Académique Documents
Professionnel Documents
Culture Documents
Comprehensive Two-Phase
Pakistan U. of Tulsa Petroleum Ltd.; N .D.
Mechanistic Flow
Sylvester, U. of Akrow
Model Wellbores
and C. Sarioa, O.
Shoham,
for
Upward
A.M. J.P. Anssri, Brill,
in
and
s
Summary. a model pressure cases for dmp covering A comprehemive prediction slug, variety model. performed and of model is and annul= field Overall the best. data. formulated a set flow. Model model of to predict the flow behavior models model is is in also good for for upward predicting by six using commonly the data. In two-phase such a well flow. flow data used
%
This
40630
model is composed as up of holdup 1,712 and methods, of and well the the
mechanistic
compared agreement
correlations other
Hasan-Kabr comprehensive
comparison
Introduction Two-ph.a.se cal, and flow nuclear of, is commonly indushies. understanding, encountered This frequent in the PeVO1eum, presents two-phase VSg=0.25V, of the complex through modeling approach Vwious Separate is nature of empirical approach, the two-phase methods. Tbe flow, The fundamental of been developed holdup resulting and flow developed for each patterns to the problem trend has postuor preflow drop. can be Dispersed breakkwge and Hasan modeling to develop and Kab@ of published flow that studies in vertical describe the on exceeding Bubble gas 0.25. ltansition. bubbles This down yields the Athigh into small liquid ones, rates, even turbulent at void bubble forces fractions flows: This is shown gc7L(pL-pG) v, = 1.53 where v. is the +0.333USL, slip or bubble-rise . . . velocity . given by . (2) chefithe of about 0.25. Using of this vatuc of void and fraction, slip velocities we can express tbe
transition
in terms
superficial
challenge systems. Because was shifted late flow dict pattern By of first
analyzing,
%
A in
the
modeling
theories mcdels
[1
---r as Transition Fig. 2. transition
(3)
pressure models
condkions
other
than
those.
development. et al.]
to dispersed
mechanistic work is needed more of this model fmt predicts vtiables the range U, of and one Fluid the mcdel mechanistic predicted of study for
two-phase models
rigorously. is to upw%d the by flow existing taking pattern. attd formulate two-phase flow into The field a detailed flow. patmn account model data weU with in the six field. sition for no-slip d%persed bubble flow as (5) is available data bank. comprehenThe and then the comprecalcuactu.at Thii At mum concluded The coris shown gas as = 0.725 + 4.15 0,5 vs. + SL B in this to Fig. 2. is governed Scott of 0.76, by and giving the maxiKouba7 the tram
()
.% bubbles occurs
. . . . . . . . . . . . . . . . . . ...(4)
evaluated in the
Transition
high packing
velocities, of
transition give
mdcscence. fraction
empirical
that
this
at a void
model
%E=3J7%L. FlowPattern Taitd et al? Prediction presented Wmsitiom flow and et of the prediction into can one be slug, patterns evaluated al.4 model later the basic work on mechanistic flow. slug, transition the flows. chum, and modeling Theyidentiled annular among to then extend combined inclination works, !J~g = 3.1 flow) them the gives apphcabdity flow-pattern angle flow among ranges pattern bubble, to inclined models unified predicted and annular model. by Bamea5 to on different these transition different applicable Based detining flows. shown Bubble/Slug at which bubble Emsition.Taitel flow occurs as. et a[.3 gave the minimum diameter of fdm stability at dtin= 19.01 low Bamea5 film as the transition of This is shown as
. . . . . . . . . . . . . . . . . . . . . . . . . . . Tramitio C in Fig. 2.
. .
Transition
flow trained is based tiquid
to
Anmdar on the
criterion to tbe
for prevent
boundaries transitions
droplets
gas
stream.
modified
%
(6) in Fig. same transition. at high which bridging required to 2. transition One liquid causes by effect rates. considering is that The other flow is liquid governed slug . . . . . ...(7) by mechanism Lockhart the liqof the a thick effect of the by effects liquid is inthe
boundaxks
[1
the liquid rates.
guL(pLpG) P:
Transition modified
thickness
on gas
downw%d
film the
mechanism forma
[1
than this, tie coalescence was found of of Petr&a.m
. . . . . . . . . . . . . . . . . . . . . . .
(1)
liquid .12,
holdup
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . fraction no can be of pipe cress in in and section the core. of the occupied The
For to
pipes slug
basic small
transition gas
for large
bubble Taylor fraction where uid film is the assuming entrainment expressed Q
bubbles.
experimentally
at a void
instability Martinelii
terms YM,
modified
C-2pyriaht
1994 SC&W
Engineers
and
parameters,
SPE
Production&
Facilities,
May
1994
143
tt
B:::&E _sLUG FLOW
Fig. lFlow patterns in upward
t
CHURN
. . . . . . . . . ,.. . D
To account for the
effect as
of the
liquid
enmainment
in
the
gas
core,
Eq.
7 is
modified
here
()
,zfw+aLc*
>0.12
. . . . . . . . . . . . . . . . . . . . . . . ..
(12)
Annular by
flow
exists
if vsg
is greater criteria be
than am
that
at tbe
givthe Hm flow
en
Eq.
6 and
if the Eq.
two
Bamea first
Bamea
criteria,
8 must Eq. be
solved
is then exists.
calculated Eq.
from
annular a se.mnd-ordm as
8 cm
usually
Newton-Rapbso
approach.
EqT8
expressed
F&)
YM-
2-15H~ WJI1.5HH)
FM
. . . . . . . . . . . . . . . . ..
(13)
and
. . .
