Solar is now cheaper than nuclear. Even in the UK.

The Breakthrough Institute, a Californian environment and energy research unit, has put out an eye-catching report about German solar subsidies. According to Breakthrough’s assessment, the feed-in tariffs paid since the start of the solar boom make PV four times as expensive as nuclear power, even using the inflated costs suggested by the construction of the reactor at Olkiluoto in Finland.

Breakthrough should have made the point – but didn’t – that the initially generous feed-in tariff rates in Germany have been repeatedly cut. The correct analysis would have not have compared today’s nuclear costs with PV of a decade ago but the current costs of both technologies. At 2013 prices, solar PV in mid-latitude countries is now cheaper than new nuclear. Put in the UK context, the proposed EdF power station at Hinkley is now more expensive per unit of electricity generated than solar farms in the south of England.  The implications of this need a great deal more consideration than they are getting.

By itself, the cost crossover  doesn’t mean that countries shouldn’t invest in nuclear power. Nuclear delivers electricity reliably throughout the year. This baseload power is more valuable than PV’s high levels of output at midday in summer when demand levels are low in most of Europe. And nuclear power stations take up little space compared to the land needs for solar farms. Nevertheless nuclear proponents, such as Breakthrough, should recognise the truly staggering improvement in the economics of solar power around the world, mostly driven by the German government’s commitment to PV a decade ago.  Costs have fallen by approximately 75%. By contrast, it probably doesn’t need saying, nuclear has nearly doubled in price.

The analysis

The ‘cost’ of the many options for generating electricity is difficult to calculate. For both nuclear and for PV, the underlying expense  of generating electricity is dominated by the required payment to the providers of the capital needed to build the plant. PV farms, for example, have operating costs close to zero and nuclear power operates at no more than £15 per megawatt hour. Whether nuclear electricity therefore  ‘costs’ £80 or £100 per megawatt hour crucially depends on the rate of interest demanded by financiers on the huge amounts of money needed to construct new power stations. This is even truer for solar farms.

We do know what EdF, the owner of the Hinkley site, thinks it needs to pay its capital providers. Press reports, not denied by the company, suggest that it believes that it needs a minimum price of £97 per megawatt hour in order to achieve a required 10% return on the capital used to build the plant. Agreement has yet to be reached with the UK government that such a price will be written into law as the ‘strike price’ which EdF will be paid for the output from Hinkley. Nevertheless, £97 is consistent with the calculations of outsiders looking at the £14bn financing challenge faced by EdF for the two proposed Somerset reactors.

The question I therefore asked was this: would a ‘strike price’ of £97 per megawatt hour (just under 10p per kilowatt hour) be enough to incentivise developers to build PV farms in reasonable locations on flat land in southern England with nearby grid connections? My extremely simple modelling assumptions were as follows.

  Assumption Notes
Capital cost £800,000 per megawatt installed In line with recent quotes from UK developers. This may rise as a consequence of the possible tariff wars between China and the rest of the world
Operating cost 1p per kilowatt hour produced A large PV farm needs monitoring and some security provision
Inflation 0% I understand that EdF’s requirement for £97 per MWh is index-linked. My assumption is therefore consistent for PV.
Life of the PV farm 30 years  
Yearly loss of power producing capacity as a result of panel decay 0.3%  
Output per kW installed 1000 kWh per year This is achievable across Cornwall, south Devon, and some parts of the rest of the south coast and the Isle of Wight.


These rough calculations suggest that a ‘strike price’ of £97 for solar electricity would yield a return of 11.3% on the funds committed.[1] This is more than the 10% return achieved by EdF on its proposed investment at Hinkley. Electricity from solar PV is therefore cheaper – in good locations – than nuclear.

This can be put another way. Developers of solar farms should be willing to accept a strike price of less than £100 per megawatt hour, if their required return is similar to EdF. My approximate calculations suggest that a figure of £88, indexed to price inflation as with the nuclear company, will give returns of 10% on PV investments. Perhaps as importantly, the financial risk attached to a solar farm is tiny compared to the roll of the dice at Hinkley. Investors will actually need a much lower return on PV than nuclear.

