Surrogate Model for Geological CO2 Storage and Its Use in Hierarchical MCMC History Matching

TL;DR

The paper tackles history matching of geological CO$_2$ storage models under substantial geological scenario uncertainty. It extends a 3D extended recurrent R-U-Net surrogate to handle geomodel realizations defined by metaparameters and embeds this surrogate in a hierarchical, dimension-robust MCMC framework to jointly infer metaparameters and geomodels from monitoring data. Key contributions include training the surrogate on $2000$ GEOS realizations, achieving median errors of $4.5\%$ in saturation and $1.3\%$ in pressure across $500$ tests, and attaining substantial uncertainty reduction in $\mu_{\log k}$, $\sigma_{\log k}$, and $a_r$, with posterior 3D fields closely matching true-model responses. The approach yields large computational speedups (approximately $3000\times$) enabling tens of thousands of likelihood evaluations, thereby enabling practical, scalable history matching for large-scale CO$_2$ storage with uncertain geology.

Abstract

Deep-learning-based surrogate models show great promise for use in geological carbon storage operations. In this work we target an important application - the history matching of storage systems characterized by a high degree of (prior) geological uncertainty. Toward this goal, we extend the recently introduced recurrent R-U-Net surrogate model to treat geomodel realizations drawn from a wide range of geological scenarios. These scenarios are defined by a set of metaparameters, which include the horizontal correlation length, mean and standard deviation of log-permeability, permeability anisotropy ratio, and constants in the porosity-permeability relationship. An infinite number of realizations can be generated for each set of metaparameters, so the range of prior uncertainty is large. The surrogate model is trained with flow simulation results, generated using the open-source simulator GEOS, for 2000 random realizations. The flow problems involve four wells, each injecting 1 Mt CO2/year, for 30 years. The trained surrogate model is shown to provide accurate predictions for new realizations over the full range of geological scenarios, with median relative error of 1.3% in pressure and 4.5% in saturation. The surrogate model is incorporated into a hierarchical Markov chain Monte Carlo history matching workflow, where the goal is to generate history matched geomodel realizations and posterior estimates of the metaparameters. We show that, using observed data from monitoring wells in synthetic `true' models, geological uncertainty is reduced substantially. This leads to posterior 3D pressure and saturation fields that display much closer agreement with the true-model responses than do prior predictions.

Figures (25)

• Figure 1: Model domains for flow simulations. Full domain including surrounding region shown on the left, central storage aquifer shown on the right.
• Figure 2: Schematic of the 3D extended recurrent R-U-Net architecture. 3D R-U-Net architecture for pressure and saturation prediction at a specific time step is shown in (a). Incorporation of 3D R-U-Net within recurrent neural network shown in (b). Geomodels are defined by three input channels.
• Figure 3: Two-phase flow curves. Capillary pressure curve in (b) is for $\phi=0.1$ and $k=10$ md.
• Figure 4: Locations of injection wells (I1 -- I4) and observation wells (O1 -- O4) in the $80 \times 80 \times 20$ storage aquifer. All wells are fully penetrating.
• Figure 5: Histograms of relative errors for the 500 test cases.