Deep Learning-Accelerated 3D Carbon Storage Reservoir Pressure Forecasting Based on Data Assimilation Using Surface Displacement from InSAR

Fast forecasting of the reservoir pressure distribution during geologic carbon storage (GCS) by assimilating monitoring data is a challenging problem. Due to high drilling cost, GCS projects usually have spatially sparse measurements from few wells, leading to high uncertainties in reservoir pressure prediction. To address this challenge, we use low-cost Interferometric Synthetic-Aperture Radar (InSAR) data as monitoring data to infer reservoir pressure build up. We develop a deep learning-accelerated workflow to assimilate surface displacement maps interpreted from InSAR and to forecast dynamic reservoir pressure. Employing an Ensemble Smoother Multiple Data Assimilation (ES-MDA) framework, the workflow updates threedimensional (3D) geologic properties and predicts reservoir pressure with quantified uncertainties. We use a synthetic commercial-scale GCS model with bimodally distributed permeability and porosity to demonstrate the efficacy of the workflow. A two-step CNN-PCA approach is employed to parameterize the bimodal fields. The computational efficiency of the workflow is boosted by two residual U-Net based surrogate models for surface displacement and reservoir pressure predictions, respectively. The workflow can complete data assimilation and reservoir pressure forecasting in half an hour on a personal computer.

[1]  Vincent Dumoulin,et al.  Deconvolution and Checkerboard Artifacts , 2016 .

[2]  Omar Ghattas,et al.  Bayesian Poroelastic Aquifer Characterization From InSAR Surface Deformation Data. Part I: Maximum A Posteriori Estimate , 2020, Water Resources Research.

[3]  Albert C. Reynolds,et al.  Ensemble smoother with multiple data assimilation , 2013, Comput. Geosci..

[4]  Leandro Passos de Figueiredo,et al.  Uncertainty quantification in Bayesian inverse problems with model and data dimension reduction , 2019, GEOPHYSICS.

[5]  Howard A. Zebker,et al.  Inverse modeling of interbed storage parameters using land subsidence observations, Antelope Valley, California , 2003 .

[6]  B. Jafarpour,et al.  Dynamic characterization of geologic CO2 storage aquifers from monitoring data with ensemble Kalman filter , 2019, International Journal of Greenhouse Gas Control.

[7]  M. Biot.  General Theory of Three‐Dimensional Consolidation , 1941 .

[8]  Lorenzo Torresani,et al.  Learning Spatiotemporal Features with 3D Convolutional Networks , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[9]  Louis J. Durlofsky,et al.  Deep-learning-based surrogate flow modeling and geological parameterization for data assimilation in 3D subsurface flow , 2020, Computer Methods in Applied Mechanics and Engineering.

[10]  S. Bachu,et al.  Permeability and Relative Permeability Measurements at Reservoir Conditions for CO2-Water Systems in Ultra Low Permeability Confining Caprocks , 2007 .

[11]  Jichun Wu,et al.  Deep Convolutional Encoder‐Decoder Networks for Uncertainty Quantification of Dynamic Multiphase Flow in Heterogeneous Media , 2018, Water Resources Research.

[12]  Zhenhao Duan,et al.  An improved model calculating CO2 solubility in pure water and aqueous NaCl solutions from 273 to 533 K and from 0 to 2000 bar , 2003 .

[13]  Javier E. Santos,et al.  Machine learning assisted history matching for a deepwater lobe system , 2021 .

[14]  D. McLaughlin,et al.  Data inversion in coupled subsurface flow and geomechanics models , 2012 .

[15]  Pengcheng Fu,et al.  A Deep Learning-Accelerated Data Assimilation and Forecasting Workflow for Commercial-Scale Geologic Carbon Storage , 2021, International Journal of Greenhouse Gas Control.

[16]  Don W. Vasco,et al.  Reservoir monitoring and characterization using satellite geodetic data: Interferometric Synthetic Aperture Radar observations from the Krechba field, Algeria , 2008 .

[17]  Howard A. Zebker,et al.  Seasonal subsidence and rebound in Las Vegas Valley, Nevada, observed by Synthetic Aperture Radar Interferometry , 2001 .

[18]  Matthew W Becker,et al.  Potential for Satellite Remote Sensing of Ground Water , 2006, Ground water.

[19]  Nicholas Zabaras,et al.  Integration of adversarial autoencoders with residual dense convolutional networks for estimation of non-Gaussian hydraulic conductivities , 2020 .

[20]  Wesley D. Peck,et al.  Risk‐based area of review estimation in overpressured reservoirs to support injection well storage facility permit requirements for CO 2 storage projects , 2021, Greenhouse Gases: Science and Technology.

[21]  Louis J. Durlofsky,et al.  3D CNN-PCA: A Deep-Learning-Based Parameterization for Complex Geomodels , 2020, Computational Geosciences.

[22]  Louis J. Durlofsky,et al.  Data-space inversion using a recurrent autoencoder for time-series parameterization , 2020, Computational Geosciences.

[23]  Hewei Tang,et al.  Predicting Thermal Performance of an Enhanced Geothermal System From Tracer Tests in a Data Assimilation Framework , 2021, Water Resources Research.

[24]  D. Vasco,et al.  Coupled reservoir-geomechanical analysis of CO2 injection and ground deformations at In Salah, Algeria , 2010 .

[25]  Joshua A. White,et al.  Geomechanical behavior of the reservoir and caprock system at the In Salah CO2 storage project , 2014, Proceedings of the National Academy of Sciences.

[26]  Charles D. Gorecki,et al.  Quantifying the effects of depositional environment on deep saline formation co2 storage efficiency and rate , 2018 .

[27]  M. Balhoff,et al.  Optimization of subsurface CO2 injection based on neural network surrogate modeling , 2021, Computational Geosciences.

[28]  Zhiming Lu,et al.  Reducing uncertainty in geologic CO2 sequestration risk assessment by assimilating monitoring data , 2020 .

[29]  Louis J. Durlofsky,et al.  Deep-learning-based coupled flow-geomechanics surrogate model for CO$_2$ sequestration , 2021, 2105.01334.

[30]  Louis J. Durlofsky,et al.  A New Differentiable Parameterization Based on Principal Component Analysis for the Low-Dimensional Representation of Complex Geological Models , 2014, Mathematical Geosciences.