International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27

Contents lists available at ScienceDirect

International Journal of
Rock Mechanics & Mining Sciences
journal homepage: www.elsevier.com/locate/ijrmms

Geomechanical modeling using finite element method for prediction

of in-situ stress in Krishna–Godavari basin, India
Dip Kumar Singha, Rima Chatterjee n
Dept. of Applied Geophysics, Indian School of Mines, Dhanbad 826004, India

art ic l e i nf o a b s t r a c t

Article history: This study provides quantitative rock mechanical properties and analyses of in-situ stress and pore
Received 12 June 2014 pressure in several oil/gas fields of East and West Godavari sub-basins. High pore pressure gradients
Received in revised form varying from 11.85 to 12.80 MPa/km exist within these oil/gas fields. Vertical stress (SV) gradients in the
1 October 2014
range 21.00 to 22.85 MPa/km are seen to exist. Minimum horizontal principal stress (Sh) magnitude is
Accepted 6 October 2014
found to vary from 64% to 76% of SV, while maximum horizontal principal stress (SH) magnitude is
observed to vary from 90% to 92% of SV within normally pressured to over-pressured sediments. The
Keywords: breakout derived SH orientation from two well varies from N141E to N22.51E in the Krishna–Godavari
Geomechanical modeling basin. Rock mechanical properties such as Young’s modulus, Poisson’s ratio and unconfined compressive
Krishna–Godavari basin
strength have been estimated from logs of compressional and shear wave travel time. Two-dimensional
FMI log
(2D) stress modeling using finite element analysis has been carried out for some important oil/gas fields
Finite element modeling
In-situ stress situated within East and West Godavari sub-basins as a part of the current study. Regional SH orientation
has been used for application of stress at the model boundary. Discontinuities in the stress pattern which
can be associated with interfaces between weak and competent layers have been commonly observed
and especially where silici-clastic and volcanic inter-bedded sequences are encountered. The model
predicted stress orientations are verified with the Formation Micro Imager (FMI) log data of wells at the
above-mentioned sub-basins.
& 2014 Elsevier Ltd. All rights reserved.

1. Introduction sufficiently accurate geomechanical model of the oil/gas fields

Geomechanical studies are of significant interest in the areas of
geological sciences and engineering. Accurate modeling of the
several geomechanical properties is crucial in order to safely and
efficiently drill a well in technically and economically challenging
reservoirs [1–3]. Knowledge of the orientation and magnitude of
the principal stresses along with pore pressure and fracture
pressure is essential for building comprehensive geomechanical
models. Directions of principal stress orientation hardly remain
stable over lengthy intervals, and can be often seen to be rotated in
the presence of faults, salt diapirs, mountains or other complex
structures [4–8]. Subsurface sediments respond to applied stresses
through deformation, accompanied by changes of rock mechanical
properties. Faulting, lithological changes and contrasts in rock
mechanical properties within a geological formation can lead to
stress perturbations and produce local stresses that can signifi-
cantly deviate from the regional stress field [9]. A valid and
n is capable of incorporating the specific geometry as well as the
Corresponding author. Tel.: þ 91 326 2296658;
fax: þ91 326 2296563, þ 91 326 2296618.
E-mail address: rima_c_99@yahoo.com (R. Chatterjee).
would allow us to address the many challenges that characterize
basin studies and exploration, exploratory well drilling, appraisal
drilling and development drilling that could necessitate drilling of
highly deviated wells, which cannot be planned without a geo-
mechanical model of sufficient accuracy in place. Well engineering,
assurance of wellbore stability during drilling, efficient well completion
and delivery, initial testing and production profiles plans are heavily
dependent upon the availability of robust geomechanical models of
the sub surface, to start with. Even a reasonably accurate prediction of
long-term response of a reservoir to pore pressure depletion would be
impossible without valid geomechanical models of the sub surface
[10–13]. Geomechanical models of the sub surface do have great
relevance and importance in the areas of basin studies and exploration
as such models yield valuable insights into oil or gas reservoirs
presence in the subsurface and often provide huge synergies when
integrated with models.
Geomechanical reservoir models in present study attempt the
pre-drilling prediction of the local variations in stress magnitude
and orientation. This requires a numerical modeling approach that
distribution in space of mechanical properties of the subsurface
reservoir [9]. Therefore for the purpose of effective geomechanical

