A decision-support model for managing the fuel inventory of a Panamanian generating company

Abstract

Bahia Las Minas Corp (BLM) is a fuelpowered generating company in the Panamanian power system. The purpose of this thesis is to design and evaluate a decision-support model for managing the fuel inventory of this company. First, we research BLM and its fuel replenishment methods. Then we define the problem, its objective function, assumptions, parameters and constraints. After identifying the most important given information (fuel price forecast, demand forecast, and current inventory levels), we define the equations that relate these inputs with the order sizes, and the availability and reserve constraints. Due to the large number of constraints, we devise a mechanism to calculate lower limits for the aggregate order sizes that prevent violations of the constraints beyond user-defined limits. We prepare a model in Excel for use with a single fuel type. This model takes stochastic forecasts of demand and fuel prices, and determines the best size for the weekly fuel order. After testing the model under several different scenarios, we conclude that it responds correctly to changes in price and demand. The complete discussion of these results can be found in the body of the thesis. Finally, we present some recommendations for BLM, both in relation to this replenishment problem and to its supply chain in general.