Group Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Title: 10.3.5 - Film Measurements in Two-Phase Gas-Liquid Flow by Conductance Techniques in Vertical Large Diameter Pipes
Full Citation
Permanent Link:
 Material Information
Title: 10.3.5 - Film Measurements in Two-Phase Gas-Liquid Flow by Conductance Techniques in Vertical Large Diameter Pipes Experimental Methods for Multiphase Flows
Series Title: 7th International Conference on Multiphase Flow - ICMF 2010 Proceedings
Physical Description: Conference Papers
Creator: van der Meulen, G.P.
Zangana, M.
Zhao, D.
Azzopardi, B.J.
Publisher: International Conference on Multiphase Flow (ICMF)
Publication Date: June 4, 2010
Subject: gas-liquid flows
conductance probes
film thickness
measurement techniques
Abstract: Flush-mount conductance ring probes are often calibrated in an ideal environment, with the assumption that the liquid film is uniform about the pipe periphery and in particular contains no gas bubbles. However, bubbles are often encountered in the liquid film, the effect on the accuracy of data obtained by flush mount conductance probes was examined. A new approach to the calibration of flush-mount conductance probes is suggested by using beads in the annulus between a cylindrical insert and the pipe wall. The outcome shows that errors of >25 percent and relative errors between different bead diameters of up to >600 percent were obtained. In addition, the present study shows a churn/annular flow boundary in the large diameter pipe based on visual observations by high speed photography. The Wallis and Kutateladze parameters show poor agreement with experimental data. Therefore, data on the huge/disturbance wave transition of Sekoguchi and Mori (1997) was converted to allow for the larger pipe diameter. This shows a reasonable agreement but the error between the boundaries may be due to a strong effect of the Reynolds and/or Weber numbers.
General Note: The International Conference on Multiphase Flow (ICMF) first was held in Tsukuba, Japan in 1991 and the second ICMF took place in Kyoto, Japan in 1995. During this conference, it was decided to establish an International Governing Board which oversees the major aspects of the conference and makes decisions about future conference locations. Due to the great importance of the field, it was furthermore decided to hold the conference every three years successively in Asia including Australia, Europe including Africa, Russia and the Near East and America. Hence, ICMF 1998 was held in Lyon, France, ICMF 2001 in New Orleans, USA, ICMF 2004 in Yokohama, Japan, and ICMF 2007 in Leipzig, Germany. ICMF-2010 is devoted to all aspects of Multiphase Flow. Researchers from all over the world gathered in order to introduce their recent advances in the field and thereby promote the exchange of new ideas, results and techniques. The conference is a key event in Multiphase Flow and supports the advancement of science in this very important field. The major research topics relevant for the conference are as follows: Bio-Fluid Dynamics; Boiling; Bubbly Flows; Cavitation; Colloidal and Suspension Dynamics; Collision, Agglomeration and Breakup; Computational Techniques for Multiphase Flows; Droplet Flows; Environmental and Geophysical Flows; Experimental Methods for Multiphase Flows; Fluidized and Circulating Fluidized Beds; Fluid Structure Interactions; Granular Media; Industrial Applications; Instabilities; Interfacial Flows; Micro and Nano-Scale Multiphase Flows; Microgravity in Two-Phase Flow; Multiphase Flows with Heat and Mass Transfer; Non-Newtonian Multiphase Flows; Particle-Laden Flows; Particle, Bubble and Drop Dynamics; Reactive Multiphase Flows
 Record Information
Bibliographic ID: UF00102023
Volume ID: VID00254
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: 1035-vanderMeulen-ICMF2010.pdf

Full Text

Paper No 7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

Film Measurements in Two-Phase Gas-Liquid Flow by Conductance Techniques in Vertical
Large Diameter Pipes

G.P. van der Meulen*, M. Zangana, D. Zhao and B.J. Azzopardi

Process and Environmental Engineering Research Division, Faculty of Engineering,
University of Nottingham, University Park, Nottingham, NG7 2RD, United Kingdom.
*Corresponding author:

Keywords: Gas-liquid flows, conductance probes, film thickness, measurement techniques.


Flush-mount conductance ring probes are often calibrated in an ideal environment, with the assumption that the liquid film is
uniform about the pipe periphery and in particular contains no gas bubbles. However, bubbles are often encountered in the
liquid film, the effect on the accuracy of data obtained by flush mount conductance probes was examined. A new approach to the
calibration of flush-mount conductance probes is suggested by using beads in the annulus between a cylindrical insert and the
pipe wall. The outcome shows that errors of >25 percent and relative errors between different bead diameters of up to >600
percent were obtained. In addition, the present study shows a chur/annular flow boundary in the large diameter pipe based on
visual observations by high speed photography. The Wallis and Kutateladze parameters show poor agreement with experimental
data. Therefore, data on the huge/disturbance wave transition of Sekoguchi and Mori (1997) was converted to allow for the larger
pipe diameter. This shows a reasonable agreement but the error between the boundaries may be due to a strong effect of the
Reynolds and/or Weber numbers.


