Impact of Coordinated Capacity Mechanisms on the European Power Market.

• Author: Bucksteeg, Michael

1. INTRODUCTION

   Over the last two decades, power markets in Europe have been liberalized and competition has been introduced into the wholesale markets. At the same time a European emissions trading scheme has been introduced and renewable technologies have been supported in most European countries. As a consequence previously regulated utilities face a more dynamic world with various uncertain and stochastic parameters. In this new environment, investment decisions in generation facilities are based on revenues gained in the energy-only market and are therefore dependent on adequate price signals. Moreover, they are subject to insecure future conditions notably with respect to the development of renewable capacities, electricity demand, C[O.sub.2]- and fuel-prices as well as regulatory uncertainties (cf. Hasani and Hosseini, 2011).

   In theory, competitive energy-only markets incentivize optimal investments in new generation capacities in line with the peak load pricing approach (cf. e.g. Boiteux, 1960 or Stoft, 2002). In practice, market imperfections however lead to a violation of the assumptions behind the peak load pricing theory. Especially, the long lead times for generation investments and the absence of demand response prevent reaching a situation of market equilibrium. Moreover, if price caps in the energy markets are set too low, peaking units are not able to cover their investment costs during critical hours with peak demand, leading to the missing money problem (cf. Joskow, 2008). As a result there are reasonable concerns about insufficient investment incentives in restructured electricity markets, which would lead to supply shortages and endanger system adequacy. While in theory the adequacy problem can be avoided by sufficient short-term demand elasticity, in practice electricity markets are characterized by low demand flexibility at present. Thus, in order to ensure adequate long-term allocation of generation capacities, different capacity mechanisms have been proposed and implemented in recent years (cf. e.g. Cramton and Ockenfels, 2012).

   In general, volume- and price-based capacity mechanisms may be distinguished. Under the latter mechanisms, revenues for (firm) capacities are increased through capacity payments set by the regulator. To induce sufficient investments in line with the intended reliability level, the set capacity remuneration should not exceed the Value of Lost Load (VoLL) reduced by the revenues from the energy market (with capped prices). The VoLL can be defined as the willingness to pay of consumers to avoid a supply disruption (cf. e.g. Cramton et al., 2013). Under volume-based mechanisms, the regulator on the contrary stipulates the required capacity that optimizes the overall system costs including costs of not being serviced. This capacity is then procured through auctions (or contracts). The last bid accepted to cover the required capacity would then typically become the capacity price paid to all generators, on top of the revenues from the energy market. For a more detailed discussion on fundamentals of capacity mechanisms the interested reader is referred to Cramton et al. (2013), Cramton and Ockenfels (2012), Joskow (2007), and Joskow (2008).

   The restructuring of power markets worldwide has led to a variety of capacity mechanisms. In Latin America, capacity payments have been predominantly implemented (cf. Galetovic et al., 2015). Primarily due to the large share of hydro generation, the Colombian regulator however introduced auctions for firm energy obligations that require the providers delivering firm energy in scarcity situations (cf. Mastropietro et al., 2017). In the United States, mainly capacity markets have been introduced from the beginning of the liberalization, e.g. in ISO New England and PJM. Other countries (e.g. New Zealand and Australia) that introduced capacity mechanisms at a later stage have opted for strategic reserves (cf. Galetovic et al., 2015). Likewise, strategic reserves operated by the system operator have been implemented in several European countries (e.g. Belgium, Sweden, and Finland, see also Figure 1). In order to avert a capacity shortage, the system operator contracts a certain volume of generation capacity. This reserve capacity is made available to the electricity market at a price significantly higher than the marginal cost of generation during scarcity hours. While in many European countries (e.g. Germany and Belgium), there is ongoing debate on the introduction of full-fledged capacity markets, only a few countries (United Kingdom, France, and Italy) recently implemented or are implementing capacity markets (cf. e.g. Creti et al., 2012; Newbery, 2011). In this context, capacity markets can be understood as a market instrument to allocate and provide the level of generation capacity that optimizes the duration of supply shortages in line with a specified reliability standard. Other countries (Ireland, Spain, and Portugal) implemented price-based capacity mechanisms with capacity payments. Moreover, capacity markets are considered in Spain and Poland, but have not yet been specifically planned. While capacity mechanisms are in place in several European countries, only few of them defined a reliability standard. In Belgium, France, Italy and United Kingdom the duration of supply interruptions, also referred to as Loss of Load Expectation (LOLE), should not exceed 3 hours per year. Greece set a LOLE of only 2.4 hours per year, whereby Portugal, Ireland and Hungary aim for a LOLE of 8 hours per year (cf. ACER, 2016).

