No abstract

Coal is the second-largest source for electricity generation in the United States. However, the burning of coal produces dangerous gas emissions, such as carbon dioxide and Green House Gas (GHG) emissions. One alternative to decrease these emissions is biomass co-firing. To establish biomass as a viable option, the optimization of the biomass supply chain (BSC) is essential. Although most of the research conducted has focused on optimization models, the purpose of this paper is to incorporate machine-learning (ML) algorithms into a stochastic Mixed-Integer Linear Programming (MILP) model to select potential storage depot locations and improve the solution in two ways: by decreasing the total cost of the BSC and the computational burden. We consider the level of moisture and level of ash in the biomass from each parcel location, the average expected biomass yield, and the distance from each parcel to the closest power plant. The training labels (whether a potential depot location is beneficial or not) are obtained through the stochastic MILP model. Multiple ML algorithms are applied to a case study in the northeast area of the United States: Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), and Multi-Layer Perceptron (MLP) Neural Network. After applying the hybrid methodology combining ML and optimization, it is found that the MLP outperforms the other algorithms in terms of selecting potential depots that decrease the total cost of the BSC and the computational burden of the stochastic MILP model. The LR and the DT also perform well in terms of decreasing total cost.

This work proposes a hybrid scheme that combines a simulation model and a mathematical programming model for designing logistic networks for co-firing biomass, specifically switchgrass, in conventional coal-fired power plants. The advantages of co-firing biomass include: (1) the creation of green jobs; (2) the efficient use of current power plant infrastructure; (3) fostering the penetration of renewable energy into power networks; and, (4) the reduction of greenhouse gas (GHG) emissions. The novelty of this work lies in the inclusion of (1) the inherent variability of biomass supply at the parcel level, and (2) the effects of climate change on future biomass supply when designing a feedstock logistic network. The design optimization is conducted at the farm/parcel level (most, if not all, previous works have used county level average data) and integrates the crop growth predictions employing United States Department of Agriculture’s (USDA’s) Agricultural Land Management with Numerical Assessment Criteria (ALMANAC) simulation model; the output of the simulations is input into the mixed integer linear programming (MILP) hub-and-spoke model to minimize the overall cost of the logistic network. Specifically, the MILP-based model selects the parcels and depot locations as well as biomass transportation flows by taking into consideration different types of soil, land cover characteristics, and predicted yields, which account for both historical and forecasted weather data. The hybrid methodology was tested by solving realistic situations, which considered varying weather conditions. The gross results indicate that the optimized logistic network enabled meeting a 20% biomass co-firing rate demand, which reduced 1,158,867 Mg per year in GHG emissions by co-firing with biomass.

No abstract