Cross-border Effects between the Spanish and French Electricity Markets: Asymmetric Dynamics and Benefits in the Light of European Market Integration.

• Author: Urquijo, Ignacio Mas

1. INTRODUCTION

   Since mid 1990s, the EU has fostered the development of an internal market for electricity (IME) for socioeconomic and environmental reasons. In particular, in December 1996, the EU set the foundations for the creation of the IME after Directive 96/92/EC (First Energy Package) was passed (European Parliament, 1996). The IME pursues the liberalization of domestic electricity sectors, an energy transition, and market integration. By definition, electricity market integration is the process of progressively harmonizing the rules of two or more markets. An integrated market fosters competition between players across Europe, enhances security of supply, and increases efficiency, leading to a higher welfare. Interconnection capacity is considered to be a critical factor to ensure market integration (Figueiredo and da Silva, 2013). Additionally, a well interconnected grid is crucial for the decarbonization of the energy mix, as it allows to absorb a higher infeed from renewable sources and to share excess production with neighboring countries. Through market coupling, the electricity market is aligned with the physical electricity flows since neighboring electricity grids are physically interconnected and electricity follows all possible paths in reciprocal proportion to the resistance of each path.

   In 2014, the EU set a minimum interconnection target of 10% of installed electricity generation capacity for each member country before 2020 (European Council, 2014), and of 15% by 2030. Back then, there existed energy islands in the EU which lacked sufficient cross-border infrastructure. The status in 2014 is represented in Figure la, which relies on data from ENTSO-E (2014b). Spain was the European country with the highest interconnection deficit in Europe in 2014, with additional 6.8 GW required (mostly with France) to meet the 2020 goals (Martin Rivals, 2014). In this setting, and despite Spain's import dependence from France, electricity could not flow freely across borders. Although at EU-level the situation improved in the following years, by 2020 Spain was still the group's straddler with just 6% interconnection capacity. The status in 2020 is represented in Figure lb, relying on data provided in EC (2017).

   The Pyrenean geography represents one major barrier for the development of cross-border infrastructure, which adds to the limited interest of French state-owned EDF (Electricite de France) and its subsidiary, TSO RTE (Reseau de Transport d'Electricite), that already enjoy low generation costs given the high percentage of nuclear-based power. Furthermore, since France complied with the minimum 10% interconnection level already in 2014 through connections with the remaining adjacent countries, Spain would be the main beneficiary of further cross-border infrastructure developments.

   After more than 30 years without new interconnectors commissioned between Spain and France (Reuters, 2015), a key step towards market integration to Spain was the new underground interconnector passing the Pyrenees, commissioned in October 2015 by Inelfe--a 50-50 joint venture between Spanish TSO REE (Red Electrica de Espana) and RTE. The location is represented in Figure 2, relying on information provided by Inelfe (2015a) represented on a map by Google Maps (2019). Although the project was not sufficient for Spain to reach its interconnection target, it represented a key milestone for the integration of the European electricity market. It held a power record of 2,000 MW from VSC (voltage source converter) technology, and with 64.5 km length also set a record as the world's longest underground high-voltage cable at that time. The initiative required an investment of EUR 700 m with EUR 225 m obtained from European funding (Inelfe, 2015b). Two 1,000 MW cables boosted transmission capacity from 1,400 to 3,400 MW, thereby increasing the interconnection level of Spain to 6% (Inelfe, 2015b) and allowing France to become Spain's main trading partner (Table 1).

   In addition to the progress with respect to physical integration, as part of the Price Coupling of Regions (PCR), the Multi Regional Coupling (MRC) couples South-Western Europe (SWE, including Spain) with North-Western Europe (NWE, including France) since May 2014. PCR seeks to develop a single coupling solution for European day-ahead markets through implicit auctioning, thereby leading to higher price convergence (EPEX SPOT, OMEL and NORD POOL, 2010). It couples day-ahead markets of more than 20 countries from Portugal to Finland (ENTSO-E, 2014a).

