SPE

International

Society of Petroleum Engineers

SPE 106339

Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs
A. F. Olaberinjo, University of Lagos, Nigeria.

This paper was prepared for presentation at the 2006 SPE Technical Symposium of Saudi
Characterization of gas condensate reservoirs is a difficult task, since
Arabia Section held in Dhahran, Saudi Arabia, 21-23 May 2006. multiphase flow in the reservoir and change of the mixture
Copyright 2006 Society of Petroleum Engineers composition as fluid flows towards the well complicates the
This paper was selected for presentation by the Technical Symposium Program Committee
following review of information contained in full manuscript submitted by the author(s). interpretation of well tests considerably. Analysts of gas condensate
Contents of the paper, as presented, have not been reviewed by the Society of Petroleum fields have generally assumed single-phase flow and the solution of
Engineers and are subject to 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. the diffusivity for slightly compressible fluid applied to such fields,
that is, the P method for gas well testing. The result obtained for such
simplified assumptions can be misleading in a gas condensate field
Abstract
with a low liquid drop – out.
Gas condensate reservoirs exhibit complexities due to the production The pseudo gas pressure method has also been applied for analyzing
of gas, condensates and many times, in situ oil phase. As the average condensate fields. This method could be satisfactory where the liquid
reservoir pressure in a gas condensate reservoir continues to decline drop – out is low. But in a very rich gas condensate fields where the
on production, condensate drop out occurs across the reservoir. liquid drop – out is high enough to permit the flow of condensate, the
Pressure decline between the reservoir boundaries and the well bore method could give inaccurate results. This could be attributed to the
could lead to condensate formation near the well bore. fact that the pseudo gas method assumes single-phase flow of gas.
The effect of oil phase is neglected and the relative permeability to
This paper considers the flow behavior in gas condensate reservoir
gas is taken as one. Mass transfer effects and the compositional
and presents an approach for calculating the pressure depletion
changes in the system are neglected. These assumptions can affect
performance of a gas condensate reservoir from the original reservoir
the accuracy of the results obtained and thus the total effective
fluid composition using the vapor – liquid equilibrium calculations,
management of the reservoir.
the fluid properties, dimensionless properties, and retrograde liquid
accumulation as guides. This work considers the flow behavior in gas condensate reservoir
with special consideration to drawdown behavior as presented by
The results correspond to drawdown analysis of a condensate
Fevang and Whitson1. Phase saturations are obtained by flashing the
reservoir where flow of the phase occurs. It is also a compositional
reservoir fluid at given pressure value. The equilibrium constants are
analysis and includes the composition of the hydrocarbon fluid up to
computed using the vapor – liquid equilibrium constants for natural
and including C7+. The presence of non-hydrocarbon impurities such
gas and billing crude oil at temperature values of 120oF and 200oF for
as Carbon dioxide, Nitrogen and Hydrogen Sulphide are also
pressure range of 1000 psia to 10000 psia. Pressure and Temperature
considered. The comparison with other standard work on transient
outside this range were approximated.
pressure analysis from gas condensate reservoirs showed that the
approach is satisfactory. Liquid phase molal volume, isothermal compressibilities and density
were calculated using Alani and Kennedy2 and Harvey Kennedy et
A principal use of the method would be for the initial estimation of al.3 correlations for hydrocarbon mixtures and modified to include
pressure depletion pattern after discovery of a gas condensate non – hydrocarbons. Also the liquid phase viscosity is calculated
reservoir, where the compositional analysis of the reservoir fluid may using Lorenz and Bray4 concept.
be only approximate and the dew point if not known, is assumed to
be the reservoir temperature and pressure. It is also a valuable tool Gas compressibility factor is calculated using Hall – Yarborough5
during the production period for reservoir surveillance and equations developed using the Starling – Carnaham equation of state
monitoring. while the isothermal compressibility of gases is calculated by means
of an eight-coefficient Benedict – Webb – Rubin type equation of
Introduction state6. Gas phase viscosity is calculated using Lee et al7 approach.
The work employs Kay’s mixing rule6 to determine the molecular
Gas condensate reservoirs differ essentially in their behavior from weight of liquid and gas.
conventional gas reservoirs, and the optimization of the recovery Relative Permeability data were computed using Naar and
needs careful reservoir analysis, planning and management. Henderson8, Jones and Raghavan9 equations and correlations for two-
phase relative permeability saturation relationship.
At the time of discovery, gas condensate reservoirs are often found
containing single-phase gas vapor. As the reservoir is being The solution of the diffusivity equation for the simultaneous flow of
produced, the pressure decreases from the reservoir to the wells and both gas and oil in a two – phase problem is for an infinite reservoir
to the surface installations, leading to condensation of liquid out of and the result obtained apply to drawdown analysis of a condensate
the gas. This isothermal condensation as the pressure drops below the reservoir where the liquid phase occurs. It is a time and pressure
dew point pressure of the original fluid is known as retrograde conscious compositional analysis and includes the composition of
condensation. hydrocarbon fluid up to and including C7+ with bias for the presence
of impurities – CO2, H2S and N2.
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE 106339

