Document Type : Original Article
Authors
Department of Industrial Engineering, K. N. Toosi University of Technology (KNTU).
Abstract
Air pollution, fossil-fuel dependency, and water scarcity have increased the need for cleaner and regionally adaptable energy alternatives in Iran. This study develops a single-period location–allocation–production–distribution model for designing a linseed-based biodiesel supply chain. The main contribution of the study lies in the decision-oriented integration of data-driven crop-yield estimation with strategic biodiesel supply chain design, whereby spatial yield heterogeneity is translated into biomass availability, cultivation, facility-location, capacity, and distribution decisions.Unlike conventional models that rely on fixed or average biomass-yield assumptions, the proposed framework estimates regional linseed yield using agricultural and climatic data and integrates the predicted yield values into the optimization model to calculate biomass availability more realistically. The model determines cultivation allocation, biorefinery and distribution center locations, biodiesel production, transportation flows, and unmet demand while minimizing the total system cost. A case study in Iran is conducted to evaluate the proposed framework. Using year-grouped five-fold cross-validation, the Random Forest model achieved an R^2 of 0.682, RMSE of 0.0995 ton/ha, and MAE of 0.0630 ton/ha. The optimization results show that cultivation and production costs are the dominant cost components. Sensitivity analysis confirms that demand growth increases total system cost and shortage pressure. The Random Forest model also significantly outperformed a leakage-free historical-average benchmark, producing a lower annual MAE in 43 of the 44 evaluated years.These findings demonstrate that integrating crop-yield prediction into biodiesel supply chain design can improve biomass supply assessment, reduce planning inefficiencies, and support more reliable strategic decisions for linseed-based biodiesel development in semi-arid regions.
Keywords
Main Subjects
)