| This dissertation describes the development of a model for comparing alternative bioenergy production locations based on raw material transportation costs. Although the focus of the developmental research is wood-energy production in Michigan's Northern Lower Peninsula, only minor changes would be required to adapt the model to crop-biomass production of synfuel. This research is unique because it integrates a regional forest biomass inventory based on Landsat satellite imagery and a computerized model which forecasts network biomass transportation costs based, in part, on this inventory. It is believed to represent the first integrated model of this type.; The model developed here consists of four major components: (1) a specification of the biomass supply location and quantity, (2) estimation of the cost of transporting biomass to each of the consumers, including both existing consumers and the hypothetical bioenergy plant, (3) application of a linear programming algorithm to select, from among all the possible combinations of biomass supply points and consumers, those combinations which produce the lowest network transportation costs for the entire network, and (4) a time simulation structure for components 1.2 and 3.; Four sets of experiments were conducted using the model: (1) a series to develop preliminary statistical parameters for the network cost estimates, (2) a benchmark series of model runs every five years from 1982 through 2012 which included only the existing biomass consumer network, but no wood-energy plant, (3) a series covering the same time period simulating a small (25 MW) wood-energy plant at Rose City, Michigan, and (4) a series simulating the same size plant at Idlewild, Michigan. These experiments indicate information produced using this model will be valuable in bioenergy industrial location decisions. |
