See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/374045512

Reservoir Geomechanics: A Data-driven Approach

Article in Journal of Energy Research and Reviews · September 2023

DOI: 10.9734/jenrr/2023/v15i2306

CITATIONS READS

0 170

3 authors, including:

Izuchukwu Korie
Nnamdi Azikiwe University, Awka
4 PUBLICATIONS 0 CITATIONS

SEE PROFILE

All content following this page was uploaded by Izuchukwu Korie on 16 January 2024.

The user has requested enhancement of the downloaded file.

 Journal of Energy Research and Reviews

Volume 15, Issue 2, Page 47-56, 2023; Article no.JENRR.102937

ISSN: 2581-8368

Reservoir Geomechanics: A Data-

driven Approach
Izuchukwu Josephmartin Korie a*,
Chudi-Ajabor Ogochukwu a
and Onwuagba Kenechi Innocent a
a Department of Geological Science, Nnamdi Azikiwe University, Awka, Nigeria.

Authors’ contributions

This work was carried out in collaboration among all authors. All authors read and approved the final
manuscript.

Article Information
DOI: 10.9734/JENRR/2023/v15i2306

Open Peer Review History:

This journal follows the Advanced Open Peer Review policy. Identity of the Reviewers, Editor(s) and additional Reviewers,
peer review comments, different versions of the manuscript, comments of the editors, etc are available here:
https://www.sdiarticle5.com/review-history/102937

Received: 08/05/2023
Accepted: 10/07/2023
Review Article
Published: 19/09/2023

ABSTRACT

Reservoir geomechanics is a crucial aspect of optimising and developing oil and gas activities,
especially in maximising production. Recent technological advancements have revolutionised
reservoir geomechanics studies, including integrating data-driven approaches. This review
examines and integrates machine learning, data science, and data twin in reservoir studies. The
primary aim is to identify the benefits, limitations, significant advancements, potential challenges,
opportunities, and research gaps of data-driven approaches to reservoir geomechanics.
Additionally, this study aims to create opportunities for further research to address these
challenges. The review identifies cost-effectiveness, improved reservoir characterisation, and
reduced operational risks as the benefits of integrating data-driven approaches in reservoir
geomechanics. However, the review also highlights the significant challenges of data-driven
approaches, such as insufficient datasets, limited interpretability, and limited transferability of
models. By shedding light on these issues, this review provides a foundation for future research
toward finding solutions to these challenges.
_____________________________________________________________________________________________________

*Corresponding author: E-mail: martinsjoseph675@gmail.com;

J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023

 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

Keywords: Reservoir; geomechanics; machine learning; data science; digital twin.

