Vol. 9 No.

2 317~325 A C T A S E I S M O L O G I C A SINICA May,1996

Inversion of breakout data from inclined bore-

holes for stress state of the upper crust in
Jizhong depression"
YAN-XIANG Y U ( ~ / ~ h ~ ) and ZHONG-HUAI XU (~,~,~]~)
Institute o f Geophysics, State Seismological Bureau, Beijing 100081, China

Abstract

The forward and inverse problems of studying crustal stress state from breakout data of inclined boreholes are
concisely stated. Direction of the maximum horizontal principal stress (compressive) and relative magnitudes of
the horizontal stresses to the vertical stress in the upper crust in two regions of the Jizhong depression, the North
of China, are obtained by analyzing the breakout data of 6 inclined wells. To get stable results in the analysis we
searched for the unknown parameters both forwardly and inversely. The results show that the maximum az-
imuths of the horizontal principal compressive stresses in the central and southern part of the Jizhong depression
are N86°E and N77°E, respectively, while the relative magnitudes of the three principal stresses in the upper
crust (about 1000~4000 m) of the depression are variable. In the central parJ:of the Jizhong depression we have
found S u z S v ~ Sh-----1.38 : 1.00 : 0.57, where S H , S v and Sh are the maximum horizontal, vertical and mini-
mum horizontal stress, respectively. This indicates that the present stress regime in this area is of strike-slip
faulting type. In the southern part of the depression we have obtained SH : S v • Sh= 0. 80 z 1.00 , 0. 62, indi-
cating a normal faulting stress regime in the shallow part of the crust.

Key Words: breakout, stress state, Jizhong depression, inclined borehole.

Introduction
In recent years well bore breakouts were widely used as indicators of the direction of hori-
zontal principal stresses, because breakout analysis is a simple, convenient and reliable method,
its observation points are usually dense and relatively deep (can reach 4 0 0 0 ~ 5 0 0 0 m) (Bell and
G o u g h , 1979; Zoback e t a l . , 1985~ Gap et a l . , 1991, 1993). Previously well bore breakouts
were only used to deduce the directions of horizontal principal stress axes because of the limitation
of data and method. It is difficult to acquire the information of crustal stress magnitudes, which
is vital to many geophysical researches. Although some researchers have estimated the magnitude
of maximum and minimum horizontal principal stress based on the depth at which the radially un-
symmetric breakouts of borehole wall initiate (Mops and Zoback, 1990; Huang et a l . , 1994),
the lack of knowledge on t h e / n s i t u rock strength restricts the precision of the estimation. Re-
cently, some researchers have tried to find the direction of the minimum horizontal principal
stress and the relative magnitudes of all these three principal stresses simultaneously by inversion
of breakout data from inclined boreholes (Qian and Pedersen, 1991). In this paper, we study
the stress state at two sites of Jizhong depression, North China, by both forward calculation and
inverse analysis of breakout data from inclined well bores.

• Receivedfirst draft February 6, 1995; Receivedin final form March 9, 1995; Accepted March 30, 1995.
Contribution No. 95A0054, Institute of Geophysics, SSB, China.
 318 ACTA SEISMOLOGICASINICA Vol. 9

The Jizhong depression is a large depression in North China basin. In North China basin,
there exists a series of half-grabens bounded by normal faults on one side, and the faults may ex-
tend to 6 "~ 8 km. This leads some geologists to believe that North China basin is a typical multi-
ply extended basin (Ye et al. , 1983), which implies the stress state in the basin should be of
normal faulting type. Both 1966 Xingtai earthquake and 1976 Tangshan earthquake are the types
that the main shock occurred within North China basin. Their mechanisms show that the fault
plane solutions are strike-slip type, indicating a strike-slip faulting stress regime in focal depth
range (12"--20 km). The tectonic stress field inferred from a large number of small earthquakes
also shows the strike-slip faulting type in this region (Xu et al. , 1983). Besides, there are other
evidences showing that the shallow part of upper crust in some places of North China basin is also
in the strike-slip faulting stress regime. For example, Ding eta/. (1985) obtained a result of SH
> S v > S h based on hydra-fracturing data in oil-wells in this area, where Sv, Sn and Sh denote the
vertical, maximum and minimum horizontal principal stress, respectively. But by hydra-fractur-
ing it is difficult to determine Sn precisely and impossible to measure Sv. We need new evidences
and further researches to answer what on earth the stress state is in the North China basin. In
this paper we investigate the present stress state in the upper crust of Jizhong depression, North
China, by analyzing well wall breakout data from inclined boreholes.

