You are on page 1of 23

DESIGN AND PERFORMANCE OF GAS LIQUID CYLINDRICAL CYCLONE

SEPARATORS
by
Gene E. Kouba, Chevron Petroleum Technology Company
Ovadia Shoham and Siamack Shirazi, The University of Tulsa
Presented at the BHR Group 7
th
International Conference on Multiphase 95,
Cannes, France, June 7-9, 1995
ABSTRACT
Current separation technology based on the conventional vessel-type separator is several
decades old. The vessel-type separator is large, heavy and expensive to purchase and
operate. Recently, the petroleum industry has shown interest in the development of
innovative alternatives to the conventional separator that are compact and low weight, and
have low capital and operational costs. One such alternative is the Gas Liquid Cylindrical
Cyclone (GLCC) separator.
Very few studies are available on the optimum design and performance of GLCC's.
Consequently, the sizing of GLCC's and estimation of the operational envelope are based
on limited experience, without a high degree of confidence.
This paper presents new experimental laboratory data and limited number of field data on
the flow behavior and performance of the GLCC separator. Also, initial mechanistic
models that describe the flow behavior in the GLCC are presented and compared with the
experimental data. These provide the state of the art for the design of GLCC's for the
industry.
308
INTRODUCTION
The Petroleum Industry has relied heavily on the conventional vessel-type separation
technology, which has not changed substantially over the last several decades.
Conventional separators are bulky, heavy and expensive in capital and operating costs.
These limitations are felt most severely in offshore operations where platform costs are
escalating. The high costs associated with conventional separators have motivated the
Petroleum Industry to explore the development and application of alternative technologies
such as compact separator systems.

