Multiphase Flow and Heat Transfer

Proc. of the Fourth International Symposium'

Aug; 22-24, 1999, Xi'an China

SLUG FREQUENCY AND LENGTH INCLINED LARGE

DIAMETER MULTIPHASE PIPELINE

J.Y:CAI, R W. WANG, T.HONG and W.P.JEPSON

NSF IlUCRe, INSTITUTE FOR CORROSION AND MULTI PHASE TECHNOLOGY
.omo UNIVERSITY
.
340 Yz W.STATE ST., ATHENS, OH 45701
U.S.A

E-MaD: calra:bobcat.~nt.ohiou.edu
Fax: 001-740-593-9949

ABSTRACT
ExperimentaIinvestigationson
slug frequenc:y and slug length are carried out in 10-em ID., 18-:mlong
pipe at incIIDa:tions offlo at atmosphere pressure .with. tap water ~d carl>ondioxide ~ working fluids ..
The superficial liquid velocity ranged from 0.1 to 1.7 m/swith supetficialgas velocity between 0.5 mls to
15 mls. -New results for slug frequenc:y arid slug unit length are obtained. The effects. of superficial liquid'
velocity, superficial gas velocitj andiilclination on slug frequenc:y and slug unit length are considered .

.. ~:.-.

.-From theexperimentaIresults,
it is shown that slug frequency increases With increasing liquid flow rate
for all test conditions. Interesting results are Seen with increasing the gas flow rate. In upwardflow,at
lower liquid flow rate, slug frequency decreases with increasing gas flow rate atlower superfiCial gas
velocity and then increases with an increase of gas flow rate. Slug frequenc:y is almost independent ot
superficial gas velocity at moderate liquid flow rate. However, slug frequenc:y increases With increasing
gas velocitY at high superficial liquid' velocity. In downward flow, slrigfrequenc:y is almost independent'
o~ superficial gas\relocity at mOderate liquid flow rate. However, at higher superficial liquid velocity,
, slug frequency decreases with increasing super:ficiatgas velocity at lower superficial gas velocity and
then increases with increasing gas velocity.
.'
..
Siug unit length decieaseswith incteasingliquid flow rate at same gas flow rate. However; at same liquid .
flow rate, slug unit length will increase with increasing gas flow rate. Inclination has significant effect on
slug unit length. Increasing the inclination .leadsto a decrease of slug unit length.
.
.,'

'.A new tnodelis used to predict the slug frequency; good agreement is achieved with experimental results.
,

.t

1. INTRODUCTION
The flow ofliquid and gas is bei~g eilco~tered mor~ and more in the offshore production of oil and gas.
Although there is a large number of experimental and theoretical results on gas-liquid flow, known results
for gas-liquid milltiphase flowinlargediameter
pipelines are small. In particular, the region of slug flow
is of great interest. in the offshore production of hydrocarbons and also in some other areas, such as in the
power plant and chemical industry.
.

195

The firStcomprehensive experi~entadlinVesti?ationsI"

Don3S15Ugl::q~~~~:~erp~pe'
~rf:;:n~~.
h"
. ti ted the flow of at! an water man"
" . "
"easured

:~b~~ ~~
1

ptcsSt
into tl
is m
meter
.then
appaI1
the n
gas-Iii
at the
de-em
sepan

;~u:';
or liquid s!u? 'mum"
creas~
o;~:r g:l;~~. "~e:
results~:C~
as~~~o:~~~~o~;.~~~~al .
frequencies have a numm
.
.
..
uk! [3] G
confirmed by the experimental investigations of Gregory & Scott [2] and T~tel &
er
. regory
and Scott [2] proposed a model for predicting the slug frequency based on therr exnernnental results.

'v

.36m2 / S2

f. ~O.0157g~(

V,

1.2

+V,

..