I The forward problem

The forward problem is to determine the azimuth of the greatest compressive stress on the
wall of a uniformly loaded borehole when the direction and magnitude of the loading principal
stresses, as well as the trend and deviation of the borehole are known.
Suppose one of the principal stress axes is vertical. For simplicity, we assume that the re-
gional stress tensor is on diagonal form in an ( x , y , z) coordinate system with the x axis point-
ing upwards. Then, we may write the stress tensor as

X| Xl
a = ° °0]
Sn
0 Sh
(1)

Ptane perpendicular to
o1¢ axis where Sv, Sn and Sh are the vertical, maximum
Magnetic horizontal and minimum horizontal stress, re-
orth
spectively. Here we define compressive stress
rend of
borehol¢ being positive.
Let (xl, xz, x3) be the geomagnetic coor-
t direction dinate system where xl is vertically up, x2 is in
Horizontal plane reako

magnetic east, and x3 is in magnetic north. Set

up a borehole coordinate system (X'l, x'z, x'3)
which varies with depth as the borehole deviates
Figure 1 Borehole coordinate system (x'l, x'2, from the vertical, where x'l coincides with the
x'3 ) and geomagnetic coordinates borehole axis and positive up, x'z and x'3 are in
(xl, x2, xa). the plane perpendicular to the borehole axis, and
x'3, when projected to the horizontal plane, co-
incide with the trend of borehole. The coordinate systems are sketched in Figure 1.
The transformation of coordinates from (xa, xz, x3) to (x'l, x'z, x'3) can be obtained as
 No. 2 YU,Y. X. eta/. : STUDY ON INCLINED BOREHOLE BREAKOUTS 319

/ [x
Qian and Pederson (1991 ) :

x2 -- Kb x'2 (2)
•273 X3

COS~ 0 sine ]
K b ---- -- sine sin~ cos3 cos¢ sinai (3)
-- sin~ cos~ - - sin3 cos¢ cos3J

where ~ is the deviation of the borehole with respect to the vertical and d is the azimuth of the
horizontal projection of the borehole trace measured clockwise from magnetic north eastwards.
I{ the least horizontal principal stress Sh makes an angle v with the x3 direction (from north
to e a s t ) , the stress tensor in the geomagnetic coordinate system ( x l , xz, x3) can be expressed as

og = R~oRh (4)

Rh = cosy o I
-- sinv (5)
sinv cosy )

The stress tensor in the borehole system (X's, x ; , sc;) can then be found as

'SI! $12 $13}

a' =
I$12 $22
[$13 $23 $33
Sz3 (6)

o J --- R'~agRb = (RhRb)Ta(RhRb) (7)

T h u s , if we have found a ' , we can calculate the stresses at the borehole wall

a= = Sn - - 2P($22 -- S33)cos2a - - 4psin2a

t a.. = $22 + Sa3 - - 2(Sz2 -- S3a)cos2a -- 4sin2a

r,, = 2(S13cosa - - Slzsina)
(8)

where a is the angle measured from x'z towards x'3 counterclockwise to the studied point at the
borehole wall in the borehole coordinate system, and p is Poisson's ratio of the medium, the val-
ue taken here is 0.25.
The greatest compressive stress on the borehole wall is then given by

1
,~ = -~- {a~ + a.o + [-(a.~ -- a~.) 2 + 4r~Z,]'A} (9)

When a takes an angle a M , ,~"reaches the m a x i m u m ,v . aM determines the breakout direction in

the borehole coordinate system. T h e breakout direction in the geomagnetic coordinate system
( x l , x2, x3) can then be obtained by projecting the unit vector (0 COSa M sinaM) T from the
( x ] , x'2, x'a) system to the geomagnetic ( x ] , Xz, x3) system
 320 ACTA SEISMOLOGICA SINICA Vol. 9

