International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1211–1216

Technical Note
Influence of elastic constants on the horizontal in situ stress
P.R. Sheorey*, G. Murali Mohan, A. Sinha
Central Mining Research Institute, Barwa Road, Dhanbad 826 001, Jharkhand, India
Accepted 19 October 2001

1. Introduction relationship between the horizontal stress and the elastic

constants.
A theory for in situ stresses published earlier [1] gives It is proposed to show in this Technical Note more
an equation for estimating the mean horizontal in situ comprehensively that the in situ horizontal stress does
stress which, as per this equation, depends on the elastic depend on elastic constants. The purpose of such an
constants, the geothermal gradient and the coefficient of exercise is to try and estimate the horizontal stress in
thermal expansion. The applicability of this equation is soft rocks in which reliable stress measurement is often
shown for several stress measurement results from not possible. For example, stress measurement in coal
different parts of the world. It is also shown that in measures is generally confined to the rocks and not the
the presence of a mountainous topography and major coal seam itself and we have practically no knowledge of
geological features, the scatter with respect to the in-seam horizontal stresses. While modelling coal pillars
predicted mean stress becomes unacceptably large. Since using the FLAC software developed by ITASCA of
stress measurements are generally carried out in USA, the authors faced considerable difficulty in
relatively competent formations, a proposition in terms arriving at realistic results which were found to be
of the predictive equation is put forward here for rather sensitive to the in situ horizontal stresses.
estimating both horizontal stresses in a softer formation,
if the in situ stresses measured in a competent formation
at the same location are available. 2. Equation for horizontal stress
A large number of publications are available to date
on in situ stress measurements but many of these have A theory for in situ stresses published earlier [1]
reported measurement results graphically rather than represents the earth as a spherical shell, including the
numerically. This somewhat limits the scope of further crust and the mantle, with a variable geothermal
research in this important area. Even so, the numerical gradient and shows that the dependence of the
results available are by no means insignificant and can horizontal stress on elastic constants exists because of
be utilized for further analysis, perhaps leading to some the temperature gradient. If the gradient is taken to be
important conclusions. zero, the horizontal stress ceases to depend on the elastic
Several pointers showing the possibility of depen- constants. This theory proposes two equations for the
dence of the in situ horizontal stresses on the elastic mean horizontal stress sh :
modulus exist in the literature, as reviewed by Amadei Isotropic rock,
and Stephansson [2]. While the vertical in situ stress n bEG
may, in general, be taken to equal the cover pressure sh ¼ gH þ ðH þ 1000Þ; ð1Þ
1n 1n
without causing much error in most instances, the
horizontal stress can be affected significantly by plate Transversely isotropic rock,
tectonics, major geological features and mountainous nhv b Eh G
sh ¼ gH þ h ðH þ 1000Þ: ð2Þ
topography. These factors may considerably garble the 1  nh 1  nh
Eq. (2) is for horizontally disposed transversely isotropic
*Corresponding author. Tel./fax: +91-326-204-339. rock and shows that the modulus, which is the most
E-mail address: psheorey@[Link] (P.R. Sheorey). important single parameter, should be measured in the

1365-1609/01/$ - see front matter r 2002 Published by Elsevier Science Ltd.

PII: S 1 3 6 5 - 1 6 0 9 ( 0 1 ) 0 0 0 6 9 - 7
1212 P.R. Sheorey et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1211–1216

Nomenclature nh horizontal Poisson’s ratio

nhv poisson’s ratio for the effect of horizontal
H depth of cover stress on vertical strain
E elastic modulus b1 horizontal coefficient of linear thermal
n poisson’s ratio expansion
g unit rock pressure sv in situ vertical stress
b coefficient of linear thermal expansion shmax major horizontal in situ stress
G geothermal gradient shmin minor horizontal in situ stress
Eh horizontal elastic modulus (parallel to sh mean horizontal in situ stress
stratification)

horizontal direction even when Eq. (1) is used as a From Table 1, b ¼ 8  106 =1C appears to be a
general approach in practical application. reasonable representative value for different rock types
In some cases, the measured value of the vertical but not for coal. The thermal gradient for crustal rocks
stress sv is also available, especially when overcoring has is taken as 0.0241C/m while for coal measure rocks it is
been employed for measurement. To make Eq. (1) 0.031C/m.
sufficiently general, such values of sv should be Poisson’s ratio reported along with some overcoring
substituted. Thus, data is very high, i.e. 0.4–0.49. Such values are difficult
n bEG to accept and therefore a uniform value of n ¼ 0:25 has
sh ¼ sv þ ðH þ 1000Þ: ð3Þ been assumed.
1n 1n
Where the vertical stress has not been measured, sv ¼
gH: 4. Application to observed data

