Professional Documents
Culture Documents
by
Mutaz A. Daas
August, 2001
OHlOUNIVERStTY
LIBRARY
2001
Mutaz A. Daas
All Rights Reserved
111
TABLE OF CON-TENTS
LIST OF FIGURES
vi
NOMENCLATURE
xiv
CHAPTER ONE
INTRODUCTION
CHAPER TWO
LITERATURE REVIEW
15
17
39
CHAPTER THREE
EXPERIMENTAL SETUP AND TEST MATRIX
51
51
54
55
CHAPTER FOUR
MODELING PROCEDURE
58
58
62
64
65
iv
CHAPTER FIVE
RESULTS AND DISCUSSION
68
5.1 Quantitative Analysis of the Flow of2.5 cP Oil in Horizontal Pipes .....68
5.1.1 The Components of the Pressure Drop
69
78
82
83
89
97
97
u
l05
5.4 Quantitative Analysis of the Flow of 50 cP Oil in Horizontal Pipes .... 109
5.4.1 The Components of the Pressure Drop
110
116
120
120
140
CHAPTER SIX
CONCLUSIONS
149
BIBLIOGRAPHy
156
159
172
ABSTRACT
186
vi
LIST OF FIGURES
Figure 1.1:
Figure 1.2:
Figure 1.3:
Unit Slug
Figure 1.4:
10
Figure 1.5:
12
12
Figure 2.1:
19
Figure 2.2:
Fan et al. (1993) Definitions for Variables for a Slug in the Reference
Frame Moving With Velocity C F ................................ 26
Figure 2.3:
37
40
Figure 3.1:
System Setup
52
Figure 4.1:
Computational Procedure
68
Figure 1.6:
Figure 2.2.1 :
Figure 5.1.1.1: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 0.5 mis, 0 ppm DRA, Horizontal Flow
71
Figure 5.1.1.2: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1 mis, 0 ppm DRA, Horizontal Flow
71
Figure 5.1.1.3: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1.5 mis, 0 ppm DRA, Horizontal Flow
72
vii
Figure 5.1.1.4: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 0.5 mis, 20 ppm DRA, Horizontal Flow
73
Figure 5.1.1.5: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1 mis, 20 ppm DRA, Horizontal Flow
73
Figure 5.1.1.6: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1.5 mis, 20 ppm DRA, Horizontal Flow
74
Figure 5.1.1.7: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 0.5 mis, 50 ppm DRA, Horizontal Flow
74
Figure 5.1.1.8: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1 mis, 50 ppm DRA, Horizontal Flow
75
Figure 5.1.1.9: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1.5 mis, 50 ppm DRA, Horizontal Flow
75
Figure 5.1.2.1: Effects ofDRA on Frictional Pressure Drop, 1OOO~ Oil (2.5 cP),
Vsl = 0.5 mis, Horizontal
80
Figure 5.1.2.2: Effects ofDRA on Frictional Pressure Drop, 100% Oil (2.5 cP),
Vsl = 1 mis, Horizontal
80
Figure 5.1.2.1: Effects of DRA on Frictional Pressure Drop, 100% Oil (2.5 cP),
Vsl = 0.5 mis, Horizontal
81
Figure 5.1.2.3: Effects ofDRA on Frictional Pressure Drop, 100% Oil (2.5 cP),
Vsl = 1.5 mis, Horizontal
81
40
Figure 5.1.2.4: Effects ofDRA on Acceleraional Pressure Drop, 100% Oil (2.5 cP),
82
Vsl = 0.5 mis, Horizontal
Figure 5.1.2.5: Effects ofDRA on Acceleraional Pressure Drop, 100% Oil (2.5 cP),
Vsl = 1 mis, Horizontal
82
Figure 5.1.2.6: Effects ofDRA on Acceleraional Pressure Drop, 100% Oil (2.5 cP),
Vsl = 1.5 mis, Horizontal
83
Figure 5.2.1.1: Pressure Drop Vs. Superficial Gas Velocity, 1000/0 Oil (2.5 cP),
Vsl = 0.5 mis, 0 ppm DRA, 2-Degree Upward
86
viii
Figure 5.2.1.2: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
VsI = I mis, 0 ppm DRA, 2-Degree Upward
86
Figure 5.2.1.3: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
VsI = 1.5 mis, 0 ppm DRA, 2-Degree Upward
87
Figure 5.2.1.4: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
VsI = 0.5 mis, 20 ppm DRA, 2-Degree Upward
87
Figure 5.2.1.5: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1 mis, 20 ppm DRA, 2-Degree Upward
88
Figure 5.2.1.6: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1.5 mis, 20 ppm DRA, 2-Degree Upward
88
Figure 5.2.1.7: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
VsI = 0.5 mis, 50 ppm DRA, 2-Degree Upward
89
Figure 5.2.1.8: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
VsI = 1 mis, 50 ppm DRA, 2-Degree Upward
89
Figure 5.2.1.9: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (2.5 cP),
Vsl = 1.5 mis, 50 ppm DRA, 2-Degree Upward
90
Figure 5.2.2.1: Surface of Liquid Film Before and After the Addition ofDRA
91
Figure 5.2.2.2: Effects ofDRA on Frictional Pressure Drop, 100% Oil (2.5 cP),
Vsl = 0.5 mis, 2-Degree Upward
92
Figure 5.2.2.3: Effects ofDRA on Frictional Pressure Drop, 100% Oil (2.5 cP),
VsI = 1 mis, 2-Degree Upward
93
Figure 5.2.2.4: Effects ofDRA on Frictional Pressure Drop, 100% Oil (2.5 cP),
VsI = 1.5 mis, 2-Degree Upward
94
Figure 5.2.2.5: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (2.5 cP),
VsI = 0.5 mis, 2-Degree Upward
95
Figure 5.2.2.6: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (2.5 cP),
95
VsI = 1 mis, 2-Degree Upward
ix
Figure 5.2.2.7: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (2.5 cP),
96
Vsl = 1.5 mis, 2-Degree Upward
Figure 5.2.2.8: Effects ofDRA on Gravitational Pressure Drop, 1000/0 Oil (2.5 cP),
Vsl = 0.5 mis, 2-Degree Upward
97
Figure 5.2.2.9: Effects ofDRA on Gravitational Pressure Drop, 100% Oil (2.5 cP),
Vsl = 1 mis, 2-Degree Upward
97
Figure 5.2.2.10: Effects ofDRA on Gravitational Pressure Drop, 1OO~/~ Oil (2.5 cP),
98
Vsl = 1.5 mis, 2-Degree Upward
Figure 5.3.1.1: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 0.5 mis, 0 ppm DRA, 2-Degree Upward
99
Figure 5.3.1.2: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 1 mis, 0 ppm DRA, 2-Degree Upward
100
Figure 5.3.1.3: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 1.5 mis, 0 ppm DRA, 2-Degree Upward
100
Figure 5.3.1.4: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
VsI = 0.5 mis, 10 ppm DRA, 2-Degree Upward
102
Figure 5.3.1.5: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 1 mis, 10 ppm DRA, 2-Degree Upward
102
Figure 5.3.1.6: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 1.5 mis, 10 ppm DRA, 2-Degree Upward
103
Figure 5.3.1.7: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 0.5 mis, 50 ppm DRA, 2-Degree Upward
103
Figure 5.3.1.8: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 1 mis, 50 ppm DRA, 2-Degree Upward
104
Figure 5.3.1.9: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (26 cP),
Vsl = 1.5 mis, 50 ppm DRA, 2-Degree Upward
104
Figure 5.3.2.1: Effects ofDRA on Frictional Pressure Drop, 100% Oil (26 cP),
Vsl = 0.5 mis, 2-Degree Upward
106
Figure 5.3.2.2: Effects ofDRA on Frictional Pressure Drop, 100% Oil (26 cP),
Vsl == 1 mis, 2-Degree Upward
106
Figure 5.3.2.3: Effects ofDRA on Frictional Pressure Drop, 100% Oil (26 cP),
Vsl == 1.5 mis, 2-Degree Upward
107
Figure 5.3.2.4: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (26 cP),
VsI == 0.5 mis, 2-Degree Upward
108
Figure 5.3.2.5: Effects ofDRA on Accelerational Pressure Drop, 1000/0 Oil (26 cP),
Vsl == 1 mis, 2-Degree Upward
109
Figure 5.3.2.6: Effects ofDRA on Accelerational Pressure Drop, 1000/0 Oil (26 cP),
Vsl == 1.5 mis, 2-Degree Upward
109
Figure 5.3.2.7: Effects ofDRA on Gravitational Pressure Drop, 1000/0 Oil (26 cP),
Vsl == 0.5 mis, 2-Degree Upward
110
Figure 5.3.2.8: Effects ofDRA on Gravitational Pressure Drop, 100% Oil (26 cP),
Vsl == 1 mis, 2-Degree Upward
111
Figure 5.3.2.9: Effects ofDRA on Gravitational Pressure Drop, 100% Oil (26 cP),
Vsl == 1.5 mis, 2-Degree Upward
111
Figure 5.4.1.1: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
Vsl == 0.5 mis, 0 ppm DRA, Horizontal Flow
113
Figure 5.4.1.2: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
VsI == 1 mis, 0 ppm DRA, Horizontal Flow
114
Figure 5.4.1.3: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
VsI == 1.5 mis, 0 ppm DRA, Horizontal Flow
114
Figure 5.4.1.4: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
VsI == 0.5 mis, 20 ppm DRA, Horizontal Flow
115
Figure 5.4.1.5: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
VsI == 1 mis, 20 ppm DRA, Horizontal Flow
116
Figure 5.4.1.6: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
Vsl == 1.5 mis, 20 ppm DRA, Horizontal Flow
116
Xl
Figure 5.4.1.7: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
VsI = 0.5 mis, 50 ppm DRA, Horizontal Flow
117
Figure 5.4.1.8: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
Vsl = 1 mis, 50 ppm DRA, Horizontal Flow
117
Figure 5.4.1.9: Pressure Drop Vs. Superficial Gas Velocity, 100% Oil (50 cP),
Vsl = 1.5 mis, 50 ppm DRA, Horizontal Flow
118
Figure 5.4.2.1: Effects ofDRA on Frictional Pressure Drop, 100% Oil (50 cP),
VsI = 0.5 mis, Horizontal Flow
119
Figure 5.4.2.2: Effects ofDRA on Frictional Pressure Drop, 100% Oil (50 cP),
Vsl = 1 mis, Horizontal Flow
119
Figure 5.4.2.3: Effects ofDRA on Frictional Pressure Drop, 100% Oil (50 cP),
VsI = 1.5 mis, Horizontal Flow
120
Figure 5.4.2.4: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (50 cP),
VsI = 0.5 mis, Horizontal Flow
121
Figure 5.4.2.5: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (50 cP),
121
VsI = 1 mis, Horizontal Flow
Figure 5.4.2.6: Effects ofDRA on Accelerational Pressure Drop, 100% Oil (50 cP),
122
Vsl = 1.5 mis, Horizontal Flow
Figure 5.5.1.1: Effects of Viscosity on Total Pressure Drop, 100% Oil (2.5 & 26 cP),
124
VsI = 0.5 mis, 2-Degree Upward
Figure 5.5.1.2: Effects of Viscosity on Total Pressure Drop, 100% Oil (2.5 & 26 cP),
VsI = 1 mis, 2-Degree Upward
125
Figure 5.5.1.3: Effects of Viscosity on Total Pressure Drop, 100% Oil (2.5 & 26 cP),
125
VsI = 1.5 mis, 2-Degree Upward
Figure 5.5.1.4: Effects of Viscosity on Accelerational Pressure Drop, 100% Oil
(2.5 & 26 cP), Vsl = 0.5 mis, 2-Degree Upward
127
127
xii
128
130
130
100~/o
Oil
131
131
134
134
Figure 5.5.2.1: Effects of Viscosity on Total Pressure Drop, 100% Oil (2.5 & 50 cP),
VsI = 0.5 mis, Horizontal
136
Figure 5.5.2.2: Effects of Viscosity on Total Pressure Drop, 1000/0 Oil (2.5 & 50 cP),
VsI = 1 mis, Horizontal
136
Figure 5.5.2.3: Effects of Viscosity on Total Pressure Drop, 100% Oil (2.5 & 50 cP),
VsI = 1.5 mis, Horizontal
137
Figure 5.5.2.4: Effects of Viscosity on Accelerational Pressure Drop, 1000/0 Oil
(2.5 & 50 cP), VsI = 0.5 mis, Horizontal
138
138
139
140
xiii
141
141
Figure 5.6.1 :
143
144
144
146
146
148
150
150
151
Figure 5.6.2:
Figure 5.6.3:
Figure 5.6.4:
Figure 5.6.5:
Figure 5.6.6:
Figure 5.6.7:
Figure 5.6.8:
Figure 5.6.9:
xiv
NOMENCLATURE
cross-sectional area, m 2
diameter, m
friction factor
Fr
hEFF
hI
I, L
R =
Re
Froude number
local gravity, mls2
effective height of liquid film, m
height of liquid film, m
length, m
liquid holdup
Reynolds number
perimeter length, m
VI
Vs
Vsg
VsI
Vt
density, kg/rrr'
xv
Vs
L1L
LiP
pressure drop, Pa
Subscripts
a =
accelerational
body
slug body
.DRA
withDRA
f
fe
film
frictional
liquid film just prior to pickup
liquid film
gas
gravitational
gas-liquid
L
MZ
NoDRA
liquid
mixing zone
withoutDRA
s, slug
slug
total
WG
wall-gas
WL
wall-liquid
1
CHAPTER ONE
INTRODUCTION
In remote areas such as sub-sea installations, the production of oil and natural
gas involves multiphase flow. Maturing oil wells produce more seawater and natural
gas, which consists of mostly methane and carbon dioxide, than new wells. With aging,
the oil well pressure depletes. At this point of the oil well life, natural gas together with
all gases present in the production process are injected into the reservoir gas cap (the
region above the oil phase) or into the water region (below the oil phase) as shown in
Figure 1.1. As the reservoir pressure drops further, this gas re-injection is not sufficient
to maintain reservoir pressure; thus water in injected to replace the removed crude oil.
At this point an increasing amount of water sweeps into the well from the higherpressure surroundings to replace the exiting crude oil. Thus, maturing oil wells usually
produce higher water cut (volumetric percent of water). Another typical source of gas
is enhanced recovery processes. As oil well ages, enhanced oil recovery is applied by
injecting gas into the well tubing down the pipeline. This injection of gas reduces the
density of the mixture, thus lowering the pressure required to lift the liquid. As this
mixture moves upward toward the wellhead it loses pressure, more gas flash out from
the mixture and gas pockets grow due to compressibility as the gas density decreases.
The most frequent flow patterns that may take place in two-phase oil-gas and
multiphase oil-water-gas flows are illustrated in Figure 1.2. These flow patterns are
categorized in three groups; stratified, intermittent, and annular.
Gas
Oil
Water
.,.",
.
./
.,.",
0''''''''''''
.,.",
o'
.",0/ .
.. ~:"'O
.//
./
-- ------------
---
------~
""
""
""
",
"
Flow Direction
----~~
Gas
Oil
Water
Smooth Stratified
Wavy Stratified
Rolling Wave
Plug
Slug
... . ,...
, ".' ..., ., ..
"
.'
'.
Annular
4
At low liquid and gas velocities, the three phases flow in smooth stratified
pattern with the water flowing on the bottom, gas flowing along the top, and oil flowing
in between water and gas phases. Two-dimensional waves at the gas-oil interface start
to propagate as a result of increasing gas velocity, while the three phases still remain
stratified. Further increase in gas velocity causes these waves to grow and rollover
leading to rolling wave regime.
If the overall liquid flow rate is increased, plug flow is reached. The three-phase
plug flow is characterized by a wavy, yet mostly stratified interface between the oil and
water phases.
The oil level in plug flow reaches the top of the pipe with regular
intermittent gas pockets passing, which remove the oil from the top of he pipe. (Robert
Wilken, 1997)
An increase in gas velocity in the plug regime will result in slug flow. Jepson
(1989) proposed new transition criteria for slug flow. Rather than wave growth to reach
slug flow from stratified flow, Jepson proposed that a slug was a hydraulic jump which
was propagating down the pipe. A hydraulic jump takes place when there is a transition
from subcritical to supercritical flow. The Froude number is usually used to measure
the turbulence of hydraulic jumps. A flow that has a Froude number less than unity is
considered subcritical, while supercritical flow has Froude number greater than unity.
There are many forms of Froude numbers. In this study the film Froude number will be
used. This is the ratio of the velocity difference between the liquid film and the jump to
the square root of the film height times the gravity.
(1.1)
Visual observations indicate complete mixing between oil and water phases in
slug flow. The slug front travels with a velocity many times greater than the pocket of
gas ahead of it resulting in gas penetration into the slug front.
front travels faster than the liquid film ahead of it, thus the slug front picks up and
accelerates this slow moving film to slug velocity. The force required for pick up and
acceleration manifests itself as accelerational pressure drop. The velocity of the liquid
within the slug body is less than slug translational velocity, and this velocity decreases
thus liquid sheds at the rear of the slug in an equal rate of liquid pickup at the front. A
schematic of a unit slug is shown in Figure 1.3.
L MZ
IF'll<n>w:EE:>
i,.,,,,,,,,,,,.,.'.,,,,,,,:,,,:.".,..
,.,...
,.,.,.:.:.:.:.:.,..
Each slug constitutes of four regions as shown in Figure 1.3. These regions are
highly turbulent froth zone known as the mixing zone, slug body with gas bubbles
entrained in, slug tail that results from liquid shedding, and stratified liquid film
between the tail of this slug and the successor aIle. A further increase in gas velocity
results in a similar flow pattern called pseudo slug. Pseudo slug is shorter and frothier
with more gas content. In pseudo slug flow pattern, the liquid in the body never fully
bridge the pipe cross section. Since pseudo slug flow occurs as a result of increasing
gas velocity, then mixing region increases until it exceeds the length of slug body.
Further increase in gas velocity results in an annular flow. At high velocities,
where gas is entrained in the body of the slug, the gas does 110t distribute
homogeneously in the slug body but is collected mostly in the upper half of the pipe.
Consequently, as the gas velocity is increased, the upper part of the slug body becomes
frothier and eventually blow-through occurs leading to the transition to annular flow
(Jepson, 1993).
