Abstract
In a CO2 transport pipeline, decompression can occur either due to planned maintenance or accidental rupture. We investigate the impact of modelling simplifications and assumptions regarding the purity of CO2 on the decompression of an industrial scale transport pipeline. Using industrially relevant compositions of impurities we both calculate simplified (isentropic) decompression curves as well as perform full-scale fluid simulations using models for heat transfer, friction and choked flow. Herein, we compare the use of the Peng–Robinson cubic equation of state (EOS) with the use of the highly accurate EOS-CG and GERG-2008 EOS.
We find that the saturation pressure can change significantly when certain impurities are present, even when in small quantities. A simplifying assumption that CO2 is pure can therefore lead to significant underestimation of the fluid pressure during the decompression, with consequences for the prediction of running ductile fractures. The choice of EOS was found to mainly affect the velocities of the initial decompression wave, with the long-time evolution of the pipeline decompression remaining relatively unchanged. The temperature minimum did neither depend significantly on the choice of EOS nor the presence of impurities.
We find that the saturation pressure can change significantly when certain impurities are present, even when in small quantities. A simplifying assumption that CO2 is pure can therefore lead to significant underestimation of the fluid pressure during the decompression, with consequences for the prediction of running ductile fractures. The choice of EOS was found to mainly affect the velocities of the initial decompression wave, with the long-time evolution of the pipeline decompression remaining relatively unchanged. The temperature minimum did neither depend significantly on the choice of EOS nor the presence of impurities.