Results in Engineering 24 (2024) 103016

Contents lists available at ScienceDirect

Results in Engineering
journal homepage: www.sciencedirect.com/journal/results-in-engineering

Wellbore stability analysis based on the combination of geomechanical and

petrophysical studies
Abdolmanan Ounegh a , Atefeh Hasan-Zadeh b,** , Mohammad Mohammadi Khanaposhtani b,
Yousef Kazaemzadeh c,*
a
Chemical Engineering, Fouman Faculty of Engineering, College of Engineering, University of Tehran, Fouman, Iran
b
Fouman Faculty of Engineering, College of Engineering, University of Tehran, Fouman, Iran
c
Department of Petroleum Engineering, Persian Gulf University (PGU), Bushehr, Iran

A R T I C L E I N F O A B S T R A C T

Keywords: Determining the physical and mechanical properties of rocks is crucial for planning drilling operations. Ignoring
Geo-mechanical modeling petrophysical charts geomechanical studies and tools has led to millions of dollars in additional costs related to drilling, meshwork
RFT test optimization, production optimization, casing collapse, and other issues. These costs can be managed with a
thorough geomechanical study. Additionally, some costs may not be apparent in the short term but will become
significant over the life of the field. Long-term reservoir management is essential to control these costs. Issues
such as reservoir compression, rock cover fractures, and gas storage, if neglected in the short term, can lead to
substantial and irreparable damage to the reservoir over time. The "geomechanical model" is fundamental to
geomechanical studies, particularly regarding wellbore stability. Given the importance of reservoir geo­
mechanics and petrophysics, this research investigates wellbore stability analysis by combining geomechanical
and petrophysical studies through the construction of a one-dimensional geomechanical model of the well. First,
the dynamic geomechanical coefficients of the rock for the studied depth range were calculated from wells and
petrophysical charts. Then, the static geomechanical coefficients of the rock were estimated based on laboratory-
developed relationships. Subsequently, rock mechanics test results were used to estimate rock resistance pa­
rameters. The pore pressure was estimated using the compressibility and Eaton methods for the studied range
from Well No. 5. Comparing the results of these two methods with RFT test data indicated that the compress­
ibility method is more accurate than Eaton’s method. By estimating the in situ and induced stresses in the well’s
studied area, the lower and upper limits of the safe mud weight were calculated using the Mohr-Coulomb and
Hooke-Brown criteria. Comparison of the results showed that the Mohr-Coulomb criterion is more accurate and,
moreover, more conservative than the Hooke-Brown method.

1. Introduction stresses through one-dimensional geomechanical modeling of the well,

both before drilling and while monitoring changes during or after the
A comprehensive understanding of the state of stress before pro­ operation. Ignoring geomechanical studies and tools can result in mil­
duction and its changes over time is crucial for the safe and economical lions of dollars in additional costs related to drilling, casing collapse, and
extraction of hydrocarbons from reservoirs and underground storage. other issues. However, with a minimal amount of study, these costs can
Subsurface stresses impact various aspects, including well wall stability, be managed. Moreover, some costs may not be apparent in the short
rock cover integrity, changes in rock properties due to induced stresses, term but will become significant over the life of the field. These
fault reactivation, reservoir compression and settlement, and engineer­ long-term costs can only be controlled through a comprehensive view of
ing operations within the radius of influence [1–3]. Therefore, it is reservoir management. Issues such as reservoir compression, cap rock
essential to determine the state (magnitude and direction) of these fractures, and gas storage, if neglected in the short term, can lead to

* Corresponding author.
** Corresponding author.
E-mail addresses: [email protected] (A. Ounegh), [email protected] (A. Hasan-Zadeh), [email protected] (M. Mohammadi Khanaposhtani),
[email protected], [email protected] (Y. Kazaemzadeh).

https://doi.org/10.1016/j.rineng.2024.103016
Received 15 August 2024; Received in revised form 13 September 2024; Accepted 26 September 2024
Available online 27 September 2024
2590-1230/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
A. Ounegh et al. Results in Engineering 24 (2024) 103016