Annular flow is characterized as a less dense fluid (gas) flowing in a core along
the center of the pipe while the more dense fluid (oil/water) flowing as an annular ring
around the pipe wall.
thickness of the liquid film at the bottom of the pipe is greater than at the top due to
gravity.
Pressure loss in single-phase flow occurs due to friction between the flowing
fluids and pipe wall.
pressure drop is achieved along a pipeline of smaller diameter. Fluid density has less
influence on pressure drop except in inclined pipes where the flowing fluid loses part of
it momentum due to gravity force. Therefore, fluids of greater density show more
momentum loss than fluids of lower density in inclined pipes at the same flowing
conditions.
Pipe geometry and roughness plays an important rule in determining pressure
loss in single-phase flow. Acid gases such as hydrogen sulfide and nitrogen are present
along with carbon dioxide and natural gas in the gas phase in oil pipelines.
The
Frictional loss occurs in both the slug body and the stratified liquid film between
each two successive slugs. Pressure loss due to friction in slug body is a strong function
of slug characteristics; more gas content in slug body means less dense flowing fluids
and hence less pressure loss, a longer slug body is always accompanied by greater
pressure loss than shorter slugs. Frictional pressure loss in slug body is proportional to
the square of slug velocity. Finally, slug frequency results in more slugs and multiplied
amounts of total pressure loss.
Height of stratified liquid film between two successive slugs determines the area
of the pipe occupied by the liquid and interfacial area between liquid and gas. Pipe
wall-liquid and liquid-gas friction are the main sources of pressure loss in stratified
liquid film in slug flow. Gas velocity has significant influence on the shape of the gasliquid interface. At low gas velocities smooth stratified presents with low relatively
pressure loss. Increasing gas velocity causes the formation of interfacial waves at the
gas-liquid interface. These waves increase in size and degree of turbulence as gas
velocity is increased. Gas-liquid interfacial friction factor increases dramatically with
the presence of interfacial waves resulting in greater amounts of frictional pressure loss.
It is important to mention that for the flow of low viscosity oil, accelerational
component of pressure drop is dominant and frictional loss is considered minor
contributor to total pressure drop.
Gravitational contribution to total pressure drop occurs in inclined pipes due to
gravity.
Fluid density is the main factor that determines the magnitude of this
contributor. Therefore, as gas velocity is increased higher void fractions are expected in
Sllig body. This results in less dense slugs and lower gravitational pressure loss. Gasliquid surface tension determines the degree of mixing between the two phases. Lower
surface tension between the gas and liquid requires smaller force to mix the two phases
resulting in greater gas content with in the liquid bulk.
The drag reduction phenomenon was discovered accidentally by Toms (1947),
W}lO
reported
that
very
dilute
solutions
of polymethyl
methacrylate
In
10
thixotropic solution. Depending on its application and the nature of treated oil-water
mixtures, DRA's are either oil-soluble or water-soluble. When oil percent is high in the
liquid phase then oil-soluble DRA is to be injected into the pipeline, otherwise watersoluble DRA is to be used.
Drag reducing agents found many applications in oil and gas industry. They
made it possible to design and utilize smaller pipes when bringing an oil field to
production. Figure 1.4 describes the production rate of an oil field with time. As new
oil wells are brought on line in the first few years of the oil field lifetime, peak
production is usually expected. Production rate decreases with time until it reaches the
economic level. Pipelines are usually designed to utilize the maximum production,
which last only for few years. Instead, DRA's can be injected into the pipeline to
increase its capacity; by decreasing pressure loss along the pipeline, without further
mechanical modifications at the peak period, hence smaller pipes can be used and
significant capital savings can be achieved.
PRODUCTION RATE
11
When higher throughput is required at certain time of an oil field lifetime, drag
reducing agents offer less expensive alternative. Neither boosters nor pump stations or
larger pipes are to be in.stalled along the pipeline. DRA's offer an attractive way to
utilize greater production rates without additional large capital expenditures since the
polymer system is compact and transportable. If the system is no longer needed, no
costly pipelines or pumps have to be dismantled.
Experimental studies indicate that drag reducing agents not only reduced
pressure loss in all flow patterns but also helped the transition from high turbulent flow
patterns to lower and smoother ones. For example, DRA's increased the liquid velocity
required for the transition from stratified flow pattern to the highly turbulent slug flow
by decreasing the height of the liquid film, hence more liquid was needed to bridge the
pipe cross section which can be reached at higher liquid velocities. Figures 1.5 and 1.6
show the effect ofDRA on flow regime transitions.
The use of DRAs has been considered the most successful application of drag
reducing agents in Trans-Alaska pipeline system. Little or no loss of effectiveness was
reported over pipeline lengths of hundreds of miles. Polymer dosages in the order of 10
ppm not only increased production rates significantly, but also made it possible to stop
construction in two pumping stations along the pipeline.
12
~
--..
E
2
1
SkJg
stratified
0.1
1
7 8 9 10
2
Plig
1.L
Sug
stratified
0.1
1
7 8 9 10
13
Applications of DRA's in Crude-oil pipelines are increasing. Gasoline, diesel
fuel, and fuel-oil pipelines are also possible fields of drag reduction applications. Field
trails indicate excellent drag reduction, together with substantial operating economies.
Five pump stations in Iraq-Turkey (1-T) crude oil pipelines have been in operation. The
throughput increased from 0.7 to 1.0 million barrels of oil per day when 40---50 ppm
DRA were injected at each pump station (J. F. Motier et aI., 1989).
Applications of drag reduction in free surface flows include the increase of the
capacity of sewer systems during peak flow periods and the reduction of the water level
during flood periods in natural streams. The need for the construction of larger sewer
systems, at least in the short term, may be avoided.
Drag reducing agents find many applications In the military.
The most
14
reduce pressure loss in both accelerational and gravitational components as well due to
the changes they cause in flow characteristics.
A quantitative analysis of total pressure drop and its components in slug flow is
introduced in this study. Results from computational studies corresponded very well
with experimental data. Current results would be helpful in establishing a quantitative
understanding of the contribution of each component to total pressure loss in slug flow
and the effect of both oil viscosity and pipe inclination on these components. This work
gives also a quantitative description of drag reduction achieved in each component. A
quantitative presentation of these components is important to understand flow
phenomena associated to slug flow.
Quantitative analysis of pressure drop will help introduce new mechanisms for
drag reductions in multiphase flow.
15
CHAPTER TWO
LITERATURE REVIEW
Few attempts have been carried out to predict pressure drop is slug flow. Oil
viscosity may change significantly due to changes in temperature. Pipe inclination also
experience frequent changes due to irregularity in land topography. A small number of
these works have studied the effect of oil viscosity and pipe inclination on each
component of slug pressure drop. Moreover, literature is short of any computational
analysis of pressure drop in slug flow with the presence of drag reducing agents.
Therefore, there is a lack in understanding the effect of drag reducing agents in reducing
growing need to study the influence of oil viscosity and pipe inclination on the
contribution of each component to total pressure drop and the performance ofDRA.
This work is the first attempt of its kind to predict pressure drop in slug flow
with the presence of DRA. A quantitative analysis and experimental work have been
carried out to estimate each component of the pressure drop in oil-gas slug flow.
Experimental results corresponded very well with calculated values. Similar work have
been undertaken after the addition of DRA to study the effects of DRA on each
component of the pressure drop and to find out where most of the drag reduction came
from.
Different kinds of oil of markedly different viscosities were tested in an attempt
to correlate variations in oil viscosity with the magnitude and contribution of each
16
component to total pressure drop. The system used in this study has been set to two
inclinations; namely horizontal and 2-degree. Measurements of pressure drop under
slugging conditions were taken for the two different inclinations.
Again, these
measured values agreed very well with calculated ones. A clear understanding of the
in.fluence of pipe inclination on the contribution of each component to total pressure
drop and the effectiveness of the DRA in reducing each component individually and
overall pressure drop was approached.
It is important to point out to some technical and experimental issues that one
should be aware of when referring or using data developed by other workers or when
running experiments under slugging conditions. The process of slug flow is a highly
complex unsteady phenomenon because the special distribution of the two phases
(liquid and gas) is difficult to specify quantitatively. As a result, processes of heat and
mass transfer are also unsteady with substantial fluctuations in temperature and
concentration.
For the measured data to be meaningful and of practical importance, slug length
should not exceed the distance between pressure taps. Moreover, the distance between
pressure taps should not be filled only with air space. Readings under such conditions
never reflects reality and deviates significantly from calculated values. It is important
to have the test section a distance down the stream enough to achieve fully developed
slug flow.
questionable. Slug flow in capillary or small pipes is equivalent to bubble flow with
17
Taylor type bubbles having bullet shaped noses separated by liquid plugs moving in
laminar flow.
~Pf,
shear in the back section of the slug. The total pressure drop across a slug is thus
(2.1)
Pressure drop in the gas phase above the liquid film was neglected in their
model. This pressure drop in the gas phase above the liquid film was taken into account
and estimated in the current study.
A slug that has stabilized in length can be considered as a body receiving and
losing mass at equal rates. The velocity of the film just before pickup is lower than that
in the slug and a force is then necessary to accelerate this liquid to slug velocity. This
force manifests itself as pressure drop given by (Hubbard & Dukler, 1975):
18
(2.2)
It is evident that accelerational component is a strong function of both slug
velocity and liquid film velocity.
Behind the mixing zone in the body of the slug pressure drop takes place due to
wall friction. For the calculation of this term, the similarity analysis for single-phase
frictional pressure drop developed by Dukler and others in 1964 was applied. Within
this part of the slug the two phases were assumed to be homogeneously mixed with
negligible slip.
becomes:
(2.3)
The similarity analysis showed that for "non-slip" conditions the slug friction
factor,
[slug,
Re
slug
PL R +PG(I-R s )
=DV - -s - - - s
R s)1G
+ (1- R s )
)1L
(2.4)
19
Refer to Figure 2.1. Consider a slug whose outline is designated by the solid
line and whose front is located at the plant A-A at a specified instant in time. In the
interval Lit, the front of the slug shown dotted moves to plane C-C. The film formerly
located at A-A, moves only to B-B because of its lesser velocity. The amount of film
sh.own crossed-hatched is picked up and mixed into the slug. Thus the mass rate of
pickup is calculated using the following equation:
(2.5)
................................................................
Vfe
--7
ABC
Figure 2.1 Dukler and Hubbard (1975) Schematic Diagram of a Unit Slug
In their study, Dukler & Hubbard discarded the pressure drop that was
developed from the stratified liquid film behind each slug. Results obtained from the
computational analysis of the current work show that the contribution of the frictional
pressure drop in the stratified liquid film behind a slug to total pressure drop is minimal
for the flow of low-viscosity oil. However, as oil viscosity increased the frictional
component became significant in value and could not be omitted.
20
Petalas and Aziz (1996) developed a new model that can be used for all pipe
geometries and fluid properties.
computer program in that a significant number of calculations are required and several
of theses require iterative procedures. A large amount of database has been used to
develop empirical correlations.
These
empirical correlations were required even when a mechanistic approach was used.
Many mechanistic models developed by other workers were utilized in their study.
These mechanistic models are based on fundamental laws and thus can provide for
more accurate modeling of the geometric and fluid property variations, unlike empirical
correlations.
Most of the mechanistic models used in their study began by assuming that a
particular flow regime was present. By solving the momentum balance equations for
certain quantities that determine its characteristics, the stability of the flow pattern was
examined.
If the chosen flow pattern was shown to be stable, the procedure was
terminated, the pressure drop and phase volume fractions being obtained directly from
the momentum balance equations.
specified conditions, a new flow pattern was assumed and the procedure was repeated
until a stable flow pattern was detennined.
The basic flow patterns considered in their approach are:
Stratified smooth
Stratified wavy
21
Bubble
Dispersed bubble
Froth or churn
incompressible flow and that the film thickness is uniform. Writing an overall liquid
mass balance over a slug bubble unit resulted in the following equation:
(2.6)
Where R L is the total liquid holdup in a unit slug, V Gdb represents the velocity of
the dispersed bubbles, V, is the translational velocity of the slug, and R, is the liquid
volume fraction in the slug body. All of these quantities were determined by empirical
correlations.
In their model, Petalas and Aziz calculated the liquid volume fraction in the slug
body based on the Gregory et al. correlation:
22
1
(2.7)
J1.39
_m_
V
1+
( 8.66
It is
approximately equal to 1.2 and approaches 2 as the flow become laminar. The bubble
drift velocity, Vb, was calculated utilizing Zuboski correlation:
(2.9)
s.;
(2.10)
Bendiksen gave the bubble drift velocity at high Reynolds numbers as:
(2.11)
23
b h :
]
= 0.54- B1.76
o.56
o
(2.12)
(p - PG)
L
gD
b v oo
(2.13)
PL
Finally, the velocity of the dispersed bubbles in the liquid slug was obtained
from the correlation of Ansari:
(2.14)
(2. I 5)
24
Carrying out momentum balance over a slug-bubble unit resulted In the
following equation:
(2.16)
The frictional pressure drop in the gas bubble is small compared to that in the
slug body. Xiao et al. have modeled this part of intermittent flow by assuming it to be
analogous to stratified flow.
. + [(R
- ( -dP ) -_ Pm g SIn
1]
L
dL
D PL Vm
(2.17)
ilL
taken as 1.
Equation (2.17) shows that the pressure drop constituted of the frictional loss in
the slug body and the gravitational loss due to gravity.
accelerational component of pressure drop was obviously omitted in this model that
makes it questionable.
25
Fan, Ruder, and Hanratty (1993) achieved another approach to predict pressure
drop in slug flow. In their model, the pressure drop over a slug was considered as being
composed of three parts. A diagram of a slug moving with a velocity of CF is shown in
Figure 2.2 below.
Cr
4
0
0
0
1
I
,:CF
I
I
I
I
I
,,
,
I
~UG3-CF
,
I
I
I
I
I
~UL3-CF
I
I
I
UGI-C F
~ UL1-C F
Figure 2.2 Definitions for variables for a slug in the reference frame moving with
velocity C F, Fan et al. (1993)
A new contributor to total pressure drop in slug flow was defined and suitable
equations to estimate such contributor were developed from mass and momentum
balances over a unit slug. A comprehensive understanding of the behavior of individual
slugs and the prediction of the number of slugs in a given length of pipe were essential
in developing their model to predict pressure drop in intermittent flow.
The change associated with the hydraulic jump in front of the slug from station 1
to 3,
~Ph;
~Pf;
and a
26
pressured variation associated with the velocity change at the rear of the slug from
stations 4 to 5, ~Pr, as shown in Figure 2.2.
Thus the total pressure change in horizontal pipes was given as:
(2.18)
The front of the slug is traveling with a velocity of CF Gas and liquid velocities
at locations 1, 3 and 4 were assumed uniform and represented by
ULi
and
UGi
respectively. The liquid film ahead of the slug is air-free and travels with a velocity of
ULl
ill laboratory framework. The flows of the liquid and gas in the slug body were
assumed uniform after a short distance behind the slug front and had the same
velocities,
UG3
= UL3.
Fan et al. (1993) classified slugs into two categories; stable and unstable. A
stable slug is a slug that has a constant length during its lifetime. This requires equal
rates of liquid pickup and liquid discharge at the front and rear of the slug, respectively.
Correspondingly, a slug that grows or decays during its lifetime is considered as an
unstable slug.
For a stable slug, consider a reference frame moving with velocity C F .
control volume was chosen from stations 1 to 3. The velocities in and out of the slug at
stations 1 and 3 had negative values, since a positive velocity was defined in the
direction in which the slug I traveling. These velocities were designated by -(C F and -(C F -
UGi).
ULi)
27
Conservation of mass for the liquid between stations 1 and 3 gives:
(2.19)
(2.20)
Where
ALl
is the area of the liquid film ahead of the slug, A 3 is the pipe cross-
section area, R A is the rate at which gas is occluded into the slug, and
&3
-U Gl )2(A 3 -
-U G3)2A 3&3
(2.21 )
Where PI and P3 are the pressures at the top of the pipe, PL and Po are the
densities of liquid and gas,
hZ, hg
and
h;
hydrostatic pressure in the liquid layer, the gas pocket in the front of the slug and the
slug body.
28
Substituting (2.19) and (2.20) into (2.21) gives the pressure drop caused by the
hydraulic jump (accelerational component):
(2.22)
Where L1Ph = P, - Pl. The last two terms in (2.22) are very small compared to
the first two because PG is a small number.
PL
For a stable slug the velocity of the tail, C T , is equal to C F The following
assumption was made to approximate the pressure change at the rear of the slug. The
rear may be assumed as Benjamin bubble. This requires that the top of the slug tail is a
stagnation point and that the pressure change associated with the acceleration along the
top wall can be calculated from Bernoulli equation. Because surface tension effects are
negligible, the pressure of the liquid at the top of the tails equals the gas pressure, P6.
Thus a pressure increase is calculated between stations 4 and 6 as
(2.23)
Where UL4
= ULj.
Fan et al. considered an unstable slug to be growing when its length was
increasing with time, CF > CT. Again, if station 1 moved with velocity C F and stations 3
and 4 were chosen to move with a velocity C T, then Bernoulli equation could be used
between stations 4 and 5. The control volume between stations 1 and 3 extends in
29
length at an equal rate of the slug increase in its length, CF
reference frame at stations 1,3 and 5 are, respectively, -(CF -
-(CF -
UL3)
and -(C F
- C T) . The tail of the slug is not a stagnation point, unlike the tail ofa stable slug.
(2.24)
The last term in (2.24) is the accumulation of liquid within the control volume.
The following relation can be obtained from (2.24):
(2.25)
If the momentum of the gas phase was neglected, the momentum balance would
be:
1~A3 + PL(C F
-U L1) 2
-U L3) 2
A3(1- &3)
-U L3)
(2.26)
Where, again, the last term is the accumulation of momentum within the control
volume. By substituting (2.25) into (2.26), the following equation was found:
30
(2.27)
(2.28)
(2.29)
For a decaying slug, C F < CT , the reference frame was chosen to move at a
velocity of C T, so that the fluid in the different parts of the slug move in the same
direction. Station 1 moves with velocity CF and stations 3 and 4 were chosen to move
with velocity CT. The length of the control volume between stations 1 and 3 changes at
a rate equal to (C F - CT) . The flow at the top of the slug tail was a stagnation point.
