Rock Stress and Earthquakes – Xie (ed.
)
© 2010 Taylor & Francis Group, London, ISBN 978-0-415-60165-8
Downloaded from http://onepetro.org/ISRMISRS/proceedings-pdf/ISRS10/All-ISRS10/ISRM-ISRS-2010-069/1716747/isrm-isrs-2010-069.pdf/1 by University of Baghdad, Ali Mahmoud on 17 June 2023
A pseudo-3D model with 2D flow of hydraulic fracture propagation
in thin interbedded sandstone reservoir
Yang Xiujuan, Wang Tongtao & Yan Xiangzhen
College of Storage & Transportation and Architectural Engineering, China University of Petroleum, Dongying,
Shandong, China
Wang Xin
CNPC Research Institute of Petroleum Exploitation and Development (Langfang Branch), Langfang, Hebei, China
ABSTRACT: During the large-scale hydraulic fracturing of thin interbedded sandstone reservoir (short for
TISR), the fracture is propagated greatly along the fracture height direction, leading the length/height ratio of
fracture usually smaller than 4. Because the interlayer is thin and weak, this can not afford strong restrictions
on the height of fracture propagations. In this condition, the traditional pseudo-3D model only considering the
fluid flowing along the length direction of fracture does not agree with the actual situation. In this study, a new
pseudo-3D model of predicting the fracture development with 2D flow is built up according to the hydraulic
fracturing characteristics of TISR. In the new model, the flow along the fracture height direction is considered.
The precise solutions of fracture length and height dimensions are calculated by the new model. The numerical
calculating software is obtained based on the new model, which is used to stimulate the fracture development
during hydraulic fracturing in TISR. The results show that the new model has a perfect performance in analyzing
the fracture development during large-scale hydraulic fracturing in TISR.
Keywords: pseudo-3D model with 2D flow; large-scale hydraulic fracturing; thin interbedded sandstone
reservoir (TISR); fracture; length/height ratio
1 INTRODUCTION In this paper, a new psudo-3D with 2D fluid flow is
developed to stimulate the fracture propagations of
Stimulation of TISR by massive hydraulic fracturing TISR during massive hydraulic fracturing. In the new
has been established as a very successful technology modal, the flow in fracture is presumed running along
for improving the petroleum production performance both the fracture length and height directions. The pre-
(XU Yun1ong et al., 2005; Dou Ranglin et al, 2001). cise solutions of pressures, fracture length and height
In the TISR, interbeds and storage layers appear by are deduced and the calculating software is complied.
alternation and their strengths are weak, leading a
weak constrain to the fracture propagation in height
2 THE PSUDO-3D WITH 2D FLUID FLOW
direction. So, during massive hydraulic fracturing, the
fracture propagates effectively in vertical and horizon-
The fracture of TISR by massive hydraulic fracturing
tal directions. To properly simulate the fracture and the
can be presumed as an elliptical fracture in infinite
reservoir, the engineers have used three-dimensional
plane according the available literatures (Qiu Weide
(3D) or pseudo-3D hydraulic fracture propagation
et al., 2002; M.M.Rahman et al., 2003; M.K.Rahman
models. At present, during the simulations of frac-
et al., 2007). In order to analyze the proposed new
ture propagations by 3D or pseudo-3D model, the fluid
model, the following assumptions are made:
flow is considered as 1D flow along the fracture length
direction, neglecting the flow along the fracture height 1) The material of storage layers is ideal elastic.
direction, e.g. Palmer pseudo-3D (Palmer ID et al., 2) The pumping rate is constant during hydraulic
1983). These 3D and pseudo-3D are suitable in cas- fracturing.
ing of fractures length/height >4, illegal for the TISR 3) The length/height of fracture is small, less than 4.
hydraulic fracturing calculating for the fracture propa- 4) The minus horizontal in-situ stress of the lower and
gation in height direction effectively.The length/height upper bed are uniform and symmetric with storage
of TISR hydraulic fracturing is usual smaller than 4. layers.