Latest trends in oil and gas production show an increase in
the exploitation of large diameter, dynamic risers for the
transport of multiphase flows. Dynamic risers allow for
motion when configured in a catenary and are therefore
beneficial in deep water oil and gas production. Annular
gas-liquid flows are found in the power generating industry
and natural gas production and processing plants.
Understanding of the evolution of vertical annular
gas-liquid flows in large diameter pipes is however less
comprehensive compared to smaller diameter tubes
(Azzopardi et al. (1983) and Azzopardi (1997)) and,
consequently, data is still very limited. The disadvantage of
large diameter test facilities is that they occupy larger
installation areas, are costly and one of the most important
challenges is the achievement of gas flow rates covering all
regions of the flow patterns; high capacity gas compressors
are thus required. These challenges limits ones experimental
capabilities. It might not be surprising that recent studies
performed in large diameter pipes, concentrate on bubbly
and chur flow regimes. Popular well performance and
mechanistic void fraction models do often perform
reasonable against experimental data. However, an overall
pararameter such as void fraction hides a wealth of detail; it
does not contain information about the details of the film
interface and entrainment. In annular flow, three types of
waves are identified. At very high gas flow rates the
gas-liquid interface is of capillary nature, often referred to
as ripples. Ripples are small in magnitude, wavelength and
velocity and have a short lifetime. At lower liquid flow
rates disturbance waves occur. These are characterized by
a height several times mean film thickness. In addition, the

wavelength is greater than that of ripples and the velocity is
greater than that of the liquid film. In addition to ripple and
disturbance waves, Sekoguchi and Takeishi (1989) and
Sekoguchi and Mori (1997) reported waves that were even
larger in amplitude, wavelength and velocity than
disturbance waves; they defined these as huge waves.
However, still little is known about these waves and more
investigation is necessary to explain the hydrodynamic
mechanisms in combination with wisps. However, at some
conditions ripples, disturbance and huge waves co-exist and
interact. Individual disturbance waves and huge waves can
have slightly different velocities and so coalesce with one
another and cause a temporary increase in wave height.
Although some scholars attempted to develop linear
mathematic models to predict wave growth, it is worth
noting that that in practice the growth and life of waves is
non-linear and finite. Evidence shows that liquid
entrainment occurs at the onset of disturbance waves in
annular flow. Azzopardi (1997) pointed out that
non-dimensional analysis of the onset of liquid entrainment
result in data collected for different pipe diameters to give
better agreement. However, the effect of surface tension and
pipe diameter still causes a large relative deviation, where
especially for large pipes (125mm) this onset is several
orders of magnitude lower. The wave crests of disturbance
waves are subject to the momentum, hence shear, of the gas
phase flowing over it. (Parts of) the waves crests are torn off
to form small strings of droplets which further disperses
into single drops in the gas core. Drops in annular flow are
known to be major contributors to pressure drop. For
instance, Fore and Dukler (1995) found that mass transfer
between the two phases can contribute up to 20% of the
pressure drop. There have been two types of break up

Paper No

reported, namely ligament break up (figure 1, (LHS)) and
bag break up (RHS)).

Figure 1: Break up mechanisms in vertical annular flow
(LHS) ligament break up (RHS) bag break up

Ligament break up occurs when the crest of the wave breaks
in the direction of the gas flow. Bag break up occurs when
shear of gas velocity creates a bag of the wave crest and
disperses this fraction into accelerating drops. Azzopardi
suggested that the boundary between bag and ligament
break up can be described by a Weber number