X2
x3
1 /0i[
. ~ R b COSOtM :
[.sinaM J
sin~ sina u
cos3 cosaM -+- cos@ sin3 sina u
-- sin3 cosaM -+- COS~COS3 sina M
]
(10)

Thus the breakout direction in the x2-x3 plane with respect to the geomagnetic north x3 is

ub = t a n - ' x--! = t a n - ' cos3 cosaM ~1_ COS~sin3 sina M (11)

x3 -- sin3 COSaM+ COS¢COS3 sinaM

It is difficult to maximize the greatest compressive stress E with respect to the angle a ana-
lytically. So we have to use a numerical method to maximize E. Firstly we search for a maximum
of E with one degree step for the angle a, then progressively refine the step length until satisfac-
tory precision is reached.
If the borehole is vertical ( ~ = 0 ) , the greatest compressive stress oa the borehole wall is
2= {a={Sv--2p(SH--Sh)cos(2a+2v)= (12)
¢7. S H "-~ S h - - 2 ( S H -- Sh)cos(2a + 2v)
,Y is equal to either a= or am, depending on the relative magnitude of a= and a**. Maximium ~ can
be obtained when angle a gives

aM . . . . 11 (13)
2

Projection of the breakout direction in the geomagnetic coordinate system ( x l , xz, x3) gives
v~ = u (14)
This means that the breakout direction of a vertical borehole coincides with the direction of the
least horizontal principal stress. So we can find the orientation of horizontal principal stress from
the breakout direction.

2 The inverse problem

When a set of breakout directions, borehole deviations and borehole azimuths are given, we
can determine the relative magnitudes of all three principal stresses and their directions. This is
the inverse problem.
Let us define the term "data", d = (dz, dz, "'" , dN) w, where di is the breakout direction as-
sociated with the ith borehole deviation (¢i) and azimuth (3~). we define the term "model" m =
(v, SH/Sv, Sh/Sv) T where v is the angle between the minimum horizontal principal stress and
magnetic north (clockwise). SH/Sv and Sh/Sv are the ratios between the maximum and mini-
mum horizontal principal stresses and the vertical principal stress.
The following nonlinear forward relations exist between the data and the model
d = f(q~,3,m) (15)
Which is presented in last section (eq. (11)) where

3 = (31, 3Z, °°°, 3N) T (16)

To solve this equation, we use the Marquardt-Levenberg method (Computation center of

 No. 2 YU,Y. X. et a2. : STUDY ON INCLINED BOREHOLE BREAKOUTS 321

Chinese Academy of Science, 1987).

Let the objective function is

Q : ~ [di -- f(~,, ~i, ra)-]2 (17)

i=l
T o minimize Q, a iteration method is used. The iteration formula is
m (k+l) = m (*) + A ~k) k = 0, 1, 2, ...... (18)

where m (k) is the model parameter of the kth iteration. An initial value m (°) must be given for the
iteration. A(k) = (A~k) ' ~2A(k), " " , A ~ ) ) T , A (k> is the correction and must satisfy
(j(k)Tj(k) + A(k)l)A(k) = j(k)T(d _ f ( ~ , ~ , m ( k ) ) ) (19)
where ,~(k)is the damping factor, I is the unit matrix and j(k) is the Jacobian matrix with

j(i~) a f ( ~ i , ~,, m (k))

= ~ns (20)

When a satisfactory fitness is reached or the maximum iteration steps have been exceeded,
the iteration stops. The m at the last step is the approximate solution of our inverse problem.

