SPE 97978

Temperature Distribution in Natural Gas/Condensate Pipelines

Using a Hydrodynamic Model
A.A. Sadegh and M.A. Adewumi, Pennsylvania State U.

Copyright 2005, Society of Petroleum Engineers

operational difficulties arise. Multiphase flow significantly
This paper was prepared for presentation at the 2005 SPE Eastern Regional Meeting held in reduces the volumetric capacity of pipelines and necessitates
Morgantown, W.V., 14–16 September 2005.
installation of separation stations and implementation of
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as
expensive pigging schedules.
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to Basically the design engineer needs to know the size
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at pipe to be used which is obviously related to the expected
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
pressure drop. In addition it is necessary to estimate the liquid
for commercial purposes without the written consent of the Society of Petroleum Engineers is dropout location and quantity in order to design and place
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous efficient separation units. The operator of the pipeline needs
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
to establish an efficient pigging schedule in order to minimize
operating cost and therefore maximize profit. In addition,
Abstract transmission companies periodically perform optimization
A hydrodynamic model for predicting the flow of natural gas studies aimed at improving the overall technical and economic
and condensate in transmission pipelines has been developed. soundness of there operations.
This steady state model was used to study the temperature The aforementioned requirements make it a necessity
profile along a transmission pipeline. The model is based on to have tools for multiphase flow analyses. Traditionally these
formulating the problem in terms of mass, momentum, and analyses have been performed using empirical correlations.
energy balance equations. Thermodynamic properties are Among the more popular of these methods are the Beggs and
determined using flash and thermo-physical properties Brill [1973] method, and the Aziz et al. [1972] method. Brill
calculators based on a Volume Translated Peng-Robenson and Mukherjee [1999] present a detailed discussion of these
EOS (VTPREOS). methods.
This new model will provide gas transmission In order to supplement these methods, hydrodynamic
engineers with a versatile tool for modeling multiphase flow in modeling has been used at Penn State [Adewumi et al, 1993]
pipelines. It will also provide approximations to the for the passed twenty years to provide a more detailed and
parameters of most importance to them specifically, the theoretically sound analysis of multiphase flow phenomenon
impact of temperature and terrain effects on liquid dropout. of gas and condensate flow. Successive development was
Liquid formation is a major concern for transmission made by Vincent [1988], Mucharam [1990], Boriyantoro
companies due to the significant decrease in gas flow capacity, [1994], Martinez [1994], Carrillo [1999], Antonini [2000],
inaccuracy in metering, and potential damage to Ayala [2001], and Eltohami [2003].
instrumentation and equipment. This model simply tells the The present model incorporates several contributions to
field engineer where and how much liquid will form. This the multiphase flow research. First a more accurate PVT
information is valuable for locating and sizing liquid model is used in order to better predict density of the liquid.
collection tanks, and in the design of a cost effective pigging Second, the energy equation is coupled to the mass and
schedule. momentum equations which is a more theoretically sound
The results of the systematic numerical studies showed approach compared to the normal approach of using a
that the temperature of the gas decreased from the inlet correlation to determine the temperature distribution. The
condition to that of the surrounding temperature within 3 miles purpose of this paper is to address in details temperature
of the inlet. In addition a temperature rise of up to 5 degrees variations in gas / condensate flow.
Fahrenheit was observed when an undulating terrain was Formulation of the Hydrodynamic Model
simulated (in downhill flow). The Steady-State Formulation
Introduction The derivation of the hydrodynamic model is based
Multiphase flow phenomenon occurs in several on the conservation of mass, momentum and energy. The
hydrocarbon production systems. This phenomenon poses derivation is based on the conventional control volume
some serious problems for the design and operation engineers approach as detailed by Bird et al., [2002], Ayala [2001], and
and personnel. The frictional pressure losses necessary to size Eltohami [2003]. We present here the formulation of the
pipes become much more difficult to calculate. In addition model based on the two-fluid approach. The steady-state
2 SPE 97978

