Nodal or Zonal?

The UK is reviewing a shift in how it prices electricity from a single wholesale market, with two locational alternatives under consideration: nodal or zonal. This week’s pod addresses the pros and cons of locational prices and whether they will help or hinder the UK’s quest for net-zero emissions.  

Host: Snjólfur Richard Sverrisson, Editor-in-Chief, Montel

Guests: Stephen Woodhouse,  Director Afry Management Consulting and Phil Hewitt, Director, Montel EnAppSys.

Transcript

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 0:18

Hello listeners and welcome to the Montel Weekly podcast. Bring energy matters in an informal setting. In today's pod, we discuss nodal or locational, or even zonal pricing. The UK is currently undertaking a review of its electricity market arrangements, also known as rema, and the government has sought out views on whether to introduce sharper locational signals. Into the wholesale market. The issue is contentious and has proved quite polarizing with many diverse views. But what could be the benefits of locational, nodal, or zonal pricing? Could it drive increased renewable generation where it is required or will it drive away investors? Helping me, Richard Sverrisson to address these issues and much, much more are Stephen Woodhouse of Afry and Phil Hewitt of Montel EnAppSys. A warm welcome to you gentlemen. Stephen, if I can start by you. When we talk about locational pricing, what do we mean?

Stephen Woodhouse, Director Afry Management Consulting: 1:10

The conventional use of the phrase locational marginal pricing. LMP has been captured for nodal pricing where there's a separate price potentially for every node on the transmission network. It's still a zone below that level, but I am also using the phrase to mean in the British context, a separation into separate price zones. So the idea. You would have different energy prices in different locations. That would be a step change in Britain. I think the distinction between zonal and nodal would be a further step change.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 1:45

Phil, why choose nodal pricing? Why go down this route? What? What are the benefits of having such a system in place?

Phil Hewitt, Director, Montel EnAppSys: 1:51

The situation we have in Britain is that we have a big demand concentration in the south of the country. And a large concentration of generation, renewable generation in the north of the country, and the capacity on the cables is not sufficient for when the wind is blowing Scotland, essentially to move the power down to demand center in the Southeast. So broadly what happens is that National Grid has to constrain the network and has to turn down a lot of expensive renewable generation in the. And then turn up a lot of fossil generation in the south to balance the system. And that's because we have a national price. Everybody clears at the same price. So in a locational system, whether that's multiple zones or all the way down to individual nodes, you have prices. So if the system becomes constrained in the north, then the price for power in the north goes down to encourage people to stop generating. The price in the south will increase to encourage people to generate more in the south. So you transfer the cost in the market from the system operator who's using something like the balancing mechanism to turn down and bring up power stations. You transfer that risk onto the generators who then may not be able to generate because there's literally not enough capacity on the wires. And then the argument further goes on that. Because you've got this differential in price, there's an encouragement to build transmission between the two zones to allow more generation to come from the zone, which has got too much to the zone, that's got too little. And that is essentially the principle behind any kind of locational pricing, whether it's zonal or nodal.

Stephen Woodhouse, Director Afry Management Consulting: 3:24

So the location of anything you buy matters, whatever it is you buy. If it's a physical thing, you care about how quickly it will arrive, whether it'll arrive safely, and how much it would cost to deliver. So location, nobody can dispute that location is an important part of the economics of buying and selling things. And if you describe, the model of locational pricing, essentially what you're doing is. At the moment when the market runs, and that's very important. At the same time, you dispatch and price everything according to the grid congestion, and you take the congestion into account in the grid, so that sounds brilliant. The trouble is there aren't enough times when the markets run to get the right price. For flexibility and I'll come back to that. And the other problem is you want a market to incentivize investment in the right locations, not just to get efficient. Dispatch. Now, if we compare the worlds that we just described, where today generators in the north would like to run, we have to compensate them not to do and we then have to pay more expensive generators in the south to run to fill the gap under a locational market. Exactly the same dispatch would happen. We don't have enough grid to run all of the generation in the north, and therefore we have to replace it with more expensive generation in the south. So at face value, a nodal pricing system wouldn't make dispatch better. It might make the process of dispatch better. It might change the compensation that some generators get because of the rights they have. In itself, it doesn't automatically improve the efficiency of dispatch or get a lower cost of generation. In practice, you would get some improvements because we don't have efficient re-dispatch of all of the assets on the system. So some things need quite a lot of notice. The interconnectors as well are not perfectly flexible right up to the balancing market. In principle A no nodal oral market doesn't automatically even give you a better dispatch. The biggest impact is it changes the rights to some of the generators and you may have to give them access rights and compensate them anyway,

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 5:49

Phil, if the issue is mainly then congestion, why not just build more grids, more wires?