We (1)
which depends on the wave height. It is true to argue that
this boundary criterion holds since the onset of drop
atomisation is strongly dependant on surface tension.
Consequently, the height, shape and uniformity of the
disturbance wave could be a measure of drop entrainment
by bag or ligament break up. It has been found that
disturbance waves in large diameter pipes are of
non-uniform nature, contrary to smaller diameter pipes
where they appear as uniformly distributed along the
periphery (Hewitt and Lovegrove (1969)). The relation
between entrainment and deposition was proposed by
Leman et al. (1980). By injecting a tracer into the liquid film
they found that the concentration of tracer in the film
decreased with axial distance relative to the initial value
because of entrainment and secondly, decreased more due to
non-tracer drops depositing onto the film. They found that
both the rate of entrainment and deposition increased
sharply with axial distance to a maximum and hereafter it
decreases gradually to adopt a semi-constant value.
Azzopardi et al. (1983) employed an experimental entrained
fraction measurement technique first proposed by Whalley
et al. (1974). By extracting the liquid film from the pipe wall
through a porous wall section, and measuring the fraction of
liquid and gas taken off, a mass balance between inlet and
measurement section can be calculated. Liquid entrainment
is of crucial importance in annular flow. The gap in the
literature confines to slug and annular flow and especially to
the latter since studies performed until present did not
succeed to establish developed slug flow in large diameter
pipes as it is known to occur in smaller tubes.
Omerbere-Iyari (2006) found that flow regime transitions in
large diameter pipes deviate considerably from those
observed in smaller diameter tubing by comparing cine film
and void fraction data with existing models and flow pattern
maps derived from smaller diameter tubes. Moreover, large
quantities of entrained gas in the liquid film were also
reported in the latter study. Since (transition) boundaries
have shifted and empirical models are unable to accurately

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

predict the evolution of gas-liquid flows in large vertical
pipes, it is necessary to systematically shed new light on the
phenomena occurring. Summarising, regardless of pipe
diameter, it can be concluded that the gas-liquid interface
and mass transfer are the most important parameters to
consider in order to understand the annular flow regime.
The first aim of this paper is to present a new approach to
the calibration method of flush-mount conductance probes
in annular gas-liquid flow used to measure phase
distribution by allowing gas entrainment into the liquid film.
The second aim of this paper is an attempt to address the
main question arising from the different perceptions of
annular flow in the literature and experimental work: "What
is annular flow in large diameter pipes". Preliminary
answers to this question are given by analysed cine film and
a data set, consisting of 600 liquid hold-up measurements.






Area (m2)
Diameter (m)
Resistivity (-)
Product modified Bessel function
gravitational constant (ms-1)
Height (mm)
Bessel functions
Apparent conductance O/Bessel functions
Elliptic integral (-)
Kutateladze number (-)
Length (m)
pressure (Nm 2)
Resistance 0
Thickness or width (mm)
Velocity (ms-1)
Dimensionless gas number (-)
Voltage (V)
Weber number (-)
Density (kg-m-3)
Viscosity (Pa-s)
Surface tension (N-m-1)
Angle (o)
Liquid conductivity (Sm-1)
Fraction (-)

Is Liquid superficial
gs Gas superficial
g Gas
1 Liquid
app Apparent
e Film thickness
w Wave
L Cross sectional area of liquid/wetted perimeter
x Two phase flow resistance or voltage
out Output
Probe Probe resistance
ref Reference resistance
full Full pipe
TP Two Phase
max Maximum
mix Mixture
* Dimensionless gas number

:. 0

Paper No

Experimental Facility

The experiments conducted in the present study are
primarily carried out on a closed loop facility, containing a
127mm id, 11 m (86.6 L/D) tall riser (Figure 2). Liquid is
stored in the main separator and pumped into the riser base.
The gas phase is compressed by two liquid ring vacuum
pumps operated in parallel and delivered to the riser base.
The phases come together in the mixer and from this point
the flow develops along the riser. The flow is then directed
horizontally into the downcomer and back into the separator.
Here the gas is separated from the liquid and the fluids are
fed to the compressors and pump. The flow of both fluids
can be regulated by valves and the flow rates monitored by
flow meters. System pressure can be up to 5 barg. Gas and
liquid superficial velocities of up to 17 m/s and 1.5 m/s,
respectively, can be achieved. Temperatures and pressures
are measured at various points throughout the experimental
facility. The vertical pipe is equipped with a transparent
acrylic resin section (Figure 2 Insert) wherein pressure drop
and void fraction measurements are acquired and which is
11.8 L/D. The conductance probe rings, used for phase
distribution measurements in the present study, are located
at 62.7, 63.5 and 65.5 pipe diameters from the riser base,
respectively. The stainless steel rings are flush mounted
with the pipe wall. They were located using cylindrical
dowels placed at either side of wall sections

Figure 2: Closed loop facility. (Insert Transparent Riser
Section) ((van der Meulen et al. (2" i))

The thickness s of the rings is 3mm and the distance De
between the probes is 25 mm, insulated by non-conducting
acrylic resin. This is a electrode separation distance to pipe
diameter ratio, De/Dt of 0.20. The total, time averaged,
pressure drop is being measured by an electronic differential
pressure detector/transmitter (Rosemount 1151 smart
model), with a range of 0- 37.4 kPa and an output voltage
from 1 to 5 V, i.e. a resolution of 9.35 kPa per volt. Two
pressure tappings, separated by an axial distance of 12.9
pipe diameters, across the transparent section, are connected
to the differential pressure device via stainless steel tubes.
The tubes were filled with water to keep the density

