7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Hydrodynamics of Lube oil-water downflow through a vertical pipe


S.Ghosh1, G.Das1 and P.K.Das2

'Department of Chemical Engineering, IIT Kharagpur
2Department of Mechanical Engineering, IIT Kharagpur
Kharagpur 721302, India
E-m ail: g..! i lc, ii, 10 cl, r.l i I


Keywords: core annular flow, pressure drop reduction




Abstract

In the past few decades there has been an increasing interest in cross-country transportation of high viscous oils. Researchers
have reported that core annular flow (where the viscous oil core is circumscribed by a thin layer of water) could be an energy
efficient tool in this regard. However a review of past literature suggests that very less information is available on vertical
downflow of high viscous oil and water. The present study aims to provide a flow pattern map for vertical downward flow of
high viscosity oil (lube oil) and water through a glass pipe of 0.012 m diameter and 2.5 m length as well as to estimate the
pressure drop during different flow regimes. It is noted that the minimum frictional pressure drop occurs for the thick core
pattern and the water cut required to achieve the minimum frictional pressure drop decreases as the oil velocity increases.


Introduction

The past few decades have observed various studies, both
experimental and theoretical, on the simultaneous flow of
oil and water. The investigations are primarily confined to
the estimation of flow patterns and pressure drop for low
as well as high viscous oils and water. One of the major
reasons of increasing interest in oil water flow is the
enhancing importance of heavy oils in world economy.
However, the difficulty arises during the transportation of
heavy oils, which require exceedingly high pumping power.
Researchers have suggested water-lubricated transport
could be an economic way to circumvent the difficulty. In
this technique, water is injected in the line transporting the
oil such that it flows as an annular film along the pipe wall
while oil flows in the core region. As a result, the wall
friction arises due to flow of water only and the pumping
power is drastically reduced.

Several studies, both experimental and theoretical are
performed on this topic. One of the earliest studies was
reported by Russell and Charles (1959). Charles et al.
(1961) have provided a flow regime map in terms of
superficial oil and water velocity and noted that core flow
could not be established below a critical oil velocity for a
fixed water fraction. Miesen et al. (1993) have reported a
stable flow was observed when oil superficial velocity
varied from 0.5-2 m/s and the input water fraction was in
the range of 0.04-0.14. Amey et al. (1993) have performed
experiments with emulsified waxy crude oil (t= 0.6 Pa-s
and p = 985kg/m3) and No. 6 fuel oil (t= 0.27 and p =
989kg/m3) in a horizontal pipeline. Parda and Bannwart
(2001) demonstrated that the vertical up lift of oil (p =
963 kg/m3, t = 17.6 Pa-s) was possible using very small
amount of water, through a 0.025m pipe. Bensakhria et al.


(2004) have experimented with heavy oil (t= 4.74 Pa-s)
and water in a horizontal pipeline of 12 m length and
0.025m inner diameter. They reported a maximum of 90%
reduction in pressure drop during annular flow. Flow
regimes of high viscous oil water are also reported by
Grassi et al. (2008) and Sotgia et al. (2008).

In order to understand the mechanism of such flows
several studies are reported. Ooms et al. (1984) used the
hydrodynamic lubrication theory to analyze core annular
flow in a horizontal pipe. Rovinsky et al. (1997) have
attempted an analytical prediction of the various
characteristics of eccentric, laminar annular flow. Ooms
and Poesio (2003) have studied core-annular flow through
a horizontal pipe. They analyzed how the buoyancy force
on the core is counterbalanced in case of snake wave and
bamboo waves. Rodriguez and Bannwart (2006) tried to
analyze the shape of the liquid-liquid wavy interface and
express it in terms of pipe diameter, fluid flow rates and
properties. They predicted the wavelength, amplitude and
hold up ratio using the analytical model.
A survey of past literature reveals that majority of the
investigations on oil-water flow were conducted either on
horizontal (Charles et a.l (1959), Arney et al. (1993),
Miesen et al. (1993), Chakrabarti et al. (2005), Mandal et
al. (2007)Grassi et al. (2008), Sotgia et al. (2008) ) or
vertical upward (Parda and Bannwart (2001), Jana et al.
(2006 a, b) ) orientation. Very less information is
available regarding vertical downflow of oil-water flow.
Bai et al. (1992) have reported the simultaneous flow of
cylinder oil (t=0.6 Pa-s and density 905 kg/m3) and water,
through a 0.009525 m diameter glass pipe shaped as an
inverted U tube. They identified new types of flow
patterns namely bamboo waves in upflow and corkscrew









waves in downflow.
The objective of the present study is to provide a complete
flow pattern map for vertical downward flow of high
viscous oil and water and also to estimate the pressure drop
encountered during different patterns.

