Full length articleVariability-based optimal design for robust plastic recycling systems

- We develop a framework for the robust design of a plastic sorting facility.
- The variability of sales prices of plastics is considered as external changes.
- We obtain the optimal solutions that can maximize the gross operating profit.
- We provide a recommendation on which resin types should be recovered.
- The developed framework is useful for avoiding selection of a risky solution.

The demand for recovered materials affects the profitability of the recycling processes and at the same time is susceptible to changes in external factors such as resource prices. Therefore, variability of such external parameters should be considered, when designing a recycling system. In this study, we develop a framework for the variability-based optimal design of plastic recycling for constructing a robust recycling system over external changes in the market such as fluctuations of material prices. The subject of this study is a sorting facility of post-consumer plastics equipped with automated sorting machines. The gross operating profit (GOP) is formulated based on a process model of the plastic sorting facility, internal and external parameters such as recovery ratios and sales prices of recovered plastics by resin type. By combining integer programming and Monte Carlo simulation, we obtain the potential optimal solutions that can maximize the GOP depending on the variability of plastics sales prices and their correlations between resin types, which are estimated based on the statistics for the interval between 2007 and 2011. Distributions of the optimal solutions and variability of the GOP are visualized according to the import price of naphtha for supporting decision-makers to determine resin types that should be recovered. Based on the results, we recommend the solution where polyethylene, polypropylene, and polystyrene are recovered. Moreover, the recommended solution is validated with the observed sales prices of plastics by resin type between 2012 and 2014.