The Social Efficiency of Electricity Transition Policies Based on Renewables. Which Ways of Improvement?

• Author: Villavicencio, Manuel

1. INTRODUCTION

   Over the past ten years, several European countries, engaged in climate and energy policies as prescribed by the European Commission, (1) have launched official goals promoting renewables energy (RE) sources and C[O.sub.2] emissions offsets in joint policy packages. Real concerns with reducing C[O.sub.2] emissions exists, but climate policies struggle to translate into effective implementation. In practice very few reforms have been made to use direct carbon policy instruments and only indirect instruments are effectively implemented (e.g., technology-oriented policies), so offsetting C[O.sub.2] emissions no longer appears as the priority goal. This justifies the showcasing of policies in the power sector targeting ambitious development of renewables in many countries: 80-100% of total production in 2050, passing by a level close to 50% in 2030-2035.

   The objective of promoting renewables is reinforced by the fact that this responds to other tangible goals on the industrial sector in Europe, i.e. job creation, reinforcing industrial tissue, or ethical and cultural positions (energy self-sufficiency, supply decentralization, but also damming up existing nuclear production, maybe eventually abandoning it, due to technological risks that are perceived as too high). The German case is a good example: after deciding a complete nuclear energy phase-out by 2022, the objective of developing REs at 80-100 % responds to a clear need, especially if confidence is lacking with respect to methods that directly reduce C[O.sub.2] emissions without limiting fossil fuel generation. Yet it is well worth questioning the environmental effectiveness and economic efficiency of such policies targeting very high levels of renewables, mainly through variable wind and solar.

   The pursuit of this objective requires the use of support mechanisms that guarantee the long-term revenues of investors in variable renewables (VREs) forcibly has an opportunity cost, which is that of the de-optimisation of the electricity mix with respect to a first-best solution that would result from a the market coordination or the optimisation of a social planner (equivalent in theory) introducing a carbon emission constraint to account for environmental externalities (Abrell, Rausch, and Streitberger 2019b; Abrell, Kosch, and Rausch 2019; Percebois and Pommeret 2019). The opportunity cost should grow in function of the level of the renewable energy shares targeted by the mandates.

   Because of the growing need for providing firm capacity and flexibility as back-up to variable renewables, flexible dispatchable technologies are required (de Sisternes, Jenkins, and Botterud 2016; Abrell, Rausch, and Streitberger 2019a). In the current market landscape, flexibility is mainly provided by fossil fuel technologies (mainly gas-fired power plants, but also coal-fired plants), which could translate into less C[O.sub.2] emissions savings than expected. Electricity storage and demand response are also possible solutions for providing system flexibility, but their optimal deployment remains uncertain due to still important challenges dealing with planning, market design and business models of such technologies (Zerrahn, Schill, and Kemfert 2018; Debia, Pineau, and Siddiqui 2019; Burger et al. 2020; Percebois and Pommeret 2019). This invites us to question the coherence of policies targeting C[O.sub.2] reductions (ceiling on total emissions) with high shares of renewables specific to the electricity sector, to correct for eventual divergencies.

   The addition of economic instruments for the pursuit of a single objective is criticized in theory as a source of inefficiency with reference to the Tinbergen's golden rule (2) (Tinbergen 1952), according to which, in political economy, the realisation of a given number of objectives requires the utilisation of the same number of instruments, and not more. Any policy that goes around this rule can only introduce additional costs with respect to a policy that is based on a single instrument. Thus, combining a RE obligation with a carbon constraint leads by definition to second-best solutions of which the results will be distant from the first-best found through the implementation of direct C[O.sub.2] policies like cap and trade mechanisms. Theoretical discussion on first and second-best solutions have been recently provided by Abrell, Rausch, and Streitberger (Abrell, Rausch, and Streitberger 2019b). Therefore, the results from combining climate--energy policies could be ranked on a hierarchical equilibrium framework.

