Meta-heuristic optimization for a high-detail smart management of complex energy systems

Abstract

Distributed generation and, in particular, cogeneration and trigeneration are generally considered viable solutions to reduce energy consumption and mitigate the environmental impact of developed economies. Nonetheless, such systems need to be carefully designed and managed to effectively meet all the economic and environmental expectations. The design of a distributed generation plant and the choice of its proper management policy are complex tasks that require effective support methodologies and tools.In this paper, we develop a methodology to determine the optimal control strategy for a trigeneration plant. The model enforces mass end energy balances and accounts for the nonlinear and the basic dynamic behavior of each energy converter, for the time varying energy prices and environmental conditions, for maintenance and cold start costs, and for the possibility to store energy. We built on a methodology previously developed and we dramatically broaden its field of application to complex smart grids with a very high temporal detail, by cutting down its computational costs. To this aim, we implement an heuristic procedure that reduces the computational complexity of the non linear optimization problem. The total cash flow, the primary energy consumption, the plant efficiency, and the CO2 emissions, besides the instantaneous set-point of the plant, are among the most relevant results of the model.The model is first validated through 11 test-cases specifically designed to stress the possible weaknesses of the heuristic procedure. The validation evidences that the proposed procedure does not introduce further approximations to the mathematical model. The global optimum is retrieved for all the considered cases. Afterwards, we apply the proposed methodology to a realistic energy management scenario: the assessment of a fuel cell based trigeneration plant for a civil building for a whole year. The discussion highlights the effectiveness of the proposed method for different applications including the optimization of the control strategy for existing plants, the design of new distributed generation systems, the assessment of innovative energy conversion technologies, and the evaluation of national energy policies.