An Application of the Hub Location Problem in the South African Coal Transport Sector

Abstract

The South African inland logistic systems are heavily reliant on existing road networks. The bulk transportation of goods, such as coal, increases the loads the roads currently have to carry and lead to damaged roads that in turn increase logistic costs. For these and other socio-economic reasons the strategic decision has been made by one of the major coal transportation companies in South Africa to migrate the transportation of coal as far as possible from road to rail. To support this strategy, coal consolidation centres are to be optimally located, whereby road-based coal will be diverted to the hubs and transported by rail to the demand point. The purpose of this project is to develop and evaluate mathematical models that could be used by decision makers to determine optimal locations of coal consolidation centres. To ful ll this aim two mathematical models are developed: a network-based model in which a limited number of prede ned nodes along the existing rail network can be chosen as hub locations, and the continuous model in which all points on the plane are considered. The two main objectives are to decrease total operational cost and to migrate the transportation as far as possible o road and onto rail. A test case is used to highlight strengths, weaknesses and underlying behaviours of the models. The most noteworthy endings involve the major weaknesses of both models. The greatest weakness experienced by the network-based model is the limitation of the possible hub locations; improved locations not on the rail network are excluded and could yield improved results. The continuous model eliminates this weakness, however, it does not consider the distance along the rail network in calculating optimal locations. This chosen locations the continuous model yields are numerically inferior to the network-based counterparts. For these reasons the decision is taken to use both models in order to achieve optimal results; the continuous model is used to determine additional hub locations, while the network-based model is used to determine nal results. In the procedure described above, mathematical models are developed with the capability to guide decision makers in determining optimal coal consolidation centres. The results with regard to the case study are analysed to give recommendations to guide the decision-making process.