1. INTRODUCTION 2. DATA-DRIVEN APPROACHES

(DIGITAL TWIN, MACHINE LEARNING,
A better estimation of the reservoir rock elastic AND DATA ANALYTICS)
and failure properties is instrumental to
minimizing wellbore instability problems [1,2], The hydrocarbon industry is increasingly aiming
avoiding differential sticking, improving hole to leverage the power of data to improve
cleaning, improving casing placement, improving performance, increase efficiency, and lower
hydraulic fracturing operations, minimizing costs. It makes the idea of a new generation of
subsidence, and many more. Carrying out digital instruments more advantageous,
mechanical rock tests such as triaxial particularly in the mature and brown fields or
compression, uniaxial compression, scratch, and developed manufacturing fields. Digitalization
impulse hammer is an accurate way to determine has aided the hydrocarbon industry in integrating
these properties. These tests are usually carried and automating all its manoeuvres like
out on the downhole samples retrieved from exploration, drilling, production, and more.
some depth of interest.
Digital Twin is an advanced 3D model that can
However, without core samples and well-log data be utilised to aid the design phase of an asset.
[3,4], analytical and empirical models determine Digital Twin is a combination of cloud computing,
rock's mechanical properties. In the last two machine learning techniques, and high
decades, predicting the mechanical rock computing power which has made the concept of
properties using AI tools were thoroughly integrating all data a viable reality. Digital Twin is
investigated. computer software that takes real-world data
about a physical entity or system as inputs and
Production of hydrocarbons leads to changes in gives outputs. Data is required for analytics,
reservoir pore pressure, resulting in changes in prediction, and automation in a Digital Twin. The
the stress acting on the reservoir and data must be of excellent quality, confirmed, and
surrounding rocks [4,5]. The increasing stress referenced to make a usable twin. To operate in
acting on the rock framework may also lead to real-time, the Digital Twin needs existing data
the compaction of the reservoir. It may result in and models to be up to date with the current
problems such as surface subsidence, casing asset status and changes.
deformation, reactivation of faults, or bedding-
parallel slip. Machine learning (ML) is a subset of AI that
provides statistical tools to explore and analyse
Reservoir geomechanics plays a vital role in big data. Machine learning comprises subsets
optimizing and developing any oil and gas such as supervised, unsupervised, and
activities, especially those related to maximizing reinforced learning (Fig. 1). Supervised learning
oil/gas production [6,7]. It is well understood that is the data learning technique applied when
the applications of reservoir geomechanics are some past or labelled data is available for future
entirely relevant to reservoir subduction, forecasting by function approximation. The
reservoir compaction, water flooding unsupervised learning technique is the machine
management, permeability/porosity reduction, learning technique when the pastlabelledd data
hydraulic fracture, fault reactivation, sand control, is unavailable and is usually used for clustering
and wellbore instability [8,9]. purposes. Reinforced learning combines
supervised and unsupervised learning
The alteration in reservoir rock mechanical techniques in which some parts of the data are
properties and the state of stresses during the labelled, and others are not.
production should be considered since they
affect reservoir performance to deliver 3. RECENT ADVANCES AND EVOLUTION
hydrocarbon. Therefore, stresses and strain OF DATA-DRIVEN APPROACHES IN
balance are required to mitigate the reservoir RESERVOIR GEOMECHANICS
properties alteration by sustaining an equilibrium
between the extracted hydrocarbons and the Due to the high degree of variability in natural
amount of the fluid being injected concerning the materials and the physical environment in which
state of the stresses and the mechanical they are located, rock mechanical behaviour
properties of the reservoir rocks. always exhibits a high degree of nonlinearity.

48
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

Therefore, in practice, experience, knowledge, But from previous applications, we can know that
and numerical modelling methods have been it is often difficult to integrate all the data
important tools for understanding rock mass collected into empirical and numerical models
behaviour and predicting its response to its effectively. Furthermore, frequent extrapolation
environment and changes in situ stress and interpolation techniques are likely to distort
conditions [10,11], hence, crucial in reservoir the data. Since the 1990s, scholars have noticed
stress studies (Fig. 2).

Fig. 1. A comprehensive classification of machine learning showing statistical frameworks [14]

Fig. 2. Recent progress and future advancement of artificial intelligence in reservoir

geomechanical studies [12]

49
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

the advantages of new computer science understand the reservoir's geology and
research methods applied to the simulation of architecture and the potential for deformation and
rock mechanical behaviour. failure during production. A study by Zoback and
Gorelick [15] demonstrated the importance of
By integrating machine learning into rock geomechanical modeling in improving reservoir
mechanics, we could obtain more meaningful characterization and reducing uncertainty in
conclusions from data that are collected relatively geologic models. Also, Khatibi & Aghajanpour
inexpensively [13]. As a result, we use [16], employed to predict Shear velocity,
computers to find patterns in data, not compressional velocity, and vertical stress
"manipulate" data. relationship of offshore gas fields of which the
predicted drilling events matched quite well with
4. THE BENEFITS OF RESERVOIR the observed drilling reports fromt the Machine
GEOMECHANICS Learning algorithm (Figs. 3 & 4).

