Market-Based Redispatch May Result in Inefficient Dispatch.

• Author: Grimm, Veronika

1. INTRODUCTION

   Today, in most industrialized countries electricity markets are partly liberalized. Typically, markets govern supply and demand while regulated transmission system operators (TSOs) operate the network. As electricity wholesale prices at the day-ahead spot markets often do not entirely reflect network constraints, the TSO is typically engaged in congestion management ex-post spot market trading. (1) In the case where spot market allocations yield infeasible network flows, the TSO intervenes ex-post and adjusts the traded quantities to restore network feasibility. These short-run operations are called redispatch operations; see for instance Oggioni and Smeers (2013).

   To determine allocations and reimbursements of firms in the redispatch procedure two types of redispatch systems are mainly implemented and discussed in the literature. Under a cost-based redispatch system (CBR)--as it is applied in Austria, Switzerland, or Germany (2)--variable cost of production is the basis for redispatch payments; see Trepper et al. (2015), Grimm et al. (2016a). or Grimm et al. (2016b) for more details. As a CBR compensation is purely based on the incurred short-run cost of redispatched producers, it clearly aims at minimizing congestion management cost of the TSO. In contrast, under market-based redispatch or counter-trading (MBR) the TSO procures redispatch quantities at the different nodes in a market environment. Depending on the specific pricing rules (locational marginal pricing or pay-as-bid pricing) the compensation of market participants can be different from short-run cost. Versions of counter-trading are applied in the Nordic market (comprising Denmark, Finland, Norway, and Sweden), in the Netherlands, and in the United Kingdom; see Bj0rndal et al. (2003), Dijk and Willems (2011), Glachant and Pignon (2005). Green (2007), Inderst and Wambach (2007), or Mirza and Bergland (2015).

   There is an intensive academic and political debate regarding the desirability of the different redispatch systems. On the one hand, the different redispatch systems induce different revenues for the market participants. One intensively discussed argument in this context relates to potentially improved long-run investment incentives of market participants in case of MBR; see Inderst and Wambach (2007) or Haucap and Pagel (2013). Private investments in generation capacity are typically driven by expected future profits. In Grimm et al. (2016a). Grimm et al. (2016b), Grimm et al. (2017), Grimm et al. (2019), or Weibelzahl and Martz (2020) it is shown that the chosen market design highly influences optimal long-run investments. In contrast to a CBR regime, where firms cannot realize profits by being redispatched. redispatch prices under MBR often exceed the firms' variable production cost; see de Vries and Hakvoort (2002). Note that our formal analysis does not directly contribute to the debate regarding long-run investment incentives of market participants but focuses on the short-run perspective. Nevertheless, we jointly discuss those issues at the end of the article (see Section 6).

   On the other hand, the different redispatch systems induce different levels of redispatch cost for the TSO. The present analysis focuses on potentially changed incentives of the TSO to chose redispatch adjustment quantities. TSOs in electricity markets are typically regulated based on an incentive regulation which establishes limits for the prices that can be charged to customers for transmission services. A reduction in cost incurred by the TSO allows for larger profits, providing incentives for an efficient cost reduction. In many cases, TSOs thus have incentives to minimize their spendings, including those resulting from the redispatch process. (3) For example, in the UK the TSO National Grid uses redispatch measures that account for the cost-effectiveness of the used redispatch resources; see Neuhoff et al. (2011), Konstantinidis and Strbac (2015), and National Grid (2020). Also in Germany. TSOs are incentivized to minimize their network costs, including redispatch cost, according to EnWG (2017) [section] 1, the "Incentive Regulation Ordinance" (ARegV, 2017). (4) and the "voluntary self-commitment for redispatch" by the TSOs (German TSOs. 2018). As our analysis indeed shows, an incentive regulation inducing the minimization of redispatch cost as the objective of the TSO can be highly problematic in case of MBR since it may result in distorted redispatch choices. In contrast, in case of CBR, incentives of the TSO to minimize redispatch cost yield undistorted redispatch decisions.

