Effective and scalable modelling of existing non-domestic buildings with radiator system under uncertainty

Abstract

Effective and scalable methods for modelling existing non-domestic buildings and their HVAC systems under uncertainty continue to be instrumental in risk-conscious building performance assessment, recommissioning, and retrofit practice. This study makes such an attempt for large buildings with radiator system with a modelling method that builds upon detailed state space models of radiator-heated spaces, an archetype-based spatial reduction approach to modelling an entire building, a steady-state model of heat distribution subsystem, and explicit quantification of uncertainties in the above models. The capability and efficacy of the method were demonstrated by a case study on a building section on campus. The results show that the proposed method can effectively capture the detailed dynamic building heat transfer phenomena in individual spaces and is scalable to large complex buildings with moderate model complexity and computation cost.