You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/223774674

Liquid holdup in horizontal two-phase gas—liquid flow

Article  in  Journal of Petroleum Science and Engineering · August 1996


DOI: 10.1016/0920-4105(95)00069-0

CITATIONS READS

48 2,033

1 author:

Ghassan H Majeed
University of Baghdad
49 PUBLICATIONS   386 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Two-phase Fluid Flow in Pipes View project

IOT,security,lightweight cryptography ,AES View project

All content following this page was uploaded by Ghassan H Majeed on 06 December 2017.

The user has requested enhancement of the downloaded file.


Journal of Petroleum Science and Engineering 15 ( 1996) 27 l-280

Liquid holdup in horizontal two-phase gas-liquid flow


G.H. Abdul-Majeed
Computer Center, Uniuersity ofBaghdad, P.O. Box 47101, Jadiriyah, Baghdad, Iraq
Received 1 August 1995; accepted 15 October 1995

Abstract

The purpose of this study was to simplify and improve the mechanistic model developed by Taitel and Dukler (1976) for
estimating the liquid holdup in horizontal two-phase flow. An experimental study was first conducted to develop a data bank
used for evaluation and improvement. The holdup data were obtained using an air-kerosene mixture flowed through a test
section consisting of a horizontal pipe 2-in (50.8 mm) in diameter and 118 ft (36 m> long. The liquid holdup ranged from
0.009 to 0.61 and the flow patterns observed were stratified, slug and annular.
It was shown that the Taitel-Dukler model, which consists of several equations and requires an iterative method to solve,
can be accurately represented by a single explicit equation.
Based on the measured data, it was found that Taitel-Dukler model tends to overestimate liquid holdup for stratified
wavy, slug and annular flow patterns, whereas it tends to underestimate the liquid holdup for stratified smooth flow. An
empirical modification, therefore, is proposed which results in a significant improvement in predictions compared to
experimental data.
Based on statistical results, it is observed that the proposed model gives excellent results and clearly outperforms the
original model and all the existing correlations when tested against the present data (89 points) and against data from the
literature (111 points).

1. Introduction when using the Taitel-Dukler model. First, it re-


quires an iterative method to solve. Second, multiple
Several correlations have been published for pre- roots may occur in some cases (Crowley and Rothe,
dicting liquid holdup in horizontal pipes. Most of 1986; Baker et al., 1988).
these correlations are empirically developed from The purpose of this study was to simplify and
experimental data. Also, various theoretical models improve the Taitel-Dukler model. The author first
with different degrees of complexity have been pro- performed an experimental study to obtain measured
posed in the literature. Among them, the physical data for evaluation and improvement.
model developed by Taitel and Dukler (1976) has
gained great popularity. All the theoretical models
proposed later are based on this approach (Andritsos
and Hanratty, 1987; Baker et al., 1988; Xiao et al., 2. Experimental equipment and testing procedure
1990). Many investigators evaluated the Taitel-
Dukler model and found that it tended to underesti- A schematic flow diagram of the equipment used
mate holdup values. Two problems are encountered is shown in Fig. 1. The test section consisted of a

0920-4105/96/$15.00 Copyright 0 1996 Elsevier Science B.V. All rights reserved.


SSDIO920-4105(95)00069-O
272 G.H. Ahdul-Majeed/Journal of PetroleumScience and Engineering 15 (19961 271-280

