ISPI Energy Watch – Gennaio 9, 2015
On December 16th, the first-ever capacity market auction in Europe has been held in UK. Great Britain is the first Member State implementing a capacity market to ensure supply security in the long-term, for its electricity sector. The design of the British capacity remuneration mechanism, and the results of the opening auction, provides useful insights to the debate currently interesting Europe about resource adequacy. In the following, once illustrated the theoretical and empirical background concerning the adoption of capacity remuneration mechanisms, a discussion of the features and results of the British auctions is developed. The analysis highlights important aspects to be kept in mind by those countries in the process of adopting a capacity market, including the contradictions between European and national energy policies that may arise even in the case of a well-designed mechanism such the UK ones.
Capacity markets in theory and in practice
The UK has just held the first auction for its newly-introduced capacity market. A capacity market is a subset of a broader set of policy instruments, known as capacity remuneration mechanisms, that are aimed at providing a solution to the so-called “missing money problem”, a market failure that can occur in electricity markets.
The latter refers to the inability of day-ahead and ancillary services markets – perhaps the most important sessions in electricity markets as they are designed under EU liberalization directives – to provide appropriate price signals to promote the optimal level of investment in generating capacity. The adoption of caps to electricity spot prices, to avoid the exercise of market power, and of non-market mechanisms to procure reserve resources (as e.g. the reliability must run contracts), are some of the determinants of the missing money problem (here a more detailed discussion of the missing money issue). More specifically, these factors prevent electricity spot prices to reach the level of the so-called Value Of Lost Load (VoLL – i.e. the price that makes end-customers indifferent between consuming electricity at that price or not consuming it) during scarcity hours.
For the electricity industry, scarcity hours are of crucial importance. In particular, day-ahead markets work under a system marginal price rule, i.e. at any given moment, electricity is priced at a level that corresponds to the marginal generator’s marginal cost. Therefore, only when dramatic imbalances between electricity demand and supply occur, generators having invested in peaking generating capacity are able to recover a large portion of their fixed operating costs. Flaws in price formation during scarcity hours – as the adoption of caps to electricity spot prices or of out-of-market procurement mechanisms of reserve – or in the number of scarcity hours, could influence generators’ profitability to a significant extent.
Moreover, due to political concerns, the adoption of an administratively set price equal to the VoLL is unlikely to be feasible. End-customers would consider paying for price spikes as unfair. Estimated values of VoLL, indeed, range from 150 €/kWh to 3000 €/kWh. For this reason, to deal with the missing money problem – which, as seen, mostly concerns regulatory and political failures rather than a market failure – policy makers often prefer to adopt regulatory solutions aimed at setting capacity targets, more or less explicitly. That is, they prefer to opt for capacity markets – as well for other CRMs such as capacity payments and strategic reserves (see here for a taxonomy and description of CRMs) – instead of adopting an administratively set value of the VoLL, to promote investments in generating capacity and ensure long-term supply security.
The increasing and rapid penetration of intermittent renewable sources of electricity generation (such as solar and wind power), due to the pursuance of European climate and energy targets, has broadened the concept of the missing money problem and has often assigned to CRMs a different goal than that originally justified by the economic literature: CRMs have become a tool to support the survival of conventional power plants whose profitability has been largely impacted by the penetration of renewable generation.
The lack of programmability of, and limited ability to forecast, the electricity generation from wind and sun, indeed, requires the availability of a consistent amount of thermoelectric generating production – characterized by specific operating flexibility and efficiency standards – suitable to provide the back-up capacity needed to balance the electricity system in presence e.g. of unexpected reduction in eolic or photovoltaic generation. However, the higher the importance of conventional generation becomes (especially for the provision of balancing services), the narrower the room for worthy investments in thermoelectric plants turns to be. Actually, the almost null marginal cost of generation that characterizes intermittent renewables allows them to be strongly competitive in the day-ahead market merit order curve (here an illustration of the merit order curve) in a significant number of hours during the year. The higher the penetration of renewables, the higher the frequency of hours during which intermittentsourcesdisplaces thermoelectric power plants in supplying demand. For any given level of demand, by shifting the merit-order curve rightwards, subsidized renewables result on the one hand in reducing the volume of dispatched thermal energy, on the other hand in lowering the price and therefore the margins (see Figure 1 for an exemplification).
