SPE 84226 A Mechanistic Heat Transfer Model For Vertical Two-Phase Flow
SPE 84226 A Mechanistic Heat Transfer Model For Vertical Two-Phase Flow
SPE 84226 A Mechanistic Heat Transfer Model For Vertical Two-Phase Flow
Abstract
Convective heat transfer for vertical gas-liquid two-phase flow
was investigated experimentally and theoretically.
Experimental data on convective two-phase heat transfer
were acquired with a crude oil-natural gas system at cooling
conditions using a large diameter (2.067-in I.D.), high
pressure (450 psia) test facility. Flow pattern dependencies of
convective heat transfer with changing liquid and gas
velocities were revealed.
A comprehensive mechanistic heat transfer model was
developed by flow pattern dependent approach for bubbly,
intermittent and annular flow in vertical pipes. The model is
capable of predicting flow pattern first and then predicting
hydrodynamics and heat transfer based on the
predicted flow-pattern.
Comparing with experimental data, the model is found to
predict two-phase flow heat transfer coefficient within 26%
error for all flow patterns showing a better overall
performance than existing correlations.
Introduction
As oil and gas production moves to deepwater environments,
production systems with subsea completions and tiebacks to
existing platforms have become common occurrences. In such
systems, long-distance transportation of unprocessed reservoir
fluids, which are normally multiphase systems, from the
reservoir to downstream process facilities, must be assured. In
order to optimize the design and operation of such systems,
engineers must understand how petroleum fluids behave, both
hydrodynamically and thermally, during transportation. In
particular, wax deposition during transportation of waxy
crudes is found to be sensitive to convective heat transfer1.
Experimental data2-5 on two-phase heat transfer were
reported for a variety of fluids, gas and liquid flow rates, pipe
diameter, inclination, and flow patterns. There are a number of
SPE 84226
section. The inlet and outlet temperatures for both the twophase mixture and glycol were measured during the test
duration after the flow became steady hydrodynamically
and thermally.
Average convective heat transfer coefficients are
determined by a heat balance based on heat flux
measurements14. The total heat loss of the two-phase mixture
between the inlet and the outlet of the measurement interval,
Q, is determined by
Q = (m& L C pL + m& g C pg )Tm . ........................................ (1)
(dL )TLM
. .......................................................... (2)
The system pressure was fixed at about 350 psig. The inlet
glycol temperature was fixed at 120oF. The inlet temperature
of the two-phase mixture was set as high as possible to reduce
measurement uncertainty. The fluid inside temperatures varied
from 139oF to 157oF. The calculated relative uncertainties in
the convective two-phase heat transfer coefficient varied
between 2.5 and 13.0 percent of the reading14.
Experimental Results
A total of 55 tests were carried out, including 5 tests for
single-phase liquid, 6 tests for single-phase gas, 8 tests for
bubbly flow, 27 tests for intermittent flow, and 9 tests for
annular flow.
Single-Phase Liquid Flow. The experimental convective
heat transfer coefficients for single-phase liquid flow, hSL,
were measured to verify the facility and evaluate the existing
model. The convective heat transfer coefficients increase
linearly with increasing superficial liquid velocity, except for
the data point in laminar flow (ReL=828).
Single-Phase Gas Flow. The experimental convective heat
transfer coefficients for single-phase gas flow, hSg, were
measured. All the data were acquired under a fully-developed
turbulent flow (Reg > 40664). The convective heat transfer
coefficients increase linearly with increasing superficial
gas velocity.
Vertical Two-Phase Flow. Experimental convective twophase heat transfer coefficients, hTP, are plotted against
superficial gas velocities at fixed superficial liquid velocities
in Fig. 4. In general, higher hTP values are obtained for higher
vSL values for a given vSg.
Nu = 1.86 Re Pr
L
1/ 3
0.14
............................... (4)
45
Pr b
w
1 3
f
Re Pr
8
Nu =
f
1.07 + 12.7
Pr 2 / 3 1
8
0.14
.................................. (5)
0.25
........................... (6)
SPE 84226
Bubbly Flow
A pseudo single-phase approach is used to predict the
convective heat transfer coefficient, as well as hydrodynamics,
for bubbly flow. The flow is assumed to be fully turbulent and
the heat transfer in the gas phase is neglected. Moreover, the
bubbles are assumed to have no influence on the boundary
layer and the increase in hTP is only due to the increase in the
mean velocity of the two-phase mixture.
