Printer Friendly

UK Electricity Market Reform and the Energy Transition: Emerging Lessons.

1. INTRODUCTION: 'MODEL OR WARNING?'

The UK was widely seen as one of the world's leaders in electricity deregulation in the early 1990s. The move from a centrally dispatched Pool model to an energy-only market in 2001 seemed to embrace the trend to more decentralized market-determined pricing, so it came to some observers as a surprise when in 2010 the new UK Government embarked on a fundamental reform to the architecture of UK electricity markets. Some claimed it abandoned the principles of market competition seen as defining the UK approach (e.g. Timera Energy, 2011; Darwell, 2015), with widely divergent views as to whether it represents a potential model which others could follow, or a warning of the perils of--apparently--returning to greater state involvement in the market (Pollitt and Haney, 2013). Britain's electricity reform is therefore of central interest concerning electricity market design, and particularly in relation to the case for reforming the EU Target Electricity Model with its insistence on energy-only markets (Keay, 2016).

The proximate causes of Electricity Market Reform (EMR) were the impending closure of old fossil and nuclear plant with a lack of willingness to invest in new gas-fired generation, and the need to decarbonize the electricity sector without raising consumer costs excessively. The analysis provided to the Government proved controversial, and the required legislation took most of the 5-year Parliamentary term to complete. The first auctions under the new system only took place in December 2014.

This paper summarizes the background to and reasons for EMR, its structure and the results to mid-2018, commenting on the extent to which it has met the objectives of a secure, sustainable and affordable electricity system, and how it might be improved. Even in this relatively short period, substantial policy changes have been enacted, the regulator, Ofgem, has responded to criticisms of inefficient network tariff setting, and the auction outcomes have been substantially better than expected, so there is every reason to hope for further improvements, providing they can be effectively motivated.

Simulation studies of the EMR (Franco, Castanedo and Dyner, 2015) have indicated that such a 'comprehensive intervention or a similar one that includes the promotion of low carbon electricity generation through the simultaneous implementation of various direct and indirect incentives, such as a carbon price floor, a Feed in Tariff (FIT) and a capacity mechanism' were needed to deliver the evolving objectives of UK policy. But while there are reports to Government evaluating EMR (e.g. Grant Thornton and Poyry, 2015), and studies of electricity market reforms driven by EU Directives (e.g. Pollitt, 2012), to our knowledge, there has been no academic empirical assessment of the emerging lessons from the reforms. Four years after its enactment, we aim to provide such an assessment.

2. UK ELECTRICITY IN CONTEXT

2.1 The Evolution of the UK Electricity Supply Industry 1947-2001

The UK's electricity industry was state-owned from 1947-1990, and until 1955, almost the entire output was generated from coal, supplied by the state-owned National Coal Board. Under pressure from the Treasury, oil-fired power stations were then built, and the first generation of gas-cooled Magnox nuclear power stations started producing (followed with a long delay by Advanced Gas-cooled Reactors).

Figure 1 shows generation output by fuel with some of the key events. The share of oil peaked at 34% just before the oil shock in 1972, and thereafter coal and nuclear power gradually replaced oil, whilst the nuclear share rose to 20% by 1990.

By 1989, just before restructuring for privatization, around 90% of the conventional thermal generation was from coal, and thereafter the share of oil fell from 7% to 1% in 2002 (the remainder of thermal generation is largely from industrial by-product gases). Shortly after privatization, the coal share rapidly declined as imported electricity and nuclear power increased. It continued declining with the 'dash for gas', which was all new entry despite the considerable spare capacity. At the end of the century, consumption fell with deindustrialization and increased demand efficiency, while renewables displaced gas and/or coal, whose shares depended on the very volatile clean (gas) and dark green (coal) spark spreads (the margins between the wholesale price and the gas or coal cost including the cost of C[O.sub.2]--see Figure 5).

Privatization replaced the state-owned companies in England and Wales with two fossil and one nuclear (initially state-owned) generation companies, with an unbundled National Grid (initially collectively owned by the privatized distribution and supply Regional Electricity Companies, RECs). In Scotland the two vertically integrated companies were sold bundled, while in Northern Ireland three generation companies were sold with long-term power purchase agreements. The network companies were subject to price-cap regulation with the basket of tariffs changing in line with the Retail Price Index, minus an annual efficiency factor X (the "RPI-X" incentive regulation).

The wholesale market took the form of the mandatory gross Electricity Pool, into which all plant had to be offered (with sub-50 MW exceptions). This was centrally dispatched with a System Marginal Price (SMP) set by the marginal price offered by the most expensive unconstrained generator required. To this price was added a capacity payment, equal to LoLP*(VoLL--SMP), where LoLP is the Loss of Load Probability in that half-hour and VoLL is the Value of Lost Load ([pounds sterling]5,000/MWh in 2016[pounds sterling]). This would have been the efficient price if the SMP were equal to the System Marginal Cost (SMC), but the restructuring had left two large fossil companies (National Power and PowerGen) setting the price in the Pool with the ability to raise the wholesale price above the SMC.

Figure 1 shows the dramatic 'dash for gas' with its share growing from next to nothing in 1992 to almost a third of generation by 2000; a result of a combination of reasons. A legal ban on using gas for power generation had been lifted and the newly developed Combined Cycle Gas Turbines (CCGTs) were cheap, quick to build and offered high efficiencies, which, with falling gas prices, offered low average costs. The Pool allowed new entrants to sell at the same price as incumbents and the transparent system-wide price facilitated contracts. With energy policy leaving the market to guide choices, political risk was considered low and substantial entry by 'Independent' Power Producers (IPPs) occurred. These entered on the back of long-term fixed-price contracts (and often share ownership) with the RECs, who could pass on their costs to the captive franchise domestic market.

The combination of long-term gas contracts, long-term IPP contracts, regulated pass-through and performance guarantees on the CCGTs all reduced risk, whilst an added incentive for the RECs to sign such contracts was to exploit their new independence from centralized generation. The two fossil generators dominated the England & Wales Pool and clearly had considerable market power (Newbery, 1995; Tashpulatov, 2015), which the regulator negotiated down by encouraging them to divest 6 GW of coal plants to a third generator in 1996. The resulting triopoly was subject to less regulatory constraint in exercising market power, with an incentive to do so as they wished to divest coal plant before the "dash for gas" eroded their market share too drastically (Sweeting, 2007). Indeed, by 2000, coal-based generation had shrunk by more than a third (and increasing amounts of coal were imported rather than domestically produced).

1.2 The Electricity Industry Structure after 2001

Once they had divested enough plants, the generation companies were free to buy the supply (retailing) businesses originally integrated with distribution in the RECs. The market evolved towards the current Big Six generators plus retailers. (1) The market power of the original triopoly led to an increasing gap between cost and price in the Pool between 1996-2000, and encouraged the Government to replace the Pool with New Electricity Trading Arrangements (NETA)--just at the date (2001) when the price-cost margin collapsed under the weight of competition and excess capacity (Newbery, 1998; 2005).

NETA replaced central dispatch and the Pool with a self-dispatched energy-only market (abolishing capacity payments). The argument put forward was that getting rid of the Pool in favour of direct bilateral trading would encourage competition. To meet the physical need to balance supply and demand, NETA created a two-priced Balancing Mechanism which penalized under-delivery of promised power with a high penalty. The claimed logic for the reform was that self-dispatch required generators to submit a balanced offer (output matched by contracts to purchase), requiring them to contract all output, thus removing the incentive to manipulate the spot market (under-contracting encourages sellers to increase the spot price above the marginal cost, over-contracting to reduce the price below marginal cost, Newbery, 1995).

In practice, the balancing mechanism was so flawed that it required numerous painfully negotiated modifications to approximate an efficient balancing market. In addition, the risk of incentives to manipulate the spot market was replaced by a clear incentive to vertical integration: the merger of retailing and generation companies ensured that they were automatically hedged against electricity price uncertainties, since they would then be selling wholesale to themselves. However, this in turn created major barriers to entry, and a perception of the electricity system as an oligopoly of major power companies controlling the entire system from generation to consumption.

Despite evidence that transmission constraints requiring expensive redispatch could be exploited by generators, in 2005 the retrogressive principles of NETA were expanded to incorporate Scotland in BETTA--British Electricity Trading and Transmission Arrangements, creating a single Great Britain electricity market. National Grid acted as the National Electricity Transmission System Operator (NETSO) for GB, owning transmission south of the border but acting as an Independent System Operator in Scotland, where the two incumbents remain Transmission Owners. BETTA created a single price zone despite serious congestion on the Scottish border, where redispatch costs were high and growing as wind energy was increasingly deployed in Scotland. The EU Target Electricity Model that came into effect in 2014 mandates that separate price zones are created when there are significant boundary constraints. Had this been followed, Scottish consumers would frequently enjoy lower prices than the rest of GB, and the costs of redispatch would have been avoided. These costs rose to hundreds of millions of pounds annually, amounting to [pounds sterling]60 million in October, 2014, for a single (high cost) month. (2)

3. THE CASE FOR GB3 ELECTRICITY MARKET REFORM

Arguments for reforming the electricity market go back to earlier criticisms of the energy-only self-dispatched market model introduced in 2001 (Newbery, 1998; Grubb, Jamasb and Pollitt, 2007). As time passed, older coal and nuclear stations were scheduled for closure, leading to a growing concern about investment and security. An energy-only market encourages generators to mark-up their offer prices during periods of scarcity. Theoretically investors would predict future scarcity with higher prices, which would encourage them to start investments now for delivery at the time of predicted higher prices (if they could sell forward on sufficiently distant futures markets).

