Optimization

Abstract: To ensure effective biomass feedstock provision for large-scale ethanol production, a three-stage supply chain was proposed to include biomass supply sites, centralized storage and preprocessing (CSP) sites, and biorefi nery sites. A GIS-enabled biomass supply chain optimization model (BioScope) was developed to minimize annual biomass-ethanol production costs by selecting the optimal numbers, locations, and capacities of farms, CSPs, and biorefi neries as well as identifying the optimal biomass fl ow pattern from farms to biorefi neries. The model was implemented to study the Miscanthus-ethanol supply chain in Illinois. The results of the baseline case, assuming 2% of cropland is allocated for Miscanthus production, showed that unit Miscanthus-ethanol production costs were $220.6 Mg–1, or $0.74 L–1. Biorefi nery-related costs are the largest cost component, accounting for 48% of the total costs, followed by biomass procurement, transportation, and CSP related costs. The unit Miscanthus-ethanol production costs could be reduced to $198 Mg–1 using 20% of cropland, primarily due to savings in transportation costs. Sensitivity analyses showed that the optimal supply chain confi gurations, including the numbers and locations of supply sites, CSP facilities, and biorefi neries, changed signifi cantly for different cropland usage rates, biomass demands, transportation means, and pre-processing technologies. A supply chain composed of large biorefi neries with the support of distributed CSP facilities was recommended to reduce biofuels production costs. Rail outperformed truck transportation to ship pre-processed biomass. Ground biomass with tapping is the suggested biomass format for the case study in Illinois, while high-density biomass formats are suggested for long distance transportation.

An efficient and sustainable biomass feedstock production system is critical for the success of the biomass based energy sector. An integrated systems analysis framework to coordinate various feedstock production related activities is, therefore, highly desirable. This article presents research conducted towards the creation of such a framework. A breadth level mixed integer linear programming model, named BioFeed, is proposed that simulates different feedstock production operations such as harvesting, packing, storage, handling and transportation, with the objective of determining the optimal system level configuration on a regional basis. The decision variables include the design/planning as well as management level decisions. The model was applied to a case study of switchgrass production as an energy crop in southern Illinois. The results illustrated that the total cost varied between 45 and 49 $ Mg
1 depending on the collection area and the sustainable biorefinery capacity was about 1.4 Gg d
1. The transportation fleet consisted of 66 trucks and the average utilization of the fleet was 86%. On-farm covered storage of biomass was highly beneficial for the system. Lack of on-farm open storage and centralized storage reduced the system profit by 17% and 5%, respectively. Increase in the fraction of larger farms within the system reduced the cost and increased the biorefinery capacity, suggesting that co-operative farming is beneficial. The optimization of the harvesting schedule led to 30% increase in the total profit. Sensitivity analysis showed that the reduction in truck idling time as well as increase in baling throughput and output density significantly increased the profit.