Reservoir geomechanics is a discipline that Enhanced Reservoir Performance:

combines geology, geophysics, and engineering Geomechanics can also help optimize reservoir
to study the mechanical behavior of subsurface performance by providing insight into the effects
rock formations that contain hydrocarbons. It is of fluid injection, production, and other operations
an interdisciplinary field that studies the on the subsurface. For example, geomechanical
mechanical behavior of rocks in the subsurface, models can predict the formation of fractures,
including their deformation, failure, and fluid flow shear velocity and faults, and other deformation
properties. Integrating geomechanics into features that can enhance or impede fluid flow.
reservoir engineering and management has Several studies have shown that incorporating
proven beneficial in several ways, including geomechanics into reservoir simulation can lead
improved reservoir characterization, enhanced to more accurate predictions of reservoir
reservoir performance, and reduced operational behavior and improved production rates
risks. [17,18,16].

Improved Reservoir Characterization: Reduced Operational Risks: Drilling,

Geomechanical models can provide valuable completion, and production risks can also be
information about the mechanical properties of assessed using geomechanical models. For
reservoir rocks, including their strength, stiffness, example, they can help identify potential drilling
and stress state. This information can help better hazards, such as lost circulation, wellbore

Fig. 3. Machine learning employed to predict shear and compressional velocoities of gas
reservoir [16]

50
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

Fig. 4. Predicted vs. True shear velocity for Well-2 using (left) Simple linear and (right) multiple
linear regression methods from Well-3 [16]

instability, and casing collapse. Geomechanical reservoir engineering data. This enables
analyses can also help optimize hydraulic more accurate predictions of reservoir
fracturing operations by predicting the extent and behavior and helps mitigate the
orientation of fractures and minimizing the risk of uncertainties associated with reservoir
induced seismicity. For example, a study by characterization.
Maxwell et al. [19] showed that geomechanical 2. Enhanced Reservoir Management: Data-
modeling could optimize hydraulic fracturing and driven approaches can optimize reservoir
reduce the risk of micro-seismic events. management strategies by predicting
future reservoir behavior based on current
Reservoir geomechanics has several benefits for and historical data. For example, predictive
reservoir engineering and management, models can identify the most productive
including improved reservoir characterization, well locations and the best strategies for
enhanced reservoir performance, and reduced managing water and gas injection [20].
operational risks. These benefits are supported 3. Improved Well Design and Completion:
by a growing body of research and case studies Data-driven approaches can help
demonstrating the importance of geomechanics engineers design and complete wells more
in modern reservoir management. effectively by predicting the behavior of
reservoir rocks under different conditions.
5. BENEFITS OF DATA-DRIVEN For example, machine learning algorithms
APPROACH IN RESERVOIR can predict the mechanical properties of
GEOMECHANICS rock formations and identify the most
effective completion methods [21].
Research on reservoir geomechanics has been 4. Increased Efficiency and Cost Savings:
revolutionized by machine learning, data Data-driven approaches can help reduce
analytics, and digital twin technologies in recent the time and cost associated with reservoir
years. As a result, data-driven approaches have management by providing more accurate
become increasingly crucial in reservoir and timely information. For example,
geomechanics due to their ability to provide more predictive models can optimize production
accurate and reliable predictions of subsurface schedules and reduce the frequency of
behaviour. Some of the benefits of data-driven well interventions [22].
approaches in reservoir geomechanics include
the following: 6. LIMITATIONS OF USING DATA-
DRIVEN APPROACHES IN RESERVOIR
1. Improved Prediction Accuracy: Data- GEOMECHANICS MODELING
driven approaches can incorporate vast
amounts of data from various sources, Several technologies, such as machine learning,
including geological, geophysical, and data analytics, and digital twins, are being

