Session: 30-06 Fluid Dynamics

Paper Number: 122056

122056 - Supercritical Carbon Dioxide Mixing Loss Characteristics Near the Critical Point 

Supercritical carbon dioxide (sCO₂) has recently emerged as a promising working fluid candidate for the next generation of power cycles. Research has demonstrated that operating these cycles near the critical point of sCO₂ has the potential to enhance cycle efficiency. However, the substantial nonlinear variations in thermal properties in the vicinity of the critical point introduce difficulties in modelling and designing sCO₂ turbomachinery. The impact of the non-ideal fluid behaviour on turbomachine performance remains to be determined. One common scenario involves mixing two coflowing streams with distinct thermodynamic states, such as tip leakage flows and trailing edges in turbomachines.

 

This paper aims to have an evaluation of the mixing loss characteristic of sCO₂ near the critical point. A simple case of mixing in a constant area adiabatic duct involving two parallel flows is studied by control volume analysis. Two uniform flows are specified to have different thermodynamic states and their pressure, temperature, and velocity are varied as independent variables of this study. Assuming uniform outlet flows, the entropy rise coefficient can be calculated as an indicator of mixing loss. The independent and coupled effects of the independent variables on mixing loss can hence be investigated.

 

To understand the non-ideal fluid mixing loss characteristics, two different calculation cases will be set up for each specified average mixing state: a perfect gas CO₂ case, and a non-ideal fluid CO₂ case. Although having the same inlet conditions, each pair of cases was calculated using different principles: one relied on the perfect gas assumption to analytically solve the conservation equations, while the other used the Span-Wagner equation of state (EOS) to solve the conservation equations iteratively. The comparison of the entropy rise coefficient during the mixing process was then carried out. It highlights the differences between the calculations for the perfect gas and non-ideal fluid sCO₂, particularly in the vicinity of the critical point. The primary sources of deviations are the polytropic isentrope assumption and the speed of sound calculation. In this work, we find that the error introduced by the first source increases when the specific entropy of the average state decreases because the sCO₂ is becoming a liquid-like supercritical fluid and behaves more like an incompressible fluid. The error from the speed of sound calculation peaks near the critical point as a result of sparsely distributed isobars along the isentrope on a TS diagram.

 

From the study of independent effects from pressure, temperature and velocity differences, it is notable that the non-ideal fluid mixing loss is sensitive to the static temperature difference between the mixing flows, and the loss for the non-ideal fluid case is ten times higher compared to the perfect gas case near the critical point. For the same temperature difference, the loss coefficient is also contingent on the average state of the two mixing streams: the loss coefficient is greater when the average state is closer to the critical point. In contrast, mixing losses introduced by velocity differences are solely dependent on the magnitude of velocity difference and remain unaffected by the average states. The loss characteristics of the mixing case with pressure vary significantly if either stream is near the critical point. As for the combined effects of the independent variables, the effect of temperature and pressure differences on non-ideal fluid mixing loss is significantly coupled while their effects are almost independent for perfect gas mixing losses.

 

This study offers insight into the mixing behaviours of non-ideal fluids, particularly in the context of supercritical carbon dioxide in the vicinity of the critical point. It contributes to the understanding of mixing loss mechanisms in sCO₂ turbomachinery and holds significance for the development of mean-line models.

Presenting Author: Jinhong Wang Imperial College London

Presenting Author Biography: Jinhong is a PhD candidate in the Sustainable Energy Technology and Turbomachinery Lab at Imperial College London. His current research involves non-ideal compressible fluid dynamics, computational fluid dynamics (CFD), and supercritical turbomachinery design for sCO2 with a focus on the effects of thermodynamic critical point anomalies.

Authors:

Jinhong Wang Imperial College London
Teng Cao Imperial College London
Ricardo Martinez-Botas Imperial College London