Experimental Study of the Slug Flow

O. C. Benítez-Centeno, O. Cazarez-Candia and S. L. Moya-Acosta

Abstract Slug flow is the flow pattern that more often is presented in two-phaseflow. It has a complex physical configuration which has not been totally under-stood. For decades slug flow has been modeled by mechanistic approach with theuse of the slug-unity concept. For this, slug length must be known. This parameteraffects the determination of shear stresses and then the pressure drop calculations.In this work data are presented from experiments carried out in a two-phase flowequipment. Equipment has a pipe of 12 m length and a diameter of 0.01905 m,which can be inclined from 0 to 90�. The measured data were: (1) angle for whichthe Taylor bubble breaks contact with the pipe wall, (2) the characteristic lengthsof the slug-unit, (3) pressure drop, and (4) presence of bubbles by means of opticalsensors. It was found that Taylor bubbles break contact with the wall pipe at 45�.With the voltage signal from optical sensors it was possible to quantify velocities,lengths and frequency for the Taylor bubbles.

O. C. Benítez-Centeno (&) � S. L. Moya-AcostaCentro Nacional de Investigación y Desarrollo Tecnológico,Prolongación Av. Palmira esq. Apatzingan, Col. Palmira,62490 Cuernavaca, Morelos, Méxicoe-mail: omarcbc@gmail.com

O. Cazarez-CandiaInstituto Mexicano del Petróleo, Eje Central Lázaro Cárdenas 152,07730 Col San Bartolo Atepehuacan, Méxicoe-mail: ocazarez@imp.mx

O. C. Benítez-Centeno � O. Cazarez-CandiaInstituto Tecnológico de Zacatepec, Calzada del Tecnológico No. 27,62780 Zacatepec, Morelos, México

J. Klapp et al. (eds.), Experimental and Theoretical Advances in Fluid Dynamics,Environmental Science and Engineering, DOI: 10.1007/978-3-642-17958-7_23,� Springer-Verlag Berlin Heidelberg 2012

287

1 Introduction

Slug flow is perhaps, between all the two-phase flow patterns, the one that appearswith a greater incidence (Taitel and Dukler 1977). It presents a complex geometry,faults in tubes and equipment, high pressure drop and instability, and then this flowis an undesirable phenomenon. Therefore it is necessary to study it in more detail.Slug flow has been modeled using mechanistic approaches and the slug-unit con-cept. This work presents an experimental study about the slug flow characteristics(Taylor bubble length and velocity, pressure drop). Experiments were made in atwo-phase flow equipment constituted by a tube of 12 m length and internaldiameter of 0.01905 m. The equipment can be inclined from 0 to 90�. In Fig. 1 thescheme for the two-phase flow equipment is shown.

2 Experiments

Using flow patterns maps, it was determined the water and air volumetric fluxesfor which the slug flow appears for 0, 3, 15 and 30� of inclination pipe. Theseinclinations are the four cases of tests whose parameters appear in Table 1.By means of rotameters (Fig. 1) the flow rates of each phase are introduced inthe mixer. The water–air mixture goes to out from the mixer and flow throughthe pipe.

The detection of Taylor bubbles is made by two infrared optical sensors fromwhich a voltage signal is obtained. Voltage varies from 4.6 to 0 volts. In Fig. 2, thephysical interpretation of the obtained voltage signals from sensors 1 and 2 can be

Fig. 1 Experimental equipment

288 O. C. Benítez-Centeno et al.

observed. When a liquid-slug is detected, a sensor obtains a voltage signal of 4.6volts and when a bubble passes voltage diminishes.

