none
10.1190/geo20190650.1
Society of Exploration Geophysicists
Society of Exploration Geophysicists
186
116891496
5981
2021031006521100339
10.1190
20221026T20:23:50Z
20200605T13:39:07Z
69
GEOPHYSICS
GEOPHYSICS
00168033
19422156
09
01
2020
85
5
Petrophysical properties prediction from prestack seismic data using convolutional neural networks
Vishal
Das
Shell Exploration and Production Company, Houston, Texas 77079, USA and Stanford University, Department of Geophysics, Stanford, California 94305, USA..
https://orcid.org/000000015815537X
Tapan
Mukerji
Stanford University, Department of Geophysics, Stanford, California 94305, USA; Stanford University, Energy Resources Engineering Department, Stanford, California 94305, USA; and Stanford University, Department of Geological Sciences, Stanford, California 94305, USA..
https://orcid.org/0000000317111850
We have built convolutional neural networks (CNNs) to obtain petrophysical properties in the depth domain from prestack seismic data in the time domain. We compare two workflows — endtoend and cascaded CNNs. An endtoend CNN, referred to as PetroNet, directly predicts petrophysical properties from prestack seismic data. Cascaded CNNs consist of two CNN architectures. The first network, referred to as ElasticNet, predicts elastic properties from prestack seismic data followed by a second network, referred to as ElasticPetroNet, that predicts petrophysical properties from elastic properties. Cascaded CNNs with more than twice the number of trainable parameters as compared to endtoend CNN demonstrate similar prediction performance for a synthetic data set. The average correlation coefficient for test data between the true and predicted clay volume (approximately 0.7) is higher than the average correlation coefficient between the true and predicted porosity (approximately 0.6) for both networks. The cascaded workflow depends on the availability of elastic properties and is three times more computationally expensive than the endtoend workflow for training. Coherence plots between the true and predicted values for both cases show that maximum coherence occurs for values of the inverse wavenumber greater than 15 m, which is approximately equal to 1/4 the source wavelength or λ/4. The network predictions have some coherence with the true values even at a resolution of 10 m, which is half of the variogram range used in simulating the spatial correlation of the petrophysical properties. The Monte Carlo dropout technique is used for approximate quantification of the uncertainty of the network predictions. An application of the endtoend network for prediction of petrophysical properties is made with the Stybarrow field located in offshore Western Australia. The network makes good predictions of petrophysical properties at the well locations. The network is particularly successful in identifying the reservoir facies of interest with high porosity and low clay volume.
09
01
2020
N41
N55
10.1190/geo20190650.1
1
10.1190/crossmarkpolicy
library.seg.org
true
20190930
20200503
20200504
20200817
Stanford School of Earth, Energy and Environmental Sciences
http://dx.doi.org/10.13039/100010864
10.1190/geo20190650.1
https://library.seg.org/doi/10.1190/geo20190650.1

https://library.seg.org/doi/pdf/10.1190/geo20190650.1

https://pubs.geoscienceworld.org/geophysics/article/85/5/n41/589719/petrophysicalpropertiespredictionfromprestack
Abadi, M., A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro, G. S. Corrado, A. Davis, J. Dean, M. Devin, S. Ghemawat, I. Goodfellow, A. Harp, G. Irving, M. Isard, Y. Jia, R. Jozefowicz, L. Kaiser, M. Kudlur, J. Levenberg, D. Mane, R. Monga, S. Moore, D. Murray, C. Olah, M. Schuster, J. Shlens, B. Steiner, I. Sutskever, K. Talwar, P. Tucker, V. Vanhoucke, V. Vasudevan, F. Viegas, O. Vinyals, P. Warden, M. Wattenberg, M. Wicke, Y. Yu, and X. Zheng, 2016, TensorFlow: Largescale machine learning on heterogeneous systems: 12th USENIX Symposium on Operating Systems Design and Implementation, 265–283.
10.1190/INT20180250.1
10.1111/j.13652478.1982.tb01328.x
10.1190/tle36030208.1
10.1190/geo20160083.1
10.1190/geo20160083.1
Avseth, P., T. Mukerji, and G. Mavko, 2010, Quantitative seismic interpretation: Cambridge University Press.
10.1093/gji/ggy511
10.1190/geo20150104.1
10.1007/9783319532011
10.1190/1.2235991
00166731
Genetics
Beaumont M. A.
162
2002
10.1093/genetics/162.4.2025
10.1201/9781420050646.ptb6
Bishop, C. M., 1995, Neural networks for pattern recognition: Oxford University Press.
10.1190/INT20180236.1
10.1080/01621459.2017.1285773
10.1007/9789401117395_27
Bortoli, L. J., F. A. Alabert, A. Haas, and A. G. Journel, 1993, Constraining stochastic images to seismic data: Proceedings of the 4th International Geostatistical Congress, Kluwer Academic Publishers, 325–337.