model is formulated so that the liquid volume fraction α l , the ⎛ ∂ρ g ∂ρ ∂h ⎞

⎜ α g hg vg + α l hl vl l + α g ρ g vg g ⎟
gas volume fraction α g , pressure, temperature, gas and liquid ∂P T ∂P T ∂P T ⎟ dP
⎜
velocities are the principal unknowns. The mass conservation ⎜ ∂h ⎟ dx
equation for the gas phase is then given by: ⎜ +α l ρl vl l ⎟
⎜ ∂P ⎟
⎝ T ⎠
∂ρ g dP dvg dα g dvg dα g
α g vg + α g ρg + ρ g vg dv
∂P dx dx dx +α g ρ g hg + α l ρl hl l + ( ρ g hg vg − ρl hl vl )
T
[1]
dx dx dx
∂ρ g dT ⎛ ∂ρ g ∂ρ ∂hg ⎞
+α g vg = M gl ⎜ α g hg vg + α l hl vl l + α g ρ g vg ⎟
∂T P
dx ⎜ ∂T P ∂T P ∂T P ⎟ dT
+
⎜ ∂h ⎟ dx
And is given for the liquid phase by ⎜ +α l ρl vl l ⎟
⎜ ∂ T ⎟
⎝ P ⎠
∂ρl dP dv dα g dα g P
α l vl + α l ρl l − ρl vl dα l P
∂P dx dx dx = vg + vl + Qg + Ql + hg M g + hl M l
T
[2] dx dx
∂ρ dT [5]
+α l vl l = − M gl The Closure Relationships
∂T P dx The Mass Continuity Equation. Consider the right hand
The gas and liquid momentum equations are given by: M
gl
side of equation [1]. The mass transfer term is a
generation term that is unique to natural gas/condensate
⎛ ∂ρ g ⎞ dP dvg dα g
⎜⎜ α g vg
2

∂P T
+αg ⎟
⎟ dx
+ 2α g ρ g
dx
+ ρ g vg2 + P( dx
) transportation system. Differences in pressure, temperature or
composition of the flowing gas/condensate mixture lead to
⎝ ⎠ deviation from equilibrium state. This is shown schematically
∂ρ g dT in Figure 1. The system reacts by either transferring mass
+α g vg2 = from gas phase to condensate phase (condensation) or vise
∂T P dx versa (evaporation).
− Fmgl − Fgg − Fwg − Fdgl Quantitatively, the mass fraction of liquid increase upstream
of x in Figure 1 can be expressed as (Mucharam [1990])
[3]
M gl =
(
W f mgx + ∆x − f mgx ) [6]
⎛ 2 ∂ρl ⎞ dP dv dα l AP ∆x
⎜ α l vl
∂P T
+ αl ⎟ + 2α l ρl l − ρl vl2 + P( ) The Momentum Equations. In the two-phase system
⎝ ⎠ dx dx dx
under consideration, phases act upon each other and are acted
∂ρ dT upon by the pipe walls and by the force of gravity. Newton’s
+α l vl2 l = [4] third law of dynamics states that each action on a given body
∂T P dx causes an equal and opposite reaction from that body, thus it is
Fmgl − Fgl − Fwl + Fdgl sufficient to consider one phase in this discussion. Let’s
consider the liquid phase, specifically the right hand side of
And the combined energy equation for the two fluids is given equation [4]. We will analyze the force resulting from mass
by
Fmgl
transfer between the two phases , the friction force

exerted on the liquid by the pipe wall,

Fwl , and the force
Fdgl
exerted on the liquid due to friction with the gas phase ,
F
as well as force of gravity gl .
Mass Transfer Force. When mass is exchanged between
the two phases, momentum is exchanged with it as well. This
occurs in the sense that if mass is transferred from the phase
with higher velocity then its transfer will cause a speed up of
the slower phase. This force is estimated using:
Fmlg = M lg ( vl − vg ) = − Fmgl [7]
SPE 97978 3

