A. (2013). Energy intensities, EROIs (energy returned on invested), and energy payback times of electricity generating power plants. Energy, 52, 210–221. doi:10.1016/

    3. “Nuclear gets subsidies for 40 years, renewables only for 15 years and they’re therefore cheaper.” True as long as you conveniently ignore the fact that the lifetime of a renewable generator is likely to be no more than 25 years, while nuclear plants are designed today for 60 years and seem to be quite capable for 80 years. In other words, if the renewables don’t suddenly become very much cheaper, you’d have to rebuild the generators and therefore pay the renewable subsidies at least twice during the lifetime of the nuclear plant.

    4. “Nuclear, too, needs backup and grid connections.” True, but the real question is the scale of backup and grid connections required. As Andy Dawson nicely explained, the difference between backup requirements is significant. Another issue is that renewables require storage and/or backup basically all the time, not just when something goes wrong. If this backup is generated from fossil fuels, the aforementioned CO2 statistics are even more tilted in favor of nuclear. Similarly, renewables have to be built in geographically diverse locations, and therefore require a lot more grid kilometres per unit of energy produced.

    5. “Nuclear needs lots of cooling water and that’s difficult to come by.” The first part is correct, although out of all renewables, only wind and solar PV are significantly less thirsty in terms of “used” or “diverted” water; other options, including geothermal, concentrated solar and especially hydro, are by far thirstier. However, we’ve had for decades solutions to these problems, one such being the icon of nuclear power plants from Simpsons to Sim City: those cooling towers. If desired, nuclear plants could even use closed circuit cooling, although this would have a rather undesirable impact on their efficiency. But far easier, considering that even worst case estimates suggest rather slow sea level rise, is to put nuclear plants close by the sea, and prepare to build higher barriers as sea levels rise in the order of centimeter or three per year (if worst case scenario comes true).

    I’ve stated many times and likely need to state again that I’m all for the deployment of renewables as well, but refusing to even consider the one energy source with by far the best track record for actually cutting fossil fuel use and CO2 emissions (re: France and Sweden; compare and contrast to Germany and Denmark) seems a bit foolhardy to say the least.

    1. Clyde Davies

      Game, set and match to Mr. Korhonen there.

    2. Pete Brace

      “Nuclear has also, according to IPCC figures, only some 1/3th of the carbon footprint of solar PV”

      The IPCC number is based on the 50th percentile figure of all total emission studies – which is pretty meaningless, as this doesn’t look into the quality of data and the bias of the studies.

      Sovacool’s approach of reviewing several studies and looking into the omissions seems like a more thorough method (although even that study did receive some criticism):

      Sovocool gives the following figures:

      Wind – 9g / kWh
      Solar – 32g / kWh
      Nuclear – 66g / kWh


      All of which is fairly irrelevant, as none of them, even combined, are going to be able to produce a significant proportion of the energy most people in the developed world consume. You just need to stop consuming so much.

    3. J. M. Korhonen

      Pete Brace, thanks for the comment. You are absolutely correct in that the IPCC figures by themselves are relatively meaningless, although they’re a good ballpark figure in absence of better information; it’s the situational details that determine which energy source would have the lowest CO2 emissions in any particular case.

      However, it needs to be stressed that the IPCC figures do not include system effects. This is a valid decision since system effects are very significantly dependent on the details, but it does paint a somewhat overly rosy picture of wind and solar in particular. We know from practical experience that when accounting for necessary backup, the CO2 intensity per kWh is significantly higher. How high, exactly, depends, but ballpark figures for “true” CO2 intensity for wind for example range from anywhere between 25gCO2/kWh if all the backup is from already built hydro (with no significant methane emissions from decaying biological matter) to more than 600gCO2/kWh when average production-demand correlation is poor and the backup comes from coal.

      Nuclear, of course, has its own system costs and CO2 impacts, but these are significantly lower, for obvious reasons.

      Sovacool’s figures are likewise incomplete. I stress that this is not a fault of the study – the system effects need to be computed on a case by case basis. However, Sovacool’s meta-study does suffer from the fact that it includes several studies based on one, highly suspect set of studies (so-called Storm & Smith study) commissioned by anti-nuclear activists and suffering from extremely dubious assumptions such as mine covering methods never used in reality, and a spurious method of counting CO2 intensity of construction based simply on financial value and figures for average energy intensity of construction sector in the 1970s. This latter method, counting the energy intensity of construction based on dollar value (and using 40-year old figures at that), is particularly problematic in case of nuclear construction, since we know a significant cost components in nuclear construction are regulatory and financing costs – whose CO2 footprints are effectively very close to zero. Such problems biase Sovacool’s results against nuclear.

      More recent studies that have attempted to correct for these biases include

      Beerten, J., Laes, E., Meskens, G., & D’haeseleer, W. (2009). Greenhouse gas emissions in the nuclear life cycle: A balanced appraisal. Energy Policy, 37(12), 5056–5068. doi:10.1016/j.enpol.2009.06.073

      which gets a result of 17.6 gCO2/kWh for nuclear power in average European energy mix. (Notably, Sovacool’s study is an outlier among several more recent meta-studies of CO2 intensities.) The electricity used in uranium enrichment and plant maintenance contributes significantly; in a less CO2-intensive electricity environment, such as that of France, the CO2 intensity would very possibly be less than half of that.

      As to your suggested solution, that we simply consume less, I must point out that this has not been achieved in practice, contrary to decarbonization of electricity generation which has been achieved successfully in several countries with the help of nuclear energy. While you may be correct in that this would not be the perfect solution, it is at least reducing damages faster than yours.

    4. Pete Brace

      J. M. Korhonen – thanks for the further info.

      I agree that it is also worth making changes to our electricity generation – but this is never going to be anything like enough change.

      Based on the work of Kevin Anderson and Alice Bows at Tyndall Climate Research (for example, “Reframing the climate change challenge in
      light of post-2000 emission trends”), we need to reduce emissions by around 20% per year until we have reached 20% of current levels – or face catastrophic climate change.

      Based on an optimistic 5% growth in nuclear AND renewables, we still need to reduce consumption by at least two thirds within the next ten years. Even if we could achieve the required 700% growth in nuclear over the next 10 years, uranium is still a finite resource – which wouldn’t last long at that rate.

      Just because reducing consumption has not worked so far does not necessarily imply that it won’t work in the future. Many people believe that non-fossil energy sources can solve everything; when they realise that this is not the case – and the choice is change or mass extinction – they might actually consider reducing their consumption.

      Or perhaps everyone will just carry on consuming like there’s no tomorrow…

    5. J. M. Korhonen

      Pete Brace,

      seems we’re pretty much on a same page here, then. I, too, doubt that we will ever build enough carbon-free energy to stave off the climate crisis in time – but I don’t believe that’s a reason not to try. There are degrees in any catastrophe, and with luck, we might be able mitigate the worst of impacts. And long time ago, a lesson was pounded into my thick head, and it hasn’t gone away since: “we don’t expect you to always win, but we do expect you to always fight to the finish” :).

      I’m however doubtful about whether people will radically change their habits. I’ve been involved with environmental issues and their communication for quite some years now, and I seriously doubt it’s the lack of information per se that’s the problem. The real problem seems to be the long-delayed feedback: even if the climate crisis would result to end of the world as we know it, the delay between the effect and the individual decision to, say, not purchase something is decades at the least. Our brains simply aren’t equipped to reliably deal with such problems.

      For that reason, I believe the most likely course is that we’ll continue to burn fossil fuels and the planet while bickering whether this or that technology is acceptable due to possible long-term problems or pure NIMBYism. What I’m doing is more of a rearguard action and documentary; I chiefly hope to prevent further foolishness á la Germany and to prevent people who now claim 100% renewables are possible and even easy from claiming, in 30 years, that no one told anything. Yet I still believe, or hope, that reversing the course is possible. Technically speaking, it certainly is: the planet could be fossil fuel free in 20 years if the industry is put to anything resembling a war footing to churn out modular reactors, wind turbines and solar panels.

      Heck, converting one major shipyard for building modular reactors as replacement for the 1000 largest coal-fired boilers would alone do most of the job in a few years. 18 U.S. shipyards built 2710 Liberty ships in four years, and IFR-type reactor is not much more complicated than that, especially if you could reuse the turbine plant.

      As far as uranium supplies are concerned, the relatively recent MIT study “The Future of Nuclear Power” (last updated in 2009) concluded that even with 300% increase in light water reactor capacity, not even fuel reprocessing would be economically viable before 2050. Fissile materials are non-discrete resources we’re not exactly short of: the erosion alone deposits annually much more uranium in world’s oceans than we could ever possibly hope to use. Extracting uranium from seawater is possible, and that’s what caps the price of uranium. Recent developments in seawater extraction put the price ceiling at somewhere possibly around $300/lb U3O8, nearly 8 times the current spot price but only 2.5 times the high price few years ago. Because reactors use so very little of the stuff, even $300/lb would practically have no effect in electricity prices.