substantial and irreparable damage to the reservoir over time. The Table 1
"geomechanical model" is fundamental to geomechanical studies, Evaluation of the complete log graphs of the studied well.
particularly in terms of ensuring stability. Parameters Statistical Parameter Studied Well
To build a geomechanical model, data from porosity, density, and
Depth (Meter) Minimum 2552
acoustic logs are typically used, and the model is calibrated with core Maximum 2853/2943
samples and well tests. This research emphasizes the importance and Driilling Bit Diameter (Inches) – 8/5
application of geomechanics and petrophysics in the oil and gas in­ Caliper (Inches) Minimum 8/198
dustry. A comprehensive workflow and flowchart were presented for Average 10/1596
Maximum 23/87
geomechanical-petrophysical analysis to examine stable and unstable Gamma Ray (GAPI) Minimum 1/906
intervals. Average 18/7874
Several studies have contributed to the development of geo­ Maximum 79/24
mechanical models. In Ref. [4] presented a mathematical model for Porosity (Percent) Minimum 0/001
Average 0/0476
determining the geomechanical properties of a field in Saudi Arabia.
Maximum 0/1841
Also in Ref. [5] developed a geomechanical model to analyze stability Density (G/Cc) Minimum 2/187
and determine the mud window for a sand reservoir in Saudi Arabia. In Average 2/4922
Ref. [6] constructed a model to predict wellbore instability. In Ref. [7] Maximum 2/768
designed a model for wellbore stability analysis. Bagheri et al. [8] pre­ Pressure Wave Slowness (Us/Ft) Minimum 50/88
Average 71/0187
sented a model for analyzing wellbore stability in a carbonate formation, Maximum 99/53
and Almasi and Mohsenipour [9] determined the mud window and Shear Wave Slowness (Us/Ft) Minimum 93/8685
analyzed wellbore stability through analytical and numerical methods Average 133/6112
for a well in southwestern Iran. Maximum 199/7395
This study addresses one of the key topics in the oil and gas industry,
namely wellbore stability, by integrating geomechanical and petro­
study, no tests were conducted to measure the minimum in-situ stress;
physical approaches within the Geolog software. Typically, wellbore
only pore pressure measurements were performed, which will be
stability is analyzed using geomechanics; however, in the current
described below. Information from the RFT pore pressure test for Well
research, a comprehensive method combining reservoir geomechanics
No. 5 at 17 points is presented in Table 2.
and petrophysics has been utilized to construct a geomechanical model
and analyze wellbore stability. Furthermore, this study details the pro­
4. One-dimensional geomechanical modeling of the well
cess of deriving dynamic and static coefficients and uses various
methods to compare and demonstrate the superiority of each method in
Creating a one-dimensional model of the well before drilling is
estimating pore pressure and determining the safe mud window (SMW)
crucial, as it allows for informed decisions regarding the stabilization of
in the studied formation. In this research, rock dynamic coefficients for
the wellbore under a specific drilling fluid density, considering the
the studied depth range were initially calculated using petrophysical
geomechanical conditions of the studied area. Fig. 2 illustrates the
charts. The static geomechanical coefficients were then estimated based
process of one-dimensional geomechanical modeling for the well.
on laboratory-developed relationships. Subsequently, results from
In this process, the dynamic elastic coefficients of the formation are
various rock mechanics tests were used to estimate rock resistance pa­
first determined based on petrophysical data. Then, using established
rameters. Pore pressure was estimated using different methods for the
relationships between dynamic and static properties, the static elastic
studied area from Well No. 5. Finally, by estimating the local and
coefficients of the formation are calculated, which require calibration
induced stresses in the well’s studied range, the lower and upper limits
through laboratory studies on core samples. Rock resistance parameters,
of the safe mud weight were calculated and compared using various
such as uniaxial compressive strength, confined triaxial compressive
criteria.
strength, internal friction angle, and tensile strength, are determined
using relationships developed from laboratory studies.
2. Research zone
In the next step, pore pressure and in-situ stresses are estimated and
calibrated using in-well tests. Finally, based on the induced stresses
The studied field is located in the northern part of the Dezful subsi­
generated during drilling, the safe mud weight window is determined
dence area. This field contains four reservoirs: Asmari, Ilam, Saruk, and
using the rock fracture criterion. By adjusting strain values in horizontal
Khami. The research primarily focused on the Asmari formation, which
directions, the safe mud window can be aligned with the wellbore
is the main reservoir of the field. The data used in this study includes
events.
information from Well No. 5, located in an oil field in southwestern Iran.
The collected information will be analyzed based on petrophysical logs
5. Development of rock properties estimation models based on
and in-well tests.
laboratory results

3. Research method To determine geomechanical parameters from petrophysical logs

and build geomechanical models, it is essential to use the results of
In this section, a statistical evaluation of the comprehensive logs laboratory studies conducted on core samples taken from the reservoir.
based on depth is provided, including the following: diameter log (in Several empirical relationships have been developed to estimate rock
inches), drill diameter (in inches), gamma ray log, porosity log, depth properties based on laboratory results, but these relationships are only
log, shear wave travel time log, and pressure wave travel time log, as valid if the rocks share similar lithology and reservoir conditions. In this
shown in Table 1. The well-studied was drilled with an 8.5-inch (21.59 study, laboratory data from cores taken from the same formation were
cm) section. The changes in petrophysical values along the depth profile used to calibrate the geomechanical model and develop empirical
for the reservoir area of the well are presented in Fig. 1. This figure il­ relationships.
lustrates that the variations in petrophysical profiles align well with the Laboratory studies conducted by the South Oil-Bearing Zones Com­
lithology changes depicted in the left column of the figure. pany measured dynamic and static elastic moduli, uniaxial compressive
The availability of data from pore pressure tests and minimum in-situ strength, density, and internal friction angle for 40 core samples. Using
stress measurements is crucial for constructing a one-dimensional geo­ these values and applying the linear regression method, Equation (1)
mechanical model of wells. It is important to note that in the well under