Wall Friction Force. The friction force on the wall is Interfacial Drag Force. The interfacial drag force acting
expressed in the conventional way of a drag force [Agrawal et on the gas phase due to contact with liquid at the interface is
al. 1973, Taitel and Dukler 1976]. This force can be given by:
expressed as ρ g (v g − vl ) v g − vl
ρ l v l vl Fdg = Agl f gl [13]
Fwl = Awl f wl [8] 2
2g c In the above expression, Agl is the area of the interface
Where:
Fwl Is the friction force per unit volume [lbf/ft3] between gas and liquid and f gl is the friction factor
associated with interfacial drag. The gas also exerts an equal
Awl Is the contact area between the liquid phase and and opposite force on the liquid phase. This force is given by
the pipe wall [ft2/ft3] Fdl = − Fdg [14]
f wl Is the Fanning friction factor The Energy Balance Equation. Energy can be transferred
ρl Is the density of the liquid [lbm/ ft3] between the two flowing phases or between each phase and
the surroundings.
vl Is the actual liquid velocity [ft/sec] Convective Heat Transfer. Heat is exchanged between
g c Is a conversion factor (32.174 [lbm-ft/lbf-sec2]) both of the flowing phases and the surroundings based on the
temperature gradient between the fluids and the surroundings.
We note here that this method of determining wall
Heat will therefore flow to the surroundings when this
friction assumes that the liquid and gas phase can be
temperature gradient is positive. On the other hand, if the
considered as flowing separately in two pipes such that the
surroundings are at a higher temperature, heat flow will be in
cross section of the pipe conducting the liquid plus the cross
the opposite direction. Mathematically, this heat exchange
section area of the pipe conducting the gas is equal to the cross
can be quantified as
sectional area of the actual pipe occupied by both fluids. This
approximation is depicted in Figure 2.
The phase friction factor is a function of the phase (α g u g + α l u l )(T − Ts )
Reynolds number where Reynolds number is for a particular
QT = [15]
DP
phase i is given by
Where u g is the heat transfer coefficient between gas phase
i ρ i vi d hi
N Re = [9]
and surroundings and u l is the coefficient of heat transfer
µi
from liquid to surroundings. The temperature of the
Where d hi is the hydraulic diameter given by Dukler and
surrounding is denoted by Ts .
Taitel [1976].
For laminar flow the Fanning friction factor is given by: Latent Heat of Vaporization (or Condensation). When
vaporization occurs, excess heat in the system is absorbed by
16 the liquid phase, this latent heat is quantified by
f wl = [10a]
N Re l ∆h l = hl − hg [16]
And for turbulent flow, N Re l > 2,100 , it is given by applying Where:
Chen’s [1979] correlation as: hg Is gas phase enthalpy per unit volume [BTU/lb.ft3], and
⎡ ε 5.0452 ⎤ hl Is liquid phase enthalpy per unit volume [BTU/lb.ft3].
⎢ 3.7065d − N ⎥
1 ⎢ hl Re l
⎥
= −4.01log On the other hand, when condensation occurs, excess heat is
f wl ⎢ ⎛ ( ε / d hl ) 5.8506 ⎞ ⎥ released from the gas, in this case, latent heat is given by:
⎢× log ⎜ + 0.8981 ⎟ ⎥
⎣⎢ ⎝ 2.8257 N Re l ⎠ ⎦⎥ ∆h l = hg − hl [17]
[10b] Flow Regime Determination and Transitions
Where: Preamble
ε Is relative roughness of pipe [in/in]. When a natural is flowing in a pipe, in the gas phase, changes
Gravity Force. The force of gravity per unit volume of in pressure, temperature, or composition may lead to the
fluid is given by: formation of condensate. From the onset, condensate is
Fgg = α g ρ g g sin φ [11] formed as a dispersed mist of droplets; such flow regime is
typically known as mist flow. As liquid accumulation
For the gas phase, and given by:
increases, liquid will start to accumulate at the bottom of the
Fgl = α l ρ l g sin φ [12] pipe. However, at low liquid levels, the flowing gas stream
For the liquid phase. Note that φ is the inclination angle for a tends to smear the liquid phase leading to the formation of an
annulus in the pipe. Such flow regime is normally referred to
particular pipe segment from the horizontal. as annular/mist flow.
4 SPE 97978

