Professional Documents
Culture Documents
Relationships for
Damaged or Improved
Wells Producing Under
Solution-Gas Drive
Mark A. Kilns, SPE, Chevron U.S.A. Inc., and Mark W. Majcher,
SPE, Conoco Inc.
1 1
0.8 0.8
I
I
I
0.6 0.6
~
Q.
Q.
t
Q.
0.4 0.4
0.2 0.2
~
- Pr
.
= VOGEL
= 1990
.
0
Pr
Pr
Pr
= 1500
= 1000
=500
o 0
o 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
qo/(qo)max
Fig. i-Effect of bubblepoint pressure on dimensionless IPR Fig. 2-Effect of reservoir depletion on dimensionless IPR
curves. curves.
0.8 0.8 -i
0.6
.0.6
Q.
\
Q.
0.4 0.4
r
I
0.2 0.2
o API =26 o Sor = 0.20
.. API
=30 sor. = 0.30
API =45 .. Sor = 0.40
0
0 0.6 0.8
o
0.2 0.4 1 o 0.2 0.4 0.6 0.8 1
qo/(qo)max
Fig. 3-Effect of API gravity on dimensionless IPR curves. Fig. 4-Effect of ROS on dimensionless IPR curves.
ticular interest because Vogel limited him- depletion significantly affect dimensionless in lower producing rates and a straighter
selfto relatively low bubblepoint pressures IPR curves. curve. In this study, IPR curves were cal-
in the construction of his reference curve. Figs. 3 through 8 show the effects of oil culated for a wide range of skin values that
In this study, the pressure was varied from gravity, ROS, critical gas saturation, abso- included both positive (damage) and nega-
1,000 to 4,000 psia. The resulting IPR lute permeability, relative permeability ex- tive (stimulation) values, from +6 to -4.
curves showed that at lower bubblepoint ponent, and reservoir extent, respectively. Fig. 9 plots regular IPR curves for several
pressures, the curves straightened. As ex- These plots indicate that although the curves skin values. The flow rates do indeed
pected, the curves for 1,000 and 2,000 psia are not identical, they generally exhibit a decrease, and the curves straighten as well-
are similar to Vogel's reference curve; how- similar shape and also demonstrate much bore damage increases. Fig. 10 represents
ever, the higher-pressure curves are well less variance than the bubblepoint pressure the curves as dimensionless pressure vs. the
outside it. and depletion plots. Therefore, these vari- ratio of flow rate to the maximum, un-
Fig. 2 depicts dimensionless IPR curves ables have only a minor effect on calculat- damaged flow rate. This plot is similar to
for several reservoir depletion stages. As the ed, dimensionless IPR curves. Standing's,3 except it is based on skin
reservoir depletes, the curves tend to Although he investigated only one skin rather than flow efficiency. Fig. 11 shows
straighten. These plots strongly indicate that value, Vogel noticed that the inclusion of the skin effect on dimensionless IPR curves
both bubblepoint pressure and reservoir positive skin in his IPR calculations resulted when they are normalized to their individu-
1 1
0.8
0.6
Q.
0.4
0.2
Sgc .. 0.00
o Sgc = 0.05
.. Sgc = 0.10
o
o 0.2 0.4 0.6 0.8 1
Fig. 5-Effect of critical gas saturation on dimensionless IPR Fig. 6-Effect of absolute permeability on dimensionless IPR
curves. curves.