http://dx.doi.org/10.1016/j.ijrmms.2014.10.003
1365-1609/& 2014 Elsevier Ltd. All rights reserved.
16 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
modeling, we focus on (a) determination of the in-situ stress field,
pore pressure gradient and fracture initiation pressure gradient of
selected oil/gas fields located in the Krishna–Godavari (K–G) basin,
(b) computation of mechanical properties such as Young’s mod-
ulus, Poisson’s ratio and unconfined compressive strength from logs of
compressional/shear wave velocities (Vp/Vs), (c) construction of finite
element model (FEM) which represents the subsurface geometry
and deformation distribution, using the above well point data
(d) quantitative assessment of model’s (response) sensitivity to
variation in material properties and boundary conditions and,
finally (e) validation of stress orientation from breakout data of
Formation Micro Imager (FMI) logs.
2. Study area
The K–G basin is a passive margin pericratonic basin situated
on the Eastern Continental Margin of India (ECMI) and encom-
passes large areas both on land and offshore including those
located in deep waters. The basin itself came into existence
following rifting along ECMI craton during early Mesozoic. Both
on land part of the basin and its offshore host a large number of
structural traps that have been mapped and a large number of
them has been established through drilling [14]. The basin was
created as a result of tensional basement tectonics and is char-
acterized by ENE-WSW to NE-SW trending horsts and sub-basins/
grabens overlying a rifted basement structure. K–G basin is
subdivided into three sub-basins namely; Krishna, West Godavari
and East Godavari which are separated by Bapatla and Tanuku
horsts respectively (Fig. 1) [15,16]. It is characterized by a normal
faulting tectonic regime. West Godavari sub-basin is further
subdivided into the Gudivada, Bantumilli grabens separated by
Kaza–Kaikalur horst. The Mandapeta graben and Kavitam–Drak-
sharama high [14] are situated on either side of Tanuku horst in
the East Godavari sub-basin. In addition to the basin margin fault,
three more regional faults developed: Matsyapuri–Palakollu fault
Fig. 1. Illustrates location of six wells including two geological sections AA0 and BB0 in K–G basin [42]. Major oil/gas fields are shown near the six wells [14].
of Eocene age on land, a Miocene structure building fault in shallow
water near to the coastal area and a Pliocene structure building fault in
deeper water [17]. The thickness of sediments varies from 3 km
(Krishna graben) to over 7 km (Godavari offshore) and commercial
hydrocarbon is found in the Permo-Triassic to Pliocene reservoirs.
Geological cross-sections such as, AA0 (56 km long) and BB0 (160 km
long) (Fig. 2) show the disposition of the sedimentary formations top
and basement across N–S and SW-NE directions respectively traver-
sing the K–G basin. Subsidence in the south-Eastern part of the East
Godavari sub-basin may contributes to the formation of a steep step-
fault zone in early Paleocene Razole Formation (basalt). This fault zone
is known as the Matsyapuri–Palakollu fault zone [14]. The AA0 section
shows the sediment deposition pattern from onshore to offshore.
Razole basalts are deposited at depth in the offshore areas but occur at
shallower depth in on land.
The candidate wells chosen for the current study (and named
as KM, KA and KR respectively) are located near producing gas
fields in the basin. Sedimentary formation penetrated in the
Suryaraopeta and Mahadevapatnam fields at the West Godavari
sub-basin include Raghavapuram Shale at the base, followed
by Tirupati Sandstone, Razole volcanics and Nimakuru Sandstone
at the top, ranging in age from Early Cretaceous to Paleocene
age [14]. Raghavapuram–Tirupati is the dominant petroleum
system in the West Godavarisub-basin [17]. The wells KG, KS and
KK are located between the Rangapuram gas field and Ravva oil
field at the East Godavari sub-basin and penetrate Vadaparru Shale
at bottom followed by Matsyapuri Sandstone and Godavari Clay at
top, ranging in age from Eocene to Pleistocene. Gamma ray (GR),
resistivity (LLD), density (ρ) and sonic travel time (DT) logs have
been used to identify the top of the formations encountered in
these wells. Multi well correlation through KG–KS–KK at the East
Godavari sub-basin (Fig. 3a) identifies the top of Matsyapuri
Sandstone Formation which overlies the Vadaparru Shale and
capped by Godavari clay. The Vadaparru–Ravva/Godavari clay
system is one of the relatively younger petroleum systems in
K–G Basin of Tertiary age which includes a narrow coastal strip of
 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
Fig. 2. Interpreted geological sections AA0 and BB0 across the K–G basin [42] BMP LST, Bhimanapalli limestone and PASP Fm., Pasarlapuri formation. The blocks H and L have
been referred for finite element model set-up.
Godavari graben and a large offshore area. Matsyapuri Sandstone
Formation is the main commercial reservoir in these wells.
Similarly multi well correlation through KM–KA–KR at the West
Godavari sub-basin (Fig. 3b) identifies the top of Raghavapuram
Shale Formation underlying the Tirupati Sandstone. Raghava-
puram shale is the main source rock here. Alternating sand and
silt within Raghavapuram shale forms the reservoir [17]. The sands
within Raghavapuram Shales are sealed by overlying and under-
lying shales. Tirupati Sandstone reservoirs are capped by the
overlying Razole Formation.
3. Determination of in-situ stress, pore pressure, fracture
pressure and rock mechanical properties
In order to characterize the stresses in the main formations in
the East and West Godavari sub-basin, it is essential to determine
the orientation and magnitude of principal stresses. The procedure
for estimating in-situ stress, pore pressure, fracture pressure, rock
mechanical properties and their utilization in stress modeling is
given in the flowchart diagram (Fig. 4). For purposes of geome-
chanical modeling, the in-situ stress, pore pressure, fracture
pressure and rock mechanical properties such as Young’s modulus,
Poisson’s ratio and unconfined compressive strength are to be
obtained from well logs namely; density, sonic, Vp, Vs as described
in the flowchart. The inputs for construction of model geometry
are well log correlation and geological section. The finite element
model predicted outputs are stress magnitude and orientation
including modeling error. The model predicted stress orientation is
further to be validated by stress orientation obtained from FMI logs.
3.1. Estimation of vertical stress, pore pressure and fracture pressure
The principal stresses occurring in the earth crust acting in the
mutually orthogonal directions are the vertical stress and the
17
maximum and minimum horizontal stresses [18]. For the case of a
vertical well (as in the present study) and for a normal faulted
stress regime, the first principal stress is considered to be vertical,
corresponding to the weight of the overburden (SV), the second
principal or intermediate stress as well as the third principal
stresses are referred to SH and Sh indicating the maximum and
minimum horizontal principal compressive stresses respectively
[19–21]. At shallower depths, it is the vertical stress that consti-
tutes the major principal stress since at deeper levels, the tectonic
ones take this role and the vertical stress becomes the minor
principal component. The first principal stress i.e. vertical stress
(SV) is assumed to coincide with the vertical axis (z axis) of the
vertical well. The magnitudes of SV of six wells (KG, KS, KK, KR, KA
and KM) (Fig. 5) have been estimated by cumulative sum of rock
densities obtained from density logs (Fig. 3) from surface to the
depth of interest. Vertical stress is calculated using bulk density of
the rock which is force per unit area applied by load of rock above
the point of measurement. The required equation given [18] is
Z Z
SV ðZÞ ¼ ρðzÞgdz ð1Þ
0
where Z is the depth at point of measurement, ρ(z) is the bulk
density of the rock which is function of the depth, and g is the
acceleration of gravity.
Using the magnitude of vertical stress data, computation of
pore pressure (PP) and fracture pressures (FP) has been carried out
using Eaton’s sonic [22] and Matthews–Kelly’s equations [23],
respectively. The PP can be calculated from the vertical stress data
and can be validated with repeat formation tester (RFT) data.
In normal compacted shale, compressional wave travel time is
observed decreasing with depth due to porosity decay. Following
Eaton’s equation, the exponent is set to 3 for the K–G basin [24].
The normal compaction trend (NCT) against low permeable shale
has been identified for each of the six wells for initial pore
pressure prediction for each of the six wells. Pore pressure (PP)
18 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27

Fig. 3. Showing the log correlations for three wells in (a) East and (b) West Godavari sub-basins. Log correlation displays the sedimentary formation with geological age.

