Simultaneous optimization of design and operation of an air-cooled geothermal ORC under consideration of multiple operating points

We simultaneously optimize both the design and operation of an air-cooled geothermal organic Rankine cycle, maximizing total annual return (TAR), while considering multiple operating scenarios based on different ambient temperatures. In order to accurately capture realistic off-design behavior of the heat exchangers and turbine, as well as for the overall system, we incorporate component models that consider performance variations with both size and operating conditions. We employ a hybrid mechanistic data-driven modeling approach, involving artificial neural networks (ANNs) as surrogate models for accurate fluid properties, as well as for intermediate expressions for which ANNs improve tractability of the optimization problem. We demonstrate the importance of considering multiple operating conditions within design problems and propose a methodology for formulating and solving such problems globally, using our open-source solver MAiNGO. (c) 2022 Elsevier Ltd. All rights reserved.