Ep94: Julia Pyke "The Case for GW Nuclear"

ML Before we start, if you're enjoying these conversations, please make sure that you like or subscribe to Cleaning Up, it really helps other people to find us. Cleaning Up is brought to you by Capricorn Investment Group, the Liebreich Foundation and the Gilardini Foundation. Hello, I'm Michael Liebreich, and this is Cleaning Up. My guest today is Julia Pyke. She’s the Director of Financing for Sizewell C, the next gigawatt-scale nuclear power station that EDF wants to build on the UK’s Suffolk Coast. Please join me in welcoming Julia Pyke to Cleaning Up. So, Julia, great to see you. Thanks for joining us here on Cleaning Up.

JP It's very nice to be here. Thank you, Michael.

ML Let's start by you explaining what your role is because there's Sizewell C, and there’s EDF. Exactly where do you sit within that? And what is it you do?

JP I spent most of my career in law in the City and I was the Head of Power for Herbert Smith Freehills. I was inspired, having done so many years of work to get Hinkley to financial close, by the idea that we must make use of all that work and get a second project over the line. So, I left the law and came to EDF in order to develop Sizewell. My job is to agree with the government how it's going to move forward, and to raise the considerable quantities of money that we need to hit financial close and build the power station.

ML You are now financing director for Sizewell C, which is different from the finance director who presumably doesn't exist yet? But you're the financing director?

JP That's right. There is a finance director who looks after the money that we have, but it's my job to agree with government how we're going to finance nuclear, get it on a stable footing and to then raise the money.

ML That makes you one of the great experts in the world on gigawatt nuclear, gigawatt scale nuclear? That's the big plants, the sort that that deliver most nuclear power today.

JP That's right. That's the sort of reactor for which there are designs available today, which we can get on with building.

ML Right? So perhaps just walk us through the economics because I have in the past, and quite frequently, said that the economics of those gigawatt scale plants has been tested to destruction, to the extent that we should not be building them. Because they're too darn big, they're too darn complicated. They take too long; they go over budget. And so, we should move on. Now, you presumably, given your role, would disagree, but tell us why.

JP I certainly would disagree. And what I'd say is that the economics are complicated by the comparisons that people make. What is a nuclear power station to be compared to: it's not the same as wind, it's not the same as solar, it's not doing the same job. So, the economic justification for nuclear in the UK and the economics in different countries are different. But in the UK, if you build the right amount of nuclear, gigawatt or SMR, but nuclear, then it actually makes household bills go down. The reason it makes household bills go down is because, although people look at the megawatt hour price of wind or solar, people don't generally calculate the whole cost of the system and its impact on the electricity bill. We've done modeling, which is based on the modeling the government uses, and that modeling shows that if you build the right amount of nuclear, if you build, for example, 20 gigawatts, then household bills will go down by round about £280 a year. If you go ahead and build Sizewell itself, then household bills will go down by a range - because of course, it will depend on what Sizewell will ultimately cost - but a range of say between £30 and £50 a year. It will be cheaper to have nuclear than not have nuclear, even though the nuclear power station in itself is, as you say, Michael, expensive to build.

ML That's £30 to £50 just for Sizewell, or is that including the impact of Hinkley?

JP It’s the impact of adding Sizewell.

ML Okay, I'm paraphrasing, feel free to disagree if I get it wrong. It is more expensive on a per megawatt hour basis. It doesn't compete with the wind and solar and the cheap stuff, it competes with other things that could be very, very expensive. Therefore overall, it pushes the price down, not up. Is that a good summary?

JP That's exactly right. I feel passionately that in order to have a successful nuclear industry, we must identify the right role for nuclear. It’s not the case that nuclear is competing with renewables; nuclear should be supporting the growth of renewables fulfilling a different role.

ML How does that work in practice? Because what you can't do with nuclear is have it sit there doing nothing. When the wind drops, and the sun isn't shining – what they call in Germany, the Dunkelflaute, the dark doldrums - you fire up the nuclear and it only works for that short period. That would be enormously expensive, orders of magnitude too expensive, right?

JP That would be bananas. You're quite right, that's not what we're proposing. What we're proposing is to make the nuclear effectively flexible in two ways. The first way is to extract heat for the first time from a UK nuclear power station. You'll be able to extract steam at around 270 degrees before it hits the turbines. You can extract around 400 megawatt thermal of steam from a UK EPR without significantly impacting electrical output. If you want to, you can make the safety case to extract more steam, but that would be more work to make that case. The second thing is to make sure that the power station is connected for non-grid uses of electricity. In an ideal world, where you didn't need all that electricity on the grid, from time to time, it would be working in harmony with renewables, co-powering, for example, green electrolysis for hydrogen.

ML Let me let me come back to the first point as we're going to talk about this quite a bit. You said 400 megawatt thermal of steam at 270 degrees. That strikes me as being not a byproduct steam at that temperature, that's presumably at secondary generating stage? It’s surely costing you electricity to do that

JP Well, I'm not a physicist, and I don't want to muddle the exact details of this, but I am reliably informed that at 400-megawatt thermal does not significantly impact electrical output, because of the innate efficiency or inefficiency of the processes in the power station itself.

