Vous êtes sur la page 1sur 20

SPE

Society of Petroleum Engineers

SPE 16204
Inflow Performance Relationships for Solution-Gas-Drive
Reservoirs
by R.G. Camacho and R. Raghavan, U. of Tulsa
SPE

Members

Copyright 1987, Society of Petroleum Engineers


This paper was prepared for presentation at the SPE Production Operations Symposium held in Oklahoma City, Oklahoma, March 8-10, 1987.
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
presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Permission to copy is
restricted to an abstract of not more than 300 words. Illustrations may not be copied. 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
that the exponent of the deliverability curve is known.
In this theoretical study, a numerical model was

Although, relative permeability and fluid property data

used to examine the influence of pressure level and skin

are required, the Muskat mat~rials balance equation

factor on the Inflow Performance Relationship (IPR) of

and the assumption that gas-oil ratio is independent

wells producing solution gas drive systems. Examina-

of distance can be used to predict future production

tion of the synthetic deliverability curves suggests that

rates.

the exponent of the deliverability curve is a function

other methods in the literature and always yields reli-

of time, and that the exponent is usually greater than

able results.

This method avoids problems associated with

unity. The implication of this observation to field data


is discussed. The accuracy of procedures given in the
literature to

~redict

New methods to modify the IPR curve to incorpo-

oil-well deliverabilities is also ex-

rate changes in skin factor are presented. A new defini-

amined. It is shown that the methods given in the litera-

tion of flow efficiency that is based on the structure of

ture can be used to predict future performance provided

the deliverability equations for solution gas drive reser-

that the exponent of the deliverability curve is known

voirs is proposed. This definition avoids problems that

and extrapolations over large time ranges are avoided. If

result when the presently available methods are applied

single point tests are used to predict future performance

to heavily stimulated wells.

(such tests assume that the exponent of the deliverability curve is constant), then errors in predictions will be
minimized if test conditions correspond closely to ac-

INTRODUCTION

tual producing conditions. Under these circumstances


errors in rate predictions will be directly influenced by

This paper addresses three concerns pertinent to in-

the magnitude of the error that results in assuming

flow performance relationships under solution gas drive.

199

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

SPE 16204

First, via numerical simulations, we examine the inflow

to modify the IPR curve can result in inconsistent pre-

performance relationship {IPR) when a skin region ex-

dictions if the skin factor, s , is negative. Whitson 9 says

ists around the wellbore.

This phase of our work is

that Standing's method leads to nonphysical results if

intended to address the question whether the generalized form of the IPR curve is governed by the exis-

< 0. These observations suggest that Standing's procedure should be incorrect if s > 0. In this paper we

tence of a skin region. Although the skin region has a

show that that inconsistent or nonphysical predictions

dominant effect on the well response, the influence of a

are a result of the definition of flow efficiency used by

skin region for all practical purposes has been ignored.

Standing. Standing defined flow efficiency based on liq-

{The seminal work of Vogel

however, does briefly ad-

uid flow and not on the quadratic form suggested by

dress the influence of a skin region.). Second, we ad-

Vogel's equation. In this work, we redefine flow effi-

dress the prediction of future IPR relationships from

ciency based on the quadratic forms of the deliverabil-

test data. Many methods have been proposed in the

ity equations suggested by Vogel 1 and Fetkovich 3 , and

literature 2 -

show that the inconsistencies mentioned by Brown can

to predict future performance. The meth2

ods of Standing and Fetkovich are based on empirical

be eliminated. Unfortunately, the curves developed by

observations of solution gas drive reservoir performance.

-standing need to be corrected for all values of skin fac-

All methods presented in the literature assume that the

tor. Incidently, the definition we present is consistent

mobility function, kro/ (p, 0 B 0 ) , is a linear function of

with the definition of flow efficiency used for gas wells.

pressure,p, where kro is the relative permeability to oil,

A new chart to modify the IPR curve based on the ap-

J.Lo is the viscosity of oil and Bo is the formation vol-

propriate deliverability equation {Vogel or Fetkovich) is

ume factor of oil. Although the methods of Standing

presented, and represents the third contribution of our

and Fetkovich can be justified on the basis of this as-

work.

sumption, this requirement {mobility function is a linear


function of pressure) is not a necessary condition for the
NUMERICAL MODEL

methods given in Ref. 2 and 3 (or for that matter the


other references cited above) to be valid. We examine

We consider a homogeneous closed circular reser-

the validity of the precedures suggested by Standing and

voir with the well located at the center of the reservoir.

Fetkovich via simulations for a wide range of producing

The well is capable of producing at either constant oil

conditions. We show that the success of these methods

rate or at constant wellbore pressure. Gravity effects

rests primarily on the fact that the deliverability equa-

are considered to be negligible. Initially the reservoir is

tions suggested by Vogel and Fetkovich incorporate fea-

assumed to be at the-bubble point. An annular region

tures that reflect the non linear aspects of flow in the

that is concentric with the wellbore, with a permeability

reservoir, and more importantly are calibrated via mea-

different from formation permeability is used to incor-

surements. Their procedures attempt to only predict

porate the influence of a skin region.

the end point - the final answer. We demonstrate this


point by comparing mobility function profiles predicted

Several sets of relative permeability and fluid prop-

by methods given in the literature.

erties were used in this study. Figures 1 and 2 show the


fluid properties for two of the data sets we have con-

The third part of this paper examines procedures

sidered, and figure 3 presents the relative permeability

to modify the IPR curve to incorporate the influence of

data used in this study. Table 1 presents other infor-

the skin factor. Brown 7 shows that Standing's 8 method

mation regarding reservoir properties used in this work.


200

SPE 16204

R. G. CAMACHO-V and R. RAGHAVAN

The results given in.this work, however, can be applied

reports that Vogel's equation yields n

over a wider range of conditions.

discuss later, this observation is important. In terms of

1.24. As we

Qo,maz, equation (2) may be written as:

A finite difference model was used to obtain the


results presented in this communication.

~=
Qo,maz

Procedures

2 )
1-Pwf
(

fi

(3)

followed to ensure that results are accurate are similar

Fetkovich (see also Handy 12 ) also suggested that if the

to that presented in Refs. 10 and 11 and will not be

variation of kro/ (p, 0 B 0 ) with pressure were known, then

considered in here.

well deliverability, for a well located at the center of a


circle may be calculated by the following equation:

Non Darcy flow effects are not considered.

The

p(t)

f (p) dp.

Qo = J {

influence of non Darcy flow for wells producing by solu-

} Pwf

tion gas drive represents a formidable analytical prob-

(4)

(t)

where
Jkh
- 141.2 [In (re/rw) - 3/4 + s]'

lem. Many fundamental considerations are involved and


these considerations are beyond the scope of this work.

and

The influence of non Darcy flow will be considered in a

f (p) = kro/ (p, B


0

0 ).

( 5)

All symbols in equations (4)

and (5) have their conventional meaning (see Nomencla-

future communication.

ture). Theoretical confirmation of equation (4) is given


in Ref. 13.