Phil Hewitt, Director, Montel EnAppSys: 5:54

So one of the reasons for locational marginal pricing or zoneal pricing is that you get a clear price signal on where you should create, where you should connect the the cables to the different zones. So at the moment, a national market that just is done by planning and engineers saying I think this is the best way to route it. Whereas within a market signal, you get a market incentive and then that means that ultimately you could have third parties who decide to build cables 'cause they see there's a revenue opportunity to move power from the north to the south. So the opponents of locational pricing say that this encourages. More efficient building of connections between the different zones versus a kind of national market where there is no price incentive to do that. And what's your view on this? Do you agree? So now you get an economic debate versus kind of physical reality? The problem, yes, it does give efficient, potentially it gives efficient price signals. That's what some proponents say. The issue is you then have to change the market to deliver that. And people are saying it's gonna take between five and nine years to deliver the zone a nodal market. And then after that you have to give a few years for those price signals to appear. And we are up against it. We're in a climate emergency, we need to build renewables and we need to connect them to the network. So you're going to land up building the cable anyway in the interim. So therefore, by the time you actually get to your nodal market, you may have already built a lot of cables, inefficiently. And anyway, the market has changed, which means that maybe. All that money and effort you've spent. So I did a bit of research last night and I found out that need cost around 600 million pound to implement across the whole industry. So that was in trading arrangements that went live in 2001. So you can consider that the step change in that cost to consumers ultimately of changing the market. Would you be able to recover that when you actually get to the point of implementation because you've already had to build a lot of these cables anyway to keep the national market running.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 7:47

Absolutely. What's your view, Stephen?

Stephen Woodhouse, Director Afry Management Consulting: 7:48

It's not a question of building grid or managing congestion. We will have to do both. LMP is one means of managing congestion, and there are plenty of others. I looked at LMP, I modeled it for Ireland 20 years ago, and at the time I was one of these naive believers. The economic sounds right. It must be good, right? I've got a lot more experience of the practical realities now. We've done a study. We got some people from Ercot as guest speakers to speak to our study group. The network planning in Ercot is done using traditional network planning. The network planning is not done using LMPs. They're too volatile and ephemeral, so their network planning is in exactly the same way as ours'. Although there has been theory about merchant investment in transmission, it is nothing more than theory. It has never happened, it's never been practical. So I don't think there would be any improvements on the network planning and network build process arising from LMP National Grid ESO today sees congestion as a cost. Actually under an LMP world congestion would be a financial cash flow, which would allow them to reduce network charges. So you can argue about their incentives, but I would say they already have pretty strong financial incentives to avoid congestion under today's national market. If you look at the invest incentives to invest for, let's say generators or storage units or demand, yes, you can have some short term based incentives based on prices, but again, if they're so volatile and unpredictable that you can't. Really use them as an investment signal. What's the point? And I think that's a step change between a Zal system where things might be reasonably predictable to a nodal system where I simply don't believe that people would have perfect foresight of those nodal prices. The real issue then is whether the investment signals would be more stable and more usable under that world compared to today's status quo in which we have. Zonal loss factors, which are plus minus up to about 5% of your gross revenue. And also very strong locational to incentives. So there are other ways of providing those incentives over investment timeframes. And I think I would argue that LMP, at least at the nodal level, perhaps isn't a great way of getting those incentives to people who want to connect.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 10:16

Stephen, you mentioned Ercot, and that's a US based grid, isn't it? Is that where they run the nodal pricing system there? Is it Texas? Isn't it if I'm not completely wrong.

