Financing the electricity system: Part 1 Back to the Future?

Interested readers may have noticed a recent notice that energy suppliers will be subject to a levy of £3.45 per MWh of electricity supplied from November 1^st onwards to pre-fund the development of the Sizewell C nuclear power plant. This is linked to the decision to adopt what is called the Regulatory Asset Base (RAB) arrangement to fund this project. Dieter Helm was the most prominent advocate of this arrangement for electricity projects – particularly nuclear plants – with high capital costs, lengthy development periods, and substantial uncertainty about future construction costs.

In narrow terms, most of his arguments in favour of the arrangement are correct. However, it highlights the complete shambles that financial arrangements for the electricity sector have become because of the adoption of piecemeal and sometimes inconsistent arrangements to solve specific policy goals.

The outcome resembles the tangled Schleswig-Holstein question which preoccupied mid-19^th century European diplomacy. The eminent Victorian statesman Lord Palmerston was alleged to have said that only 3 people had ever understood the issue, of whom one (the Prince Consort) was dead, one (a Danish diplomat or a German professor) had gone mad, and he himself who forgotten all about it. What most people who refer to this quip forget is that the question was settled after the Second Schleswig War in 1864 when Prussia annexed a large swathe of formerly Danish territory including the Duchies of Schleswig and Holstein. A similarly brutal solution may be required to establish a more efficient and less cumbersome set of arrangements to finance investment in the electricity sector from the 2030s onward.

The founding myth of the current trading system for the British electricity system emerged gradually over nearly 15 years following the breakup of the CEGB and the privatisation of distribution companies in 1990. The restructuring involved was chaotic and painful. It included compulsory divestment of generating assets by the two main power generators – National Power and Powergen – to reduce their monopoly power in pool trading. There were multiple mergers and demergers of generating and supply businesses, major bankruptcies, and the infamous “dash for gas” by which coal generation was displaced by heavy investment in gas combined cycle plants (CCGTs).

The core element of this founding myth is that investment in new generating capacity would be driven by market incentives provided by wholesale market prices. This is a model of what is usually called merchant generation that is contrasted with the regulatory investment model that is standard in the water sector. Under the regulatory investment model, at intermittent intervals regulated companies propose investment programs to meet a variety of service goals. The regulator decides which elements of these investment programs will be funded in the sense that the regulated companies will be permitted to recover their costs through their regulated charges. In US terminology this is known as inclusion in the rate base, while in the UK we refer to additions to the regulatory asset base.

Like most founding myths, the merchant generation model was a product of specific circumstances and technical options. Even as it was nominally adopted, the government undermined the core framework because market prices could not sustain the growth in renewable generation required to meet EU targets that it had agreed to. An increasingly elaborate set of price premia was introduced to provide additional incentives to invest in a whole range of renewable generation technologies. Initially, the idea was that renewable generation would earn a single market-determined premium. That arrangement was rapidly abandoned under pressure from lobbyists for a range of renewable technologies who all argued the need for special treatment of their preferred options.

Two crucial factors underpinned the merchant generation model. First, in the 1980s the CEGB and SSEB completed the construction of the program of AGR nuclear plants and commenced construction of Sizewell B. In addition, by agreeing to implement the EU’s Large Combustion Plants Directive, the government had ensured that most coal plants would either close or operate for a restricted number of hours from the early 2000s. Therefore, after privatisation there was a clear and large requirement for new generating capacity that could operate either on baseload or as flexible mid-merit generation. Second, improvements in the performance of gas turbines and heat recovery units from manufacturers in the US, Europe and Japan as well ample supplies of North Sea gas offered the option of building CCGT plants quickly (less than 3 years) and at a low cost per MW of capacity.

The result was the “dash for gas” – the construction of 12.5 GW of CCGT capacity in the 5 years from 1992 to 1996 – plus a further 9.4 GW completed from 1998 to 2002. Few, if any, of these plants were conceived as merchant generators. The orders were placed from the last 1980s to the mid-1990s under completely different market conditions. Several CCGTs completed between 1992 and 2002 were either mothballed or reduced in capacity after no more than 20 years of operation. Changes in environmental requirements and grid constraints meant that the investments required to refurbish plants could not be justified.

The total amount of new CCGT capacity built under the merchant generation model has been 10.3 GW. Only 1720 MW (Carrington and Keadby 2) of that total was completed after 2013. The investment decision for Carrington was taken in 2009. So, for practical purposes the merchant generation model functioned as envisaged by the founding myth for little more than 5 years. The only new CCGT built in the last decade was Keadby 2, which was commissioned in 2023.[1]

Of course, idealised versions of how markets could or should operate are not unusual, but the collision between myth and reality came very quickly in the case of the merchant generation model. The reason lies in the nature of the alternative forms of generation that policymakers and lobbyists wanted.

