Adaptive Surrogate Estimation with Spatial Features Using a Deep Convolutional Autoencoder for CO2 Geological Sequestration

This paper develops a reliable deep-learning framework to extract latent features from spatial properties and investigates adaptive surrogate estimation to sequester CO2 into heterogeneous deep saline aquifers. Our deep-learning architecture includes a deep convolutional autoencoder (DCAE) and a fully-convolutional network to not only reduce computational costs but also to extract dimensionality-reduced features to conserve spatial characteristics. The workflow integrates two different spatial properties within a single convolutional system, and it also achieves accurate reconstruction performance. This approach significantly reduces the number of parameters to 4.3% of the original number required, e.g., the number of three-dimensional spatial properties needed decreases from 44,460 to 1920. The successful dimensionality reduction is accomplished by the DCAE system regarding all inputs as image channels from the initial stage of learning using the fully-convolutional network instead of fully-connected layers. The DCAE reconstructs spatial parameters such as permeability and porosity while conserving their statistical values, i.e., their mean and standard deviation, achieving R-squared values of over 0.972 with a mean absolute percentage error of their mean values of less than 1.79%. The adaptive surrogate model using the latent features extracted by DCAE, well operations, and modeling parameters is able to accurately estimate CO2 sequestration performances. The model shows R-squared values of over 0.892 for testing data not used in training and validation. The DCAE-based surrogate estimation exploits the reliable integration of various spatial data within the fully-convolutional network and allows us to evaluate flow behavior occurring in a subsurface domain.

[1]  Fabrice Rossi,et al.  Mean Absolute Percentage Error for regression models , 2016, Neurocomputing.

[2]  Keiron O'Shea,et al.  An Introduction to Convolutional Neural Networks , 2015, ArXiv.

[3]  Joe M. Kang,et al.  Data-driven inverse modeling with a pre-trained neural network at heterogeneous channel reservoirs , 2018, Journal of Petroleum Science and Engineering.

[4]  W. F. Yellig,et al.  Determination and Prediction of CO2 Minimum Miscibility Pressures (includes associated paper 8876 ) , 1980 .

[5]  Ming Zhong,et al.  A Data-Analytics Tutorial: Building Predictive Models for Oil Production in an Unconventional Shale Reservoir , 2018 .

[6]  Sven Behnke,et al.  Evaluation of Pooling Operations in Convolutional Architectures for Object Recognition , 2010, ICANN.

[7]  Behnam Jafarpour,et al.  Convolutional neural networks (CNN) for feature-based model calibration under uncertain geologic scenarios , 2020, Computational Geosciences.

[8]  Yimin Liu,et al.  A Deep-Learning-Based Geological Parameterization for History Matching Complex Models , 2018, Mathematical Geosciences.

[9]  Marco Maggipinto,et al.  A Convolutional Autoencoder Approach for Feature Extraction in Virtual Metrology , 2018 .

[10]  Richard K. G. Do,et al.  Convolutional neural networks: an overview and application in radiology , 2018, Insights into Imaging.

[11]  Yoshua Bengio,et al.  Why Does Unsupervised Pre-training Help Deep Learning? , 2010, AISTATS.

[12]  Ilsik Jang,et al.  Restoration of Missing Pressures in a Gas Well Using Recurrent Neural Networks with Long Short-Term Memory Cells , 2020 .

[13]  Louis J. Durlofsky,et al.  A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems , 2019, J. Comput. Phys..

[14]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[15]  Yoshua Bengio,et al.  Convolutional networks for images, speech, and time series , 1998 .

[16]  Kyungbook Lee,et al.  Deep neural network coupled with distance-based model selection for efficient history matching , 2020 .

[17]  Baehyun Min,et al.  Determination of an infill well placement using a data-driven multi-modal convolutional neural network , 2020 .

[18]  Roland N. Horne,et al.  Simulating the Behavior of Reservoirs with Convolutional and Recurrent Neural Networks , 2020 .

[19]  Turgay Ertekin,et al.  Artificial Intelligence Applications in Reservoir Engineering: A Status Check , 2019, Energies.

[20]  Priya Ravi Ganesh,et al.  Building Statistical Proxy Models for CO2 Geologic Sequestration , 2014 .

[21]  Marco Aurélio Cavalcanti Pacheco,et al.  Towards a Robust Parameterization for Conditioning Facies Models Using Deep Variational Autoencoders and Ensemble Smoother , 2018, Comput. Geosci..

[22]  Morteza Haghighat Sefat,et al.  Development of an adaptive surrogate model for production optimization , 2015 .

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

[24]  Changhyup Park,et al.  Surrogate Model with a Deep Neural Network to Evaluate Gas–Liquid Flow in a Horizontal Pipe , 2020 .

[25]  Nicholas Zabaras,et al.  Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification , 2018, J. Comput. Phys..

[26]  Shahab D. Mohaghegh,et al.  Modeling pressure and saturation distribution in a CO 2 storage project using a Surrogate Reservoir Model (SRM) , 2014 .

[27]  Jürgen Schmidhuber,et al.  Stacked Convolutional Auto-Encoders for Hierarchical Feature Extraction , 2011, ICANN.

[28]  M. Kramer Nonlinear principal component analysis using autoassociative neural networks , 1991 .

[29]  D. Grana,et al.  Petrophysical characterization of deep saline aquifers for CO2 storage using ensemble smoother and deep convolutional autoencoder , 2020 .

[30]  Qiang Chen,et al.  Network In Network , 2013, ICLR.

[31]  Samy Bengio,et al.  Understanding deep learning requires rethinking generalization , 2016, ICLR.

[32]  Rob J Hyndman,et al.  Another look at measures of forecast accuracy , 2006 .