BACKGROUND
To predict future performance using equation (1),
Based on numerical simulations, Vogel showed that

Standing suggested that Qo,maz in the future can be ob-

the deliverability curve for a well producing under solu-

tained by the relation:

tion gas drive is given by:

~ = 1- 0.2P~!
Qo,maz

Qo,maz,f _ PJf (PJ)


Qo,maz,p - Pp/ (Pp).

0.8P~J.
P

(1)

Here the subscripts

(6)

f and prefer to future and present

conditions, respectively.

Here Qo,maz is the maximum possible flow rate (rate


corresponding to Pwf = 0), and q0 is the flow rate cor-

Fetkovich proposes that future performance can be

responding to average pressure, p , and wellbore flowing pressure, PwJ In 1973, based on field experiments

predicted by the relation: J~ 1 /J~P

Fetkovich suggested that the deliverability curve for so-

case we obtain (for n

tion:

2 )n
Jo1 ( t ) (-2
P - Pwf

to do so is to assume, as proposed by Handy, that

f (p)

varies linearly with pressure; that is:

Ref. 3 indicate that the exponent n, should be in the

(7)

sented by Fetkovich or Standing, a convenient method

ing under solution gas drive. Field tests presented in

= 1):

If we wish to establish a basis for the equations pre-

(2)

Here J~ (t) is the productivity index for a well produc-

range 0.5

and in this

Paf
qo,maz, f
=-a
Qo,maz,p
Pp

lution gas drive systems is given by the following rela-

Qo

= fiJ/fip,

f (p)

1. Based on our knowledge of gas well

= f (p)- [/ (p)-= f (0)] (p- p).


p

behavior, Fetkovich attributes exponent values less than

Substituting the right hand side of equation (8) for

unity to non Darcy flow, although he does acknowledge

(8)

f (p),

in equation (4) we obtain:

that the exponent, n, can be less than unity strictly as

Qo,maz,f
P! [fJ (p)
Qo,maz,p = Pp [/p (p)

a result of variation in fluid properties. Fetkovich also


201

+ fJ (0)]
+ /p (0)]'

(9)

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

SPE 16204

f (0) is proportional to f (p) (in Vogel's case

sions for J' and J* (productivity index at zero draw-

this constant of proportionality is 1/9; in Fetkovich's

down) which are derived in the literature on the basis of

f (0) is assumed to be zero), we obtain equation

the expression for J given by equation (5) and the as-

As long as

case

f (p) is a linear function of pressure will

(6). Fetkovich, then proceeds to make the assumption

sumption that

f (p) = ap ; that is f (p) = app, where a is a constant

in general be incorrect. We make this observation on the

that is independent of time, and under these circum-

basis of several sets of computations we have made 10

stances equation (7) results (for n =F 1, in the general

In fact, on the basis of a number of computations it can

ft (p) //p (p) = p}n' fp;nP),

case this may imply that

be concluded that the success of the methods given in


Refs.2 and 3 can be attributed to the fact that the

It is easy to show that equation (2) for n

=1

can

expression for the productivity index, J , given by the

be derived from equation (4) if equation (8) is valid. As

right hand side of equation (5) is never calculated ex-

f (0) = 0.

plicitly, although the expression for J given by equation

mentioned by Fetkovich, we can assume that

Fetkovich 's equation can also be derived if we assume


that

f (0)

{3 (p + Pwf)

, where

(5) is used in deriving the final result.

{3 is a constant of

proportionality. Inspection of equation (1) suggests that


it can be derived if we assume that

f (0) = {3p.

RESULTS

It should be noted that the assumption given in

In this section we first consider the influence of the

equation (8) is merely a convenient way to establish a

skin factor and pressure level on the shape of the deliv-

= 1), or equations (6)

erability curve. Methods to predict future performance

Equation (8) represents only

are then examined. Third, we consider modifications to

basis for equations (1) and (2) (n


and (7) (n

= 1) to be valid.

a sufficient condition to establish these results. It does

the IPR curve to incorporate changes in skin factor.

not represent a necessary condition for these results to


be valid.

For example we can readily show that the


The Inflow Performance Curve

relation:
q.

=J

[t (i') (i'- PwJ) - [/ (i') - f (0)] (i'-;;,wf)']

Figure 4 is a log-log plot of the deliverability curve

(10)

for a well producing under solution gas drive. The vari-

although equations (1)

able of interest is the fraction of initial oil produced,

and (3) may be valid. This point may become more

Np/N , (or average pressure, p ). The data shown here

readily apparent if we note that equation (4) will pre-

are for s

will in general be incorrect

13

> 0. Figure 5 presents a similar plot for s < 0

f (p)

for identical values of Np/N . Set 1 data are used in

f (p) functions can yield iden-

both plots. For each specific value of Np/N well de-

tical values of the integral. The point we wish to make is

fined straight lines (within engineering accuracy) are ob-

that the success of the methods of prediction of Stand-

tained. Careful examination of these responses suggests

f (p)

that there is some curvature at high rates and the curva-

function for the reservoir is a straight line or that we can

ture is concave downwards. Considerable care was taken

predict the f (p) function from measured data (q 0 , Pwf,

to estimate

and p).

downwards in these figures denote the value of qo,maz

dict flow rates accurately if the area under the


curve is accurate. Many

ing, Fetkovich and others does not imply that the

q 0 ,maz

for each case. The arrows that point

All data shown here are for boundary-dominated flow,


At this stage, it should be noted that the expres-

the alignment of the deliverability curve is insensitive to


202

R. G. CAMACHO-V and R. RAGHAVAN

SPE 16204

production mode. Unlike single-phase flow, however, a

It is difficult to make further generalizations re-

family of curves is obtained during boundary-dominated

garding the mobility function profile without additional

flow. Results similar to that shown here were obtained

assumptions. If we assume for convenience that the mo-

for Set 2 data. In this case, however, values of n are

bility function is given by Eq. (8) and note that

greater than unity even for small values of Np/N and

must be zero in order for n to be unity, then n can

this result can be attributed to the existence of a critical

be greater than unity under three circumstances: (i)

gas saturation.

f (0) > 0; (ii) f (0) < 0 and If (0) I< a (p+ PwJ)i and
(iii) f (0) < 0 and If (0) I > a (p + PwJ), where a is the

From a practical view point, the most important

f (0)

slope of the straight line (see Appendix).

points to note are: (i) the value of n is a function of


time or depletion, (ii) the change inn is not monotonic
with time, and (iii) for most cases the value of n

Although on the basis of our results and those of

is

Vogel, we can conclude that the exponent of the deliv-

is approximately

erability curve will be usually greater than unity, this

unity only for exceedingly low values of gas saturation.

observation does not imply that the deliverability equa-

greater than unity. The value of n

tion for single phase flow will be valid; that is, data
The above results indicate that the exponent n , is

should not be plotted in terms of the pressure differ-

a function of system compressibility or gas saturation.

ence (p - Pw 1) (In the gas case if n

The change inn with time reflects the change in the mo-

result would suggest that the deliverability equation is

bility function as a function of pressure. The fact that

of the form q0

n > 1 in many cases suggests that the variation in f (p)

are a replot of the data shown in Fig. 4 in terms of the

with pressure (distance) is much shallower than the pro-

pressure difference tip . These plots exhibit an upward

file suggested by Fetkovich (! (p)

= app).