has been calculated using Eaton’s [22] sonic equation, where Ph is the hydrostatic pressure, Δtpn is the compressional
 3 wave travel time in low permeable zone calculated from normal
Δt pn compaction trend for the six wells, and Δtp is the observed
PP ¼ SV –ðSV –P h Þ ð2Þ
Δt p compressional wave travel time. Hydrostatic pressure gradient is
 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
Fig. 4. A flowchart diagram showing procedure for geomechanical modeling over the study area.
considered as 10 MPa/km for the six wells in this basin [8]. As a
next step, predicted PP has been calibrated with the RFT data to
obtain the calibrated pore pressure distribution. The RFT is
operated by an electrically driven hydraulic system so that it can
be set and retracted as often as necessary to pressure test up to
two fluid samples in a single trip in open hole [25]. Two sample
chambers are available in the tool. The tool has two pre-test
chambers of 10 cm3 volume which are opened automatically after
the tool is set to withdraw formation fluid at two differing rates.
A strain gauge pressure transducer is located in the flowline to
monitor the pressure during the test. Pressure drawdown during
this periods and the following build-up curves are recorded at the
surface and provide data for formation pressure. The predicted
pore pressures from Eq. (2) are found to be close to RFT data
acquired against specific depth levels in each of these six wells
(Fig. 5). The Predicted PP is validated by the RFT data as indicated
in Fig. 5.
The fracture pressures (FP) have been determined from Mat-
thews–Kelly’s equation [23]:
FP ¼ K i ðSv –PP Þ þ PP
where Ki is the matrix stress coefficient, defined as the ratio Sh/SV.
3.2. Estimation of horizontal stress magnitude
The minimum horizontal stress (Sh) magnitude is calculated
from the vertical stress and pore pressure data for the above-
mentioned wells in this basin. The minimum horizontal stress
magnitude for the above-mentioned six wells has been calculated
from the following equation [26–28] and can be validated with the
leak-off test (LOT) data.:
ðSv –PP Þν
Sh ¼ PP þ
ð1  νÞ
where ν is Poisson’s ratio of the rock in the K–G basin, and ranges
from 0.24 to 0.28 [29].
In practice, another valuable source of Sh data is a LOT, which is
performed at selected depths [30] providing an approximate
magnitude of least horizontal stress. A LOT is performed by
pressurizing the uncased section of a borehole until fracture open
and begin to take in fluid [31]. In a LOT, pressure is plotted as a
function of mud volume pumped into the hole. A deviation from
linear trend in this plot indicates the onset of fracture opening and
19
gives a rough estimate for the least principal stress. However, a
consistent analysis of LOT throughout the study area of K–G basin
is lacking. Thus, we have plotted limited numbers of LOT values
from these wells (Fig. 5) to improve the knowledge of Sh obtained
from Eq. (4) as well as to verify the acceptable estimate of Sh
magnitude.
The magnitude of SH can be obtained from breakout interval in
image or in dipmeter logs with prior information on Sh value and
rock strength [1]. Due to lack of hydro fracturing data for these
wells, it is not possible further to estimate or validate the
magnitude of SH with depth. Image log evaluation is preferred
to constrain maximum horizontal stress (SH) at the breakout
interval [32]. Breakouts are wellbore enlargements caused
by stress-induced failure of a well occurring 1801 apart. Stress
induced borehole breakouts observed on FMI logs are shear
failures of the borehole wall in response to horizontal compres-
sion. This failure occurs at the azimuth of Sh which is the azimuth
at which the circumferential compressive stress (Hoop Stress) is
highest [10,33,34,35]. In vertical wells, breakouts occur at the
azimuth of Sh and have a consistent orientation within a given well
ð3Þ
or field. The magnitude of SH has been computed at the breakout
intervals such as: 1297–1310 m and 2800–2830 m from FMI log,
only for two wells; KS and KA at East and West Godavari sub-
basins respectively (Fig. 6). Owing to non-availability of image or
dipmeter logs for the other wells falling under the study area, the
magnitude of SH has not been computed for these wells.
The magnitude of SH at the breakout intervals has been
computed from Sh, PP and unconfined compressive strength (C)
data using Eq. (5) where maximum circumferential compressive
stress (hoop stress) is equal to C as [33]:
3SH –Sh –2PP  ΔP ¼ C ð5Þ
ð4Þ where ΔP is the difference between wellbore pressure and forma-
tion fluid pressure.
Traditionally, rock strengths are obtained directly from labora-
tory tests on core samples. But in absence of core samples, rock
strength parameter and elastic moduli can be obtained from
empirical relationship based on well log data. For determination
of SH, the value of unconfined compressive strength (C) is to be
obtained from Matysapuri Sandstone and Raghavapuram Shale
formations in wells KS and KA at the breakout intervals, respec-
tively. Previously we had reported laboratory test results on rock
strength and elastic rock moduli from wells distributed mainly at
20 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27

Fig. 5. Earth stress profiles for (a) well KG, (b) well KS, (c) well KK,(d) well KM, (e) well KA and (f) well KR. Lines 1, 2, 3, 4 and 5 indicate hydrostatic pressure, high pore
pressure, minimum horizontal stress, fracture pressure and vertical stress. The star and triangle symbols represent the leak off test (LOT) and RFT pressure values.

the Kaza–Kaikalur horst, Bantumilli high, near Razole and Man-
dapeta oilfields including K–G offshore [36,37]. The rock strength
parameters for core samples of Raghavapuram Shale were not
tested previously. Few core samples of offshore wells were from
Matsyapuri Sandstone and Vadaparru Shale and the value of C
ranges from 16 to 45 MPa [36]. In order to compute C from well
logs, it is necessary to first compute bulk compressibility from
compressional wave travel time (Δtp) and shear wave travel time
(Δts), and density (ρ). Bulk modulus (Cb) can be computed [38] as
1000ρ
Cb ¼
½1=ðDt p Þ2 –4=3ðΔt s Þ2 
and the cohesive strength (S0) can be computed as
" #
10  6 E
S0 ¼ 0:025   ½0:008V sh þ 0:0045ð1–V sh Þ
Cb
where E is Young’s modulus and Vsh is the volume of shale
computed from gamma ray log. The unconfined compressive
strength is therefore given by
S0
C¼ ð8Þ
0:289
The value of C from Eq. (8) has been estimated as 30.93 MPa for
Matsyapuri Sand, and 62.20 MPa for Raghavapuram Shale, respec-
tively, for the above-mentioned depth interval. The estimated SH as
ð6Þ computed at the breakout intervals only for two wells is found to
be 25.01 MPa in depth 1297–1311 m and 59.03 MPa at depth
2800–2830 m.
Fig. 5 displays the variation of Ph, PP, Sh, FP and SV with depth
ð7Þ for six wells under the East and West Godavari sub-basins. Top of
overpressure zone (OPZ) has been identified for all these wells.
 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27 21

Fig. 6. FMI log indicating the breakout interval and SH direction for well KS ((a) and (b)) and for well KA ((c) and (d)) in the East and West Godavari sub-basins, respectively.

Table 1
Top of OPZ, in-situ stress, pore pressure, fracture pressure and stress ratio computed from six wells in K–G basin.

