UKERC Energy Data Centre: UKERC Publications

UKERC Home >> EDC Home >> UKERC Publications >> Back to Results >> Two Models for Benchmarking UK Domestic Energy Demand

Two Models for Benchmarking UK Domestic Energy Demand

Citation Summerfield, A.J., Lowe, R.J. and Oreszczyn, T. Two Models for Benchmarking UK Domestic Energy Demand. 2010.
Cite this using DataCite
Author(s) Summerfield, A.J., Lowe, R.J. and Oreszczyn, T.
Opus Title Building Research and Information
Pages 12-24
Volume 38

From publicly available data, two simple models are developed to help identify the trajectory of total delivered energy to UK households and provide benchmarks for the UK domestic sector. Both models can help to inform policy-makers and the public whether delivered energy in the domestic sector is on track and whether the reductions correspond with the expected impact of a more efficient domestic sector. The annual delivered energy, price, and temperature (ADEPT) model uses multiple linear regression to fit consumption data since 1970 (R2=0.76). Findings indicate that with typical recent heating season temperatures of 7C and at 2005 energy prices, average household delivered energy is estimated at 21.7 MWh (95% confidence interval=20.8, 22.6). For every 1C increase in heating season temperature, average household delivered energy drops by approximately 1 MWh/year. Energy price elasticity is estimated at 0.2, so that a 50% rise in energy prices corresponds to an approximate 10% decline in energy demand. To fit quarterly delivered energy data since 1998, a seasonal temperature energy price (STEP) model is also developed using polynomial multiple regression (R2>0.99). At 2005 energy prices, the average household has a minimum power demand of 1.2 kW (including approximately 0.4 kW electricity) at about 16C, which rises as external temperatures drop to 5C, where it begins to plateau at approximately 3.8 kW. The average heat loss coefficient is estimated at 240320 W/C. Both models find no evidence to date to indicate that changes in delivered energy are beyond those expected due to variations in external temperature and energy price. The seasonal model also provides benchmarks for delivered energy as a response to external temperature and energy price that will be useful for a comparison of building performance evaluations with the national stock, without requiring data for a whole year.