Advanced power cycle modelling

Abstract

Renewed interest in the sCO2 power cycle has necessitated further work to understand its dynamic behaviour in order to better understand operational aspects of this cycle. The aim of this thesis is to create a steady state model of a sCO2 printed circuit heat exchanger (PCHE) to be integrated into a model of a complete dynamic sCO2 power cycle being developed by the UQ Renewable Energy Centre (REC). The ability of the model to approximate inlet transients using a quasi-steady approach was also investigated. In the sCO2 application, conventional methods of steady state analysis are not valid, given the varying properties of the fluid as it passes through the heat exchanger. As a result, a nodalised approach to steady state modelling was applied, dividing the heat exchanger channel into discrete elements or nodes. Correlations for heat transfer and pressure drop from the literature were reviewed and suitable correlations coded into the solver. The output of this solver was validated using the commercial software IPSEpro for a sCO2 test case defined in previous research conducted by the Queensland Geothermal Centre of Excellence.The developed model is capable of solving for the internal and outlet states of the low and high temperature recuperators. The model was found to accurately capture real gas behaviour, such as the large variations in specific heat, density and heat transfer coefficient.These distributions matched those reported in the literature. The calculated outlet temperatures were in agreeance with those calculated by IPSEpro, with a maximum relative error of 3.5%. The pressure drops calculated by the model were lower, with a maximum relative error of 15.38%. The quasi-steady model was developed to determine whether a simpler, faster quasi-steady analysis can be used in place of a full dynamic analysis under certain circumstances. Results from the simulation of fast and slow transients were compared to those from the fully dynamic model currently under development at UQ.Relative errors ranged from 0.62% to 19.2%. However, the time varying profile of the response was not captured accurately by this model, and it was concluded that full dynamic modelling is required for all simulation of dynamic behaviour. The significance of this thesis is its contribution to the overall dynamic power cycle model under development, for which this model is a building block. This will contribute to the advancement of the sCO2 power cycle through deepening our understanding of its operational behaviour.ii