2
A. Ounegh et al. Results in Engineering 24 (2024) 103016

Fig. 1. Input parameters (Well Logs) used in the analytical method for constructing the geomechanical model.

relationship between static Young’s modulus and uniaxial compressive

Table 2
strength on a cross-plot, where the relationship effectively predicts the
Values of pore pressure measured in the studied well (well 5).
uniaxial compressive strength of the rock.
Row Depth of Pore Pressure Value of Pore Pressure Value of Finally, the internal friction angle demonstrated the strongest cor­
Measurement (M) RFT Test (Psi) RFT Test (Mpa)
relation with rock density, expressed in Equation (3). Fig. 5 evaluates
1 2548/3 3748 25/8415 this relationship on a density versus internal friction angle cross-plot. As
2 2564/5 3765 /25/9587 shown, the model is highly accurate in predicting the internal friction
3 2595/5 3833 26/4267
4 2603/5 3842 26/4778
angle.
5 2620/3 3902 26/9033
Esta = 0.7278 × Edyn + 18.752 (1)
6 2648 3887 26/7999
7 2670 3906 26/9308
8 2692 3924 27/056 UCS = 3.5153 × Esta − 48.821 (2)
9 2701 3928 27/0825
10 2723 3954 27/2549 ∅ = 65.704 × RHOB − 140.61 (3)
11 2747 3977 27/4203
12 2767 3998 27/5584
13 2773 4004 27/5996 6. Calculation of dynamic coefficients
14 2786 4014 27/6756
15 2797 4026 27/7584 As previously mentioned, the first step in constructing a one-
16 2812 4041 27/8617
dimensional geomechanical model of a well is determining the dy­
17 2822 4051 27/9307
namic coefficients of the rock. These dynamic coefficients include
Young’s modulus, Poisson’s ratio, shear modulus, and bulk modulus,
was developed to estimate static Young’s modulus from its dynamic which can be calculated using the relationships provided in Table 3 and
counterpart. Fig. 3 shows a cross-plot of dynamic versus static Young’s the petrophysical logs (see Table 4).
modulus with a fitted regression line. The figure illustrates that the fitted
model accurately estimates the static Young’s modulus. 7. Calculation of static coefficients
The regression method was also applied to estimate uniaxial
compressive strength, showing the best correlation with static Young’s The static elastic coefficients of rock essential for one-dimensional
modulus, as expressed in Equation (2). Fig. 4 presents the regression geomechanical modeling include Young’s modulus and Poisson’s
ratio. To estimate these coefficients, relationships between dynamic and

3
A. Ounegh et al. Results in Engineering 24 (2024) 103016

Fig. 2. Flowchart of the process of one-dimensional geomechanical modeling of the well to determine the safety window of mud weight.

Fig. 3. Results of laboratory tests and the regression relationship between static and dynamic Young’s modulus.

static coefficients derived from laboratory results are utilized. In this 8. Estimation of rock resistance parameters
study, the static Young’s modulus can be estimated using laboratory
results from cores of the same formation, as described in Equation (1). Resistive parameters include uniaxial compressive strength, internal
Since laboratory measurements of Poisson’s ratio were not con­ friction angle, and tensile strength. The uniaxial compressive strength
ducted, the best empirical relationship for estimating this parameter and internal friction angle of the rock for the studied area in the well
must be selected from those presented for carbonate reservoirs by other were estimated using Equations (2) and (3), respectively. Typically, the
researchers. Recently developed empirical relationships for estimating tensile strength is approximated as one-tenth of the rock’s uniaxial
the static Poisson’s ratio in carbonate reservoirs in southwestern Iran are compressive strength.
listed in the table. Anemangely et al. [10] indicated that for fields su­ Fig. 6 illustrates the variations in rock resistance parameters across
pervised by the South Oil Regions Company, Equation (8) is used to the depth range studied in Well No. 5. As depicted in the figure, the
estimate the static Poisson’s ratio. Therefore, this relationship was trends in these parameters align well with the lithology changes shown
applied to estimate the static Poisson’s ratio for the Ahvaz field, which is in the left column of the figure.
also under the supervision of the South Oil-Bearing Zones Company.
Furthermore, a comparison of the results from these two methods 9. Estimation of pore pressure
showed that the model presented by Anemangely et al. [10] provides
superior accuracy compared to the other models. In porous media, part of the pressure from overburden layers is
supported by the pore fluid, which exerts pressure on the pore walls,
known as pore pressure. Accurate knowledge of pore pressure is crucial
for exploratory and drilling projects to determine the appropriate mud

4
A. Ounegh et al. Results in Engineering 24 (2024) 103016

Fig. 4. Results of laboratory tests and the regression relationship between unconfined compressive strength and static Young’s modulus.

