Estimating cooling demand flexibility in a district energy system using temperature set point changes from selected buildings

Energy demand flexibility from buildings remains a largely untapped resource in electric power systems, in spite of its potential to be a low-cost substitute for investments in rarely used generation capacity. We develop a general methodology to estimate the aggregate cooling demand response for a large number of co-located buildings to thermostat temperature set point increases using empirically estimated building-level demand reductions in a subset of these buildings. For this subset of buildings, estimates were previously computed from real-world experimental data. The response of each remaining building is estimated as a different weighted sum of the empirical estimates, where the weights depend on observable characteristics of the buildings. We apply our method to a district energy system at a university campus that is roughly equivalent to a city of 30,000 people. Cooling is produced at a central energy facility with electric heat pumps and distributed to 124 commercial buildings through a chilled water loop. The response of six of these buildings to 1.1 °C (2 °F) daily temperature set point adjustments was previously estimated. Our methodology provides estimates for all 124 buildings and an estimate of the campus-wide demand response potential by leveraging a dataset including both structural (e.g. age, square footage) and operational (cooling loads and types of building operation) features for the full set of buildings. We estimate a 13.47% reduction in the campus energy system capacity needs under a 1.1 °C daily set point increase in all campus buildings during the 10 highest system demand days in 2020. On the highest demand day of 2020, we find that our predicted demand reduction could provide services equivalent to those provided by a lithium-ion battery with $4.6–$8.0 million installation cost at current prices and a storage capacity of 35.6–52.6 MWh.