PETROLEUM EXPLORATION AND DEVELOPMENT

Volume 40, Issue 2, April 2013

Online English edition of the Chinese language journal

Cite this article as: PETROL. EXPLOR. DEVELOP., 2013, 40(2): 216–223. RESEARCH PAPER

Derivation of water flooding characteristic curve for high

water-cut oilfields
SONG Zhaojie1,*, LI Zhiping1, LAI Fengpeng1, LIU Gang2, GAN Huohua3
1. School of Energy Resources, China University of Geosciences, Beijing 100083, China;
2. Research Institute of Yanchang Petroleum (Group) Co. Ltd, Xi’an 710054, China;
3. China National Oil-Gas Exploration & Development Corporation, Beijing 100034, China

Abstract: The linear relationship between relative permeability ratio (Kro/Krw) and water saturation (Sw) on the semi-log coordinate in
the stage of middle water-cut is the theoretical basis for the derivation of traditional water flooding characteristic curve. However, the re-
lationship of Kro/Krw versus Sw deviates from the straight line in the high water-cut stage, which results in the upwarping of water flooding
characteristic curve. In order to accurately predict the production performance and recoverable reserves in the late development stage, the
relative permeability curves of actual cores were averaged. Furthermore, using the core data in the reference, a new expression of Kro/Krw
versus Sw was obtained by regression for the high water-cut stage. On the basis of the frontal-drive equation and the average water satura-
tion equation proposed respectively by Buckley-Leverett and Welge, a new water flooding characteristic curve was derived which is more
applicable for high water-cut oilfields. The calculation results indicate that the recoverable reserve calculated by the new approach is al-
most equal to the result of the production decline method, proving it is a practical tool in the prediction of production indexes in the late
development stage of oilfields.

Key words: water flooding characteristic curve; high water-cut; relative permeability; water saturation; production decline; recoverable
reserves

Introduction because of their distinct theoretical foundations. Generally,

Water flooding characteristic curves have a widespread ap-
plication in China because of its ability to predict recoverable
reserves based on the cumulative oil production, cumulative
water production, cumulative liquid production and water-oil
ratio of target oilfields. Type-A, type-B, type-C, and type-D
curves, among tens of water flooding characteristic curves, are
the most traditional methods that have been derived theoreti-
cally by Chen [1] and Yu [2], respectively. Besides, a new type
of water flooding characteristic curve proposed by Zhang [3] is
applicable for a wider scope. In actual application, traditional
water flooding characteristic curves are always upwarping in
the late development stage of oilfields. In order to solve this
problem, based on the principle of oil and water flow in po-
rous media, both type-A and type-B water flooding character-
istic curves were attested to exhibit a second straight line of
different slope in the late development stage of oilfields, and
the second equation was derived for both type-A and type-B
water flooding characteristic curves [4]. The recoverable re-
serve results predicted by various methods are much different
the prediction results of type-A and type-C water flooding
characteristic curves are more reliable while that of type-D
water flooding characteristic curve is much bigger [5]. The
same method should be selected and utilized in different de-
velopment stages of a target oilfield to predict the recoverable
reserves for comparison analysis [6]. Additionally, the optimal
prediction result should be in accordance with the production
performance rather than the average value of different meth-
ods.
This study focuses on the applicability of traditional water
flooding characteristic curves for the purpose of deducing a
new water flooding characteristic curve for high water-cut
oilfield. By analyzing the non-linear plot of relative perme-
ability ratio Kro/Krw versus water saturation Sw on semi-log
coordinate, a new correlation of Kro/Krw versus Sw was ob-
tained by regression in high water-cut stage, and further, a
new water flooding characteristic curve was derived that is
more applicable in high water-cut oilfield. The calculation
results have proved the convenience and applicability of the

Received date: 28 Dec. 2011; Revised date: 26 Oct. 2012.