As the amount of liquid increases further, it accumulates at the P⎛ v 1⎞

bottom of pipe, however if the gas velocity is high enough, ln φi = ∫ ⎜ i − ⎟dP i = 1, n
shear at the interface causes a wavy surface to form on the 0
⎝ RT P ⎠
liquid. This pattern is known as stratified wavy flow. [19]
Reduction in gas flow rate leads to shear reduction leading to Where vi is the partial molar volume with respect to
smoothing of the interface. This is known as stratified smooth
flow. Note that flow regimes are conceived due to dominance composition and is symbolically given by:
of one of several competing forces. These forces are the ⎛ ∂V ⎞
gravity force, surface cohesion force, and shear forces. vi = ⎜⎜ ⎟⎟ j ≠ i; i = 1, n [20]
Modeling of regime dependent forces ⎝ ∂z i ⎠T ,P, z j
Wall shear force. The general formula for this force has It has been reported by several authors that cubic
been given in equation [8], it is restated here for clarity equations of state inaccurately predicts the volume of heavy
ρ l v l vl hydrocarbons ( [Peneloux et al., 1982], [Gmehling and Wang,
Fwl = Awl f wl [8] 1999]). It is therefore of important to correct this volume
2g c discrepancy.
Here we have chosen the liquid for illustration purposes. In Peneloux et al. proposed the concept of volume
order to evaluate the wall friction force, it is necessary to translation. They argued that the volume obtained using
determine the wetted area Awl from Table 1 and the phase SRKEOS is a “pseudo-volume” which is given symbolically
by:
friction factor f wl which is a function of the hydraulic vactual = vSRKEOS − c [21]
diameter d hl from Table 2. They claimed that a translation along the volume axis will
In Tables 1 and 2, θ is the wetted angle in stratified smooth improve volume prediction without changing the equilibrium
and stratified wavy flow regimes. This angle is depicted in conditions. They provided the following proof:
Figure 3. They defined a “pseudo volume” as:
n
Interface drag force. When gas and liquid travel at ~
different velocities, the slower phase exerts a drag for on the V = V + ∑ ci z i [22]
i =1
other phase. This force is given by equation [13] as:
ρ g (v g − vl ) v g − vl
A translated functional relationship can then be used to
describe pressure in the following form:
Fdg = Agl f gl [13] ~ ~
2 P = F (T , V , z1 ,..., z i ,..., z n ) [23]
In order to evaluate [13] , it is necessary to first compute the A “pseudo partial volume” can be defined as:
interface area. This area is dependent of the flow regime. ~
⎛ ∂V ⎞
Table 3 summarizes the formulae used to calculate this area. v~i = ⎜⎜ ⎟
⎟ = v i + ci j ≠ i; i = 1, n
Second, interface friction factor is computed. This is ∂z
⎝ i ⎠T ,P,z j
performed according to formulae in Table 4.
Flow regime transition criteria. The criteria given by [24]
Ayala [2001] are used in the present model. For the sake of With this definition “pseudo fugacity coefficients” can be
briefness the reader is referred to the original reference. defined as:
Thermodynamic and Thermo-Physical Properties ~ ~
P⎛ v 1⎞
Model ln φi = ∫ ⎜ i − ⎟dP i = 1, n
Thermodynamic properties as well as thermo-
0
⎝ RT P ⎠
physical properties needed for model closur are determined [25]
according to the methods outlined by Erdogmus [2000]. The A short manipulation of equation [25] shows that:
Volume Translated Peng-Robinson equation of state
(VTPREOS) is the basis of semi-theoretical computations. P⎛ v 1⎞ P⎛ v 1⎞ P c
When these calculations are not mathematically tractable, ln φi = ∫ ⎜ i − ⎟ dP = ∫ ⎜ i − ⎟ dP + ∫ i dP
empirical correlations are used. Here we will only discuss the
0
⎝ RT P ⎠ 0
⎝ RT P ⎠ 0 RT

volume translation technique, since PREOS is discussed in cP

details elsewhere [Peng and Robinson, 1976]. = ln φi + i
Theoretical development RT
An equation of state is a functional relationship . [26]
between the pressure, temperature, volume, and composition “Pseudo” Phase equilibrium is reached when the when the
of a system. In general, this relationship can be stated as: concentration and activity of a given compound in the liquid
phase is balanced by its composition and activity in the gas
P = F (T ,V , z1 ,..., z i ,..., z n ) [18]
phase. This can be stated symbolically by:
The fugacity coefficients used to establish equilibrium ~ ~
conditions are given by: xi φ i L = y i φ i V [27]
Equation [27] can be rewritten using equation [26] as follows:
SPE 97978 5

