Bauer Human-Centered AI Lab

Implications of Market Impact and Price Forecast Accuracy on Energy Arbitrage for Electricity Merchants with Storage and Renewable Power Plants

Abstract:

Problem definition: We study the joint optimization policy for electricity merchants with both energy storage and renewable power plants. We also identify the conditions when it is beneficial for the merchant to let independent system operators (ISOs) dispatch her energy storage directly. Methodology: We approximate market impact via a linear function of the electricity traded by the merchant, and employ dynamic programming, adaptive decisions making (ADM) based on rolling price forecast, and strong duality theory to study merchants’ energy operating policy while incorporating physical characteristics of storage systems, market impact, and uncertainty and stochastic evolution of price forecasts. Results: We show that, for a merchant with both PSH and wind farm, there exist three optimal SOC (state of charge) reference points such that the SOC range is divided into four subranges, each of which corresponds to one of four distinct decisions. We investigate the trade-off between increasing the unit energy profit and lowering the transaction quantity. We further find that the optimal scheduling decisions from merchant’s profit-maximizing align with ISOs’ social welfare-maximizing when the ISO sends the cleared prices to the merchant based on the social welfare-maximizing solutions. Managerial implications: Our study demonstrates that market impact substantially alters the optimal SOC reference points and energy operation decisions. We examine the conditions that merchants with energy storage and renewable power plant have the incentive to follow ISO’s schedule to achieve maximum profit.