The assumed U-values of solid walls represent a significant source of uncertainty when estimating the energy performance of dwellings. The typical U-value for UK solid walls used for stock-level energy demand estimates and energy certification is 2.1 Wm⁻² K⁻¹. A re-analysis (based on 40 brick solid walls and 18 stone walls) using a lumped thermal mass and inverse parameter estimation technique gives a mean value of 1.3 ± 0.4 Wm⁻² K⁻¹ for both solid wall types. Among the many implications for policy, this suggests that standard UK solid-wall U-values may be inappropriate for energy certification or for evaluating the investment economics of solid-wall insulation. For stock-level energy modelling, changing the assumed U-value for solid walls reduces the estimated mean annual space heating demand by 16%, and causes a proportion of the stock to change Energy Performance Certification (EPC) band. The analysis shows that the diversity of energy use in domestic buildings may be as much influenced by heterogeneity in the physical characteristics of individual building components as it is by variation in occupant behaviour. Policy assessment and guidance material needs to acknowledge and account for this variation in physical building characteristics through regular grounding in empirical field data.

In the context of performance-based building regulations, the benefits provided by building energy performance prediction tools in facilitating the exploration of innovative approaches to achieving performance requirements has been recognized. In accordance with the Directive on Energy Performance of Buildings (EPBD), the UK has introduced a methodology that allows the use of accredited third-party software for demonstrating compliance with energy performance criteria outlined in the relevant technical guidance for the non-domestic sector. This study investigates the issue of predictive inconsistencies between accredited tools available at the time of writing and presents the final results of an inter-model comparative analysis that examines the extent of variability in key compliance outputs. An overview of tool applicability limitations and issues associated with results variability are presented. Implications concerning the credibility of the methodology and recommendations to address current shortcomings are noted.