* Corresponding author. E-mail: songzhaojie1210@163.com
Foundation item: Supported by the National Natural Science Foundation of China (51174178); National Science and Technology Major Project of China
(2011ZX05016-006).
Copyright © 2013, Research Institute of Petroleum Exploration and Development, PetroChina. Published by Elsevier BV. All rights reserved.
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

Table 1 Equations of traditional water flooding characteristic curves and relevant formulas to predict recoverable reserves
Equation of traditional water flooding
Type of water flooding characteristic curves Formula to predict recoverable reserves
characteristic curves
1⎡ f w max ⎤
Type-A water flooding characteristic curve [1] lgWp = a1 + b1 Np N pmax = ⎢ lg − a1 ⎥
b1 ⎣ 2.303b1 (1 − f w max ) ⎦
1
Type-B water flooding characteristic curve [1] lgWOR = a2 + b2 N p N pmax = [lg WORmax − a2 ]
b2
Lp 1⎡
Type-C water flooding characteristic curve [2] = a3 + b3 Lp N pmax = 1 − a3 (1 − f w max ) ⎤⎦
Np b3 ⎣
Lp 1⎡ (a4 − 1)(1 − f w max ) ⎤
Type-D water flooding characteristic curve [2] = a4 + b4Wp N pmax = ⎢1 − ⎥
Np b4 ⎣ f w max ⎦
Wp Wp a5 (1 − f w max )
ZHANG Jin-qing water flooding characteristic curve [3] = − a5 + b5 N pmax = b5 − b5
Np N p2 f w max + a5 (1 − f w max )
1 ⎡ f w max ⎤
lgWp = A1 + B1 N p N pmax = ⎢ lg − A1 ⎥
CHEN Yuan-qian two straight lines water B1 ⎣ 2.303B1 (1 − f w max ) ⎦
flooding characteristic curve [4] lgWOR = A2 + B2 Np N pmax =
1
[lg WORmax − A2 ]
B2

new approach that is beneficial to the evaluation method of

water flooding in a high water-cut oilfield.

1 Evaluation of traditional water flooding

characteristic curves
Table 1 summarizes the equations of traditional water
flooding characteristic curves and the relevant formulas for
recoverable reserves prediction. Water flooding characteristic
curves are utilized to match the straight line to predict recov-
erable reserves, therefore, it is important to determine the
initial point of the straight line precisely. It is generally con-
sidered that all the water flooding characteristic curves would
exhibit a straight line when water cut exceeds 50% [6]. For the
application standardization of water flooding characteristic
curves, the initial point of the straight line should be the same
one when this method is applied to predict recoverable re-
serves in different development stages of a target oilfield so as
to avoid the possible phenomenon that the prediction results
change from large to small [6].
The hypothesis of traditional water flooding characteristic
curves is the linear relationship of Kro/Krw versus Sw on
semi-log coordinate in the intermediate water saturation
stage [7]. The formula is
K ro
= ne − mSw (1)
K rw
The traditional water flooding characteristic curves are
generally upwarping in the late development stage of oilfields.
As Fig. 1 depicts, a steam tube method was proposed to cal-
culate and plot the water flooding characteristic curve [8]. It
makes clear that the plot is upwarping in the later stage. When
the economic limit of water-oil ratio is assumed 49, the re-
covery factor predicted by type-B water flooding characteris-
tic curves is clearly higher than the actual result. Therefore,
traditional curves are not applicable for the high water-cut
oilfield.
Fig. 1 Type-B water flooding characteristic curve in the late
development stage of a bulky sandstone reservoir in Shengli Oil-
field
2 New correlation of Kro/Krw versus Sw in high
water saturation stage
From Equation (1), the linear relationship of relative per-
meability ratio Kro/Krw versus water saturation Sw on semi-log
coordinate is very typical in the intermediate water saturation
stage. However, the plot of Kro/Krw versus Sw deviates from
the straight line in high water saturation stage [4, 9], which can
not be correctly described by Equation (1). Applying the
standardized method [10], the relative permeability data of
eight cores in a block of Beier Oilfield and five cores in Dong
14 Block of the Yushulin Oilfield were normalized as shown
in Table 2. Besides, referring to the core data in Zhuang 19
Block of Xifeng Oilfield [11], the semi-log plots of Kro/Krw
versus Sw were drawn in Fig. 2.
Fig.2 shows that the plots of Kro/Krw versus Sw using the
core data from different oilfields all deviate from the straight
line in high water saturation stage. In order to deduce the
equation of water flooding characteristic curve for high wa-
ter-cut oilfield, a new regression formula of Kro/Krw versus Sw
was proposed to match the deviated core data in high water
saturation stage. Fig. 3 and Table 3 show the regression re-
sults.