J]

(1.1)

is ven
In equation (1.1), Vs1 is superficial liquid velocity, VI is trans!ational velocity and d is the diameter of
pipeline.

liquid
is inc
upwaJ
of the

Tronconi [4] considered the slug frequency based on his experimental resul~. He.a.s~med that the waves
on a liquid surface would grow according tb the Kelvin-Helmholtz mstabilities, but oruy waves
characterized by a critical growth rate cause the formation of a stable liquid slug. Based on the theory of
Mishima & Ishii [5], Tronconi [4] obtained a correlation for calculating the slug frequency:

The t1
meast
Inthi
. supen
0.1-

(1.2)

atmos
analy.
3. EX

In Equation (1.2), ha is the height of the gas phase layer in the.stratified flow, and Va is the average gas
velocity within the gas layer cross section of the pipe. Po and Pi denote the densities of gas phase and
liquid film The proportionality factor C". in equation (1) is equal to 2.

3.1R

Fig. 2
gas v.
veloci
at the

Hill and Wood [6] also created a commonly used slug frequency model based on their experimental
results. Their model was based upon the equilibrium film height
..

Is = 0.275 Vm 10

h \,
( 2.68.-1
d)

(1.3) .

In Equation (1.3), Vm is the velocity of mixture, d is diameter of pipeline and also h is film height.
'yv'~ens [7] investigated the slug frequencies with carbon dioxide and sea water in large diameter
mclined pipelines. The aim of this investigation is to develop a model to predict the slug frequency for
~arg~ ~ameter pipeline and also to determine the effects of gas and liquid velocities and pipeline
mclmatlOnon slug frequency.'
.
.

2.EXPE~NTALSETUP'
Fig. 1 shows the experimental setup of the flow loops. The test section is a lO-em J.D. pipe made of
transparent glass. The volume of the tank is 1.44 m3 which is filled with tap water. The fluid is pumped
through a 7.62~m J.D. PVC pipe and is metered using an orifice plate; Carbon dioxide gas from high196

frequ

I],

tle
ilg

re
ry

pressure stornge tanks is introduced

into the system and the gas velocity
is measured using in-line flow
meters. The multi phase mixture
then flows through a lO-cm I.D.
apparent pipe test section where all
the measurements are made. The
1. npantor
gas-liquid mixture reenters the tank
2.d nt~ln",.nt plah
at the top of through the pipeline. A
uparalor
. no. 00'.
5."''''
de-entertainment plate is used to
t .mp.ratu O'.uge
l.oantriNoa' pump
separnte the gas and liquid. The gas
Fig. 1 Flow Loop
is vented to the atmosph~re and the
liquid is recycled. When the system
is inclined, measurements in both
upward and downward flows can be made at the same time. A back pressure regulator is fitted on the top
of the tank and is connected to the exhaust to control and maintain the required system pressure.
3.CII11I'1ng

tank

<t,

10 .

of
es

, otlon

es
()f

The flow patterns were determined with a special technique using differential pressure fluctuations. The
measurements were made with 0 - 35 KPa OMEGA PX-750 heavy-duty differential pressure trnnsducers.
In this study, carbon dioxide was used for gas phase. Tap water was used for the liquid phase. The
superficial gas velocity was varied from 0.5 - 15 mls. while the superficial liquid velocity ranged from
0.1 -1.7 mls; The system temperature remained constant.at 25C, and the system pressure was kept at
atmosphere pressure. Two tape recorders were used to record all the flow phenomena, which were used to
analyze the results later.
3. EXPERIMENTAL

RESULTS

15

ld

3.1 Results of Slug Frequency

Fig. 2 shows the results of slug frequency at different superficial liquid velocities at the same superficial
gas velocity in upward flow. It is seen that slug frequency increases with increasing superficial liquid
velocities for all conditions tested. The results of slug frequencies at different superficial liquid velocities
at the same superficial gas velocity in do\\nward flow are shown in Fig. 3. It is also seen that slug

1.00
~

0.80

i.,

0.60

e
~
fij
loo.

:r
Ir

.e

1.00

L
~

-!!