⎛c P⎞ ⎛c P⎞ Table 6. A 15 mile, multi-elevation pipeline is used in the

xiφiL exp⎜ i ⎟ = y iφiV exp⎜ i ⎟ [28] study. The inlet stream is at 1500 Pisa and is at a temperature
⎝ RT ⎠ ⎝ RT ⎠ of 120ºF. The flowrate is 500 MMSCFD. The pipe has a
Equation [28] is the same as: uniform nominal diameter of 26 inches and an absolute
roughness of 8.0x10-5 inches. Table 7 provides the elevation
xiφiL = y iφiV [29]
profile and inclination angle for each segment of pipe.
Which is the equilibrium criterion for the PREOS. Temperature Distribution for Flat Terrain
For this study we used the volume translation The system under investigation here is a complex
techniques proposed by Gmehling and Wang [1999]. The one. Energy changes in the system occur due heat exchange
reason for this is that their correction accounts for temperature with the surroundings, to latent heat of vaporization and
variation and also is based on a large data bank incorporating condensation, to work performed by the fluid to overcome
inorganic compounds normally present in natural friction, and due to potential energy changes due to the force
gas/condensate fluids. of gravity. All these changes in work and heat are in tern
Volume Correction Procedure manifested as internal energy changes. This is what causes the
Gmehling and Wang provide the following relations for temperature of the system to change. The gas/condensate
determining the translated volume using SRKEOS for mixture enters the pipe at 120°F and the surrounding ground
mixtures normally found in natural gas. Here their method is temperature is 60°F. This temperature gradient causes
extended to PREOS. The following steps are taken to correct convective heat transfer from the mixture to the surroundings.
volume: This is manifested as the sharp temperature decrease for a
1. The experimental molar volumes at the critical relatively short distance from the inlet, as shown in Figure 4.
pressure and temperature are compared to the The Temperature seems to stabilize at about 60°F within 3.32
volumes obtained using PREOS and the volume miles.
difference is estimated. After convective heat transfer is no longer significant
2. The mixture critical volume and volume deviation due to equilibrium between the inside and outside temperature,
terms are determined with: it can still be seen that a slight decrease in temperature occurs.
n
This decrease is approximately 1.2°F. To explain this drop we
c m = ∑ z i ci [30a] examine the pressure profile in Figure 4. The pressure of the
i =1 gas decreases due to friction with the wall of the pipe and due
n
to the drag at liquid-gas interface. This decrease is fairly
∆v mc = ∑ z i ∆vci [30b] uniform. When the pressure of a gas drops, the gas expands.
i =1
For expansion to occur; internal energy of the system is
3. The parameter B2 m is calculated according to: reduced and as a result, temperature decreases. It should be
n noted that temperature reduction due to expansion occurs
B 2 m = ∑ z i B2 i [31] within 3.32 miles of the inlet. However its effect is not
i =1 observed because it is small in magnitude compared to heat
loss to the surroundings.
Where B2i correlation parameters are reported [Gmehling
Temperature Distribution for Undulating Terrain
and Wang, 1999]. The calculation formulae are given in Figure 5 shows a steady temperature decrease up to
Table 5.1. the first downward sloping point. From this point onward,
temperature increases steadily up to the point of alteration to
4. Next compute the parameter B1m : an upward sloping terrain. First, we note that this change
occurs after the temperature of the mixture inside the pipe and
⎛ cm 2 ⎞
temperature of the surrounding have equalized. Now we can
⎜⎜ − 1 − B1m 0.3 3 ⎟⎟
∆v proceed to explain this trend. For the following analysis it is
B1m = ⎝ cm ⎠
1 [32] assumed that the system under investigation consists of the
0.3 3
pipe and the fluids inside it. We note that this system is open
Finally the temperature-dependent volume correction can be and dynamic. Never the less, the system must still satisfy the
determined using: first law of thermodynamics. The first law is stated here in
c m′ = ∆v cm ⎛⎜1 − B1m 1 − Trm ⎞⎟
1 2 terms of the following conservation of energy equation:
3
+ B21m 1 − Trm 3
⎝ ⎠
[33] d ⎡ ⎛ 1 ⎞ ⎤ dQ
⎢ ρv ⎜ e + v2 ⎟⎥ = − Ws − Wb [34]
Where ω is the accentric factor and Trm is the reduced dx ⎣ ⎝ 2 ⎠ ⎦ dx
temperature of the mixture. Where:
e Is the specific internal energy
Q Is the heat exchange with
Results and Discussion
surroundings.Assumed positive for transfer
Model validation case
into system
The model was validated using the wet gas
composition from Iran [Mocharam, 1994] summarized in Ws Is the work performed by surface forces on
6 SPE 97978