= C (t) (p- PwJ)

> 1 , then this

.) The results in Fig. 7

It should be

bend at higher flow rates; that is, _the decline in oil rate

noted that Vogel also obtained exponent values greater

will be much faster than that predicted by the equation

than unity (n = 1.24). The results in Fig. 6 demon-

for single phase flow. It is possible to fit straight lines

strate this point where the variation off (p) with pres-

through portions of the data. The dashed lines shown in

sure for a specific case is shown for both transient and

Fig. 7 represent a fit through data at lower rates. The

boundary-dominated flow. Here, the symbol

rep-

exponent corresponding to these straight lines are ap-

resents dimensionless time based on the drainage area,

proximatly equal to 0.9. For reference straight lines with

A, and is given by:

slopes equal to unity are also shown (unbroken lines).


0.000263 7kkroit
<f>ctiJ.loiA

tDA=------

tDA

If we perform an analysis similar to that shown in the

(11)

Appendix under the assumption that

f (p) is a straight

line and that the deliverability equation is of the form


These profiles are typical of cases where the critical gas

q0

saturation, Bgc, is zero and the well is produced at a


constant rate. If Sgc

C (t) (p- PwJ t(t) , then we can show that n (t)

will be less than unity. In this case, as expected, n = 1

> 0 , and/ or the well is produced

only

at a constant wellbore pressure, then the profiles are

~f

the slope of the mobility function is zero. Note

that even if n is unity, estimates of q 0 ,maz obtained by

shallower than those shown here. The dashed lines rep-

extrapolating these straight lines will not be correct.

resent profiles predicted by the Fetkovich hypothesis.


It is clear that the Fetkovich hypothesis predicts mobil-

The results shown above substantiate Fetkovich's

ity function profiles that are much steeper than those

claim that a near wellbore effect is the probable cause

obtained via simulations.


203

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

SPE 16204

for field curves exihibiting slopes much less than unity.

observation also suggests that the correction suggested

The fact that all field experiments presented by Fetkovich

by Couto and Golan5 to incorporate the change in skin

yield values of n ::; 1 , and that simulations yield esti-

factor is unnecessary. We return to this point later in

mates of n 2: 0.9 , leads to the following observations:

the paper.

(i) For wells under solution gas drive the non Darcy flow
component should not be ignored, (ii) the value of n in
Prediction of Future Inflow Performance Curves

general will change with depletion, unless this change is


completely overshadowed by the non Darcy flow effect,
and (iii) if field experiments suggest that n

The data we have generated enables us to test the

1 , then

validity of the equations suggested by Standing 2 and

this result does not necessarily imply that non Darcy

Fetkovich 3 to predict the shift in the deliverability curve

flow is insignificant. (Of course one could argue that

with time. As mentioned in the Introduction, to our

the field results of Fetkovich may not be representative

knowledge, verification of these suggestions is yet to be

of other cases, but the evidence in Ref. 3 is convincing.

published. Figure 9 presents data that tests the validity

Similar conclusions can be derived if the results in Figs.

of equation (6), which is denoted by the unbroken line.

9 and 10 of Ref. 14 are examined).

Data for several values of NpjN and s are presented.


Results for both sets of data we have considered are

H we compare the results in Figs. 4 and 5, there

shown here. The data points reflect results obtained

is nothing in these figures to suggest that the shape of

from log -log deliverability plots similar to figures 4

the deliverability curve is influenced by the skin fac-

and 5 and estimates of f (p). The alignment of the data

tor. Figure 8 presents results we have obtained for both

points with the unbroken line is good. For purposes of

positive and negative skin factors and solutions are pre-

discussion, most of the data points are labelled. The

sented along the lines presented by Vogel 1 . The un-

nomenclature used here is as follows. The first number

broken line represents the solution of Vogel 1 and the

refers to present conditions and the second to future

dashed line represents the deliverability equation given

conditions. Thus, 1

by Fetkovich 3 (n = 1). Figures a,b, and c present infor-

2 represents values at pressure

levels 1 and 2, respectively and 2

mation for several values of skin factor for fixed values of

4 represents values

and pressure level 2 and 4, respectively. The symbol

Np/N (set 1 data), and fig. d presents data for several

values of Np/N, for a fixed value of s (set 2 data). The

4 implies that data at pressure level 2 were used

to predict the value of q0 ,maz at pressure level 4. The

results shown here and in Figs. 4 and 5 are intended to

results given here suggest that if data are used to pre-

demonstrate that the shape of the deliverability curve

dict responses over a short time range, then Standing's

is essentially unchanged whether results are presented

method should provide an adequate engineering approx-

along the lines presented by Fetkovich (log -log plots)

imation to the actual value of q 0 ,maz,f For example,

or Vogel (Cartesian plots). The results shown in Figs.

consider the points 4

4, 5, and 8 clearly show that there is no basis for the

for Set 1 data. Points 4

contention in Refs. 1 and 8 that the deliverability curve

5, 2

5 and 3

5, 3
~

the unbroken line than the point 2

changes from a quadratic form to a linear form as the

5, and 1

5 are closer to
5 and all three

of these points are closer to the unbroken line than the

skin factor increases (see discussion on this point in Ref.

point 1

15). In fact these results show that equations (1) and


(3) can be used to analyze data if s

5. Thus, predictions over a longer time span

are less reliable.

=f 0, provided that

we assume that the appropriate relations are valid. This


If we consider the alignment of the data points with
204

SPE 16204

R. G. CAMACHO-V and R. RAGHAVAN

the unbroken line, we note that all data points fall on

First, the information given in Figs. 9 and 10 assumes

or to the left of this line. This result implies that if

that deliverability curve is well defined; that is, n

f (p) values are estimated accurately, then estimates

known.

of qo,maz,f

will be lower than the correct values of

ing and Fetkovich provide no information on the future

Thus, values of q 0 predicted by this method

value of n . One would have to assume n to be unity

q 0 ,maz,f

will be lower than actual values of q0

is

Second, the procedures suggested by Stand-

or that n is unchanged (implicitly assumed in Refs. 2


and 3). In the following, we consider the consequences

As mentioned by Fetkovich, Standing's method requires that the kro

of assuming that n is constant (unity or unchanged) in

curve be well defined. To avoid

performance prediction.

the need for relative permeability data, based on empirical observations, Fetkovich suggested that equation

Fig. 11 presents results that compares rates pre-

(7) be used. Figure 10 presents information that can

dicted by various methods. For convenience, we also

be used to verify this hypothesis. The unbroken line in

present results obtained by the two-rate methods sug-

Fig. 10 represents equation (7). The data points reflect

gested in Refs. 4 and 6. The methods in Ref. 4 and 6 re-

information extracted from the deliverability curves for

quire data at two pressure levels. (In this work we make

several values of p . Except for the fact that results

no distinction between the two Kelkar-Cox6 methods

are presented in terms of average pressure the infor-

since the differences in rates predicted by the two meth-

mation given here is identical to that given in Fig. 9.

ods are negligible.) All results given here are normal-

The results shown here suggest that if predictions are

ized on the test rate at the first time level q0 , 1

made over a short time span, then the recommenda-

arrows denote times corresponding to the two tests).

tion of Fetkovich should yield an accurate estimate of

Predictions for the single-point methods were made us-

For example, points 1 -+ 2, 1 -+ 3, 3 -+ 4, etc.

ing Test 1 data. Since these methods have never been

qo,maz,/ .

