Linear Programming As A Decision Tool

Mark S. Huber, Utah State University


This thesis considered the potential benefits of employing linear programming in cheese manufacturing plants as a decision tool for management. Its potential has been enhanced by the recent approval of acid orange 12 as a chemical for testing the percent protein in milk; therefore, a practical test is now available for monitoring protein as well as milk fat in milk manufacturing and fluid milk plants. Seven models, each one differing only in the milk fat and protein percentages or means of standardizing the cheese milk, were manipulated individually and simultaneously to test the managerial benefits of linear programming under various plant and market conditions. Each model consisted of five cheese activities or variables, two butter activities, three powder activities, and a selling activity for each product produced. The maximum price that could be paid the farm producer per hundred-weight of milk and the minimum wholesale rice per pound of manufactured product, to cover variable costs were determined for each variety of cheese and composition of milk. There was a definite interaction between each of the activities. This caused the cost to produce a Pound of cheese to vary according to the alternative uses for milk, cream, skim milk, and whey. When the simulated plant was being utilized at or near full capacity and the cheese milk was standardized with non fat dry milk powder, total cheese yield increased as did total profits. When the plant was not being utilized to full capacity, profits were higher by not standardizing.