Are these conclusions consistent with the evidence from sunny counties? Yes, they very definitely are. Applications to build large PV farms are flying in to planning authorities. And what is the current price achieved for solar PV? A developer of large farm will receive 1.6 ROCs (Renewable Energy Certificates) worth today around £65-£70. In addition, they will sell the electricity, perhaps for £40 per megawatt hour, meaning that their total income will be just over £100 per megawatt hour. In other words, developers are rushing to build solar farms today at prices only very slightly higher than demanded by EdF for nuclear.

These farms are not always even in particularly good locations, such as the one that the comedian Griff Rhys Jones is currently complaining about in Suffolk. The marketplace is therefore saying that solar power is now cost-competitive with nuclear. I’ll try to address what I think are the enormous implications of this for energy policy, here and around the world, in a note on this web site soon. As we’re coming to realise, the fact that PV is now cheaper than retail electricity (and therefore doesn’t actually need any subsidy at all if the electricity is all used on site) has the potential to really upset many of the assumptions we’ve made about renewable energy. Electricity markets have yet to understand the disruption that is likely to be caused.






[1] This is the Internal Rate of Return of a solar PV projects selling their electricity for £100 a megawatt hour.

  1. Robert Wilson’s avatar


    Any claim that solar is cheaper than nuclear, gas or anything else must deal with the rather obvious problem that there is absolutely no solar at 6 pm in Winter when UK electricity demand peaks. Even if we get large storage online solar is close to useless in December (consider that Germany’s solar capacity factor was below 2% last December, a figure the UK is not likely to exceed.)

    So, any cost calculation needs to factor in the system impacts, which in the case of solar is effectively the need for all electricity during the winter peak to come from something other than solar.

    There is also a rather stark social problem here. Rooftop solar has always been something that does not help the urban poor. Exactly how solar can be rolled out in a way that doesn’t help the rich, and predominantly harm the poor seems to me to be very unclear.

  2. Damon Hart-Davis’s avatar


    Nukes don’t work when they trip out either, and E7 had to be invented to soak up electricity when it wasn’t otherwise needed. So let’s agree that mismatch of generation to demand is an issue, but not absolutely unique to PV.

    Community generation schemes seem to be one way to avoid PV being a “rich only” solution, though it doesn’t really matter who provides the capital and the space if the power is getting injected into the grid. I see no substantive difference between getting people to putt their money directly into (distributed) generation on their roof and via their pension funds and the Big Six into centralised generation.

    So I think that you are clutching at two straw men there.



  3. Chris Goodall’s avatar


    Storage is an enormously important issue. In my opinion, *the* issue. I hope to deal with how we might organise hundred of terawatt hours of storage in the next post. No easy answers, of course.

    Your second point: cost to less well-off consumers. My argument is that nuclear is more expensive than solar. Therefore solar makes things better.

    Energy prices are going to have to go up. The political issue is how poorer households are protected. My preferred solution is for a US New Deal-style approach that employs people to carry out insulation measures for free. This has the advantage of protecting us all from high gas prices (more important than electricity) and also giving large amounts of employment.


  4. Robert Wilson’s avatar


    The mismatch for solar is a far more serious problem than for nuclear. It’s basic arithmetic.

    Let’s say you wanted to get the same amount of electricity from solar as you would from 10 GW of modern nukes. To get this in the UK you would need about 100 GW of solar. Now, electricity demand in the UK never goes above 45 GW in summer. So, what exactly can we do with the excess solar? We either curtail it or store it. Both options will add significantly to the cost of solar. These problems are clearly not as stark with a max of 10 GW coming from nukes.

    You may call these straws, I simply think they are basic engineering realities.


  5. Proteos’s avatar

    I do not know if large solar PV stations are really close to £100/MWh in Britain, but they certainly are in sunny places like the Gulf countries and south of Spain. In the Gulf countries, it’s probably a good financial idea to install panels to avoid burning oil that fetches $100 a barrel.