D
e
v
e
l
o
p
m
e
n
t
GLCCs
FWKO
Cyclones
Emerging
Gas Cyclones
Conventional
Horizontal and Vertical
Separators
Growth
Finger Storage
Slug Catcher
Vessel Type
Slug Catcher
Maturity
Time
Fig. 1: S Curve for Development Ranking of Separation Technology
Hydrocyclones
Fig. 1 shows schematically the "S" curve for the developmental ranking of some of the
various separation technologies. As shown in the figure, conventional vessel-type
horizontal and vertical separators are very mature, with only minor improvements coming
from new developments of internal devices and control systems. Hydrocyclones for
cleanup of produced water are in the growth region. Although large diameter separators
are relatively mature, recent developments have given rise to an emerging class of vertical
separators known as Gas Liquid Cylindrical Cyclones (GLCC).
The GLCC, shown in Fig. 2, is a simple, compact, low-cost separator that can be used as
an economically attractive alternative to the conventional separator. The wide variety of
309
applications of GLCC's may have different performance requirements, varying from only
partial separation to a complete phase separation. Potential applications include: control
of GLR for multiphase flow meters and pumps, portable well test metering, steam quality
metering, flare gas scrubbing, primary surface or subsea separation and pre-separation
upstream of slug catchers or primary separators.
GAS-LIQUID
INTERFACE
L
Fig. 2: Gas Liquid Cylindrical Cyclone Configuration
A representation of the available literature on cyclone separators and related physical
phenomena is given in the bibliography section. A review of the literature reveals that very
little information is available about the optimum design and performance of GLCC's.
Furthermore, existing mathematical models of cyclone separators have been limited to
single phase flow with low concentration of a second dispersed phase. No reliable models
are available for cyclones (conical or cylindrical) that are capable of simulating full range of
multiphase flows entering and separating in a cyclone. Moreover, no models are available
to account for the range of fluid properties that occur
in the oil industry.
310
Despite the lack of performance information, GLCCs are beginning to attract some
interest for numerous applications. Several cases of successful application of gas/liquid
cyclone separators were reported for multiphase separation, metering and pumping, as
follows:
Two studies carried out for British Petroleum by Davies and Watson
8
and Davies
9
indicated size, cost and performance benefits (foam handling) of a modified cyclone over a
conventional separator in offshore applications. This was confirmed by comparative tests
conducted by Oranje
20
.
In a different application, BHRG has developed a GLCC to control the gas liquid ratio to
optimize efficiency of a multiphase pump
19
. Similarly, BHRG with AGIP are investigating
using a GLCC in conjunction with multiphase meter for subsea metering applications.
GLCC's are also utilized for well testing meters (Paul Munroe Engineering; Spartan
Controls) and for limited range of gas/liquid separation such as for wet gas (Gasunie; CE
Natco; Porta-Test).
Arco
14
has developed a GLCC with spiral vane internals. The GLCC was tested in Alaska
and exhibited gas carryunder between 2% to 18%, depending on crude foaminess.
Alternatives for level control were explored in lab tests by Kolpak
14
, including throttling
floats and throttling diaphragm valves operated by the vessel hydrostatic head. Finally,
sensitivity of the liquid level in the GLCC to the pressure drop in the liquid and gas legs
was explored.
A gas liquid cylindrical cyclone was considered by the Naval Weapons Lab
4,5
to remove
gas from electrolyte used in large batteries. The GLCC used for this purpose had both
tangential inlet and outlet. Bandopadhyay
4
found that the stability of the vortex core is
influenced by the rotational angle between the inlet and the outlet. Also, the optimum
angle for the most stable core was found to be function of liquid flow rate and GLCC
geometry.
The GLCC can also be very useful for small production or stripper well operations. One
application is the measurement of fluids for production allocation and reservoir
management. Typically, a battery of wells are cycled through a test separator where gas
and liquids are measured before transport. A metering system based on a simple compact
separator costs less and requires fewer control systems than a typical test separator. It is
also expected to operate for long periods with little or no maintenance. Another impact of
compact separators in this case is to reduce the cycle time between well tests, as smaller
separators require less purge time. This translates into less down time for wells needing
attention.
311
Chevron
15
has successfully built and operated several GLCC's for use in a low GOR flow
metering applications. In Chevron's configuration, gas and liquid streams are separated in
a simple GLCC, metered by gas and liquid flow meters and recombined for transport.
Kouba
15
found that simple improvements to the GLCC resulted in greatly increased
performance of the tested prototype. Figure 3 shows a field prototype GLCC operated