3 Data and processing

Jizhong depression,-North China, a long and narrow block elongating in N E - S W direction,
is a second-order tectonic element in northwestern part of North China basin (Figure 2). Hun-
dreds of wells in the depression were logged with 4-arm dipmeter since 1980's. The researchers
have derived the orientation characteristics of the horizontal principal stress axes in detail in this
region by analyzing data from 67 oil wells (Yu and Xu, 1994). Now we process the 4-arm dip-
meter logging (DIPLOG) data of the well Majian 3, Magu 4 and Ma 351 in Raoyang sag, and
Yuxie 21, Yuxie 22 and Yuxie 23 in the east of Shenxian sag. Locations of these 6 wells are
sketched in Figure 2. The logging interval and
hole deviation for the well Majian 3 are 3705
4086 m and 7 ° ~ 2 7 °, respectively. For Magu 4,
1 4 5 0 ~ 2 4 0 0 m and 1 ° ~ 1 9 ° ; for Ma 351, 1 0 0 0 ~
2045 m and 1 ° ~ 8 ° ; for Yuxie 21, 2 0 0 0 ~ 3 0 3 4 m
and 5 ° ~ 2 2 ° ; for Yuxie 22, 1 2 0 0 ~ 3 1 3 0 m and 8 °
~ 3 8 ° ; and for Yuxie 23, they are 1000"-'1185 m ~ ' p.,)' Tianjin
¢- / o
and 26° ~ 28 ° , respectively.
We take the weU Majian 3, Magu 4 and Ma
351 in Raoyang sag as the first group, Yuxie 21, /Baoding ^Rcnq,u q_~
/ ~ .Magu4 ~, /
Yuxie 22 and Yuxie 23 , located in the east of
Shenxian sag, as the second group. We digitized
those sections of the logging curve where obvious
/ Yuxie23 ,.~.'/
stress-induced breakouts and stable breakout orien- (Oek:::..k. Sl~ttxitu s|U*sYuxie22 ~ /
atj~aL~u~ Yuxie2f"
tation can be seen in more than 10 meters logging V
interval. After interpolation and averaging, 36 da-
ta are compiled for the first well group and 56 for Figure 2 Jizhong depression and well locations.
 322 ACTA SEISMOLOGICA SINICA Vol. 9

Table la T h e a z i m u t h s of b r e a k o u t s , deviations a n d a z i m u t h s of boreholes of t h e first well g r o u p

Logging Azimuth of Deviation Azimuth of Logging Azimuth of Deviation Azimuth of

intervals breakouts g/(*) borehole intervals breakouts ~/(*) borehole
/m ubl(°) ~1(°) /m ub/(°) ~1(°)
Magu 4 M@hn 3 (continued)
1480~1490 4 1 313 3925~3935 16 20 154
1490~1500 6 1 314 3935~3950 12 21 151
1500~1510 2 1 313 3955~3965 7 21 151
1510~1520 0 1 321 3985~3995 --2 24 152
1570~1580 14 1 305 3995~4005 9 25 151
1705~1715 14 4 301 4015~4025 --11 27 152
1715~1725 5 4 299 4025~4040 3 27 155
1835~1845 8 3 270 4055~4065 2 25 153
2240~2250 4 17 274 4075~4085 1 27 154
2250~2260 9 18 271 Ma 351
2260~2280 1 17 272 1125~1150 5 5 255
2280~2290 3 17 272 1150~1175 --5 5 255
2290~2300 5 17 272 1265~1280 0 6 270
2300~2310 I 17 272 1290~1310 5 5 270
Majian 3 1500~1520 --I0 6 280
3755~3765 --4 11 183 1520~1550 --5 5 285
3820~3830 --15 16 170 1550~1575 --lO 5 285
3865~3875 16 19 164 1575~1600 --5 5 283
3915~3925 12 20 158 1600~1625 --5 5 280

Table l b T h e a z i m u t h s of b r e a k o u t s , deviations a n d a z i m u t h s of boreholes of t h e second well g r o u p

