7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Paper No
Introduction
Natural gas lwdrates are crystalline compound of water and
gas. The gas molecules (guests) are trapped in water cavities
(host) that are composed of hydrogenbonded water
molecules (Sloan, 2007). They have been known to plug
transportation pipeline in oil and natural gas industry. With
the exploration of the deepwater oil fields and construction
of the subsea system in petroleum industry, the operation
conditions are ideal for the formation of natural gas lwdrates.
How to prevent lwdrate plugs in pipelines is becoming a
major serious concern for flow assurance in offshore and
deep offshore oilfields.
Traditionally, the pipelines are designed to operate outside
the lwdrate formation region by adding inhibitors or
insulation. Bulk chemicals such as methanol or glycol are
mixed with the liquid water phase to lower the lwdrate
formation temperature. The application of electrically
heated flowlines system can eliminate the hydrate and wax
problems (Bell & Chin, 2005; Sinquin et al, 2004).
However, both of these solutions have significant
economical impact leading to high capital expenditure and
technical limitations. The problem is aggravated by the
large elevation differences involved in deep offshore fields.
Hydrate formation will not be prevented by these solutions
upon startsups following emergency longterm shutins
(Sloan, 2007).
Replacing the thermodynamic hydrate inhibitors with ten to
hundred times more effective nonthermodynamic inhibitors
offers a significant coat reduction to gas companies and
pipeline operators. The nonthermodynamic hydrate
inhibitors called Low Dosage Hydrate Inhibitors are further
categorized into AntiAgglomerant (AA) and Kinetic
Hydrate Inhibitor (KHI) (Lovell & Pakulski, 2002).
Numerical Study the Flow Characteristics of GasHydrate Slurry Two Phase Stratified
Flow
Jing Gong*, Bohui Shi*, Wei Wang*'and Jiankui Zhaos
Beijing Key laboratory of Urban Oil and Gas Distribution Technology, China University of PetroleumBeijing (CUPB),
No. 18 Fuxue Road, Changping District, Beijing, 102249, China
PostDoctoral Research Center of China Petroleum Pipeline Bureau (CPPLB), CNPC, Langfang, China
SChina National Oil and Gas Exploration and Development Corporation (CNODC), No. 61 Fuchengmengbei Avenue,
Xicheng District, Beijing, 100034, China
ydgi acup.edu.cn and sbhtwh ill ;!hloo coml enI
Keywords: gaslwdrate slurry, two phase stratified flow, thermodynamic phase equilibrium, hydrate shell growth model,
compositional model
Abstract
Flow assurance topic of twdrate formation in pipelines in offshore and deep offshore oilfields has been paid more and more
attentions. Formation of twdrate slurry flow is a new technology to avoid the occurrence of twdrate blockage during
multiphase transportation process, especially when the fluids phase contain a dramatically percentage of methane (CH4) and
water phase. AntiAgglomerate Low Dosage Additives allow the formation of hydrate solid particles, while the lwdrate
particles are evacuated with the oil phase as pseudofluid like slurry (Peysson 2005).
This work demonstrates the application of thermodynamic phase equilibrium (Ma et al. 2005) and hydrate shell growth model
(Yapa et al. 2001) with two phase flow simulations in stratified pipeline flow. The schematic of flow with and without lwdrates
is shown in Figure 1.With the strict thermodynamic phase equilibrium calculation, the cross sectional phase distribution can be
determined, thus the thermodynamic quantities are concluded with proper relation expression. The lwdrate shell growth model
considering the kinetics, mass transfer and heat transfer is solved by Euler method. Figure 2 shows the schematic diagram of
twdrate shell growth model and the concentration profile of gas phase. The compositional model is used to simulate two phase
flow, including couple mass, momentum, energy equation and equation of state. All the parameters in all the equations are
interacted. The lwdrate shell growth model is coupled with the compositional model to determine the flow characteristics of
gaslwdrate flow system. Liquid holdup and pressure drop are simulated with this method. Parameters are analyzed that
diffusion coefficient is found to be the key parameter in hydrate forming process. A great agreement is attended between the
results calculated by this study and the data in the literature.