Fig. 5. Results of laboratory tests and the regression relationship between internal friction angle and density.

Table 3 Table 4
Equations for calculating rock dynamic coefficients from petrophysical logs Experimental equations presented for estimation of static Poisson ratio of cor­
collected in the well. responding dynamic type in research background.
Equations Unit Parameter Name Equation Equations Equation Number Reference
Number
νstatic = νdyn 8 Anemangely et al. [10]
Edyn = GPa Elastic modulus or Young’s 3–4 νstatic = − 0.8461νdyn + 0.5106 9 Mohamadian et al. [11]
( )
3Vp2 − 4Vs2 modulus
ρVs
Vp2 − Vs2
–
V2 − 2Vs2 Poisson’s ratio 3–5
νdyn = (p ) rely on physical and petrophysical characteristics and their deviations
2 Vp2 − Vs2
from normal conditions. This approach typically involves analyzing
Gdyn = ρVs2 GPa Shear modulus 3–6
( ) deviations in petrophysical profiles, such as acoustic and resistivity
4 GPa Bulk modulus 3–7
Kdyn = ρ Vp2 − Vs2 profiles, to infer changes in pore pressure.
3
In this study, pore pressure was directly measured in the well.
Experimental methods were employed to estimate pore pressure for Well
weight, thereby preventing wellbore instability, blowouts, drilling fluid No. 5, and these methods demonstrated superior performance in this
leakage, sand production, and reservoir damage. context. While all experimental methods for estimating pore pressure
Pore pressure can be determined using either direct or indirect were originally developed for sandstone reservoirs, their application to
methods. Direct methods involve tools that measure pore pressure carbonate reservoirs—due to the high heterogeneity of these for­
directly, such as borehole pressure measurement tests. Indirect methods mations—requires caution in drilling process design and geomechanical

5
A. Ounegh et al. Results in Engineering 24 (2024) 103016

Fig. 6. Estimated geomechanical parameters for well5.

studies [12]. Among these methods, Eaton’s method [13,14] has been method using effective stress parameters (σeff ), compressibility volume
widely used by industry experts for estimating pore pressure in car­ (Cb ), pore compressibility (CP ), rock porosity (ϕ) and experimental
bonate reservoirs. Therefore, this study utilizes both Eaton’s method and constant (γ) [15]. The key factor in the accuracy of the compressibility
the compressibility method [15], based on research conducted on car­ method is how to determine the pore compressibility coefficient (CP ).
bonate reservoirs in Iran, to estimate the pore pressure. Laboratory results are usually used to determine this parameter, and its
value based on background studies has a strong dependence on rock
porosity values, and it can be estimated with good accuracy from rock
9.1. Eaton method
porosity with equation (12).
( )γ
In 1968, Eaton developed a relationship between velocity and pore (1 − ϕ)Cb σeff
PP = (11)
pressure, which he revised in 1976 to the form presented in Equation (1 − ϕ)Cb − ϕCP
(10).
( )
( )k 1
( ) Vp CP lim estone = × 10− 6
(12)
Pp = Pob − Pob − Phyd × (10) 1.022− 2 + 1.681− 2 ϕ1.05
NCT
( )
1
where Pp is the estimated pore pressure, Pob is the overburden pressure, Cb = × 10− 6
(13)
0.367 + 0.099 ln(ϕ)
Phyd is the hydrostatic pressure at the desired depth and k is Eaton’s
constant, which is considered equal to 3 for sandstone reservoirs. Also, Fig. 7 Estimated Pore Pressure Using the Compressibility Method for
NCT is the normal velocity in Chile zone with normal compression and Well No. 5.
Vp is the pressure wave velocity at the desired depth. As shown in the figure, at the points where pore pressure has been
Fig. 7 Estimated Pore Pressure Using Eaton’s Method for Well No. 5. measured, the curve of estimated values using the compressibility
The figure shows that the estimated pore pressure in most depth ranges method closely aligns with the measured values. In contrast, the curve
is higher than the values measured by the Repeated Formation Test obtained using Eaton’s method exhibits significant fluctuations
(RFT) method. throughout the studied area of the well. This variation appears incon­
sistent with the generally linear trend observed in the measured values.

9.2. Compressibility method

Equation (11) to calculate pore pressure based on the compressibility

6
A. Ounegh et al. Results in Engineering 24 (2024) 103016

10. Estimation of stresses

Accurate estimation of stress distribution on-site, combined with

knowledge of rock mechanical properties, is crucial for evaluating fault
slip, induced earthquakes, and optimizing drilling conditions to enhance
wellbore stability, rock cover integrity, grid orientation, and hydraulic
fracturing [16–21]. During drilling, the stress balance around the well­
bore becomes disrupted, leading to potential instability in the well wall.
Therefore, it is essential to estimate both the in-situ stresses in the
excavated area before drilling and the induced stresses during the dril­
ling process. The following sections will detail the estimation of these
stresses.