Logging Azimuth of Deviation Azimuth of Logging .Azimuth of Deviation Azimuth of
intervals breakouts borehole intervals breakouts borehole
/m ~J(°) ~/(°) 8/( ° ) /m ~b/( ° ) ~/(°) ~/(°)
Yuxie 21 Yuxie 22
2000~2010 --15 14 357 1215~1225 --55 38 195
2020~2030 --15 16 357 1225~1235 --45 37 195
2030~2040 --5 16 358 1235~1250 --40 36 200
2040~2050 4 17 2 1250~1280 --30 36 199
2050~2060 3 18 0 1480~1F~00 --41 36 190
2060~2075 --6 18 2 1500~1525 --40 34 195
2075~2085 --18 18 0 1525~1550 --30 32 200
2085~2095 --13 19 2 1550~1570 --42 31 196
2095~2105 --16 20 2 1600~1625 --4O 28 195
2105~2115 --12 21 4 1675~1690 --26 24 195
2115~2125 I0 22 6 1800~1825 --19 I8 180
2125~2135 --8 22 2 1825~1850 --20 18 210
2160~2170 4 21 0 1850~1865 --5 18 215
2170~2180 1 21 359 1865~1880 --23 18 22O
2180~2190 3 21 1 1880~1900 --20 17 220
2200~2210 2 20 0 1900~1925 --20 17 225
2225~2235 0 19 1 1925~1950 --19 16 215
2245~2255 17 18 2 1990~2010 --5 14 215
2275~2285 4 18 4 2025~2035 --10 14 210
2220~2230 9 17 2 2080~2095 15 12 195
2375~2385 1 16 358 2145~2160 5 12 195
2405~2415 5 15 1 2390~2400 --2O 10 2OO
2525~2535 II 14 5 2445~2460 10 10 190
2535~2545 16 14 1 Yuxie 23
2565~2575 13 13 3 1140~1150 --37 28 191
2585~2600 --3 13 2 1150~1160 --39 28 190
2660~2675 16 12 6 1160~1170 --35 27 190
2685~2695 12 12 5
2705~2715 i3 11 2
2715~2735 17 11 4
 No. 2 YU,Y. X. et al. : STUDYON INCLINEDBOREHOLEBREAKOUTS 323

the second as shown in Table la and lb.

Inversion of above two data sets were undertaken with the Marquardt-Levenberg ( M - L )
method. We found in the inversion that the result of u is more stable than that of Sn/Sv and Sh/
Sv. T o avoid unstable result of the later two, we jointly used the M - L inversion and grid search.
Firstly we searched for v using the M-L method because u was more stable during inversion pro-
cess. Then we fixed the value of v. Since the range of Sn/Sv and Sh/Sv values are limited (about
0 ~ 3 generally), we made a grid test in the parameter space of Sn/Sv and Sh/Sv in a step of
0.01. The parameters minimizing the objective function are taken as the best solution to our
problem.
The final results are as follows. Orientation of the maximum horizontal principal stress axis
is, after correction of magnetic declination, N86*E in Raoyang sag and N77°E in eastern Shenxi-
an sag. This result is similar to that of our previous work (Yu and Xu, 1994). The ratio SH/Sv
and Sh/Sv representing the relative magnitudes of the three principal stresses are variable in the
two sags. In Raoyang sag, Su/Sv= 1.38 and Sh/Sv = O. 57, or Sn>Sv~Sh. This indicates that
1/he present stress regime in this area is of strike-slip faulting type. In the east of Shenxian sag,
SH/Sv=O. 80 and Sh/Sv=O. 62, i.e. , Sv>SH>Sh, this suggests that the present stress here is
in normal faulting type.
Figure 3 shows the contour plots of objective function Q with respect to model parameters
Sn/Sv and Sj,/Sv, while u is fixed. Q is defined in equation (17). If we consider that the values
of Sn/Sv and Sh/Sv are acceptable when Q is within some limits (e. g. , less than the value repre-
sented by the most inner contour line, the shaded area in Figure 3 ) , we find that the value of Sh/
Sv would always be less than 1, i. e. , the thrust faulting type stress state is impossible at the two
sites studied. At the first site (Raoyang sag) the probability of SH/Sv>I is much higher than
that of S u / S v < l . This means that the first site is more likely to be in the strike-slip faulting
type stress regione than normal faulting type. At the second site (eastern Shenxian sag) the
probability of SH/Sv<I is much higher than that of S H / S v ~ I , indicating that a stress state of
normal faulting type is more probable. So the conclusion about the stress state at these two sites
is believable.

