Policy and Theoretical Implications of the Zero-subsidy Bids in the German Offshore Wind Tenders.

• Author: Greve, Thomas

1. INTRODUCTION

   Wind energy has been an important technology in the aim to reach the renewable energy targets, mostly to be met by 2030. (1) The exploitation of offshore wind is recent when compared to that of onshore wind. Nonetheless, it increasingly contributes to the European Union renewable energy targets (Dedecca et al., 2016). Total offshore installed capacity in Europe has increased from about 1,471 MW at the end of 2008 to about 18,499 MW in 2018 (Wind Europe, 2018a). Although most of the existing installed offshore capacity is concentrated in a few European countries (Belgium, Denmark, Germany, Netherlands and the UK), the interest in offshore wind farms is widening rapidly.

   Auctions have been widely used as an allocation mechanism to procure offshore wind projects. By 2017, six European Union Member States (Denmark, France, Germany, Italy, Netherlands and the UK) had introduced tendering procedures for procuring offshore wind projects (CEER, 2018). The recent offshore wind auctions incorporated some types of subsidy from the regulator/government to the developers, as an incentive mechanism. One example is from Germany that launched its first round of competitive tenders for offshore wind projects in April 2017, to construct and operate offshore wind farms. This tender is considered to be a turning point for the industry, given that of the four winning projects, three requested zero subsidies, relying only on the wholesale price (Huebler et al., 2017). As noted by the Executive Vice President and CEO of Orsted (2):

   "The zero subsidy bid is a breakthrough for the cost competitiveness of offshore wind, and it demonstrates the technology's massive global growth potential as a cornerstone in the economically viable shift to green energy systems." The German energy regulator launched its second round in April 2018. As was the case for the 2017 tender, the 2018 tender also had bidders requesting zero subsidies. The 2017/18 tenders were part of the 2017 Offshore Wind Act (EEG 2017), aimed at auctioning existing offshore wind installations to be commissioned between 2021 and 2025.

   From the perspective of mechanism design, the outcome of the German offshore wind tenders (hereafter, German tenders) can be regarded as being not far from optimal, if one defines optimality as the payment of the lowest amount of non-negative subsidies to the winning bidders. (3) Optimality is one of the key properties in mechanism design and auction theory, along with efficiency. (4) In terms of efficiency, the German tenders used a discriminatory (or pay-as-bid) auction format which, according to theory, does not achieve efficiency when units are identical and bidders have multiple demands. Indeed, optimality and efficiency are generally considered to be conflicting goals and a regulator usually needs to opt for one of them when designing a mechanism (Ausubel, 2003; Zhan, 2008).

   The main purpose of this paper is to analyze the theoretical and policy implications of the zero-subsidy bids submitted in the German tenders. Our analysis is threefold. Firstly, we describe the design, study the bidding strategies and the outcome of the German tenders.

   Secondly, we study an alternative design - a modified Vickrey-Clarke-Groves (VCG) mechanism - with the aim of exploring how the German auction could be re-designed to deliver both efficiency and optimality/zero subsidies. Our design is based on a key feature revealed in the 2017 German tender and its implications. The 2017 German tender gave bidders the opportunity to be awarded contracts of different time periods (contract lengths, i.e. 20, 25 and 30 years). The design of this feature was, however, not entirely clear to the bidders pre-auctioning, due to the lack of clarity regarding which bid(s) qualified for a 25- or 30-year contract. Nonetheless, this feature influenced the bidding strategies of the bidders. Our design incorporates the bidding strategies revealed in the tender, while allowing bidders to bid against a subsidy for periods of varying contract lengths with clear qualification rules. The mechanism asks each bidder to submit a vector of bids (i.e. a bid for each contract length/version) that is matched with the vector of willingness to pay (WTP) of the regulator. The mechanism chooses the winners and versions where social value is maximized (i.e., the sum of the regulator and bidders' surpluses is maximized).

   Lastly, we discuss key theoretical and policy implications relating to our analysis of the German tenders and our proposed re-design. In terms of theoretical implications, our re-design contributes to the literature on mechanism design and auction theory. Our design uses a VCG mechanism composed of a zero-subsidy benchmark and a vector of bids to achieve optimality and efficiency, together with strategy-proofness (i.e., truth-telling is optimal) and individual-rationality (i.e., utility from participation is non-negative). It obtains a social welfare optimum.

