Sorry, Nuclear Power is Not Expensive
I’ve been looking at energy policy in somewhat more depth than usual over the last week or so.
I responded to the panic, sorry “fast-track” consultation on feed-in tariffs (FITs), which I mentioned earlier in the year (maybe more about this later); I attended a Climate Change Campaign (CCC) debate on nuclear power; and, today, masochist as I am, I downloaded the Climate Change Committee’s (also CCC, damn, can’t use that one!) 4th Carbon Budget (let’s call it “the 4CB”), for 2023-7, which apparently we’re all now committed to.
I have to say that participating in debates on energy policy is to enter a parallel universe where the veracity of statements seems to be entirely optional. Especially if numbers are involved. I find it physically painful. Blood vessels in my head threaten to burst.
Just as one example, here’s what the 4CB says on p.254 (Joslin’s 25th Law: the accuracy of the content of a report is inversely proportional to its length):
“Solar PV could play an important role in global power sector decarbonisation, with the IEA estimating that this could generate around 11% of global electricity by 2050. However, the importance of this technology in the UK is unclear given relatively high costs:
• Solar PV is expected to cost around 28 p/kWh in 2020 for large applications (around 5MW) and 45 p/kWh for small residential-scale deployment, compared to around 7 p/kWh for nuclear and between 11-13 p/kWh for offshore wind.”
It’s usual to quote such figures in today’s prices, ignoring the uncertainties of inflation, and this is what appears to have been done here for nuclear power. But the figures for solar PV are bizarre. They are of the order of the current UK FITs, which could probably be halved to something approaching the level in other European countries and still give the intended 5-8% return. And the whole point of the FITs is to build economies of scale to bring PV costs down in the future. 2020 was in the future last time I checked.
I believe the cost of PV is too high now. That’s why I object to subsidising home-owners installing solar panels with absurdly expensive FITs. Nevertheless, I appreciate the whole point of the FIT scheme is to build up economies of scale in order to rapidly bring unit costs down. Presumably whoever buried the above paragraph on p.254 of the report is also sceptical. But few observers of the industry would doubt that the cost will be much less by 2020. Jeremy Leggett is claiming (though somewhat implausibly) that “grid parity” will be reached by 2013.
The reason I’m sceptical about FITs is that, for a relatively small amount of electricity – maybe 1GW peak output – the FITs scheme will cost around £8bn (and that’s just up until 2030), according to the impact statement (PDF) on DECC’s page for the 2009 consultation on the proposal. That makes sense. There’s been talk of a “budget” of £400m, which Osborne wants to cut by 10% (it’s not really his budget as the costs of the FIT subsidy are added to electricity bills). If the £400m is the annual subsidy (it’s none too clear what it is), that would be the equivalent of about 400,000 PV schemes of around 2kW capacity (let’s be generous and call it 1GW in total), each subsidised by around £1000 a year (that is, at 40p/kWh for an average of (1,000/0.4 = 2,500/365 or around 7kWh/day). £400m over 20 years is around £8bn.
£8bn. Interesting figure that.
Coincidentally it’s the same figure I heard from Darren Johnson (Green Party, anti) at the nuclear power debate. He noted that £8bn is the cost of disposing of the waste from 8 nuclear reactors. I spoke briefly to Darren after the meeting, querying the figure. He said he’d heard it from Caroline Lucas and sure enough it’s all over the internet. I was surprised, because £8bn is peanuts. The output of a single commercial nuclear reactor is typically around 1GW (potentially quite a bit more in some of the latest models). And, unlike solar PV, nuclear power is 24×7. So, to decommission 8 nuclear reactors will cost a similar amount to the FIT scheme, which will provide peak power output equivalent only to that of 1 reactor! And the sun don’t shine all the time!
Let’s look at the £1bn waste disposal cost for each nuclear reactor in a slightly different way. How much electricity does it represent? Let’s say we sell it for 7p/kWh wholesale (10p/kWh for consumers would be easier, but I don’t want to be accused of being optimistic – hell, let’s be pessimistic and say 5p/kWh). Now, we’re producing 1 million kWh of electricity every hour (that’s what 1GW means). At 5p each, that’s £50,000 of kerr-chang each and every hour. Still, £1bn is a lot. In fact, it’ll take our reactor 20,000 hours to earn £1bn. Call it 1,000 days, or 3 years to allow for a bit of downtime. But nuclear reactors last 40-60 years. So the waste disposal cost is less than 10% of the value of the output of the reactor. Or to put it another way, less than 0.5p/kWh, according to Caroline Lucas’ figures.
