Energy efficiency modelling and estimation in petroleum refining industry- A comparison using physical data

Abstract

In this paper, energy efficiency in petroleum refining industry has been analyzed using the non-parametric data envelopment analysis approach with physical data. The proposed model for evaluation of energy efficiency does not require energy data in the operational levels for considering the structural effects. The average weighted boiling point of refined products is used as structural indicator in the proposed model. Also, data envelopment analysis (DEA) is used as a sensitivity analysis tool for determination of the potentials and means of energy consumption improvement and optimization. The proposed model is applied to a set of data from Iran and some OECD countries. The results show that the potential of savings in fossil fuels are greater than that of electricity consumption in refineries. The unique feature of this study is based on the utilization of DEA for separation of structural effects from other effects which are referred to as activity and intensity effects. The proposed model is also verified and validated by principle component analysis (PCA) and non-parametric Spearman correlation technique.