51
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

adopted in reservoir geomechanics to analyze use of modeling and simulation techniques for
and predict subsurface reservoir behavior. subsurface porous media can aid in decision-
However, their use is subject to several making for managing oil and gas reservoirs.
limitations.
While utilizing these approaches, several
1. Lack of High-Quality Data: Machine challenges have been encountered. Some of
learning and analytics require high-quality these challenges have been overcome through
and reliable data to generate accurate the advancement of research and technological
predictions. However, the data available in capabilities. Nonetheless, certain challenges are
reservoir geomechanics often need to be being managed, while others, such as the effects
completed, noisy and uncertain. Therefore, of bias, remain unaddressed [27].
in order to avoid inaccurate predictions, it
is essential to preprocess the data before This review aims to consolidate the challenges of
analyzing it [23]. data-driven approaches in reservoir mechanical
2. Computational Complexity: Digital twin studies, with the objective of providing a
models can be computationally expensive, foundation for future research to explore ways of
and the simulation time can be extended. addressing and managing these challenges. As
This limits the applicability of digital twin such, this study comprehensively summarizes
models to real-time monitoring of reservoir these challenges and provides detailed
geomechanics [24]. discussions of each of them.
3. Limited Interpretation: Machine learning
and data analytics tools can use large Previous literature, including Tariq et al. [27] and
datasets to identify patterns and make Khatibi and Aghajanpour [16], have identified
predictions. However, they do not explain insufficient and poor-quality data as a major
why a particular outcome was reached. challenge to data-driven approaches, particularly
This may limit the ability to interpret results machine learning, data science and digital twin,
and make informed decisions. in reservoir geomechanics. The effectiveness of
4. Dependence on Historical Data: Machine these approaches is heavily reliant on the quality
learning algorithms rely heavily on and quantity of historical data provided for future
historical data for training models. This predictions. Since these approaches are data-
may limit their effectiveness in predicting oriented, the accuracy of the models can be
events that have yet to occur or in significantly impacted by the quality and quantity
analyzing new reservoirs. of the data used. Limited and noisy data pose
5. Black Box Nature: Machine learning challenges to the accuracy of the models, and
models can be challenging to interpret and the data wrangling process may introduce
explain, leading to a lack of transparency overfitting of the dataset, another critical
in decision-making. This can create trust hindrance to data-driven approaches in reservoir
issues among stakeholders [25]. geomechanics.
6. Limited Applicability to New Scenarios:
Also, Tariq et al. [27] identified the recurrent
Machine learning models are trained on
challenge of inadequate data for training
historical data, and their performance can
machine learning models, particularly in the
deteriorate when applied to new scenarios.
context of reservoir characterization studies.
Therefore, it is necessary to retrain the
Furthermore, the study identified the
models regularly to ensure their accuracy
unavailability of large datasets as a concern in
in new situations [26].
training models. In line with the above statement,
7. CHALLENGES OF DATA-DRIVEN Khatibi and Aghajanpour [16] opined that the
APPROACHES IN RESERVOIR data-driven approaches have the highest
accuracy in predicting reservoir geomechanics
GEOMECHANICS STUDIES
parameters. However, the study further identified
Data-driven approaches, including machine the poor and unavailability of data to be a
learning, data analysis, and digital twin, have significant hindrance to the data-driven
become increasingly popular in reservoir approaches to reservoir mechanics.
geomechanics studies. The late popularity of
these approaches is due to their ability to Again, limited interpretability is another critical
analyze and interpret large datasets generated challenge of data-driven approaches in
from subsurface measurements and simulations geomechanics studies. Data-driven models,
[27-30]. Also, Sun and Zhang [31] agree that the especially machine learning and data science