      And then we have thorium, which is 3-5x more abundant than uranium. And then we have fast reactors like IFR, which are so U-efficient that extracting uranium from ordinary granite (where it exists in around 3 ppm concentrations) would be economically feasible with current energy prices and result to net energy gain.

      In short, I’m not concerned the world is running out of U.

  15. SteveK9

    It would be no contest, except for the absurdly high price to build reactors in the West now. This was not the case even 20 years ago when France completed when France completed its nuclear buildout. It won’t be in the future as more ‘standard’ reactors are built. 2 EPR’s like those proposed for Sizewell are more than half complete in China at about 1/4 the cost discussed for the UK.

  16. Edmund Ireland

    The thing about the Hinkley Point contract is, EDF know precisely what all the alternatives to nuclear are and approximately how much they cost.
    Which means they know exactly how much they can charge for electricity without losing the business.

    At £95/MWh a plant that costs ten billion pounds to build will produce something on order of ~£2.4bn in electricity EVERY YEAR.
    The whole thing stinks.

  17. Bas

    At the start of Hinkley Point in 2023 the cost price for solar (PV panels) will be half the present cost prices. So ~60GBP/MWh.
    Hinkley then gets ~93GBP/MWh during the next 25 years.

    Either this brings fantastic returns for investment in solar at that time, or the solar strike price goes down to 70…
    That will make Hinkley Point very expensive for the British tax-payer.

    PV panel prices went down ~7%/year during the last 30years and there are many signs that this price fall will continue for at least a decade.

    What do you think will happen??

    1. Christopher Willis

      The cost of panels are not the cost of deploying a solar farm. Drops in solar panel prices, therefore, do not represent a linear drop in deploying solar installations. Further, it doesn’t account for external backup systems or other abstract system effects of renewable intermittence, for which the OECD has a small investigation into (though hardly a conclusive set of figures).

      Further studies like this are required, but cost externalities like voltage support, and system stability need to work their way into LCOE metrics for them to be a more accurate representation of the larger effects in play.

    2. Bas

      The prices I gave are based on all installation costs including maintenance, etc. unsubsidized in Germany. But even if Chinese panel installations are 10% more expensive in UK due to less efficiency etc., that will not change the grim picture.

      The Germans calculate their guaranteed feed-in tariff on the cost-price +6% margin. Those went down with ~15%/year during the last 10years:

      They are widely predicted to continue their way down e.g:

      At ~2033 wholesale prices will be ~0,03GBP, which is below the operating costs (=costs of only fuel, staff, maintenance) of Hinckley. You can see the first signs of that situation in the German market (read this Bloomberg):

      So by that time, UK’s tax-payer has to pay more than 50% of Hinckley’s costs to the owner (or the electricity consumer via a surcharge).

    3. Bas

      Btw. The big margin between the Hinckley guaranteed price of ~93/MWh and the solar cost price in ~2033 of ~30/MWh deliver enough to adapt the grid and improve storage facilities.
      In combination with wind, waste burning PP’s, Hydro, smart grid, etc. those storage capacities are not such a big issue.

      – install power lines to Norway, they have an enormous unexploited potential for pumped storage and like to do more (e.g. check
      Denmark (~35% generated by wind), Germany and NL do that too.
      – the Scottish are busy with a pilot plant that converts electricity into fuel (which can be stored an used to fill the gaps)
      – the Germans with a pilot plant that converts electricity to gas that can be injected in the gas distribution system without changing that system.
      – improve your own pumped storage / hydro, and use it only to fill the gaps

  18. Andrew Gould

    At last a sensible forum! Somewhere in all these comments are the beginnings of a solution… Correct me if I am wrong, and there will be many much much more knowledgable than I am, but the new generation of fast breeder nuclear reprocessing reactors actually use nuclear waste as fuel and the waste gets re-processed. There’s still readiocative waste, but it has a much shorter half-life meaning it won’t still be radioactive in thousands of years. This is low carbon, and there’s enough nuclear waste already in the country to last out a very long time indeed. Combine that with a significant amount of offshore wind, say 80% nuclear, 20% wind to cover all the UK’s energy requirements. I believe David Mackay’s book is one of the best ever written on the subject. Given the urgency outlined in Mark Lynas’ 6 degrees, I just wonder if there is time for the political environment to change to allow large scale nuclear?

    1. Bas

      “… fast breeder nuclear reprocessing reactors…”
      Germany build one at Kalkar. When finished, it was decided not to start up as it showed to be to dangerous (a loss of many billions)…

      The problem is that they are also very fast if something goes wrong. Three Mile Island (TMI) and Fukushima showed that your have hours to correct, but with these fast breeders you have only minutes… And then they will/can explode.. Something that the Fukushima & TMI Light Water Reactors do not

      Note that the explosion at Fukushima was not the reactor but hydrogen gas inside the building that build up as the spent fuel pool was overheated and all ventilation in the building was stopped.

      US had a small test thorium reactor in the sixties at Oak ridge. It showed to be rather difficult to manage, especially decommissioning was difficult.
      The China and India (not friends…) decided to cooperate in order to develop one. Development time still at least 10 years.
      So it seems that many hazards show already in the design phase …

    2. George Lerner


      Fast breeder reactors aren’t “fast” like “you have only minutes”, they use neutrons that are literally traveling fast. The change in reactivity is controlled by other things (especially control rods). Chernobyl was a thermal (slowed neutron) reactor, with a design where steam bubbles could make the reactivity increase rapidly (and other design problems, no containment building at all, and a management culture that ignored safety). The main risk in some types of fast reactors is the use of sodium or sodium/potassium (NaK) coolants, that react explosively with water or air.

      Several of the Gen-IV Reactors are fast reactors, and all have much better safety than LWR, and much less long-term nuclear waste than LWR. In addition to the Molten Salt Reactor (salt cooled, molten fueled), other Gen-IV reactors might be salt cooled: no risk of chemical explosions.

      The Molten Salt Reactor Experiment (Oak Ridge National Laboratories, 1960s) was not difficult to manage, it was shown to be a highly stable reactor, with high inherent control and inherent safety — no water so no high pressure; molten fuel self-adjusts fission rate (more fission expands the fuel, making it less dense = less fission); no build-up of neutron poisons e.g. krypton (very difficult to balance krypton on reactor start in LWR), no chemically reactive materials, and reactor materials can handle the hottest the reactor could possibly get. (LWR materials can’t, e.g. zirconium fuel rods oxidize at ~500C, only 45% above normal temp, releasing hydrogen from water, and further heating the fuel rods).

      MSR has inherent safety (atmospheric pressure, stable reactivity, simple control systems, materials that don’t melt or chemically react) and passive safety (evacuate fuel to cooling tanks, triggered e.g. by remote earthquake sensors or any overheating, by a “freeze plug” melting, to tanks where fission is impossible and no power and no water is needed).

      MSRE decommissioning was only difficult for one reason: politicians denied funds for that, so it was Not decommissioned for decades, allowing the expected chemical reactions to take place over time. (These chemical reactions are part of normal operation, and easy to handle during a “normal” decommission.) Who in their right minds would leave a non-decommissioned nuclear reactor “just sitting there” for decades??? (Oh, right, politicians aren’t in their right minds!) Only when the scientists who kept pushing for decommission, pointed out it could start fissioning soon, did politicians finally approve funding. Normal MSR decommissioning would be easier than LWR: fuel is easily separated from the coolant and fission products, no tons of water to deal with, no spent fuel rods, no steam containment building.

      India’s thorium reactors are water cooled, solid thorium+uranium fueled, barely different than a LWR. Problems balancing reactivity in solid fuel with U-235, thorium turning into U-233, and U-238 turning into Pu-239, at different rates. That is a completely different technology than a Molten Salt Reactor, you can’t compare them like minor variations of the same technology.

      China is both working on solid-fueled water cooled reactors with India, and is developing Molten Salt Reactors (probably LFTR).


  19. John Logsdon

    The problem with all these discussions is that Rome is burning. We need power and we need it now, not in 20 years time when everyone has finished complaining about it.

    Personally I favour a mix of solar, wave/tidal, marine current and nuclear. Wind is too variable and no good when there is no wind and no good when there is too much. Solar is at least a bit predictable, marine current is much more so and we need to have the capacity to handle the 10:1 or larger range of electricity demand through the day/year.

    If we are to replace all UK power requirements with electricity (including road transport) then the installed capacity will need to be much more than the approx 60GW (and falling) currently available. I have seen 300GW quoted. There is no way you can supply this without nuclear.

    I have two questions:

    First is that I see no strike price estimates for marine current turbines. Is it that they.are as yet unproven? I would have thought no more than tidal or wave power. They have the advantage over offshore wind (which can be co-located) of predictable flows.

