Smart predictions of petrophysical formation pore pressure via robust data-driven intelligent models

Publikationen: Beitrag in FachzeitschriftArtikelForschung(peer-reviewed)

Standard

Smart predictions of petrophysical formation pore pressure via robust data-driven intelligent models. / Krishna, Shwetank; Irfan, Sayed Ameenuddin; Keshavarz, Sahar et al.
in: Multiscale and multidisciplinary modeling, experiments and design, Jahrgang 7.2024, Nr. 6, 24.07.2024, S. 5611-5630.

Publikationen: Beitrag in FachzeitschriftArtikelForschung(peer-reviewed)

Bibtex - Download

@article{6eb5a1b0837445e4ac639e052cb22591,
title = "Smart predictions of petrophysical formation pore pressure via robust data-driven intelligent models",
abstract = "Predicting pore pressure in the formation is crucial for assessing reservoir geomechanical characteristics, designing drilling schemes/mud programs, and strategies to enhance oil recovery. Accurate predictions are vital for safe and cost-effective exploration and development. Recent research has seen the emergence of intelligent models utilizing machine learning (ML) and deep learning (DL) algorithms, offering promising outcomes. However, there remains a need to identify the most accurate and dependable model among these. This study aims to address this gap by comparing the performance of various ML and DL models, as reported in existing literature, to determine the optimal approach for pore pressure prediction. The sorted machine learning (ML) and deep learning (DL) regression algorithms used for the comparative analysis are decision tree (DT), extreme gradient boosting (XGBoost), random forest (RF), recurrent neural network (RNN), and convolutional neural network (CNN). A total dataset of 22,539 is gathered from five wells (15/9-F-1 A, 15/9-F-1 B, 15/9-F-11 A, 15/9-F-11 T2, and 15/9-F-14) drilled at North-sea Volve oil field, Norway. The first four wells are used to train and test the ML and DL algorithm, and the remaining well (15/9-F-14) is used to evaluate the best-performing algorithm{\textquoteright}s universality in predicting pore pressure at the field of study. Seven different petrophysical parameters are used as input parameters to develop the predictive models. Statistical performance metrics are carried out to analyze the applied ML and DL performance. Based on performance indicators, the RF algorithm showed superior results compared to other predictive models with R 2 and RMSE values of 0.97 and 2.70 MPa, respectively. Furthermore, the best-performing predictive model with low prediction error RMSE value is applied to the other well dataset from the field of study to access the universality of the RF algorithm to predict pore pressure in the field of study. The results of the universality analysis show a satisfactory prediction accuracy with R 2 and RMSE values of 0.905 and 6.48 MPa, respectively.",
keywords = "Deep learning, Machine learning, Petrophysical properties, Pore pressure, Prediction",
author = "Shwetank Krishna and Irfan, {Sayed Ameenuddin} and Sahar Keshavarz and Gerhard Thonhauser and {Umer Ilyas}, Suhaib",
note = "Publisher Copyright: {\textcopyright} The Author(s) 2024.",
year = "2024",
month = jul,
day = "24",
doi = "10.1007/s41939-024-00542-z",
language = "English",
volume = "7.2024",
pages = "5611--5630",
journal = "Multiscale and multidisciplinary modeling, experiments and design",
issn = "2520-8179",
publisher = "Springer Science + Business Media",
number = "6",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - Smart predictions of petrophysical formation pore pressure via robust data-driven intelligent models

AU - Krishna, Shwetank

AU - Irfan, Sayed Ameenuddin

AU - Keshavarz, Sahar

AU - Thonhauser, Gerhard

AU - Umer Ilyas, Suhaib

N1 - Publisher Copyright: © The Author(s) 2024.