Several factors undermine this theoretical hope. The first is that electricity futures markets are either very illiquid or absent for much more than a year ahead, while it takes 4-8+ years from final investment decision to plant commissioning. Investors therefore need to be confident that the market conditions over the next 20-30 years are moderately predictable on the basis of existing laws and policies, and that demand and supply conditions are set by commercial, not political factors (Newbery, 2015b). The alternative to futures markets are long-term Power Purchase Agreements (typically of 15-25 years tenor) but with the ending of the domestic retail franchise, supply companies could no longer be confident about their future market and so there were no willing counterparties to sign such contracts, as there had been in the early days of the Pool.

Even without other considerations, it would be a brave investor to commit billions of pounds to a project against the prospect of electricity prices rising to reflect growing scarcity, on highly uncertain timescales, to unknowable levels, but set against the predictable political pressures that would likely curtail price rises. The early 2000s already saw a growing debate between economists, largely cast between abstract theory and the practical realities of likely 'missing money' in the calculations of cautious and risk-averse investors. (4) To the unavoidable economic uncertainties--associated not only with future market conditions but also the likely level, timing and frequency of scarcity pricing--was added political uncertainty. Investment requires some confidence in the future political landscape and the determinants of the wholesale electricity price, which one could at least plausibly estimate or hedge. However, UK energy policy had been in turmoil for most of the post-1997 period when the Labour Party came to power, with arguments over the role of coal, gas, renewables, and especially nuclear power. There were four Energy White Papers from 2003-2011 (the last being the precursor to EMR). The lack of any futures market combined with these multiple and often inestimable economic and policy uncertainties clearly deterred new investment in the UK's energy-only market.

Second, in theory, the growing imperative towards mitigating climate change and decarbonizing was to be driven by carbon pricing. The European Commission, persuaded by the success of the US sulphur cap-and-trade scheme, (5) created the EU Emissions Trading System (ETS) to deliver the EU's Kyoto targets with an EU-wide carbon price covering half total emissions. However, the EU ETS had signally failed to deliver an adequate, durable and credible carbon price signal. By the end of the first trading period in December 2007 the emissions allowance price had fallen to zero, and although it reached a more realistic [euro]30/tonne C[O.sub.2] in the second period in early 2008, it crashed to [euro]15/tonne with the financial crisis, oscillated around that for two years, and then sank further to well below [euro]10/tonne. The choice between coal, gas and zero-carbon generation investment depends critically on the future level of the carbon price, and investors had watched the EU carbon price collapse three times within five years.

Third, UK renewables policy was similarly unstable and hard to predict. The EU's Renewables Directive (2009/28/EC) (6) raised the required share of renewable energy (not just electricity) from 12% in 2010 to 20% of final energy demand by 2020, with each country agreeing its target share. The UK signed up to a particularly challenging share; starting from one of the lowest contributions (barely 1%), its target of 15% implied a dramatic growth of renewables. With electricity the easiest sector to tackle, this implied foreclosing much of the electricity market to conventional generation (at least, measured by output). The Directive also failed to remove allowances now displaced by renewables from the EU ETS, putting further downward pressure on the carbon price. To these conflicting signals was added a slowly growing realization that massive renewables entry would, if delivered, crash the wholesale market electricity price (an outcome predicted in falling utility share prices and realized most obviously in the German wholesale market, see Hirth, 2018). The case for conventional investment was thus further undermined and beset with uncertainty.

The imperative for low carbon investment became the other driving concern for EMR. Domestically, the UK Climate Change Act 2008 (7) was passed and provides a legal framework for ensuring that Government meets its climate change commitments. The Act requires that emissions be reduced by at least 80% by 2050 compared to 1990 levels, with the Government committed to a series of 5-year carbon budgets. (8) Yet UK renewables support policy was a shambles (Gross and Heptonstall, 2010; Grubb et al., 2014, box 9.3), and after a decade of political efforts to rehabilitate the reputation of nuclear power, the Government also wanted nuclear stations built by the private sector in this liberalized electricity market.

Britain faced two additional problems. First, the EU Large Combustion Plant Directive and then the Industrial Emissions Directive set tighter emissions limits that would force the retirement of older coal plant unless refurbished--a prospect that seemed risky and uneconomic. Second, Britain's first two generations of nuclear power stations (the Magnox and Advanced Gas-cooled Reactors) were coming to the end of their lives. It was expected that some 12 GW of the older coal-fired plant (about 20% of peak demand) would close by 2015 and an additional 6.3 GW of nuclear plant by 2016.

As fossil fuel prices rose towards their peak of 2008, the UK electricity model seemed increasingly untenable, as underlined by two official assessments. First, the UK Climate Change Committee--the body set up to guide implementation of the Climate Change Act--concluded that a market structure built purely around competition for buying and selling electrons could not deliver low carbon investment (CCC, 2008). Added to the generic concerns about investability of the market at all, and the inadequacy of carbon pricing, electricity prices driven by short-run generating costs could not support capital-intensive low variable cost zero-carbon generation, whether renewables or nuclear. Gas generation can be hedged by passing through fuel prices into the market; zero-carbon investments in contrast would take all the price risk of both fossil fuel and carbon price uncertainties. The energy-only market model was in direct conflict with the fundamental aim of the Climate Change Act, whose core rationale was to give strategic certainty for low carbon investments and hence reduce their high financing cost.

Then the regulator, Ofgem, concerned over the impending threat to energy security, launched Project Discovery in June 2009. The institution seen by many as the guardian of the liberalized energy model concluded that that '[t]he unprecedented combination of the global financial crisis, tough environmental targets, increasing gas import dependency and the closure of ageing power stations has combined to cast reasonable doubt over whether the current energy arrangements will deliver secure and sustainable energy supplies.' (Ofgem, 2010). Ofgem recommended 'far reaching energy market reforms to consumers, industry and government.'

Shortly thereafter, the Labour Government lost to a Conservative and Liberal Democrat coalition, and the newly formed Department of Energy and Climate Change (DECC) consulted on EMR (DECC, 2010). It concurred with Project Discovery that the carbon price was now too low to support unsubsidized nuclear power and the wholesale electricity price was set by fossil fuel prices (and the ETS) that ensured that fossil generators had a natural hedge as electricity prices mirrored gas and coal prices. Non-fossil generation faced volatile wholesale and renewable obligation certificate (ROC) prices. DECC was concerned about security of supply and that the market was not delivering the required volume of renewables.

In conclusion, the electricity market was not well suited to delivering either secure or sustainable electricity and 'affordable' was doubtful as retail electricity prices continued to rise, while industry warned about the high financing costs from the multiple risks facing the sector. Britain's model of liberalization was seen to be failing on all three key Government objectives.

4. THE EMR PACKAGE

The resulting White Paper (DECC, 2011) set out an intellectually coherent basis for electricity market reform through a combination of four mechanisms. The lack of a credible carbon price would be addressed by a Carbon Price Floor, almost immediately enacted by HM Treasury in the Budget in March 2011. Fossil fuel used to generate electricity would be taxed (through the Carbon Price Support, CPS) to bring the minimum price of C[O.sub.2] up to [pounds sterling]16/tonne in 2013, rising linearly to [pounds sterling]30/tonne in 2020, and projected to rise to [pounds sterling]70/tonne by 2030 (all at 2009 prices). (9)

When EMR legislation was being developed in 2010-11, the ETS forward price had hovered around [euro] 15/tC[O.sub.2] ([pounds sterling] 12/tC[O.sub.2]) for about two years, and the rate was set in relation to these levels. This implied a CPS top-up of just a few [pounds sterling]/tC[O.sub.2] in 2013, expected to rise slowly. However, with the collapse of the ETS price during 2011, the CPS required when written in to the legislation by 2013 actually escalated very rapidly.

As any tax could be changed with every budget (and the Carbon Price Support was indeed subsequently capped, as explained later), this policy was buttressed by an Emissions Performance Standard (EPS) that would limit C[O.sub.2] emissions from any new power station to 450 gm/kWh "at base load", intended to rule out any unabated coal-fired station (with exemptions for the demonstration Carbon Capture and Storage, CCS, stations which would only require a third or less of output to be subject to carbon capture). (10) The EPS had followed on from experience of a long battle over plans for a new coal plant at Kingsnorth in Kent, which E.On had proposed in 2006, and served to remove any ambiguity about UK policy towards coal. (11)

In terms of policy design, these two steps were relatively straightforward. The more difficult issues concerned how best to support low carbon investment, and how to ensure system security. The UK's carbon and renewables targets were estimated to require over [pounds sterling]12 billion investment per year (compared with less than [pounds sterling]5 billion in 2008). (12) This was considerably above financial analysts' estimates of the capacity of the Big Six (see footnote 1) to finance, requiring new sources of finance. All zero-carbon generation has high capital costs and low variable costs, making their cost highly sensitive to the Weighted Average Cost of Capital (WACC). By 2020 the cumulative investment in generation alone would amount to [pounds sterling]75 billion (DECC, 2011) and if the WACC could be reduced by 3% (as the auction discussed below demonstrated), the consumer cost would be reduced by [pounds sterling]2.25 billion per year (if all attributed to households, this is about 15% of a typical electricity bill). Lower risk enabling higher debt made this eminently feasible. As the Renewable Obligation scheme placed all the market price and policy risk on developers, replacing this by a fixed-price contract would considerably reduce risk and hence encourage new finance and entry.

