Optimisation of beer production at Tanzania Breweries Limited - an application of linear programming

This study is an attempt to apply Operations Research to a practical problem, and it deals with the optimisation of beer production at Tanzania Breweries Ltd. The company produces and markets five brands of beer whose demand vary from region to region, and the beer production which involves several process stages, is carried out at two plants owned by the company. The profit margins for each brand differ depending upon the production costs. A linear programming model has been developed for maximising the contribution thereby the profit, by determining the optimal brand - mix. The problem has been solved using Linear Programming Mark 2 an the ICL 16K - computer of the University of Dar es Salaam. This model has also been used to investigate the improvement in profit by relaxing immediate bottlenecks and to find out the capacity requirements to meet the demand in 1982. The results show that the optimum brand-mix under present conditions yield a 8.97% increase in production with associated increase in profit by 40.11%. A releasing of the immediate bottlenecks in this situation only two, an increase in profit of 30.80% (with an associated increase of 29.18% in production), compared to the optimal results under present condition, could be obtained. The stage wise increases in process capacities are also determined to meet the 1982 demand for beer. Finally, a generalised model, to include the distribution aspect, has been developed and discussed.