An analytical approach to designing a circular waste management system

The increasing generation of medical waste, driven by higher consumption levels, changing lifestyles, and natural disasters, threatens both the environment and human health. Medical waste from healthcare centers is particularly concerning due to its hazardous nature, necessitating effective management strategies. This study addresses key challenges in Medical Waste Management Systems (MWMS), including fluctuating waste generation, diverse waste types, incompatible handling practices, container and truck management, and the need for sustainable circular waste management. To tackle these issues, we developed a two-stage Stochastic Mixed-Integer Linear Programming (MILP) model to optimize MWMS network design. The model incorporates revenue generation from recycling, Waste-to-Energy (WTE) conversion, and container reuse while minimizing costs and environmental impacts. The model’s robustness is enhanced through data-driven parameter estimation, treatment technology selection, and revenue forecasting. To efficiently address the computational complexities associated with large-scale stochastic optimization, we employed a combination of the Sample Average Approximation (SAA) technique and a novel Hybrid algorithm that integrates deterministic optimization with metaheuristic methods, enhancing solution robustness and scalability. The model’s efficacy was validated through a case study in Hamilton, Ontario, Canada, where results demonstrated a 90.5 % reduction in computational time and a 56.7 % reduction in binary variables compared to the original model. The optimized solution achieved an annual waste disposal capacity of 300,000 tons, with an average revenue of $55 million, including $24.1 million from waste disposal, $17.1 million from recycled products and electricity, and $12.3 million from container reuse. Additionally, the network design reduced operational costs to $29.6 million and transportation costs to $6.4 million. This research contributes to the field by addressing gaps related to waste-to-container compatibility, revenue generation from reused materials, and uncertainty management. Future work may focus on enhancing predictive models for waste generation, integrating real-time data analytics, and expanding the framework to other regions with diverse waste management challenges.