− 217 −
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

Table 2 Relative permeability data of different blocks

A block in Beier Oilfield Dong 14 Block of Yushulin Oilfield Zhuang 19 Block of Xifeng Oilfield
Relative permeability/f Relative permeability/f Relative permeability/f
Water saturation/f Water saturation/f Water saturation/f
Oil phase Water phase Oil phase Water phase Oil phase Water phase
0.622 0 1.000 0 0 0.477 9 1.000 0 0 0.367 0 1.000 0 0
0.632 8 0.871 7 0.019 2 0.488 6 0.862 4 0.005 9 0.383 1 0.883 2 0.053 7
0.643 6 0.754 2 0.027 5 0.499 4 0.737 8 0.010 9 0.399 1 0.774 8 0.072 4
0.654 3 0.647 2 0.034 0 0.510 1 0.625 6 0.015 8 0.415 2 0.674 6 0.086 3
0.665 1 0.550 2 0.039 4 0.520 8 0.525 2 0.020 5 0.431 2 0.582 5 0.097 7
0.675 9 0.462 9 0.044 3 0.531 6 0.435 9 0.025 0 0.447 3 0.498 2 0.107 6
0.686 7 0.384 8 0.048 7 0.542 3 0.357 2 0.029 5 0.463 3 0.421 6 0.116 5
0.697 4 0.315 5 0.052 8 0.553 0 0.288 4 0.033 9 0.479 4 0.352 3 0.124 5
0.708 2 0.254 7 0.056 6 0.563 7 0.228 9 0.038 3 0.495 4 0.290 2 0.131 9
0.719 0 0.201 7 0.060 2 0.574 5 0.178 1 0.042 6 0.511 5 0.235 1 0.138 8
0.729 8 0.156 3 0.063 6 0.585 2 0.135 3 0.046 8 0.527 5 0.186 6 0.145 2
0.740 5 0.117 9 0.066 8 0.595 9 0.099 8 0.051 0 0.543 6 0.144 6 0.151 4
0.751 3 0.086 0 0.069 9 0.606 7 0.071 0 0.055 2 0.559 6 0.108 7 0.157 2
0.762 1 0.060 1 0.072 9 0.617 4 0.048 3 0.059 3 0.575 7 0.078 7 0.162 7
0.772 9 0.039 8 0.075 8 0.628 1 0.031 0 0.063 4 0.591 7 0.054 2 0.168 0
0.783 6 0.024 4 0.078 5 0.638 9 0.018 3 0.067 5 0.607 8 0.034 8 0.173 1
0.794 4 0.013 4 0.081 2 0.649 6 0.009 6 0.071 5 0.623 8 0.020 3 0.178 0
0.805 2 0.006 2 0.083 8 0.660 3 0.004 2 0.075 6 0.639 9 0.010 1 0.182 7
0.816 0 0.002 1 0.086 4 0.671 0 0.001 3 0.079 6 0.655 9 0.003 8 0.187 3
0.826 7 0.000 3 0.088 8 0.681 8 0.000 2 0.083 5 0.672 0 0.000 7 0.191 7
0.837 5 0 0.091 3 0.692 5 0 0.087 5 0.688 0 0 0.196 0

Fig. 2 Plot of Kro/Krw versus Sw in different blocks Fig. 3 Matching plot of deviated data in high water saturation stage

Table 3 Regression results of the deviated data of plots of Kro/Krw versus Sw

Oilfield Block Regression formula Correlation coefficient
A Block in Beier Oilfield ln(Kro/Krw)=－876.47Sw2+1 311.5Sw－490.58 0.991 2
Zhuang 19 Block of Xifeng Oilfield ln(Kro/Krw)=－548.1Sw2+639.25Sw－187.56 0.995 0
Dong 14 Block of Yushulin Oilfield ln(Kro/Krw)=－1 160.1Sw2+1 419.7Sw－434.69 0.993 4

As shown in Fig. 3 and Table 3, applying binomial expres- be represented exactly by the binomial equation in high water
sion to match the data of ln(Kro/Krw) versus Sw enabled us to saturation stage. The binomial equation is given as
obtain a good agreement in the high water saturation stage, K
ln ro = fS w 2 + gS w + h (2)
which indicates that the relationship of Kro/Krw versus Sw can K rw