The following relations were obtained:
31
(2.30)
and
(2.31 )
(2.32)
Fan et al. (1993) used similar approaches of different workers to predict the
pressure drop due to friction.
successive slugs, they relied extensively on the correlations developed by Andritsos &
Hanratty (1987) for stratified flow. On the other hand, pressure loss in the slug body
due to wall shear was estimated using a general approach followed by most workers in
which equations similar to those in single phase flow were incorporated. This approach
32
will be discussed later in this chapter when dealing with pressure drop in the stratified
liquid film behind each slug.
Lockhart and Martinelli developed the first widely used correlation for pressure
drop and holdup in stratified flows (1949). This correlation over predicted pressure
drop by a factor of 100%. Recently, considering the momentum balance for fully
developed flow for the gas phase has developed improved correlations for pressure drop
and holdup.
Taitel and Dukler (1976) developed new mechanistic model to predict flow
regime transitions in horizontal and near horizontal gas-liquid flow. In their study,
stratified flow was central to their analysis. Taitel and Dukler considered a smooth
stratified flow with a flat gas-liquid interface where no waves existed. A momentum
balance on each phase yields:
(2.33)
(2.34)
(2.35)
33
= r
t
wG
JG
PGUG
'
(2.36)
The liquid and gas friction factors were evaluated using Blasius equation:
(2.37)
Where C ==0.046, n == m == 0.2 for the turbulent flow and C == 16, n == m == 1.0 for
laminar flow.
D L and Do are the hydraulic diameters for the liquid and gas phases,
D = 4A L
L
'
(2.38)
34
(2.39)
(2.40)
h
SL =D[rc -COS- l (2.-!:--1)]
.
D
(2.41)
(2.42)
(2.43)
=U SL
(2.44)
Uc; =U
(2.45)
SGA
Taitel and Dukler assumed that the velocity in gas phase was much greater than
the velocity of the fluid at the gas-liquid interface,
UG
sh.ear tress was evaluated with the same equations as the gas wall shear. Moreover, the
flow was assumed smooth stratified with no waves at the gas-liquid interface, fi =. !G.
The error introduced by making this assumption was small.
35
Andritsos and Hanratty (1987) studied the influence of interfacial waves in
stratified gas-liquid flows. In a previous paper (Andritsos and Hanratty 1986) defined
two types of interfacial waves, these were:
1.
2.
instability.
They found that large amplitude waves could cause large increases in gas-liquid
shear. Although Taitel-Dukler method was found to be good for predicting the liquid
height and the frictional pressure drop, considerable improvement was possible if the
influence of waves on
better relation for
'WI
1j
'WI
and (2.34) from their measurements of pressure gradient in each phase and the height of
the liquid film. They suggested equation (2.36) to estimate !G. A modification of the
Cheremisinoff-Davis correlation to predict characteristic stress has been used in their
research. The characteristic stress in the liquid was taken as:
(2.46)
36
(2.47)
(2.48)
for
-4-=1+15(h
JG
U SG
<5
(2.49)
)0,5 [U (~)O.5
SG
-1]
for
U SG
>=5
(2.50)
PGO
37
between the liquid and the gas and made liquid film climb the pipe wall resulting in a
concave
._.-*h
o
LWL,
is assumed to be uniformly
LWL:
38
t WL (e)
( ---1
r WLO
---=1+
'WG
where
J{l-exp(()
-0 J}
-m-l-l0
'WG
'WLO
(2.51)
is the liquid to wall shear stress at the pipe bottom (0 = 0); 8 is defined in
'WG
over the tube perimeter in contact with the gas phase, and equal to the liquid-wall shear
stress value at the angle 8.
regression methods and found to be a strong function to of gas and liquid superficial
velocities:
-2
-0.4
(2.52)
m= 70 U SG u L
'WI0,
UL,
were
PL U L
'wLo=fLO--
(2.53)
Values of fio, obtained from (2.53), were fitted satisfactory with a Blasius type
equation:
39
t LO =.
0 2Re -0.25
film
U h,
where Re film - -L - ,
vL
(2.54)
Vlachos et al. (1999) used Andritsos-Hanratty correlation (2.49) & (2.50) to estimate
interfacial friction factor,fi.
laminar sublayer and buffer zone beyond which turbulent core exists. Laminar flow
occurs in the laminar sublayer where the relationship between point velocity and the
distance between this point and the pipe wall is linear. The contribution of the laminar
sublayer to total pressure loss arises in viscous shear generated in the sublayer.
In the
buffer zone the velocity profile is not well understood, but it is believed to be similar to
the one in laminar sublayer at the outer border.
turbulent core, the velocity profile has a trend similar to that in the turbulent core.
Pressure loss in the buffer zone is due to viscous shear, while Reynolds stresses are
responsible for the pressure loss in pipe turbulent region. Lester (1985)
Since drag reducing agents are presumed not to act on Reynolds stresses, due to
their degradable effects, it is concluded that drag reducing agents work in the buffer
zone. The forces, in the wall region, tend to deform, stretch and unfold the polymer
molecules of the DRA. These forces are unidirectional. Hence, the deformation that
40
changes the shape of polymer molecules from spherical to near linear is expected to be
larger. The solution viscosity in the wall region becomes greater due to the distribution
of deformed polymer molecules in the flowing fluid. Such high viscosity would be
expected to oppose the intensification of vortices near the wall and hence decrease the
rate of bursting.
Figure 2.2.1 shows the changes in velocity profile in a full-pipe turbulent flow in
cylindrical coordinates. It is believed that DRA's help reduce cross flows, maximize the
thickness of buffer zone and eventually make velocity profile flatter.
Pipe Center
After Addition
efore Addition
Virk (1975) carried studies that involved mean velocity profile measurements.
According to his experimental results and observations, one can find that mean velocity
profiles during drag reduction have three zones from the wall outward: (1) a viscous
41
sublayer, (2) an elastic sublayer, characteristic of drag reduction, and (3) a Newtonian
plug where turbulence is due to Reynolds stresses. On the other hand, his gross flow
measurements indicated three flow regimes. III laminar flow, Re < 2000, the polymer
solution and solvent exhibited identical behavior.
polymer solution data are identically the same as for solvent for 3000< Re <12000; at
R.e == 12000, he witnessed a rather distinct onset of drag reduction and, for Re > 12000,
the polymer solution data diverged from solvent in the direction of lower friction. In
the current study, no attempt was undertaken to study the relationship between flow
characteristics and Reynolds number.
this study.
Since DRA's are believed to begin reducing drag after critical shear stress, most
of the data reported to date in the literature are for flows at high Reynolds numbers.
Furthermore, solvents used so far are fairly of low viscosity. Hom, Wu, Prilutski, and
Motier (1986) investigated drag reduction in a high viscosity crude, 48 cP, over a range
of Reynolds numbers of 1,790 to 3,000, during which transition from laminar to
turbulent flow takes place. They concluded from their experimental results that with.
highly effective DRA's, effective drag reduction could occur as soon as the transition to
turbulent flow begins and drag reduction was not inhibited by high solvent viscosity.
Usui and Sano (1989) studied the effect of polymer threads on a turbulent flow
of water. Their flow visualization results provided an evident that the injected polymer
solution exists as polymer threads. Furthermore, they added these polymer threads have
high viscoelasticity and they may cause
all
42
Sana claimed that the results obtained from mean velocity measurements show clearly
that local velocity of polymer threads was higher than that of water phase throughout all
radial positions. Their experimental results demonstrated a significant effect of polymer
threads on the turbulent motion exists in the core region. In other words, suppressing
the large motion in the turbulent core region can control the wall turbulence structure.
The precise mechanism describing how a DRA works to reduce friction has not
been established.
absorbing and later returning to the flowing stream energy which otherwise would have
been wasted in producing the cross flows which comprise turbulence. (Lester, 1969)
The following are turbulent structures that take place in the wall region as defined
by Eckelmann (1984):
regions with high and low velocities randomly distributed in space and time.
It is generally
Ejections: abrupt movements of fluid away from the wall occur frequently
near the wall (Corino and Brodkey, 1969). The reason for the ejections is not
known, but it is closely associated with the lift-up phase during bursting.
Sweeps: high-speed motion of fluid directed toward the wall. Sweeps also
43
It is
generally assumed that vortex pairs are responsible for the lift-up of fluid from
the wall. It is found that the ejections provide the major contribution to the
stream-wise vorticity.
A mechanism for drag reduction must specify the basic polymer-turbulence
interaction responsible for the phenomenon and detail how this interaction affects the
energy balances in turbulent pipe flow to yield the observed reduction in specific energy
requirement relative to the solvent alone.
Walsh (1967) associated drag reduction with visco-elasticity by linking it to
turbulent transport of macromolecular strain energy.
assumption.
Lumley (1969, 1973) provided a comprehensive and physically reasonable drag
reduction mechanism. According to his assumption, sequences of changes in mean and
turbulent flow structures were initiated by macromolecular elongation.
When injecting the flowing stream with DRA, polymer molecules may be
adsorbed on the pipe wall and the following question comes to the surface: what
44
molecules are responsible for drag reduction, the molecules on the wall or the molecules
traveling in the moving fluid?
Wells and Spangler (1967) injected polymer solution into water at different
points, at the wall region and at the axis of the pipe. In the former case, the wall shear
stress was reduced directly downstream of the injection point, whereas in the latter case,
the drag reduction was observed relatively much further downstream, presumably only
after the macromolecules had diffused to the pipe wall region.
Sedov, Ioselevich, and Pilipenko (1984) studied the structure of wall turbulence
and the mechanism of drag reduction by polymer additives. They concluded that drag
reduction is not dependent on the thickness of the adsorbed layer on the wall and that
substitution of the material has nothing to do with drag reduction. Their results showed
clearly that the flow drag is a result of interaction between DRA molecules traveling in
the fluid with the turbulent disturbances.
Ejection of the fluid to the outer region of the boundary layer occurs at the wall
region, meanwhile an invasion of the accelerated fluid to the wall region takes place. In
addition, counter-rotating stream wise vortices pairs are involved in that region. It is
generally assumed that vortex pairs are responsible for the lift up of fluid from the wall
region. The ejections provide the major contribution to the stream-wise vorticity.
At rest, each polymer molecule balls up around it self. Under conditions of
steady shearing flow, the molecules start to deform, stretch and unfold and after a while
they become oriented in the direction of the flow. This results in an increase in fluid
viscosity due to the presence of polymer molecules in the flowing stream.
The
45
unfolding and flow orientation take place immediately after certain, critical, value of
sliding velocity. Polymer molecules become stretched to their maximum permissible
length, forming a system of threads in the flow. Such high viscosity retard the intensity
of ejections, lower their frequency, retard vortices and therefore reduce the friction
resistance and prevent further development of jet flow which is responsible for counterrotating vortices.
In their theory, an explanation for drag reduction mechanism was based upon
the changes in fluid properties such as viscosity. Such increase in solvent viscosity is
accompanied by a decrease in turbulence intensity and an overall pressure loss. Two
factors would influence the validity of this explanation. These are the solubility of the
polymer additive in the treated liquid and the dependence of treated liquid viscosity and
other properties upon polymer molecular concentration.
Given a good
dispersion of the DRA in the bulk of the treated oil, the influence of the DRA on the oilwater mixture properties becomes an important factor in determining the validity of
above mechanism of drag reduction.
Bewersdorff,
Gyr,
Hoyer,
and
Tsinober (1993)
investigated
possible
Their
experiments were carried out to find out whether the drag reduction is partially due to
dissolved polymer molecules disposed from the thread. Based on their experimental
work and previous knowledge about heterogeneous drag reduction, they concluded that
drag reduction is a result of two effects, namely:
46
Small amounts of polymer are removed from the thread volume and enter
The viscoelastic thread can interact with the flow turbulent structures
47
reduction increased with increasing liquid rate at a fixed gas rate. Pressure gradient
reductions up to 37% were obtained.
Sylvester, Dowling, and Brill (1980) investigated drag reduction in co-current
horizontal natural gas-hexane pipe flow experimentally. Their test section consisted
mainly ofthree100-ft pipes 1",2", and 3" diameter. A flow pattern map was provided.
This map agrees very well with a developed generalized flow regime criteria presented
by Taitel and Dukler. Their results showed that in the annular-mist flow regime, drag
reduction increased with decreasing gas flow rate for a given liquid rate. They added
that drag reduction decreased as the liquid-gas ratio approached zero. Attempts to
predict two-phase drag reduction were unsuccessful as it was found to depend upon
friction velocity, liquid and gas flow rates, liquid-gas flow rate ratio, and additive
concentration.
These findings agree with the results I developed at the Flow Improvement
Center at Ohio University for the annular mist flow of 2 cP oil-C02 mixture in a 4" ill
horizontal pipes. Superficial liquid velocity had the values of 0.003, 0.005, 0.01, and
0.1 mls whereas superficial gas velocity had the values of 12, 15, and 20 mls. The
experimental results obtained at the Center showed that liquid-gas ratio not only
affected the performance of the DRA but also the flow pattern of the mixture. Similar
to Sylvester and Brill's results, drag reduction was found to increase with increasing
liquid-gas ratio. The DRA had little or no effect on the flow pattern but it did decrease
the flow fluctuation and reduced flow turbulence.
48
Greskovich and Shrier (1971) investigated the effect ofDRA on pressure drop in
two-phase (water/air and kerosene/nitrogen) flows in 3.81-cm I.D. pipelines. The
pressure drop reduction of up to 50% and 45% in water/air flows was achieved in plug
flow and slug flow, respectively, with addition of 50 wppm DRA. In kerosene/nitrogen
flow, a Drag reduction of29% was obtained in slug flow with the same amount of
D.RA.
DRA's were believed to work only on frictional drag, hence most of the work
has been carried out to study drag reduction phenomenon in single-phase flow. DRA's
were found to effectively reduce pressure drop in all flow patterns in multiphase flow.
This indicates that DRA's reduces not only frictional drag but also accelerational
component of pressure drop in slug flow too.
Kang, Vancko, and Jepson (1998) studied the effect of polymer additives in both
full pipe and multiphase flow. Experiments were carried out in both horizontal and 2degree inclined, 4" ill pipelines. The fluids tested were 2 cP oil, water and Carbon
dioxide as the gas phase. The results indicated that the DRA used in the study was
effective in reducing pressure gradient in both single and multiphase flow. Pressure
gradient reductions of up to 42% in full pipe flow, 81% in stratified flow, and 35% in
annular flow were achieved in horizontal pipes. In 2-degree upward multiphase flow,
drag reduction for slug flow had values as high as 380/0 at DRA concentration of 50
ppm. Furthermore, they added that the DRA had some influence on flow regime
transition. The transition to slug flow with the addition of DRA was observed to occur
at higher superficial liquid velocities. For example, slug flow took place at superficial
49
liquid velocities above 0.2 mls. Adding 25 ppm of the DRA shifted the transition to
slug flow regime to superficial liquid velocity above 0.5 mls. At DRA concentration of
75 ppm, the transition to slug flow occurred at superficial liquid velocity above 0.56
mls. these findings can be seen in Figures 1.5 and 1.6. In contrary to many results in
single -phase flow, Kang indicated that the effectiveness of the DRA in full pipe flow
of2.5 cP oil did decrease with the increase in liquid flow rate, Re. No explanation was
given for this phenomenon.
This study will show similar results to these obtained by Kang et al. (1998).
Drag reduction in two-phase 27 cP oil-C0 2 flow was achieved regardless of flow
pattern. Drag reduction did occur in multiphase flow of oil-water-Cfi, at water cut
below 10%. Mixtures of oil and water of water content more than 10% seemed to form
some kind of emulsion. The properties of such mixtures changed dramatically after the
addition ofDRA causing the pressure drop to increase rather than decrease.
Sifferman and Greenkom (1981) have studied drag reduction of three polymers
(sodium carboxymethylcellulose, polyethylene oxide, and guar gum) in three different
fluids (single-phase polymer/water, two-phase solid/liquid suspension, and three-phase
immiscible liquid/liquid/solid solutions) in 2.7-cm diameter pipelines. Drag reduction
was observed for all three-flow systems studied. At water Reynolds numbers exceeding
105 , drag reduction of up to 80% was observed for both guar gum (Jaguar) and
Polyethylene oxide at DRA concentration of 0.3 wt%. For the polymer/sand system,
drag reduction of95~98% was achieved, indicating an additive drag reduction effect for
polymer solutions with suspended solid particles.
50
There still a conspicuous absence of work involving drag reduction in oil-watergas multiphase flow.
51
CHAPTER THREE
52
A. Storage Tank
E. Pressure Taps
B. Pump
F. Flexible Hose
c. Mixing Tee
G. Gas Flow Meter
F
D. Gas Tank
E
E
Flow
Meter
Outlet Gas
i t
Figure 3.1 System Diagram
53
The gas flow rate is determined by means of a HLIT208 HEADLAND variable
area flow meter (G) located after the ball valve. The gas then flows, through a pressure
regulator of 0.035-0.862 MPa capacity, into a 40-m long, 5.08-cm ill PVC pipeline
before it enters the system at the tee section. Pressure is monitored with a pressure
gauge, located between the ball valve and the pressure regulator, with a range of 0 to
2.07 MPa, whereas temperature is measured by means ofTEL-TRU thermometer with a
range of -50 DC to 50 DC.
The mixture then flows upward through a 20-m long, 10.16-cm ill Plexiglas
pipeline and is fully developed before it reaches the upward test section. The mixture
then flows downward through another 20-m long, 10.16-cm ill PVC pipeline before it
reaches the downward Plexiglas test section. Upward and downward test sections are
identical. The flow then returns to the tank (A) where gas and liquid are separated. The
gas is vented to the atmosphere and the liquid is recycled.
The system pressure is measured by means of pressure gauge with a range of 0
to 0.103 Mpa. Two sets of taps, the upward 6-m apart and the downward 9-m apart, are
used to measure pressure drops across upward and downward test sections respectively.
Two identical SENSOTEC 882-12A 5D differential pressure transducers, with a
range of 0 to 0.0345 MPa, are connected to the pressure taps.
A 12-BIT 12
over a period of time set at each measurement by an interactive entry of the time
required for the current measurement. The response is taken to a Pentium PC where the
54
average pressure and pressure fluctuation are determined by a data acquisition program
. C++ Compi1 ere
using
The oil is
pumped from the storage tank into a 10.16-cm ID Plexiglas pipeline and the liquid flow
rate is controlled using digital-speed control of the pump.
