Numerical Simulation of Two-Phase Slug Flow From Horizontal to Downward Inclined Pipe Using a Hybrid Code Based on Slug Tracking and Two-Fluid Methodologies

Author(s):  
Vinicius R. Almeida ◽  
Marco G. Conte ◽  
Fausto A. A. Barbuto ◽  
Cristiane Cozin ◽  
Rigoberto E. M. Morales

The slug flow regime is probably the prevailing pattern in the oil and gas industry, appearing in the nuclear industry as well. As a consequence, several studies have been conducted in order to understand the physics of this flow regime and to obtain a model to predict its properties. This work presents a transient hybrid methodology to simulate the gas-liquid slug regime in pipes with a change of direction from horizontal to downward inclined flow. The simulation initiates with a slug tracking model assuming the pipe to be filled with liquid, and follow the unit cells while they flow through the horizontal section; the information about the unit cells entering the pipe are obtained from experimental data. Near the elbow and beyond, the unit cells are simulated by a simplified two-fluid Lagrangian model, capable of providing flow details with the change of direction, and the dissipation of the slug flow to the stratified regime in a descendent slope. Simulations for a 4.862-m pipe were carried out, with the change of direction from horizontal to −3° and −5°. The results were compared to experimental data, showing that the model can successfully predict the flow behaviour for the given conditions.

2013 ◽  
Vol 135 (12) ◽  
Author(s):  
A. O. Nieckele ◽  
J. N. E. Carneiro ◽  
R. C. Chucuya ◽  
J. H. P. Azevedo

In the present work, the onset and subsequent development of slug flow in horizontal pipes is investigated by solving the transient one-dimensional version of the two-fluid model in a high resolution mesh using a finite volume technique. The methodology (named slug-capturing) was proposed before in the literature and the present work represents a confirmation of its applicability in predicting this very complex flow regime. Further, different configurations are analyzed here and comparisons are performed against different sets of experimental data. Predictions for mean slug variables were in good agreement with experimental data. Additionally, focus is given to the statistical properties of slug flows such as shapes of probability density functions of slug lengths (which were represented by gamma and log-normal distributions) as well as the evolution of the first statistical moments, which were shown to be well reproduced by the methodology.


Author(s):  
Stamatis Kalogerakos ◽  
Mustapha Gourma ◽  
Chris Thompson

Severe limitations of the use of three-dimensional computational fluid dynamics codes (CFD) arise when trying to simulate multiphase flow in long pipes due to time constraints. 1D codes for two-phase flow, based on two-fluid models, are fast but are known to be accurate only when the velocities are within the Kelvin-Helmholtz inviscid limit [1]. An alternative is to carry out a two-dimensional CFD simulation of a channel based on the Volume of Fluid (VOF) model. 2D CFD has a wider applicability range compared to 1D, it does not have the issue of ill-posedness and it also has better turbulence models built in. Again compared to 1D the 2D VOF model has a better interface description and wall treatment. In this paper a novel method is introduced that allows swift simulations of pipeline two-phase flow in the stratified and slug flow regime, by approximating the pipe as a channel and with a methodology that solves the problem of the interfacial velocity differences, inherent in the volume of fluid model. An initial validation using the wave growth problem has already been carried out [2]. Here a set consisting of 92 experimental cases in the slug flow regime has been simulated with 2D CFD, and the simulation results showed a good agreement with experimental results. Discussions in the paper include also the question of the range of applicability for 2D CFD, and the advantages and disadvantages compared to 3D CFD and also to 1D code based on the two-fluid model. Shear stresses are then extracted from the 2D CFD simulations and used to recalibrate the friction factors [3] used in the 1D code.