− 218 −
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

Transforming Equation (2) gives 2 (6aE + 3b) Soi

where, A = 9aSoi , B = ,
K ro 2
9.212 4.606
= d e − ( aSw + bSw ) (3)
K rw μo Boγ o aE 2 + bE
C = lg + .
where, d=eh, a=−f, b=−g. It can be seen that the new correla- d μ w Bw γ w 2.303
tion of Kro/Krw versus Sw in high water saturation stage (Equa- Substituting R = N p /N into Equation (9) obtains the rela-
tion (3)) is different from that in the intermediate water satu- tionship of WOR versus cumulative oil production Np in high
ration stage (Equation (1)). water-cut oilfield as
3 Theoretical derivation of the new water flooding lgWOR = A′N p 2 + B′N p + C (10)
characteristic curve A B
where, A′ = , B′ = .
N2 N
Based on the new regression formula of Kro/Krw versus Sw in
When the water-oil production ratio reaches the economic
high water saturation stage discussed previously, the equation
limit WOR max, the water flooding recoverable reserve can be
of water flooding characteristic curve in high water-cut oil-
computed as
field is derived theoretically in this section.
− B′ + B′2 − 4 A′(C − lg WORmax )
On the basis of the regression of relative permeability data N pmax = (11)
2 A′
in high water saturation stage, the relationship of relative
permeability ratio at the exit-end of the core Kroe/Krwe versus 4 Applications and discussions
the exit-end water saturation Swe is given as
4.1 Case of Craig F. F. Jr.
K roe 2
= d e − ( aSwe + bSwe ) (4)
K rwe In order to evaluate the applicability and feasibility of the
Ignoring gravity and capillary force, in the steady state flow new water flooding characteristic curve in a high water-cut
condition of oil and water, the relationship of relative perme- oilfield, the actual production data from References [5] and
[16] were selected to conduct calculation and result analysis.
ability ratio Kroe/Krwe versus the transient production rate of oil
Table 4 summarizes the relationship of relative permeabil-
and water at the exit-end of the core is expressed as [12]
ity of oil and water, and the fractional flow of water, i.e., wa-
K roe Qμ Bγ
= o o o o (5) ter cut, versus water saturation, while the fist four rows of
K rwe Qw μ w Bwγ w
Table 5 show water flooding performance of an example res-
Substituting Equation (4) into Equation (5) gives the corre- ervoir after water breakthrough [15]. The analytic equation of
lation of water-oil ratio WOR as: recovery degree in a water flooding reservoir was computed
Q μ Bγ 2
in Reference [1] as
WOR = w = o o o e aSwe + bSwe (6)
Qo d μ w Bw γ w N S − S wi S wa − S wi
R = p = wa = (12)
Applying the frontal-drive equation and the average water N Soi 1 − S wi
saturation equation proposed by Buckley-Leverett [13] and By applying Equation (12), the recovery degree, cumulative
Welge [14], respectively, and combining the experimental the- oil production and cumulative water production are calculated
ory of Ilflos, the relationship of the exit-end water saturation at different exit-end water saturations. In the calculation, the
versus recovery degree was presented in Reference [1] as initial water saturation is 0.1. The exit-end water saturation is
3⎛ N S ⎞ 1 3 1 0.469 and the cumulative water production is 0 at the time of
S we = ⎜ p oi + S wi ⎟ − (1 − Sor ) = ( Soi R + S wi ) − (1 − Sor ) water breakthrough.
2⎝ N ⎠ 2 2 2
(7) The plot of lgWOR versus Np is drawn in Fig. 4 using the
By logarithmic transformation of Equation (6), and substi- calculated data above. The first half of the plot is a straight
tuting Equation (7) into it, we get Table 4 Relative permeability and fractional flow of water at
⎛3 ⎞ ⎛3
2
⎞ different water saturations in an example reservoir [15]
a ⎜ Soi R + E ⎟ + b ⎜ Soi R + E ⎟
μo Boγ o
+ ⎝
2 ⎠ ⎝ 2 ⎠= Water Relative permeability /f Fractional flow
lg WOR = lg
d μ w Bwγ w 2.303 saturation /f Oil phase Water phase of water /f
9aSoi 2 2 (6aE + 3b) Soi 0.10 1.000 0 0
R + R+
9.212 4.606 0.30 0.373 0.070 0.272 9
μ Bγ aE 2 + bE 0.40 0.210 0.169 0.616 8
lg o o o + (8)
d μ w Bwγ w 2.303 0.45 0.148 0.226 0.753 3
1 0.50 0.100 0.300 0.857 1
where, E = (3S wi + Sor − 1) . The new equation of water
2 0.55 0.061 0.376 0.925 0
flooding characteristic curve in high water-cut oilfield is de- 0.60 0.033 0.476 0.966 5
rived as 0.65 0.012 0.600 0.990 1
lgWOR = AR 2 + BR + C (9) 0.70 0 0.740 1.000 0