Wells Top of Sv Predicted PP gradient Predicted FP gradient (MPa/km) Sh gradient (MPa/km) Sh/Sv
OPZ gradient (MPa/km)
(MPa/km)
In Hydrostatic In In normal pressured In In normal pressured In In normal pressured
OPZ gradient OPZ sediment OPZ sediment OPZ sediment

KG 1919.16 21.00 12.80 10 18.85 17.47 15.80 14.16 0.76 0.68

KS 1431.46 21.10 12.30 10 18.75 17.40 15.47 14.08 0.72 0.66
KK 2424.00 21.80 12.32 10 19.00 17.80 15.60 14.40 0.71 0.65
KM 1650.00 22.37 12.18 10 19.42 18.05 16.04 14.75 0.72 0.66
KA 2280.00 22.85 12.30 10 19.78 18.30 16.20 14.79 0.70 0.64
KR 1870.00 21.80 11.85 10 18.90 17.89 15.61 14.52 0.72 0.67

The top of OPZ, the gradient of SV, the gradient of PP and FP of
these wells are listed in Table 1. The gradient of SV for these six
well is found to vary between 21.00 MPa/km in well KG and
22.85 MPa/km in well KM. High PP gradient in Raghavapuram
Shale as observed from wells of West Godavari sub-basin varies
from 11.85 MPa/km at well KR to 12.30 MPa/km at well KA.
Similarly, a high PP gradient is observed in Vadaparru Shale and
ranges from 12.30 MPa/km to 12.80 MPa/km from well KK to KG in
the East Godavari sub-basin. It is observed from this PP analysis
that the Vadaparru Shale in East Godavari sub-basin and Ragha-
vapuram Shale in West Godavari sub-basin are characterized by
high PP gradient. The gradient of PP varies in the overpressure
zone (OPZ) from 11.85 to 12.80 MPa/km whereas the gradient of FP
ranges 18.75–19.78 MPa/km. It can also be inferred from Fig. 5 that
the ratio of Sh/SV ranges from 0.64 to 0.68 at the normal pressured
sediment whereas it is increased and varies from 0.70 to 0.76 at
the overpressured sediments. The stress ratio SH/SV varies from
0.90 at normal pressured sediment in well KS to 0.92 at over-
pressured sediment in well KA. This results also indicates that the
ratio of SH/SV increases with the increase of pore pressure gradient.
Therefore, in this normal faulted stress regime, we can assume the
magnitude of SH equal to the magnitude of SV in absence of image/
dipmeter log as well as hydro fracturing data.
3.3. Estimation of rock mechanical properties
Mechanical properties of rocks vary significantly between
reservoirs and within a reservoir due to the wide variety of
material composition and intrinsic anisotropy exhibited by
shales. Heterogeneous rock properties in terms of layering
and complex infrastructure of fault zones are typical phenom-
ena in sedimentary basins. This section deals with the mea-
surement of rock mechanical properties for only two wells; KK
and KA, from well logs. Estimation of rock mechanical proper-
ties in the industry is associated with measurement of com-
pressional wave velocity (Vp) and shear wave velocity (Vs) at
laboratory and from well logs [39]. The Vp/Vs and density log
data are available for selected depth intervals from two wells
only. The rock mechanical parameters have been estimated for
depth intervals 1994–2577 m and 2800–2830 m for wells KK
and KA, respectively.
The unconfined compressive strength (C), elastic moduli such
as: Young’s modulus (E) and Poisson’s ratio (ν) have been
estimated from Vp, Vs and density (ρ) log data using the equations
as follows [38,40]. The shear modulus (G) is computed as
G ¼ ρV 2s ð9Þ
22 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
Fig. 7. Rock mechanical profiles for (a) well KS and (b) well KA in the East and West Godavari sub-basins.
Poisson’s ratio (ν) is given by as
2 correlations in Fig. 3 for setting up the simplified structural model.
½0:5ðV p =V s Þ –1
ν¼
½ðV p =V s Þ2  1
and Young’s modulus (E) is computed as
E ¼ 2Gð1 þ νÞ
The values of C, ν and E have been calculated based on Eqs. (8),
(10) and (11) only for Raghavpuram Shale from KA well and for
Vadaparru Shale and Matsyapuri Sand from KK well, respectively.
Fig. 7 illustrates the variation of C, E and ν with depth for these
two wells. The variation of E and ν ranges from 27 to 31 GPa and
0.26 to 0.29 respectively in Raghavapuram Shale in well KA. The
rock mechanical properties E and ν in well KK ranges from 6 to
38 GPa and 0.18 to 0.40 in Matsyapuri Sandstone as well as from
6 to 26 GPa and 0.23 to 0.38 in Vadaparru Shale Formations. The
variations in E and ν for well KA are less as observed for a short
depth interval (30 m) in Raghavapuram Shale in well KA compared
to the variations in E and ν for well KK for a larger depth interval
(583 m). The large variations in E and ν account for the change of
lithology, mainly due to occurrence of sand, limestone, claystone,
shaly sand, siltstone, shale, shale/sand alteration and reservoirs in
the Matsyapuri formation. Low value of ν (0.18 or less) indicates
gas sands in the Matsyapuri formation. The low value of E (6 GPa)
and high value of ν (0.38) are observed due to unconsolidated
sediments in Vadaparru Shale formation. The sand lenses in the
Vadaparru Shale characterize with Y of 26 GPa and ν of 0.23.
Previous laboratory test results on core samples from this forma-
tion showed variation from 15 to 42 GPa due to change of lithology
at offshore well [36].
4. Geomechanical modeling of stress using finite element
method
4.1. Model set up
Two-dimensional (2D) geomechanical modeling entails deter-
mining the distribution of stresses in the reservoir and non-
reservoir sections and hence the generalized geological sections
referred in Fig. 2 have been considered along with well log
ð10Þ Section AA0 of length 56 km parallel to N–S direction traverses
near wells KG, KS and KK in the East Godavari sub-basin in the
vicinity of Matsyapuri–Palakollu growth fault. The other section
(section BB0 ) of length 160 km striking SW-NE passes through
ð11Þ
wells KR, KA and KM, located in between Tanuku horst and
Bantumilli high. These aforementioned geological sections along
with the well correlation information, and data sets of in-situ
stress pore pressure and rock mechanical properties with respect
to sedimentary rocks exposed in the respective well sections, have
been used to create the model set-up for the two sections; KG–KK
and KM–KR respectively in East and West Godavari sub-basins.
The model-section created on the section KG–KK is 15 km long, is
aligned along NE-SW, and has a depth of 3.5 km (Fig. 8). The other
model-section created on the section KM–KR, 14 km in length with
depth coverage of 4.5 km and includes wells KM, KA and KR
(Fig. 9). The average of SH gradients from three wells (assuming,
magnitude of SH ¼ magnitude of SV, refer Table 1 for SV) has been
applied as load at the mid of layers occurring at left and right
boundaries. Rock mechanical parameters associated with each the
different layers, which have been estimated previously by [36] are
provided at Table 2, along with log data of wells; KK and KA in East
Godavari and West Godavari sub-basins respectively (Fig. 7). To
facilitate model developments, each layer in the models has been
assumed as homogeneous and isotropic. Direction of recent
compression, more or less parallels the trend of coast as well as
to the horst/graben alignment. This coincidence is a good start for
modeling compression along KG–KK and KM–KR lines, respec-
tively. Geomechanical modeling would bring out, the effect of
applied stress along regional SH direction towards NE-SW [29]
underlying the East and West Godavari sub-basins. Previous
publications on 2D models in the K–G basin show model deforma-
tions along seismic lines transgressing this basin passing through
onshore to offshore. Model deformations are observed under the
application of compressive minimum stress along Sh direction
(NW-SE) at the offshore boundaries constraining the boundary
at the continental side [37]. The model deformation goes up to
3–4 km in the vertical direction, bearing in mind that the original
vertical dimension of the model was 4–5 km. The deformations
 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
Fig. 8. (a) 2-D model for KG–KK section in the East Godavari sub-basin. Five layers (digits 1 through 5) are identified in reference to Table 2, (b) von Mises stress for the
deformed 2-D FEM, and undeformed edge, (c) vector plot of SH for the KG–KK section, (d) vector plot of SH for block P and (e) vector plot of SH for block Q.
upto 3–4 km were not explained during 2D modeling because of
the absence of log data and limited knowledge on geologic history
of the study area. In 3D model for Mandapeta field, the stress
magnitudes, SV, and horizontal stresses were applied vertically and
horizontally respectively and the model was constrained at the
surfaces at one side parallel to the Sh direction [5]. The 3D model
set-up for Mandapeta field indicated the stress modification
around normal fault and acceptable model deformation due to
the model boundary conditions. Now here, the models represent
the current geometry based on the geological section and well log
data. No deformation or movement is reported at the surface of
the study area in the tectonically passive K–G basin. Growth faults
and subsidence are reported in the offshore sediments [14,17].
Hence, in both of these models along KG–KK and KM–KR lines, the
upper surface is restricted from movement horizontally and
vertically (Figs. 8 and 9). Boundary conditions for these models
are displayed in Figs. 8 and 9. This satisfies the implicit demand
that the current geometry or structure is not allowed to deviate
too much from its initial geometry, because the model is con-
structed on the basis of present day geometry.
Based on the above, finite element models (FEMs) for the
horizontal and stratified sediments are generated and stress
perturbations with stress vector are calculated using ANSYS finite
element software package. Boundary constraints are the other key
elements of the model. These lead to bounds on the solution space
that play a crucial role in assuring robust and meaningful model
solution to be generated by the process. Model predicted results
generally produce an error value at the edge due to boundary
conditions and model constraints. The geological section across
N–S in the East Godavari sub-basin (Fig. 2a) is showing the five
different formations within block H down to a depth of 3.5 km.
Accordingly with the help of well log correlation, the first model
23
for line KG–KK has been considered as consisting of five lithologic
layers with cumulative sedimentary column of 3.5 km length. The
lithology of layers 1, 2, 3, 4 and 5 is broadly defined as follows:
Layer 1 is claystone, layer 2 is dominantly sandstone and a part of
Matsyapuri sand stone reservoir development. Layer 3 is mostly
limestone dominated, layer 4 is a part of Vadaparru Shale which is
a known source rock followed by layer 5 which is sand/shale
alteration. The 2-D model geometry has been constructed with
length 15 km along NE-SW direction and depth 3.5 km. It is
divided by 10,246 numbers of 6-noded triangular elements con-
sisting of 20,755 nodes. Loading, has been implemented, as
horizontal stresses (SH) are applied on the left side of the model
as 10.65 MPa at 500 m, 25.56 MPa at 1200 m, 36.21 MPa at
1700 m, 57.51 MPa, 2700 m and 69.22 MPa at 3250 m and stresses
applied on the right side of the model as 10.6 5 MPa at 500 m,
42.6 MPa at 2000 m and 68.16 MPa at 3200 m, respectively. The
average associated pore pressure as applied is 2.5 MPa in Layer 1,
10 MPa in Layer 2, 15 MPa in Layer 3, 31.7 MPa in Layer 4 and
25 MPa in Layer 5. Since Layer 4 lies in the Vadaparru Shale, it is
observed to have an elevated PP while the others are at normal
hydrostatic values.
The generalized geological section across SW-NE in the West
Godavari sub-basin (Fig. 2b) shows the six different formations
within block L down to a depth of 4.5 km. Accordingly with the
help of well log correlation, the second model for KM–KR section
of length 14 km with NE-SW direction has been divided into six six
stratified lithologic Layers 1, 2, 3, 4, 5 and 6. This model in ANSYS has
been divided with 10,712 numbers of six-noded triangular elements
consisting of 21,735 nodes. Layer 1 is considered as claystone overlying
the sand layer 2. Layer 3 is a part of Razole volcanics (basaltic). This is a
known regional cap rock that overlies the Tirupati sandstone reservoir
development. Tirupati Sand Formation represented by Layer 4 overlies
24 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27