52
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

algorithms, are often black boxes, making it geomechanics studies presents significant
difficult to understand how they arrived at their opportunities for improving the understanding
predictions [26]. This is particularly challenging in and management of hydrocarbon reservoirs.
reservoir geomechanics, where understanding These opportunities are based on the capabilities
the underlying physical processes is crucial. of these approaches to analyze and interpret
Integrating physical and data-driven approaches large, complex datasets and to identify patterns
is the best bet in solving this challenge. and insights that may not be apparent through
traditional methods [27-29].
Furthermore, Virginia [32] highlighted that the
limited transferability of models is another However, Busetti [35] agrees that a key
obstacle to using a data-driven approach in advantage of these approaches is the ability to
reservoir geomechanics. The study emphasized analyze vast amounts of data, including seismic
that AI models have a narrow range of data, well logs, and field measurements, to gain
applications and cannot be effectively insights into reservoir properties and behavior
generalized to predict outcomes from various under stress. In addition, machine learning
datasets. Similarly, Ramamoorthy and algorithms can identify patterns and relationships
Yampolskiy [33] pointed out that models struggle in the data that may not be apparent through
to provide precise predictions when utilized on a traditional analysis, allowing for more accurate
dataset that differs from the one on which they reservoir geomechanical studies.
were trained. The authors explicitly underlined
that the reusability of machine learning and data Moreover, digital twin models can provide real-
science models poses a significant challenge, as time monitoring and visualization of reservoir
models trained on one geological field may not behavior, enabling operators to optimize
be as dependable when applied to other production and reduce costs. Digital twin
geological fields. technology can also facilitate decision-making
processes by simulating various scenarios and
However, Mohaghegh [34] strongly advises predicting the impact of different parameters,
implementing the model only when the input including stress/pressure, compaction, etc., on
parameters of the given dataset fall within the reservoir performance [36].
range of input parameters for which the model is
designed. Mohaghegh's [34] suggestion is Furthermore, integrating machine learning, digital
reasonable; however, researchers should strive twin, and data analysis approaches can enhance
to find solutions to such issues. In fact, one of the reservoir modeling and simulation, improving the
primary objectives of this review is to identify accuracy of predictions and reducing uncertainty
potential challenges and research gaps in data- [16]. In addition, by incorporating physical
driven approaches to reservoir geomechanics to constraints into machine learning models and
create opportunities for further research toward digital twins, these approaches can provide more
finding solutions to these challenges. realistic representations of reservoir behavior
The challenges associated with using machine and enable better estimation of reservoir
learning, digital twin, and data analysis properties [37].
approaches in reservoir geomechanics studies
include data quality and quantity, complexity and Another opportunity is the potential for cost
interpretability, computational resources, savings, as these approaches can improve
integration and interoperability, and regulatory production efficiency and reduce the need for
challenges. Researchers are developing novel expensive well interventions [27]. Additionally,
approaches to overcome these challenges and integrating data from multiple sources can
leverage the benefits of these approaches to provide a more comprehensive understanding of
improve the understanding and management of the reservoir, reducing the risk of costly errors
hydrocarbon reservoirs. and increasing the potential for successful drilling
and production [27].
8. OPPORTUNITIES OF INTEGRATING
DATA-DRIVEN APPROACHES INTO In summary, integrating machine learning, digital
twin, and data analysis approaches in reservoir
RESERVOIR GEOMECHANICS
geomechanics studies presents opportunities for
WORKFLOWS more accurate reservoir characterization,
Integrating machine learning, digital twin, and improved prediction of reservoir performance,
data analysis approaches in reservoir enhanced reservoir modeling and simulation,

