Sie sind auf Seite 1von 12

SPE

Society of Petroleum Engineers

SPE 14256

Application of Pseudotime To Estimate Average


Reservoir Pressure
by S. Aanonsen, Norsk Hydro a.s.
SPE Member

Copyright 1985, Society of Petroleum Engineers

This paper was prepared for presentation at the 60th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in Las
Vegas, NV September 22-25, 1985.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the
author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum 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. Papers
pres~nted at SPE meetings are subject to publication revi~w by Editorial Co~mittees of the Society of Petroleum Engineers. Permission to copy is
restncted to an abstract of not more than 300 words. Illustrations may not be cop1ed. The abstract should contain conspicuous acknowledgment of where
and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.

ABSTRACT due to the introduction of the pseudopressure


function by Al-Hussainy et al. in 1966.1
Several methods taking into account vari- This idea was later extended to two-phase
ations in compressibility and viscosity have flow by Fetkovich2 and Raghavan.3 The trans-
recently been proposed to improve the formed equations are, however, still
estimate of average reservoir pressure from non-linear implying that one of the most
pressure buildup tests in gas reservoirs. important properties of linear equations -
This paper demonstrates how well-known the possibility of generating complicated
methods developed for oil reservoirs can be solutions from simple ones by superposition -
applied directly if the pseudopressure trans- is a priori not valid. Consequently, if the
formation is combined with pseudotime. It is well-known solutions of the heat equation are
shown that the method is equally applicable used to analyse more complicated cases, the
both for gas reservoirs and oil reservoirs validity of the solutions has to be
where substantial gas saturation has considered separately for each type of test
developed during production. The method may (drawdown, buildup, multirate, etc.).
also improve the estimates of formation
permeability and skin factor. Kazemi4 showed in 1974 that the variations in
the compressibility-viscosity product, ~c,
By introducing different flow regions where could cause errors if the Matthews-Brons-
simplifying assumptions can be stated, it is Hazebroek (MBH) method 5 was used to estimate
shown that pseudotime is valid for buildup, average pressure in gas reservoirs and
but not for drawdown. In addition, it is
tp >> tpss. To eliminate this error, he used
shown that the use of the producing GOR at an iterative method, which was based on the
shut-in to calculate buildup pseudopressure use of tpss instead of tp in the Horner plot.
in solution-gas-drive reservoirs also depends hater, other methods have been proposed to
on these assumptions. account for variations in the ~c product when
estimating average reservoir pressure from a
Numerical simulations demonstrating the buildup test. 6, 7 However, the theoretical
validity of the method is presented for a basis for both these methods is questionable.
circular as well as different rectangular
geometries. In 1979, Agarwal 8 introduced a new time
transformation, pseudotime, to account for
INTRODUCTION variations in ~c. Although the theoretical
verification of this transformation has been
In contrast to the diffusivity equation for limited, it has been used with success to
slightly compressible fluids, the equations analyse effects of storage and fractured
describing gas flow and two-phase flow in wells .8 - 11 Recently, several investigations12 · 15
porous media are not easily linearised, and have shown that pseudotime does not linearise
no exact transformation between these the gas flow equation for drawdown. However,
equations and the linear diffusion equation for buildup, pseudotime can be a very useful
exists. A majpr theoretical improvement was tool. This has been verified numerically by·
Reynolds et al., 15 and in Refs. 12 and 13,
References and illustrations at end of paper. it is shown that the pseudotime trans-
2 APPLICATION OF PSEUDOTIME TO ESTIMATE AVERAGE RESERVOIR PRESSURE SPE 14256