ML  What people can't see who are listening on the podcast, and it may even be edited and cropped out on the YouTube version, is that behind me, there is a book called “Elements of Nuclear Power” by D.J. Bennett, which is to prove that I actually did study nuclear. Although I did study it, I must be completely open and transparent - my studying of nuclear finished in 1984... So, although I probably could have masqueraded as a bit of a physicist and a nuclear expert, I'm not anymore at all. So, we're pretty much on a level playing field there. It just strikes me that with 270 degrees steam, there are processes- the organic Rankine cycle, for instance - that you could put that through and extract more electricity. The fundamental point that you're making though is that it's going to be a flexible resource, in the way it's used within the energy system of the UK, right?

JP Yes, that's right. This is steam, which, broadly speaking, might otherwise go to waste. Instead, we want to use it. Likewise, there are things that you can do with the warm water which emerges at the end of one of the processes. Water emerges at temperatures which are still useful, for example, for district heating. If you to absolutely make the most use of your power station, alongside all the social benefits of building nuclear, then you can make more of it. You can make the most of it for the community and for UK consumers as a whole.

ML  There are probably people listening to this who are not quite familiar with the way nuclear is used, who will say ‘wait a minute, warm water going straight from a nuclear power station into the district heating and then circulating around every home, what could possibly go wrong?’ But there are actually secondary loops and separations. Infact, certainly in China, there's discussion about potentially building nuclear that does nothing but district heating, that never generates any electricity at all.

JP Absolutely, and in countries which have historically valued heat more highly - in Finland, Sweden and Russia - nuclear has been used for district heating for a very long time.I must stress the steam is not radioactive steam, it's steam.

ML  So, we’ve established that it can be used flexibly, and that the heat is a sort of key weapon. But nevertheless, I'm going to push you on the cost of the power because there are other ways of generating power. Ways that are dispatchable, that can be used when there's no wind and when there's no sun. There's lots of discussion around using CCS, around hydrogen, around using pumped storage,  more connections to Norway, or even building more pumped storage of our own. Of course, people talk about batteries, although the volume there is quite small. You still do need to be affordable, do you not?

JP Yes, absolutely. There's an absolute imperative to be affordable. Alongside saying that I don't think we're competing with renewables, I personally think that we need to try everything. I'm fully in favor of the development of CCUS and indeed, pumped storage, if people will accept the consequences of all of these things. However, I think you shouldn't rule out a technology which you know works. We foresee this huge growth in the need for electricity, and for low carbon electricity, and nuclear has a part to play in it, which is to be in the mix.

ML  Right... But there's a big jump between saying, well, we should rule nothing out, which sounds very collegial. and saying we should spend literally tens of billions keeping it in. We're going to have to talk about Hinkley. From 2006 a big part of your life was pushing through Hinkley. The deal was done in 2016, which was £92.50 in 2012 money per megawatt hour, for 35 years, growing with inflation. The total amount paid in real terms, not undiscounted, is nearly £100 billion. We committed to spending £100 billion. I think the number is 96 billion on electricity from Hinkley C and the National Audit Office thought that about 30 billion of that was a subsidy. That’s 3.2 gigawatts? Not that huge of a supply and a £30-billion subsidy on a £96 billion purchase? How could that possibly make sense?

JP Well, I'm not here to comment on the exact deal, but the nuclear industry needed to get going again. It was the first build of its kind for a generation. The costs of building nuclear are something which most definitely need attention. They need attention from the industry, where they do get due attention, but they also need attention from all the people around the industry: the government, the regulators, the supply chain. The whole system needs attention, and of course, the financing mechanism needed attention. That's what we've given particular attention to for Sizewell C, whilst others are looking at how to practically reduce the build costs. I think it's really important to explain to people that of the £92.50 only sort of £11 to £13 or so actually represents the cost of construction. Around £20 per megawatt hour represents the cost of operating, including fuel, and the rest of it represents the cost of money. The cost of money was made up partly of running an interest bill from the point at which work started in earnest. The power station will turn on in 2027. And that's a long time to have your money out without any interest being paid. The cost of capital reflected that the whole of the construction risk sat with EDF, the developer.

ML Do we get a sense of what that average cost of capital, the unleveraged project cost of capital was for Hinkley? Is there a figure that's out there that you can share with us?

JP I think it was all published at the time. I think it was around about 9%. Lots and lots of things are said about Hinkley. I imagine that at any moment, you're going to quote the Vincent de Rivaz 2017 turkey quote. Vincent said that Hinkley can be cooking UK turkeys by Christmas 2017. So people forever say that it is late. But Hinkley is not late other than delays due to COVID. It's a great success story. The issue with Hinkley is that it took 10 years to do all the paperwork.  If we wanted to pursue the turkey analogy. we could say that government didn't build the oven until 2016. In 2017, that turkey would have been quite raw. Now, Hinkley is a huge success story. There are 1000s of people completely dedicated to building nuclear for the first time in a generation. If you want a flavor of it, you can look at the BBC Two documentary. COVID has impacted Hinkley as it’s impacted everything so it's running around about six months behind schedule. I'd just like to pay tribute to my colleagues at Hinkley because I think the people who build net zero infrastructure are not always identified as climate heroes, but really they are. They're there day in day out all weathers building this infrastructure.