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

constant. This was assured by an efficient purging
procedure which eliminated the risk of gas fractions in the
pressure lines. The latter procedure was repeated at the start
of each set of the experiments. Data acquisition was
performed through a PC equipped with a National
Instrument (NI) DAQ card. An existing data acquisition
programme in Labview (Guglielmini et al. 1,21l 2)) was
adapted to convert the voltage output of the probes into a
cross section averaged void fraction. The data acquisition
rate was 1 kHz

Conductance probe theory

In gas-liquid annular-type flow, the instantaneous wall film
thickness can be determined by measurements of the
electrical conductance between two electrodes in contact
with the liquid film. Different types of electrodes such as
parallel wires, flush-wires, flush-mounted pins and
flush-mounted rings have been adopted (Brown et al.
(1978), Koskie et al. (1989), Kang and Kim (1992), Conte
et al. 'l2 1"'), Asali et al. (1985), Andreussi et al. (1988),
Tsochatzidis et al. (1992), Fossa (1998), Coney (1973)).
Among these probes, the flush-mounted parallel ring probe
is attractive to researchers because it provides non-intrusive
measurements, can pick up small impedance and allows
electric field to be efficiently confined. Coney (1973)
proposed that the apparent conductance Kapp for two
parallel strips embedded flush onto a flat surface can be
defined as;
K K* Ly (2)
Kapp = KappLr (2)
with k(m)
app k(-m) (3)
and sinh 2 (s/2h)
m = (4)
sinh2 [7(s + De)/2h]

Where, Kpp is the dimensionless apparent conductance,
L electrode length, s electrode width, y liquid
conductivity h liquid film thickness and Function k the
complete elliptic integral of the first kind,
k r/2( 2 -0.5 (5)
k(m)- = 2 -msin 2 dO
Furthermore, it was proposed that the conductance of
annular liquid film can be quantified by above equations by
replacing h to a liquid equivalent thickness hE.
D (6)
hE = n(1-2h/D)
hE can also be defined as hE =AL PL (Andreussi et al.
(1988)), where AL is the cross-sectional area occupied by
liquid and PL the wetted length of electrode for the
application of the annular and stratified flow. The analytical
solution to the apparent conductance of the ring probe is
(Tsochatzidis et al. (1992)),


7* 37 2s
app 32 (2+1

b, = cos[(De + s)] cos[(De s)0

Paper No

and f, is the product of modified Bessel functions
Io,11,Ko,K ,
1 I,[(D- 2h)O]Ko(DO)
lo(DO) K,[(D-2h)0Io(DO)
(D) II 11 [(D- 2h)0\K(Do) (10)
SKI[(D -2h)0]I1(DO)
(2i +1)7;
2L (12)

A different expression for b, was adopted by Fossa (1998),
b, = cos[(De 2s)]- cos[(De)O] (13)

Conventional conductance probe calibration

The unique resistive characteristics of the individual
conductance probes need to be identified. When the probe
pair is subject to an electrical current, the relationship
between the dimensionless resistivity E of the probe ring
and its voltage output Vou should be linear in the form of;
E= aV,,t +b (14)
Gradients a and b can be obtained from this relationship.
The resistance Rx of the two phase flow is being simulated
by a variable resistor in between each of the probe pairs.
Hence, the probe resistance Rprobe is the product of the
applied voltage Vpp, the voltage measured subject to the
applied variable resistor V, and Ref, the internal circuit
variable resistance, by using;

By plotting the

Probe app ref

S(ppRref i x)
Sx/ Rref -1)
~*=R,/ Rre+ l)

against V0,t Equation 14 was deduced. The regression
data for the relationship was, a =3.21 b =0.042 and
R2 = 0.99 Equation 14 can then be substituted into the
expression for the dimensionless conductance Ge of the
two phase flow
(1+ E fu )/(1- Efui)
Ge (1+ Ep )/(1 E )(17)
In which E fun is the dimensionless resistance for the flow
domain fully occupied by liquid and ETp the dimensionless
resistance for two-phase flow. Ultimately, by decreasing the
diameter of the insert and subsequently repetition of a
significant number of calibration steps, one can relate the
phase distribution at each probe to the dimensionless
conductance. The relationship of the calibration curve can
be mathematically expressed in the form;
El =c(Ge*)3 +b(Ge*)2 +a(Ge*)+d (18)