   Cramton and Ockenfels (2012) note that strategic reserves would not reduce investment risks and might distort efficient energy market pricing in scarcity situations. Accordingly, strategic reserves would likely suppress new investments and might even compromise long-term reliability of a power system. Cramton and Ockenfels (2012), de Vries and Heijnen (2008) and Khalfallah (2011) note that capacity markets perform best in terms of a lower price volatility and system reliability. Consequently, we focus on capacity markets within this paper and analyze the impacts on the European power market.

   In recent years mainly national debates brought out several and heterogeneous capacity market models (cf. e.g. ACER, 2016; Creti et al., 2012; Newbery, 2011). However, as the EU member states aim for an internationalization of the electricity sector and an integration of national power systems, the implementation of unilateral national capacity markets is not expedient. The target model for the implementation of the internal market focuses on an integrated energy market but does not foresee a capacity market design (cf. European Commission, 2011; European Commission, 2016). While several studies investigate the optimal design of capacity mechanisms using simplified test systems (e.g. Lynch and Devine, 2017; Khalfallah, 2011), so far few studies have addressed the impacts of capacity markets in interconnected power markets. Cepeda and Finon (2011) study the impacts of heterogeneous market designs on generation adequacy using a stylized two-country model. Their results contradict the intuition that in the case of unilateral capacity mechanisms countries without adequacy policies would free-ride in the long-term. Heterogeneous capacity mechanisms rather lead to negative externalities compromising the efficiency of neighboring energy-only markets. Meyer and Gore (2015) apply a game-theoretic two-country model and analyze cross-border effects of strategic reserves and reliability options. They find that unilateral capacity mechanisms lead to competition effects increasing benefits for consumers in both countries but reducing producer surplus in the country without capacity mechanism. Accordingly, the missing-money problem would be exacerbated leading to decommissioning of generation capacities. However, in both studies the impacts on a large scale power system like the European one and the resulting level of security of supply are not investigated.

   Tasios et al. (2014) deploy a model of the European electricity market and compare EU-wide symmetric (homogeneous national) and asymmetric (unilateral national) capacity mechanisms. They also find that an asymmetric implementation of capacity mechanisms induces market distortions and inefficiencies. However, they do not model generation investments endogenously and apply capacity payments based on the missing money resulting from a wholesale market simulation, hence ignoring interactions between energy-only and capacity markets.

   The present paper analyses the impacts of different capacity market designs on the European power market and adds to the current literature by filling the following gaps: First, in addition to symmetric and asymmetric capacity markets we also consider two EU-wide coordinated capacity market designs, which are in line with the internal electricity market. To the authors knowledge so far several studies have addressed the effects of symmetric and asymmetric capacity markets (cf. Tasios et al., 2014; Cepeda and Finon, 2011), but not the impacts of coordinated capacity markets. Second, we apply a probabilistic approach for a consistent determination of capacity requirements. For the cases of coordinated capacity markets this approach is extended to a multiregional or transnational level. Third, we assess the long-term development with an electricity market model (E2M2s) in order to consider interactions between energy-only and capacity markets. The model covers the European power system and generation investments are modelled endogenously (cf. Spiecker et al., 2013 and Spiecker et al., 2014). Besides economic effects, the impacts on security of supply utilizing the probabilistic approach are also analyzed.

   The remainder of this article first describes the applied methodology in section 2. A probabilistic approach for the determination of capacity requirements...