The drag reducing agent is added to the mixture by preparing a master batch
from the required amount ofDRA based on the total volume of the liquid.
The following equation is used to calculate the required volume ofDRA for
certain concentration:
55
= C DRA
X VTotal
1 X 10 6
DRA
(3.2.1 )
Where:
batch. The solubility of DRA in the oil is monitored. This master batch is then poured
into the tank through an inlet valve fixed at the top of the tank. The mixture is then
circulated in the system for 30 minutes to assure the dissolution of the DRA in the
liquid phase before starting experiments.
A SUPER VHS PANASONIC camera and an AG-190 Panasonic camcorder are
positioned 6 and 2 meters back from the upward and downward test sections,
respectively. The shots are recorded by high speed VCR and saved in VHS tapes. The
flow is viewed on a TV screen. Flow characteristics are then analyzed.
56
and on the effectiveness of DRA's in reducing each of these components. For these
reasons, three different types of oil of markedly different viscosity were examined. The
system used in this study was set to two different inclinations so that effects of pipe
inclination on DRA effectiveness and the contribution of each component to total
pressure drop can be measured. The DRA was added into the flowing fluids in dosages
ranging between 0 ppm and 75 pp. Table 3.3.1 shows the pipe inclinations set for
experiments carried out on each oil, while Table 3.3.2 describes the flow conditions.
All Measured quantities are shown in tables in Appendix A.
Appendix B
includes tables of selected calculated quantities; these are film Froude number,
Frj;
slug
Froude number, Frslug , slug holdup, Rs, and the components of pressure drop.
Table 3.3.1 Pipe Inclinations for Experiments Carried Out on Each Oil
Oil
Pipe Inclination
2.5 cP
26 cP
Two-degree upward
50 cP
Horizontal
57
Oil Type
Gas Phase
Carbon Dioxide, CO 2
0,20,50,75
System Pressure
5 Psi
0.5, 1, 1.5
2~14
58
CHAPTER FOUR
MODELING PROCEDURE
The pressure drop across single slug consists of two components; The first, ilP a ,
is the pressure drop that results from accelerating the slow moving liquid film, ahead of
the slug, to slug velocity. The second, ilPf, is the pressure drop required to overcome
wall shear in the back section of the slug. In inclined pipes, a third component emerges
due to resistance of gravity, LlP G The total pressure drop across a slug is thus:
(2.1)
(2.39)
(2.40)
59
(2.41)
(2.42)
(2.43)
(2.38)
The actual bulk velocity of each phase was evaluated using the following
equations:
(2.44)
(2.45)
60
The liquid and gas friction factors were evaluated using Blasius equation:
(2.37)
1.0 for
laminar flow.
To count for the gas-liquid interfacial effects on the gas-liquid interfacial
friction factor.ji, the proposed model by Andritsos and Hanratty (1987) was considered:
JG
SG
PGO
for
U SG
<5
(2.49)
for
U SG
>=5
(2.50)
(2.36)
61
The pressure drop due to friction in the stratified liquid film can be calculated
k:nowing the pressure gradient in the film and the length of such film:
_(dP)
dX
fitm -
film
~-l
v
s
(4.1.1)
slug
60
The quantity in brackets represents the length of the slug stratified liquid film.
Following Taitel and Dukler (1976), a momentum balance on each phase yields:
(2.33)
(2.34)
In smooth stratified flow, pressure gradient of both phases are equal. For the
purpose of determining the pressure gradient in our case, equation (2.34) was modified
as shown below:
(4.1.2)
62
Hence, pressure loss due to friction in the liquid film in each slug is:
Mt,film
= If
dP)
(dX
-
(4.1.3)
fi/m,gradient
becomes:
(2.3)
The similarity analysis showed that for "non-slip" conditions the slug friction
factor, !slug, could be correlated as a unique function of the slug Reynolds Number,
63
(2.4)
Where !slug is slug friction factor and calculated using the following equation:
r; =0.0791 (Re
/
s ug
rO. 25
(4.2.1)
Gregory, Nicholson, and Aziz (1978) developed a correlation for liquid holdup
within the slug. The following correlation was developed for liquid holdup, R; within
the slug:
1+
]1.39
(2.7)
_5_
[ 8.66
Where Vs is the velocity of the mixture in the slug body, equal to the sum of
superficial velocity of both the liquid and the gas.
The length of the mixing zone, L MZ was estimated using an equation developed
by Kouba and Jepson (1990) based on Froude number of the liquid film right behind the
slug body, Frj; was used:
(4.2.2)
64
(1.1)
h =A L
eff
S,
(4.2.3)
Total frictional pressure drop associated to each slug is then given as:
(4.2.4)
(2.2)
For simplicity, velocity of liquid film just prior pickup by slug front, Vie, was
given the value of the measured average film velocity, Vf
Rate of mass pickup was calculated by carrying out mass balance over the slug
front, Hubbard & Dukler ( 1975) used the following equation:
65
(2.5)
Where Rf e was given the value liquid holdup in the stratified film, RJ-
(4.3.1)
Gravitational
LiPg =
LiPg,bOdY
+ APg,film
(4.4.1)
~Pg,bodY == PsJug
g x Is x sin(e)
(4.4.2)
66
The density of the fluid in the slug body was already estimated in modeling the
frictional loss in the slug body. Pressure loss in the liquid film is given below:
Mg,jilm
== R f
X PL
xlf x sin(e)
(4.4.3)
The group of measured quantities consists of the height of liquid film, hL, stratified film
velocity, VI, slug length, Is, slug translational velocity, VI, and slug frequency, Vs. This
group of parameters was fed in the equations and models used in this study, as
described in the current chapter, and computations proceeded until each individual
component of the pressure drop was predicted. It is evident that this group of measured
key parameters is essential to any computations of pressure drop.
To incorporate this procedure in industry rather than laboratories, appropriate
models should be used to predict each of these parameters individually. Fortunately,
literature is rich of mechanistic, as well as empirical, correlations that fit many
conditions under which oil pipelines are operated. On the other hand, the best majority
of these models are valid without the presence of drag reducing agents. The addition of
drag reducing agents causes many changes in the flow characteristics. Among these
important changes is the slug frequency. A main reason behind the reduction in total
pressure drop is the reduction in slug frequency. A flow with fewer slugs per unit time
is always accompanied with less total pressure drop
67
e:nhancement factor to account for the influence of DRA's on flow characteristics. The
enhancement factor would be proposed upon our the bank of data developed from the
extensive course of experiments undertaken in the Center for Corrosion and Multiphase
Technology utilizing different fluids and DRA's in both horizontal and inclined pipes.
It is clear that this work is capable of producing an excellent quantitative
description of the pressure drop in slug flow and the influence ofDRA on pressure drop
components. This work also paves the road for further understanding of the effects of
oil viscosity and pipe inclination on the structure of pressure drop and the performance
of DRA's. All of that was possible when key parameters such as slug properties were
known and appropriate models were used.
68
CHAPTER FIVE
RESULTS AND DISCUSSION
The pressure drop in horizontal slug flow using 2.5 cP oil was calculated by
estimating both frictional and accelerational components. It was found that the main
component of pressure gradient is the accelerational component where its percentage to
69
the total pressure gradient ranges from 770/0 to 890/0. The remainder was the frictional
contributor.
The drag reducing agent, DRA, used in this study was found to be effective in
reducing the pressure drop at all superficial liquid and gas velocities. A dosage of 20
ppm of DRA caused the pressure drop and both of its components to decrease. Further
addition of DRA to a concentration of 50 ppm was accompanied with more drag
reduction and higher effectiveness of the DRA.
In sharp contrast to what is believed, most of the gained drag reduction took
place in the accelerational component, while smaller fraction was attributed to the drag
reduction gained in the frictional component.
70
4000
I:
'V
Accelerational
Frictional
co
c.. 3000
l-
'iJ
c.
~
...
~
Q)
Ii
'iJ
\l
c. 1000
I-
o
I
0
2
Fi gure 5.1 .1.2 Pres sure Orop Vs. Sup erfic ial Gas Velocity
100% Oil (2.5 cP), Vsl= 1 m/s,O ppm, Horizontal Flow
71
10
8 --
T ~aI ru1easured
Total Calculated
sr Acceleratiooal
f-
Frictional
o
'V
4 2 -
'V
It can be seen that total pressure drop increased with increasing superficial gas
velocity due to the increase in both components. For example, Figure 5.1.1.1 shows
that, at superficial liquid velocity of 0.5 mls and DRA concentration of 0 ppm, the total
pressure drop increased from 571 to 1630 Pa as a result of increasing superficial gas
velocity from 2 to 6 mls.
frictional components were from 491 to 1433 Pa and from 79 to 197 P, respectively.
Similar behavior was noticed after the addition ofDRA. Figures 5.1.1.4 through
5.1.1.6 describe the changes in total pressure drop and its components at DRA
concentration of 20 ppm and at superficial liquid velocities of 0.5, 1, and 1.5 mis,
respectively. The corresponding values at DRA concentration of 50 ppm are shown in
Figures 5.1.1.7 through 5.1.1.9, respectively.
72
2000
t':I
~ 1500
..!!..
lotalCalculated
'iJ Accelerational
...o
...~
lotal fJleas....ed
I-
c.
1000
500
Frictional
...
~
c..
9
o
o
I
4000
:. 3000
JC
c...
...
Total Measured
Total CaIculated
sr Accelerational
\J
Frictional
2000
CL)
CL)
~
0-
1000
73
8000
~ 6000
Tot~ Measured
Total Calculated
'V
Acceleratimal
c-
o Frictional
~ 4000
'V
...
Q.'I
,.,
~
'iJ
~ 2000
~
0
c
6
2000
c.. 1500
TotaiMeasured
Total C~cul~ed
'V
AcceIe rational
Frictional
c-
...e
c 1000
'iJ
f-
...,.,
Q.'I
~
Q)
c:
500 ~
Fi gure 5.1 .1.7 Pres sure Oro p Vs. Sup erfic ial Gas Velocity
100~b Oil (2.5 cP), Vsl=O.5 m/s,50 ppm, Horizontal Flow
74
3000
Total Measured
2500
Total Calculated
:I 2000
...o
c 1500
...
'I
Accelerational
C.
C-
o Frictional
-
Q)
- 1000
~
...
Q)
c.
500 -
o
2
o
I
Fi gure 5.1 .1.8 Pres sure Oro p Vs. Sup erfic ial Gas Velocity
100~b Oil (2.5 cP), Vsl= 1 m/s,50 ppm, Horizontal Flow
75
This phenomenon can be explained by recalling the following equations:
(2.1)
The accelerational component of the pressure drop was the dominant contributor
and formed around 80% to 90% of the total pressure drop in slug flow. The percentage
of accelerational component was noticed to increase with increasing superficial gas
velocity.
(2.2)
(2.5)
76
(1.1)
(2.4)
Thus, an increase in the frictional component will occur due to the increase in
the slug friction factor as can be noticed from equation 2.3:
(2.3)
The
77
the increase in the slug translational velocity and hence the force necessary to accelerate
the liquid in the film just before pickup to the slug velocity. This force manifests its self
as the accelerational pressure drop.
caused the liquid film height to decrease so that decreasing the wet area of the pipe and
the frictional part of the total pressure gradient.
One can find from Figure 5.1.1.2 that at superficial liquid velocity of 1 mls and
I)RA concentration of 0 ppm, the percentage of accelerational component increased
from 81% to 90% when increasing superficial gas velocity from 2 to 6 mls. Similar
results were found at DRA concentrations of 20 and 50 ppm. Figure 5.1.1.8 indicates
that at superficial liquid velocity of 1 mls and 50 ppm DRA, an increase in
accelerational component from 238 Pa to 1969 Pa, that is an increase in its percentage
from 840/0 to 90%, occurred as a result of increasing superficial gas velocity from 2 to 6
m/s.
The fractions of both accelerational and frictional components of total pressure
gradient did not change much with increasing superficial liquid velocity holding all
other variables constant, despite the increase in the liquid film height and hence the
frictional component. Remember that an increase in liquid velocity results in increasing
the height of liquid film, hence rate of mass pickup (2.5), and the wet perimeter, hence
frictional pressure gradient in the liquid film (2.33).
For example, Figures 5.1.1.1 through 5.1.1.3 shows that the percentages of the
fiictional component at 0 ppm DRA and superficial gas velocity of 6 mls were around
12%, regardless of superficial liquid velocity. The corresponding percentages at DRA
78
concentration of 50 ppm were around the same value as shown in Figures 5.1.1.7
through 5.1.1.9.
At superficial liquid velocity of 1.5 mls and DRA concentration of 20 ppm, total
pressure drop increased from 1286 to 4954 Pa when increasing superficial gas velocity
from 2 to 6 mls as shown in Figure 5.1.1.6.
accelerational and frictional components were from 1070 to 4344 Pa and from 215 to
611 Pa, respectively.
Similar results can be seen in Figures 5.1.1.7 through 5.1.1.9 at different DRA
concentrations and superficial liquid velocities.
5.1.2 Effectiveness of DRA's in Reducing Pressure Drop
DRA was added into the flowing stream to examine its effect in reducing total
pressure drop as well as each one of its components. Figures 5.1.2.1 through 5.1.2.3
describe the effect of DRA on frictional component at superficial liquid velocities of
0.,5, 1, and 1.5 mis, respectively. The addition of DRA was found to decrease frictional
component at all superficial liquid velocities and DRA concentrations. On the other
hand, the addition of 50 ppm of the DRA was accompanied with a transition in the flow
pattern from slug to wavy stratified at superficial liquid and gas velocities of 0.5 and 2
mls respectively. Figure 5.1.2.1 shows that at superficial liquid and gas velocities of 0.5
&~
6 mis, respectively, frictional component had the values of 197,152, and 121 Pa at
79
250
c.
200
_0
C&..
Q.'I
20ppm
o 50ppm
Oppm
150 -
&..
-J
Q.'I
&..
c..
100 -
C
Q
';
50
o
o
~
&..
u,
0
2
500
~
c.
400
c0
&..
Q
Q.'I
300
&..
-J
Q.'I
&..
200
100
Oppm
20ppm
50ppm
C.
t'C:'
C
Q
';
o
o
0
&..
u,
80
1000
c.
:II:
800
~D
600
400
Oppm
20ppm
C-
Q
(1)
50ppm
-'
(1)
~
c.
o
o
(':J
c
0
200
(,J
U-
Accelerational component was also found to decrease with the addition of DRA.
Figures 5.1.2.4 through 5.1.2.6 indicate the influence of the addition of the DRA on the
accelerational component.
81
2000
Oppm
I:t'
a- 1600
~D
jj
......
c.
0
1200
20ppm
50ppm
Q)
~
~
Q)
800 -
0
0
al'tI
400 . .
';
~
u..
4000
~
c,
:Ie
c.
Oppm
20ppm
50ppm
3000 -
...
Q
0
.....
Q)
2000
Q)
c.
(10
s:: 1000
;::
u
...
u.
8
2
Fi gure 5.1 .2.5 Effects of DRA on Ac ce lerati onal Pre ssu re Drop
100% Oil (2.5 cP), Vsl= 1 mis, Horizontal
82
(t1
6000
c.
......
Oppm
-D
20ppm
4000
-0
50ppm
3000
JL 5000
CQ
Q
Q)
U;
~
OJ
~
C.
~
$:
0
2000
1000
0
0
;::
(t1
~
Q)
(.)
(.)
0
2
83
ppm are shown in Figures 5.2.1.4 through 5.2.1.6, while Figures 5.2.1.7 through 5.2.1.9
describe the component of pressure drop at DRA concentration of 50 ppm.
It is evident that the calculated and measured values were in good agreement
except at high superficial liquid and gas velocities of 1.5 & 6 respectively at which the
height of the liquid film was over estimated. Figure 5.2.1.1 describes the changes in
pressure drop and its three components with superficial gas velocity at DRA
concentration of 0 ppm and superficial liquid velocity of 0.5 mls.
All components
different superficialliquid velocities of 1 and 1.5 mls and at all DRA concentrations.
84
One can see that accelerational component was dominant. Figure 5.2.1.3 shows
that at superficial liquid and gas velocities of 1.5 & 6 respectively, 860/0 of total pressure
drop came from the accelerational contribution, 8 % from gravitational contribution and
the rest from frictional contribution.
An increase in superficial liquid velocity was accompanied with significant
concentration of 50 ppm were 613, 555, and 485 Pa as shown in Figure 5.2.1.8.
85
5000
(tI 4000
c..
-0
'OJ
I
~
3000
ae
..~ 2000
-0
Total treasured
Total Calculated
Grav itatio nal
Accelerational
iJ. Frictional
tJ.)
a-
tJ.)
a-
e..
1000 0
\J
\J
.+.
6000
Total rv1eas....ed
5000
-o
T(ial Calculated
c.
c.
e
Q
..~
"i]
4000
r-o
3000
1i
Accelenrtional
Frictional
~ 2000 ~
E
c.
Gravitational
1000 -
o
2
86
12....-------...-----------------...
I ol~ Measured
o lolalCaiculated
9 - 'V Gravitational
o Acceleratimal
6-
Frictional
3-
~
2
o--.----.....----------------.......---~
2
Fi gure 5.2.1.4 Pres sure Oro p Vs. Sup erfic ial Gas Velocity
1000/0 Oil (2.5 cP), Vsl= 0.5 m/s, 20 ppm, 2-0egree Upward
87
6000
5000
-0
Total Measured
TutaI Calcul ated
c,
:IC
e,
...Q
...
'V
Gravitational
4000 -
o AcceleratiJnal
3000
-t,.
Frictional
Q,)
- 2000
~
...
c,
Q,)
1000 ....
0
o
..L
.L
'V
'V
10
9-
8 ..
Totll Measured
Total Calculated
0.
ii~
......
-c
c- e
...
c
Q
~
Q
... ..c
Q,)
=
to~'-'
...
0.
f-'V
Gravitational
.... 0
Accelerational
~t,.
Frictional
Q,)
2 ..
1~
0
1
t
'V
4~
3f-
88
2 4 6
4000
T ~al Measured
Total Calculated
(\'J
c. 3000
I:
c-
...
......
r-~
Gravitational
Acc:eleratimal
Q
OJ
2000
...
c.
