Paper Conference

Proceedings of eSim 2020: 11th Conference of IBPSA-Canada


Multi-objective Battery Energy Storage Control Modeling for Residences Within Utilities

Byrne Campbell, Lukas Swan
Dalhousie University, Canada

Abstract: In this study, a residential battery system was modelled in MATLAB. The size and characteristics of the battery were held constant and the control algorithm altered for a range of objectives. The objective functions were designed to benefit residential homeowners, via energy arbitrage (EA Control), and/or the utility system, by peak sharing demand (PS Control). The models were tested using ten, one-year duration, measured 5-minute time-step residential load profiles and the measured local electrical utility load profile for the same year. ‘EA Control’ was designed purely for the benefit of the homeowners. A ‘stacked’ control (EA+PS Control) was designed to provide benefits to both homeowners and the utility. Finally, a rule-based real-time control (RT Control) strategy was designed to further optimize benefit sharing between the homeowners and the utility without relying on explicit load forecasting. EA control raised the utility’s bulk system demand charge by $4,365 while only saving the homeowners $5,408. EA+PS saved the utility $2,910 and the homeowners $4,459 for an overall savings of $7,369 per year. RT control reduced the utility peak load by the maximum possible 50 kW (aggregate system discharge power 50 kW) achieving utility savings of $7,275; combined with the homeowner savings of $5,077 the total annual savings becomes $12,352 for the system as a whole. EA, EA+PS and RT control resulted in 277, 250 and 265 complete-discharge-equivalent cycles for the battery over the year, respectively. The results of this analysis better the understanding of energy storage system control for multiple stakeholder operation.