plastic pipe with inside diameter of 2-m (50.8 mm> simple gravimetric methods were utilized for the in
and length of 118 ft (36 m>, supported along its situ kerosene content of the test section. For each
length in a horizontal position by angle iron sections. test, air and kerosene were allowed to flow through
Liquid holdup was measured using three pneumati- the separator until a steady-state condition was
cally actuated quick closing ball valves. Two valves reached. The pressure, temperature and flow rates
were installed at the ends of the plastic pipe, and the were recorded when steady-state conditions were
third one was placed in the bypass line. The three reached. Then the holdup valves were actuated and
valves were full-opening ball valves, with inside the amount of liquid trapped in the test line (between
diameter equal to the inside diameter of the pipe, so valves) was measured to obtain holdup. To get accu-
that the flow stream was not disturbed by passing rate results, the last step was repeated at least 5
through the open valves. The valves were controlled times, and the average value was then determined.
simultaneously by a single switch. The physical properties of the kerosene were cal-
Kerosene was used as a liquid phase and air as a culated using the following equations, which are
gas phase. Air was supplied by a compressor rated at based on laboratory measurements:
7.50 Mscf/D (21.23 m3 D-l> at 120 psig (827.4
p, = 62.4(0.8252 - 0.000727)
kPa). A pressure regulator was used to maintain the
pressure at 130 psig. The temperatures were read p, = 3.473004exp( -0.02016T)
from a thermometer. Pressure transducers and pres-
o, = 29.9776 - 0.13176T
sure gauges were located at several points on the
plastic pipe. where T is the temperature in “C.
Kerosene and air were supplied to the test line The total number of measured data was 89. The
through 3-in lines in which orifice plates and rotame- liquid holdup covered by these data ranged from
ters were installed to measure the flow rates. Control 0.009 to 0.61. Based on Mandhane et al.‘s (1974)
and measurement of the variables were made with flow pattern map, 20 points were in stratified smooth
the aid of calibrated orifice plates and rotameters for flow, 13 in the stratified wavy flow, 33 in the slug
the air and kerosene flow rates. Calibrated dial ther- flow, and 23 in the annular flow. The data are
mometers were used for measuring temperatures, and presented in Table 1.

X CHECK VALVE
frq GATE VALVE
@I PNEUMATICALLY
ORIFICE
ACTUATEDVALVE
METER
8 PRESSUREGAUGE
AIR
sUppLY = DO I; I :: =_ ,QJ TEMPERATUREGAUGE

PUMP

Fig. 1. Schematic flow diagram of the apparatus used by the author.


G.H. Abdul-Majeed/Journal of Petroleum Science and Engineering 15 (1996) 271-280 273

3. Taitel-Dukler model Taitel and Dukler (1976) assumed that f, =fg,


and concluded that h, and E, are functions of the
The final dimensionless equation of the Taitel- parameter X only. Eq. (1) is an implicit function of
Dukler model is: h, and requires an iterative method for solution.

4. Simplification

zz
0 (1) To eliminate the problems associated with the
where the superscript (-) over a variable designates Taitel-Dukler model, an attempt was made to re-
it as dimensionless; and X is the Lockhart-Martinelli place this implicit model by a single explicit equa-
parameter (Lockhart and Martinelli, 1949) defined as tion. Eq. (1) is solved for the parameter X by
follows: assuming values of h, in the range of O-1.0. The
liquid holdups corresponding to the assumed values
of h, are calculated using Eq. (14). Using regression
(2) analysis with E, values as dependent variable and X
values as independent variable, the following equa-
For turbulent flow, m = 0.2, whereas for laminar
tions were obtained:
flow, m = 1.0. The f, and fg are the interfacial and
For turbulent flow:
gas wall friction factors, respectively. The other
dimensionless variables in Eq. (1) can be determined E, = exp( -0.9304919 + 0.5285852 R
as follows: -9.219634 X 1O-2 R2 + 9.02418 x 10m4 R4)
h, = h,/d (3) (16)
F=(2h,-1) (4) For laminar flow:
sg = cos-‘( F) (5) E, = exp( - 1.099924 + 0.6788495 R

s, = ?r - s, -0.1232191 x 1O-2 R2 - 1.778653 x 10-j R3


(6)
+ 1.626819 X 1O-3 R4) (17)
ji = (1 - qo.5 (7) where
Kg = 0.25( $ - Fs,) (8) R = ln( X)

&=0.25(~4&) Detailed calculations show that the Taitel-Dukler


(9)
model and the findings of the author give approxi-
v, = 0.25+& (10) mately the same results, with an average percent
difference of u 0.01%. Fig. 2 shows excellent
VP = 0.25~r/‘A, (11) agreement between the Taitel-Dukler model and the
0, =4x,/s, new equations. Due to their accuracy and simplicity,
(12)
the new equations are recommended to be used
Dg = 4/i/( sg + s,) (13) instead of the implicit model developed by Taitel
and Dukler.
All these variables are functions of h, only. The
liquid holdup can be determined from (Xiao et al.,
1990):
5. Modification
E = 6 - sin 4
1 2r
( 14) On deriving their model, Taitel and Dukler as-
sumed that the ratio of the interfacial to gas wall
where
friction factor <j/f,> is constant and equal to unity.
4= ~COS-I( -F) (15) It has been reported in the literature that this assump-
214 G.H. Ahdul-Majeed/ Journal of Petroleum Science and En~ineerinl: 15 (19961 273-280