    The second is that all the discussion on nuclear has centred on the the Hinkley C Euro design which elsewhere has proven to be substantially over schedule. Guaranteeing the price that EdF are demanding is maybe more to do with filling their 8 billion Euro black hole than supplying the UK with the power it needs.

    I favour using our indigenous expertise from Rolls-Royce and build collections of small civilanised submarine designs. These can be largely mass-produced, are well understood and put into small farms with maintenance bays for refuelling, the expiring reactor being craned out and replaced with a refurbished one. Such a programme would be largely self-funding after the first few reactors are installed and provide an interesting export opportunity.

    These reactors have worked pretty well for RN. Of course the dynamics, reactivities etc will be a bit different but nothing that a company like RR could not solve. A big advantage would be that they can load follow. As has been noted, the waste from these small PWRs is very little other than the spent fuel which can be reprocessed by a fast reactor.

    1. Bas

      “..Wind is too variable …”
      Check Denmark. About ~35% of their electricity is produced by wind, and they plan for 50% by 2020 (it’s an electricity exporting nation).

      Regarding predictability/continuity: never heard of (pumped) storage? In Schotland they are busy with a pilot plant to convert electricity into (car?)fuel. The Germans build a similar plant near Hamburg that converts electricity into gas that can be injected into the natural gas (piping) system (so we don’t have to change our heating devices).

      Btw. Danmark plans all energy (also heating and transport) to be renewable by 2050. So no need for nuclear or coal or oil….
      In 2015 you get only a license to build a new houses in Denmark, if you show that that house is energy neutral (so solar on the roof, a share in wind turbine cooperation, design with high isolation using the sun optimal, etc…).

      All new less dangerous nuclear power plants are very expensive. Only if you build a dangerous one (such as the old ones) that cannot withstand a small plane crash, etc. it becomes cheaper. The USA AP1000 may cost a little less but that one is clearly more dangerous than the French EPR.

      These new nuclear power plants are more expensive that solar, wind turbines at land, etc.
      If you take away / calculate the huge liability subsidy (if disaster, the tax payer pays) and the waste storage subsidy (after 100years your grand children pay the storage costs via tax) then they are even more expensive than Wind turbines at sea incl. electricity storage costs.

  20. John Logsdon


    I wasn’t suggesting the AP1000.

    I know all about pumped storage, having been around when Dinorwic was built. It was a massive investment to smooth electricity supply because AGRs couldn’t load follow unlike all modern designs – turns 1800MW storing into 1500MW generation in 10-15 seconds and once caused shutdowns at Heysham 1 if I recall because they had the droop coefficient wrongly set on their turbines (the grid frequency changes led HRA to try and correct it in its own when Dinorwic kicked in or out, I don’t recall which).

    Yet Dinowic was split away from the nuclear element under the botched privatisation which has led us to the position we are in now – no forward planning because the market will decide. It did and we are burning gas.

    Good luck to the Danes but I say again, no wind or too much and either the turbines don’t turn or they have to be feathered to avoid accidents which have happened recently in the Lake District. The UK has a much larger power requirement than Denmark.

    As for exporting, it depends on the demand. One of the follies of the Yes Scotland brigade is that they expect to be able to export surplus (mainly wind) power to England. At the moment the renewable feed in tariff in the UK is paid for by all UK consumers but under independence, why should Englad pay to Scottish power generators? If Scottish consumers have to pay the feed in tariff to export electricity they will soon close them all down. If Danish consumers are prepared to pay a feed-in tariff (I presume they have one or there is neglible return on investment) then I am sure Germany is happy to receive. And what happens when there is no renewable – no wind at night etc? You have to have some spinning reserver or quick-start diesel/gas turbine.

    If some fingers were pulled out, England could install much more renewable, which I fully support as we need to move completely away from carbon-based generation. It is just that wind power is not really a good idea.

    Nuclear is an option that must remain on the table and costs based on historic designs are just rubbish. Newer technologies are so much more cost-effective and IFR and other gen 3/4 technologies offer the opportunity to reduce massively the waste problem as well as burning much more fuel.

    Stupidities like Chernobyl (unsupervised ‘experiment’ on a fundamentally dangerous design), TMI (only 3 utility employees who failed to recognise the characteristics of a stuck PORV) and Fukishima (ancient design in a stupid position that should have been closed years ago) blot an otherwise good safety record.

    We can all make mistakes, the issue is whether we learn from them. Closing all nuclear in the world is an option but that is not the only way to learn. Engineering out failure occurs in all other industries – witness the civil aircraft industries, space industry etc. Why nuclear should be the subject of so much emotional objection escapes me. But then some people in the UK didn’t give their children the MMR because they listened to Wakefield’s rubbish that was based on very faulty research and experimental design. So I am not suprised, just saddened.

    1. Andy Dawson

      I’m curious that you seem to have accepted that there are issues with the AP1000 – certainly the GDA process found nothing fundamental, and if anything seemed to rate it ahead of the EPR.

      Some of the features are just plain elegant – going “two-loop” reduces vulnerability to component failure, the evaporative cooling from the inner containment wall, and particularly the idea of putting the reactor itself in a floodable plenum.

      The idea of that latter is simply that by doing so, you ensure effective heat removal from the RPV walls (through local boiling), such that even if fuel slumps or melts inside, burn through isn’t credible. The flooding can be entirely passive (and happens anyhow at refuelling). It does away with the need for a “core catcher” and makes even internal release within the containment unlikely in the extreme.

      By the way, there’s a small irony in your enthusiasm for “submarine type” reactors – decay heat removal through reactor walls into a flooded, vented compartment is part of the safety case for the UK’s submarine reactors. And, since there’s never been any detectable contamination from any of the various nuclear subs lost at sea, it presumably works….

      That’s, together with the evaporative containment cooling is the core of the AP1000 safety concept. As steam rises from the flooded plenum it condenses on the inner walls of the steel inner containment, and dumping water from a passive tank gives a heat removal mechanism from the outer wall of that inner containment (the spacing between the inner containment and the outer shield wall/containment is designed to promote upwards airflow).

      The rooftop tank is good for 72 hours passive cooling (enough for decay heat production to drop to around 2% of normal power output) and can be topped up by something as simple as a diesel firepump (piping is pre-mounted to enable it)

      It’s not my preferred option of the generation III+ plants (that’s the ESBWR/EU-ABWR, which have even better passivity), But I can see little credible criticism outwith the imagination of Arnie Gundersen…

  21. John Logsdon


    BTW I agree that the strike price for Hinkley C demanded by EdF is stupidly high and that the govt should not sign off on that deal. It was never AFAIK raised when EdF took over British Energy and is probably as much to do with the black hole in their accounts. So why the British consumer should end up paying to subsidise the French consumer completely escapes me.

    I also agree that solar PV will reduce over time as more efficient panesl, possible graphene based – will become avaible. In fact we are having our roof done this autumn and will install some PV on it when complete but I want to ensure that the installation and wiring can cope with the larger capacity when we install more or replace the panels.

    But none of these account for the daily variation in solar energy down to 0 at night. Ideally you need some base load but the peak demand is early evening in the winter when the solar power is low and decreasing. So you have to have some way of bridging that gap.

    Installation of massive amount of stored energy on this island is not really feasible – Dinorwic was chosen because it was the best site in the UK. Other sites would store/generate rather less. Tidal barrages, lagoons or whatever are predictable, marine current also but if we want to move away from the carbon economy, then one way or anpther we need much much more electricity.

    New technologies may make the production of hydrogen a lot cheaper but at the moment it is a very expensive fuel to produce. And they are not even on the horizon yet.

    PS I meant HYA in my previous reply of course, not Hartlepool!

    1. Bas

      “….Ideally you need some base load…”
      Within ~15years that is totally wrong.
      You see that already coming in Germany now.

      This summer Solar (and wind) produce almost all what is needed during sunny days. So power plants are only used to fill in the gaps. That delivers a lot of power plants with load factors below ~25% which imply they make a loss.
      So now the German authority got lot of requests for an allowance to close the plants (even one that was became operational in 2010).

      As Germany extents its solar (3GW/a) and wind (~2GW/a) at a slower pace, this implies that within ~10years power plants can only fill in the gaps left by renewable.

      Especially since no power plant (also nuclear) can produce for 1GBP/MWh, due to their variable costs, while solar and wind can as they have no variable costs at all. So for solar and wind it is better to deliver for 1GBP/MWh then to stop delivering…

      So it is for sure that base load power plants are become a stumble block for utilities within ~10-20years (including nuclear).

      That may also the reason EDF asks a strike price during such a long time (35years), as the nuclear plant cannot compete at all after ~2025 (and the building period is ~10years). Just as in Germany, the whole sale price will sink during parts of the day/year towards 1GBP/Wh level due to the competition of solar.

      Within ~20year the cost price of solar will sink towards <20GBP/MWh…
      It's a paradigm change

  22. Bas

    “…Why nuclear should be the subject of so much emotional objection escapes me…”

    Because nuclear radiation affects our genes and hence the quality of generations after us. Even low level (0.5mSv/a) does that already as shown in the rock-solid study of the German environmental institute in Munich.

