Biomass inventory optimization for willow based pellet production integrating seasonal supply and stochastic demand

Abstract

Renewable biomass energy is a critical component in the global transition to sustainable energy systems. Its contribution to reducing greenhouse gas emissions and fossil fuel dependence has become increasingly vital. However, biomass supply chains face unique challenges due to material characteristics such as bulk, low energy density, and seasonality, making effective inventory management critical to their economic viability. This study develops a Mixed Integer Linear Programming (MILP) model to optimize inventory management policies for a willow-based wood pellet production facility. Built upon an (s, S) inventory control system, the model integrates internal operational decisions like inventory adjustment frequencies, while incorporating external uncertainties such as seasonal biomass harvesting capacity constraints and stochastic pellet demand patterns. Results demonstrate that the stochastic approach could maintain zero stockout probability when the lost-sale-cost is assumed to be higher across our modeled stochastic demand scenarios while achieving lower overall inventory management costs compared to solutions from deterministic methods. Critically, the stochastic framework provides substantially lower cost variability (over 55% reduction in coefficient of variation) and greater operational robustness under demand uncertainty, with the performance advantage potentially increasing as uncertainty levels rise. Weekly inventory policy adjustments provide a balance between efficiency and complexity, with total inventory management costs of approximately 2.30 million USD with 72,000 tons annual production capacity. The model successfully accounts for seasonal supply constraints and stochastic market demand and facilitates a multi-feedstock strategy that offers additional supply chain resilience and associated cost reduction. The framework's computational efficiency and broad applicability make it suitable for adoption by diverse biomass industries with variability in their supply chains.