Optimizing Demand Response of Aggregated Residential Energy Storages

Abstract

The increase in uncontrollable volatile renewable generation leads to a growing need for flexibility also on the consumption side of electricity. Residential consumers, especially with energy storages such as water heaters and batteries, could shift and curtail their energy consumption. However, in order to provide suitable changes to the consumption profiles of the consumers, appropriate methods for control and incentives have to be developed. This thesis presents several developed optimization methods for an aggregating retailer to aggregate multiple mostly independently acting consumers for diverse market participation. The first set of contributions included methods for planning aggregate consumptions schedules for the day-aheadmarket, as well as for taking into account the potential intra-day flexibility of the optimized schedules. In addition, analysis were made on how much the aggregator can shift the consumption of its consumers depending on whether the consumption is responsive to changes in prices or if the aggregator can directly alter the consumption. Secondly, the participation of residential consumption in frequency control was considered. Centralized control was simulated utilizing a proposed agent-basedmodel, under uncertain communication latencies. Furthermore, methods were proposed for optimizing day-ahead schedules for participation in the frequency reserve market. The optimization frameworks were extended to include various uncertainties in the resulting electricity demand and market prices, as well as coordination of charging and reserve participation plans between the consumers under the different conditions. In order to test the various developed methods, multiple stochastic programming models are devised and simulated for large populations of residential consumers. The results indicate that flexibility of residential consumption could potentially be utilized for participation in the wholesale market, intra-day trading and frequency reserves. Furthermore, the performed simulations illustrate the benefit of considering flexibility during the day-ahead planning of electricity consumption.