successfully by Chevron in the Fox Deese Springer field in southern Oklahoma. The
construction and installation cost for the field prototype have totaled about $2,500 apiece.
Fig. 3: Gas Liquid Cylindrical Cyclone Field Prototype
The above examples demonstrate the applicability of the GLCC concept and the
pronounced impact it can have in the Petroleum Industry . However, the GLCC's in these
applications were utilized without a complete understanding of the hydrodynamic flow
behavior. Additional research and development are needed to produce models capable of
predicting the performance of GLCCs in different configurations and applications. The
design tools resulting from these models will not only allow more predictable and reliable
312
implementation of GLCC's, but will also guide the way for additional design
improvements. Realizing the potential of GLCC's and the need for more technology
development in this area, the Tulsa University Separation Technology Projects (TUSTP)
was formed. TUSTP focuses on collecting experimental laboratory and field data on the
performance of the GLCC and developing mathematical models for the prediction of the
hydrodynamic behavior of the flow in the GLCC.
Fig. 4: Schematics of Chevrons Multiphase Metering Loop
EXPERIMENTAL DATA
One of the most enthusiastically explored applications of the GLCC is for gas separation in
multiphase measurement system. Multiphase meters that measure the full stream flow
without separation suffer from size and accuracy limitations in gas dominated flows. The
size of the multiphase meter can be kept small and measurement accuracy improved by
simply pre-separating most of the gas. It was recently observed
21
that many, if not most,
of the multiphase metering applications are, in fact, gas dominated and fall outside the
313
operational envelope for a typical full stream multiphase meter. Also, it was reported that
the accuracy and range of application of such meters deteriorate for in-situ gas volume
fraction greater than 70%. In low gas flow applications, complete gas separation in the
GLCC allows use of conventional metering technology. Fig. 4 is a schematic of Chevron's
multiphase metering loop, using a GLCC. A vortex shedding meter is used to measure the
gas flow, while the oil and water are measured simultaneously with a Coriolis based net oil
computer. The coriolis based net oil computer measures the total mass flow rate and the
mixture density. These measurements combined with the knowledge of the oil and water
densities allow the determination of the water cut and the water and oil mass or volume
flow rates
16
. This configuration requires more space than a full stream multiphase meter,
but is significantly cheaper and potentially more accurate. Another advantage of this
configuration is that for a wide range of flow conditions no active liquid level control
system is required. Initial testing have confirmed this concept and defined the operational
envelope for adequate (99%) gas/liquid separation, under limited operating conditions.
Test data have been acquired on three laboratory metering loops, incorporating GLCCs of
0.0254, 0.0508 and 0.0762 m I.D, respectively. The two smaller diameter GLCC's were
built and tested by Kouba
15
, and the larger 0.0762 m GLCC was built by TUSTP
2,12
. Fig.
5 shows the operational data obtained for three different 0.0508 m I.D. laboratory GLCC's
prototypes. The fluids used were air and water and the operating pressures were
approximately 205, 308 and 446 kPa (absolute). The x and y axes represent the in-situ gas
and liquid superficial velocities, respectively. The different data sets indicate the onset of
liquid carryover in the gas stream. The data were obtained by establishing a liquid flow
rate and increasing the gas flow rate until the onset of liquid carryover was observed. The
mechanism for liquid carryover is generally churn flow, except for very high gas rates and
low liquid rates, where the liquid is carried over by mist flow. Therefore, the appropriate
operating region for a complete gas/liquid separation lies to the left of the data points for
each lab prototype. The region to the right of each data set represents liquid carryover for
the corresponding GLCC prototype.
The lower left data set was generated for a laboratory prototype GLCC with a tangential
inlet oriented perpendicular to the axis of the separator. This configuration is typical for
conventional vertical separators with a tangential inlet as commonly found in industry
applications. The performance for this configuration suffers because the swirling liquid
passes in front of the GLCC inlet, forcing gas to blow through and entrain liquid. The
result is that liquid is carried over at relatively low gas flow rates.
The performance of the second single stage prototype is significantly improved by simply
inclining the GLCC inlet downward. The downward inclined inlet promotes stratification
and initial pre-separation of gas and liquid before entry into the GLCC. Kouba
15
observed
that the optimum inclination angle of the inlet is approximately -27, which causes the
314
swirling liquid to pass below the inlet, allowing the gas unobstructed passage into the
upper part of the GLCC. This resulted in more than a two-fold increase in the
performance envelope, as shown in the figure.