V'g=4m1.

t
~
"

0.40

0.40

0.20

r;;

0.20

&
&

&

& &
&

&

0.00
0.00

0.00
0.50
Superlldal

"'Ig. 1 Results

1.50

1.00
L1quldVelodty

or slug freqency

liquid velcx:ity at Vsr4

0.00

2.00

0.50

1.00

Superlklall.lquld

mI,

Flg. 3 Results

along superfidal

of

5 lug

freqen~y

1.50
Velocity
along

2.00

mi.
superfklal

liquid vdodty at V'I:m/s. -1 dtEre

mls. +2 degree

If
d
1-

mi. i

0.60

"'"
~

.."

V.g=4

1&

0.80

frequency increases with increasing superficial liquid velocities.

197

.,

j.

.
'.
ult. fi r slug frequency. At lower
Increasing superficial .gas velocity generates very i:~e~~~~s
th: su~erfiCial gas velocity at lower
superficial liquid velocity, slug fr~quency de~~es
Wi 'g m superficial gas velocity in upward flow (See
. superficial gas velocity and then mcreases Wi mcreasm
..
.
t oderateand
Fi . 4). Fig.5 shows the distribution of slug freq~e~cy along superfiCial gas v~locity a.m
ndent of
hi~ superficial liquid velocities. From this picture, it is seen that slug frequency. is almost ~:~creasing
gas velocity at moderate liquid ve~ocity o~ O.9m/s. However, slug frequency mcr;eases Wi
.
. gas velocity athigh superficialliqmd velOCity.
.
.
0.30

;;

.. ..

I
~
~
~

0.20

""

+
A

_0
0

..

+ 0+

0.10

'- "

1.50

.....

1.20

0
0

t.

o.

0.'

O~

..~

t::

0.5

0
0

0
00

~
0.1 rnls

;;:;

"

.... + +

0.90
0
6 .

0.60

..

..

0
6

00

.6

.6

0.30
0.00

0.000.00

2.50

5.00

7.50

10.00

12.50

or slug rreqency

ak'Jng superllcbl

Jowsuperficialliquid

SuperftcJal

I
I.

1
I'

Fl2. S Promeor~ug

gas velocity at

at bigh

velocitieSt +2 degree

12

15.00

Sup~rndaJ Cas Vdodty m/s

Fig. 4 Profile

fnqmq

5 uperficial.liquid

Gas

Y doclty

along

supalldal

~eJo~iti1s

m/s
gasvdodty
+2 degree

Fig.6 shows the profile of slug frequency along superficial gas velocity at different superficial. liquid
velocities in downward flow. It is seen that slug frequency is almost independent of superfiCial gas
velocity at moderate liquid flow rate. However; at higher superficial liquid velocity, slug frequency
decreases with increasing superficial gaS velocity atlower superficial gas velocity and then increases with
increasing gas velocity.

1.50

;;

j.

.::

0.90
0.60

DO

1.20

0.30

eo
-+

0
6" 6
+ .,.

+
0

~"

+
+

O.5m1s

+.

1.7

..

00
6,,"

0.00
.9

Superfk:tal Gas Vdoclty

Fig. ~ ProrDe of slug rreqency
at dJ1ferent super1ldal

.higher 1
unit fen
Interest
superfi<
increasi
velocit)
length.
liquid,
Sliperfi(
.increas,
ratio of
, veloci~
slug wi
decrea~
of slug

'-

remains
fromth
superfi~
heighti

12

15

m/s

~uper1k:1a1gas'Veloc~

along superOclal gas vekJctly

Fig. 7

UquJ~ velodUes, -2 degree

Etfct of'lndlnatlon

mls

on slug 1ftqCIICJat supatkia

liquid veloc ity of 1.3 m/s

. The in
.inClinal
same s

frequer
. mass.c1,

Fig. 7 shows the effects of inclination of test section on slug frequency. From this picture, it is seen that
inclination of test section has significant eff~ct on slug frequency. Slug frequency increases with
increasing the inclination of test section.
3.2 Results of Slug Unit Length
In all cases tested in this paper, slug unit length was found to decrease with increasing superficial liquid
velocity. Fig. 8 shows the distribution of slug unit length along superficial liquid velocity at superficial
gas velocity of 5 m/s in +2 degree upflow. Fig. 9 shows the results of slug unit lengt.l) along superficial
liquid velocity at superficial gas velocity of 5 mls in -2 degree downflow. Usually, the slug length
198

'
T
,

*~

'..