10.1. In situ stresses

In situ stresses have three main orthogonal components, including

vertical stress (σ v ), minimum horizontal stress (σ h ) and maximum hor­
izontal stress (σH ). Among these three components, the vertical stress
can be calculated by having the density (ρ) of overburden rocks for each
depth point using equation (14). The density for the reservoir area can
be determined from the density graph; But for the upper area of the
reservoir, this pattern is not taken. Therefore, several methods for esti­
mating density or overburden stress have been presented by researchers;
But in this study, equation (15) was used to estimate the vertical stress
with the help of porosity parameters (ϕ), matrix density (ρma ), depth (z)
and pore fluid density (ρf ).
By estimating the value of vertical stress and also determining the
pore pressure (Pp ), Poisson’s ratio (ν), static Young’s modulus (Esta ) and
Biot’s coefficient (α, the value of this parameter was determined based
on lithology), the minimum (σh ) and maximum horizontal stress values
(σH ) was estimated by relations (16) and (17) respectively. It should be
noted that in these relationships, the strain values (εx and εy ) are
adjusted based on the safe window of the mud weight based on obser­
vations in the well. If there are tests in the well to measure the minimum
stress, the minimum stress should be calibrated with these data.
Figure (8) shows the estimated principal stresses for the studied
depth range from well No. 5. As it is evident in this figure, the stress
regime in the investigated area is strike-slip, which is consistent with the
stress map and geological reports of this field. The results of this well are
also consistent with field reports.
∫z
σv = P0 + ρ(z) dz (14)
0

[ ]
σ v = 0.433 (1 − ϕ)ρma + ρf ϕ × z (15)

ν ( ) Esta ( )
σh = σV − αPp + αPp + εx + νεy (16)
1− ν 1 − ν2

ν ( ) Esta ( )
σH = σ V − αPp + αPp + εy + νεx (17)
1− ν 1 − ν2

10.2. Induced stresses

At the time of digging a well, the rocks around the well carry the
amount of load that has already been placed on the removed rocks. This
causes stress concentration around the well. This concentration of stress
eventually causes the instability of the well, and as a result, the employer
or contractor incurs a lot of operational costs. The stress concentration
around the well is characterized by three components of tangential (σ θ ),
radial (σ r ) and axial (σ z ) stresses. For isotropic and isotropic elastic
rocks, the induced stresses around the well are calculated by the re­
lations provided by Kersh (Equations (18)–(20)).
( ) ( )
Fig. 7. Comparison of measured pore pressure and estimates from Eaton’s and 1 a2 1 a2 a4 a2
σ r = (σ H + σh ) 1 − 2 + (σH − σ h ) 1 − 4 2 + 3 4 cos 2θ + 2 Pw
compressibility methods for well no. 5. 2 r 2 r r r
(18)

7
A. Ounegh et al. Results in Engineering 24 (2024) 103016

Fig. 8. Mud safe window for well 5 based on Mohr-Columb and Hooke-Brown criteria.

( ) ( ) 11.1. Mohr-Columbus criterion

1 a2 1 a4 a2
σ θ = (σ h + σ H ) 1 + − (σ H − σ h ) 1 + 3 cos 2θ − 2 Pw (19)
2 r2 2 r4 r
The Mohr-Coulomb shear failure criterion is widely used in engi­
neering operations. According to this criterion, failure occurs in a plane
a2
σ z = σ v − 2ν(σH − σ h ) cos 2θ (20) where the normal stress (σ n ) and shear stress (τ) follow the equation:
r2
The above equations are used to calculate induced stresses at a point τ = c + μσn (21)
at a distance r from the center of a well with radius a, which is equal to
Pw with the fluid pressure inside it. The angle θ is clockwise to determine in this equation, μ represents the coefficient of internal friction, and c is
the position of the point in polar coordinates relative to the σ H direction. the cohesion.
Considering that the critical values of these induced stresses are The Mohr-Coulomb criterion in the space of principal stresses is
important in the instability of the wellbore, therefore the minimum and expressed in the form of Equation (22). As shown in this relation, the
maximum values of each of these induced stresses are determined average principal stress does not influence the shear failure of the rock.
around the wellbore. σ 1 = σ c + Nσ 3 (22)

11. Evaluation of well stability and determination of mud- In this regard; σ c is the uniaxial strength of the rock, σ 1 is the maximum
weight-safe window principal stress and σ 3 is the minimum principal stress. The values of
maximum and minimum main stresses are determined based on the
To determine the safe mud weight window (SMWW), rock failure stress regime of the area or depth range under investigation. Since the
criteria must be used. In other words, this section examines whether internal friction angle (ϕ) is known, the parameter N in this relationship
collapse or tensile failure in the wellbore will occur at the assumed can also be calculated with equation (23).
downhole fluid pressure based on the critical induced stresses calculated ( )
π ϕ
for the study area. The results of this section are then compared with in- N = tan2 + (23)
4 2
well observations – such as image logs, caliper logs, and mud loss data.
For this purpose, this study employs two criteria: Mohr-Coulomb and To determine the minimum mud pressure that leads to collapse in the
Hoek-Brown. Each of these criteria will be explained for a vertical well. direction of the minimum horizontal principal stress, the minimum and
maximum values of the induced stresses in this direction are calculated.
Based on their placement in Equation (22), the results are summarized in