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Paper No
Extensive laboratory testing and formulation work are
performed to ensure that the AA is compatible with system,
and that the products will not plug nor cause upsets in the
transportation system (Frostman, 2000). AA can allow the
formation of hydrate particles, while the hydrate particles
are evacuated with the oil phase as pseudofluid like slurry
(Peysson, 2005). Cold flow technology are proposed by
Gudmundsson (2001) and Tumner & Talley (2008)
respectively, which concerns costeffective flow of oil, gas
and water mixtures in deepwater production pipelines 
from wellhead to processing facility without the constant
(steadystate) use of chemicals to prevent hydrate or wax
deposition.
The thermodynamics and kinetics of hydrate formation have
been studied extensively, and many reliable models are
developed by many researchers (Chen & Guo, 1998:
Englezos et al, 1987a,b: Jamaluddin et al, 1991: Ma et al,
2005; Tumner et al, 2009; Yapa et al, 2001). The mechanism
of hydrate slurry fow suggested by Tumner and his
coworkers (2005, 2009) is that water droplets can be
entrained in an oil phase, and hydrate form and growth as a
shell outside the water droplets. Lots of experiments about
hydrate slurry flow are carried out by many laboratories
worldwide (Andersson & Gudmundsson, 1999; Boxall et al,
2008: Dellecase et al, 2008: Nuland & Tande, 2005;
Pauchard, 2007; Peysson & Nuland, 2003; Peysson, 2005;
Sinquin, 2004). However, all of these theoretical hydrate
slurry flow studies are assumed it is a single phase (hydrate
slurry). In fact, there are four phases in the pipeline flow,
which are oil, gas, water and hydrate.
The experimental and theoretical researches about
oilgaswater flow in pipe have been studied by many years
(Acikgoz et al, 1992: Ghorai et al, 2005; Lee et al, 1993).
Khor et al. (1997) developed threefluid model to estimated
phase holdups in threephase stratified flow. Tentative
recommendations were made on the choice of friction
relationships to provide the best representation of the data.
Bonizzi & Issa (2003) presented a mathematical model to
simulate threephase stratified and slug flows, which was
based on the onedimensional transient twofluid model. As
a result of the formation of hydrate in pipeline, the
multiphase flow is more complex. The multiphase flow with
hydrates is not studied deeply.
Pipeline simulations are now a common tool in multiphase
hydrate slurry flow. With the additives of AA, the hydrate
particles are evacuated with the oil phase as pseudofluid
like slurry. In this paper, a hydrate shell model is established
in gasslurry pipeline flow, which is described by
onedimensional twofluid model. The hydrate shell model
is coupled with pipe flow model to study the flow
characteristics of gashydrate slurry flow system. The
consumption of gas and water caused by hydrate formation
is solved by Euler method and the analysis of influence
factors is performed. It is found that the diffusion coefficient
is a key parameter in hydrate forming process. Considering
the thermodynamic phase equilibrium and hydrate shell
growth model with twofluid model, the simulations of the
gashydrate slurry two phase in stratified pipeline flow is
more close to its practical situations.
Nomenclature
A cross area of the pipeline, nr
cm specific heat of mixture, Jmol K'
C concentration of gas, molm
Deff diffusion coefficient, m2s'
Do outer diameter of pipeline, m
f fugacity of gas, MPa
F friction coefficient
AH the enthalpy of hydrate formation, Jmol'
K kinetic rate, molm MPa's'
mLmaSs transfer from gas to liquid phase, kgs'm3
mLG maSs transfer from liquid to gas phase, kgs 'm3
n mole number, mol
P pressure, MPa
R radius of the water droplet, m
S wetted perimeter, m
t time, s
T temperature, K
Te temperature of environment, K
U overall heat transfer coefficient to environment
of mixture, Wm K'
v velocity, ms'
vLoss aVerage velocity of loss mass phase, ms
Wm mass flow rate of mixture, kgs
xdistance to the inlet of the pipeline, m
(3 JouleThomson effect coefficient, KMPa'
h hydrate structure constant
p density, kgm
cp fugacity coefficient
Shear friction, N
Subsripts
dinner radius of the hydrate shell
eq equilibrium condition
G gas phase
j component index of natural gas
i index of the pipe segment
I gasliquid interface or node index
L liquid phase
s outer radius of the hydrate shell
W water
wt water cut
Numerical Scheme
Hydrodynamics Model
Fluids from subsea wells are multiphase, including oil, gas,
water, even hydrate. In multiphase pipeline simulations,
multifluid model is more complex than two or threefluid
model. It requires much faster computers. In our present
studies, gasslurry flow is described by onedimensional
twofluid model. For simplification, the model assumes that
the flow is in stratified regime whether the hydrates form or
not. The schematic of flow with and without hydrates is
shown in Figure 1. It is assumed that the liquid phase is
waterinoil emulsion in the inlet of the pipeline. With the
additives of AA, the hydrate particles formed from the water
droplets are evacuated with the liquid phase as pseudofluid
like slurry. The governing equations of the gasslurry flow
are expressed by continuity equations, momentum equations
and energy equations. For steadystate flow, neglecting the
rate of change of momentum across the control volume, the
momentum balances are reduced to force balances (Shoham,
Paper No
2005). There must be mass transfer between the gas phase
and liquid phase, which is caused by the formation of
hydrates.
