Professional Documents
Culture Documents
com
BCIIINCII@DIRIlCT.
PERGAMON
Intem.tio.... Journal or
Rock Mechanics
and Mining Sciences
of Civil and Environmental Engineering, University of Alberta, Edmonton, Albta, Canada 1'6G 2G7
Geomechanics Research Centre, Laurentian University, Sudbury, Ont., Canada
e Swedish Nuclear Fuel and Waste Management Company, Stockholm, Sweden
b MlRARCO,
Abstract
When the stress-induced risks to a projects warrant it, in situ stress must be measured. However, as the stress-induced risks
increase, i.e., the stress magnitudes approach the rock mass strength, the confidence in commonly used stress measurement
techniques decrease. The design of underground openings at depth requires knowledge of the in situ stress state, yet it is for these
design conditions where our confidence in stress measurement techniques is at its lowest. To quantify the stress state for these
conditions, elements of the Observational Design Method have to be used. These elements rely on the development of a geological
site model, documented observations of over stressed rock in pillars or near the boundary of underground openings, and iterative
two- and three-dimensional numerical modelling calibrated with observations. Examples are provided to illustrate how the
philosophy of Observational Design Method can be used to infer the in situ stress state.
2003 Elsevier Ltd. All rights reserved.
Keywords: In situ stress; Instability; ModeUing
1. Introduction
C.D. Marlin el al. I International Journal of Rock Mechanics & Mining Sciences I (aU) __ _
Nomenclature
cohesion
normal stress on a plane
friction angle
stress (tangential) level around boundary of
SL
underground openings
geological Strength Index
GSI
depth of stress-induced failure
dr
released strain energy
ARE
COV coefficient of variation
statistical mean
J1.
statistical standard deviation
0"
Swedish Nuclear Fuel and Waste ManageSKB
ment Company
AECL Atomic Energy of Canada Limited
Aspo HRL SKB's Aspo Hard Rock Laboratory
Sweden
URL AECL's Underground Research Laboratory
Canada
O"n
experience WIth
similar conditions
Predictions of rock
mass responce to
excavation
Feedback loop
n"n
;;1 "n
L...-i=l
( 1 "n
;; L...-i=l
1 L...-i=l O"EEj
O"NE j
O"UEj
1 "n
1 "n
;; L...-i=l O"UNj ;; L...-i=l O"UUj
1 L...-i=l
(1)
C.D. Martin et al. I International Journal of Rock Mechanics & Mining Sciences. (00) .....m
Table I
Stress measurement techniques tried at AECL's Underground
Research Laboratory su=arized from Martin (42] and Martin et aI.
(45]
Method
Technique
Indirect
Hydraulic fracturing
Modified CSIR
CSIRO
Borre probe
CSIR door stopper
USBM gauge
Borehole slotter
Maximum stress
Direct
Hydraulic fracturing
Minimum stress
Large-scale
back-analysis
Convergence
Under-excavation
Radial displacements
EaSI-Norlh-Up
(XYZ)
NortlJ-Easl-lJown
(XYZJ
Table 2
Example of 4 triaxial overcore test results presented as principal
stresses
uI{fr/PI
(MPatt>
u2/Tr/PI
(MPatt>
u3/TrfPl
(MPatt>
33.1/237/25
26.3/238/17
33.1/233/29
34.1/244/09
18.9/339/23
14.5/136/36
17.0/142/02
18.8/145/45
16.2/106/55
7.9/349/49
14.7/049/61
13.0/324/44
Mean
oX
0"1
0"2
b..
0"3
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences I (UO) __ m
>.
80
Table 3
Summary of measured vertical stress gradients in various rock types
60
Vertical stress
gradient (MPa/m)
Location
(rock type)
Depth
(m)
0.0249
0.0266 0.0028
0.0270
0.0265
0.0260.0324
0.02660.008
0.027
0.0285
0.0260
0.0264
0.0249 0.00025
900
0-2400
0-3000
100-3000
0-2200
0-2200
Reference
c:
Q)
:::>
IT
l!!