8
A. Ounegh et al. Results in Engineering 24 (2024) 103016

Table 5 Table 7
Relationships for determining the minimum mud weight to maintain the sta­ Proposed values of m parameter for rock types [25,22]
bility of the good wall based on the Mohr-Columbus criterion [22]. Rock Type Amount of m
Minimum amount of mud pressure required Different Modes of Induced stresses parameter

E − σc σr ≤ σθ ≤ σz Carbonated stones with well-developed cliff surfaces (such as 5 to 8

PW(BO) =
N dolomite, limestone and marble)
D − σc σr ≤ σz ≤ σθ Clay stones (such as mudstone, shale, slate and Lith) 4 to 10
PW(BO) =
1+N Very fine crystal sandstones and poor crystalline surfaces (such 15 to 24
PW(BO) = D − σc − NE σz ≤ σr ≤ σθ as sandstone and quartzite)
( )
D = 3σH − σh E = σV + 2ν(σH − σh ), N = tan2
π
+
ϕ Fine crystalline igneous rocks with several types of minerals 16 to 19
4 2 (such as andesite, dolerite, diabase and rhyolite)
Metamorphic and igneous rocks of coarse grains with several 22 to 33
types of minerals (such as amphibolite, gabbro, genius,
Table 5. Consequently, according to the state of the induced stresses, one granite, neurite and quartz diorite)
of the relationships listed in the table will be used to calculate the
minimum mud weight (PW(BO) ) at each depth point based on the Mohr-
Coulomb criterion. Table 8
To determine the maximum mud weight, the process is similar to the Equations for determination of minimum mud weight based on Hook-Brown
previous one, but with the focus on the maximum mud weight that criterion [28].
causes tensile failure in the direction perpendicular to the minimum Minimum amount of mud pressure required Different Modes of Induced
horizontal stress. Thus, the minimum and maximum values of the stresses
induced stresses in this direction are calculated and substituted into √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
σ r ≤ σ θ ≤ σz
Equation (22). The relationships derived from this process are summa­ (2D + p) − (2E + p)2 − 4E2 + q
PW(BO) =
2
rized in Table 6. Based on the state of the induced stresses at each depth √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
σ r ≤ σ z ≤ σθ
point, one of the relationships provided in this table is used to calculate (4D + p) − (4D + p)2 + 16q − 16D2
PW(BO) =
8
the maximum mud weight. PW(BO) = σ z ≤ σr ≤ σθ
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
2(D − E) − 4(D − E)2 − 4(D − E − pE − q)
11.2. Hooke-Brown criterion 2
D = 3σH − σh , E = σV + 2ν(σH − σh ) , p = mσc , q = σ2c
The Hoek-Brown failure criterion, introduced in the 1980s, is used to
estimate the strength of intact rock and rock mass [23]. Subsequent
modifications to this criterion have been made [24–27]. This empirical Table 9
criterion employs the uniaxial compressive strength of intact rock as a Equations of maximum mud weight based on the Hooke-Brown criterion [28].
scale parameter and introduces two dimensionless parameters, m and s,
Minimum amount of mud pressure required Different Modes of Induced
which are determined based on the type of rock. The proposed values for stresses
the parameter m are presented in Table 7. Using these parameters, Hoek √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( ) σr ≤ σθ ≤ σz
and Brown developed the relationship given in Equation (24), based on 2B + 2B2 − 4 B2 − pB + q
PW(Frac) =
minimum and maximum stresses, after extensive studies on laboratory 2
PW(Frac) = σr ≤ σz ≤ σ θ
data. √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( )
(4A − p) + (4A − p)2 − 16 A2 − pA − q
( ( ) )1 8
σ3 2
σ1 = σ3 + σc m +s (24) PW(Frac) = σz ≤ σ r ≤ σ θ
σc √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( )
(4A − p) + (4A − p)2 − 16 A2 − pA − q
Based on the minimum and maximum values of the induced stresses 8
relative to the minimum and maximum horizontal stresses, the lower A = 3σh − σH ، B = σV − 2ν(σH − σh )، p = mσc ، q = σ2c
and upper limits of the safe mud weight are determined. Accordingly,
the equations presented in Table 8 are used to calculate the minimum
closely with observations from the caliper graph. It is important to note
mud weight, while the equations in Table 9 are used to determine the
that the Mohr-Coulomb criterion is more conservative compared to the
maximum mud weight.
Hoek-Brown criterion. Therefore, it is advisable to use the Mohr-
Coulomb criterion for particularly sensitive depth ranges.
12. Failure criteria results
13. Conclusions
Fig. 8 shows the safe mud weight window obtained for well number 5
based on both the Mohr-Coulomb and Hoek-Brown criteria. As illus­ In this research, a one-dimensional model of the studied well was
trated in the figure, the Mohr-Coulomb criterion demonstrates better developed. Initially, rock dynamic coefficients for the depth range were
performance than the Hoek-Brown criterion, with results that align more calculated using petrophysical logs. Next, static coefficients were esti­
mated based on laboratory-developed relationships. Rock resistance
Table 6 parameters were then determined using results from rock mechanics
Equations for determining the maximum mud weight to maintain the stability of tests. Pore pressure was estimated using both the compressibility and
the good wall based on the Mohr-Columbus criterion [22].
Eaton methods. Comparison with RFT test data indicated that the
Minimum amount of mud pressure required Different Modes of Induced stresses compressibility method provided more accurate results than the Eaton
PW(Frac) = σc + NB σr ≤ σθ ≤ σz method. The key factor in the accuracy of the mentioned method is how
σc + NA σr ≤ σz ≤ σθ the coefficient related to pore compressibility (Cp ) is determined. This
PW(Frac) =
1+N parameter is typically obtained from laboratory results and is strongly
σc − B σz ≤ σr ≤ σθ
PW(Frac) =
N
+A dependent on the rock’s porosity. It can be estimated with reasonable
accuracy from rock porosity using specific equations. Additionally, by
( )
π ϕ
A = 3σh − σH B = σV − 2ν(σH − σh ), N = tan2 +
4 2 estimating in situ and induced stresses, the safe mud weight window was

