A simulation-based model for optimal demand response load shifting: a case study for the Texas power market

Jacob R. Schaperow, Steven A. Gabriel, Michael Siemann and Jaden Crawford

Abstract

This paper describes a prototype Monte Carlo simulation tool that we use to evaluate retail demand response (DR) programs for the Texas power market (the Electric Reliability Council of Texas/ERCOT), but which could also be applied to other regions. The model simulates a direct load control technique in which customers’ power consumption is adjusted during certain time periods called DR events. During a DR event, smart thermostats automatically increase participating customers’ temperature setpoints, reducing air-conditioning loads and thus potentially shielding retail electric providers (REPs) from financial losses during times of peak load. This study identified a Pareto optimal load control schedule based on forecasted load, settlement point prices and weather variables, taking into account stochastic load and prices as well as gray-box thermodynamic modeling. A Pareto optimal schedule is defined as the schedule that maximizes profit to the REP and minimizes its risk of low profits. We show how much benefit REPs stand to gain by following an optimal load schedule and how much they stand to lose by enacting DR events at the wrong times.