1 .5HLWM
,.
IV&)
. ... .
. . .
wJ1-1.5ffu)
. .
(2-1.5HM)VwH~(~5.5HW)
. . . . . ..
(14)
Wm(l-1.5HW)2
FLOW
fwo-phase
ANNULAR FLow
flow.
The
minimum
dimensionless
film
thickness
is
then
determined
it-
eratively
from
) . . . . . (15)
=0.25.
FIow-Sehavior After physical resulted Chum is treated patterns tbe flow models in flow as ze
Prediction patterns for the me flow predicted, behavior for modeled The the in each bubble, bemuse models next flow step is to develop This step flow. and flow
slug, of developed
its
slug
other
discussed
bubble The
flow bubble
model flow in
is had and
on dispersed the
regimes bubble
separately
developing
Because and no
of
gas
bubbles
in bubble
the flow
liquid can
0.01
approximated two-phase
a pseudmingle can
simptificmim,
pammetm
expressed
GAS
(M/S)
PIP=PLh
Fig.
+P&aiL.......
. . . . . . . . . . . . . . . . . ..(l6)
. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. (17) (18)
wellbores.
/%=#LaL+##-&).
YM
wJ1-1.5ffm)
2-15H~
x~,
. . . . . . . . . . . . . . . . . . . . . . . . .
(8)
andvrP=v~v~L+v~g,
where
).L=vJvm.
. .
. .
. . . . . . . . . . . ..
(19)
where
XH
J3SL
~ . . . . . . . . . . . . . ...(9) Sc the ing
For
bubble
flow,
the
by
taking velocity.
into By rising we
bubble-rise a turbulent
with pipe
the waif,
.[
~w = g sin O(p=-pd ,. dp () and
slippage
. . . . .
. . . . . . . . .
(10)
To
(20) 3).
~c
Fem of minimum geometric dimensionless considerations, film ffLF thickness, can ~ti, be tied
effect
swarm,
mod~.
B=(lFE)2(fr&SL). in term
expression
expressed as
~~ Hm=@tin(l+tin). . . . . . . . . . . . . . . . . . . . . . . . . .. (11)
153
[1
P;
WAPL-%)
%
H;,
. . . . . . . . . . . . . . . . . . . . . (21)
144
(a)
DEVELOPED
SLUG Fig.
UNIT
S-Schematic of slug flow.
(b)
DEVELOPING
SLUG
UNIT
where study,
the
value
from
one the
study
to another. Thus,
b Eq.
the 20
present yields
of for of
this slug
model. void
Tbe fraction.
basic
was an
the
use
of
?/=0.5
%
@5
to give
best
results.
fully the
developed concept of
introduced
= k_l,2yM, . . . . . . . . . . . . (22
developing of developed
1.53
[1
gives The
g%(!%k)
flow-pattern P; L I-HL ometty, rately. an implicit flow equation Patzmeters for the can actual now holdup be cafctdated for bubble tlom For mass a fuly balances the
Because fully
in flow flow
This flow.
developed give
slug
unit
(Fig.
3a),
the
overall
gas
andfiqtdd
two-phase
P=P=P.H. md#=p=pLH+j@-ffJ.
+P,(l-HJ
. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . ..
S$
k%s(%ts) -~VL#LrB, . .
. . .
. . . .
(29) (30)
v~L
The TbUs,
two-phase
pressure
gm.dient
is made
up
of three
components.
respectively,
B=.%JLsw
(a ($). (J (d.
+ ~ + ~
balances
(32) (33)
()
z,
dp
=ppgsiutl.
. . . . . . . . . . . . . . . . . . . . . . . . ..
(26)
bubble-rise
in a stagnaut
()
&f where~p defined by N&r*.
=fTPPrP% , 2d
Um =
. . . . . . . . . . . . . . . . . . . . . . . . . . . .. (27) Similarly, from a Moody diagram for a Reynolds number gu=
,A
. 1/.
1.2vm
0.35
[1
= of -. [1
. . . . . . . . . . . . . . . .
(34)
the
veloci~
the
is obtained
1.2,,+
1.53
. . . . . . . . . ..
(35)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
(28) where velocity the s=ond defined in term Eq.21. falliigfimcanb+ the Brotz13 expression, correlatedwitbthe film on the bubble-rise
flow is no the
by change
incompressible density of the resulting Therefore, compared with flowin the the
keeps pressure
constant,
acceleration.
(36) ex-
pressure drop
neglected,
components.
in terms
Slug physical
Flow
et al.lo Sylvesterl
developed 1 presented
the
fmt
a simplified
SPE
Prodnctio
& Facilides,
May
1994
145
slug fmm
void
fraction
can et al.
be slo
obtained and
by Schmidts
Sylvestersll 14 data,
cor-
TIE
geommy
of
the
film 2
flow
gives
HNUB
in
terms
of
8N
as
Femandes Vss
=l()
1-$
. . . . . . . . . . . . . . . . . . . . . . . . . ..
(46)
Sm Eqs. to obtain
= 29
35,37,
&d
38
can that
be Min. and be
determine
vNgTB,
the
net
flowrate
of &
can
be
used
to obtain
&
HLm,
VI,LS, can
Ngm=vrB_
(T8_gM)-.
. . . . . . . . . . . . . . . . .