40
II..
20
200
400
Depth (m)
0-440
0-2300
0-2200
150-420
140-1830
[48J
[49]
[SOJ
[51J
[52J
[53J
[42J
[54J
[10J
[55J
[56]
(2)
-1
(-J ___
n<~~r.T~Tnnr~TMOT~
:5
g.250
300
350
400
25
,
\
20
100
X,
l~ 15
:g
,g
120
Standard DevioIlonl100tn
A Meanll00m
X Coetllclent ofVariatlonll00m
,,
,
,,
X,
10
CI)
,,
80
60
Sl.
40
Q)
:::!.
g,
o
:l
20
100
200
300
400
500
600
700
800
Depth (m)
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences. (Ba) __ _
50
120
40
~100
10.08 + 0.041=
n.
~ 30
J:
~ 60
~
.12'
0
0
0 0
117.1-111e-o. 00052<
'"
~80
n.
4.55 + 0.036z
in'"
20
en
300
400 500
Depth (m)
(a)
600
_0-
20
BOO
700
500
(b)
50
120
40
100
1000
1500
Depth (m)
m
n.
2500
5.37+0030:
~80
~ 30
2000
J::
40
10
0-
~60
'", 20
'"
O40
in
10
20
600
(0)
BOO
700
500
(d)
Depth (m)
5
J::
~4
.12'
en
2000
2500
5
0
'6
.12'
;:3
'"
~2
~3
&
_ _0
--.:.
~--~
0
0
-0-
200
300
400
500
Depth (m)
600
as
,2
in
.Q
_0_
=1
Q:
2 16 -1 14e"0.0053<
100
as4
0 '
.Q
(e)
1000
1500
Depth (m)
700
800
(f)
1.89 - 0.0003:
500
1500
1000
Depth (m)
2000
2500
Fig. 7. Comparison of in situ stress data from Scandinavia and Canada. In Canada 0'1 is approximately horizontal hence 0'1 ::::: 1THm.. and 0'2 ::::: CTb"." .
(a) Scandinavian data: O'H..... (b) Canadian data: 0'1. (c) Scandinavian data: CTb".". (d) Canadian data: 0'2 . (e) Scandinavian UH.... /O'hmm ratio.
(f) Canadian 0'1/0'2 ratio.
In contrast to the "de-stressing" explanation discussed above, " pop-ups" are an indicator of high
horizontal stress at very shallow depths. Fig. 9a is a
'pop-up' observed at a granite quarry in Southeastern
Manitoba. Stress measurements (10) were made at
depths ranging from 5.75 to 16.2 m using the USBM
borehole deformation gauge (Fig. 9b). The maximum
horizontal stress ranged from 16.9 to 21.3 with an
average of 18.1 MPa and the minimum horizontal stress
ranged from 8.0 to 10.6 with an average of 9.2 MPa.
While the maximum to minimum horizontal stress ratio
is similar to that given in Fig. 7, the magnitude of the
stresses is outside the range that could be expected from
the stress databases. Figs. 8 and 9 highlight the pitfalls
of using databases for estimating stress magnitudes,
particularly at shallow depths.
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences I (00) __ m
;f
80
Transition Zone
Stress
~ 70
'"'"
~0
.t!
50
40
0 30
J:
E 20
:J
E
10
(1!
II
.at~.
II
II
II
rIP
......
II
l!! 60
Ci5
...
Maximum stress
Zone ...
I.,
Relaxed
Zone
...
+lIP
:2
200
400
600
800
1000
Depth (m)
(3)
~20
I)
(JHmax
It
ltI
a.
III
19.5 - 0.095D
e.
0.
