A bioethanol production optimization model is developed and implemented to assess the feasibility of producing Taiwan's target volume of cellulosic ethanol. The most strategic placement of biorefineries based on feedstock supply is also explored to optimize the allocation of agricultural wastes. To simulate different supply targets and varying maximum biorefinery capacities, the nonlinear planning model utilizes the following information: ethanol yields, quantities of feedstock materials, potential refinery locations, and transportation networks. This simulation is undertaken to satisfy a minimum cost objective function and discover the most cost-effective factors of production. Sensitivity analysis explores the possible effects of technological advancements in cellulose conversion. This study estimates that a usage rate of 90% on two million metric tons of agricultural waste would produce a potential maximum 410 million liters of bioethanol, enough to meet 40% of Taiwan's demand for transportation fuel. (C) 2015 Elsevier Ltd. All rights reserved.