    After Chernobyl some districts in the south of Germany (Bavaria) got rainfall containing fall-out and nearby other districts did not get any rainfall.
    Furthermore Bavaria had an accurate population registration that included also all stillbirth, Down syndromes, Congenital malformations, etc. that operated at district level since ~1980 (~6years before Chernobyl)

    That delivered the unique situation of comparable districts (same economic situation, etc) where the frequency of stillbirth, etc could be compared:
    – between clean districts and contaminated (~0.5mSv/a) districts; and
    – the period before Chernobyl and thereafter.
    No selection of cases the whole population could be included!

    So this allows to compensate all for confounding factors.
    – If the population started to drink more in 1986 the number of stillbirth etc would also rise in not contaminated districts.
    – If the population in non-contaminated districts behaved better, those districts had already better stillbirth, etc. statistics before Chernobyl.

    The results showed with high significance that:
    – the frequency of stillbirth rise with ~30% per extra mSv/a
    – the frequency of misfits (congenital malformation, Down, etc.) rise with ~60% per mSv/a.
    Check yourself:

    Note that background radiation levels are 1-2mSv/a, while in USA medical radiation adds a lot so US citizens receive ~3mSv/a (Houston even ~6mSv/a).

  23. John Logsdon


    “Because nuclear radiation affects our genes and hence the quality of generations after us. Even low level (0.5mSv/a) does that already as shown in the rock-solid study of the German environmental institute in Munich” and “Note that background radiation levels are 1-2mSv/a”.

    Clearly Steven Hawking is right. We need another planet to live on as we will soon end up like aliens with two heads and no arms, nuclear or not. 🙂 In fact people have lived in Cornwall etc for quite a long time.

    Issues of radiation for example at Sellafield were debunked by Rick Wakeford in the 1990s – much the same increase in leukemia was noted at major construction plants for example in the oil indusrt and was shown to be virally initiated due to migrant workers interacting with a local population.

    I don’t doubt for one moment that the explosion at Chernobyl contributed to major problems although a lot of the thyroid problems would have been avoided by the timely issuance of iodine tablets. And these problems will be with us for many generations to come. The sarcophagous is being replaced, the radiolytic erosion if the concrete inside is causing it to collapse. The initial filters had to be removed because not enough cooling air would circulate, etc etc. It was a complete disaster and the poor Soviet soldiers who were sent in to dump sand and stuff on what was a hydrogen explosion (not a nuclear one) paid with their lives.

    Equally the excess radiation seen in Bavaria plus a number of countries neighbouring what is no Belarus is subject to a lot of uncertaintly. Something clearly happened although association is not of course causality. However this only goes to amplify the stupidity of the RBMK1000 design and the way it was run. You cannot completely rule out the latter problem of course but the engineering and physics are pretty well understood so designing a reactor that cannot fail in that catastrophic way is central to all contemporary reactor designs.

    At the same time as the Chernobyl incident, careful modelling of natural circulation cases had been done on UK reactors. A proposal for a tests at Trawsfynnd had to be abandoned – the physics showed that with the boilers so high up and no two-phase to consider, the reactor could still be cooled if all circulators failed. Public understanding of the engineering difference reminds me of the initial response to the invention of electricity – many people were fearful of it because someone had got a shock. My own grandparents wouldn’t have it (although they had radio). Now we think nothing of it.

    But that is the whole point of not comparing like with like – there is no way that the RBMK1000 reactor system with no secondary containment, fuel within the pressure tubes which on fracture would spread superheated steam onto a graphite moderator (clearly the designers had never heard of water gas aka hydrogen), with control systems that were completely overidden, can compare with a much more modern system. RBMK1000s were designed for onload refuelling as source for the Soviet weapons programme.

    There are of course other failures – Windscale was one which led to a public panic because the authorities to demonstrate their ‘competence’ decided to tip all the mik down the drain. A much better thing to have done would have been to dry the milk and store it for 6 months or so when the I131 component would have decayed.

    PWRs which despite their poor thermal efficiency and high internal pressures are the core of most gen 2 reactors, have performed very well world wide. And AGRs despite not being able to load follow and being physically massive, therefore costly to build and decommission, have not had too many problems, none of them nuclear related.

    But I do agree with you about cost. The amounts demanded for the initial investment are far too high and destroy the case for nuclear. There are better ways to skin that cat which include basing the next generation on known and proven designs which have worked pretty well for the military and about which we (ie RR) know a lot. Future reactor systems, if any ever come to pass, depend critically on investment now. The thing which will kill it in the west is the money yet it doesn’t have to be so. I don’t think the Chinese will be so worried – they are far more likely to see nuclear as a solution to their awful environmental problems.

    1. Bas

      …Issues of radiation for example at Sellafield were debunked by Rick Wakeford in the 1990s – much the same increase in leukemia was noted…

      Apparently you did not read the publication I linked. Leukemia was/is a non-issue in this study (also less serious).
      The issue is a rise of 30% in the frequency of Down syndrome, serious congenital malformations (heart disease, non-functioning limbs, etc) because the radiation level raised 0.5mSv/a (while those areas are >1000mile from Chernobyl). Also a very significant rise in the number of stillbirth.
      Other showed a change in the male/female ratio in the newborn.
      These all show that the genes of these newborn was affected.

      So it is not strange that Swedish research found a significant dip in the intelligence of babies born in the months after Chernobyl.

      To my opinion, these are far more serious than leukemia or some more deaths, as it affects the abilities of the generations after us.

      Your statement:
      “…Something clearly happened although association is not of course causality…”
      If another factor influenced the unborn, then that influence would also affect the unborn in the 10 not contaminated districts. And those districts would also show a rise of born misfits. But they did not.

      Your statement:
      “…PWRs … have performed very well world wide ..”
      The three PWR’s at Fukushima distributed more radio-activity than Chernobyl! Luckily almost all went towards the ocean (thanks for the winds).

      No nuclear Power Plant in UK can withstand a jumbo-jet collision …
      Even the new to be build Hinckley point C reactor of EDF (the EPR) may not withstand such an attack despite its double hull!
      EDF calculated it only to withstand an F16 (compare the impact of a 500kg racing car versus a 20ton truck).
      So there is a clear chance a new Fukushima will arise in UK (e.g. an intelligent boy that works for 25 years to become a good jumbo pilot in order to revenge his killed family in Afghanistan).
      And the major winds in UK are not towards the ocean but…

  24. John Logsdon

    1 Fukushima’s 4 reactors are/were BWRs not PWRs. They went critical between 1981 and 1986.

    2 I did read the CSF article and I don’t necessarily disagree with the qualitative conclusions. I found the change-point analysis interesting but I do have issues with overall (logistic) regression sums where results are quoted giving p-values and confidence intervals (although in the authors’ defence I know some journal editors still demand them). At least overall likelihood ratios should be shown as well as the more detailed distributional assumptions – not everything rare is Poisson nor are effects additive in the log field. I could find no mention of the inevitable hiearchical and cross-classification which occurs in such data or how this was accomodated. The only reference that I could see that was really statistical was the David Cox’s excellent (undergraduate) Analysis of Binary Data which is still relevent but methodologies have moved on a lot from that.

    3 I mentioned the leukemia work of Wakeford because it is a classic example of how populist ‘science’ perverts arguments not because it necessarily had relevence to Chernobyl, although the many suffers of childhood leukemia that arose would dispute this. David Clayton did some very interesting work on this issue in the early 90s if I recall.

    4 To my memory, a lot of work was done on the risks associated with a jumbo jet colliding into an AGR (and I think also a Magnox) station. It was one of the 1-in-10^6 calculations carried out along with the probability of the core restraint system collapsing, overpressurisation faults and so on. Many scientist years of work were put in and as far as I can recall, no serious problems arose. These had to satisfy the NII and anyway I do not question the integtity of the scientists involved. No doubt today’s calculations would be more exact but enormous engineering margins were built in to the construction of the AGRs. Therefore your bald statement that no UK nuclear plant could withstand being hit by a jumbo is just that – a bald statement.

    5 All the same, concentration on the problems of 30, 40 or even 50 year old designs remains irrelevent except to note that I would expect the designers of new reactor systems would know about them and take appropriate action. This is what progress should be about. Doing the same thing and expecting different results is one of the definitions of insanity. I would contend that the difference between 40-50 year old and new nuclear designs is so large as to be clearly not insane, unlike burning fossil fuel with only slightly improved efficiencies. What is clearly insane is to march blindly into the next decade without a clue how our power generation will arise other than ‘the market will provide’. This is a recipe for national and economic disaster with a probability of 1, unlike any of the scenarios postulated by opponents of nuclear power with respect to newer engineered designs.

