Вы находитесь на странице: 1из 6

SPE 71111

Slug Flow Model for the Prediction of Pressure Drop for High Viscosity Oils in a
Horizontal Pipeline
Colmenares J., Ortega P., Padrino J., Trallero J.L.
PDVSA-INTEVEP, VENEZUELA

Copyright 2001, Society of Petroleum Engineers Inc.


This paper was prepared for presentation at the 2001 SPE International Thermal Operations
and Heavy Oil Symposium held in Porlamar, Margarita Island, Venezuela, 1214 March 2001.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

ABSTRACT
A modified mechanistic model is formulated to
predict the pressure drop in horizontal slug flow for
two-phase flow (viscous liquid and air). The model
is evaluated by using accurate PDVSA INTEVEP
experimental data for liquid with viscosity of 480
cP. A comparison between the modified model and
experimental data shows that the absolute average
relative error in pressure drop prediction is less than
6%.
INTRODUCTION
Venezuela has the world largest heavy oil
reserves. PDVSA has launched several projects to
develop the technology for optimum exploitation
and production schemes. Special attention have
been focused on multiphase flow along the
production system, which includes horizontal &
multilateral wells, vertical wells (tubing & annular
flow), pipelines and production networks.

Multiphase flow is characterized for the


existence of flow patterns. There are different types
of them, where the most common one is called slug
flow, see Fig 1. Therefore, proper production
system design requires of reliable pressure drop
models for slug flow. Current pressure drop models
for slug flow, have been developed and validated
for low viscosity oils. Fluid properties affect the
slug flow characteristics as well as the behavior of
the pressure losses. Available pressure drop models
estimate the pressure gradient with average errors
about 30% as can be seen in Fig 2. This uncertainty
might affect CAPEX and OPEX up to 10%.
The interest of this work is to develop a rigorous
pressure drop model that can be applied for both
light and heavy oils. The model should be validated
initially with lab data and then with field data. Due
to the lack of high quality laboratory data for
pressure drop in heavy oils, PDVSA INTEVEP
built a multiphase production laboratory.
The experimental facility and the slug flow
model will be described next.
EXPERIMENTAL SETUP
Test facility description
Experiments were carried out in a 2-in test loop
facility at PDVSA INTEVEP. Lube oil (480 cP)
and air were testing fluids.

COLMENARES J., ORTEGA P., PADRINO J., TRALLERO J.L.

Figure 3 shows a diagram of the experimental


facility. It consists of the following modules:
Liquid phase system handling
Gas phase system handling
Test section
Non conventional gas/liquid separation
Liquid was pumped using a gear pump with a
variable speed control. The liquid flow rate was
measured using a mass flow meter. The liquid flow
range is 0.4 - 40 m3/hr.
The air was supplied using two compressors.
The flow rate was controlled by three flow control
valves and measured with an orifice plate and a
vortex meter. The maximum gas flow rate is 1400
sm3/hr at 125 psig.
A regular Tee was used as gas injection point.
The test loop is 64-m long where the first 42-m is
the entrance region. The next 13-m are the test
section where pressure (gauge, differential @ 5 and
500 Hz & piezoelectric @ 500 Hz) and temperature
are measured. The final 9-m section is equipped
with a set of quick closing valves and a viewing
section.
A combination of high-speed video camera and
fast signal for pressure drop and wall pressure
fluctuations was used measure the slugs
characteristics.
At the end of the loop, the gas/liquid mixture
was separated in a non-conventional separator. The
oil is returned to the storage tank and the air is
vented.
Brief experimental observations
There are major differences between the slug
features in light oils and heavy oils. Liquid phase in
light oil is commonly in turbulent flow while in
heavy oil is mostly laminar. In heavy oils, the

SPE 71111

liquid height in the film region increases


notoriously. Translational velocity also increases
due to a larger pick up rate. Slugs are much more
frequent and shorter. The 30D assumption seems to
be no valid for heavy oil. Liquid holdup in the slug
body is relatively high.
It was observed that the larger the viscosity, the
larger the slug flow region in the flow pattern map.
Some flow pattern transitions will not be reached in
most of the practical application found in the
industry (Trallero et al. (2000)).
MODELING
In the last 30 years, industry, universities and
R&D centers has developed a rich variety of models
for slug flow as the ones published by the Duckler
and Taitel school (Dukler & Hubbard (1795), Xiao
et al. (1990), Taitel & Barnea (1990), etc). Even
though, the physical modeling has been improved
continuously, these models were only validated
with low viscosity data.
A good model should have a strong foundation
in physics and it must use a minimum of closure
relationships based on experimental data.
A
comprehensive review of all available slug flow
models indicates that Taitel & Barnea (1990) model
is the best candidate to be used. This model can be
extended to heavy oil applications. It is the most
rigorous treatment published for slug flow, as far as
we know. The model is based on mass balances on
the slug body and momentum balances on the slug
unit (see Fig. 1), as well as, closure relationships for
the slug length (Ls), the liquid holdup in the slug
body (HLLS), and the translational velocity (VTB).
Mass and momentum balances are independent
of the oil viscosity. Therefore, we will focus on
modifying the closure relationships of the original
model. Please notice that for the pressure drop
calculation, the model is insentive to the slug length