10.1029/1998JB900014
10.1190/1.1801944
10.1190/1.3148410
10.1190/1.2783766
10.1190/1.3043796
10.1190/1.3478209
10.1109/MSP.2017.2776602
10.1190/geo20150348.1
10.1190/1.2147935
10.3997/22144609.201901148
Das, V., and T. Mukerji, 2019a, Traditional feature based vs direct machine learning based AVO classification: 81st Annual International Conference and Exhibition, EAGE, Extended Abstract, WeR0203.
10.1190/segam20193215122.1
10.1190/geo20180838.1
10.1190/geo20170463.1
10.1190/geo20180839.1
GSLIBGeostatistical software library and user’s guide
Deutsch C. V.
1992
10.1093/gji/ggz226
10.1190/1.1438618
10.1190/1.1442404
10.3997/9789073781771
10.1190/1.9781560801962
Dumoulin, V., and F. Visin, 2016, A guide to convolution arithmetic for deep learning: arXiv preprint arXiv:1603.07285. https://arxiv.org/abs/1603.07285.
10.1190/1.1778241
10.2118/88574MS
Ementon, N., R. Hill, M. Flynn, B. Motta, and S. Sinclair, 2004, Stybarrow oil field — From seismic to production, the integrated story so far: Asia Pacific Oil and Gas Conference and Exhibition, SPE, Extended Abstracts, 88574.
10.1190/geo20170239.1
First Break
Francis A.
24
2006
First Break
Francis A.
24
2006
Gal, Y., and Z. Ghahramani, 2016, Dropout as a Bayesian approximation: Representing model uncertainty in deep learning: 33rd International Conference on Machine Learning, ACM, 1050–1059.
Regional geology of the Northern Carnarvon Basin
Geoscience Australia
2013
10.1016/j.petrol.2017.03.013
Glorot, X., and Y. Bengio, 2010, Understanding the difficulty of training deep feedforward neural networks: Proceedings of the 13th International Conference on Artificial Intelligence and Statistics, 249–256.
10.1190/1.2803748
Deep learning
Goodfellow I.
2016
10.1190/1.3386676
10.1190/geo20160161.1
10.1016/j.cageo.2003.10.013
10.3997/13652397.1994034
10.1190/1.1438620
The Journal of Machine Learning Research
Hoffman M. D.
14
2013
Ioffe, S., and C. Szegedy, 2015, Batch normalization: Accelerating deep network training by reducing internal covariate shift, arXiv preprint arXiv:1502.03167, https://arxiv.org/abs/1502.03167.
Kingma, D. P., and J. Ba, 2014, Adam: A method for stochastic optimization, arXiv preprint arXiv:1412.6980, https://arxiv.org/abs/1412.6980.
10495258
Advances in Neural Information Processing Systems
Krizhevsky A.
2012
Machine learning to enhance the vertical resolution of seismic geostatistical inversion
Liu Y.
2018
10.1190/1.1527087
10.1017/CBO9780511626753
10.3997/22144609.201901609
Mosser, L., O. Dubrule, and M. Blunt, 2019a, Deep stochastic inversion: 81st Annual International Conference and Exhibition, Extended Abstracts, EAGE, ThR1104.
10.3997/22144609.201901510
Mosser, L., R. Oliveira, and M. Steventon, 2019b, Probabilistic seismic interpretation using Bayesian neural networks: 81st Annual International Conference and Exhibition, Extended Abstracts, EAGE, ThR0313.
10.1190/1.1438938
10.1190/1.1487078
Nair, V., and G. E. Hinton, 2010, Rectified linear units improve restricted Boltzmann machines: Proceedings of the 27th International Conference on Machine Learning, 807–814.
10.1515/9780691206837
10.1190/INT20180202.1
10.1190/segam20182998259.1
Pradhan, A., and T. Mukerji, 2018, Seismic estimation of reservoir properties with Bayesian evidential analysis: 88th Annual International Meeting, SEG, Expanded Abstracts, 3166–3170.
10.1093/oxfordjournals.molbev.a026091
10.1017/CBO9781139150019
10.1214/aos/1176346785
10.1190/1.9781560802303
10.1190/1.1993269
10.2118/57260MS
Sams, M. S., D. Atkins, N. Siad, E. Parwito, and P. van Riel, 1999, Stochastic inversion for high resolution reservoir characterization in the central Sumatra basin: Presented at SPE Asia Pacific Improved Oil Recovery Conference, 57260.
Global optimization methods in geophysical inversion
Sen M.
1995
10.1190/1.2542449
10.1190/INT20180235.1
10.1190/1.2760162
Journal of Machine Learning Research
Srivastava N.
15
2014
Inverse problem theory: Methods for data fitting and model parameter estimation
Tarantola A.
1987
10.1190/tle37070529.1
10.1190/tle36110910.1
10.1109/TGRS.2002.800278
10.1190/geo20170666.1
10.1190/geo20180249.1
10.1007/s0036500606632
10.1190/segam20182997085.1
10.1190/tle38070526.1
MathematischPhysikalische Klasse
Zoeppritz K.
1919