Bi-level stochastic programming for optimal modular construction yard deployment based on Benders decomposition

Abstract

To promote wider adoption of modular construction (MC), many governments in high-density regions are planning to establish module storage yards (MSYs) to support local contractors in achieving just-in-time module supply chain. Given the limited availability of developable land and government budgets, an optimal MSY deployment plan is urgently needed. This paper represents the first attempt at capturing the fundamental government–contractor interactions and formulating a bi-level stochastic program to maximize MSY utilization and minimize MC logistics costs. To address the computational challenges posed by a hierarchical model structure, a solution method based on Benders decomposition is designed to solve the problem to optimality. Benchmarked against particle swarm optimization through extensive numerical experiments, the solution method shows a 15% average improvement in solution quality (in medium- and large-scale instances), highlighting its superior computational performance. A real-world Hong Kong case is conducted as methodology validation and application that provides governments with optimal decisions on MSY deployment including the area of the MSY to be established and module storage service pricing.