Yesterday EDF announced a further delay to the schedule for its troubled flagship European Pressurised Water Reactor (“EPR”) at Flamanville: fuelling is now expected to start at the end of Q2 next year, delayed from late 2022. The project costs are also forecast to increase from €12.4 billion to €12.7 billion – a far cry from the original estimate of €3.3 billion.
Among the reasons for the new delay is the need to repair various non-compliant welds, which will be fixed by the end of August rather than by the end of April, as previously expected, and a requirement from the French nuclear regulator, ASN, to provide assurances in relation to the problems at Taishan 1, which EDF says are not indicative of a design flaw despite claims to the contrary made by a whistle-blower.
“Taishan shows there are a few corrections, a few adaptations, to be made,
but in no way does it question the EPR (as a whole)”,
– Xavier Ursat, head of new nuclear projects at EDF
In the meantime there is better news at Olkiluoto which reached criticality in late December and is currently in the commissioning phase – its output can be viewed here. The plant’s power level will be gradually increased with commissioning tests carried out at every stage. Electricity production will start when a 30% power level is reached, which is expected at the end of January, and regular electricity production is expected to start in June, 20 years after the project was first approved by the Finnish parliament. Once fully operational, the plant is expected to produce around 14% of Finland’s electricity.
Despite this, I am still of the view that the EPR is a difficult and un-proven technology, and would urge the Government to consider alternatives to the EPR at Sizewell C for its next large-scale nuclear project. My preference is for an Advanced Boiling Water Reactor (“ABWR”), as originally proposed for Wylfa Newydd, a proven technology which has been delivered on time and on budget in Japan, and with much lower construction costs and timescales than the EPRs.
Although Hitachi withdrew from the Wylfa project and the site is now being considered for the Westinghouse AP 1000, the AP 1000, another pressurised water reactor, is facing similar challenges in the US to the EPRs in Europe. The Government would do better to approach other ABWR providers to gauge their interest, and be willing to pay a higher price to ensure the security of supply which is needed in the energy transition. It seems unlikely that a pressurised water reactor whether an EPR or AP 1000 could be delivered before 2035, the date at which the UK has committed to a net zero electricity system. That commitment is subject to security of supply, which is unlikely to be achieved in the absence of significant new nuclear capacity.
The Government needs to avoid past mistakes, diversify its technology risk, and opt for a more reliable alternative.
An interesting post. Perhaps our much criticised home-grown AGR programme needs to be reassessed. With hindsight their construction problems and delays look mild and their operational record good in both output and safety terms.
I agree we need significant new nuclear capacity to reach net zero- but how much do we need? Unbelievably no modelling has been done to assess this. Nuclear and intermittent renewables do not mesh well together. An optimum mix must be determined before we commit to building more capacity of any type – or risk a huge and costly overcapacity.
Replying to Richard, it is worth noting that the world has changed since we were building AGRs, so you can’t directly compare the construction delays. The over-capacity point is interesting. The CEGB ended up with over-capacity in the 70’s and 80’s because they were building to match Treasury forecasts of economic growth which turned out to be optimistic.
The EPR story is getting interesting, and I don’t believe the whole story is out there yet. The implication of the Taishan problems (derived from the whistleblower) is that fuel cladding became damaged in service because of vibration and/or fluid dynamics issues. This resulted in noble gases being released into the primary circuit. The fact that Okiluoto has just been taken critical suggests to me that they do not believe that problem applies there (unless they can envisage a relatively simple fix). The delays at Flamanville currently seem to be being attributed to non-compliant welds in the containment building (although I have also seen media reports of issues with primary pressure circuit welds). Taken together, I regretfully tend to agree with Katharine that there is a risk in proceeding rapidly with Sizewell C although, with Hinkley Point C well advanced and the technical skills of the UK and France behind it, I believe we should be able to overcome any current issues. I don’t have any real knowledge of Advanced BWRs, but if I were placing bets on a future system I’d be looking at Hualong 1. To my mind, the current western presentation of China as a bogeyman is a re-run of the way the Soviet Union was demonised in the post-war decades. China will need more electrical energy, and they have the technical capability to build these for their domestic market. I don’t see why they should not be as successful as the Westinghouse-derived PWRs that were built in France following the 1973 oil crisis. (They are, of course, also wisely trialing a wide range of other options).
I think there’s two ways to look at the EPR vs X decision of Sizewell C.