− 219 −
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

Table 5 Water flooding performance of an example reservoir after water breakthrough

Exit-end water Fractional Water-oil Average water Recovery Cumulative oil Cumulative water
saturation /f flow of water/f ratio saturation/f degree/f production/104 m3 production/104 m3
0.469 0.798 3.950 0.563 0.514 2.284 0
0.495 0.848 5.579 0.582 0.536 2.378 0.523
0.520 0.888 7.928 0.600 0.556 2.467 1.227
0.546 0.920 11.500 0.617 0.574 2.550 2.191
0.572 0.946 17.518 0.636 0.596 2.644 3.833
0.597 0.965 27.571 0.652 0.613 2.723 6.009
0.622 0.980 49.000 0.666 0.629 2.792 9.394
0.649 0.990 99.000 0.681 0.646 2.866 16.719
0.674 0.996 249.000 0.694 0.660 2.930 32.688
0.700 1.000 ∞ 0.700 0.667 2.960
Note: Original Oil in Place is 4.440×104 m3.

enables us to get a better agreement for the upwarping data in

Fig. 4 Plot of water-oil ratio versus cumulative oil production of
an example reservoir in Reference [15]
line, while the latter half exhibits upwarp. Applying type-B
and the new water flooding characteristic curve to match the
first and latter half of the plot, respectively, regression formu-
las and the predicted recoverable reserves are presented in
Table 6. Besides, the recoverable reserves predicted by the
type-A, type-C, type-D, Zhang Jinqing and Chen Yuanqian
two-straight-line water flooding characteristic curves are
listed in Table 6 for comparison.
As seen in Fig. 4, good match is obtained by applying the
type-B water flooding characteristic curve in the first half of
the plot, while the upwarp of plot in the latter stage indicates
poor match. Applying the water flooding characteristic curve
for the high water-cut oilfield with a pattern of a parabola
the latter half of the plot. In this case, the production data cal-
culated by Equation (12) was matched by different water
flooding characteristic curves to predict recoverable reserves;
therefore, the cumulative oil production calculated by Equa-
tion (12) when water cut reaches 0.98, 2.792×104 m3, was
considered the standard value of water flooding recoverable
reserves to conduct comparison with results of different water
flooding characteristic curves. In Table 6, the recoverable
reserves predicted by the water flooding characteristic curve
in the high water-cut oilfield perfectly matches the value cal-
culated by Equation (12), which indicates the new method is
better applicable in high water-cut oilfields.
In order to conduct a theoretical analysis of the upwarping
of water flooding characteristic curve in the late development
stage of oilfields, the plot of relative permeability ratio Kro/Krw
versus water saturation Sw in this case is drawn in Fig. 5 to
conduct a comparison analysis with the upwarping of water
flooding characteristic curve in Fig. 4.
Because the relative permeability data provided by the ref-
erence and the cumulative oil production calculated by Equa-
tion (12) are scattered points, it is difficult to determine ex-
actly from which point the plots deviate from the straight line
in Fig. 4 and Fig. 5. However, what can be determined is that
the deviating point in Fig. 4 is somewhere between the cumu-
lative oil production interval of 2.550×104m3 and 2.644×

Table 6 Water flooding recoverable reserves predicted by different water flooding characteristic curves
Types of water flooding characteristic curves Regression formulas Predicted recoverable reserves/104 m3
Type-A water flooding characteristic curve lgWp=－6.399 2+2.992 5Np 2.734
Type-B water flooding characteristic curve lgWOR=－3.153 5+1.891Np 2.871
Type-C water flooding characteristic curve Lp/Np=0.257 6+0.385 6Lp 2.727
Type-D water flooding characteristic curve Lp/Np=1.042 2+0.412 4Wp 2.665
Zhang Jinqing water flooding characteristic curve Wp/Np=－0.045 3+2.422 7Wp/Np2 2.661
Chen Yuanqian two straight lines water flooding lgWOR=－3.816 6+2.175 7Np
2.841
characteristic curve (the second straight line)
Water flooding characteristic curve for high
lgWOR=3.112 5Np2－12.292Np+12.98 2.806
water-cut oilfields

