Article
Influence of slug flow on flow fields in a gas–liquid cylindrical cyclone separator: A simulation study

https://doi.org/10.1016/j.cjche.2020.03.026Get rights and content

Highlights

  • Method of simulating slug flow based on the VOF model is developed.

  • Evolution of liquid slug dissipation in inlet pipe of GLCC is revealed.

  • Effect of liquid slug on the flow field of GLCC is analyzed.

  • Existence of optimal expanding diameter ratio and inclination angle is confirmed.

Abstract

A simulation method for slug flow based on the VOF multiphase flow model was implemented in ANSYS® Fluent via a user-defined function (UDF) and applied to the dissipation of liquid slugs in the inlet pipe of a gas–liquid cylindrical cyclone (GLCC) separator while varying the expanding diameter ratio and angle of inclination. The dissipation of liquid slug in inlet pipe is analyzed under different expanding diameter ratios and inclination angles. In the inlet pipe, it is found that increasing expanding diameter ratio and inclination angle can reduce the liquid slug stability and enhancing the effect of gravity, which is beneficial to slug flow dissipation. In the cylinder, increasing the expanding diameter ratio can significantly reduce the liquid carrying depth of the gas phase but result in a slightly increase of the gas content in the liquid phase space. Moreover, increasing the inclination angle results in a decrease in the carrying depth of liquid in the vapor phase, but enhances gas–liquid mixing and increases the gas-carrying depth in the liquid phase. Taking into consideration the dual effects of slug dissipation in the inlet pipe and carrying capacity of gas/liquid spaces in the cylinder, the optimal expanding diameter ratio and inclination angle values can be determined.

Introduction

Gas–liquid separation technology has become an indispensable link in the gathering and transportation of oil fields. The gas–liquid cylindrical cyclone (GLCC) separator is widely used in gas–liquid separation processes as it offers the advantages of high separation efficiency, low volume footprint, and simplicity of operation. The structure of a typical GLCC consists of a vertical cylinder with a downward inclined tangential inlet pipe and two outlet pipes [1]. The inlet pipe parameters include angle of inclination, length, diameter, and expanding diameter ratio, all of which have a direct impact on the flow patterns therein. Various flow patterns such as slug flow, annular flow, and stratified flow appear in the inlet pipe, which affect the cyclonic flow field distribution in the cylinder and thereby the separation efficiency.

From the literature regarding GLCCs, it is generally believed that an appropriate downward inclination of the inlet pipe is conducive to the formation of stratified flow, which expands the effective working range of the equipment. Jeyachandra [2] distinguished the inlet flow into stratified smooth, elongated bubble, and slug flow patterns. Moreover, they established a predictive model for downward inclined pipe flows. The diameter of the inlet pipe is designed to guarantee stratified flow in the inlet nozzle, and is calculated using the Taitel and Dukler model for predicting flow regime transitions [3]. An inlet pipe inclination of 25° to 30° ensures that the liquid flows along the lower surface of the pipe and does not block the entry of gas into the upper part of the GLCC [4]. Kouba [5] observed that the optimal inlet pipe angle (i.e. which achieved the best gas–liquid separation) was approximately 27°. When the heat transfer coefficient was contained [6], the optimal inclination leading to the highest heat transfer coefficient for stratified flows was found to be 15°. In addition to proper downward inclination, an increase in inlet pipe length can promote preliminary gas–liquid separation in the pipe itself; however, the allowable length is limited by the equipment installation space. The optimal length of the inlet pipe has been found to be 1.0–1.5 m [5].

While there are plenty of studies regarding the inlet pipe conditions in the literature, there has been relatively little research to date on the flow in the cylinder itself due to the presence of complex, irregular three-dimensional unsteady flow fields with rotation therein. In the cylindrical coordinate system, the three components of the cylinder flow field are the tangential, axial, and radial velocities. The tangential velocity is composed of a quasi-free vortex near the cylinder wall and a quasi-forced vortex near the axis [7], which promotes gas–liquid separation. The axial velocity can be divided into the outer downstream flow and the inner upstream flow [8,9]. The radial velocity is difficult to measure experimentally as it is an order of magnitude lower than the other velocities. In experiments, it was clearly observed that the vortex core in the center of the cylinder alternated between laminar and turbulent states, showing complex dynamic characteristics [10]. The vortex core directly affects the gas carrying capacity, and hence the separation efficiency. Marti [11] attempted to establish a mechanism model to predict gas carrying in GLCCs, which was capable of predicting the gas–liquid interface near the inlet of cylinder according to the radial distribution of the tangential velocity.

The developments in numerical simulation methods in recent years have allowed separator structures to be optimized through the use of CFD simulations, which makes up for the shortcomings of experiments. However, for numerical simulations to be useful it is crucial to select an appropriate turbulence model. Comparing between three commonly used turbulence models—RANS, URANS and LES—it was found that LES has a higher simulation accuracy but requires a longer calculation time [12], whereas RANS speeds up the calculation time while maintaining acceptable simulation accuracy and is therefore widely used. Based on 2D and 3D CFD simulations of GLCC flow fields [13], the existence of short-circuit and circulating flows can be observed directly from the velocity field distributions [14]. Additionally, by tracking the trajectory of a single bubble, it is possible to analyze the effects of bubble size, viscosity, Reynolds number, and inlet tangential velocity on bubble carrying [15,16]. Hence, the numerical simulation results can be used to optimize the structure of GLCCs. It was found that using a rectangular inlet can reduce vortex distortions [17,18], which improves the flow characteristics by comparison with a circular inlet.