HYDRATE ilii1
Figure 1: Schematic of gasslurry stratified flow and
without hydrates.
(1) Continuity Equation
For the gas phase the continuity equation is given by
(AG G G )= AGD2GL(1
dx
For the slurry phase the continuity equation is given by
(AL LL ) = AL LG (2)
(2) Momentum Equation
For the gas phase the momentum equation is given by
~dpl ~S S = An (3)
For the slurry phase the momentum equation is given by
AL IIL TI ln~ (4)
dYI = mG Loss
The required closure relationships for stratified flow are the
shear stresses, including the interfacial shear stress and the
gas and liquid wall shear stresses. Refer to Equations (5)
and (6) for the definitions of the shear stresses.
pG G
T wo = FG
2
pL L
zwI = FL
2
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
W~c,,W,c ,
+ ~ /7 +exp Ar,
Phase Behaviour Model
Most of the exiting thermodynamic models for predicting
hydrate formation condition are various modifications of the
van der WaalsPlatteeuw model (1959). Based on the
statistical mechanics and twostep hydrate formation
mechanism, Chen & Guo developed a new model (1998) to
predict the formation of the hydrate, which had more clearly
physical sense, simpler mathematical and more accurate
result.
An algorithm for performing gasoilwaterhydrate
multiflash calculation proposed by Ma et al. (2005) is used
in this work. The equation of the state is used PREOS
(Peng & Robinson, 1976). The hydrate phase equilibrium is
calculated by ChenGuo hydrate model (1998). This method
is simple as it avoided the complexity of simultaneous
solution of a sophisticated group of equations. In differential
unit, the authors consider the phase equilibrium is
CStablished instantly (Gong & Zhao, 2008).
With the strict thermodynamic phase equilibrium
calculation, the cross sectional phase distribution can be
determined, thus the thermo physical property can be
estimated with proper correlations. The concentration of the
natural gas in the liquid phase can be defined
simultaneously. The model to defmne the viscosity of the
slurry used in this work is mentioned by Zhao (2009). It is a
function of the hydrate fraction, the viscosity of the oil and
the shear rate of the flow.
Hydrate Shell Growth Model
To predict the volume Fraction of the hydrate is a complex
process. If water contact with gas at the flow condition and
in the hydrate formation region, hydrates form immediately
from the thermodynamic theory. However, kinetic, mass
transfer and heat transfer are the key control factors and the
limitations during the water droplet conversed to hydrate
shell in the pipeline flow. The schematic diagram of hydrate
shell model and concentration profile of gas is shown in
Figure 2. Analogous to a shrinking core model (Homma et
al, 2005), the hydrate growth outside of the water droplets
can be described by hydrate shell growth model.
c
c~c~t
RR R,
Figure 2: Schematic diagram of hydrate shell model and
concentration profile of gas.
The kinetic model proposed by Englezos et al. (1987a, b) is
I, = F, (6)
The velocity of the interface can be approximated by the
velocity of the liquid phase. And, the summation of the mass
transfer from gas to liquid phase and mass transfer from
liquid to gas phase is equal to 1. The initial value of the
hiquid holdup can be defined by dimensionless equation
developed by Taitel & Dukler (1976). Researches on the
friction coefficient calculation have done by many years.
Lots of empirical, sennempirical correlations have been
proposed. The detailed correlations used to define all the
friction coefficients in this work can be found in Zhao
(2009).