Table I
Experimental liquid holdup data

EL Flow regime

-
I I .246 0.329 898.4 46.7 0.4034 slug
2 0.366 0.032 829.4 47.2 0.541 I stratified
3 20.660 0.121 277.9 47.8 0.0350 annular
4 0.892 I.531 795.7 43.3 0.5180 slug
5 0.75 I 0.982 781.2 45.0 0.505 I slug
6 7.882 0.104 374.4 46.7 0.0945 slug
7 0.344 0.006 843.2 34.4 0.31 IO stratified
8 0.958 0.402 863.9 40.6 0.4545 slug
9 0.685 0.701 919.1 44.4 0.5 I45 slug
IO 0.653 0.120 760.5 35.6 0.5188 stratified
II 0.413 0.095 767.4 42.8 0.6061 stratified
12 7.061 0.102 388.2 35.0 0. I040 \lug
I3 3.148 I .ooo X84.6 45.0 0.299 I slug
I4 18.992 0.40 I 307.5 47.2 0.077 I slug
I5 2.720 0.989 822.5 40.0 0.3494 slug
I6 12.334 0.1 I2 296.5 31.1 0.053 I annular
17 1.400 I .200 786.7 44.4 0.5406 AU&
I8 16.62 I 0.555 328.9 45.0 0.07 14 slug
I9 3.085 0.700 914.2 43.3 0.2742 slug
20 IO.291 0.301 336.5 27.8 0.1 I91 <lug
21 6.149 0.350 775.7 35.0 0.1527 \1ug
22 0.787 0.03 I 766.0 44.4 0.3364 stratified
23 0.733 0. I28 752.9 37.8 0.4619 stratified
24 0.315 0.004 810.1 35.0 0.3527 stratified
2s 0.459 0.102 768.8 42.2 0.5225 stratified
26 I.084 0.109 547.4 46.7 0.3655 stratified
27 I.951 I.321 590.2 41.1 0.4020 slug
28 1.064 0.710 649.5 35.6 0.4617 slug
29 0.196 0.002 426.8 3x.3 0.3636 stratified
30 1.022 0.991 597.8 45.6 0.47 I8 slug
31 6.612 I .o IO 384.0 36.7 0.2138 slug
32 0.59 I 0.055 549.5 42.2 0.4503 stratified
33 0.621 0. I I I 549.5 38.9 0.5205 stratified
34 0.993 0.1 I2 527.4 35.6 0.4126 stratified
35 3.383 1.000 863.9 32.6 0.3214 alug
36 1.203 1.042 S38.5 43.3 0.434 I slug
37 13.01 I 0.121 377.1 3 I .7 0.062 I annular
38 0.634 0.095 605.4 46.7 0.5046 stratified
39 15.820 0.52 I 349.6 46.7 0.1030 Slug
40 22.067 0.045 246. I 42.2 0.0 159 annular
41 13.023 0.3 1 I 410.9 45.6 0.0979 slug
42 I 1.545 0.410 452.3 40.6 0.1 I I9 slug
43 10.925 0.306 47 I .6 40.0 0.1135 slug
44 IS.712 0.0062 274.4 43.3 0.0092 wave
45 IS.821 0.035 282.7 46. I 0.0303 wave
46 33.264 1.825 429.5 35.6 0.0920 slug
47 48.537 0.100 197.2 47.2 0.0130 annular
48 48.908 0.179 211.0 42.2 0.0187 annular
49 47.665 0.323 246.8 45.0 0.0265 annular
50 3.334 0.324 512.3 42.2 0.2567 slug
51 17.861 0.01 I 237.9 38.9 0.0 109 wave
52 3.412 0.119 569.5 38.9 0.1910 slug
G.H. Abdul-Majeed/Journal ofPetroleum Science and Engineering 15 (1996) 271-280 215

Table 1 (continued)
No. “v v,, El Flow regime
(In s-1) (m s- ‘)