(47)
showed to give
equations
algebraTbe length of the liquid slug can be calculated empirically from (48) this
(9.916
@)(l-_~H,,HVTB(l-H,TB)
+x=
0, (39)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . found16to the Taylor vary from bubble 16t045. length as Weuse C=30 in
. . . . . . . . . . . . . . . . . . . . ..
gives
Lm=[LH/(l#)I&
. . . . . . . . . . . . . . . . . . . . . . . . . . . .. comparison flow. This emherfor bubble of& requires developed volume can and LTB, new if & values flow. be used ? LTB, for the L~, flow f&,
(49)
.[vm.HgH[l.53[-~(1-H,u)O}].
.. . . . . . . . . . . . . ..... . .. (40) With be with readily an VB and Hgf-s given from by Eq. method. Eqs. 40. 34 Eq. and 39 38, is then the left respectively, used side to fmd ~ can
From
is develand
Taylor
iB A~(L)dL, ~ . where A& i This turn gives can can be be expressed expressed in in tennsof terms of local velocities holdup by hLTB(L), using Eq. . . . . . . . . . . . . . . . . . . . . . . . . . . .. (50)
HLm 39 as
De fting
of Eq.
F(ffLT,s),
F(HU,)
(9.916
@(
I--)05HLrurB(I-HLm)
+.l.
which 32.
.........
Taking the derivative of Eq. 41 with respect to HLTB yields
(41)
F(ffm)
VTB
(9.916@
~JL)=[l-(vr=i (51)
The HLTB volume w be expressed in terms of flow gmmetry as .. . . . . . . . . . . . . . . . . . . . ..
(1--0+ 4J~
Lu -vgDAP(lHuJ:. (52)
md52into
Eq.50
gives
HLTB,
the
root
of
Eq.
39,
is
then
determined
iteratively
flom
LB -V@(lHIM) ~
The ablesis
step-by-step as follows. WB
pmcedme
for
determining
all
slug
flow
vari-
and
HgLS
from 43,
Eqs. determine
34
and
1[ .
F.q. 53 can where
~_
(v,,be
vu)Hm Jzz
do . . . . . . . . . . . . . . . . . . . . . 1
and then simplified to give . . . . . . ..
(53)
Eqs.40tbrough
H~TB.
integrated
is HLTB=O.15. Solve Solve Solve Solve Solve Assuming of~. model developing of such with for flow, the the total cap slug flow, This as requires length length of is12
2
37 32 35 33 29
Note Note
that that
that&
in Fig.
3b,
we
~nd
calculating a developed
. . . . . . . . . . . . . . . . . . . . . . . . . . . ..
(57)
expression
calculating
L$8,
the
other
local
parameters
can
be
calcu-
LC=&II~a+_
[
where (called WgTB NusseIt and fhn
V:rB(O=
~-VIZ
. . . . . . . . . . . . . . . . . . . . . . . . . . ..
(58)
(vTrvU)HU . : In L; pressure and Jz gradiems, neglect the we effect consider of friction the effect the of varyTaylor . . . (59)
113
~N= ;dVNL.B#L(l-ffNLIB)
[
146
8(PL-P,)
. . . . . . . . . . . . . . . . ..
(45)
ing
along
bubble.
SPE
production&Facilities,
May
1994
For slug
flow, by
the
elevation
component
occurring
across
()
y dLe where The
=[(l-i$pts
+~pglgsinO,
. . . . . . . . . . . . . . . . ..
(60)
ST
. 1.
pU=pLHm
+pJ1-Ifu).
. . . . . . . . . . . . . . . . . ..
(61)
GAS
CORE
elevation
component
for
developing
slug
flow
is
given
by
LIQUID
FILM
()
where section P,B.
A dLe
[(l~*)pU
+B*pm.lgsinO,
. . . . . . . . . . ..
(62)
prm witi
is based varying
on film
average Wlckness.
void
fraction It is given
in by
the
Taylor
bubble
=p,HLm+P,(l-Hma),
. . . . . . . . . . . . . . . . . ..
(63)
where giving
HLTLM
is obtained
by
integrating
Eq.
59
and
dividing
by
L*m,
VF
2(uTrYw)Hm HL7BA = &G%The veloping This is friction slug given component flows as because is the it same occurs for boththedeveloped only across the liquid and deslug. . . . . . . . . . . . . . . . . (64)
rF
, . . . . . . . . . . . . . . . . . . . . . . . .. (65)
dp
(.)
m, where calculated R,U
=f_(l~),
~should by
bereplaced using
by,O*for
developing
flow.
f~canbe
i
=pUvmd/pB. . . . . . . . . . . . . . . . . . . . . . . . . . . .. (66)
For @lm eration ity exits uid occur celeration flow stable eration pressure is of
the must
pressure be
due
to liquid
he
velocity
in
the
of VLLS changes atso experiences VLTB two then tmh.f used of for a net no over into
the
a velocity the
an in
upward the
changes no net
within
unit, slug
pressure
-e
Dc i of annular flow. liquid enmined in the core, given 1.5)1, . . . .; . . . . . . . . . . . . . . . . . . . .. can be expressed as diagram for a Reynolds
2
8
by (71) (72) (73) number (74)
exist.
component
gradient
is considwed
a slug
FE Annular h flow Flow was tie as gave flow. ModeL presented classic A discussion by Wallis. on the hydrodynamics with this, and and of WafIis interracial Hall-Tayin an on this anapannualso wallis
is the as =
fraction
of
the
total
17 Along for
FE
l-sxP[4.125(u.,+~
where
uCtir=lO,
OOO~(~)fi.
analysis models
involved based
followed
The
shear
stress $
in
the
film
developed
annular applied
flow separately
is shown to the
in Fig core
momentum
A.%
()
~ core
-zj~t-p.A.gsin6=0 c
. . . . . . . . . . . . . . . . . . ..