ML  You got your turkey reference in early. The reference that Julia was making there  was that Vincent de Rivaz, who was the CEO of EDF UK. In 2007, he said that electricity from Hinkley will be cooking our Christmas turkeys by 2017. I think you're right to point out that the building of Hinkley is being well-managed. But what was extraordinarily naive of Vincent was to think that there would be a decision within 10 years; if he thought that he was going to get a decision in 2006, or 2007, under the Blair-Brown regime, whose only goal in life was to kick this sort of thing into the long grass, he must have been extremely naïve. I want to come back to the costs if you don't mind. The £92.50 really is £118 in today's money. I think it's really important that we understand just how expensive Hinkley is, it's £118 per megawatt hour. For those in the audience who are not familiar, generating power from gas in normal times would be about £8. With offshore wind, we've got the Contracts for Difference at £50, also in 2012 money. We're talking about Hinkley being essentially two-and-a-half times what wholesale electricity prices have been until recently. Right now it's of course much higher because of the situation with the price squeeze coming out of COVID, and also the Russian aggression on Ukraine, which has also pushed the price of gas up. But mostly, the wholesale electricity prices are £30 to £50, and Hinkley is £118. That's just to put it all in perspective.

JP So I think it comes back to something we touched on earlier. Today, prices are more like £200. Today, if Hinkley was on, it would be saving consumers around about a billion pounds a year. Of course, I realized I've not offered you the open goal to point out to you that it's not on.

ML What I would say is that the current wholesale price of electricity, which is driven up horrendously at the moment is not going to be there for 35 years. Yes, it's an example of an insurance perhaps paying off, but it's paying off much, much less than the cost of the policy.

JP When you buy nuclear, you're buying greater energy security. That's not something which has habitually been priced in. You're buying an alternative to currently very expensive forms of storage. The reason that I am so passionate about Sizewell being flexible to fit in with the future system is that I'm very keen that nuclear is fulfilling the role you need it to fulfill. £118 is a lot more expensive than offshore wind, but as I said at the start, building Sizewell will literally bring household bills down, and that's what people ultimately really care about.

ML The Hinkley deal I've called the worst deal since Russia took the wrong side of the Alaska purchase, that was how I characterized it. One of the justifications is that you get the nuclear industry restarted, and all the benefits start to flow. That brings you to Sizewell: how much cheaper will Sizewell be than Hinkley? Are we talking about a third of the price? We talk about half the price? 20% cheaper? What are your figures telling you?

JP Well, I think it's going to be 20% cheaper or more. The reason that we can't give you a figure is not because we are attempting to be evasive, but because we're negotiating build costs with the supply chain. The capital costs of the power station absolutely must come down. However, as we've been through with the Hinkley division of cost, that's not the main driver of cost to consumers. We also have a new financing mechanism. The other components of the price, which we don't currently know, are what will be the cost of equity, which will be set competitively. So unlike Hinkley which was bilaterally negotiated with the government, the cost of capital for Sizewell will be set in competition, and therefore should reflect the lowest price that the market is willing to accept to invest.

ML So when you talk about 20% cheaper, that's the overnight cost? That's the building of the thing. Do you think there'll be further savings from the finance?

JP The cost to the consumers will reflect the cost of equity, which will be set in competition. The cost of debt will be quite a dominant feature of the actual outturn and cost to consumers.

ML You said that the build cost for Hinkley was only about would you say £12 out of the £92. Let's say £15 out of my £118, in today's price is about £15, and that will become 20% cheaper. So that's going to save you about three out of £118. And then you've got the finance costs, which maybe the 9% the cost of capital becomes, let's say, 6%, let's be really ambitious, and save a third of the cost of financing. And that should give you another let's say £20, maybe £25. So you'll be lucky to get this down below £90 per megawatt hour.

JP Well, we think it will come down below that. And if the government doesn't, doesn't consider it to be value for money. So if in the government's modeling, this does not represent value for money, then it won't go ahead. And that's completely understood.

ML  Yes, but wait a minute, because that's government modeling. And don't forget, it's government modeling that gave us Hinkley C and it gave us HS2 and all sorts of other things. I want to really push on the economics of this because we had a guest episode five of Cleaning Up Kirsty Gogan, whom I'm sure you know. She’s a tremendously articulate and, and charismatic advocate for nuclear power. She said that Sizewell C should hit a price point of £40 per megawatt hour. And I pushed back immediately and said is that really credible? I asked her and obviously the implication being I don't think that's remotely credible. So you're telling me that it's got to be lower than the £90 that I think is credible here. But surely you're going to admit that it's higher than £40?

JP Yes, I'm not at all claiming that it's going to be £40. I don't have to hand exactly what Kirsty was basing the £40 on. But no, it's not. It's not going to be £40. What I would say about the real economics is, if you go back to fundamentals, France actually has cheaper and greener electricity than we do. You look at things like the article in The Economist last year, which said that if nuclear had been built out in the way which was planned in the 80s, then we wouldn't have global warming. Yeah, how much money would that be saving on the whole? I think the other thing I'd refer to when you look at the real comparisons is that the waste product of nuclear is more expensive and difficult to deal with than the waste products of some of the ways of making electricity. However, this is planned in and priced in. And so it is very hard to make straight economic comparisons between the impact benefit and disbenefit of building nuclear with the impact and benefit of building other things. It's hard to compare on a completely straightforward basis.

