A mixed integer programming model was developed using General Algebraic Modeling System (GAMS) to forecasts grain facility replacement in Oklahoma. The results indicated regionalization in grain storage with fewer but larger capacity structures. The results of sequential replacement overtime indicated that there would be some abandonment of facilities and some shift to larger capacity structures. Producer’s transportation cost did not increase with sequential replacement as expected because storage were added in places to the current deficits. The results were not sensitive to crop production, fuel and construction cost and amortization factors. Cost comparison per bushel between configuration after sequential replacement and unrestricted replacement show that transportation cost was $0.04 lower in sequential replacement but total cost was $0.02 higher than unrestricted replacement. The findings of the study are important to grain firms and producers considering replacement of obsolete facilities.