Repository logo

Masters Degrees (Agricultural Engineering)

Permanent URI for this collectionhttps://hdl.handle.net/10413/6652

Browse

Browsing Masters Degrees (Agricultural Engineering) by Author "Bezuidenhout, Carel Nicolaas."

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Issues pertaining to cane supply reliability and stockpiling at the Umfolozi sugar mill - model development and application.
    (2011) Boote, Gordon L. N.; Bezuidenhout, Carel Nicolaas.; Lyne, Peter William Liversedge.
    The co-owned Umfolozi Mill area has developed as an integrated supply chain. Cane supply reliability was identified as a potential area for productivity improvement at Umfolozi. It is important that the cane supply to a sugar mill arrives at a steady and reliable rate. A reliable cane supply ensures that the mill can operate at an optimum efficiency. Sugarcane supply reliability depends on how the mill area adapts to unforeseeable changes in the supply chain. An important aspect to this is the weather and how it affects the harvesting regimes. The sugarcane supply chain at Umfolozi is divided into two branches, road transport and tram transport. The trams account for 70 % of the cane delivered to the mill and the can is sourced from a climatically homogenous region. In the occurrence of a rainfall event of above 5 mm, infield harvesting cannot take place on the Umfolozi Flats; hence 70 % of the mill‟s supply is halted for one or more days. To address the problem, a stochastic model was created to simulate the effectiveness of an enlarged cane stockpile if it were maintained on the current tram sidings outside the mill and were crushed when wet weather prevented further harvesting. The stockpile was simulated on a first-in first-out principle and was able to supply the mill with enough cane to continue running for 24 hours. The model was then used to conduct a series of Monte Carlo simulations on which sensitivity analyses and economic feasibility assessments were carried out. Results show that the stockpile was effective in reducing the length of milling season and the number of no-cane stops. However, on further analysis into the implications of creating a stockpile it was found that 1% recoverable value (RV) was lost during the 24-hours that the cane is stored outside the mill. The loss in revenue as a result of the RV reduction had a negative impact on any savings created with the implementation of the stockpile. This result made apparent the negative impact of deterioration to the whole supply chain. Further research is required to determine more accurately the rate of deterioration, and therefore, quantify more accurately the losses that occur in the supply chain. A significant outcome of the study was the development of a mechanistic tool which drove decision making at Umfolozi Sugar Mill. It lead to the development of the modelling framework LOMZI, a simulations based framework which places more emphasis on environmental factors and risks.
  • Thumbnail Image
    Item
    Modelling sugarcane quality in the context of mill scale supply chain logistics.
    (2014) Jenkins, Edwin Peter Garland.; Bezuidenhout, Carel Nicolaas.; Ortmann, Gerald Friedel.
    The length of milling season (LOMS) refers to the length and timing of sugarcane crushing operations at a sugar mill. LOMS is central to the competitiveness and profitability of any sugar mill supply area (MSA). Conflicting interests between supply chain stakeholders can make adjusting the LOMS difficult. The LOMS should take into account weather conditions, cane quality, milling capacity, supply chain capabilities and other interrelated issues, such as agronomics. Previous LOMS models in South Africa were developed over a decade ago and there was scope to improve the calculation of risks by using a stochastic modelling approach. Recently, a stochastic model named LOMZI was developed to evaluate stockpiling options at Umfolozi. In this study, LOMZI was adapted and expanded to allow the LOMS for any MSA in South Africa to be investigated. However, mill area specific applications of the updated model fell outside the scope of this study. As it currently stands, LOMZI simulates a sugarcane supply chain from the point where sugarcane is cut, up to delivery at the mill. During the process of adapting LOMZI, the simulation of sugarcane quality was identified as an important area for improvement in the model and this became the focus of the study. A predictive MSA scale cane quality model was developed, based on recent weather conditions and a mechanistic understanding of sugarcane quality. The quality model was developed to simulate the daily average brix %, pol % and fibre % of sugarcane delivered to the mill. The preceding 11 weeks’ rainfall and temperature values were used to predict cane quality. A total of 98 mill-specific coefficients were calibrated from historic milling data and, for model demonstration purposes, the quality model was applied at two mills, namely Sezela and Umfolozi. Independent verifications yielded R2 values between 0.56 and 0.74. A useful method to estimate the average burn/cut to crush delay for a MSA was also identified. The quality model has been successfully integrated with LOMZI. Future work is envisaged to expand LOMZI and to model the operations of sugar mills and the RV cane payment system.