Here we go again. European energy markets are reeling from another gas shock, and politicians, consumer groups and many others are asking again why gas sets the price of electricity, when it now accounts for less a third of our annual generation (down from about 40% in the last crisis, in 2022).
Back in Autumn 2022, the then-conservative government introduced “…new powers to help sever the link between high global gas prices and the cost of low-carbon electricity…” – though this took the form of an expensive ‘Cost-Plus Revenue Limit’, rather than actual structural reform. Three and a half years later, the Secretary of State Ed Miliband has announced the same objective, caveated with recognition that it “will not be simple.”
As observed in a recent deconstruction of Britain’s long-running energy price crisis, the fact that competitive electricity markets result in wholesale electricity prices being set almost entirely by fossil-fuel generators is “known and natural to most energy economists, but raises disbelief or even outrage to many others”, given its minority role in overall generation.
This Commentary is intended for those interested to understand those radically divergent views, but has a particular aim to clarify some conceptual issues – including for economists to whom the centrality of marginal cost pricing seems intuitively obvious. It focuses on one of the main categories of proposals (not the only one) for structural reform to reduce or avoid the conflation of gas generation costs with renewables output – namely the concept of Green Power Pool(s) (GPP). This proposed arrangements to aggregate renewables through an electricity pool mechanism (potentially analogous to the overall electricity pool used in the 1990s), available directly to consumers at prices reflecting their long-run costs.
The intellectual foundations for GPP approaches were laid out in a series of four Working Papers. The first two outlined the empirics of what drives wholesale electricity prices across European power systems (WP1), and analysis of the profits accruing to both renewables and gas generators in the UK during the 2022 crisis (WP2) (both now confirmed with more recent studies). The third and fourth Working Papers summarised the economic principles, underpinning the current wholesale market and its limitations (WP3), as backdrop to a specific illustrative option of how a GPP based on CfD-backed renewables could work (WP4).
One essential point to emphasise is that the proposal retained a key role for the existing wholesale market, with the GPP system buying from or selling to it, at times of shortfall or surplus respectively. The ‘scarcity price signalling’ would thus be retained, but only for the volume of backup required when renewables output was insufficient to meet the demand on the pool – not on all its electricity consumption. It would also thereby give clarity to the cost of such backup – not necessarily welcome by renewable energy generators, but I think important for transparency as the share of renewables rises. We used the term ‘dual markets’, given this interaction.
‘Dual markets’: a Gedanken experiment
This Commentary does not repeat these analyses, or the proposal as also summarised in a 2023 OEF article Disentangling the debate on electricity market (re)design and ‘split markets’. It considers briefly some of the specific objections, but first I focus on an underlying conceptual issue: the apparent inherent aversion to the idea that the market should be ‘split’, even with subsequent interaction, which to most energy economists seems intrinsically seems inefficient.
There are two basic responses to this instinctive reaction.
The first is to observe that the market is already split, and quite profoundly. It is split between consumers and generators, particularly regarding the renewables on contracts-for-difference (CfD contracts). The arguments for these CfDs – which guarantee a fixed ‘strike price’ price for output sold – are well understood and well justified, in both theory and practice. However this means that the wholesale market is increasingly disconnected from the cost of renewables. Indeed the relationship is partly inverse: the higher the strike price, the more that generators will bid low-to-zero prices in the day-ahead market (or indeed, negative prices in the UK electricity short-term balancing market). This is because get ultimately get the strike price as long as they can find a buyer anywhere, anyhow, as a top-up if needed to the difference between that and the market price. The disconnect between renewables costs and wholesale market will rapidly get bigger as more CfD-backed renewables come online.
Moreover, most consumers cannot access renewables at the cost of those CfDs. Whether to split the market that way; whether instead to rely on direct contracts (Power Purchase Agreements, see below) for driving renewables investment; or whether to organise the system so that consumers can reconnect with the cost and characteristics of renewables in aggregate, is a legitimate public policy debate, not one to be dismissed with rhetoric about ‘splitting markets’.
The second response is to delve more deeply into what actually seems like the fundamental issue. Given the vested mindsets in the current way of organising electricity, consider an analogy, in the form of what Einstein called a “Gedanken experiment” – a “What If …” experiment, to challenge and clarify the thinking.
If the underlying presumption is that it is inefficient to split a market of the ‘same product’ (e.g. electrons), the question is – what defines a ‘same product’ – what makes products alike? So, for example there are plenty of pears and apples that look almost indistinguishable and have quite similar nutritional values – to the extent, indeed, that hybrid pearapple fruit do exist. What if we considered pears and apples to be so similar as to count as ‘like products’?