(3) Energy Equation
For the pipe segment with the endpoint node (i1, i) and the
length ax, (Figure 1), Zhao (2009) modified the explicit
temperature equation (7) in the pipeline flow with hydrate
formation, which was developed by Deng (2005). The
enthalpy of hydrate formation of simple natural gas hydrate
former (AH) is defined from Sloan (2007).
Paper No
based on the cr stallization theory and twofilm theory.
Since the concentration of the gas can be written in terms of
its fugacity, a modified and simplified kinetic model for the
pure gas component is established as follow:
dnG x/ P P C ()
dt )CL c L eq
The onedimensional concentration distribution in the
hydrate shell can be described by the Fick's second law of
diffusion (Tunner et al., 2009; Yapa et al. 2001; Zhao, 2009).
The gas molecules diffusion rate can be expressed by the
Fick's first law as follow:
"4'iD 1 1
R R
Water consumption rate is in proportion to the radius of the
water droplet, which can be given by
"r, =4;2K rR (10)
dt "1 My dt
In the quasisteady condition during the unit intervals, the
gas molecules diffusing through the hydrate shell are
balanced with the gas molecules consumed around the
surface of the water droplets. Based on the structure
characteristics of hydrates, the water molecules consumed
during the hydrate shell growth are also in proportion to the
total gas consumption (Zhao, 2009). Combining equations
(8), (9), and (10) gives the following expression:
pP pP
dRC 'CL s L eq
.(11)
dt p 1 R1 1 apP
+
K D ,Rc R CL
The initial boundary condition of equation (11) is that the
thickness of the hydrate shell is zero. The improved Euler's
method algorithm is used to solve the above equation. Then,
the water consumption rate can be determined. And the
water conversion rate and the volume fraction of the hydrate
can be defined. This result is used to be modified the liquid
holdup in the computer calculation.
Integrated Model
The major components of the integrated model are the
twofluid model, the phase behaviour model and hydrate
shell growth model. All these are coupled together in a
unified manner to permit integration in the pipeline flow.
The detailed description of the associated assumptions and
the solve procedure are found in Zhao (2009).
Results and Discussion
Owing to particularity of operating conditions, submarine
temperature and transporting technique, condensate gas
hydrate is easy to form in the submarine pipeline. This
integrated model mentioned above is used to simulate the
practical submarine pipeline with the JZ202 offshore field
data (Sun, 2008). There is only a little terrain undulation. It
can be considered as a horizontal pipeline flow. The
diameter of this pipeline is 370 mm with about 50 km in
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
distance. The coefficient of roughness in the pipeline is 0.15
mm. As inlet boundary conditions, a temperature is 200C
and pressure is 6 MPa. The ambient temperature is 40C with
11.62 W/(nr2oC) in coefficient of heat transfer. The flow
rates of the pipeline are up to 981651.6 m3 gas and 353.8 m3
gas condensate oil per day. The composition of the natural
gaS from JZ202 is listed in Table 1. The relative density of
the gas condensate oil is 0.89266. With this high gas oil rate,
hydrates will not formation during the production. To
analyze the flow characteristics in the gashydrate slunry
flow, we modified the inlet fluid. The flow rate used in our
simulation is 15.79 kg/s. We assume that the water cut of the
flow is 10% with 20% gas condensate oil. So, the inlet
holdup is assumed 30%. The components of the gas
condensate oil are also given by Table 1.