(67)
defined
and
AP
* ()
at the
, ~L
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
vF=_=w
. . . . . . . . . . . . . . . . . ..(75)
density,
PO
because en fmined
the
core
a homogeneous same
fiquid
velocity.
PC=
PL&C+P&aLC), FEv~L
. . . . . . . . . . . . . . . . . . . . . . . ..
(69) f_
where
,IK
. SX
. . . . . . . . . . . . . . . . . . . . . . . . .. (70)
+
EVSL
ZF=.
@-FJP~
[-
&4J
. .
. . . . . . . . . .
(77)
SPE
production&
Facilities,
May
t994
147
Eq.
77
reducks
to by
To
simplify et
equation, is used. in
Alves
~F _
(l-FE)z
f~
4 [4C!(ldJ]2f~L
()
dL
d~
, ~L
. . . . . . . . . . . . . . . . . . . ..
(78)
mensionless Lockhart
Mutinelli
parameters,
where
the
superfmid
liquid
friction
pressure
grad~ent
is
given
by
~;=
(dp/dL)c-gpcsin8 (dp/dL)xc
....... ......
..
(94)
dp
()
z
fiL
_ .fsLPLv& , 2d SL
,,, .,,,,,
,,,
is the
fdiO.
f?iCtOr
for diagram
Supefi.ial for
liquid a Reynolds
velocity number
and
can
be by
and&
obtained
from
a Moody
defined
using to
the
moditied
Lockftart
MartineHi
parameters,
Eq.
93
re-
For
the
shear
stress
at the
interface,
Ti=~jPc!J~/8,
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
(81)
y~-4d(l&.5(14J]25 +
l%. above F(&), eqwtions then taking can the be 96 is derivative
. . . . . . ..
(96)
solved
obtain
&If
Eq. to ~
where
vC=v~c/(1-2c3)
z . . . . . . . . . . . . . . . . . . . . . . . . . . ..
(82)
with
respect
yields and~=fScZ, where thickness. p~ssion as low the for WhMey Z is . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. a correlating Based Zworks and on the well factor for in ferfacial of films the friction model, tie aid the Wallis (83) .zt4(l-21)] fibn ex@ [@(14J12[14_(14JI~5 performance for thin
or high is
Hewitt19 fbu.s,
expression
FE>
0.9
. . . . . . . . . . . . . . . . . . . . . . . . .
mdZ=l+24($13~
for
FE
<0.9.
. . . . . . . . . . . . ..
(85)
-@(l-@
[14@d)]35-
Combining
Eqs.
81
through
83
~,_d
Z 4(1-2@4
()
~ friction
dp
. SC
. . . . . . . . . (86)
dj+, =b.-~.
-J F@j) Once &is from known, the the fained followhg
. .. .. .. . .. . .. . ... . .. .:
dimensionless form of Eqs. groups 91 and #F 92
. . . . . . ..
(98)
The
supetilcial
pressure
gradient
in the
core
is given
(*)sc=fwherefSc defined by =pcv#i//4, ., . . . . . . . . . . . . . . . . . . . . . . . . . . .. is obtained from a Moody diagram for a Reynolds N% vsc and#c=pJU Tbe stituting pressure the =
(87)
@~=_Z.number (1-tiJ5
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(99)
(88)
FEW
VS8, +p,(l-,lLc),
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . .. for annulw into flow Eqs. 67 can and be 68. calculated llns, by
gradient above
equations
. . . . . . . . . . . . . . . . . . .. {drz=l.
(100)
Alves20stated that Eq. 100canbe expresseda$
($JC=*(*),C
~d () @ dLF _ (1-FJ2 *-3(1~~3 f, f, ()()
+pwsin6
. . . . . . . . . . . . . ..(9l)
~; = %+-YM
~
& d ~ The or 95 dp +p&slne. -4&f&@3 The film equating basic thickness, Eqs.91 unknown ~. An and92. () m in ,C the above equations for ~ is the can dimensionless be obtained by . . . . . . . . . . . . . . . (92) total pressure the
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..(101)
gradient pressure
can
then
bc
obtained film
either must
F.q. be
94 the
because
gradient
in the
same.fhus,
implicit This
equation gives
dp 4XIJWW5
()
= fF dp W ()
(l-FE)z _ 64d3(ldJ3fSL
=0.
~L
. . . . . ..
(93)
found range
except
a limited result-
of high
accelerational
component
SPE
Production&
Facilities,
May
1994
TABLE Nominal Source Old Bank Govier Fogaras% Asheim23 Chierici Prudhoe .Includes
ESPmOl,m
lRANGE Oil
OF Rate
WELL
Diameter (in.)
(STBID) 01010,150
TUFFP
Data
I108
70,5
and
2t04
8 to 1,600
114
tO 27,400
17t0
112
tO
740
to 55,700 6 to 27,914
35 8.3 24
7homas,28
to S6 to 46 to 66
,29
27L3 to 5 5% to 7
200
to 110,000
and
and Catwnter,=
and field
and Brown,%
seveml 03
Baxendall
0rkiszews!4
MMSU18.M,3*
cnmPanias.
ing film
from is
the
exchange
of
liquid
droplets
between
the
core
and
the
the
degree
of
scattering
of the
errors
shout
their
average
negligible.
ermc
Evahtation The evaluation tie updated with of another indus~. of pressure TUFFP a wide the model mechanistic range is the comprehensive drop well of also model from data data, compared tlmt am the bank as given with mcdel that in model with is the cank?d measured 1,712 1. six used The out hy data well perforwhere E4
dzop.