(the

are much closer to the unbroken line than points 1 -+ 5

tested when

or 2-+ 5 (the nomenclature used here is identical to that

and are typical of the results obtained in this study.

used in Fig. 9). On the basis of the results shown here

In general for constant pressure production, we have

we can conclude that Fetkovich's suggestion will yield

found that the Vogel-Standing (n

estimates of q 0 ,maz,f that are good approximations to

dicts rates that are lower than the Fetkovich method

the actual values of qo,maz,f provided extrapolations

(n

are conducted over a short time range. From a prac-

in most cases the Fetkovich method predicts rates es-

tical viewpoint, this method will provide estimates as

sentially identical to the two-test methods of Uhri and

accurate as the method suggested by equation (6), al-

Blount 4 or Kelkar and Cox6 , although the former only

though the procedure recommended by Standing (Fig.

requires one data point. Unfortunately for the specific

9) will result in more consistent predictions. Based on

case considered here errors using the Couto-Golan 5 pro-

the information given above, we can conclude that the

cedure are exceedingly high. Rate predictions using Eq.

suggestions of Standing or Fetkovich can be used to pre-

(4) and the constant gas-oil ratio assumption suggested

dict future performance.

by Levine and Prats 16 are also shown. The gas-oil ratio

= 1).

=f. 0 , results shown here are for

= 20

1.24) method pre-

Although not of direct interest to this work,

for each specific average pressure was obtained from the


Care should be taken in applying the information

Muskat 17 materials balance equation. Unlike the other

given in Fig. 9 and 10 to field data particularly if sin-

methods 2 -

gle point tests are used to predict future performance.

well after the onset of boundary-dominated flow and in


205

6,

this procedure predicts actual rates very

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

most of the cases it yields the best results. A similar


Prats 16

SPE 16204

methods and single-point methods are for engineering

for

purposes equally good, (ii) predictions based on two-

case. Note that this approach does not re-

test methods are, in general, better than single-point

quire rate or pressure measurements, however, relative

methods and predictions becomes more accurate if the

permeability and fluid property data are needed. (The

gas saturation in the reservoir is greater than the critical

method we have used is slightly different from that used

gas saturation, (iii) in general, Standing's method pre-

in Ref. 16. The authors in Ref. 16 assumed that the

dicts rates that are lower than all other predictions, (iv)

pressure and the gas-oil ratio obtained from the dif-

in many instances, Fetkovich's method predicts rates as

ferential materials balance equation corresponds to the

accurate as the two-test methods, and (v) predictions

values at the outer boundary. In Ref. 13, we show that

based on the Muskat materials balance equation and

results from the differential materials balance equation

the assumption that the gas-oil ratio is independent of

correspond to the average reservoir pressure and hence

distance are reliable.

conclusion was also reached by Levine and


the s

=0

in this work we assume the gas-oil ratio given by the


Muskat materials balance corresponds to the average

Although, the results presented in Figs. 11 and 12

reservoir pressure, p . In some cases, these differences

and other computations not shown here suggests that

can be significant.)

future production rates can be predicted with reasonable accuracy, one cannot conclude that th.e mobility

I (p) , can be predicted by any of these methFig. 13 we plot the mobility function, I (p) ,

function,

An unfortunate aspect of the prediction methods


given in the literature is that it is not possible to de-

ods. In

rive definite conclusions regarding the accuracy of these

predicted from some of the methods discussed in the

methods. If test conditions are different, then the differ-

literature at a given instant in time. The results shown

ences in rate predictions can be significant. This point

here demonstrate the point we made earlier: the success

is demonstrated in Fig. 12 where all conditions except

of these methods in predicting future production rates

the Test conditions are different. In this particular case,

does not require that the

the two-point methods are superior to the single-point

curately. The principal reason for the inability of these

methods. Nevertheless in this case also, the single-point

methods to predict the mobility function is that the ex-

methods are reasonably accurate if short time spans are

pression for J given by the right hand side of equation

considered. This example emphasizes the need to obtain

(5) is, in general, not valid if it is assumed that the mo-

data at regular intervals and not rely on extrapolations

bility function is a straight line. Note however, that the

over long time ranges. The prediction methods cited

constant gas-oil ratio assumption predicts the mobility

above can be used only for constant pressure produc-

function fairly well. In summary, neither the value of

tion. We have not considered production at a constant

I (p)

rate because this mode of production is not pertinent

any instant in time can be predicted on the basis of test

to rate predictions. For the constant rate case, only the

data and the assumption that

constant gas-oil ratio method yields consistent results.

of pressure.

I (p)

function be known ac-

as function of pressure nor the value of

I (p)

I (p)

at

is a linear function

In summary, based on the results in Figs. 11 and

The results presented in Figs. 11 and 12 and in

12 and others we conclude the following: (i) all methods

the literature assume that the well will be produced

predict rates that are reasonably accurate for times im-

under conditions identical to test conditions. This as-

mediately following test times; in this regard two-test

sumption may not be always true. Since the exponent


206

R. G. CAMACHO-V and R. RAGHAVAN

SPE 16204

n can be different from unity ,and since it also changes

lation:

with time, additional errors can be introduced if test

/.v) (

FE= ( 1 + Vp:0 1
1- P:0f/P)
(1 + VPwJ/P) {1- PwJ/P) .

conditions are widely different from actual producing


conditions. Figure 14 presents results that address this

P:0 1 is the well pressure when

values of J' obtained from three points on a deliverabil-

= 1.