On the one hand, the existing EPR builds at Flamanville and Olkiluoto have been ridiculously troubled, partially because they started construction while many detailed design decisions had yet to be made, at Flamanville also because of workmanship problems. (Who knows, it might be so delayed that they can put a new RPV lid on before criticality rather than after a few years).
The questions around what is going on at Taishan also need to be answered.
However it does seem to be the case that each reactor started has been less troubled with more cost certainty (but also higher cost estimate) than the one before it.
So the question I have is this: is the EPR a fundamentally unbuildable design (i.e. if you built 50, the 50th would be equally troubled)?
Is this a complex design, built by a company that hasn’t built a reactor in many decades, in many cases without a mature nuclear supply chain, which they are only just learning how to build? The builders seem to think this (but they would, wouldn’t they?) in their statements on productivity improvements at HPC unit two.
If it’s the latter, there could be a case for sticking with the EPR as we’ve now effectively spent-down the nastiest bits of the learning curve and it would be a shame to give up just as we’re getting the hang of it!
I would love to see a detailed analysis of the real performance of the ABWR as well, I think that a few bad startup years has depressed the uptime numbers and it doesn’t help that many units had to shutdown after Fukushima just as they were hitting their stride but would be great to read an independent view of the technology.
I agree with this, but am also mindful of research done in the US which showed that the argument that building subsequent reactors of the same design will be at lower cost tend not to be borne out in practice since site-specific issues tend to require significant adjustments to the design. While I sympathise with EDF that regulatory changes since Fukushima have made the engineering challenges higher – and arguably the regulations are overly conservative – there is really no excuse for the workmanship and product control issues (I think they were given 7 years to replace the RPV lid, so it may well have to be done before firing up). The Fnnish and French regulators clearly have a different view on Taishan.
There were issues with some of the early ABWRs with Hitachi having to replace components, but as you say, they were all shut down after Fukushima and so far only PWRs have been allowed to re-start since the ABWRs apparently need some additional containment. That wouldn’t be an issue in the UK given the massively reduced earthquake risks.
But in the end, the build time for the EPR is around double that of the ABWRs that were built in Japan, and costs have more than trebled while the ABWRs came in on budget. Delivering projects on time and on budget suggests they were less complex and less prone to issues. I doubt Sizewell C could be built in anything close to the 39-43 months that it took to build the Japanese ABWRs, so, given the need to replace the aging AGRs, and retiring coal, I think they are a better bet.
Thank you for your very hard hitting but accurate comments regarding this dreadful Electric Power situation. I have an “ONC Heavy Electrical Certificate” from the 1950s but worked for the GPO Telephones & BT for 35 years when “0dB – “1milliWatt” is High Power but am interested in our Power Situation now.
I was at college and sat next to the fellow who found the heat problem with Berkeley Magnox in 1963 and resulted in Derating every Nuclear Station at that time.
I love looking at http://www.gridwatch.templar.co.uk daily and worry what will happen if we get a “1947 or 1963” winter now.
I notice in the last few months I have noticed huge pulse Demands about tea time and can only assume it is people charging their Electric Cars. What do you think. They are so short duration there is no way they could be allowed for if it gets bigger.
The idea of generally removing gas boilers and providing Heat Pumps on domestic houses is total rubbish – even though I have a simple Air Con installation.
[Mod: edited to correct Gridwatch URL]
All the proposed reactors types are safe but some are more inherently safe than others. All are much safer than the publics and politicians perception but to increase public support we must not cut any corners to save either time or money. There is no excuse for workmanship issues and I hope you are right about earthquake risks. To improve resiliance to type faults a mix of design types is perhaps better than selecting the ‘best’ one.
Thank you for posting my comment but unfortunately I quoted the Grid Watch Web Site incorrectly.
Could you please amend it to show: –
http://www.gridwatch.templar.co.uk It displayed the power situation of the UK and France.
The short duration power surges I noticed yesterday were in the order of 3 GWs around 5pm.
I would be interested in the views of people from the Grid side of the industry on Mr Bowen’s suggestion that the tea,-time peak is from electric cars. My experience is all from the power station side but pre-privatisation there was often details of the demand profile in the GEGB house newspaper Power News, and also in the public literature. There is little doubt that as electric vehicles become more common it will be important to manage this load away from that winter peak, but I suspect the current risk comes more from the green pressure against coal than from electric vehicles (which I have long believed to be the way of the future). It is of course profoundly ironic that the early closure of German nuclear plant leads, in the winter, to the burning of lignite.