the system
Wb Is the work performed by body forces on the References
Adewumi, M.A., Nor-Azlan, N. and Tian, S: “Design Approach
system Accounts for Condensate in Gas Pipelines,” SPE Paper 26904
The gravity force is the only body force acting on the presented at the SPE Eastern Regional Conference and
system. If the system does work to counteract the force of Exhibition, Pittsburgh, Pennsylvania, November 2-4, 1993
gravity as is the case in uphill follow then energy is extracted Agrawal, S.S., Gregory, G.A. and G.W. Govier: “An Analysis of
from the system. This energy may be extracted in the form of Horizontal Stratified Two-Phase Flow in Pipelines,” Canadian
kinetic, potential, or internal energy. When the flow is Journal of Chemical Engineering, v. 51, 1973.
downhill, then, work is performed by the gravity force on the Ayala, L.F.: A Unified Two-Fluid Model for Multiphase Flow in
Natural Gas Pipelines, MS Thesis, the Pennsylvania State
system and as such this excess energy is stored in one form as
University, University Park, Pennsylvania, 2001.
internal energy. As more energy is stored, molecules within Aziz, K., G.W. Govier, and M. Fogarasi: “Pressure Drop in Wells
the system vibrate faster and this in tern lead to increased Producting Oil and Gas,” Journal of Canadian Petroleum
temperature. Technology, July-Sept. 1972.
Liquid holdup and liquid drop out Baker, A., Nielsen, K. and Gabb, A.: “Pressure Loss, Liquid Holdup
Figures 7 and 8 shows the liquid holdup and liquid Calculations Developed,” Oi l & Gas Journal, March 14, 1988.
drop out respectively. It can be seen that the liquid holdup is Beggs, H.D. and Brill, J.P.: “A study of Two-Phase Flow in Inclined
strongly dependent on terrain effect. Let’s first focus on the Pipes,” SPE Paper 4007, Journal of Petroleum Technolog, May
first uphill rise in the pipe. The phenomena taking place here 1973.
Bird, R.B., Steward, W.E.and Lightfoot, E.N.: Transport Phenomena,
are complex indeed. As the pressure drops (Figure 6)
Second Edition, John Wiley & Sons Inc., 2002.
liquid condenses, this is a purely thermodynamic condition. Boryiantoro, N.H.: An Integrated Single / Multiphase Flow Model for
However there are the hydrodynamic effects to be considered Gas Condensate Pipelines, MS Thesis, The Pennsylvania State
as well. However from the beginning of the first inclined pipe University, University Park, Pennsylvania, 1994.
to about its center point condensation dominates. After this Brill, J.P. and Mukherjee, H.: Multiphase Flow in Wells, SPE
point however, the velocity gradient between the liquid phase Monograph No. 17, Richardson, Texas, 1999.
and the gas phase predominates, this is influenced by the fact Carillo, S.: A Compositional Network Pipeline Model as a Tool for
that the gravity drag on the liquid is much higher than its pull Decision Making, MS Thesis, The Pennsylvania State
on gas. This leads to liquid evaporation because an optimum University, University Park, Pennsylvania, 1999.
Chen, N.H.: “An Explicit Equation for Friction Factor in Pipe,”
contact surface area between the two phases must be
Industrial and Engineering Chemistry Fundamentals, v. 8(3),
maintained to ensure no-slip at the interface. 1979.
Eltohami, E.S.: Modeling of PCB Removal Processes From Natural
In the downhill section, the situation is reserved. Gas Transmission Lines, PhD Thesis, The Pennsylvania State
That is the pressure rises due to potential energy rise. This University, University Park, Pennsylvania, 2003.
leads to further evaporation. Also the liquid accelerates much Erdogmus, M.: Development of a Modefied Patel-Teja Equation of
faster now due to its high density compared to the gas. This State, PhD Thesis, The Pennsylvania State University,
second phenomenon also favors evaporation to increase the University Park, Pennsylvania, 2000.
interfacial area between gas and liquid. Gmehling, J. and Wang, L.: “Improvement of the SRK Equation of
State for Representing Volumetric Properties of Petroleum
Figure 4.5 also shows the cumulative liquid dropout
Fluids Using Dortmund Data Bank,” Chemical Engineering
in the pipe. It is noted that the rate out dropout in the downhill Science, v. 54(17), 1999.
segment of the pipe is lower compared to the uphill segments. Martinez A., F.F.: Two-Phase Gas-Condensate Flow in Pipeline
Conclusion Open-Network Systems, MS Thesis, The Pennsylvania State
It is hoped that this paper has demonstrated the utility University, University Park, Pennsylvania, 1994.
of hydrodynamic modeling of gas/condensate flow to pipeline Mucharam, L.: One-Dimensional Compositional Modeling of Gas
design and operating engineers. The model presented in this and Condensate Flow in Pipelines, PhD Thesis, The
paper can predict the pressure drop, liquid holdup, and liquid Pennsylvania State University, University Park, Pennsylvania,
dropout, for a multi-componenent wet gas being transported in 1990.
Peneloux, A., Rauzy, E. and Freze, R.: “Consistent Correction for
a hilly terrain at non-isothermal conditions.
Redlich-Kwong-Soave Volumes,” Fluid Phase Equilibria, v.
It has also been demonstrated that this model can 8(1), 1982.
determine the temperature profile along a multi-elevation Peng, D. and Robinson, D.B.: “A New Two-Constant Equation of
pipeline by coupling the energy equation to the mass and State,” Industrial Engineering Chemical Fundamentals, v.
momentum equations. It is observed that regardless of the 15(1), 1976.
type of terrain, the temperature of the flowing mixture drops to Vincent, P.A: One-Dimensional Compositional Modeling of Gas
the surrounding temperature value within 3.3 miles from the Condensate Flow in Pipelines, MS Thesis, The Pennsylvania
inlet. State University, University Park, Pennsylvania, 1988.
It is also observed that after this point, the gravity Taitel, Y. and Dukler, A.E.: “ A model for Predicting Flow Regime
Transitions in Horizontal and Near Horizontal Gas Liquid
effect dominates. Therefore the temperature increases when
Flow,” AIChE Journal, v. 22(8), 1976
mixture is flowing downhill and decreases otherwise. This
temperature increase was about 5 degrees Fahrenheit. This
rise is attributed to the conversion of excess potential energy
to internal energy.
SPE 97978 7