This paper aims at understanding the flow behavior in gas condensate Hawthorn16 carried out additional work on two-phase flow model.
reservoirs and, in particular, the entire phenomenon that contribute to He based his solution of flow on Dietz model where viscous and
the optimization of gas condensate wells. It also presents a systematic gravitational displacements in a two-phase flow model were
method for calculating pressure depletion performance of gas considered. In each of these cases, numerical approach was used to
condensate reservoir with special considerations to the effect of arrive at the final result. The model considers simultaneous flow
compositional changes on pressure during production. The neglecting the effect of mass transfer, properties of the liquid and
interpretation techniques require only the measurements of vapor phases and compositional changes.
dimensionless time, dimensionless pseudopressure, bottom hole
flowing pressure and liquid-vapor phase properties. Stone and Gardner17 studied the effect of two-phase flow in gas – cap
drive, dissolved gas drive and pressure maintenance by gas injection.
Literature Review Properties of the fluid were considered as a function of pressure only.
Relative permeability relationship was used to define flow properties
Reservoir engineers have a responsibility to ensure optimum of the rock. He assumed equilibrium between the flowing phases and
production for each reservoir. For a well to be effectively tested and considered the effect of phase changes during decline of reservoir
all parameters that control production fully utilized, the relationship pressure.
between the flow parameters has to be established. Multiphase flow
and mixture composition change in the reservoir make the Attra18 considered the effect of mass transfer on two-phase flow. His
interpretation of well tests in gas condensate reservoirs a real model was based on the effect of mass transfer on gas displacement
challenge. Relationship between reservoir pressure, flow rate and the processes in a one-dimensional model using vaporization and
flowing time are of significant importance in the determination of the equilibrium constant and included redistribution of composition of
flow capacity and efficiency of the formation. liquid and gas saturation.

Relationships for single-phase flowing wells have been obtained Welge et al.19 developed an equilibrium displacement process in a
analytically for both oil and gas reservoirs. With reservoirs that depict reservoir produced by gas drive mechanism. Recovery of liquid phase
two phase flow; the solution of the diffusivity equation has been was enhanced by re - vaporization and condensation in the reservoir.
intractable especially by analytical means. It even becomes more The method was especially applicable to cases where injected gas is
difficult to solve when mass transfer and compositional changes are partially miscible with liquid.
taken into consideration. Till recently, only few literature have A two-phase model for light and volatile hydrocarbon was developed
advanced numerical solutions to two-phase flow problems and fewer by Naville and Kniazeff20. They assumed constant composition under
still have solved analytically two-phase flow problems. differential liberation with only one phase flowing in the reservoir.
Raimondi and Torcaso21 studied a similar process studied by Naville
In early development of reservoir models for two phase flow; and Kniazeff20 and also assumed single-phase flow. They reported
Schilitus10 presented an equation that calculates reservoir changes that miscibility is attained if equilibrium exists between the phases.
that occur during oil production where gas is a second phase present. They saturated a media with two phases (carrier and stationary) in the
This was a simple material balance equation that involved oil and gas laboratory and also reported that existence of equilibrium also
accounting. He assumed complete equilibrium with the initial gas in depends on relative velocity of the two phases.
place being only free gas. Reservoir production was assumed to be at
atmospheric pressure at constant volume. Eilert et al.22, 23 developed the solution to a multi-component partial
differential equation for both the radial and linear geometries of a gas
Later, Buckley – Leverett11 included the effect of pressure and condensate reservoir where he reported condensate build up high
saturation gradients in a one-dimensional model. He assumed viscous enough to permit flow around the well bore. Liquid phase
flow and solved a steady state equation where flow rates were high. composition was determined by distillation analysis and liquid vapor
He also assumed parallel and simultaneous flow of reservoir fluid. volume ratios calculated at different pressures. Results were
presented for different saturations and pressure profiles.
West et al.12 developed a solution for an unsteady state two-phase Roebuck et al.24 used compositional representation of the
model to predict behavior of an oil and gas reservoir. Numerical hydrocarbon system for a one-dimensional flow study of a reservoir.
technique was employed to solve both the radial and linear reservoir He considered the effect of mass transfer and changing composition
geometry. The results consider pressure and saturation changes. in the numerical model.
Gravity effects were neglected.
An analytical approach for a gas drive solution gas reservoir was
Reudelhuba and Huds13 applied Depletion analysis method to solve a presented by Raghavan25. His model assumed that flow depends only
volatile oil system and assumed that production of both phases has no on the properties of the dominating fluid present although he
influence on each other and that surface liquid is just that which is considered the effect of saturation increase of the second phase on
produced by the reservoir. This implied that liquid solubility is relative permeability of the dominating phase. He presented pseudo
neglected. Laboratory data was used to carry out depletion drive. The pressure techniques for both build up and drawdown cases and
data permit continuous material accounting and provides hydrocarbon pointed out differences in both approaches.
compositional model to calculate pressure depletion performance of a
Understanding the multiphase flow in condensate reservoirs is vital in
gas condensate field. Laboratory technique was used to determine
characterizing the condensate dropout and subsequent blockage
pressure production history and retrograde liquid accumulation. The
effect. Saeidi and Handy26 studied the flow and phase behavior of gas
procedure showed that it is only adequate for gas phase composition
condensate (methane – propane) in porous medial (sandstone core).
determination.
They indicated that interstitial water shifts the oil relative
permeability to an appreciably lower saturation. In addition, no flow
Douglas et al.14 considered immiscible displacement with cross flow
of condensate was observed for this system even with an 18%
between layers of different permeability by analytic and numerical
volumetric dropout in the presence of 30% interstitial water
methods respectively. Martins15 presented a two-phase model for oil
saturation.
and gas flow and replaced single-phase mobility and compressibility
by multiphase mobility and compressibility. He assumed small Asar and Handy27 investigated the influence of interfacial tension on
pressure drop and applied a partial linearization to arrive at a relative permeability of gas/oil in a gas condensate system. They
solution. postulated that the irreducible gas and liquid saturations approach
zero as interfacial tension approaches zero. In addition, they observed