    The issue with solar power is more that it cannot supply more than 10% of electricity in Northern Europe with the current technologies. And at such a scale, it will make the wholesale prices crash when it produces at full swing. It’s already visible for wind power in Germany.
    The issue is about how the remaining 90% will be produced.

  6. Robert Wilson’s avatar


    This is the real issue. We can throw LCOEs around to suggest solar is this and that. In fact we could have a LCOE of zero, and would still suffer from the rather obvious problem of solar’s dreadful seasonal profile, and spiky daily profile. There is no way around these problems, so we should recognise solar will have limits.

    The wholesale price crash is another huge problem. And as you point out the German situation is not one to give us confidence. Gas plants like to pull in the big bucks during the day. Then solar comes along and takes it from then. Unfortunately there is no solar when we really need it, so we still need these unprofitable gas plants. Consequence: we pay them extra to keep running, which Germany has started doing (

    Your 10% figure is probably about right. This is the point where they will start having days where solar production is more than 100% of demand. Beyond this point you are starting to look at lots of curtailment in summer, either of solar or wind. I’ve even seen people from the German Green party say getting beyond 10% is going to be tough.


  7. Michael Knowles CEng MIMechE’s avatar

    Oh dear Chris, to compare solar pv costs with nuclear is like Sam Laidlw CEO Centrica saying in an Energy World article three years ago that 1GW of wind power is equal to 1 GW of nuclear, forgetting that new nuclear plants have load factors of 80 and 90% whereas wind is anywhere between 16% and 35% for wind plus/minus 15% annually. Solar PV LF is 10% in a good UK year with the sun behind you.

    You are right about EDF Hinkley C costs that have rocketted. The sooner the Government drops the EPR and looks at more economical designs like Westinghouse AP1000 and Hitachi ABWR the better. The latter has been successfully built in under 5 years in Japan not 10years as the EPR is likely to be.

  8. Samuel’s avatar

    “In line with recent quotes from UK developers. This may rise as a consequence of the possible tariff wars between China and the rest of the world”

    Well, talk about an euphemism :) I think you should be aware that the EU plan to impose an average tariff or 47% on Chinese PV import.

    Of course with this in line your nice calculation is not so rosy for PV investment anymore (not counting the fact that you are completely forgeting the externalities of maintaining a baseload infrastructure for when the sun is not shining).

  9. Proteos’s avatar


    whatever 0 carbon technology wins the day, I think capacity payments would still need to be paid. Because be it nuclear, wind or solar, these technologies do not fit well with emergency production. Emergency producers are the real target of such schemes: having to run a week every ten years is not enough to entice someone to invest to guarantee the grid stability. The question is how much do we pay for the whole system, and what does the system provides.

    The wholesale price issue is indeed a real one. The average price that wind power gets in Germany is lower than the baseload price. And the LCoE of wind is twice higher than the baseload price. After more than 10 years of subsidies, there is no end in sight. This is in part because during the build up, there is some over capacity. But other reasons include the fact that wind production is random & has 0 marginal cost, and this may very well prevent wind to fetch a price enough to support itself.

  10. Oliver Tickell’s avatar

    The key question here is how UK can best invest ~£100bn on low carbon electricity. Chris has deftly demonstrated that solar PV is a lower cost solution than Hinckley C – it will produce many more MWh for the money. And of course that is at current prices. If you wait the 10 years it would take to build Hinckley C, the cost will be under half what it is now. But of course SPV would be complemented by other technologies to produce a smoother production profile, including onshore wind – already about half the cost of nuclear – and offshore wind, also cheaper than nuclear power even at this early stage of its development. These technologies are not only cheaper to build than nuclear, but also attract investment at 5% rate of return, while nuclear investors demand 10% owing to the far higher risk profile and the long period of pre-return tie-in. All in all, nuclear is about the worst possible investment UK can make: mad, bad and stupid.

