Simulation of the Sasol Nitro Watergel Plant using Rockwell Software’s Arena

Abstract

Sasol Nitro Ekandustria manufactures a variety of explosives and related accessories. Among these products are watergel products, which are water based explosives produced as cartridges and shipped once gelled. The product line has experienced a recent increase in demand and as such the company requires a tool that will help with decision making for possible projects that have been recommended and could implement as well as the possible identification of additional projects to expand or to debottleneck the facility to increase productivity. To this end simulation will be performed of the current state of the plant with possible improvements being simulated on this model. Via the simulation, analyses on the possible increases in productivity will be explored against determination of financial implications. Finally a recommendation will be made from the analyses and the data available. To simulate the plant the system first needs to be studied, including data on the throughput rates of equipment such as the mixers and packaging machines. Breakdowns and interruptions were determined and a detailed methodology for the data gathering is given. A Literature study was performed giving a brief overview and history on simulation. It also shows possible forms of practical application as wells as the advantages to be gained through simulation. Three weeks were spent gathering and analysing. The data gathered includes the mix times for each different type of batch, including the separate mixing steps’ times and the speed at which the KP (cartridge packaging machine) produces cartridges from this mix. Additional data gathered includes the breakdown times for each separate machine as well as additional mix times from the mix houses’ log books. The data was analysed and recommended mixing times were given and speeds for cartridges that was not measured was calculated by using the straight line method. With the data the bottlenecks in the system was determined and once the bottlenecks were determined the material handling speed of the plant was determined. It was found that the bottleneck depends on the mix type and the mixing time associated with this mix type as well as the product size of the product being produced. In general the KP is the bottleneck for all cartridges of 29x200mm and smaller, but depending on mix types all cartridges of 29mm diameter and smaller was found to be close to being the bottlenecks with a calculated time of 3 – 6 min/h idle time on the KP at most. For most cartridge sizes the bottleneck was found to be the mix time and the maximum speed that a plant can produce (for V6 mix) was calculated to be 74 boxes per hour. The analysis of the breakdowns for December found that the average breakdown time for all the KP’s are 5.22 min/h of the total operating time (less lunch times etc). The breakdown times were analysed and a statistic distribution was estimated using the Arena Input Analyser. An in depth analysis of the model is given with all the logic explained. The detailed steps for expanding the model is given and explained. Once the model was validated for a moth’s sales the different model setups were run. With the results available they are interpreted and analysed with a conclusion given. The recommendation to analyse the long term cost of replacing the existing KPs versus the cost of maintaining these KPs is given after simulation proved that mix time is the true bottleneck for most sizes and a faster KP will not improve the plant production rate. Finally a summary of the project and the value of simulation are given.