429
2.1 Computation of fracture width and stress
intensity factor
According to the methods developed by Shah and
Kobayashi (Shah, R.C et al., 1971), the fluid pressure
Downloaded from http://onepetro.org/ISRMISRS/proceedings-pdf/ISRS10/All-ISRS10/ISRM-ISRS-2010-069/1716747/isrm-isrs-2010-069.pdf/1 by University of Baghdad, Ali Mahmoud on 17 June 2023
of fracture can be written as:
Where, Aij is undetermined constant.
The geometrical dimensions and stress distribution
Figure 1. Stress on the fracture and related parameter dia-
in upper and lower bed are showed in Figure 1. By gram when stress distribution in upper and lower bed is
using assumptions (4), the fluid pressure of fracture is symmetrical.
obtained as
2.2 Computation of fluid pressure in elliptical
fracture with 2D flow
The with of fracture in Figure 1 is obtained as The fluid pressure in the elliptical fracture with a line
source can be described by Laplace equation ∇ 2 P = 0.
In this paper, the length of line source is equal to per-
forated casing length in storage layers. Close to the
well, the solution given by Muskat (Muskat, M, 1982),
which assumes a line source and a region with an infi-
nite boundary, is adopted.Away from the well and close
to the elliptical boundary, the solution is that of a point
source with an elliptical boundary at constant pressure
given by Prats (Prats, M., 1961; S. Amini, 2007). The
During the fracture propagation, the stress intensity two solutions are matched along a circle of radius rm ,
factors are expressed as where the discrepancy between the two solutions is
minimal.
In the circle of radius rm , the fluid pressure of
fracture Pfm solution given by Muskat, is obtained as
6µq
Where, λ = .
πb3
Substituting θ = 0 into equation (4), the stress inten-
sity factor of fracture end in length direction (x = Lf ),
KIL , is expressed as the following simplified forms
Where, µ is the viscosity of fracturing liquid mPa·s; q
is pump rate, m3 /min; b is average width of fracture,
m; hs is the half length of line source, m; K is constant,
Similarly, the value of θ = 90◦ and y = hf are sub- determined by matching the Muskat calculating fluid
stituted into equation (4) to give the stress intensity pressure and Prats’.
factor of fracture end in height direction Away from the circle of radius rm , the fluid pressure
of fracture Pfm solution can be written as equation (8)
according Prats method.
Where, Lf is the half length of elliptical fracture, m;
hf is the half height of elliptical fracture, m; G is rock
shear modulus, GPa; θ is the angle parameter of ellipse
to define the position of calculating point, ◦ ; C00 , C20 , Where, A, B, C and m are constants; C1 is constant
C02 are constants determined by A00 , A20 , A02 . solved by equations (7), (8) and (2).
430
2.4 Computation of effective viscosity in
non-Newton fluids
The above calculations and discussions are mainly
used to Newton fluids. The fracture fluids are usual
Downloaded from http://onepetro.org/ISRMISRS/proceedings-pdf/ISRS10/All-ISRS10/ISRM-ISRS-2010-069/1716747/isrm-isrs-2010-069.pdf/1 by University of Baghdad, Ali Mahmoud on 17 June 2023
non-Newton fluids, so the effective viscosity in non-
Newton fluids should be computed, given as
Where, R = (Lf + hf )/2, kp is denseness coefficient of
Figure 2. Filtration length calculation parameter diagram. fracture fluids, mPa·snp ; np is flow exponent parame-
ter; b0 is the maximum fracture width, m.
2.3 Computation of fracture propagation time and 3 SOLUTION OF PSEUDO-3D WITH 2D FLOW
leakoff
It is assumed that the relationship of fracture length 3.1 Computation of fluid pressure in fracture
propagation rate and time may be modeled with a The fluid pressure at x = Lf and y = hf are approxi-
power-law equation of the form (Palmer ID et al., mately the same according to the assumption of upper
1983) and lower bed stress symmetric with storage layers,
which causes the fluid pressure at y = hf , in higher
stress (S2 ) zone, to be less than the fluid pressure
at x = Lf , which in the lower stress zone (S2 ) (see
Where, t is the demanded time of fracture growing to Fig. 1). The difference of the fluid pressure at x = Lf
length Lf , min; m , n are undetermined constants. and y = hf , PLh , is the function of S2 −S1 , written as
The leakoff volume of fracture liquid is computed as
Where, ht is the thickness of storage layer, m; S1 is the
Where, ht is leakoff height, equal to the thickness of
in-situ stress of storage layer, MPa; S2 is the in-situ
storage layer, m; CL is comprehensive leakoff coef-
stress of upper and lower bed, MPa.