ii
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

that condensate could flow at a low condensate saturation (Scc = strongly nonlinear partial differential equation. Hence, Equation 1
10%). Finally, it was concluded that liquid could flow at a very low may not be valid to represent the flow of such fluids.
liquid saturation at low interfacial tensions in a condensate reservoir.
This is significant as regions with two-phase (gas and liquid) The continuity equation for component i in a gas condensate system
conditions have low interfacial tension. can be presented as:
1 ∂
Hinchman and Barree28 studied the effect of the fluid characteristics
r ∂r
{r (ρ oU o x i + ρ g U g y i )} = φ ∂∂t {ρ o S o x i + ρ g S g y i } . . . . . . . . . . . .. .2
in the predicted productivity decline of a gas condensate well. They
demonstrated that the amount of gas condensate accumulation near ∀1 ≤ i ≤ ni
the wellbore depends greatly on the richness of the gas condensate,
the relative permeability data and the liquid viscosity. The Darcy’s equation that describes the flow of oil and gas phase
Sognesand29 discussed the condensate build up in vertical fractured when each phase is flowing at constant rates Ug, Uo are given by:
gas condensate wells. He showed that the condensate build up K g dP ........................3
depends on the relative permeability characteristics and production U = −
µ
g
g dr
mode, increased permeability to gas yields reduced amount of
condensate accumulation, and constant pressure production yields the K dP . . .. . . . . . . . . . . . . … . . . . . 4
g
largest near fracture condensate buildup. He presented a method to U o = − f
µ dr
correct the effect of condensate blockage using the concept of time – g

dependent skin factor. Where f is the flowing liquid vapor volume ratio or the mobility of
The productivity loss associated with condensate build up can be the flowing phases defined as:
substantial. Afidick et al.30 accounted for several instances in which K µ
well productivities have been reported to decline by a factor of two to f =
o g . . . . . . . . . . . . . . . . . . . . . .. .5
four as a result of condensate accumulation. µ o K g

Fevang1 and Ali et al.31 showed that, when reservoir pressure around Substituting Equations 3 and 4 into 2:
a well drops below the dew point pressure, retrograde condensation
occurs and three regions are created with different liquid saturations. 1 ∂    ρo Ko xi dP   ρg Kg yi dP   ∂ …………….6
r  +  = φ {ρoSo xi + ρg Sg yi }
According to Economides et al.32 and Fussel33 there may also exist a r ∂r    µodr   µg dr   ∂t
 
fourth region in the immediate vicinity of the well where low
interfacial tensions (IFT) at high rates yield a decrease of the liquid From the definition of compressibility:
saturation and an increase of the gas relative permeability.
1 ∂ ρ
Bennion et al.34 worked on optimizing production from a gas c =
ρ ∂ P
condensate reservoir. Their work seeks to describe some of the
phenomena that are at work in rich gas condensate reservoirs. In this
context, specific parameters such as interfacial tension, mobility ∂ ρ
effects, pore size distribution and compositional changes all = c ρ . . . . . . . . . . . . . . … . . . . . . . . .7
contribute to the optimization of gas condensate wells. ∂ P

Cho et al.35 presented a correlation to predict maximum condensation Introducing mass transfer effect in terms of mole concept, the molar
for retrograde condensation fluids and its use in pressure depletion density will be given by:
calculations. The correlation presented is a function of the reservoir ρ1 ρ2 ρn ρ
temperature and the heptanes plus mole fraction. + +L+ =
m1 m2 mn m
Jatmiko et al36. described a method of analyzing wellbore pressure
response of a gas condensate reservoir under transient conditions. Therefore,
The heavier component, C7+ is splitted into three groups and the
ρ 1 c1 ρ 2c2 ρ ncn ρc .....................8
pseudopressure function is calculated based on the saturation + +L + =
m1 m2 mn m
distribution.
Despite a large number of reported studies on gas condensate Equation 6 now becomes:
reservoirs 37, 38 & 39 in addition to those cited here, none of them
indicated a reasonably systematic and economical approach for 1 ∂   ρo Ko xi dP   ρ g K g yi dP  ∂P  ρoco So xi ρ g cg S g yi  ……..9
r  + =φ  + 
calculating pressure depletion performance of gas condensate r ∂r   mo µo dr   mg µ g dr  ∂t  mo mg 
reservoirs with consideration to the effects of compositional changes
and mass transfer during production. But from equation 5:
Model Development K K g
o
= f
µ µ
All direct interpretation of pressure transient from a well test is based o g

on the Horne40 linear diffusion equation: Therefore,

φµ c t ∂ P …………………………………..1
∇ P− =0 1 ∂  K g  fρ o xi dP ρ g yi dP 
2
 ρ cS y
K ∂t r  +  = φ ∂P  ρ o co So xi + g g g i  …….10
r ∂r  µ g  m dr m dr  ∂t  m mg 
 o g   o
Equation 1 applies essentially to a single – phase slightly
compressible oil reservoir (i. e. with Pr > Pbubble). Introducing equation of state for gases and the relative permeability
relationship:
In gas condensate reservoirs, retrograde liquid condensate, and gas
may be present together and multiphase flow may also occur. The PM g
ρg = . . . . . . . . . . . . . . . . . 11
flow of such fluids in porous media can only be represented by a
ZRT