0 . Note that this

commonly used in gas wells, where the quadratic form


of the deliverability equation is usually valid. For this

using data at the highest rate. Curve 2 presents predic-

definition of flow efficiency, the flow rate when FE =/;1

tions using an intermediate rate and Curve 3 represents

J;

definition of flow efficiency is consistent with definitions

The actual

value of n was 1.362. Curve 1 represents predictions

predictions corresponding to the

12

( )

Here V takes on values 0.8 (Vogel) or 1 {Fetkovich) and

question. Here we compare predictions based on three

ity curve under the assumption that n

can be obtained by the following relation:

value at the low-

est rate. Differences in the predictions are significant

{13)

and these differences are a result of using different values for qo,maz,p If it is the intent of the operator to

Figure 15 presents solutions of PwJ/P vs. q 0 fq:,!:,=;.~

produce the well at conditions similar to that of Test

The unbroken lines are solutions for the V

2, then the value of

J;

0.8 case

{Vogel's relationshlp of the deliverability curve) when

corresponding to Test 2 should

equation {12) is used to define flow efficiency and the

be used in predicting future production rates. Simula-

dashed lines are solutions when Standing's definition of

tor results corresponding to well conditions identical to

flow efficiency is used. The curves are distinct for all val-

that in Test 2 are also presented in Fig. 14. From these

ues of

results it is clear that it is preferable to test wells at

or FE, except for FE= 1. For the FE= 1.2

and 1.6 cases, as indicated by Brown, the dashed line

conditions that reflect actual producing conditions.

bends over and indicates that flow rate decreases as Pwf


is lowered. This incorrect behavior is avoided if equa-

Modifications to IPR curve to incorporate

tion {12) is used. It is also interesting that all unbroken

changes in skin factor

curves predict that q 0 fq:,:a=:.~ = FE when PwJ = 0 .


This is the expected result. However, none of the dashed

Since Vogel's 1 deliverability equation was developed for

lines reflect this result, and is further confirmation that

= 0 , Standing 8 suggested a simple method

the liquid flow definition of flow efficiency should not be

to incorporate the effect of a change in skin factor. His

used. Similar results are obtained when the Fetkovich

method is based on the assumption that the definition

deliverability equation is used. The above observations

of flow efficiency for single phase flow will also be valid

are also applicable to the results presented in Ref. 18.

for solution gas drive. As mentioned earlier, Brown7


and Whitson9 have shown that inconsistent results are

At this stage, two points regarding the use of Stand-

obtained when this method is used if sis less than zero.

ing's method deserve clarification. For purposes of dis-

The deliverablity plots shown in Figs. 4, 5, and 8 con-

cussion let us assume that Standing's definition of flow

clusively suggest that quadratic forms of the deliverabil-

efficiency is correct. In this case it can be shown that

ity curves suggested by Vogel and Fetkovich also apply


if

=I 0

the point at which q0

. Thus, these results imply that the definition

PwJ/P

of flow efficiency must also reflect the quadratic form

begins to decrease is given by

(FE -1.125) /FE. If we examine this rela-

tionship, it is clear that problems with this approach

of the deliverability equation. Thus, in the following

will arise if FE > 1.125 . It can also be shown that

discussion we define flow efficiency by the following re-

the ideal well response, P:OJ , will be less than zero,


207

SPE 16204

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

10

formation in Column 2. Differences between the results

when Pwt/P <(FE -1) /FE. Again, this relationship


suggests that no problems will be evident, if FE

<

obtained from the deliverability curve and the equations

(damaged wells). Note that the point at which the flow

of Vogel and Fetkovich reflect the influence of the ex-

rate is a maximum is not identical to the point at which

ponent, n . Vogel's method predicts larger values of

p~ 1

0 . It should also be noted that computations

qo,maz than Fetkovich's method, since Vogel's equation

> 1 and Fetkovich 's procedure is used

which suggest that p~ 1 ~ 0 , are consistent with the

assumes that n

van Everdingen19 -Hurst 20 concept of skin factor.

with the assumption that n = 1 . If we now compute


q~~az/ q~~:_~

for a specific data set, then this ratio

Fig. 16 compares responses for both the Vogel and

will be equal to the appropriate value of FE (Columns

Fetkovich deliverability equations under the assumption

4 and 5). These results confirm that the definition of

that the quadratic form of the definition of flow effi-

flow efficiency given by equation (13) is correct.

ciency is valid. The unbroken lines are the Vogel solutions and the dashed lines are solutions corresponding

On the basis of the results presented in Figs. 15

to the Fetkovich deliverability curve. In this case all

and 16 and Table 2, we can again conclude that the

curves indicate that q 0 fq~~:_~


0

general shape of the deliverability curve is unchanged if

behavior or

s =/= 0 , and there is no basis for the conclusion that the

FE , when Pw!

and more importantly the incorrect

curve will be linear if s

non physical result is not evident. (Actually, the sug-

>0

1 8 ,18

gestion of Brown 7 , that a log -log plot be used to determine q0 ,maz when incorrect behavior is evident, tac-

Results in Fig.

17 demonstrate the effect of us-

itly assumes the quadratic definition of flow efficiency;

ing the quadratic definitions of flow efficiency. The test

however, the procedure recommended by him must be

conditions are noted on the figure; also noted are the

modified since he assumes that the liquid definition of

values of q:,~:_~. As indicated by Brown, the results ob-

flow efficiency is valid prior to the incorrect behavior

tained by Standing's method suggests that the flow rate

[see Figs. 2.22 and 2.23 of Ref.7]. Also, the solutions in

will decrease as the wellbore pressure approach zero. If

Fig. 2.25 of Ref. 7 do not show that q0 / q~~:_~

= FE

the quadratic definition of flow efficiency is used, then

at Pwf = 0 .)

the maximum flow rate is obtained, as expected, when


Pw 1

=0

. Also shown in this figure is the response pre-

Results summarized in Table 2 demonstrate the ad-

dicted by Harrison's equation (see Eq.2.39 of Ref. [7]).

vantages of our recommendations. The log-log deliver-

At lower pressures both the Standing and the Harrison

ability curves generated in the course of this study were

methods, which are based on the definition of flow ef-

used to obtain the results given here. Three data sets

ficiency for single phase flow, predict rates much lower

are considered in Table 2. Columns 1 and 2 list val-

than the responses predicted by the quadratic definition

ues of the skin factor, s, and the pressure ratio, Pwt/fi,

of flow efficiency and these differences are significant.

for each value of the skin factor.


qo,maz

Column 3 presents

obtained from the deliverability plots (values


CONCLUSIONS

in parenthesis in this column are the exponent, n .)