E,=
compming in the
()
}~ej e; =
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
(108)
i= 1
comprises Table of
(109)
that
commonty
Absolute
average
errcx
for
with using
. -()
is ASO magnitude Standard
j
;~lefl
i= 1
. . . . . . . . . . . . .
(1 10)
E,=
. ()
+~eri r. , ed = Apiw the pressure average
of
the
measured error.
pressure
drop
and
indicates
average
E,=
,= , APM-APi. . . . where . . . . . . . . . . . . . . . . . . . . . . .. (105) E6 sured EI indicates overatt drop. percentage error trend of the perfommce, relative to the Criteria Correlations The
r
the
(e,;_J2 ..
. .
. .
(111)
indicates pressure
scattering drop.
of
the
results,
independent
of
the
mea-
measured Absolute
for
Comparison and and Bmwn,z7 Beggs Models models Duns and et al..
With
Other
ccmelations and
used
for
the
comparison
area
Hagedorn x1OO. . . . . . . . . . . . . . . . .
E,=
()
.
~ j= I
~~le.1 i.1
........
(lo6)
gia heriee
Aziz
RISLUI
pmison Ez indicates Percent how standard large the errors are on the average. The defined
accomplished involves
statistical performance
parameters. factor
comparison by
deviation
E3.
J__
n-1 EDB 1712
(eri
-El)z . . . . . . . . . . . (107) F. =
IE,I-IE,J 1?21
m
E2-E2 .1
nun
I-IE1
+-
TASLE
WV
2-RELATIVE VNH__@~ 755 0.081 0.B76 0.803 1.711 1.836 3,321 5,836 3,909 ANH 1381 0.000 0,774 1,062 1.792 1,780 3,414 4.688 4.601
VNH.VB,liC4 with slug 100% flow and slug Wm.! R..
FACTORS SNH 1052 1.295 0.386 1.798 2.056 2,575 2,S83 3.12S 5,342
H.g8dom well cams dalx and and with Brown 100%
VSNH 387 0.142 0.939 1.486 2.298 1,996 3,262 4.403 4.683
cases w.U Wi!hout Hagedorn s[ug
and flow
VS=vetical and
cases annular
Hagtiorn and
with
Hagedorn com!atiow
Bmn
,&AN4] Kabir
100%
DUNRS=Dun3 correlation.
HASKA=H.Sa.
mechmlstic
BEG8R=Bww
corml%x
)RKIS.OrWzews.ki
cerrela!iox
E,-E, + E3mm-E3dn +
IE41-IE4 IE4JIE,
I I
6.
Several and
variables film
in the thickness, to
mechanistic we
modeI, dependent
such cm pipe
as
bubble inclination on
rise
nun
include improve
inclination model
angle performance.
effects
these
ETE5mh + E5m-E5ti +
E=E6Mn . . . . . . . . . . . . . . . . . . ,. E6muE6& We thank used Ltd. the TUFFP part fnancial member of fbis companies reseacb whose projecf, provided and A.M. membership Pakistan Ansari. fees Petrowere Ieum in TabIe 2, to fund for the (112) Acknowledgments
The
and best
vafues
for
Fw
are
O and
6,
respectively. of Fw is given
support
References L Ozon, Flow P.M., Model presented Sept. A.R. Ferschmider, Pdicm al the G., Resmre 1987 and and SPE Chwekoff, Temperature Offshore Emope A.: A New M.kiphme paper SPE Aber-
value
being
The so as
ov?rall to study
involves
the
entire of evalin
petionnance together The resulting for deviated comparison datsba.w and vertical for 5 of combined Table 2. vertical
S1 1, 19S7. and Kabir, SPEPE D.. and C, S,:A Smdy 1988) of Mulfiphase 263. Moddline Flow in VeI%l Flow Paftem Tubes, TmnAIChE flow Behavior in
pattern by using
bank, checked
(M&y
Dnkter.
A.E.:
resulting in
and make
for the
weU
Upward
Gas-Liquid
. . .. . . . . .,-.. .
9$4)>
D.,
Y:
F1.aw
VeI74I.
331
Chem Predicting
Flow-Pattern Intl.
Tmnsiti.m Flow
shown and
Col. weil
inclinations;
J. Mulliphase
deviated
cases
shown
of Lacge AJChEJ.
Drops (1960)
and
i Media
of Inti-
of Individual flow one pattern particular earlier. existing order for For
~e of
perfmmance data by well length These of Rable for selected 9 and eliminated 10 the cases we that m
7. SCOR, for
G.E.:
Advances Imfimd
i Slug P@efines?
Slightly SPE
dominant tions bubble sidered are results tbe the 11 well entire were shown for
bansiwifh con.
esented New
Annw.t
Technical
Cmference
and Exbibitim,
23-26. Vertical U. of Tulsa, L: Sfeady llvcHimse T!m, Stale and Limits. OK ROW T&o.gh an .&mu-
more an Table
Upward ion,
to have 6 of
dissmta[
N. and
Hench, and
flow
exist
2fmir
operating
62GL1OO Dukfer,
A.S.:
Gas-Liquid N,D.:
in Vemicaf
Tubes, for
the data,
bank
IL
Mechanistic
the those
ASMEJ, and
Energy Whalley,
(1987) 161.