iii
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

K g retrograde condensation and compositional changes on the pressure

K =
rg
K transient in the gas condensate reservoir, equilibrium flash
vaporization calculations were carried out on the hydrocarbon
Kg = KKrg . . . . . . . . . . . . . . . . . . . .12 components at the initial reservoir pressure and temperature and also
at other pressures as the production continues isothermally. The flow
Equation 10 then becomes, coefficient of gas phase to oil phase for Cases 1, 2 and 3 at 4000psia
and 200oF are 0.45, 3.386 and 47.6 respectively. This conforms very
1 ∂  ∂P KKrg  fρo xi Pyi  ∂P  ρocoSo xi cg Sg Pyi  . . ……..13 well with the nature of the feed composition in each category. Also,
r  +  = φ  + 
r ∂r  ∂r µg  mo ZRT ∂t  mo ZRT  the variation of pressure with flowing liquid – vapor phase is as
shown in APPENDIX A, Figure A1.
Equation 13 is a strongly non-linear diffusivity equation for gas
The predicted liquid molal volume, isothermal compressibility and
condensate with component i in a two-phase flow reservoir in the
the liquid phase density were calculated based on the liquid phase
radial geometry derived under the assumptions that the initial
composition in the overall mixture at different production intervals.
reservoir pressure is equal to the dew point pressure, well produces at
From a practical point of view, the liquid phase viscosity, molal
constant rate (that is, flow rate of each component in a particular
density and volume affect the pressure. The liquid phase fraction
phase is constant) and fully penetrate the formation, gravitational
determines the response of a gas condensate system. Figures A2 and
effects and capillary effects are ignored and the reservoir is
A3 shows the variation of pressure with the liquid phase molal
homogeneous.
volume and compressibility factor respectively. The liquid phase as a
A classical method to handle Equation 13 is to define a variable m(P) function of pressure is measured and it determines the degree of
named pseudopressure and pseudo compressibility, cgascon (P): reduction in gas relative permeability.
TABLE 1: Gas Condensate Feed Composition
Pwf  fρ x Py  KKrg 
mgascon(P) = ∫  o i + i  dP MOLE % - MOLE % - MOLE %
Pdew
 mo ZRT µg  COMPONENT CASE1 CASE 2 - CASE 3

m (P )= 2 RT (m gascon
(P )) C1 58.77 73.190 88.060
C2 7.57 7.800 3.770
⇒ C3 4.09 3.550 1.970
 fρ o x i C4 3.0 2.160 1.170
Py i  KK rg 
m(P ) = 2 RT ∫
Pwf

Pdew  m
+  dP C5 1.92 1.320 0.530
 o ZRT  µ g  C6 1.75 1.090 0.320
 2 fRT ρ o x i 2 Py i  KK rg  C7+ 21.76 8.210 2.910
m (P ) =
Pwf
∫Pdew

 mo
+
Z

 µg
 dP

N2 0.21 0.310 0.040
CO2 0.93 2.370 1.230
⇒ H2S 0.00 0.000 0.000
TOTAL MOLE 100.00 100.000 100.000
 2 Pyi 2 fRTρ o xi  KK rg  ……………..14
dm(P ) =  +  dP
 Z mo  µ g 
The gas deviation factor, viscosity and the formation volume factor
Equation (13) becomes: (Bg) are functions of pressure only in situ, although the deviations are
usually small. The leaner the retrograde gas, the smaller the
−1
1 ∂  ∂ m (P )  φ ∂ m (P )  ρ o S o co xi c g S g Py i    Py i f ρ o xi  KK rg  deviations. Only the sensitivity of gas relative permeability was
r =  +  ∗   +  
r ∂r  ∂r  ∂ t  m o ZRT 
   ZRT mo  µg


tested. The deliverability loss due to condensate blockage is dictated
by the near well bore region contribution to the pseudopressure
Also defining pseudo compressibility, cgascon(P): integral. The contribution is solved by finding the relationship
between Krg and the ratio Krg / Kro.
 ρ oco S o xi c g S g Py i 
 +  Basically, on a semi-log plot of dimensionless pressure, PD and
mo ZRT . . . . . . . . . . . . 15
c gascon (P ) =   dimensionless time, tD a straight line represents the liquid solution.
 Py i fρ o xi  Therefore, if a pseudopressure is the right liquid analog, its semi log
 +  representation of dimensionless pseudopressure, PD and
 ZRT mo 
dimensionless time, tD should give exactly the same straight line
Substituting Equations (14) and (15) into (13) gives the equation (Drawdown Test). This is evident in Figure A4 where the pattern of
governing pressure transmission in gas condensate reservoir with bias the plot agrees well with the work of Xu and Lee37 where laboratory
for compositional changes and mass transfer effect as: constant volume depletion (CVD) data, gas and condensate oil
relative permeability, and producing GOR were used to interpret
1 ∂  ∂ m (P )  µ g c gascon (P ) ∂ m (P ) . . . . . . . . . . . . . 16 transient pressure data from gas condensate reservoirs. The flow
r  =φ capacity, kh (md-ft) can be estimated from the pseudopressure
r ∂r  ∂r  KK rg ∂t
method using Equation 18.
The solution method for this non-linear diffusivity equation is
kh(True ) . . . . . . . . . . . . . . . . . . . . 18
expressly analysed in Appendix B. kh(Estimated ) =
dPD
Analysis of Results dt D