The UK was reluctant to adopt the relative simplicity of the technology-specific German feed-in-tariff (FiT) model except for very small scale renewables (for which anything else would be unreasonably burdensome) but achieved the same risk reduction with 'Contracts-for-Difference' (described as a 'CfD with FiT'). Government would pay the difference between the reference wholesale electricity price and an agreed 'strike price' (or receive the excess over this strike price). This was initially done by publishing a set of strike prices for the CfDs based on inflated estimates of the required hurdle rate of return (i.e. the WACC) derived by asking the financial sector what they needed (DECC, 2013), combined with estimates of costs for different technology bands. Unsurprisingly, there was an enthusiastic uptake. As part of EMR, DECC had appointed an independent Panel of Technical Experts (PTE) to comment on the delivery of policies. (13) The PTE's first report (DECC, 2014) criticized the over-generous hurdle rate that resulted in high strike prices for the 15-year contracts oflTered to renewable generators. The stakes were even higher for nuclear power, in which the first (and possibly only) such contract was awarded for the Hinkley Point nuclear station on overly generous terms of a 35-year contract at [pounds sterling]92.5/MWh, roughly twice the then wholesale price.

For several reasons (including pressure from the EU Directorate-General for Competition concerning State Aids), after this initial round of 'administered' contracts, DECC moved to auctions for allocating specified volumes of renewables, divided into one 'pot' for developed technologies, and one for less developed technologies. As described below, Newbery (2016a) estimated the resulting clearing prices for on-shore wind lowered the WACC by 3% real. Unfortunately, the Conservative Government, in its 2015 election manifesto appealing to its rural constituencies, ruled out supporting on-shore wind--and along with it, all the other developed 'pot 1 ' renewable technologies--so the dramatic reduction in support prices for on-shore wind only survived one auction round.

The final strand of EMR was directed at security of supply through a Capacity Mechanism. After extensive internal debate and exploration of international experience, the Government rejected the idea of payments targeted to new entrants, or to retiring plant out of regular market operation (a 'Strategic Reserve'), in favour of system-wide payments to all generators who could contract to generate whenever called upon by the System Operator, National Grid. Wielding the fear of 'the lights going out', DECC overcame Treasury skepticism about the need for any capacity mechanism, whilst Ofgem amongst others argued that targeted supports for new entrants would create perverse incentives, for example, for a company to close down one plant (and many fossil plant were near life-expired but still useful to meet occasional peaks) in order to get subsidies to open another. The prevailing view became that capacity payments would in effect be a market for reliable capacity, with a fixed payment (the clearing price of the 'descending clock reverse auction') to all who could provide it. The assumption behind the design, however, was that GB's main need was for new efficient flexible CCGTs, and the system was designed accordingly with auctions held for delivery 4-years ahead--allowing both for major refurbishment and new plant, with the latter being offered 15-year capacity contracts.

The auction volumes would be decided by the Minister on the basis of advice from National Grid on the capacity needed to meet the GB security standard--of a Loss of Load Expectation of 3 hrs per year (on average over a large number of years)--together with estimates of the 'de-rating factor' to reflect technology-specific plant availability.

The institutional set-up behind this structure was itself a challenge. The Government created a separate, Government-backed body (the Low Carbon Contracts Company) to be the counterparty for CfD contracts, while National Grid is charged with both running the Capacity and the CfD auctions. Transparency was underpinned by publishing National Grid's analysis and the PTE's critique annually (see e.g. National Grid, 2016; DECC, 2014). The Minister chooses the de-rated capacity to procure in a winter auction for delivery four years' hence (hence the 'T-4 auction'), supplemented by year-ahead ('T-1') auctions for additional resources (including demand-side response), and, critically, to allow otherwise retiring plant to remain available for a further year.

5. RESULTS TO DATE

This paper is written (mid 2018) four years after the UK's EMR was enacted and the first administered contracts awarded, and more than three years after the first auctions.

5.1 CfD Allocation and Auctions

With the legislation so long in the making, by the time it was in the final stages in 2013, both the nuclear and renewables industries were impatient and warning of declining confidence, interest and capabilities in the UK market. In parallel with the legislative timetable, the Government negotiated a preliminary round of contracts for renewables and what was intended to be the first of a fleet of new nuclear power stations.

5.7.7. From negotiated contracts to competitive auctions

The first 'Administered contracts' for renewables summarized for Table 1 involved 15-year contracts for wind energy at strike prices of [pounds sterling]95/MWh (onshore) and [pounds sterling]140/MWh (offshore). (14)

The latter was almost three times the estimated cost of CCGT generation, and divided opinion deeply between those who saw offshore wind as the UK's great zero-carbon prospect--with almost unlimited resource--and those who saw it as a very expensive way to cut emissions. At this price, the contract value for each GW of offshore wind was over [pounds sterling]7bn (and they were expected to generate at load factors of only around 35%, so roughly three times the capital cost per 'derated' GW of nuclear power). The industry argued that given scale and commitment, it would be able to engineer costs down to [pounds sterling]100/MWh by 2020--a claim greeted with considerable skepticism.

The scale of commitments of EMR--and the low-carbon transition overall--were becoming clear. The long process of negotiating the contract for the 3.2 GW Hinkley Point C nuclear station finally emerged with a price of [pounds sterling](2012) 92.5/MWh indexed for a 35-year contract--with a total value (in present money, undiscounted) over [pounds sterling]70bn--along with extensive underwriting of some key risks (mainly of the CfD). This was substantially above most estimates of the generating cost assumed by the Climate Change Committee in recommending a new fieet of nuclear as part of its decarbonization strategy. (15) More than anything else, it all underlined the centrality of the finance challenge--those opposing feared that the [pounds sterling] 15-20bn construction cost would bankrupt the company before commissioning--along with the complete implausibility of any private entity building nuclear without massive government involvement.

EMR, however, delivered a considerably better outcome with the first competitive auction of renewable CfD contracts, held barely six months after the administered contracts, with the results shown in the final columns of Table 1. Newbery (2016a) argued that the close juxtaposition of these contracts provides an ideal natural experiment. Although both involved 15-year contracts, the first were conducted in parallel with the operation of the ROCs system, and companies could use projects constructed under this regime as their evidence for costs and required rates of return. With the move to auctions, this no longer applied; the contracts would go to those offering the best value, including lowest cost of capital, irrespective of costs under the far more volatile and uncertain ROC system. Using the results in Table 1, Newbery estimated the move to long term contracts awarded through competitive auctions lowered the cost of capital from about 6% to 3% (real)--which, applied to the [pounds sterling]75+bn expected investment required, translates into [pounds sterling]2.25bn annual saving for 15 years.

5.1.2 The Levy Control and the Second CfD Auction

Shortly after these first renewables auction contracts were awarded, a Conservative Government replaced the coalition, with resulting policy uncertainty. The previous coalition Government had placed a cap on the overall levy that could be charged to consumers, rising to [pounds sterling]7.6bn/yr (2011/12 prices) by 2020/21. This cap was now predicted to be breached (Lockwood, 2016). Overly generous feed-in-tariffs led to an unexpected explosive growth of solar PV (almost 10 GW compared to an expected 1.5 GW) before future tariff reductions were finally imposed. The post-2014 fall in gas and hence wholesale electricity prices increased the subsidy element in the CfD contracts. Finally, the offshore wind farms were generating substantially more output than expected, increasing their payouts, and underlying the undesirable way of paying subsidies per MWh rather than on capacity (see [section]7.1 and Grubb, 2015; Newbery et al., 2017).

The rising cost of the energy transition, and particularly, the CfD contracts for offshore wind, put EMR under considerable political pressure. Gradually, however, the arguments that led to the EMR won out, buttressed by the fact that breaching the levy cap was a sign of success in delivering renewables--more capacity (PV) with higher output (wind) than anyone expected. Indeed, the renewable energy target for electricity (30% by 2020), initially widely viewed as impossibly ambitious, looked increasingly plausible. Figure 2 shows the percentage increase in the share of generation from renewables since 2005 (the starting points were very different, due partly to pre-existing levels of hydro and biomass), for the 10 EU countries whose increase was higher than the EU as a whole. Until 2010 the UK lagged most of this pack but has since accelerated.

With industry arguing for policy stability and no credible alternative to EMR being proposed, the new Government finally announced its intent to continue. Nevertheless, after the first CfD auction of January 2015, it was over two and half years before the next took place, in September 2017. The 'pot 1' auctions for developed technologies legally had to include onshore wind (due to the 'technology neutral' principles embodied in the State Aid clearance), so the Government adopted the simple if ironic fix of declaring that no money would be made available in auctions for the cheapest renewables, and the second auction would focus entirely on the less developed 'pot 2'--which included offshore wind.

The outcome was unexpected but welcome. Figure 3 contrasts the administered prices and first auctioned prices (of summer 2014 and Jan 2015 respectively), with the prices obtained in the September 2017 auction for three major offshore wind farms. The most striking were two wind farms, totaling 2,300MW capacity, scheduled for delivery by 2022/23 at a contract price of just [pounds sterling]57.50/MWh--far below any expectations, at half the price in the first auction, and allowing the Government to secure 57% more capacity for 44% less estimated subsidy, compared to round 1. (16) Renewable UK (2017) also estimated the UK had regained ground in the associated industries, with almost 50% of the supply chain value expected to go to British business.

