Putting renewable energy auctions into action – An agent-based model of onshore wind power auctions in Germany

Abstract

The following analysis looks into auctions for renewable energy, specifically onshore wind power in Germany. Following an agent-based modeling approach, the two most commonly applied auction pricing rules are compared (uniform and pay-as-bid) and first conclusions on outcomes are drawn for future policy design. The auctions are modeled to closely represent the auction design foreseen in the German Renewable Energy Sources Act (EEG, 2017) and replicate their parameters.The analysis draws on auction theory. For both pricing schemes, individually rational agents with independent valuation are assumed. As support for renewable electricity through auctions is to be established permanently and auction rounds will be held multi-annually, a further focus lies on agents learning over time by adapting their behavior to new information.The model results show that pay-as-bid exhibits lower prices and thus support costs than uniform pricing, whereas allocative efficiency suffers under pay-as-bid. Over time, one can observe a decline in the strike price, which is due to learning effects, whereas agents' profits increase in the course of the auctions. Furthermore, smaller actors will experience difficulties and agent diversity is likely to suffer in the long term, if this is not accounted for in other ways.