Mapping Fuel Poverty Risk at Household Scale using Infrared Thermography and a Fuzzy Set Approach

Abstract

This project investigates to what extent infrared thermography can be applied to improve the identification of households at risk of fuel poverty. As a method for qualitatively assessing the thermal efficiency of buildings, infrared thermography cannot be used to obtain estimates of indoor temperatures or allow inferences about the energy consumption behaviour of the occupants. However, front elevation thermal surveys can complement existing datasets like local house condition surveys and Reduced Data SAP assessments, where these datasets include assumptions on house condition based on housing archetypes. This allows for more accurate and individual identification of homes with poor thermal performance which might put the household at risk of fuel poverty. A front elevation thermography dataset was combined with data from a local house condition survey and Census variables to build a fuel poverty risk index for a small study area in Aberdeen. In the absence of consensus in the fuel poverty literature about the weighting of fuel poverty indices, and due to the lack of empirical evidence linking infrared thermography classifications to fuel poverty risks, a data-driven fuzzy set approach to multidimensional poverty measurement was used to weight and aggregate the index. This method also overcomes the limitations of defining fuel poverty vulnerability as an absolute and dichotomous condition, by permitting relative degrees of risk. The index performs well in reflecting the known interactions of fuel poverty risk factors, but could not be verified since data on the actual level of fuel poverty in the study area was not available. The study area contains mostly blocks of flats, which required a novel 3D GIS approach to map the fuel poverty risk index at the sub-building scale.