PY - 2024/7/24

Y1 - 2024/7/24

N2 - Predicting pore pressure in the formation is crucial for assessing reservoir geomechanical characteristics, designing drilling schemes/mud programs, and strategies to enhance oil recovery. Accurate predictions are vital for safe and cost-effective exploration and development. Recent research has seen the emergence of intelligent models utilizing machine learning (ML) and deep learning (DL) algorithms, offering promising outcomes. However, there remains a need to identify the most accurate and dependable model among these. This study aims to address this gap by comparing the performance of various ML and DL models, as reported in existing literature, to determine the optimal approach for pore pressure prediction. The sorted machine learning (ML) and deep learning (DL) regression algorithms used for the comparative analysis are decision tree (DT), extreme gradient boosting (XGBoost), random forest (RF), recurrent neural network (RNN), and convolutional neural network (CNN). A total dataset of 22,539 is gathered from five wells (15/9-F-1 A, 15/9-F-1 B, 15/9-F-11 A, 15/9-F-11 T2, and 15/9-F-14) drilled at North-sea Volve oil field, Norway. The first four wells are used to train and test the ML and DL algorithm, and the remaining well (15/9-F-14) is used to evaluate the best-performing algorithm’s universality in predicting pore pressure at the field of study. Seven different petrophysical parameters are used as input parameters to develop the predictive models. Statistical performance metrics are carried out to analyze the applied ML and DL performance. Based on performance indicators, the RF algorithm showed superior results compared to other predictive models with R 2 and RMSE values of 0.97 and 2.70 MPa, respectively. Furthermore, the best-performing predictive model with low prediction error RMSE value is applied to the other well dataset from the field of study to access the universality of the RF algorithm to predict pore pressure in the field of study. The results of the universality analysis show a satisfactory prediction accuracy with R 2 and RMSE values of 0.905 and 6.48 MPa, respectively.

AB - Predicting pore pressure in the formation is crucial for assessing reservoir geomechanical characteristics, designing drilling schemes/mud programs, and strategies to enhance oil recovery. Accurate predictions are vital for safe and cost-effective exploration and development. Recent research has seen the emergence of intelligent models utilizing machine learning (ML) and deep learning (DL) algorithms, offering promising outcomes. However, there remains a need to identify the most accurate and dependable model among these. This study aims to address this gap by comparing the performance of various ML and DL models, as reported in existing literature, to determine the optimal approach for pore pressure prediction. The sorted machine learning (ML) and deep learning (DL) regression algorithms used for the comparative analysis are decision tree (DT), extreme gradient boosting (XGBoost), random forest (RF), recurrent neural network (RNN), and convolutional neural network (CNN). A total dataset of 22,539 is gathered from five wells (15/9-F-1 A, 15/9-F-1 B, 15/9-F-11 A, 15/9-F-11 T2, and 15/9-F-14) drilled at North-sea Volve oil field, Norway. The first four wells are used to train and test the ML and DL algorithm, and the remaining well (15/9-F-14) is used to evaluate the best-performing algorithm’s universality in predicting pore pressure at the field of study. Seven different petrophysical parameters are used as input parameters to develop the predictive models. Statistical performance metrics are carried out to analyze the applied ML and DL performance. Based on performance indicators, the RF algorithm showed superior results compared to other predictive models with R 2 and RMSE values of 0.97 and 2.70 MPa, respectively. Furthermore, the best-performing predictive model with low prediction error RMSE value is applied to the other well dataset from the field of study to access the universality of the RF algorithm to predict pore pressure in the field of study. The results of the universality analysis show a satisfactory prediction accuracy with R 2 and RMSE values of 0.905 and 6.48 MPa, respectively.

KW - Deep learning

KW - Machine learning

KW - Petrophysical properties

KW - Pore pressure

KW - Prediction

UR - http://www.scopus.com/inward/record.url?scp=85199358804&partnerID=8YFLogxK

U2 - 10.1007/s41939-024-00542-z

DO - 10.1007/s41939-024-00542-z

M3 - Article

VL - 7.2024

SP - 5611

EP - 5630

JO - Multiscale and multidisciplinary modeling, experiments and design

JF - Multiscale and multidisciplinary modeling, experiments and design

SN - 2520-8179

IS - 6

ER -