SPE 71111

SLUG FLOW MODEL FOR THE PREDICTION OF PRESSURE DROP FOR HIGH VISCOSITY OILS IN A HORIZONTAL PIPELINE

(Ls), because it cancels out in the mass and


momentum balance.
Taitel and Barnea propose three-different ways
or cases to treat the film region. Case 1 is the most
rigorous and general formulation for one
dimensional channel flow approximation. In case 2,
the film is treated as a free surface channel flow
where the pressure drop along the film zone is
neglected. Finally, case 3 assumes an equilibrium
film thickness but it takes into account the pressure
drop along the film zone. There was no major
difference between the case 1 and case 3 results
(Shoham (2000) and Taitel (2000)). Case 3 was
validated against experimental data (Ortega et al.
(2000)). Model output is the pressure gradient
across the slug unit.
Next it will be described the improvements in the
closure relationships of the Taitel & Barnea model.
Slug length
Original Taitel & Barnea model uses a constant
value of 30 diameters for the slug length (LS), as
was first reported by Nicholson in 1978. However,
during the experiments carried out at PVDSA
INTEVEP with 480-cp oil, it was noticed average
slugs length much shorter than 30D. Since the slug
length (Ls) cancels out in the mass and momentum
balance, it has no influence on the pressure gradient
prediction in this model. However, LS will affect
the calculation of the slug unit length and its
frequency. In a future work, a detailed calculation
of the slug length based on first principles combined
with correlated data will be given.
Translational velocity
The translational velocity, VTB is calculated
using an expression presented by Bendisken (1954):
VTB = C .V M + 0 .54 g .D

(1)

where, VM is the mixture velocity; C is a constant,


which depends on the liquid velocity profile in the

slug body. It takes a values of C=1.2 for turbulent


flow and C=2.0 for laminar flow. This expression
was validated experimentally, even for very low
Reynolds number.
Liquid holdup in the slug body
The liquid holdup in the slug, HLLS is estimated
using Gregory et al. (1978) correlation.
(2)
1
H LLS

1 + (V M / 8 . 66

)1 . 39

Equation 2 follows fairly well the hold up


measurements.
Unit length
The length of the slug unit (LU) is the sum of the
length of the slug (LS) and the film region (LF).
The film region extent is calculated in the model
using mass and momentum balances while the slug
length is estimated as it was mentioned before.
Slug frequency
The slug frequency , is calculated, using the
following relationship:

VTB
LU

(3)

where LU is the length of the slug unit.


RESULTS
Model performance and experimental results are
compared based on pressure gradient. In Fig. 4,
model results and experimental measurements for
the 480-cp oil/air system are put side by side. All
data is contained in a 10% error band. Similar
results were found if the Gomz et al. (2000)
correlation to estimated the liquid holdup in the slug
body is used. The same data was used to explore
the performance of the Beggs & Brill (1973)

COLMENARES J., ORTEGA P., PADRINO J., TRALLERO J.L.

correlation (see Fig. 2), where the error spread is


30% average. Definitely, the modified model has a
better performance. In table 1, the figures of the
relative and absolute errors are given.
CONCLUSIONS
In high viscosity gas-liquid flow it is necessary
to adjust the closure relationships of existing
models. Barnea & Taitel model seems to be the
best option to calculate the pressure drop in slug
flow for heavy oils. Modified model provided
results with in a 6% absolute average error in lab
conditions while current models are 30% off.
It was observed that slug frequency () and
liquid film height (hf) increases while the slug
length (Ls) decreases as the liquid viscosity rises.
NOMENCLATURE
DPU Pressure drop in the slug unit, [Pa/m]
C
Constant coefficient for flow regimen
D
Pipe diameter, [m]
Liquid holdup in the film zone
hF
HLLS Liquid holdup in the slug zone
Film length, [m]
LF
Slug length in the slug zone, [m]
LS
Length of the slug unit, [m]
LU
Mixture velocity, [m/s]
VM
VTB Translacional velocity, [m/s]
GREEK LETTERS