Columns 4 and 5, present appropriate values of flow

A finite difference model was used to evaluate in-

efficiency, that are computed using results in Column

How performance relationships in solution gas drive

2. Columns 6 and 7 present values of q~~az obtained

reservoirs for a wide range of producing conditions. The

from the Vogel and the Fetkovich relations using the in-

effects of pressure level, well condition and production


208

SPE 16204

R. G. CAMACHO-V and R. RAGHAVAN

mode were studied. The consequences of using the

Vogel 1

11

3. The suggestions by Standing and Fetkovich can

and Fetkovich deliverability equations (under identical

be used to predict future IPR curves provided that ex-

conditions) are examined. Many of the methods pro-

trapolations over long time spans are avoided. Predic-

posed in the literature for analyzing IPR curves have

tions using single-point tests will be as accurate as pre-

been derived by extending ideas based on single phase

dictions based on two-point tests. If single-point tests

flow. Although these extensions demonstrate consider-

are used to predict future production rates then test

able insight regarding the mechanics of fluid flow under

conditions should reflect actual producing conditions.

solution gas drive, inconsistences exist particularly re-

Under these circumstances, errors in future production

garding the influence of the skin factor, the value of the

rates will be primarily a function of the error in the

exponent of the deliverablity curve, and the shape of

initial estimate of the exponent, n .

the mobility function

I (p) . One of the objectives of

this work is to attempt to resolve some of these incon-

4. The method of prediction that uses the Muskat

I (p)

sistencies by examining well performance in a rigorous

materials balance equation to compute the function

manner and to lay a foundation for furthur work. For

and the constant gas-oil ratio assumption, always yields

example, we have shown that the skin factor does not

reliable results for boundary-dominated flow for all val-

influence the shape of the deliverability curve and that

ues of s . This conclusion is an extension of the results in

problems encountered in modifying the IPR curve to

Ref. 16 to skin factors different from zero. This proce-

study the effect of ~ change in the skin factor are a re-

dure requires relative permeability and PVT data, but

sult of ignoring the shape of the deliverability curve.

it does not require any data point as the other meth-

On the basis of this work, the following conclusions are

ods and thus concerns pertaining to extrapolations over

warranted:

long time spans are not pertinent.

The exponent, n , of the deliverability curve

5. Problems that arise in applying Standing's mod-

is a function of time or pressure level and changes in

ification to the IPR proposed by Vogel to incorporate

the exponent are not monotonic with time. In most

changes in the skin factor can be avoided if flow effi-

of the simulations we have conducted the exponent is

ciency is defined after consideration is given to the fac-

greater than unity. Even for early stages of depletion the

tors that govern the pressure distribution in solution

exponent can be different from unity.Since many of the

gas drive reservoirs; that is, the quadratic form of the

synthetic deliverability curves predict exponents greater

deliverability equation must be used to define flow effi.-

than unity (this conclusion can also be reached on the

ciency,and this definition should be used for both posi-

basis of Vogel's work) and field results suggest that n <

tive and negative skin factors. Differences in predicted

1 , it appears that non-Darcy flow is the norm rather

rates by the Vogel and Fetkovich (n = 1) relations do

than the exception in solution gas drive reservoirs.

not appear to be significant.

1.

2. The shape of the deliverability curve does not


change from a quadratic form to a linear form if the

NOMENCLATURE

well is damaged. Methods proposed in the literature


that assume that the shape of the curve is a function of

A= drainage area, ft 2

skin factor are incorrect.

Bo =oil formation volume factor (RB/STB)


209

12

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

C A = geometric shape factor

r = radial distance, feet

I (p) =mobility function (Eq. (4))

r e = external drainage radius, feet

FE = flow efficiency

rw = wellbore radius, feet

h = formation thickness, feet

R = producing gas-oil ratio, SCF /STB

J =productivity index, defined by Eq. (5)

R 8 = solution gas-oil ratio, SCF /STB

J' =stabilized productivity index STB/D/(psi 2 )n

s = skin factor

J* = productivity index at zero drawdown,

So = oil saturation

SPE 16204

STB/D/psi
Bgc

= critical gas saturation

Swi

= initial water saturation

k = absolute permeability, md

kro = relative permeability to oil.

t =time, days
m(p)

pseudopressure

function

(Eq.(A9));

(psifcp RB/STB)

tvA = dimensionless time based on A and initial

conditions.
n = exponent of inflow performance curve

V =constant (0.8 or 1), see Eq. (12).


N = initial oil in place, STB
Vp =pore volume MMbbl

N p = cumulative oil produced, STB


Z = gas compressibility factor.

p = pressure, psi
'Y = Euler's constant, 0.57721. ..

Pi = initial pressure, psi


J. 0
Pwf

= oil viscosity, cp

= wellbore flowing pressure, psi


=porosity

p = average reservoir pressure, psi


Subscripts
q 0 = oil flow rate (STB/D)

I = future conditions
q 0 ,ma:z: = maximum oil rate (STB/D)

p = present conditions
210

R. G. CAMACHO-V and R. RAGHAVAN

SPE 16204

Annual Technical Conference and Exhibition of the So-

initial conditions

ciety of Petroleum Engineers held in Las Vegas, Nevada,


Sept. 22-25, 1985.
ACKNOWLEDGEMENTS
7. Brown, K. E.: The Technology of Artificial Lift
We thank the advice, suggestions and comments

Methods, Vol. 4, Penn Well Publishing Company, Tulsa,

of Messers. M.J. Fetkovich, M.D. Bradley, D. Epps,

Oklahoma, 1984, pages 13-21.

S. Joshi and T. Thrasher. We acknowledge the assistance of our colleague Dr. Kermit E. Brown who has

8. Standing, M. B.: "Inflow Performance Relation-

patiently explained the significance of this topic to one

ships for Damaged Wells Producing by Solution Gas

of us (R.R.) for the past ten years.

Drive", J. Pet. Tech. (Nov., 1970) 1399-1400.

9.

REFERENCES

Whitson, C.S.: "Reservoir Well Performance

and Predicting Deliverability", paper SPE 12518 - submitted for publication.

1. Vogel, J. V.:"Inflow Performance Relationships


for Solution Gas Drive Wells", J. Pet. Tech. (Jan. 1968),

10. Camacho-V., R. G.: Well Performance Under

83-92.

Solution Gas Drive, Ph.D. Dissertation, University of


Tulsa, 1987.

2. Standing, M.B.: "Concerning the Calculation


of Inflow Performance of Wells Producing Solution Gas

11. Camacho-V., R. G. and Raghavan, R.:

Drive Reservoirs", J. Pet. Tech. (Sept. 1971)

"Performance of Wells in Solution Gas Drive Reser-

1141-1142.

voirs under Transient Flow", to be submitted to SPE


of AIME.

3. Fetkovich, M. J.: "The Isochronal Testing of Oil


Wells", Paper SPE 4529 presented at the 48th Annual

12. Handy, L.L.: "Effect of Local High Gas Satu-

Fall Meeting, Las Vegas, Nevada, Sept. 30- October 3,

rations on Productivity Indices", Drilling and Prod.

1973. (SPE Reprint Series No. 14, 265.)

Practice , API (1957) 111-122.


4.

Uhri, D.C. and Blount, E.M.: "Pivot Point


13. Camacho-V., R. G. and Raghavan, R.: "Some

Method Quickly Predicts Well Performance",

Theoretical Results useful in analyzing the Well Perfor-

World Oil, (May 1982) 153-164.

mance under Solution Gas Drive", to be submitted to


SPE of AIME.

5. Dias-Couto, L.E. and Golan M.: "General Inflow Performance Relationship for Solution-Gas Reser-

14. Millikan, C.V. and Sidwell, C.V.: "Bottom-

voir Wells", J. Pet. Tech. (February 1982) 285-288.

hole Pressures in Oil Wells", Petroleum Development


and Technology 1931 Petroleum Division A.I.M.E.