formation for buildup-is consistent with the To establish the connection between m(r,ta)
solution obtained by using a regular pertur- and the corresponding liquid solution, we
bation method. have found it useful to introduce different
flow regions as shown on Fig.1. Region I and
In this this paper, it is shown that the II contain the drawdown period and Region III
applicability of pseudotime for buildup also the buildup period.
may be explained by introducing different
flow regions, and it is demonstrated how The following assumptions are then made:
pseudotime may be used to estimate average
reservoir pressure from buildup tests in gas 1. In region I, the flow is approximately
reservoirs. It is also shown that neither of radial. In addition, the time
the two methods of applying pseudotime derivatives in the flow equations are
presented in Ref. 15 are rigorously correct. small and may be neglected.
However, the difference between the Horner
time used here and "HR3" of Reynolds et al.15 2. In Region II, all terms in the flow
will be very small, and as pointed out in equations are of the same order of magni-
Ref.15 the "HR4" Horner time ratio will give tude. However, the variations in the
good results for long producing times. dependent variables are relatively small,
Reynolds et al. report that equally good and if the coefficients are slowly
results are obtained by using a normalised varying functions of the dependent
time in the Horner plot. This method is not variables, they may be assumed constant.
considered here.
3. In Region III, which is the buildup
In 1981, a method to calculate buildup period, quadratic gradient terms may be
pseudopressure in solution-gas-drive neglected.
reservoirs slightly different from the one
proposed by Raghavan3 was ~uggested by B¢e et The relative size of Region I and II will
al~6 B¢e et al. used the pressure- change with the degree of nonlinearity.
saturation relation at wellbore to calculate However, this effect seems to be negligible,
buildup pseudopressure, and this method and it will be shown in the following that it
seemed to have a better theoretical is possible to generate approximate solutions
foundation than Raghavan's method. Despite by matching solutions in the different
this, the method of B¢e et al. yields a regions.
highly rate dependent pseudopressure function
and is inferior to Raghavan's method. 17 Single-Phase Flow
We show that the reason for this also may be
explained using flow regions. A precise For single-phase flow of real gases through
definition of two-phase pseudopressure is porous media, Darcy's law, the continuity
given which combines the definitions of equation, together with an equation of state
Fetkovich 2 and Raghavan, 3 and a method for may be combined to:
calculating pseudotime for two-phase flow is
proposed.
jlC Clm
Simulated examples are presented both for 0
, ............... . (1)
single-phase gas flow and two-phase flow. (]lc)i CltDi
The latter examples are limited to solution-
gas drive. However, with some minor changes
the method should be applicable to any two-
phase system provided the fluid flow can be where m0 = m0 (p) is the dimensionless real
described by the "beta"-formulation, i.e., gas pseudopressure function of Al-Hussainy et
formation volume factors and solubility al. 1 defined by:
factors being functions only of pressure.

THEORY kh pi p' dp'

The objective of introducing the pseudo-


pressure and pseudotime transformations are
~(p)
711 Q T
g
J
p
ll(p')Z(p')
•••••••••••• (2)

to transform the wellbore pressure as a


function of time, Pw(t), into a new function,
mw(ta), in such a way that the new function (Field units are used consistently.)
behaves as the corresponding liquid solution.
Subscript i is used to indicate that
In order to achieve this, not only the dimensionless quantities are based on initial
equations, but also the initial and boundary
values. In Eq. (1), ¢and k are assumed to be
conditions have to be identical to the corre-
constants, but a variation in these
sponding equations for liquid flow. Since
quantities with pressure may be accounted for
superposition generally is invalid, the
in the definition of pseudopressure. 18
drawdown profile at shut-in has to be
considered as an initial condition for
To account for the nonlinearity still
buildup. Hence, it is not possible to
remaining in Eq. (1) due to variations in 11 c,
explain the validity of the buildup solution
Agarwal 8 introduced a pseudotime function, t ,
without considering also the foregoing draw- defined by: a
down.
SPE 14256 S. I. AANONSEN 3
As noted by Finjord, 14
an extra term
0.0002637 kt involving 3ta/3r will also appear in the
a boundary conditions. However, since 3m/3r
(3)
is zero both at the well and at the outer
boundary for buildup, it is seen that this
term may be neglected as long as ~c is
continuous. In order to verify the validity
However, since ~c depends on both r and t of the pseudotime transformation for buildu~,
through pressure, ta is a function of both r it then only remains to show that also the
and t. Hence, Eq. (1) is still not initial condition for buildup, i.e. the
linearised, but transforms in radial geometry drawdown profile at shut in, is the same as
to: for liquid flow.

1 d(
--- r--
d~) =d~
--
A gas reservoir will not reach a pseudosteady
state (PSS) characterized by a constant
r 3r D3r 3t pressure decline at every point in the
D D D aD reservoir. However, it is well known that
the dimensionless pseudopressure profile in

flb) _2(3taDV 3 ~ 2
2 2 PSS to a very good approximation is identical
_ (3taD) (3 ) to the profile of the liquid solution. 1 That
3rD~rD3taD is:
2
3rD 3ta D

