Abstract/Summary

Robust estimation of the environmental impacts of commercial buildings during early building design can support the ability to choose a design with demonstrably lower carbon emissions than other alternatives. However, gaps currently exist in the methods, tools and supporting data required to estimate embodied carbon emissions, especially for building services systems. This thesis aims to help address those gaps by developing an embodied carbon estimation method able to inform choices of those systems during building design. Estimation of embodied carbon of a proposed building services design requires knowledge of raw material types and quantities of all its components as well as the relevant carbon emission factors for each material applicable to the life cycle stage of the component being considered. The relative scarcity of available data on raw material types and quantities used in building services systems makes the estimate subject to considerable uncertainty. A method is therefore needed to generate the estimate using limited source data whilst quantifying this uncertainty. Previous research on embodied carbon estimation has mainly used deterministic methods, which can generate misleading conclusions when comparing products. This study investigates the applicability of first-order analytical uncertainty propagation (AUP) combined with a parametric approach to estimate embodied carbon of heating, ventilation and air conditioning (HVAC) components and systems, using a series of empirical case studies. The AUP approach is tested against Monte Carlo simulation to determine whether it can offer a computationally efficient alternative for supporting rapid decisions during early building design. The resulting method is outlined in a specification for a decision support tool. The study also examines – and provides depth of understanding on - choice mechanisms used by designers to select systems of building services and explores the ways in which information on embodied carbon emissions might influence the choice. The readiness of designers to reduce embodied carbon is also informed by societal barriers and drivers associated with an environmental intervention not widely used by industry, in that it prioritises environmental impacts of building services systems that happen prior to their operational use in completed buildings. Decision-making is investigated via a qualitative study of mechanical services designers and wider social factors are investigated using a survey of construction practitioners. Theories of rational choice and bounded rationality are explored as a framework to explain decisions by practitioners and to inform possible decision support measures. The results of both studies are used to inform the design of the embodied carbon estimation method and the specification for the decision support tool. The parametric estimation method is shown to support comparisons of embodied carbon of HVAC components and systems for which input parameters are uncertain and can be continuously improved by an iterative process as additional empirical data becomes available. Insights into decision making by practitioners highlight the scope of influence of mechanical services designers on decisions that may reduce embodied carbon as well as identifying interventions able to improve decision support in this area. Recommendations for further research include expansion of the estimation method to cover additional building service systems and estimation of operational carbon emissions.