A Decision Methodology for Coal Inventory Optimization

The procurement of coal as a power plant fuel is formally treated in a decision analysis methodology to assess optimal purchasing and stockpiling strategies. The uncertainties which characterize future prediction of fuel burn, receipt, and market price are expressed in a probabilistic multiperiod inventory model. A dynamic programming optimization procedure of the decision problem is developed. The effects of forecasting uncertainties and price variations are examined in several case studies and the value of perfect information is estimated.