A Machine-Learning Methodology Using Domain-Knowledge Constraints for Well-Data Integration and Well-Production Prediction
- Jorge Guevara (IBM Research) | Bianca Zadrozny (IBM Research) | Alvaro Buoro (IBM Research) | Ligang Lu (Shell Exploration and Production) | John Tolle (Shell Exploration and Production) | Jan W. Limbeck (Shell Exploration and Production) | Detlef Hohl (Shell Exploration and Production)
- Document ID
- Society of Petroleum Engineers
- SPE Reservoir Evaluation & Engineering
- Publication Date
- November 2019
- Document Type
- Journal Paper
- 1,185 - 1,200
- 2019.Society of Petroleum Engineers
- predictive modeling, data mining, sweet spotting, statistical models, machine learning
- 60 in the last 30 days
- 300 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 12.00|
|SPE Non-Member Price:||USD 35.00|
In this paper, we propose a machine-learning methodology using domain-knowledge constraints for well-data integration, prior/expert-knowledge incorporation, and sweet-spot identification. Such methodology enables the analysis of the effects of the main variables involved in production prediction and the evaluation of what-if scenarios of production prediction within geological zones. This methodology will allow streamlining the process of data integration, analytics and machine learning for better decisions, saving time, and helping geologists and reservoir and completion engineers in the task of sweet-spot identification and completion design. We tested the proposed methodology with production, completions, and petrophysical data from a field within a geological target zone. For instance, using local Kriging, we estimated gamma ray features from gamma ray measurements from vertical and horizontal wells, and we integrated those features into the production- and completion-well data, generating an integrated data set for machine-learning modeling. Besides the usual black-box machine-learning models, we used generalized additive models (GAMs) and shape-constraint additive models (SCAMs) for predictive modeling. Those models permit the incorporation of prior/expert knowledge in terms of interaction terms and mathematical constraints on the shape of the effect of the covariates, such as petrophysical and completion parameters, resulting in greater accuracy and interpretability of the predicted production vs. classical black-box machine-learning modeling. We also defined hypothetical what-if scenarios of oil production, such as by estimating the empirical distributions of production estimates using hypothetical settings for completions within the region of interest.
|File Size||1 MB||Number of Pages||16|
Ambroise, C. and McLachlan, G. J. 2002. Selection Bias in Gene Extraction on the Basis of Microarray Gene-Expression Data. Proc Natl Acad Sci 99 (10): 6562–6566. https://doi.org/10.1073/pnas.102102699.
Assael, Y. M., Shillingford, B., Whiteson, S. et al. 2016. LipNet: End-to-End Sentence-Level Lipreading. arXiv preprint arXiv: 1611.01599.
Bibor, I. and Szabó, N. P. 2016. Unconventional Shale Characterization Using Improved Well Logging Methods. Geosci Eng 5 (8): 32–50.
Bowman, T. 2010. Direct Method for Determining Organic Shale Potential From Porosity and Resistivity Logs to Identify Possible Resource Plays. Oral presentation given at the AAPG Annual Convention and Exhibition, New Orleans, 11–14 April.
Breheny, P. and Burchett, W. 2013. Visualization of Regression Models Using visreg. The R Journal 9 (2): 56–71.
Cawley, G. C. and Talbot, N. L. 2010. On Over-Fitting in Model Selection and Subsequent Selection Bias in Performance Evaluation. J Mach Learn Res 11: 2079–2107.
Dimitrakopoulos, R. and Luo, X. 2004. Generalized Sequential Gaussian Simulation on Group Size ν and Screen-Effect Approximations for Large Field Simulations. Mathemat Geol 36 (5): 567–591. https://doi.org/10.1023/B:MATG.0000037737.11615.df.
Fernandez-Concheso, J. E. 2015. Characterizing an Unconventional Reservoir With Conventional Seismic Data: A Case Study Using Seismic Inversion for the Vaca Muerta Formation, Neuquen Basin, Argentina. PhD dissertation, Colorado School of Mines, Golden, Colorado.
Flaxman, S., Sejdinovic, D., Cunningham, J. P. et al. 2016. Bayesian Learning of Kernel Embeddings. arXiv preprint arXiv: 1603.02160.
Friedman, J., Hastie, T., and Tibshirani, R. 2001. The Elements of Statistical Learning, Vol. 1. New York City: Series in Statistics, Springer.