1.50

1.40 1.20
1.30
1.00
1.20
0.80
LIO
LII!
0.6C

0.90~i(a)iRaoyanglIsag I l t [ ~40
0.50 0.60 0.70 0.80 0.90 L00 1.10 1.20 0.O0 0.20 0.40 0.60 0.80 1.00
skis, sds~

Figure 3 Contour plots of objective function.

 324 ACTASEISMOLOGICASINICA Vol. 9

4 Discussion
1. There are different views on the stress state in upper crust of the North China basin.
Based on hydraulic fracturing data, Ding and Liang (1985) obtained a empirical relation for the
variation of principal stress magnitude with depth in Tangshan, Tianjin and Cangzhou region,
North China.
The formula is
S,=0.7+0.023H Sh=--0.5+0.018H Sv = O. 0 2 1 H
where H is in meter, and S , , Sh and Sv is in MPa. From this relation the ratio SH/Sv and Sh/Sv
at depth of 1500 m should be 1.10 and 0.85, respectively, while at 4000 m depth SH/Sv and Sh/
Sv be 1.10 and 0.85, respectively. This implies that the stress state in this region should be of
strike-slip type. Zeng et al. (1991) analyzed the geological and deep seismic reflection data, as
well as the focal mechanisms and surface deformations of the 1966 Xingtai and 1976 Tangshan
earthquakes. They believed that the upper crust of the North China basin is in the stress state of
normal faulting type, while the middle crust is in state of strike-slip faulting type. They proposed
a model to explain the genesis of this stress state: an additional stress field due to injection of
magma from uppermost mantle to middle crust is superimposed on the horizontal tectonic stress
field, producing a considerably large extensional force in the upper part but a horizontal shear
stress in the middle and lower parts of the crust (Zeng et od. , 1991). Nevertheless, more reliable
observations are needed to verify the real stress state in the North China basin. The results of
this paper indicate that stress state in the shallow part of upper crust of the Jizhong depression,
North China basin, is variable. It may be of strike-slip faulting type in some places and normal
faulting type in others.
In general, if the deviation and azimuth of boreholes vary significantly and the angle of devi-
ation is large, the information about the stress tensor included in the breakout data would be rich
(Moos and Zoback, 1990). Although the data we used are not so good due to the limited varia-
tion range of the hole deviations careful analysis shows that the data do include information about
the state of stress underground. When we changed the data set a little bit in the inversion pro-
cess, the stress state of Sn>Sv>S~ in Raoyang sag and Sv>Sn~:>Sh in eastern Shenxian sag did
not changed, although the ratio between SH, Sh and Sv changed slightly. This indicates that the
conclusion on the different stress state between two sites whould be reliable.
2. The direct way to know the stress state underground is making stress measurement.
There have been some methods to measure the stress near the ground surface. To know the
stress at depth (thousands of meters), at present we can only analyze the hydraulic fracturing
data of oil-wells. These data are usually unreliable due to strong disturbance. In general the mag-
nitude of vertical principal stress is not measured, but estimated by the overburden load. In some
places, however, the vertical stress may deviate significantly from the overburden as computed
from the rock density and measurement depth (Cornet and burlet, 1992). In China, up to now,
the deepest borehole specially for hydrofracturing stress measurement (HSM) is only about 800
meters deep (Li e t a / . , 1993). HSM is very expensive and impossible to be widely used. The
method we used in this paper is prospective for stress measurement because it uses only the log-
ging data which many oil field already have and can deduce reliable state of stress at deeper
depth.
3. In this paper we only used the azimuths data of borehole breakout, so we have only ob-
tained the relative magnitude of three principal stresses instead of their absolute magnitude. If we
 No. 2 YU,Y. X. et od. : STUDY ON INCLINED BOREHOLE BREAKOUTS 325