9
A. Ounegh et al. Results in Engineering 24 (2024) 103016

calculated using both the Mohr-Coulomb and Hoek-Brown criteria. The [7] R. Ramjohn, T. Gan, M. Sarfare, 3D geomechanical modeling for wellbore stability
analysis: starfish, ECMA, Trinidad and Tobago, in: SPE Trinidad and Tobago
comparison of these methods revealed that the Mohr-Coulomb criterion
Section Energy Resources Conference?, SPE, 2018, June D031S027R002.
is more accurate and conservative compared to the Hoek-Brown [8] H. Bagheri, A. Ayatizadeh Tanha, F. Doulati Ardejani, M. Heydari-Tajareh, E. Larki,
method. This is because the results show a high correlation with ob­ Geomechanical model and wellbore stability analysis utilizing acoustic impedance
servations from the caliper log. Thus, employing this criterion for highly and reflection coefficient in a carbonate reservoir, J. Pet. Explor. Prod. Technol. 11
(2021) 3935–3961.
sensitive depth ranges (complex lithologies and well collapse zones) is [9] A. Almasi, A. Mohsenipour, Determining the mud window, geomechanical model
recommended. (MEM), and well wall stability analysis, using analytical and numerical methods in
one of the wells in Iran’s southwest fields, Advanced Applied Geology 12 (1)
(2022) 1–11.
CRediT authorship contribution statement [10] M. Anemangely, A. Ramezanzadeh, B. Tokhmechi, Safe Mud weight window
determination using log based methodology, in: 79th EAGE Conference and
Exhibition 2017, European Association of Geoscientists & Engineers, 2017,
Abdolmanan Ounegh: Writing – review & editing, Writing – orig­
pp. 1–5.
inal draft, Visualization, Software, Methodology, Investigation, Funding [11] N. Mohamadian, H. Ghorbani, D.A. Wood, M. Mehrad, S. Davoodi, S. Rashidi,
acquisition, Formal analysis, Data curation, Conceptualization. Atefeh A. Soleimanian, A.K. Shahvand, A geomechanical approach to casing collapse
Hasan-Zadeh: Writing – review & editing, Writing – original draft, prediction in oil and gas wells aided by machine learning, J. Pet. Sci. Eng. 196
(2021) 107811.
Validation, Supervision, Project administration, Investigation, Concep­ [12] A.R. Huffman, Geophysical pore pressure prediction in complex geologic
tualization. Mohammad Mohammadi Khanaposhtani: Writing – re­ environments, offshore Libya, in: Offshore Technology Conference. Offshore
view & editing, Writing – original draft, Validation, Resources, Project Technology Conference, 2013.
[13] B.A. Eaton, Graphical method predicts geopressures worldwide, World Oil 183
administration, Methodology, Investigation, Funding acquisition, Data (1976).
curation, Conceptualization. Yousef Kazaemzadeh: Writing – review & [14] B.A. Eaton, Fracture gradient prediction and its application in oilfield operations,
editing, Writing – original draft, Visualization, Validation, Supervision, J. Pet. Technol. 21 (1969) 1–353.
[15] V. Atashbari, M.R. Tingay, Pore pressure prediction in carbonate reservoirs, in: SPE
Software, Resources, Project administration, Methodology, Investiga­ Latin America and Caribbean Petroleum Engineering Conference, Society of
tion, Funding acquisition, Formal analysis, Data curation, Petroleum Engineers, 2012.
Conceptualization. [16] W.K. Abdelghany, A.E. Radwan, M.A. Elkhawaga, D.A. Wood, S. Sen, A.A. Kassem,
Geomechanical modeling using the depth-of-damage approach to achieve
successful underbalanced drilling in the Gulf of Suez Rift Basin, J. Pet. Sci. Eng. 202
(2021) 108311.
Declaration of competing interest
[17] M.A. Addis, Reservoir depletion and its effect on wellbore stability evaluation, Int.