15
Ci5
]i 10
c:
(Jhmin
0
-'0
.~
0
J:
0
0
4
(b)
9.02 + 0.0025D
015
0
10
12
14
16
18
Depth (m)
-r = c + Un tan <p,
(4)
(-J __ _
.j
:t:1 ;.
b b 3
cr'" 2
100
200
300
400
(a)
500
600
700
800
100
(b)
Depth (m)
200
300
400
500
600
700
800
Depth (m)
Fig. 10. Horizontal to vertical stress ratios determined from the Scandinavian stress database. (a) Ratio uHaa./uvert. (b) Ratio Ub"../uvert.
t5
~
I
t> 2
o
100
200
300
400
500
Depth (m)
600
700
800
Peak Strength
8\
en..-eloP
3.5
3
2.5
.:.
.. --- 0
::~. -;
In situ Stress
0.1
0.2
0.3
0.4
0.5
C.D. Martin et al. I International Journal of Rock Mechanics & Mining Sciences. (BU) __ m
SL = CTmax =
CTci
CT ci
Table 4
Observed stress level changes at two location near an advancing
stope [15)
or ~O"I
or ~0"2
0"3 or ~0"3
Max. SL
Min. SL
0"2
Gmax/Gci
\
\
0.8
\
\
~ 0.6
6'
\
\
\
\
(S)
0"1
Initial
stress
(MPa)
Stress
change
location 1
Stress
change
location 2
35.5
26
15.5
0.6
0.07
45.5
16
5.5
1.48
-0.12
-15.5
-6
-15.5
0.4
-0.13
0.2
0.4
0.6
08
cr3/crci
Fig. 13. Schematic stress path for excavation affected by a mininginduced stress change (~O"I).
(1 _D) VTOO
.{Cici
2
X IO(GSI-I0)/40
'
(6)
10
CD. Martin et all International Journal of Rock Mechanics & Mining Sciences 1(_) __ _
100
i.
~ 10
III
1_.-.-
100
:::I
:;
"8
~
-~ 1
c
::I
!
~ 0.1
~
(a)
f
i
I
I
I
I
I
I
I
I
I
I
I
I
I
g-
::I
0.1
"i5
~
~ 10
III
.a:::I
~
~
"8
6"
::I
s:
::I
lD
!~ 0.1
"i5
~
E"
0.1
0.1
1
10
100
Laboratory Uniaxial Compressive Strength (MPa)
..-
...--
~-
i.
(b)
100r-----------~~--~~~~~
Ii
0.1
1
10
100
laboratory Uniaxial Compressive Stranglh (MPa)
100
-.-y-
.,,-
Ci)
~
g
~
~ 10
!
"8
"3
:E
i1
::I
"i5
"11
i~ 0.1
0.1..!!!.
(c)
as
11.
0.1
1
10
100
Laboratory Uniaxial CompnlSSive Strength (MPa)
(d)
0.1
1
10
100
Laboratory Uniaxial Compressive Stnmglh (MPa)
Fig. 14. Relationships between Deere's classification for O"ci, laboratory Young's modulus and geology. Superimposed on the plots is the relationship
between rock mass deformation modulus and GSI. (a) Igneous rocks. (b) Metamorphic rocks. (c) Sedimentary rocks. (d) Weak rocks.
C.D. Martin et al. I International Journal of Rock Mechanics & Mining Sciences I
(~O)
__ m
11
= (3k - l)li3.
(7)
12
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences I (UU) ID-W
Risk/Hazard Assessment
Weak & Soft Rocksl Brittle & Hard Rocks
I
GSI <30
GSI >40
Stress-induced brittle
spalling. Usually found
in blocky and massive
rock masses
(a)
~E
"Q;
!::
.!!!
iE 3
.!!!
"~
tl
.!!!
a.
(b)
0.1
02
0.3
0.4
0.5
0.6
07
08
Rock mass strength I Far-field maJCimum wess ("= I",)
2.25
2.5
2.75
(0)
,G51=80
/IG51=60
"
$,,01
/
20
!!