2 gives m
results be in
12. McQnilIan,
Flow (1985) da
in Vecricd
Tvm-
data well
bank length.
flow
100%
13. Brotz, results of each (El, model E6) are or correlation give in the 14. keit
Vmausberechmmg
performance statistical
this
Gasen 26,470.
in Stmrnemkm
Ftumiekeitsscbicbkn.
parameters
paper.m
Study
Flow
inaPipeline-Riser
U. of TM.% 0.:
*A Note
Existence ASME
for
Conclusions From and Cofs. other 1. The rior a.nces Hasan the in last these to afl of 1 through empiricaf overall other the 11 of Table 2, the performance the fcdSowing, mcdel the aL, that of of by made and best Brown models results found The with model model slug flow when bank and well of methods the (Cols. Brow cases annufar (Col. 11). Brown tie to ovemfl and the is speperfofmRm, model. mechanisms Hagedorn data and used In are com 20. 19. and For of the mcdel 16.
i VertimJ
111,64. MaroD, D. M., Stable and Branmr, Slug La@? N.; C<APhysical Chem, E.g. Model Set. forpr~ (1985)
indicates the
Minimum
me fhods Hage.dmn
. ...,
G.B.: New G.F. One-Dinemimxd York and Citv (1969). N.S.: Annular Two-Plw.w F@ PergaTwo-Phase Flow, McGmw-Hill Book Inc.,
and three,
Kabir this
comparable to
attributed
mo
performance cmly
T&
Cormladmof
Liquid
Entrainment
Erdminmmt 187,
i Annular (1978).
Two-Phase
Flow
URAEA
the
data
ModdinS 19S8
WeJls,;,pa. Chicago,
(Cols.
m fbe 2. and
correlation for (Cccl. over75% performed for model the bubble is deviated 3),
betnone 21.
than the 3.
Dukfer,
A.E.:
Droplet to Mommtwn
Entmimnmt Tram
in Verdcaf fer~A!ChEJ.
Aruu(1986)
fbods 29
Its Ccmtibution
of the
well
15rNl Goviq and G.W. Codemate. H.: and Fogarasi, J, Cdn, M.: Pet. pressure Tech Dmp i Wells 1975) Well producing 28. Bawd Gas
to be and
i bubble Kabir
mechanistic of the
Ash.im, on Phase
MONA,
FlowModel
correlation data
24.
Chierici, Flow
Two-Phase
in Oil
Hagedorn vertical
included 10). is
25.
Poemmmn, and
of Gas,
(Ccds. flow
9 and models
W.mfTbrougb Gas-Lift
Strings
significantly
and 257.
& Prod
150
26.
F?acher, Multiphase
G.H,,
and
Brown,
K.!2
TrvXctio. AfME
for
NW= ~ =
Reynokk pressufe, flowrate, wetted velocity, volume, Lockhart Lockhart empirical Iengfb fdm ratio ratio,
number m/Lt2, L3{t perimeter, Ut L3, and and factor mls m3 Martielli hfardnelli defining in m to diameter parameter parameter interracial F.q. 31 friction psi m3fs L, m
Tram..
27.
q = s. v =
Con.
1
L, md or m &g
28.
Baxendell, Flov?ig
in High
Rate V= X= Y=
Wells;
29.
Two-Phase
DrOPS
in
VtiCal
30.
Three
Methods
Flow,
fore?.lculating thesis.
aFms-
Z = ~ =
MS
U. of2Msa.Tul-
defined L, thickness
C5 = Redicting
thesis.
thickness, of fti
31.
Messufam, phase
MukiU. oPDd-
C3 = X s = =
difference absolute angle no-slip dynamic kinematic density, Solace shear from pipe m@mss, horizontal, fraction kg/ins. L21t, kghr$ m21sq
(1970) comparison Gas-Liquid Ratio of Correlations Vedicrd for PTedcdng fhesii. Ressure
@ = WelisYMS U. ofTulsa, A = Vertical Pet. in the in Wells, Con% Ftow of Gas and Liquid 451. Correlation Tech, (M.xch, for predict1989)41, Mixtures P = (1970).
holdup viscosity,
(1963)
v = p = ~ r = =
Discontimilies
Orfi=wsid J. E.eqyRes.
H.D.
J.P.:
Sfudy
of Two-Phase
Flow
in fnctiied
@ =
dimensionless
Pipes, - JPT(?vfay 36. Palmer, Correlations OK 37. (1975). H. and Flow, Govier, Gil C.P., Mcdel A.M. and and C.M.:
6+77. of fnclimd Data, Pipe MS Twc-Phase thesis, U. Liquid oflldsa. HoldIIp Tu@ Subscripts a = amelera,tion average Taylor critical elevation
friction
Expetimmcal
Muklmjee, IVc-Phase
Brill,
Drop fDec.
for
fnctined
bubble
cap,
core
38.
tiIz, ducing
K,
pressure (July-gept.
Gas, Hasan,
39.
Kabir, Flow
f.
F= g.ga3 H = i = I= L IS= =
i Vmical et d,:
40.
to paper
A Comprehmin WeUbores,paTX.
film
Mcdd
available
Richardson,
interracial liquid liquid . mixture modified maximum . Inil-hum Nimselt = pipe relative slip superficial slug total Taylor two-phase bubble unit slug
Nomenclature
m = = coefficient cross-sectional coefficient cceffkient constant number defined area defined defined factor for in of in in Eq. pipe, Eq. Eq. 56 57 factor to Reynolds 55 L, m2
M.
a A
mu= mill N= p
b = c-= C =
relating smoofh in m
r = s = S = SU = t. error, % 3B = TP.
coefficient pipe error average absolute standard average absolute standard friction . = fraction relative gravity local average Iengfb = number exponent rise diameter, function
defined L,
error, mlLt2,
mlL12, psi
f=
FE FT
in
gas
defined
g = h = H. L. n z=
T.mvmhn
factor
is exact
SPRPF
pips, cases
m
OIIOi.4 SPE 2s, 1993. ma.u!ctipt Paw? 1990 SPE mceiv@d acmpted Annual lor for review Sept. DWC 2, 6, 1930. 1993. 4 Revb%d Fawr mm.szrht (SPE held 2Ce30) i New -bed fimt Orbs.,, Pr6-
to account velociw
for
the
swarm
effect
on
bubble
Sept. ~#:2~
Lmbl!cation
Tecim@d
Cnnfemnm
Ei+M+io
SPE
production.9
Facitides.