Table 1: The compositions of the natural gas and gas
condensate oil
Component Mol% Component Mol%
C1 79.822 C6 2.98
C2 7.583 C7 3.9
C3 4.029 C8 8.55
IC4 0.693 C9 7.07
NC4 1.367 C10 6.46
IC5 0.433 C11 5.39
NC5 0.493 C12 5.64
NC6 0.645 C13 6.91
C7+ 3.653 C14 53.1
N2 0.945
CO2 0.337
For simplification, we assume the kinetic rates and diffusion
rates of each hydrate former component of the inlet fluid are
equalled. In general, the hydrate former of natural gases
include CH4, C2H6, C3H8, iC4H10, nC4H10, N2, H2S
and CO2 (Sloan, 2007). The kinetic rate proposed by
Englezos et al. (1987a, b), 0.65x10'5mol/(m2MPas), and
the diffusion rate presented by Jamaluddin et al. (1991),
7.25 x107 cm2/s are used in this work. From Figure 3, it
shows that the flow is not in hydrate zone after inlet. The
first appearance of hydrate is at 1.5 km from the inlet as
simulated. We could examine the existence of hydrate and
its location along the pipe from the PT diagram. Figure 4
shows the results of steadystate pressure and temperature
distribution and the pressure drop changes about 0.25 MPa
along this 50 km pipeline. Since the hydrates form in the
flow, the liquid velocity decreases and the gas velocity rises
sharply as shown in Figure 5. Then, both of them remain
constant with little changes. The liquid holdup increases
after the hydrate formation and then remains constant as
shown in Figure 6. The reason of this phenomenon is that
the viscosity of liquid phase with hydrate increases. Water
cut along the pipeline decreases to convert into hydrate.
To predict the initial water droplet size distribution requires
a distribution function (Simmmons & Azzopardi, 2001). An
easier way to compare droplet size distributions is to use
characteristic mean diameters. The Sauter mean diameter
d32 is often used to characterize dispersions formed in
dnG
dt
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Paper No
pipelines (Angeli & Hewitt, 2000; Simmons & Azzopardi,
2001). The variety of the water droplet radius implies the
extent of the conversion of the water droplet to hydrate. The
hydrate fraction also increases as presented in Figure 7. The
size of the initial water droplet radius determines the
conversion time. The smaller radius of the droplet is, the
larger contact surface would be between gasliquid. With
same driving forces of reaction, the larger contact surface
would lead to the increase of gas consumption.
* holdup
.water cut
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure6: Distributions of liquid holdup and water cut
along the pipeline.
* hydrate formation curve ~.
=PT along pipeline *
e.............................,......... ........... 
2
1 _....a
~.

* water roplet raduis
hydrate fraction
't<*..
,a~
...***
..**
., 3.5
,E. 3
2
S1.
5.s
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 17 18 19 20
Temperature [C]
Figure 3: Pressure and temperature along the pipeline and
the hydrate formation curve calculated by ChenGuo model.
0 5 10 15 20 25 30
Distance [km]
35 40 45 50
6.2 r
6.1
"5.9
s.s
5.6
 pressure
 temperature
~~ .. L
cr. w
CLL~,, ....
Figure 7: Changes of the water droplet radius and
distribution of the hydrate fraction along the pipeline.
Influence of Initial Water Cut and Diffusion Coefficient
The initial water cut of this simulation is assumed in this
work. To find it how to affect the simulation, we change the
initial water cut to 20% with other condition unchanged.
Me8HWhile, diffusion coefficient is proportional to the
Squared velocity of diffusing particles, which depends on
the temperature, viscosity of the fluid and the size of the
particles. The real diffusion coefficient can not be measured
easily. Most of the diffusion coefficients proposed by other
researchers are regressed by experiment data, such as
5.00 x10 "5.00 x106 c2/S (Makogon, 1981) and 1.40 x1012
m2/S (Turner, 2009a). To study the influence of the diffusion
coefficients to the gashydrate slurry flow, we simulate it
again with these different assumed conditions. The results
are shown in Figure 815.
The first appearance of hydrate is also at 1.5 km from the
inlet as simulated with 20%/wt. With the increase of the
initial water cut, the pressure drop increases to 0.44 MPa as
shown in Figure 8. The temperature changes a little. The
trend of the liquid velocity, gas velocity, liquid holdup,
water cut and hydrate fraction are identical between
different initial water cuts. The increase extent of liquid
holdup with 20%/wt is smaller. The reason of this
phenomenon is that the initial liquid holdup remains 30%.
With the increase of the initial water cut, the initial fraction
of the gas condensate oil decrease. The liquid holdup
depends on combined action of water cut and the fraction
gas condensate oil. The total water conversion rate is 33%
with 20%/wt, and 48% with 10%/wt. This indicates that the
conversion rate will not increase with the increase of the
initial water cut. However, the total hydrate fraction will
increase with the increase of the initial water cut as shown
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 4: Distributions of steadystate pressure and
temperature along the pipeline.
a s
  liquid velocity
.gas velocity
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 5: Distributions of liquid and gas velocity along the
pipeline.