6. Kelkar, B.G. and Cox, R.: "Unified Relationship


to Predict Future IPR Curves for Solution Gas-Drive
Reservoirs", paper SPE

142~9

15. Raghavan, R.: "Well Test Analysis for Multi-

presented at the 60th


211

13

14

INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION GAS DRIVE RESERVOIRS

phase Flow", SPE 14098, presented at the International

SPE 16204

et. al. present a similar analysis for gas wells.

Meeting on Petroleum Engineering, March 17-20, 1986,


Beijing, China.

Let the mobility function be given by:

f (p) = a 1p + b'.

16. Levine, J.S. and Prats, M.: "The Calculated


Performance of Solution- Gas Drive Reservoirs",

(A1)

then equation (4} can be written as:

Soc. Pet. Eng. J. (Sept. 1961) 222, 142-152.

17.

Muskat, M.:

"The Production Histories of

Oil Producing Gas-Drive Reservoirs", Jour. of Applied

where a , and b can be functions of time but independent of

(p P!J) .
2

Physics (March 1945) 147-159.


If we follow Fetkovich3 the deliverability curve is

18. Nind, T.E.W.: "Skin Factor and Flow Effi-

given by:

ciency in Wells Producing Under Solution Gas Drive",

(A3)

Petroleum Society of CIM, Paper No. 85-36-9 presented


at the 36th Annual Technical Meeting of the Petroleum
Society of CIM held in Edmonton, June 2-5, 1985.

Combining equations (A2) and (A3), and assuming C (t)

(p P!J), we obtain:
2

and n (t) to be independent of

n(t) = [a +b dt- Pwt))] [a+ b/ (p + PwJ)]-

19. van Everdingen, A. F.: "The Skin Effect and

d P2

Its Influence on the Productive Capacity of a Well",

P!f

(A4)

Trans., AIME (1953) 198, 171-176.

Equation (A4) suggests that n will not remain constant


20. Hurst, W.: "Establishment of the Skin Effect
and Its Impediment to Fluid Flow Into a Wellbore",

for isochronal curves if the mobility function is given by


equation ( A1).

Pet. Eng. (Oct. r1953) 25, B-6.


In the following discussion we will first assume that
21.

Carter, R.D., Miller, S.C. and Riley, H.G.:

"Determination of Stabilized Gas Well Performance

> 0. Equation (A4) suggests that if b = 0 (i.e.

f(O) = 0), then n = 1.

From Short Flow Tests", J. Pet. Tech. (June 1963) 651We also have:

658.

-1

d (p- Pwf) _
_
) d ( p2 -p2
wf

APPENDIX

_
P + Pw f

+ (p -

Pw f)

1+

2
d_
~
dpwf

-1

(A5)
From equation (A4) we note that the value of n will be

DISCUSSION ON THE VALUE OF THE EXPO-

dictated by one of the following inequalities:

NENT OF THE DELIVERABILITY CURVE


d(p- Pw!)
d

In this section we examine the value of the exponent, n , when the mobility function
line, and

f (p) is a straight

f (p) is independent of pressure. Carter

(p

P!f)

d (p- Pw!)

21

d
212

(p

<

P!f)

>

1
P+Pwf

1
P+Pwf

(A6)

(A7)

] ]

R. G. CAMACHO-V and R. RAGHAVAN

SPE 16204

15

If neither of these inequalities is valid then n = 1 . The

If a= 0 , then for fixed value of p, equations (A4)

inequality given in equation (A6) will apply if the well

and ( A5) suggest that n 2:: 1 . This result is well known

is produced at a constant pressure or if dp/dPwf = 1 .

and this behavior also results when data that should be

(Note that dp/dPwJ

approximated by the liquid flow equations is plotted in

1 for single phase liquid flow,

constant rate production and that dp/dPwf

1 for

terms of

the production of real gas under the assumption that

p,Z is proportional to p ). Inequality (A7) will apply


if dp/dPwf

< 1 or if we study the shape of the deliver-

ability curve for fixed values of p.

Case A (equation (A6) is valid):

Combining inequality (A6) and equation (A4) we


have:

i)- if b > 0 , then 0

< n < 1,

ii)- if b < 0 , then there are five possibilities. Only


two are of interest to this work which yield values of n
which have been observed in practice.

If

__,1'---b1'-P+Pwf

<a

then

if

> 1,

then

and

O<n<l.

Case B (equation (A7) is valid):

Combining equation (A4) and inequality (A7), we


have:

i)- If b > 0 then n

>

1.

ii)- If b < 0 five possibilities exist. We will restrict our attention to the cases which yield values of n
observed in practice.

If

and

then

if

lbl
P+ Pwf

>a
'

then

0<n

< 1,

n>l.
213

(p

P!f) .

TABLE 1
RESERVOIR PROPERTIES USED IN SIMULATION

Set 1
Drainage Radius, re, Feet

Set 2

656.17

1000

Porosity, $, Fraction

0.3

0.119

Pore Volume, Vp' MMbbl

1.124

1.665

Permeability, k, md

10

Well Radius, rw' Feet

0.32808

Initial Pressure, pi, psi

1500

0.3

W1

Initial System Compressibility, cti' psi


~oi'

0.5

5704.78

initial Water Saturation, S .

Initial Oil Viscosity,

10

-1

cp

1. 085xl0
0.298

Thickness, h, Feet

15.55

0.0
-5

1.334xl0

-4

1. 7645
25

TABLE 2
EFFECT OF DEFINITION OF FLOW EFFICIENCY ON FLOW RATE PREDICTION
Maximum Production
FE
Rate, qo,max
Vogel
Fetkovich

Skin
Factor

Pressure
Ratio,
pwf/p

0
-1
11.51

0.8993
0.9174
0.6619

440(0.93)
530(0.93)
151(0. 964)

1.0
1.2082
0.3348

1.0
1. 2068
0.3404

508.37
614.21
170. 20

460.23
555.44
156.64

0
-1
11.51

0.7942
0.8273
0.2822

238(0.986)
275(0.984)
102.6(1.05)

1.0
1. 1726
0.3825

1.0
1. 1700
0. 4011

261.52
306.66
100.02

238.36
278.87
95.61

0
-1

0.3130
0.4444

100. ( 1. 04)
131(1.204)

1.
1. 1405

1.
1.1240

102.44
116.83

97.56
109.65

Maximum
Rate,qo,max
log-log

Flow Efficiency
(FE)
Vogel Fetkovich

Set 1, rDe = 2000


Flow Rate for all Computations; q

88 STB/D

214

SPE 1 6 2 0 4.
SET 2

fLg X400(cp)

1.1

FIG. 2 - FLUID PROPERTY DATA, SET 2

FIG. 1 - FLUID PROPERTY DATA, SET 1

- - S0 r=Sgc =O,Swi =0.3

>=
..._

SET I
s = 11.51

S0 r=0.234,Sgc=0.07
rt)

4
Q 10

.....J

co

.....