53
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

real-time monitoring and visualization, cost unreliable models, which can negatively impact
savings, and reduced risk. operational planning and decision-making.
Another limitation is the interpretability of data-
9. CONCLUSION driven models, which can be difficult to
understand, leading to skepticism and mistrust
Reservoir geomechanics plays a crucial role in
among stakeholders. Finally, the transferability of
optimizing and developing oil and gas activities,
models between different reservoirs can be
particularly in maximizing production rates and
limited, reducing the generalizability and
improving recovery efficiencies. Recent
practicality of these approaches.
technological advancements have led to the
emergence of novel and innovative data-driven Overall, this review suggests that data-driven
approaches in reservoir geomechanics studies, approaches in reservoir geomechanics hold
such as machine learning, data science, and significant potential for optimizing and developing
data twin. These approaches have the potential oil and gas activities. By addressing the identified
to revolutionize the field by enabling a more challenges and limitations, further research can
comprehensive understanding of reservoir expand our understanding of reservoir behavior
behavior and optimizing operational planning and and enhance operational planning, leading to the
decision-making. A multitude of literature has maximization of oil/gas production.
applied, examined, and integrated these
approaches in reservoir studies, highlighting their ACKNOWLEDGEMENT
potential benefits and limitations.
Completing this work could not have been
This review aims to provide a comprehensive possible without the participation of the authors
and critical overview of data-driven approaches enumerated in this review. However, the authors
in reservoir geomechanics by analyzing and would like to express their deep appreciation and
synthesizing past literature. Specifically, the indebtedness to Dr. Vivian O. Oguadimma for
objectives of this study are to identify the her assistance and in-depth guidance throughout
benefits, limitations, significant advancements, the review process. Your contributions are
potential challenges, opportunities, and research sincerely appreciated and gratefully
gaps in data-driven approaches to reservoir acknowledged.
geomechanics. Additionally, this study aims to
create opportunities for further research to COMPETING INTERESTS
overcome the identified challenges and expand
Authors have declared that no competing
the knowledge of data-driven approaches in
interests exist.
reservoir geomechanics.
REFERENCES
The integration of data-driven approaches in
reservoir geomechanics offers several potential 1. Joshua Pwavodi, Ibekwe N Kelechi,
benefits. Firstly, these approaches provide a Perekebina Angalabiri, Sharon Chioma
more cost-effective and efficient alternative to Emeremgini, Vivian O Oguadinma. Pore
traditional methods that rely on extensive and pressure prediction in offshore Niger delta
expensive fieldwork. Secondly, data-driven using data-driven approach: Implications
approaches can improve and enhance reservoir on drilling and reservoir quality, Energy
characterization, leading to a more accurate and Geoscience. 2023;4(3):20-23.
detailed understanding of reservoir properties 2. Oguadinma, et al. Lithofacies and Textural
and behavior. This allows for more precise Attributes of the Nanka Sandstone
decision-making and operational planning. (Eocene): Proxies for evaluating the
Finally, data-driven approaches can reduce Depositional Environment and Reservoir
operational risks by providing real-time Quality. Journal of Earth Sciences and
information and insights into reservoir Geotechnical Engineering. 2014;4(4).
performance, enabling timely intervention and 1-16ISSN: 1792-9040 (print), 1792-9660.
management of potential hazards. DOI: 10.13140/RG.2.2.33124.07042
3. Oguadinma, et al. Study of the Pleistocene
However, data-driven approaches are not without submarine canyons of the south-eastern
limitations. A significant challenge is the Niger delta basin: Tectonostratigraphic
availability of sufficient and reliable datasets, evolution and infilling Conference/Reunion
particularly in emerging oil and gas regions. des sciences de la terre, Lyon, France;
Insufficient data can lead to inaccurate and 2021.