Fig. 9. (a) 2-D model for KM–KR section in the West Godavari sub-basin. Six layers (digits 1 through 6) are identified in reference to Table 2, (b) Von Mises stress for the
deformed 2-D FEM and undeformed edges, (c) vector plot of SH for the KR–KM section, (d) vector plot of SH for block R and (e) vector plot of SH for block S.

Table 2 55.85 MPa at 2500 m, respectively. The hydrostatic pressure applied

Mechanical rock properties employed in the 2-D FEM for KG–KK and KM–KR are as 1.5 MPa, 7.5 MPa, 11 MPa and 15 MPa in respect of Layers 1, 2,
sections in East and West Godavari sub-basins. 3 and 4 whereas abnormal PP of 36.33 MPa is applied in Layer 5. Layer
5 lying in Raghavapuram Shale in West Godavari sub-basin is
Line Layers Lithology Young Poisson’s Density
modulus ratio (g/cm3)
characterized by high PP gradient whereas other layers are at normal
(GPa) pressured condition. PP values are not applied at the basement
(Layer 6).
East Godavari 1 Clay stonea 15 0.18 1.96
KG–KK line 2 Sanda 12 0.34 2.25
3 Limestone dominateda 50 0.25 2.70
4 Shalea 35 0.23 2.40 4.2. Analysis and discussion of model result
5 Sand/Shale 13 0.35 2.47