Many useful stress measurement data are available

3. Estimation of properties for application region- or country-wise. Although some of the con-
cerned publications are rather old, they are sufficiently
Many of the published measurement data are not comprehensive and usable. In order to supplement these
accompanied by the elastic constants, especially when data by later observations, especially from other parts of
hydrofracturing has been used. Results from overcoring, the world, a few authors/organizations were contacted,
on the other hand, mention the elastic constants since but to no avail.
their estimation is an integral part of the method. While Presented below is the performance of Eq. (3) with
applying Eq. (3) to published stress data, the given respect to different groups of stress measurement results.
modulus values (which are to be taken in MPa) are The data, which are already readily available numeri-
directly used, whether obtained from biaxial tests on cally in publications, are not reproduced here.
overcores or from uniaxial compression. Also, it was not
possible to verify if the modulus was determined in the 4.1. North America
horizontal direction.
The coefficient of linear thermal expansion b is not This is a comprehensive compilation of stress
very well documented but Table 1 shows the values for measurements carried out till the late 1970s in the
some rocks available from published literature [3–5]. North American continent [6]. Of the large number of
data, 46 sets of values could be found gainful in showing
the dependence of the mean horizontal stress on elastic
Table 1 modulus [1]. However, in this earlier exercise, the
Coefficient of linear thermal expansion b of some rocks measured vertical stress was ignored and it was taken
Rock type b  106/1C)
as gH: Using the measured values of sv ; we see a
reasonable relationship in Fig. 1 between the mean
Granite [3] 6–9 measured horizontal stress and that predicted by Eq. (3).
Limestone [3,4] 3.7–10.3
Marble [3] 3–15
Sandstone [3] 5–12
4.2. United Kingdom
Schist [3] 6–12
Dolomite [4] 8.1 The results of overcoring tests in British coal measure
Conglomerate [4] 9.1 rocks published recently by Cartwright [7] offer another
Breccia [4] 4.1–9.1 source of comprehensive stress data. In this work, a
Coal [5] 30
definite correlation is shown to exist between the major
 P.R. Sheorey et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1211–1216 1213

80 30
70
25
60
σh calculated, MPa

σh calculated, MPa
20
50

40 15
30
10
20
5
10
0 0
0 10 20 30 40 50 60 70 80 0 5 10 15 20 25 30
σh measured, MPa σh measured, MPa
Fig. 1. Performance of Eq. (3) for the mean horizontal in situ stress for Fig. 3. Comparison of the measured and calculated mean horizontal
North American data. stress as per Eq. (3) for UK coal measures.

50 40
45
35
40
σhmax measured, MPa

σh calculated, MPa 30
35
30 25
25 20
20
15
15
10
10
5 5
0 0
0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40
σhmax calculated, MPa σh measured, MPa

Fig. 2. Comparison of the measured major horizontal stresses with Fig. 4. Performance of Eq. (3) for the stress data from the Japanese
those calculated using Eq. (4) developed for UK coal measures [7]. islands.

in situ horizontal stress shmax and the elastic constants. 4.3. Japan
The empirical relation fitted to 26 test data sets is given
as The earlier published Japanese stress data [8] have
n been supplemented by two more recent measurement
shmax ¼ 0:0092H þ 0:779E  [Link] ð4Þ results [9]. Fig. 4 shows that Eq. (3) fits these data very
1n
well.
Fig. 2 shows that this correlation is very significant
(index of determination 0.94). Cartwright, however,
states that the correlation for the minor horizontal stress 4.4. India
shmin is not as significant as the above relation.
Eq. (3) again gives a good estimate of the mean Many stress measurements have been done in India so
horizontal stress (Fig. 3). It was seen, however, that the far for geophysical and mining purposes and for
earlier value of b¼ 8  106 =1C somewhat overesti- underground civil excavations, the most preferred
mated the stress. Therefore, b¼ 6  106 =1C was taken method being hydrofracturing. Because of this, elastic
to get a better fit, this still being within the range of modulus values are in many cases not available, unless
values for rocks. The earlier value of b has, however, they were required for other purposes.
been retained for the plots from other countries which The stress measurement results are detailed in Table 2
follow. and the performance of Eq. (3) is shown in Fig. 5.
1214 P.R. Sheorey et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1211–1216

