Petrophysical characterization of deep saline aquifers for CO2 storage using ensemble smoother and deep convolutional autoencoder

Liu, Mingliang; Grana, Dario

doi:10.1016/j.advwatres.2020.103634

Cited by 47 publications

(15 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another challenge in hierarchical Bayesian inversion is to invert high‐dimensional large spatial fields. We often perform dimension reduction techniques before the actual inversion to make those problems more tractable: principal component analysis (Fouedjio et al., 2021; Grana et al., 2019; Kitanidis & Lee, 2014; Lee & Kitanidis, 2014; Scheidt et al., 2018; Sun & Durlofsky, 2017; Yin et al., 2020), kernel principal component analysis (Sarma et al., 2007, 2008; Scheidt & Caers, 2009), optimized‐based principal component analysis (Vo & Durlofsky, 2015), convolutional autoencoder (Liu & Grana, 2020) and convolutional variational autoencoder (Canchumuni et al., 2019). These dimension reduction techniques are mostly bijective or we can use approximations (Scheidt & Caers, 2009) to project back to high‐dimensional spatial fields.…”

Section: Introductionmentioning

confidence: 99%

Hierarchical Bayesian Inversion of Global Variables and Large‐Scale Spatial Fields

Wang

Kitanidis

Caers

2022

Water Resources Research

View full text Add to dashboard Cite

Estimating hydrological properties and quantifying their uncertainty from observed measurements is vital for groundwater exploration and exploitation decision-making processes. Bayesian inversion is now commonly applied to obtain posterior distributions (Tarantola, 2005). Two popular Bayesian inversion approaches are (a) the Markov chain Monte Carlo (MCMC) method, (b) the Monte Carlo ensemble method.Many Markov chain Monte Carlo (MCMC) methods (Brooks et al., 2011) have been well studied to solve Bayesian inversion problems in hydrology and other subsurface reservoirs:

show abstract

Section: Introductionmentioning

confidence: 99%

Hierarchical Bayesian Inversion of Global Variables and Large‐Scale Spatial Fields

Wang

Kitanidis

Caers

2022

Water Resources Research

View full text Add to dashboard Cite

show abstract

“…A comprehensive characterization is often performed through the inversion of indirect hydraulic or geophysical data, such as hydraulic and tracer testing data (Berkowitz, 2002;Chen et al, 2013;Somogyvári et al, 2017;Vogt et al, 2012;Wu, Fu, Hawkins, et al, 2021), electrical resistivity (Johnson et al, 2021;Wu et al, 2019), seismic (Emerick, 2018;Liu & Grana, 2020), and so on. A key component of hydraulic/geophysical inversion is a reliable model that can properly simulate the underlying physical processes and output model responses for given model parameters.…”

Section: Introductionmentioning

confidence: 99%

Selecting appropriate model complexity: An example of tracer inversion for thermal prediction in enhanced geothermal systems

Jin

et al. 2022

Preprint

View full text Add to dashboard Cite

A major challenge in the inversion of subsurface parameters is the ill-posedness issue caused by the inherent subsurface complexities and the generally spatially sparse data. Appropriate simplifications of inversion models are thus necessary to make the inversion process tractable and meanwhile preserve the predictive ability of the inversion results. In the present study, we investigate the effect of model complexity on the inversion of fracture aperture distribution as well as the prediction of longterm thermal performance in a field-scale single-fracture EGS model. Principal component analysis (PCA) was used to map the original cell-based aperture field to a low-dimensional latent space. The complexity of the inversion model was quantitatively represented by the percentage of total variance in the original aperture fields preserved by the latent space. Tracer, pressure and flow rate data were used to invert for fracture aperture through an ensemble-based inversion method, and the inferred aperture field was then used to predict thermal performance. We found that an over-simplified aperture model could not reproduce the inversion data and the predicted thermal response was biased. A complex aperture model could reproduce the data but the thermal prediction showed significant uncertainty. A model with moderate complexity, although not resolving many fine features in the "true" aperture field, successfully matched the data and predicted the long-term thermal behavior. The results provide important insights into the selection of model complexity for effective subsurface reservoir inversion and prediction.

show abstract

“…With improving computational power and emerging novel machine learning techniques, several works have applied models such as convolutional neural networks as surrogate models to mimic CO2 plume evolution [13][14][15][16][17]. However, the more traditional reduced order models still fail to use in the actual field setting because it is difficult to parameterize heterogeneous material properties by a few parameters [18][19][20].…”

Section: Introductionmentioning

confidence: 99%

Predicting the CO2 propagation in geological formations from sparsely available well data

Ferreira¹,

Stępień²,

Hosseinzadehsadati³

et al. 2022

SSRN Journal

View full text Add to dashboard Cite

Petrophysical characterization of deep saline aquifers for CO2 storage using ensemble smoother and deep convolutional autoencoder

Cited by 47 publications

References 77 publications

Hierarchical Bayesian Inversion of Global Variables and Large‐Scale Spatial Fields

Hierarchical Bayesian Inversion of Global Variables and Large‐Scale Spatial Fields

Selecting appropriate model complexity: An example of tracer inversion for thermal prediction in enhanced geothermal systems

Predicting the CO2 propagation in geological formations from sparsely available well data

Contact Info

Product

Resources

About