The UK auction was the culmination of several such auctions by countries bordering the North Sea over the previous year, which had yielded declining costs for various reasons. Off-shore rig costs fell as North Sea oil and gas declined, while greater confidence in future prospective demand justified developing supply chains and learning from earlier installations led to better (and larger) designs with lower costs. These provided reassurance that the auction prices were not just an example of the winner's curse. Given the huge scale of the UK offshore wind resource, it seemed that the gamble of committing to North Sea wind development, based on initial Government contracts followed by competitive auctions, had paid off, and opened up a major new and zero-carbon national (and regional) energy resource.

5.2 Capacity Market

The first capacity auction held in December 2014 was for almost 50 GW de-rated capacity by winter 2018/19. (17) Figure 4a shows the types of all (existing, refurbished and new) capacity procured in the first four annual auctions by delivery year (essentially the winter periods), including the "Early Capacity Auction" (i.e. T-1) held in Feb 2018 for delivery the following winter, with the resulting total capacity secured for winter 2018/19 ("Early 2018/19"). Based on the estimated 'net Cost of New Entry' (net CoNE)--which was interpreted as the price required to support a new CCGT investment above the revenue earned in the market--the Government projected the likely clearing price in the first T- 4 auction to be [pounds sterling]49/kWyr, (18) which it used to help define a demand curve, with a price cap of [pounds sterling]75/kWyr (1.5 x net CoNE). (19)

In the event, the first T-4 auction cleared at [pounds sterling]19.40/kWyr (figure 4b). Only one CCGT company (with two turbines) won a contract (but withdrew after failing to raise finance). Figure 4a shows the amount surviving, which was hence was less than intended, a shortfall then compensated by procuring more in the year-ahead T-1 auction for 2018/19 (whose clearing price is the single 2018/19 diamond in Figure 4b). The major beneficiaries were existing coal, gas and nuclear generators. This was as expected, but led to protests about the Government subsidizing the coal plant that it claimed to be trying to phase out.

More concerning was that interconnectors were excluded (the UK had about 3 GW of connections to continental Europe, with more planned). (20) The evidence was unambiguous that interconnectors contributed to security, with imports even more likely in stress periods when GB wholesale prices would be very high (Newbery and Grubb, 2015). The European Commission intervened, ruling that excluding interconnection was discriminatory, and only gave state-aid approval for the first auction provided interconnectors were included in subsequent rounds (Figures 4a). Although absent in the calculation, their physical contribution made up for the shortfall from the withdrawn new CCGT plant, and they were included in the Early Auction for 2018/19.

More problematic still was the large volume of small open cycle gas turbines (OCGT) and reciprocating diesel generators (arrowed in figure 4a), with an average size of about 10 MW, connected to the distribution network. Diesel was clearly both a carbon-intensive fuel and one with dangerous air pollutants like nitrogen oxides and particulates, which came into sharp public focus with the VW vehicles scandal. In principle, these plants are unlikely to be used much--most of this new build is of the cheap capital (BEIS, 2016), high running cost plant appropriate to a role of just meeting extreme system needs, though their very limited running could not be guaranteed. Politically, paying for polluting diesel instead of relatively clean and efficient CCGTs was highly problematic.

The reason for the rush to connect small generators to distribution networks was, in hindsight, obvious. Transmission-connected generation pays a Transmission Network Use-of System (TNUoS) G (for Generation) charge that varied in 2017 across the country from about [pounds sterling]20/kWyr in the far North to -[pounds sterling]5/kWyr in London (i.e. paid to deliver peak power), signaling where new generation is encouraged or discouraged. In addition Load (actually the Distribution networks that pass it through to final consumers) pays TNUoS on peak demands: lower where G charges are high, higher where G charges are low, so that the sum of the two is roughly constant across the country at about [pounds sterling]50/kWyr.

A small part of the TNUoS charge represents the marginal cost of expanding the grid, but the overwhelming part is the residual charge to collect the regulated transmission revenue set in the periodic price controls. This residual was low (about [pounds sterling]10/kWyr in 2006) but was projected to rise to over [pounds sterling]60/kWyr by 2020.

Distribution-connected generation was credited with the avoided TNUoS charge, and hence gained a relative advantage of about [pounds sterling]50/kWyr compared to transmission-connected generation, so that the [pounds sterling]20/kWyr auction price translated into a [pounds sterling]70/kWyr benefit for these small generators. The efficient saving in transmission they deliver is not the average but the marginal cost of grid expansion, and, worse, the lost revenue by reducing peak demand on the grid must be recovered from the remaining grid users, further encouraging generators to migrate to local networks. The cost of the delay in remedying these tariff distortions (already clear in December 2015 and recognized in DECC, 2016, [section]33-34) was apparent when 8.7 GW of 'embedded generation' registered for the third Capacity Auction. It took Ofgem, charged with ensuring the transmission tariffs are cost-reflective, nearly three years to remove this biased embedded benefit. (21) The removal of the embedded benefit was immediately subject to litigation, although tariffs are routinely revised every year and have no long-term durability. Auctions that offer 15-year contracts in one afternoon of bidding need the regulated price signals to be fit for purpose before the auction, not changed several years later, after investments have been committed.

The second T-4 auction in December 2015 confirmed that UK electricity demand was falling, not rising (at least at transmission level), and the capacity procured for the second auction was lower (wind is excluded as it has already been paid for its capacity, but its equivalent firm capacity contribution is netted from total demand to determine the amount to procure). However, coal plants were beginning to close apace, for reasons indicated below (in addition to the low value of capacity payments)--including some which had capacity contracts, thus prompting the government into holding a 1-year-ahead auction earlier than planned (the Early Auction), and increasing the volume to be procured in the next 1-year-ahead auction to cover for cancelled capacity contracts (Figure 4).

The experience after these auctions underlined the unexpected: gross capacity requirements so far have turned out lower than the auction volumes set, and yet the system had become somewhat more dependent on year-ahead auctions than originally envisaged because of cancelled contracts for new build. The 2017 PTE report (DECC, 2017) argued there needed to be more attention to demand side response (DSR) and the 'latent capacity' of the system to handle stress events (Newbery and Grubb, 2015), (22) to get a better balance of costs, and hence reduce the inevitable institutional and political pressures to over-procure.

The Government had designed the Capacity Mechanism to deliver reliable generating capacity at the cheapest price, which the auctions did. Unfortunately, prices and true economic cost were not aligned as transmission and distribution tariffs failed to give efficient location decisions, resulting in small polluting diesel securing long-term contracts. The auctions discriminated against demand-side management, undermining their main potential market of responding to scarcity pricing in the wholesale market. The Government expected the Capacity Mechanism to deliver large flexible gas-fired generation (which it has not). Incumbent generators criticized the unfair competition from decentralized generation, which was effectively subsidized due to the exemption from the now very high residual transmission charges.

In partial response, the Government effectively barred diesel from the third auction using environmental regulation, and the price in that auction (Dec. 2016) rose somewhat, bringing another surprise with the scale of storage coming forward. Embedded generation still dominated the winning bids (Ofgem had not yet published its intention to end embedded benefits), but despite higher procurement volumes, cleared at a price again much too low to support new CCGTs, which many still regarded as necessary for providing flexible bulk power through the 2020s and beyond. Moreover, it also became clear that much of the 500MW of battery storage in the most recent auction has storage lifetime much shorter than the potential duration of 'stress events', but was being accredited as if firm--leading to another revision of rules. (23)

Aside from concerns about the lack of a 'level playing field', the Capacity Mechanism faces two other criticisms. One is that the incentives on the Minister and National Grid are to over-procure capacity--no-one wants to be held responsible for the 'lights going out', as the newspapers periodically announce is imminent. (24) As Ministers and National Grid do not pay (and National Grid may benefit if more transmission investment is required), and consumers do not see the capacity payment in their bills, there is an additional bias to over-procurement. (25)

This exacerbates the other concern about the potential perverse consequences of over-procuring capacity. Existing generators pass the capacity payments through in reduced wholesale prices (Ofgem, 2017) which increase the capacity payments required to support new investment--and in addition the net cost of the other major pillar of EMR, the CfD supports--whilst the dampening effect on peak-load pricing reduces the revenues available to demand-side management, for which the Capacity Mechanism as it stands is not a credible substitute. The Irish model discussed in [section]7.2 avoids many of these problems.

The judgement is thus mixed. The positive case is that the Capacity Mechanism is delivering capacity to maintain security, and has uncovered many options previously not seriously considered at prices far lower than expected. In so doing it raised several challenges, only some of which are, slowly, being resolved as running annual auctions allows the system to be fine-tuned to avoid future problems. A more skeptical view is that the Capacity Market is an overly complex mechanism that mainly rewards existing capacity and less desirable forms of small-scale new generation, which risks impeding market-based DSR, and that Ofgem has repeatedly failed to adequately address failures in the locational element of network charges, in large part as it is unwilling to grandfather charges to existing power stations (which cannot relocate, but can exit) and confine tariff revisions to fiiture connections.

The final observation concerns the remarkably low (and most recently falling) auction clearing prices, despite concerns over plant retirements and tightening margins. One explanation is that the System Operator is increasingly procuring various balancing services, most recently enhanced frequency response for plant "capable of responding within one second to frequency deviations and operate in frequency sensitive mode". (26) These additional ancillary services do much to alleviate the "missing market" (and thereby the missing money) problem for firm capacity (Newbery, 2015b), and in part are also responsible for the entry of batteries capable of providing such fast response.

