YKYong Do Kim

Yong Do Kim

*This is an algorithmically generated author record based on publicly available information.

Yong Do Kim has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Diffusion-Model-Based Multifidelity Reservoir Simulation Surrogates for Upscaled Model Error Correction

    by Yong Do Kim, Satomi Suzuki, Xiao-Hui Wu, Gauthier Becker

    Summary: Upscaled models have become integral tools for the rapid prediction of dynamic subsurface flow within reservoir simulations. Despite their ability to produce simulation results more swiftly than fine-scaled models, upscaled models often suffer from inaccuracies in the output quantities of interest. In this work, we develop multifidelity surrogates based on diffusion models to correct errors in upscaled model simulation solutions. The multifidelity surrogates developed here accept pressure and saturation solutions from upscaled model simulation runs as conditional inputs and generate pressure and saturation at fine scale. To accommodate reservoir models of different sizes and enhance flexibility, we develop two different multifidelity surrogates: diffusion-model-based and latent-diffusion-model-based multifidelity surrogates (DM-MFS and LDM-MFS). The DM-MFS and LDM-MFS are tested using a 40×40 Gaussian model and a 60×220 SPE 10 top layer model, respectively. Both surrogates are trained using pressure and saturation solutions from a set of 80 fine-scaled and upscaled simulation runs that involve different injection well rates and production well bottomhole pressure (BHP) schedules. After the training, the multifidelity surrogates are shown to be capable of generating pressure and saturation at fine scale accurately for given upscaled model simulation pressure and saturation solutions.


  • Improving Reservoir Characterization and Prediction Via Machine Learning-Driven Integration of Subseismic Geologic Concepts, Geophysical Attributes, and Wells

    by D. Stojkovic, M. Shringarpure, Y. Kim, H. Alqassab, T. Rapstine, J. Fedele, S. Fullmer, R. Sain

    Summary: Reservoir dynamic performance is affected by subseismic (below the resolution of seismic imaging) geologic features such as submeter thin barriers and baffles, or high-permeability streaks. Here, we present an application of machine learning that has the potential to improve and significantly accelerate the workflow for generating reservoir realizations and scenarios with relevant subseismic features consistent with both well-log and 3D/4D seismic data and/or attributes. We leverage unsupervised machine learning methods, in particular, a modification of diffusion probabilistic modeling (DPM).


  • Optimization of Subsurface Flow Operations Using a Dynamic Proxy Strategy

    by Zhiwei Ma, Yong Do Kim, Oleg Volkov, Louis J. Durlofsky


  • Open Access

    Convolutional-Recurrent Neural Network Proxy for Robust Optimization and Closed-Loop Reservoir Management

    by Yong Do Kim, Louis J. Durlofsky

    Production optimization under geological uncertainty is computationally expensive, as a large number of well control schedules must be evaluated over multiple geological realizations. In this work, a convolutional-recurrent neural network (CNN-RNN) proxy model is developed to predict well-by-well oil and water rates, for given time-varying well bottom-hole pressure (BHP) schedules, for each realization in an ensemble. This capability enables the estimation of the objective function and nonlinear constraint values required for robust optimization. The proxy model represents an extension of a recently developed long short-term memory (LSTM) RNN proxy designed to predict well rates for a single geomodel. A CNN is introduced here to processes permeability realizations, and this provides the initial states for the RNN. The CNN-RNN proxy is trained using simulation results for 300 different sets of BHP schedules and permeability realizations. We demonstrate proxy accuracy for oil-water flow through multiple realizations of 3D multi-Gaussian permeability models. The proxy is then incorporated into a closed-loop reservoir management (CLRM) workflow, where it is used with particle swarm optimization and a filter-based method for nonlinear constraint satisfaction. History matching is achieved using an adjoint-gradient-based procedure. The proxy model is shown to perform well in this setting for five different (synthetic) `true' models. Improved net present value along with constraint satisfaction and uncertainty reduction are observed with CLRM. For the robust production optimization steps, the proxy provides O(100) runtime speedup over simulation-based optimization.

  • A Recurrent Neural Network–Based Proxy Model for Well-Control Optimization with Nonlinear Output Constraints

    by Yong Do Kim, Louis J. Durlofsky

    Summary: In well-control optimization problems, the goal is to determine the time-varying well settings that maximize an objective function, which is often the net present value (NPV). Various proxy models have been developed to predict NPV for a set of inputs such as time-varying well bottomhole pressures (BHPs). However, when nonlinear output constraints (e.g., maximum well/field water production rate or minimum well/field oil rate) are specified, the problem is more challenging because well rates as a function of time are required. In this work, we develop a recurrent neural network (RNN)–based proxy model to treat constrained production optimization problems. The network developed here accepts sequences of BHPs as inputs and predicts sequences of oil and water rates for each well. A long-short-term memory (LSTM) cell, which is capable of learning long-term dependencies, is used. The RNN is trained using well-rate results from 256 full-order simulation runs that involve different injection and production-well BHP schedules. After detailed validation against full-order simulation results, the RNN-based proxy is used for 2D and 3D production optimization problems. Optimizations are performed using a particle swarm optimization (PSO) algorithm with a filter-based nonlinear-constraint treatment. The trained proxy is extremely fast, although optimizations that apply the RNN-based proxy at all iterations are found to be suboptimal relative to full simulation-based (standard) optimization. Through use of a few additional simulation-based PSO iterations after proxy-based optimization, we achieve NPVs comparable with those from simulation-based optimization but with speedups of 10 or more (relative to performing five simulation-based optimization runs). It is important to note that because the RNN-based proxy provides full well-rate time sequences, optimization constraint types or limits, as well as economic parameters, can be varied without retraining. NOTE: This paper is published as part of the 2021 Reservoir Simulation Conference Special Issue.