Paper No
in Figure 15.
With different diffusion coefficients at 10%/wt, the trend of
all the values mentioned in this work are identical. The flow
characteristics of the gaslwdrate slurry change a little with
the different diffusion coefficients, such as pressure.
temperature, liquid and gas velocity, liquid holdup. The
influence of the diffusion focuses on the lwdrate shell model.
The more increase of the diffusion coefficient is, the more
increase of water conversion rate is. The lwdrate fraction
increase with the increase of diffksion coefficients. With lots
of analysis and comparison, it is concluded that diffusion
coefficient is the key parameter in the lwdrate shell growth
model.
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
B ii * 7.25E11 10%
> 5.00E12 10%
t  * 5.00E14 10%
0.s  a 1.40E16 10%
e 7.25E11 10%
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 11: Distributions of gas velocity along the pipeline
with different simulated conditions.
70
6.02
~5.78
S5.7
 * .5E1 0 *, s"=u
= 50 E 1 1 %e
. 5.0 14 1 %* .
. 0 1 0 *
7.25E11 20% e.,
** * 7.25E11 10%
 5.00E12 10%
*o 5.00E14 10%
m 1.40E16 10%
*7.25E11 20%
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 8: Distributions of steadystate pressure along the
pipeline with different simulated conditions.
0 5 10 15 20 25 30 35 40
Distance [km]
45 50
Figure 12: Distributions of liquid holdup along the pipeline
with different simulated conditions.
35
* 7.25E11 10%
30  50E1 0
25 * 5.00E14 10%
....  1.40E16 10%
S20 +~_~ 7.25E11 20%
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 13: Distributions of water cut along the pipeline
with different simulated conditions.
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 9: Distributions of steadystate temperature along
the pipeline with different simulated conditions.
''':g~ .u:gg:gg: ::" g ggg ggg gggI:gII:gIII g~ ~:gr~:
* 7.25E11 10%
= 5.00E12 10%
e 5.00E14 10%
 1.40E16 10%
 7.25E11 20%
2.5
1a 2
0.5
0
a0.4 ...... 7.25E11 10%
. 5.00E12 10%
0.2 * 5.00E14 10%
 1.40E16 10%
 7.25E11 20%
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure10: Distributions of liquid velocity along the
pipeline with different simulated conditions.
0 5 10 15 20 25 30
Distance [km]
35 40 45 50
Figure 14: Changes of the water droplet radius along the
pipeline with different simulated conditions.
* 7.25E11 10%
 5.00E12 10%
I 5.00E14 10%
o1.40E1 6 10%
a ~7.25E11 20%
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
181191 (1999)
Angeli, P. & Hewitt, GF. Drop Size Distributions in
Horizontal OilWater dispersed Flows. Chemical
Engineering Science, Vol. 55, 31333143 (2000)
Bell, J.M., Chin, Y.D. & Hanrahan, S. State of the Art of
Ultra Deepwater Production Technologies. OTC Paper
17615 (2005)
Bonizzi, M.& Issa, R.I. On the Simulation of ThreePhase
Slug Flow in Nearly Horizontal pipes using the multifluid
model. International Joumnal of Multiphase Flow, Vol. 29,
17191747 (2003)
Boxall, J., Davies, S., Nicholas, J., Koh, C.K., Sloan, E.D.,
Turner, D. & Talley, L. Hydrate Blockage Potential in an
OilDominated System Study Using a Four Inch Flow Loop.
Proceedings of the 6th Intemnational Conference on Gas
Hydrates, Vancouver, Canada (2008)
Chen, G J. & Guo, T.M. A New Approach to Gas Hydrate
Modeling, Chemical Engineering Joumnal, Vol. 71, 145151
(1998)
Dellecase, E., Geraci, G, Barrios, L., Eatanga, D.,
Domingues, R. & Volk, M. Hydrate Plugging or Slurry Flow:
Effect of Key Variables. Proceedings of the 6th Intemnational
Conference on Gas Hydrates, Vancouver, Canada (2008)
Deng, D.M. Modeling GasCondensate TwoPhase Flow in
Pipelines. [Ph.D. Thesis]. China University of Petroleum,
Beijing, China (2005)
Englezos, P., Kalogerakis, N., Dholabhai, P.D. & Bishnoi,
P.R. Kinetics of Formation of Methane and Ethane Gas
Hydrates. Chemical Engineering Science, Vol. 42,
26472658 (1987a)
Englezos, P., Kalogerakis, N., Dholabhai, P.D. & Bishnoi,
P.R. Kinetics of Gas Hydrate Formation from Mixtures of
Methane and Ethane. Chemical Engineering Science, Vol.