− 220 −
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

Fig. 6 illustrates the first half of the plot corresponds with

Fig. 5 Plot of Kro/Krw versus Sw in Reference [15]
104m3 and the corresponding water cut interval is between
0.920 and 0.946. The deviating point in Fig. 5 is in the water
saturation interval between 0.55 and 0.60 and the correspond-
ing water cut interval is between 0.925 0 and 0.966 5. There-
fore, the two deviating points in Fig. 4 and Fig. 5 are almost
located in the same water cut interval, which indicates that the
semi-log plot of Kro/Krw versus Sw deviating from the straight
line in high water saturation stage results in the upwarping of
traditional water flooding characteristic curves in the late de-
velopment stage of oilfields. Meanwhile, the case result
proves that the hypothesis of the new water flooding charac-
teristic curve is rational.
4.2 Case of Sha 1 Reservoir in the Pucheng Oilfield
The actual production data of Sha 1 Reservoir is given in
Reference [16]. This reservoir has an oil volume factor of
1.429 9 m3/m3 and a water volume factor of 1 m3/m3; there-
fore, the cumulative oil production, cumulative water produc-
tion, cumulative liquid production and water cut on the sur-
face can be calculated (Table 7), of which the value of the
cumulative water production on the surface equals that inside
the reservoir.
Based on the calculated production data above, the plot of
lgWOR versus Np is drawn in Fig. 6 where the production data
in 1984 was not included because the water cut was just 0.180
4 and the water flooding characteristic curve has not yet ex-
hibited the characteristic of a straight line.
the traditional relationship of straight line, while the latter half
is upwarping clearly. Applying water flooding characteristic
curve in high water-cut oilfield to match the latter half of the
plot yielded the expression of water-oil ratio versus cumula-
tive oil production as
lg WOR = 0.000 023 2 N p 2 − 0.019 507 N p + 4.369 (13)
Substituting the relevant coefficient into Equation (11)
gives the water flooding recoverable reserves of 667.95×104
m3 in Sha 1 Reservoir in the Pucheng Oilfield.
There are various evaluation methods for water flooding
performance [17−20] of which the production decline method is
the most popular one. In order to verify the result of the new
approach, the expanded Arps decline method of generalized
decline types is utilized to conduct a comparison analysis. The
expanded Arps decline method was proposed by Chen Yuan-
qian in 1994 that considers the exponential and harmonic de-
clines as two special types of hyperbolic decline, and further,
was derived based on hyperbolic decline to predict the recov-
erable reserves [21−22].
From the production performance of Sha 1 Reservoir in the
Pucheng Oilfield, we can see the production started to decline
in 1986 (Fig. 7). However, for precise prediction of the recov-
erable reserves, the year of 1988 is determined as the initial
point to calculate the maximum cumulative oil production in
the selected decline period. Table 8 summarizes the produc-
tion data in the selected decline period of Sha 1 Reservoir in
the Pucheng Oilfield.
Applying linear regression trail and error method and as-
suming decline exponent as 0, we can obtain a straight line
with the correlation coefficient of 0.9999 (Fig. 8), which illus-
trates the annual oil production corresponds to an exponential
decline in the decline period. The correlation of cumulative oil
production in the selected decline period versus annual oil
production is given by regression as
N p = 141.88 − 3.069Q (14)
Reference [22] provided the time conversion coefficient of
1.0; further, we can work out the maximum cumulative oil

Table 7 Production data of Sha 1 reservoir in the Pucheng Oilfield

Cumulative oil production Cumulative water Cumulative oil production Cumulative liquid production
Year Water cut/f
(inside reservoir)/104 m3 production/104 m3 (on the surface)/104 m3 (on the surface)/104 m3
1983 210.85 10.77 147.46 158.23
1984 296.28 23.92 207.20 231.12 0.180 4
1985 443.66 60.52 310.27 370.79 0.262 0
1986 592.50 149.53 414.36 563.89 0.461 0
1987 692.23 291.27 484.11 775.38 0.670 2
1988 758.31 461.90 530.32 992.22 0.786 9
1989 808.06 657.94 565.12 1 223.06 0.849 3
1990 846.14 904.44 591.75 1 496.19 0.902 5
1991 873.79 1 181.27 611.08 1 792.35 0.934 7
1992 895.56 1 452.87 626.31 2 079.18 0.946 9
− 221 −
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

