LAUSR

Abstract This study synthesizes and optimizes a polygeneration system for a dairy industry located in Northeast Brazil. Based on the energy demands (steam, hot water, electricity, and chilled water), a superstructure was built to represent all the possible processes and connections of the energy supply system, using commercially available equipment and locally available energy sources. The optimization model is based on mixed-integer linear programming and its objective function is to minimize the total annual costs. The solution of the optimization problem determines the optimal combination between pieces of equipment, connections, and energy resources within the superstructure. A reference system is established, which does not include the possibilities of cogeneration nor biomass. The annual cost of the reference system is R$ 2,780,550/year. When the optimization model is freely solved, the optimal economic system yields an annual cost of R$ 2,074,793/year (25% lower). This reduction is reached by using diesel and natural gas cogeneration modules to substitute electricity from the grid at peak hours. Only 10% of the cogenerated heat is employed within the system (due to low thermal energy demands), and the remainder is evacuated by a cooling tower, Sensitivity analyses were carried out to assess the system behavior under variations of energy demands, fuel costs, and capital recovery factor, and the optimal economic solution presented good performance under fuel price variations, demonstrating robustness in most scenarios.