ML When you say is hard, it is indeed hard. And this example of France always get trotted out. I could come back with two important points. One is that right the French nuclear fleet has never produced or, not in recent history, so little power. Of the 56 plants, nearly half of them are down as we speak, some for planned maintenance, but some for unplanned maintenance: corrosion in six of them and suspected corrosion in five or six more. This reliability of nuclear, which gives it its value, its claim to be worth more is actually built on sand. And then the other point that I could make is that for all the French green electricity and it's great statistics on grams per kilowatt hour of carbon. It’s also a major utility which is repeatedly being rescued by the government. This is about the third or fourth year, we're now talking about this project Hercules, and I realize it's probably a bit delicate for you, because this is EDF. Not your employer right now because that’s Sizewell, but...

JP My very close connection

ML Your very close connection. Exactly, your major backer. They have a market cap of €31 billion, debts of €40 billion, undiscounted liability for decommissioning of plants, which depends on what you think they will cost to decommission. Potentially somewhere around €30 or €40 billion unprovisioned. It's got a pension liability, which is unprovisioned by another maybe 20 billion. It’s got another €50 billion or so liability to upgrade the French plants, the old ones it wants to keep open, and then another liability to build new renewables, which is something else it wants to do. It's great having cheap nuclear power, but if you're not really paying the correct price for it, which is what all of that tells me, then that's not much use is it?

JP Well, I think that the financial condition of EDF is a very complicated topic on which I'm not qualified to speak. But I would point out that EDF sells nuclear electricity at below its cost as part of its deal with the French government. And making as with many things, making straight comparisons is difficult. I mean, in terms of reliability, if you if you look at what happened in the UK, the UK fleet has been very reliable until actually they didn't invest it sufficiently in maintenance in the 90s, at which point, reliability went down. As soon as actually EDF acquired the stations and there was much more invested in, in maintenance. It's had fantastic reliability and around the world Pressurized Water Reactors, which is the sort of reactor which the EPR is a variety of they have 91% availability. So on a global basis, they are very well available. Obviously, everybody needs to make sure that enough is spent on maintenance to maintain full availability as with anything, as with any way of making electricity.

ML But when you talk about reliability running and I'm happy to concede that point because I think what's going on in France right now is unique. As to the second part of my point, I totally accept that you can't really comment on EDF’s financial status. I understand that entirely. What I would say is you can't claim that nuclear is cheap, because look at France, it's cheap, but then come back and say, well, it's only cheap because it's being sold under cost. I can make lots of things cheap if I'm allowed to sell them at a loss.

JP I don’t think the selling at a loss is part of the cheapness of nuclear power. It's a particular deal, which happens inside France. But I agree, it's probably too complicated to compare the French electricity system with the UK electricity system, what we can see is that the French electricity system has very low carbon emissions and generally lower than the UK electricity system.

ML  Okay, but now let's talk about a different aspect of reliability, which is, if you are shutting other power stations, they may be coal, we've already largely shut our coal but now the UK wants to shut old nuclear, which is no longer safe to run, reaching the end of… end of life nuclear, we’ve got a number of power stations there, but also it's going to have to get rid of this gas to hit the requirements of the Climate Change Act and Net Zero 2050. For that, you have to be able to deliver your power stations on time, not just within budget. You can extol the virtues of the EPRs. But the but the French EPRs are not reliable or they run enormously over. And there's some good examples of that. You've got Olkiluoto, which is now working in Finland. 12 years behind schedule. The went from €3 billion to €11. Taishan, China, that one construction began in 2000 in nine finally started operation 2019 instead of 2013. So despite being in China, a place where you can get stuff done quickly, six years late. Budget about 160% of what was planned. And then Flamanville which surely is the flagship for this technology. I mean, if they can't do it in Flamanville, why on earth would they be able to do it anywhere else in the world? I don't know. Supposed to cost €3.3 billion for Flamanville III, now expected to cost nearly 20 billion, supposed to be commissioned in 2012. Now thought to be commissionable perhaps in 2023. I mean, it's a devastatingly bad record, isn't it?

JP Well, if I take each of those in turn, and there are good records of nuclear built around the world, and it tends to be when there's a program and when there's fleet builds, and people go from one to another to another because it's only human to learn by doing. Instead, what has happened is that in in different countries, there's been a gap in building nuclear. And after a long time, a unitary single project has started up and it hasn't gone well. But if we look at Olkiluoto, for example, it's on, it can provide 14% of Finland's electricity reduces their dependence on Russia. And it's branded as being Finland's greatest climate act. It's obviously had a late delivery and cost too much, but who now wishes it wasn't built? I think in lots of ways, it's analogous to a lot of UK, big infrastructure where there's lots of criticism of things being late and over budget. And that's not to say that they shouldn't be both on time and within budget. They should and everyone should make every effort. If think about HS1, or you think about the current press coverage of Crossrail as an engineering marvel in comparison with the coverage two years ago, you think historically about the 1858 sewerage system or the London Underground? No one can imagine life without these things. And so absolutely, we must make every effort to get cost down to get these things in on budget and to get them build on time. But they're very long-term projects and how people see that varies over time. In Finland, there's a recent poll, which has 74% of the population in favor of nuclear, and only 18% oppose, and even the Green Party in Finland is pro nuclear. So not trying to dismiss your point about Olkiluoto being late and over budget. It was both, but it's also a great thing. Taishan, they obviously had an unrealistic time estimate. It took them about 10 years. That's probably how long it should have taken them. The mistake was ever estimating four, the mistake wasn't the 10 that it took. It's got a problem with the design of the fuel rods is a relatively contained problem. It's not uncommon to have problems with fuel when reactors start up, the problem will be fixed and we should all be delighted that China a second successful nuclear program, because the alternative is in an industrialized country, they're going to burn even more coal or gas unless they have a successful nuclear program. And Taishan, broadly speaking is a success. It's got a problem at startup with its fuel, it will be corrected, it will be a great thing for 60 years. Flamanville has had a lot of problems. And far from being the flagship, it was the prototype, it was the first one to be designed. I think the way we can look at Flamanville in the UK is that we are lucky that we can learn from it, we can avoid its errors, Hinkley has already shown that it can avoid a big proportion of the errors Flamanville made. If you look at the errors in Flamanville you can see that Hinkley has advanced beyond the point at which a lot of these errors had already occurred. Ultimately, it will be fixed, and it will be a good thing. But the main point for us is that Hinkley is going well. It's benefiting from all this learning around the other EPRs, and Sizewell will benefit from Hinkley. And you can see Hinkley between unit one and unit two, the level of learning and the productivity improvement. I'd say again, humans learn through doing and this sort of stop/start approach to doing one nuclear project per generation is part of what leads us to the position which we find ourselves in in which it's slow and takes too long and costs too much money.