3 Results

The voltage signals data are treated statistically, obtaining their frequency bymeans of Fourier transform. Slug frequency is used to calculate slug intermittence(Eq. 1), mixture velocity (Eq. 2), slug length (Eqs. 3 and 4) and Taylor bubblevelocity (Eq. 5). The intermittence and mixture velocity are given by (Woods andHanratty 1996):

I ffi Usl

Usl þ Usgð1Þ

Table 1 Test cases

Parameters Case 1 Case 2 Case 3 Case 4

Diameter, (m) 0.01905 0.01905 0.01905 0.01905Pressure, (Pa) 107760 110976 111321 12046Water flow rate, (m3/s) 2.83E-04 0.94E-04 1.80E-04 1.38E-04Air flow rate, (m3/s) 4.53E-05 5.66E-05 8.18E-05 7.18E-05Water Temperature, (�C) 26.0 25.0 24.0 24.0Air Temperature, (�C) 26.0 25.0 24.0 24.0Inclination angle, (�) 0.0 3.0 15.0 30.0

0 2 4 6 8

2

3

4

5

Vol

tage

[vol

ts]

Time [s]

sensor 1sensor 2

Fig. 2 Typical signals of infrared sensors

Experimental Study of the Slug Flow 289

C ¼ C0 ðUsl þ UsgÞ ð2Þ

In Eqs. 1 and 2Usl is the liquid superficial velocity, Usg is the gas superficialvelocity, and C0 is a coefficient that is refered at the maximum average gasvelocity in the pipe center, which has a value of 1.2 for ascending flow and 1.12 fordescending flow (Hasan and Kabir 1988). The slug length can be determined by(Woods et al. 2006):

LLS ¼CI

fSð3Þ

LLS ¼1:2ðUsl þ UsgÞ

fs

Usl

Usl þ Usg

� �¼ 1:2

Usl

fs

� �ð4Þ

The taylor bubble velocity can be determined by (Chew-Chen 2001)

UBT ¼ LLSfS ð5Þ

In Eqs. 3–5, fs is the slug frequency, I is the slug intermittence, LLS is sluglength, and UTB is the Taylor bubble velocity. Figure 3 shows the frequencyspectrum obtained by using Fourier transform on voltage signals for sensors 1 and2 (Fig. 1). The frequency is of 3.5 Hertz. Once the slug frequency is determined,the liquid-slug length is calculated. In Table 2 the results are compared againstliquid-slug lengths determined by direct readouts from 30 photography’s foreach case.

0 1 2 3 4 5 6 70.0

0.1

Frequency (Hz)A

mp

litu

de

Fig. 3 Frequency for Case 3(Frequency = 3.5 Hz,Amplitude = 0.1219)

Table 2 Slug length (m) Angle (�) Measured Length(Photographs)

Computed Length(Voltage Signals)

0 0.2120 0.26563 0.2113 0.193415 0.2178 0.215030 0.1863 0.1860

290 O. C. Benítez-Centeno et al.

In Table 3 slug (LLS) and Taylor bubble (LTB) average lengths are shown, d isthe pipe diameter.

Table 4 shows the comparison of the slug length obtained in this work againstdata from literature. It is evident that it exist certain discrepancy among the data.

Another measured parameter is the pressure drop. It was measured by a digitalmanometer for water-to-air simultaneous flow. Figure 4 shows the behavior ofpressure as function of time for Case 3 (15�). To determinate the pressure drop themedian of the data (Fig. 4) was calculated. The results, for the cases stipulated inTable 1, are 344.73 Pa (Case 1), 345.79 Pa (Case 2), 689.47 Pa (Case 3) and690.02 Pa (Case 4).

