A Novel A.I Enhanced Reservoir Characterization with a Combined Mixture of Experts -- NVIDIA Modulus based Physics Informed Neural Operator Forward Model

Read original: arXiv:2404.14447 - Published 4/24/2024 by Clement Etienam, Yang Juntao, Issam Said, Oleg Ovcharenko, Kaustubh Tangsali, Pavel Dimitrov, Ken Hester

🧠

Overview

Developed an advanced workflow for reservoir characterization to address challenges in reservoir history matching
Integrated a Physics Informed Neural Operator (PINO) as a forward model within a Cluster Classify Regress (CCR) framework
Enhanced the process with an adaptive Regularized Ensemble Kalman Inversion (aREKI) for rapid uncertainty quantification in reservoir history matching
Parameterized unknown permeability and porosity fields, capturing non-Gaussian posterior measures using techniques like a variational convolution autoencoder and the CCR
CCR synergizes with the PINO surrogate to accurately simulate the nonlinear dynamics of Peaceman well equations
The integrated PINO-Res-Sim model outputs crucial reservoir parameters including pressures, saturations, and production rates

Plain English Explanation

This research paper presents an advanced workflow for characterizing oil and gas reservoirs. Reservoir characterization is the process of understanding the properties and behavior of an underground rock formation that contains oil and gas. One of the key challenges in this field is "history matching" - accurately predicting how a reservoir will perform over time based on past data.

The researchers tackled this challenge by developing a novel approach that integrates two powerful machine learning techniques. The first is a Physics Informed Neural Operator (PINO), which is a type of artificial neural network that can accurately simulate the complex physical processes happening in a reservoir. The second is a Cluster Classify Regress (CCR) framework, which allows for flexibility in applying different machine learning algorithms to different stages of the workflow.

This integrated approach, called PINO-Res-Sim, is further enhanced by an adaptive Regularized Ensemble Kalman Inversion (aREKI) method. This helps the system rapidly quantify the uncertainty in its predictions, which is crucial for decision-making in the oil and gas industry.

The key innovation in this research is the way it parameterizes the unknown properties of the reservoir, such as permeability and porosity, using advanced techniques like variational autoencoders. This allows the system to capture the complex, non-Gaussian patterns in the data, leading to more accurate predictions.

Technical Explanation

The researchers developed an integrated workflow that combines a Physics Informed Neural Operator (PINO) as a forward model with a sophisticated Cluster Classify Regress (CCR) framework. The PINO serves as a surrogate model that can accurately simulate the nonlinear dynamics of Peaceman well equations, which describe the fluid flow in a reservoir.

The CCR approach allows for flexibility in applying different machine learning algorithms across its various stages, such as parameterizing the unknown permeability and porosity fields. Techniques like a variational convolution autoencoder and the CCR itself are used to capture the non-Gaussian posterior measures of these unknown fields.

The CCR synergizes with the PINO surrogate, with the PINO's updates driven by a loss function derived from supervised data, initial conditions, and residuals of the governing black oil partial differential equations. The integrated PINO-Res-Sim model outputs crucial reservoir parameters like pressures, saturations, and production rates for oil, water, and gas.

The researchers validated their methodology through controlled experiments on synthetic reservoirs and the Norne field, demonstrating remarkable accuracy compared to traditional simulators. Additionally, the PINO-Res-Sim model in the aREKI workflow was able to recover the unknown fields with a computational speedup of 100 to 6000 times faster than conventional methods. The training of the PINO-Res-Sim model was conducted efficiently on an NVIDIA H100 GPU, making it compatible with ensemble-based methods for complex computational tasks.

Critical Analysis

The researchers have presented a comprehensive and innovative workflow for reservoir characterization, addressing key challenges in the field of history matching. The integration of the PINO as a forward model and the CCR framework, along with the adaptive Regularized Ensemble Kalman Inversion (aREKI), represents a significant advancement in the state of the art.

One potential limitation of the research is the reliance on synthetic and controlled experiments, as well as the Norne field dataset. While these provide a solid validation of the methodology, it would be valuable to see the workflow applied to a wider range of real-world reservoir scenarios to further assess its robustness and generalizability.

