GB Grid Constraints and Renewables Roll-out | Louis Stansfield, AFRY

Hi there, hopefully you can hear me, my name is Louis Stansfield, I'm here from A3 Management Consulting, I have a couple of slides on A3 which I'll go through quite quickly in the interest of time, A3 is a big company, we're part of Management Consulting which is a smaller subsection of that, but we still cover many many countries and many continents and there's about 700 of us across the globe, so we're quite a big company, please come and ask us about any of this that you want to come and find a bit more out about. Specifically within that, specifically within that, specifically within that, and the area that I'm sort of primarily involved in is around market reports and market projections of power markets, so I'm particularly responsible for what we call our GB Flex Deliverable, which is looking at sort of the closer to real-time dispatch parts of the GB market, but we have a wealth of experience across many markets across the globe. I'll leave it for that on the introduction on A3, you'll get more on it when it gets sent out, but I'll get straight into the detail if that's okay. So, the topic, and it's been touched upon a little bit already, which is fantastic, is grids versus renewable build-out, and I guess the question that we're kind of asking is, can grids build-out keep at pace with renewables? And I think there have already been a couple of views on the answers to this, and I don't think I disagree with anything that's been said, but let's dig into it in a bit more detail. Grid constraint costs have become a material part of energy policy discussion and general views on affordability of energy. I've pulled out some relevant but sort of a scatter selection of comments that have been made over the last sort of year around the points of balancing the grid, the points of the way that energy is used, the changes that might need to be happening to the way that renewables are subsidised and the reconciliation for them. But it's a very hot topic. It gets discussed a lot. It has already come up several times today. But it is definitely a very relevant conversation to be having, and in fact, since I was asked to prepare for this session, there have been material announcements in the last week around the changes to renewable subsidies, particularly auction round nine, which is the... We talked about auction round eight happening next week, or the budget for that being announced next week, but that would be the next one. I'll come on to this later, but there's the potential for a pause and potentially fairly quite significant changes to the way that contracts for differences are allocated. But it is, yeah, a very big topic of conversation at the moment and something that's being talked about quite a bit. And over the course of what I'm going to try and talk about in the next 10 minutes or so, I'll explain some of the drivers of this and explain some of what's happening. To start off with, it's worth pointing out that a lot of the grid constraint costs that we see and are going to talk about are driven by renewables and, in particular, wind. And it was touched on by Ben in the last section that there's a lot of renewables in Scotland, and there isn't necessarily always the demand in Scotland, and that creates some issues. So, here on this slide on the left, I've got sort of a map with capacity size scaled of offshore wind in England and Wales and Scotland. And you can see that it's naturally on the coastlines around the country, but with quite a lot of it in the North Sea, and that exists across both England and Scotland. In this case, it's not so much in Scotland, but there is still a significant capacity in Scotland, about 4 gigawatts currently. And the locations of this are defined by seabed lease sites, so it's not an entirely sort of, it's not something individual parties have entire control of, over where these things will happen. But they are distributed across particularly the east side of the country. And then within that, the drivers are things like wind speed to NOOS, which is, you know, a locational cost that's relevant for these projects that are coming ahead. And then things like distance to shore, some of these projects are easier or less easy to build. But generally speaking, contracts for difference are the auctions of those, the driver of where this wind ends up and which projects end up being successful. For onshore, much more distributed, many more of them, they are smaller in size all over the countries. But it is worth pointing, and maybe this graph doesn't, the map doesn't give it full detail, they are quite heavily clustered into Scotland. So about two-thirds, something like that, of the capacity, onshore wind capacity in GB is located in Scotland. That is a combination of factors. Typically the wind speeds are higher in Scotland, but most importantly, probably, until fairly recently, there was a de facto ban on onshore wind in England and Wales, which meant that onshore wind projects were, by definition, going to be built in Scotland. That is now lifted, and we are now seeing that a combination of merchant economics and contracts for differences are the drivers of these capacities being built going forwards. This leads to a structural problem, which is that the GB transmission network, and we often talk about it as a day-ahead price, one homogenous thing. The truth of the matter is that it isn't, and a lot of the electricity is produced in parts of the country where the demand is not, and the electricity needs to flow there to basically achieve its objective. Structurally speaking, there are what I would call key transmission boundaries, transmission bottlenecks in the system, and these historically are linked to the sort of the evolution of the GB power network. Until 2005, there wasn't a homogenous GB power market at all. So it's actually a relatively recent thing, and we are still seeing some of the hangovers of that, which is specifically that the boundaries between Scotland and England, or the boundary between Scotland and England and the boundaries in Scotland as well, are not currently capable of resolving the imbalance between electricity being produced in Scotland and the majority of the land being in England. And so on the left, I pulled out some of the kind of key boundary constraints. In AFRA's modelling, we model around these constraints. But then also for the key ones, which is really B2, B4, B6, and B7a, the top four most of these constrained boundaries, is how much of the time in hours per year that they were constrained over the last three years, including year-to-date this year. And you can see that they vary a bit, and this is a combination of different projects coming online, different levels of reinforcement. It is, to some extent, a dynamic equilibrium, if not quite an equilibrium. But what you can see is that the amount of time that they're constrained is really quite significant. And in fact, in 2026, year-to-date, the two uppermost boundaries, B2 and B4 in Scotland, have been constrained basically half the time. So, you know, one in every two days or one in every two hours, the renewables that are being produced in Scotland are not able to flow down to where they're needed, which is a very, you know, significant number. What this means is that this combination of the distribution of renewables and where they're generating and the transmission boundary constraints of where the energy can flow, the electricity can flow, mean that NISO, in the balancing mechanism, needs to take action to effectively resolve what are called these thermal constraints, literally stopping the wires getting too warm, by limiting the amount of electricity that can flow over them. So they take action. And the way they take action most of the time, and you can see this on this graph, is to turn off, turn down a bid action in the balancing mechanism renewables in Scotland and then generate that electricity somewhere else to keep the system balanced because the electricity was still needed. It was still bought, but it just can't be produced where it was being produced. And you'll have to take my word for it, but this graph is all of the zones that I was modelling based on those transmission constraints on the previous slide with the volumes by year over the last seven years. And you can see that the colour of those graphs is very blue-dominated, which is the top three of those zones, which is the Scottish zones, and you can see that nearly all of that action is being taken in the balancing mechanism to turn off wind in Scotland because that's where the constraints exist. There are some other constraints, and you can see at other times those actions need to be taken elsewhere for a variety of reasons, but the majority of that is in Scotland. And you can also see that if you were to draw a trend line through that graph, the lines are increasing, but significantly in 2024 and 2025, those volumes stepped up, and that was a result of specific large onshore and offshore projects coming in line in Scotland, such as Seagreen and Viking onshore wind farm. And effectively, the issue has become bigger over the last few years, and more of these constraints need to be managed. It is worth pointing out, B7a is also quite often constrained, I'll just go back quickly, which is the fourth lowest on these three maps that I've shown, but because there actually isn't a lot of transmission generation between that zone and Scotland, most of the actions need taken in Scotland. So the constraints aren't always in Scotland, but often the action is taken in Scotland to resolve it. This leads to some very interesting activity, of which this will either shock you, or you'll be very aware of it, depending on what your interest is and where your crossover is. But this is an example of a wind farm, it's an offshore wind farm in Scotland. It is a specific one, but I haven't named it, because actually it's representative of many of the large projects in this area, so it doesn't seem fair to pick on one particular project, but it is kind of representative of what we're seeing for offshore and even some onshore wind farms. I'll talk you through the graphs quickly. What we have is the same graph for two different years. The bars are the load factor by month of this offshore wind farm, and you can see, as is broadly typical for a Scottish offshore wind farm, that the load factor is somewhere between, well, it varies quite a bit, but, you know, an average of 50%, 60% load factor. That's the potential load factor. That's what the asset would be generating if left available in the day-ahead market to produce what it could produce. And then the grey bars are what it actually ends up producing. And then just interlayed over the top of that, because it's kind of interesting, is the wind speed on average for that month, which obviously there's a link between the wind speed and the amount of renewable generation. What you can see is that the potential load factor is materially higher than the actual load factor, and, in fact, in some months, take August in 2024, astronomically different. The wind farm is basically not being used, even though it has the potential to produce a lot of energy. And this is basically because the transmission constraints can't hold the cumulative amount of renewables that are being generated in Scotland, and therefore something needs to be taken as an action, and, in this case, and is often the case, NISO are turning down a subsidised renewable in Scotland to manage that situation. So, on the right, I've just got the tables for comparison. I just think they're quite eye-catching. You know, the average annual potential load factor across those two years was pretty similar, 51%, 49%, but the actual load factor of the assets in terms of metered generation that they produced was 14% and 17%. So, subsidised wind being built, not actually being used anywhere near its full potential. It's also, I think, interesting to spot that in some of the months where the drop-off is biggest, it tends to correlate to the highest wind speed. So, when the wind speeds are average, some curtailment needs to happen, but when the wind speeds are particularly high, large projects often need to be curtailed the most. The driver of this, and the look forward to this, is that we are in a period of significant renewable build-out, and we will continue to be in a significant period of renewable build-out. There is still a very strong decarbonisation policy in GB. There is still a strong appetite for renewables, and there's a long pipeline of projects either already procured or already tendered for that we expect to go forward. So, I thought for interest, it might be useful to show AFRA's projections of some of these technologies. So, we've got, from top left round, offshore wind, onshore wind, solar, and batteries. I'll come on to why I put batteries on here in a second. We have, and apologies, it's a little bit small. I wish I had made this font slightly bigger. But the dashed lines across horizontally are the target range announced in the Clean Power 2030 announcement, a sort of a guide as to what level of renewables would be needed by 2030 to hit the sort of general decarbonisation metrics of Clean Power 2030. And the first bars, as I was going to say, darker, but it's not for solar, but the left of the two bars is AFRA's in-house projection, our most recent in-house projection. And then for context, next to it is NISO's future energy scenarios holistic transition, which is probably one of the more ambitious decarbonisation scenarios out there, or certainly an interesting benchmark for where it gets to. And generally, we see that offshore wind will continue to grow significantly, will probably even hit Clean Power 2030 targets, but not by 2030. That's too ambitious. I don't think that's a hugely controversial statement. But it will hit it later down the line. But it won't get anywhere near the sort of upper bound of that holistic transition scenario where there is potentially 40% more capacity of offshore wind by 2040. On onshore wind, we have not seen significant levels of onshore wind through CFDs so far. If you look at the mix, very heavily weighted towards offshore wind. And when you look at the merchant economics of onshore wind, we don't see the Clean Power 2030 targets being particularly reasonable without a heavy increase in subsidy and intent to do so. So our views stay below the Clean Power 2030 range at least out to 2040. And again, well below the holistic transition scenario. And on solar, even the holistic transition scenario doesn't get anywhere near the Clean Power 2030 range by 2030. But we still see maybe a more conservative view than the holistic transition as, again, the merchant economics of solar start to see too much cannibalization to make a huge amount of build-out commercially viable. So these things are coming. More renewable capacity will get built. And in general, these are going to exacerbate these grid constraints. In fact, our in-house view may not be as aggressive as some other views out there in terms of what can happen with the renewables, but it will still happen. These constraints will still be an issue. The reason I include batteries is that these have an interesting interplay with renewables, and it happens to be kind of my particular area is looking at batteries in the GB market. They can, if used appropriately, help to alleviate some of these constraints. They are, in themselves, a form of sort of demand-side response. They can turn up when the prices are appropriate. And so, to some extent, the level of build-out of batteries can, to some extent, alleviate those constraints. But with the durations of batteries you see in the GB market, there is a limit to how much that can help. Somebody else earlier, I can't remember who, apologies, mentioned the LDES, and there's been a lot of information around LDES in the GB market last year, a lot of interesting developments there. And a lot of that is in Scotland, or the minded-to-announcement put a lot of capacity in Scotland, which also will probably help. But again, the amount of gigawatt hours that that can store will only do so much to alleviate these constraints. Oh, that has not mapped over from the slide I was looking at to this format, so apologies, there's a bit of blurring on the text, but I can talk through it. In short, this slide is really talking about the solution, or at least the partial solution, which is that NISO have, in their energy tenure statements, a view on what level of grid reinforcement will happen, particularly through the lens of high-voltage direct current transmission capacity, effectively bootstraps that will go around the country, typically offshore, to connect pressure points in the transmission network to alleviate some of these constraints. In general, you will see that a lot of these are going from Scotland into various parts of England or Wales, and it is designed to alleviate that congestion. That is primarily because you just can't, for a variety of reasons, put the level of onshore transmission network cable needed to alleviate these constraints. So you end up with these sort of inter -market interconnectors to help manage this. 14 gigawatts is a very ambitious and very, you know, intentful attempt to resolve this, and you can see that the dates of when these assets are expected to come online, according to the energy tenure statement, have changed a little bit since the last vintage to the most recent one. In the last one, they were varied over sort of a seven-year period, from 2029 to 2035, sort of staggering of these projects coming online. The latest one has grouped them all together from basically 2030 to 2033, which really means we will see a very, if delivered on time, a very material reinforcement period of the network over those years, as a lot of these constraints get eased. I think it is worth pointing out there is some healthy scepticism about how achievable these dates are, and while it's in many people's favours to achieve these dates, there is some concern of whether that's achievable. Historically, these projects used to take 10 to 14 years from conception to delivery. Some of these dates would suggest more like kind of six to seven to nine. So there is a concerted effort to invest in this and make it happen, but it's a significant step. In terms of what this means in costs, and in terms of what this actually means to consumers, and the impact on anyone in GB who's paying for the electricity, I've got some data points on a graph over time. These effectively show, at least on the left, the cost of, for want of better phrase, wasted wind. This is the cost, in green, of turning off wind farms that want to generate and have the availability to generate but can't get their electricity out. And then in grey, the cost of turning up something somewhere else in the system. And I think this often gets misrepresented. There's often a lot of talk of the cost of turning down renewables, and that is a cost, and those numbers are not insignificant. But the vast majority of that cost that is seen by the consumer is actually compensating for the renewables that have to be turned off to manage the constraint in the first place. They generally are thermal assets, generally gas, with a much higher marginal price than maybe the wholesale price would have been in that hour. And therefore, the cost of replacing that energy is much higher. You can see from 2024 to 2025, and even 2026, year to date, which isn't a full year in comparison, that cost is ramping up in line with things we saw earlier with the level of build-out and the reinforcement, the transmission network keeping up. We then jump forward to two numbers that are not, per se, AFRI's view, or not even the first ones are AFRI's view, but kind of real data, but projections from other people. So NISO themselves have a view of constraint costs, which isn't exactly the same as wasted wind, but wind curtailment is a significant driver, and that could be as high as £3.2 billion in the sort of August 26th to July 27th, 12-month period. And then DESNAS themselves, in the Reform National Pricing Delivery Plan, hint at a range for these constraint costs as well, which, if no acceleration in the build-out of network happens, could range from £5 billion to £10 billion. That is a huge number. That is roughly a quarter of what total electricity costs were last year. So that is a material part of the bill that could grow, and for obvious reasons, there are actions and needs to do something about that. It was touched upon earlier that there are changes to the Contract for Difference process. Contract for Difference reform has been on the cards since about 2022, or in its current state since 2022. It was announced in 2024 that this wouldn't happen until AR9, which is the next one to be announced after the one that will happen in weeks or so's time. And even last week, while I was preparing these slides, the Chancellor came out with a statement that basically says there is likely to be a pause before AR9 to build in these significant changes to the scheme. So there's a bit of uncertainty over what that's likely to look at. AR8 is expected to run as it was. I will skip these two slides. You can look through them. In short, there is another part of the issue of building out renewables, which is not just grid constraints, but is quality of energy and stability services, which also need to come with the consequence of building out renewables and replacing thermal generation. But I think in the interest of time, I won't go through that in detail. Please reach out if you want to talk to me about any of this, and I think you'll get it in the slides anyway. So I think I'll leave with just two final slides, which is this is not a new problem. This is not an unforeseen problem. The scale of it maybe is higher than anyone hoped it would be and more significant than it was hoped it would be. But there are many, many things being done to tackle these problems, of which I've listed some here, all designed on getting the network and the subsidies in a way that is actually efficient for consumers, which is hoped to solve many of these problems, but is a challenging task. I do not envy Desne's on ESA. And so I'll just leave with some final thoughts, which is that grid balancing costs have become a material issue for consumers and energy policy, and we're seeing a lot of that at the moment. Balancing costs are likely to rise despite this grid reinforcement, and there is expected to be a period where it goes up before it probably comes down in the mid-2030s with these grid reinforcements that I mentioned earlier. And energy policy is going to have to balance these sort of three things, a slightly different trilemma if you want, which is reinforcing the transmission benefit to maximise the benefit of existing and future subsidised renewables. That is needed. Supporting renewable build-outs to meet decarbonisation targets, noting that the power system is taking a lot of the weight of wider energy decarbonisation and minimising consumer bills, given the focus on affordability, wider energy costs. And I'll leave it there. If you want to speak to me or my colleague Fleur, who is in the background, please reach out to us on the session or following today. But thank you very much for that. Perfect. Wasted wind, eh? Well, that's clearly unsustainable. Some of the numbers you picked out here were mind-boggling. Louis, just a question, really. We have to move on because of time constraints. But the UK government clearly rejected zonal pricing. Do you expect that to resurface or to be re-examined? Because surely that would deal with some of these issues. Not all, clearly. Yeah, I mean, so Rima, as you say, came out with the conclusion that zonal pricing was being dismissed. And I think as soon as that statement had been made, there was an immediate revision of some of those kind of conversation points. It has definitely been discussed in the context of AR9 and whether some of that is needed. So I don't think the conversation's gone away. It comes with its own challenges and its own issues. So while it would resolve some of these issues, it does create some other problems. So I don't expect it to come back as the solution to this single-handedly. But I definitely think it's likely to become back on the table, particularly as sort of political counterviews want to get presented. Once again, Louis, thank you very much indeed. Thank you. And here's a little present from us. Thank you. Thank you. Thank you. Thank you. Thank you.