Hello, and welcome to another series in our series of webinars from AFRY Management Consulting. We've been running this series of webinars since early 2020 we started, and all of them are available on the website. If you were to Google or use some other search engine, AFRY energy transition webinars, you'll be able to find access to them all. Today, I'm very excited that we're gonna be presenting some of our at least high-level results of the work we've been doing on the review of electricity markets in Great Britain, and we've got a great team lined up to talk about it. If we move on to the next slide. Lovely. Thanks. Thanks for that. Just to say, before we start, we encourage questions throughout the webinar. We will do our best to cover as many as we can at the end. The slide pack and recording will be available, and attendees will be able to access that, as well. With that, I will introduce our two main presenters, that's Stephen Woodhouse, who's a director based out of the UK at AFRY Management Consulting. I'm sure Stephen will introduce himself, and Tom Williams, Principal, both of them heavily involved in this. We also have a special guest appearance later on, and that's Simon Dawes, who's from... I'm not sure I'm allowed to say DESNZ, but I've said it. And there's policy lead on REMA, I believe, you correct me if I'm wrong, Simon, and we'll be able to answer a few questions later on as well. So with that, I will, I'll hand it over to Stephen and Tom. Thank you, Matt, and Simon, thank you for agreeing to join us. We got caught up in the wave of enthusiasm around market reform about March last year, when I think I, among other people, were quite surprised that ESO was so bullish about the proposed move to locational marginal pricing and co-optimization. So I've worked on market design for more than 25 years. Well, no, I was very junior at the time. I was in the team that took out the centralized market that we had in the 1990s, and over those years, I've blown hot and cold on the idea of centralizing everything. So we set up a project with 12 clients and four observers, which included Simon's government department, BEIS, I think, at the time, as well as National Grid ESO, NGET and Ofgem as observers, and we had 12, 12 paying clients. And if you remember back then, everything was on the table. So the REMA consultation came out and basically threw absolutely every aspect of our trading and transmission arrangements into the air, including the concept that we might separate somehow the trading of, you know, let's use colors, green and brown energy into two separate products. So we started a project. We concluded phase I, which was entirely qualitative, last September, in time for the first REMA consultation. In summary, our conclusions there were, much to my disappointment, given the need to achieve 2035 decarbonization of the UK power sector, which remains a government target, we concluded that we would continue to need government-backed financial support, both for the zero carbon generation and flexible generation that enabled it. We thought further thought needed to be given to the form of support, so we concluded that CFDs, which currently distort dispatch, needed to be moved to another model, and we were recommending the deemed volume CFD, and that the form of capacity mechanism, which I personally have never been much of a fan of, needed to move on to something more akin to a flexibility mechanism, and we've not gone into that in any more depth. We concluded that there should not be split markets for different products, although curiously, with the 24/7 green initiative, that might come back through another route, but we concluded that shouldn't be a top-down market design. We concluded that we should look for evolution rather than revolution, and, you know, that in itself, you know, we did look at the centralization issues around nodal pricing, just in a qualitative way, and we concluded that moving to full-blown nodal pricing was likely to be more of a problem than a solution. The current phase, we focused on the aspects that related to location. So we set up a proper modeling study in which we covered national, zonal, and nodal markets, and Tom will walk us through the detail. You can't isolate location. It's not just about the definition of the energy price and the boundaries around it. We found we also had to look at the transmission charges, also the reference price for whatever form of CFD, and we were using a deemed CFD. Whatever form of CFD, you have to define the reference price, which is currently a national price. We concluded you might also need to change the mechanism for the award of the CFD. So not just chasing the lowest levelized cost, but chasing the lowest, lowest top-up value. So even the locational modeling was quite all-encompassing, and we've just concluded that modeling phase of the study. So I'll hand over to Tom to talk us through what we've been doing for the last nine months. Thank you, Stephen, and good morning, everyone. So yes, as we've seen, the REMA reform process is very broad, and we did indeed need to narrow the focus of our quantitative analysis in order to deliver it. So we've, as Stephen said, been focused mainly on the locational aspects of the proposed reforms. Well as national zonal or nodal energy market granularities, we've picked up locational grid charges and also locational aspects of the support regime for renewables. So I'm going to walk you through quickly the main model cases and the scenarios that we've modeled, just so you've got some background in terms of the results that will then follow. So we've modeled three main market design cases, which reflect coherent choices on those locational elements. So it's a national business as usual baseline, a zonal case, and a nodal case. So the national business as usual case retains the existing market arrangements. So it's a national market with locational grid charges and existing support arrangements for renewables. So in other words, 15-year CFD contracts awarded based on lowest price and payment based on metered output. The zonal case comprises an 11-zone energy market and grid charges, which are non-locational. So the assumption here is that the locational energy market gives you a sufficient locational signal, so the locational element of grid