Three-phase artificial intelligence-geographic information systems-based biomass network design approach: A case study in Denizli

Renewable energy sources are of great importance in protecting the environment by reducing greenhouse gas emissions and global warming. Recently, more efficient conversion technologies and different types of resources have been introduced to generate energy in a sustainable way. Biomass-to-energy systems need to be designed efficiently due to high volumes of raw material flows and transportation costs. This study proposes a novel integrated approach for the solution of biomass supply chain network design using artificial intelligence, geographic information systems, multi criteria decision making and mathematical modelling. First, the five-year forecasts of biomass raw materials including animal waste, agricultural residues, and municipal solid waste were made using support vector regression. Alternative biogas facility locations were determined spatially based on various criteria using geographic data and a decision-making technique. Then, considering annual net present value streams, the bioenergy system has been configured using a mixed integer linear programming model. The proposed methodology was applied on a real case in the city of Denizli, Turkey. The results showed that nine conversion facilities could be opened with 2000 kWh capacity each, and approximately 83.2% of net income could came from electricity sales, with the remainder from fertilizer sales. Furthermore, a sensitivity analysis was carried out to see the variations in model parameters such as biomass purchase cost, transportation cost, fertilizer sales prices and discount rate. It was found that the most dominant factor affecting net present value was fertilizer sales price, which was followed by discount rate. © 2023 Elsevier Ltd