Development of a spatial sugarcane transport infrastructure-planning model.

Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2008.

Due to the significant cost of transport in the sugar industry, a model, named FastTrack, was developed to investigate infrastructure planning opportunities. The model mathematically incorporates road construction and maintenance costs, terrain and land-use maps, vehicle performance specifications and annual sugarcane volumes to determine the most cost effective route, per vehicle type, from a production region to a mill. Route planning using geographical information systems (GIS) is a standard approach for determining the optimum alignment for pipelines, roads and canals. Theory of this approach was reviewed to create a foundation for the development of FastTrack. A small portion of the Noodsberg sugar mill region in the KwaZulu-Natal midlands was selected as a case study area to test the capabilities of FastTrack. A start location was identified as a natural flow point for 70 000 tons of sugarcane hauled from an area south of the mill. Currently this volume is transported along a 9.3 km stretch of national road from the start location to the sugar mill, while the Euclidean distance is approximately 7 km. Three vehicle types, differing in payload, fuel consumption and road speed were assessed. Two common and currently utilised vehicles, the tractor hilo and interlink combinations, were aligned by FastTrack along existing national roads. A financial penalty for driving on national roads was assumed for the third vehicle type considered, land trains, as these are currently not permitted to operate on national roads in South Africa. This high bulk vehicle was selected to test the capabilities of FastTrack and to identify if cost savings could be realised through increased consignment capacity as has been achieved in Australia, Malawi and Brazil. Utilising the model a new and more direct theoretical route was generated for the land train with a length of 7.4 km. Existing farm roads which would require upgrading made up 34 % of this proposed route. An economic analysis was conducted and showed that under current conditions, the private route generated by FastTrack for land train use, would be the most cost effective, with a system cost of R 57.50 t" . The tractor hilo and interlink had system costs of R 59.58 t" and R 60.98 t"1 respectively. Repeating the economic analysis with projected fuel prices indentified that the cost saving advantage of the land train system over the other two vehicle configurations increases with increasing fuel costs. A rigorous validation process, including a sensitivity analysis of results from FastTrack, revealed that the model performs predictably under a wide range of input conditions and could be a valuable tool for decision making in the sugar industry. However, further research is required to combine more economic and logistical aspects into FastTrack and to increase its usability.