ficient, m/min0.5 ; Sp is leakoff volume coefficient,
The half height of fracture, hf , is assumed as given
m3 /m2 ; L is leakoff length of fracture, m. and PLh could be obtained from equation (15). In
The leakoff length of fracture L in equation (10) is the same time, the stress intensity factor of frac-
obtained by equation (11) (see Fig. 2). ture ends (x = Lf and y = hf ) are satisfied with the
boundary condition, KI |θ=0 = KIL and KI |θ= π2 = KIh ,
respectively. The fluid pressure distribution in fracture
are obtained from the equations as follow
ht 2
Where, Lh = Lf 1− .
2hf
The total pump volume of fracture fluid is equal
fracture volume and leakoff volume according to mass
balance theory, written as
Where, Vf fracture volume, is calculated from the
length, height and width of various segments of
the fracture. The time of fracture propagation is
expressed as
The fluid pressure and width of fracture can be
computed from equation (16).
431
3.2 Computation of fracture height
In order to get the required solutions, the half height
of fracture, hf , should be iterated to satisfy following
conditions:
Downloaded from http://onepetro.org/ISRMISRS/proceedings-pdf/ISRS10/All-ISRS10/ISRM-ISRS-2010-069/1716747/isrm-isrs-2010-069.pdf/1 by University of Baghdad, Ali Mahmoud on 17 June 2023
1. The fluid pressure over the elliptical fracture sur-
face obtained from Muskat and Prats (short for MP)
is equal to that obtained from Shah and Kobayashi
(short for SK), satisfying the width and pressure
coupling conditions.
2. The difference between the average fluid pressure
along both x and y axes in the MP and SK calcula-
tions is a minimum, thus satisfying the asymmetry
requirement along the two axes. The differences are Figure 3. Relationship between height and length of half
expressed as fracture.
bigger than that of pseudo-3D with 2D flow modal and
The difference of two squares is written as measured values (see Fig. 3). The results show that
the pseudo-3D with 2D flow modal is more precise
than pseudo-3D with 1D flow modal comparing with
measured values, illustrating the flow along the height
The precise solution of half height hf is gotten direction of facture cannot be neglected in simulation
when is minimum. of TISR hydraulic fracturing.
The width and fluid pressure along fracture length
direction can be determined by the half height hf and 4.2 Influence factors of fracture height
other known parameters. The fracture is divided into
a number of segments, knowing the width and height In this section we provide different examples to
at the ends of each segment; we can calculate the area demonstrate the influences of pump rate, matrix frac-
and volume values for each segment. Thus, the total ture toughness, matrix Young’s modulus and in-situ
volume of the elliptical fracture, Vf , and the fracture stress on the fracture height and width. In the given
growing time t are obtained. pump time, the computed results are shown in Figure 4.
The fracture height grows as the pump rate grow-
ing at the same fracture length position (see Fig. 4
(a)). When the matrix fractures toughness increases,
4 NUMERICAL STUDY OF PSEUDO-3D WITH
stronger constraints are loaded at the fracture grow-
2D FLOW AND DISCUSSION
ing along height direction, leading the decrease of
fracture height (see Fig. 4 (b)). In contrast, the frac-
Prior to the demonstration of pseudo-3D with 2D
ture height is inverse proportion with matrix Young
flow model, the calculating software has been com-
modulus (see Fig. 4 (c)). From Figure 4 (d), it can
piled by Visual Basic computer language. An actual
be observed that the in-situ stress difference has a
TISR in Shengli oilfield is stimulated as example.
remarkable influence on the fracture height, which is
The average thickness of the storage is 67.97 m,
decreasing with the in-situ stress difference of storage
and stress difference between storage layer and
and upper/lower bed increasing. From the magnitude
upper/lower bed of 0.7 MPa. The fracture toughness,
of fracture height changes, matrix Young modulus and
shear modulus
√ and Passion of storage material are in-situ stress difference are more significant than other
1.098 MPa· m, 2.154 GPa and 0.2, respectively. The
two factors.
pump rate is 6.359 m3 /min and fracture fluid viscosity
of 3.352 Pa·s. The calculating results obtained by the
software are shown in Figure 3–4.
