Application of the extended two-stage network DEA model for the biomass-biofuel logistics network design

Abstract

Network data envelopment analysis (N-DEA) models have been applied to measure efficiency scores for decision-making units (DMUs), where DMUs represent network processes. The biomass-biofuel logistics network (BBLN) design problems with the risk of biofuel facility shutdown have been approached by applying the regular two-stage network DEA (R-TSN DEA) model. This paper proposes and demonstrates how to apply the extended TSN (E-TSN) DEA model for designing efficient BBLN systems more accurately and consistently than the R-TSN DEA approach. We apply a weighted goal programming (WGP) model for the BBLN design problem by considering five performance metrics simultaneously. Various BBLN configurations are generated by solving the WGP model with multiple weight values assigned to five performance metrics. Decision makers are usually interested in the top efficient DMUs before deciding to select the most appropriate option. The proposed E-TSN DEA approach more consistently identifies top-notch BBLN schemes than the R-TSN DEA. A case study utilizing available data in South Carolina, USA, shows that the proposed E-TSN DEA suggests more accurate, consistent, and robust.       

Article history: Received (June 14, 2024); Revised (October 19, 2024); Accepted (October 30, 2024); Published online (February 14, 2025)