53 12.225 0.221 341.3 44.4 0.0890 slug


54 6.553 0.009 307.5 44.4 0.0284 stratified
55 2.119 0.044 556.4 46.1 0.1891 stratified
56 I 1.670 0.058 304.7 45.6 0.04554 wave
51 15.330 0.092 287.5 41.1 0.0433 annular
58 16.771 0.056 289.6 46.7 0.0274 wave
59 19.154 0.037 253.0 37.2 0.0 169 annular
60 18.555 0.337 248.2 44.4 0.0725 annular
61 29.824 0.395 212.3 46.7 0.0465 annular
62 6.907 0.057 408.2 47.2 0.0806 wave
63 14.133 0.037 241.3 45.6 0.0261 wave
64 39.144 0.041 230.3 45.6 0.0114 annular
65 1I.833 1.144 347.5 45.0 0.1504 SlUg

66 16.333 0.982 273.1 43.3 0.1290 slug

67 13.444 0.502 268.9 42.8 0.1166 slug

68 20.225 0.391 280.6 43.9 0.0716 annular


69 38.187 0.112 255.1 41.7 0.0171 annular
70 44.111 0.331 279.9 38.3 0.0266 annular
71 43.227 0.560 294.4 40.6 0.0363 annular
72 5.334 0.0274 231.7 47.8 0.0710 wave
73 10.607 0.091 265.4 48.3 0.0680 wave
74 15.300 0.018 293.1 48.9 0.0170 wave
75 14.780 0.06 1 298.5 34.4 0.0410 wave
16 14.510 0.040 293.7 40.0 0.0310 wave
17 2.890 0.006 437.1 42.8 0.0410 stratified
78 1.600 0.0061 431.6 37.0 0.0910 stratified
79 10.400 0.520 424.0 38.3 0.1420 slug
80 I 1.200 0.0061 216.5 41.1 0.0102 wave
81 1.800 0.058 548.8 35.6 0.2900 stratified
82 1.010 0.0062 554.3 47.2 0.1330 stratified
83 22.010 0.0305 211.0 43.3 0.0190 annular
84 30.880 0.09 I 282.0 38.9 0.0245 annular
85 39.140 0.043 231.7 42.8 0.0110 annular
87 45.870 0.677 248.9 40.0 0.0520 annular
88 30.600 0.305 286.8 43.3 0.0570 annular
89 27.550 0.240 293.7 37.2 0.0350 annular

tion resulted in underestimation of liquid holdup and, equations ( Eq. (16) and Eq. (17)), the following
hence, in pressure drop for stratified flow (Xiao et equation is proposed to predict the true liquid holdup:
al., 1990). Several alternative correlations, therefore,
were proposed in the literature (Lopes and Dukler, (18)
1986; Andritsos and Hanratty, 1987; Baker et al.,
1988) for estimating this ratio. These correlations, The theoretical holdup value can be determined from
however, require a trial-and-error solution and, thus, either Eq. (16) or Eq. (171, and C is the correction
increase the degree of complexity of the Taitel- factor to account for the errors due to using <f,/f,>
Dukler model. = 1. Using Eq. (1 S), the correction factor is calcu-
In order to allow for the effects of a wavy inter- lated for each data point (89 values). Analysis of
face without giving up the simplicity of the new results shows that the correction factor appears to be
276 G.H. Abdd-Mujeed/ Journal of’ Petroleum Sciencr and Engineering 15 (1996) 271-280

Lockhart-Matiinelli X Parameter

Fig. 2. Comparison between Taitel-Dukler model and that devel- Fig. 4. Relationship between correction factor and (V\? V,,) for the
oped by the author. present data.