0.00
0.20
0.40
0.60
0.80
1.00
1.20
0 5 10 15 20
Superficial Gas Velocity (m/s)
S
u
p
e
r
f
i
c
i
a
l

L
i
q
u
i
d

V
e
l
o
c
i
t
y

(
m
/
s
)
One Stage Perpendicular Inlet 308 kPa (abs)
One Stage Inclined Inlet 336 kPa (abs)
Two Stage Inclined Inlets 308 kPa (abs)
Two Stage Inclined Inlets 205 kPa (abs)
Fig. 5: Operational Envelope Defined by the Gas Capacity Limits to Avoid
Configuration Pressure
Theory, [Eq.(13)] 205 kPa (abs)
Theory, [Eq.(13)] 308 kPa (abs)
////////// Region of field
Liquid Carryover in a 0.0508 m I.D GLCC
The third data set in Fig. 5 was obtained from a two stage configuration in which a second
GLCC was placed in series with the first one. In this configuration the gas outlet of the
first GLCC was connected to the inlet of the second GLCC. Liquid was carried over in a
mist flow for the near vertical segment of the operational envelope. This is the onset to
annular mist flow, which represents the theoretical limit to the performance of the GLCC.
Data for the third set of runs were collected at 308 kPa (abs.). The effect of pressure on
the performance of the GLCC was studied by lowering the pressure in the GLCC. The
fourth set of data was taken at 205 kPa (abs.). Comparison between the third and fourth
315
sets of data reveals that increase in the GLCC pressure results in a decrease of the
operational envelope. This indeed is the effect of pressure on the transition to annular mist
flow, which is the mechanism for liquid carryover for high liquid loadings in the two-stage
GLCC. The data for low liquid loadings seem to indicate little advantage of two stages
over one stage under similar conditions.
The shaded region in the lower left corner of Fig. 5 indicates the region where Chevron's
field prototypes have successfully operated. The boundaries for this region were based
average flow rates. Instantaneous flow rates, i. e., superficial velocities were generally
more than double the average rates.
A critical parameter for the prediction of liquid carryover is the maximum allowable liquid
holdup in the GLCC above the inlet, as a function of the gas velocity. Measurements of
this liquid holdup were made for zero net liquid flow conditions in the upper region of the
GLCC. The GLCC was first filled with water and the liquid exit blocked. A low gas flow
rate was then established, which initially removed part of the liquid from the upper region
of the GLCC. This resulted in a special two phase flow above the inlet, where the gas
flows through the liquid, but no liquid is carried over, i.e., zero net liquid flow conditions.
The experiments were repeated for increasing gas flow rates. Each gas flow rate resulted
in a unique maximum allowable liquid holdup which decreased as the gas flow rate
increased, as shown in Fig. 6. The liquid holdup above the GLCC inlet is calculated as the
ratio of the liquid volume above the inlet to the volume of the GLCC from the inlet to the
highest point reached by the liquid. At the gas capacity limit prior to liquid carryover, the
highest point that the liquid reaches is the GLCC gas exit. Gas flowing conditions are
under the capacity limit when the highest excursion point of the liquid falls below the gas
exit.
Two different methods were used to measure the volume of the liquid above the inlet. The
liquid volume was measured statically in the first method by simultaneously stopping the
gas flow and trapping the liquid in the GLCC. This method ignores the small amount of
gas entrained below the inlet. In the second method the differential pressure across the
GLCC was measured under flowing conditions and attributed entirely to the hydrostatic
head across the GLCC.
The differential pressure was converted to an equivalent fluid column height, and the liquid
holdup in the upper part of the GLCC was calculated based on the amount of liquid above
the inlet. As can be seen from Fig. 6, the two sets of data fall close together, which implies
that the assumptions made in the two methods are valid. These are: neglecting entrained
316
gas below the inlet in the first method, and neglecting frictional pressure drop in the
second method.

0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6
Superficial Gas Velocity (m/s)
L
i
q
u
i
d

H
o
l
d
u
p

i
n

G
L
C
C

A
b
o
v
e

I
n
l
e
t
Gas flow, trapping (capacity)
Gas flow, trapping (under capacity)
Gas flow, pressure difference (capacity)
Two-phase, pressure difference (under capacity)
Fig. 6: Liquid Holdup Above GLCC Inlet vs. Superficial Gas Velocity
The fourth set of data was obtained under flowing conditions for both gas and liquid. The
liquid holdup was measured by the second method. The close agreement between the data
sets indicate that the method is a reasonable means to obtain liquid holdup in the upper
part of the GLCC.
MECHANISTIC MODELING
The models presented in this section represent initial mechanistic models developed for the
prediction of the hydrodynamic flow behavior in the GLCC. The models enable the
prediction of the equilibrium liquid level, bubble trajectory and onset of annular mist flow
317
L
g
2
L
g
1
L
vc
L
in
L
eq
P
1
L
l
2a
L
l
2m
L
l
2b
D
2
D
1
L
l
3
P
2
L
g
3
D
3
GAS
METER
Fig. 7: Gas Liquid Cylindrical Cyclone Nomenclature for Mechanistic Model
in the GLCC. The equilibrium liquid level and the onset of annular mist flow are needed
for the prediction of liquid carryover. The bubble trajectory will enable the prediction of
gas carryunder. The mechanistic models are necessary to properly scale the results from
the laboratory experiments to field conditions. In particular the occurrence of slug flow
which can be handled easily in the laboratory, might be a potential problem under field
conditions, where long slugs may occur in large diameter flowlines.
Equilibrium Liquid Level
For proper operation of a GLCC the liquid level should be maintained below the inlet, in
order to avoid gas blowing through the liquid phase and carrying liquid in the gas stream.
Under static no-flow conditions, the liquid phase will have the same level in the GLCC and
in the recombination section. As the liquid flow increases, the liquid level in the GLCC
will rise up due to hydrostatic head gain needed to compensate for frictional losses in the
liquid section. For two phase flow the liquid level in the GLCC may be above or under the
318
inlet, depending on the operational conditions. Thus, it is essential to be able to predict the
liquid level in the GLCC.
The proposed model determines the liquid level from a simple pressure balance between
the inlet and outlet of the GLCC. This simple model neglects any hydrodynamic
interactions between the gas and the liquid phases. Refer to Fig. 7 for the geometrical
parameters.
The pressure drops in the liquid and gas sections are given, respectively, by
( ) ( )
P g L L g L L
f L
D
v
l l l l g in l l
l l l l
+ +