..~.
.-,'

Trl!
~,-

,
1 ti ely constant at about 15 to 20 times the pipe diameter. In a slug unit, mass is conserved
remam~ r~ a ~.
and liquid flow rates, If the superficial liquid velocity is increased at a constant
from
~almpu ~l 'ty the ratio of liquid to gas in the slug unit must be increased, Although the film
SUperfiCI gas ve OCI ,
, f l"d
. till
height in film region is increased with increasing the liquid flow rate, the ratio 0 lqUl to gas IS s

t lower

It lower
IW

'

(See

ate and
Ident of
:reasing.

40

Ir

"..~
.:l

~
;"
;;
CI

30

..

'

il

v "".

10

..~

20

S
;;
c

;>

"ii
i"

'0

i
!

o '
OAO

0.80

Superficial

J.20

L60

Liquid Velocity

0.40

0.00

2.00

Fig. 9 Res

L60

L20

0.80

Superficial

mls

FIg. 8 Results of slug uolt length along snperflclalllquid

velodtlesatVsg=S mJo,+2 degreeupOow

12

20 ,

o I"
0.00

~~I

..l

..

30

2.00

Liquid "!.'eloclty II1Is

of s lug unit leneth. along s uperficialliquid

liltS

veloctUes

atVsg=5

mis, -2 degree

dowoftow

higher than that in the :film region, Since the slug length remains relatively constant, a decrease of slug
unit length corresponds to a higher ratio of liquid to gas,
Interesting results were obtained for slug unit length by increasing the superficial gas velocity at constant
superficial liquid velocity, Atlo"er
superficial liquid velocity, the unit length increases greatly with
increasing superficial gas velocity. Fig. 10 shows the results of slug unit length along superficial gas
velocity at different superficial liquid velocities. At moderate superficial liquid velocity, the slug unit
length increases slightly to a maximum with increasing superficial gas velocity, At higher superficial
liquid velocity, the slug unit length increases slightly to a maximum and then decreases with increasing
superficial gas velocity. These results are expected. The ratio of gas to liquid in the slug unit will be
increased with increasing superficial gas velocity at constant superficial liquid velocity. Usually, higher
ratio of gas to liquid leads to larger mixing zone for same slug unit length. So. increasing superficial gas
velocity, the unit length will be increased. At higher gas and liquid velocities, the void fraction in the
slug will be much greater. This is due to higher turbulence leads. To some extent, the mixing zone will be
decreased, however the length of film region will be remained relatively constant. In this case, the length
of slug unit will be decreased.

liquid
ial gas
quency
es with

The inclination of pipeline has a slight effect on the slug unit length. Fig. 11 shows the effects of
inclination on the slug lengtlL Increasing the inclination causes the slug unit length to be decreased at the
same superficial liquid and gas velocities. This is due to the fact that the slugs are shorter and more
frequent in upward flow than that in downward flow at the same gas and liquid velocities. Based on the
mass conservation in the slug unit, shorter slug leads to shorter film region.
:n that
; with

100

'I

-,

liquid
:rficial
:rficial
length

..

10 ~

;>

~
;;;
1

y
0

;;
c

"
~

..:I

..

0000

.3

'E

.......

"

e
-=

lO~~

Vi

5
10

20

:ddllTermt~"lquIdvdodUes.+2

..........

! ......
j

Super1'ldal Cas vclDd!y mill.

F1a:.10 Results orslq"unIt It~slong

15

1.7

~11

::;

::-'-1

~:::::

:a

..

:::::n-' ~ I

Ir

20

Superflclal

superlldaJzws velodty

FIg. 1]

degRe upn01'r

Errs{ or lndDa aon

Oq

12
m/s

~ue unit Imgtb at W-O.9

and dift"ert:'ntsuperficiaJ

199

Gas velocity

gas velocities

mIll.