    Mark’s original and proper question on this post seems to have been hijacked – my apologies for this! Perhaps readers bored with this unfruitful thread may care to return to the intented path. 🙂

    1. Bas

      … work was done on the risks associated with a jumbo jet colliding into an AGR…
      That may be true for AGR.
      But it is very clear that BWR/PWR reactors (the great majority in EU) are not capable to withstand such an attack.

      As Fukushima showed, there is no need to damage the reactor itself. It is enough to damage all cooling in order to create disaster.

      Most BWR/PWR reactors have a spent fuel pool in the dome. If the plane hits at that place then the pool will loose its water, and the spent fuels rods will become overheated within hours producing enough radio-activity to stop any fire brigade. After a day or so the reactor is so overheated it produces a big radio-active plume. Enough to make evacuation up to ~100km downwind necessary.

      “,i>…except … expect the designers of new reactor systems would know about them and take appropriate action……”
      Yes they do, unless those appropriate actions make the design so expensive that nobody buys it.
      And that happened with new designs such as AP1000!
      And, to a lesser extent with the EPR.

      …What is clearly insane is to march blindly into the next decade without a clue how our power generation will arise…
      Agree. But that is not the thing e.g. Germany or Denmark do.

      Germany agreed a transition scenario in 2000 with the goals:
      – all nuclear phased out in 2023
      – 80% of all electricity generated by renewable at 2050.
      Intermediate goals were defined for every decade. E.g. 35% renewable by 2020 etc. Now 23% of all their electricity consumption is renewable, so they are ahead of their transition scenario (and slow down next years as grid adaptations are not ahead).
      German economy goes best of the EU. Germans attribute much of their low unemployment the transition towards renewable.

      Denmark defined 100% renewable at 2050. Now about 35% of their electricity is generated by wind (target for 2020 is 50%).

      If these countries can do it…

  25. John Logsdon

    “Germans attribute much of their low unemployment [to] the transition towards renewable”

    This they may do but any consequence of renewable energy employment will be short term and due to relatively low tech building/installation work. The financial consequences are still have to be seen in the feed-in tariffs that have to be paid by someone and the maintenance of largely unused reserves in still-new power stations that will remain unused.

    The reality is that Germany has been the main beneficiary of the common currency while ensuring continued control over their mainly southern neighbours to whom they have sold massive amounts of goods backed by soft loan guarantees, as has France.

    For Germany the Euro is nicely low (rather like the Yuan is for China but that is maintained by strong currency controls against the USD), while banks such as Deutsche Bank and SocGen are stuffed with loans which they cannot afford to write down. An example is arms sales to Greece, which was at one point the worlds 5th biggest arms importer – neither Germany nor France who were the main suppliers would let the Greeks off the deals. More recently of course Cyprus has been a target with illegal haircuts to depositors’ money because Germany will not sanction the ECB to become a proper central bank unlike all other central banks in the world.

    Meanwhile Frau Dr Merkel, in her foreign policy, seems not to understand that excess austerity will never solve the economic problems of some of her customers. A generation is arising which regards Germany with hostility.

    The full employment in Germany is a consequence of this, plus of course their excellent attitude to quality. It is an enviable platform but one of the two legs is built on dishonesty. Not that Mr Osbourne is any better but that’s another story.

    1. Bas

      Your statement regarding German renewable:
      …financial consequences are still have to be seen in the feed-in tariffs..
      These feed-in tariffs are now far below consumer prices, even near the grid delivery prices of regular power plants!

      E.g. Solar feed-in is now near euro100/MWh, going down fast on a monthly basis (in line with the costs decreases of solar electricity).

      So that burden is falling away…
      And it is much lower than new nuclear.
      E.g. EDF asks price guarantees during 35years(!) of euro120/MWh for Hinckley Point C…

  26. John Logsdon


    Solar panel costs are reducing from your figures at 8%pa but I suspect installation costs are not going to reduce as fast. At the moment UK feedin tariffs re about £140/MW so it will take about 8.5 years to halve the costs. But a substantial part at least of retro-fitting solar to existing buildings is pure building costs and anyway as the capacity will be increasing per installation, larger feed cabling and inverter capacity will be needed. If these costs even stay the same, that means 17 years before the price falls to £70/MW and that is still too high IMHO.

    I agree that the strike price requested by EdF is ridiculously high and should not be signed off (although with the government and fingers in the till. anything is possible). As I said earlier this may have more to do with EdF’s black hole. But I still think an economic – as well as carbon – case can be made for small PWRs along the lines of naval reactors with additional safety features that are not possible in a confined submarine – better containment, boron rods etc. I suspect the later reactors generate several hundred MWt. I wouldn’t advocate BWRs as they circulate radioactive water outside the pressure vessel.

    BTW I am pretty sure that the CEGB et seq would have done the same sort of sums for Sizewell B as for the AGRs in their fault studies – it was almost an industry within the industry at the time. That station was completely re-engineered from the original Framatome design, including rewriting the control software which was originally written in C unbelievably. This is part of the reason whiy it was a lot more expensive but that always happens in what are essentially prototypes and the UK industry is full of prototypes which is one reason why we have the problems today.

    1. Bas

      “Solar panel costs are reducing from your figures at 8%pa but I suspect installation costs are not going to reduce as fast. At the moment UK feedin tariffs re about £140/MW so it will take about 8.5 years to halve the costs…”
      German experience shows that installation costs, etc. go down just as fast (more economy of scale; better yields imply less panels per KW). And new thin films will make installation much faster / easier.

      German feed-in tariff now varies from €101/MWh for bigger, to €145/MWh for small consumer installations.
      Check the table and see how fast these go down:
      Looking to the UK feed-in tariff (€165/MWh), I think UK installers can improve efficiency.

      Building a new nuclear plant costs ~8.5 years. So at the moment the new NPP starts, the UK feed-in is ~€83/MWh (assume UK installers are still less efficient compared to the Germans).
      And that rate is half at the time the new NPP runs 10years!

      And that UK solar feed-in rate is €21/MWh at the moment the new NPP runs 20years still receiving ~€100/MWh (assume UK get 20% off the requested rate)…

      A really enormous extra subsidy above the huge liability subsidies NPP’s already get (not liable in case of disaster, not liable for the cost of long term waste disposal; those two have already a value of €10/MWh).

    2. Andy Dawson

      I’ve been following this conversation for a while, but it’s the first time I’ve had a few minutes to chip in.

      1 – on aircraft impact.

      Yes, it was most decidedly “factored in” on the AGR design and Sizewell B (I did my “Graduate Apprenticeship” with the old UK “National Nuclear Corporation” the builders of the latter AGRs and of the Sizewell B “nuclear island”). Indeed, I worked on some aspects of the safety qualification for this.

      Three points re the AGRs (of which I’m not a fan – having been involved in on-site work for Heysham II and Torness, they were utter pigs to build and operate).

      a – penetration of the primary circuit by aircraft parts is a non-issue. The combined Pressure Vessel / Biological Shield is a prestressed concrete structure with a minimum wall thickness of five metres. One of my jobs (typical of the sort of things that junior engineers get stuck with) was observing tests firing the spindles of F4 Phantom engines (Rolls-Royce Speys) at Mach 2 into much lighter structures simulating the walls of the shutdown buildings. These were about 1 metre thick, and reinforced, rather than prestressed. We could achieve no more than 10-15cm of penetration. The engine spindles were used as they’re just about the most penetrative thing on a plane.

      b – similarly, fire damage resulting from the plane impact. Key safety systems are always multiply redundant, and are required to be physically separated and segregated. The AGR shutdown buildings (which include things like the decay-heat removal back-ups) are at the four corners of the reactor/turbine buildings, so at least 200 metres away from the nearest of the others, and with a substantial structure between them. Sizewell B, and every plant I know of has similar separation/segregation.

      c – The only potential hazard identified for an AGR for plane impact was if the plane managed to hit the reactor hall while the refuelling machine was attached. The worst case then would be a loss of seal integrity from the standpipe for the single stringer being refuelled (damage to the refuelling machine itself wasn’t an issue – they’re big and VERY tough beasts). In that circumstance, the hazard would be of venting coolant through a hole of at most about 0.1M^2 (well within design parameters for a safe shutdown and depressurisation) and damage to the single stringer in transit.

      Note that there’s no analogue for BWR/PWR, since they don’t refuel “on-load”.

      A few other comments.

      No, Sizewell B wasn’t “almost completely reengineered” – in fact, most of the key components are interchangeable with the P4 class PWRs, of which it’s a slightly de-rated version. Most of the re-engineering was in peripheral systems – for example, the CEGB’s f****g stupid insistence of running the station with 2x660MWe turbogenerator sets simply because that was all that the UK turbogenerator manufacturers were equipped to build.

      RE Bas’s comment that BWR/PWR are vulnerable to aircraft impact, it’s hard to know where to start re the errors in that.

      First, BWR’s don’t feature a “dome”.

