Advancing simultaneous component and cycle optimization: A holistic cycle-optimized collaborative design concept for Brayton cycle heat pumps

To achieve globally optimal performance of industrial heat pumps, matching both thermodynamic cycle and component parameters is a complex challenge. The integration of multiple disciplines and design stages is required. Holistic design strategies, which simultaneously optimize thermodynamic cycle parameters together with detailed component optimizations are able to overcome the shortcomings of conventional sequential design. However, the complexity significantly increases with high dimensionality integrating all disciplinary analyses. Four optimization architectures are analyzed, integrating multi-stage compressor optimization and heat exchanger sizing with cycle optimization. Distributed structures that solve the component subproblems separately, yield higher reliability, better objectives, and faster convergence, though at increased computational cost. Surrogate models trained with active learning strategies are able to reduce function evaluations enabling integration of complex models. A collaborative optimization concept with distributed cycle optimization is identified as the most effective, achieving convergence in fewer than 500 evaluations and designing a Brayton cycle heat pump with a COP of 1.82 at a 260 K temperature lift. The trade-off with engine size is demonstrated in a multi-objective approach highlighting the potential of holistic optimizations for high-temperature heat pumps.