Frictional
0
OJ
1000
r-
0
0
~
\J
2*
11
A
I
89
90
I>RA was found to decrease substantially and the surface of the liquid film kept most of
its semi-flat shape except at high DRA concentrations and superficial gas velocities.
Increasing superficial gas velocity resulted in a force-buildup upon the liquid film, thus
decreasing its height and reshaping its surface. The combined effects of the addition of
DRA and increasing superficial gas velocity were enlarged at higher liquid flow rates
during which more liquid presented. Figure 5.2.2.1 shows, qualitatively, the effect of
the addition of DRA on the shape of a liquid film at certain superficial liquid and gas
velocities.
Figure 5.2.2.1 Surface_of liquid film before and after the addition of ORA
91
At low superficial gas velocities, DRA was found to slightly decrease frictional
component. Figures 5.2.2.2 through 5.2.2.4 describe the etTect of DRA on frictional
component at superficial liquid velocities of 0.5, 1, and 1.5 mls respectively. It can be
seen from Figure 5.2.12 that at superficial liquid and gas velocities of 1 & 2 mls
respectively, frictional component had the values of 190, 108, and 148 Pa at DRA
concentrations of 0, 20, and 50 ppm respectively.
superficial gas velocity of 6 mls were 173, 384 and 287 Pa respectively.
250....-----.----------------------.
Oppm
c,
I:
200
~ 0
20ppm
50 ppm
C-
e
c
OJ
150 -
--~
OJ
~
e,
c
0
;:
100 50 -
u,
92
500
(0
c..
:I
:c..
0
...
Q
400
...
300
Q,)
-~
...
Oppm
20ppm
50ppm
1
6
Q,)
c..
200
(0
c
0
+::
(.)
...
u.
100
!
Q
900
t.tI
c.
:I
e,
0
...
Q
...
Q)
800 -
Q)
t.tI
:c:
600
500 400 -
200 -
...
100 '"""
u,
20 ppm
50 ppm
0
0
300 '"""
-..=
(.)
Oppm
100 -
...
c.
93
Accelerational component was found to decrease significantly with the addition
of DRA especially at higher superficial gas velocity of 6 mls.
associated decrease in liquid film height, hence the rate of mass pickup by slug, and
slug frequency.
c.
......
..!L
3000
c.
Oppm
20ppm
50ppm
Q)
~
2000
t;;
~
Q)
c..
e 1000 0
.';:;
~
~
Q)
e
Q
..L
Q)
(.)
(.)
<
94
6000
c..
JL 5000
Oppm
_0
20Rlm
50Rlm
C-
Q)
4000 -
t;;
VJ
Q)
3000 -
c..
s:: 2000 0
~
Q)
(,J
(,J
eX
.J..
;:
~
1000 -
~ 9000
c,
jL 8000 ~ 7000 -
Oppm
0
20ppm
'; 6000
- 0 50ppm
(;)
VJ
5000 -
Q)
c,
~
4000 3000 -
+:
2000 -
Q)
Q)
(,J
C,.'t
eX
1000
95
mis, gravitational component decreased slightly from 609 to 586 Pa when adding 20
ppm DRA. At DRA concentration of 50 ppm, it had the value of 555 Pa.
700
c,
600 ~
0
0
c.
0
500
f-
C;; 400
0
0
0
0
Q)
Q)
300
c 200
0
Q.
~
.....
~
100 ::(10
~
C!)
Oppm
20ppm
50ppm
I
96
(tI
C......
JL
C0
900
800
700
Q
Q)
600 -
;)
(I)
Q)
~
c..
f0-
.....
200
(tl
(tl
e
0
.t:
>
...
(tl
(.!)
100
Oppm
20ppm
50ppm
fo-
1000
('0
e,
......
900
JL 800 e,
...
...
Q
Q)
V;
(I)
Q)
...
c..
('0
c
Q
-.
400 300
.t::
>
100 -
...
('0
Oppm
20ppm
50ppm,
97
Calculated and measured values were in very good agreement at all superficial
liquid and gas velocities and DRA concentrations. Plug flow replaced slug flow pattern
on the flow regime map at superficial liquid and gas velocities of 1.5 & 2 respectively
regardless of the DRA concentration.
5000
~
C.
I:
...c~
..~
4000
,...
Total f\&!asured
Total Calculated
'V
Gravitational
0 Acceleratilnal
3000 8
Q.'l
'"'-
2000
Frictional
Q.'l
'"'-
C.
1000
0
!
6.J.
!
[5
.J.
Q
'V
'V
'V
A
98
e,
8000
7000
-0
6000
_v Gravitatiooal
ii
~ 5000
...
...
l-
Total Calculated
Accelerational
Frictiooal
0
3000
...
c. 2000 -
1000 -
\I
Q)
4000
Q)
.""
Total Jvle8S~ed
20
Total fu1easured
Total Calculated
e,
15
iI
-\7
~i
Q s::
...
Gravitational
AcceleratiJnal
l't1
UJ
=
=.c
Q)
10 -A Frictional
.... 0
~t:,
...
c.
Q)
99
Similar to results found when testing 2.5 cP oil, all components increased with
increasing superficial gas velocity except gravitational component, which decreased
linearly.
superficial liquid velocity of 1 mls and DRA concentration on 0 ppm. It can be seen
from this Figure that accelerational component increased from 2176 to 4636 Pa as a
result of increasing superficial gas velocity from 2 to 6 mls.
Frictional component
increased from 396 to 468 Pa for the same increase in gas velocity. On the other hand,
gravitational component was found to decrease linearly when increasing superficial gas
velocity at certain liquid flow rate and DRA concentration due to the decrease in slug
liquid holdup and liquid film height behind the slug. Figure 5.3.1.1 indicates that at
DRA concentration of 0 ppm and superficial liquid velocity of 0.5 mls a decrease in
gravitational component from 762 to 612 Pa occurred as a result of increasing
superficial gas velocity from 2 to 6 mls.
superficial liquid velocities of 1 and 1.5 mls and at all DRA concentrations.
One can see that accelerational component was dominant. Figure 5.3.1.2 shows
that at superficial liquid and gas velocities of 1 & 6 mls respectively, 80% of total
pressure drop came from the accelerational contribution, 12% from gravitational
contribution and 8% from frictional contribution. Similar behavior was noticed at DRA
concentrations of 10 and 50 ppm. Figures 5.3.1.4 through 5.3.1.6 describe the changes
in pressure drop and its components at DRA concentration of 10 ppm and superficial
liquid velocities of 0.5, 1, and 1.5 mls respectively. The corresponding values at DRA
concentration of 50 ppm are shown in Figures 5.3.1.7 through 5.3.1.9, respectively.
100
5000
e.~ 4000
Total Measured
r;;;
Gra vitation al
,..-
I:
a
~ 3000 ~
1l
Q)
~
~ 2000 -
Q)
~
e.
1000
i
1
2;
.1
"V
'iJ
.+.
4a.
1l
10
e.
ii-.
--",
c.-c
QS::
8
6
Total Measured
Total CaIculated
'V
l-
,..-0
~t':1
Q(I.l
Acceler &tional
Frictional
Q)=
~o
=-=
~C
Q)
l-
e.
2 -
~
.1
:
i
Q
\J
~
'V
101
15
12 -
Tot~
Measured
o Total C~culated
11
Gravitational
Acceleratimal
Frictional
o
6
4000
Tmal f\IIeas~ed
o Tot~ Calculated
~ 3000
I:
c.
...
c
...
~V
Gravitational
o Accelerational
2000
Q,)
-.I
...
Q,)
c.. 1000
f-
Frictional
,.
!
T
.+.
102
8000
('0
c.
:=
Total Measured
Total Calculated
:Ie
c:.
4000
"""A
c
Q)
....
Frictional
II
....~
Acceler ational
2000 -
c.
o
4
9000
8000
('0
e.
-0
7000
f0-
6000
fo-
~ 5000
....
'; 4000
TotalMeasured
Total Calculated
\J
Gravitational
Acceleratiooal
Frictional
fo-
....
3000
f-
2000
f-
I1
Q)
c:
1000
103
Gravitational component was found to increase when increasing superficial
liquid velocity regardless of DRA concentration.
indicate that at superficial gas velocity of 4 mls and DRA concentration of 10 ppm, an
increase in superficial liquid velocity from 0.5 to 1 mls was accompanied with an
increase in gravitational component from 613 to 689 Pa. At superficial liquid velocity
of 1.5 mls the gravitational component further increased to 756 Pa. On the other hand,
an increase in superficial gas velocity caused gravitational component to decrease at all
DRA concentrations.
velocity of 0.5 mls and DRA concentration of 50 ppm, gravitational component had the
values of 653, 598, and 496 Pa at superficial gas velocities of 2, 4, and 6 mls
respectively. The corresponding values at DRA concentration of 10 ppm were 722,
613, and 541 Pa as shown in Figure 5.3.1.4.
liquid and gas velocities of 0.5 & 6 mls respectively, frictional component had the
values of 394, 208, and 187 Pa at DRA concentrations of 0, 20, and 50 ppm
respectively.
104
500
~
C-
JI:
400 -
Oppm
10ppn
o 50ppn
C-
...
Q
.....
300 -
...
c.
200
Q)
Q)
o
o
(tI
~
;:
CJ
100 -
0
0
...
U-
1000
~
c,
800
JI:
ce
...
.....~
600
Oppm
10 ppm
50 ppm
Q)
...
Q)
400
c.
c:
Q
200 -
.J.
J.
.';:
U
...
u,
105
800
~
C-
:I
700
600
Oppm
o 10ppm
o 50 ppm
CI-
500
Q)
....
400
Q)
300
200
100
I-
I-
CL.
(to
6
-l
i
1
;:
~
I-
U.
The corresponding values at superficial liquid velocity of 1 mls were 468, 295,
and 321 as shown in Figure 5.3.2.2.
Accelerational component was found to decrease significantly with the addition
of DRA, especially at higher superficial gas velocity of 6 mls due to the decrease in the
rate of mass pickup by the slug. Figures 5.3.2.4 through 5.3.2.6 indicate the decrease in
accelerational component as a result of increasing DRA concentration. For example,
Figure 5.3.2.6 shows a decrease in accelerational component of pressure drop from
7424 to 6811 Pa at superficial liquid and gas velocities of 1.5 & 6 mls respectively after
the addition of 10 ppm DRA.
106
2 5 0 0 . . . . - - - - . . . . - - - - - - - - - - - - - - - - -.......
JL 2000 _ 0
c.
Oppm
10ppm
~
~
50ppm
1500 ...
tjj
V)
Q)
'-
c..
1000-
c
o
500 -
.4. . .- --------"---.
..
6
o~-----------I
6000
c..
...-
JL 5000
c-
Oppm
--0
10ppm
50ppm
...
o
Q
Q)
4000 ...
'-
(i)
VJ
Q)
'-
3000 --
c,
('Q
2000 -
...
0
0
.~
1000
0-
Q)
u
u
<
107
(tt
c.
9000
JL 8000
c-
Oppm
10 ppm
7000
Q)
6000
~o
5000 -
Q
r-
Q
rtJ')
Q)
r-
C.
4000 -
(tt
:::
0
50 ppm
11
3000 -
';:
(tt
r-
2000 -
ei)
Q)
u
u
eX
1000 0
10 ppm DRA. At 50 ppm, it had the value of 735 Pa. The corresponding values at
superficialliquid velocity of 0.5 m/s were 762, 722, and 653 Pa respectively.
108
1000
~
c.
JL
.--
900
800
C-
0
Q)
~
tj)
~
Q)
~
c.
....ttl
~
500 -
-.
t
1
400 300
200
100
~o
(..!j
600 -
.=:
::-
700 -
Oppm
10ppm
50ppm
I
1000
c.
.--
..!L
c.
800
r-
Q
Q)
r-
;;
600 -
o
o
o
o
~
Q)
r-
C.
400 -
c
0
.....~
.'t::
200
~D
Oppm
10ppm
::-
...
(.!J
50ppm
I
109
1000
~
c.
....-
JL
...
800 -
c..
~~
!}
...
...
Q)
600 -
V;
tJ)
Q)
...
c.
400 -
=
0
....
200
~D
Oppm
10ppm
,;t=
>
~
...
50ppm
I
This section consists of two parts. In the first, a quantitative description of the
contribution of each component to the total slug pressure drop will be provided. In the
second part, the effect ofDRA on each component shall be discussed for the flow of the
50 cP oil. It is important to mention that the data obtained from the flow of the 50 cP
oil were developed through an experimental work at the Center for Corrosion and
Multiphase Technology carried out by Derek Bleyle, who also measured flow
characteristics, such as slug frequency, velocity and length. These data were fed into
the different models used in this study to break down the pressure drop into its
110
components and to estimates the changes in the components of pressure drop as a result
of adding the DRA.
calculated and measured values were in good agreement. The error percent ranged from
4% to 20%.
Figure 5.4.1.1 describes the changes in pressure drop and its three
component increased with increasing superficial gas velocity, while the frictional
component had fairly constant values.
velocity from 2 to 6 mls was followed with an increase in total pressure drop from 1865
Pa to 4730 Pa, and accelerational component from 1067 Pa to 3589 Pa.
The frictional component increased slightly from 798 Pa to 1140 Pa. Similar
results were found at different superficial liquid velocities of 1 and 1.5 mls.
The
addition of the DRA minimized the increase in the pressure drop and its components
due to the increase in superficial gas velocity as will be seen later.
111
6000
5000
Total Measured
\J
Frictional
(10
CL.
I: 4000
e,
\J
~ 3000 -
Q)
l-
~ 2000
rq
Q)
l-
e.
1000
0
I
15
12
(10
0-
-0
il
~i
Q
I-
s::
(10
=
=.:
~c,
I-
'V
AcceIerational
Frictional
o
0
6 -
If
Q,)
l-
0-
9 -
Q~
Q,)
TotaiMeasured
3 -
\l
\J
112
15
12
.. 0
v
Q.
i1-.
.... ~
c.c
...Q
Q =
Q
Q)
=1~'-'
...
Accelerational
Frictional
[I
6 'V
Q)
e.
Total Calculated
~o
~
~
... .c
Total Measured
3 -
mis, respectively.
Increasing superficial liquid velocity had insignificant influence on the
contribution of both components to total pressure drop. For example, at superficial gas
velocity of 2 mls and DRA concentration of 0 ppm, frictional component had the
fractions of 43%,42%, and 42% at superficial liquid velocities of 0.5, 1, and 1.5 mis,
respectively as shown in Figures 5.4.1.1 through 5.4.1.3. Similar results were found at
all superficial gas velocities and DRA concentrations.
113
o
6
9000
8000
c. 7000
~ 6000
Q 5000
~
~
~
C-
4000
...
Total Measured
Total Calculated
Accelerational
Frictional
3000
o
2000
f-
1000
f-
\l
114
15
12 -
('0
C.
~
II
Tmal t4easlI"ed
Total Calculated
Accelerational
~~
c.tJ'J
o-C
.... s:::
-0
l-
Frictional
('0
Q) ~
0
"';.s:::
~~
Q)'-'
....
c..
rq
0
2
4000
('0
~ 3000
foo-
..!L
c-
....Q)
Total Measured
Total Calculated
rq
Accelerational
o Frictional
....0
Q
2000 -
".;
o
o
~
Q)
....
c. 1000 -
V'
0
115
10
9 ...
Total Measured
Total Calculated
r0-
J[
ro-
c-C
ro-O
e,
c.~
c. . =
...-Q)
i.e
Q)C
...
xr Acceler ational
Frictional
3 -
0.
2 -
1I
15
12
(10
e.
ii~
tJ1)
Total MeaslIed
Total Calculated
\J
Accelerational
Frictional
.......
c..-C
c
0
...
... ~
:-=
Q)
~c
'1
Q)
'1
l-
e.
3
0
6
4
Superficial GasVelocity [=) m/s
Figure5.4.1.9 Pressure DropVs. Superficial GasVelocity
100% Oil (50 cP), Vsl= 1.5 mIs, 50 ppm, Horizontal Flow
116
5.4.2 Effectiveness of DRA's in Reducing Pressure Drop
DRA was added into the flowing stream to examine its effect on each
component. Figures 5.4.2.1 through 5.4.2.3 describe the effect of DRA on frictional
component at superficial liquid velocities of 0.5, 1, and 1.5 mis, respectively. The
addition of DRA was found to decrease frictional component at all superficial liquid
velocities and DRA concentrations. Figure 5.4.2.1 shows that at superficial liquid and
gas velocities of 0.5 & 4 mis, respectively, frictional component had the values of 1151,
926, and 800 Pa at DRA concentrations of 0, 20, and 50 ppm, respectively.
The
corresponding values at superficial liquid and gas velocities of 1 and 6 mls were 2885,
2350 and 2004 Pa, respectively.
1500
('0
Q.
I:
c.
...
c
...
1200 -
CI,)
900
Oppm
20 ppm
50 ppm
...
CI,)
c.
('0
C
Q
';:
...
u-
117
4000...----..--------------------.
(10
o 20ppm
Q.
JC
Oppm
3000.- 0
50 ppm
c.
Q
'-
-Q)
2000
o
o
f-
Q)
c.
(10
1000 . .
;:
CJ
U.
5000
Oppm
20ppm
50ppm
(10
Q.
JC
c.
...
Q
...
4000 ...
3000
f-
Q)
Q)
0
0
2000 -
c.
Q
.';::
1000 -
u.
n
Superficial Gas Velocity (=] m/s
118
Accelerational component was found to decrease with the addition of DRA.
Figures 5.4.2.4 through 5.4.2.6 indicate the influence of the addition of the DRA on the
accelerational component.
(10
4000
Oppm
20 ppm
-0
50 ppm
e,
JL
ce
...
...
3000
Q)
V;
fA
...
c.
Q)
2000
(10
0
.';=
...
1000 -
(10
Q)
Q)
(.)
(.)
eX
0
2 4 6
119
8000
Q.
JL
c..
...0
c
...
Q)
VJ
...
c..
Q)
Oppm
7000
-0
20ppm
6000
-0
50ppm
5000
4000
3000 -
0
.';::
ttl
2000
1000
...
Q)
(,J
(,J
eX
10
(0
Oppm
20ppm
f-O
50ppm
="'t::C
... ~
f-
,..
(10'"
f-
cC
Q
3f-
...
c.
JL
c..
...
c
0
Q)
...
..-..
VJ
VJ
VJ
(0
c..
Q)
;:
(10
f-
2.