./
/
/ "
/' ".. '"
~ 10
g0
(ij
0.1
0.2
0.3
0.4
05
(e)
'" G51=40
'"
G51=20
:::::---:.--
.:::'"'-1.5
2.5
Fig. 15. Illustration of the two co=on modes of failure and associated hazards observed around under-ground openings. (a) Decision tree.
(b) Depth of plastic zone, data from [26]. (c) Depth of spalIing, data from [28]. (d) Deformation hazard, data from [27]. (e) Strain energy release
hazard.
(O"cm)
as
(8)
O"cm
+ (2b 2 + ab -
(9)
Fig. 16. General layout for the 62-m span Olympic Ice Hockey Cavern
at Gjevik, Norway. Also shown are the high horizontal stresses, the
rock mass modulus Em and the joint patterns, modified from Ref. (30).
from www.rocscience.com
14
C.D. Martin et al. I International Journal of Rock Mechanics & Mining Sciences I
'l
~
(~O)
t;Q"
*0 ZEDEX
Min&-by Test Tunnel
III-m
E
m
= 10(GS"OV40
I '
~III
III
.:! 60
"8'"
E m =2RMR-l00'
~ 40
~
r~
.~
---0'--'-'-'-"20
20
,i.
20
40
60
80
Geomechanics Rock Mass Rating (RMR)
Fig. 17. Stress trajectories around the Olympic Ice Hockey Cavern at
Gjffi1ik, Norway upon completion of the excavation using the
continuum model in Phase2.
220
Extensometer
0-15
Extensometer
Extensometer
0+00
0+15
210
,,
~ 200
"
\
I
\
190
,,
180
\
\
170
oJil
,.til.
_ ,
!lO
Distance (m)
.!!l
& 20
140
~>
J6
,4A
9p
\~Dp
:2
:2
4 0
4 0
Fig. 18. Extensometer results from the Olympic Ice Hockey Cavern at
Gjffi1ik, Norway compared to Phase2 results with a rock mass modulus
of 5 GPa. The extensometer results were taken from Ref. [30].
100
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences. (BU) __ m
15
Fig. 21. General view between 1250 Level and 1800 Level looking
towards mine north.
Mine North
4
Stope
2900
CD
fStope-:
: 3200 :
.
LC3~Lj
Proposed
Fig. 22. Plan view of open stopes and stoping sequence (circled
numbers). Stope 3300 was stable until mining of 3400 stope
commenced.
16
C.D. Martin et aL !International Journal of Rock Mechanics & Mining Sciences I (UU) Ul-IU
Table 5
GSI and Hoek-Brown strength parameters
GSI
O'ci
(MPa)
m;
mb
Em (GPa)
O'cm
0'00
(MPa)
(MPa)
60
85
230
28
6.71
0.012
27
25
-0.40
230
28
16.39
0.189
80
100
-2.65
CO'ci
Factor of Safety = yD(3k _ 1)'
4 Available
from http://\\'WW.map3d.com/
(12)
6 Available
from http://www.rocscience.com/
from Palisade Corporation http://www.palisade.com/
G.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences. (~O) Dl-W
1260
1260
1260
1280
1300
1300
1320
1320
1340
1340
1360
1360
1380
1380
1225
1250
1275
(a)
1300
1325
1350
1375
1225
(b)
(m)
17
1250
1275
1300
1325
1350
1375
(m)
1280
1300
1320
1340
.. Stable
o Failed
1360
-3
1
2
3
03 (MPa)
1380
1225
(c)
1250
1275
1300
1325
1350
1375
(m)
Cd)
Fig. 23. Vertical section through 3300 Stope looking towards mine west showing the contours for Ul, U2, U3 and the results from the cavity survey.