_
,


1 3 1
1 1 1
1
2
2
(1)
P g L L
g g g g g
( )
3 1
(2)
where
l
and
g
are the frictional pressure losses in the liquid and gas sections,
respectively, as given by

l
l i i i
i
i i
l
f L v
D
K v +

_
,

2
2 2
(3)

g
g
i i i
i
i i
g
f L v
D
K v +

_
,

2
2
2
(4)
The first terms in the parentheses of Eqs. (3) and (4) represent the frictional losses in the
different pipe segments of the loop. Note that consistent with Eq. (1), the first term in Eq.
(3) does not include the frictional losses in the GLCC itself. The second terms in the
parentheses represent the resistive losses such as through reductions and elbows .
Equating the pressure drop in the liquid and gas sections, one can solve for the liquid level
in the GLCC, as follows
L
gL g L L L
g
v f
D
l
l g l l g in g g
l g
l l
l
1
3 1 3
1
2
1
1
2

+ +

_
,




( )
( )
(5)
319
The prediction of Eq. (5) for the liquid level in the 0.0508 m lab GLCC prototype as a
function of the operational conditions is presented in Fig. 8. As will be discussed later, the
equilibrium liquid level can be used for the prediction of liquid carryover in the GLCC.
320
321
-0.2
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8 1.0
Superficial Liquid Velocity ( m/s)
L
i
q
u
i
d

L
e
v
e
l

i
n

G
L
C
C

w
r
t

o
u
t
l
e
t

(
m
)
vsg = 0.30 m/s
vsg = 3.0 m/s
vsg = 6.1 m/s
Fig. 8: Equilibrium Liquid Level for 0.0508 m I.D GLCC
with Air - Water Flow
Bubble Trajectory
The swirl created below the GLCC inlet causes a radial separation of the gas and the liquid
phases due to the buoyant and centrifugal forces acting on them. The liquid phase moves
towards the GLCC wall and the gas bubbles towards the centerline. Prediction of the
bubble trajectory can be used to determine the "fate" of the bubbles, i.e., whether the
bubbles are carried under with the liquid or caught in the updraft with the gas core. Some
of the bubbles will move radially inward sufficiently to merge with the gas core, and will be
carried upwards in the gas stream. Other bubbles will not merge with the gas core but will
be carried by the liquid stream out of the GLCC.
The radial motion of a gas bubble can be determined from a force balance on the bubble.
In this simple model the forces acting on a bubble in the radial direction are the centrifugal,
buoyant and drag forces. Equating these forces yields the radial velocity distribution of a
gas bubble, as follows
322
( )
( )
( )
( )
( )
v r
r
v r
r
D
Cd r
r
m g
l
t b


4
3
2

(6)
where the mixture density distribution
m
(r) and the circumferential velocity distribution v
t
(r) are approximated, respectively, by
( )

m g l g
s
m
r
r
R
( ) +

_
,
(7)
v r v
r
R
t ts
s
n
( )

_
,
(8)
In Eqs. (7) and (8) r/R
s
is the ratio of the radial location to the GLCC radius, and the value
used for both the exponents m and n is 0.9.
The drag coefficient proposed by Turton and Levenspiel
10
is used in this study, as follows
( )
( )
( )
( )
Cd r
r
r
r
( )
. Re
Re
.413
, Re
.
.

