Search

Abstract: In this paper, an agent-based simulation of a Financial Transmission Rights (FTR) auction for both types of point-to-point FTRs, obligation and option, is presented. Each auction participant is simulated as an adaptive agent who has the ability of evaluating its environment and acting accordingly, following the decision rules of the naïve reinforcement learning algorithm presented in the paper. Initially, a Locational Marginal Pricing (LMP) based ex-ante energy market is assumed. From the solution of the ex-ante energy market the agents calculate their initial FTR bid prices. Subsequently, the ISO solves the FTR auction problem and the agents profits equal the reimbursement they receive in the ex-post energy market for holding the FTR. The agent-based simulator repeats the FTR auction under the same conditions; in each repetition the agents update their bids according to the implemented algorithm and they “learn” their bidding strategies, through exploration by repetition. A five bus test system with five agents is used to illustrate the presented method. The results demonstrate the impact of contingency constraints and the effect of speculation in FTR auctions. Furthermore, the difference between bidding for FTR-obligations and FTR-options is shown. [Copyright &y& Elsevier]