A case study on the impact of fixed input parameter values in the modelling of indoor overheating

Abstract

Global efforts to reduce greenhouse gas emissions from buildings while also improving their environmental resilience have intensified. These efforts are often supported by building stock models which can inform policymakers on the impact of policies on energy consumption, greenhouse gas emissions and the indoor environment. The input values of such models are commonly informed by reference tables, which can result in inaccurate specification and incomplete representation of the distribution of possible values. In this modelling case study of a semi-detached dwelling archetype, the influence of using a reference U-value (2.1 W/(m2K)) for solid walls in England on heat-related mortality rate is compared to a probabilistic specification based on empirical evidence (median = 1.7 W/(m2K)). Using the theoretical reference U-value generally resulted in a lower indoor overheating risk compared to the use of the empirically derived U-values pre-retrofit, but a larger increase in heat-related mortality rate following internal wall insulation (1.20 %) than the use of the empirical median (0.94 %, 95 % Confidence Interval = 0.87-0.99 %). This highlights the potentially significant implications of using fixed reference values. Future work will employ this probabilistic framework on multiple influential parameters.