West Godavari 1 Clay stonea 15 0.18 2.04

This study builds a model by analyzing geological sections of
KM–KR line 2 Sand 20 0.25 2.30 the upper crust (up to 3.5 to 4.5 km) and well log correlation
3 Basalta 70 0.25 2.76 considering unstressed state. In-situ stress magnitude and orien-
4 Compacted sanda 42 0.17 2.32 tations are not known spatially through the vertical section in this
5 Shale 30 0.28 2.50
initial geometry. To meet the objective for predicting the local
6 Basementa 80 0.25 2.74
variation of stress magnitude and orientation, the model deforma-
a
Material properties are assigned from previous studies [36]. tions is kept as minimum as possible with the application of
boundary conditions. Model deformation is restricted within
1.25 km from the edge of the vertical boundaries. The scale
Raghavpuram Shale (which is known to be source rock), and is of deformation goes up to 0.1 km in the vertical direction and
modeled as Layer 5 here. Basement has been considered as Layer 6. 0.8–1.25 km in the horizontal direction. The model deformation is
While the cumulative length of the sedimentary column is 3.9 km and practically negligible at the central section of the models. The
total depth considered in this model is 4.5 km. Horizontal stress response to variation of material properties and application of far-
loading has been implemented on either side and the applied stress field stress to these models show the distribution of stress
magnitudes are 16.75 MPa at 750 m, 30.15 MPa at 1350 m and trajectories of the present day geological model. Initial geometry
 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
at the central part of this model remains unchanged through the
modeling. Therefore, the modeling results can be verified with the
present day stress orientations obtained from FMI log data.
The numerical modeling results indicate the deformation of
sedimentary rocks under the above mentioned boundary condi-
tions. The von Mises stress (which is the equivalent stress
computed from the three principal stresses) is displayed for the
model along KG–KK in Fig. 8a. The value of von Mises stress ranges
from 4.3 to 31.6 MPa at the central part of the model. The reservoir
sands are located at the Layer 2 which corresponds to Matsyapuri
Sandstone Formation. The layer is showing equivalent stress
magnitude of 10–20 MPa (Fig. 8b). The Layer 3 is mainly comprised
of limestone and having high E value compared to the other layers.
Due to material property contrast, stress concentration is more
(Fig. 8b–d) and stress magnitude at the Layer 3 is also high
compared to that of the other layers. Due to presence of high PP,
the von Mises stress in the Vadaparru Formation as the Layer 4 is
more than the Layer 5 but less than that of Layer 3. The stress
vector plots for these models under K–G basin illustrate the
orientation of principal stress (SH) in the vertical plane. The
models predict rotation of stress about horizontal axis at the
location of strong gradient in elastic properties. The stress vector
plot at Layer 2 shows variation of SH from N101E to N201E in
Matsyapuri Sand and FMI log data for this interval also indicates
the SH orientation of N141E (Fig. 8c and d). The stress orientation is
rotated from Layer 2 to Layer 3 by an amount of 201 due to contrast
of material property. The SH orientation has been ascertained to
rotate at the interface between Layers 2 and 4 as well as between
Layers 3 and 4. The SH orientation is N301E at the Layer 4 with a
variation of 5–101 at the interfaces. Stress rotation is observed in
the high pressure zone of Layer 4.
Similarly the next model (Fig. 9a) is analyzed in term of stress
pattern and stress vector from FEM generated model. The value of
von Mises stress displays value ranging from 1.3 to 15.3 MPa in the
first two layers and then it increases gradually from Layer 3 to Layer
5 varying from 8.5 to 30.2 MPa (Fig. 9b). In basement (defined as
Layer 6), the magnitude of von Mises stress has attained very large
value compared to the stress magnitude at above layers. It is known
that the local stress is perturbed at the interfaces between strong
and weak layers. This hypothesis has been verified by observing a
rotation of SH of about N201E in Layer 3 comprised of strong basaltic
rocks than the relatively weak Layers 2 and 4. The SH orientation is
observed as N15–201E at Layer 4 in the Tirupati Sandstone which
holds the reservoir in the wells under section KM–KR. The stress
orientation at the Layer 5 in the high pressure zone is rotated
from 7 to 151. The direction of the SH at Layer 5 especially in middle
part of the model is varying from N221E to N301E.
The model predicted SH orientation has been validated with the
local stress orientation obtained from FMI log for one well in each of
two models. Breakout intervals have been observed in Matsyapuri
Sandstone in well KS (KG–KK section) and in the Raghavapuram
Shale in well KA (KM–KR section). The orientations of SH have been
displayed in rosette diagram at the breakout intervals for 1297–
1310 m and for 2800–2830 m in wells KS and KA respectively
(Fig. 6b and d). The SH orientation is N141E for well KS and
N22.51E for well KA at East and West Godavari sub-basins, respec-
tively. The model predicted SH orientation at Layer 2 ranges from
N101E to N201E in Matsyapuri Sand. FMI log data at well KS
indicating SH orientation of N141E supports the model predicted
stress orientation at the East Godavari sub-basin. Similarly, the
direction of the model predicted SH orientation at Layer 5 under the
KM–KR section in West Godavari sub-basin is varying from N221E
to N301E which is validated by breakout derived SH orientation from
KA well in the depth interval 2800–2830 m. Therefore model
predicted SH orientation provides the present day stress direction
in the reservoirs belonging to Matsyapuri Sandstone, Tirupati
25
Sandstone and Raghavapuram Shale. The estimated SH orientation
corroborates well with the northeastward movement of India. More-
over, this result confirms the previous SH orientation predicted by four-
arm dipmeter log by [29].
4.3. Stress error analysis
The finite element analyses for the stress modeling are subject
to several types of errors that can compromise the validity of
results. The errors in any type of finite element modeling include
incorrect model set-up, incorrect assumptions and simplifications
used in the model as well as insufficient mesh discretization. The
material properties like ρ, E and ν for each lithologic layer may be
varying laterally and vertically within each layer. Due to lack of
well log, core and seismic data, it is not possible presently, to know
the variation of material properties within each lithologic layer.
Errors due to incorrect input of material properties have been
prevented using average material properties from well logs and