Stephen Woodhouse, Director Afry Management Consulting: 10:25

Yeah. Ercot is Texas. There are nodal pricing systems run in various markets, so they tend to be us or US influenced. So you've got. California under Kaiso. You've got Texas, you've got PJM, Pennsylvania, Jersey, Maryland. I think New England is nodal pricing. At one point there was a proposal for a standard market design in the US, which eventually got withdrawn, but that would've implemented nodal pricing. For me, it's version 1.0 of the market, so New Zealand does it as well. Singapore does it as well. Singapore does nodal real time pricing and that. I can see advantages of that Ercot and PJM do day ahead and real time. And the trouble is with the day ahead, you've gotta find ways of dealing with all of the complicated things like start costs that you can't really allocate to an individual time period. And they don't price them properly. They bung them under the table as make old payments and they don't go into the market prices. And they have a day ahead and a real time market and nothing in the middle. So there is no intraday though. If you look on the discussion of those markets, you will find that they admit freely that they don't value flexibility. There was a discussion which turned into a mini war about PJM. Allowing fast starting plants to set price to allow some more value for flexibility, and it ended up in a dispute and a federal ruling over whether the definition of a fast start is one hour or two hours. So you just end up codifying a lot of the price setting into algebra and parameters that people tussle over. It doesn't seem to me that this is a way forward. It seems to me that version 1.0 of a market design is what we have here in England and Wales. From 1990 where you use the same optimization, calculate some prices for it, and the grid operator dispatches everything. And we went to version 2.0, which we took from Norway, which was is where we have free price formation and people take control of their own assets in their own balance position. America went from 1.0 to 1.1, where they make more sophisticated the algorithms. Still retain central control, and I think we've moved way beyond that.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 12:40

I sense in a way here that the nodal pricing system isn't fit for the purpose that we have in place at the moment. Phil, would that be a valid assumption to make here?

Phil Hewitt, Director, Montel EnAppSys: 12:48

So nodal pricing down at the transmission node level, or maybe even further down to distribution network, requires central dispatch, and that means that you no longer have an intraday market. And currently at the moment, the intraday market is quite active in Britain and it's a good source of revenue for flexibility. Which means that the market kind of delivers a lot of the flexibility that National Grid needs close to real time. If you then move to a central dispatch system. You can't have an intraday market because you've got 300 mini markets maybe, and those mini markets, there's not enough liquidity for people to trade on them, which means that the intraday market disappears and the responsibility for everything rests on the system operator. So that's a big risk for National Grid to move away from having a crowdsource solution to the flexibility problem where it might get it wrong sometimes, but the market gets it wrong, but punishes itself to them being completely in control. So it's quite a high risk strategy. You can have some degree of intraday market, but they're nothing like what we recognize in Europe. So in Europe we have market time-based interstate markets. You might have an hourly market and a quarter hourly market that runs all the time and people are buying, selling power in their price zone, which might be Germany or it might be Netherlands, or it might be one of the five price zones in Norway. People are buying, selling power in that zone. What happens with a a local an LMP market is that. Is that you have to bunch a load of nodes together to create something called a hub, but those intraday markets actually run primarily as trading for base load. So basically you are just trading a daily price. So there's no like granularity down to that individual small unit, which means it's a lot less flexible and it's not what we would recognize in Europe where we have. Quite an active intraday market that solves a lot of the problems for the grid.

Stephen Woodhouse, Director Afry Management Consulting: 14:38

I'd like to pick up on that theme. So even in the centralized markets, not everything is centrally dispatched, but the whole basis of the market is that you get critical mass and enough is centrally dispatched that you're setting prices. The assumption that everything else doesn't really matter very much. So you should be dispatching. The things that are at least are at the point of where you set prices. Now that might make sense in a world where you're dispatching a load of big generators, although there are none of those markets actually deals properly with startup and no load costs. So I might even dispute that. But there is no market that deals with lots and lots of small resources. I think none of those markets correctly optimizes the types of decentralized resource that we're looking for in the future, including batteries and demand side management. They simply don't have the parameters in the optimization to deal with those assets over the timeframes that the decisions need to be taken. And even the idea of aggregation starts to break down if you need to bid individually every one of the nodes on the transmission network, and there are 10,000 in California and 4,000 in Ercot. So I think the capability of those markets to deal with the range of resource types and the number of different assets, I think is highly questionable. And I would echo Phil's point that it would take a while to move to that kind of system. The basis for locational pricing starts to break down as the grid build catches up with the location of the generation. We've got quite an aggressive grid build program. Coming forward and hopefully, implementation of the results of the Windsor Review and the latest discussions on connections will help the grid to catch up with the location of the generation and the importance of congestion will fall. So by the time you implement this stuff, it's probably gonna be too late for it to have any real value, and it will deter investment because of the risk that it puts in place. So I think a nodal market is not the right answer. Now, I saw a comment on a webpage a chatter on some of the work we've put out, and it said the right time to have done this was 20 years ago. And I think I would agree with that.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 16:46

So if a, then a nodal pricing model is not fit for purpose in, a decentralized world where we're moving to decarbonize quite rapidly. Phil, what about a zonal model? You mentioned the five zones in Norway, Sweden has several Italy, et cetera, et cetera. There's a big debate in Germany where you have the similar, where you have the demand center in the south and the generation mainly in the north. Is that a model that could be applicable for the uk?

