A facility location model with fixed capacities for recycling facilities’ locations - a case study in Seremban, Malaysia

Malaysia encounters considerable challenges in waste management and recycling, driven by population growth, urbanization, and evolving lifestyles. Despite the government's goal of a 40% recycling rate by 2025, over 80% of recyclable waste continues to be disposed of in landfills. This is largely due to low public participation in source separation activities, especially among households, which is a critical step in promoting recycling practices. A key obstacle to enhance recycling rates is the insufficient accessibility and availability of recycling infrastructure, including collection points and appropriately sized containers. To address this, optimal location of recycling facilities is critical to encourage public participation and efficiently managing recyclable waste. Hence, this research focuses on a variant of the Facility Location Problem (FLP) model, utilizing the set covering problem to guarantee that areas generating recyclable waste have access to at least one operational facility within an acceptable travel time. Moreover, the uncapacitated model of Maximal Expected Coverage Location Problem is unrealistic for real-world applications, as it fails to account for varying container sizes and demand constraints. Therefore, the proposed model of Maximal Covering Location Problem incorporates capacity considerations to ensure more practical and effective coverage in real-life scenarios. The proposed methodology was validated using both randomly generated datasets and small-scale real-world data to demonstrate its applicability and practical relevance. The model was then applied to Seremban, which comprises approximately 315 household areas with a population of 422,710 in 2024. These figures indicate a growing volume of recyclable waste, highlighting the need for a strategic approach to locating recycling facilities with optimal capacity level to support a better waste management system of the city. Using the proposed methodology, nine recycling facility locations were identified to serve selected areas in Seremban. Of these, six are the existing facility locations, while the remaining three are newly proposed locations. Scenario testing was also carried out by imposing limitations on the authorities’ capacity to serve users. The analysis indicates that operating a single facility, which serves only 10% of total demand, results in significantly lower coverage efficiency compared to operating eight facilities, which serve 90% of demand. The findings emphasize that the commitment of both users and authorities plays a vital role in the success of recycling efforts. This reflects the strength of proposed methodology in accounting for stakeholder perspectives and operational capacities in the process of identifying optimal locations for recycling facilities. To conclude, the proposed mathematical model for the FLP shows significant potential to improve the current Malaysia’s recycling system. By optimizing facility locations, it enhances service coverage, reduces infrastructure costs, and supports user engagement while accounting for authority constraints. For future research, it is recommended to apply the proposed model to other urban settings across Malaysia to assess its scalability and adaptability. Further research could also explore dynamic demand patterns, integration with smart waste technologies, and multi-objective optimization to enhance decision-making in sustainable waste management.