... introduces an innovative methodology, building on recent work on the "Hybrid Newton" method, that leverages machine learning for non-linear preconditioning to accelerate the history matching process, with a specific focus on incremental learning for near-well prediction. Our approach integrates machine...

...-logs. We employ a local Gauss-Newton (GN) optimizer that requires about half the number of simulations compared to equivalent gradient-based or serial optimizers. We approximate the sensitivity matrix by interpolating the resulting predictions with Duchon splines, i.e., radial basis functions, opposite...

... is essential to predict how much CO2 will be effectively trapped and how quickly this trapping occurs, and may necessitate fine grids in regions with steep gradients, while other areas with minor variations can be more coarsely resolved. Additionally, the expected gravity-driven convective mixing at field...

... variables from a high-dimensional space to a low-dimensional space, a linear transition network for predicting the evolution of state variables in the latent space, and a linear transition output for extending predictions to well output evolution over time. The E2CO framework is implemented using Proper...

... to use in history matching, optimization and uncertainty quantification workflows. We propose a novel machine learning architecture to auto-regressively predict state variables and well outputs of a reservoir for an ensemble of cases with varying well configurations. This new approach constructs a Graph...

... This paper presents a carbon dioxide (CO 2 ) storage application of a data-driven approach in forecasting reservoir performance (via the prediction of state variables and well outputs) of a compositional fluid system and multi-objective optimization under geological uncertainty using a deep-learning-based...

... DSI is an approach that conditions the model's predictions for either history matching (HM) or future projections through statistical models or machine learning, without repeatedly running the reservoir models (models are run only to create the prior. Sun and Durlofsky (2017 ) delineated...

... and empirical correlation predictions of the gas compressibility factor increases at higher pressures and temperatures, leading to uncertainties in reservoir performance and production modeling. This work includes experimental determination of Z-factors across pressure and temperature ranges, selection...

... Spatial-temporal modeling plays a critical role in understanding and predicting the behavior of complex systems, such as geological CO2 storage in our study. These systems involve interactions across both spatial domains and temporal scales, making accurate and efficient modeling crucial...

... part of this paper therefore describes a new phase labeling method that uses both the critical temperature and saturation pressure predictions from the ML models to generate accurate labels. Results are presented for CO 2 rich fluids. We show that this ML approach can result in accurate labeling...

... Abstract Integrated reservoir studies for performance prediction and decision-making processes are computationally expensive. In this paper, we develop a novel linearization approach to reduce the computational burden of intensive reservoir simulation execution. We achieve this by introducing...

... component. The mass accumulation term is then given by (6) ∑ β = A ,   G ,   I ,   H ϕ S β ρ β X β κ ,       κ = w ,   m ,   h ,   i reservoir characterization flow assurance flow in porous media ada cluster prediction fluid dynamics...

... are presented that illustrate homogeneous and drift-flux flow model differences for various well scenarios. thermal method reservoir simulation well model steam-assisted gravity drainage prediction well trajectory light blue circle enhanced recovery upstream oil & gas triangle homogeneous...

... in this work. machine learning reservoir simulation artificial intelligence deep learning fluid dynamics neural network upstream oil & gas prediction geomodel es-mda procedure application time step evaluation flow in porous media history matching parameterization surrogate model...

... successfully captured the physics in the high-fidelity model. The Bayesian-LSTM MCMC produces an accurate prediction with narrow uncertainties. The posterior prediction through the high-fidelity model ensures the robustness and precision of the workflow. This approach provides an efficient and high-quality...

... ( Rewieński and White 2003 ). Several people used this POD-TPWL in subsurface applications. See ( Cardoso and Durlofsky 2010 ), ( He and Durlofsky 2014 ), and ( Jin and Durlofsky 2018 ). Reservoir simulators are the most widely used tools in reservoir management and subsurface flow predictions. Accurate...

... features in the CNN structure which are crucial for achieving high accuracy of predictions for the DPDP model closures, and put forward the corresponding CNN architectures. Obtaining a suitable training dataset is challenging because i ) it requires a dedicated effort to map the fracture geometries; ii...

... is investigated. The results show that the PIML performs well, giving good results comparable to analytical solution. Further, we examined the potential of PIML approach in handling fluxes (sink and source terms). Our results demonstrate that the PIML fail to provide acceptable prediction for no-flow boundary...

... and convergence behavior. flow in porous media reservoir simulation case 2 modeling & simulation complex reservoir average ratio timestep fracture network locality equation of state fluid dynamics prediction iteration iterate newton method convergence behavior localization method...

... Abstract We present a novel reservoir simulator time-step selection approach which uses machine-learning (ML) techniques to analyze the mathematical and physical state of the system and predict time-step sizes which are large while still being efficient to solve, thus making the simulation...