A hybrid multi-criteria decision analysis framework for optimal biodiesel feedstock and facility location selection: A case study of Botswana

Résumé fourni par la source

The development of sustainable biodiesel supply chains in landlocked developing countries like Botswana is constrained by feedstock supply uncertainty, high logistics costs, and the absence of systematic decision-support tools. This study develops and applies a hybrid multi-criteria decision analysis (MCDA) framework that integrates the weighted sum model (WSM) for criteria weighting with TOPSIS and VIKOR methods for ranking. Using primary data from 22 experts and secondary sources, five potential feedstocks and six candidate locations in Botswana were evaluated. A transparent hybrid consensus score, calculated as the average of the normalised TOPSIS closeness coefficient and the benefit-transformed VIKOR index, was used to produce final rankings. Waste cooking oil emerged as the preferred feedstock, followed by beef tallow and morula nuts, while Francistown ranked first among locations, followed by Palapye and Gaborone. Robustness was confirmed through one-at-a-time weight perturbation (±10–30%), successive elimination of lower-ranked alternatives, and a multi-level Monte Carlo simulation comprising of 1500 iterations at each perturbation level. The top three feedstocks and locations retained an overall high stability (78–100%) across all tests. This study provides a practical and replicable hybrid MCDA methodology for data-scarce environments and offers evidence-based priorities to guide multi-feedstock biodiesel development in Botswana and similar developing economies.