Slug frequency, [slugs/s]

SPE 71111

REFERENCES
1.- Trallero, J., Mata C., Pereira A., Ahow, M.,
Cabello, R. & Gonzlez, P. Estudio de la
Transiciones en Patrones de Flujo Bifsico Aireaceite en Tubera Horizontal, INT-8125- (2000).
2.- Dukler, A. E. & Hubbard, M. G., A Model for
Gas-liquid Slug Flow in Horizontal and Near
Horizontal Tubes, Ind. Chem. Fund., 14 pp 337347, (1975)
3.- Xiao, J., Shoham, O. & Brill J. P., A
comprehensive Mechanistic Model for Two-phase
Flow in Pipelines, SPE 20631,(1990).
4- Taitel, Y. & Barnea, D., Two Phase Slug Flow,
Academic Press Inc. (1990).
5. - Shoham, O. Private communications,
PDVSA-INTEVEP, July (2000).
6.- Taitel, Y. Private communications, PDVSAINTEVEP, July (2000).
7- Ortega, P., Colmenares, J., Padrino J., Trallero J.
Modelo para la Prediccin de la Cada de Presin
para Flujo Tapn en una Tubera Horizontal, INT8123, (2000).
8- Bendiksen, K., An experimental Investigation of
the Motion of Long Bubbles in Inclined Tubes. Int.
Journal Multiphase Flow, Vol. 10, No 4, pp 467483 (1984).
9- Gregory, G.A., M. K. Nicholson & K. Aziz,
Correlation of the liquid volume fraction in the
slug for horizontal gas-liquid slug flow. Int.
Journal Multiphase Flow, Vol. 4 pp 33-39 (1978).
10- Beggs, H. D.& Brill, J. P. Bendiksen, K., A
study of Two-phase Flow in Inclined Pipes.
Trans. AIME, pp. 607 (1973).

SPE 71111

SLUG FLOW MODEL FOR THE PREDICTION OF PRESSURE DROP FOR HIGH VISCOSITY OILS IN A HORIZONTAL PIPELINE

APPENDIX TABLES

Model
B&B
Modified

Relative Error
(%)
28
1

Absolute Error
(%)
30
6

Table 1. Pressure gradient model errors.

COLMENARES J., ORTEGA P., PADRINO J., TRALLERO J.L.

SPE 71111

APPENDIX FIGURES

P/dx PIPESIM-Beggs & Brill, [Pa/m]

20000

LU

LS

LF

G
VGTB

VLLS

VTB

i
VGLS

VLTB

hf

30%
B eggs & B rill

16000

12000

8000

4000

0
0

4000

8000

12000

16000

20000

P/dx Measured, [Pa/m ]

Figure 1. Schematic diagram for a slug unit.

TT PT
007 005

DPT FT
003 002

FT FT
001A 001B

PT TT
003 001

Figure 3. Pressure gradient comparison for Beggs and


Brill correlation (1973).

AIR

PT
010

LIQUID

1 80 0 0

PT
001

SEALES
RPIDAS

DPT
017

AT
011A

PT
011

DPT
011

FAST RESPONSE
SENSORS

DPT
013

AT
011B

DPT
018

PE

DPT

PE

PE

DPT

014

016

015

016

019

DPT

PE

014

013

AT
012A

AT
012B

DPT
012

DPT
016

PE AT AT
013 012A 012B

AT
013A

TO SEPARATION

AT
013B

PT
015

LEYEND
FT FLOW TRANSMITER
PT PRESSURE TRANSMITER

DPT
015

DPT DPT PE
017 018 014
AT AT DPT
011A011B 014

PT
012

TT
012

DPT
015

TT TEMPERATURE TRANSMITER
PE PE
015 016

DPT
019
AT AT
013A 013B

P/dx Modified Model, [Pa/m]

M o d ifie d M o d e l

TT
011

1 50 0 0
10%

1 20 0 0

9 00 0

6 00 0

3 00 0

DPT PRESSURE DROP TRANSMITER


AT HOLD UP TRANSMITER

0
0

PE DINAMIC PRESSURE TRANSM.

3 00 0

6 00 0

9 00 0

1 20 0 0

1 50 0 0

P /d x M e a s u re d , [P a /m ]

Figure 2. System setup for the test loop.


Figure 4.
model.

Pressure gradient comparison for modified