To be more precise, suppose pears and apples could only reach the market through a network (say, deciduous fruit pipes) that form a natural monopoly, and hence had to be regulated. And it was then observed, as economists did for electricity in the 1980s, that it would still be possible to have competition in production (planting trees and picking the fruit) and consumption (purchase at the fruit market); and designed a unified PearApple wholesale market? Both pear and apple growers would submit offers to sell, and fruit markets would bid to buy, for the next-day delivery, with the market clearing for all at the price of the most expensive fruit required to meet the overall demand.
Would most agrarian economists then be outraged at the idea of ‘splitting’ this PearApple market between pears and apples?
To pursue this Gedanken experiment further, what if pears were predominantly grown in the fertile soils of south-eastern Europe, whilst apples grew better elsewhere. If a war in Ukraine curtailed supplies of pears, would we accept it as a natural consequence that apples should also become far more expensive? Would we accept the argument that this was a market working as it should – signalling PearApple scarcity and providing apple-growers with large windfall profits (more technically, unexpected inframarginal rents) as a signal to plant more apple trees?
This proposition sounds absurd, but its purpose is to highlight that part of untangling the debate is to be clear about what we mean by a ‘like product’ that we expect to be sold close to a common price, through a single integrated ‘wholesale market’. That, in part, is a matter of social choice and legal definition (which has indeed tormented the WTO trading regime over many years).
Some misunderstandings
Our Green Power Pool proposal was considered in depth by the government Review of Electricity Market Arrangements. REMA’s second Consultation document (March 2024) acknowledged that this “could be workable”, but rejected it on the grounds that it would be too disruptive to investment, had not been tried anywhere before, and that it would not really be needed because gas would soon be reduced to a trivial part of the system.
Concerning the last point – relevance – quite aside from the current crisis, the issue will not resolve itself any time soon. The growth of renewable energy remains impressive but no analysis now projects that the role of gas in setting the electricity price will become trivial, even within the next decade: projections of declining average gas generation have only a very weak relationship to how often gas sets the price. Indeed that is the crux of the issue: the wholesale market is designed around the economics of precisely the kind (gas) that the government is trying to phase out.
Projections by the National Energy System Operator (NESO), the UK Energy Research Centre, and indeed our own Centre using NESO scenarios suggest that with the current market structure, gas will continue to drive the electricity price more than half the time still in 2030 even as its overall contribution continues to fall. If demand growth outstrips the deployment of other flexible generation assets, that will continue to be the case through much of the 2030s, even if and as its share of total generation shrinks well below 10%.
The concern that green power pool approaches would necessarily disrupt renewables investment also seems to reflect misunderstanding. The proposition was primarily about the arrangements through which the cost of renewable generation reaches consumers – a generic question, in principle quite separate from the contractual terms on which renewables investment itself is financed. Nothing in our GPP proposals necessitated changing the terms of CfDs, for example, though it could potentially enable them to be simplified – and rendered more visible.
The motivation for the proposal was not primarily a desire to separate ‘green’, from ‘fossil’; nor even, as a way to reduce energy prices. It is reflection of the fact (as stressed in our Working Papers #3 & #4) that there are fundamental, structural differences between the economics of gas-based generation on the one hand, and wind and PV generation – our biggest renewable sources – on the other. The former can run ‘on demand’, with costs dominated by fuel and carbon emissions (underpriced in most markets) and highly exposed to geopolitical uncertainties. PV and wind in contrast are ‘as available’, with costs dominated by capital investment, with output determined by mostly domestic, free but ‘as available’ energy from the fluctuating sun and winds, with strong locational diversity in terms of output characteristics and value to the system.
The Oxford Institute (Keay and Robinson 2017) was right to draw a functional dividing line between these two categories of ‘as available’ and ‘on demand’ sources. That distinction is further underlined by a recent contribution (Vergés-Jaime 2025) which critiques the theoretical foundations of marginal cost pricing in electricity markets, identifying four ways in which the classical theory does not match up to the realities of the sector and is thus inappropriate for electricity. The furore around market design is a consequence of this, being hugely aggravated by the fact that gas markets are “a kind of international oligopoly .. creating instability and subject to speculative swings”, so that gas price shocks do not actually reflect any change in the real cost of supply but rather reflects “what the captive-depending buyers have to bear at the time (speculative suppliers’ prices)”. Consequently Vergés-Jaime concludes “there is not any rationale for a pricing design dooming (electricity market players) to amplify the ensuing speculative effect coming from MWh of gas power plants, to [all electricity] traded in a session.”