The Petukhov-Kirillov correlation is utilized with liquid
physical properties, L, L, CpL and kL, and mixture
velocity, vm,
NuTP
f
ReTP PrL
8
=
f
1.07 + 12.7
PrL 2 / 3 1
8
L ,b
L, w
0.25
. ................. (8)
L vm d
............................................................. (9)
L
Nu TP k L
............................................................ (10)
d
Annular Flow
Kaminsky9 proposed the modeling of convective heat transfer
separately for turbulent and laminar. In this study, Kaminskys
method is utilized only for annular flow liquid films with
slight modifications. The definition of the critical liquid film
Reynolds number, ReF, for the laminar-turbulent transition is
still an unsolved issue. Many studies have been carried out on
this topic. It was found that the critical liquid film Reynolds
number varied in the range between 1,000 and 1,200.18 In this
study, a value of the critical liquid film Reynolds number of
ReF = 1100 is used.
Turbulent Liquid Film. The Prandtl-Taylor analogy6 is
applied for the turbulent liquid film with the
following assumptions:
Heat transfer between the liquid film and gas core
is negligible.
Flow in the film is divided into two parts: laminar
boundary layer and turbulent core.
Laminar boundary layer is much thinner than the
liquid film thickness and there are no bubbles in the
liquid film.
Temperature drop occurs only in the laminar
boundary layer.
q/ is constant over laminar boundary layer.
The momentum transfer and heat transfer in laminar
boundary layer are expressed as, respectively;
v
dv
= lb , ........................................................ (11)
lb
dy
q = k
T Tw
dT
............................................ (12)
= k lb
dy
lb
qw
k
= L
L
(Tlb Tw )
....................................... (13)
vlb
or
hTP =
k
qw
= L
Tw Tb L
vlb
. .................................... (14)
vlb = w . ........................................................... (15)
F =
d dP
........................................................ (17)
4 dL TP , f
hTP
k L L dP
=
d
10 L dL TP , f
1/ 2
.................................. (18)
hSF
k L L dP
=
d
10 L dL SL, f
1/ 2
.................................. (19)
SPE 84226
hTP
h SF
dP
dL TP. fric
=
dP
dL SL, fric
1/ 2
= L ..................................... (20)
= 1.86 ann
Nu LS
Nu TP k L
d hF
kL
d
Re SF PrL
d
L
Intermittent Flow
The convective heat transfer in vertical intermittent flow can
be analyzed separately for the heat transfer coefficient in the
liquid slug body, hLS, and in the Taylor bubble region, hTB.
Thus, the average coefficient over a slug unit, hTP, is
expressed as,
1/ 3
. ... (22)
0.14
H LF
2 / 3
f
Re LS PrLS
8
=
f
1.07 + 12.7
PrLS 2 / 3 1
8
LS ,b
,
LS
w
0.25
. ......... (27)
Where the Reynolds number for the liquid slug body, ReLS,
and the Prandtl number of the liquid slug body, PrLS, are
respectively defined as,
Re LS =
PrLS =
LS v m d
,........................................................ (28)
LS
C PLS LS
. ....................................................... (29)
k LS
Nu LS d
. ........................................................... (30)
k LS
Model Evaluation
The performance of the mechanistic two-phase heat transfer
model is evaluated using both sensitivity and statistical error
analyses. The model is also compared with the existing
correlations recommended by Kim et al.10(OSU Model). The
statistical error analysis results are summarized in Tables 1-4
and discussed separately for respective flow patterns.
SPE 84226
SPE 84226
Nomenclature
Cp = heat capacity
d = pipe diameters
f = friction factor
HL = Liquid Holdup
h = convective heat transfer coefficient
k = thermal conductivity
L = length
m = mixture
m& = mass flow rate
Nu = Nusselt number
Pr = Prandtl number
q = heat flux
Q = heat flow
r = pipe radius
Re = Reynolds number
U = overall heat transfer coefficient
y = distance
= ratio of Taylor bubble to slug unit length
= shape factor (constant)
= dimensionless pressure group defied by Lockhart
and Martinelli
= viscosity
= density
= shear stress
Subscript
ann. = annular
b = bulk
circ. = circular
F = film
g = gas
gl = glycol
i = inside
L = liquid
lb =at laminar boundary layer
LS = liquid slug
LTB = liquid in Taylor bubble region
o = outside
p = pipe
SF = superficial film
Sg = superficial gas
SL = superficial liquid
TB = Taylor bubble
TP = two-phase
w = wall
Acknowledgments
Authors thank Tulsa University Fluid Flow Project (TUFFP)
and Tulsa University Paraffin Deposition Project (TUPDP)
member companies for their supports. They also acknowledge
Tony Butler and Mr. Emmanuel Delle Case for their
dedicated work.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Aggour, M.A.: Hydrodynamics and Heat Transfer in TwoPhase Two-Component Flow, Ph.D. Dissertation, U. of
Manitoba, Canada. (1978).
