Simulating the Behavior of Reservoirs with Convolutional and Recurrent Neural Networks

Abstract

Summary Recent experience in applying recurrent neural networks (RNNs) to interpreting permanent downhole gauge records has highlighted the potential utility of machine learning algorithms to learn reservoir behavior from data. The power of the RNN resides in its ability to retain information in a form of memory of previous patterns and information contained in the previous behavior of phenomena being modeled. This memory plays a role of informing the decision at the present time by using what happened in the past. This property suggests the RNN as a suitable choice to model sequences of reservoir information, even when the reservoir modeler is faced with incomplete knowledge of the underlying physical system. Convolutional neural networks (CNNs) are another variant of the machine learning algorithm that have shown promise in sequence modeling domains, such as audio synthesis and machine translation. In this study, RNNs and CNNs were applied to tasks that traditionally would be modeled by a reservoir simulator. This was achieved by formulating the relationship between physical quantities of interest from subsurface reservoirs as a sequence mapping problem. In addition, the performance of a CNN layer as compared with an RNN was evaluated systematically to investigate their capabilities in a variety of tasks of interest to the reservoir engineer. Preliminary results suggest that CNNs, with specific design modifications, are as capable as RNNs in modeling sequences of information, and as reliable when making inferences to cases that have not been seen by the algorithm during training. Design details and reasons pertaining to the way these two seemingly different architectures process information and handle memory are also discussed.

Application of nature-inspired algorithms and artificial neural network in waterflooding well control optimization

“…There are also other interesting literatures (Nait Amar et al 2018Navrátil et al 2019;Alakeely and Horne 2020;Ng et al 2021) that discuss and present the use of ML methods in the establishment of proxies of numerical models in petroleum domain, especially for reservoir engineering. Regarding this, there is a riveting insight being provided by Nait Amar et al (2018) about the modeling of proxies, which is the difference between static and dynamic proxy models.…”

Computationally Efficient Multiscale Neural Networks Applied to Fluid Flow in Complex 3D Porous Media

“…This is accomplished with learnable kernels which can be identified with small windows that have the same dimensionality as the input data (i.e., 2D for images, 3D for volumes). They have been successfully applied in many tasks regarding earth science disciplines (Alakeely and Horne 2020;Jo et al 2020;Pan et al 2020), and particularly in the field of digital rocks (Guiltinan et al 2020a;Mosser et al 2017a, b;Chung et al 2020;Bihani et al 2021;Santos et al 2018b). These data-driven models have also been useful for solving flow (Santos et al 2020b;Wang et al 2021a, b;Alqahtani et al 2021), successfully modeling the relationship between 3D microstructure and flow response much more accurately than empirical formulas that depend only on averaged properties.…”

Development of Deep Transformer-Based Models for Long-Term Prediction of Transient Production of Oil Wells

“…The authors analyzed different feature combinations and highlighted the optimal set providing the maximum quality of predictions. In the study [5], the authors compared the predictive capabilities of convolutional (CNN) and recurrent (RNN) neural networks to simulate the reservoir behavior. The authors demonstrated that both networks could capture the interference between multiple wells at the same reservoir.…”