In other words, with the wild price swings of natural gas, and given the timescales of building most forms of energy infrastructure, there no economic rationale for saying that the gas price at any particular moment is an efficient incentive for low carbon development – any more than we would expect a poor year for pears to dictate the planting of apple trees. Obviously in reality, electricity has unique properties and the analogy is highly imperfect, but this fundamental point remains valid.
Realities of the current market and its theoretical underpinnings
At this point, those familiar with our electricity market would (and should) doubtless jump up and say that the idea of one unique wholesale electricity with just one price is a myth: the UK and EU electricity markets are in fact unfettered, with a whole range of options for buyers and sellers, from the intraday market to forward and futures contracts, and bilateral Power Purchase Agreements (PPAs) between major generators and big buyers.
Despite the plethora of markets and contracts, most are so time-limited as to be largely irrelevant to the question of smoothing major price shocks, with most being limited to a year or two ahead, typically designed to hedge against short-term fluctuations and secure purchases across seasons. Moreover, the price of forward and futures contracts typically rapidly adjusts to reflect the current, gas-driven market sentiment.
The partial exception is bilateral Power Purchase Agreements (PPAs) between major generators and big buyers, and in particular, many point to the role of Corporate Power Purchase Agreements (CPPAs).
This is important (even neglecting the fact that many such CPPAs are still partially pegged to the day ahead price) – but partly because the analogy with pears and applies can be taken to almost invert the way the question is framed. For CPPAs are, predominantly, contracts signed between single companies, and specific generating plants. The analogy is not that I am free to go to the fruit market and buy apples irrespective of the price of pears; it is that to insulate myself from shocks to pear growers, I need to sign a long-term contract to buy apples from a particular orchard, for the next several years.
Put that way, it seems unlikely that the CPPA market can be a sufficient answer to the problems with electricity markets. CPPAs involve large transaction costs (typically taking a year or two to negotiate) and lose both the physical diversity and risk-spreading benefits of aggregating output from different sites, and indeed, across a variety of different consumers.
The core question then is actually about what scope of aggregation is appropriate, and how it could best be achieved. In my view, these are now some of the most fundamental questions in electricity markets.
But … don’t CfDs solve the problem anyway?
At this point, experts on UK renewables could reasonably jump up with a second objection to the argument, namely that in fact consumers do gain the benefit of price stability, from those renewables on CfDs.
Yes; and no. The current mechanism involves generators with CfD contracts bidding in to the market, and then settling books with the Low Carbon Contracts Company. In three-month sequential windows, this contractually tops up their revenues to reflect their strike price when there is a shortfall, with the cost added as a charge on all bills. Conversely, when the wholesale market has delivered prices above their strike price – the contracts return the surplus revenues to suppliers; the revenues thus returned are supposed to then be passed through to consumers, though this is only guaranteed for those on the standard tariff regulated under Ofgem’s price cap.
Whilst much ‘better than nothing’, this has multiple limitations. Being added to bills in general, the top-up payments are somewhat regressive. Payments back to suppliers are also belated, with the actual transfer payments coming some months later. And they are largely invisible; the positive returns only arise, a few months after a gas price surge, in the form of bills that have increased somewhat less than they would otherwise have done. CfDs provide no way for ordinary consumers to access these renewables at anything related to the strike price.
CfDs have been wonderful at securing renewable energy investment at low cost. But the subsequent financial flows – being regressive, belated, largely invisible, and inaccessible – do not seem like the best we could possibly do.
So …
The purpose of this Commentary has been to clarify, more than to propose. Reflecting on the issues, my own view on this, in barest outline, has developed to the conclusion that a mature electricity market design needs to reflect more directly the radically different characteristics of PV and wind compared to fossil fuel generation: otherwise, we are indeed inappropriately forcing ‘apples and pears’ together through our wholesale market.
There are three other important elements relevant to good solutions. First, the economics of renewables at present are defined by the kind of contracts they are on. In particular, generation supported by CfDs have fundamentally different properties from those which face the wholesale price, whether or not this is topped up by Renewable Obligation certificates. For the UK system it is entirely plausible to argue therefore that we have at least three main categories of domestic generation (gas, RO renewables, and CfD generation – which technically includes the Hinkley Point nuclear station), each with overall generating capacities through the late 2020s in the range 30 – 40GW.