   In terms of climate--energy policy, a couple of instruments promoting electricity technologies that aim to reduce the total emissions necessarily leads to economic inefficiency, because it turns its back on the principle of equimarginality between alternative low carbon options. In addition, the complexity introduced by the development of VREs in power systems may not have the effect of leading to a simple and monotonous reduction of C[O.sub.2] emissions. [A.sub.t] this stage it is worth taking into consideration what can be brought by new sources of flexibility, different types of electrical storage and demand response (including direct load control), as these can reduce the environmental effects of policies targeting high levels of VRE and facilitate its integration into the power system. Promoting storage and direct load control can be an important element in policies that prioritize the development of VREs.

   Most studies that evaluate climate-energy policies in the power sector concentrate on the feasibility of objectives defined in terms of renewable energy penetration and/or energy efficiency goals, off to the side of the primary objective of offsetting C[O.sub.2] emissions. But they do not question the economic rationality of these objectives nor their effectiveness in limiting total emissions. Based on the seminal analyses on economic policy developed by Tinberghen (Tinbergen 1952) and Thiel (Theil, Barten, and Van den Bogaard 1964), we propose a framework of analysis to design and test energy policies aiming at decarbonization objectives through renewable energies on the power sector. The methodology consists on finding the best set of instruments (policies and/or incentives) to attain a set of goals (targets) at the lowest system cost. In the case of climate-energy packages the targets are commonly defined as C[O.sub.2] emissions savings, security of supply compliance and cost affordability, while the policy instruments are the application of reliability standards, "cap and trade" mechanisms on C[O.sub.2] emissions (or taxes) and renewable energy standards (or mandates). Therefore, the power system is represented as a partial equilibrium problem under system cost minimization subject to climate-energy policy constraints, so targets are obtained from the resulting equilibrium states.

   Apart from the technical constraints required to represent the operational limits of the different technologies, by optimizing the system subject to a policy constraint (e.g., a goal on C[O.sub.2] offsetting) this approach leads to first-best solutions with respect to an unconstrained system with environmental externalities. When an additional constraint is simultaneously targeted, which is here that of a mandate on renewable energy shares, our methodology allows to calibrate the joint policies such that their combination can be the most effective in terms of environmental performance and the closest from first-best solutions.

   To appropriately evaluate climate--energy policy packages targeting the electricity sector at the 2050 horizon we use a detailed expansion model of the power system which allows us to seize the challenges introduced by the large-scale development of variable renewable energies (VREs) in terms of investment and system operations.

   The complexity of power systems subject to large-scale deployment of VRE is a major difficulty in sweeping through the different combinations of climate-energy policies to identify hierarchical equilibriums in terms of economic and environmental efficiency. Multiple dependencies and non-linear relations between variables and targets exist as a result of non-convexities and jointness in the constraints. Investment choices and market coordination on the operation of different types of technologies to guarantee long-term security of supply and ensure the stability of the system are rendered very complex by the variability of VRE infeed. The interactions within the operations of different generation technologies multiply rapidly. Furthermore, the development of VRE capacity triggers dynamic effects that entail technical externalities (3) and increases integration costs (Giulietti et al. 2018; Hirth, Ueckerdt, and Edenhofer 2016), effects that are not present with technologies available on-demand as with fossil and nuclear technologies. When evaluating a climate-energy policy package, all these effects must be resituated in a modelling framework.

   Indeed, precision and detail while representing these interactions become an imperative necessity to proceed to rank the hierarchical equilibriums obtained so as to define and calibrate the right set of policy instruments to efficiently achieve the targets while accounting for interdependencies. This is what is pretended with the methodology here presented. (4)

   As D'haeseleer et al. (D'haeseleer et al. 2017) judicially underline on the complexity of the aftermaths of policies shaping the power system:

   "... policy and regulation often have unexpected and, possibly counterproductive effects on overall system performance. It should, therefore, be a part of good policy making to first study the overall system by modeling its different parts, with much emphasis on the interactions among the different subparts as well as among different policies. As the behavior of the system regulated by electricity market will be strongly...