dicted by the water flooding characteristic curve, 667.95×104

m3, matches well with the result of the expanded Arps decline
method, 672.2×104 m3, which proves the practical applicabil-
ity of the new method.
However, for the production decline rules of which the de-
cline exponent does not equal 0, 0.5 or 1.0, computer program
is needed in the expanded Arps decline method to conduct a
linear regression trail and error by setting different values of
decline exponents between 0 and 1 according to a fixed step,
such as 0.01, which makes the computing complex and
Fig. 6 Plot of water-oil ratio versus cumulative oil production of time-consuming. Applying the water flooding characteristic
Sha 1 reservoir in the Pucheng Oilfield curve can avoid a complicated process of determining decline
exponent, meanwhile solve the problem of predicting recov-
erable reserves caused by the upwarping of traditional water
flooding characteristic curve in the late development stage of
oilfields.
5 Conclusions
The theoretical hypothesis of traditional water flooding
characteristic curves is the linear relationship of relative per-
meability ratio Kro/Krw versus water saturation Sw on semi-log
coordinate in the intermediate water saturation stage, so the
approaches are more applicable to predict the production in-
Fig. 7 Production performance of Sha 1 reservoir in the dex for middle water-cut oilfields. However, applying tradi-
Pucheng Oilfield tional approaches in high water-cut oilfields would make a
significant error in the predicted recoverable reserves, which
Table 8 Production data in the selected decline period of Sha 1 limits the application of the traditional approaches in high
reservoir in the Pucheng Oilfield water-cut oilfields.
Annual oil Cumulative oil The semi-log plot of Kro/Krw versus Sw deviates from the
Year t /a
production/(104 m3⋅a−1) production/104 m3 straight line in the high water saturation stage, which is the
1988 0 46.21 0 main reason of the upwarping of traditional water flooding
1989 1 34.79 34.79 characteristic curves in the late development stage of oilfields.
1990 2 26.63 61.42 The equation of water flooding characteristic curve in high
1991 3 19.34 80.76 water-cut oilfields is a binomial expression, and the pattern of
1992 4 15.22 95.99
the curve is a parabola, while the pattern of traditional water
flooding characteristic curves is a straight line. Therefore,
good agreements can be obtained by applying new approach
to match the upwarping data of water flooding characteristic
curve in the late development stage of oilfields.
When water flooding characteristic curves exhibit upwarp-
ing in the late development stage of oilfields, it is suggested to
apply new approach to predict the production index that can
solve the problem of predicting recoverable reserves caused
by the upwarping of traditional water flooding characteristic
curves in the late development stage of oilfields. The calcula-
tion results indicate that the new approach has a good practi-
cability.
Fig. 8 Linear regression trail and error result of expanded Arps
decline method
Nomenclature
production of 141.88×104 m3 in the selected decline period.
Adding the cumulative oil production of 530.32×104 m3 in the Kro, Krw—relative permeability of oil and water, f;
previous period enables us to predict the recoverable reserves Sw—water saturation, f;
of 672.2×104 m3 in Sha 1 Reservoir of Pucheng Oilfield. Wp—cumulative water production, 104 m3;
By the comparison analysis, the recoverable reserves pre- Np—cumulative oil production, 104 m3;

− 222 −
 SONG Zhaojie et al. / Petroleum Exploration and Development, 2013, 40(2): 216–223

