Two-phase slug flow is a common occurrence in wells, riser pipes and pipelines of crude oil and natural gas systems. Current predictive tools for two-phase flow are based on either the mixture model or the mechanistic two-fluid model. The latter one, also called phenomenological model, requires the use of closure relations to solve the transfer of mass, momentum and energy between the phases, in the respective conservation equations, so that integral flow parameters such as liquid holdup (or void fraction) and pressure gradient can be predicted. However, these closure relations carry the highest uncertainties in the model, since they are obtained empirically or through the use of overly simplified assumptions. In particular, significant discrepancies have been found between experimental data and closure relations for the Taylor bubble velocity in slug flow, which has been determined through an in-house study to strongly affect the pressure gradient and liquid holdup predicted by the mechanistic models of (Orell and Rembrand, 1986), (Ansari et al., 1994), and (Petalas and Aziz, 2000). In this work, Computational Fluid Dynamics (CFD) and the Level Set (LS) interface tracking method (ITM), implemented in the commercial code TransAT®, are employed to simulate the motion of Taylor bubbles in slug flow. Therefore, a numerical database is being generated to develop a new, high-fidelity closure relation for the Taylor bubble velocity as a function of the fluid properties and flow conditions, rendered non-dimensional through the use of the Froude, Reynolds, Eötvös and Morton numbers, and pipe inclination angle. The simulations suggest that in inclined pipes the Taylor bubble velocity is strongly reduced if there is no lubricating liquid film between the bubble and the wall. A simple analytical model predicting the drainage of this lubricating film is also presented.
Two-phase slug flow is a common occurrence in wells, riser pipes and pipelines of crude oil and natural gas systems. Current predictive tools for two-phase flow are based on either the mixture model or the mechanistic two-fluid model (Brill and Mukherjee, 1999). In the latter one, slug flow is modeled as a sequence of fundamental units, also called slug units. Each unit contains a long bullet-shaped bubble, known as Taylor bubble, and a liquid portion with smaller homogeneously distributed bubbles, known as liquid slug.