related to the type of flow regime. For stratified


smooth flow, the correction factors are greater than whereas for other regimes they are > I .O. Many
unity. For other flow regimes (stratified wavy, slug investigators reported that the values of the ratio
and annular), the correction factors required to sat- f,/f, for stratified wavy, slug and annular flow
isfy the measured E, data are always smaller than regimes, are always greater than unity (Wallis, 1969;
unity. This leads to the conclusion that the Taitel- Chen and Spedding, 1981; Andritsos and Hanratty,
Dukler model (with the assumption of J/f, = 1) 1987; Tronconi, 1990).
tends to underestimate E, for stratified smooth flow, Because the correction factor values are found to
and to overestimate E, for stratified wavy, slug and be dependent on the flow regime, it is necessary to
annular flow regimes. Based on Eq. (l), E, is in- correlate these values with the parameters that con-
versely related to the ratio j;/,&, as shown in Fig. 3. trol the flow regime. Most of the existing flow-reg-
Thus, it is expected that the actual values of j;/.fi ime maps reveal that the flow regime can be accu-
for stratified smooth flow are, in general. < I .O, rately predicted by the superficial gas and liquid
velocities. In this study, the values of f,/,fg ratio
required to satisfy the measured E, data increase
with increasing product of the superficial gas and
liquid velocities (V&V,,>. Therefore, it is expected
that an inverse relationship exist between the correc-
tion factor and ( V’gV,,). Fig. 4 is a plot of these data.
The solid line is the best fit, which can be repre-
sented by the following empirical equation:

c = 0.528( ~ev\,)P”~?10’2’ (19)

6. Calculation procedure

Based on the equations developed in this study,


liquid holdup can be estimated from the following
Fig. 3. Liquid holdup as a function of f; / ji. procedure:
G.H. Abdul-Majeed/ Journal of Petroleum Science and Engineering 15 (I9961 271-280 211

(1) Determine the parameter X from Eq. (2). standard deviation, SD. The definitions of these pa-
(2) Obtain the correction factor C from Eq. (19). rameters are:
(3) Determine the theoretical liquid holdup from
(estimated) - (measured)
Eq. (16) or Eq. (17). PE = x 100
(4) Using the obtained value of C, determine the (measured)
true liquid holdup from Eq. (18).
CPE,
APE=‘=
n
7. Evaluation

The validity of the proposed model was tested AAPE = i=l


using two sets of data: the present data (89 points) n
and Minami and Brill (1987) data (111 points). Most 2 I.5

of the existing correlations are included for compari-


n 2 (PEi)2 -
son purposes. The correlations considered are those i= 1
of Armand (1946), Hughmark and Pressburg (1961); SD =
n2
Hughmark (1962); Eaton and Brown (19671, Guzhov
et al. (19671, Beggs and Brill (1971), Gregory et al.
(1978), Brill et al. (1981), Chen and Spedding (1981,
1983), Mukherjee and Brill (19831, and Minami and
Brill (1987). Although the last two correlations were
developed for vertical flow, they are widely used for 8. Present data
horizontal flow. The statistical parameters used in
the evaluation are the average percent error, APE; As mentioned earlier, data obtained by the author
the absolute average percent error, AAPE; and the cover stratified, slug and annular flow regimes. Sta-

Table 2
Statistical results using present measured data
Prediction method APE AAPE SD
Eaton and Brown (1967) -40.83 44.56 29.14
Brill et al. (1981) - 50.96 82.76 160.87
Gregory et al. (1978) 494.86 494.86 656.15
Mukherjee and Brill(l983) - 34.58 41.55 36.82
Minami and Brill ( 1987) (method 1) 1.26 13.61 19.15
Minami and Brill(1987) (method 2) 4.60 14.84 17.39
Beggs and Brill ( 197 1) 7.06 25.58 31.76
Guahov et al. (I 967) 316.33 322.56 459.77
Hughmark and Pressburg (1961) 14.87 78.22 116.32
Hughmark ( 1962) 63.41 75.54 92.84
Taitel and Dukler (1976) 100.39 108.02 95.82
Armand (I 946) (slug data) 49.74 5 1.47 45.47
Chen and Spedding (1981) (annular data) 33.72 35.85 30.97
Chen and Spedding (1983) (annular data) 130.24 130.24 51.15

Proposed (all data) -0.38 10.30 13.30


Proposed (slug data) - 6.37 8.46 9.43
Proposed (annular data) 9.56 16.19 16.05
278 G. H. Abdul-M&d/ Journal of Petroleum Science and Eqineering I5 ( I9961 271-280

tistical results are listed in Table 2. This table reveals


several important points:
(1) The proposed modification causes a signifi- - Eq. 19

cant improvement in the performance of the Taital- .


