Hybrid Model for Micro-channel Heat Exchangers used in sCO2 Brayton Power Blocks

Abstract

Studies have shown that Micro-channel Heat Exchangers (MCHEs) commonly referred to as Printed Circuit Heat Exchangers (PCHEs) are suitable candidates for recuperators and gas coolers used in sCO2 power blocks. CFD and 1-d unit-cell-based models have been proposed in the literature for the design and analysis of MCHEs. To estimate the heat exchanger size and arrive at an optimum channel configuration, CFD models are found to be computationally expensive and time-consuming, especially when full-scale MCHEs are to be modelled. On the other hand, 1-d models are inadequate for correctly predicting the performance. The thesis aims to address this gap by proposing a hybrid model. The hybrid model incorporates a Thermal Resistance Network (TRN) framework combined with a unit-cell CFD model to investigate the thermo-hydraulic performance of the complete MCHE stack. CFD-based unit-cell models are developed for straight and non-straight flow paths to obtain the variation of local heat transfer coefficient and Fanning factor along the channel length. A stack optimization scheme based on the rate of heat loss from the external surfaces of the MCHE core is proposed and incorporated in the hybrid model to arrive at the optimum stack width, height, and number of rows. Additionally, the hybrid model also includes a model for the inlet and exit manifolds utilizing the concept of flow resistance balance to obtain optimum pressure drop across the entry and exit manifolds. The proposed manifold pressure balance scheme facilitates uniform flow distribution among the channels in the MCHE stack. The efficacy of the hybrid model is presented for MCHE based recuperator and a gas cooler used in a 1-MW scale sCO2 Brayton power block. For the recuperator, optimum stack volume and corresponding pressure drop are presented for a 5˚C pinch temperature differential. The improvement over the straight channel is demonstrated by using sinusoidal and zigzag flow path configurations. Channel pressure drop values obtained from the optimum stack and channel dimensions are used as inputs in the manifold model to obtain optimum overall pressure drop. The optimum overall pressure drop ensures uniform flow distribution among the channels in the stack. Subsequently, the hybrid approach is extended to model the gas cooler performance and obtain stack volume with water as the secondary heat transfer fluid. Unlike the recuperator where the mass flow rates across the hot and cold sides are identical, the water flow rate and the corresponding Reynolds number in the case of a gas cooler are dictated by the sCO2 side. For the gas cooler, a range of Reynolds numbers for both water and sCO2 are obtained, ensuring temperature pinch at the cold inlet. The effect of Reynolds numbers (both sCO2 and water) on stack volume and performance of gas cooler is demonstrated for a temperature pinch of 3˚C pinned at the cold inlet. Multi-objective optimization is performed with water pumping power and gas cooler stack volume as objective functions to arrive at optimum channel dimensions and stack geometry. The dimensions are used to optimize the manifold flow passages in a similar way as done for the recuperator. The procedure is extended to double and quadruple banking configurations to assess the impact of banking on gas cooler performance and stack volume. Multi-objective optimizations are performed to arrive at an ideal banking arrangement for the gas cooler which provides minimum stack volume without compromising performance.