Currently EVs will be having negligible affect on the demand curve. The peak at 5pm is the traditional winter darkness peak. The highest demands of the year. A combination of people returning home putting the lights on and cooking the evening meal. After spring clock change the darkness peak moves well away from the 5pm teatime peak which then becomes less significant. In future EV charging will become very significant and will need to be carefully managed away from the winter peaks and indeed any peak or periods of supply shortage .
Agree – this time has always seen peak demand, well before EVs were a thing. As EV penetration grows it will be important to prevent people from immediately charging their cars because this will hugely increase the evening peak. It will be interesting to see how this plays out, and I think people will need to be offered incentives to delay their charging, beyond simply making evening electricity more expensive which would harm vulerable consumers who might then elect not to cook a hot dinner.
Agree – but the charging regime needs to be discussed now and pre-planned and not allowed to play out. The new charging points will need to administer these arrangements. Again the uncomfortable choice is between market incentives (which may not be driven by supply realities) and a measure of command and control (like the fixed hours of Economy 7 – that also created its own real world problems). Underlying these choices should be the need to maintain an equitable electricity supply not one that favours the better of or the canny gameplayers.
2022-01-22 @0920 Looking on my favourite http://www.gridwatch.templar.co.uk
Noting between 1630hrs & 2330hrs on the 21st January a series of 5GW pulses.
These were accommodated by repeated power pulses from CCGT & Pumped Hydro so somebody knows what is happening. It even affected our power relationship with France.
If this was repeated in severe weather conditions at the 1745hrs peak it may crease our National power supply.
Traditionally the big teatime/evening pulses were associated with lighting, cooking, and TV programmes, typically people putting on kettles and taking a loo break (hence peaking the water demand) during commercial breaks, also at the end of Eurovision and Match of the day, etc. These must all be flattened somewhat by streaming, pause capability, and greater choice of viewing. It’s where pumped storage comes into its own, also big thermal plant that can load follow and has inertia to help with frequency control. Batteries and interconnectors are good for this too. I’m not a gambling man but as electric vehicles grow and smart meters enable demand control, perhaps with back-feeding, maybe the grid can be made to cope with more intermittency from renewables and less stability from big thermal plant. Maybe not cheap, maybe not as expensive as some make out. My background is nuclear, but I don’t see it coming back in a big way in the western world on the decade timescale. I’d love to be proved wrong.
I am at risk of straying off from Kathryn’s original subject matter and turning this into a general discussion forum.
In reply to John Bowen. The spikes he observed on the demand curve are of no significance being merely errors on the data feed to Gridwatch. They do however highlight a possible future problem with data driven automatic systems (particularly commercial innovations – such as charging batteries at low demand or cheap periods). Several passenger aircraft have crashed recently due to unconsidered data errors in ill engineered automatic flight control systems.
In reply to Steve. I hope he is wrong about Nuclear. With my background in system operation I have developed a simple but accurate model to test future scenarios (unbelievably I have not seen any modelling done elsewhere) that shows that without baseload nuclear we will not achieve net zero. Back-feeding may be useful for short term issues (like TV pickups) but for lingering high pressure zones (as the current one) it will not help unless you have at least ten fully charged cars in your drive and another one to actually use as transport.
I’m very happy to see discussions here broadening – it’s great to stimulate discussion and share ideas.
Interestingly there are also errors in Elexon data, or more specifically, missing data. I noticed this when using downloads from Gridwatch and then saw the same in BM Reports…you can see it when converting the time stamps into a time format that Excel understands. I don’t understand the underlying cause (it’s not daylight-saving related), but you’re right that it highlights the challenges with data-driven systems.
I have seen both arguments on nuclear – some people see it as finished as a technology, while others believe net zero will be impossible without it. I agree with the latter, but the wild card really is timing – if legacy nuclear closes before new nuclear comes online, we may end up having to build more gas generation to guarantee security of supply simply because it is quick and simple to build and we may just run out of time to do anything else. I have no confidence that CCS is going to play any meaningful role.