* *
Dukler model. The proposed model yields statistical 2.0 -

results - 80% lower than those obtained by using


the original model.
(2) For the entire data, the proposed model shows
the best performance compared to the other correla-
tions considered.
(3) For slug flow, the modified model yields
results much better than those obtained by the Ar-
mand correlation. Also, for annular flow it gives
results superior to those obtained by using Chen-
Spedding correlations. Fig. 5. Relationship between correction factor and (V,BV,,)
(4) The correlations of Eaton and Brown (1967), Minami-Brill data.
Brill et al. (1981) and Mukherjee and Brill (1983)
underestimate liquid holdup, whereas all others over-
estimate the holdup data. of the proposed and existing correlations are listed in
Table 3. The model proposed by the author gives
results superior to those obtained by using Minami-
9. Minami-Brill data Brill correlations and all others considered. For slug
Bow, the modified model yields results much better
Minami and Brill (1987) data were obtained using than those obtained by using the Armand correlation.
air-water and air-kerosene two-phase mixtures. The Also, Table 3 shows that the correlations of Eaton
liquid holdup ranged from 0.009 to 0.45. Using and Brown and Mukherjee and Brill tend to underes-
Mandhane et al.‘s (1974) flow map, 5.5 points fell in timate holdup, whereas other correlations overesti-
the stratified smooth flow region, 26 in the stratified mate.
wavy flow, and 30 points in the slug flow. Minami The values of the correction factor for Minami-
and Brill used these data to develop their own corre- Brill data were calculated using Eq. (18) and plotted
lations (correlations 1 and 2). The statistical results versus (V&V,,). as shown in Fig. 5. Minami-Brill

Table 3
Statistical results using Minami-Brill data
Prediction method APE AAPE SD

Eaton and Brown (I 967) -48.10 4R.10 20.73


Brill et al. (198I) 34.00 s9.00 97.50
Gregory et al. (I 978) 722 722 860
Mukhejee and Brill(1983) - 35.70 39.10 3 I.35
Minami and Brill (1987) (method I) 4.35 14.6.5 19.35
Minami and Brill (1987) (method 2) -4.64 15.74 18.47
Beggs and Brill(1971) - 10.70 26.26 30.70
Guzhov ct al. (1967) 252. IO 254 434.60
Hughmark and Pressburg (1961) 2.60 72.70 IO9.40
Hughmark (1962) 51.10 66.40 IO2.70
Taitel and Dukler (1976) 40.56 56.X4 59.39

Proposed I .46 13.93 17.00


G.H. Abdul-Majeed/Joournal of Petroleum Science and Engineering 15 (19%) 271-280 219

data show the same trend as those shown by the Subscripts:


present data. Also, there is a good agreement be-
g gas phase
tween Minami-Brill data and values obtained by
interface
using Eq. (19).
; liquid phase
S superficial
10. Conclusions

(1) It has been shown that the implicit model Acknowledgements


developed by Taitel and Dukler can be accurately
represented by a single explicit equation. The author is grateful to Dr. James P. Brill of
(2) Based on experimental data, the Taitel and Tulsa University for his suggestions and help.
Dukler model tends to underestimate the holdup for
stratified smooth flow and tends to overestimate the
holdup for stratified wavy, slug and annular flows. References
(3) A new empirical modification is proposed
Andritsos, N. and Hanratty, T.J., 1987. Influence of interfacial
which caused a significant improvement in the per- waves in stratified gas liquid flows. AIChE J. (Mar.), 33:
formance of the original model. 44-453.
(4) The proposed model contains no complex Armand, A.A., 1946. The resistance during the movement of a
dimensionless expressions and is, therefore, easier to two-phase system in horizontal pipes. Izv. VT1 1, AERE
Trans., 13: 16-23.
apply than the original model.
Baker, A., Nielsen, K. and Gabb, A., 1988. Pressure loss, liquid
(5) The proposed model is found to be applicable holdup calculations developed. Oil Gas J. (Mar.), 11: 55-59.
to stratified (smooth and wavy), slug and annular Beggs, H.D. and Brill, J.P., 1971. A study of two-phase flow in
flow regimes. inclined pipes. J. Pet. Technol. (May), 4: 607-617.
(6) The proposed model shows excellent results Brill, J.P., Schmidt, Z.R., Coberly, W.A., Herring, J.D. and
Moore, D.W., 1981. Analysis of two-phase tests in large
and clearly outperforms all the existing correlations,
diameter flow lines in Prudhoe Bay field. Sot. Pet. Eng. J.
when tested against the present measured data and (June), 3: 363-378.
against data from the literature. Chen, J.J. and Spedding, P.C., 1981. An extension of the Lock-
hart-Martinelli theory of two-phase pressure drop and holdup.
Int. J. Multiphase Flow, 6: 659-675.
11. Notation Chen, J.J. and Spedding, P.C., 1983. An analysis of holdup in
horizontal two-phase gas-liquid flow. Int. J. Multiphase Flow,
2: 147-159.
A area occupied by fluid, m* (ft*) Crowley, C.J. and Rothe, P.H., 1986. State of the art. Report on
C liquid holdup correction factor Multiphase Methods for Gas and Oil Pipelines. Vol. 3, Theo-
D hydraulic diameter, m (ft) retical Supplement, prepared for Project PR-172-609 of
liquid holdup, dimensionless Pipeline Research Committee, A.G.A. (Dec.), 44 pp.

