Professor David Newbery and Professor Michael Grubb – Written evidence

The Final Hurdle?

Security of supply, the Capacity Mechanism and the role of interconnectors

Authors: David Newbery[1] and Michael Grubb[2]

We have recently completed a paper of this title and wish to draw the main themes to the Committee’s attention as they relate to the topic of your enquiry into the Resilience of Electricity Infrastructure.  Our note addresses the “generation adequacy” of the system, not the resilience of transmission or distribution systems.  Historically, the latter has been the dominant source of problems in electricity supply adequacy; our analysis may be interpreted as concluding that this is likely to remain the case.

Resilience of UK electricity to 2020

Much of the policy effort and discussion around UK Security of Supply has focused on the risks, and the potential need for, and design of, a Capacity Mechanism.  We consider here a third aspect: assessment of the amount to be procured, specifically in the context of the UK announced intent (30 June) to procure 53.3 GW through its Capacity Auctions for 2018-19. DECC estimated the associated gross payments at £2.6bn annually, but with a much smaller net cost to the extent that generating companies pass through most of the subsidies as lower wholesale prices.

A conservative approach to procuring capacity is understandable, but we argue costs can be substantially reduced by deferring some of the associated auctions. At the heart of this is the (somewhat unfashionable) conclusion that the UK electricity is more resilient to the risk of “capacity shortfall” than widely assumed.  Our analysis concludes that 53.3GW is likely to be excessive, particularly but not exclusively in its (lack of) assumed contribution from interconnectors.  Political fear of ‘the lights going out’ can easily become a catch-all argument for excessive procurement, and associated subsidy to incumbent generators. The risk of over-procurement, particularly of new conventional capacity on long-term contracts, is that it drives up the costs to consumers; undermines renewable energy by implicitly transferring financial support from renewables to conventional generators; and impedes the European Single Market’s aim at a single pan-Eu electricity market, including by weakening the business case for other options, including future interconnectors that are widely agreed to be increasingly important as the share of intermittent electricity rises.

UK electricity demand has been on a declining trend for several years. As the New Balancing Services have served to remind us, the decline in capacity margins has been driven in large part by thermal plant retirement due to uneconomic conditions; much of this physical capacity (particularly gas and oil) remains potentially available given modest incentives.

In addition, there is large capacity of industrial backup generation, mostly diesel – the only estimate we found was an estimate of 20GW, a huge volume which if correct, and made available at times of peak need, would negate any significant risk of capacity shortfall; the apparent lack of any official estimate of this capacity appears to be an important lacunae which should be corrected as a priority.  We also note an indicative lesson from international experience; the US PJM capacity mechanism procured 4.8 GW of new generation, 11.8 GW of demand response, and 6.9 GW of increased net imports (all data cited in Newbery and Grubb, note 11).

One fundamental point is confusion of terms: the traditional measure of ‘loss of load’ risks is increasingly divorced from any risk of the ‘lights going out’. It is an estimate of the probability that demand, under normal market operating conditions, exceeds domestic ‘derated’ generating capacity of plant connected and licenced to generate on the system. It is a statistical measure of (mostly existing) contractual relationships; it is only weakly related to the potential resilience of the physical system. This is because it takes no account of mothballed generation or industrial backup (which appear as demand-side management, if companies use it to cut their load on the grid), other forms of demand-side management, or use of interconnectors – which collectively can be termed ‘latent capacity’.  Nor does it allow for the various ‘emergency’ measures could be invoked in times of system stress. 

The ‘Loss of load probability’ is also set on the basis of security standard which in terms of the estimated Value of Lost Load (VoLL) is likely to be excessive from a purely economic standpoint, as we explain in our paper, because it reflects estimates of domestic VoLL but is then applied in practice to industrial VoLL.

Thus there is no ‘cliff edge’ at which the lights go out, but rather an increasing array of options for managing tight conditions.

These options include the regional pooling of capacity implied by interconnectors, which should be recognised through a more appropriate treatment of interconnectors in security assessments, and/or their participation in the Capacity Mechanism.  As with other commodities (including food and gas) international trade supplements domestic production capacity, and security is not synonymous with self-sufficiency. In our paper we delve in some detail into the statistics and modelling assessments, all of which point to the fact that in any credible probabilistic assessment (which is what the ‘Capacity Margin’ concern is ultimately about), interconnectors make a positive contribution. The relevant issue is the risk of interconnectors being unable to supply when needed (most likely due to shortages abroad at precisely the same moment as the UK); excepting the Irish interconnectors, we question whether the probability of this is any higher than non-availability of conventional domestic generation. 

Indeed all three key institutions responsible – DECC, National Grid, and Ofgem – have made numerous statements that interconnectors make a clearly positive contribution to the resilience of UK electricity. This should be appropriately reflected in the methodologies of official security assessments, which is not yet the case.  For the present, having not taken account of the contribution of interconnectors in the basic assessment of margins for this December’s Capacity Auction, the overall capacity procured in the first auction should be adjusted to reflect the likely future contribution of interconnectors.

To conclude, our paper argues that there is considerable ‘latent capacity’ in the electricity system, including but by no means confined to interconnectors, which could be brought into play in the next few years and thus help to maintain security in the face of uncertain trends in electricity demand. Given this, the potential costs of the (probably excessive) caution implied by the decision to procure 53.3GW for 2018-19 could be substantially mitigated by deferring a much greater proportion of this to subsequent, shorter-term auctions. Consequently we consider it would be prudent to reduce the capacity proposed for procurement in the December 2014 auction, and leave more space for contributions that can be procured from a wider range of options once we have experienced national Grid’s management of the coming tight winter conditions, and once DECC has agreed the role of interconnectors in the Capacity Mechanism - all without detriment to the generating resilience of the UK system.

Beyond 2020

For the medium term (to 2030), we believe that a great deal may be learned about the options for ensuring continued generating resilience of the UK electricity system. The more active participation of demand side, mediated with ‘smart meters’, and a more systematic integration of industrial and local back-up capacities may extend.  Moreoever there is large scope for greater interconnection, including to more diverse regions (such as Scandinavia), contractual and institutional development in Europe, and the advent of electric vehicles which could be connected (or their batteries later utilised) to provide some degree of short-term storage capacity.  All these offer potential longer term efficient resources for enhancing system resilience. 

The Lord’s Enquiry addresses a crucial topic, and we trust that it will take full note of the way that the UK electricity system is changing and could evolve further – and the extent to which this can introduce new options for ensuring the continued generating resilience of the UK electricity system.

Reference

Newbery, D. and M Grubb (2014) “The Final Hurdle? Security of supply, the Capacity Mechanism and the role of interconnectors” EPRG Working paper (in press) at http://www.eprg.group.cam.ac.uk/research/publications-and-information/eprg-working-papers/eprg-working-papers-2014/

19 September 2014