   Given the increasing investment in offshore wind and the use of auctions to select the developers, it is vital to draw some policy implications of our analysis. In particular, our paper contributes to the policy discussion on how governments or regulators may design their offshore wind tenders to obtain zero-subsidies, while not giving up on efficiency and without the use of reservation prices. This is particularly relevant given that recent offshore wind tenders are now being designed with the aim of ensuring only no-subsidy bids. The recent Dutch offshore wind tender, opened in March 2019, introduced a no-subsidy requirement. (5) This design choice - with a zero-subsidy requirement - may force an optimal outcome from the outset, but may have significant implications in terms of efficiency, as this may easily be lost. Hence, our paper also contributes to the discussion on the design of future offshore wind tenders. This is particularly important given that around 10 GW of offshore installed capacity is already set in the plans of a number of European countries to be procured via auctions during the period May 2019 to 2022 (Wind Europe, 2018b).

   The rest of the paper is organized as follows. Section 2 introduces the German tenders. In Section 3, we present our modified VCG mechanism. Section 4 discusses the theoretical and policy implications of the zero-subsidy bids in the German tenders and our proposed design. Section 5 concludes.

2. THE GERMAN OFFSHORE WIND TENDERS

   As of January 2017, new legislation in Germany came into force to expand the use of offshore wind energy through competitive tenders (German WindSeeG 2017, 2017), as part of a wider reform of the Renewable Energy Sources Act (German EEG 2017, 2017). (6) This legislation envisages two different stages of tenders: (i) auctions for existing projects (first stage); and (ii) auctions for sites (7) which have been subject to a preliminary investigation (second stage). (8) The first stage, also known as the transition period, involves the 2017/18 auctions with projects to be commissioned between 2021 and 2025. The second stage will involve the projects to be commissioned from 2026 onwards, where the bidders will compete for specific sites which have been subject to the preliminary investigation. The focus of the present paper is the two tenders held during the transition period - the 2017/18 German tenders. Below, we discuss the details of these tenders, in particular, the design and the outcome.

   Design. In both tenders, the German regulator asked each bidder to submit a volume in MW (a project) and a subsidy per kWh to be received in order to construct and operate the project. More specifically, each bidder was asked for a volume in MW and a subsidy per kWh in addition to wholesale electricity prices. The wholesale price was not part of the tender. Subsidies were capped at [euro]12 ct/kWh in 2017 and at [euro]10 ct/kWh in 2018, and negative subsidies were not allowed. (9)

   The format used was a discriminatory auction where the bidders submitting the lowest bids won the projects (or group of projects) they were interested in and thus, received the subsidy requested. In addition, the tenders provided the regulator with enough flexibility to award contracts of different lengths. A base contract was for a period of 20 years, however contract lengths of 25 years and 30 years were also available (German WindSeeG 2017, 2017; Huebler et al., 2017). According to Dannecker and Kerth (2017), in the 2017 tender documents, the regulator specified that these lengths were linked to the size of the subsidy submitted, but did not otherwise specify the qualifying criteria for these contract lengths. The 25-year and 30-year contracts were subject to an evaluation by the regulator after the end of the auction.

   Results. In 2017, a total of 1,490 MW was procured. (10) Four projects were accepted, three of which requested zero subsidies (amounting to 1,380 MW out of 1,490 MW). Zero subsidies meant that the developers have pledged to develop the projects without any subsidies, relying on the market rate for the energy they sell (Huebler et al., 2017; Huebler, 2018). The developer Orsted won three projects, two of which are to be developed without any subsidies and one with a subsidy of [euro]6 ct/kWh. The closing contracts, including the contract lengths, were, by 2018, yet to be decided (Orsted, personal communication, 7 February 2018). (11)

   In 2018, a total of 1,610 MW was procured. There were six winning projects. The developer Orsted won two, Iberdrola two and KNK Wind and Innogy won one project each. (12) The bids of Innogy and KNK have not been published, but two projects were awarded with zero-subsidy bids (one from Orsted and one from Iberdrola; Musgens and Riepin, 2018). The average winning bid was [euro]46.60 /MWh, which was 10 times higher than the average successful bid in the 2017 auction (Huebler, 2018). In contrast to the 2017 German tender, all...