Most forms of renewable generation are highly capital intensive with low operating costs. Unfortunately for potential investors, the average margin between expected wholesale prices and their operating costs was insufficient to recover their capital costs and earn a reasonable risk-adjusted rate of return. Hence, a guaranteed price premium was required to secure investment in renewable capacity. Over a period of a decade that turned into a system with an ever-changing collection of price premia for different technologies and generators.

Even the nearest equivalent to conventional coal-fired generation - burning wood rebranded as “renewable” biomass combustion – was unable to cover the costs of converting former coal plants without substantial support. Partly, this was due to the costs of maintaining adequate supplies of fuel, and partly because biomass is not well-suited for mid-merit operation involving regular warm-up and cool-down periods.

The priority given to increasing renewable generation undermined the merchant generation model for reasons that are more basic than just the costs involved. Since the dominant forms of renewable generation are intermittent rather than dispatchable like gas, coal and even nuclear, they have the effect of increasing the variability of market prices as the share of intermittent generation in total supply increases. That variability spills over to increase the risks of investing in any kind of dispatchable generation.

Greater variability in market prices means that gas plants must cover their capital costs from profits made in a small number of hours in the year – in effect from occasional price spikes. However, for both policymakers and the public this is seen as price-gouging, leading to strong pressures to impose either price caps or taxes on “excess” profits. Both groups tend to rely upon a mental framework in which cost-plus is the only “fair” way of setting prices, but that is fundamentally antithetical to markets in which trading profits are random and extremely variable.

The combination of lower operating hours for dispatchable generation and greater variability of market prices due to greater reliance on intermittent renewables also means that gas plants are forced to rely on short-term trading to buy gas to run their plants. This reverses the conventional logic that gas prices determine electricity prices. The reality is that high electricity prices – due to low levels of intermittent generation – drive up demand for gas and thus wholesale prices for gas. Storage of gas is easier and cheaper than storage of electricity, but it is still expensive. Consequently, gas prices tend to vary less in the short term than electricity prices, but everyone must pay the cost of maintaining buffer supplies of gas to smooth the intermittency of renewable generation.

While the compression of the margins earned by gas plants due to the increasing reliance upon intermittent renewable generation has been gradual, it was predictable for more than a decade. Thus, potential investors in new gas CCGTs could foresee that they were unlikely to earn adequate margins to cover the capital cost of building new gas plants, unless they could rely on revenues on top of what they could earn from merchant generation.

This logic led to the introduction of capacity payments to reward generators able to supply “firm” rather than intermittent capacity. Unfortunately, the mechanism adopted to award capacity contracts has strongly favoured the extension of the lives of existing CCGTs rather than the construction of new plants which would operate up to the 2040s.

The evolution of the electricity market over the last two decades is a story in which the founding myth of market-driven investment in generating capacity has been systematically sabotaged by the pursuit of the policy goal of promoting intermittent renewable generation. This broad trend has been reinforced by an ever-changing collection of special incentives and regulatory interventions so that, as suggested above, the whole structure of policies has become so complex that almost no-one understands either its extent or its impacts. The current government has shown every intention of making the system more complex by adopting special incentives for almost every major project – from new nuclear plants to carbon capture.

Stepping back from the details, it is obvious that all attempts at “reform” – for example, changing the terms of CfD contracts for each Allocation Round – merely make the overall system more complex and less efficient. As with the Schleswig-Holstein Question, the only solution may be a brutal restart rather than ineffective tinkering.

This is where my title of “Back to the Future” comes in. What successive governments have done is to recreate a restructured version of the CEGB. Generation, transmission and system operation are effectively controlled by a combination of political and regulatory intervention. The only major difference is the existence of a wholesale market for trading electricity, but this is little more than a veneer on a government-controlled set of incentives for investment.

The current system is baroque and inefficient. However, the legacy of incentives created to favour investment in intermittent renewable generation is inescapable. The only way of cleaning up the mess that has been created over nearly two decades is to acknowledge the reality of central control over generation. The aim should be to create a simpler regime based on capacity and dispatch payments. In this way, the legacy of commitments to existing generators can be run-off in an efficient manner, while ensuring that investments in generation for the 2040s are based, as far as possible, on reasonable economic criteria rather than ideology.

In subsequent articles in this series I will try to sketch the considerations should shape the transition as well as future investment decisions.

[1] Edit: I am grateful to the commenter who pointed out that in the original version of this article I had omitted the Keadby 2 CCGT (840 MW) in Lincolnshire that was commissioned by SSE in 2023. (I had used an outdated list of active power plants.) The core point about the dearth of investment in major CCGTs over the last 15 years remains valid.