Influence of district features on energy consumption in non-domestic buildings

The spatial variability of gas consumption is investigated in non-domestic buildings across districts of Greater London, UK. It is argued that the energy consumption of a building is to some extent influenced by where the building is located in a city, due to contextual features of its own district as well as those of neighbouring districts. Using Bayesian spatial models, the analysis suggests the energy consumption due to the influence of district features can be quantified and dissociated from the energy consumption associated with the physical features and operational characteristics of the buildings. An important distinction is made between extrinsic values of energy consumption (district features) and intrinsic values (building characteristics, management and operation). The results indicate that 90% of the mean value of extrinsic gas consumption across districts in London is between -42 and 87 kWh/m(2). Buildings located in districts that have positive values of extrinsic gas consumption consume more gas over and above their expected intrinsic value of gas consumption. The novel features of this study are in the quantification and propagation of district-scale features to their influence on buildings, and the reduction in uncertainties around the mean value of gas consumed by different building types.