News & insights

By Lewis Heather & Attila Hajos

The UK government launched its much-anticipated review of the electricity capacity market (CM), which was introduced in 2014 to quell fears of inadequate supply capacity. CEPA sets out some of the key questions that stakeholders should be asking as part of this review in a Capacity market briefing note.

Is a capacity market necessary to attract sufficient investment in new generation assets?

One of the main questions in BEIS’s Capacity Market and Emissions Performance Standard Review is to assess whether the CM is still needed in the future. One way to approach this is whether the concerns that prompted the Electricity Market Reform (EMR)—lack of investment in new capacity and resulting threats to security of supply—would still be present today in the absence of the CM. In other words, whether an energy-only market (EOM) without capacity payments could incentivize enough new investment to ensure security of supply. BEIS’s working assumption is that the CM will be needed in the future, implying that BEIS considers that going back to an EOM would not ensure capacity adequacy.

The questions posed in this review have been discussed in the context of an ongoing global debate about the ability of EOMs versus markets with CMs to ensure security of supply. Some argue that EOMs in practice cannot ensure supply adequacy because they rely on hard-to-predict scarcity periods with high prices to incentivise new investments. Even if such prices were high enough to recoup capital costs, investors often find such revenue streams too uncertain and choose not to invest instead. This is even more pronounced when investors have the opportunity to invest in multiple markets and CM markets are available. A recent news story has highlighted investor perceptions of EOMs vs CMs in two US markets: an energy-only market (ERCOT) and a wholesale market with CM (PJM). (1) Most investors prefer to invest in a market with a CM even when forward electricity prices are less than one fifth of those in the EOM. This is not surprising, given that CMs provide guaranteed revenue streams, but also suggests that investors require much higher rates of return in EOMs due to the volatility and uncertainty of the revenue streams in such markets.

Is the capacity market really technology-neutral? If not, how can it become so?

While the GB CM is touted as a technology-neutral mechanism—where existing and new generating capacity, demand-side response (DSR) and interconnectors can compete on a level playing field—in the early phases, separate Transitional Arrangements auctions have been held targeted at DSR. In addition, interconnectors could not participate in the first T-4 auction. While these have been explicit design features of the CM that will not continue, unintended consequences of other polices have created distortions that tilted the playing field. Most notable are the embedded benefits which provided distribution-connected generators with a cost-advantage of about £50/kW-year. (2)

In its call for evidence, BEIS poses the question of whether “the CM needs to adapt in light of recent changes in the energy system and wider energy policy”, (3) citing participation by DSR, aggregators and smart system services. This raises a potential concern, since if the CM were truly technology-neutral, it would not need to be adapted as new technologies emerge.

BEIS notes that technology neutrality does not necessarily mean that each technology faces identical requirements, arguing that “sometimes differential treatment can be justified on technical grounds to achieve equal access to the CM and fair competition”. (4) It may be true that a level playing field is created by comparable but not necessarily equal treatment of the different resource types. For example, DSR may provide verifiable load reductions even if it does not satisfy the same metering requirements that generators have historically faced, so identical metering requirements could create an undue barrier to DSR participation. Equally important are any differences in incentives and penalties faced by the various resource types. The playing field may be tilted if some resource types received a more favourable treatment than others. When considering requirements, it is important to return to first principles to consider what the requirement is trying to achieve to understand its relevance from one technology to the next.

How can the capacity makret be structured to attract greater volumes of DSR to provide capacity more cost-effectively? Are there lessons from other markets?

Although the GB CM has attracted some DSR, including through aggregators, much of this has been through the temporary DSR-only Transitional Arrangements auctions where DSR did not face competition from other capacity types. Furthermore, the GB CM has been less successful than other comparable forward capacity markets in attracting new DSR in similar stages of market development.

In the capacity auctions held for the first five delivery years, the GB CM has proportionately attracted only about one third as much capacity as two long-established US capacity markets (PJM and ISO New England). The first capacity auction in the integrated single electricity market I-SEM (T-1) in Ireland attracted even more capacity: 7% of total capacity cleared in the auction was from DSR. So why does the GB CM seem to be less attractive to DSR?