ML Your argument for learning is a kind of double-edged sword because how you've got Sizewell C is going to be your fifth project. Right, the fifth project for broadly the same technology. Hinkley essentially is on track and on budget only because the track was allowed to be the full 10-12 years and the budget was so enormously royal that you could hardly go over budget because it was such a lush, contract for difference, right? What you're really saying is that we're this technology has proven to be really bad. It’s kind of like saying ‘we're really crap at projects, so please give us more projects, so we could get less crap.’ But it's not a very compelling argument, is it?

JP Well, obviously, Michael, I would not put it like that. I would say that nuclear power stations are great things, they can be great additions to you to a country's energy infrastructure. That's something that people don't understand about the and there's no reason why they should, or the nuclear industry has not been the best at communicating. But nuclear is quite heavily nationalized. Each of Taishan, Olkiluoto, Flamanville is first of a kind in its own country. If you use the analogy of a car, the engine, the basic technology, across all PWRs is quite similar. In the EPR, that's quite similar. But if you use analogy of the body of the car, the UK wants the steering wheel on a different side from France, it wants slightly different braking systems. So, you end up nationalizing the design and each of those countries. The first time you will see what the learning impact is is on Hinkley Unit Two, and the next time you will see what's actually the learning impact, it will be Sizewell which we hope will be treated in learning terms as Hinkley Unit Three. And I think if we can't demonstrate learning at Sizewell your comments could be proved correct, but I'm very confident that we will.

ML Okay, so a few more tens of billions. And we should know whether you are going to learn enough to justify the whole thing, on top of the billions on Hinkley. But I'm smiling, as I say this, for those who are following on the podcast, they may not see, there's a good bit of banter around this as well, because I am in favor of giving technologies a chance to prove their place. It just is this one is so expensive, upfront. But I do feel on the economics, I want to come back to the regulated asset base model, because the learning that you've talked about is the physical learning, it's the thing that will turn Hinkley's £13 or whatever it was for the build to £9 or something like that. Yeah, but the financing costs the operating costs, then presumably there'll also be learning but a lot of that is fuel

JP That’s right. There'll be similar.

ML There'll be similar. Okay, so there'll be similar, but the regulated asset base model. Let's spend a little time on that. I think we've got certainly in the audience some people are very financial, they'll know all about it, and then there's some that will never get it, because that's not what they do. But there's a big chunk in the middle who probably kind of oh, that sounds interesting. So, could you explain what is the regulated asset base approach particularly has it's going to apply to Sizewell C?

JP Yes. So the regulated asset-based approach was devised for infrastructure in the UK, in order to privatize it, essentially. It applies to the National Grid, it applies to lots of distribution, electricity distribution networks, it's applied to the water industry, it’s applied to airports in particular. What was achieved in the water industry is that something which had always applied to a water network was applied to a specific project, being the Thames Tideway Tunnel. So that's what we're largely looking to replicate for Sizewell. So, construction, in the construction period, investors, equity and debt investors will receive a return on capital, a certain amount regulated, set in advance. That stops interest building up in the way that interest builds up and up and up. For Hinkley it rolls up and up and up, like a credit card, that part of the cost is taken out. That's a big chunk in why the cost are down. And then once it's operating, then it's paid it's CAPEX, and it's paid. It's OPEX, as we generate, but the formula will not be the Hinkley formula, which incentivizes the maximizing of generation. You're paid by reference to how many megawatt hours you've put onto the grid, it's going to pay Sizewell to be available, so that Sizewell can fulfill the role over the decades in which it will be a useful asset, which is most useful to consumers from time to time.

ML  Okay, so broadly speaking, and let me see if I've understood it, there's two phases, there's the construction phase, where you've got no revenues as the project. If it's perceived as risky, and therefore the cost of capital is high, then you get this ballooning of the financing amount and therefore cost, which then has to be paid off through the life of the project. So that is kept down by the government underwriting everything. And it's a return, but it's not some super normal return.

JP Yeah. Yeah, I have to I have to stress again, the government doesn't underwrite everything, the equity remains heavily incentivized to control construction costs. So, construction cost overruns are shared between equity and consumers. And that's, that's essentially what happens in all regulated asset bases.