      Just to run over the fundamental differences; PWRs use a large-volume containment, typically god for a 2-3 bar overpressurisation. That’s the structure usually known as the “dome” – in it, the reactor and primary circuit occupies no more than 10-20% of the contained volume (with the reactor taking up maybe 5%). There are pressure suppression systems within that mean even in the case of a primary circuit break, overpressurisation is kept well <1bar.

      Refuelling therefore takes place IN the containment (although the spent fuel pond is usually external, in it's own massive concrete building).

      The containment building is typically a prestressed or reinforced concrete building of around 1 metre wall thickness, usually with a pressure (but not load) bearing inner steel liner. See above for experimenting with what can penetrate that.

      BWRs are built differently (IMHO, in many ways they're better – far easier to maintain passive cooling, but we'll come to that). They use a much smaller containment (the reactor is maybe 25% of the contained volume, and there are no heat exchangers etc). After the Mark I/2 designs, which used a steel containment all subsequent stations use a reinforced concrete system. Pressure supression is inherent, in that on depressurising, the reactor vents through a large volume of water in the lower parts of the containment.

      The containment sits in a "secondary containment" which is actually designed mostly to maintain negative pressure relative to atmosphere. In refuelling, the containment proper is flooded, which provides several metres of shielding above the reactor head. The containment and reactor heads are removed, and refuelling takes place – the idea in having the pond nearby is to minimise operator dosage). Later designs actually separate the pond and refuelling hall. The idea of the negative pressure is that any air leakage takes place INWARD, and hence any theoretical contamination is retained within.

      The "aircraft vulnerability" claim arises from the idea that an aircraft strike could penetrate the secondary containment; which may, or may not be the case (certainly claims that a light aircraft or small airliner could do so or utter rubbish) – these are still substantial concrete structures with 30cm+ wall thinknesses – by comparison, the WTC was mostly glass, and the perimeter columns were steel of thickness around 7-8mm at the points where the aircraft impacted.

      Were an aircraft strike to do safety critical damage it'd have to

      a) penetrate the containment in order to destroy any primary safety system – to do which it's have to penetrate the outer walls of the upper part of the secondary containment, then penetrate at least 2 metres of concrete shield and about 150mm of steel (on a Mark 1) or 3 metres plus of reinforced/prestressed containment wall (on later models).


      b) happen to hit during refuelling when the containment and reactor heads were open – however, note that the reactor is shutdown at that point by having high boron concentrations in both the containment floodwater and the reactor vessel, so there's minimal heat production there.


      c) in a model with a reactor-hall level fuel pond, manage to damage the pond in such a way that it couldn't be refilled before there was sufficient heating of the rods to cause damage.

      Now, that's unlikely itself, for several reasons.

      The fuel is held in fabricated cages as individual assemblies about 1 metre long – obviously, they're at the bottom of the pond. Depending on the design, they've got 4-5 metres head of water above them. That's not for cooling reasons per se (although it does slow down the time it takes for the pond to warm up if there's a loss of one of the duplicated or triplicated heat-removal systems). It's there for shielding against "gamma shine" from the fuel. I've stood on a gantry over the pool at a UK station, that shows how effective the shielding is, and the fuel's a long way down…

      Which, if you think about it….means any damage to drain the part of the pond that's doing the cooling needs to extend to about 5-6 metres below the level of the reactor hall. That's well down in the massive concrete structures, as are the pumps, intakes/outlets, etc. for the circulation systems. Note also, the pool is typically a steel liner in a concrete structural matrix – not easy to break!

    3. David Walters

      Andy, I’ve always been curious about aircraft impact. Basically large tin cans spread out of dozens of meters at impact. The only serious part of an aircraft are the turbine shafts which could be an issue. But that is not why I’m writing.

      In your opinion suppose said aircraft hits the main bank transformers and associated bus equipment? I’ve wondered about this as it would create a permanent station black out. Your thoughts?


    4. Bas

      Thank you for your extensive story. I don’t know about specific UK reactors.

      They seem to be far more secure compared to e.g. the one here in NL (Borssele), which does not have 3meter armed concrete and less safety.
      While Borssele is 6meter below sea level if the dikes brake (which they do once in 5000years according to the dutch authority), the air intakes of the emergency diesels were only 2 meter above ground (so 4 meter below water level after dike brake). And those diesels are needed as all electricity will fail after dike brake. This was corrected thanks to Fukushima (still we are not sure the diesels and pumps will work wile under water; hence the EU recommended to raise the dikes. Which is still not done.

      Just two remarks:
      1. The plane crash at the dome creates also a fuel fire. I assume that that fire together with the impact damage stops the cooling pumps (it will damage electric wiring, etc).
      And that this stop creates the melt down.
      Not so much that the reactor is damaged.

      Your story assumes that more damage is required.
      I don’t think so, as Fukushima showed that stopping all cooling is enough.

      2. The new EPR (Olkiluoto, Flamanville, and perhaps Hinckley point)
      has a double dome.

      EDF ran simulations that show that it can withstand an (unarmed) F-16.
      Now EDF says that it probably can also withstand a jumbo…
      Others studied it and concluded that it cannot…

      I tend to think the others are right, as EDF did not do a 747 (or an A380) while executing a study with a jumbo will not cost much more.. And a terror attack with that type of plane is far more likely.
      In addition, EDF used the word ‘probably’….

      What are your considerations comparing the strength of EPR compared to a normal or UK PWR?

    5. David Walters

      Well…I am sure I speak for everyone, pro-nukes and anti-nukes on this: don’t ya’ just wanna see them build a reactor dome, pressurize it, and remote fly a jumbo jet into it? Don’t ya’ though???!!!


    6. Mark Lynas (Post author)

      The reactor dome is not pressurized – and some reactors work at or close to atmospheric pressure (not PWRs, obviously). I think you mean the pressure vessel inside the dome?

    7. Andy Dawson


      I can’t reply directly, but assume this will appear in sequence.

      I can only talk for UK practice, but that’s back the the separation/segregation issue I discussed in my first post.

      We’d never have a single grid connection or transformer bank – and they’d be well separated, at opposite ends of the site usually. One strike couldn’t take out both.

      Similarly, for back-up generation; on Heysham II and Torness, our back up was four Rolls Olympus naval gas turbines (basically, Concorde engines). They were at the four corners of the site, with perhaps 700-800 metres between them. In a s**t and fan situation, we could also cross-connect onto the Heysham I station next door.

      I described the physical separation of the shutdown buildings – opposite corners of the station proper, so four buildings for a pair of reactors. I suspect the most vulnerable thing on those was the secondary decay heat cooling heat exchangers (i.e those dumping the heat to air), which are on the rooves of those buildings – but again, any one could service both reactors, so to cause a problem you’d have to take out four units separated by between 200 and 300 metres from its nearest twin.

    8. David Walters

      Thanks Andy…this gives me more confidence. I am a power plant worker but in a conventional thermal unit (and GT units as well) though recently retired. I wondered how the bus system works, both internally (transmission) and back up power locations…given the stupidity of siting back up diesels at Fukushima with the fuel tanks AT the intake structure!! So with all the talk of AP1000s, I have to assume that main bank placement at these units to are segregated into bus zones so that if one fails the other can automatically flop over and provide incoming power from the grid from the other side of the bus.



    9. Andy Dawson

      I’m mainly from a mechanical background, David, so can’t say to much about bus design – I do recall from instrumentation commissioning that triplication with no common modes of failure was the established standard, so I’d assume the same on electrical supply. The EU standard nowadays for any safety critical system is “N+2” – that is there is always at least three-fold redundancy.

    10. John Logsdon

      It makes sense to de-pressurise the dome a little below atmospheric . Then any small activity within the dome, should there be a leak to the outside, will not escape.

  27. John Logsdon

    @Andy – as anyone who has ever flowin in an aircraft, let alone made or repaired one, will know, in order for them to fly they have to be very light and fragile. Apart from the engines, an aircraft would essentaily collapse when if it hit such a monster as a power station – nuclear or otherwise. [Conventional plant also have some sort of containment against flying turbine blades of course – Ince A if I remember correctly once shed blades which flew a few miles.]

    The aircraft impact is a non starter yet frequently quoted by people who know very little about the engineering. Unfortunately this dramatic quote gets picked up by journalists and the like who have a (political) axe to grind. As I said earlier I fully support renewable energy – I just think that wind power is the least reliable of them all. It is fossil fuel of any description that needs to be replaced and unfortunately since the dash-for-gas stupidity following privatisation we are still churning out far too much CO2.

    I don’t know enough about AP1000 or EPR to comment on which is better.

    The problem is purely economic. The Hinkley C design is massive with corresponding construction and decommisisoning costs. By using small buried (for safety from passing aircraft 🙂 ) reactors, possibly in a bath of boronated water (for the decay heat issue) all within the containment vessel with a maintenance bay, a lot of the issues and cost are avoided. This is using well understood technology that are much smaller and therefore cheaper and quicker to bulld. The earlier reactors once operational would fund the later ones.