42, 26592666 (1987b)
Frostman, L.M. AntiAgglomerant Hydrate Inhibitors for
Prevention of Hydrate Plugs in Deepwater Systems. SPE
Paper 63122 (2000)
Ghorai, S., Suri, V. & Nigam, K.D.P. Numerical Modeling
of ThreePhase Stratified Flow in Pipes. Chemical
Engineering Science, Vol. 60, 66376648 (2005)
Gong, J. & Zhao J. K. Numerical Simulation of
GasHydrate Slurry Two Phase Flow. Proceedings of the 6th
International Conference on Gas Hydrates, Vancouver,
Canada (2008)
Gundmundsson, J.S. Cold Flow Hydrate Technology.
Proceedings of the 4th Intemnational Conference on Gas
Hydrates, Yokohama, Japan (2002)
Homma, S., Ogata, S., Koga, J. & Matsumoto, S. GasSolid
Reaction Model for a Shrinking Spherical Particle with
Paper No
. 5.00E12 10%
S30
* 5.00E14 10% ..** .
25 a1.40E16 10% .*
20 * 7.25E11 20% .*~..
0' .. * ' ........
0 5 10 15 20 25 30 35 40 45 50
Distance [km]
Figure 15: Distributions of the lwdrate fraction along the
pipeline with different simulated conditions.
Conclusions
In this work, the application of thermodynamic phase
equilibrium and hydrate shell growth model with two phase
flow simulations in stratified pipeline flow. With the strict
thermodynamic phase equilibrium calculation, the cross
sectional phase distribution can be determined, thus the
thermodynamic quantities are concluded with proper
relation expression. The lwdrate shell growth model
considering the kinetics, mass transfer and heat transfer is
solved by Euler method. The compositional model is used to
simulate two phase flow, including couple mass, momentum,
energy equation and equation of state. All the parameters in
all the equations are interacted. The lwdrate shell growth
model is coupled with the compositional model to
determine the flow characteristics of gaslwdrate flow
system. Liquid holdup and pressure drop are simulated with
this method. Enthalpy balance equation is substituted by
explicit formulation of temperature. So the calculation is
greatly speeded up.
The influence of the initial water cut and the diffusion
coefficient is performed. This indicates that the conversion
rate will not increase with the increase of the initial water
cut. However, the total hydrate fraction will increase with
the increase of the initial water cut. The influence of the
diffusion focuses on the hydrate shell model. It is found that
the diffusion coefficient is a key parameter in the process of
lwdrate formation. Considering the thermodynamic phase
equilibrium and hydrate shell growth model with twofluid
model, the simulations of the gaslwdrate slurry two phase
in stratified pipeline flow is more close to its practical
situations.
Acknowledgements
The authors would like to acknowledge the financial support
provide by the National Science & Technology Major
Project (No. 2008ZXO502600403).
References
Acikgoz, M., Fiarlca E. & Lalley, R.T. An Experimental
Study of ThreePhase Flow Regimes. Internal Joumnal of
Multiphase Flow, Vol. 18, 327336 (1992)
Andersson, V. & Gudmundsson, J.S. Transporting Oil and
Gas as Hydrate Slurries, BHR Group Hydratransport 14,
7th International Conference on Multiphase Flow
ICMF 2010, Tampa, FL USA, May 30June 4, 2010
Sloan, E.D. & Koh, C.A. Clathrate Hydrates of Natural
Gases, 3rd Edition, CRC Press, New York, 1, 552, 645
(2007)
Sun, B.B. Study on Process Calculation of Natural Gas and
Condensate Pipeline. [MI.D. Thesis]. China University of
Petroleum, Beijing, China (2008)
Turner, D.J., Kleehammer D.M., Miller, K.T., Koh, C.A.,
Sloan, E.D. & Talley, L.D. Formation of Hydrate
Obstructions in Pipelines: Hydrate Particles Development
and Slurry Flow. Proceedings of the 5th Intemnational
Conference on Gas Hydrates, Houston, Texas, USA (2005)
Tuner, D.J., Miller, K.T. & Sloan, E.D. Methane hydrate
formation and an inward growing shell model in
waterinoil dispersions. Chemical Engineering Science, Vol.