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

New approach to conductance probe calibration

The conductance probe need calibration before it can be
employed for measurements. Annular-type flow is usually
simulated by placing a non-conductive rod inside the pipe
and the conductive liquid is filled in the annulus between
the rod and pipe wall. This is an "ideal" situation that no
bubbles exist in the liquid film. In contrast, in gas-liquid
annular-type flow, the continuous trapping and folding
actions of the disturbance waves transport gas bubbles into
the liquid film. The presence of a considerable amount of
bubbles in the liquid film was reported in air/water
horizontal and vertical annular-type flow systems (Jacowitz
et al. (1964), Hewiit et al. (1990), Omerbere-Iyari (2006).
More recently, the bubble size distribution, bubble mean
diameter and bubble number concentration in the wall film
of a horizontal annular flow in a pipe of 15.1 mm diameter
was quantified (Rodriguez et al (2" I4)). Bubble size has an
exponential distribution with the average of diameters
between 15% and 45% of the film thickness at the gas
superficial velocity ranging from 28 to 65 m/s and the liquid
superficial velocity from 0.019 m/s to 0.14 m/s. Around 100
bubbles/cm2 exist in the film at the gas superficial velocity
28 m/s. Therefore, a new approach for probe calibration was
devised. In order to simulate gas bubbles in the liquid film
during annular-type flows, packing of spherical glass beads
was used. While the diameter of the insert decreased with
calibration steps, beads with a larger diameter were used to
occupy a fraction of the annulus between the insert and pipe
wall. The diameter of beads used was 3, 4 and 6mm. The
procedure, carried out off-line, of calibration was performed
according the following steps; (1) a conventional void
fraction measurement was first done with the single insert
only; (2) the test section was then emptied, dried and the
insert re-installed; (3) a known theoretical volumetric
amount of beads were added to the annulus; (4) in order to
verify the theoretical bead voidage, the height of beads in
the annulus was returned to a water volume. The annulus
was then filled with water until the level of water was equal
to the level of beads. The water left was then weighted and
the in situ bead voidage calculated (5) a void fraction
measurement was then performed. The difference between
the void fraction measured with the single insert and the
single insert-beads combination was verified by repeating
the calibration with an additional cylindrical insert which
represented a volumetric fraction equal to the sum of the
initial insert-beads combination. (6) the final step in the
calibration procedure, for one particular single insert
diameter, single insert-bead combination and the additional
insert, was taken the void fraction measurement of the latter.
Annular flow calibration data obtained from the 127mm
pipe was plotted and comparison was made between a
theoretical 70mm flow arrangement, solved by using the
model derived by Tsochatzidis et al. (1992), In addition,
also other models proposed were solved and plotted for the
127mm pipe. The results are shown in Figure 3. The data
obtained with the new calibration approach shows
deviations from the initial calibration points by using the
single insert for ideal annular-type flows. The error was
determined using Equation 17. This returned the void
fraction as theoretically responded by the probes according
the calibration curve and thus for an "ideal" annular-type
flow regime. These results were plotted against

Paper No

the experimental Ge obtained during calibration with
single insert-bead combination. Figure 4 shows the absolute
error. It can be noted that this not directly explains the effect
of bead diameter. Therefore the relative error between the
bead diameters was determined, shown by Table 1.

SPlobc I P ob 2 X Probc 3
A Addtionill iRt *B ianuu* inOiimtniupalio n
+ Coiny Todmatll[dI A nel bOilyri
-Calrstoantomnrc nular

0.5 -
090 02 .2 0.4 0.6 0.8 1
Figure 3 : Calibration curves



Figure 4

I *Without beads WithBeads

.20 0.25 0.30 0.35 0.40 0.45 0.50
:Error in Ge between beads and conventional
calibration with 25 percent error bars.

Bead diameter Change in bead Change in Error
(mm) voidage (%) (%)
3 20 624.9
4 23.2 48.8
6 12.9 3.5
Table 1: Error effect of bead diameter

Experimental results

In order to address the main question "What is annular flow
in large diameter pipes", two models for the onset of flooding,
which is believed to be an indicator for the chum/annular
transition, are tested. In addition data obtained by Sekoguchi
and Mori (1997) was corrected to allow for larger pipe
diameter used in the present study. The two equations are the
Wallis (1961) parameter and the Kutateladze number. The
Wallis parameter and the transition between the two regimes
is proven to be around ug =1.0, the Kutateladze number on
the other hand proved to be more popular for larger pipe
diameters and the transition criterion lies around Kug=3.2.
Visual verification was done by analysing cine film data
taken simultaneously at 1kHz. 600 experiments at 2 barg
were performed. The liquid flow rate was kept constant and
gas flow rates varied. The range of superficial gas and liquid