May
19P4
151
Pressure 111, 34
ASME
JERT,
26Chlerlol, G. L,, Culeol, Q, M, ,and Soloccl, G,: Two.Phase Vertical Flow In 011 Wells . . Predlotlon of Pressure Drop, SPE J. Pet, Tech. (Aug. 1974), 927-938. 26 Poet? mann, Multlphase Tht ough Application Installations, Prlictlaes, F, H., and Car penter, P. G,: The Flow of Gas, 011 and Water Str Ings with Vertical Fiow Design of Gas. Lift to the API Drllllng and Produotlon 257317 (1962).
Study Jilts
of
Evacuation of inc!!ned Two- Ph#jse Liquid Hoidup Correlations Experimental Data, M, S, Thesis, University of Tuisa (1975).
K. E,: Pr edlc4 Ion 27 Fanoher, G. H., and Brown, of Pressure Gradients for Multi phase Flew In Tubing, Trans. AM4E(1 963), u, 59-69. 28 Hagedorn, ~lRa.d ~, Dissertation, Austin (1964), Ph, D. Texas at A, R,: ~~
38 Mukherjee, H. and Brill, J. P.: Drop Correlatior?s for inclined ASM5 JERT (Dee., Flow, Trans.
3gAziz, Y.,
G. W, in Welis
The
University
of
lWGr.@h a A b c
of Pr es$ure 28 Baxendell, P, B,: The Caloulatlon SPE Gradients in High Rate Flowing Well), J. Pet, Tech, (@S, 1961), 1023. 300rklszew sKI, ..t,: Predicting Two-Phase Pressure Drops in Vertical Pipes, SPE J. Pet, Tech. (June 1967), 829=838. 31 Espanoi, ~~
lRMLWU4ulikp~ZSe FI ou ,
c c
d D e El 132 B3 E4 Es E6 L 8 h If
H,
J.
H.:
Q21tlfJJf~ff M, S. Thesis
of Tuisa A,
(1968). Q,: ~ o~
The A,:
University @mg6111~
of Tuisa
(1970),
L
M.S* of Tulsa (1970), n Th@~is, The n
34 Duns, H,, Jr, FIGW of Gas ~Worid (1063), 35 Briii, J. Qkiczewski P,:
and Ros, N, C, J,: Vertiaal and Liquid Mixtures in Welis, pet, Congr es% 451,
N P
in the Predicting
Q Re RPF s
coefficient defined in Bq. 46 cross-sectional area of pipe, mz coefficient defined in Hq. 47 coefficient defined in Eq. 48 constan: factor reldtg friction factor to Reynolds number for smooth pipm coefficient defined in Eq. 3$ differential change in a variable pipe diameter, m error function avarago percentage error, % absolute average porcontage error, % standard deviation, % averago error, psi absoluto average error, psi standard deviation, psi friction factor fraction of liquid entrained in gas coro grswity aoceleratlon, m/s2 local holdup fractton averago holdup fraction longti~ along the pipe, m exponent relating friction factor to Reynolds numbu for smooth pipm expentmt to account for the swarm effect on bubble volooity number of well cases successfully traversed pressure, psi [ N/m2 1 flow rate, m9/s Reynolds number Rolat{vo Porfortnmtao Factor, defined in Iiq. 92 wetted porimotor, m
riso
180
v x
Y
velocity, mls volume, m Lockhart and Martinelli parameter Lockhart and Martinelli parameter empirical factor dcfinitig interfaclal
friction
Icmgth ratio, defined in I@ 27 film thickness, m ratio of fihn thickness to diwnetcr dlfferenoe aiwolute pipe roughness, m dimensionless groups, defined in Eqs, no-slip holdup fraction dynamic viscosity, kg/m+i kinematic viscosity, m2/s angle from hortizontnl, rad or deg density, kg/m3 surface tcttsion, dyntdcrn shear stress, N/m9
79 and f?f)
.yw%m.ii
A c crit 9 f P o H i I L LS
M
mn N r s S w t m Tp
acceleration avorago Taylor bubble cap, com critical elevation friction film ga8 hydrauIie ith element interfacing liquid liquid slug mixture min{mum Nussolt relative slip sttporfioial slug unit total Taylor bubble two.phase
dtwelopin~
slug flow
TArKE MXGE
1 DiSVi
TABLE 3 STAT1871CAL RESUEE5 USING ALL VSKIICAL E2 w i3AGBR WIODEL 10.s 14.5 34.G Es WELL C!MES
OFWELL
mm.