charges can be removed. And the nodal case comprises a 511 node grid representation, also with grid charges made non-locational. We also had to give thought to locational signals for new, new renewable capacity with CFD contracts. So in the main zonal or nodal cases, the CFD-supported plants are indexed to their zonal or nodal reference price, so they're not exposed to the locational price risk from that. So in the revised market design cases, we amended the award mechanism so that it's based on the least cost of support payments rather than the lowest price. Now, award based on least cost of support or least cost of top-up is a way of introducing a locational price signal at the investment stage, 'cause the lower costs of top-up payments are associated with locations where prices are higher and capacity is needed, and vice versa. Yeah, just to say also that the CFDs in all cases other than the national business as usual case are also based on deemed output rather than metered output to incentivize efficient dispatch. We've also modeled market redispatch with respect to transmission constraints, so that's based on an 11-zone redispatch for national business as usual, and nodal redispatch for the zonal cases with redispatch. Those cases have been modeled against two scenarios, which are based on the Consumer Transformation and System Transformation scenarios from National Grid's Future Energy Scenarios 2022. We chose those scenarios because they both achieve a net zero power grid by 2035, but under different decarbonization pathways. And we also added a requirement to both of those scenarios that 50 GW of offshore wind was built by 2035. In terms of capacity expansion, that's based on exogenous input assumptions from the FES 2022, with locations based on the Electricity Ten Year Statement 2022. And we then have additional economic capacity built on top of that, within the model, subject to constraints. Network expansion is based on exogenous assumptions from the Electricity Ten Year Statement 2022, and NOA 7 until 2030, plus economic assessment of holistic network development plan options. From 2035, there's also endogenous network expansion, you know, and output from the model based on assessing the costs of reinforcement against the cost of congestion, or the cost of redispatch, in the national cases that it relieves. For the nodal case, our approach to capacity and network expansion was a bit different. In the nodal case, we basically applied the same capacity mix by location, and the network boundary capabilities that we obtained in the zonal case to a nodal network representation. So effectively, what we're assuming there is that the zonal market is providing a sufficient locational signal for investment, and we effectively examine the nodal case for any additional operational benefits. In terms of assessment metrics, we've looked at several metrics related to investment effectiveness, operational efficiency, and allocation of cost, risk, and reward. But we've used a socioeconomic welfare assessment as an all-in way of comparing the benefits or disbenefits of different locational market designs. Okay, good. All right. So our first overall message is that there is only a small potential economic welfare benefit from moving to a locational market. So if we look firstly at the chart on the left, what we're showing is that the net present value of total economic welfare benefits in our zonal and nodal cases relative to the national business as usual case for both the Consumer Transformation scenario, which is labeled CT on the X-axis, and the System Transformation scenario, which is labeled ST. So we're seeing a total benefit of GBP 4.2 billion in both scenarios for the zonal case and GBP 4.4 billion or GBP 4.5 billion for the nodal cases. Just to be clear on what these numbers are showing, so the net present values for economic welfare for each market design have been calculated over an appraisal period from 2028 to 2050, assuming a 3.5% discount rate. We've then taken the difference between the total economic welfare in the revised market design case and that in the national business as usual case, to show the change in welfare in the new case relative to the business as usual. If we now look at the chart on the right, what that's doing is comparing the benefit achieved by the nodal case in the consumer transformation scenario with total consumer bills in the consumer transformation scenario over the same discount period. As you can see, the total benefit of GBP 4.5 billion is about 1% of the total consumer bills of GBP 466 billion, which is a small benefit given the change of the market arrangements that would be required to deliver it. And that figure excludes any cost of implementation, by the way. So the benefits of locational markets appear small, both in absolute and relative terms. And whether it's worthwhile undertaking such a big market reform to try and obtain those is a question policymakers will have to address, particularly in the case of nodal, where the additional increase in complexity seems unlikely to adjust by the additional economic benefit. All right. In addition to being small, the potential welfare gains from locational markets risk being overshadowed by distributional issues. So what this chart is showing is the build-up of additional economic welfare in the zonal case across producers and consumers, compared with the national business as usual. And the total GBP 4.2 billion welfare benefit on the right corresponds to the GBP 4.2 billion we saw for this case on the previous slide. Now, one thing that stands out on this breakdown is the size of the individual components. So if we look at the intra-GB congestion rent, which is the big green bar, that's the largest bar on this chart, with a value of GBP 10.6 billion. And just so we're clear what we're talking about there, intra-GB congestion rent is revenue captured in the zonal case by the grid operator as a result of price differentials between the zones, and it doesn't exist in a national market. So in theory, you can allocate intra-GB congestion rent across the spectrum, ranging from all being allocated to consumers, all being allocated to producers or somewhere in between. What we've done on this chart is we've allocated it all to consumers. So effectively, what that's