1
]
1
+
+

24 1 0173
0
1 16 300
0 657
1 09
(9)
where the Reynolds number is determined from
( )
( ) ( )
Re r
r v r D
m r b
l

(10)
For a short time increment dt, the differential distance traveled by the gas bubble in the
axial direction is
dz
dr
v r
v
r
z

( )
(11)
where v
z
is the axial liquid velocity in the GLCC. Thus, Eqs. (11) and (6) enable the
determination of the bubble trajectory, i. e. z(r). The total axial distance traveled by the
gas bubble in the GLCC is determined from the integration of Eq. (11), yielding
323
Z
v
v r
dr
t
z
r R
r
s

( )
(12)
Figure 9 shows the prediction of a 0.5 mm bubble trajectory in the 0.0508 m laboratory

Fig. 9: Bubble Trajectory for Air-Water Flow in a 0.0508 m I.D GLCC
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.000 0.005 0.010 0.015 0.020 0.025 0.030
Radial Distance From Wall (m)
A
x
i
a
l

D
i
s
t
a
n
c
e

F
r
o
m

I
n
l
e
t

(
m
)
G
a
s

c
o
r
e

b
o
u
n
d
a
r
y
C
e
n
t
e
r
l
i
n
e
T = 27 deg. C
q
g
= 1,700 sm^3/d
D
c
= 6.35 mm
D = 0.0508 m
q
l
= 63.6 m^3/d
P = 345 kPa (abs.)
D
b
= 0.5 mm
GLCC operating with air and water. For these conditions, the GLCC length below the
inlet should be at least 0.4 m long in order to avoid gas carryunder.
Onset of Annular Mist Flow
The onset to annular mist flow represents the theoretical gas capacity limit to the
performance of the GLCC. Under these gas flow rates conditions the first fine droplets of
liquid are atomized and carried over in the gas stream. The criterion for this transition is
similar to the one suggested by Taitel et al.
23
for the transition boundary between slug or
annular flow in vertical pipes, as follows
v We
a m
l g
g

_
,
2 3351
2
0 25
.
.

(13)
324
Taitel et al.
23
used a value of We=20 for large droplets encountered in the transition
boundary between slug to annular flow. In this study a value of We=7 is used to represent
the fine droplets occurring in annular mist flow. As shown in Fig. 5, an excellent
agreement is obtained between the prediction of Eq. (13) and the experimental data for the
high liquid loading region for the two sets of data.
FUTURE WORK
The models presented in this study provide qualitative guidance for GLCC design.
However, enhancements of these models are underway that will enable prediction of the
GLCC performance. For example, the proposed equilibrium liquid level model can predict
the operational envelope (Fig. 5) for liquid carryover, if provided with the following: the
actual liquid holdup in the region above the GLCC inlet, and the maximum allowable liquid
holdup prior to liquid carryover. The actual liquid holdup will be determined from the
prediction of the shape of the gas liquid interface, applying free vortex theory. The
maximum allowable liquid holdup will be predicted from zero net liquid flow analysis for
the gas region above the inlet. This is shown empirically in Fig. 6. The bubble trajectory
model will be improved to account for bubble swarm and turbulent diffusivity effects.
Additional work being planned for this project include:
1. Field testing of GLCC prototypes.
2. Collection of laboratory and field data into a data bank.
3. Investigate passive and active controls for the GLCC.
5. Analyze response of the GLCC to severely fluctuating flows.
6. Development of a dedicated multiphase multidimensional CFD simulator for the
GLCC.
CONCLUSIONS
GLCC's are finding a broad range of applications in the Petroleum Industry. Successful
applications have been reported despite of lack of understanding of the optimum design
and performance of the GLCC. Potential applications abound but are dependent upon a
more complete understanding of the hydrodynamic behavior of the flow in the GLCC.
New data and rudimentary models have been presented which are the building blocks for
the design tools that will allow better performance prediction and application of the
GLCC. The operational envelope developed in the laboratory prototype was found to be
exclusively governed by the liquid carryover in the gas stream. The operational envelop
was found to be significantly enhanced by using an inclined inlet. Furthermore, using two
325
GLCC's in series reached the theoretical performance limit due to the onset of annular mist
flow. Mechanistic models were proposed for the prediction of the equilibrium liquid level,
bubble trajectory and onset of annular mist flow in the GLCC. Further enhancements of
these model were discussed that will enable the prediction of the GLCC operational
envelope.
NOMENCLATURE
C
d
= drag coefficient
D = diameter
f = friction factor
g = acceleration of gravity
K = resistance coefficient for elbow or tee
L = length
m = mixture density exponent, m = 0.9
n = tangential velocity exponent, n = 0.9
q = volumetric flow rate
r = radial coordinate
Re = Reynolds Number
R
s
= radius of GLCC
T = temperature
v = velocity
z = axial coordinate
Z
t
= axial distance
We = Weber Number, W
e
= 7
P = pressure drop