54
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

Oguadinma, et al. The art of integration: A Niger delta, Nigeria. International Journal
basic tool in effective hydrocarbon field and Advance Geosciences. 2018;6:51-56.
appraisal, Med-GU Conference, Istanbul. Available:http://dx.doi.org/10.14419/ijag.v6i
Turkey; 2021. 1.8776
4. Nwaezeapu, et al. Hydrocarbon Reservoir 12. Fahad IS, Abdulla A, Amirmasoud KD,
Evaluation: a case study of Tymot field at Neghabhan S. Application of ML & AI to
southwestern offshore Niger Delta Oil model petrophysical and geomechanical
Province, Nigeria. Nanoscience and properties of shale reservoirs: A systematic
Nanotechnology. 2018;2(2). literature review. Petroleum. 2022;8:158-
DOI: http://dx.doi.org/10.18063/nn.v0i0.618 166.
5. Oguadinma, et al. An integrated approach 13. Zhou YX, Wu XP. Use of neural networks
to hydrocarbon prospect evaluation of the in the analysis and interpretation of site
Vin field, Nova Scotia Basin. S.E.G. investigation data. Comput. Geotechnics.
Technical Program Expanded Abstracts; 1994;16:105–122.
2016. DOI: 10.1016/0266-352x(94)90017-5.
DOI: 10.1190/segam2016- 13843545.1 14. Diksha S, Neeraj K. A review on machine
6. Oguadinma, et al. Advanced Study of learning algorithms, tasks and applications.
Seismic and Well Logs in the Hydrocarbon International Journal of Advanced
Prospectivity of Siram Field, Niger Delta Research in Computer Engineering &
Basin. Geological Society of America Technology. 2017;6(10):1550-1551.
Abstracts with Programs. 2017;49. 15. Zoback MD, Gorelick SM. Earthquake
DOI: 10.1130/abs/2017AM-296312 triggering and large-scale geologic storage
7. Oguadinma O Vivian, Ibekwe N Kelechi, of carbon dioxide. Proceedings of the
Lanisa Ademola, et al. Reservoir and National Academy of Sciences. 2012;
sequence stratigraphic analysis using 109(26):10164-10168.
subsurface data. ESS Open Archive; 16. Khatibi S, Aghajanpour A. Machine
February 09, 2023. Learning: A Useful Tool in Geomechanical
Oguadinma O Vivian, Ibekwe N Kelechi, Studies, a Case Study from an Offshore
Lanisa Ademola, et al. Submarine canyon: Gas Field. Energies. 2020;13(14):
A brief review. ESS Open Archive; 35-38.
February 09, 2023 Available:https://doi.org/10.3390/en131435
8. Denney D. Coupled Reservoir- 28
Geomechanics Model for Wellbore Stability 17. Rahman SS, Haque A, Islam MR.
and Sand Prediction. Society of Petroleum Integrating geomechanics into reservoir
Engineers; 2001. simulation for improved performance
DOI: 10.2118/0501-0062-JPT prediction. Journal of Petroleum Science
9. Ibekwe, et al. Enhanced hydrocarbon and Engineering. 2017;155:330-340.
recovery using the application of seismic 18. Wang J, Li X, Wang H, Xu Y, Sun Y, Liu H.
attributes in fault detection and direct Integrating geomechanics into reservoir
hydrocarbon indicator in Tomboy Field, simulation: A review. Journal of Petroleum
western-Offshore Niger Delta Basin. ESS Science and Engineering. 2020;190:
Open Archive; January 24, 2023. 107102.
Ibekwe KN, Oguadinma VO, Okoro VK, 19. Maxwell SC, Urbancic TI, Zoback MD, Das
Aniwetalu E, Lanisa A, Ahaneku CV. I. Reservoir geomechanics and hydraulic
Reservoir Characterization Review in fracturing: From reservoir characterization
Sedimentary Basins. Journal of Energy to full-field modeling. SPE Journal. 2015;
Research and Reviews. 2023;13(2):20–28. 20(3):549-561.
Available:https://doi.org/10.9734/jenrr/2023 20. Zhang C, Huang W, Liu X. A data-driven
/v13i2259 approach to optimizing water injection
10. Morgenroth J, Khan UT, Perras MA. An strategy in heterogeneous reservoirs.
overview of opportunities for machine Journal of Petroleum Science and
learning methods in underground rock Engineering. 2019;176:609-622.
engineering design. Geosciences. 2019; 21. Song L, Zhang H, Bai B, Shi C. A data-
9(12):504. driven approach for predicting rock
DOI: 10.3390/geosciences9120504 mechanical properties in shale reservoirs.
11. Aniwetalu, et al. Spectral analysis of Journal of Petroleum Science and
Rayleigh waves in Southeastern part of Engineering. 2019;175:408-421.

