A TWO-LEVEL OPTIMIZATION APPROACH FOR THE SYNTHESIS, DESIGN AND OPERATION OF MULTI-ENERGY SYSTEMS INTEGRATED WITH ENERGY NETWORKS

Achieving decarbonization targets requires optimizing the synthesis, design, and operation of multi-energy systems that meet end-user demand for different forms of energy, with the goal of minimizing costs. This means contextually finding the number, type, location, size and operation conditions of the conversion and storage units that supply the required energy, as well as the topology and capacity of the networks that deliver it to the end users. However, this inherently integrated problem includes several nonlinearities and an extremely large number of continuous and binary decision variables, making it impractical to solve. Linearization allows the computational complexity to be drastically reduced. Nevertheless, the combinatorial nature of binary variables exponentially increases computational time. To simplify the problem while ensuring good accuracy of the results, this study presents an evolutionary algorithm that decomposes the overall problem into two levels: i) selection and placement of components to be included in the system (synthesis) and ii) sizing and scheduling of the included components (design and operation). This allows for a drastic decrease in combinatorial complexity and, in turn, in computational time. The proposed two-level algorithm is validated here against the traditional mixed integer linear programming (MILP) approach for solving the complete, non-decomposed problem. Finally, application examples considering electrical and heating networks coupled with both renewable and conventional energy conversion and storage systems show that decomposition reduces computational time by up to 70% and provides solutions to problems that otherwise cannot be solved in an acceptable time.