4. SLUG FREQUENCY

MODELING

The Hill & Wood' model[6] includes the film height asa parameter .. It is difficult to measure this
accurately. It is also very hard to detennine the shape of gas~liquid interface. Same problem can also be
found in Tronconi's model[4]. Also, Tronconi's model [4] doesn't include the effect of pipe diameter on
slug frequency. From the Gregory & Scott' model [2], it is seen that there is only one parameter VI. which
is sensitive to the results. In this case, how to calculate the value of VI for large diameter pipeline is the
key to set up the slug frequency model. This model doesn't reflect the effect of inclination on slug
frequency. In this paper, a new model. which reflects the effects of diameter and inclination on slug
frequency, will be proposed based on the eX1>erimentalresults and Gregory & Scott' model.
A common way to consider the translational velocity is as a ratio with superficial mixture velol.oity. At
low gas flow rate, Kouba & Jepson [8] and Jepson & Taylor [9] pointed out that the value of V; IV", was
about two based on their experimental results in large diameter pipeline. At higher gas flow rate, tIns ratio
drops to a steady value of around 1.2 to 1.3. Also Kouba and Jepson [8] specified V/IV",= 1.25 for mixture
velocities above 3 mls. Jepson and Taylor [9] showed the ratio tapering to 1.25 as the superficial gas
velocity was increased. The new model can be expressed as follows:

.( = K

[V d (36mV
2

S1

{}

g.

. /

S2 + V J]1.2

bigt
mcr
Ani
achi

(4.1)

. ~/

In equation (4.1), K{} is the function of inclination.. which reflects the effect of inclination
frequency. Based on experimental results. K{}can be calculated as follows.

K{}

I
I

It is
resu
deCI
iner,
vela
.vela
Iiqu
gas
supe
len~
slug

= 0.018 * exp(sin

B)

on slug

(4.2)

In Equation (4.2), Vlcan be calculated with the following equation:

(4.3)

Fig. 12 and Fig. 13 show the comparison between the experimental

down~ard flows. It is seen that the experimental results have a good
and Fig. 15 show the comparison between the experimental results
and Gregory & Scott's model. It is clear that the Gregory & Scott's
at low and high gas flow rates.

and predicted results in upward and

agreement with the predicted. Fig.l4
and the predicted by the new model .
model under predicts slug frequency

5. CONCLUSION

1.

In this paper, experimental investigations on slug frequency and film height are carried out in lO-cm I.D.,
18~m long pipe at inclinations of:t2 at atmosphere pressure with water and carbon dioxide as working
flwds. New results of slug frequency and slug unit length in large diameter pipeline were obtained.

2.

200

~t

It is seen that slug fr:equencyincreaSes with increasing liquid flow rate for all test conditions. Interesting
results are seen with increasing gas flow rate. In upward flow, at lower liquid flow rate, slug frequency
decreases with increaSing gas flow rate at lower superlicial gas velocity and then it increases with an
increase of, gas flow rate. Slug' frequency is almost independent of gas velocity at nioderate IlqUid
velocity. However, slug frequency increases with increasing gas velocitY at high superficial liquid
velocity. In downward flow, slug frequency is almost independent of superficial gas velocity at moderate
liquid flow rate. At higher superficial liquid velocity, slug frequency decreases with increasing superficial
gas velocity at lower superficial gas velocity and then increases with increasing gas velocity. At the same
superficial gas velocity, slug unit length decreases with increaSing superficiaJliquid velocity. Slug unit
length increases with increasing superlieial gas velocity at'lower liquid flow rate. At' moderate flow rate,
slug unit length' increases slightly t" a maXimum with increasing superficial gas velocity. However, at
higher liquid flow rate, slug unit length increases slightly to a maximum 'and then decreases slightly with
increasing gas flow rate. Increasing the inclination leads to the decrease of slug unit length.

as
io
re

A new model is propOsed to predict the slug frequency for large diameter pipeline, and good agreement is
achieved with experimental results.
'
'

I
~S

:Ie
:)0

;h
le
Ig
Ig

IS

...........

2.00,
A

1.60

!
:t

;s

This . d"
A
'Vd O.5.w
----.T.is:_d .

