Richard A. R. Kilpatrick
Heriot-Watt University, Edinburgh, Scotland
Phillip F. G. Banfill
Heriot-Watt University, Edinburgh, Scotland
Download articlehttp://dx.doi.org/10.3384/ecp110571008Published in: World Renewable Energy Congress - Sweden; 8-13 May; 2011; Linköping; Sweden
Linköping Electronic Conference Proceedings 57:34, p. 1008-1015
Published: 2011-11-03
ISBN: 978-91-7393-070-3
ISSN: 1650-3686 (print), 1650-3740 (online)
The energy consumption associated with non-domestic buildings represents 11% of the UK’s total energy consumption; 11% of Europe and 18% of the USA’s. Annual non-domestic building energy consumption is often presented in the form of average benchmarks; such as 450kWh/m²/year for a large airconditioned building and 200kWh/m²/year for a small naturally ventilated office. Benchmark values give very little insight into how and where a building consumes energy. While some benchmarks provide a breakdown of energy use by energy category (lights; IT; cooling; heating); these data still fails to demonstrate how the energy associated with each category varies throughout the year. To further understand building energy use; a more detailed data breakdown and analysis is required. The electricity demand data for a variety of school buildings (secondary; primary; specialised) in Scotland has been made available for analysis. This consists of half hourly resolution data spanning several years for 50 schools; allowing key trends and patterns in energy use to be identified. These trends can include differences between annual profiles; differences between winter and summer months; and differences in weekday and weekend energy use. Additionally; the effect of other variables such as climate; user behaviour and general building data on the buildings energy consumption can be investigated. A database of half-hourly school energy demand data; with corresponding building details has been set up and a preliminary analysis preformed. Alternative method of pattern recognition in non-domestic energy usage are discussed; and the variables necessary to calibrate this information. This demonstrates the possibility of creating generic energy profiles and hence new benchmarks.