Friedman, J. H. 2001. Greedy Function Approximation: A Gradient Boosting Machine. Ann. Statist. 29 (5): 1189–1232. https://doi.org/10.1214/aos/1013203451.
Goldstein, A., Kapelner, A., Bleich, J. et al. 2015. Peeking Inside the Black Box: Visualizing Statistical Learning With Plots of Individual Conditional Expectation. J Comput Graph Stat 24 (1): 44–65. https://doi.org/10.1080/10618600.2014.907095.
Grace, K., Salvatier, J., Dafoe, A. et al. 2017. When Will AI Exceed Human Performance? Evidence From AI Experts. arXiv preprint arXiv: 1705.08807.
Guevara, J., Kormaksson, M., Zadrozny, B. et al. 2017. A Data-Driven Workflow for Predicting Horizontal Well Production Using Vertical Well Logs. Proc., DM4OG 2017, Houston, 27–29 April.
Guevara, J., Kormaksson, M., Zadrozny, B. et al. 2018. A Hybrid Data-Driven and Knowledge-Driven Methodology for Estimating the Effect of Completion Parameters on the Cumulative Production of Horizontal Wells. Presented at the SPE Annual Technical Conference and Exhibition, Dallas, 24–26 September. SPE-191446-MS. https://doi.org/10.2118/191446-MS.
Hastie, T. J. 2017. Generalized Additive Models. In Statistical Models in S, ed. J. M. Chambers and T. J. Hastie, 249–307. Abingdon-on-Thames, UK: Routledge.
Kormaksson, M., Vieira, M. R., and Zadrozny, B. 2015. A Data Driven Method For Sweet Spot Identification in Shale Plays Using Well Log Data. Presented at the SPE Digital Energy Conference and Exhibition, The Woodlands, Texas, 3–5 March. SPE-173455-MS. https://doi.org/10.2118/173455-MS.
Liu, X. 2013. Workflows for Sweet Spots Identification in Shale Plays Using Seismic Inversion and Well Logs. Presented at the Integration GeoConvention 2013, Calgary, 6–12 May.
Matta, C. F., Massa, L., Gubskaya, A. V. et al. 2010. Can One Take the Logarithm or the Sine of a Dimensioned Quantity or a Unit? Dimensional Analysis Involving Transcendental Functions. J. Chem. Educ. 88 (1): 67–70. https://doi.org/10.1021/ed1000476.
Meyer, B. and Nederlof, M. 1984. Identification of Source Rocks on Wireline Logs by Density/Resistivity and Sonic Transit Time/Resistivity Crossplots. AAPG Bull 68 (2): 121–129. https://doi.org/10.2307/1162357.
Mitchell, T. M. 1997. Machine Learning, first edition. New York City: McGraw-Hill.
Nunez, G. J., Sandoval Hernandez, G., Stundner, M. et al. 2010. Integrating Data Mining and Expert Knowledge for an Artificial Lift Advisory System. Presented at the SPE Intelligent Energy Conference and Exhibition, Utrecht, The Netherlands, 23–25 March. SPE-128636-MS. https://doi.org/10.2118/128636-MS.
Passey, Q., Creaney, S., Kulla, J. et al. 1990. A Practical Model for Organic Richness From Porosity and Resistivity Logs. AAPG Bull 74 (12): 1777–1794.
Pya, N. and Wood, S. N. 2015. Shape Constrained Additive Models. Stat Comput 25 (3): 543–559. https://doi.org/10.1007/s11222-013-9448-7.
Rajpurkar, P., Irvin, J., Zhu, K. et al. 2017. CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays With Deep Learning. arXiv preprint arXiv: 1711.05225.
Rassmussen, C. E. 2004. Gaussian Processes in Machine Learning. In Advanced Lectures on Machine Learning, eds. O. Bousquet, U. von Luxburg, G. Rätsch, ML 2003 Lecture Notes in Computer Science, Vol. 3176, 63–71. Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-540-28650-9_4.
Samuel, A. L. 1959. Some Studies in Machine Learning Using the Game of Checkers. IBM J Res Develop 3 (3): 210–229. https://doi.org/10.1147/rd.33.0210.
Sharma, R. K. and Chopra, S. 2016. Identification of Sweet Spots in Shale Reservoir Formations. First Break 34 (9): 43–51. https://doi.org/10.3997/1365-2397.2016012.