LAUSR

This paper presents an approach based on stochastic data envelopment analysis (SDEA) and clustering analysis to assess and optimise energy consumption in manufacturing sectors. Fossil fuel consumption, electricity consumption, total weighted production, and average weighted boiling point are considered as key performance indicators in this study. SDEA is tailored and used to alleviate data uncertainty and randomness for energy consumption problem. Clustering analysis is used to achieve homogeneity between decision-making units (DMUs). Noise and sensitivity analyses are performed to select the best α of SDEA model and also to identify the most important shaping factor. The results show that total weighted production is the most influential shaping factor in this study. Also, the distance between ideal and real value of each factor is estimated in order to help decision makers in improving performance. Finally, the proposed model is validated and verified through a robust analysis. The proposed approach would help decision makers to have a comprehensive understanding of energy consumption in manufacturing sectors. To the best of our knowledge, this is the first study to assess and optimise energy consumption of manufacturing sectors by clustered stochastic data envelopment analysis.