0.8 0.8
0.6 0.6
Co -
'J
Co
0.4 0.4
I
r
0.2 0.2
0 ra = 526.6
...
ra = 744.7
ra =1053.2
o 0
o 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Fig. 7-Effect of relative permeability exponent on dlmension- Fig. a-Effect of reservoir extent on dimensionless IPR curves.
less IPR curves.
al, maximum flow rates. These curves are Fig. 12 compares the two curves. Clear- no effect on dimensionless, normalized
all virtually identical, which indicates that ly, they are quite similar. Noting this, and curves, Fig. 13 indicated that redid indeed
for normalized, dimensionless plots only, that bubblepoint pressure and reservoir affect the unnormalized curves, particular-
skin has no effect. depletion playa primary role in IPR curve ly those with negative skin values. With this
shape, a new regression model was devel- in mind, an equation was developed through
Development of Empirical oped to relate dimensionless flow rate with nonlinear regression of the data that related
IPR Equations dimensionless pressure, bubblepoint pres- dimensionless flow rate not only to skin and
Because most of the IPR curves exhibit ap- sure, Pb, and depletion, Pro The resulting dimensionless pressure but also to re/ rw.
proximately the same shape, with little var- equation is This equation is given as
iance, it seemed desirable to fit the 19,492
j
data points to an empirical equation that was qo _ =M[ 1.0-0. 1225( PW )
simple and could be applied as a general (qo)max s - O Pr
reference curve for all solution-gas-drive
reservoirs. Nonlinear regression techniques
were used to determine the regression pa-
rameters and to make inferences from them.
-0.705(:~r, ............... (6) -0.8775(:~) 2l ............. (7)
Pr re
Type 1 Curves (Dimensionless, Normal- where n=(0.28+0.72 ) 0 476
ized Form). Initially, the data were fit to Pb r
where M= (_In_re-w_-__.__ _ ) .
a Vogel-type equation (second-degree poly-
nomial). The reSUlting equation is x(1.235 +O.OOlpb)' In - -0.476+s
Table 2 provides statistical information rw
~ = 1.0-0.1225( Pwj ) on each of the three equations. Eq. 4 and If a value of re is uncertain or unknown,
Vogel's equation are statistically compara-
r, . . . . . . .
(qo)max Pr the multiplier, M, may be approximated by
ble. However, Vogel's curve tends to un- 6.886/(6.886+s). This and all succeeding
-0.8775(:~ (4)
derpredict flow rate by approximately 1.8%,
with a maximum error of 54 %. The bubble-
multiplier approximations were determined
by regressing the data without considering
point-pressure/depletion relationship (Eq. 6) re/rw' Fig. 14 provides the predicted
where qo is the oil-producing rate, Pwf is yields a much better data fit, reducing the curves from this equation using the approx-
the corresponding flowing sandface pres- average absolute error 365 % and the maxi- imation. The multiplier approximations
sure, Pr is the average reservoir pressure, mum error by nearly half. Besides provid- match fairly well with the actual multipli-
and (qo)max is the maximum oil-producing ing a much-improved fit, Eq. 6 allows a ers. The maximum error when using the ap-
rate, assuming 100% drawdown. To provide simple, more universal application as a proximations is 9 %, which gives the user
a comparison and to evaluate any improve- dimensionless IPR curve. a good degree of confidence.
ments made, Eq. 4 and Vogel's equation The same type of equation was fit with
(Eq. 5) were plotted. Type 2 Curves (Dimensionless, Unnor- Vogel's curve and is given as
malized Form). In fitting the data to the
~=1.0-0.2(Pwf) dimensionless, unnormalized form (dimen-
sionless pressure vs. the ratio of flow rate qo =M[1.0-0.2(Pwf)
r ......... rl . . . . . . . .
(qo)max Pr to the theoretical, maximum, undamaged (qo)maxs=O Pr
flow rate), it seemed that a multiplier could
-0.8( :~ (5) be applied to the dimensionless, normalized
curves. Although Fig. 8 showed that re had
-0.8(:~ (8)
1500
Ii 1000 0.6
.e,
;}
CL
0.4
500
0.2
o 0.0 .. ~ -
o 1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 1.4 1.6 1.8
q 0 (atilo/d)
Fig. 9-Effect of skin on regular IPR curves. Fig. i0-Effect of skin on dimensionless, unnormalized IPR
curves.