(1)
(J
(J
eX
120
An experimental studies
utilizing 2.5 cP and 26 cP and 50 cP oils were carried out in the system described in
chapter three. Computational analysis was also done. Measured and calculated values
were compared.
121
total pressure drop was greater in the 2.5 cP oil than in the 26 cP one. This is because
the DRA was more effective in reducing accelerational component, the dominant
contributor, in the 2.5 cP oil than in the 26 cP oil as will be explained later.
For
example, It can be seen from Figure 5.5.1.3 that at superficial liquid and gas velocities
of 1.5 & 6 mis, respectively, increasing DRA concentration from 0 to 50 ppm caused
41 % reduction in total pressure drop for the 2.5 cP oil from 8529 to 5033 Pa. The
corresponding reduction in total pressure drop for the 26 cP oil was only 21 % from
8764 to 6944 Pa. Figure 5.5.1.1 shows that at superficial liquid and gas velocities of 0.5
The
corresponding reduction in total pressure drop for the 26 cP oil was 16% from 2740 to
2307 Pa.
As oil viscosity increased, height of liquid film was noticed to increase at the
same liquid and gas flow rates resulting in greater accelerational component for the 26
cP oil than for the 2.5 cP oil. This difference in accelerational component due to oil
viscosity was even much greater after the addition of DRA. In addition to that, slug
translational velocity increased with increasing oil viscosity resulting in a greater force
required to accelerate the slow liquid film ahead of the slug to the slug velocity, thus
increasing accelerational pressure drop.
122
4000
oppm
2.5cP,
26cP,Oppm
CtI
c.
3000
C-
...
Q
e
...
Q)
..-I
2000 -
...
Q)
.L
C.
.....Q
26cP,50ppm
1000
iy
T
-l
Jo-
7000
c. 6000
j[
c. 5000
...
('10
\l
26cP, 50 ppm
...
(1)
4000 -
..-I
...
C.
3000 -
....e
2000 -
(1)
('10
Jo-
(J
\J
1000
~
I
123
10.....- - - - -.......- - - - - - - - - - - - - - - - - .
8 _. 26cP,Oppm
o 2.5cP,50ppm
6
26 cP,50ppm
i
1
'T
!
Q
o..l..
Figures 5.5.1.4 through 5.5.1.6 describe the effects of oil viscosity on DRA
efficiency to reduce accelerational component and its contribution at superficial liquid
velocities of 0.5, 1, and 1.5 mls respectively. Figure 5.5.1.5 indicates that at DRA
concentration of 0 ppm and superficial liquid and gas velocities of 1 & 4 mis,
respectively, accelerational component increased from 2514 to 3147 Pa when increasing
oil viscosity from 2.5 to 26 cl'. The corresponding slug translational velocity increased
from 6.2 to 7.6 mls as a result of increasing oil viscosity from 2.5 to 26 cl'.
The
corresponding increase, due to an increase in oil viscosity from 2.5 to 26 cP, in the
accelerational component at DRA concentration of 50 ppm was from 1110 to 2868 Pa.
124
ttl
3000
c.
JL
e,
...
...
Q
Q)
26 cP, Oppm
26 cP,50ppm
2000 -
V;
2.5cP,Oppm
fA
...
c..
Q)
ttl
c:
1000 -
.1
...
(1;1
.,.
.1.
~'
~
g.
';
Q)
Q)
Co.)
Co.)
6000...------......- - - - - - - - - - - - - - - - - . .
2.5cP,oppm
ttl
c.
---jL--5000
~.
--26-cP;-0'ppm-
Q..
::
Q
4000
v;
fA
...
c..
Q)
ttl
C
3000
2000
...
'7
26cP, 50 ppm
'V
'OJ
';
ttl
1000
Q)
Q)
u
u
-=0::
125
9000
~
c.
8000
c.
e
7000
6000
JL
~
Q)
~
t;
5000
2.5cP,Oppm
26 cP,Oppm
2.5cP,50ppm
26cP,50ppm
VJ
Q)
4000
3000 -
c.
e
,.
J.
-.;
~
2000 -
Q)
Q)
(.)
(.)
<
1000 -
Co
The associated change in slug velocity was fonn 6.5 to 7.72 mls. Such increase
in slug velocity requires greater force to accelerate liquid film ahead of a slug to the
slug velocity. The greater force manifests itself as greater accelerational pressure drop.
Similar to results found for the total pressure drop, DRA was found more
effective in reducing accelerational component for the oil of lower viscosity. At certain
superficial liquid and gas velocities and DRA concentration, the height of liquid film in
the 2.5 cP oil experienced more decrease than in the 26 cP oil. This cut-down in the
height of the liquid film was accompanied with proportional spread of the liquid film
around the pipe circumference. Such decrease in the height of the liquid film was
responsible for the reduction in the rate of mass pickup by the slug and hence the
126
accelerational component of total pressure drop. It was noticed that the height of the
liquid film decreased continuously at higher DRA concentrations and superficial gas
velocities, regardless of oil viscosity, until it reached minimum value after which
transition in the flow pattern from slug to annular could take place.
Figure 5.5.1.6 shows that at superficial liquid and gas velocities of 1.5 & 6 mis,
respectively, increasing DRA concentration from 0 to 50 ppm caused 46% reduction in
accelerational component of pressure drop for the 2.5 cP oil from 7334 to 3982 Pa. The
corresponding reduction for the 26 cP oil was only 21 % from 7424 to 5849 Pa. Similar
results were found at all superficial liquid velocities.
indicates that at superficial liquid and gas velocities of 0.5 & 6 mls a reduction of 42%
in the accelerational component occurred when increasing DRA concentration from 0 to
50 ppm for the 2.5 cP oil.
component for the 26 cP oil was only 6%, from 1734 to 1624 Pa.
Oil density has significant impact on gravitational component. The 26 cP oil has
a density of 820 Kg/m 3 whereas the 2.5 cP oil has a density of 800 kg/m'. Comparing
results of gravitational component for both oils shows that this contributor was much
greater for the 26 cP oil than the 2.5 cP oil as it should be if density was the only factor
determining gravitational contribution to total pressure drop. Slug liquid holdup was
found greater in the 26 cP oil than its corresponding values in the 2.5 cP oil at all
superficial liquid and gas velocities and DRA concentrations, possibly due to its higher
surface tension. This gives another reason why gravitational component was, markedly,
greater in the 26 cP oil than in the 2.5 cP oil.
127
Figures 5.5.1.7 through 5.5.1.9 describe the effect of oil viscosity on both DRA
effectiveness and the magnitude of accelerational component at superficial liquid
velocities of 0.5, 1, and 1.5 mls respectively.
One can find that at DRA concentration of 0 ppm and superficial liquid and gas
velocities of 1 & 4 mis, a 25% increase in gravitational component took place form 609
to 759 Pa, as shown in Figure 5.5.1.8, whereas the increase in oil density did not reach
2.5%. The corresponding increase in gravitational component at DRA concentration of
50 ppm was 17%, from 555 to 647 Pa. Similar results were found at all superficial
liquid velocities and DRA concentrations.
1000
~
c..
900
JL
800
700
......
c.
Q
Q
Q)
~
(fJ
600 500 -
Q)
~
c..
400
300
c
0
.....
~
::>
~
~.
200
100
(.!:)
--
2.5cP,Oppm
26cP, 0ppm
2.5 cP,50 ppm
v 26cP,50ppm
2
128
1000
('0
c.
..!L
C-
...e
...
800
Q)
\l
600
;;
\l
CIJ
...
c.
Q)
400
~.
200
(0
c
0
....
(0
::
...
26cP,Oppm
\l
26cP,50ppm
io
2.5cP,Oppm
::('0
1000
~
c,
,....
JL
c.
...
...
800 -
Q)
600 -
;;
CIJ
...
Q,)
c.
400
(t1
=
....
Q
200
.t::
:>
...
~
-.
~
-0
\J
26cP,~Oppm
2 4 6
129
DRA was found more effective in reducing gravitational component of pressure
drop for the oil of higher density and viscosity. One can see form Figure 5.5.1.8 that at
superficial liquid and gas velocities of 1 & 6 mis, respectively, increasing DRA
concentration from 0 to 50 ppm caused 28% reduction in gravitational component of
pressure drop for the 26 cP oil from 687 to 555 Pa. The corresponding reduction for the
2.5 cP oil was only 8% from 525 to 485 Pa. Similar results were found at all superficial
liquid velocities.
Frictional component of pressure drop was greater in the flow of the 26 cP oil
since the height of liquid film, thus Froude number, where found greater. As shown in
Figure 5.5.2.1, At certain liquid and velocities and DRA concentration, the height of the
liquid film ahead of a slug in the case of 26 cP oil was greater than its corresponding in
the flow of 2.5 cP oil. The difference in frictional drag for the two types of oil was
minimized at higher DRA concentrations and superficial gas velocities since both oils
had almost equivalent height of liquid film below which a transition in flow pattern
could happen.
Figures 5.5.1.10 through 5.5.1.12 show the effect of oil viscosity on frictional
loss and DRA efficiency in minimizing this loss at superficial liquid velocities of 0.5, 1,
and 1.5 mls respectively. Figure 5.5.1.10 indicates that at DRA concentration of 0 ppm
and superficial liquid and gas velocities of 0.5 & 6 mis, respectively, frictional
component increased significantly from 163 to 394 Pa when oil viscosity increased
from 2.5 to 26 cl'. At superficial liquid velocity of 1 mis, the corresponding increase
was from 173 to 468 Pa as shown in Figure 5.5.1.11. Similar results were noticed with
130
the presence ofDRA. For example, Figure 5.5.1.12 shows that at DRA concentration
of 50 ppm and superficial liquid and gas velocities of 1.5 & 4 respectively, frictional
component increased from 374 to 391 Pa for the same increase in oil viscosity. One can
see also that the percent of increase in frictional component, due to the increase in oil
viscosity, was less with the presence ofDRA.
500
:I 400 -
2.5cP,Oppm
Q.
e,
...
Q
26cP,Oppm
2.5cP,50ppm
c 300
-v 26cP,50ppm
Q)
...
.",
...
Q.
Q)
0
0:;
200 -
100 ~
(.)
...
u..
\l
'7
131
1000
I.'t:I
e.
800 -
c.
Q
....
e
......,Q)
~
....
Q)
e.
600
2.5cP,Oppm
26cP,Oppm
\J
26cP,50ppm
200 -
c
0
....
u.
.l
400 -
,.
n
Superficial Gas Velocity (=] m/s
Figure5.5.1.11 Effect of Viscosity on Frictional Pressure Drop
100% Oil (2.5 &26 cP), Vsl= 1 mis, 2-Degree Upward
1000
~
e.
....c.
800 ...
e
Q)
....
600 -
ii
2.5cP,Oppm
26cP,Oppm
\J
26cP,50ppm
....Q
..
l"
....OJ
e.
400
..L
C
Q
200
....
u.
132
DRA had greater influence on frictional component for the 26 cP oil than for the
2.5 cP oil. Figure 5.5.1.10 shows that at superficial liquid and gas velocities of 0.5 & 6
The
corresponding reduction in frictional pressure drop for the 2.5 cP oil was 9% from 163
to 149 Pat Another example can be seen in Figure 5.5.1.11. This Figure shows that at
superficial liquid and gas velocities of 1 & 4 m/s respectively, an increasing in DRA
concentration from 0 to 50 ppm was accompanied with 240/0 reduction in the frictional
component for the 26 cP oil, while it almost did not change for the 2.5 cP oil.
viscosity on total pressure drop and overall DRA effectiveness for the flow of 2.5 and
50 cP oils at superficial liquid velocities of 0.5, 1, and 1.5 mis, respectively. Comparing
results of total pressure drop for both oils in horizontal flow indicates that total pressure
drop for the 50 cP oil was always greater than its corresponding value for the 2.5 cP oil.
Figure 5.5.2.3 shows that at superficial liquid and gas velocities of 1.5 & 6 m/s and at a
D~RA
concentration of 0 ppm, total pressure drop for the 50 cP oil was 12030 Pa
concentration of 50 ppm were 7399 and 3591 Pat Similar results were found at all
superficial liquid velocities and DRA concentration. One can notice also that the DRA
133
effectiveness in reducing total pressure drop was greater in the 2.5 cP oil than in the 50
cP one. For example, Figure 5.5.2.2 shows that at superficial liquid and gas velocities
of 1 & 4 mis, respectively, increasing DRA concentration from 0 to 50 ppm caused 490/0
reduction in total pressure drop for the 2.5 cP oil from 1814 to 885 Pa.
The
corresponding reduction in total pressure drop for the 50 cP oil was 36% from 7191 to
5449 Pa. Figure 5.5.2.3 shows that at superficial liquid and gas velocities of 1.5 & 4
m/s, respectively, increasing DRA concentration from 0 to 50 ppm resulted in 480/0
reduction in total pressure drop for the 2.5 cP oil from 3842 to 1841 Pa.
The
corresponding reduction in total pressure drop for the 50 cP oil was 31% from 10109 to
8474 Pa.
6000
(tI
5000
c.
-.
-0
.,.
50 cP, 0 ppm
()
'iJ
50 cP, 50 ppm
4000 -
c.
c 3000 _
Q)
...
\J
~ 2000 ~
C.
'V
134
9~
V'
50cP, 50 ppm
.....
6-
'"
!
3-
(I)
15
~
c..
12
to-
E
c.
...
...
Cl
Q)
'"
50cP,Oppm
f-V'
50cP,50ppm
fA
2.5cP,Oppm
.,.
~
0
~ .c
Q)
6 -
....
3-
...
e,
(10
'V
t-
o
0
2 4 6
Superficial Gas Velocity (=] m/s
Fi gu re 5.5.2.3 Effect of Vi seo sity on Total Pres su re Drop
100% Oil (2.5 &50 cP), Vsl= 1.5 mis, Horizontal
135
Figures 5.5.2.4 through 5.5.2.6 describe the effects of oil viscosity on the DRA
efficiency to reduce accelerational component and its contribution at superficial liquid
velocities of 0.5, 1, and 1.5 m/s respectively.
Similarly, DRA was found more effective in reducing accelerational component
for the oil of lower viscosity. At certain superficial liquid and gas velocities and DRA
concentration, the height of liquid film in 2.5 cP oil decreased more significantly than in
the 50 cP oil. This cut-down in the height of the liquid film was accompanied with
proportional spread of the liquid film around the pipe circumference. Such decrease in
the height of the liquid film was responsible for the reduction in the rate of mass pickup
b)' the slug and hence the accelerational components of total pressure drop.
5000
.....
JL
c-
4000
...o
Q
...
Q)
3000
'9
50cP,O ppm
-'\j
fy
2000 -
C
Q
-..:
...
~
1000 "'"
Q)
CJ
(J
Q)
50cP,50ppm
fA
Q)
...
c.
I
.1
136
~
c. 8000
JL 7000 -
.--
c.
Q
6000 -
2.5cP,Oppm
'"
50 cP,O ppm
Q)
~
;; 5000 - \} 50cP,50ppm
fA
Q)
4000
c.
:= 3000 ,..
2000 ....
'V
..2:
eX
1000 -
Q)
(.)
(.)
'V
'"
Fi gure 5 .5.2.5 Effect of Vise osity on Ac cel eratio nal Press ure Drop
100% Oil (2.5 &50 cP), Vsl= 1 mIs, Horizontal
10
c..
--.
JL
9~
.
Q)
U;
~
Q)
~
c.
~
c:
"'C
c:
~
.c
CI
......
-\}
4~
Q)
Q)
(.)
(.)
eX
50 cP, 0 ppm
.l
50 cP, 50 ppm
53
..
;:
.,.
8~
c.
1-
!
\}
137
Figure 5.5.2.5 shows that at superficial liquid and gas velocities of 1 & 4 mis,
respectively, increasing DRA concentration from 0 to 50 ppm caused 49% reduction in
accelerational component of pressure drop for the 2.5 cP oil from 1605 to 824 Pa. The
corresponding reduction for the 50 cP oil was 32% from 4190 to 2863 Pa.
results were found at all superficial liquid velocities.
Similar
indicates that at superficial liquid and gas velocities of 1.5 & 2 mis, respectively, a
reduction of 61% in the accelerational component occurred when increasing DRA
concentration from 0 to 50 ppm for the 2.5 cP oil. Whereas the corresponding decrease
in accelerational component for the 50 cP oil was 35%.
Frictional component of
pressure drop was greater in the flow of the 50 cP oil than in the flow of 2.5 cP oil due
to the difference in viscosity.
Figures 5.5.2.7 through 5.5.2.9 show the effect of oil viscosity on frictional loss
and on the DRA efficiency in minimizing this loss at superficial liquid velocities of 0.5,
1, and 1.5 mls respectively. Figure 5.5.2.7 indicates that at DRA concentration of 0
ppm and superficial liquid and gas velocities of 0.5 & 6 mis, respectively, frictional
component increased significantly from 197 to 1141 Pa when oil viscosity increased
from 2.5 to 50 cP. At superficial liquid velocity of 1 mis, the corresponding increase
was from 367 to 2885 Pa as shown in Figure 5.5.2.8. Similar results were noticed with
th.e presence ofDRA. For example, Figure 5.5.2.9 shows that at DRA concentration of
50 ppm and superficial liquid and gas velocities of 1.5 & 4 respectively, frictional
component increased from 223 to 2665 Pa for the same increase in oil viscosity.
138
1500
ttl
C.
--
ii 1200
50 cP, 0 ppm
'\J
50 cP!,.50 ppm
c.
Q
Q)
...
900
c.
'i
Q)
.,.
600 ....
C
Q
;::
300
CJ
~
U.
Fi gure 5.5.2.7 Effe ct of Vis cos ity on Friction al Press ure Drop
100% Oil (2.5 &50 cP), Vsl=0.5 mis, Horizontal
5000..------.....------------------.
I:
4000
2.5cP,Oppm
.,.
50cP,Oppm
2.5cP,50ppm
\J
50 cP, 50ppm
C-
C
Q)
3000 -
...
~
~
Q)
C.
2000 -
1.'0
C
Q
u..
1000
o.....-----'+~------------------'--11".
2 4 6
...
139
6000
tt'
c.
ii
..-..
e.
...
...
4800
...
50 cP, 0 ppm
'\l
50 cP, 50 ppm
f0-
Q)
3600
fo-
2400
fo-
1200
...