   In this paper, we shed light on the implications of EU energy policies concerning the integration of Spanish and French electricity markets by analyzing changes in the price formation process in each country, in the presence of cross-border effects. In particular, we show the merit order effect (MOE) and the integration of renewable energies (RE) and their impact on prices. An understanding of the joint market dynamics is highly relevant, given the historical and still prevailing isolation of the Iberian Peninsula: Spain's interconnection capacity in end-2020 remained at 6%, thus missing the 10% target, and no new interconnection projects have been commissioned since 2015 as of February 2022.

   Several researchers have provided evidence for the impact of European energy policy on the creation of an integrated IME. One example in this sense is the study of Zachmann (2008), who applied a principal component analysis of wholesale electricity prices in eleven European electricity markets, including Spain, and rejected the hypothesis of full market convergence against bilateral cointegration only. Similarly, Frauendorfer et al. (2018) confirmed the cointegration of electricity prices in the German and Swiss markets between 2011 and 2016. Gugler et al. (2018) analyzed 25 European electricity spot markets between 2010 to 2015, and controlling for cross-border capacity, congestion, and market coupling, found an increasing harmonization until 2012. The degree of cointegration weakened since then due to the increasing share of renewables. However, Spain appears to be less integrated than the rest of its European peers. For instance, Robinson (2007) applied [beta]-convergence and cointegration tests to electricity prices of nine European countries between 1978 to 2003, and concluded convergence between most countries except for the United Kingdom, France and Spain. Similarly, Bower (2002) analyzed 15 European electricity markets in 2001, and proved cointegration between 14 of them except for Spain, due to the country's low cross-border transmission capacity and congestion. More recently, Keles et al. (2020) examine the interdependencies between the Swiss electricity market and those of neighboring countries in the light of new grid connections and market coupling regulations, finding a seasonal correlation with the German and the French prices in summer and winter, respectively.

   Previous studies have shown the effect of RE on electricity prices in different markets. For instance, Paraschiv et al. (2014) employed a dynamic fundamental model and disentangled the effect of wind and solar power generation in the German market, finding evidence for a continuous electricity price adaption process to market fundamentals and a negative marginal effect of renewable energies. They found that the promotion of renewables in Germany entailed higher costs for final consumers. Clo et al. (2015) employed Prais-Winstein estimation regressions in order to analyze the effect of wind and solar on Italian wholesale electricity prices, and concluded a higher marginal effect of solar infeed. Similarly, Denny et al. (2017) found evidence of positive net savings resulting from wind generation in the Irish electricity market. Also for the case of Spain, Saenz-de-Miera et al. (2008) and Azofra et al. (2014) confirmed net savings and the MOE from wind in the day-ahead market from 2005 to 2007 and for 2012, respectively. Further, Bockers et al. (2013) used a structural vector-autoregressive model to test the MOE of solar and wind power on the Spanish market between 2008 and 2012, finding that the effect is mainly driven by wind power. A series of recent studies has been focusing on cross-border effects on electricity prices in integrated European markets. For instance, Frauendorfer et al. (2018) applied a dynamic fundamental model and found cross-border effects of German wind and solar generation on Swiss prices. This has also been explored in Keppler et al. (2017), who analyzed cross-border effects on the spread between electricity prices in France and Germany between 2009 and 2013. Although significantly mitigated by market coupling, they found a negative effect of intermittent electricity production on price convergence, caused by the limited transmission capacity.

   The literature review shows limitations of Spain's energy market integration with the rest of continental Europe. However, the related research is limited, as it does not take into account the most recent developments in the market integration to France. We aim at closing the literature gap and provide revised results. As far as we know, no previous research has investigated the cross-border effects at the Spanish-French border.

   Building on the findings in previous research presented above, in our study we will test the following hypotheses:

   * A) Price convergence: Convergence between Spanish and French electricity prices has increased in the wake of market integration.

   A.1) Both enhanced grid infrastructure and market coupling increase price convergence. Especially enhanced grid infrastructure leads to higher convergence.

   A.2) Price convergence is achieved gradually across the three main periods of Spanish-French market integration.

   A.3) Joint price dynamics on a granular scale are still weak, given the still limited integration between the markets.

   * B) Price formation: Given the...