OIL]
QLEak2 Smom
Gad@& @W/Dj
01 Grat&z
{%1
15.1 19-2 19.3 21.9
[&i)
-75 -17.7 -18-6 232 52-0 509 ?8.0
{Ril
959 81.3 98.4 102.0 121.7 154.9 1472
[Ri]
173.9 144.9 M2.5 176.3 Has 298.8 211.0
A@n
1-8
0-10150
1-5-10567
9-5-70.5
Kuz
8-?600 720-27000
114-27400 740-55700
0-3-534702-68 600-23000
: 8.346
MUiiR
Uiitrmecal=
TABLE 4 24-86 !STATISTZCAL RE-SULX?5X7*SING ALL V8E?lTCAL WELL CASES WTHOUT HM3EDORW AN= BROWN= OATA E2 {%] MODEL %chadcs da?a&om Fmmmaml and &qxo*&r26. F~&a ~d ~27. Hagedmn amd Brmm~BaxcndeU arid 3bri9as29. OrEszewsl@O. E5pafio13u MeSSaam=. and ~cbc@ !Md data &om smera? oii HAGRR A212 DI)NROS 10.Z 12.2 128 15.0 182? MUR%R ORKIS 20-6 27.4 ES
~OAJ ~~i) [Ri)
g)
1722 207.2 216.1 209.2 2356 239.5 362-4
(-)
5.000 6.801 8.459 10.814 19-166 21s01 22.400
STA77S77CAL RES13Z3S USRIG N.Lmw VEm?ciLmm3. cAsEs E2 :%] MODEI. fmz Iz&Gss m?NiROs 0R81S 121 X22 92 122 16.1 14.4 17s 17.1 168 13.6 185 3+72 202 20.2 93 -20-8 -285 33-4 122 41-3 78.7 IOL3 116.S 10243 1109 XW3 1343 158s 163-2 190-4 37s.4 177.7 2733 207s 217-2 5373 MODEL 7349 HWBR 7-101 DWWROS 8.470 Jlz12 8X53 M3KSR 10.102 24.5 14.751 oRms 60-7 71.9 295.6 453-5 538.1 118.515 25.7 152.6 215S) 193.0 58.808 la2 19.8 110.3 176.5 191.3 43.140 102 14.7 -909 154.6 280.5 35-685 18.1 27.1 -6.4 165.8 216-7 9261 10.6 14-8 13.1 122.1 166.2 8334 8.6 12.3 -3.0 109.0 164.4 5.000 Es (%)
[Ei)
lsj)
(Ri)
(-- )
~ ~ Sl&k WnH
T=IS
--:1
SrA=iXAI.. RESETS USING Ail. WELL CASES WITH K)(E% SLUG FLOW
.. .-
E2
[%] IKQDEL Aziz omaS 3-2 32 3-3 3.6 ~) 3.7 a7 42 40 (Ril -*kx-8 -30-3 -269 -47-8 (Ril 67-0 68S) 69-4 77.5 78.7 792 155.6
l%
[psi] 76-9 79-1 90.6 8.2 80-1 102-6 83-3 n 5.000 5286 5.493 6.374 6.511 6.842 12.852 IXmROS .
E2 [%] Az32
MODEL HAGBR 14.8 162 10-1 14.6 15.5 15-1 213 :9.8 20.4 14.8 26.3 21-3 21.4 213 5.6 13.0 -19-7 17.4 43.7 56.6 99-1 1023 1012 80.4 116.3 114.8 m82 153.2 173-8 160.8 176.8 212.9 184.9 170.7 1972
RPF (-1
6.016 7.413 7.CW5 8.820 13.181 15276 24.146
D?mRos HAG8Ra8
8EG3R MEKBR
Oms
4s
3-8 7-3 4.S 3-8
-44-9
46-6 -154.0
ii
TA%LE8
-MBLE 9
SBmSKXL
UELLCA6ESWTIH
ALL VEK7iCAL R8S?iEXS t.SIXG mLY%sLz??Fl nwwr7HoliT HAGEDmN A??= BRtwn.= Dfsm Es [%)
E4
SmnSTicilL WITH RPF (-) 5.s31 5.896 7.118 22.694 24.619 25.873 32.319 MODEL J!2Jz Z2 f%) 9.7 12.4
E2
).
~o~)
Q-)
-7s 59 14.4 M3Lxo 101s 118-9 152.3
RPF (-)
5.000 5-896 8.652 11283 17.409 20.515 45.810
BIODEL
A212
EL%G8R
DUNROS MUKBR SEE
1?5.1
20-0 2s-5 322 78.7
mmRos
ORKXS
KGK3R
332
or?ms
~
I
t.k ,w
t:
!
0000 0000
0:.:0:
t ~
t
..$. . . . . . . . $.. ,, . ,..
0,
.n {?=
201 I I I I [ I
u ,
Do o
Q:? 33.%7 ~o 00
!2 Q
3
z A a
6 ii
BUBBLT
BARNEA TRANSITION
0.1
/
0.1
I [ I
D I
ANNULAR
SLUGOR CHURN
.*
. .*
.$
K u.! o.
a u)
0.0
I
II
1 I
II
i
I
I I 1 I I
}
SLUG FLOW
Fig. l-Flow
t t C1-llol; A;lW&AR
0.00i
02
10
100
SUPERFICIAL
FIu. 2-Typlcel
,aso~
V~TB DEVELOPING J TAYLOR BUBBLE -
.-1
~
0
v L O?do
4L
GTB
LyB
1
NTGTB
164
1, ,,, ,,.
I ,,
!. ,,
I :.
,
1
10
,
.
9.0
GAS CORE ,, LIQUID FILM ~ ENTRAINED LIQUID DROPLET ~ 1 _ ,, . ,, ., . .
I ?2(
+
x
CALCULATED
PRESSURE
ANNhLAFl
MEASURED PRJNWRE
J.
\
\
,, 1: I ::, ;:: ~F
. . IC
h
:,:
,4
,,
S1
,:
i, ,. i , ,,,
,.,
F19.4-Sc,hemellc diagram o! annular flow, ~ Fig. 5.-Performanoe
PRESSURE
of tho co~praltan$lve
( PSI ) :
model-typloal prenaure profile,
1
,,,
16s