(3 taD\~3mD )
2
_[2_ (3tac) + (4)
r 3r 3r 2} 3t
D D D aD
- (t ) + l(lnt - p (t .) + 0.8091) .. (6)
ron DAi 2
Di DMBH DAl
Note that the last term on the right hand
side of Eq. (4) is missing both in Refs. 9 and where PD is the constant terminal rate
14. The first two nonlinear terms in Eq. (4) solution of the radial diffusion equation.
is seen to be quadratic gradient terms and
may be neglected according to assumption 3. From Eq. (6) together with the linearised form
Neglection of the last parenthesis requires: of Eq. (4) it now follows that in any units:
s
m* - m = -_ - p DMBH (t DAi ) , . . . . . . . . . . . . . . . ( 7)
2 303

where s in the slope and m* the extrapolated


value of the Horner straight line obtained by
plotting mw 5 VS. log[~ta/(tp/(~c); + ~ta)J.
The details in this derivation may be found
<< 1 •• (5) in the Appendix. Note that pseudotime is
used in all terms involving ~t but not in the
term involving the production time. However,
except for very short production times the
The dominating term in E~(5) during buildup difference between this Horner time and the
will be the last term in the integrand, and Horner time denoted "HR3" in Ref. 15 will be
it follows that [(~c) i/ (~c) 2 ] [d(~c) /dmD] has very small. Note also that from the
to be small. This condition may be a derivation in Appendix it follows that other
limitation, but at least for the cases we methods to estimate average pressure can be
have considered, the term is of an order of used as for liquid flow provided ~t is
magnitude 10- 1 maximum at shut-in and replaced by ~ta in the analysis.
decreases very fast with increasing pressure.
The integrated valuewill thus be considerably
smaller, and consequently, it is reasonable When t > t 25 ~ PD -2rrtDAi is independent of
to neglect the nonlinear terms in Eq. (4) time, and Eq. (6) expresses that mD and Pb
during buildup. In Eq. (3) ~ta is then has the same radial dependence. It is seen
obtained by integrating from tp to tp + ~t at that this also may be explained from
r = rw. assumption 1 and 2. It is also seen that the
value of ~c used in the definition of
In Region I, however, the gradient terms will dimensionless time in Eq. (6) is arbitrary as
be much larger than the time derivatives, and long as toAi > toAPss Consequently, the
even if all the terms on the right hand side derivation in Appendix may be carried out for
of Eq. (4) can be neglected to get the ·radial any value of dimensionless production time
variation of m, the nonlinear terms will larger than tpss This corresponds to the
contribute to the time variation. Hence, well-known fact that different production
pseudotime is not valid for drawdown. This times may be used in the MBH-analysis, and
may also be shown by a perturbation explains why an analysis using the "HR4"
technique 1 ~ 13 in ac~ordance with the results Horner time of Reynolds et al. 15 gives the
of Finjord 14 and Reynolds et al. 1 s correct result for long production times.
4 APPLICATION OF PSEUDOTIME TO ESTIMATE AVERAGE RESERVOIR PRESSURE SPE 14256
However, as pointed out also by Reynolds et <P a
al. 15 , the same dimensionless time has to be oil: V•{ a(p,S )Vp} = - - S(p,S)
o k at o
' ... (9)
used both in the Horner time ratio and the
MBH function.

It is here important to realise that this s0 1 - sg s.l.W (10)


result is not influenced by the deviation
between mn(tnAi) and Pn(tDAi) or mwn<tnAi)
and Pwn<tnAi) in PSS. Hence, the only error The problem when defining a pseudopressure
made when estimating average pressure without function for two-phase flow is that satu-
using pseudotime is due to the variations in ration, and hence a and a, are not uniquely
~c during buildup. It can be shown that defined functions of pressure. 12 However, in
Eq. (6) breaks down and predicts a newative the drawdown period, pressure will be a
wellbore pressure at late times, 12 • 1 but strictly monotone function of radius and
this should not be any large problem in time, and it is then possible to express
_practice since other constraints on pressure saturation as a function of pressure and time
probably will imply a decrease in production or pressure and radius, respectively:
rate before this limit is reached. Note that
since Kazemi's method 4 is based on replacing S S (r,t)
0 0
tp with tpss in the Horner plot, this method
also relies on the validity of Eq. (6) • S (r,t(p,r)) S (p,r)
0 0

Note also that the application of pseudotime s0 (r(p,t) ,t) s 0 (p,t) ( 11)
to buildup analysis is not based on super-
position, but on the fact that mn(rn,~tan> to
a very good approximation satisfies the same Note that Eq. (11) will not be valid for a
equation with boundary and initial (buildup) general rate history.
conditions as the corresponding solution for
liquid flow (see Appendix). Eq. (11) may now be substituted for S 0 in the
expressions for a, b, a, and S making these
In order to use Eq. (7) to estimate average quantities functions of pressure and radius
reservoir pressure, Eq. (6) must be valid for or pressure and time similarly. A pseudo-
a general geometry with r being radius pressure function can then be uniquely
vector. This is reasonable since the outer defined for all r and t ~ tp by integrating
boundary only affects the pressure profile in first over time for r = rw and then over
Region II where the variations in ~c are radius. That is, for drawdown:
small. Another condition is that p(m) has to
be equal to p. As noted in Refs. 1 and 15, p(r,t)
this seems to be the case for all practical
purposes. m(r,t) + J a(p' ;t)dp'