Table 3 Taylor bubble andslug lengths obtained bymeasuring

Angle (�) 0 3 15 30

LTB (m) 0.4491 0.3497 0.2974 0.2651LLS (m) 0.2120 0.2113 0.2178 0.1860LTB.d 23.57 18.35 15.61 13.91LLS.d 11.12 11.09 11.43 9.76

Table 4 Slug length from literature for horizontal pipes

Author d (mm) Angle (�) LLS (da)

Dukler and Hubbard (1975) 38 0 12–30Nicholson et al. (1978) 25.51 0 20–30Gregory et al. (1978) 25.51 0 30Andreussi and Bendiksen (1989) 50 0 22Nydal (1992) 53, 90 0 15–20Manolis (1995) 78 0 10–25This work 19.05 0 9–11.4a d is pipe diameter

38:04 38:54 39:44 40:34 41:24 42:14 43:04

-500

0

500

1000

1500

2000

Time [min]

Pre

ssur

e [

N/m

2]

Fig. 4 Pressure drop forCase 3

Experimental Study of the Slug Flow 291

On the other hand, experiments were done to determinate the angle for whichTaylor bubble break contact with the pipe wall. In the present work it wasobserved that such angle is 45�. In Table 5 a comparison is shown for the angleobserved in this work against data from literature.

4 Conclusions

The differential pressure drop has been measured for 0, 3, 15 and 30� of inclinationfor water–air slug flow. Also liquid-slug lengths were measured. It was observedthat at 45� of inclination the Taylor bubble breaks contact with the pipe wall.The liquid-slug lengths and the angle of 45� measured in this work should be usedinto the calculation of shear stresses. This in order to obtain better predictions ofpressure drop when mechanistic models are used to simulate liquid–gas two-phaseslug flows.

References

Andreussi P, Bendiksen K (1989) An investigation of void fraction in liquid slugs for horizontaland inclined gas–liquid pipe flow. Int J Multiphase Flow 15(6):937–949

Chew-Chen L (2001) Slug development/dissipation in an inclined pipeline with changing pipeID. (SPE 68827)

Dukler AE, Hubbard MG (1975) A model for gas–liquid slug flow in horizontal and nearhorizontal tubes. Ind Eng Chem Sci Fundm 14(4):337–347

Gomez LE, Shoham O, Schmidt Z, Chokshi RN, Zenith R, Brown A, Northug T (1999) A unifiedmechanistic model for steady-state two-phase flow in wellbores and pipelines. Society ofPetroleum Engineers (SPE 56520)

Gregory GA et al (1978) Correlation of the Liquid volume fraction in the slug for horizontalgas–liquid slug flow. Int J Multiphase Flow 4:33–39

Hasan AR Kabir CS (1988) Predicting multiphase flow behavior in a deviated well. SPEDE 474,Trans AIME 285

Kaya AS, Sarica CY, Brill JP (2001) Mechanistic modeling of two phase flow in deviated Wells.Society of Petroleum Engineers (SPE 72998), pp 156–165

Manolis IG (1995) High pressure gas–liquid slug flow. PhD thesis, Department of ChemicalEngineering and Chemical Technology, Imperial College of Science, Technology andMedicine, UK

Nicholson R et al (1978) Intermittent two phase flow in horizontal pipes: predictive models. Can JChem Eng 56:653–663

Table 5 Angle for whichTaylor bubble breaks contactwith the wall pipe

Author Angle (�)

Kaya et al. (2001) 10Gomez et al. (1999) 86Zukoski (1966) 45This work 45

292 O. C. Benítez-Centeno et al.

Nydal OJ, Pintus S, Andreussi P (1992) Statistical characterization of slug flow in horizontalpipes. Int J Multiphase Flow 18(3):439–453

Taitel Y, Dukler AE (1977) A Model for slug frequency during gas–liquid flow in horizontal andnear horizontal pipelines. Int J Multiphase Flow 3:585–596

Woods BD, Hanratty TJ (1996) Relation of slug stability to sheding rate. Int J Multiphase Flow22(5):809–828

Woods BD, Fan Z, Hanratty TJ (2006) Frequency and development of slugs in a horizontal pipeat large liquid flows. Int J Multiphase Flow 2(20):902–925

Zukoski EE (1966) Influence of viscosity, surface tension, and inclination angle on motion oflong bubbles in closed tubes. J Fluid Mech 25:821–837

Experimental Study of the Slug Flow 293