Vijay, M.M.: A Study of Heat Transfer in Two-Phase TwoComponent Flow in a Vertical Tube, Ph.D. Dissertation, U. of
Manitoba (1978).
Rezkallah, K.S.: Heat Transfer and Hydrodynamics in TwoPhase Two-Component Flow in a Vertical Tube, Ph.D.
Dissertation, U. of Manitoba (1987).
Kakac, S. and Yener, Y.: Convective Heat Transfer, 2nd Edition,
CRC Press (1995).
Ravipudi, S.R. and Godbold, T.M.: The Effect of Mass
Transfer on Heat Transfer Rates for Two-Phase Flow in Vertical
Pipe, Proc. 6th Int. Heat Transfer Conf., 1, 505-510 (1978).
Rezkallah, K.S. and Sims, G.E.: An Examination of
Correlations of Mean Heat Transfer Coefficients in Two-Phase
and Two-Component Flow in Vertical Tubes, AIChE Symp.
Series, 83, 109-114. (1987).
Kaminsky, R.D.: "Estimation of Two-Phase Flow Heat Transfer
in Pipes," ASME J. Energy Res. Tech., 121, 75-80 (1999).
Kim, D., Ghajar, A.J., and Dougherty, R.L.: Development of
Improved Two-Phase Two-Component Pipe Flow Heat Transfer
Correlations from Existing Correlations and Published Data,
presented at ASME/JSME Joint Thermal Engineering
Conference (1999).
Ansari, A.M., Sylvester, N.D., Sarica, C., Shoham, O., and Brill,
J.P.: A Comprehensive Mechanistic Model for Upward Flow in
Pipes, SPE J. Prod. & Fac., 143-152 (May 1994).
Xiao, J.J., Shoham, O. and Brill, J.P.: A Comprehensive
Mechanistic Model for Two-Phase Flow in Pipelines, SPE
20631, presented at 1990 SPE ATCE, New Orleans (1990).
Kaya, A.S., Sarica, C. and Brill, J.P.: Comprehensive
Mechanistic Modeling of Two-Phase Flow in Deviated Wells,
SPE 56522, presented at 1999 SPE ATCE, Houston (1999).
Manabe, R.: A Comprehensive Mechanistic heat Transfer
Model for Two-Phase Flow with High-Pressure Flow Pattern
Validation, Ph.D. Dissertation, U. of Tulsa (2001).
Sieder, E.N. and Tate, G.E.: Heat Transfer and Pressure Drop
of Liquids in Tubes, Ind. Eng. Chem, 28, 1429-1453. (1936).
Petukhov, B.S.: Heat Transfer and Friction in Turbulent Pipe
Flow with Variable Physical Properties, Advances in Heat
Transfer, 6, 505-564 (1970).
Dittus, F.W. and Boelter, L.M.K.: Heat Transfer in Automobile
Radiators of the Tubular Type, Univ. Calif. Pub. Eng., 2,
443 (1930).
SPE 84226
1,hTP
2,hTP
3,hTP
4,hTP
5,hTP
6,hTP
(-)
(-)
(-)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
This model
-0.157
0.157
0.038
-27.041
27.041
14.709
OSU Model
-0.139
0.139
0.034
-23.178
-23.178
11.202
This model
OSU Model
1,hTP
2,hTP
3,hTP
4,hTP
5,hTP
6,hTP
(-)
(-)
(-)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
-0.147
-0.092
0.3
0.188
0.312
0.198
-19.27
-17.222
48.083
33.416
50.555
67.628
1,hTP
2,hTP
3,hTP
4,hTP
5,hTP
6,hTP
(-)
(-)
(-)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
This model
-0.203
0.221
0.177
-20.439
22.053
19.867
OSU Model
0.336
0.434
0.637
54.471
64.555
135.094
1,hTP
2,hTP
3,hTP
4,hTP
5,hTP
6,hTP
(-)
(-)
(-)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
(Btu/hr-sq.ft-F)
Bubbly
-0.157
0.157
0.038
-27.041
27.041
14.709
Intermittent
Annular
Overall
-0.203
-0.147
-0.164
0.221
0.3
0.258
0.177
0.312
0.26
-20.439
-19.27
-17.139
22.053
48.083
34.375
19.867
50.555
40.999
SPE 84226
TT4
5
TT 50
TT40
TT30
TT35
250
Glycol
1/2
200
TW1
TT49
TW 3
2 o
TT29
2.00 in.