ML  Okay, but this pretense that it's a sort of fully commercial asset with only the private sector taking risk that's gone in exchange for lower financing cost during that period. Then there's the operating period, when you are producing either electricity or some other benefit, even if it's just waiting to produce the heat or hydrogen or whatever. But you're ready to go and produce electricity to keep the lights on. And there's a stream of revenues of various sorts. And that is the second phase those will be covered by some kind of contract for difference, but it will allow Sizewell C to be flexible to switch between the different uses of its output.

JP Yes, that's right. And as for other regulated assets, the return for equity will be set by OFGEM periodically. So every five years OFGEM will set the return for Sizewell investors in the same way the OFGEM sets the return for investors in grid and distribution networks.

ML  Okay. Well, I think that that point about flexibility is very important, because the way that Hinkley works is it is paid to produce electricity. And it wants to do that whether that electricity is needed or not needed. And that's similar to offshore wind, and we've seen offshore wind or in fact renewables overall being paid not to produce, they're not paid to go off and produce hydrogen or to do something else. They're just paid to produce electricity. So that's what they demand being allowed to do, or to be paid not to do, and this will be different, right?

JP Exactly so. I think there's the problem of lumpy electricity production and constraint payments is one which I know that people in government and OFGEM are looking at very closely. What we want to do is to make the most of the opportunities to maximize the use of the electricity and not be paying people to produce electricity that can't be used.

ML Right, now, one of the things if we go back to your statement about this is going to save people money. You have to have a kind of counterfactual for the other ways to do everything that you want Sizewell C to do and those things. There are a number of ways that you could fulfill that demand. You could use pumped storage and connections to Norway, you could overbuild. I mean If the levelized cost for Sizewell C is £90, you could build nearly twice as much offshore wind, and just hope that some of it somewhere is working when you need it. In economic terms, not in market regulation terms. I have very small investment in something called Xlinks, which is building 10.6 gigawatts of wind and solar in North Africa and some batteries, and it will have 90% dispatchability by bringing it around the coast, never lands in Europe comes around the coast into the UK via high voltage DC under, under, under the ocean under the seafloor, on the seafloor cable. You can do energy efficiency, we could save however many gigawatts Sizewell C is going to be we could take that money and put it into reliably reducing demand. So, there's lots of things we could do that I'm guessing would cost, let's call it £50, £60 per megawatt hour. And Sizewell C is going to cost we sort of came out at somewhere around below 90, but let's call it maybe £80 to £90 per megawatt hour. And then are you going to say that we should still do it because we'll learn so much that we will then be able to use that in future gigawatt scale nuclear?

JP I'm saying I'm saying that you must take account everything that you're buying. You have to accord a value to energy security, of having something which can be made almost entirely once it's operational onshore. We are very interested in using the existing stocks of uranium already in the UK and re-enriching so that we're not causing new uranium mining. And energy security has a value, which I think historically hasn't really been taken into account. We also contribute an awful lot to social value. I'm not suggesting that creating better lives for people in remote coastal communities is in itself a justification of a choice of electricity technology. But it is an absolutely huge benefit of nuclear power stations, as are all of the skills and training that nuclear gives rise to. One of the huge skill shortages in the UK is welding. We've historically imported our welders. You can bring it back to the fact that further education colleges haven't been able to afford to run welding courses. There's just been a dearth of welders. And through nuclear, because we need a lot of UK welders, we have welding academies. Our trained welders who will be able to weld to very high spec are then available to weld for anything which needs welding to a very high spec. These things don't really get valued in things like government business cases, but they do have a value. Nuclear pays for this in its price. The whole cost of decommissioning and waste management, tends to be treated as a negative in discussions of nuclear. You could look at them as a positive, the whole lifecycle is priced in. And the other thing I'd say about Sizewell C is it will have absolutely exceptionally low lifecycle carbon emissions. We've had an independent study done by Ricardo and verified by WSP. It's really granular, it looks at the content of everything that we're using, and it comes out at 5.5 grams per kilowatt hour, that's around about half the level attributed even to wind by the IPCC. So, for people who are looking at how to lower the carbon content of their portfolio, an investment in Sizewell C will be very helpful. So, all of these things I realize are softer justifications than then the straight ‘well is it the cheapest thing we can do?’ It depends on what you put into your column of what you're needing to provide.

ML  Right, and I buy that I buy, I buy the story about skills. I buy the welding story. I've been up in Barrow and Furness, looking at some of the welding going on on submarines and being good at welding being really, really good at welding. It's not something that sounds very glamorous in Islington or Notting Hill Gate, but it's really important for a country. I buy the low carbon and I buy the decommissioning costs if they have indeed been properly accounted for. I buy the thing which you didn't mention which is that nuclear is not just about nuclear power stations, nuclear technology plays a role in health. And it plays a role in the food industry. And it plays a role in how you measure things. There are all sorts of little bits of the economy, which rely on engineers knowing how to handle and produce nuclear isotopes and so on. So, I think there's a lot of co-benefits. On jobs more generally, that does make me smile, because there was this fantastic tweet Hinkley C tweeted in May 2019. Somerset Larder has got this huge contract to do catering for the Hinkley building site. And they said ‘did you know we sell 7,800 sausages every day to feed Hinkley C workers, and we deliver 316,000 kilos of beans every year’. I calculated using the National Audit officers estimate of a £30 billion subsidy. I worked out that if you took those 7,800 Sausages per day for 10 years during the building of Hinkley C, it worked out as a subsidy of £1,430 per sausage.