literature survey, whereas load specifications are known from
density logs at particular depths. Boundary constraints have been
applied to this model as predicted previously from regional stress
orientation in K–G basin [29]. Further the model geometry has
been simplified using interpreted geological section and well logs
correlation for East and West Godavari sub-basins. The main
source of error in this finite element analysis arises due to
insufficient mesh discretization of the model. The error due to
mesh discretization has been remedied by evaluating the quality
and adequacy of the mesh and by developing and utilizing criteria
that characterize the accuracy of solution [41]. The stress contour
for each element in finite element mesh is approximated based on
the displacement of the nodes of the element. The stress contour
will not necessarily be continuous from one element to the next.
Consequently, the discontinuity of the stress contour from one
element to another engenders a discretization error. The size of
each element is to be reduced in such a way that error becomes
close to zero. For each element the stress error at each node is
approximated by the following equation:
Stress error at node n of element i, {ΔSin} ¼{San}  {Sin}, where Sin
is the stress vector of node n of element i and San is the averaged
stress vector at node n. The stress error for these two models
located at East and West Godavari sub-basins are displayed in
Fig. 10a and b, respectively. The stress error is ranging from
0.123e  14 to 0.76e 04 MPa at the mid part of model to a
maximum of 0.137e  03 MPa at the model boundary for line
KG–KK at East Godavari sub-basin. The stress error plot for line
KM–KR at the West Godavari sub-basin is varying from 0.109e  15
to 0.234e  04 MPa at the mid part of the model with maximum
error at the model edges. Limestone dominated layer in the first
model and basalt as well as basement layers in second model have
shown the stress error in the order of 10  4 MPa which is more
than the stress errors observed at the adjacent layers. The stress
error increases at these layer interfaces due to high Young’s
modulus values. The modeling output with less stress errors in
the East and West Godavari sub-basins may be considered as one
of the best possible solution for these geological situations
under study.
4.4. Impact for the reservoirs
Magnitudes and orientation of the model predicted stress in
reservoirs occurring in Tirupati Sandstone, Matsyapuri Sandstone
and Raghavapuram Shale Formations of six wells near to the oil/
gas fields namely; Mahadevapatnam, Suryaraopeta, Razole and
Rangapuram may be used for several aspects of hydrocarbon
exploration like; well design, well location, production optimization,
26 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
Fig. 10. Stress error plot (a) for 2-D model for KG–KK section in the East Godavari sub-basin and (b) for 2-D model for KM–KR section in the West Godavari sub-basin.
designing fracture stimulation treatments, understanding and
controlling casing deformations. Model predicted stress data may
be used for assessing caprock seal capacity (and therefore seal
failure analysis) and determining stress dependent reservoir
properties such as permeability, porosity and compressibility.
In a high stress regime, the time bound changes in reservoir
pressure can result in changes in total compressibility, voidage and
permeability, having adverse impact on assurance of optimal
withdrawal during the life cycle of a producing field. The stress
distribution data will allow robust forward modeling and realistic
reservoir simulation for timely intervention and a structured plan
of pressure maintenance and voidage compensation to yield
maximum fruit. The information on in situ stress distribution
can be used for developing further insights into the drainage
pattern of the reservoir.
5. Conclusions
The methodology presented above can be used to build
geomechanical finite element models for various types of reser-
voirs ranging from field-scale models to smaller, highly detailed
submodels of specific geometry. Geomechanical parameters such
as the in-situ stress, pore pressure, fracture pressure and mechan-
ical properties of rocks have been estimated from log data of East
and West Godavari sub-basins. In-situ stress has significant impact
on reservoir properties and reservoir performance. This makes the
modeling crucial for developing a comprehensive simulation of
reservoir dynamics and plan necessary inputs for assurance of
optimal withdrawal for maximizing the economic returns as the
fields are exploited. The in situ stresses predicted by the finite
element model have been calibrated against stress orientation
data actually observed, e.g. borehole breakouts. Changes in rock
mechanical properties have led to stress perturbations and pro-
duce local stresses showing significant deviation from the regional
stress field. The local stress vectors of K–G basin are oriented sub-
parallel and parallel with NE-SW in the direction which is the
significant direction characterizing basinal horst graben morphol-
ogy. Present day stress direction observed in two wells of K–G
basin corroborates the NNE movement of the Indian plate.
Although the rock mechanical properties of reservoir and stress
orientation have been addressed through two wells there is still a
need for more laboratory and/or log data for a more realistic
spatial distribution rock mechanical properties under reservoir
conditions for which stress sensitivity is significant, as well as to
quantify model predicted variations with field data. A robust
model of rock mechanical properties distribution in three dimen-
sions in the sub-surface is the foundation of stress field distribu-
tion in the sub-surface. The current geomechanical reservoir
model that has been calibrated against well data can be used for
stress predictions in the inter-well space and undrilled parts of the
reservoir.
Acknowledgement
Authors express sincere gratitude to ONGC, India for providing
the concern well data and geological information to carry out this
scientific work. Authors are highly obliged to Mr. K. M. Sundaram,
Former Executive Director, ONGC and Advisor, ONGC for careful
editing of the manuscript. The editing by Mr. Anshuman, ISM is
also thankfully acknowledged. Authors acknowledge the financial
support from the Ministry of Earth Science through the R&D
project MoES /P.O./(Seismo)/1(138) 2011 dated 9.11.12.
 D.K. Singha, R. Chatterjee / International Journal of Rock Mechanics & Mining Sciences 73 (2015) 15–27
References
[1] Plumb R. The mechanical earth model concept and its application to high-risk
well construction projects. In: Proceedings IADC/SPE drilling conference.
Louisiana; 23–25 February 2000.
[2] Moos D, Peska P, Finkbeiner T, Zoback M. Comprehensive wellbore stability