Second, not all demand is the same. There are industrial, commercial and residential consumers with widely different time horizons, price sensitivities, and capabilities to adjust their demand (‘flex’). There are low and high consuming households, driven partly by poverty vs wealth, but also many other factors. Demand is also becoming more, not less, diverse, with the adoption of electric vehicles, heat pumps (including pioneering companies building heat networks with heat storage), and varied forms of industrial decarbonisation. Innovative supply companies have started tapping into this potential, but most consumers remain constrained in how they can access electricity – many on green tariffs abandoned them in the last crisis, when they discovered they were still mostly paying the price of gas generation (plus a ‘green’ addition). Except for the biggest companies, the wholesale market continues to offer the equivalent of undifferentiated pearapples, even to those wanting to buy purely renewable electricity.
Third, both generation and supply have significant, and growing, regional dimensions. After intense debate, the UK government in July 2025 decided against general locational pricing in the wholesale market, e.g, different prices in different zones. Yet the value of generation – and the system cost of meeting demand – varies with location, given transmission costs and constraints. Again, this will become more important as the output from renewables rises, and most economists are rightly concerned about the disjuncture.
Notably, the decision to retain a single national wholesale price does not change the physical reality that renewable output greatly exceeds demand in Scotland, with the converse in England, and with many divergences also at the level of the seventeen zones considered in REMA analysis. In no scenarios do the transmission and distribution network build-out expand to the point of making this irrelevant. The government’s promised to identify alternative ways to signal locational value, but nine months later we are still waiting for details of the ‘reformed national pricing’ proposals to emerge.
An interesting variant would be if green power pools could be developed in different zones, to reflect the value particularly of local renewables generation for consumers in that zone, and to incentivise better siting and use of storage and demand-side flexibility in relation to transmission constraints.
Who and how?
The key question then is whether, and if so how, the government, or the regulator Ofgem, have a role in disentangling the economics of gas and renewables, to facilitate better matching with diverse demands.
At one end there is the view that that there are no legal impediments to private companies aggregating electricity from any mix of sources, to sell on to consumers as they wish (albeit, constrained by the decision to retain a single national wholesale price, without any dynamic pricing of transmission use). At the opposite end, others are now arguing for direct government takeover of parts of the market, most prominently either:
- gas generation to be taken into a ‘strategic reserve’ dispatched centrally by the System Operator and financed as a Regulated Asset Base, proposed by Greenpeace and Stonehaven (2025)
- for the government to become the Single Buyer for the Renewable Obligation generators, to prevent these making further windfall profits (Brown 2026 for CommonWealth) potentially with some elements of the Greenpeace / Stonehaven proposal as well.
For the UK, these are radical proposals that must face the fact that direct government control of the energy sector, and of energy prices, has rarely gone well, for obvious reasons. Our own GPP proposal oriented towards an electricity pool based on private generation but shaped by government around the long-run contractual or marginal cost of key categories of renewables.
Another intermediate approach – with less direct government involvement but the closest operating analogy to a GPP – is the Green Power Trading platform developed by Beijing Power Exchange (a subsidiary of China State Grid). This has established a platform for physically-validated matching of bids and offers for renewable electricity; initially mostly short term, but with a standardised longer-term contract design now being launched.
There are thus multiple routes to consider (and important details still to resolve). The common theme that as the physical structure of our energy system involves growing dominance by renewables, there is a case to review and reform the market structure to fit – including to consider options that have been – somewhat pejoratively? – termed ‘splitting the market’.
************* Further information and References *************
The literature on electricity markets is of course now huge. Aside from the links in this Commentary, two Special Issues of the quarterly Oxford Energy Forum offer a really useful way in to a wide range of issues and viewpoints:
- Electricity prices and the energy transition: economic efficiency, social equity, and climate alignment, Oxford Energy Forum Issue 147, March 2026:
And
- Electricity market design during the Energy Transition and the Energy Crisis, Issue 136, May 2023
Core analysis of our own work on electricity markets can be found at our Centre Website Output and Impacts, notably the Working Papers series (the first four, being those produced in 2020/23 under the theme ‘Navigating the Energy and Climate Crises’), and also some of the Briefing Papers. For general queries please email cnzmd@ucl.ac.uk. My own journal, other and earlier publications are available at the Electricity Markets section of https://profmichaelgrubb.com/publications/?filter=markets.
Other (non-exhaustive) references to recent proposals on ways of ‘splitting’ the electricity market are as linked in the text.