SYNTHSEP: A general methodology for the synthesis of energy system configurations beyond superstructures

Highlights: • A new general methodology is proposed to build complex energy systems configurations. • The building blocks of the methodology are elementary thermodynamic cycles. • The methodology does not build a superstructure in advance. • The paper shows how an artificial intelligence is able to replace designer's experience. The proper choice of the energy system configuration and design parameters, generally named "synthesis/design problem", is only rarely straightforward because of the many variables involved. The goal of a standard for the generation of new system configurations has recently led to superstructures that potentially include all possible configurations, among which the optimum one, yet the ability of defining in advance such superstructures is a fundamental limit of this technique. To overcome this problem a bottom-up methodology is proposed, which relies on the basic idea that the system configuration is certainly based on one or more thermodynamic cycles that may share some processes or be combined in a cascade form. Accordingly, all the possible ways of combining elementary cycle processes into meaningful system configurations are first identified using a comprehensive and rigorous set of rules. An optimization is then performed in which the search space consists of all the obtainable configurations and associated design parameters. The paper shows all steps of this original synthesis/design optimization methodology and its effectiveness in the search for the best two-pressure level ORC system configuration. The optimum results obtained using different working fluids and temperatures of the heat source allow general design guidelines to be identified.