120 1500

f x
mg f mgx+ ∆x 110 1490

100 1480

Temperature [F]

Pressure [Psia]
90 1470
M gl
80 1460

70 1450

60 1440

x x + ∆x 50 1430
Figure 1: Schematic depiction of the mass transfer 0 20000 40000 60000 80000

mechanism and its dependence on the phasial mass Distance [ft]

fraction. Figure 4: Temperature and Pressure Profiles for Flat
Terrain Pipeline

Ag 120 10000

Al 110
8000

100

Temperature [F]

Elevation [ft]
6000
90
Figure 2: Conceptual representation of the equivalent
diameter in the two-fluid model.
80
4000

70

2000
60

50 0
0 20000 40000 60000 80000

Distance [ft]
Figure 5: Effect of Undulating Terrain on Temperature
Profile

1520 10000

1500
8000
1480
Pressure [Psia]

1460

Elevation [ft]
6000

1440

4000
1420

Figure 3: Schematic of the conceptualization of 1400

2000
stratified smooth flow regime. Note the definition of the
1380
wetting angle.
1360 0
0 20000 40000 60000 80000

Distance [ft]
Figure 6: Pressure Profile for Undulating Terrain
8 SPE 97978

0.06 10000

0.05
8000

0.04

Elevation [ft]
6000
Holdup

0.03

4000
0.02

2000
0.01

0.00 0
0 20000 40000 60000 80000

Distance [ft]
Figure 7: Liquid Holdup Profile for Undulating Terrain

2000 10000

1500 8000
Liquid Dropout [Gallons]