4.3 Carbon Price Floor and Emissions Performance Standard

As described, the other two elements of the EMR targeted coal more directly. With the Performance Standard effectively removing any prospects of new coal investment, the issue really concerned the existing fleet and the incentives for keeping coal power stations open. Before the introduction of the carbon price support (CPS), the carbon price was insufficient to have much operational impact. However, the CPS rose from [pounds sterling]4.94/tC[O.sub.2] in 2013 to [pounds sterling]18.08 in 2015-16, but was then frozen at an [pounds sterling]18 add-on to the EU ETS price. (27) At this level the CPS is a tax raising around [pounds sterling]1.5bn/yr.

In addition, gas prices began to decrease at last. The combination made it economical to start base-loading gas instead of coal. Figure 5 shows that the carbon-inclusive cost of gas-fired generation fell below that of coal from April 2014 and, for high efficiency CCGTs, has remained below since. Indeed, coal has been frequently unprofitable to operate since mid-2015 (negative spread, below zero in figure 5), prompting a number of large coal plant closures.

The overall impacts on the GB electricity system and its emissions have been dramatic. As the combination of fuel and carbon prices increasingly made gas plants cheaper than coal to run, this made coal the marginal plant. Gas increasingly displaced coal, as figure 1 illustrates, and on several days there has been no coal generating for the first time since the 1890s. Coal bum has now decreased dramatically from 41% of UK generation in 2013 to less than 8% in 2017, so that total UK C[O.sub.2] emissions are now lower than a century ago. (28)

6. POPULAR CARICATURE: RETURN OF THE "CENTRAL ELECTRICITY GENERATING BOARD"?

The original vision that motivated privatization was, to quote the then energy minister Lawson: 'the business of Government is not the government of business' (Lawson, 1992, p211). As to energy policy, Lawson stated at a British Institute for Energy Economics conference in 1982 'I do not see the Government's task as being to try and plan the future shape of energy production and consumption. It is not even primarily to try to balance UK demand and supply for energy. Our task is rather to set a framework which will ensure that the market operates in the energy sector with a minimum of distortion and energy is produced and consumed efficiently.' (29)

Critics (e.g. Darwell, 2015) have argued that EMR represents a reversal of this ideal, with the Government now planning the future shape of energy production and trying to balance demand and supply through the capacity auctions. Renewable technologies are procured through CfD auctions; nuclear power is procured by a bilateral contract underwritten by the Government. Critics further argue that long-term contracts are replacing the market as a mechanism to attract new investment into the industry, seemingly moving back to the Single Buyer Model that the French, with their state-owned electricity industry, pressed unsuccessfully for in the first EU Electricity Directive.

So, is EMR an admission of a failure of the liberalized electricity market model, or is the Government, though the Energy Act 2013, attempting instead to better correct market failures? We would argue the latter. Long-term contracts (only for new investment) replace the absent futures markets, all the more necessary given the unpredictability of future energy policy. Many renewables (and certainly PV and wind) create learning spill-overs that are unrewarded by the market, which justify subsidy. (30) As learning spill-overs depend on technology and the state of the technology's maturity, the subsidies should also be technology specific (although the form of subsidy provided by EMR is not particularly well-directed to addressing the learning market failure--see e.g. Newbery, 2018).

It is moreover wrong to confuse government-led auctions with central planning. The auctions created new markets, and, as is common with new markets, both unearthed and stimulated the unexpected. But the new markets--and investments and learning--could not have occurred without the Government recognizing there were market failures that had to be corrected, with strategic and public benefit dimensions that spot markets could never plausibly value sufficiently. (31)

Providing a long-term contract for nuclear power also reflects the lack of a durable credible carbon price, as well as the lack of futures and risk markets for future power prices and nuclear policy changes (such as the Energiewende in Germany). While the particular form of underwriting for Hinkley Point is highly unsatisfactory, it seems inconceivable that private companies would take on nuclear risk without some Government-backed guarantee to facilitate financing. The UK, like many other countries, has struggled to find cost-effective ways to support nuclear power, and yet it remains contested whether or how the UK will meet its ambitious goals to almost entirely decarbonize the power system, well before 2050, without it. (32)

It is also worth remembering that the massive entry of new CCGTs in the 1990s by 'Independent' Power Producers was based on long-term power purchase agreements with the Regional Electricity (distribution) Companies, many of whom were co-sponsors and shareholders in the projects. The development of those CCGTs was in turn heavily subsidized by the defence industry developing jet engines.

The important difference with the new contracts is they are competitively secured at auction, and so market tested in a way that is central to the idea of a liberalized market. Holding periodic auctions also allows flaws in the market design to be detected and corrected in a timely fashion. In contrast, the period of the Electricity Pool from 1989-2001 was marked with great difficulties in reforming the Pool, a multilateral contract that was intended to be hard to change in order to offer greater stability to the liberalized market.

Recent interventions by the Government (such as banning any subsidies for on-shore wind) (33) can also be contrasted with the earlier period of the intended 'hands-off' energy policy. In the 1990s, the coal industry was (temporarily) saved with coal-backed contracts forced on the retailers. The incoming Labour Government imposed retrospective windfall taxes on the privatized utilities. Gas-fired generation was also proscribed for a period, again to save the coal industry. (34)

The New Electricity Trading Arrangements (NETA) that ended the Pool was a blunt market redesign to address market power--a problem that had already been solved by the time the reforms took place. The problems of NETA were exacerbated by the expansion to include Scotland (which benefited Scottish power generators at the expense of both Scottish and English consumers). The Renewables Obligation Scheme was a poor substitute for the earlier auctioned Feed-in Tariff support scheme, and the alphabet soup of interventions to enhance energy efficiency, stimulate new technologies, and reduce C[O.sub.2] emissions at various levels in the system were poorly coordinated, lacking a clear consistent intellectual framework to guide their choice, design and relationships.

Electricity, delivered to each voter's home and critical to modem existence, is inevitably politicized. The main question is how to reduce the adverse effects of inevitable interventions. The move to auctions, fixed price contracts with the price set at auction for renewables and firm capacity, backed by a government-guaranteed credible counter-party, and even the Carbon Price Support, seem steps in the right direction. Compared to most of their predecessors, they are arguably better policies to address market failures, and do more to shape the evolution of the electricity system in directions consistent with the goals of public policy.

7. FUTURE CHALLENGES

7.1 The limits of fixed price contracts for renewables

After two decades of lagging behind renewable energy in Europe, the UK turnaround is remarkable (figure 2), as is the policy progression from rather ineffective, through inefficient, to the effective and relatively low-cost structures of the EMR. Yet this success in delivering the target quantity of renewables will, ironically, start to highlight the limitations of the chosen support approach, and the need for further reform to reduce costs. As renewables mature, the objectives and conditions change. The fixed price auctioned contracts of the EMR reduced the cost of capital and attracted investment to emergent and potentially risky new industries. However, this simple attraction carries strategic weaknesses.

First, the subsidy element amplifies incentives to inefficient siting. Going to resource-rich regions makes sense, but the windiest locations get a full 15 years of guaranteed income at a rate defined by the most costly plant in the auction--much more than they need. A move to contracts which guarantee support for a given volume (MWh/MW) of total output, rather than for a given number of years, could solve this problem (Newbery et al., 2017).

As renewable volumes grow, the value of power will become more volatile over time and more variable according to location, so a support system based on fixed prices will be increasingly at variance with the underlying economics (Mills & Wiser, 2013). The need will be less for subsidy, and more for pricing which is capital-efficient but better reflects system value and takes account of the 'system costs' variability imposes. Most estimates suggest these are modest at low levels of penetration (Heptonstall, Gross, & Steiner, 2016), but the EMR provides little incentive to manage these sensibly as penetration increases.

Under the EMR, renewables do face short-run balancing costs--the contract strike price is averaged in relation to the wholesale price, and generators pay imbalance prices in the balancing mechanism. This however is not the same as pricing that reflects the overall value to the system, and its dependence on location and time of generation. The real value depends on location including transmission costs / constraints, which are not adequately reflected in the CfD system, amplifying the incentive to inefficient siting. As capacities rise, system value will increasingly reflect time of operation: more PV added to a system which already has enough to meet demand on sunny days; or wind expanding in existing regions to feed more power at the same time into a system which already has enough to meet demand on windy nights, is obviously of declining economic (and even environmental) value (Baldick, 2012). Within about a decade, the UK power system may have growing periods in which zero carbon sources with CfD or other renewable energy contracts are able to meet all the power needs (e.g. on sunny weekend days, or windy winter nights). In these conditions, the existing holders of CfDs will bid down wholesale prices to deeply negative levels to keep receiving a subsidy--with adverse effects on other zero-carbon generation lacking CfDs such as existing nuclear power. Interconnection will alleviate but not remove these problems, not least as local constraints become more important.

Newbery et al.'s (2017) proposal to pay an auctioned price for a fixed number of full operating hours, while confronting developers with nodal prices, hedged with contracts on a reference node, would solve this problem and also improve the incentives. That would hedge against system wide variations in wholesale price caused by fuel and carbon price changes, while encouraging developers to choose locations that can export power without local transmission constraints. Already UK distribution networks are offering more suitable contracts to encourage sensible locations (Anaya and Pollitt, 2015). Grubb et al. (2018) suggest an additional approach, to aggregate unsubsidized renewable energy contracts through a 'green power pool' which would have to buy backup and balancing services--from its customers and from the rest of the system--and charge them on to the renewable generators that supply the pool.