Additionally, the paper does not delve deeply into the specific machine learning algorithms used within the CCR framework, nor does it provide detailed comparisons to other state-of-the-art history matching techniques. Further exploration of the trade-offs and nuances of the different components of the workflow could strengthen the overall contribution.

Nevertheless, the researchers have demonstrated the potential of their integrated approach to significantly accelerate the reservoir characterization process while maintaining high accuracy. The application of techniques like variational autoencoders and the synergistic coupling of the PINO and CCR models are particularly noteworthy and could inspire further advancements in this important field.

Conclusion

The research presented in this paper represents a significant step forward in the field of reservoir characterization, addressing longstanding challenges in reservoir history matching. By integrating a Physics Informed Neural Operator (PINO) as a forward model within a sophisticated Cluster Classify Regress (CCR) framework, the researchers have developed a powerful workflow that can accurately simulate the complex nonlinear dynamics of oil and gas reservoirs.

The integration of advanced techniques like variational autoencoders and the adaptive Regularized Ensemble Kalman Inversion (aREKI) has enabled the system to capture non-Gaussian posterior measures of unknown reservoir properties, leading to significantly more accurate predictions. Importantly, the researchers have demonstrated the computational efficiency of their approach, with the PINO-Res-Sim model achieving speedups of up to 6000 times compared to conventional methods.

This research has important implications for the oil and gas industry, as it has the potential to revolutionize the way reservoir characterization and history matching are conducted. By providing a more accurate and rapid understanding of reservoir behavior, industry professionals can make better-informed decisions, leading to more efficient and sustainable resource extraction. Additionally, the flexibility and adaptability of the CCR framework suggest that this workflow could be extended to other complex physical systems beyond the scope of reservoir engineering.

This summary was produced with help from an AI and may contain inaccuracies - check out the links to read the original source documents!

Follow @aimodelsfyi on 𝕏 →

Related Papers

🧠

A Novel A.I Enhanced Reservoir Characterization with a Combined Mixture of Experts -- NVIDIA Modulus based Physics Informed Neural Operator Forward Model

Clement Etienam, Yang Juntao, Issam Said, Oleg Ovcharenko, Kaustubh Tangsali, Pavel Dimitrov, Ken Hester

We have developed an advanced workflow for reservoir characterization, effectively addressing the challenges of reservoir history matching through a novel approach. This method integrates a Physics Informed Neural Operator (PINO) as a forward model within a sophisticated Cluster Classify Regress (CCR) framework. The process is enhanced by an adaptive Regularized Ensemble Kalman Inversion (aREKI), optimized for rapid uncertainty quantification in reservoir history matching. This innovative workflow parameterizes unknown permeability and porosity fields, capturing non-Gaussian posterior measures with techniques such as a variational convolution autoencoder and the CCR. Serving as exotic priors and a supervised model, the CCR synergizes with the PINO surrogate to accurately simulate the nonlinear dynamics of Peaceman well equations. The CCR approach allows for flexibility in applying distinct machine learning algorithms across its stages. Updates to the PINO reservoir surrogate are driven by a loss function derived from supervised data, initial conditions, and residuals of governing black oil PDEs. Our integrated model, termed PINO-Res-Sim, outputs crucial parameters including pressures, saturations, and production rates for oil, water, and gas. Validated against traditional simulators through controlled experiments on synthetic reservoirs and the Norne field, the methodology showed remarkable accuracy. Additionally, the PINO-Res-Sim in the aREKI workflow efficiently recovered unknown fields with a computational speedup of 100 to 6000 times faster than conventional methods. The learning phase for PINO-Res-Sim, conducted on an NVIDIA H100, was impressively efficient, compatible with ensemble-based methods for complex computational tasks.