Nomenclature


D
g
dp/dz
Ap
Q
Us


Pipe diameter
gravitational constant (ms-1)
Pressure gradient (Pa/m)
Pressure drop
Volumetric flow rate
Superficial phase velocity


Greek letters
P Density (kg/ m-3)
li Viscosity (Pas)
a Insitu volume fraction
P Inlet volume fraction of water


Subsripts
f frictional
diff differential
o Oil
r Reduction ratio
w Water


Experimental Facility


A schematic of the setup is presented in figure 1.


Figure 1: Schematic of experimental set up.


The vertical downflow test rig comprises of a glass pipe of
0.012 m diameter and 2.5 m length. The test fluids used in
the experiments are lube oil (p=960 kg/m3, yt=0.22 Pa-s)
and water. Lube oil is pumped by a high head gear pump
while centrifugal pump is used for water. To measure the
flowrates Electromagnetic flowmeter (E) and Coriolis
mass flow meter (M) are used for water and lube oil


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

respectively. The Electromagnetic flowmeter measures
water flow rate from 0 to 4.17x10 4m3/s with a least count
of 4.17x10 6 m3/s. The range of accuracy is within 0.5 %
as supplied by the manufacturer. The range of Coriolis
mass flow meter is 0 to 3.33X10 4 m3/s with a least count
of 3.33X10 m3/s. The accuracy of flow rate measurement
using the mass flow meter lies within 0.8 %. The
experimental errors in measuring the superficial velocities
have been estimated to be within 0.47% and 0.167% for
lube oil and water respectively.

Lube oil is introduced into the central portion of a
specially designed copper nozzle while water is injected
into the adjoining annular passage. The fluids enter the
separator (S) after flowing through the test and exit
sections. They are gravity separated and recycled back to
the respective storage tanks.
The flow regimes are visually observed and photographed
at a distance of 1 m from the inlet. A digital camera
(DSCH9, SONY) is used for photographic recording of the
flow phenomena at different superficial velocities of both
the phases. A viewbox filled with water is placed at this
position to correct the optical distortions induced by pipe
wall curvature and to allow easy observation and camera
recording. The pressure drop is measured between two taps
0.7 m apart by differential pressure transmitter (Honeywell
-STD 120 Type) with a pressure range of 70 kPa and full
scale accuracy of +0.5%.


Results and Discussion


Lube oil-water flow patterns:

A number of unique flow patterns are observed during
co-current downward flow of water and lube oil.

a) Core annular regime-This is the dominant flow
regime under the present range of experimental conditions
in the test section. In this distribution oil flows as a central
core and water moves as an annular film between the wall
and oil core respectively. Two different types of core flow
are observed.

Thick coreflow:- This pattern is characterized by the long
waves formed at the interface of oil and water (Figure 2).
The waves are sinusoidal in nature. In this case oil
occupies the major portion of the pipe cross-sectional area.
This regime is restricted to low water superficial velocity
(Usw<0.45 m/s) for entire range of oil superficial velocity.
However, there exists some lower critical value of oil
superficial velocity (Uso.>0.075m/s) below which no core
flow is noted. This flow regime resembles core flow with
corkscrew waves at interface as observed by Bai et al.
(1992).

Thin coreflow This is observed when the water velocity
is increased further. The oil core is observed to be very thin
unlike the previous one (Figure 3). Also the core becomes
more and more wavy and characterized by short irregular
waves at the interface of oil and water. The waviness
increases with the increase in slip between the phases.







Uso =0.3 m/s,
Usw = 0.075 m/s


Uso = 0.3 m/s, Uso = 0.53 m/s,
Usw = 0.15 m/s Usw = 0.3 m/s




II I


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010
(Usw > 1.0 m/s) velocity these slugs are characterized by
the presence of waviness at the interface. These are
denoted as sinuous slug (Figure 5)


Uso = 0.6 m/s,
Usw =1.5 m/s


Uso = 0.15 m/s,
Usw = 0.7 m/s


Uso = 0.4 m/s,
Usw =1.2 m/s


Figure 2: Representative snapshots of thick core flow


Uso = 0.3 m/s,
Usw = 0.5 m/s


Uso = 0.6 m/s,
Usw = 0.9 m/s






I


Uso = 0.45 m/s,
Usw = 0.7 m/s


Figure 3: Representative snapshots of thin core flow

b) Transition This regime is observed with a further
increase in water velocity (Usw >0.6m/s). In this regime
the oil core is very wavy and intercepted by water at few
points (Figure 4). Between two discontinuous zones of oil
a narrow bridge of water is noticed. However, due to very
high viscosity of the oil it is not possible for the water film
to break the core completely in the corresponding range of
operating parameters.