55
 Korie et al.; J. Energy Res. Rev., vol. 15, no. 2, pp. 47-56, 2023; Article no.JENRR.102937

22. Wang K, Gao H, Wang H, Gao Y. Data- Saudi Arabia annual technical symposium
driven production optimization based on a and exhibition 2017. Society of Petroleum
hybrid neural network model. Journal of Engineers. 2017b;1428–1440.
Petroleum Science and Engineering. Available:https://doi.org/10.2118/187974-
2020;193:107350. MS
23. Acar E, Unler A. Machine Learning for 30. Anifowose FA. Advances in hybrid
Reservoir Characterization: A Review. computational intelligence application in oil
Computers & Geosciences. 2020;135: and gas reservoir characterization. In:
104398. Society of petroleum engineers—SPE
24. Guo Z, Li H, Li L, Li X, Li G. Digital twin- Saudi Arabia section young professional's
based reservoir geomechanics simulation technical symposium 2012, YPTS 2012.
and application in oilfield development. 2012;1–8.
Journal of Petroleum Science and Available:https://doi.org/10.2118/160921-
Engineering. 2021;196:108017. ms
25. Koehrsen W. The Limits and Ethical 31. Sun S, Zhang T. A 6M digital twin for
Considerations of Machine Learning. modeling and simulation in subsurface
Medium; 2018. reservoirs. Advances in Geo-Energy
Available:https://medium.com/@williamkoe Research. 2020;4(4):349-351.
hrsen/the-limits-and-ethical-considerations- DOI: 10.46690/ager.2020.04.01.
of-machine-learning-1c914b1c869e 32. Virginia D. Responsible artificial
26. Wang X, Zhang Y, Ma J. A Review of data- intelligence: designing AI for human
driven approaches for reservoir values. ITU J. 2018;1(1):1–8.
characterization and Modeling. Journal of 33. Ramamoorthy A, Yampolskiy R. Beyond
Petroleum Science and Engineering. 2020; map: The race for artificial general
188:106842. Intelligence. ITU J. 2018;1(1):77–84.
27. Tariq Z, Aljawad MS, Hasan A, Murtaza M, 34. Mohaghegh SD. Shale analytics: Data-
Mohammed E, El-Husseiny A, et al. A driven analytics in unconventional
systematic review of data science and resources. Springer International
machine learning applications to the oil Publishing, Cham; 2017.
and gas industry. Journal of Petroleum Available:https://doi.org/10.1007/978-3-
Exploration and Production Technology. 319-48753-3
2021;11:4339–4374. 35. Busetti S. Guest Editorial: Five innovation
Available:https://doi.org/10.1007/s13202- themes for integrated geomechanics
021-01302-2 technology. Journal of Petroleum
28. Tariq Z, Elkatatny S, Mahmoud M, Ali AZ, Technology. 2019;71:14–15.
Abdulraheem A. A new technique to DOI: https://doi.org/10.2118/1019-0014-
develop rock strength correlation using JPT
artificial intelligence tools. In: Society of 36. Sircar A, Nair A, Bist N, Yadav K. Digital
petroleum engineers—SPE reservoir twin in hydrocarbon industry. Petroleum
characterization and simulation conference Research; 2022.
and exhibition, RCSC 2017. Society of Available:https://doi.org/10.1016/j.ptlrs.202
Petroleum Engineers. 2017a;1340– 2.04.001
1353. 37. Chen Y, Zhao L, Pan J, Li C, Xu M, Li K,
Available:https://doi.org/10.2118/186062- Zhang F, Geng J. Deep carbonate
MS reservoir characterization using multi-
29. Tariq Z, Elkatatny S, Mahmoud M, Ali AZ, seismic attributes via machine learning
Abdulraheem A. A new approach to predict with physical constraints, Journal of
failure parameters of carbonate rocks Geophysics and Engineering. 2021;18(5):
using artificial intelligence tools. In: Society 761–775.
of petroleum engineers—SPE Kingdom of DOI: https://doi.org/10.1093/jge/gxab049
_________________________________________________________________________________
© 2023 Korie et al.; This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.
Peer-review history:
The peer review history for this paper can be accessed here:
https://www.sdiarticle5.com/review-history/102937

56

View publication stats