Phil Hewitt, Director, Montel EnAppSys: 17:11

It could be, a model that's applicable and it might be where we move towards, 'cause it's not as big a change. You'd have to design your zones to be big enough to be able to provide some kind of intraday markets that the markets to could still participate. And you're probably looking at four or five zones. And that would then solve the, reduce the congestion issue potentially to a degree. Although I can sometimes say energy markets like bouncy castles you push down on one part of the bouncy castle to fix a problem and it, and the cashflow pops up somewhere else. So you chase the cash flows around the market, but you as a ZR market still gives you those signals. It's provides signals in the Nordics for, building additional transmission, but I'm not sure exactly how much additional transmission is being built in response to those deltas between the different price zones. So it is a kind of form of locational pricing that maintains a lot of the features of a national market. You make smaller zones and those provide you still with the flexibility in the intraday market. It's not a massive seismic change. You could probably do it quite quickly. It's not such a massive change that it would frighten investment and stall investment to a degree. You could probably model it quite easily. One of the problems with locational marginal pricing is an for energy model, is to model it. It requires a hell of a lot of computation. To actually model all these different zones and then model different build out of everything. So it's a lot easier to just model five zones versus 300 nodes. It'd probably be easier to make a wise investment decision as well to build new generation in the, in a zal market. Versus a locational market where it's fairly difficult to model going forward. So it's, it should be less damaging in terms of delaying investment, which is what we need right now. We need a lot of generation to be built.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 18:59

Stephen what's your view?

Stephen Woodhouse, Director Afry Management Consulting: 19:00

I would echo some of the comments on modeling LMP. If you go to nod. I did a nodal pricing study for Ireland, 254 nodes in 2002, and we delivered the results in six weeks. And it felt easy, mainly because I didn't understand the realities of some of the things. Then we've just done an LMP study. It was intended to take three months and it took nine. And one of the issues now with any kind of location, and this is, this applies to zonal as well. You are not just looking at optimizing where people put the next generator. And then the grid, you're also optimizing where the hydrogen infrastructure and the CCS infrastructure is where the offshore connections come in. So there are many more degrees of freedom and the ability to model and predict is much, much harder than it was when these markets were developed originally, where it's. Primarily about where you put, as I said, the next generator. That's a problem for a zonal market to some degree as well. I still have the comments ringing in my ear from one of our investors who was a member of our study saying, I don't care if it's nodal or zonal. The price in Scotland will always be zero under either system, so the merchant tailor my wind farms is worthless and I could never build another plant there again. Now that is overstating the case. There are many issues that are the same under a zonal and a nodal system, but I think many issues are also different. A zonal model is what we were in while we were part of the internal market for energy. It just happened that we had one zone for Britain, so a zonal market would actually give you a reentry. It would still leave the door open mechanically to a reentry to the internal market if we choose to go down that route, which I think a nodal market wouldn't. We are much more nuanced about the benefits of going zonal rather than staying national. If you go zonal, you do get some efficiency benefits in dispatch, and a lot of those are around how you handle transmission constraints. It's inefficient to deal with transmission constraints under today's. Balancing arrangements so that we found that there would be some efficiency gains. But the difficulty with that is you would place a lot of risks on investors based on their location, and there would be very big wealth transfers, and there would also be a need to hedge those risks in the future. Now, if you went zonal, unless you dealt with those risks, I think you could end up adding more cost to the system through. Increased hurdle rates and cost of capital than the efficiency gains. And critically where transmission rights exist, they last one or two years and they're base load. Do you know anybody who deals base load anymore? It might be the trading product, but it doesn't reflect any kind of reality. And a really big example is if you wanna strike a renewable PPA. Between a producer and a consumer, they will not be at the same location. They will need a locational hedge for that energy that follows the production profile of that energy. So I think there needs to be some development around the nature of the products we trade, not just. The energy products, but also the point to point products, the transmission rights before we could be convinced that a move to a zal market would be a step forward. But there are problems with today's national market as well, about the efficiency of dispatch, particularly around the constraints and the units that they can't efficiently dispatch close to real time because they've got log lead times and interconnectors are an example, as I've said. So there would need to be changes made either to design a bizo market or today's market designed for the national market to deal with this future world in which congestion is gonna be much more of a reality for at least a period until the grid catches up.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 22:42

Stephen, so what are the next steps? This was apart proposals in a bigger set of, looking at the electricity market arrangements in the uk. Do you expect it to put forward some measures that include some degree of zonal or even nodal pricing?