A derivation similar to the one presented in pw(t)


the Appendix shows that the presented method
may be used to estimate initial pressure in t ~ t .... (12)
p
the infinite-acting period provided
mn(rn,tni) = pn(rn,tni> at shut in. The error
in estimated initial pressure (dimensionless)· In Eq. (12), it is assumed that the boundary
will be equal to mn<rn,tni> - Pn<rn,tni> at condition is a specified constant surface oil
shut in. For gas flow, this difference is of rate. However, any linear combination of a
order 10- 2 • 13 If pseudotime is used for and a corresponding to a linear combination
production time in this case, the error will of Og, and q 0 can be usedJ 6 Note that since
be equal to the difference between mD(rD,tan> saturation generally is a different function
and Pn<rn,tan> at shut in,which can be of pressure at different points in the
substantial.1s reservoir, pseudopressure, when defined by
Eq. (12), is not a uniquely defined function
Two-Phase Flow of pressure. Eq. (12) inserted in the oil
equation, Eq. (9), now yields the following
We assume that the fluid flow can be equation:
described by diffusivity equations based on
"beta"-formulation, i.e., formation volume p(r,t)
factors and solubility factors being
functions only of pressure. If only oil and
gas are assumed to be mobile and the
irreducible water is assumed to be incompres-
v2m =~=-)*am-~=-)*
k\A at k\A
J {~ at
a(p' ;t)}dp'

sible, the flow equations are in the notation pw(t)


of B¢e et al.16:
t <t .... (13)
p
Here:
a
{ ~aS0[asol +~}/a
<P
gas: V•{ a(p,S )Vp} = --b(p,S) '
o k at o
... (8)
GJ ap JR ap
' ••••••••••• (14)
SPE 14256 S. I. AANONSEN 5

with [as 0 /ap]Rbeing the derivative of S0 with replaced by a kind of average function of
respect to pressure when written as a pressure.
function of pressure and radius as in
Eq. ( 11) • Pseudotime has been used for solution-gas-
drive reservoirs by Verbeek 11 who reports
In Region I, both terms on the right hand good results when using an average of ct!At
side of Eq. (13) may be neglected implying a between initial and wellbore values. For the
logarithmic pseudopressure profile as for case of buildup from PSS, however, it is
liquid flow. In Region II, the variations in doubtful that this will be a good solution.
a will be relatively small, and in addition, We have found that good results in this case
most of the variations will be, through can be obtained by using the relation between
pressure making I( a/at) (a (p;t) >I << 1. Ct/Atand p for drawdown. This relation is
Consequently, the pseudopressure profile shown in Fig. 3 and it is seen that Ct/At
during drawdown, and thus the initial then is relatively independent of position
condition for buildup, will be approximately and a reasonable average to buildup (c/A)*.
equal to the pressure profile for liquid Shut-in pseudotime for two-phase flow is
flow, and it follows that Eq. (6) is a good then defined by:
approximation also for two-phase flow. Note
also that Eq. (12) combines the definitions of 6t
pseudopressure given by Fetkovich 2 and
Raghavan. 3

If Eq. (11) is valid in the buildup period, a


pseudopressure function may be defined by
6t
a
f
0
~t
1\t
d(6t')
[Pws (lit >]
I
' .......•....... (17)

Eq. (12) also for buildup, and it is seen that where


this would be equivalent to the method of B¢e
et al. 16 However, for buildup, neither the
time derivatives nor the variations in a will
be small, and Eq. (13) cannot be linearised.
To get a buildup equation that can be
linearised, pseudopressure may be defined as
a function of pressure using the m(p) profile So(B dRso _ dBo)+ SG(B0 drsG _ dBG)
at shut in. That is: __ Bo Gdp dp BG dp dp
' .. (18)
( k /A + k /A
ro o rg g

m(r,t)
and the pressure-saturation relationship used
in Eq. (18) is thus drawdown SJp) at wellbore.

