Application of predictive data analytics to model daily hydrocarbon production using petrophysical, geomechanical, fiber-optic, completions, and surface data: A case study from the Marcellus Shale, North America

Abstract Predicting gas production from stimulated unconventional reservoirs has always been a challenge to oil and gas companies because the physics of fluid flow through such reservoirs is not well understood. On the other hand, modeling gas production from shale reservoirs is a complex multi-variate problem that requires proper integration of data from multiple disciplines, such as geology, petrophysics, geomechanics, completions, and surface-based measurements, etc. This study demonstrates the application of data-driven machine learning algorithms, integrating geoscientific, distributed acoustic sensing (DAS), distributed temperature sensing (DTS) fiber-optic, completions, flow scanner production log, and surface data to model daily gas production from a 28-stage stimulated horizontal well drilled in the Marcellus Shale of the Appalachian basin in North America. In addition, this study aims to utilize the data from a fiber-optic monitoring system, such as DAS and DTS to evaluate the well performance. We build supervised data-driven machine learning models using Random Forest (RF), Artificial Neural Network (ANN), and Support Vector Machine (SVM) algorithms to understand the well performance and forecast the daily gas production. The study compares the forecasted results, tests the level of accuracy, and addresses different issues with these machine learning algorithms. A spatio-temporal database is constructed and used to perform sensitivity analyses to identify the key drivers for gas production. The results show that RF and ANN algorithms can be used to predict daily gas production with significantly high accuracy; however, RF algorithm is the best predictor in terms of highest accuracy (96%), less computational time and cost. Based on sensitivity analysis, it appears that Poisson's ratio, minimum horizontal stress, DAS, casing pressure, gamma log are the most important parameters to predict gas production.

[1]  Ângelo Cardoso,et al.  Generalising Random Forest Parameter Optimisation to Include Stability and Cost , 2017, ECML/PKDD.

[2]  Timothy R. Carr,et al.  Comparison of supervised and unsupervised approaches for mudstone lithofacies classification: Case studies from the Bakken and Mahantango-Marcellus Shale, USA , 2016 .

[3]  Colin Perkins,et al.  Micro-Seismic Detection using Distributed Acoustic Sensing , 2013 .

[4]  Susan Schrader,et al.  Performing Reservoir Simulation with Neural Network Enhanced Data , 2013 .

[5]  R. Rickman,et al.  A Practical Use of Shale Petrophysics for Stimulation Design Optimization: All Shale Plays Are Not Clones of the Barnett Shale , 2008 .

[6]  T. Ahmed,et al.  Application of Neural Network Parameter Prediction in Reservoir Characterization and Simulation - A Case History: The Rabbit Hills Field , 1997 .

[7]  Erkan Fidan,et al.  First Downhole Application of Distributed Acoustic Sensing for Hydraulic-Fracturing Monitoring and Diagnostics , 2012 .

[8]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[9]  Richard A. Startzman,et al.  Predicting Natural Gas Production Using Artificial Neural Network , 2001 .

[10]  Timothy R. Carr,et al.  Petrophysical Analysis and Sequence Stratigraphy of the Utica Shale and Marcellus Shale, Appalachian Basin, USA , 2013 .

[11]  W S McCulloch,et al.  A logical calculus of the ideas immanent in nervous activity , 1990, The Philosophy of Artificial Intelligence.

[12]  G. F. Hughes,et al.  On the mean accuracy of statistical pattern recognizers , 1968, IEEE Trans. Inf. Theory.

[13]  M. D. Rudnicki,et al.  Organic matter–hosted pore system, Marcellus Formation (Devonian), Pennsylvania , 2013 .

[14]  S. Amini,et al.  Correlating distributed acoustic sensing (DAS) to natural fracture intensity for the Marcellus Shale , 2017 .

[15]  T. Parker,et al.  A VSP field trial using distributed acoustic sensing in a producing well in the North Sea , 2013 .