saying is that the congestion that's captured by the grid operator, used to pay for the cost of the grid, and then that is reducing the cost recovery required through grid charges. And if you're in a zonal market where you don't have any locational grid charges, for generators, that's a benefit, to consumers. And we can see that's the biggest driver of the consumer surplus of GBP 7.7 billion, which is the black bar in the middle there. But we can also see there's a red bar of GBP 2.9 billion, showing that producers are, in aggregate, worse off. That's gonna include, obviously, a range of producers, some of whom are going to be much worse off than the aggregates, due to being in a particularly unfavorable location or not having a support contract, and so on. To manage those risks, generators would need locational hedges, but suitable locational hedges for intermittent generators are not available, as they are invariably base load products. And another issue is that the energy transition also poses particular new risks for generators in a locational market, because not only is there a risk of another generator locating nearby and disrupting your revenues, which has always been the case for locational markets, there are now also policy-driven infrastructure risks, such as the advent of hydrogen or CCS clusters, which are coming from outside the power market and are even harder to predict. So in practice, risk management measures are needed for at least some producers, for example, through grandfathering of transmission rights. Now, as a simple example of trying to get the balance right between consumers and producers, we can see that if we take the GBP 10.6 billion of inter-GB congestion rent and allocate that 50/50 to producers and consumers, so we allocate GBP 5.3 billion to risk management to producers, both consumers and producers would then be better off by GBP 2.4 billion. But the GBP 5.3 billion transfer that you're then doing through those risk management measures is obviously larger than the actual social economic welfare benefits across all of society. That's one issue. And then the other issue is that risk management measures may also have the potential to soften locational signals, and that means you may actually reduce the total welfare benefits that you achieve across the whole of society as well. So in other words, you, you wouldn't get GBP 4.2 billion anymore, it would be something a bit less. So the size of the transfers could be overshadowing the overall gains that are achievable. Okay, that brings us neatly onto the topic of locational signal strength. Now, the national business as usual case already has a significant locational signal in the form of the wider tariff component of TNUoS. So any consideration of signal strength between different cases has to consider the effect of those charges, as well as the locational granularity of the energy market. As I mentioned earlier, we've assumed in our national case that the locational TNUoS charges continue, but in the zonal and nodal cases, they are removed. What we've then done is look at the strength of the overall locational signal faced by different technology types, based on the change in their financial position, between the different market design cases. The table that you see here is summarizing those results, with a larger number of balls indicating a sharper signal for demand and generation to locate close together. Now, what we've found in our scenarios is that for many technologies, and if we look at onshore wind there, offshore wind, and gas CCS plants, for example, locational signals under a zonal or nodal market design are sharper until 2030. So we've got more balls in the zonal case in 2030, but weaker by 2035, so fewer balls showing a weaker locational signal strength after that. And what's happening is that in 2030, the rapid deployment of renewable capacity, which includes the 50 GW of offshore wind coming in by 2035, particularly in the north of GB, is resulting in high levels of network congestion and lower revenues for lots of generators in the north compared to the south of the country. which gives a strong locational signal to locate nearer demand, and that's a stronger signal than is coming from the national price with the TNUoS charges, which are also significant. But then, as network build is catching up for 2035, the congestion rent in the zonal case is reducing, whereas TNUoS continues to provide a strong differential between different parts of the country. So what we found, which was quite interesting and perhaps contrary to conventional wisdom, is that for many technologies, in the longer term, today's national market with locational bid charges actually provides a stronger signal for demand and generation to locate closer to each other than a locational market with... Yeah, as I say, is a bit unexpected in terms of conventional wisdom. Okay, just in terms of some of the sensitivities that we did. So, as well as the main national and zonal cases, we also ran a sensitivity based on an unanticipated delay of grid build. So we applied a five-year delay to all transmission reinforcements from 2030 onwards, and wanted to see what the effects were for producers and consumers and the different market designs as a result of that. So the first thing to say is that overall, transmission build delay is negative for overall economic welfare, whether it occurs in a national market or a zonal market. But the impact falls more on consumers if it happens in a national market, and more on producers if it happens in a locational market. Just thinking about the producers for a moment, the chart is showing the results of the transmission delay sensitivity for an unhedged onshore wind plant. If we look at it, going from left to right, we've got the 11 geographical areas, which correspond to the 11 zones in the zonal market. On the vertical axis, we've got the post redispatch energy market gross margin in the transmission delay case, relative to the base case. If we take the orange line as an example, that's showing how much higher or lower the annual energy market gross margin is in the zonal case with a transmission delay, compared to the zonal case with no transmission delay. So in a zonal market, what we see is that the impact of a delayed grid build is severely negative for