Deep Learning Framework for History Matching CO2 Storage with 4D Seismic and Monitoring Well Data

TL;DR

This work develops a dual deep learning surrogate framework to enable efficient history matching of geological CO$_2$ storage scenarios using early-time monitoring data and 4D seismic observations. A 3D U-Net subsea seismic surrogate and a 1D U-Net monitoring surrogate predict interpreted seismic saturation and vertical well saturations, respectively, from high-resolution geomodel inputs. These surrogates feed a hierarchical MCMC history-matching scheme that samples PCA latent variables and geomodel metaparameters, achieving substantial uncertainty reduction in key parameters and improved CO$_2$ plume predictions, demonstrated on synthetic cases with and without seismic data. The approach offers a scalable, data-type-specific pathway to quantify the value of different data streams and can be extended to incorporate additional measurements and direct seismic-data assimilation for practical CO$_2$ storage operations.

Abstract

Geological carbon storage entails the injection of megatonnes of supercritical CO2 into subsurface formations. The properties of these formations are usually highly uncertain, which makes design and optimization of large-scale storage operations challenging. In this paper we introduce a history matching strategy that enables the calibration of formation properties based on early-time observations. Early-time assessments are essential to assure the operation is performing as planned. Our framework involves two fit-for-purpose deep learning surrogate models that provide predictions for in-situ monitoring well data and interpreted time-lapse (4D) seismic saturation data. These two types of data are at very different scales of resolution, so it is appropriate to construct separate, specialized deep learning networks for their prediction. This approach results in a workflow that is more straightforward to design and more efficient to train than a single surrogate that provides global high-fidelity predictions. The deep learning models are integrated into a hierarchical Markov chain Monte Carlo (MCMC) history matching procedure. History matching is performed on a synthetic case with and without 4D seismic data, which allows us to quantify the impact of 4D seismic on uncertainty reduction. The use of both data types is shown to provide substantial uncertainty reduction in key geomodel parameters and to enable accurate predictions of CO2 plume dynamics. The overall history matching framework developed in this study represents an efficient way to integrate multiple data types and to assess the impact of each on uncertainty reduction and performance predictions.

Figures (21)

• Figure 1: Model and wells used in all simulations.
• Figure 2: Original SGeMS and reconstructed PCA fields.
• Figure 3: Schematic showing the workflow for generating interpreted seismic data from high-fidelity simulation results. Images on the left are $x$-$z$ cross sections from the full 3D model and those on the right are $y$-$z$ cross sections. Interpreted seismic data can provide estimates of either seismic-scale $S_g$ or plume location.
• Figure 4: Schematic of U-Net surrogate model used to predict interpreted seismic saturation data.
• Figure 5: Architecture of 1D U-Net surrogate for monitoring data.