J. Rock Mech. Min. Sci. 34 (1997), 4-e1.
The authors have no conflicts of interest to declare. [18] R. Baouche, S. Sen, M. Sadaoui, K. Boutaleb, S.S. Ganguli, Characterization of pore
pressure, fracture pressure, shear failure and its implications for drilling, wellbore
stability and completion design–A case study from the Takouazet field, Illizi Basin,
Data availability Algeria, Mar. Pet. Geol. 120 (2020) 104510.
[19] A. Morris, D.A. Ferrill, D.B. Henderson, Slip-tendency analysis and fault
No data was used for the research described in the article. reactivation, Geology 24 (1996) 275–278.
[20] A.E. Radwan, W.K. Abdelghany, M.A. Elkhawaga, Present-day in-situ stresses in
Southern Gulf of Suez, Egypt: insights for stress rotation in an extensional rift
References basin, J. Struct. Geol. 147 (2021) 104334.
[21] J. Zhao, D. Tang, W. Lin, Y. Qin, H. Xu, In-situ stress distribution and its influence
on the coal reservoir permeability in the Hancheng area, eastern margin of the
[1] K. Fischer, A. Henk, A workflow for building and calibrating 3-D geomechanical
Ordos Basin, China, J. Nat. Gas Sci. Eng. 61 (2019) 119–132.
models &ndash a case study for a gas reservoir in the North German Basin, Solid
[22] A. Al-Ajmi, Wellbore Stability Analysis Based on a New True-Triaxial Failure
Earth 4 (2013) 347–355.
Criterion, 2006.
[2] J. Herwanger, Seismic geomechanics: how to build and calibrate geomechanical
[23] E.T. Brown, E. Hoek, Underground Excavations in Rock, CRC Press, 1980.
models using 3D and 4D seismic data, in: Fourth EAGE CO2 Geological Storage
[24] E. Hoek, C. Carranza-Torres, B. Corkum, Hoek-Brown failure criterion-2002
Workshop, European Association of Geoscientists & Engineers, 2014 cp-439.
edition, Proc. NARMS-Tac 1 (2002) 267–273.
[3] M.D. Zoback, D. Moos, L. Mastin, R.N. Anderson, Well bore breakouts and in situ
[25] E. Hoek, P.K. Kaiser, W.F. Bawden, Support of Underground Excavations in Hard
stress, J. Geophys. Res. Solid Earth 90 (1985) 5523–5530.
Rock, CRC Press, 1995.
[4] M.Y. Al-Qahtani, R. Zillur, A mathematical algorithm for modeling geomechanical
[26] E. Hoek, D. Wood, S. Shah, A modified Hoek–Brown failure criterion for jointed
rock properties of the Khuff and Pre-Khuff reservoirs in Ghawar field, in: SPE
rock masses, in: Rock Characterization: ISRM Symposium, Eurock’92, Chester, UK,
Middle East Oil and Gas Show and Conference, 2001, March, pp. SPE–68194. SPE.
14–17 September 1992, Thomas Telford Publishing, 1992, pp. 209–214.
[5] M.A. Mohiuddin, M.R. Awal, A. Abdulraheem, K. Khan, A new diagnostic approach
[27] E. Hoek, E.T. Brown, The Hoek-Brown failure criterion–a 1988 update, in:
to identify the causes of borehole instability problems in an offshore Arabian field,
Proceedings of the 15th Canadian Rock Mechanics Symposium. Toronto, Canada:
in: SPE Middle East Oil and Gas Show and Conference, SPE, 2001, March,
Proc. 15th Can. Rock Mech. Symp, University of Toronto, 1988, pp. 31–38.
pp. SPE–68095.
[28] R. Gholami, A. Moradzadeh, V. Rasouli, J. Hanachi, Journal of Rock Mechanics and
[6] I. Abalioglu, H. Legarre, C. Garland, B. Sallier, J. Gao, M. van Galen, Q. Chou,
Geotechnical Engineering Practical application of failure criteria in determining
B. Neil, H. Soroush, H. Qutob, Z. Mahli, The role of geomechanics in diagnosing
safe mud weight windows in drilling operations, Rock Mech. Geotech. Eng. 6
drilling hazards and providing solutions to the northern Iraq fields, in: SPE Middle
(2014) 13–25.
East Oil and Gas Show and Conference, SPE, 2011, September, pp. SPE–142022.

10