Table 1 shows the condensate feed compositions for the three cases dPD = slope of the Semi log Plot of PD and tD
considered – Case 1: Gas Condensate with very high liquid dropout, dt D
Case 2: Gas Condensate with low liquid dropout and Case 3: Gas
Condensate with very low liquid dropout. To show the effect of During the early producing time of practical interest the pressure
behavior can be described as transient. As time becomes larger the

iv
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

pressure decline throughout the reservoir becomes a linear function of can be represented by the solution of the slightly
time. When this flow regime occurs, it is commonly spoken of as compressible fluid flow.
semi-steady state or pseudo – steady state behavior. 7. The approach works best when pressure differences Paverage
The high liquid dropout effect results in a bigger pressure drop with a – Pdew and Pdew – Pwf are large. It is also applicable to oil
pressure decline rate that varies considerably for Cases 1, 2 and 3. and gas reservoirs with compositions.
For gas condensate with low dropout, the change in pressure
difference with time is not significant since the flow coefficient of Recommendation:
gas phase to oil phase and relative permeability to gas is considerably
high. This is unlike the case rich gas condensate where retrograde A principal use of the model would be for the initial estimation of
condensations impair the relative permeability to gas. pressure depletion pattern after discovery of a gas condensate
reservoir, where the compositional analysis of the reservoir fluid may
Conclusively, a new method for calculating the pressure depletion in be only approximate and the dew point if not known, is assumed to
Gas Condensate reservoirs is demonstrated. It takes into be the reservoir temperature and pressure.
consideration the compositional changes of reservoir fluid and mass
transfer effects during production. It also takes care of the impurities It is also a valuable tool during the production period for reservoir
that may be present in the mixture, the dimensionless properties, the surveillance and monitoring.
pseudopressure analysis, and the properties of the liquid and vapor
phases. Acknowledgements:
The approach shows an excellent agreement between the liquid flow
solution during transient conditions and the real gases flow solution. I am grateful to Flotto Nigeria Limited for their logistic support. I
Hence it can be used to estimate the transport properties of gas kindly thank TUTI, Ajala Deolu, Ibironke Adebola and Edu
condensate reservoirs. Temitope for stimulating discussions and insights.

Validation Nomenclature

Figures A6 to A8 presents semilog plot of the dimensionless well C = Compressibility, Psia-1

bore pseudopressure, PD versus the dimensionless time, tD. The C (P) = Pseudo Compressibility
deviation in the slope between the pseudopressure and the constant
rate liquid solution measures the error in the permeability estimation C1 = Conversion Constant, 0.00633 ft3 / ft
using the pseudopressure and the vertical shift measures the error in Ct = Total Compressibility, psi -1
the skin estimation.
f = Liquid Vapor Volume Ratio
Conclusion
h = Reservoir Thickness, ft
1. A quick relatively inexpensive and approximate
compositional approach for calculating the pressure K = Reservoir Absolute Permeability, md
transient flow of gas condensate systems has been
presented. The model could be used to interpret results of M = Molecular Weight, lbm / lbmol
any two-phase flow reservoir tested especially for a gas M (P) = Pseudo Pressure Function
condensate field or any gas oil flow system.
2. The overall composition of the flowing mixture at any Pi = Initial Reservoir Pressure, psia
location in the reservoir changes as the production Pwf = Bottom Hole Flowing Pressure, psia
continues and hence it is a function of time and pressure.
The vapor composition and the liquid composition, along Uo = Velocity of Condensate
with the mole fraction of liquid and vapor can be calculated
Ug = Velocity of Gas
directly from vapor – liquid equilibrium calculations. The
model provides a basis for understanding the mechanisms qg = Gas Flow Rate, MMscf / D
and factors, which control gas condensate well production
during depletion. qo = Oil Flow Rate, STB / D
3. The fluid system considered was limited to multi- R = Gas Constant, 10.73 (Psia ft3/mol R)
component systems; however, the correlations used to
calculate fluid properties may be used to account for the r = Radius, ft
presence of impurities in the systems. S = Fluid Saturation, Fraction
4. During reservoir depletion the vapor z – factor, molar
density, and viscosity are functions of pressure and vapor t = Time, hours or days
phase compositions. Also, during reservoir depletion the
T = Formation Temperature oF
condensate molal volume, density, and viscosity are
functions of pressure and liquid phase compositions. z = Gas Compressibility Factor
5. The most critical region affecting productivity is
immediately around the well bore. Near wellbore ρ = Density, lbm / ft3
condensate accumulation causes productivity loss that µ = Viscosity, cp
affects the production and pressure transient response.
6. The pressure distribution in the gas condensate reservoir φ = Porosity
expressed in terms of pseudopressure satisfies the classical
liquid equation for steady state flow. The interpretation of St = Total Skin
the pseudopressure leads to correct estimation of the flow xi = Liquid Composition
capacity, kh and skin effect. It is also useful for well
deliverability calculations and sensitivity analysis. Also, the yi = Vapor Composition
correlation between dimensionless pseudopressure and time Krg = Gas Relative Permeability