5 CONCLUSIONS
(1) The pseudo-3D with 2D flow modal is built up to
4.1 Comparison of results
stimulate the hydraulic fracturing propagation of
The flow along height direction is neglected in the TISR and the calculating software is compiled by
pseudo-3D with 1D flow modal and energy in frac- Visual Basic computer language.
turing fluid are all used in fracture propagation along (2) The pseudo-3D with 2D flow modal can pre-
length direction, which causes the length of fracture dict the hydraulic fracturing propagation of TISR
obtained from pseudo-3D with 1D flow modal is much effectively comparing with the measured values.
432
Downloaded from http://onepetro.org/ISRMISRS/proceedings-pdf/ISRS10/All-ISRS10/ISRM-ISRS-2010-069/1716747/isrm-isrs-2010-069.pdf/1 by University of Baghdad, Ali Mahmoud on 17 June 2023
Figure 4. Different factors influence on the fracture height.
(3) The comprehensive results show that the hydraulic M.K. Rahman, M.M. Rahman, and A.H. Joarder. 2007. Ana-
fracture of TISR grow effectively both along the lytical production modeling for hydraulically fractured
height and length direction, usually length/ height gas reservoirs. Petroleum Science and Technology, 25(6):
<4. The maximum width is at the center of well 683–704.
Muskat, M. 1982. The flow of homogeneous fluids through
bottom fracture. The fracture height increases as porous media. IHRDC Publications, Boston: 181–186.
the pump rate and matrix Young modulus increas- Palmer ID, and Darroll HB Jr. 1983. Three-dimension
ing. In contrast, it decreases as matrix fracture hydraulic fracture propagation in the presence of stress
toughness and in-situ stress difference increasing. variations. SPE 10849-PA.
Matrix Young modulus and in-situ stress differ- Prats, M. 1961. Effect of vertical fractures on reser-
ence have more significant influence on fracture voir behavior-incompressible fluid case. SPEJ(June):
height than pump rate and matrix Young modulus. 105–118.
Qiu Weide, and Lu Lianjun. 2002. A new pseudo three-
dimension flow field model for predicting height of frac-
ture. Journal of the University of Petroleum China (edition
ACKNOWLEDGMENT of natural science, 36(5): 48–51.
S. Amini, D. Ilk, and T. A. Blasingame. 2007. Evaluation
The authors are very much indebted to the China of the Elliptical Flow Period for Hydraulically-Fractured
National Science and Technology Major Project (Con- Wells in Tight Gas Sands—Theoretical Aspects and Prac-
tract No. 2008ZX05037-004 and 2008ZX05036-001). tical Considerations. SPE 106308.
Shah, R.C., and Kobayashi, A.S. 1971. Stress intensity fac-
tor for an elliptical crack under arbitrary normal loading.
REFERENCES Engineering Fracture Mechanics, 3: 71-96.
XU Yun1ong and GE Hongkui. 2005. Status of the study on
Dou Ranglin, Xu, Long, Zhang, and Baotuan, et al., 2001. the technology of controlling fracture in near wellbore.
Application of Large Scale Hydraulic Fracturing in Wen- Chinese Journal of Rock Mechanics and Engineering,
dong Oilfield. Journal of Jianghan Petroleum Institute, 23: 24(Supp.1): 5189–5193 (in Chinese).
74–75 (in Chinese).
M.M. Rahman, M.K. Rahman, and S.S. Rahman. 2003.
Multicriteria hydraulic fracturing optimization for reser-
voir stimulation. Petroleum Science and Technology,
21(11/12): 1721–1758.
433