= frictional losses

l
= liquid viscosity
= density
= surface tension
Subscripts
b = bubble
c = core
g = gas
326
l = liquid
m = mixture
r = radial
s = slot
t = tangential
z = axial
REFERENCES
1. Arato, E.G., Barnes, N.D., "In-line Free Vortex Separator Used for
Gas/Liquid Separation within a Novel Two-Phase Pumping System",
Hydrocyclones - analysis and application. Editors: L. Svarovsky, M. T. Thew.
Brookfield, VT: Kluwer-Academic (1992), pp. 377-396.
2. Arpandi, I., "Gas Carryunder in Gas Liquid Cylindrical Cyclones Separators -
Experiments and Modeling", M.S. Thesis in Progress, The University of Tulsa
(1995).
3. Baker, A.C., Entress, J.H., "VASPS (Vertical Annular Separation Pumping
System) Subsea Separation and Pumping System", Trans I Chem E., v. 70,
Part A (Jan. 1992), p. 9-16. Paper presented at the Institution of Chemical
Engineers Conference "Subsea Separation and Transport III", London, (23-24
May, 1991).
4. Bandopadhyay, P. R. , Pacifico, G.C and Gad-el-Hak, M., Sensitivity of a
Gas-Core Vortex in a Cyclone-Type Gas-Liquid Separator, Advanced
Technology and Prototyping Division, Naval Undersea Warfare Center
Division, Newport, Rhode Island (1994).
5. Bandopadhyay, P.R., Pacifico, G.C., Weapons Technology and undersea
Systems Department; Gad-el-Hak, M. University of Notre Dame, Gas Core
Configurations in a Cyclone-Type Gas-Liquid Separator, NUWC-NPT
Technical Report 10, 308, 3 (January 1994).
6. Boyson, F. Ayers, W.H., Swithenbank, J., : "A Fundamental Mathematical
Modelling Approach to Cyclone Design", Trans. IChemE., v. 60, No. 4 (July
1982), pp. 222-230.
327
7. Cowie, D., "Vertical Caisson Slugcatcher Performance", Trans. IChemE, v. 70,
part A (Jan. 1992), p. 25-31. Paper presented at the Institution of Chemical
Engineers Conference "Subsea Separation and Transport III" held in London,
(23-24 May, 1991).
8. Davies, E. E., Watson, P., "Miniaturised Separators for Offshore Platforms."
1st New Technology for Exploration and Exploitation of Oil and Gas Reserves
Symposium. Luxembourg, Apr. 1979. London, Eng.,BP Research Centre,
(1979), pp. 75-85.
9. Davies, E.E., "Compact Separators for Offshore Production", 2nd, New
Technology for the Exploration & Exploitation of Oil and Gas Resources
Symposium. Luxembourg, 5 Dec. 1984. London, Eng.,BP Research Centre,
(1985), Proceedings, v. 1, pp. 621-629.
10. Dimitar G. Karamanev and Ludmil N. Nikolov, Free Rising Spheres Do Not
Obey Newtons Law for Free Settling", AIChE Journal, (November 1992), vol.
38, No. 11, pp. 1843-1846.
11. Forsyth, Ray A., "Cyclone Separation in Natural Gas Transmission Systems -
the Design and Performance of Cyclones to Take the Debris out of Natural
Gas", Chemical Engineer, (London) (June 1984), pp. 37-41.
12. Joshi, A., "Liquid Carryover in Gas liquid Cylindrical Cyclones Separators -
Experiments and Modeling", M.S. Thesis in Progress, The University of Tulsa