c) Slug flow: This pattern occurs over a narrow range of
flow conditions at lower phase velocities (Uso < 0.15 m/s).
Under this condition, oil slugs occasionally traverse
through the continuous water phase and the distribution
resembles the conventional slug flow pattern. Oil slugs
are of regular shape surrounded by a smooth and thin
circumferential film of water. While the diameter of the oil
slug is comparable to that of the pipe, its length is several
time the diameter. Both its nose and tail are blunt.
Occasionally small oil droplets are found in the water plug
entrapped between two successive oil slugs.
At relatively lower oil (Uso < 0.4 m/s ) and higher water


Figure 4. Representative snapshots of transition regime

Uso = 0.075 m/s, Uso = 0.15 m/s, Uso = 0.15 m/s,
Usw = 1.2 m/s Usw = 1.2 m/s Usw = 1.5 m/s













Figure 5. Representative snapshots of slug flow

The interfacial configurations of oil slugs are distinctly
different from sinuous slug. This implies that the
mechanisms of the development of oil slugs are different
in these two flow regimes.

Flow pattern map:

The interfacial distributions characterizing each flow
pattern are described below. Photographs of each flow
pattern and their range of existence are presented in figure
6. In the figure Usw and Uso denote the respective
superficial velocities of water and lube oil and the legends
denoting the different patterns are shown below the
corresponding flow pattern.






7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Ulo-0.15m'i
--- Uso-O.3mrs
- Uo=. .6 mIn

* Thickcore
SThin ct
i Trnsition
o Sinuous iL


P
t'


Q-IQI
Figure 7: Frictional pressure gradient of lube oil water as a
function of input volume ratio of water to oil with oil
superficial velocity as parameter


V From Figure 7 it is noted that there exists a minimum
0 0.2 0.4 0.6 0.8 I
pressure drop for each oil velocity. It is also noticed that
minimum pressure gradient is observed during core flow
regime. A close observation of the figure reveals one of the
6: Flow pattern map of lube oil water downflow. unique features of the core annular flow. As the oil velocity
increases the water cut require to achieve the minimum
frictional pressure gradient decreases. This indicates it is
economical to operate heavy oil water at a higher oil
mental estimation of frictional pressure velocity once the core flow is established. This trend of core
nt: flow is also reported by Oliemans et al.(1 987) and Arney et
al (1993) for horizontal oreintation


Extensive measurements of pressure drop have been carried
out over a wide range of flow velocities to estimate the
frictional pressure gradients. The frictional pressure
gradient encountered during simultaneous flow of lube oil
and water can be estimated by subtracting the gravitational
pressure gradient of two-phase mixture from the measured
total pressure gradient as:


(-)dpf = (dp iff -a,(P. Po)g
dz dz


Where (P)df can be measured using the pressure
dz'^
transmitter, ao is the insitu volume fraction of the oil

and p,, po are the densities of water and oil respectively.

In the present case ao for both the systems is calculated
using the empirical correlation proposed by Arney et al.
(1993) as:
a, = [1 +0.35(1-f)] (2)
ao =l-aw


p is the inlet volume fraction of water. The frictional
pressure gradient of lube oil-water is presented in Figure 7.

The figure depicts (-dp/dz)f as a function of water to oil
volume ratio with oil superficial velocity used as a
parameter. The ranges of existence of the flow patterns have
been superimposed on the figure in order to understand the
influence of flow pattern on the pressure drop.


Pressure drop reduction factor:

Next an interest is felt to estimate the amount of energy save
during transportation of the high viscous oil by injecting
water. The energy efficiency can be predicted by reduction
factor (Ap,), introduced by Russell and Charles (1959). It
is defined as:


Ap0


Apo~ is the measured two-phase frictional pressure drop
and Ap, is the estimated frictional pressure drop if the oil
flows alone in the pipe with the same superficial velocity as
the two-phase. Figure 8 depicts the variation of reduction
factor with inlet volume fraction of water with oil
superficial velocity as parameter.

It shows that the pressure drop can be reduced to as much
as 0.0011 times that of single-phase flow of oil only for 0.8
m/s of oil superficial velocity. It also shows that as the
velocity of oil increases lesser inlet water fraction will be
needed to achieve the minimum reduction factor. The thick
core flow has a lower value of pressure drop reduction
factor than the thin core


*Slug Thick core 1 Thin core Transition o Sinous slug


Figure




Exper
gradie









0.045

0.04

0.035

0.03

0.025

0.02

0.015

001

0.005


--- Uso=0.4 m/s
------- Uso=0.53 m/s
- Uso=0.6 i/s
Uso=0. Sms
SThick core
T Thin core
x Transition


0 0.2 0.4 0.6 0.8
P


Figure 8: Variation of pressure drop reduction factor with
inlet water fraction

Conclusions

The present work reports the different flow patterns during
lube oil-water downflow through a vertical pipe From the
study the following conclusion can be made:
> The core annular flow is the dominant flow regime
within the present range of experimental
conditions for lube oil-water downflow.
> Different types of core flow are observed during
the experiments namely thick core and thin core.
However, the interfacial configuration of these two
core flow is different.
> It is observed that the minimum frictional pressure
gradient occurs for thick core flow pattern.
> It is noted that the water cut required for achieving
the minimum frictional pressure gradient
decreases as the oil velocity increases.
> Hence for transporting high viscous oil using
water, it is economical to operate at higher oil
velocities once the core flow is established.