   To analyze those important issues in more detail, the paper at hand introduces a model that allows us to assess the short-run impact of the different redispatch regimes on the redispatch decisions taken by the TSO in liberalized electricity markets. We consider a day-ahead, energy-only spot market with a uniform market price for the case of elastic spot market demand and multiple generation technologies at the different network nodes. Subsequent redispatch is applied to deal with network congestion. Assuming that the TSO is incentivized to minimize redispatch cost, we explicitly compare CBR to two different variants of MBR. The model is applied to a simple two-node network as it is often used in the literature (see for instance de Vries and Hakvoort, 2002, Dijk and Willems, 2011, Holmberg and Lazarczyk, 2015, or Hirth and Schlecht, 2018), and also to a setting with three nodes and more complex physical network constraints.

   For the case of only two-node networks, we show that both CBR and MBR result in identical welfare-maximizing outcomes, which is in line with the existing literature (see, e.g., de Vries and Hakvoort, 2002, and Holmberg and Lazarczyk, 2015). As a main result, for networks with at least three nodes we demonstrate that in contrast to a CBR mechanism, redispatch cost minimization of the TSO may not always imply welfare maximization in the case of MBR. We show that the TSO might have incentives to decrease MBR redispatch cost at the expense of market efficiency. Based on this finding, we finally emphasize the importance for energy markets that use MBR to establish a regulation where the TSO is obliged to implement the welfare maximizing (instead of the redispatch cost minimizing) dispatch. This would result in efficient outcomes as under a CBR regime. Observe that our results do not require the assumption of strategic firms, but already hold for the standard case of perfect competition. (5)

   Note that the adoption of properly functioning redispatch regimes as analyzed in the present article is of highly political relevance. In Germany, for example, the German Monopolies Commission repeatedly discussed the introduction of alternative redispatch mechanisms including MBR during the last years; see Monopolkommission (2009), Monopolkommission (2011), or Monopolkommission (2013). In 2015, the Oberlandesgericht Dusseldorf (Higher Regional Court of Dusseldorf) also concluded that in Germany the current CBR rules must be adjusted in order to account for foregone revenues of redispatched firms; see Oberlandesgericht Dusseldorf (2015). (6) In contrast, the Dutch Ministry of Economic Affairs introduced an MBR system, arguing that one reason for the introduction is that the corresponding long-run investment signals would induce additional market entry. This might then increase the level of competition in the market and thus increase social welfare by reducing price-cost margins; see Dijk and Willems (2011). In the Clean Energy Package of the European Union, MBR is defined as a fundamentally binding principle in congestion management, from which deviations are only permitted under certain conditions. (7) For this reason, in Germany the project "Study on the procurement of redispatch" commissioned by the Federal Ministry of Economics and Energy examined the effects of a change of the existing CBR to an MBR regime and argued against such change; see Consentec (2019). (8) These examples highlight the relevance of the two redispatch systems in current political and scientific discussions on electricity markets. Furthermore, they present possible arguments for a change from CBR to MBR. In this context, our analysis most importantly highlights that switching the redispatch regime may have to be accompanied by a careful adjustment of the incentive scheme of the system operators.

   1.1 Literature

   Despite the fact that a lot of studies deal with the effects of redispatch (see, e.g., Oggioni et al., 2012, Oggioni and Smeers, 2012, Kunz, 2013, Oggioni and Smeers, 2013, Perninge and Soder, 2014, Kunz and Zerrahn, 2015, Holmberg and Lazarczyk, 2015, or Kunz and Zerrahn, 2016), only few authors have explicitly analyzed the differences between CBR and MBR mechanisms. (9) An exception is Knops et al. (2001), who consider different congestion management methods and compare MBR to CBR. The authors mainly focus on institutional and basic system requirements that may be quite similar for the two redispatch systems. In an extension, de Vries and Hakvoort (2002) explicitly focus on the question of efficiency of the two redispatch mechanisms. In a two-node network with inelastic demand, the authors find that CBR and MBR are both economically efficient in the short run. More recently, these results are affirmed in Holmberg and Lazarczyk (2015). The authors show that MBR results in an efficient dispatch, assuming inelastic demand and continuous, strictly increasing marginal cost of production at each network node. Given this short-run equivalence findings, existing studies often use clear-cut CBR models to formulate MBR mechanisms; see for instance Oggioni et al. (2012) or Oggioni and Smeers (2013). Observe that our model may be seen as an extension to the above-mentioned studies that focus on a simpler network and generation structure or that assume the TSO to maximize welfare or...