    The advantage of small reactors is that they are much cheaper to build and it also implies a production line approach which again would save costs. I would obviously support Gen 3 and 4 reactor designs as long as they meet the proper criterion of being completely fail-safe. The GE-Hitachi IFR looks interesting as it would burn up a lot of the ‘waste’ that we already have.

    My comment on Sizewell B engineering was that some aspects was a bit over the top! I was aware of some changes particularly to the control system and a lot of NDT inspection was added but most of my knowledge came from the AGRs.

    You are right about the turbines though – it was a silly decision.

    1. Andy Dawson

      “some sort of containment against flying turbine blades of course – Ince A if I remember correctly once shed blades which flew a few miles.”

      There are dozens of examples – including at least one that came close to sinking a ship!

      In fact, another of my more amusing jobs when working for NNC in my university vacations was coming up with a fortran model to assess the probability of damage to key components and systems from a turbine burst. It was quite fun – you have to understand the distribution of the parts about the plane of each disc, then look at both high and low trajectories, the distribution of the energies of the missiles/parts and so on. It kept me happy for about 8 weeks…

      “By using small buried (for safety from passing aircraft 🙂 ) reactors, possibly in a bath of boronated water (for the decay heat issue) all within the containment vessel with a maintenance bay, a lot of the issues and cost are avoided.”

      Far from it – first, you’ve got to overcome the basic “economy of scale” issue – the amount of materials/fabrication etc. per MW of generating capacity is inherently higher as you go down the size scale; that only gets offset if mass manufacturing effects kick in. You’ve got to go a LONG way up the volume curve for that to happen – look at aircraft manufacture, which still entails huge amounts of highly skilled manual labour, even though the lines for (say) the B737 or Airbus 320) turn out several hundred aircraft per year.

      So yes, a 200MW SMR won’t cost the £3-4 Bn of a 1600MW EPR or EU-ASBWR; it’s most unlikely to be 1/8th the cost.

      Secondly, I think you’re working on an assumption that smaller reactors can dispense with some of the redundancy that’s required for large plant – given regulation, I think that’s a very dubious assumption!

      Above or below-ground is hardly an issue – as you yourself point out, aircraft aren’t penetrative things. And building below-ground is expensive compared to above. Plus we’ve seen a rather strong example of the increased vulnerability of equipment mounted below ground to flooding and water ingress issues in the last 2 1/2 years.

      No, the good engineering approach is to work towards “passivisation” and simplification of the plant. At the most obvious level, a 2-loop AP1000 is likely to be significantly cheaper per MW of capacity than a 4-loop EPR of the same capacity; and indeed, anyone who’s done reliability engineering will show you that it’s likely to be more reliable and safer too. The Chinese seem to have no doubts – they’re proposing to base their programme primarily on a 1400MWe derivative of the AP1000, with later stretches to 1700 and 2100MWe on the table!

      The AP1000’s one good example of that. The ABWR and it’s ESBWR offspring are good examples, too, perhaps better.

      They dispense with the whole of the steam generator/primary circuit. Rendering them entirely passive post accident is easier, too – they don’t need a large mass of water at the top of a tall containment, with all the challenge that brings for seismic qualification. They don’t need the complication of (theoretically) vulnerable primary circuit or recirculation piping – the ESBWR dispenses with circulation pumps completely, the ASBWR moves them “in-vessel” with only external drives. (the downside of dispensing with the pumps is a shorter core, hence worse fuel utilisation, and loss of inherent load-following). Both now feature passive heat removal from the containment, allowing indefinite stable post accident conditions needing nothing more than a small external water supply.

      I can see a little “cross-fertilisation” happening – I suspect there are people at Hitachi and Toshiba looking carefully at some aspects of AP1000 like the floodable reactor plenum!

  28. John Logsdon

    “In fact, another of my more amusing jobs when working for NNC in my university vacations was coming up with a fortran model to assess the probability of damage to key components and systems from a turbine burst. It was quite fun – you have to understand the distribution of the parts about the plane of each disc, then look at both high and low trajectories, the distribution of the energies of the missiles/parts and so on. It kept me happy for about 8 weeks…”

    That sounds like fun and would be a suitable tool for some sophisticated statistics these days. Extreme values anyone? But in the old days I guess it was essentially a deterministic solution that you developed. These days the two can be combined with monte carlo approaches but frequently for all the sophistiication we have available, all we do is to show what good work was done 30 and 40 years ago, which is very reassuring.

    My time was with the CEGB et seq so I wasn’t so well acquainted with the design and construction aspects as work was generally firefighting after some issue had arisen. I am acquanted with some features of the boilers and reactors of gas cooled reactors, mainly corrosion science and stuff and did a lot of statistical modelling. Not much about Sizewell B though!

    The omni-present fortran is still widely used in high performance computing work in physics. Nag are still doing pretty well largely on the backs of early algorithms developed within the nuclear (mainly weapons) industry. Gill, Murray and Wright spring to mind. Not to mention the late John Nelder, also famous in statistics for gneralised linear models.

    Andy clearly was much more involved with the real engineering and it is good to have someone in this discussion who can refute some of the wild assertions made.

    1. Andy Dawson

      I was lucky in that I got to straddle the heavy numerical end, and the “muck and bullets” (I also spent time on Eggborough, so really do know the grimy end of the business).

      Your points on the modelling are good – we simply lacked the computing power to do proper monte-carlo type work, so deterministic approaches were about all we could do. I remember building the first finite element model to look at stressing in the heat-exchanger support skirts for the proposed CFR; days of marking out so-ordinates for the element corners via punched cards, and running the model on bought time on Boeing’s Cray. It seemed pretty mind-bending then to be doing that from Warrington in Cheshire. In fact, the clever bit was that was the first time we’d ever been able to work out what happened over time through a transient.

      The heaviest bit of statistical work I ever did probably seems a touch mundane by your standards. There was a problem with AGR fuel stringers that they got rattled against the walls of the standpipes as they came out, during on-load refuelling. They were surrounded by a graphite sleeve, which had to hold differential pressure (the bottom end was in the core, and then passed through the lower pressure area above the gas baffle). And the local strength of the graphite was variable.

      So, I got introduced to a concept called “convolution” – basically taking a point on one of the probability distribution functions, applying a pdf over that, then doing that again to take account of the “third dimension” of variability – then producing an aggregate curve that let us work out the probability of a weak point in the graphite hitting the standpipe hard and a moment when it was bearing high pressure – and hence how likely it was to fail. All in FORTRAN.

      We used it to decide how tight the inspection process for the stringer sleeves needed to be.

      Nowadays, I can’t do long multiplication in my head…;-(

      (Oh, and you’d not have done much to do with Sizewell B because it’s been a largely trouble free beast – 88% cumulative capacity factor over life to date (and it’d have been over 90% had here not been the pressuriser cracking problem).

  29. John Logsdon

    @Andy – that’s fascinating.

    I was working down the road from you in Manchester probably at similar times. I did a lot of boiler modelling on the pod boilers in particular, which if you recall had really unfortunate temperature tilts at the UTJ. In the end I dug into the model (by a trojan horse that would today be called hacking!) and wrote a bottom-up optimiser based on the Broyden rank 1 update algorithm that worked (incredibly well if O may say so!) so that the essential coefficients for the modelling could be devised from a plant experiment. It is still used today for referruling the pods. Good old Fortran – difficult to do that in C++ and impossible in any supposedly robust object orientated programming language I guess!

    Other stuff I did was to look at high temperature creep in various steels and propose a statistically more rigorous replacement approach – in fact the only statistically sensible approach as was commented. This has subsequently been updated as BS PD66o5, some of the coding for which I wrote in GLIM (an excellent and ignored program). This incorporated variance heterogeneity – the variation in ttf changed with log(stress).

    It is interesting that a similar approach has now been adopted in fatigue work in the US – only about 10 years later as I discovered talking to Bill Meeker some years ago.

    And lots more. These days I am just web building and coding stuff – not a lot of statistics done!

    The industry was then full of people with one motivation – scientific excellence. Yes, there was economic input as well but fortunately we could always appeal to the fact that a safety case had to be robust to be presented to the NII. All hands to the pump and get the station back if and only if it was justified.

    It was really a university of engineering and its demise following privatisation will I think be one of the great losses, along with the sale of some of the government laboratories.

    The fragmentation in the nuclear industry has led to work being done by XYZ agencies who have employed all sorts of people, some good, some bad but generally the project management time well exceeds the scientific thinking time.

    Just like the rest of UK industry and universities really – HR and management rule, sod the innovators. Which is why many of them go overseas.

  30. John Logsdon

    @Bas “While Borssele is 6meter below sea level if the dikes brake (which they do once in 5000years according to the dutch authority), the air intakes of the emergency diesels were only 2 meter above ground (so 4 meter below water level after dike brake). And those diesels are needed as all electricity will fail after dike brake. This was corrected thanks to Fukushima (still we are not sure the diesels and pumps will work wile under water; hence the EU recommended to raise the dikes. Which is still not done”

    I though the NL authorities worked on a return period of 1000 years – on this sloppy island we work I think to 100 years and accept the occasion water ingress, It’s much cheaper!

