A Fiber-optic Assisted Multilayer Perceptron Reservoir Production Modeling: A Machine Learning Approach in Prediction of Gas Production from the Marcellus Shale

Ghahfarokhi, Payam Kavousi; Carr, Timothy R.; Bhattacharya, Shuvajit; Elliott, J. Kelly; Shahkarami, Alireza; Martin, Keithan

doi:10.15530/urtec-2018-2902641

Cited by 20 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The aim is to provide new representations of the input data that are useful for achieving accurate prediction. Multi Layer Perceptron (MLP) [ 104 ], Binary Descriptor [ 105 ], and multi-stage CNNs [ 37 ] are considered as the non-handcrafted engineering techniques that can be used for multiphase estimation. Dimensional reduction algorithm, such as Principle Component Analysis (PCA), can be used to reduce the number of features while selecting highly relevant features for multiphase estimation objective [ 57 , 106 ].…”

Section: Machine Learningmentioning

confidence: 99%

“…Ghahfarokhi et al [ 104 ] used an averaged daily data from 1320 DTS measurements along the lateral of the gas-producing well in the Marcellus Shale, in Northern West Virginia to forecast daily gas production. An MLP model was trained and deployed, and Sensitivity Analysis (SA) was conducted to analyse weight behaviour.…”

Section: Machine Learningmentioning

confidence: 99%

“…An MLP model was trained and deployed, and Sensitivity Analysis (SA) was conducted to analyse weight behaviour. Similar to [ 104 ], Bhattacharya et al [ 112 ] used DAS and DTS among other datasets to predict daily gas production using ANN, SVM, and RF. A high accuracy (96%) was achieved by employing 18 features to RF model for prediction.…”

Section: Machine Learningmentioning

confidence: 99%

“…MLP [104] 2018 Gas producing well with DAS and DTS Mainly focus on using DTS for forecasting gas production while DAS data was only recorded during hydraulic fracturing of the well.…”

Section: Modelling Techniquementioning

confidence: 99%

See 3 more Smart Citations

A Survey on Distributed Fibre Optic Sensor Data Modelling Techniques and Machine Learning Algorithms for Multiphase Fluid Flow Estimation

Arief

Wiktorski

Thomas

2021

Sensors

View full text Add to dashboard Cite

Real-time monitoring of multiphase fluid flows with distributed fibre optic sensing has the potential to play a major role in industrial flow measurement applications. One such application is the optimization of hydrocarbon production to maximize short-term income, and prolong the operational lifetime of production wells and the reservoir. While the measurement technology itself is well understood and developed, a key remaining challenge is the establishment of robust data analysis tools that are capable of providing real-time conversion of enormous data quantities into actionable process indicators. This paper provides a comprehensive technical review of the data analysis techniques for distributed fibre optic technologies, with a particular focus on characterizing fluid flow in pipes. The review encompasses classical methods, such as the speed of sound estimation and Joule-Thomson coefficient, as well as their data-driven machine learning counterparts, such as Convolutional Neural Network (CNN), Support Vector Machine (SVM), and Ensemble Kalman Filter (EnKF) algorithms. The study aims to help end-users establish reliable, robust, and accurate solutions that can be deployed in a timely and effective way, and pave the wave for future developments in the field.

show abstract

Section: Machine Learningmentioning

confidence: 99%

Section: Machine Learningmentioning

confidence: 99%

Section: Machine Learningmentioning

confidence: 99%

“…MLP [104] 2018 Gas producing well with DAS and DTS Mainly focus on using DTS for forecasting gas production while DAS data was only recorded during hydraulic fracturing of the well.…”

Section: Modelling Techniquementioning

confidence: 99%

See 2 more Smart Citations

A Survey on Distributed Fibre Optic Sensor Data Modelling Techniques and Machine Learning Algorithms for Multiphase Fluid Flow Estimation

Arief

Wiktorski

Thomas

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…One hypothesis is that all wells produce at a constant bottom hole pressure (BHP). Ghahfarokhi et al (2018) used a multi-layer perceptron neural network to predict gas production using multi-point formation temperature monitoring in distributed temperature sensing and flowing time data. Khan and Louis (2021) used the bottom node and top node pressures of wells as ANN inputs to forecast shale gas production as the targeted output.…”

Section: Introductionmentioning

confidence: 99%

A deep-learning approach for reservoir evaluation for shale gas wells with complex fracture networks

Chu

Lee

2022

Adv. Geo-Energy Res.

View full text Add to dashboard Cite

The complex fracture networks in shale gas reservoirs bring greater challenges and uncertainties to the modeling in reservoir evaluation. As the emerging potential technology, deep learning can be usefully applied to many aspects of reservoir evaluation. To further conduct the reservoir evaluation in rate transient analysis, this work proposes a data-driven proxy model for accurately evaluating the horizontal wells with complex fracture networks in shales. The production time, variable bottom hole pressure, and the fracture networks properties are used as input variables, while the output variable refers to the production for the forecast time period. The data from boundary element method is used to generate the proxy model for the learning process. The method of shuffled cross-validation is used to increase the model's accuracy and generalizability. The proxy model is coupled with recently developed deep learning methods such as attention mechanism, skip connection, and cross-validation to address the time series analysis problem for multivariate operating and physical parameters. Results demonstrate that the attention mechanism is robust. The operating parameters analysis shows that the attention mechanism has the ability to analyze variable pressure drop/flowrate data. Sensitivity analysis also indicates that the model takes into account the geometric characteristics of fracture network. The model reliability is proved by a case study from Marcellus shale. The computation time of the trained attention mechanism model is approximately 0.3 s, which equates to 3.8% of the physical model's running time.

show abstract

A Literature Review

Li¹,

Karrenbach²,

Ajo‐Franklin³

2021

Geophysical Monograph Series

View full text Add to dashboard Cite

A Fiber-optic Assisted Multilayer Perceptron Reservoir Production Modeling: A Machine Learning Approach in Prediction of Gas Production from the Marcellus Shale

Cited by 20 publications

References 0 publications

A Survey on Distributed Fibre Optic Sensor Data Modelling Techniques and Machine Learning Algorithms for Multiphase Fluid Flow Estimation

A Survey on Distributed Fibre Optic Sensor Data Modelling Techniques and Machine Learning Algorithms for Multiphase Fluid Flow Estimation

A deep-learning approach for reservoir evaluation for shale gas wells with complex fracture networks

A Literature Review

Contact Info

Product

Resources

About