Another number was thrown into the air at the CCC debate. Someone said the Fukushima accident would cost “hundreds of billions of pounds”. Sorry, it’s in the tens of billions (like the Deepwater Horizon oil-spill). It’s a disaster, sure, but – even if we call it £10bn per reactor (there are 6 in total, 4 badly damaged) and take account of the less than 1GW output of most of the reactors (they’re quite old) – call them 500MW units – the clean-up cost is still only of the same order as the value of the electricity produced over the lifetime of the reactors (0.5bn kW * 0.5p/kWh is £25,000 per hour, so earning £10bn takes 400,000 hours or around 20,000 days or about 60 years, allowing for some downtime). And there are 100s of nuclear reactors around the world. It turns out that the cost of a Fukushima or a Chernobyl every couple of decades is in fact insignificant compared to the value of the electricity produced. Sorry, that’s just how it is.
I’m not trying to make an argument for nuclear power here. There are clearly potential grounds for objection other than the cost.
All I’m saying is that the facts do not support the claims that nuclear power is expensive that you hear so often.
And unfortunately most forms of renewable energy are more expensive at the moment. Possibly excepting onshore wind, but no-one seems to want that.
How about reading Caroline Lucas’s report?
“The price of waste disposal for a new build programme has been
estimated, by a pro-nuclear industry consultant, at around £8.2 billion.”
Hello, noticed that? “pro-nuclear industry consultant”! So, that number is the MINIMUM, the real number is probably much higher. Bad job by the green party to let this number stand unopposed with a more reasonable estimate, of course. Regardless, you can’t take that number for real. Just use common sense: Do you REALLY think decomissioning a reactor and storing the waste its generated during its lifecyle is CHEAPER than building it? D’oh.
Btw, the green party report also says “A variety of government estimates from the 1990s, have
put the cost of the 1986 Chernobyl accident, over two decades, at hundreds of billions of dollars.” And they included a source for that statement. Fukushima is at least as bad, probably worse than Chernobyl.
Thanks for the links to the sources, but you should take more care in reading and evaluating that data. Of course, you look fine compared to that idiot Darren Johnson who cites pro-nuclear numbers without noticing it. Nice example how not to educate the public about green issues. What a jerk.
Thanks – I’d just heard the £8bn from the Green Party, as described in the original post. I queried it with Darren Johnson because I thought I’d misheard! It seemed too low.
But the £8.2bn is for 8 new reactors to be built in the UK, of at least 1GW power output each. So the cost is £1bn or so per reactor which would generate ~£30bn worth of electricity (@7.5p/kWh) over its 60 year lifetime. Even if the decommissioning figure were too low by a factor of (say) 5, it wouldn’t massively undermine the case for the reactors.
If the Green Party don’t believe the figure they shouldn’t quote it.
I was referring to costs of compensation and clean-up by TEPCO for Fukushima from articles (e.g. in the FT) which give figures in the 10s of £bns.
The report informing the GreenFacts site you quote from is a reasonably authoritative source but also takes account of opportunity costs e.g. the electricity not generated by the reactor (which is silly) and the farmland (which is now providing valuable ecosystem services including biodiversity – actually that’s what I went to Chernobyl to see!).
For example, p.33 of the report notes that:
and also:
The cost of the new sarcophagus for Chernobyl is expected to be around 10 billion euros, so I expect you could reach £100bn all told.
And Fukushima is worse than Chernobyl only to the extent that 3 reactors have melted down. At Chernobyl only one was affected (several others continue to operate at the site). The explosive loss of containment at Chernobyl made the clean-up more expensive.
The health effects of Chernobyl have been 1000s of times worse than those of Fukushima, and it’s not clear to me that the effect on the “liquidators” (conscripts from all over the Soviet Union) is fully documented.
Having said all this, it has become apparent since I wrote this post that reactors 1 to 3 at Fukushima have all melted down and are probably all leaking, too, which will significantly inflate the long-term clean-up costs. I now expect the bill for the direct costs of Fukushima to reach around £100bn (by, say, doubling the decommissioning cost figure of around $15bn currently being quoted – note compensation claims are estimated at “up to” Y11trn which is around £83bn).
My “tens of billions of pounds” may, on reflection, err to the low side, but even if the final bill for Fukushima is £1-200bn, it’s misleading to characterise that as “hundreds of billions of pounds”, which most people would take as several hundreds, which was the statement I was criticising.
Nevertheless, the question I asked myself towards the end of the post was whether accidents undermine the business case for nuclear as a whole (obviously they’re likely to be terminal for the operating company). Obviously we don’t want a repeat of Chernobyl or Fukushima – and future reactors will be designed to avoid similar accidents – but there may be other things to go wrong that we haven’t thought of. Even at the current rate of serious accidents (surely the worst case – compare the decline in the accident rates for say passenger jets), the economic costs are not significant in comparison to the value of the electricity generated by the nuclear power industry as a whole.