  11. Jani-Petri Martikainen’s avatar

    I mentioned my reservations regarding assumed capital costs already in twitter, but let me repeat it here. Recently constructed PV parks in Germany have had a cost around 2000 euros/kW and if they deliver around 10% capacity factor that amounts to something around 20 billion euros/average GW. US costs seem to tell a similar story. The capital cost assumed here seems to be lower by a factor of about two. I will believe it when I see it. For comparison the ballooned cost of the EPR in Finland is around 5.6 billion euros/delivered GW with a lifetime that is much longer than PV lifetime. Since we need deep emissions cuts and the decarbonization will (let us hope) occupy us for the next century, this lifetime difference should not be ignored.

    Links :

  12. Chris Goodall’s avatar

    Dear Jani-Petri

    Thank you for your comments. I apologise for the terseness of my reply to your tweet of yesterday.

    If I put ‘solar PV’ into Google I am offered a page of solar installers proposing prices to put 4kW on a house in Oxford. Many quote prices of just less than £6,000 for this size of installation, or about €1,750 per kW. This includes panels, inverters, cabling and the full costs of installation, including scaffolding. This is lower than the quoted cost of the huge German farm that you mention.

    The price of PV has fallen dramatically, even in the past year since the German farm was finished. I agree that the figure of £800,000 per MW may well not be sustainable but it does appear to be the current market rate in the UK for a large multi-MW farms on flat land such as disused airfields.

    Chris Goodall

  13. Jani-Petri Martikainen’s avatar

    Oh, I didn’t think you were “terse” any more than what is imposed on you by Twitter. However, I am still sceptical. The thing is that as far as I have understood the price of modules has declined so that we are now in a situation were most of the expenses are from other things. Land costs, labour costs etc. are probably harder to lower than the module cost. The German example that I quoted was just one example. They have similar costs in other recent solar parks as well.

  14. Gage’s avatar


    Are we not forgetting what an American described as ‘socket parity’? For a householder with 4kW of solar PV who uses all the 4,000 kWh generated per year fo 30 years who has paid £5,000 for the installation (to which should be added perhaps another £1,000 for inverter replacement), the cost is just 5p/kWh. This replaces a BG daytime tariff (incl daily standing charge) of 20p/kWh. If the householder is a 20% taxpayer, the pre-tax cost of BG’s 20p/kWh is 25p/kWh. A saving of 20p/kWh. Put another way, each kWh of home brewed solar PV that is used to substitute a BG daytime supplied kWh earns 20p. The householder is paying minus 20p/kWh.

    No other externally supplied and centrally generated electricity supply can ever match this. And this is without any FIT.

    As a 20% taxpayer, my return on my 4kW solar PV that cost £6,000 has been 51% in Year 1 when the FITs that are income tax free are included – and that was in the wettest 12 months ever recorded. The Green Deal could be used for solar PV so as to eliminate fuel poverty while also providing the required insulation. The cost of capital needed would be paid through the ‘golden rule’ with interest payments paid through the household’s electricity bills.

  15. Oliver Tickell’s avatar

    On solar PV cost in UK: it is now firmly below £1/W capacity for large installations, all included.

  16. Mike Lloyd’s avatar

    I shall await Chris’s forthcoming article on storage with much interest.

    In the meantime, I have a question regarding nuclear power (probably for Michael Knowles?).

    Is there a technical ‘showstopper’ that prevents nuclear power to be used in a variable manner compared with its use as base load?

  17. Samuel’s avatar

    There’s no technical showstopper. All new plants (EPR, AP1000) have load follow capabilities. EPR is sold with a capability to fluctuate between 60% to 100% nominal by 5% increment per minute which is quite impressive.

    Now given that nuclear is capital intensive it is not necessarily the best investment for load following capability. Though there is some margin given the cheap price of nuclear energy: France do a bit of seasonal load following with it’s old nuclear plants this is why they have an average load factor of 85% only. and they’re still at around 55€ MWh.

