Agent-based model for electricity consumption and storage to evaluate economic viability of tariff arbitrage for residential sector demand response

• Storage-based demand response (loadshifting) is underutilized in residential sector.
• Economics (arbitrage savings versus equipment cost) are not well understood.
• Stochastic demand models and real-life tariffs can illuminate economic viability.
• A range of available storage options provide economically viable DR.
• Daily/seasonal stochastic demand variations crucial to understanding optimum capacity.

Demand response (DR) is one of many approaches to address temporal mismatches in demand and supply of grid electricity. More common in the commercial sector, DR usually refers to reducing consumption at certain hours or seasons, thus reducing peak demand from the grid. In the residential sector, where sophisticated appliance-level controls such as automatic dimming of lights or on-demand lowering of air conditioning are less common, building-based electricity storage to shift grid consumption from peak to off-peak times could provide DR without requiring consumers to operate their appliances on shifted or reduced schedules: Storage would be dispatched to appliances as needed while still shaving peaks on the grid. Technologically, storage and two-way-inverters are readily available to enable such residential DR. Economically, however, the situation is less clear. Specifically, are time-varying electricity tariffs available such that electricity cost reduction via arbitrage could offset manufacturing, financing, and installation costs of the required storage? To address this question we (i) devise an agent-based appliance-level stochastic model to simulate the electricity demand of an average U.S. household; (ii) loadshift the demand via simple dispatch strategies; and (iii) determine potential profits to the building owner, i.e. reduced electricity cost of the modified demand with realistic tariffs (Con Edison, NY) minus storage cost. We determine the economic viability for a range of traditional and advanced storage technologies as well as their optimum storage capacities to maximize profits. We find that (i) profits can range from <1% to 48% of annual electricity costs of a typical household; and (ii) optimum capacities, while approximately equal to households’ kWh consumption during peak hours, is affected by stochastic variations in daily and seasonal consumption. Future improvements to storage technology, arbitrage strategies, and tariffs are discussed. Details of the storage technologies, agent-based model, testing, and benchmarking are supplied as Supplementary Data.