can know the magnitude of one principal stress we may find out the whole stress tensor. For ex-
ample, HSM can determine the minimum horizontal principal stress rather precisely. If HSM is
combined with the method of this paper, we would be able to estimate the whole stress tensor.
Using the HTPF (Hydraulic Tests on Pre-existing Fractures) inversion (Cornet and Burlet,
1992) the whole stress tensor can also be determined, but the estimation of maximum horizontal
principal stress is uncertainty. The three-dimensional stress measurement recently used in engi-
neering reguires to drill several (usually 4) holes with different deviation and azimuth. Its result
may be much affected by local environments such as the stress disturbance induced by the excava-
tion of tunnels. The method in this paper, if combined with other method of stress measure-
ment, provides a new way for determining the whole stress tensor.
4. This paper assumes that the vertical stress is a principal stress. Generally there must be a
principal stress whose axis is vertical near the ground surface. The Jizhong depression area is geo-
graphically a plain without rugged topography. So it is unlikely that the axis of vertical principal
stress significantly deviates from the vertical. The directions of horizontal principal stresses in
Jizhong depression derived from wellbore breakout data show a regular pattern (Yu and Xu,
1994). Therefore it is reasonable to assume that one principal stress is vertical. As to which prin-
cipal stress is vertical, it depends on the stress regime in that region.

The project is sponsored by the National Natural Science Foundation of China.

The logging curves were provided by the Logging Company of Huabei Oil Field. The au-
thors thank A-JIA GAO and DONG-NING ZHANG for their participation in data collection.

References
Bell, J. S. and Gough, D. I. , 1979. Northeast-Southwest compressive stress in Alberta: evidence from oil wells. Earth P h n a .
Sci. L e t t . , 45: 475~482.
Comet, F. H. and Burlet, D. , 1992. Stress field determinations in France by hydraulic tests in boreholes. J. Geophys. R~s.,
97: 11829~11849.
Computation Center of Chinese Academy of Sciences, 1987. A package for statistical analysis of stochastic data. 282pp (in Chi-
nese).
Ding, J. M. and Liang, G. P. , 1985. Stress measurement by hydraulic fracturing in oil-wells of North China. Acts S¢/smo/og/-
ca Sinica, 7 : 3 6 3 ~ 3 7 3 (in Chinese).
Gao, A. J. , Xu, Z. H. and Chen, J. G. , 1991. Horizontal principal stress axes in Sichuan basin deduced from oil-well break-
outs. Acts Seismologica Sirffca, 4: 209~217.
Gao, A. J. , Xu, Z. H. and Zhang, D. N. , 1993. Present tectonic stress field of Karamay oil field. Acts Seismologica Sinica,
I$ (Supp.) : 598~606 (in Chinese).
Huang, Y. R. , Xu, Z. H. and Gao, A. J. et al. , 1994. Study on tectonic stress of Zhongyuan oil field with wellbore breakout
data. Acta Seismologica Sinica , 7: 251~261.
Li, F. Q. , Zhang, B. C. and Su, K. Z. , et aL , 1993. Reservor induced earthquake risk in the Yangtze gorges dam area in
situ tests in borehole at Maoping and analysis. Beijing: Seismological Press, 210pp (in Chinese).
Moos, D. and Zohack, M. D. , 1990. Utilization of observation of well bore failure to constrain the orientation and magnitude of
crustal stresses: Application to continental, deep sea drilling program boreholes. J . Geophys. Res. , 951 9305~9325.
Qian, W. and Pedersen, L. A. , 1991. Inversion of borehole breakout orientation data. J . Geophys. Res. , 96: 20093~20107.
Xu, Z. H. , Yah, M. and Zhao, Z. H. , 1983. Evaluation of the direction of tectonic stress in North China from recorded data
of a large number of small earthquakes. Acts Seismologica Sinica, 5.. 268~279 (in Chinese).
Ye, H. et al. , 1985. The North China basin: An example of a multiply rifted intraplate basin. Tectonics, 4t 153.
Yu, Y. X. and Xu, Z. H. , 1994. A study on orientations of horizontal principal stress in Jizhong depression using borehole
breakout data. Petroleum Explorcaion and Development. 21(2): 48~55 (in Chinese).
Zeng, R. S. , Zhu, L. P. and He, Z. Q. et al. , 1991. A seismic source model of the large earthquakes in North China exten-
sional basin and discussions on the genetic processes of the extensional basin and earthquakes. Acts Geophysica Sinica, 34:
288~301 (in Chinese).
Zohack, M. D. et al. , 1985. Well bore breakouts and in situ stress, d. Geophys. Res. , 90: 5523~5530.