JP Only if you were only getting the benefit of sausages from this machine, rather than all of the benefits it will bring. If we talk about Somerset Larder, it is actually a great example of the sort of thing a big project can do, not just a nuclear project. But some of that larger is a collection of small Somerset businesses who aggregated in order to cater for the site. One of the things I do feel really strongly about is if you're going to build something like Sizewell, because you want the electricity and the heat and you need it in your system and it’s affordable and offers value for money, you definitely should do it in a way which brings the most benefit. So if you're going to do very large-scale catering, make sure you're doing very large-scale catering in a way which really works for Suffolk, which reduces food miles which gives people opportunities. It's not really money; it requires a lot of thought about how you put your project together.

ML  Obviously I'm making a humorous point by ascribing the entire subsidy cost for Hinkley to the sausages. By the way, I'm sure they are absolutely fantastic sausages and if anybody from Somerset Larder is listening, I in no way impugn the quality of your sausages, I would love to tuck into one at some time…

JP Even though you're talking about it, I don't know if you recall an episode of Bagpuss where they're producing chocolate biscuits, and the mice are just taking the biscuit and running around the back of the machine pretending to do something and then coming out with the same biscuit. That's obviously not what Hinkley is doing, it is not a sausage machine, it is making electricity.

ML No, but the serious point is that those sorts of subsidies figures that are being thrown at Hinkley, potentially at Sizewell C as well, that's a really inefficient and expensive way to provide an economic stimulus. So, I would sort of push back on the generic jobs argument. The specific skills argument I buy, and I more than buy, but the generic ‘oh, well, it doesn't really matter because we'll just overcharge people for electricity and put it into local businesses in Somerset.’

JP I completely agree with you. If Sizewell is given the go ahead it's because it is considered to be the best thing to do for the UK in order to buy what Sizewell brings, which other ways of delivering electricity don't bring.

ML  I want to give you a chance to deal with two other things that are I suspect, often thrown at you. I suspect you'll have brilliant answers for them. One is, oh, nuclear shouldn't be built because it can't get commercial insurance. So, let's talk about how nuclear is insured and which bits can't be insured. And then the other thing is, what about radioactive waste? Because we don't have a solution for radioactive waste. Perhaps this whole discussion we shouldn't have been having at all because there's no solution for long term storage of radioactive waste. So, if you can, rapid-fire, insurance and waste.

JP So insurance: Sizewell is commercially insured, including for nuclear third party or will be commercially insured, including for nuclear third-party liability. There is international legislation which has extended the ambit of nuclear third-party liability into a couple of categories where it's possible there won't be commercial insurance in which case the government will provide, be the insurer of last resort as it is for example, for Tideway Tunnel, if Tideway Tunnel can’t obtain a particular form of insurance, but the vast majority of Sizewell’s liabilities will be commercially insured, on the markets. Radioactive waste, everything produces waste. And nuclear deals with its waste. So, I use my Coke can analogy, if all your electricity for all of your life came from nuclear, you would use a Coke can of uranium. At the end of your life, the waste resulting from your whole life's use of nuclear electricity would fit back in that Coke can. With the radioactive waste from a nuclear power station, we encase it in concrete. It sits completely safely, you can go and touch it, if you wanted to, you could hug it. It sits in a Sizewell dry store, it's volumetrically small, and it can stay there. I think there's a safety case for 100 years. If it is needed to stay there for another 100 years, you could re-concrete it. It's government policy to build a geological disposal facility and they've built a geological disposal facility in Finland at Olkiluoto. They've got the consenting to build facilities in Sweden and in France. If ultimately, the government doesn't build a geological disposal facility, the waste is safe, it sits cooling safely. It's not subject to terrorism. It's not the Simpsons, it's not sitting in green ponds of sludge minded by Homer. It's sitting encased safely in concrete.

ML  Does it not start off in a pool of water, which can leak or boil over all the rest it doesn't it spend some time in a pool first. Then it goes into this concrete?

JP Yes, yes, it spends about 18 months in the in the, in the fuel pond in the reactor. But But it's, as you would imagine, Mitel designed not to leak to an extremely high standard of safety and continually checked and monitored, and then it and then it goes into the Dry Store. And although there's been obviously a big cost issue associated with the decommissioning of historic nuclear in the UK, new nuclear power stations designed to be decommissioned, is understood how to decommission them. And you can actually D fuel an EPR very quickly in around about 24 hours. Whereas if you look at something like the existing advanced gas cooled reactors, the ones that have reached the end of their lives now, that sort of three to five years to D fuel and if you looked at the magnet stations, they spend a long time cooling down. You can see that that is actually great progress in nuclear design.

ML  Right, because we have in this country, a liability for decommissioning of past nuclear, which is somewhere if you add the nuclear decommissioning agency together with the fund that EDF co-managers for those Magnox reactors, and started the advanced gas-cooled reactors, rather, I think it is that you add that together you get somewhere between £150 billion real terms undiscovered, discounted, not undiscounted, £100 and £50 billion and it depends what discount rate you use. It could be as high. They used to think it was £250 or £270 billion. And then they decided to use a different discount rate because the number was too scary. But what about half of that? I think is Sizewell is the old Sellafield site, which is an absolute mess right.