  18. Mike Lloyd’s avatar


    Thank you. So it is an economic rather than technical matter. Given that wind and solar are also capital intensive and involuntary sources of supply, I guess there could be a wider scope for nuclear in a low carbon future.

    I am also watching the plans for smart meters closely. Looks to me that we can expect both ‘function creep’ and ‘mission creep’ with these. Plenty of potential for greater variance in cost to the consumer?

  19. Alex Trembath’s avatar


    Thanks for replying to our analysis. I’d first refer you to a comment I left at the Energy Collective comparing the cost of the EPR to German solar panels installed in 2016 — the EPR is still less costly on a $/kWh basis, and this is before considering difficulties arising in intermittency, integration, and forced curtailment:

    I’d also point you towards DECC’s figures on solar PV. According to DECC, the capex on solar PV installations between 4-10kW range from £2139-2332/kW, or £2.1-2.3 million per MW. DECC also pegs the current LCOE of solar at between £110-160/MWh, and expects a solar plant built in 2018 to have an LCOE over £120/MWh.

    Solar has certainly come down in costs, but we need to look at the unsubsidized capital cost of plants across entire countries.

  20. Chris Goodall’s avatar

    Dear Alex,

    The point I was making that big solar farms – 10s of MW, not the 4-10 kW you mention – are now being built in increasing numbers in the UK for revenues below the proposed support levels for the EPR. The market, rightly or wrongly, is saying that PV (at scale) is cheaper than the over-engineered and complex EPR.

    DECC and other sources are simply out-of-date on this. PV prices in the UK have plummeted in the last six months to a year (but may rise again as a consequence of anti-dumping duties). Your figures from Germany were, in my opinion, also not current.

    The EPR is reportedly budgeted at £14bn for the two units in Somerset. This works out at about £4.5k a kilowatt. Solar is 20% of this. Solar yields in the South West of the UK are about 12% of rated capacity. So, yes, solar’s capital cost is more per lifetime kWh but it can be built in six months, not seven years and has no fuel costs or waste problems.


  21. Carlo Ombello’s avatar


    You’re spot on with your article. This is something I had in mind to write about for quite some time. And not only is what you say true, it will also be outdated over time with further decreased PV costs (duties on Chinese panels will boomerang within a year), increasingly accessible battery storage, and further bad news from the nuclear economics front.

    A few solar skeptics still exist, and they point at intermittency to make a case, despite the fact that solar output blends naturally with a daily national electricity demand curve, and that any traditional huge power plant doesn’t per se follow local demand but is part of a nationwide network. What will they say in 2 years, when solar will be cheaper still with storage? When solar will even qualify for capacity payments whilst being the cheapest option of them all?

    Maybe they’ll say that solar panels are ugly, or that batteries are bulky, or they will hire comedians to say that drilling the countryside for shale gas is better than laying dark blue solar panels on it.


  22. HNE’s avatar

    Hi Chris,
    Hi All,

    The benefit for residential small roof PV mounted systems is higher FIT and for large scale MW’s ground mounted solar farms is lower unit costs/mass procurement & installation but I’m actually interested in smaller scale PV ground mounted systems.

    I need more clarification on solar systems in the range of 50-150kwp, does this still gain revenue from the ROC’s scheme ? what is the main difference between export tarriffs & ROC’s ?

    I know that for stand alone ground mounted systems the value from the solar system after installation is currently :
    1) Generation Tarrif for Stand Alone systems : £0.0685/kWh
    2) Export Tarrif for Stand Alone systems : £0.0464/kWh
    Researching different websites/articles I found that currently costs will be around 1500/kWp which makes the return on investment slightly less than 5%. Are the costs in your article above for the full development costs (i.e. planning, land lease, engineering, procurement, installation, …etc) or just installation/O&M costs ?


  23. RAJENDRA KANPHADE’s avatar

    The environmental issues, risk factors, hazards due to unforeseen events and sustainability are to be considered. If you consider all these factors the solar energy is much much cheaper than the energy obtained from any other source of energy


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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>