Unsupervised learning elastic rock properties from pre-stack seismic data

Abstract As for a typical machine-learning method, large volume of labeled examples is required to train and update neural parameters that are later used on new datasets. Well logs usually provide true values of elastic rock properties and enough amount of their correct answers may not often be available, due to budget limit and consumed time. In order to solve this problem, an unsupervised machine-learning approach is proposed to estimate rock properties based on pre-stack seismic data. The pre-designed Convolutional Neural Networks is capable to transform input gathers to relative rock properties, which are going to be added with a low-frequency component. Simulation of wavefields is then incorporated in this novel system, and neural weights and biases are altered to minimize a data-to-data loss function. The synthetic model of Book Cliffs is used for an application of this new method, in which both subsurface structures and property values are recovered very well. The proposed approach is further applied to a real dataset for the geothermal exploration in Denmark. Attributed to a relatively large distance between the well location and seismic line, inverted results are suboptimal.

[1]  Andrew Curtis,et al.  Fast probabilistic petrophysical mapping of reservoirs from 3D seismic data , 2012 .

[2]  Runhai Feng,et al.  A Bayesian Approach in Machine Learning for Lithofacies Classification and Its Uncertainty Analysis , 2021, IEEE Geoscience and Remote Sensing Letters.

[3]  Runhai Feng,et al.  Lithofacies classification based on a hybrid system of artificial neural networks and hidden Markov models , 2020 .

[4]  S. Poulsen,et al.  Inverse geothermal modelling applied to Danish sedimentary basins , 2017 .

[5]  George A. McMechan,et al.  Feature-capturing full-waveform inversion using a convolutional neural network , 2018, SEG Technical Program Expanded Abstracts 2018.

[6]  Tim Salimans,et al.  Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks , 2016, NIPS.

[7]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[8]  Tapan Mukerji,et al.  Convolutional neural network for seismic impedance inversion , 2019 .

[9]  P. Jilek,et al.  Density extraction from P-wave AVO inversion: Tuscaloosa Trend example , 2010 .

[10]  S. Mallick,et al.  Amplitude-variation-with-offset and prestack-waveform inversion: A direct comparison using a real data example from the Rock Springs Uplift, Wyoming, USA , 2015 .

[11]  R. Feng.  Estimation of reservoir porosity based on seismic inversion results using deep learning methods , 2020 .

[12]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[13]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[14]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[15]  How reliable is statistical wavelet estimation , 2011 .

[16]  Erika Angerer,et al.  Reservoir lithology classification based on seismic inversion results by Hidden Markov Models: Applying prior geological information , 2018 .

[17]  Dries Gisolf,et al.  Obtaining a high-resolution geological and petrophysical model from the results of reservoir-orientated elastic wave-equation-based seismic inversion , 2017, Petroleum Geoscience.

[18]  Erika Angerer,et al.  Reservoir Lithology Determination by Hidden Markov Random Fields Based on a Gaussian Mixture Model , 2018, IEEE Transactions on Geoscience and Remote Sensing.

[19]  A. Buland,et al.  Bayesian linearized AVO inversion , 2003 .

[20]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[21]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[22]  T. Mukerji,et al.  Rock-physics diagnostics of depositional texture, diagenetic alterations, and reservoir heterogeneity in high-porosity siliciclastic sediments and rocks — A review of selected models and suggested work flows , 2010 .

[23]  Mrinal K. Sen,et al.  A Pre-Stack Basis Pursuit Seismic Inversion , 2012 .

[24]  P. Avseth,et al.  AVO inversion in exploration — Key learnings from a Norwegian Sea prospect , 2016 .

[25]  Mrinal K. Sen,et al.  Global Optimization Methods in Geophysical Inversion , 1995 .

[26]  Mingliang Liu,et al.  Time-lapse seismic history matching with an iterative ensemble smoother and deep convolutional autoencoder , 2020 .

[27]  Guangmin Hu,et al.  Unsupervised seismic facies analysis via deep convolutional autoencoders , 2018 .

[28]  Roel Snieder,et al.  To Bayes or not to Bayes , 1997 .

[29]  Ernesto Della Rossa,et al.  Probabilistic petrophysical-properties estimation integrating statistical rock physics with seismic inversion , 2010 .

[30]  Ran Bachrach,et al.  Joint estimation of porosity and saturation using stochastic rock-physics modeling , 2006 .

[31]  Zehui Huang,et al.  Permeability prediction with artificial neural network modeling in the Venture gas field, offshore eastern Canada , 1996 .

[32]  Mingliang Liu,et al.  Accelerating geostatistical seismic inversion using TensorFlow: A heterogeneous distributed deep learning framework , 2019, Comput. Geosci..

[33]  Victoria Stodden,et al.  Making massive computational experiments painless , 2016, 2016 IEEE International Conference on Big Data (Big Data).