F
h
friction factor
equilibrium elevation, m (ft)
Eaton, B.A. and Brown, K.E., 1967. The prediction of flow
patterns liquid holdup and pressure losses occurring during
continuous two-phase flow in horizontal pipelines. J. Pet.
s wetted periphery, m (ft) Technol. (June), 12: 815-828.
V velocity, m s- ’ (ft s- ’ > Gregory, G.A., Nicholson, M.K. and Aziz, K., 1978. Correlation
X Lockhart-Martinelli parameter, dimension- of the liquid volume fraction in the slug for horizontal gas
liquid slug flow. Int. J. Multiphase Flow, 4: 33-39.
less
Guzhov, A.I., Mamayev, V.A. and Odishariya, G.E., 1967. A
study of transportation in gas liquid systems. 10th Int. Gas
Greek symbols: Congr., Hamburg, 8 pp.
gas viscosity, Pa s (cP) Hughmark, G.A., 1962. Holdup in gas liquid flow. Chem. Eng.
Prog., 58: 62-64.
oil viscosity, Pa s (cP) Hughmark, G.A. and Pressburg, B.S., 1961. Holdup and pressure
gas density, kg rnp3 (lbm ftC3) drop with gas liquid flow in a vertical pipe. AIChE J. (Dec.),
liquid density, kg rnp3 (lbm ftC3) 7: 677-680.
280 G.H. Ahdul-Majeed/ Journal qfPetroleum Science and Engineering 15 (19961 271-280

Lockhart, R.W. and Martinelli, R.C., 1949. Proposed correlation Taitel, V. and Dukler, A.E., 1976. A theoretical approach to the
of data for isothermal two-phase, two-component flow in Lockhart-Martinelli correlation for stratified flow. Int. J. Mul-
pipes. Chem. Eng. Prog. (Jan.), 45: 39-43. tiphase Flow, 2: 591-595.
Lopes, J.C.B. and Dukler, A.E., 1986. Droplet entrainment in Tronconi, E.. 1990. Prediction of slug frequency in horizontal
vertical annular flow and its contribution to momentum trans- two-phase slug flow. AIChE J. (May), 7: 701-709.
fer. AIChE J. (Sept.), 6: 1500-1515. Wallis. G.B.. 1969. One Dimensional Two Phase Flow. McGraw-
Mandhane, J.M., Gregory, G.A. and Azir. K.. 1974. A flow Hill. New York. NY, 385 pp.
pattern map for gas liquid flow in horizontal pipes. Int. J. Xiao. J.J.. Shoham, 0. and Brill, J.P.. 1990. A comprehensive
Multiphase Flow (Oct.), I : 537-553. mechanistic model for two-phase flow in pipelines. 65th Annu.
Minami. K. and Brill, J.P., 1987. Liquid holdup in wet gas Tech. Conf., Sot. Pet. Eng., New Orleans. LA, Pap. SPE
pipelines. Sot. Pet. Eng. Prod. Eng. (Feb.), 5: 36-44. 20631. 14 pp.
Mukherjee, H. and Brill, J.P.. 1983. Liquid holdup correlations for
inclined two-phase flow. J. Pet. Technol. (May), 4: 1003-1008.

View publication stats

You might also like