A rapid assessment method for the thermal mass energy storage capacity of building clusters

Abstract

Buildings, particularly building clusters, can utilize their thermal mass as a virtual energy storage (VES) resource by modulating HVAC operation strategies to achieve flexible power load regulation. Accurate quantification of building VES capacity is essential for effective integration into generalized energy storage systems for coordinated planning and supporting grid demand response programs. However, existing methods—such as detailed simulations or field experiments—entail high computational costs and scaling difficulties, while simplified cluster-level models often suffer from insufficient parameter characterization, leading to poor accuracy. To address these issues, this paper proposes a framework for rapid VES capacity assessment of building clusters. The framework comprises: (1) simulation-based excitation-response experiments to determine equivalent thermal capacity by analyzing cooling load response during temperature setback; (2) efficient generation of a large-scale VES capacity dataset through stochastic parameter sampling and batch simulation; (3) development of a multiple linear regression model that quantifies the relationship between equivalent thermal capacity and key parameters using feature engineering; and (4) practical application requiring only few key parameters to rapidly estimate building VES capacity using the trained model. Case study validation shows model errors within 20% (below 10% for mid-floor rooms), demonstrating good accuracy. Applied to residential buildings under different energy efficiency standards in Tianjin, the framework successfully quantified VES capacity ranges across building phases, confirming its feasibility and practical potential for rapid, large-scale VES capacity assessment at the cluster level.