<(

C\J~
0.

Wo
~~0.6
wt,

C\J

a..

a...
w~

10.

10 3

n=I.05

>

~
.....J

0:::

0.2

0.2
0.4
0.6
0.8
OIL SATURATION, S0

FIG. 4 - LOG-LOG DELIVERABILITY PLOTS

FIG. 3 - RELATIVE PERMEABILITY DATA

Np/N
I 0.001

2 0.026
3 0.066
4 0.093
5 0.108

rt)Q
X

nl.2;#!

~09~
n=0.984

SET I
s =-I

10
q 0 ,STB/D
FIG. 5 - LOG-LOG DELIVERABIJ.ITY PLOTS

215

L
I

Np/N

SET I

s= 0, r0 e=500, q 0 =80 STB/D

I 0.001

(.)

I ~ - - FETKOVICH, f(p) =app.

LL

c;;

>!::: 5:0.6

Q.
~

....J-

:Eel.

0;:

~/

ffi-

Q.
I

0~

IQ.

0.4

-n=l

2 0.026
3 0.066
4 0.093
5 0.108

LLI
N

:::i

<(

~o/

7/~
/ ~
o/

o/

Q)

:E

a:

z
qo,STB/010
FIG. 7 - LOG-LOG DELIVERABILITY PLOT - 6p MODE OF PLOTTING
FIG. 6 - MOBILITY FUNCTION PROFILES AS A FUNCTION OF TIME

1\)

"'

a,b,c

(b)

0.8~

0.8
lc:l..

......

0.6

Jo.4
0.2
00

IQ..

......

A~o

"i
Q.. 0.4

0.4

0.6

'~

0
0
A

0:~ .~ ~~~0.0~6

Np/N =0.001
0.2

0.6

0.8

0.2

qo lqo,max

0.4

0.6

-VOGEL EON.
- - FETKOVICH EON.
(n =I)
O.B

-.
~

0.8
IQ..

......

"i

Q..

Np /N = 0.108
0.2

0.4

0.6

0.6
0.4
0.2

O.B

qo /qo, max

0
I

.. ...
~

~~

0.4

.!E!:...

20

..........
Q.

3.7, 7.3
3.7

1SET I

2.8

SET 2

STANDING'S
EQUATION

10

0[)4

I~;;{ooJ3-5
en

Q.

-c

IQ.

....s._
0.6

I Q.

IQ.

Np/N

r 0 e=2000
s=20
ros=2.8
0.2

o 11.51
.0. -1

,...._..

qo lqo,mox

r-r-rr0

0
11.51
-1

~-

..

-VOGEL EON.
- - FETKOVICH EON.
O.B

qo lqo,mox

FIG, 8 - DELIVERABILITY PLOTS - CARTESIAN COORDINATES

rri

0.0037
0.038
0.078
0.105

(n =I)

1-'

aI

[\.)

lo-

10

qo,max,p / Qo,max,f
FIG. 9 - PREDICTION OF FUTURE PERFORMANCE - THE STANDING METHOD

102

0
~

0
D

11.51 3.7, 7.3


-1
3.7
20
2.8

JSET I

!:;.

.,_

"

It) Q.

10.

TEST 2= toA = 0.122, Pwflp = 0.5376

0
0'"

SET 2

FETKOVICH'S
EQUATION

10.

SET 2, roe= 2000, s =20, r 05 = 2.8, Pwf /pi= 0.5333


TEST 1: foA = 0.058, Pwf /p = 0.5354

45

a:

10

:::i
<(
~

a:

1
1
10-l

, , ,,,,,,/
I

, , ,,,,,,
10

, , ,,,,,,
__

qo,2 lqo, 1 =0.9838

'
:;~

0.50

Qo,max,p /Qo,max,f

DIMENSIONLESS TIME, t 0A

FIG. 10 - PREDICTION OF FUTURE PERFORMANCE - THE FETKOVICH METHOD


FIG. 11 - PREDICTION OF RATE RESPONSES - COMPARISON OF VARIOUS APPROACHES

..,.
"'

SET I, s = 0, roe= 2000


toA = 10.27, Pwf/Pj = 0.5941

1.1 r--r----r----,r--.--~-~--

SET 2, r 0 e = 2000, s= 20, r 05 = 2.8, Pwf /pi =0.5333


TEST I: t 0 A =0.2903, PwfiP = 0.5429
TEST2= t 0 A =0.5230, Pwf /p = 0.5498

w
~

a:
0

:J

STANDING
FETKOVICH
PIVOT-POINT

<(
~

a:
0

q0,2/q0,1 = 0.9659

z
0

(I')

..-

DIMENSIONLESS TIME, t DA
FIG. 12 - PREDICTION OF RATE RESPONSES - INFLUENCE OF TEST CONDITIONS

0. 34 L...-_ _ _....___ _ ____.__ _ ____,


0.85
0.9
0.95

NORMALIZED PRESSURE, PwtiP


FIG. 13 - MOBILITY FUNCTION PROFILE PREDICTIONS

0"-

f\.J
C>

.J;:e

1.1

.------...---~----,..--"""'T'"---r---.,-----,

10.

0.75

Q.

w
1--

a::

a::
LaJ
a::

<(

0
LaJ

!::::!
....J

::::>

C/)
C/)

Pwflp Jp,STB/D/psi 2

<(

::iE

LaJ

a::
a.

I 0.0679 1.054XI0-5
2 0.5429 8.43XI0-6
3 o.9162 1.21 x1o-s

a::

z
0.40

0'

I
2
DIMENSIONLESS TIME, tDA

0.2

0.4

0.6

\1

0.8

II

1.2

FIG. 14 - INFLUENCE OF TEST CONDITIONS ON RATE PREDICTIONS

FIG. 15 - MODIFICATIONS TO IPR CURVE - EFFECT OF THE DEFINITION OF FLOW EFFICIENCY

"'
"'
TEST DATA: q0

II)

10.

:=
Q.

0.75

Pwt= 1500 psi

'; 1500

j5 =2000 psi
FE= 2

Q.

FE= I

LaJ

Qo,max

a::

~
a::

::::>

1000

LaJ

LaJ

a::
a.
LaJ
a::

a::

::::>

C/)
C/)

LaJ

a::
a.

=200 STB/D

Q.

STBID
VOGEL
250.0
FETKOVICH 228.6
HARRISON 281.7
STANDING 285.7

en

500

....J
....J
LaJ
~

00

0 2

04

0.6

0.8

.
FE=I
FLOW RATE RATIO, Q0 /qo, max

I-A

o-

FIG. 16 - IPR CURVES USING QUADRATIC DEFINITION OF FLOW EFFICIENCY, INFLUENCE OF V

FIG. 17 - COMPARISON OF METHODS ON RATE PREDICTIONS

N
0
-};:-

Vous aimerez peut-être aussi