t >tp .... (15) The remaining problem is now the saturation,


which is needed both for the calculation of
pseudopressure and pseudotime. In the first
With this definition, Eq. (9) transforms to: integral in Eq. (12) , only saturation at
wellbore is needed, and the same is the case
with pseudotime. This saturat~on may be
found from the producing gas-oil ratio (GOR)
as shown by Raghavan, 3 or in the infinite-
acting period also from the relation of B¢e
et al. 16 Note, however, that Raghavan's
method does not give the saturation when gas
t >t (16)
saturation is less then critical. By
p
neglecting ab/at and as/at in Eqs. (8) and
Now the nonlinear term is a quadratic (9), it is easily seen that GOR is
gradient term, which may be neglected independent of radius in Region I and equal
according to assumption 3. (a(p,S 0 ) is here to the producing GOR. This is slightly
the value corresponding to the actual values
of pressure and saturation, while a(p;tp) is different from Levine and Prat's method, 19
the value obtained when the actual pressure since they assumed GOR to be constant for all
is inserted in the relation between a and p at r and correspond to the pressure and
shut in.) Consequently, when defined by saturation at the outer boundary. It may be
Eq. (15), the dimensionless buildup pseudo- shown that GOR varies significantly with r in
pressure (defined in the nomenclature) will Region II, but because of the small
to a good approximation satisfy an equation variations in p and S0 in this region, a good
similar to Eq. (1) with ~c replaced by (c/A)* approximation to the pseudopressure profile
and with Eq. (6) as the initial condition for seems to be obtained from the producing GOR
6t = 0. (c/A)* will, however, not be a for all r, at least for the solution-gas-
function only of pressure as for gas flow. A drive case. Note also that when producing
typical example is shown in Fig. 2, where GOR is used to calculate the integral in
(c/A)* is plotted vs. pressure at different Eq. (15) this method is equivalent to the
points in the reservoir. Hence, an adaption method for calculating buildup pseudopressure
of the multiphase flow effects to the gas proposed by Raghavan. 3 We have thus shown
flow solution requires that (c/A)* can be that conditions 1-3 about flow reqions also
6 APPLICATION OF PSEUDOTIME TO ESTIMATE AVERAGE RESERVOIR PRESSURE SPE 14256
arenecessary and sufficient conditions for the theory presented and the validity of the
Raghavan's method to be valid. fundamental assumptions. Some additional
examples may be found in Ref.12.
Application of this method to estimate
initial pressure from a Horner analysis will
depend on the accuracy of the pseudopressure CONCLUSIONS
profile calculated in Region II. For the
solution-gas-drive case the producing GOR 1. For buildup, pseudotime together with
seems to give accurate results, but this may pseudopressure linearises the flow
not be the case generally. This problem is equations both for single-phase gas flow
outside the scope of this paper, but some and two-phase flow described by the
examples can be found in Ref.12. "beta"-model to a very good approximation.
Pseudotime does not linearise the flow
equations for drawdown.
EXAMPLES
2. Provided correctly used, application of
Four simulation runs were chosen to examplify pseudotime in a Horner/MBH analysis
the presented theory, one gas-reservoir improves the estimate of average pressure
example and three examples of solution-gas and also the calculated formation
drive. Test data for the single-phase case permeability and skin factor.
were generated by solving Eq. (1) numerically
with a NAG-routine 20 for partial differential 3. For two-phase flow, pseudopressure is not
equations. The examples of two-phase flow a uniquely defined function of pressure,
were generated with a two-dimensional, but can be uniquely defined for every r
three-phase reservoir simulator with variable and t if pressure is a strictly monotone
saturation pressure developed at Rogaland function of radius and time for drawdown.
Research Institute. 21
4. For two-phase flow, the pseudotime used in
Reservoir and production data for the the buildup analysis can be calculated
simulated examples are shown in Table 1. Gas from Ct/At as the function of ~t obtained
properties are taken from Ziauddin, 6 and by using drawdown S0 (p)-relation at
example 1 is identical to Kazemi's example wellbore.
1, 4 except that the production rate is kept
constant and the production time is slightly 5. The drawdown S0 (p)-relation needed to
different. Fluid properties and relative calculate pseudotime may be found from the
permeabilities for the solution-gas-drive producing GOR. The S 0 (p)-profile needed
cas~are taken from Ref. 16, the only to calculate buildup pseudopressure may
difference being initial pressure and initial (at least for solution-gas drive) be found
bubble-point pressure. from the producing GOR at shut in.

Results 6. To study a drawdown/buildup process, it is


useful to introduce different flow regions
Horner plots corresponding to the simulated where simplifying assumptions can be
examples are shown in Figs.4-7. ~ws is stated.
plotted both vs. inverse Horner t1me and
inverse Horner pseudotime, and the results NOMENCLATURE
from an analysis of both plots are presented
in Table 2. In the table, calculated A Drainage area, sq.ft.
absolute permeability and skin factor are a = k /~ B + R k /~ B
rg g g so ro o o
presented in addition to average pressure.
It is seen that in all cases, the standard a = k /~ B + r k /~ B
procedure gives too low values for k, S, and ro o o sg rg g g
p, a result that agrees with the analytical
solution presented in Refs. 12 and 13 based b = S /B + R S /B
g g so 0 0
on a perturbation technique.
8 = S /B + r s /B g
sg g
0 0
In all cases the pseudotime procedure gave Bg, B0 Formation volume factors for gas
average pressures close to the simulated and oil, rcf/scf, rb/stb
values. The relative error was less than 1%
except for example 4 where a slightly larger c Compressibility, 1/psi
error may indicate problems with very assym-
metrical geometries. GOR Gas-oil ratio, scf/stb

