Construction Quality and Reducing Energy Demand: Lessons for UK Construction

Introduction: Buildings account for one-third of energy consumption, making reductions in energy demand essential for achieving net-zero targets by 2050. However, the UK continues to experience a gap between predicted and actual building energy performance, known as the energy performance gap. Measured energy use in non-domestic buildings is 34% higher than design predictions, while domestic buildings can consume up to five times more energy than anticipated. This paper argues that addressing this challenge requires construction quality control alongside enhanced predictive capabilities through machine learning and data-driven modelling.

Method: A systematic literature review of academic studies, industry reports, and UK construction case studies was undertaken. The selected literature was analysed thematically to identify causes of the energy performance gap, examine quality assurance practices, and evaluate the potential of machine learning and data-driven approaches for improving energy prediction accuracy.

Results: Findings from UK and international case studies indicate that current project approaches often prioritise visible aesthetic defects over hidden performance issues, including discontinuous insulation, damaged air barriers, and unresolved thermal bridges. Reliance on third-party approval processes may also reduce accountability for thermal integrity. Emerging machine learning and artificial intelligence techniques offer opportunities to enhance energy modelling by integrating datasets, including IoT sensor information, as-built geometric data from LiDAR, and historical performance. Compared with conventional compliance-based simulations, these approaches can generate realistic energy demand predictions and identify performance sensitivities and defect risks before they occur during operation.

Conclusion: Reducing the energy performance gap in the UK construction sector requires standardised in situ fabric assessment, including thermographic surveys and repeated air permeability testing. It also requires a transition from compliance-focused design towards performance-based delivery, supported by workforce development and digital twins that maintain data continuity throughout the lifecycle. Combining quality governance with predictive tools can strengthen better alignment between design intent and operational performance.