Comparing system-marginal-price versus pay-as-bid auctions in a realistic electricity market scenario
Résumé
A realistic agent-based computational model of the day-ahead market session of the Italian wholesale electricity market is simulated to compare market performances between system-marginal-price and pay-as-bid clearing mechanisms. An empirical validation of computational results at a macro-level is performed to test for accuracy of simulated outcomes with historical ones. The level of prices are accurately reproduced except for few peak hours. As far as concerns pay-as-bid auction, the computational experiments point out that it results in higher market prices than the system-marginal-price auction. In the pay-as-bid mechanism, sellers' endeavours to maximize their profits are more costly thus leading to higher price levels