For example 3 and 4, the pseudotime analysis, h Reservoir height, ft


gave a positive skin factor. The reason for k
this is not quite clear, but some of this Absolute permeability, rnD
skin may be explained from a possible Relative permeabilities
inaccuracy in the calculation of wellbore
flowing pressure from well-block pressure. Pseudopressure, psi 2 /cp (gas
flow), psi/cp (two-phase flow)
However, the pseudotime procedure gave the rn*, rn'*
best results in all examples, and supports Extrapolated values on Horner
·plots
SPE 14256 S. I. AANONSEN 7
Dimensionless pseudopressure: ACKNOWLEDGMENTS
kh
(mi - m) gas I wish to thank all members of staff at the
1422QgT
Institute of Mathematics, University of
kh (mi - m) solution-gas drive
Bergen and Rogaland Research
mo 141.2qo Institute/Rogaland Regional College,
Stavanger for help during the progress of
p, p Pressure, average pressure, psi this work, and especially Dr. Svein M.
Skjreveland for helpful discussions and
Po Constant terminal rate solution suggestions. I also thank Norsk Hydro a.s.
of radial diffusion equation for supporting the publication of this paper.
(liquid solution) Financial support from the Royal Norwegian
Council for Industrial and Scientific
Qg Surface gas production rate,
Mscf/d Research is gratefully acknowledged.

q.o Surface oil production rate, REFERENCES


stb/d

r, rD r/rw Radius, dimensionless radius 1. Al-Hussainy, R., Ramey, H.J. Jr., and
s Skin factor Crawford, P.B.: "The Flow of Real Gases
Throuqh Porous Media", Trans. AIME
s g , s o , s. ~w
Saturations (1966) 237, 624-636.
s Slo~e on Horner plot,
psi /cp-log~ or psi/cp-log~
2. Fetkovich, M.J.: "The Isochronal
Testing of Oil Wells", Paper SPE 4529
t Time, hrs presented at the 1973 SPE Annual
Meeting, Las Vegas, Sept. 30 - Oct. 3.
t.t Shut-in time, hrs
3. Raghavan, R.: "Well Test Analysis:
Pseudotime, shut-in pseudotime, Wells Producing by Solution Gas Drive",
hrs/cp/psi Soc. Pet. Eng. J. (Aug. 1976) 196-208,
Trans. AIME, 261.
Dimensionless time based on rw:
4. Kazemi, H.: "Determining Average
0.0002637 kt gas Reservoir Pressure From Pressure Buildup
toi= <j>(Jlc)irw2 Tests", Trans., AIME (1974) 255, 55-62.
0.0002637 kt
toi= <j>(c/A)*irw2 solution-gas drive 5. Matthews, c.s., Brons, F., and
Hazebroek, P.: "A Method for
Determination of Average Pressure in a
Dimensionless time based on
drainage area (r~ replaced by A)
Bounded Reservoir", Trans., AIME (1954)
201, 182-191.
Dimensionless pseudotime
6. Ziauddin, Z.: "Determination of Average
Production time, hrs. Pressure in Gas Reservoirs From Pressure
Buildup Tests", Paper SPE 11222
Porosity presented at the 1982 SPE Annual
Technical Conference and Exhibition, New
Jlg' Jlo Viscosities, cp Orleans, Sept. 26-29.
Total mobility
7. Toh, C.H., Farshad, F.F., and LeBlanc,
Generalized compressibility- J.L.: "A New Iterative Technique With
(c/A)* Updated Curves for Estimating Average
mobility ratio, defined by
Eq. (14), cp/psi Reservoir Pressure of Gas Wells from
Buildup Tests", Paper SPE 13235
Subscripts presented at the 1984 SPE Annual
Technical Conference and Exhibition,
D Dimensionless Houston, Sept. 16-19.
f Flowing
8. Agarwal, R.G.: "'Real Gas Pseudotime' -
g Gas A New Function for Pressure Buildup
Analysis of MHF Gas Wells", Paper SPE
i Initial 8279 presented at the 1979 SPE Annual
Technical Conference and Exhibition, Las
MBH Matthews-Brons-Hazebroek Vegas, Sept. 23-26.
0 Oil 9. Lee, W.J. and Holditch, S.A.:
PSS Pseudosteady.state "Application of Pseudotime to Buildup
Test Analysis of Low-Permeability Gas
s Shut in Wells with Long-Duration Wellbore
Storage Distortion", J. Pet. Tech. (Dec.
w Well 1982) 2877-2887.
8 APPLICATION OF PSEUDOTIME TO ESTIMATE AVERAGE RESERVOIR PRESSURE SPE 14256
10. Bostic, J.N., Agarwal, R.G., and Carter, 12/82, Rogaland Research Institute,
R.D.: "Combined Analysis of Stavanger (1982).
Postfracturing Performance and Pressure
Buildup Data for Evaluating an MHF Gas APPENDIX
Well", J. Pet. Tech. (Oct. 1980) Buildup Equations
1711-1719.