Table 2
In situ stresses measured in India

Sl. no. Location Rock type Depth shmax shmin sv Elastic modulus
(m) (MPa) (MPa) (MPa) (GPa)

Hindustan Zinc Limited

1 Balaria Dolomite 96 8.20 5.60 F 19.3
2 Balaria Dolomite 315 20.40 11.80 F 19.3
3 Mochia Dolomite 500 28.60 21.45 F 42.0
4 Zawar Mala Dolomite 235 17.70 9.60 F 33.0
5 Rajpura Dariba Dolomite 277 27.00 15.20 F 35.0
6 Rajpura Dariba Dolomite 318 28.40 15.50 F 35.0

Uranium Corporation of India Limited

7 Narwapahar Schistose formation 104 7.40 4.40 F 27.3
8 Narwapahar Schistose formation 154 9.80 5.60 F 27.3

Ghatghar Hydel Project

9 Ghatghar Compact Basalt 151 5.40 3.60 F 26.8
10 Ghatghar Compact Basalt 295 6.20 3.40 F 26.8

Koyna Stage IV Hydel Project

11 Koyna Compact Basalt 87 4.40 2.80 1.27 32.0
12 Koyna Amygdaloidal Basalt 295 6.20 3.40 4.30 3.0

Hindustan Copper Limited [12]

13 Khetri Quartzite 203 8.40 5.60 6.96 37.5
14 Khetri Quartzite 268 8.89 5.93 7.87 37.5
15 Khetri Quartzite 300 15.00 7.88 8.82 37.5
16 Kolihan Quartzite 364 12.65 7.70 10.70 22.1

Eastern Coalfields Limited

17 Porascole Sandstone 135 5.21 4.46 F 15.5
18 Jhanjra Sandstone 110 5.84 4.68 F 10.2
19 Chinakuri Sandstone 255 8.94 7.15 F 10.4

30 which are relic mountains of the Deccan Trap caused by

lava flow. The mines of Hindustan Zinc Limited are
25 located in the Aravali Range system of Precambrian age
with folds and thrusts indicating tectonic activity.
σh calculated, MPa

20
4.5. Italy and Austria
15
Stress measurements in Italy [10] and Austria [11]
10
carried out in mountainous regions show the influence
of topography, major geological features and tectonics.
5
The scatter with respect to the predictions using Eq. (3)
is clearly seen in Figs. 6(a) and (b).
0
0 5 10 15 20 25 30
σh measured, MPa
5. Estimation of major and minor stresses
Fig. 5. Comparison of the measured and calculated mean horizontal
stress as per Eq. (3) for the Indian data of Table 2.
As is well known, the two horizontal in situ stresses
are more often than not unequal. This inequality or
The last three measurements were done in Ranigunj stress anisotropy occurs because of tectonics, geology
coal measures using Serata stressmeter, which is a sleeve and/or topography. If a soft and jointed formation
fracturing method. The others were done using hydro- exists at a location where in situ stresses have been
fracturing but the elastic modulus was determined for measured in a relatively more competent rock, the
some associated rock mechanics studies. Ghatghar and results so obtained clearly should not be used in the soft
Koyna Hydel Projects are located in the Western Ghats, formation. Although the average horizontal stress may
 P.R. Sheorey et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1211–1216 1215

50 6. Conclusion
45
Eq. (3) fits the observed in situ stress measurement
40
data reasonably if the measurements are not seriously
σh calculated, MPa

35 influenced by topography, major geological features or

30 tectonics. The equation, however, shows a definite
25 dependence of the horizontal stress on elastic constants.
20 It is proposed that a formula like Eq. (5) may be used
to estimate both horizontal stresses in a softer rock
15
formation if the stresses measured in a competent
10 formation at the same location are available. It can
5 also be used to estimate the stresses in any particular
0 rock type, where measurement has not been practically
0 5 10 15 20 25 30 35 40 45 50
possible. It may thus be possible to estimate the in-seam
(a) σh measured, MPa horizontal stresses in coal mines. If in-seam measure-
50 ments are not possible, careful numerical modelling of
cases may be done to check such a derivation.
45
40
σh calculated, MPa

