A deep learning-accelerated data assimilation and forecasting workflow for commercial-scale geologic carbon storage. (December 2021)

• Record Type: Journal Article
• Authors: Tang, Hewei; Fu, Pengcheng; Sherman, Christopher S.; Zhang, Jize; Ju, Xin; Hamon, François; Azzolina, Nicholas A.; Burton-Kelly, Matthew; Morris, Joseph P.
• Is Part Of: International journal of greenhouse gas control. Volume 112(2021)
• Publication Date: 2021-12
• Subjects: Deep learning -- Data assimilation -- Geologic carbon storage -- Ensemble-based methods -- Surrogate modeling; Greenhouse gases -- Environmental aspects -- Periodicals; Air -- Purification -- Technological innovations -- Periodicals; Gaz à effet de serre -- Périodiques; Gaz à effet de serre -- Réduction -- Périodiques; Air -- Purification -- Technological innovations; Greenhouse gases -- Environmental aspects; Periodicals; 363.73874605
• DOI: 10.1016/j.ijggc.2021.103488
• Languages: English
• ISSNs: 1750-5836

Abstract

Highlights: Developed a data assimilation workflow to predict CO2 plume migration in GCS. Deep learning-based surrogate models are used to accelerate the workflow. An ES-MDA framework is adopted for inversion and uncertainty quantification. Complete history matching and forecasting in less than one hour on a workstation. Intelligent treatments are applied to make the workflow relevant and practical. Abstract: Fast assimilation of monitoring data to update forecasts of pressure buildup and carbon dioxide (CO2 ) plume migration under geologic uncertainties is a challenging problem in geologic carbon storage. The high computational cost of data assimilation with a high-dimensional parameter space impedes fast decision-making for commercial-scale reservoir management. We propose to leverage physical understandings of porous medium flow behavior with deep learning techniques to develop a fast data assimilation-reservoir response forecasting workflow. Applying an Ensemble Smoother Multiple Data Assimilation (ES-MDA) framework, the workflow updates geologic properties and predicts reservoir performance with quantified uncertainty from pressure history and CO2 plumes interpreted through seismic inversion. As the most computationally expensive component in such a workflow is reservoir simulation, we developed surrogate models to predict dynamic pressure and CO2 plume extents under multi-well injection. The surrogate models employ deep convolutional neural networks, specifically, a wide residual network and a residual U-Net. The workflow is validated against a flat three-dimensional reservoir model representative of a clastic shelf depositional environment. Intelligent treatments are applied to bridge between quantities in a true-3D reservoir model and those in a single-layer reservoir model. The workflow can complete history matching and reservoir forecasting with uncertainty quantification in less than one hour on a mainstream personal workstation.