11. Verbeek, C.M.J.: "Analysis of The basis for the buildup theory is that the
Production Tests of Hydraulically nonlinear terms in Eq. (4) are negligible
Fractured Wells in a Tight Solution during buildup, and that Eq. (6) is valid at
Gas-Drive Reservoir", Paper SPE 11084 the instant of shut in. Define now a new
presented at the 1982 SPE Annual dimensionless pseudopressure function
Technical Conference and Exhibition, New mD(rD,lltaD> by:
Orleans, Sept. 26-29.

12. Aanonsen, S.I.: "Nonlinear Effects


During Transient Fluid Flow in
Reservoirs as Encountered in Well-Test roD will then satisfy the following initial-
Analysis", Report No. 79, Department of boundary-value problem:
Applied Mathematics, The University of
Bergen, Bergen 1985.
llt
aD
>0 , . . . . . . . . . (A-2)
13. Aanonsen, S.I.: "A Study of a
Quasilinear Diffusion Equation with
Application to Well-Test Analysis", to
be published.
rnD(rD,lltaD I • • • • • • • (A-3)
14. Finjord, J.: "A Study of Pseudotime",
Paper SPE 12577, Preliminary version,
submitted to Soc. Pet. Eng. J. (1984).

15. Reynolds, A.C., Bratvold, R.B., and 0 , lltaD > 0 (A-4)


Ding, W.: "Semilog Analysis of Gas Well
Drawdown and Buildup Data", paper SPE boundary
13664, presented at the California
Regional Meeting of SPE, Bakersfield, The problem, Eqs.(A-2)-(A-4~ is identical to
March 27-29 (1985). the corresponding problem for liquid flow,
and the solution ffiD(rD, lltaD> will thus be
16. B¢e, A., Skj~veland, S.M., and Whitson, equal to the corresponding liquid solution
C.H.: "Two-Phase Pressure Test with lltD replaced by lltaD· That is:
Analysis", Paper SPE 10224 presented at
the 1981 SPE Annual Technical Conference
and Exhibition, San Antonio, Oct. 5-7.

17. Whitson, C.H.: "Topics on Phase


Behavior and Flow of Petroleum Reservoir Note that Eq. (A-5) is not based upon super-
Fluids", Thesis for the Degree Doctor of position of two pseudopressure solutions.
the Technical Sciences, Norwegian Note also that this shows that pseudotime
Institute of Technology, University of should be used in all terms involving llt
Trondheim, Trondheim (1983). but not in the terms involving production
time-.--Introducing the MBH functions and the
18. Raghavan, R., Scorer, J.D.T., and Horner approximation, Eq. (A-5) may be written
Miller, F.G.: "An Investigation by at rD = 1 as for liquid flow:
Numerical Methods of the Effect of
Pressure-Dependent Rock and Fluid t + llt
Properties on Well Flow Tests", Soc. :mwD (lltaD ) _ ln Di aD
Pet. Eng. J. (June 1972) 267-275, 2 llt
Trans., AIME 253. aD

19. Levine, J.S. and Prats, M.: "The 1


Calculated Performance of
Solution-Gas-Drive Reservoirs", Soc. + 2TitDAi - ~ pDMBH(tDAi) . . . . . . . (A-6)
Pet. Eng. J. (Sept. 1961) 142-152,
Trans. AIME 222.

20. NAG FORTRAN Library Mark 9, Numerical 1 tD. + flt


Algorithm·s Group, Oxford (1983). m (llt ) _ 1n 1. aD
wD aD
2 llt
21. Tjetland, G., Litlehamar, T., and aD
Skj~veland, S.M.: "'TODVARS' -A
Three-phase, Two-Dimensional Implicit 1
Reservoir Simulator with Variable
Saturation Pressure", Report No. T + mD(tDAi) - ~ pDMBH(tDAi) ..... (A-7)
SPE 14256 S I AANONSEN 9
or Eq. (7) now follows directly by introducing m*
as the extrapolated value of mws as 6t + oo
kh in Eq.(A-8).
(m (6t ) - m)
ws a
1422 Q T
g

