Optimization Of Biomass Logistics System Using Genetic Algorithm And Particle Swarm Optimization For Biofuel Production

As time goes by, Renewable Energy keeps proving to be an important and potential replacement for fossil fuels. All the different types of Renewable Energy offer a relief to the environmental aftermaths of the prolonged reliance on fossil fuel energy. Bioenergy is one of the types of Renewable Energy that can help by minimizing the emissions of fossil fuels. The Energy Independence and Security Act mandates the use of 21 billion gallons of advanced biofuels including 16 billion gallons of cellulosic biofuels by the year 2022. Biomass and Biofuels can clearly become a significant aid to sustainably supply energy in the future. Nevertheless, the sustainable supply of energy has proven to be quite challenging. The logistic challenge of supplying biomass to biorefineries while being efficient and keeping costs low is one that demands to be tackled. The motivation for this thesis is to provide an approach to the biofuel supply chain challenges along with a cost-effective solution. In order to meet the mandate set by the government, corn ethanol production has significantly increased in the last few years. Conversely, the production of advanced biofuels, such as the ones obtained from biomass, are not meeting the target amount of biofuel production set in the mandate. This can be attributed to the significant economic and logistical challenges for regional planners and biofuel entrepreneurs in terms of feedstock supply assurance, supply chain development, biorefinery establishment, and setting up transport, storage and distribution infrastructure. With the high logistics operation cost, is it crucial that an optimal logistics system is designed in order to allow for the smooth transition from fossil fuels to biofuels. The thesis presents two different approaches to the optimization of a biomass-to-biofuel logistics system through the use of evolutionary algorithms that mimic nature, Genetic Algorithm and Particle Swarm Optimization. The performance of these two metaheuristic methodologies is compared in this work. In the past, these types of problems have been solved using mathematical methods such as Linear Programming and Mixed Integer Programming. Metaheuristic methods can provide near-optimal results significantly faster than mathematical optimization methods for complex problems such as this one. For that reason, this thesis presents two metaheuristic approaches for the optimization of a biomass-to-biofuel logistics system design considering multiple types of feedstock and demonstrates that metaheuristic optimization methods are suitable to solve combinatorial problems such as the one tackled in this research work