A Stochastic Linear Programming Model for Corn Residue Production

Abstract

This paper presents the results of a stochastic linear program for estimating the supply of corn residue for use as raw material in an ethanol plant. The model is based on the production capacity of an average Illinois farm, and considers the feasibility of three mutually exclusive residue harvesting alternatives-own baling, custom baling, and cob collection. Since the potential for residue use in animal feed may be even more promising, these results are directly useful for the feed industry. They also indicate the profitability of investing in residue harvesting equipment. From a methodological point of view, the paper contrasts the results of three OR approaches, a deterministic LP approach, a stochastic LP approach, and a chance-constrained approach. Because of the stochastic nature of the problem both Monte Carlo simulation and chance-constrained programming are found to be computationally viable, even though they differ in the way they incorporate risk information.