Browsing by Author "Sanjika, Thawani M."

Now showing 1 - 1 of 1

Development of network theory approaches to analyse cause and effect relationships in complex integrated sugarcane supply and processing systems.
(2013) Sanjika, Thawani M.; Lyne, Peter William Liversedge.; Bezuidenhout, Carel Nicolaas.; Bodhanya, Shamim Ahmed.
Network theory has been widely and successfully used to model, analyse and visualise complex systems. This study aimed to develop approaches to analyse complex integrated sugarcane supply and processing systems. A literature review includes network theory, complex systems, the Theory of constraints, indicator analysis and root cause analysis. The cause-and-effect networks of four sugarcane milling areas in South Africa; viz. Eston, Felixton, Komati and Umfolozi were developed, where the factors that negatively affected the performance of the milling areas were represented by vertices, the relationships among the factors by arcs and the strength of these relationships by weights. Three network theory based analytical tools namely; (a) primary influence vertex analysis, (b) indicator vertex analysis and (c) root cause vertex analysis were developed to analyse the networks. The results from the analyses indicate variations in the numbers and strengths of primary influence factors, problem indicator factors and root causes of problems between the four milling areas. Rainfall, drought and high soil content in sugarcane were identified as the strongest primary influences in the respective milling areas. High crush rate variability, low cutter productivity, running behind allocation and increases in operating costs were identified as the strongest indicators of poor performance in the respective milling areas. Rainfall was found to be the most dominating root cause of poor performance in all the milling areas. Since the South African integrated sugarcane production and processing system is complex, it is likely that the unique approaches developed in this study can be used successfully to also analyse other relatively complex systems. It is recommended that these approaches be tested within other systems. The main contribution of this study is in the form of a relatively easy-to-use network theory based comprehensive systems analyses tool. This analytical approach has, to the author's knowledge, not been used in any agri-industrial application previously.