In the long-run, the decline in profits and the difficulty to cover fixed costs, may accelerate the closure of existing thermoelectric power plants – with consequent political and environmental issues such as unemployment and the recovery of power plants’ sites – while discouraging the development of new installations with possible effects with respect to supply security. This latter problem derives from the nature of intermittent renewables, that cannot reliably provide energy as it is demanded. Therefore conventional, programmable power stations may be, at the same time, needed to supplement renewables when they are not able to generate enough energy, and not able to operate profitably. Capacity markets are a possible answer to this gridlock. They involve the procurement of an amount of target capacity needed to ensure a given standard of system reliability – often defined in terms of hours of Loss of Load Expectation (see e.g. here for a definition of Loss of Load Expectation) – in a given future period called ‘delivery period’. Few years in advance, usually 3-4 years, the Transmission System Operator (TSO, henceforth) runs auctions to procure the target capacity from existing and new generating power plants as well as from technologies such as Demand Side Reduction programs (DSR, henceforth) and storage. Procurement occurs in the form of descending clock auctions, an iterative auction procedure in which the TSO announces a price – which in the first round corresponds to the price-cap of the auction – and bidders indicate whether they wish to supply their capacity at that current price. If at that price excess supply occurs, a new round of the auction is held and bidders express a lower price at which they would be willing to provide their availability during the delivery period. This process is repeated until supply equals demand. The clearing price of the auction determines the fixed premium bidders will receive during the delivery period to make their capacity available during those hours in which the electricity system experiences the risk of service disruptions. The obligation to deliver the committed capacity is enforced by means of monetary penalties, as it is the case of the UK and thePJM capacity market in the US, or implicitly, as it is the case for the Italian and the ISO-New Englandcapacity markets where the obligation has the form of a reliability option. The latter implies that the capacity contracts offered to generators in auctions have the form of a call option anchored to a strike price. The strike price is usually defined as the marginal cost of generation of the peaking technology of a wholesale market. Capacity providers forego the potential (but uncertain) revenues in hours in which the market price in the wholesale market is above the strike price, in exchange for the certain revenues of the option. Consumers pay the option premium and in return avoid prices above the strike price.
The European debate on capacity remuneration mechanisms
In Europe, the unintended consequences of renewables’ penetration, together with the launch of nuclear decommissioning programs in several countries, prompted a lively debate about the introduction of capacity markets in several Member States.
This debate led the European Commission to launch a set of Guidelines on Capacity Markets to precisely address the circumstances under which public intervention would be needed. The concerns of the European Commission are essentially twofold: to avoid the introduction of further unneeded incentives in electricity markets, and the introduction of heterogeneous national CRMs such to jeopardize the development of the internal energy market.
At present, a capacity market has been approved in France and Italy, while it is under discussion inGermany. Other countries, such as Spain, Portugal, Ireland and Italy (until the coming into force of the new capacity market) rely on a capacity payment to ensure long-term security of supply. Other countries, such as Finland, Sweden and Denmark rely on strategic reserves to pursue the same goal (Figure 2 below provides a detail on CRMs across Europe).
UK is the only Member State that actually implemented a capacity market to prevent the serious risks of unexpected service disruptions predicted by DECC and Ofgem in future years. The British mechanism – which covers Scotland, England, and Wales but not Northern Ireland – involves the management by National Grid of descending clock auctions 4 years in advance a delivery period of 1 year which runs from October 1st to September 30. Participants to the auctions – new, existing and refurbishing power plants, as well as DSR (included small storage and embedded generation) and storage – are admitted to participate only if they do not benefit of other types of incentives like the Renewable Obligation (RO), Contracts for Difference (CfDs), or small scale Feed in Tariffs (FIT).
