The tightening of Emission Control Area (ECA) and Vessel Speed Reduction (VSR) regulations poses significant challenges for biomass exporters, particularly from emerging economies. These regulations raise transportation costs and reduce competitiveness, threatening the resilience of the biomass supply chain (BSC). This study develops an integrated optimization framework that combines machine learning (ML) (Bayesian XGBoost and Long ShortTerm Memory (LSTM) models) with discrete-event simulation (DES) to enhance supply chain (SC) efficiency under regulatory constraints. Using operational data from twenty-four Vietnamese wood pellet plants and international shipping records (2023–2024), the framework optimizes procurement, inland logistics, maritime transport, and demand planning. Simulation results show that full SC optimization reduces costs by nearly 40%, while combining SC optimization with power plant efficiency innovations achieves up to a 43.1% reduction in total costs and doubles exporter profits compared to the baseline. Beyond economic gains, the approach enables compliance with international environmental regulations without imposing prohibitive costs on exporters. This study contributes a novel methodological framework that bridges ML and simulation for sustainable SC optimization, offering practical guidance for exporters in emerging economies to remain competitive under tightening global environmental policies.
Global Trade and Customs Journal