35
Acknowledgements
30
25 Thanks are due to the director, CMRI for his
20 permission to publish this paper. The views expressed
15 in the paper are those of the authors and not necessarily
of the CMRI. The authors are grateful to Mr. P.K.
10
Rajmeni of Hindustan Zinc Limited for the elastic
5 modulus values from their mines. Stress measurement
0 results from the mines of Hindustan Copper Limited are
0 5 10 15 20 25 30 35 40 45 50
(b) σh measured, MPa
due to the National Institute of Rock Mechanics, Kolar.
All other results in Table 2 are due to Dr. A. Sinha of
Fig. 6. The scatter with respect to the calculated mean horizontal the CMRI, Dhanbad, who is a co-author of this paper.
stress using Eq. (3) in the Alpine regions of (a) Italy and (b) Austria.

be estimated in the softer rock by measuring its elastic References

constants and coefficient of thermal expansion, it will
[1] Sheorey PR. A theory for in situ stresses in isotropic and
not be possible to estimate both the horizontal stresses.
transversely isotropic rock. Int J Rock Mech Min Sci Geomech
It is then proposed that if at a location the measured Abstr 1994;31:23–4.
major and minor horizontal stresses are available, the [2] Amadei B, Stephansson O. Rock stress and its Measurement.
corresponding stresses in a softer rock at the same London: Chapman & Hall, 1997. p. 490.
location are given by [3] Haas CJ. Static stress–strain relationships. In: Touloukian YS,
Ho CY, editors. Physical properties of rocks and minerals. IIF2.
New York: McGraw Hill, 1981. p. 409–83.
[4] Schafer K. In situ strain measurements in Libya. Rock Mech
shmax1 ð1  n2 Þ½n1 sv þ b1 E1 GðH þ 1000Þ 1980;(Suppl 9):49–61.
¼ ; ð5Þ
shmax2 ð1  n1 Þ½n2 sv þ b2 E2 GðH þ 1000Þ [5] Van Krevelen DW. Coal. Amsterdam: Elsevier, 1961. p. 514.
[6] Lindner EN, Halpern JA. In situ stress in North America: a
compilation. Int J Rock Mech Min Sci Geomech Abstr
where the suffixes 1 and 2 stand for the softer and the 1978;15:183–203.
[7] Cartwright PB. A review of recent in situ stress measurements in
competent formations, respectively. The depth H is to United Kingdom coal measure strata. In: Sugawara K, Obara Y,
be taken corresponding to the soft formation. A similar editors. Proceedings of the International Symposium on Rock
equation can be written for the minor stress shmin : Stress. Rotterdam: Balkema, 1997. p. 469–82.
The underlying assumption in such a proposition is [8] Kamagawa T, Hibino S, Ishida T, Hayashi M, Kitahara Y. In situ
stress measurements in Japanese Islands: overcoring results from
that the factors which give rise to horizontal stress
a multi-element gauge used at 23 sites. Int J Rock Mech Min Sci
anisotropy at that location remain the same, which is Geomech Abstr 1986;23:29–39.
reasonable if the soft rock formation is not far from the [9] Matsui H, Sato T, Sugihara K. Comparison of the results of stress
site of measurement. measurements by various methods at the Kamaishi mine. In:
1216 P.R. Sheorey et al. / International Journal of Rock Mechanics & Mining Sciences 38 (2001) 1211–1216

Sugawara K, Obara Y, editors. Proceedings of the International [11] Kohlbeck F, Roch K-H, Scheidegger AE. In situ stress measure-
Symposium on Rock Stress. Rotterdam: Balkema, 1997. ments in Austria. Rock Mech 1980;(Suppl 9):21–29.
p. 95–100. [12] Hindustan Copper Ltd. Personal communication referring to the
[10] Martinetti S, Ribacchi R. In situ stress measurements in Italy. test reports of the National Institute of Rock Mechanics, Kolar,
Rock Mech 1980;(Suppl 9):31–47. 2000.