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

velocities covered are Ugs 2.57 16.25 and

Uls 0.0128 -0.406 m/s respectively. Ugs,, however

decreased with increasing Us Cross-section averaged and
time-resolved film thickness was calculated for the three
flush-mount conductance ring probes by using
s= 0.5Dt l g )0.5 19

,where a is the film thickness and eg the void fraction, and
plotted against the Wallis parameter.

ug -1.0 20
Figure 5 illustrates a selection of data points on their
significance and the evolution of normalised film thickness
with the Wallis parameter.
[Liquid Superficial Velocity
*0014 0O 020 AO 033 X 0040 *0 060 0 301


0.00 0.15 0.30 0.45 0.60
Dimensionless Gas Velocity (-)
Figure 5: Dimensionless liquid film thickness.


From figure 5 it is apparent that there is a systematic effect of
the superficial gas velocity on the film thickness. There is a
steep decrease over Ug of 0.15 to 0.45 and for 0.01, 0.02,
0.03 and 0.04 m/s for the liquid superficial velocities. At
higher gas superficial velocities there is a gradual increase.
This also applies to U1l = 0.06 m/s but the curve merges
with the other conditions at higher superficial gas velocities.
For higher superficial liquid velocities, the data has a
negative exponential distribution. It is suggested that the
change of slope is linked to a transition of flow pattern from
churn to annular flow since the pressure drop, wall and
interfacial shear stress are proportional to the film thickness
(dP/dz, ,,r, oc 6 ) (Hewitt et al. (1990)). Van der Meulen et
al. (2'""') found that also for large diameter pipes
dP/dz oc 6 holds but in contrast to the smaller diameter cases
(Hewitt and Lovegrove (1969)) the pressure gradient for this
larger pipe increases much slower with gas velocity.
Therefore the expected minimum was not very clear.
Moreover, the Wallis parameter under predicts the
churn/annular transition for large diameter pipes. It suggests
that if ugp 1.0 in the present study, the superficial gas
velocity did not meet the flooding criteria.
The cine film data was analysed and mapped for estimation
of the annular flow boundary. However, the boundary
criterion was defined somewhat different. In past studies,
flooding was defined as the boundary of chum to annular
flow. In the cine film data it is observed that here is still a
rather brought region between the first flow reversal


_--- -:


Ax A
S *.AA
1 * *
*' :^'tWW :*.

Paper No

phenomenon (waves travelling in negative direction) and
bridging of the liquid film to form churn flow. In fact, in the
large diameter pipe it is observed that bridging of the liquid
film occurs when the entrained liquid fraction is very high
and coalesces with huge waves travelling at the pipe wall.
These waves have a pulsating nature and travel in positive
and negative directions normal to the flow, indicating a
transition between frictional and gravitational dominated
flow. Similarities were sought between available data sets.
Sekoguchi and Mori (1997) suggested a transition from
chum to annular based on the unity of the frequency of
disturbance and huge waves in a 26mm pipe. For the present
study, data from Sekoguchi and Mori (1997) was corrected to
allow for the change in pipe diameter and predict the
chum/annular flow boundary. Figure 6 illustrates the map.

I Data set Kug=3.2 Cinefilm Sekoguchi &Mori (1997)

.. .. ..... ........ ..

..... ..... I.....

. . .. ....... .... . ...

t~~~ ~ ~ ~~~~ -- ? i F #d I1-

0 3 6 9 12 15 18
Superficial Gas Velocity (m/s)
Figure 6: Data set including Kug = 3.2 corrected data
from Sekoguchi and Mori (1997) and visually observed
annular flow boundary.

From figure 6 it can be observed that the boundary criterion
Kug = 3.2 over predicts the conditions that are in the annular
flow region in the present study. The criterion suggests that
45 percent of the data points are of annular nature with a
probability of 2 percent of Kug > 3.2, indicating a wide
distribution, against 14 percent of data points representing
annular flow predicted by visual observations. The 14
percent figure agrees well with the data points representing
liquid superficial velocities of 0.01 to 0.06 m/s. An
unsuccessful attempt was made in order to allow for the
effect of liquid flow rate to get better agreement between
visual observations and the data set. Furthermore it can be
seen that the boundary as predicted by the corrected
Sekoguchi and Mori (1997) data is higher than predicted by
visual observations, but gives reasonable results compared
with the other methods. It showed that the average error
between the two boundaries equals 4.43. Perhaps the
influence of the Reynolds number and or Weber may account
for this but more investigation of these dimensionless
parameters needs to be examined. Moreover, the structure
velocity was determined by cross correlating two successive
conductance probe signals. The structure velocity is known
as a reliable measure to extract more details from the
experimental data. Annular flow normally shows a structure
velocity with a positive slope when plotted against the
mixture velocity U,, = Ugs + Uls Figure 7 illustrates the
structure velocity calculated.

