The Hinkley Point C cost rise: a mixed bag

On 23 January EDF released a press release with a new estimate for Hinkley Point C (HPC), a project that has become a poster child for the anti-nuclear community to claim its a lost cause. Do they have a point? Well, it's a mixed bag.

It's a short press release, classically sandwiched in a positive beginning and ending, but with the meat in between. Those are:

There is a big delay and 2027 is no longer feasible as the delivery date. There are now three scenarios where unit 1 is brought into production in 2029, 2030 or 2031.
Cost estimates have gone up from £25 to £26 billion to between £31 and £34 billion.

Yikes.

Ok, so what does this all mean? First we're not given an explanation why there's a big delay. In 2021 EDF was still confident to say that production would start in June 2026, which would put build time, which started in December 2018, at 7.5 years, a global average. Now it'll potentially be 13 years? Will this be the final estimate? We're a bit left in the dark.

The BBC, among others, give more information:

In a letter to staff, seen by the BBC, Stuart Crooks, the managing director of Hinkley Point C, said there were 7,000 substantial design changes required by British regulations that needed to be made to the site, with 35% more steel and 25% more concrete needed than originally planned.

And

The project has also faced severe delays because of supply chains being hit during the pandemic, as well as labour shortages.

I think these two are worth the emphasis. In this light I would like to point to this excellent recent interview on Decouple where James Krellenstein makes a first presentation on what went wrong at Vogtle, another headache case. Two points keep being mentioned by James:

The AP1000 design at Vogtle wasn't actually finished when they started building. A lot of changes had to be made along the way. This added time and cost.
Many supply chain companies had no experience with nuclear build projects. This led to a situation where one part of the chain couldn't continue because the previous part had a holdup, because the part before that was inexperienced, and so on and so forth.

Sounds familiar? Yeah. As managing director of HPC, Stuart Crooks, said in an internal memo to staff: "Going first to restart the nuclear construction industry in Britain after a 20-year pause has been hard". Hard indeed Stuart, hard indeed.

So, anti-nuclear people can rejoice, as this proves that nuclear power is always going to be 'slow', right? Not quite. On unit 2 the existing supply chain and living experience already resulted in a 25% decrease in building time. So, the real lesson ought to be that you don't just build one nuclear power plant. You build many.

In addition you ideally build them in what is called 'fleet mode' where you leverage the experienced workforce to maximum effect. Ideally, you start building a reactor in, say, 2025, then another in 2026, then another in 2027, then another… and another… and another. We need many units for the energy transition after all. So, when the workforce is done on one part of the project they can move on to the next reactor doing the same thing, over and over.

Ideally, you keep doing this forever. The lifetime of an EPR reactor is 80, possibly 100 years. So, this is going to give generations of workers a secure, well paid job. Of course, the UK doesn't quite need 100 units, so this approach implies an international framework. I might write more on this at a later point.

Now, as for cost, £34 billion already sounds bad, but this is in 2015 Pounds! As per the inflation calculator of the Bank of England, that's £46 billion in today's money. So, what does that mean for the consumer or tax payer? Well…

Not much.

In 2013 the UK government agreed on a price strike with EDF on £92.50 per MWh. That's in 2012 money, so that's £130 in today's money value or 13p per kWh. This is a price agreed for 35 years. Is production cost below this, EDF has a profit, if it is higher EDF makes a loss. Now, with these increases costs are going to be 11p per kWh for EDF, so they're still making a profit, but even if it was higher it would still be EDF's problem, not for the tax/rate payer's.

But what does this hold for the future? Is Sizewell C also going to be this expensive? Hard to say, because a huge variable here is the interest rate. Joris van Dorp made the point in 2019 already that two thirds of the HPC costs had nothing to do with the building, but with compounded interests. The longer the build takes, the more interest is compounded as no electricity is sold after all.

If we can either lower the interest rate and/or speed up build time, costs automatically get much lower. If the two thirds ratio is still correct (it might be more by now, but I can't say without more data), the actual building costs are around £7.5 billion per reactor, the so-called overnight costs.

Some people with an anti-nuclear bent gloat over all this and put the figure of build costs at beyond $20,000 per kW. This is trickery with numbers. Build costs are usually taken from dividing overnight costs by kW installed. In the case of HPC this would be $4500/kW, which puts nuclear in the same ballpark as offshore wind. But people working at IRENA have a (perceived) material interest in obfuscating this fact.

Mind that this is still expensive for nuclear and is not only a result from laggy supply chains and a regulator demanding 7,000 changes to the design. The EPR itself is a design that's a compromise between French and German design philosophies. It accommodates both, resulting in a monster of a design.

Concluding, this news is a mixed bag. It's certainly a bad look, but the moving parts aren't completely in EDF's hands. But they're also set to improve. Supply chain vendors gain experience and commit to nuclear orders if more projects are done. Interest rates likewise are influenced by many factors, but an important one is confidence in this going through. If confidence grows, rates drop.

EDF has to deliver, that much is clear. But public opinion, crucially the political landscape, needs to commit to a nuclear future. We need not one more project in the form of Sizewell C, but indeed many more projects. On towards ten more Hinkley Point C's!

Do you value this free article? Please consider a onetime donation or subsribe to read all of my content. Thank you!