Richard: agreed, this is all getting rather general. Personally, I don’t believe we will see net zero this century, it would just be too expensive. Like you, I’d only see back-feeding from domestic batteries (initially cars, but maybe as Lithium Ion becomes cheaper, we’ll see Tesla Walls or equivalent as a useful domestic adjunct to roof-top solar) as part of short term demand management. We are always going to need something other than storage and renewables for lingering highs. Software issues are nothing new, before the 737 max there were problems in grid and share trading systems. But they are likely to get better with experience.
I also think net zero is highly unlikely in our lifetimes…it will be too expensive and the technology to support it doesn’t really exist. Lithium is unlikely to become cheaper since it is farily scarce, and it’s actually not a very good technology: it’s dirty/polluting, and short durations make it actually quite a poor fit for electricity system applications. Even for EVs it’s not great because the batteries degrade too quickly, but car makers are moving away from repurposing EV batteries for static storage, so it looks like materials recycling is the only option, and it’s not very efficient.
The other thing people overlook is that it’s all very well talking about how EVs and heat pumps will provide DSR/flexibility, but a lot of properties will need their grid connections to be upgraded to have both (we couldn’t have either on our existing connection). If you have to do a whole load of grid reinforecement in order to connect assets that will avoid the need for grid reinforcement I think there’s a breakdown in the logic…
I’m more relaxed about lithium, but then I was steered into nuclear power in the 60’s by the argument that oil was going to run out in the 1980’s. I think the current lithium argument is flawed for the same reason, as are the arguments about cobalt, tantalum, or whatever for mobile phones. You point out that lithium batteries have limited life, but the lithium does not disappear! In spite of WEE, we are not terribly good at recycling small electronic devices, but once there is, say, £5 worth of rare metals in a mobile phone they will be easy to recycle. Equally, I am not so worried about grid reinforcement. Heat pumps only really start to make real sense in new build with underfloor heating, Scandanavian levels of thermal insulation and heat-recovery ventilation. So it will take a while to penetrate. I expect to see 100A domestic connections become the norm, and that should leave plenty of capacity for EVs and electric heating without changing the grid infrastructure much. As Buckminster Fuller pointed out long ago, technology means that we do more with less, LED lighting being a good example. I see it as being evolutionary rather than revolutionary change; demand shifting from heavy industrial sites to domestic (as it already has). Going back to net zero, I don’t see CCS making big inroads. It was always going to be expensive. Perhaps it could make some sense operated together with enhanced oil or gas recovery.
Thank you for your explanation.
We should be deadly serious about reaching net zero by 2050 (wishful thinking puts this in my lifetime but certainly in my childrens). This means all out development of all currently proven technologies (are EVs proven?). Wind and solar and battery storage all have a role in a planned and tested strategy (not a drift led by business opportunities and simple-minded faith) but so do nuclear, small scale hydro and tidal schemes. Basic modelling shows that the White Paper future of “predominately wind and solar” will be very expensive and also not reach close to net zero – so why bother. Lets not give up. Modelling shows that a properly planned and diverse generation mix can reach net zero.
I agree CCS is far from proven but fossil fuel generation will not be possible without it. We should not let preconceptions lead us astray. The move from coal to gas had nothing to do with climate change. Coal and Gas are very much in the same CO2 ballpark, in fact coal with CCS is better than unabated gas and hugely better than small scale diesel generation. Perhaps strangely coal with CCS has a future as a home grown secure supply.
Domestic connection could be 100A but would the street level distribution system be able to handle that with no time diversity.
Unfortunately modelling does not show that net zero is attainable, because the modelling ignores the real world and models a toy one in its place where anything that might be truly problematic us brushed under the carpet. It makes huge assumptions about technological developments many of which are highly dubious. It ignores all manner of physical realities. I say this having delved as deep as I can into the modelling that has been done on behalf of BEIS, CCC and National Grid, mostly by an assortment of consultants whose job is to produce the answer their clients want to hear. I have also spent a long time evaluating potential technologies and running hour by hour simulations covering many years of different conditions. I do not see parallel work by these consultancies. That’s why we get policy that lurches from one thing to the next. Now we have a panic on, and Kwarteng has just authorised doling out another £400m to the T-1 capacity market that he could have saved by not blowing up coal fired power stations. There is no joined up thinking on policy.
