Translational Periodic CFD Model: A Practical Solution to Capture Hydrodynamic Slugging in Horizontal Pipes

Abstract

Abstract Prediction of erosion rate, flow-induced vibrations and fatigue assessment depends strongly on the way multiphase flow occurs through fittings such as bends, tees, manifolds, etc. Therefore, it is important to have a better understanding of multiphase flow behavior to improve confidence on erosion and fatigue prediction. Though multiphase modelling is technically possible, it cannot be confidently deployed for field scale problems due to lack of validation and excessive requirement for computational resources. When simulating multiphase flow through a long horizontal pipe, it is important to specify appropriate boundary conditions for the gas and liquid phases at the inlet which can correctly define the slug length and velocity. This paper presents a translational periodic Computational Fluid Dynamics (CFD) model which can effectively capture slug flow in horizontal pipes as the slug unit cell can be approximated as periodic in the axial direction. The simulation of the slug flow was performed in a pipe section, with periodic boundary conditions under imposed pressure gradient and fixed overall liquid hold-up. Starting from smooth stratified flow, the slug flow pattern is simulated and validated in terms of slug speed and superficial velocities for gas and liquid phases. Furthermore, sensitivity studies were performed to understand the effect of mesh resolution, turbulence model, multiphase model, and time step size on CFD predictions in translationally periodic horizontal straight pipes. Experimental datasets of Manolis (1995) on high pressure gasliquid slug flow for hydrodynamic slugging in a horizontal pipe were used to validate multiphase CFD modelling. Subsequently an efficient algorithm was developed to couple results from periodic domain model of straight pipe to a more complex pipe configuration with one (or more) 90° bends. This algorithm was successfully demonstrated on TNO FIV JIP dataset (Belfroid et Al., 2016) for prediction of multiphase flow behavior and fluid forces in horizontal pipe with bend.