Participants to the auctions cannot offer a price greater than 75£/kW/year during the first round of the auction. In subsequent rounds, until the auction clears, they may bid a price lower than 5£/kW/year below what they offered in a previous round. For existing assets a bid-cap of 25£/kW/year is established. New or refurbishing assets may bid above this threshold coherently with their different cost structure. New or refurbishing assets awarding an auction are entitled with a capacity contract of 15 and 3 years, respectively. Existing assets, differently, benefit of a capacity contract of 1 year. Contrarily than other capacity markets, such as the Italian ones and that of ISO-NE, the UK mechanism does not entail a price-floor to the bids presented by auctions’ participants.
The British capacity market: a hidden déjà vu?
The British capacity market provides important suggestions for those European countries that are in the process of implementing a CRM to deal with long-term security of supply. Specifically, the design of the UK mechanism identifies a good practice with respect to both technology neutrality and the promotion of competitive bidding behaviors.
With respect to technology neutrality, the British capacity market involved, indeed, the participation of several types of technologies (see Figure 2) such as coal, CCGT, OCGT, nuclear, hydro, and biomass generation; CHP; storage; and DSR (including small storage). Open auctions, consistently with the EC guidelines, promote the cost-effectiveness of capacity markets and, above all, avoid lock-in of fossil fuel generation which would be not consistent with the EU climate targets.
The British mechanism also displays fair characteristics with respect to the promotion of competitive auctions’ outcomes. The adoption of a simultaneous sealed bid auction and the absence of a price-floor to participants’ offers are important factors to promote competitive bidding behaviors. In particular, a price-floor may represent a focal point toward which the auction clearing price may converge and, in presence of excess capacity, a guaranteed subsidy even when the clearing price of the auction should be expected to fall to zero.
At the same time, the first auction held on December 16th provides interesting insights about the practical implications that capacity markets bring in terms of both the electricity market structure and the consistency of national and European energy policies. The results of the first auction – which procured 49.26 GW of capacity for the delivery period 2018/2019 at a clearing price of 19.40 £/kW/year – indicates that the clearing price of the auction has been mainly determined by existing CCGT (Combined Cycle Gas Turbine) power plants. Not surprisingly, this value is equal to the value of the fixed operative and maintenance cost of CCGT plants estimated by DECC in its Electricity Generation Cost Model.
The clearing price of the first British auction suggests therefore that existing CCGT power plants do not expect to realize significant margins on electricity markets and that they consider the capacity market an important further source of coverage of their O&M fixed costs. Put it differently, it seems that market participants expect that the impact of renewable generation on thermoelectric profitability will persist in the next years.
Secondly, Figure 3 and 4 reveal that other technologies, such as DSR and storage, are not yet able to compete with existing CCGTs for the provision of resource adequacy services. This issue is also proven by the fact that 70% of the DSR entered in the auction exited at a price well above the clearing price. Data on the amount and types of capacity exited by the auction also tells that almost 40% of new build assets that entered in the auction exited at prices above the clearing price.
Therefore, it seems that despite in theory the UK capacity market is not intended to support specific resources, in practice it turns out to benefit mostly existing CCGT power plants. This fact poses an intrinsic paradox with respect to the pursuance of the technology-neutrality principle and an evident contradiction with the wish of the EC to avoid “the lock-in (fossil) generation based solution that end up being stranded in the medium to long term when additional CO2 free capacity, interconnection capacity or demand and storage based solutions are expected to come on stream” (EC Guidelines on Capacity Markets).
The preliminary experience with the British CRM suggests, thus, that capacity markets may embed, even when accurately designed, the risk of a dejà vu: subsidies producing other subsidies as it has been the case for renewables incentives with respect to CRMs. The greater competitiveness of CCGTs, if supported by capacity markets – also in presence of increasing reserve margins during the delivery years following 2016/2017 – may further weaken the affordability of storage and DSR with the result that their wished contribution to the achievement of a more environmental-friendly electricity sector would be possible only if supported by new subsidies