The inlet flow pattern can affect the flow field distribution in the cylinder. One of these patterns is slug flow, which is described as the alternating flow of liquid ‘slugs’ and long bubbles which fill in the whole cross-section of the pipe [19], and is particularly disadvantageous to the gas–liquid distribution in the cylinder. Due to the instability of slug flow, numerical simulation methods are commonly used to study its characteristics. For this purpose, a one-dimensional steady state model of slug flow is widely used, but is limited to providing only the average characteristic parameters. Furthermore, Taitel [20] showed that non-physical profiles appeared in the steady-state model during the calculation of slug flow in a downward inclined pipe. In order to simulate transient phenomena, such as topographic relief and severe slugging flow, a variety of one-dimensional transient models and calculation methods of slug flow have been presented, many of which have been successfully commercialized [21]. The main types of transient simulation models for slug flow are one-dimensional two-fluid models and the liquid slug tracking models [22,23]. As the name implies, the one-dimensional models cannot be used to obtain three-dimensional parameters of slug flow. Additionally, the models are often incapable of simulating slug flows in short length pipes. Existing CFD models are also constrained in terms of the ability to simulate gas–liquid interfaces. Therefore, it is difficult to obtain an accurate picture of multiphase flow interfaces (not least slug flows with intense changes in the gas–liquid interface) in pipes using such software.

Given the limitations of current CFD models, more attention has traditionally been paid to the separation performance under stable inlet flow conditions. Consequently, research on the dissipation of liquid slugs in the inlet pipe and the effects of inlet slug flow on the flow field in the cylinder is scarce. Therefore, establishing a reasonable slug flow model to explicitly determine the effects of liquid slugs on the flow field in the cylinder is of key importance as it is more in line with actual production conditions in industrial GLCCs.

Against this backdrop, the present study employed a one-dimensional steady slug flow model in the ANSYS® Fluent CFD software package to define the initial conditions of transient slug flow. A simulation method for slug flow was developed based on the VOF multi-flow model [24], whereby liquid slug dissipation in the inlet pipe was studied, and the effects of expanding diameter ratio and angle of inclination on cyclonic flow field distributions were investigated.

Section snippets

Geometric model and mesh generation

Fig. 1 shows the GLCC structure assumed in the simulations. As the figure shows, the inlet pipe was successively composed of horizontal, riser, expanding, and inclined sections. The diameters before and after the expanding section are d1 and d2, respectively, with K being the expanding diameter ratio (K = d22/d12), while L and θ are the length and angle of the downward inclination section, respectively. The specific parameter values are shown in Table 1. The gas and liquid phase outlets of the

Dissipation process

Over the short length of the GLCC inlet pipe, liquid slug dissipation occurs abruptly, which is different from the linear decrease in slug length seen in long downward inclined pipes. Fig. 4 shows the evolution of a single liquid slug in the inlet pipe. The liquid slugs first go through the horizontal section of the inlet, Fig. 4(a), wherein a stairstepping slug head and slug tail with annular liquid film develop due to liquid collection and release by the slug head and tail, respectively. The

Conclusions

In this study, liquid slug dissipation in the inlet pipe of a GLCC separator was studied using a simulation-based approach. The effects of varying the expanding diameter ratio, K, and the angle of inclination, θ, on cyclonic flow field distributions were investigated. The main conclusions which may be drawn from the results presented herein are as follows:

  • (1)

    The dissipation of liquid plug has four processes in the inlet pipe: Firstly, stairstepping slug heads and slug tails with annular liquid

Nomenclature

    Af

    pipe cross sectional area of the liquid film, m2

    Ag

    pipe cross sectional area of the elongated bubble, m2

    Aslot

    inlet slot size, mm2

    Dsep

    diameter of the cylinder, mm

    d

    pipe diameter, m

    d1

    diameter before the expanding section, mm

    d2

    diameter after the expanding section, mm

    ff

    friction coefficient between the liquid film and the wall

    fg

    friction coefficient between the elongated bubble and the wall

    fi

    the friction coefficient at the gas/liquid interface

    HLS

    liquid holdup in slug

    HLF

    liquid holdup in the liquid film

Acknowledgements

This work is financially supported by the National Science Foundation of China (Nos. 51274233, 51574273) and the Province Natural Science Foundation (Grant No. ZR2014EEM045).

References (36)

  • T.K. Kjeldby et al.

    Lagrangian slug flow modeling and sensitivity on hydrodynamic slug initiation methods in a severe slugging case

    Int. J. Multiphase Flow

    (2013)
  • V. De Henau et al.

    A transient two-fluid model for the simulation of slug flow in pipelines

    I. Theary. Int. J. Multiphase Flow

    (1995)
  • R. Hreiz et al.

    On the effect of the nozzle design on the performances of gas–liquid cylindrical cyclone separators

    Int. J. Multiphase Flow

    (2014)
  • F. Chang et al.

    Turbulent flow field in tangentially injected swirl flows in tubes

    Int. J. Heat Fluid Flow

    (1994)
  • K.H. Bendiksen

    An experimental investigation of the motion of long bubbles in inclined tubes

    Int. J. Multiphase Flow

    (1984)
  • G.A. Gregory et al.

    Correlation of the liquid volume fraction in the slug for horizontal gas-liquid slug flow

    Int. J. Multiphase Flow

    (1978)
  • M. Cook et al.

    Slug length prediction in near horizontal gas-liquid intermittent flow

    Chem. Eng. Sci.

    (2000)
  • B.C. Jeyachandra, C. Sarica, H.Q. Zhang, E.J. Pereyra, Effects of Inclination on Flow Characteristics of High Viscosity...
  • Cited by (0)

    View full text