A decision support for multi‐level capacitated disassembly lot‐sizing under uncertain yields using chance‐constraints and two‐stage stochastic optimization

Abstract Circular economy, with a focus on disassembly processes, has gained significant attention due to the increasing acknowledgment of environmental and economic considerations. In this paper, we explore the field of disassembly lot‐sizing, which involves decisions on disassembly quantity and timing in order to satisfy demands. Our exploration extends to multi‐level product structures with yield uncertainty and disassembly capacity restrictions. We formulate a two‐stage program with first‐stage joint chance constraints and apply an extended formulation based on scenario sorting to handle the joint chance constraints. Building on this, we propose strong valid inequalities based on the structure of the problem to further enhance computational performance. Test results confirm that the proposed model can be solved efficiently, even under yield uncertainty and limited disassembly capacity. Furthermore, the strengthened extended formulation incorporating the proposed valid inequalities substantially reduces computation times compared to the classical Big‐M baseline and the standard extended formulation, thereby enabling faster and more reliable decision‐making in practical disassembly scheduling. Furthermore, the sensitivity analysis conducted on key parameters such as backlog rates, demand levels, and available disassembly capacity gives valuable managerial insights, helping decision‐makers understand how system performance responds to changes in operational conditions.