References

Amey, M.S, Bai, R., Guevara, E., Joseph, D.D, Liu, K.
Friction factor and hold up studies for lubricated
pipelining-1Experiments and correlations. Int. J of
Multiphase Flow, Vol. 19, 1061-67 (1993).

Bai, R., Chen, K., Joseph, D.D. Lubricated pipelining:
stability of core-annular flow: Part 5. Experiments and
comparison with theory. J of Fluid Mech., Vol. 240,
97-132(1992).

Bannwart, A.C. Modeling aspects of oil-water core-annular
flows. J. of Petroleum science and engineering, Vol. 32,
127-143(2001).

Bensakhria, A., Peysson, Y, Antonini, G.. Experimental
study of the pipeline lubrication of heavy oil transport. Oil
and Gas Science and Technology- Rev IFP, Vol. 59,
523-533 (2004).


7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010


Chakrabarti, D.P, Das, G.., Ray, S. Pressure drop in
Liquid-Liquid two phase horizontal flow: Experiment and
prediction. Chem. Eng. Technol., Vol. 28, 1003-1009
(2005).

Charles, M.E, Govier, G.W, Hodgson, G.W. The horizontal
pipeline flow of equal density of oil-water mixtures. Can
J.of Chem.Engg.,Vol.39,17-36 (1961).

Grassi, B., Strazza, D., Poesio, P.,. Experimental validation
of theoretical models in two-phase high-viscosity ratio
liquid-liquid flows in horizontal and slightly inclined pipes.
Int. J. Multiphase Flow,Vol. 34, 950-965 (2008).

Hasson, D., Mann, U., Nir, A. Annular flow of two
immiscible liquids I. Mechanisms. Can. J. Chem. Eng.,
Vol. 48, 514-520 (1970).

Jana, A.K., Das, G, Das, PK. Flow regime identification
of two-phase liquid-liquid upflow through vertical pipe.
Chem. Eng. Sci., Vol. 61, 1500-1515(2006a).

Jana, A.K., Das, G, Das, PK. A novel technique to identify
flow patterns during liquid-liquid two phase upflow
through a vertical pipe. Ind. Eng. Chem. Res., Vol. 45,
2381-2393 (2006b).

Mandal, T.K., Chakrabarti, D.P, Das, G. Oil water flow
through different diameter pipes-similarities and
differences. Chem. Eng. Res. Design, Vol. 85, 1-7(2007).

Miesen, R., Beijnon, G, Duijvestijn, PE.M, Oliemans,
R.VA., Verheggen, T. Interfacial waves in core annular
flow. J. of Fluid Mech., Vol. 238, 97-117(1993).

Ooms, G, Seoal, A., Vanderwees, A.J, Meerhoff, R.,
Oliemans, R.VA. A theoretical model for core-annular
flow of a very viscous oil core and a water annulus through
a horizontal pipe. Int. J. Multiphase Flow,Vol.10, 41-60
(1984).

Ooms, G, Poesio, P Stationary core-annular flow through
a horizontal pipe. Physical Review E, Vol. 68,
066301(2003).

Parda, V.J.W & Bannwart, A.C. Modeling of vertical
core-annular flows and application to heavy oil production.
Journal of Energy Resources Technology ASME, Vol.
123,194-99 (2001).

Rodriguez, O.M.H.& Bannwart, A.C. Analytical model for
interfacial waves in vertical core flow. Journal of
Petroleum Science and Engineering, Vol. 54, 173-82
(2006).

Rovinsky, J., Brauner, N., Moalem, M.D. Analytical
solution for laminar two-phase flow in a fully eccentric
core annular configuration. Int. J. Multiphase Flow, Vol. 23,
523-543(1997).

Russell, T.W.F & Charles, M.E. The effect of less viscous
liquid in the laminar flow of two immiscible liquids. Can






7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

J.of Chem.Engg, Vol. 37,18-24 (1959).


Sotgia, G.., Tartarini, P., Stalio, E.,. Experimental analysis
of flow regimes and pressure drop reduction in oil-water
mixtures. Int. J. Multiphase Flow,Vol. 34,1161-1174(2008).