4/24/2024

Reservoir History Matching of the Norne field with generative exotic priors and a coupled Mixture of Experts -- Physics Informed Neural Operator Forward Model

Clement Etienam, Yang Juntao, Oleg Ovcharenko, Issam Said

We developed a novel reservoir characterization workflow that addresses reservoir history matching by coupling a physics-informed neural operator (PINO) forward model with a mixture of experts' approach, termed cluster classify regress (CCR). The inverse modelling is achieved via an adaptive Regularized Ensemble Kalman inversion (aREKI) method, ideal for rapid inverse uncertainty quantification during history matching. We parametrize unknown permeability and porosity fields for non-Gaussian posterior measures using a variational convolution autoencoder and a denoising diffusion implicit model (DDIM) exotic priors. The CCR works as a supervised model with the PINO surrogate to replicate nonlinear Peaceman well equations. The CCR's flexibility allows any independent machine-learning algorithm for each stage. The PINO reservoir surrogate's loss function is derived from supervised data loss and losses from the initial conditions and residual of the governing black oil PDE. The PINO-CCR surrogate outputs pressure, water, and gas saturations, along with oil, water, and gas production rates. The methodology was compared to a standard numerical black oil simulator for a waterflooding case on the Norne field, showing similar outputs. This PINO-CCR surrogate was then used in the aREKI history matching workflow, successfully recovering the unknown permeability, porosity and fault multiplier, with simulations up to 6000 times faster than conventional methods. Training the PINO-CCR surrogate on an NVIDIA H100 with 80G memory takes about 5 hours for 100 samples of the Norne field. This workflow is suitable for ensemble-based approaches, where posterior density sampling, given an expensive likelihood evaluation, is desirable for uncertainty quantification.

6/4/2024

Neural Operator-Based Proxy for Reservoir Simulations Considering Varying Well Settings, Locations, and Permeability Fields

Daniel Badawi, Eduardo Gildin

Simulating Darcy flows in porous media is fundamental to understand the future flow behavior of fluids in hydrocarbon and carbon storage reservoirs. Geological models of reservoirs are often associated with high uncertainly leading to many numerical simulations for history matching and production optimization. Machine learning models trained with simulation data can provide a faster alternative to traditional simulators. In this paper we present a single Fourier Neural Operator (FNO) surrogate that outperforms traditional reservoir simulators by the ability to predict pressures and saturations on varying permeability fields, well locations, well controls, and number of wells. The maximum-mean relative error of 95% of pressure and saturation predictions is less than 5%. This is achieved by employing a simple yet very effective data augmentation technique that reduces the dataset size by 75% and reduces overfitting. Also, constructing the input tensor in a binary fashion enables predictions on unseen well locations, well controls, and number of wells. Such model can accelerate history matching and reservoir characterization procedures by several orders of magnitude. The ability to predict on new well locations, well controls, and number of wells enables highly efficient reservoir management and optimization.

7/16/2024

Residual resampling-based physics-informed neural network for neutron diffusion equations

Heng Zhang, Yun-Ling He, Dong Liu, Qin Hang, He-Min Yao, Di Xiang

The neutron diffusion equation plays a pivotal role in the analysis of nuclear reactors. Nevertheless, employing the Physics-Informed Neural Network (PINN) method for its solution entails certain limitations. Traditional PINN approaches often utilize fully connected network (FCN) architecture, which is susceptible to overfitting, training instability, and gradient vanishing issues as the network depth increases. These challenges result in accuracy bottlenecks in the solution. In response to these issues, the Residual-based Resample Physics-Informed Neural Network(R2-PINN) is proposed, which proposes an improved PINN architecture that replaces the FCN with a Convolutional Neural Network with a shortcut(S-CNN), incorporating skip connections to facilitate gradient propagation between network layers. Additionally, the incorporation of the Residual Adaptive Resampling (RAR) mechanism dynamically increases sampling points, enhancing the spatial representation capabilities and overall predictive accuracy of the model. The experimental results illustrate that our approach significantly improves the model's convergence capability, achieving high-precision predictions of physical fields. In comparison to traditional FCN-based PINN methods, R2-PINN effectively overcomes the limitations inherent in current methods, providing more accurate and robust solutions for neutron diffusion equations.

7/17/2024