Targeting and modelling urban energy retrofits using a city-scale energy mapping approach

City authorities, community groups and retrofit installers need to identify suitable local areas and dwellings for installing energy retrofit measures. This paper presents a localised Geographical Information System (GIS) based approach that utilizes publicly-available national and local datasets on housing and energy to provide targeted low carbon measures across UK cities. The study uses a rapid city-level energy assessment approach to spatially identify suitable neighbourhoods for particular retrofit measures, based on relative energy use and fuel poverty ratings. A GIS-based carbon mapping model (called DECoRuM) is then used to estimate energy use and potential for energy reduction on a house-by-house level. The improvement measures are aggregated to encourage bulk installations and drive down installation costs. To identify an appropriate neighbourhood case study area, publicly available datasets were assessed for the town of Bicester (Oxfordshire, UK), which included Ordnance Survey Mastermap, Energy Performance Certificate data (EPC) and sub-national energy statistics available at lower layer super output area (LSOA). When the EPC data for Bicester were compared with the sub-national statistics for Bicester, the average difference was found to be similar to 800 kWh. This is interesting as EPCs represent dwelling specific but modelled data whereas sub-national datasets represent actual but aggregated data. Superimposing the above datasets, a neighbourhood in southwest Bicester was selected as having the highest percentage of dwellings with energy consumption >300 kWh/m(2)/yr (EPC), most dwellings in need of wall insulation (EPC), second highest mean total energy consumption (sub-national), and third highest percentage of fuel poor dwellings (sub-national). House-level energy assessment in the selected area using DECoRuM showed that a package based approach comprising fabric and heating system upgrade and solar PVs emerged as the most effective. (C) 2017 Elsevier Ltd. All rights reserved.