Also shown are the stress paths for the hanging wall obtained using MAP3D. At the Stope wall U2 is oriented out of the plane in an East-West
direction (see Fig. 22). The irregular contours close to the stope wall are related to the numerical sampling points. (a) Sigma I contours. (b) Sigma 2
contours. (c) Sigma 3 contours. (d) Stress paths for hangingwall.
C.D. Martin et a I International Journal of Rock Mechan ics & M ining Sciences. (0_) __ m
18
Weibull(2.4B, 1.77)
0.7
06
0.5
~
~ 04
><
(/)
Loss of
Confinement
Region
gj
0.3
iii
>
03<0
0.2
0.1
0.0
5.0%
Fig. 24. The 0") "" 0 isosurface obtained from the three-dimensional
numerical program Examine3D after mining the 3400 stope.
90.0% .
-
---
5.0%>
3.07
'V
I
'VI
I
I
1.6
1.4
~
0
'"
Fig. 26. The Weibull Distribution for the stress ratio k data given in
Fig. lOa.
1.B -"
0.84
1.2
O.B
>< 1.0
I/)
Q)
:::I
O.B
0.6
>
0.6
0.4
til
0.4
0.2
0.2
0.0
'"'"o
<5.0%
147
90.0% '
2.2
5.0%>
228
827%
Fig. 27. The distribution of the factor of safety against spalling using
the distribution for the uniaxial strength, the stress ratio k and
Eq. (12).
While this technique does not provide a full distribution of the output variable, as do the Monte Carlo and
Latin Hypercube methods, it is very simple to use for
problems with relatively few random variables and is
useful when general trends are being investigated. When
the probability distribution function for the output
variable is known, for example, from previous Monte
Carlo analyses, the mean and standard deviation
values can be used to calculate the complete output
distribution.
5.4. Stress variability and depth offailure
(13)
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences. (RRJ
Table 6
Summary of the in situ stress tensor determined at the end of the
experiment, after [41]
01-.
19
3.5
30
100
90
80
Measured
2.5
Magnitudes (MPa)
Trend/plunge (deg)
603
145/11
454
054/08
114
290/77
70
60 ~
50 [
40 ~.
g 2.0
Q)
:l
1.5
LL
1.0
COV=O%,
COV=2%,
COV= 13%.
0.5
0.0
0
(a)
1.4
7Typica1 COY of
25% to 30%.
<Tci
0.4
Depth of failure
0.6
0.8
>0
90
80
70
60
50
1.0
c:
Q)
:l
!!!
0.6
2
~
iii
40 ~
30 (I)
LL
0.4
20
10
0.2
0.0
.-------~--------~,---------~100
1.2
CT
30
20
10
0
~~~~~~~~~~~~~~30
(b)
0.2
0.4
06
0.8
1
Depth of failure = 1.25 SL - 0.66 (m)
1.2
100
;R
~
!!! 80
.2
'iii
::: 60
o
.s:::.
g.40
'0
~----
::::. 20
o ~--------~~----~--~----~~~
20
30
10
20
C.D. Martin et al I International Journal of Rock Mechanics & Mining Sciences I (UU) __ _
fall outside a range of 50 < O"H < 70 MPa before differences in spalling failure could clearly be attributed to
stress. Inversely, this analysis demonstrates why it is
rather difficult to accurately establish the in situ stress
magnitude and stress ratio borehole breakouts.
In summary, while it is demonstrated that the
maximum stress near an excavation is sensitive to
variations in stress magnitude and stress ratio, it is the
variability in rock strength rather than stress, that is
responsible for the frequently observed variability in
depth of failure and failure mode in general.
6. Conclusions
References
[1] Peck RB. Ninth rankine lecture: advantages and limitations of the
observational method in applied soil mechanics. Geotechnique
1969;19(2):171-87.
[2] Herget G, Arjang B. Update on ground stresses in the Canadian
Shield. In: Herget G, Aljang B, Betournay M, Gyenge M,
Vongpaisal S, Yu Y, editors. Proceedings of the Stresses in
Underground Structures, Ottawa. Ottawa, Canada: Canadian
Government Publishing Centre; 1990. p. 33-47.