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

Liquid Superficial Velocity, Uls (m/s)

S0.01 0.06 A0.03 x0.04 0 0.3 0.02





0 0
S0 0 0 c

00 o o o
o oo

*X *;4 d. .
x ,* *


0 5 10 15 20
Mixture Velocity (m/s)
Figure 7 The structure velocity for the selected conditions

It can be observed that the overall structure velocity is
negative for the majority of conditions; negative values are
an indication of flow reversal or flooding. However, the slope
of the data is positive and this could indicate annular flow
without bridging phenomena. Interestingly, also the data for
U1l =0.06 m/s merges here but more strongly when
compared to figure 5. The selected data is in overall good
agreement with the visual observations made.


The results of the present study show that the calibration of
conductance ring probes needs to be approached carefully.
The performance and accuracy of these probe types is
strongly dependant on the calibration procedures. In chum
and annular-type gas-liquid flows one may encounter gas
bubbles entrained in the liquid film, particularly in the waves.
The comparison between the new approach suggested in the
present study for the calibration of conductance probes in
annular-type flows to the conventional method shows that the
gas bubbles entrained in the liquid film can cause erroneous
results. The inappropriate calibration of the conductance
probes can lead to an unrealistic view of the phase
distributions in annular-type flow and the transition to chum
flow. Simultaneous phase distribution measurements and
visual observation (e.g. high speed photography) may
therefore reduce the risk of experimental errors. However, it
may be concluded that conductance probes, calibrated for a
specific flow regime, are producing invalid results when used
Popular flooding boundaries to predict the transition between
churn and annular flow do not agree very well with visual
observations made in the large diameter pipe. However,
extraction of details from the raw data shows that the selected
conditions are in the annular flow region. The film thickness,
visually determined annular flow boundary and the structure
velocity are in good agreement. It can be concluded
preliminary that the annular flow boundary is roughly
determined, regardless of flow reversal but merely based on
bridging phenomena in large diameter pipes. The latter
disregards the brought region between flow reversal and
bridging phenomena. Data by Sekoguchi and Mori (1997)
and their flow pattern boundary based on unity of frequency
of huge and disturbance waves, was corrected to test against
the flow pattern boundary in the present study. Although the
boundaries are reasonably close, it is expected that the

No A





Paper No

difference may be due to the effect of the Reynolds and/or
Weber numbers. In order to establish high spatial resolutions,
more local and less averaged measurements are going to be
made in a later stage of the present study with a Wire Mesh
Sensor (WMS), a Film Thickness Sensor (FTS) based on a
design used by Belt (2" I') and Phase Doppler Anemometry


This work has been undertaken within the Joint Project on
Transient Multiphase Flows and Flow Assurance. The
authors wish to acknowledge the contributions made to this
project by the UK Engineering and Physical Sciences
Research Council (EPSRC) and the following: Advantica;
BP Exploration; CD-adapco; Chevron; ConocoPhillips; ENI;
ExxonMobil; FEESA; IFP; Institutt for Energiteknikk; Norsk
Scandpower PT; Shell; SINTEF; Statoil and TOTAL. The
Authors wish to express their sincere gratitude for this
support. The Ministry of Higher Education in the Kurdistan
Regional Government and Koya University are highly
acknowledged for their contribution to PhD candidate M.

Asali, J. C. & Hanratty, T. J. Interfacial drag and film height
for vertical annular flow. AIChE J. Vol. 31, 895-902.

Andreussi, P. Di Donfrancesco, A. & Messia, M.. An
impedance method for the measurement of liquid hold-up in
two phase flow. Int J Multiphase Flow, Vol. 14, 777-785

Azzopardi, B.J. Drops in annular two-phase flow, Int J
Multiphase Flow, Vol. 23, S1-S53 (1997).

Azzopardi, B.J.,Taylor S. & Gibbons, D.B. Annular two
phase flow in a large diameter tube. Int. Conference on the
Physical Modeling of Multiphase Flow, Coventry, U.K.

