CFD wall shear stress benchmark in stratified-to-annular transitional flow regime

摘要

Multiphase transport phenomena are an integral part of flow assurance design in subsea systems. One area of importance in subsea design is carbonic corrosion and how this affects the material selection and thickness. Corrosion prediction strongly depends upon the fluid flow regime, whereby a pragmatic interpretation of the effects of flow on corrosion mechanisms is used. It is postulated that the lack of agreement between corrosion modelling and corrosion field data is more likely due to the inadequate consideration of the multiphase flow regime during design, rather than any limitations in the corrosion models employed. This suggests a better representation of flow regime will provide greater confidence in the corrosion modelling predictions. Wall shear stress is an important parameter in corrosion models; as corrosion inhibitor removal from the wall increases with increasing wall shear stress and therefore escalates the corrosion rates. This parameter is a direct measure of the viscous energy loss within the turbulent boundary layer, and it is related to the intensity of turbulence in the fluid acting on the wall. It is not a force on the wall from the flowing fluid but, rather, a force within the flowing fluid at the wall. Therefore an adequate corrosion prediction has to take into account the flow behaviour accurately. Air-water two phase flow behaviour within a 10 inch horizontal pipe was modelled using Computational Fluid Dynamics (CFD) considering the free surface flow and the instability of the interfacial area. A very promising qualitative agreement has been obtained in comparison with the experimental observations for a stratified-to-annular transitional flow regime. Clear behaviour of the liquid film on the pipe wall and droplets entrained into the gas stream as well as the wave development/propagation can be seen in the simulations. The position of the maximum wall shear stress has been identified in the CFD simulations. After establishing good agreement on flow regime prediction, the CFD wall shear stress results were benchmarked against OLGA simulations, as well as data obtained using the NORSOK M506 method. The OLGA simulations underestimated the maximum liquid wall shear stress in comparison with CFD results (8.7 Pa vs. 96.0 Pa), whereas the maximum CFD mixture wall shear stress showed good agreement to the NORSOK M506 method (34.2 Pa to 27.0 Pa). A further study of a horizontal pipe with a 90° bend was carried out for the same input conditions. The study shows that the maximum wall shear stress for both the liquid and gas phases are higher in the bend than those predicted in the horizontal section. The wall shear stress around the bend was examined for variations in the flow development. The maximum water shear stress in the bend varied with time but this trend was not present in the horizontal section. It was concluded that the CFD wall shear stress analysis provided greater understanding of the potential for corrosion inhibitor stripping and therefore is beneficial to evaluate the maximum and variation in wall shear stress to assist in selecting the corrosion inhibitor and/or the material selection requirements for all subsea piping geometries, providing that the flow regime mechanism can be adequately modelled.