Npmax—water flooding recoverable reserves, 104 m3; high water-cut stage. Journal of Oil and Gas Technology,
fwmax—economic limit of water cut, f; 2008, 30(2): 127−131.
WOR—water-oil ratio, f; [9] Gao Li, Song Kaoping, Ma Chunhua, et al. Two straight
WORmax—economic limit of water-oil ratio, f; lines to forecast development index of high water-cut oil-
Lp—cumulative liquid production, 104 m3; field. Petroleum Drilling Techniques, 2008, 36(5): 72−74.
a1, a2, a3, a4, a5, b1, b2, b3, b4, b5, A1, A2, B1, B2, a, b, d, f, g, h, m, n, [10] Chen Yuanqian. Standardized method of average relative
A, B, A′, B′, C, E—coefficient; permeability curve. Oil Industry Standard and Measurement,
Kroe, Krwe—relative permeability of oil and water at the exit-end, f; 1990, 6(3): 6−9.
[11] Dong Dapeng. A study of percolation rules of oil-water
Swe—exit-end water saturation, f;
two-phase flow in low-permeability reservoirs. Chengdu:
Qo, Qw—transition production rate of oil and water at the exit-end,
Chengdu University of Technology, 2007: 60−62.
m3/d;
[12] Jiang Hanqiao, Yao Jun, Jiang Ruizhong. The fundamental
μo, μw—viscosity of oil and water in reservoir, mPa•s;
and practice of reservoir engineering. Dongying: China
Bo, Bw—volume factor of oil and water in reservoir, m3/ m3;
University of Petroleum Press, 2006: 239−241.
γo, γw—gravity of oil and water in reservoir, f;
[13] Buckley S E, Leverett M C. Mechanism of fluid displace-
N—original oil in place, 104 m3;
ment in sands. Trans AIME, 1942, 146(1): 107−116.
Soi—initial oil saturation, f;
[14] Welge H J. A simplified method for computing oil recovery
Swi—initial water saturation, f; by gas or water drive. JPT, 1952, 4(4): 91−98.
Sor—residual oil saturation, f; [15] Craig F F Jr. The reservoir engineering aspects of water-
R—recovery degree, f; flooding. New York: Society of Petroleum Engineers of
Swa—average water saturation, f; AIME, 1971: 112−114.
t—time, a; [16] Gao Wenjun, Xu Bingtao, Wang Qian, et al. Using water
Q—annual oil production at time t in the decline period, 104 m3/a. drive curves for determination of piston displacement de-
gree index. Xinjiang Petroleum Geology, 2000, 21(4):
References 311−314.
[17] Ji Shuhong, Tian Changbing, Shi Chengfang, et al. New
[1] Chen Yuanqian. Derivation of relationship of water drive understanding on water-oil displacement eficiency in a high
curves. Acta Petrolei Sinica, 1985, 6(2): 69−78. water-cut stage. Petroleum Exploration and Development,
[2] Yu Qitai. Characteristic of oil-water seepage flow for sev- 2011, 38(3): 338−345.
eral important water drive curves. Acta Petrolei Sinica, [18] Li Qiaoyun, Zhang Jiqun, Deng Baorong, et al. Grey deci-
1990, 20(1): 56−60. sion-making theory in the optimization of strata series
[3] Yu Qitai. Application of Zhang’s water drive curve and its recombination programs of high water-cut oilfields. Petro-
characteristics of oil-water seepage flow. Xinjiang Petro- leum Exploration and Development, 2011, 38(4): 463−468.
leum Geology, 1998, 19(6): 507−511. [19] Zou Cunyou, Chang Yuwen, Wang Guohui, et al. Calcula-
[4] Chen Yuanqian, Tao Ziqiang. Derivation of water drive tion on a reasonable production-injection well ratio in wa-
curve at high water-cut stage and its analysis of upwarding terflooding oilfields. Petroleum Exploration and Develop-
problem. Fault-Block Oil & Gas Field, 1997, 4(3): 19−24. ment, 2011, 38(2): 211−215.
[5] Chen Yuanqian. Calculation methods of recoverable re- [20] Xu Changfu, Liu Hongxian, Qian Genbao, et al. Microcos-
serves of oilfields. Xinjiang Petroleum Geology, 2000, mic mechanisms of water-oil displacement in conglomerate
21(2): 130−137. reservoirs in Karamay Oilfield, NW China. Petroleum Ex-
[6] Chen Yuanqian. Classification, correlation and evaluation of ploration and Development, 2011, 38(6): 725−732.
water-drive performance curve analysis method. Xinjiang [21] Chen Yuanqian. Simplification of hyperbolic decline and
Petroleum Geology, 1994, 15(4): 348−356. the intercept method of determining recoverable reserves.
[7] Craft B, Hawkins M F. Applied petroleum reservoirs Natural Gas Industry, 1994, 14(4): 32−37.
engineering. New Jersey: Prentice-Hall Inc., 1991: [22] Chen Yuanqian, Li Dang. Modern petroleum reservoir en-
[8] 340−341.
Yu Bo, Sun Xinmin, Yang Yong, et al. Influential factor of gineering. Beijing: Petroleum Industry Press, 2004:
upwarping moment of water-flooding characteristic curve at 137−146.

− 223 −