generators who end up being held behind an export constraint, for longer, with revenues reducing by up to GBP 50 a kilowatt on an annual basis, around 2035. So looking over at the Scottish zones there on the left. And that's, you know, severe enough that it suggests some of those generators may actually go out of business if they're unhedged, basically because they happen to be in the wrong place. So overall, what the transmission delay sensitivity showing, with respect to producers, is that locational markets do have the potential to cause quite large risks to unhedged generators. And those unhedged generators, of course, include any existing generators that would become unhedged if their long-term support contracts end. So that, yeah, is a sort of further reinforcement of the general point earlier about any move to a locational market needing to have quite strong risk management frameworks in place for generators who would be exposed. Okay, just following on from that then, another one of our key messages is that the complexity of the energy market would increase with the number of locations, and that creates additional challenges for predictability and decision making by market participants. So the graphic is showing the increased number of prices in both the zonal and nodal markets, and their evolution over time. There were 11 locations with different prices in the zonal case, and 511 in the nodal case. In addition to that, the differences between the locations are also changing over time for multiple reasons, which are quite complicated. There is a complexity and cost to working out if a location in a locational market is a good one, and if you are a smaller developer, that could be quite a significant barrier to entry. Overall, our view is that the increased complexity of locational markets may indeed create barriers to entry, which could result in a reduced number of market participants. Obviously, if that happens, you'd then have potentially a reduction in market liquidity, which would then also potentially reduce the actual overall economic welfare benefit that you might be able to achieve in practice. Okay, so following on from the risks and uncertainties that may be faced by certain generators in a locational market, it's time to talk about cost of capital. We found that the overall welfare benefits achieved in locational markets are very sensitive to increases in cost of capital. In our main zonal and nodal cases, I should say that we have already assumed increased levels of cost of capital for some generators relative to the national business as usual baseline case. So what we assumed is that it was +100 basis points for non-CFD supported capacity, or 50 basis points if it was OCGT capacity. But we didn't assume any increase for new renewable capacity, all of which we assume has a 15-year CFD contract. contract. So the small potential overall welfare benefits in the zonal and nodal cases that we've seen already assume some increases in cost of capital relative to the national case. Now, for the zonal market, we modeled actually two variants. So there's the main zonal case, which has the CFDs for renewable plants referenced to a price which is the zonal, local zonal price. And we also did a variant of the zonal, which we called the Zonal (N) variant. And there, the CFD reference price is a proxy national price. It's based on a generation weighted average of the zonal prices. So in that case, there's a basis risk between the zonal and national prices that the supported plant's exposed to, and we say that that would increase investment risk. And consequently, we assumed a one percentage point increase in hurdle rates for that type of CFD-supported capacity in the Zonal (N) case, relative to the zonal case. Now, if you look at the chart on the left, we can see that this has more than eliminated the welfare benefit in the zonal end case. So we've now got a net disbenefit compared to business as usual, of GBP 8.9 billion in the consumer transformation scenario, and GBP 6.2 billion in system transformation scenario. Now, as a variation on this, rather than applying a 1% increase to CFD capacity and seeing the disbenefit that resulted, we also looked at what smaller increase would be required to move a zonal market to being neutral on overall economic welfare. And if we take the bottom row in the table on the right, what we found was that the increase needed to do that is 52 basis points in consumer transformation and 56 basis points in system transformation in the cost. So that's the increase in cost of capital needed for new renewable support CFD plants in the zonal case, to eliminate all the benefits that we saw... we see there. So that shows a degree of sensitivity, to say the least, in terms of the benefits and cost of capital assumptions. Our view is that given the current ambitions for the pace of investment over the coming decade, you know, the potential implications of locational markets on cost of capital do require thorough investigation. Really, you know, a more robust evidence base is really needed to inform further moves towards the locational market as part of the process that's ongoing at the moment. Okay, so that concludes my section. I'd now like to hand over to Stephen to sum up and present the summary study recommendations. Thank you very much, Tom. And you've covered a lot of the detail. I'm sure there'll be some question on more of the assumptions, and in the end, most of these modeling assumptions hang on the methodology and the assumptions. There's a report on the web, which should answer quite a lot of the detail. So let us step back then to the high-level messages. We have confirmed our conclusion from the first phase of the study, which is that we recommend evolutionary rather than revolutionary change in order to maintain the investment momentum. We found only a small economic welfare benefit of changing to a locational market, whether that be zonal or nodal. We find that locational markets give better dispatch or better operational incentives, and in particular for interconnection and perhaps for plants behind constraints, where a lot of this is about managing locational constraints. However, we found that locational markets may not significantly improve the investment incentives, the locational investment incentives, and