    But maybe in that area they are working to higher standards. The problem with return periods is that they assume stationary processes. In the event of substantial sea water rising, this could be an optimistic assumption. You have no stationary process in sea level and probably never have had.

    Just looked at the map – perhaps we should be worried in the UK! It’s about as far from London as spasely populated Sizewll (Bradwell is closer but I don’t think there are any plans for that).. And Borselle is very close to large population areas in NL/BE!

  31. Bas

    …I though the NL authorities worked on a return period of 1000 years – on this sloppy island we work I think to 100 years and accept the occasion water ingress, It’s much cheaper!..

    I believe that we had a standard of once in 500years before 1953.
    In 1953 we got many broken dikes, ~1500 deaths and huge areas under sea water. It took some years to repair all dikes.

    In that period Dutch parliament decided to raise the standard 10times.So now once in 5000years.

    Taken the predicted sea level rising, we are now in a gradual process of raising the dikes further (~1-2meter) and making them stronger (wider).
    Starting with the weak spots.
    Last year parts along the coast of S-Holland were finished.
    We hope to be ready at ~2050-2080.

  32. Alex Burger

    Well, now we got the real data for Hinkley C.
    And what we can see:
    nuclear is more expensive than solar and wind energy,
    also with the data, you have given:

    92,5 pounds per MWh plus inflation.
    Hinkley C will start 2023.
    It will start in the year 2023 with (we assume inflation of 2 percent):
    92,5 * 1,02 ^ 10 = 112,75 Pounds per MWh.
    and after 35 years in 2058 it will end with:
    92,5 * 1,02 ^ 45 = 225,50 Pounds per MWh.

    a very bad deal for the people in Britain.
    I think, the people in Britain are still far behind the learning process in Germany. That experience will cost the people in Britain a lot.


    1. Alex Burger

      still a comment to my post yesterday:
      nuclear energy has no Liability Insurance.
      In Germany they computed, what such a Liability Insurance would cost:
      72 Mrd. Euro per year per Nuclear power plant.

      This would increase the cost of electricity to the forty-fold.

      In Germany majority of people is convinced, that nuclear energy isn’t cheap.


    2. Bas

      The most expensive renewable method of electricity generation is Wind turbines at sea in combination with (pumped or chemical) storage.

      But that is still far cheaper than electricity from Hinckley Point C.
      May be not quite at the start in 2023, but for sure in 2030 and onwards…

      And then I don’t calculate the risk premium subsidized by the citizens and government (refer to my response to John).

  33. John Logsdon

    I agree it’s a bad deal, particularly with Chinese involvement which is far more worrying. This probably more to do with the number of people EdF has placed within Whitehall than anything else. EdF is a Trojan Horse that has taken Cameron and his stooges to the cleaners except it is not Cameron who will pay the bill. It’s the consumers which will last well past the time when Cameron will become a bad memory.

    The Germans can calculate what they want but that doesn’t make it right. Is German maths different to British maths? Is the risk of something happening in Germany orders of magnitude different to it happening in the UK? I think not.

    In this area where no such event has happened, liability insurance is a game of trying to estimate what the other guy is estimating and bears no relationship to the actual risk. It’s gaming the system with little or no science behind it at all as long as the insurers make a profit, which they will do as long as they set the premiums high enough and there is no-one else to bid them down. If they can get an obscenely high premium then the shareholders are happy and the directors even happier because they know that it’s all profit.

    That’s what a market is all about. If there is no accident then no-one can demonstrate that the premium was too high or too low. And if ever an accident were to happen, you can bet that the insurance companies are the first on to the lawyers to wriggle out of anything.

    The few cases where someone actually tries to put some numbers in come to such low premiums that no-one will believe them so they bump them up, which is all profit. Bully for them. Garbage in, garbage out. It’s a game that anyone can play.

    There are all sorts of calculations around.

    See for example with which you will no doubt disagree.

    Enjoy your disagreement. 🙂

    1. Bas

      Just simple math give an estimation of the insurance premium the citizen / government pays / subsidizes each year per reactor.

      We forget the many smaller accidents, such as Three Mile Island which released radio-activity which has killed xx people (use LNT).

      All reactors together operated ~12.000years (UNSCEAR).
      There were 2 disasters involving 4 reactors (3 at Fukushima, 1 Chernobyl).
      So chance for disaster was once in 3000 reactor years.
      Assume reactors are now 2 times safer (are they? E.g. Borssele not).
      So chance on disaster once in 6000years.

      Cost of a disaster:
      1 – Exclusion zone (~3000square km) for ~300 years (Cs half life 30yrs)
      Note Fukushima was lucky as winds took 97% of fallout to the ocean.
      So all buildings, factories and other loose all value.
      2 – Damage control (at Fukushima now >€100billion)
      3 – low level fall-out costing extra cancers (deaths), etc. and disastrous heredity effects (~30% more stillbirth, 60% more Down, spina bifida, etc per mSv/year!):
      4 – harm to the people that have to move

      1- Hinckley point’s fall-out winds go to Bristol, London, etc.
      Not to unpopulated areas such as Chernobyl or the ocean (97%) of Fukushima. This implies a loss of ~a trillion euro. A million people out with all infra-structure, etc. is ~a million each.
      2- Assume ~€200 billion
      3- Human live costs of Chernobyl estimates are between 4k (IAEA/WHO) and 6 million despite the unpopulated area. So assume ~1million incl the low level radiation victims that come after 20-60years as shown by LSS (same latency as smoking and asbestos).
      That implies ~€200billion
      4- Those are small compared to point 3. So assume zero.
      Hence total damage: € 1400billion caused once in 6000years.

      That is an insurance premium of at least €230million per reactor per year.
      Add to that the coming costs of nuclear waste storage mainly to be paid by our grand children (simple mesh clearance at Sellafield already >100billion)
      Note that talks about reprocessing/burning in fast breeder are going on during >50years now, without any result, worse results are farther away than ever.

      So an insurance premium should be ~€300million per year per reactor.
      This premium is now subsidized by citizens and Government (tax-payers).
      Also for the two new reactors (in addition to loan guarantees, strike price, etc)!

      So it is safe to conclude that nuclear is by far (10times more?) the most subsidized method of electricity generation.
      Far more than e.g. offshore Wind turbines.

  34. John Logsdon


    I will clearly go with your company as 300 million EUR << 72 billion EUR. 🙂

    But these simple calculations are just that – simple calculations.

    Apart from the distributional assumptions inherent in any calculations of risk, it also depends on the physics. Some of the accidents are not possible with new reactor design. I might suggest all of the serious accidents so your new numerator for new design should perhaps be zero.

    It's like driving down the motorway and expecting your car to fly into orbit. It won't and no-one every expects it to.

    Even with your simple sums, the price quoted by insurance assessors for accidents where there is no real prior information is just guesswork that depends on what they believe the market will pay. Nothing more.

    Perhaps we should be taking insurance policies out against going to war. We have a lot more information on that than anything to do with nuclear power. The UK is always at war somewhere in the world.:)

    1. Alex Burger

      Please read this source:
      72 Mrd. Euro for the Liability Insurance of one nuclear power plant per year:

      this computing was done by official members of the German insurance industry. And this article is posted in the manager magazine, it is a popular magazine for German managers. If they post it, you should take it serious.

      I would say: just let some officials of your insurance industry compute this value. You will get a real value. And I think, you wouldn’t believe it before.


    2. Bas

      I took that improved security into account when applying 50% less chance and deleting all minor accidents that also created (low level radiation) damage (heredity; such as more Down, etc., as well as deaths).

      Some accidents may not be possible with the EPR. But others stay possible, quite well!
      – EPR can withstand an unloaded F16 fighter yet, but not a loaded Boeing (e.g. piloted by a son of a Taliban family UK soldiers murdered; revenge may take 30years there, but is a matter of honour).

      – EPR cannot withstand a terrorist attack, and probably also not an operator that wants to revenge towards society and creates a real famous suicide.

      – Smart EPR operator may want to revenge to society. Takes care to remove all blocking controls (incl. colleague’s) and maneuver the EPR into a great disaster


      Furthermore I estimated the size of the damage rather optimistic.
      Insurance experts will probably come with 5fold or so, which implies an insurance premium of ~1,500million/a. Now all subsidized by citizens and Government.

      Your estimation of no chance is similar to those of the sixties (by ‘experts’!) regarding the unsafe reactors…

      ‘UK always at war somewhere.’
      So what if a smart enemy drops a penetrating bomb from an (adapted) sports plane? Or sends a penetrating rocket launched from a truck?

Leave a Comment

Your email address will not be published. Required fields are marked *