Elevation [ft]

1000 6000

500 4000

0 2000

0
0 20000 40000 60000 80000

Distance [ft]
Figure 8: Liquid Dropout for Undulating Terrain

Table 1: Wetted wall area for different flow regimes

Flow regime Awl Awg

4 4
Mist Awl = α l Awl = α l
d d
4
Annular Mist Awl = 0
d
⎛θ ⎞ 4 ⎛ π −θ ⎞4
Stratified Wavy Awl = ⎜ ⎟ Awg = ⎜ ⎟
⎝π ⎠ d ⎝ π ⎠d
⎛θ ⎞ 4 ⎛ π −θ ⎞4
Stratified Smooth Awl = ⎜ ⎟ Awg = ⎜ ⎟
⎝π ⎠ d ⎝ π ⎠d
SPE 97978 9

Table 2: Hydraulic diameter for different flow regimes

Flow regime d hl d hg
Mist d hl = d d hg = d
Annular Mist d hl = α l d N/A

⎛α π ⎞ ⎛ πα g ⎞
d hl = ⎜ l ⎟d d hg = ⎜ d
Stratified Wavy
⎝ θ ⎠ ⎜ (π − θ ) + sin θ ⎟⎟
⎝ ⎠
⎛ πα g sin θi ⎞
⎛α π ⎞ d hg = ⎜ ⎟⎟ d
Stratified Smooth d hl = ⎜ l ⎟d ⎜ (π − θ ) sin θ + θ sin θ
⎝ θ ⎠ ⎝ i i i ⎠

Table 3: Gas/liquid interface area for different flow regimes

Flow regime Agl

3α l
Mist Agl =
rP

Annular Mist Agl = ( αc ) d4

⎛ θ sin θ ⎞ 4
Stratified Wavy Agl = ⎜ i ⎟
⎝ π sin θi ⎠ d
⎛ sin θ ⎞ 4
Stratified Smooth Agl = ⎜ ⎟
⎝ π ⎠d

Table 4: Interface friction factor for different flow regimes

Flow regime f gl
1
Mist, [Cliff et. al., 1978] f gl = Cd
4
Annular Mist, [Meng, 1999] f gl = 0.0089
⎡ ⎤
0.20
⎛ α l AP ⎞
⎢1.0 + 3.17 ⎜⎝ A ⎟⎠ ⎥
Stratified Wavy, [Chen, 1997] ⎢ wl
⎥
⎢⎧ ⎥
f =
gl 0.08

⎫
0.5

⎢⎨⎜ ⎛ ⎞
f ⎥
ρ ρ v
− 8.165 ⎬
l g l

⎢⎣⎩⎝ γµ l ( ρ l − ρ g ) g ⎠⎟ ⎭
wl
⎦⎥
Use Chen’s equation [10b] and
Stratified Smooth, [Baker et. al.,
Bakers equation for equivalent
1988]
interface roughness.
10 SPE 97978

Table 5: B2 parameter proposed by Gmehling and Wang for correcting PREOS

Component B2
Carbon Dioxide 0.4161
Nitrogen 1.2949
Hydrogen Sulfide 1.8341
Water 1.2460
nC1 – nC6 B2 = 1.1811 + 0.87677ω
nC7 – nC18 B2 = −0.21224 + 9.0294ω
−15.881ω 2 + 8.0731ω 3

Table 6: Chemical composition of the wet natural gas used in current study

Compound Mole Percent

N2 0.070
CO2 2.497
H2O 0.070
CH4 85.088
C2H6 5.736
C3H8 2.657
i-C4H10 0.491
n-C4H10 1.053
i-C5H12 0.411
n-C5H12 0.461
n-C6H14 0.411
n-C7H16 0.331
n-C8H18 0.281
n-C9H20 0.201
n-C10H22 0.100
n-C11H24 0.060
n-C12H26 0.030
n-C13H28 0.020
n-C15H30 0.030

Table 7: Elevation profile for pipeline used in present study

Segment ID Length(mi) Inclination Angle (Degrees)
1 5.0 2.73
2 5.0 -2.73
3 5.0 2.73