Author(s):  
Gustavo Montoya ◽  
Yixiang Liao ◽  
Dirk Lucas ◽  
Eckhard Krepper

Today Computational Fluid Dynamic (CFD) codes are widely used for industrial applications, mostly in the case of single phase flows in automotive or aircraft engineering, but multiphase flow modeling had gain an increasing importance in the last years. Safety analyses on nuclear power plants require reliable prediction on steam-water flows in case of different accident scenarios. This is particularly true for passive safety systems such as the GEKO component of the KERENA reactor. Here flashing may occur in the riser (Leyer and Wich, 2012). In such case, high gas volume fractions and the churn-turbulent flow regime may ensue. In the past, the codes for the prediction of churn-regime have not shown a very promising behavior. In this paper, a two-fluid multi-field hydrodynamic model has been developed based in the Euler-Euler framework. The main emphasis of this work has been on the modeling and applicability of various interfacial forces between dispersed gaseous phases and the continuous liquid, as well as bubble-bubble interactions, and the evolution of different bubble sizes in an adiabatic vertical pipe inside the churn-turbulent flow regime. All the expected mechanistic models that intervene in this flow pattern have been taken into account including drag force, wall force, lift force, turbulent dispersion, and bubble induced turbulence. Bubble breakup and coalescence has been defined (Liao et al., 2011), and in order to design a polydispersed model related to reality, the inhomogeneous MUSIG approach (Krepper et al., 2008) has been used to defined an adequate number of bubble size fractions which are arranged into different groups with their own velocity field. Based on these models, a series of simulations were made on the framework of ANSYS CFX 14.0, and all of the calculations were further validated with experimental data extracted from the TOPFLOW facility at the Helmholtz-Zentrum Dresden-Rossendorf. Different water and gas flow rates were used inside the churn-turbulent flow regime, as well as for the transition from bubbly to churn flow. The calculated cross-section averaged bubble size distributions, gas velocities, and time averaged radial profile for the gas fraction have shown a promising agreement with the experimental data. Nevertheless there are also clear deviations which indicate shortcomings of the present modelling. In order to further improve the modeling of this flow regime, a discussion based on the results will be used to shown a series of limitations of the actual modeling and possible solutions to be implemented in future works.


Author(s):  
Bruna Patricia Naidek ◽  
Kaique Leite ◽  
Camilla Verbiski Andrade ◽  
Cristiane Cozin ◽  
Fausto Arinos Barbuto ◽  
...  
Keyword(s):  

1994 ◽  
Vol 59 (12) ◽  
pp. 2595-2603
Author(s):  
Lothar Ebner ◽  
Marie Fialová

Two regions of instabilities in horizontal two-phase flow were detected. The first was found in the transition from slug to annular flow, the second between stratified and slug flow. The existence of oscillations between the slug and annular flows can explain the differences in the limitation of the slug flow in flow regime maps proposed by different authors. Coexistence of these two regimes is similar to bistable behaviour of some differential equation solutions.


1969 ◽  
Vol 9 (03) ◽  
pp. 293-300 ◽  
Author(s):  
J.E. Varnon ◽  
R.A. Greenkorn

Abstract This paper reports an investigation of unstable fingering in two-fluid flow in a porous medium to determine if lambda the dimensionless finger width, is unique For a viscous finger A is the ratio of finger width to the distance between the tips of the two trailing fingers adjacent to the leading finger. For a gravity finger lambda is defined as the ratio of finger width, to "height" of the medium perpendicular to hulk flow. This work confirms previous experiments and existing theory that for viscous fingering lambda approaches a value of 0.5 with increasing ratio of viscous to interfacial force. However, for a given fluid pair and given, medium, this ratio can he increased only by increasing the, velocity. Experiments on gas liquid systems show that the asymptotic value of lambda with velocity is not always 0.5. Apparently, for gas-liquid systems, the influence of the interfacial force cannot always he eliminated by increasing the velocity. For such systems lambda is a function of fluid pair and media permeability. If the gravity force normal to the hulk permeability. If the gravity force normal to the hulk flow is active, it damps out the viscous fingers except for an underlying or overlying finger. The dimensionless width of this gravity finger strongly depends on velocity and height of the medium, as well as the fluid and media properties. The existing experiments and theories are reviewed and the gravity, stable, and viscous flow regimes are described in view of these experiments and theories. The existence of a gravity-dominated unstable regime, a gravity-viscous balanced stable regime, and a viscous-anminated regime was demonstrated experimentally by increasing flow velocity bin a rectangular glass head model. Asymptotic values of the dimensionless finger width were determined in various-sized Hele-Shaw models with gravity perpendicular and parallel to flow. The dimensionless perpendicular and parallel to flow. The dimensionless finger width lambda was determined as a function of applied force, flow resistance, and fluid properties. The results are interpreted dimensionally. Some comments are made concerning possible scaling and meaningful extensions of theory to describe these regimes in three-dimensional flow. Previous description of unstable two-fluid flow in porous media is mainly restricted to studies of viscous-dominated instability. The direction of this study is to provide data and understanding to consider the more realistic problem of predicting flow in three dimensions that may result in instabilities that are combinations of all, four flow regimes. Introduction The unstable flow of two fluids is characterized by interface changes between the fluids as a result of changes in relative forces. In a given porous medium and for a given fluid pair the gravity force dominates flow at low displacement velocities. As the velocity increases the viscous forces begin to affect flow significantly, and eventually there is a balance between effects of the gravity and viscous forces. As velocity increases further, the viscous force dominates flow. In the plane parallel to gravity, four flow regimes result as the velocity is increased: a gravity-induced stable flow regime; a gravity-dominated unstable flow regime; a stable regime resulting from a balance between gravity and viscous forces; and a viscous-induced unstable flow regime. The gravity-induced stable regime is represented schematically in Fig. 1a. This general flow pattern persists with the displacing fluid contacting all of persists with the displacing fluid contacting all of the in-place fluid until the interface becomes parallel to the bulk flow. At this velocity a gravity finger forms, and the interface, is unstable in that the length of the gravity finger grows and the fluid behind the nose of the finger is practically nonmobile because of the small pressure gradient along the finger. The gravity-dominated unstable flow is shown schematically in Fig. 1b. As the injection rate is increased, the gravity finger thickens, perhaps until it spans the medium creating a stable interface where all of the in-place, fluid is again mobile. This regime would, not occur in the absence of gravity. It occurs due to the counter effects of the gravity and viscous forces (Fig. 1c). As the velocity of the displacing fluid increases, the viscous forces dominate, and, the interface breaks into viscous fingers (Fig. 1d). SPEJ p. 293