64, 39964004 (2009)
Turner, D. & Talley, L. Hydrate Inhibition via Cold Flow 
No Chemicals or Insulation. Proceedings of the 6th
International Conference on Gas Hydrates, Vancouver,
Canada (2008)
van der Waals, J.A. & Platteeuw J.C. Clathrate Solutions.
Advances in Chemical Physics, Vol. 2, 257 (1959)
Yapa, P.D., Zheng, L. & Chen, F.H. A Model for Deepwater
Oil/Gas Blowouts. Marine Pollution Bulletin, Vol. 43,
234241 (2001)
Zhao, J.K. Study on Flow Properties of Hydrate Slurry in
Multiphase Pipeline. [Ph.D. Thesis]. China University of
Petroleum, Beijing, China (2005)
Paper No
Unreacted Shrinking Core. Chemical Engineering Science,
Vol. 60, 49714980 (2005)
Jamaluddin, A.K.M., Kalogerakis, N. & Bishnoi, P.R.
Hydrate Plugging Problems in Undersea Natural Gas
Pipelines under Shutdown Conditions. Journal of Petroleum
Science and Engineering, Vol. 5, 323335 (1991)
Khor, S.H., MendesTatsis M.A. & Hewitt, GF.
OneDimensional Model of Phase Holdups in ThreePhases
Stratified Flow. Internal Joumnal of Multiphase Flow, 1997,
Vol. 23, 885897 (1997)
Lee, A. H. Study of Flow Regimes of Transition
Oil/Water/Gas Mixtures in Horizontal Pipelines. Proceeding
of 5th Intemnational Offshore and Polar Engineering
Conference, Singapore (1993)
Lovell, D. & Pakulski, M. Hydrate Inhibiton in Gas Wells
Treated With Two Low Dosage Hydrate Inhibitors. SPE
Paper 75668 (2002)
Ma, Q.L., Chen GJ., Sun C.Y. & Guo T.M. New Algorithm
of VaporLiquidLiquidHydrate MultiPhase Equilibrium
Flash Calculation. Journal of Chemical Industry and
Engineering (China), Vol. 56, 15991604 (2005)
Makogon, Y.F., Hydrates of Natural Gas. Pennwell
Publishing Company, Tulsa, OK, 85 (1981)
Nuland, S. & Tande, M. Hydrate Slurry Flow Modelling.
12th International Conference on Multiphase Production
Technology, Barcelona, Spain (2005)
Pauchard, V., Darbouret, M., Palermo, T. & Peytavy J.L.
Gas Hydrate Slurry Flow in a Black Oil. Prediction of Gas
Hydrate Particles Agglomeration and Linear Pressure Drop.
13th International Conference on Multiphase Production
Technology, Edinburgh, UK (2007)
Peng, D.Y. & Robinson, D.B. A New TwoConstant
Equation of State, Industrial & Engineering Chemistry
Fundamentals, Vol. 15, 5964 (1976)
Peysson, Y. Collision Process between Particles in the
Transport of Dispersed Hydrates in Production Lines,
Proceedings of the 5th Intemnational Conference on Gas
Hydrates, Houston, Texas, USA (2005)
Peysson, Y., Nuland, S. & Maurel P. Flow of Hydrates
Dispersed in Production Lines. SPE Paper 84044 (2003)
Simmons, M.J.H. & Azzopardi, B.J. Drop Size Distributions
in Dispersed LiquidLiquid Pipe Flow. International Joumnal
of Multiphase Flow, Vol. 27, 843859 (2001)
Shoham, O. Mechanistic Modeling of GasLiquid
TwoPhase Flow in Pipes. Society of Petroleum Engineers,
7587 (2005)
Sinquin, A., Palermo, T. & Peysson, Y. Rheological and
Flow Properties of Gas Hydrate Suspensions. Oil & Gas
Science and Technology, Vol. 59, 4157 (21 r 14)