[16]  Peter Wills,et al.  Field Trials of Distributed Acoustic Sensing For Geophysical Monitoring , 2011 .

[17]  G. Lash,et al.  Marcellus Shale Play ’ s Vast Resource Potential Creating Stir In Appalachia , 2008 .

[19]  Peter Wills,et al.  Dual-Well 3D VSP in Deepwater Made Possible by DAS , 2013 .

[20]  Palash Panja,et al.  Application of artificial intelligence to forecast hydrocarbon production from shales , 2017 .

[21]  Ramón Díaz-Uriarte,et al.  Gene selection and classification of microarray data using random forest , 2006, BMC Bioinformatics.

[22]  Robert Balch,et al.  Using Neural Networks To Estimate Monthly Production: A Case Study for the Devonian Carbonates, Southeast New Mexico , 2005 .

[23]  Giles M. Foody,et al.  Feature Selection for Classification of Hyperspectral Data by SVM , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[24]  Timothy R. Carr,et al.  Organic-rich Marcellus Shale lithofacies modeling and distribution pattern analysis in the Appalachian Basin , 2013 .

[25]  T. Engelder,et al.  Thickness trends and sequence stratigraphy of the Middle Devonian Marcellus Formation, Appalachian Basin: Implications for Acadian foreland basin evolution , 2011 .

[26]  Q. Passey,et al.  A Practical Model for Organic Richness from Porosity and Resistivity Logs , 1990 .

[27]  Vladimir Vapnik,et al.  Support-vector networks , 2004, Machine Learning.

[28]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[29]  W. Zagorski,et al.  The Appalachian Basin Marcellus Gas Play: Its History of Development, Geologic Controls on Production, and Future Potential as a World-class Reservoir , 2012 .

[30]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[31]  Vassilis Gaganis,et al.  Application of Artificial Neural Networks to Downhole Fluid Analysis , 2007 .

[32]  Maojin Tan,et al.  A dynamic adaptive radial basis function approach for total organic carbon content prediction in organic shale , 2013 .

[33]  T. Carr,et al.  Depositional Environment and Impact on Pore Structure and Gas Storage Potential of Middle Devonian Organic Rich Shale, Northeastern West Virginia, Appalachian Basin , 2017 .

[34]  Seismic Attributes Application for the Distributed Acoustic Sensing Data for the Marcellus Shale: New Insights to Cross-Stage Flow Communication , 2018 .

[35]  J. Suykens,et al.  A tutorial on support vector machine-based methods for classification problems in chemometrics. , 2010, Analytica chimica acta.

[36]  B. Biondi,et al.  Earthquakes analysis using data recorded by the Stanford DAS array , 2017 .

[37]  C. E. Weaver,et al.  Thorium-to-Uranium Ratios as Indicators of Sedimentary Processes: Example of Concept of Geochemical Facies , 1958 .

[38]  Srikanta Mishra,et al.  Applications of machine learning for facies and fracture prediction using Bayesian Network Theory and Random Forest: Case studies from the Appalachian basin, USA , 2018, Journal of Petroleum Science and Engineering.

[39]  Arthur K. Kordon,et al.  Applying Computational Intelligence - How to Create Value , 2009 .

[40]  Bianca Zadrozny,et al.  A data-driven workflow for predicting horizontal well production using vertical well logs , 2017, ArXiv.

[41]  James M. Bray,et al.  Identification of Production Potential in Unconventional Reservoirs , 2007 .

[42]  Lorenzo Bruzzone,et al.  Kernel methods for remote sensing data analysis , 2009 .

[43]  Meng Li,et al.  Current and Future Applications of Distributed Acoustic Sensing as a NewReservoir Geophysics Tool , 2015 .

[44]  Chaofeng Li,et al.  Identifying organic-rich Marcellus Shale lithofacies by support vector machine classifier in the Appalachian basin , 2014, Comput. Geosci..

[45]  Shahab D. Mohaghegh,et al.  Reservoir simulation and modeling based on artificial intelligence and data mining (AI&DM) , 2011 .