Author(s):  
Deoras Prabhudharwadkar ◽  
Chris Bailey ◽  
Martin Lopez de Bertodano ◽  
John R. Buchanan

This paper describes in detail the assessment of the CFD code CFX to predict adiabatic liquid-gas two-phase bubbly flow. This study has been divided into two parts. In the first exercise, the effect of Lift Force, Wall Force and the Turbulent Diffusion Force have been assessed using experimental data from the literature for air-water upward bubbly flows through a pipe. The data used here had a characteristic near wall void peaking which was largely influenced by the joint action of the three forces mentioned above. The simulations were performed with constant bubble diameter assuming no bubble interactions. This exercise resulted in selection of the most appropriate closure form and closure coefficients for the above mentioned forces for the range of flow conditions chosen. In the second exercise, the One-Group Interfacial Area Transport equation was introduced in the two-fluid model of CFX. The interfacial area density plays important role in the correct prediction of interfacial mass, momentum and energy transfer and is affected by bubble breakup and coalescence processes in adiabatic flows. The One-Group Interfacial Area Transport Equation (IATE) has been developed and implemented for one-dimensional models and validated using cross-sectional area averaged experimental data over the last decade by various researchers. The original one-dimensional model has been extended to multidimensional flow predictions in this study and the results are presented in this paper. The paper also discusses constraints posed by the commercial CFD code CFX and the solutions worked out to obtain the most accurate implementation of the model.


Author(s):  
Marcos F. Cargnelutti ◽  
Stefan P. C. Belfroid ◽  
Wouter Schiferli ◽  
Marlies van Osch

Air-water experiments were carried out in a horizontal 1″ pipe system to measure the magnitude of the forces induced by the multiphase flow. Forces and accelerations were measured on a number of bends and T-joint configurations for a wide range of operating conditions. Five different configurations were measured: a baseline case consisting of straight pipe only, a sharp edged bend, a large radius bend, a symmetric T-joint and a T-joint with one of the arms closed off. The gas flow was varied from a superficial velocity of 0.1 to 30 m/s and the liquid flow was varied from 0.05 to 2 m/s. This operating range ensures that the experiment encompasses all possible flow regimes. In general, the slug velocity and frequency presented a reasonable agreement with classical models. However, for high mixture velocity the measured frequency deviated from literature models. The magnitude of the measured forces was found to vary over a wide range depending on the flow regime. For slug flow conditions very high force levels were measured, up to 4 orders of magnitude higher than in single phase flow for comparable velocities. The annular flow regime resulted in the (relative) lowest forces, although the absolute amplitude is of the same order as in the case of slug flow. These results from a one inch pipe were compared to data obtained previously from similar experiments on a 6mm setup, to evaluate the scaling effects. The results for the one inch rig experiments agreed with the model proposed by Riverin, with the same scaling factor. A modification of this scaling factor is needed for the model to predict the forces measured on the 6mm rig. The validity of the theories developed based on the 6mm experiments were tested for validity at larger scales. In case of slug flow, the measured results can be described assuming a simple slug unit model. In annular and stratified flow a different model is required, since no slug unit is present. Instead, excitation force can be estimated using mixture properties. This mixture approach also describes the forces for the slug regime relatively well. Only the single phase flow is not described properly with this mixture model, as would be expected.


Sign in / Sign up

Export Citation Format

Share Document