7.2 Capacity auctions and reliability markets: the Irish example

Finally, whilst there is no obvious international example of a market structure fully appropriate for variable renewables at scale, there are many examples of approaches to system reliability that might offer improvements to the GB capacity market, and one is part of the UK. The Single Electricity Market (SEM) of the island of Ireland (comprising Northern Ireland, part of the UK, and the Republic of Ireland) is transiting to the Integrated SEM (I-SEM), under the requirement to adopt the EU Target Electricity Model. That requires replacing the centrally dispatched pool, including its mandatory requirement to bid short-run marginal costs (to which is added a regulated capacity payment), with bidding into the European auction platform, EUPHEMIA.

Along with this, the system's previous Capacity Remuneration Mechanism (CRM) has now been replaced by auctioned Reliability Options, (ROs). (35) An RO is a one-sided CfD, in which the holder pays back the excess of the market price over the strike price (set in the first auction at [euro]500/MWh). It has the advantage that prices in the electricity market are free to reach their efficient level while protecting consumers on the one hand against high prices, and providing a steady capacity payment to generators in place of the uncertainty of occasional high prices as a source of revenue. Reliability options are arguably a better alternative to capacity payments, as they address missing money problems, hedge high and uncertain prices for generators and consumers while ensuring that the wholesale market and international trade clear at efficient prices (Newbery, 2016b).

The first (T-1) Irish RO auction cleared at just over [euro]40/kWyr in December 2017, (36) although several stations received higher (regulated) prices to meet transmission constraints. Compared to the previous CRM 'pot' of [euro]516 million to be spread over all available and eligible generation, (37) the new RO auction will pay out [euro]379 million, saving [euro]137 million/yr.

Various factors may explain why the Irish RO price is so much higher than the UK Capacity Market, though one factor may be a price cap of [euro]500/MWh on the system, so that holders forego the upside still available under GB capacity agreements. (38) In addition there is uncertainty about how prices will be set in the I-SEM, as yet untested, and once the new I-SEM market has bedded down auction prices may fall, as was the case in GB. If the EMR capacity auction is to be reformed, this RO auction obviously merits attention.

8. CONCLUSIONS

The impact of more than a quarter of a century of reforms in UK electricity policy has been profound, as was already evident from the long-term evolution of fuel mix and demand in Figure 1. With the dash-for-gas during the 1990s, the UK had moved to a roughly equal mix of coal, gas, and nuclear. As the oldest nuclear plants were retired in the 2000s, the system was kept supplied by the abundance of gas, steadying demand, and the slow emergence of renewables--still barely visible in the overall statistics--whilst coal remained the mainstay of baseload generation and seasonal scheduling. Over the full period since privatization, coal fell from two-thirds of generation in 1990 to 35% in 2000, to 7% in 2017, halving C[O.sub.2] emissions from power generation over the quarter century. (39) Over the next few years, coal will be increasingly confined to meeting winter peaks.

Renewables--including conversion of some coal to biomass--began to accelerate after 2010, at a greater rate with the advent of feed-in-tariffs for the small sources and long-term CfD contracts for the large. Electricity demand on the transmission system began to fall and by 2015 the carbon price support drove coal to the margin of what remained. In 2015, the UK, an 'island of coal in a sea of oil and gas', saw the first full day without coal-fired generation for over a century. The Government is minded to phase out coal-fired generation entirely by 2025 (40)--if there is any left by then, given carbon, gas and coal price trends, and the tightening emissions standards. In the short-run, that leaves gas as the flexible dispatchable fuel to manage renewables intermittency. The earlier hostility to CCS appears to be waning (although its cost compared to alternative low-carbon options may be a greater barrier), as is hostility to on-shore wind.

Competitive auctions have proven their worth not only in revealing costs and options, but in driving down costs and prices for both renewables and firm capacity. The commitment to offshore wind, originally seen by many as a costly 'white elephant', now appears to be a way of encouraging a coherent supply chain with its cost reductions to develop and deliver. The energy-only market central to the EU Target Electricity Model is demonstrably unsuited to cost-effective new investment, while capacity auctions clearly can work--if the remaining regulated network tariffs are correctly set. Transmission pricing policy is also slowly adapting to the need to give better signals for the decentralized world of smaller generating units that can connect rapidly, but need good price guidance to locate in the right place and at the right voltage level.

At the heart of all these trends is the need for, and gradual acceptance of, credible, stable policies that encourage development and deployment, and that support leaming-by-doing in collaboration with the widest group of countries, as in Mission Innovation, (41) to continue to reap ongoing and impressive cost reductions for some technologies. Yet the regulatory journey is by no means over. The fixed-price contracts for renewables have been effective in reducing financing and technology costs, but create perverse impacts on the wholesale market, and lack any incentive to site renewables efficiently with respect to either place or generation timing, and hence the 'systems costs' they create. This mattered little when the capacities were small; it will matter far more over the next decade, when adding more renewables will increasingly serve to generate power when it is least needed, and conflict with other contracted sources (nuclear and biomass), whilst declining unit costs could be increasingly offset by rising system costs.

Similarly, the problems of the Capacity Mechanism are gradually being resolved, although arguably a Reliability Option auction would be preferable, and including demand side response remains a work in progress. Along with the small renewables feed-in-tariffs, the combination of the Capacity Mechanism and the 'embedded benefits' distortion may unwittingly have helped to launch a revolution in distributed energy resources, but the fixes to date are probably inadequate for dealing with the wider consequences and opportunities of such decentralization.

The balance between the state and private sectors is being revisited, not without dispute, and we are a long way from a credible nuclear (or even CCS) strategy. The way we support zero and low-carbon generation could benefit from further changes (supporting the learning externalities as well as the carbon saved), and better location signals are still needed for investment and dispatch. The evidence suggests that UK's Electricity Market Reform has been a major step forward, but a considerable journey remains ahead.

ACKNOWLEDGMENTS

The paper on which this was based was commissioned by the MIT Centre for Energy and Environmental Policy Research (CEEPR WP 2018-04). We are indebted to CEEPR and Tony Tran for their helpful comments, to Adonis Yatchew for encouraging us to submit and for the four excellent referees' reports, which greatly improved the paper, for which authors remain responsible, and are both writing in their independent academic capacities, drawing solely on published material.

REFERENCES

Anaya, K.L. and M.G. Pollitt (2015). "Options for allocating and releasing distribution system capacity: deciding between interruptible connections and firm DG connections." Applied Energy 144: 96-105. https://doi.org/10.1016/j.apenergy.2015.01.043.

Baldick, R. (2012). "Wind and Energy Markets: A Case Study of Texas." IEEE Systems Journal, 6, 27-34. https://ieeexplore.ieee.org/stamp/stamp.jsp?amumber=6007040. https://doi.org/10.1109/JSYST.2011.2162798.

Batlle, C., C. Vazquez, M. Rivier, and I.J. Perez-Arriaga (2007). "Enhancing Power Supply Adequacy in Spain: Migrating from Capacity Payments to Reliability Options." Energy Policy 35(9): 4545-4554. https://doi.org/10.1016/j.enpol.2007.04.002.

Bidwell, M. (2005). "Reliability Options: A Market-Oriented Approach to Long-Term Adequacy." The Electricity Journal 185: 11-25. https://doi.org/10.1016/j.tej.2005.03.010.

BEIS (2016). Electricity Generation Costs, at https://www.gov.uk/govemment/publications/beis-electricity-generation-costs-november-2016

BEIS (2017). Historical electricity data: 1920 to 2016. Available at https://www.gov.uk/govemment/statistical-data-sets/historical-electricity-data-1920-to-2011.

Cramton, P., A. Ockenfels, and S. Stoft (2013). "Capacity Market Fundamentals." Economics of Energy and Environmental Policy 2(2): 27-46. https://doi.org/10.5547/2160-5890.2.2.2.

Darwell, R. (2015). Central Planning with Market Features: how renewable subsidies destroyed the UK electricity market, Centre for Policy Studies, at http://euanmeams.com/renewable-energy-the-most-expensive-policy-disaster-in-modem-british-history/

DECC (2010). Electricity Market Reform: A consultation document. Dec, at http://www.decc.gov.uk/en/content/cms/consultations/emr/emr.aspx.

DECC (2011). Planning our electric future: a White Paper for secure, affordable and low-carbon electricity, at http://www.decc.gov.uk/en/content/cms/legislation/white_papers/emr_wp_2011/emr_wp_2011.aspx.

DECC (2013). Annex F: EMR Panel of Technical Experts Final Report for DECC. July, at https://www.gov.uk/govemment/uploads/system/uploads/attachment_data/file/223656/emr_consultation_annex_f.pdf.

DECC (2014). EMR Panel of Technical Experts 'Final Report on National Grid's Electricity Capacity Report, at http://www.eprg.group.cam.ac.uk/wp-content/uploads/2014/07/EMR_Panel_s_Final_Report on National_Grid_s_ECR.pdf

DECC (2016). Electricity Market Reform Panel of Technical Experts 2016, final report on National Grid's Electricity Capacity Report 2016, at https://www.gov.uk/government/publications/electricity-market-reform-panel-of-technical-experts-2016-final-report-on-national-grids-electricity-capacity-report-2016.

