Search

The expansion of commercial building demand response as a demand-side management resource for the electric grid necessitates new decision support resources for customers seeking to assess the benefit–risk tradeoffs of possible strategies for energy flexible building operations. To address this need, we develop surrogate models that predict the impacts of several load flexibility strategies on commercial building electricity demand and indoor temperature, focusing on offices and retail buildings at multiple scales. The surrogate models are fit to a synthetic database generated via whole building simulations, which establish the relationships between the key operational features of a given strategy and potential changes in building demand and temperature across a variety of contexts. The surrogate models are translated to a Bayesian framework to allow straightforward communication of uncertainty and parameter updating given new evidence. We find strong predictive performance across the suite of models, underscoring the usefulness of the approach in guiding decisions about implementing load flexibility strategies under a particular set of operational and environmental conditions. • A range of commercial load flexibility strategies are developed as OpenStudio Measures. • Synthetic data of Measure impacts are compiled across building types, vintages, and climates. • Models fit to synthetic data predict changes in demand and temperature under load flexibility. • A Bayesian framework facilitates model updating and communication of prediction uncertainty. • Acceptable model fit and predictive performance is demonstrated across multiple metrics. [ABSTRACT FROM AUTHOR]