###Data-driven techniques for low-carbon, healthy homes
Project ID: 2228bd1165 (You will need this ID for your application)
Research Theme: Engineering
UCL Lead department: Bartlett School of Environment, Energy and Resources
Lead Supervisor: Phil Symonds
Project Summary:
The UK must urgently retrofit its building stock to reach net-zero targets by 2050. The move to low energy homes is more urgent than ever as the population faces energy and cost-of-living crises. Each year in England and Wales, there are on average nearly 800 excess deaths associated with heat and over 60,500 associated with cold. UK housing must become more efficient and better adapted to provide thermally comfortable conditions in both summer and winter.
The primary aim of the project will be to improve our understanding of the relationship between UK housing characteristics (particularly their low-carbon credentials), indoor temperatures, and health outcomes (related to heat and cold exposure). Data driven and causal inference techniques will be used to derive relationships between outdoor and indoor temperatures for UK homes by location, dwelling characteristics, and occupancy type. A Bayesian calibration framework will be used to further validate an established UK housing stock model (UK-HSM) using a large nationally representative dataset of indoor temperatures. The health impacts of heat and cold exposure under various retrofit, adaptation and climate change scenarios will be assessed.
This doctoral research will be supervised by Dr Phil Symonds who has extensive expertise in building stock modelling and the analysis of large datasets through cutting edge statistical and machine learning techniques. Subsidiary supervisors will include Dr Giorgos Petrou and Dr Zaid Chalabi who have expertise in Bayesian calibration and mathematical methods, respectively.
The ideal candidate will come from a mathematics, physics or engineering background and have knowledge or a keen interest in: 1) building physics and thermodynamic principles; 2) building performance modelling; 3) statistical, mathematical and machine learning methods (e.g. using python or R) applicable to building-related datasets.