One possible explanation is that GB capacity prices have been lower than in other markets, although some of those markets have seen DSR growth even with low capacity prices. Experience of other markets suggests that barriers to entry and the lack of a level playing field (both against and in favour of DSR) can have a significant impact on DSR participation.

BEIS has not identified the treatment of DSR as a priority issue for the CM review, but it has recognised that some stakeholders are concerned that the participation requirements for DSR may be too complex. Others are concerned about DSR delivery assurance arrangements. These concerns merit consideration in relation to whether the CM has achieved its objectives.

Does the security standard need to be revisited? Should parameters such as VoLL be refreshed?

The existing reliability standard is for an average Loss of Load Expectation (LoLE) of three hours per year. Two parameters feed into this reliability standard – the net Cost of New Entry (CoNE) and the Value of Lost Load (VoLL).

The VoLL estimates were based on a survey of consumers carried out by London Economics in 2013. (5) The reliability standard calculation makes use of the ‘willingness to accept’ estimates of VoLL for domestic and SME consumers.

Given that both of these studies were carried out more than five years ago now, BEIS is right to consider whether they need to be updated. Importantly, the decision to only use domestic and SME VoLL estimates was predicated on the assumption that these users are less able to engage in the capacity market or to respond to price signals. It is reasonable to consider whether this assumption will continue to hold for the next five years given the greater potential for smaller consumers to directly engage in the market. Technological developments such as the increasing prevalence of battery powered equipment and self-generation may also impact on VoLL and how it is incorporated into the reliability standard.

Our own study of VoLL conducted for the Agency for the Cooperation of Energy Regulators (ACER) considered differences in VoLL for electricity consumers across Europe and explored interactions between the provision of notice and the duration of an interruption on willingness to accept. (6) While the intended purpose and broad scope of this study are quite different from that under which we carried out our study, some of the analysis involved may provide a useful platform on which to build a more detailed national study.

The CoNE estimates used in the calculation of the reliability standard have also stayed the same since 2013. However, the government has commissioned subsequent work on technology specific hurdle rates and technical lifetimes which may be drawn upon in order to update these estimates.

Is National Grid the right organisation to set capacity requirements?

Based on the reliability standard, National Grid (in its role as System Operator) takes responsibility for developing a recommendation on target capacity to government. This inevitably leads some to question whether asymmetric incentives exist given the strong reputational implications for both National Grid and government of any security of supply event. The risk is that the inherent incentives of both parties may lead to ‘gold-plating’ in which the level of procurement ends up exceeding the level which is really necessary to meet the security standard.

However, the government has acknowledged this in the design of the target capacity methodology. The methodology includes some safeguards such as scrutiny by an independent panel of technical experts. In addition, it is difficult to identify who, other than National Grid, would have the sufficient information and expertise to develop the target capacity.

We believe that, while imperfect, the most pragmatic approach is for National Grid to develop the capacity target. However, it is important that lessons are learned from the process so far, so that safeguards against the risk of gold-plating can be improved upon. The government should review the safeguards that are in place, seeking input from the independent panel of experts. As BEIS has suggested, reconsidering the balance of responsibilities between government and Ofgem for review and approval of the capacity target is sensible, though we note that Ofgem is by no means immune from the reputational risks of a security of supply incident.

The stated priority objective of allowing foreigh sources of capacity to enter is interesting in the context of Brexit

Finally, we note the focus on enabling cross-border sources of capacity to enter into the CM. BEIS identified this as a longer-term objective given its (correct) view that this is a complex issue. In the context of Brexit (which is not mentioned in the context of this objective in the call-for-evidence), and uncertainty of future cross-border policy, the stated ambition seems to require some further clarification.

Sources:

(1) S&P Global, PJM draws more investment than ERCOT, despite reserve margin differences, 11 April 2018

(2) https://www.eprg.group.cam.ac.uk/wp-content/uploads/2018/06/1817-Text.pdf, p. 18

(3) BEIS, Capacity market and emissions performance standard review, call for evidence, p. 3

(4) ibid, p. 29

(5) Ofgem website publication, "The value of lost load electricity" by London Economics

(6) See our presentation at an ACER workshop, 18 June 2018