v
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

Kro = Oil Relative Permeability Radial Flow of Gas Condensate Fluid in Porous Structure.
L = Mole Fraction of Liquid Trans Aime 234, II 141 – 152.
23. Eilerts, C. K., Sumner, E. F. (1965): Integration of Partial
V = Mole Fraction of Vapor Differential Equations for Multi - -Component, Two Phase
References Transient Radial Flow. Trans Aime 240, II 125 – 135.
24. Roebuck, I. F., Henderson, G. E., Watford, J. D. (1969):
1. Fevang, and Whitson, C. H., (1995): Modelling Gas The Compositional Reservoir Simulator Case I. – The
Condensate Deliverability, SPE paper 30714. Linear Model. Trans Aime 246, II 115 – 130.
2. Alani, G. H., Kennedy, H. T. (1960): Volumes of Liquid 25. Raghavan, R. (1976): Well Test Analysis: Wells Producing
Hydrocarbon at High Temperatures and Pressure. Trans by Solution Gas Drive. Trans Aime 261, II 196 – 208.
Aime 219, 288 – 292. 26. Saeidi, A. and Handy, L. L.: (1974), Flow and Phase
3. Harvey Kennedy et al. (1968): the Prediction of Volume, behavior of gas condensate and volatile oils in porous
Compressibilities and Thermal Expansion Coefficients of media, SPE Paper 4891.
Hydrocarbon Mixtures SPEJ 95 – 106 27. Asar, H. and Handy, L. L. (1988), Influence of Interfacial
4. Lohrenz, J., Bray, B. G. (1965): Calculating Viscosities of Tension on gas – oil relative permeability in a gas
Reservoir Fluids from their Compositions. Trans Aime 234, condensate system, SPE Paper 11740.
1171 – 1176. 28. Hinchman, S. B. and Barree, R. D.: (1985), Productivity
5. Hall, K. R., Yarborough, L. (1974): How to Solve Equation loss in gas condensate reservoirs, SPE Paper 14203.
of State for Z – Factor. Oil and Gas Journal. Feb. 18th, pp 29. Sognesand, S.: (1991), Long – term Testing of Vertically
86 – 88. Fractured Gas Condensate Wells, SPE Paper 21704.
6. Chi U. Ikoku (1984): Natural Gas Production Engineering – 30. Afidick, D., Kaezorowski, N. J., Srivinas, B.: (1994),
John Wiley and Sons. USA. Production Performance of Retrograde Gas Reservoir: A
7. Lee, A. L. Gonzalez, M. H., Eakin, B. E. (1966): The case study of the Arun Field, SPE Paper 28749
Viscosity of Natural Gases. JPT Aug. 997 – 1000. 31. Ali, J. K., McGauley, P. J. and Wilson, C. J.: (1997),
8. Naar, J., Henderson, J. H. (1961): An Imbition Model – Its Experimental Studies and Modelling of Gas Condensate
Application to Flow Behavior and the Prediction of Oil Flow near the Wellbore, SPE Paper 39053
Recovery. SPEJ, 61. 32. Economides, M. J., Schlumberger, D., Deghani, K., Ogbe,
9. Raghavan R., Chu W. C. and Jones J. R. (1995): Practical D. O. and Ostermann, R. D.: (1987) Hysteresis Effect fo
Considerations in the Analysis of Gas Condensate Well Gas Condensate Wells undergoing Buildup Tests below the
Tests, SPE Paper 30576. Dew Point Pressure, SPE Paper 16748.
10. Schilthus, R. J. (1936): Active Oil and Reservoir Energy. 33. Fussel, D. D.: (1973) Single Well Performance for Gas
Trans Aime 118 33 – 52 Condensate Reservoirs, JPT 255, 860 – 870.
11. Buckley, S. E., Leverett, M. C. (1942): Mechanism of Fluid 34. Bennion, D. B. et al. ‘ Optimizing Production from Gas
Displacement in Sands. Trans Aime 146, 107. Condensate Reservoir’, JCPT Oct., 1997, Vol. 36, No. 9
12. West, W. J., Garvin, W. W., Sheldon, J.W. (1954): Solution 35. Cho S. J., Civan F. and Starling K. E. (1985): A Correlation
of the Equation of Unsteady State Two – Phase Flow in Oil to Predict Maximum Condensation for Retrograde
Reservoir. Trans Aime 201, 217 – 229. Condensation Fluids and its use in Pressure Depletion
13. Reudelhuba, F. O., Huds, R. F. (1957): A Compositional Calculations, SPE Paper 14268.
Recovery Material Balance Method for Predicting of 36. W. Jatmiko, T. S. Daltaban, J. S. Archer (1997):
Recovery from Volatile Oil Depletion Drive Reservoir. Multiphase Transient Well Testing for Gas Condensate
Trans Aime 210 19 – 26. Reservoirs - SPE Paper 38646.
14. Douglas J., Peaceman, D. W., Rachford, H. H. (1959): A 37. Shaosong Xu and W. John Lee (1999): Two – Phase Well
Method for Calculating Multi – Dimensional Immiscible Test Analysis of Gas Condensate Reservoirs, SPE Paper
Displacement. Trans Aime 216, 297 – 308. 56483.
15. Martins, J. C. (1959): Simplified Equation of Flow by Gas 38. Sheng – Tai Lee and Marcial Chaverra (1998): Modelling
Drive Reservoirs and the Theoretical Foundation of Multi – and Interpretation of Condensate Banking for the Near
Phase Pressure Build – up Analysis. Trans Aime 216, Critical Cupiagua Field. SPE Paper 49265.
309 – 311. 39. Robert Mott, Andrew Cable and Mike Spearing (2000):
16. Hawthorn, R. G. (1960): Two Phase Flow in Two Measurement and Simulation of Inertial and High Capillary
Dimensional Systems – Effects of Rate, Viscosity and Number Flow Phenomena in Gas – Condensate Relative
Density in Displacement in Porous Media. Trans Aime Permeability, SPE Paper 62932.
219,81-87. 40. Horne, R. N. “Modern Well Test Analysis”, Petroway Inc.
17. Stone, H. L., Gardner, A.O. (1961): Analysis of Gas – Cap Palo Alto CA 1990.
or Dissolved Gas Drive Reservoirs. Trans Aime 222, II
92 – 102.
18. Attra, H. D. (1961): Non-Equilibrium Gas Displacement
Calculation. Trans Aime 222, II 130 – 136
19. Welge, G. J., Johnson, E. F., Ewing, S. P., Bimkman, F. H.
(1961): The Linear Displacement of Oil from Porous Media
by Enriched Gas. Trans Aime 222, I 787 – 796.
20. Kniazeff, V. J., Naville, S. A. (1965): Two Phase Flow of
Volatile Hydrocarbon. Trans Aime 234, II 51 – 59
21. Raimondi, P. Torcaso, M. A. (1965): Mass Transfer
between Phases on a Porous Media. Trans Aime 234, II
51 – 59.
22. Eilerts, C. K., Sumner, E. F., Polts, N. L. (1965):
Integration of Partial Differential Equations for Transient

