Document Type Doctoral Thesis Author Ssamula, Bridget firstname.lastname@example.org URN etd-07242008-093606 Document Title Strategies to design a cost-effective hub network for sparse air travel demand in Africa Degree PhD (Transportation Engineering) Department Civil and Biosystems Engineering Supervisor
Advisor Name Title Prof C Venter Supervisor Keywords
- South Africa
- air route network
- aviation industry
- network design
Date 2008-04-11 Availability unrestricted Abstract
The aviation industry worldwide is changing dynamically in reaction to trends such as globalisation and with the need to increase market share to remain competitive. The African aviation industry still faces many problems in the institutional, technical and operational areas. Despite its potential for enhancing economic development, air travel to and from Africa remains a small percentage of world air travel. The African air route network is characterised by sparse demand, with long sector distances, low frequencies and high fares. This study investigates cost-effective hub-and-spoke (H&S) network design strategies for the African route network. An H&S network would minimise the cost of air transport and improve accessibility and connectivity. The study challenges the typical characteristics of H&S networks which are usually found in denser route networks. The design methodology used was the one most appropriate for the African region, using the datasets and tools available. As a first-cut analysis for Africa, the results of the research contribute to understanding the effectiveness of H&S networks in markets with sparse demand.
A cost model previously developed by the author to calculate operating costs on a route was used. It eliminated the need to assume discount coefficients on links, as passenger demand increases, in a field with limited data. The cost indicators derived from the model were used as criteria for choosing the most efficient hubs within a cluster. These were compared with the hub location criteria in the literature which use distances and passengers. It was found that using the cost indicators gives a reasonably consistent method that lowers passenger travel time.
The optimum number of clusters and hubs was found to be four. The geo-political network design method yielded the lowest network costs. The hubs are centrally located within the clusters: Morocco in the north, South Africa in the south, Kenya in the east and Nigeria in the west. They are characterised by high passenger demand and short node-hub sectors. There are significant benefits to be gained from using this hub network design, resulting from the economies of scale with higher passenger densities on routes. Furthermore, the benefits of higher service frequencies and better connectivity outweigh the extra travel time when routing through hubs.
The study found that for sparse networks, the cheapest hub-location options have high passenger demand. The sector distance is crucial in lowering operating costs as smaller, more efficient short range aircraft can be operated. It is therefore more efficient to assign nodes to the closest hub to lower node-hub costs. The optimum number of hubs/clusters is thus determined by the distance threshold for the efficient aircraft. The effect of changing the cluster boundaries on network costs also depends on the change in node-hub distances between the clusters. As sparsity reduces, the economies-of-scale benefits outweigh the increasing operating costs of longer distances, allowing efficient operation of larger-capacity aircraft. This means that the location of the hubs and the number of clusters becomes more flexible, implying that node-hub links can become longer, reducing both the clusters and the number of hubs.
© University of Pretoria 2008D452 /gm
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access 00front.pdf 61.80 Kb 00:00:17 00:00:08 00:00:07 00:00:03 < 00:00:01 01chapters1-2.pdf 158.19 Kb 00:00:43 00:00:22 00:00:19 00:00:09 < 00:00:01 02chapter3.pdf 168.20 Kb 00:00:46 00:00:24 00:00:21 00:00:10 < 00:00:01 03chapter4.pdf 205.76 Kb 00:00:57 00:00:29 00:00:25 00:00:12 00:00:01 04chapter5.pdf 215.84 Kb 00:00:59 00:00:30 00:00:26 00:00:13 00:00:01 05chapters6-8.pdf 199.29 Kb 00:00:55 00:00:28 00:00:24 00:00:12 00:00:01 06back.pdf 1.01 Mb 00:04:40 00:02:24 00:02:06 00:01:03 00:00:05