1 6t 1
- ln t a - - pDMBH(tDAi) •• (A-8)
2 p/( ) + 6t 2
l.JC i a

TABLE 1 - RESERVOIR AND PRODUCTION DATA

gas solution-gas drive


Example 2 3 4
Porosity 0.05 0.30 0.30 0.30
Absolute permeability (mD) 1.5 10.0 10.0 10.0
Initial gas saturation 0.00 0.00 0.00
Initial pressure (psi) 7000 4000 4000 4000
Initial bubble-point pressure (psi) - 4000 4000 4000
Well radius (ft) 0.276 0.33 0.33 0.33
Reservoir height (ft) 40 15.5 15.5 15.5
Drainage area (sq.ft) 2.19•10 7 1.37•10 6 8.00•10 4 8.00•10 4
Temperature ( 0 R) 670 660 660 660
Oil production rate (stb/d) 100 100 100
Gas production rate (Mscf/d) 5500
Production time (hrs) 20200 10600 530
Connate water saturation 0.30 0.30
Maximum free gas saturation 0.21 0.19
Reservoir geometry 8 01 2
(l.JC) , (cp/psi)
~

(c/A.)*i (cp/psi)

TABLE 2 - RESULTS OF HORNER/MBH ANALYSIS

Calculated values Model values


m vs. Horner- m vs. Horner-
t:f~e plot p~~udotime plot
Example 1 2 3 4 1 2 3 4 1 2 3 4
Absolute
permeability (mD) 1.41 9.1 8.1 8.0 1.51 10.1 9.1 9.8 1.50 10.0 10.0 10.0
Skin factor -0.9 -1.7 ""'1.0 -L2 -0.02 -.06 0.5 0.7 0.0 0.0 0.0 0.0
Average
pressure (psi) 3257 2495 2680 2757 3370 2750 2920 2948 3396 2763 2909 2879
ui
tn
(I)
(.)
...0
Q.
z Q.
;:
g ::I
;g
~
-
::I
a:
c .c
1:
==
0
"C
...==ca
"C
...0ca
- tn
1:
0
·c,
~
==
-
0
;:
1:
~
(I)
c=
,....I
.~
u..
0
0
0
0

r = 10 rw
"'""'
g - - - - r = 100 rw
'=!

"iii DRAWDOWN
~
.!:!. :il
g
~ '=!

"'
~
~

0
g
g
d
500 1000 1500 2000 2500 3000 3500 4000
p (psi)

Fig. 2-(cn.)• vs. pressure at different points In the reservoir-Example 2.

0
0

~
~

~ 0

~ g
"'
.!:!. '=!
...r
~

"'~
~

0
0
g
0
d
500 1000 1500 2000 2500 3000 3500 4000
p (psi)

Fig. 3-c 1/1. 1 vs. pressure at different points In the reservoir-Example 2.

EXAMPLE 1
100

90

80

70
D.
...
"""
"iii
_g. 60

~
; 50
E

40

30

20
10-6

t,t 6t.
Inverse Horner time~ or tplC.U:lt t,t.

Fig. 4-m ws vs. Inverse Horner times-Example 1.

SPE 1 4 25.6
EXAMPLE 2
-800 m•

m''
-900

-1000

-1100

-1200

-1300 6t
0 mw. vs.
lp + 61

-1400
61.
b. mw 0 VS.~
lpl :r. i +61.
-1500

-1600
10-6 10-5 10-4 10-3 10-2 10-1 10°

6t.
Inverse Horner time _..fL!._ or
lp + 61 tP"'{f)i +6t.

Fig. 5-mws vs. Inverse Horner times-Example 2.

EXAMPLE 3
-600

-700 m•

m••
-800

-900

'ii -1000
~
'iii
a.
; -1100
E
6t
0 mw. vs.
i;:"+Tt
-1200

6t
b. mw. vs.
-1300 tpf~;+6t.

-1400
1o-& 10-2 10-1

Inverse Horner time .........::::.J or


lp + 61
f!i:fj +t, 6t.
tpf)..

Fig. 6-mws vs. inverse Horner times-Example 3.

EXAMPLE 4
-800
m•
m''
-900

-1000

-1100

'ii -1200 ~
~ 0 mwsVS. lp+ 61
'iii
.e 611
; -1300 b. mws vs.tA{); +6ta
E

-1400

-1500

-1600
10-6 10-5 10-4 10-3 10-2 10-1 10°
6t
6t.
Inverse Horner time ~ or 1 '(~)' + 61 8
p P' A j

Fig. 7-mws vs. Inverse Horner times-Example 4.

SPE 1 4 25.6

Das könnte Ihnen auch gefallen