vi
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

Appendix A:

FIGURE A1: VARIATION OF PRESSURE WITH FLOWING LIQUID - VAPOR VOLUME RATIO, f FIGURE A5: CONSTANT VOLUME DEPLETION CALCULATION FOR
0.5
RETROGRADE CONDENSATE FLUID
5.5
Case 1 New Approach
Case 2 Cho et al.
5
0.45 Case 3 Firoozabadi et al.
Poly. (Case 1) Poly. (New Approach)
Poly. (Case 2) 4.5
Poly. (Firoozabadi et al.)
Poly. (Case 3) Poly. (Cho et al.)

CONDENSATE LIQUID VOLUM

0.4 4

3.5
0.35
f

0.3 2.5

2
0.25
1.5

0.2 1
0 500 1000 1500 2000 2500 3000 3500 4000 4500 500 1000 1500 2000 2500 3000 3500 4000 4500

PRESSURE, PSIA PRESSURE, psia

FIGURE A2: VARIATION OF LIQUID PHASE MOLAL VOLUME WITH PRESSURE

4.5 FIGURE A6: SEMILOG PLOT ANALYSIS: DIMENSIONLESS PSEUDOPRESSURE, PD VERSUS
Case 1 DIMENSIONLESS TIME, tD FOR TOTAL FLOWRATE = 50000lbm/d and K = 100md
Case 2 10
Case 3
LIQUID PHASE MOLAL VOLUME, CUFT/LBMO

4 Linear (Case 3) 9

Linear (Case 2)

DIMENSIONLESS PSEUDOPRESSURE, P
8
Linear (Case 1)
3.5 7

3
5

4
Constant
2.5
3 Rate Liquid
Solution &
2 Jatmiko et
al.
2 New
1
Approach

0
1.5 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08

0 500 1000 1500 2000 2500 3000 3500 4000 4500 DIMENSIONLESS TIME, tD

PRESSURE,PSIA

FIGURE A3: VARIATION OF COMPRESSIBILITY FACTOR WITH PRESSURE FIGURE A7: SEMILOG PLOT ANALYSIS: DIMENSIONLESS PSEUDOPRESSURE, PD VERSUS
DIMENSIONLESS, TIME, tD FOR TOTAL FLOWRATE = 10000lbm/d and K = 100md
0.92 12

0.91

10
DIMENSIONLESS PSEUDOPRESSURE, P

0.9
COMPRESSIBILITY FACTOR, z

0.89 8

0.88
6

0.87

0.86 Case 1 4

Case 2
0.85 Case 3 Constant Rate Liquid Solution &
2 Jatmiko et al.
Poly. (Case 3)
New Approach
0.84 Poly. (Case 1)
Poly. (Case 2) 0
0.83
1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08
0 500 1000 1500 2000 2500 3000 3500 4000 4500
DIMENSIONLESS TIME, tD
PRESSURE, PSIA

FIGURE A4: SEMILOG ANALYSIS: PRESSURE DRAWDOWN, PDD VERSUS FLOWING TIME, t FIGURE A8: SEMILOG PLOT ANALYSIS: DIMENSIONLESS PSEUDOPRESSURE, PD VERSUS
250 DIMENSIONLESS TIME, tD FOR TOTAL FLOWRATE = 2000lbm/d and K = 10md
9

8
200
DIMENSIONLESS PSEUDOPRESSURE, P

Case 1
DRAWDOWN, PDD(PSIA

6
150
Case 2
5
Case 3
4
Log. (Case 1)
100

Log. (Case 2 ) 3

Constant Rate Liquid Solution

Log. (Case 3 ) 2 & Jatmiko et al.
50

1 New Approach

0
0 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07
1 10 100
DIMENSIONLESS TIME, tD
FLOWING TIME, t(D)

vii
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

r 2
……………………….B8
Appendix B: Method of Solution x = D

4 t D

Equation 16 can be solved by applying different boundary conditions: Equation B5 transformed to

1. Infinite Reservoir
∂ 2m(PD )  1  ∂m(PD )  ∂m(PD ) ……………….B9
2. Bounded Circular Reservoir. No flow outer boundary ∂x2
+  + 1  = α ∂x = 0
 x  ∂x 
3. Bounded Circular reservoir. At initial conditions when µ cgascon (P) is µi cgascon (Pi), α = 0 and
The solution method will consider the solution for a constant rate in Equation B9 reduces to:
an infinite reservoir ONLY. ∂ 2 m(PD )  1  ∂m(PD )
+  + 1 =0 ………………………….B10
Constant Rate at Wellbore – Infinite Reservoir ∂x 2  x  ∂x

