No economic sector incorporates more heterogeneity in the typology of energy consumers than the industrial sector, yet analysis of energy and fuel demand are normally conducted for the industrial sector as whole rather that at a more disaggregate level, a choice normally due to data availability. Motivated by the goal of developing the new industrial module adopted by the UK government Department of Business, Energy and Industrial Strategy (BEIS) for their econometric Energy Demand Model, our paper reports results obtained as part of a three step estimation process, where in the first step we estimated the determinants of economic activity for 8 industrial subsectors, in the second step we estimated the determinants of energy demand (published elsewhere) and in the third step, discussed here, we assess substitution across fuels in each of the industrial sub-sectors modelled in our research.
Using data spanning from 1990 to 2014 on fuel consumption and prices published on the Digest of UK Energy Statistics, and indices of production collected by the Office for National Statistics we implement detailed fuel demand analysis for the following sector: 1) Chemicals and Pharmaceuticals; 2) Engineering and Vehicles; 3) Food, Beverage and Tobacco; 4) Non-metallic Minerals, 5) Non-ferrous Metals, 6) Pulp, Paper, Printing and Publishing; 7) Textile, Leather, Clothing and Footwear and 8) other manufacturing. Our study constitutes the first cointegration analysis that provides evidence on fuel demand elasticities with respect to economic activity and fuel prices at a disaggregated industrial level. We model simultaneous demand for each of the different fuels as a separate cointegrating vector, a methodological innovation suggested by Pesaran et al (1999) but largely unnoticed in the energy economics literature. By implementing a Vector Autoregressive system with multiple cointegrating vectors through the Johansen approach, we find that there is considerable heterogeneity in consumers belonging to different industrial subsectors with regard to the pattern of substitution and complementarity across fuels; the existing pattern of reduction in energy consumption and consumption of specific fuels, the long-run impact of increases in fuel prices and energy price as a whole on consumption of different fuels, and the speed with which consumers in different sectors adjust back to equilibrium in response to exogenous shocks. Finally, we are also able to assess whether responses to shocks, say an increase in the gas price, implies a reduction in total energy consumption or just a reallocation of a given level of energy consumption across different fuels. Supporting the work of BEIS for their econometric Energy Demand Model, our innovative and disaggregate approach has had considerable policy impact therefore being a prime example of the way in which differences across industrial consumers are taken into account in current policy making. Agnolucci P., De Lipsis V., Arvanitopoulos T., 2017. Modelling UK sub-sector industrial energy demand, Energy Economicsm, 67, pp. 366-374  H. Pesaran, R. Smith, T. Akiyama (1999) Energy Demand in Asian Developing Economies, Oxford University Press, Oxford (1999) Angnolucci-modelling-fuel-demand-of-heterogenerous-industrial-consumers.pptx 954.57 KBAgnolucci-Modelling-fuel-demand-of-heterogenerous-industrial-consumers.pdf 585.43 KB