(1995).
13. Kanyua, J. F., Freeston, D.H., "Vertical Flow Centrifugal Separator - Effects of
Geometry." Geothermal Resources Council (N.Z.) - Transactions v. 9., part 11
(Aug. 1985), pp. 251-256.
14. Kolpak, M., : "GLCC Level-Pressure Sensitivity, Level Control and Turbulent
Diffusion" Internal Report, ARCO, (1994).
15. Kouba, G. E., Chevron Petroleum Technology Co., (1995).
16. Liu, K. T. and Kouba G. E., : Coriolis Based Net Oil Computers Gain
Acceptance at the Wellhead Oil & Gas J., vol. 92, No. 26 (June 27, 1994).
328
17. Millington, B.C., Thew, M. T., "LDA Sudy of Component Velocities in Air-
Water Models of Steam-Water Cyclone Separators", 3rd International
Conference on Multiphase Flow. The Hague, Neth., (18 May 1987).
Cranfield, Bedford, Eng.: BHRA Fluid Engineering Centre, (1987), pp. 115-
125.
18. Muschelknautz, S., Mavinger, F., "Flow Studies in the Outlet Pipeline and in a
Cyclone Separator under Pressure Relief", Chemie Ingenieur Technik v. 62, no.
7 (1990), pp. 576-577. Translated from German.
19. Nebrensky, N.T., Morgan, G.E., Oswald, B. J., "Cyclone for Gas/Oil
Separation", International Conference on Hydrocyclones, paper no. 12, held at
Churchill College, Cambridge, Eng., (1980). Organized by BHRA Fluid
Engineering Centre.
20. Oranje, I. L., "Cyclone-Type Separators Score High in Comparative Tests",
Oil & Gas Journal v. 88., no. 4, (22 Jan. 1990), pp. 54-57.
21. Perry, D., Conoco Inc., Private Communication (1995).
22. Rusak, Z., Seginer, A., "Propagation of Small-Amplitude Displacement Waves
on a Vortex in the Center of a Circular Tube", Phys. Fluids A, v. 5., no. 3
(Mar. 1993), pp. 578-587.
23. Taitel, Y., Barnea, D. and Dukler, A. E., "Modeling Flow Pattern Transition
for Steady Upward Gas-Liquid Flow in Vertical Tubes", AIChE J., (1980), vol.
46, pp. 345-354.
24. Zhikarev, A.S., Kutepov, A.M., Solov'ev, V., "Design of a Cyclone Separator
for the Separation of Gas-Liquid Mixtures", Chemical and Petroleum
Engineering v. 21., no. 3/4 (Mar. 1985), pp. 196-198.
25. "Study of the Action of a Cyclones Classifier for Separating Gas-Liquid
Mixtures", Journal of Applied Chemistry of the USSR v. 58, no. 1 (1985): pp.
169-172. Translated from: Moscow Institute of Chemical Engineering.
Zhurnal Prikladnoi Khimli, (Janujary 1985), pp. 180-183.
26. "Cyclone for Gas/Liquid Separators - Platform Equipment", Offshore Services
& Technology (Nov. 1980), pp. 12-17. Paper presented at the International
Conference on Hydrocyclones, Churchill College, Cambridge, Eng., (1980).
329
27. "Well Testing Service Uses New Mini-Separator - Platform Equipment",
Offshore Service & Technology (Jan. 1981), pp. 16-17.
28. "How Good Are Gas-Liquid Separators?", 8th International Conference on
Offshore Mechanics & Arctic Engineering., The Hague, Neth., (Mar. 1989),
The Hague, Neth., OMPEC (Offshore Mechanics & Polar Engineering
Council), (1989), pp. 297-403.