DECC (2017). Electricity Market Reform Panel of Technical Experts 2017 final report on National Grid's Electricity Capacity Report 2017, at https://www.gov.uk/govemment/publications/electricity-market-reform-panel-of-technical-experts-2017-final-report-on-national-grids-electricity-capacity-report-2017.

DUKES (2018). Digest of UK Energy Statistics, at https://www.gov.uk/govemment/statistics/electricity-chapter-5-digest-of-united-kingdom-energy-statistics-dukes

Franco, C.J., M. Castaneda, and 1. Dynera (2015). "Simulating the new British Electricity-Market Reform." European Journal of Operational Research 245(1): 273-285. https://doi.org/10.1016/j.ejor.2015.02.040.

Gmbh, M., T. Jamasb, and M. Pollitt (eds.) (2007). A low carbon electricity system for the UK: technology, economics, and policy, Cambridge: Cambridge University Press.

Grabb, M., with J.C. Hourcade and K. Neuhoff(2014). Planetary Economics: Energy, Climate and the Three Domains of Sustainable Development, Routledge / Taylor and Francis.

Gmbb M. and P. Drummond (2018). "UK Industrial Electricity Prices: Competitiveness in a low carbon world." UCL Institute of Sustainable Resources, www.ucl.ac.uk/bartlett/sustainable

Gross, R. and P. Heptonstall (2010). "Time to stop experimenting with UK renewable energy policy." ICEPT working paper, Imperial College London, at https://pdfs.semanticscholar.org/e932/cbb447c994372339a484eba7bl661d6efd7f.pdf

Heptonstall, P., R. Gross and F. Steiner (2017). "The costs and impacts of intermittency--2016 update: A systematic review of the evidence on the costs and impacts of intermittent electricity generation technologies." UK Energy Research Centre, London, at http://www.ukerc.ac.uk/publications/the-costs-and-impacts-of-intermittency-20I6-update.html

HM Treasury (2011). Carbon price floor consultation: The Government response, at http://webarchive.nationalarchives.gov.uk/20130103021036/http://www.hm-treasury.gov.uk/d/carbon_price_floor_consultation_govt_response.pdf.

Joskow, P. (2008). "Capacity Payments in Imperfect Electricity Markets: Need and Design," Utilities Policy 16(3): 159-70. https://doi.org/10.1016/j.jup.2007.10.003.

Joskow, P. (2013). "Symposium on 'Capacity Markets'." Economics of Energy and Environmental Policy 2(2): v-vi.

Joskow, P. and J. Tirole (2007). "Reliability and Competitive Electricity Markets," Rand Journal of Economics 38(1): 68-84. https://doi.org/10.1111/j.1756-2171.2007.tb00044.x.

Keay, M. (2016). "Electricity markets are broken-can they be fixed?" OIES EL17, at https://www.oxfordenergy.org/wp-cms/wp-content/uploads/2016/02/Electricity-markets-are-broken-can-they-be-fixed-EL-17.pdf. https://doi.org/10.26889/9781784670474.

KPMG (2017). Results that will blow you away: CfD allocation round two. September 2017, at https://assets.kpmg.com/content/dam/kpmg/uk/pdf/2017/09/cfd-allocation-round-two-energy-update.pdf

Lawson, N. (1992). The view from No. 11 : Memoirs of a Tory Radical, London : Bantam.

Lockwood, M. (2016). "The UK's Levy Control Framework for enewable electricity support: Effects and significance." Energy Policy 91 : 193-201. https://doi.org/10.1016/j.enpol.2016.07.026.

Mazzucato M. (2013). The Entrepreneurial Slate, Anthem 2013.

Mills, A. and R. Wiser (2013). Changes in the Economic Value of Variable Generation at High Penetration Levels: A Pilot Case Study of California. Lawrence National Laboratory. Berkeley CA, at https://emp.lbl.gov/sites/all/files/lbnl-5445e.pdf

NAO (2014). Early contracts for renewable electricity, National Audit Office, HC 172, at http://www.nao.org.uk/report/early-contracts-for-renewable-electricity/last accessed 20/7/15.

National Grid (2016). EMR Electricity Capacity Report, at https://www.emrdeliverybody.com/lists/latest%20news/attach-ments/47/electricity%20capacity%20report%202016_final_080716.pdf

National Grid (2017). Duration-Limited Storage De-Raling Factor Assessment: Final Report, at https://www.emrdelivery-body.com/SitePages/Home.aspx.

National Grid (2018a). Provisional Auction Results: T-4 Capacity Market Auction for 2021/22, at https://www.emrdelivery-body.com/Capacity%20Markets%20Document%20Library/Provisional%20T-4%20Results%20DY%202021-22.pdf

National Grid (2018b). Winter Review and Consultation, 11 June, at https://www.nationalgrid.com/sites/default/files/documents/2018%20Winter%20Review%20and%20Consultation.pdf

Newbery, D.M. (1995). "Power Markets and Market Power" The Energy Jounial 163: 41-66. https://doi.org/10.5547/ISSN0195-6574-EJ-Vol16-No3-2.

Newbery, D.M. (1998). "The Regulator's Review of the English Electricity Pool." Util. Pol. 7(3): 129-41, at https://www.semanticscholar.org/paper/The-Regulator%E2%80%99s-Review-of-the-English-Electricity-Newbery/5b58c0ba27ed-022045c4e703085332d7d7fd1461.

Newbery, D.M. (2000). Privatization, Restructuring and Regulation of Network Utilities (The Walras-Pareto Lectures, 1995), MIT Press, 2000, ISBN 0-262-14068-3.

Newbery, D.M. (2012a). "Contracting for wind generation." Economics of Energy & Environmental Policy 1(2): 19-36. https://doi.org/10.5547/2160-5890.1.2.2.

Newbery, D.M. (2012b). "Reforming Competitive Electricity Markets to Meet Environmental Targets." Economics of Energy & Environmental Policy 1(1): 69-82. https://doi.org/10.5547/2160-5890.1.1.7.

Newbery, D.M. (2015b). "Missing Money and Missing Markets: Reliability, Capacity Auctions and Interconnectors." Energy Policy 94: 401-410. https://doi.org/10.1016/j.enpol.2015.10.028.

Newbery, D.M. (2016a). "Towards a green energy economy? The EU Energy Union's transition to a low-carbon zero subsidy electricity system--lessons from the UK's Electricity Market Reform." Applied Energy 179: 1321-1330. https://doi.org/10.1016/j.apenergy.2016.01.046.

Newbery, D.M. (2016b). "Tales of Two Islands--Lessons for EU Energy Policy from Electricity Market Reforms in Britain and Ireland.' Energy Policy 105: 597-607. https://doi.org/10.1016/j.enpol.2016.10.015.

Newbery, D.M. (2018). "Evaluating the case for supporting renewable electricity", Journal of Energy Policy 120: 684-696. https://doi.org/10.1016/j.enpoI.2018.05.029.

Newbery, D.M., M. Pollitt, R. Ritz, and W. Strielkowski (2017). "Market design for a high-renewables European electricity system." Renewable & Sustainable Energy Reviews 91: 695-707. https://doi.org/10.1016/j.rser.2018.04.025.

Newbery D.M., and M. Grubb (2015). "Security of Supply, the Role of Interconnectors and Option Values: Insights from the GB Capacity Auction." Economics of Energy & Environmental Policy 4(2). https://doi.org/10.5547/2160-5890.4.2.dnew.

Ofgem (2010). Project Discovery: Options for delivering secure and sustainable energy supplies. February, at https://www.ofgem.gov.uk/ofgem-publications/40354/projectdiscoveryfebcondocfinal.pdf

Ofgem (2017). The impact of the Early Capacity Market Auction announcement on wholesale electricity prices and revenues, at https://www.ofgem.gov.uk/ofgem-publications/124340

Pollitt, M.G. (2012). "The role of policy in energy transitions: Lessons from the energy liberalisation era." Energy Policy 50: 128-137. https://doi.org/10.1016/j.enpol.2012.03.004.

Pollitt, M.G. and A. Brophy Haney (2013). "Dismantling a Competitive Electricity Sector: The U.K.'s Electricity Market Reform." The Electricity Journal 26(10): 8-15. https://doi.org/10.1016/j.tej.2013.11.001.

Renewable UK (2017). Offshore Wind Industry Investment in the UK 2017 Report on Offshore Wind UK Content, at http://c.ymcdn.com/sites/www.renewableuk.com/resource/resmgr/publications/OfFshore_Wind_Investment_V4.pdf

Schmalensee, R., P. Joskow, A.D. EUerman, and J.P. Montero (1998). "An Interim Evaluation of Sulfur Dioxide Emissions Trading." Journal of Economic Perspectives 12(3): 53-68. https://doi.org/10.1257/jep.12.3.53.

Tashpulatov, S.N. (2015). "Analysis of electricity industry liberalization in Great Britain: How did the bidding behavior of electricity producers change?" Utilities Policy 36: 24-34. https://doi.org/10.1016/j.jup.2015.07.004.

Taylor, S. (2016). The Fall and Rise of Nuclear Power in Britain: A history. Cambridge: UIT Press.