indeed, we found that the locational markets even give you weaker locational incentives than the current TNUoS arrangements, once you get to a point by 2035, where the grid build has caught up with renewables. And basically, TNUoS penalizes distance, whereas locational pricing penalizes congestion, and we find that we're moving from a world of congestion to distance being the driver. We concluded that the risks to investor confidence would be greater in a locational market. Actually, the way we modeled that, we looked at there being an increase in the cost of capital to plants, which didn't have some kind of locational hedge, and we were assuming that new build renewables somehow would have a locational hedge baked into their support contract. We weren't looking at all at the cost of the risk to existing capacity. We haven't considered that at all. But even with a small penalty to the cost of capital for new, let's say, merchant capacity, we found the overall benefits were small. If we introduced new risk to new renewables, we found the benefits were completely wiped out. So that risk issue associated with the locational markets, and as Tom said, some of those risks aren't things which are endogenous to the power market, they're given by external factors. That's a key difficulty of any move to a locational market. We concluded that moving to locational markets could create barriers to entry due to the complexity. You know, with zonal, you end up with the risk of boundary changes. With nodal, there's much more complexity in the way you have to understand the revenue streams from the market. And, you know, the best way of hedging location is to have a wide portfolio. So we're a little concerned about the barriers to entry aspect of locational markets. So stepping back to this finding that we found operational, but to a much lesser degree, investment improvements from a locational market, we started to explore whether we could achieve some of those benefits by tweaks to the transmission charging methodology. Now, we were looking at static charging, so we were only really able to focus on, you know, the incentives to site in different places, so looking at the investment issues, not the operational issues. We were able to find some of the efficiency savings through different structures of grid charges. I think if you want to explore that further, you'd perhaps have to look at more dynamic grid charging to get at some of the operational issues. We also have a locational operational incentive in GB through seasonal loss factors, which are part of, they're effectively part of people's bidding because they're volume adjustments. You know, it'd be possible to look at that and consider whether the loss factors should be applied more dynamically to get more operational benefits. So we didn't find any slam dunk there, but I think there's quite a lot of space still to explore to get both operational and investment improvements through changes to the transmission and access regime. We concluded that the window for reform is short. The ability to implement these changes in time to get in any of the operational benefits is quite limited. So that, again, concludes our recommendation to go for evolutionary rather than revolutionary change. So our conclusions move to recommendations in three spaces. Nodal pricing... Let me see if I can get back to the right slide. Nodal pricing, we haven't found a good justification to take the analysis of nodal pricing further. There are many issues there, some of which we did model, some of which we didn't. But in particular, the idea of nodal pricing is kind of going against the grain, we feel. Essentially, redispatch results in a, in the same dispatch that you might get with a locational market. You still have the same constraints to deal with, unless the process itself is inefficient. And if the process is inefficient, we should be looking at how to make the process more efficient, rather than, you know, changing the allocation of wealth between the participants. So if you move to a nodal market, you necessarily move to a central dispatch market. Essentially, what you're doing is saying, well, we want to take earlier control of people's unit commitment and dispatch decisions. So essentially, we're talking about earlier Gate Closure, and none of the discussions we have about how to make markets more suitable for renewables really go to earlier Gate Closure. They mostly talk about later Gate Closure. And I think my other concern with the move to centralized nodal pricing is the ability to optimize the diverse set of decentralized assets that we'll have in future. So I don't see working examples yet of Nodal markets where we have good optimization for things like, you know, small-scale demand response or indeed for batteries. So the intertemporal constraints that we're trying to battle with this centralized market design are not properly priced in the centralized markets. So even start-up costs for generators are not part of the price formation. They're paid under the table as make-whole payments. So we haven't found a good example of a nodal market that deals with the flexibility needs and the decentralization of a renewable power system. So we conclude that nodal pricing is not the space to explore. If we are to move to a zonal pricing regime, there are some efficiency gains that could be found there, particularly around the way interconnectors and plants behind constraints are currently treated. But in order to get a profitable move in that direction, there would have to be some real emphasis on the ways that generators in particular could hedge the locational risks. And again, I don't see any examples of locational instruments hedging risks adequately for generators. So a case in point is that cross-zonal CFDs seem to be virtually impossible. The basis risk of striking a long-term CFD for renewables between two price areas in Europe seems to be an insurmountable barrier. People talk blithely about FTRs. They tend not to last for very long, not to hedge an investment, and they tend to have a base load product shape, and there aren't very many base load generators out there that I can see. So it seems to me that there would have to be quite a bit of work on how we would manage the risks if we were to move to a locational zonal market. Conversely, if we're to continue with the national market that we have today, there's certainly scope for improvement for the operational efficiencies, and we think that would have to focus on how interconnectors are treated. So since our step out of the internal market for energy, the effectiveness of the interconnectors on the intraday timeframe has been diminished, and it's certainly the case that, you know, plants behind constraints, we see some difficulties in the way they're treated in balancing and perhaps some of the incentives for people's bidding. So we believe there should be further exploration of things like access rights, and the treatment of interconnection in the markets to exploit greater efficiencies in the dispatch in the national market. So those are our high-level recommendations. I'd like to hand over for questions at this stage. But actually, the very first question, or questions, I think I'd like to hand over to Simon Dawes to answer. So we had a number of questions which related to the REMA process itself. So the questions that I'd seen are: When is REMA due to be finally approved, and when is it supposed to come into force? What's the status on the long-term possible move towards nodal pricing? And more generally, what are the next steps in the REMA process? So Simon, would you like to have a go at answering whichever of those you choose to look at, please? Yeah. Great. Thanks, Stephen, and Tom. Very much welcome this study. It helps inform the debate around all these very complex and interesting issues. So in terms of where we are with REMA, next steps and time frames, so I think the last time that we said anything on REMA publicly is probably in March, when we published our government response. At the time, we indicated that we would aim to publish our second consultation in the autumn, which in civil service speak is pretty much any time from now until the end of the year. We're still on track to hit that very broad time window. We're currently going through the deci... Getting all the different decisions and all the different options, signed off and making recommendations to ministers and going through the cross-Whitehall clearance process, which can take quite a bit of time, which is why we're all so pretty vague in terms of exactly when we're going to publish, because it depends on how smooth that process is. And obviously, there'll probably be quite a lot of interest in what we're proposing in that second consultation. In terms of the content, because we're in the middle of that process of making decisions and recommendations, I can't really say anything about exactly what's gonna go into the consultation 'cause it... I'll probably say something, and I'll end up being wrong. We'll have to change 'cause we'll have to change our position during this process. But what I can say is that what we're aiming to do is, well, you know, as was highlighted, I think in one of the early slides, there was that illustration of all the different options that REMA was looking at in the first consultation. So what we're aiming to do now is narrow down, as far as possible, the range of options that are left on the table, so that we can then move into the much deeper assessment of the relative costs and benefits of implementing these things and think through in a much more granular way what it is that we want to proceed with. So in some circumstances, what we're aiming to do is identify preferred options that we can proceed with sooner rather than later. And then in the other areas, where it's a little bit more complicated, and the issues are more finely balanced, we might still have one or two options left on the table, where we'll have to seek further views and input. I think it's probably no surprise if I say that addressing location is one of the more complicated and complex topics within REMA. So don't be surprised if we don't come out with a preferred option in this space, but we would like to think that we can narrow the range of options, that we're looking at, on location. That said, obviously, part of the location debate is, are there alternatives to locational pricing? which is probably something that the first REMA consultation didn't look too much at, but we are certainly keen to, to look at that, in more detail. I think it's probably slightly easier in terms of investment signals, more challenging with the operational signal side of things. So I think I've probably covered, I hope, the, the three questions, Stephen, that, that you outlined. Obviously, if, if people have any further questions, either reach out to me or the team directly, and we'll, we're always happy to engage and talk to people about the, the challenges that we're collectively grappling with. Very good. Thanks very much, Simon. Thanks for that, and I think we have a few, we have a few other questions we could run through. Of course, one key question that's been posed is: How does your analysis compare with other analysis, particularly the work that FTI has done? I think we have, we have a slide. Who's gonna take us through that? Is it you, Stephen, or Tom? It's me, if I can only figure out how to show it again. One second. There. Is it now showing? Nearly. Yeah. There we are. Very good. So I think the obvious point of comparison is the work that FTI had done for Ofgem. They came out with substantially larger benefits. So I think they found on the nodal case, GBP 13.1 billion, whereas we had about GBP 4.5 billion, and on the zonal, they found about GBP 6.2 billion on a comparable case where we found GBP 4.2 billion. And in their Leading the Way scenario, which we didn't model, they found higher benefits still. Now, the start and end date are different, so we modeled 2028 to 2050. FTI modeled 2025 to 2040. Now, it's a shorter period, but they assumed the benefits would start much earlier. I think there was a difference in the monetary base year as well. We were looking at 2021 numbers, and I think they were using 2024, although I wasn't able to actually confirm that. Now, starting in 2025 is quite important. You know, we concluded that, you know, the benefits of locational pricing are greater when there's more congestion, and once you've relieved the congestion with more grid build, the benefits start to fall away. Obviously, if you assume you can have nodal pricing up and running, you know, in 18 months, and you get the benefit from 2025, you will be inflating the, the benefit that you might get. So I think that's one key difference, that start year. And actually the inflation between 2021 and 2024 could easily be 20% as well. But I think there were some more fundamental issues around the assumptions. So our study, we looked at economic optimization in our national case. So the generation locating based on signals from losses and TNUoS and building grid. At least in the early years, we were building, we were building grid according to the NOA, but we were building grid on an economic basis thereafter. And I think FTI didn't optimize the national case. So their baseline for comparison, I think, was in sort of by design, although whether it was by intentional design, but was by design worse than ours. I think the other point is that we optimized the zonal, but we didn't additionally optimize the nodal. And the justification for that is that the pattern of nodal prices is so ephemeral that it's quite difficult to make perfect foresight investment decisions based on the nodal pricing. So we optimized the location of infrastructure, including grids, based on the zonal, in both the zonal and the nodal case. So we didn't get any additional locational incentives through our nodal case. So that again is by design. Whereas FTI, I think, optimized everything with perfect foresight based on their nodal. So they've got a better nodal case and a worse national case. And I think objectively, I do believe our assumptions there are rather more justifiable. FTI assumed no increase in the cost of capital, which I find a little bit hard to believe. We assumed an increase in the cost of capital for non-CFD-supported new capacity in our zonal and nodal cases. We also tested a case in which the CFDs or the deemed CFDs give full locational protection to new renewables, and that was the one that tipped the balance into quite a significant negative benefit. We haven't looked at the increasing cost of capital at all to any existing generation. So you might argue that we even understated the impact of the cost of capital changes, but that was one differential. And I think another important differential is that we assumed that there would continue to be economic build of network based on economic need, whereas I believe the FTI figures stopped at the NOA assumption. So essentially, they built a more congested network, which gives you greater advantages of locational markets. Having said that, you know, their figures for a zonal market of GBP 6.2 billion, I don't think are night and day different to ours of GBP 4.2 billion, albeit different base year and different period models. You know, if ours was 1% of total consumer bills over the period to 2050, I'd be surprised if the GBP 6.2 billion that FTI have come up with are close even to 2% of total consumer bills over the period to 2040. So, you know, the headline numbers might look different, particularly if you look at the leading the way scenario, but I think objectively, the differences are probably explainable and not that great in the scheme of things. Very good. Thanks very much for that. I think, you know, as you said earlier, it's often about the assumptions that have been made in these sorts of analyses. I would say also that I guess the what we've presented here is focused quite a lot on locational. In the first phase of the work, there was quite a lot of thought around what could be changed positively to enable a better market design for the future. And so we're not really drilling into that here, but we'd be happy, I'm sure, to talk to people who want to discuss some of those ideas that came up in phase I as well. So with 10 minutes left, I think maybe, this question down here: When looking at energy cost to consumers, what has greater significance, system costs or cost of capital for infrastructure? I'm assuming the latter outweighs the former, hence your conclusion, but we're good to confirm. So who could take that one? Tom? Yep, thanks, Matt. So, yeah, I mean, we've found that there are some small zonal—you know—operate the overall benefits in the locational cases that are coming from some of the operational efficiency advantages. But those benefits are very easily wiped out by, you know, the cost of capital risks with relatively small increases in that, you know, overriding the benefits. So, we would say that the increases in cost of capital, you know, are very important in this comparison. Okay. Happy to discuss further that one as well. Next one: What pressure is the regulator under to allow more connections and transmission capacity? Stephen. I mean, there's a widespread recognition that we need to be better at building grid, and we need to build it faster. So I think there's now acceptance of the need for anticipatory investment in under certain circumstances. National Grid ESO will become the future system operator and will happily have responsibility for, you know, grid planning across the vectors, you know, including CCS and hydrogen, but also regional grid planning. So looking at, you know, the trade-offs between, you know, not investing in distribution grids and getting the flexibility from distributed resources available to the transmission system. So I think, and the Windsor Report as well, should give us renewed emphasis on grid build. So, you know, it seems to me that there was a period when grid build was, you know, the unpopular part of the energy transition, but I think it's widely recognized that this has to go forward faster than it has before. Okay. And then there's one more here, which I think would be interesting, and that's similarities and with and deviations from the EU market design reform. How does this compare? I know, Stephen, you're quite involved in some of the discussions there as well. Poland tried to introduce nodal pricing for balancing, and I don't think they've yet found a way to make it compatible with the internal market for energy. Something we haven't talked about is how we might integrate a nodal market, if we ever wanted to return to the internal market energy. So, you know, the central dispatch and nodal market, I think it's incredibly difficult to see how that could be made consistent with the, with the EU's market design, which is essentially about bilateral trading and decentralized, you know, decentralized trading, although there are central dispatch markets within the EU. A zonal market might be much better, to integrate to the internal market. I think the really interesting thing with the locational market is the extent to which we might end up using Ireland as our kind of loop flow. So you could see circumstances in which prices might be low in Scotland. We might export via Moyle, and they may import again via the East-West Interconnector further south. So I think there would have to be a lot of consideration how we would integrate, you know, and that would happen whether we're part of the internal market or not, if we had different prices, north and south. So I think there could be some quite big political consequences of using Northern Ireland and Ireland as a bootstrap. Okay. So, EU covered. One here, that's quite interesting. It says: Is deemed output for CFD safe? Needs to be defensible, on what consumers are paying, auditable, et cetera. What are your views on that one, Stephen and Tom? Well, I need to agree with that comment. So, you know, if you pay people per kilowatt hour of output, you can meter their kilowatt hours of output. But the trouble is, it distorts a lot of things. If, if the receipt of the subsidy is dependent on having produced at times when the value of the energy is negative, or even when the value of the reserve they could offer, if they part load, is greater than the value of the energy. So, you know, if renewables are 10% of the system, it doesn't matter if they're dispatched inefficiently. But we're looking to be able to run a completely decarbonized system by 2025. So there'll be times when renewables are a very high percentage of the system, and we can't afford to dispatch them inefficiently. So the distortions on the market of having production-based subsidies, which is what we have today, quite severe. But you're absolutely right to point out that if you pay people for things that they don't do, it has to be transparent, it has to be justifiable, and it has to be auditable and not gameable. So we haven't looked at the detailed design of that. Happy to do so, I'm sure. You know, and then there's a few questions here around demand side response, demand, how have we addressed, demand and demand side response, and, looking at in terms of modeling on that side? Tom, you could answer that one. Yeah. So, the first scenarios do contain, you know, quite significant volumes of DSR, and we've included that in the form of price threshold demand within the modeling. So to that extent, we've incorporated that. We also reflected the shape of the demand that you get in a decarbonized system from EVs and heat with the modeling that we did. Yeah, which wasn't a full dynamic modeling, hour by hour of flexible EV and heat demand, because we found it computationally quite challenging just to do what we did already. So that wasn't possible within the scope of the modeling, but we've certainly certainly reflected the shapes in a high decarbonized market from on that side as well. Thanks very much. We were looking at hydrogen particularly, so that we would, you know, there was quite a lot of demand-side flexibility in some of the resources we were modeling. And then just one, I think one last one here. It says, a lot seem to depend on your assumption that grid development catches up with res development in 2035. If that doesn't happen, what changes about your conclusion on zonal and nodal? I think it's fair to say that the operational efficiencies are more important when there's more congestion. But I would equally argue that the risks and the, you know, challenges to the cost of capital are greater when there's more grid congestion. So I think it ramps up the importance of getting the decision right. I'm not sure it pushes it particularly in one direction or another. Okay. Very good. Thank you, thank you for that. One more? Yeah. So there were a couple of questions that were asked in advance that might lead to quite a neat conclusion. Go ahead. So one of them, yeah, well, one of them was the point, sorry, about the, you know, the significance of the system cost. I think the final one does really relate to, yeah, getting the incentives for operational efficiency right. If you look at the nodal market, the kind of implicit assumption is that somebody knows best. You can put all of the assets into a box and come up with an optimal dispatch, and that optimal dispatch can somehow be updated during the course of the day as circumstances and forecasts change. And we just haven't seen any market in place where that is true. A typical nodal market has a day ahead and a real time, and anything that happens intraday isn't really a market. You know, Texas, the day ahead is a financial market. In today, what happens is out of merit redispatch under a reliability flag. So, you know, that world of nodal markets seems to be a bit of an academic's dream, but it doesn't seem to meet reality. Zonal markets are really about allocating risk, and what we're trying to find, what we think the next step should be about trying to find ways of mitigating the risk of a locational market. But really, what we think the optimum is to try and get some of the dispatch decisions that you would get in a zonal market to be taken earlier than the balancing market timeframe. I know Australia is looking at the idea of a voluntary congestion management market, so they're retaining a national market, but with a voluntary zonal market around that. So allowing the primary congestion to be traded out at an earlier stage. It seems to me that's likely to be where the more profitable and less disruptive, and less risky, solutions end up. Very good. Well, thanks very much. Thanks especially to you, Simon, for joining us and giving us an update on REMA. I know, of course, it's difficult to say too much, as you're, you know, in the middle of it all. Thanks also to Stephen and Tom, also to the client study members. We had a broad church. We also had, Stephen pointed out earlier, observers joining from a range of other industry organizations. So it really was a very, very good discussion we had. Hopefully, you'll find the conclusions interesting and useful, and look forward to discussing them further with you.