Cui, Cheng; (2025) A Probabilistic Robust Design Approach to Heat Resilience of Future Homes. Doctoral thesis (Ph.D), UCL (University College London).

Text cheng_cui_phd_thesis.pdf - Accepted Version Access restricted to UCL open access staff until 1 August 2026. Download (41MB)

Abstract

The ubiquity and the heterogeneity of uncertainty in the built environment, including social, technological and climatological aspects, have been increasingly recognised over the past few decades, bringing challenges and, more importantly, opportunities to the domain of building performance analysis. Amidst tackling climate change in the building sector, uncertainty in climate projections is of particular significance on account of the interrelationship between climatic conditions and building performance. One certainty concerning future climate is the warming trend of the global surface temperature, which, given the currently low uptake of active cooling, poses a grave risk of summertime overheating in dwellings across historically heating-dominant regions. Meanwhile, active cooling is no automatically sustainable solution to indoor heat exposure, considering both technical barriers and health implications. Addressing this adaptation urgency warrants careful consideration of the changing yet uncertain climate, so that its synergy with mitigation actions can be harnessed. On the quest for heat resilience of future homes, the philosophy of robust design, which promotes the insensitivity of a system to uncertain factors, emerged as an enabling methodology that manifests a proactive mindset towards uncertainty. Its applications have been documented in the building design area for many years, yet limited recognition is present in the existing literature, with tailoring for building applications underexplored. Likewise, probabilistic climate projections became accessible to building researchers and practitioners over a decade ago, yet they continue to receive underappreciation in today's research and practice on climate-resilient design. Thereby, this thesis aims to establish and investigate the applicability of a probabilistic robust building design framework, leveraging probabilistic future weather datasets for designing heat resilience into future homes. The framework is underpinned by the Bayesian bootstrap, a non-parametric statistical technique that accommodates small-sample weather uncertainty in predicting future building performance. The implementation of the framework was contextualised via a series of design analyses of select English dwellings, with the objective of synergising summertime thermal comfort and annual energy use. The examination of the Bayesian bootstrap demonstrated its vast potential to facilitate robustness measurement for future thermal comfort, albeit revealing a significant overheating variability of up to 11 times the regulatory threshold in future summer nights due to climate uncertainty. In the robust design optimisation exercises, the thermal comfort robustness of the case study dwellings managed to improve by over 35 % amongst the Pareto solutions compared to the regulation-compliant state, with the mean summertime overheating rate decreased by over 50 %. Findings suggested the inevitability of deploying active cooling in the future English housing stock to prevent overheating, but a future-proof revision of the currently energy- and comfort-related building regulations is still urgently needed to lower the cooling demand towards building heat resilience and energy sufficiency.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item