RENEWABLE ENERGY MANAGEMENT FOR NON-DEFERRABLE LOADS WITH DYNAMIC ELECTRICITY PRICING

Abstract

Energy-consuming facilities that have insufficient flexibility to schedule their power consumption are referred to as non-deferrable loads. In this thesis, we consider dynamic electricity pricing and propose strategies to minimize the energy expenditure incurred by a non-deferrable load facility, which is equipped with a renewable energy harvester and a rechargeable battery.

We start by designing genie-aided strategies, which assume perfect knowledge of future renewable energy arrivals. Thereafter, we propose online energy management algorithms, which do not require exact knowledge of future renewable energy arrivals, but rely on statistical forecasting. We also introduce a novel strategy, which does not require estimates of future renewable energy arrivals, and has lower computational complexity than existing strategies.

Finally, we extend our analysis to account for non-linear electricity pricing schemes, and scenarios in which multiple users share renewable energy to minimize their collective expenditure.