This is exactly true, and I think that we shouldn’t assume we need to get to net zero without properly considering whether the costs of mitigating climate change aren’t lower than the costs of trying to prevent it. We’re probably close to the amount of intermittent generation the networks can cope with absent a major increase in storage and particularly seasonal storage, but the Government is just pressing ahead with plans for more wind because that’s the easy part of the equation. But if we don’t step up with storage and/or more baseload generation, then we will lurch between periods of shortages and periods of paying to curtail renewable output that can’t be transported to where it’s needed. It will be a very expensive failure, and if we start having blackouts and/or energy costs become even more expensive then lives will be lost from cold homes.
The AR4 CFD round largely dispenses with subsidies during periods of curtailment. There is no CFD payment for any hour where prices go negative. AR4 generators will have to factor in the loss of revenue in their bid pricing. I wonder, if they do their sums, whether they will find it not possible to be viable at the administrative strike prices. As capacity increases the frequency of curtailment also rises, with volume curtailed likely to rise quadratically. That will hit the cheapest to curtail preferentially, which will now be AR4 generators.
Do they not still get curtailment payments if they are they are curtailed to manage network constraints? Renewable generators are curtailed more often because of congestion than negative pricing, although that might change in the future particularly in the summer…
Talking of “consultants whose job is to produce the answer their clients want to hear” takes me right back to the 2006 “Stern Review” which was widely welcomed in some quarters while being panned by many respected economists for picking an “unusual” discount rate.
Richard Martin, my “take” on climate change is different to yours, only time will tell which is right.
Steve – Even if some doubts remain dare we take the risk when the stakes are so high.
Ah, the precautionary principle. But are the stakes so high? The trouble with the new religion of climate change is that any sort of questioning is dismissed as apostasy. Too many empires have been built on it. Good science is based on challenge. CC is based on modelling, which is only as good as the assumptions.
Well it absolutely is a religion these days rather than a science which is a large part of the problem (burn the heretic/denier…).
But the stakes are high on both sides…those who believe in the “climate emergency” are terified that by the random date of 2050 the world will be an uninhabitable wasteland, while those on the opposite side of the debate worry that net zero policies will lay waste to the economy and create widespread energy poverty, not just in terms of affordability but also in terms of access when supplies become unreliable.
I tend to the latter argument – we’re heading down a road where energy will become expensive and unreliable, but I doubt voters will tolerate that, so we may well end up with another dash for gas in the middle of this decade when we’re scrambling to build dispatchable generation to keep the lights on.
I agree with your comments Kathryn. I feel I am watching a slow motion car crash but unable to stop it. We are heading for an expensive failure driven purely by profitable business and individual opportunities. Lets hope that the effects of climate change are not severe and can be mitigated against ( for the whole planet). To try and steer clear of this crash I have developed a simple model/game to show the public and politicians the real issues involved in moving towards a net zero electricity system. It models energy and power requirements simultaneously. Its working and assumptions are fully transparent, open to challenge and can be easily modified. All other models I have seen are opaque and model energy only.
I think the problem is that the climate will always change, and it’s difficult to strip out the degree to which human activity makes any difference. The orthodoxy, propogated by the IPCC is that climate change is 99% man-made, but I find that very unlikely to be true. The IPCC measures climate change from 1850, which was itself the end of a centuries-long climate event known as the mini ice age. I also think solar activity has more of an impact than the IPCC gives credit for. But what really makes me doubt the IPCC is its poor adherence to the scientific method…in 2014 (when it decided climate change was 99% and not 95% man-made in origin), the IPCC issued a report showing that 111/114 of its models overstated warming. That’s a massive directional bias in its models, but rather than adjusting either the models or the hypothesis, they wrote pages about why this didn’t matter. That isn’t proper science and it undermines everything else they have to say.
So while I support moves to make energy cleaner, I think we need to consider that in the round (ie lifecycle emissions and other forms of clean-ness, not just absence of CO2). And we need to be realistic (and transparent) about cost and security of supply…I suspect the reason most models are opaque is that the true outputs would never be acceptable to voters…
Kathryn, I completely agree that natural variation in climate is almost invariably ignored as being “too difficult”. While the real world is more complicated than armchair pundits like to think, it seems to me that if we had introduced an actual carbon tax (in pounds or dollars per kilogram) right at the outset, this would have gradually tipped the economics towards nuclear and gas (including fracking), while providing governments with real cash to invest in nuclear, renewables, interconnectors, and grid reinforcement. The tax rate could have started small and ramped up, to minimise impact on consumers. With reasonable consistency between different countries (as with corporation tax) this would reduce the effect of international market distortions.