1 1
~
0.8 0.8
I
I
Q.
.0.6 Q.
.0.6
CL
\ Q.
\
-i
0.4 0.4
~
-4
r-
0.2 0.2
o S =6 = VOGEL
S =0 = EQUATIO 4
... S = -4
0
1 0 0.2 0.4 0.6
qo/(qo)max
Fig. ii-Effect of skin on dimensionless, normalized IPR curves. Fig. i2-Comparison of Eq. 4 with Vogel's equation.
stimulation. One useful application is to qo2 6.81O+s 1 The effects of numerous reservoir and
write the equations in terms of present and and - = ........... (13) fluid properties on the calculated curves
future conditions and divide them by one qo 1 6.81O+s 2 were investigated. Bubblepoint pressure and
another. Assuming that P1 IPr is the same The bubblepoint-pressure/depletion equa- reservoir depletion had a significant effect
(an example may be a pumped-off well), Eq. tion and multiplier approximation are given on the curves. Skin and relrw had a serious
7 would reduce to below. influence on the unnormalized curves only.
re Several empirical equations were devel-
re In --0.467+s 1 oped that related dimensionless flow rate to
In --0.476+s 1
rw dimensionless pressure, skin effect, bub-
rw ...... (14)
- - " - - - - - , ..... (10) re blepoint pressure, depletion effect, and
re In --0.467+s2 relrw' These equations were tested and
In - -0.476+s2
rw proved to be statistically sound. They are
rw
suitable as general reference curves for
qo2 6.835+s 1 solution-gas-drive reservoirs. However, it
where the subscripts 1 and 2 refer to pres- and - = ........... (15) is recommended that the equations that ac-
ent and future conditions, respectively. qol 6.835+s 2 count for bubblepoint pressure and deple-
With the earlier approximations, Eq. 10 tion be the first choices because they
reduces to These forms of the equations provide use-
ful, accurate, and concise tools for the en- significantly improve prediction accuracy.
gineer to predict a new flow rate for a Results of this study also provide a means
6.886+s 1 desired skin improvement and vice-versa . of determining inflow performance of
. . . . . . . . . . . . . (11) damaged or stimulated wells based on skin
qol 6.886+s 2 effect, a property readily determined from
Conclusions
ill summary, dimensionless IPR curves were pressure-buildup data. These new reference
The same can be done for the Vogel equa- curves allow a simple, yet more accurate so-
tion and its multiplier approximation. calculated for 21 theoretical solution-gas-
drive reservoirs, each with eight different lution with broader applications than Vogel's
skin values and seven reservoir depletion original equation.
re
In - -0.492+s 1 stages. These reservoirs encompassed a
rw wide range of reservoir data, PVT proper- Nomenclature
...... (12)
re ties, and relative permeability characteris- Bo = oil FVF, bbI/STB, O/L3
In - -0.492+s 2 tics. This combination of conditions resulted Boi = initial oil FVF, bbl/STB,
rw in 1,344 IPR curves. L 3/L3
1.0
0.8
0.6 r-----L---+--'...vI-c---t-----1
a.
a.~
...0.6
0.4 r-----L---t---~~-t----.J
,t .e:
;}
..
0.2
~ 0.4
0
~
0
0.2
Fig. 13-Effect of reservoir extent on
dimensionless, unnormalized IPR curves.
0.0
0.0 0.5 1.0 1.5 2.0 2.5
q I(q )max_ o
o 0
Vogel-Type Inc.
Bubblepoint and
Vogel-Type Vogel Depletion
Eq.7 ~ Eq.9.
Average error 0.00298 0.01108 0.00141
Average absolute error 0.04040 0.04109 0.01144
Sum of squares (SSTO) 6.315 6,315 6,315
Error sum of squares (SSE) 85.0 88.9 7.8
Correlation coefficient (R2) 0.9865 0.9859 0.9988
Kilns Majcher