.,.
Q)
0-
\7
.,.
'I
.,.
'V
tt'
Q
.';
...
u.
I~
The addition of the DRA was followed with tremendous reductions in the
frictional component of total pressure drop in the flow of the 50 cP oil than in the 2.5
cPo
However, its effectiveness was greater in the 2.5 cP than in the 50 cPo Figure
5.5.2.7 shows that at superficial liquid and gas velocities of 0.5 & 6 mis, respectively,
in.creasing DRA concentration from 0 to 50 ppm resulted in a 28% reduction in
frictional pressure drop for the 50 cP oil from 1140 to 812 Pa. The corresponding
reduction in the frictional pressure drop for the 2.5 cP oil was 38% from 197 to 121 Pa.
Figure 5.5.2.9 shows that at superficial liquid and gas velocities of 1.5 & 6 mis,
respectively, increasing DRA concentration from 0 to 50 ppm resulted in a 23%
reduction in frictional pressure drop for the 50 cP oil from 3609 to 2781 Pa.
140
For
example, It can be seen from Figure 5.6.2 that at superficial liquid and gas velocities of
1 & 2 mis, respectively, increasing DRA concentration from 0 to 50 ppm caused 81 %
reduction in total pressure drop for the horizontal flow from 1457 to 283 Pa.
The
corresponding reduction in total pressure drop for the 2-degree upward flow was only
38% from 2675 to 1647 Pa.
141
5000
~
c..
:I
c0
...
c
...
4000
3000
,....
2-Degree, 0ppm
Horizontal,50 ppm
f-\J
2-Degree, 50ppm
(1)
-.,I
...
(1)
c..
.....~
0
....
2000 -
..L
1000 -
..L.
6000
~
c.. 5000
:I
g-
-.
2-Degree, 0 ppm
Horizontal, 50 ppm
\J
2-Degree, 50 ppm
Q)
3000 ...
Q)
'-
C.
2000
\J
(tI
....
r0-
.....
0
4000 -
c'-
......
Horizontal,O ppm
1000 -
142
10
8-
0-
il
-
c.
Horizontal, 0 ppm
2-Degree,0 ppm
(]
Horizontal, 50 ppm
or-.
t,IJ
o -c
'- c:
- \J
2-Degree, 50 ppm
~
0
..c
Q)
-c
'-
4-
....
Q)
e.
2f-
.......
,
.,.
!
Q
r-
Figure 5.6.2 shows also that for horizontal flow and at higher superficial gas
velocity of 6 mis, increasing DRA concentration form 0 to 50 ppm was accompanied
with 34% reduction in total pressure drop from 3333 to 2198 Pa. The corresponding
reduction in the inclined flow was found to be 35% from 4884 to 3176 Pa. Similar
results were found at all superficial liquid velocities.
Accelerational component of pressure drop was found to increase with
increasing pipeline inclination due to the significant increase in liquid film height,
hence the rate of mass pickup by the slug. Meanwhile, the velocity of the liquid film in
inclined flow was found to decrease, thus increasing the force required to accelerate this
slow moving film to slug translational velocity. Figures 5.6.4 through 5.6.6 describe the
effects of pipe inclination on DRA efficiency to reduce accelerational component and
143
the contribution of this component to total pressure drop at superficial liquid velocities
of 0.5, 1, and 1.5 mls respectively. Figure 5.6.4 indicates that at DRA concentration of
ppm and superficial liquid and gas velocities of 0.5 & 2 mis, respectively,
3000....-------........- - - - - - - - - - - - - - - - .
2000
Horizontal,Oppm
2-Degree,O ppm
~ 0
'V
Horizontal,50ppm
2-Degree,50 ppm
!
(10
e==
1000 ~
I
:1::
Y:
0"--------------.
. .4. -------.-..10.---.
...
2
6
I
144
c.
......
.!!...
c.
...
c
5000
4000 -
0
...
Horizontal, 0 ppm
2-Degree,Oppm
Hori zontal, 50 ppm
Q,)
VJ
...
c.
Q)
2000 -
.';::
...
(t1
1000
Q)
'V
'V
'V
Q,)
u
u
0
I
At low superficial gas velocity of 0.5 mis, DRA was found more effective in
reducing accelerational component for horizontal flow. The opposite was true at higher
superficial gas velocities of 4 and 6 mls. Figure 5.6.6 shows that at superficial liquid
and gas velocities of 1.5 & 2 mis, respectively, increasing DRA concentration from 0 to
50 ppm in horizontal flow caused 61% reduction in accelerational component of
pressure from 1989 to 768 Pa. The corresponding reduction in the 2-degree upward
flow was 40% from 2321 to 1386 Pa. The same Figure shows at higher superficial gas
velocity of 6 mis, The corresponding reduction in accelerational component in
horizontal flow was 34% from 4957 to 3247 Pa, whereas in 2-degree upward flow it
decreased 46% from 7334 to 3982 Pa. Similar results were found at other superficial
145
liquid velocities of 0.5 and 1 mls as can seen In Figures 11.4.4.4 & 11.4.4.5
respectively.
9000
~
c,
.......
JL
Q.
Q
...
...
8000 7000 -
Horizontal, 0 ppm
2-Degree,Oppm
Horizontal, 50 ppm
\J
2-Degree,50ppm
6000 -
Q)
(ij
5000 -
...
Q)
4000 -
;::
...
3000 2000
1000
Q)
T
T
C
Q
1
!
t.n
c..
-;;
Q)
'V
'V
t
.J.
Without the presence of DRA, frictional drag was greater in horizontal flow.
With the presence of DRA, this component was found greater in the 2-degree upward
flow than in the horizontal flow, at the same liquid and gas flow rates. This is because
D'RA helped the liquid film to spread around the pipe circumference, especially in the
2~degree
upward flow where the amount of the liquid in the film was greater.
Figures 5.6.7 through 5.6.9 show the effect of pipe inclination on frictional loss
and DRA effectiveness at superficial liquid velocities of 0.5, 1, and 1.5 mls
respectively,
146
3 0 0 , . . - - - - - - - - - . - - - - - - - - - - - - - - -.....
200 ...
Horizontal,Oppm
2-Degree, 0 ppm
Horizonta1,50 ppm
\l
2-Degree, 50 ppm
100 --
-l
Oa...-------I----------a...--------'-------I
6
2
4
I
500
('0
c.
:I
c.
...
c
...
Q
400 -
300
200
Q)
('0
=
-.;:
2-Degree,Oppm
Horizontal,50 ppm
100
11
2-Dearee. 50 DDm
..",
...
c.
Horizontal,0 ppm
,-'V
Q)
1
T
6i
,
'T
(.)
...
u..
147
1000
~
0-
J[
c.
800 -
l-
Horizontal, 0 ppm
2-Degree,Oppm
Hori zonta 1,50ppm
Q
Q)
I-
600
-\J
2-Degree,50ppm
.."
...
0Q)
400 -
\J
"V
'V
1."0
~
200 0
"-
u.
Figure 5.6.8 indicates that at DRA concentration of 0 ppm and superficial liquid
and gas velocities of 1 & 6 mis, respectively, frictional component decreased
significantly from 36'8 to 172 Pa when moving from horizontal to 2-degree upward
flow. At DRA concentration of 50 ppm and at the same superficial liquid and gas
velocities, the corresponding change in frictional component was an increase from 229
to 287 Pa. Figures 5.6.7 and 5.6.9 show similar results at superficial liquid velocities of
0.5 and 1.5 mls respectively.
Finally, DRA was found more effective in decreasing frictional component in
horizontal flow than in 2-degree upward flow.
concentration of free polymer molecules in 2-degree upward flow because of its lower
shear stresses. One can see from Figure 5.6.9 that in horizontal flow and at superficial
148
liquid and gas velocities of 1.5 & 2 mls respectively, increasing DRA concentration
from 0 to 50 ppm caused 73% reduction in frictional pressure drop from 465 to 127 Pa.
The corresponding reduction in frictional pressure drop for the 2-degree upward flow
was 17% from 465 to 382 Pa. It is important to remember that the frictional component
was found to increase after the addition of DRA at certain superficial liquid and gas
velocities in the inclined flow but not in the horizontal flow. These increased values
were overestimated.
149
CHAPTER SIX
CONCLUSIONS
Experiments have been carried out utilizing different kinds of oil with viscosity
values of 2.5 cP 26 cP and 50 cl'. The effects of oil viscosity and pipe inclination on
flow characteristics, each component of the pressure drop, and on the DRA
effectiveness were studied. Liquid and gas velocities were varying in the range from
0.5 to 2 m/s and from 2 to 14 m/s respectively. The DRA was added in dosages of 10,
20, and 50 ppm. All experiments were performed in 20-m long, 10.16-cm ID inclinable
Plexiglas pipeline. The pipeline was set to two inclinations; namely horizontal and twodegree.
Quantitative analysis along with experimental work has been undertaken to
estimate the contribution of pressure drop components in slug flow. The impacts of oil
viscosity and pipe inclination were studied and the influence of the addition of drag
reducing agents on each component of pressure drop and on total drag reduction was
also predicted. A quantitative analysis of the contribution of each component to total
pressure drop, the effectiveness ofDRA, and fractional drag reduction was given for the
flow of 2.5 cP oil in both horizontal and 2-degree inclined pipes, the flow of 26 cP oil in
2-degree inclined pipes, and the flow of 50 cP oil in horizontal pipes. Effects of oil
viscosity and pipe inclination on the flow hydraulics and the performance of the DRA's
in reducing total pressure drop and each of its components individually were discussed.
150
The pressure drop in horizontal slug flow using 2.5 cP oil was calculated by
estimating both frictional and accelerational components. It was found that the main
component of pressure gradient is the accelerational component where its percentage to
the total pressure gradient ranges from 77% to 89%. The remainder was the frictional
contributor.
The drag reducing agent, DRA, used in this study was found to be effective in
reducing the pressure drop at all superficial liquid and gas velocities. A dosage of 20
ppm of DRA caused the pressure drop and both of its components to decrease. Further
addition of DRA to a concentration of 50 ppm was accompanied with more drag
reduction and higher effectiveness of the DRA.
In sharp contrast to what is believed, most of the gained drag reduction took
place in the accelerational component, while smaller fraction was attributed to the drag
reduction gained in the frictional component.
Drag reduction was possible due to the decrease in the magnitude of each
component, as a result of changing flow characteristics, and the average slug frequency.
Smaller slugs merged to form longer and more stable slugs. Such merge would slightly
increase frictional loss in the slug body.
appeared. This contributor had its maximum values at high superficial liquid velocity
151
and low superficial gas velocities.
152
responsible for the reduction in the rate of mass pickup by the slug and hence the
accelerational component of total pressure drop. It was noticed that the height of the
liquid film decreased continuously at higher DRA concentrations and superficial gas
velocities, regardless of oil viscosity, until it reached minimum value after which
transition in the flow pattern from slug to annular could take place.
Slug liquid holdup was found greater in the 26 cP oil than its corresponding
values in the 2.5 cP oil at all superficial liquid and gas velocities and DRA
concentrations, possibly due to its higher surface tension. This gives another reason
why gravitational component was, markedly, greater in the 26 cP oil than in the 2.5 cP
oil.
DRA was found more effective in reducing gravitational component of pressure
drop for the oil of higher density and viscosity. Meanwhile, frictional component of
pressure drop was greater in the flow of the 26 cP oil since the height of liquid film,
thus Froude number, where found greater.
Opposite to what was previously found in the flow of the 2.5 cP oil, the
accelerational component was no longer the dominant contributor to total pressure drop
when higher oil viscosity, 50 cP, was examined. At higher superficial liquid velocities,
frictional component had fractions as high as 45% of total pressure drop.
The total pressure drop for the 50 cP oil was always greater than its
corresponding value for the 2.5 cP oil and the DRA effectiveness in reducing total
pressure drop was greater in the 2.5 cP oil than in the 50 cP one.
153
DRA was found more effective in reducing accelerational component for the oil
of lower viscosity, 2.5 cl'. At certain superficial liquid and gas velocities and DRA
concentration, the height of liquid film in 2.5 cP oil decreased more significantly than in
the 50 cP oil. This cut-down in the height of the liquid film was accompanied with
proportional spread of the liquid film around the pipe circumference. Such decrease in
the height of the liquid film was responsible for the reduction in the rate of mass pickup
by the slug and hence the accelerational components of total pressure drop.
The
reduction in the slug frequency due to the addition of the DRA was greater in the case
of lower viscosity.
While frictional component of pressure drop was greater in the flow of the 50 cP
oil than in the flow of 2.5 cP oil, the addition of the DRA was followed with
tremendous reductions in the frictional component of total pressure drop in the flow of
the 50 cP oil than in the 2.5 cP.
154
Without the presence of DRA, frictional drag was greater in horizontal flow.
With the presence of DRA, this component was found greater in the 2-degree upward
flow than in the horizontal flow, at the same liquid and gas flow rates. This is because
DRA helped the liquid film to spread around the pipe circumference, especially in the
2-degree upward flow where the amount of the liquid in the film was greater.
The DRA was found more effective in decreasing frictional component In
horizontal flow than in 2-degree upward flow.
concentration of free polymer molecules in 2-degree upward flow because of its lower
shear stresses.
Finally, I would like to point out that this work is the first work of its kind to
predict in a quantitative manner the components of pressure drop in slug flow with the
presence of DRA's. Moreover, this work will provide a clear understanding, based on
experimental findings and computational analysis, of the various effects of oil viscosity
and pipe inclination on the performance of DAR's in reducing the different components
of pressure drop in slug flow and on the contribution of each of these components to
total pressure drop.
Quantitative analysis of pressure drop will help introduce new mechanisms for
drag reductions in multiphase flow.
155
This work will help provide a new mechanistic or empirical correlation to
incorporate drag reduction in two-phase or multiphase flow.
Calculations in such
proposed model would be very helpful in design processes, eg. Predicting pumping
requirements for a specific production rate or predicting pipeline flow rates.
156
BIBIOLOGRAPHY
DukIer, A. E. and Hubbard, M. G.: "A Model for Gas-Liquid Slug Flow in Horizontal
and Near Horizontal Tubes," Ind. Chern., Fundam., (1975) 14, No.4.
Taitel, Y. and Dukler, A. E.: "A Model for Predicting Flow Regime Transitions in
Horizontal and Near Horizontal Gas-Liquid Flow," AIChe J, (1976) 22, No.1.
Dukler, A. E., Wicks, M., and Cleveland, R. G.: "Frictional Pressure Drop in TwoPhase Flow: A Comparison of Existing Correlations for Pressure Loss and
Holdup," AICheJ, (1964) 10, No.1.
Fan, Z., Ruder, Z., and Hanratty, T. I.: "Pressure Profiles for Slugs in Horizontal
Pipelines," J Multiphase flow, (1993) 19, No.3, 421.
Andritsos, N. and Hanratty, T. J.: "Influence of Interfacial Waves in Stratified GasLiquid Flows," AIChe J, (1987) 33, No.3.
Andritsos, N., Williams, L., and Hanratty, T. I.: "Effects of Liquid Viscosity on the
Stratified-Slug Transition in Horizontal Pipe Flow," Int. J Multiphase Flow,
(1989) 15, No.6, 877.
Greskovich, E. I. and A. L. Shrier: "Pressure Drop and Holdup in Horizontal Slug
Flow," AIChe J., (1971) 17, No.5.
Petalas, N. and Aziz, K.: "Development and Testing of A New Mechanistic Model for
Multiphase Flow in Pipes," Paper Presented at the 1996 ASME Fluid
engineering division conference, 236, 153.
Bamea, D. and Brauner, N.: "Holdup of Liquid Slug in Two-Phase Intermittent Flow,"
1. Multiphase Flow, (1985) 11, No.1, 43.
Rosehart, R. G., Rhodes, E., and Scott, D. S.: "Studies of Gas-Liquid (Non-Newtonian)
Slug Flow: Void Fraction Meter, Void Fraction and Slug Characteristics," The
Chemical Engineering L, (1975) 10,57.
Brauner, N., Maron, D. M., and Rovinsky, J.: "A Two-Fluid Model for Stratified Flows
With Curved Interfaces," International Journal of Multiphase Flow, (1998) 24,
975.
157
Vlachos, N. A., Paras, S. V., and Karabelas, A. J.: "Predictions of Holdup, Axial
Pressure Gradient and Wall Shear Stress in Wavy Stratified and
Stratified/Atomization Gas/Liquid Flow," International Journal of Multiphase
Flow, (1999) 25, 365.
Wilkens, R. J.: "Prediction of The Flow Regime Transitions in High Pressure, Large
Diameter, Inclined Multiphase Pipelines," Ph.D. Thesis, Ohio University (1997).
Mantripragada, V.: "A Study of The Effect of Inclination on Flow Regime Transitions,
Slug Flow Characteristics, And Corrosion Rates at Low Pressures," MS, Thesis,
Ohio University (1997).
Maley, L.: "A Study of Slug Characteristics in Large Diameter Horizontal Multiphase
Pipelines," MS. Thesis, Ohio University (1997).
Kouba, G. E. and Jepson, W. P.: "The Flow of Slugs in Horizontal Two-Phase
Pipelines," Transactions of the ASME, (1990) 20, 112.
Kang, C. and Jepson, W. P.: "Effects of Drag Reducing Agents in Multiphase Flow,
Oil/Gas Horizontal Flow," SPE 58976, Mexico, February 1-3, 2000.
Toms, B. A.: Proc. First Intern. Congo Rheology, pte 2, p. 135, North Holland,
Amsterdam, 1948.
Jepson, W. P. and Taylor, R. E.: "Slug Flow and Its Transitions in Large-Diameter
Horizontal Pipes," Int. J. Multiphase Flow, (1993) 19, No.3, 411.
Jepson, W. P.: "Modeling the Transition to Slug Flow in Horizontal Conduit," The
Canadian J. Chemical Engineering, 67, 731, 1989.
Lee, A-H., Sun, J-Y., and Jepson, W. P.: "Study of Flow Regime Transitions of OilWater-Gas Mixtures in Horizontal Pipelines," Proceedings of the Third
International Offshore and Polar Engineering Conference, II, 159, 1993.
Kang, C., Vancko, R. M., Green, A. S., Kerr, H., and Jepson, W. P.: "Effect of Drag
Reducing Agents in Multiphase Flow Pipelines," J. Energy Resources
Technology, 120, 15, 1998.