~ '1.60

TbaIDOdcl

._

130

0.80

.'.--e-~".':

0.40

1~

.+

:
" '

:-

Fie,: i2 Comparison

betwcea

model at drTereot

T.Is.-od4
'1.7

+ . +
+__ ----a

"

,- l

Cas VcIodly

iz

JDls

experimental

Iquld W;:1oclie:s,

'0.40

46.

9
Supcrlldal

-G.,

+'
-

1.20

;:; o.~o'

--+- __-.:l. __

66.-66

%.00

t-

..

Vlll-<l.5mh"

-.-;._: ~~.

ni.13

results and this

Comparison

+% upflow

mod

1.00

between

1.50

".1.7rn1.

results

aDd pa15

-1 dewwno...

aad Crecory

of expermental'resuKs

fJrq;ory a:Sl;OlI_dd

Vill-l.7

o~

0.,

wils

...............

O.~O

0.00

11

Sllper(ic~1eas Velocity

This_dd

~1.OO

....-....-~.......... - ...

t . Comparison

-'-

I.SO

~
..."

~o~

:l

vtloda~

DIls

1.00

Fie.

1%

cxpcrlmeo1a1

td;~"I!n!l{"quld

~
;:;

SuperacJal Cas Velocity

mls

Super ..
cial gas Veloti1;y

w,lh litis model

FIc.

t~ CompabJD
aad

a ScoU's Blodd at +1 decr~e upftow

~olY

11
m/s

ota:palmmtalmulbl'llJl
tIIlsmodcl
d~aadoWDt1ow

a' SmU'smodd at-2

l
I
V

REFERENCES
,I.

Hubbard, M, An analysis of horizontal gas-liquid slug flow, Ph.D thesis, University of Houston,
TX., 1965; ,
-

2.

Gregory, G. & Scott, D.S., Correlation of liquid slug velocity and frequency in horizontal co-eurrent
slug flow, AICHEJL VoLl5, pp.933-935, 1969.

201

Taitel, y. & Dukler, A.E.,A mo.del fo.r slug frequency during gas .liquid flo.Win horizo.ntal and nearho.rizontal pipes,lnt.J. Multiphase Flow Vo.l.3, pp.585-596, 1977.
4: Tro.nconi, E., Predictio.n o.f slug frequency in ho.rizontal tWO.phase slug flo.w; .AlCHE JL Vo.1.36,
pp, 70 1-709, 1990.
. '..
'.
. '.'
'.
5~. Mishima, K & Ishii, .M., Therotical predictio.n. o.f o.nset o.f ho.rizo.ntal ~lug flo.w, J.Fluids Engng
Vo.l.l02,pp,441-445,1980..
. ...
. .
6. Hill; TJ.and
D~G. WOod, A new approach to. the predictio.n o.f slug frequency, SPEAnnual"
TeehnicalConf, pp~141-149, 1990~
. 7. Wilkens; RI, Prediction 0.1' the flow regime transitions in high pressure large diameter inclined
multiphase pipelines, Ph.D thesis,OhioUIiiversity,
OR, 1997.
'.
.
. ..'
8. Kouba, G.E. and Jepso.n, W.P., The flo.W o.f slugs in ho.rizo.ntal tWo.phase pipelines, Fran. ASME,
Vo.l.l12,pp.l0-24,1990.
.
.
.....
....
.,'..
9. . Jepso.n,. W.P. and Taylo.r; RE., Slug flo.W and its transitio.nin large diameter ho.rizo.ntal pipes,
Int.J,,'\lultiphase Flow, Vo1.l9, pp41l-420, 1993 .. '
".
3.

NOMENCLATuRE
C
d

coefficient
diameter o.fpipeline, m
f
frequency, S.I'
g
acceleration of gravity, m/s2
hheight,
m
V
velocity, m/s
()
inclinatio.n o.fpipe section, degree
p
DenSity, KWm3

Subscripts.
G
'm

s
sg .
sl
t

gas phase
mixture
slug flo.w
superficial gas velo.city
superficial liquid velo.city
.translatio.nal

202

Ii

I,