Sellafield? Yeah, I mean, yes, Sellafield was used for all sorts of experimentation for both military and civil purposes through the 1940s and 1950s. What was done wasn't adequately documented. It's a huge task to clean it up. I'm in no way trying to minimize the need or cost of dealing with Sellafield. It's not analogous in any way to whether or not we understand how to decommission and the cost of decommissioning, and the cost of storing and disposing of the waste from a new nuclear power station. It's a historic problem, which, as you say, is an expensive historic problem,

ML  Right. And of course, Sellafield used to be wind scale until there was the wind scale fire. So then he got rebranded Sellafield, so a little bit of reputation management there, but it leaves the financial liability on the question of the waste. In preparation for this I rewatched, a tremendous sort of 15- or 20-minute segment of a Dutch comedy show, there's a comedian called Iron Luba, he did a section on nuclear power. It's hilariously funny, even translated into English, and I will put it in the show notes because everybody should watch it. He shows these two warehouses, one with mid-level, radioactive waste and the other with high level. And he says, ‘that's it’, so when we talk about the terrible problem for future generations, it's literally ‘don't go in that warehouse, the rest you can do.’

JP They use that warehouse to store art, because it's climatically controlled to the right level. The Dutch have been incredibly clever around the way they've communicated this. They paint they, the way they paint the outside of the building shows the cooling. You've studied nuclear ,you do understand this. These concrete canisters are inert, you can touch them. I think what people imagined about nuclear waste, they would find this very, very calming. And there are other technologies in development. I'm non-executive director of a small nuclear company called Nucleo. And there are other designs, like GE has a design called PRISM. They're designed to burn some of the waste, and that would in itself reduce the waste stocks in the UK. I think when we have to remember that what what's in our atmosphere is all sorts of other awful products of the 40s 50s 60s 70s 80s, which is all of this carbon. The cost of dealing with Sellafield is high, the cost of dealing with the waste from a new nuclear power station is actually not very high. But then we have to always compare it with the cost of making electricity as we have historically made it.

ML  You segued there actually to the last question I wanted to ask you, which is about small modular reactors, and you talked about one which consumes previously wasted nuclear materials. But there are something like 50 or 60 different designs. Is this an interim rearguard action that you're fighting on behalf of giga-scale nuclear? Because really, in your heart of hearts, you know that the economics have been tested to destruction? That really to get the learning, we need to get the smaller models, which will then encapsulate learning similar perhaps to offshore wind or even solar, all the other things that we build repeatedly on a one-year cycle or a two-year cycle, rather than maybe at best a 10-year cycle?

JP I absolutely think there's a future for both in the UK. There are not many sites suitable for gigawatts. So, I don't think there'll be more than another two or three-gigawatt power stations after Sizewell, just because of physical constraints. I think there's a great future for small modular reactors. I think they'll end up doing different things. I think that some designs like there's a Urenco <inaudible>, which is designed really for industrial process. There's, as you said, the nuclear design, which is designed to burn waste. There's Rolls Royce’s small modular reactors. I think that the small designs will find their own role in the system. In some cases, it may well not be providing electricity, they may be for heat. Because, rightly, it's a very thorough process to license a nuclear design in the UK. And then to find a way of working with the regulator to make sure that you're building with the appropriate level of quality assurance from all of your pieces of kit, it's not quick. So, if we wanted to wait for some of these small designs to be licensed and provably buildable, then I think we'd be waiting a long time. While we were waiting, we'll be burning gas. I think that we need to gigawatt to go ahead on the few sites that there are, and there's a bright future for small modular reactors.

ML  Well, thank you very, very much, it is always an absolute delight talking to you. Nuclear is such an emotional subject for so many people, both the people who support it, even as fare as thinking that they'll get it built by attacking wind and solar. But it's also emotional for the people who attack it on the other side. I've been involved in debates where former members of parliament in Switzerland have claimed that Chernobyl killed a million people, for no other reason with no basis, in fact, but simply to get so emotionalized in the discussion. it's tremendous to come back and actually talk to somebody who is coldly rational and working on the nuclear build. It's a real relief.

JP Well, thank you. As you can see, my sign says we can do it. That's an old Westinghouse sign. And I do think I would say I think it's sad that so many people, particularly in environmental NGOs, haven't come to terms with nuclear, because if we think about what we could do, if we all work together, it would be tremendous. It's incumbent on us the nuclear industry to better communicate; your examples of the Dutch are great. There should be much more of that. I would appeal to people from Greenpeace and the World Wildlife Fund for Nature, I would appeal to them to look again, in the face of the climate crisis, at what's the real detriment of nuclear. Could we all work together and make this a better place?

ML So Julia, perhaps you should invite them all, and I'll come along as well to Sizewell to be reassured. Perhaps even those ultra-environmentalists will be left chanting ‘We can do it. Yes, we can.’

JP That is a very good note to end it on. Do please come to Sizewell and see the ways we can do it.

ML  Julia, thank you very, very much. That was Julia Pyke, Director of Financing for Sizewell C the gigawatt-scale nuclear power station that EDF wants to build on the UK’s Suffolk coast. Cleaning Up is brought to you by Capricorn Investment Group, the Liebreich Foundation and the Gilardini Foundation.