analysis utilizing quantitative risk assessment. J Petrol Sci Eng
2003;38:97–109.
[3] Cook J, Frederiksen RA, Hasbo K, Green S, Judzis A, Martin JW, et al. Rocks
matter: ground truth in geomechanics. Oilfield Rev 2007;19:36–55.
[4] Ciulavu DC, Cloetingh SAPL. Late Cenozoic tectonic evolution of the Transyl-
vanian basin and north Eastern part of the Pannonian basin (Romania):
constraints from seismic profiling and numerical modeling. EGU Stephan
Muller Spec Publ Ser 2002;3:105–20.
[5] Chatterjee R. Effect of normal faulting on in-situ stress: a case study from
Mandapeta field, Krishna–Godavari basin, India. Earth Planet Sci Lett
2008;269:458–67.
[6] Moeck I, Schandelmeier H, Holl HG. The stress regime in a Rotliegend reservoir
of the North East German Basin. Int J Earth Sci 2009;98:1643–54.
[7] Heidbach O, Tingay M, Barth A, Reinecker J, Kurfeß D, Müller B. Global crustal
stress pattern based on the World Stress Map database release 2008.
Tectonophysics 2010;482:3–15.
[8] Chatterjee R, Mukhopadhyay M, Paul S. Overpressure zone under the Krishna–
Godavari offshore basin: geophysical implications for natural hazard in
deeper-water drilling. Nat Hazard 2011;57:121–32.
[9] Fischer K, Henk A. A work flow for building and calibrating 3-D geomechanical
models—a case study for a gas reservoir in the North German Basin. Sol Earth
2013;4:347–55.
[10] Bell JS. In situ stresses in sedimentary rocks (Part 1): Measurement techniques.
Petrol Geosci 1996;23:85–100.
[11] Zoback MD, Barton CA, Brudy M, Castillo DA, Finkbeiner T, Grollimund BR,
et al. Determination of stress orientation and magnitude in deep wells. Int J
Rock Mech Min Sci 2003;40:1049–76.
[12] Qin Q, Zhang L, Deng H, Su PD, Wang ZY. Determination of magnitude of
palaeo-tectonic stress and application to tectonic modeling. Chin J Rock Mech
Eng 2004;23:3979–83.
[13] Smart KJ, Ferrill DA, Morris AP, Bichon BJ, Riha DS, Huyse L. Geomechanical
modeling of an extensional fault-propagation fold: big brushy Canyonmono-
cline, Sierra DelCarmen, Texas. AAPG Bull 2010;94:221–40.
[14] Rao GN. Sedimentation, stratigraphy and petroleum potential of Krishna–
Godavari basin, East Coast of India. AAPG Bull 2001;85:1623–43.
[15] Sastri VV, Sinha RN, Singh G, Murti KVS. Stratigraphy and tectonics of
sedimentary basins on the East coast of Penninsular India. AAPG Bull
1973;57:655–78.
[16] Sastri VV, Venkatachala BS, Narayanan V. The evolution of East coast of India.
Palaeogeogr Palaeoclimatol Palaeoecol 1981;36:23–54.
[17] Gupta SK. Basin architecture and petroleum system of Krishna Godavari Basin,
East coast of India. The Leading Edge 2006;25:830–7.
[18] Zhang L. Three-dimensional numerical models of drilling induced core
fractures. MSc thesis. University of Alberta; 2011.
[19] Jaeger JC, Cook NGW, Zimmerman RW. Fundamentals of rock mechanics. 4th
ed.. Malden, Massachusetts: Blackwell; 2007.
[20] Brady BHG, Brown ET. Rock mechanics for underground mining. 3rd ed..
Dordrecht: Springer; 2004.
[21] Brown ET, Hoek E. Trends in relationships between measured in-situ stresses
and depth. Int J Rock Mech Min Sci Geomech Abstr 1978;15:211–5.
[22] Eaton BA. Graphical method predicts geopressures worldwide. World Oil
1972;182:100–4.
27
[23] Matthews WR, Kelly J. How to predict formation pressure and fracture
gradient. Oil Gas J 1967;65:92–106.
[24] Chatterjee A, Mondal S, Basu P, Patel BK. Pore pressure prediction using
seismic velocities for deepwater high temperature–high pressure well in
offshore Krishna Godavari Basin, India. In: Proceedings SPE oil and gas India
conference and exhibition, Mumbai, India; 28–30 March 2012.
[25] Combescure JP, Pallock D, Wittmann M. Application of repeat formation tester
pressure measurement in the middle east. In: Proceedings middle east oil
technical conference, Bahrain; 25–29 March 1979.
[26] Whitehead WS, Hunt ER, Holditch SA. The effects of lithology and reservoir
pressure on the in-situ stresses in the Waskom (Travis Peak) field, presented
at the low permeability. In: Proceedings reservoir symposium. Denver; 18–19
May 1987.
[27] Engelder T. Stress regimes in the lithosphere. Princeton, NJ: Princeton
University Press; 1993.
[28] Hillis R. Pore pressure/stress coupling and its implications for seismicity.
Explor Geophys 2000;31:448–54.
[29] Chatterjee R, Mukhopdhyay M. In-situ stress determination using well log
data for the oil fields of the Krishna–Godavari basin. Petrophysics
2002;43:26–7.
[30] Brent ACS, Alvin WC. Stress determination inactive thrust belts: an alternative
leak-off pressure interpretation. J Struct Geol 2010;32:1061–9.
[31] Grollimund B, Zoback MP, Wiprut DJ, Arnesen L. Stress orientation, pore
pressure and least principal stress in the Norwegian sector of the North Sea.
Petrol Geosci 2001;7:173–80.
[32] Chhajlani R, Zheng Z, Mayfield D, MacArthur B. Utilization of geomechanics for
Medusa field development, Deepwater Gulf of Mexico. In: Proceedings SPE
annual technical conference and exhibition. San Antonio; 29 September–2
October 2002.
[33] Barton CA, Zoback MD, Burns KL. In-situ stress orientation and magnitude at
the Fenton Geothermal site, New Mexico, determined from wellbore breakout.
Geophys Res Lett 1988;15(5):467–70.
[34] Zoback MD, Moos D, Mastin LG, Anderson RN. Well bore breakouts and in situ
stress. J Geophys Res 1985;90:5523–30.
[35] Tingay M, Reinecker J, Muller B. Borehole breakout and drilling-induced
fracture analysis from image logs, world stress map project stress analysis
guidelines: image logs; 2008.
[36] Chatterjee R, Mukhopadhyay M. Petrophysical and geomechanical properties
of rocks from the oilfields from Krishna–Godavari and Cauvery Basins, India.
Bull Eng Geol Environ 2002;61:169–78.
[37] Chatterjee R, Mukhopadhyay M. Stress modeling for the oil and gas fields of
Krishna–Godavari and Cauvery basins, India, using Finite Element Technique.
Petrophysics 2003;44(5):342–50.
[38] Mohammed YA, Zillur RA. Mathematical algorithm for modeling geomecha-
nical rock properties of the Khuff and Pre-Khuff reservoirs in Ghawar Field. In:
Proceedings SPE middle East oil show, Bahrain; 17–20 March 2001.
[39] Boonen P. Advantages and challenges of using logging-while-drilling data in
rock mechanical log analysis and wellbore stability modeling. In: Proceedings
AADE national technology conference. Texas; 1–3 April 2003.
[40] Potter CC, Foltinek DS, Formation elastic parameters by deriving S-wave
velocity logs. CREWES report,9; 1997.10-23.
[41] Shah C. Mesh discretization error and criteria for accuracy of finite element
solutions. In: Proceedings ANSYS 5.7 users conference, Pittsburgh; 2002.
[42] Venkatarengan R, Rao GN, Prabhakar KN, Singh DN, Awasthi AK, Reddy PK.,
et al. Lithostratigraphy of Indian Petroliferous basin, Document VIII: Krishna–
Godavari basin, KDM Institute Petroleum Exploration ONGC, Dehradun, India;
1993. p. 1–27 (unpublished report).