TW2
TF
1
45o
1/2
1.38 in.
TT4
4
TF
2
TT39
0.69 in.
TW4
TF
3
150
100
TT 34
Laminar - Sieder-Tate
Turbulent - Petukhov-Kirillov
Turbulent - Sieder-Tate
0%
-10%
10%
50
0.20 in.
TW
5
0
0
50
100
150
200
250
2 o
0.197 ft
(0.06 m)
5.0197 ft
(1.53 m)
5.0197 ft
(1.53 m)
5.0197 ft
(1.53 m)
5.0197 ft
(1.53 m)
150
300.00
Annular
Intermittent
Bubbly
+10%
250.00
120
-10%
hTP
2 o
200.00
150.00
VSL=O,2
100.00
VSL=0.5
VSL=1.0
90
60
VSL=2.0
50.00
VSL=2.5
30
VSL=4.0
VSL=5.0
0.00
0.10
1.00
10.00
100.00
0
v Sg (ft/sec)
30
60
90
120
150
100.00
300
hTP/hSL
250
10.00
VSL=O,2
VSL=0.5
VSL=1.0
200
150
100
VSL=2.0
VSL=2.5
VSL=4.0
50
Exp.data (vSL=4.0f t/ s)
This model (vSL=5.0f t/s)
Exp. data (vSL=5.0ft /s)
VSL=5.0
1.00
0.01
0.10
1.00
10.00
100.00
1000.00
vSg /vSL
0
0.10
1.00
10.00
v Sg (ft/sec)
SPE 84226
100
10
Exp.dat a ( vSL=2.5f t / s)
hTP/hSL,cal
hTP/hSL,cal
10
0.1
0.01
0.1
v Sg/v SL
Figure 9. Sensitivity of Enhancement of Convective Heat Transfer
Coefficient to Gas/Liquid Superficial Velocity Ratio for Vertical
Bubbly Flow
10
vSg/vSL
100
1000
400
300
200
Predicted h TP (Btu/hr-ft2-F)
250
This model
OSU Model
0%
+40%
-40%
350
This model
OSU model
0%
+30%
-30%
150
100
300
250
200
150
100
50
50
0
0
0
0
50
100
150
200
Experim ental hTP (Btu/hr-ft2-F)
250
100
150
200
250
300
300
50
350
400
This model (vSL=0.2f t /s)
hTP (Btu/hr-ft2-F)
300
250
250
h TP (Btu/hr-ft 2 -F)
350
300
Exp.dat a (vSL=1.0f t/ s)
This model (vSL=2.5f t/ s)
Exp. data (vSL=2.5ft / s)
This model (vSL=5.0f t/ s)
200
Exp.dat a(vSL=5.0ft / s)
150
150
100
100
50
50
0
1.00
200
0
0.10
10.00
100.00
vSg (ft/sec)
1.00
10.00
100.00
vSg (ft/sec)
10
SPE 84226
hTP/hSL,cal
10
1
This model (vSL=0.2f t/ s)
Exp.data (vSL=0.2f t/ s)
This model (vSL=0.5f t/ s)
Exp.data (vSL=0.5f t/ s)
This model (vSL=1.0f t/ s)
Exp. dat a (vSL=1.0f t/ s)
This model (vSL=2.5f t/ s)
Exp. dat a (vSL=2.5ft/ s)
This model (vSL=5.0f t/ s)
Exp. dat a (vSL=5.0ft/ s)
0.1
0.1
10
100
v Sg/v SL
Figure 15. Sensitivity of Enhancement of Convective Heat
Transfer Coefficient to Gas/Liquid Superficial Velocity Ratio for
Vertical Intermittent Flow
1000
This model
900
OSU model
800
0%
700
+40%
-40%
600
500
400
300
200
100
0
0
100
200
300
400
350
300
250
200
150
100
50
0
0
50
100
150
200
250
300
350
400