Stephen Woodhouse, Director Afry Management Consulting: 22:58

All I know is informal and maybe proved to be wildly incorrect. We're expecting the government to issue the next round in his consultation series on Rema before Christmas. And I have had confirmation recently that is gonna be before Christmas, whereas I was hearing noises that it might be afterwards. So I had also heard from them that one of the most difficult things to decide on is this issue of locational pricing. That they probably won't. We are likely to see the distinction between a national and a zonal market. There's no decision on that. It's possible they may rule out a move to a nodal market, but they certainly won't close the conversation complete. What I am a little concerned about is that they might leave the central dispatch, the central market design on the table, and I can't help thinking that a zal. Central market would be the worst of both worlds 'cause it's the centralization, which are part of my serious objections to the move to nodal. But I do think we if we were to retain a azo, a national market, as I've said, there will need to be some enhancements to the way that we achieve efficient dispatch, particularly around congestion. So I'm hoping to see something there, and it might involve looking at access rights. Again, we chose a connect and manage model. 10 years ago, which allows generators the right to connect to the transmission system ahead of the grid and get full firm access. And maybe looking at how those access rights are awarded and whether there's any trading mechanism we could put in place. That might be part of the answer. But if we move Zal, I would say it is beholden on the government to say that they would have a way of dealing with these risks. The ability to trade away your locational risk. If you leave zonal on the table and don't deal with that issue, you run the risk that investors just stop until it's been resolved. So I think if you're gonna keep zonal on, on the table, you need to deal with the way that these trading risks, the, these rights and the trading issues can be dealt with by investors.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 25:04

And Phil, what are your expectations in the next rounds of the consultation and what the government could propose?

Phil Hewitt, Director, Montel EnAppSys: 25:10

So I, I watched the individual rounds coming out to make, and you can see the noise on LinkedIn and other forums where you chat to people over a cup of coffee and there was a, ah, modal pricing is still in at the last iteration, but, so there was surprise it was still in as if it's almost considered to be not going forward because there has been quite a backlash against it in the market. I can't see nodal surviving, but I think zonal potentially is a way forward because that's seen across Europe as redesigning bidding zones as a way of. Addressing some of these issues or at grid constraints, et cetera. So I can see Zal probably staying in there. And then you then get into the detail and nitty gritty about how you implement that, because you probably have to grandfather the rights and things and come up with complicated mechanisms to allow. Units that were built in the national market to not lose out as you move to a zone market to try and not scare investment away.'cause you don't want to change the market to a point where people then say, I don't know if you're going to do move from this current market model to a new one in 10 years time. Which leaves my assets stranded.'cause it does take a long time to build these renewable assets. So it requires a long commitment to build them. So you would have to say you'll be okay if we move to a new market model for the existing generators, there'll be some form of grandfathering..

Stephen Woodhouse, Director Afry Management Consulting: 26:29

Richard, it might be worth just adding a bit there. So what does grandfathering mean for a plant with a CFD? So I'm a wind farm in Scotland with A CFD against the national price today. If we move to zonal or indeed a nodal price, the expectation I think this is written into the CFD, is that your reference price would then become the local price. So sounds good, right? No, because you only get paid if you generate the difference between a national price With re-dispatch. And where you get constrained down at a price you choose under a Z or a nodal market, the price in your area goes down and down until you self curtail so you're not producing anymore. The CFDs don't protect you when you don't produce, and they don't protect you when the price is zero, if you look at the latest CFDs. So it would be quite hard even for existing CFD holders to get a full protection. Against a move to a zal or a nodal market. So I think a huge amount of work is gonna have to be done on some of those tricky issues and just saying, oh, FTRs will solve it. It absolutely will not.

Snjólfur Richard Sverrisson, Editor-in-Chief, Montel: 27:38

No, all I can say to round up, I think it's an extremely complex issue, and I think a lot of work needs to be done to work on the issues, the concerns that you raise and the problems that you highlight. But gentlemen, thank you very much for being guests on the Montel Weekly podcast this week.