1 ∂  ∂m(P )  µgc (P ) ∂m(P )

gascon
…………………..B1 Comparing equations B9 and B10 shows that when α is small
r =φ
r ∂r  ∂r  KK rg ∂t compare to 1 + 1 B9 becomes a perturbation of B10.
x
Conditions:
The condition that α is small compared to 1 + 1 is very much satisfied
1. m(P) = M(Pi) at t=0 for all r > 0 x
for all times. An illustration of this is given below:
2. ∂m (P )  2 P 2 qo RT ρ O  qg µ g

…………….B2
r = +
∂r  z
 qg m o  2πKh
 At early times tD is small and x is large and 1 +1 ≅ 0 whereas
x
at r = rw ∀ t α < < 1 + 1.
µc gascon
(P ) ≅ µ i c gascon
(Pi ) and so α ≅ 0 . Therefore
x
3. d (P) → d (Pi) as r → ∞
At late times x is small and so 1 + 1 is large. α is seen to be small
In dimensionless form we define x
r and much smaller than 1 . Thus the perturbation analysis is
rD = + 1
rw x
applicable.
kt …………………B3 Equation B10 is an ordinary differential equation and can be
tD =
ϕµ g C gascon (Pi )rw2 integrated.
2πKh Let
m(PD ) = (m(Pi ) − m(P ))
qcT
∂ m (PD )
Y = ……………………..B11

(PD ) = µc gascon (P )
gascon
gascon ∂x
c
µc (Pi ) Substitute into B10 and rearranging, we have
dy
2qo ρ o RTsc ……………………….B4 x + (1 + x ) y = 0 ……………………….B12
qt = q g + dx
mo Psc
Separating variables we have:
Substituting B3 into B2 and B1 yields
dy
= −
(1 + x ) dx ………………….B13
1 ∂  ∂m(PD )  ∂m(PD ) . . . . . . . . . . . . . . . . . . . .B5 y x
 rD  = c gascon (PD )
rD ∂rD  ∂rD  ∂t D
Integrating the function,
Subject to:
lny = −lnx − x +C1 ………………….B14
m (PD) = 0 at tD = 0 for all rD
exp(− x )
1.

2.
rD
∂m(PD )
= −1
at y = C2 …………………B15
∂rD x
……………………….B6 Using the inner boundary
rD = 1 ∀ t D
1
3. m (P D ) → ∞ C2 = −
2
Define α
1
y = − exp
(− x) …………………….B16
µc gascon (P ) − µi c gascon (Pi ) 2 x
α=
µi c gascon (Pi )
Substituting the value of y into equation B16 yields
Therefore,
∂ m (P D ) 1
µc gascon (P ) …………………………B7
= exp (− x) ……………………B17
1+ α = = c gascon (PD ) ∂x 2
µi c gascon (Pi )

Using Equation B7 and the BOLTZMAN transformation

viii
Modeling the Effects of Compositional Changes in Transient Flow of Gas Condensate Reservoirs SPE SA -

1 exp (− x ) P − Pi , t = 0
∞
m (PD ) = ...………………….B18
2 ∫x
dx
x
x → ∞ and δ = 0
If the following relation is defined

exp (− x )
∞
− Ei (− x ) = ∫ dx Gas Equivalent Rate For Condensates
x
x
ρo
q CONDENSATE = q g + 2 .131 qo
Then equation B18 can be written as mo
1 ρo
m (PD ) = − E i (− x ) ...………………B19
qCONDENSATE = q g + 2.131* 62.43 qo
2 mo
Assuming the solution to equation B9 is given as
ρo
1 q CONDENSATE = q g + 133 .038 qo
m (PD ) = − Ei (− x ) + δ mo
2
The condensate flow rate is the gas equivalent for condensate flow
m(PD ) = m(PD ) + δ ..…………………B20 and the gas flow rate.

Substituting B20 into B9

∂ 2m(PD ) ∂ 2δ  1  ∂m(PD )  1  ∂δ ∂m(PD ) ∂δ ……..B21
+ 2 +  − 1 +  − 1 + α +α =0
∂x2 ∂x  x  ∂x  x  ∂x ∂x ∂x

From Equation B10

∂ 2 m (PD ) 1  ∂ m (PD )
+  + 1 = 0
∂x 2 x  ∂x
Neglecting the second order terms that is

∂ 2δ ∂δ using B10, equation B21 reduces to:

and α
∂x 2 ∂x
1  ∂δ  ∂m (PD )
 + 1   +α =0 ……………… B22
x  ∂x  ∂x

Rearranging

 ∂δ   x  ∂m(PD ) ……………………B23
  = − α
 ∂x   1 + x  ∂x
Considering Equation B19

 1 
∂ − E (− x) 1 −x
e
∂m(PD )  2 i = 2 ……………………B24
=
∂x ∂x x
1 −x −1
e α −x
∂δ −x 2 e ……………………..B25
= α = 2
∂x 1 + x x 1+ x x
Therefore,

1
x
αe−x
α =
2 ∫1+
∞ x
∂x …………………B26

Thus the solution of Equation B9 is given by

1
m (PD ) = − E i (− x ) + δ
2
Where

1
x
α e − x
δ =
2 ∫
∞
1 + x
∂ x …………….B27

Where

ix