Timera Energy (2011). Electricity Market Reform: An orderly transition? At https://timera-energy.com/electricity-market-reform-an-orderly-transition/

Vazquez, C, M. River and 1. Perez Arriaga (2002). "A Market Approach to Long-Term Security of Supply." IEEE Transactions on Power Systems 17(2): 349-357, at https://ieeexplore.ieee.org/stamp/stamp.jsp?amumber=1007903. https://doi.org/10.1109/TPWRS.2002.1007903.

Michael Grubb (*) and David Newbery (**)

(*) Professor of Energy and Climate Change, Institute of Sustainable Resources, UCL London. E-mail: m.grubb@ucl.ac.uk.

(**) Corresponding author Director, EPRG, Faculty of Economics, University of Cambridge, UK E-mail: dmgn@cam.ac.uk.

(1.) Centrica, SSE plc, RWE npower, E.ON, Scottish Power and EDF Energy.

(2.) National Grid October 2014 Monthly Balancing Services Summary, Table 5.1.1. at https://www.nationalgrid.com/sites/default/files/documents/37743-MBSS_OCTOBER_2014.pdf.

(3.) EMR does not apply to Northern Ireland

(4.) See e.g. Symposium on 'Capacity Markets' (Joskow, 2013), particularly Cramton, Ockenfels and Stoft (2013), and earlier papers by Joskow (2008), and Joskow and Tirole (2007).

(5.) The US system had a long-term stable plan and allowed banking of permits to encourage investments, with considerable success (Schmalensee et al. 1998), features that the ETS signally lacked.

(6.) http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32009L0028:EN:NOT.

(7.) https://www.legislation.gov.uk/ukpga/2008/27/contents.

(8.) https://www.theccc.org.uk/tackling-climate-change/the-legal-landscape/the-climate-change-act/.

(9.) HM Treasury, Budget 2011, HC 836, March 2011. The government had adopted a Social Cost of Carbon (SCC) for public policy evaluation, following the Stern Report of 2006, expecting that the EU ETS would provide a carbon price in this range. As the ETS price sank the government in 2009 shifted to a shadow price of carbon. The shadow price of emission savings outside the ETS (like households and transport) would be the SCC, that covered by the ETS would be a shadow price starting closer to the 2009 ETS price ([pounds sterling]12/tC[O.sub.2]), but rising to converge with the SCC at [pounds sterling]70/tC[O.sub.2] (c. [pounds sterling] 250t/C) in 2030. The carbon floor price was thus targeted to make this 'shadow price' real in the electricity sector. For the subsequent evolution see section 3.5.

(10.) The force of base load' effectively grants an emissions allowance per MW of capacity, hut would force coal-fired stations to operate at a capacity factor of 50% or less.

(11.) E.On argued that a new coal plant would reduce emissions by displacing older, less efficient plants; and later, that it would be built 'capture ready' (i.e. to include CCS technology as and when it became commercially viable). After three years of intense controversy, the UK government 'deferred' a planning decision, and shortly afterwards the project was abandoned.

(12.) [pounds sterling](2005) 4.3 billion (Office of National Statistics).

(13.) Both authors have been members of the PTE but this paper only draws on published evidence.

(14.) Heavily criticized by the National Audit Office (NAO, 2014).

(15.) See https://www.theccc.org.uk/2011/08/09/confused-about-costs-of-nuclear-v-renewables-read-on, where the range of costs was given as [pounds sterling]40-100/MWh by 2030, whereas renewables were expected to cost [pounds sterling]75-135/MWh.

(16.) Author calculations, based on data for Round 2 vs Round 1 auctions: Capacity 3.3 GW vs 2.1 GW, and annual subsidy [pounds sterling]176m vs [pounds sterling]315m, given assumed electricity wholesale price of [pounds sterling]45.61/MWh. Source: BEIS, CfD Round 2 Auction results (https://www.gov.uk/government/publications/contracts-for-difference-cfd-second-allocation-round-results).

(17.) Of the total projected need for around 52.5 GW, 2.5 GW was held aside to ensure some room for a 1-year ahead auction in 2017, to provide scope for nearer-term adjustment, and shorter-term options like demand-side response (DSR).

(18.) Paid per kW of de-rated capacity per year

(19.) The demand curve is applied to determine the actual capacity procured if the auction clearing price differed from this: i.e. the auction would procure the targeted volume if it cleared at the price of the assumed CoNE, but buy a little more capacity it proved cheaper than expected, or a little less if it proved more expensive than expected.

(20.) National Grid took the high end of 53.3 GW on the basis that imports could not be relied upon (National Grid, 2014, p10-11).

(21.) At https://www.ofgem.gov.uk/publications-and-updates/embedded-benefits-impact-assessment-and-decision-industry-proposals-cmp264-and-cmp265-change-electricity-transmission-charging-arrangements-embedded-generators.

(22.) Part of the problem is that a "Loss of Load Event" is not actually a situation in which load is disconnected, but one in which the system operator can invoke various measures including voltage reduction, requesting generators to exceed their rated capacity for short periods, emergency imports, etc., none of which causes the lights to go out--most of which indeed consumers would never even notice.

(23.) Batteries were previously treated with high (96%) availabilities derived from pumped storage. National Grid (2017) published new derating factors so that storage with half (or one) hour has a derating factor a quarter (or half) of that previously, rising to 'firm' (96%) for 4-hour storage. The System Operator of the island of Ireland, using the appropriate concept of equivalent firm capacity, avoided such an egregious mistake.

(24.) E.g. The Telegraph 5/9/15 at https://www.telegraph.co.uk/finance/newsbysector/utilities/11844750/Electricity-network-in-uncharted-territory-as-blackouts-loom.html

(25.) National Grid (2018b) provides a useful comparison of forecasts with out-turns.

(26.) https://www.nationalgrid.com/uk/electricity/balancing-services/frequency-response-services/enhanced-frequency-response-eir

(27.) For an excellent concise briefing on the UK carbon floor price see http://researchbriefings.files.parliament.uk/documents/SN05927/SN05927.pdf.

(28.) https://www.carbonbrief.org/analysis-uk-cuts-carbon-record-coal-drop

(29.) quoted at https://publications.parliament.uk/pa/ld201617/ldselect/ldeconaf/113/11305.htm.

(30.) Tidal lagoons are presumably an exception, as building dams in a millennium-old skill.

(31.) Mazzucato (2013) refers to the role of 'market creation and shaping' as part of 'Mission-oriented innovation', whilst Grubb (2014) argues that such strategic considerations form a distinct domain of economic processes linked to evolutionary and system innovation theories, implying a role for government-led strategic investment and similar conscious design of markets to deliver strategic benefits, including induced innovation and associated option-building.

(32.) The possibility of an additional reactor at Sizewell and one at Wylfa are under active discussion, with possibly better designed Government-backed contracts. See e.g. https://www.ft.com/content/088cldd6-e40e-11e7-97e2-916d4fbac0da

(33.) See e.g. The Guardian 4/4/2104 at https://www.theguardian.com/environment/2014/apr/04/conservatives-promise-ban-new-onshore-windfarms

(34.) https://publications.parliament.uk/pa/cml99798/cmselect/cmtrdind/404iv/ti0406.htm

(35.) See Vazquez et al. (2002); Biddle (2005); Batlle et al. (2007).

(36.) http://www.sem-o.com/ISEM/General/Capacity%20Market%20-%20Final%20Capacity%20Auction%20Results%20Report.pdf

(37.) https;//www.semcommittee.com/sites/semecmmittee.com/files/media-files/SEM-16-026%20ACPS%202017%20Consultation%20Paper%20for%20Publication_0.pdf

(38.) Prices are currently capped at [euro]3,000/MWh, As the reliability standard is 8 hrs/yr, the missing money is ([euro]3,000-[euro]500)/MWh x 8 hrs/yr = [euro]20/kWyr,

(39.) https://www.gov.uk/govemment/statistics/final-uk-greenhouse-gas-emissions-national-statistics-1990-2015.

(40.) https://www.gov.uk/government/consultations/coal-generation-in-great-britain-the-pathway-to-a-low-carbon-future.

(41.) http://mission-innovation.net/

https://doi.org/10.5547/01956574.39.6.mgru
Table 1 : Administered Renewable Energy Prices Compared to First CfD
Auction

                       Capacity  Admin Strike price
                       (MW)      2014 ([pounds sterling]/MWh)

Large solar PV            72     [pounds sterling]120
Onshore Wind            1162      [pounds sterling]95
Energy from Waste CHP     95      [pounds sterling]80
Offshore Wind            750     [pounds sterling]140
Advanced Conversion       62     [pounds sterling]140
Technologies

                       Lowest auction clearing  Maximum %
                       price Jan 2015           saving

Large solar PV          [pounds sterling]79     34%
Onshore Wind            [pounds sterling]79     17%
Energy from Waste CHP   [pounds sterling]80      0%
Offshore Wind          [pounds sterling]114     18%
Advanced Conversion    [pounds sterling]114     18%
Technologies

Source: Simplified from Newbery (2016a, Table 1).
COPYRIGHT 2018 International Association for Energy Economics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Grubb, Michael; Newbery, David
Publication:The Energy Journal
Article Type:Report
Geographic Code:4EUUK
Date:Nov 1, 2018
Words:13383
Previous Article:Impacts of Technology Uncertainty on Energy Use, Emission and Abatement Cost in USA: Simulation results from Environment Canada's Integrated...
Next Article:Cleaner Nudges? Policy Labels and Investment Decision-making.
Topics:

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters