Optimization of waste management regions using recursive Thiessen polygons

Abstract Geographic Information Systems (GIS) are commonly employed to solve problems related to landfill siting and optimization of waste collection. This research aims to develop an easily implementable tool to optimize the topology of waste management regions in various Canadian jurisdictions using ArcGIS ModelBuilder. Landfill count, populated places, and road length are minimized using standard deviation to determine optimized tessellations. In Nova Scotia, reductions in standard deviation of 9.6–30.4% are observed between original and optimized tessellations. The results suggest that an optimized tessellation of Nova Scotia's Federal subdivisions may perform better than that of their waste management regions. In Saskatchewan, reductions in standard deviation of 4.9–46.1% were observed between original and optimized tessellations. Considering all Saskatchewan Federal Subdivisions, no optimization occurred. However, partitions of Saskatchewan Federal Subdivisions yielded better results, with vertical partitions yielding a 30% decrease in standard deviation of roads, while landfills and population were reduced in the horizontal subdivision by 20.0% and 38.0%, respectively. This suggests that a different approach may be required for waste management regions in Northern Saskatchewan. Saskatchewan transportation planning committees regions had the highest standard deviation across all parameters, and optimized at the fourth iteration (landfills and populated places), and first iterations (roads), despite the fact that this tessellation was developed in direct relation to roads in the province. The proposed tool, however, showed a limited application in the City of Regina given that land use planning within City limits. This work will improve the data driven aspect of regional waste management system design.