Tullius, L.: "Study of Drag Reducing Agents in Multiphase Flow in Large Diameter
Pipes," Ph.D. Thesis, Ohio University (2000).
Nicholas, D., Sylvester N. D., and Brill, J. P.: "Drag Reduction in Two-Phase AnnularMist Flow of Air and Water," AIChE J., 22, No.3, 615, 1976.
158
Sylvester, N. D., Dowling, R. H., and Brill, J. P.: "Drag Reduction in Cocurrent
Horizontal Gas-Hexane Pipe Flow," Polymer Engineering and Science, 20, No.
7, 485, 1980.
Reddy,G. V.: "Drag Reduction by Polymers in Recirculatory Flow," J. Chemical
Engineering World, XXI, No.6, 73, 1986.
Mansour, A. R., Swaiti, 0., Aldoss, T., and Issa, M.: "Drag Reduction in Turbulent
Crude Oil Pipelines Using a New Chemical Solvent," Int. J. Heat and Fluid
Flow, 9, No.3, 316,1988.
Lester, C. B.: "The Basics of Drag Reduction," Oil and Gas J., 51, Feb. 4, 1985.
Lester, C. B.: "What to Expect from and How to Handle Commercially available Grag
Reducing Agents ," Oil and Gas J., 117, Mar. 11, 1985.
Hom, A. F., Wu, C. D., Prilutski, D. J., and Motier, J. F: "High Viscosity Crude Drag
Reduction," Pipeline and Gas J., 22, June 1986.
Seyer, F. A. and Metzner, A. B.: "Turbulence in Drag Reducing Systems," AIChE J.,
15, No.3, 426, 1969.
Virk, P. S.: "Drag Reduction Fundamentals," AIChE J., 21, No.4, 625, 1975.
Rosehart, R. G., Scott, D. S., and Rhodes, E.: "Gas-Liquid Slug Flow with Drag
Reducing Polymer Solutions," AIChE J., 18, No.4, 744, 1972.
159
APPENDIX A
160
(mls)
Vsg
(mls)
20
50
2
4
4.5
4.1
6
2
4
6
2
4.9
4.5
4.2
5.3
4.2
3.9
3.4
4.6
4.3
4
5
4.7
4.4
Stratified
0.5
1.5
3.7
4.6
3.6
3.2
4.3
4
3.7
4.7
4.5
4.2
Vsg
(mls)
20
50
Stratified
7.4
9.4
3.4
0.5
4
6
5.8
8.1
4
6
2
4
6
4
6.4
8.7
4.7
3.6
6
8.3
4.3
6.4
8.8
4.8
7.1
7.3
9.4
9.4
1.5
6
8.4
4.4
6.4
8.9
4.9
161
Vsg
(oos)
(oos)
2
0.5
1
f--..
1.5
6
2
4
6
2
4
6
0
0.65
0.83
1
1.13
1.31
1.49
1.67
1.85
2.1
0.5
1.5
Vsg
(oos)
20
2
4
6
2
4
6
2
4
6
1.14
1.2
1.25
1.38
1.26
1.36
1.28
1.32
1.51
1.13
1.13
1.24
1.33
1.1
1.35
1.08
1.15
1.52
50
Stratified
1.14
1.15
1.13
1.19
1.3
1.01
1.28
1.24
162
Vsg
(mls)
(mls)
12
12
14
29
20
26
50
40
42
0.5
6
2
1
6
2
1.5
8
12
9
16
24
31
34
39
6
12
7
12
20
24
24
31
Vsg
(mls)
(mls)
2
4
6
2
4
6
2
4
6
571
1030
1630
1457
1815
3333
2455
3842
5646
0.5
1.5
354
624
1279
410
1203
2301
1375
1913
3395
50
Stratified
451
1112
295
885
2045
997
1841
2852
163
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
4.7
4.3
4
4.9
4.7
4.2
5.5
5.2
4.9
4.4
4.1
3.7
4.7
4.5
4.1
5.4
5
4.7
50
4.3
4
3.6
4.6
4.4
4
5.2
4.8
4.5
Vsg
(m/s)
2
4
6
1.5
2
4
6
2
4
6
3.2
5.7
8.1
4
6.2
8.6
4.3
6.9
9.3
50
3.2
5.9
8.3
4.1
6.4
8.9
4.4
7
9.5
3.3
5.9
8.4
4.1
6.5
9.1
4.5
7.2
9.7
164
Vsg
(m1s)
(mls)
0.5
2
4
6
2
4
6
2
4
1.5
0
0.2
0.8
1
1.1
1.3
1.5
1.7
1.8
2.1
50
0.2
0.9
1
1.2 . 1.4
1.5
1.7
1.9
2.2
Vsg
(m1s)
(mls)
2
4
0.87
0.89
0.89
0.88
087
0.79
1
1.16
1.08
0.5
6
1
2
4
6
1.5
2
4
165
Table A.l1 Measured Slug Frequency [=] min"
2.5 cP Oil, 2-Degree Upward Flow
Vsl
(mls)
0.5
1.5
Vsg
(mls)
2
4
6
2
4
6
2
4
6
18
19
18
42
30
37
61
52
58
14
14
13
32
17
30
51
43
49
50
11
14
12
25
15
23
38
34
36
Vsg
(mls)
2
0.5
1.5
6
2
4
6
2
4
6
1875
2459
2577
2938
3048
4285
3977
4330
5548
1452
1852
1972
2238
2287
3501
3010
3238
4399
50
1207
1643
1774
1857
1609
2994
2632
2609
3815
166
Table A.13 Measured height of Liquid Film [=] em
26 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(mls)
(mls)
0.5
1.5
2
4
6
2
4
6
2
4
6
0
5.4
5
4.6
5.8
5.4
5
10
4.6
4.2
5.4
4.9
4.7
50
4.8
4.5
3.9
5.1
4.7
4.2
Stratified
Stratified
Stratified
5.8
5.4
5.3
4.9
4.9
4.5
5~2
Vsg
(mls)
(mls)
2
4
6
2
4
6
2
4
6
4.3
6.8
9.5
5
7.6
10.3
4.3
7
9.7
5.1
7.5
10.7
50
4.4
7
9.8
5.1
7.7
10.9
Stratified
Stratified
Stratified
8.2
11.1
8.2
11.5
8.5
11.7
0.5
1.5
167
Table A.15 Measured Liquid Film Velocity [=] m/s
26 cP Oil, 2-Degree Upward Flow
Vsl
(m/s)
0.5
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
168
Table A.17 Measured Slug Frequency [=] min"
26 cP Oil, 2-Degree Upward Flow
Vsl
(m/s)
Vsg
(m/s)
2
0.5
6
2
6
2
1.5
0
18
18
15
38
29
37
10
16
17
15
37
32
37
50
16
18
18
33
34
34
Stratified
Stratified
Stratified.
48
52
45
52
54
51
Vsg
(m/s)
2
0.5
6
2
6
2
1.5
0
1938
2482
2777
3857
4210
5833
10
1970
2228
2652
3568
4068
5441
50
1733
2146
2646
3312
3822
5230
Stratified
Stratified
Stratified
6666
7615
6075
7179
5868
6905
169
Table A.19 Measured height of Liquid Film [=] em
50 cP Oil, Horizontal Flow
Vsl
Vsg
(m1s)
(m1s)
50
0.5
2
4
3.8
3.6
3.3
4.4
4
3.7
4.8
4.4
4
3.7
3.6
3.4
4
3.9
3.7
4.7
4.4
4
3.9
3.5
3
4.3
4
3.6
4.7
4.1
3.9
6
2
6
2
1.5
Vsg
(m1s)
(m1s)
0.5
2
4
3.7
6.3
8.3
4.1
6.7
9.6
5.3
7.5
9.8
3.6
6.4
8.5
3.7
6.5
9.6
5.1
7.5
9.6
2
4
1.5
2
4
50
3.6
6.4
8.5
3.7
6.6
9.7
5.2
7.7
9.5
170
Vsg
(mls)
(mls)
0.5
1.5
2
4
6
2
4
6
2
4
6
0
0.52
0.77
0.89
0.98
1.2
1.41
1.71
1.83
1.98
Vsg
(mls)
(mls)
0.5
1.5
2
4
6
2
4
6
2
4
6
0
2.2
2.2
1.8
2.1
2.8
2.6
1.9
2.2
2.4
50
1.9
1.9
1.7
1.9
2.2
2.3
1.8
2.2
2.6
171
Table A.23 Measured Slug Frequency [=] min"
50 cP Oil, Horizontal Flow
Vsl
(mls)
0.5
1.5
Vsg
(mls)
20
50
2
4
6
2
4
6
2
4
6
13
15
18
31
26
29
55
41
37
14
14
17
32
25
28
52
41
38
11
13
15
25
20
24
45
33
23
Vsg
(mls)
(mls)
0.5
2
4
2138
2824
3943
4663
5982
7349
7212
8708
10648
1.5
2
4
6
2
4
6
1870
2561
3637
4395
5714
7429
6916
8630
10549
50
1822
2401
3463
4184
5449
7266
6811
8474
10379
172
APPENDIXB
173
Vsg
(mls)
(mls)
2
4
6
2
4
6
2
4
6
4.7
8.9
13.5
4.6
8.6
12.8
4.6
8.4
12
0.5
1.5
5.2
9.5
14.6
5.4
8.9
13.4
4.8
50
Stratified
10
12.6
15.3
5.7
9.3 - - 14
5.2
9.4
13
Vsg
(mls)
(mls)
2
4
6
2
4
6
2
4
6
3.4
5.8
8
4
6.4
8.7
4.7
7
9.4
0.5
1.5
3.6
6
8.3
4.3
6.4
8.8
4.8
7.3
9.4
50
Stratified
6
8.4
4.4
6.4
8.9
4.9
7.4
9.4
174
Vsg
(m1s)
(m1s)
0.5
1.5
2
4
6
2
4
6
2
4
6
0
0.7
0.59
0.52
0.74
0.63
0.57
0.76
0.68
0.62
Vsg
(m1s)
(m1s)
0.5
1.5
2
4
6
2
4
6
2
4
6
0
79
117
197
269
209
368
465
460
689
175
Table B.5 Calculated Accelerational Component, L1Pa [=] Pa
2.5 cP Oil, Horizontal Flow
Vsl
(m/s)
0.5
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
491
913
1433
1188
1605
2965
1989
3383
4957
227
570
1087
358
1210
2588
1070
2657
4344
50
Stratified
384
996
239
824 - _ . 1969
768
1773
3247
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
571
1030
1630
1457
1815
3333
2455
3842
5646
268
648
1240
437
1347
2898
1286
2935
4954
50
Stratified
444
1117
283
935
2198
895
1996
3591
176
Table B.7 Calculated Film Froude Number, Frf
2.5 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(m1s)
(m1s)
2
4
6
2
4
6
2
4
6
5
8.5
13
4.6
8.1
12.6
3.9
8
11.2
0.5
1.5
5.1
9.1
13.9
4.9
8.6
13.3
4.1
8.3
12.3
50
5.4
9.2
14.3
4.9
8.8
14
4.4
8.6
12.8
Vsl
(nlIs)
0.5
1.5
Vsg
"-(-m/s j2
4
6
2
4
6
2
4
6
."_...
50
3.2
5.7
8
4
6.2
8.6
4.3
6.9
9.3
3.2
5.9
8.3
4.1
6.4
8.9
4.4
7
9.5
3.3
5.9
8.4
4
6.5
9
4.5
7.2
9.7
__ _,
.,,, ..
.-
177
Table B.9 Calculated Slug Liquid Holdup, a,
2.5 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(m1s)
(m1s)
0.5
1.5
2
4
6
2
4
6
2
4
6
0
0.69
0.59
0.51
0.72
0.63
0.52
0.78
0.69
0.62
50
0.62
0.55
0.48
0.69
0.62
0.54
0.76
0.66
0.59
Vsg
(m1s)
(m1s)
2
0.5
1.5
6
2
4
6
2
4
6
0
78
108
163
190
165
173
463
506
548
50
28
72
149
148
169
287
382
374
475
178
Table B.ll Calculated Accelerational Component, sr, [=] Pa
2.5 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(mls)
(mls)
1112
1562
2002
1797
2514
4186
2321
4641
7334
0.5
6
2
1
6
2
1.5
808
1097
1342
1294
1389
3291
1928
3670
5905
50
616 . 972
1170
887
1110
2405
1386
2684
3982
Vsg
(mls)
(mls)
50
618
547
484
688
609
525
792
708
647
569
515
448
632
586
515
770
676
621
546
480
423
613
555
485
730
642
577
0.5
6
2
1
6
2
1.5
179
Table B.13 Calculated Total Pressure Drop, APT [=] Pa
2.5 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(m1s)
(m1s)
2
0.5
1808
2216
2648
2674
3287
4884
3578
5855
5829
6
2
1
6
2
1.5
1419
1691
1954
2035
2143
4190
3115
4869
7286
50
1189
1524
1742
1647
1834
3176
2499
3699
5033
Vsg
(m1s)
(m1s)
2
0.5
6
2
1
6
2
1.5
4
6
50
5.8
8.9
13.1
5.6
9.5
13.1
5.9
9.7
14.2
5.8
9.8
14.3
6.4
9.8
14.8
6.1
10.3
15.4
Stratified
Stratified
Stratified
9.2
13.2
9.6
14.9
10.3
15.9
180
Table B.15 Calculated Slug Froude Number, Fr
26 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(m1s)
(m1s)
0.5
1.5
2
4
6
2
4
6
2
4
6
(m1s)
0.5
1.5
. _---_..y~g
(m1s)
2
4
6
2
4
6
2
4
6
181
Table B.17 Calculated Frictional Component, L1Pf [=] Pa
26 cP Oil, 2-Degree Upward Flow
Vsl
(mls)
Vsg
(mls)
0.5
2
4
6
2
1.5
2
4
50
0
118
196
394
396
361
468
10
81
126
207
280
284
295
129
187
243
275
321
Stratified
Stratified
Stratified
590
583
552
570
391
480
58
Vsg
(mls)
2
0.5
6
2
6
2
1.5
0
1126
1586
1734
2176
3147
4636
10
955
1322
1520
1686
3115
4350
50
865
1360
1624
1448
2868
3346
Stratified
Stratified
Stratified
5355
7424
4122
6811
4326
5848
182
Table B.19 Calculated Gravitational Component, APg [=] Pa
26 cP Oil, 2-Degree Upward Flow
Vsl
Vsg
(mls)
(mls)
0.5
1.5
2
4
6
2
4
6
2
4
6
50
762
683
612
861
759
687
722
613
541
784
689
634
653
598
496
735
647
555
Stratified
Stratified
Stratified
835
757
756
680
691
615
Vsg
(mls)
(mls)
2
4
6
2
4
6
2
4
6
2005
2465
2740
3433
4268
5791
1757
2061
2269
2749
4089
5280
1576
2087
2307
2425
3790
4222
Stratified
Stratified
Stratified
6780
8764
5430
8062
5409
6944
0.5
1.5
50
.-
183
Table B.21 Calculated Film Fronde Number, Frf
50 cP Oil, Horizontal Flow
Vsl
(m1s)
0.5
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
5.9
10.7
15
5.4
10
15.6
5.8
9.7
14.3
5.7
10.8
15
4.7
9.5
15.4
5.5
9.6
13.7
50
5.2
10.7
15.9
4.4
9.4
15.6
5.4
10
13.4
0.5
1.5
Vsg
(m/s)
2
4
6
2
4
6
2
4
6
50
3.7
6.3
8.3
4
6.7
9.6
5.3
7.5
9.8
3.6
6.4
8.5
3.7
6.5
9.6
5.1
7.5
9.6
3.6
6.4
8.5
3.7
6.6
9.7
5.2
7.7
9.5
184
Table B.23 Calculated Slug Liquid Holdup, R,
50 cP Oil, Horizontal Flow
Vsl
(m/s)
0.5
Vsg
(m/s)
2
4
6
2
4
6
2
1.5
4
6
50
0.85
0.71
0.6
0.81
0.68
0.57
0.78
0.65
0.55
0.85
0.71
0.6
0.81
0.68
0.57
0.78
0.65
0.55
0.85
0.71
0.6
0.81
0.68
0.57
0.78
0.65
0.55
Vsg
(m/s)
2
4
6
2
4
6
2
1.5
4
6
798
1151
1141
2180
3000
2885
3149
3440
3609
742
926
980
1833
1788
2350
2730
3243
3393
50
600
800
812
1822
1762
2004
2504
2665
2781
185
Table B.25 Calculated Accelerational Component, LlPa [=] Pa
50 cP Oil, Horizontal Flow
Vsl
Vsg
(m1s)
(m1s)
20
50
0.5
2
4
2
4
1067
2201
3589
2794
4191
6435
4378
6669
8421
1022
1986
3443
2187
3628
572
3720
6343
8257
747
1703
2453
1746
2863
4763
2838
4358
4618
6
2
1.5
Vsg
(m1s)
(m1s)
50
2
4
6
2
4
6
2
1865
3353
4730
4973
7191
9320
7527
10109
12030
1764
2913
4423
4020
5417
8322
6450
9586
11651
1347
2504
3265
3567
4624
6767
5342
7023
7399
0.5
1.5
ABSTRACT
DAAS, MUTAZ, ABDUL. Ph.D. August 2001
Chemical Engineering
Modeling the Effects of Oil Viscosity and Pipe Inclination on Flow Characteristics and
Drag Reduction in Slug Flow ( 185 pp.)
Director of Dissertation: W. Paul Jepson
Effects of oil viscosity and pipe inclination on each component of the total
pressure drop in slug flow were determined, The impact of liquid viscosity and pipe
inclination on the effectiveness of drag reducing agents (DRA) was also investigated.
Predicted values were in good agreement with experimental results.
Results from both experiments and modeling showed that the accelerational
component of pressure drop was dominant in low and moderate oil viscosities. This
component reached values as high as 86% of total pressure drop.
reduction took place in the accelerational component and reached values as high as 88%
out of total drag reduction. As oil viscosity increased, the frictional component was
found to increase dramatically and exceeded 40% of total pressure drop in the 50 cP oil.
The DRA was found more effective in reducing both frictional and gravitational
co:mponents of total pressure drop in higher-viscosity oils than in lower ones.
For
example, at superficial liquid and gas velocities of 0.5 and 6 mls frictional loss was