Brown, R.C., Andreussi P. & Zanelli, S. The use of wire
probes for the measurement of liquid thickness in annular
gas-liquid flows. Can. J. Chem. Eng. Vol. 56, 754-757

Coney, M.W.E. The theory and application of conductance
probes for the measurement of liquid film thickness in
two-phase flow. J. Phys. E: Scient. Instrum. Vol. 6, 903-910

Conte, G & Azzopardi, B.J. Film thickness variation about
a T-junction, Int. J. Multiphase Flow, Vol. 29, 305-328
0(2' "1).

Fore, L.B. & Dukler, A.E. The distribution of drop size and
velocity in gas-liquid annular flow. Int. J. Multiphase Flow,
Vol. 21, 137-149 (1995a).

Fore, L.B. & Dukler, A.E. Droplet deposition and
momentum transfer in annular flow, AIChE J., Vol. 41,

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

2040-2046, (1995b).

Fossa, M. Design and performance of a conductivity probe
for measuring the liquid fraction in two-phase gas-liquid
flows. Flow Meas. Instrum., Vol. 9, 103-109 (1998).

Guglielmini, G. Carratterizzazione sperimentale del moto
bifase intermittent in presenza di singolarita mediante
sonde ad impedenza, Laurea Thesis, Univers ita degli studi
di Genova, Italia ('" '2).

Hawkes, N.J., & Hewitt, G.F. Experimental studies of
wispy-annular flow, Int. Conf. on Two Phase Flow
Modelling and Experimentaion, Rome (1995).

Hewitt, GF. & Lovegrove, P.C. Frequency and velocity
measurements of disturbance waves in annular two-phase
flow. UKAEA Report AERE R4304 (1969).

Hewitt GF., Lacey, P.M.C. & Nicholls B. Transition in film
flow in a vertical tube symposium on two phase flow, Exeter,
2 pp. B401-B429 (1965).

Hewitt, GF., Jayanti, S. & Hope, C.B. Structure of thin
liquid films in gas-liquid horizontal flow. Int. J. Multiphase
Flow, Vol. 16, 951-957 (1990).

Jacowitz, L.A. & Brodkey R.S. An analysis of geometry
and pressure drop for the horizontal, annular, two-phase
flow of water and air in the entrance region of a pipe. Chem.
Eng. Sci., Vol. 19, 261-274 (1964).

Kang H. C. & Kim, M. H. The development of a flush-wire
probe and calibration method for measuring liquid film
thickness, Int. J. Multiphase Flow, Vol. 18, 423-437 (1992).

Koskie, J.E., Mudawar,I. & Tiederman, W.G Parallel wire
probes for measurements of thick liquid films. Int J
Multiphase Flow, Vol. 15, 521-530 (1989).

Leman, G.W., Agostini, M. & Andreussi, P. Tracer analysis
of devoping two-phase flannular flow. Physiochem,
Hydrodyn., Vol. 6, 223-237 (1985).

Miya, M., Woodmansee, D. E. & Hanratty, T. J. Amodel for
roll waves in gas-liquid flows, Chem. Eng. Sci., Vol. 26,
1915-1931 (1971).

Omebere-Iyari, N.K., The effect of pipe diameter and
pressure in vertical two-phase flow, PhD Thesis, University
of Nottingham (2006).

Rodriguez, D. J. & Shedd, T. A. Entrainment of gas in the
liquid film of horizontal annular, two-phase flow, Int. J.
Multiphase Flow, Vol. 30, 565-583 (21 i4).

Sekoguchi, K. & Mori, K. New development of
experimental study on interfacial structure in gas-liquid
two-phase flow, ExHFT, 1177- 1188, Brussels, (1997).

Sekoguchi, K. & Takeishi, M. Interfacial structures in
upward huge wave flow and annular flow regimes, Int. J,
Multiphase Flow, Vol. 15. 295-305 (1989).

Paper No 7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30-June 4, 2010

Tsochatzidis, N. A., Karapantsios, T. D., Kostoglou, M. V
& Karabelas, A. J. A conductivity probe for measuring
liquid fraction in pipes and packed beds. Int. J. Multiphase
Flow, Vol. 18, 653-67 (1992).

Van der Meulen, GP., Zangana, M., Zhao, D. & Azzopardi,
B.J. Phase distribution measurements by conductance
probes and pressure drop in gas-liquid flows, ExHFT-7,
Krakow, Poland (i2" "').

Whalley, P.B., Hewitt, G.F., and Hutchinson, P.
Experimental wave and entrainent measurements in vertical
annular two-phase flow. Multi-Phase Flow Systems Symp.,

Wallis, GB., Flodding velocities for air and water in vertical
tubes, UKAEA Report AEEW R123.

Wolf, A., Film structure of vertical annular flow. PhD
thesis, Imperial College London (1995)

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs