CFD Analysis and Experimental Validation of a sCO2 Pocket Damper Seal

Abstract

Abstract Supercritical CO2 (sCO2) used as a closed loop fluid for power cycles in waste heat recovery or small energy production services significantly reduces the size of turbomachinery due to high operating pressure and density of sCO2 fluid above its critical point. For this reason, this OEM is working to develop the relevant centrifugal compressor technology. A damper seal like a Pocket Damper Seal (PDS) is a very effective choice for a sCO2 compressor balance piston both in design conditions and in a potential transient scenario operating in multiphase regime. The leakage and rotordynamic coefficients predictions for such a seal are very critical when this is operating close to CO2 critical point due to the large variation in gas properties. In this paper, a PDS with upstream shunt-holes is the subject of the study. A Bulk Flow Model (BFM) approach was originally used to design the PDS for both leakage and rotordynamic coefficients. To verify the design of such a seal, a Computational Fluid Dynamics (CFD) based approach was employed at a later stage. The CFD approach outlines the use of various Equations of States (EOS) like Peng-Robinson, Redlich-Kwong, and Span-Wagner (NIST-Refprop) for the computation of leakage flow. Sensitivity study on the range of NIST-Refprop look-up table was also carried out to have a trade-off between leakage accuracy and computational time. The CFD based rotordynamic coefficients were also computed using the Integral Perturbation method (IPM) by leveraging NIST-Refprop EOS. The experimental data are relevant to the sCO2 compressor prototype tested at the Authors’ internal test bench. Various parameters related to compressor performance along with balance piston leakages were measured. The leakage flows predicted by the CFD and BFM approaches were validated against the experimental results. Leakage comparison results show that BFM approach predictions are underestimating the experimental results. The CFD approach, on the other hand, overpredicts the leakage flow especially for the high-pressure operating points. The CFD approach was also validated for 6 different operating points of the compressor. Rotordynamic coefficients prediction from the CFD approach shows very good agreement with the BFM results at the design point and both agree with the test results which always showed a very stable compressor behavior.