[3] Stephansson O. Rock stress in the Fennoscandian Shield. In:
Hudson JA. editor. Comprehensive rock engineering-rock
testing and site characterization, vol. 3. Oxford: Pergamon Press;
1993. p. 445-59.
[4] Walker JR, Martin CD, Dzik EJ. Confidence intervals for in situ
stress measurements. Int J Rock Mech Min Sci Geomech Abstr
1990;27(2): 139-41.
[5] Wiles TO, Kaiser PK. A new approach for the statistical
treatment of stress tensors. In: Herget G, Aljang B, Betournay
M, Gyenge M, Vongpaisal S, Yu Y, editors. Proceedings of the
Stresses in Underground Structures, Ottawa. Ottawa, Canada:
Canadian Government Publishing Centre; 1990. p. 62-76.
[6] Martin CD, Read RS, Chandler NA. Does scale influence in situ
stress measurements? Some findings at the Underground Research
Laboratory. In: da Cunha AP, editor. Proceedings of the First
International Workshop on Scale Effects in Rock Masses, Loen,
Norway. Rotterdam: A.A. Balkema; 1990. p. 307-16.
[7] Wiles TO, Kaiser PK. In situ stress determination using the
under-excavation technique-I. Applications. Int J Rock Mech
Min Sci Geomech Abstr 1994;31(5):447-56.
[8] Martin CD, Chandler NA. Stress heterogeneity and geological
structures. Int J Rock Mech Min Sci Geomech Abstr 1993;
30(7):993-9.
[9] Christiansson R, Martin CD. In-situ stress profiles with depth
from site characterization programs for nuclear waste repositories. In: Sarlli P, Eloranta P, editors. Proceedings of the
EUROCK 2001, Espoo, Finland. Rotterdam: A.A. Balkema;
2001. p. 737-42.
[10] Arjang B, Herget G. In situ ground stresses in the Canadian
hardrock mines: an update. Int J Rock Mech Min Sci 1997;
34(3-4):652.
[11] Sheorey P. A theory for in situ stresses in isotropic and
transversely isotropic rock. Int J Rock Mech Min Sci Geomech
Abstr 1994;31(1):35-46.
[12] Hudson JA. Strategy for a rock mechanics site descriptive model:
a test case based on data from the Aspo HRL. SKB Technical
Report R-02-04, Swedish Nuclear Fuel and Waste Management
Company, Stockholm, Sweden, 2002.
[13] Hoek E, Brown ET. Underground excavations in rock. London:
The Institution of Mining and Metallurgy; 1980.
[14] Brady BHG, Brown ET. Rock Mechanics for underground
mining, 2nd ed. London: Chapman & Hall; 1993. 571p.
[15] Kaiser PK., Yazici S, Maloney S. Mining-induced stress change
and consequences of stress path on excavation stability-a case
study. Int J Rock Mech Min Sci 2001;38(2):167-80.
[16] Kaiser PK., Diederichs MS, Martin CD, Sharp J, Steiner W.
Keynote: underground works in hard rock tunnelling and mining.
In: Proceedings of the GeoEng2000, An International Conference
on Geotechnical and Geological Engineering, Melbourne, vol. I:
Invited Papers. Lancaster: Technomic Publishing Co., Inc; 2000.
p.841-926.
[17] Eberhardt E. Numerical modelling of three-dimensional stress
rotation ahead of an advancing tunnel face. Int J Rock Mech Min
Sci 2002;38(4):499-518.
[18] Hoek E, Kaiser PK, Bawden WF. Support of underground
excavations in hard rock. Rotterdam: AA. Balkema; 1995. 215p.
C.D. Martin et al. I International Journal of Rock Mechanics & Mining Sciences I (00) __ m
21