Deep net simulator (DNS): a new insight into reservoir simulation

2020 ◽  
Vol 60 (1) ◽  
pp. 124
Author(s):  
Shahdad Ghassemzadeh ◽  
Maria Gonzalez Perdomo ◽  
Manouchehr Haghighi ◽  
Ehsan Abbasnejad
Keyword(s):  
Deep Learning ◽  
Reservoir Simulation ◽  
Vital Role ◽  
Data Driven ◽  
Percentage Error ◽  
Gas Reservoirs ◽  
Gas Reservoir ◽  
Flow Equations ◽  
Reservoir Performance ◽  
Reservoir Simulations

Reservoir simulation plays a vital role as a diagnostics tool to better understand and predict a reservoir’s behaviour. The primary purpose of running a reservoir simulation is to replicate reservoir performance under different production conditions; therefore, the development of a reliable and fast dynamic reservoir model is a priority for the industry. In each simulation, the reservoir is divided into millions of cells, with fluid and rock attributes assigned to each cell. Based on these attributes, flow equations are solved through numerical methods, resulting in an excessively long processing time. Given the recent progress in machine learning methods, this study aimed to further investigate the possibility of using deep learning in reservoir simulations. Throughout this paper, we used deep learning to build a data-driven simulator for both 1D oil and 2D gas reservoirs. In this approach, instead of solving fluid flow equations directly, a data-driven model instantly predicts the reservoir pressure using the same input data of a numerical simulator. Datasets were generated using a physics-based simulator. It was found that for the training and validation sets, the mean absolute percentage error (MAPE) was less than 15.1% and the correlation coefficient, R2, was more than 0.84 for the 1D oil reservoirs, while for the 2D gas reservoir MAPE < 0.84% and R2 ≈1. Furthermore, the sensitivity analysis results confirmed that the proposed approach has promising potential (MAPE < 5%, R2 > 0.9). The results agreed that the deep learning based, data-driven model is reasonably accurate and trustworthy when compared with physics-derived models.

Technology Focus: Reservoir Simulation (July 2021)

2021 ◽  
Vol 73 (07) ◽  
pp. 43-43
Author(s):  
Mark Burgoyne
Keyword(s):  
Machine Learning ◽  
Numerical Simulation ◽  
Reservoir Simulation ◽  
Mechanistic Model ◽  
Deep Understanding ◽  
Decision Makers ◽  
Data Driven ◽  
Gas Reservoirs ◽  
New Ways Of Working ◽  
Decline Curve

In reviewing the long list of papers this year, it has become apparent to me that the hot topic in reservoir simulation these days is the application of data analytics or machine learning to numerical simulation and with it quite often the promise of data-driven work flows—code for needing to think about the physics less. Data-driven work flows have their place, especially when we have a lot of data and the system is very complex. I’m thinking shales especially, but seeing it being applied to more conventional reservoirs gave me a moment of pause. I can’t help but think that, in terms of the hype cycle as related to the application of machine learning to numerical simulation, we may be approaching the peak of inflated expectations. I say “approaching,” because many companies appear to be dipping their toes in the water, perhaps because they think they should, but few are truly committing to it. Many vocal champions of the approach exist, but most decision-makers just don’t understand it yet. If we cannot explain how something works simply, then thoughtful leaders will tend not to trust it. Whether it be numerical-simulation findings or self-organizing neural networks, the need will always exist for a deep understanding and clarity of explanation of both the discipline and method used. To decision-makers, it will be an attractive concept, but they will generally ask to validate against more traditional methods. I look forward to a future when we are through the trough of disillusionment and start climbing the slope of enlightenment to a new level of productivity. I suspect, though, that it will take at least another 5 years, as our current crop of knowledgeable evangelists become decision-makers themselves and can put in place work flows and teams to leverage the approach appropriately for their problems, intelligently leveraging their years of hard-won experience. I will lay a wager with you, though, that when that time comes, those new ways of working more efficiently will rely just as much if not more upon a deep understanding of reservoir engineering as our current methods. I hope you enjoy these papers, which include examples of both the new approach as well as tried-and-true approaches. Recommended additional reading at OnePetro: www.onepetro.org. SPE 202436 - Fast Modeling of Gas Reservoirs Using Proper Orthogonal Decomposition/Radial Basis Function (POD/RBF) Nonintrusive Reduced-Order Modeling by Jemimah-Sandra Samuel, Imperial College London, et al. IPTC 21417 - A New Methodology for Calculating Wellbore Shut-In Pressure in Numerical Reservoir Simulations by Babatope Kayode, Saudi Aramco, et al. SPE 201658 - Mechanistic Model Validation of Decline Curve Analysis for Unconventional Reservoirs by Mikhail Gorditsa, Texas A&M University, et al.

Physics-Aware Deep-Learning-Based Proxy Reservoir Simulation Model Equipped with State and Well Output Prediction

2021 ◽  
Author(s):  
Emilio J. R. Coutinho ◽  
Marcelo J. Aqua and Eduardo Gildin
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Deep Learning ◽  
Reservoir Simulation ◽  
Data Driven ◽  
Training Procedure ◽  
State Variables ◽  
Input Output ◽  
Well Control ◽  
Reduced Order

Abstract Physics-aware machine learning (ML) techniques have been used to endow data-driven proxy models with features closely related to the ones encountered in nature. Examples span from material balance and conservation laws. Physics-based and data-driven reduced-order models or a combination thereof (hybrid-based models) can lead to fast, reliable, and interpretable simulations used in many reservoir management workflows. We built on a recently developed deep-learning-based reduced-order modeling framework by adding a new step related to information of the input-output behavior (e.g., well rates) of the reservoir and not only the states (e.g., pressure and saturation) matching. A Combination of data-driven model reduction strategies and machine learning (deep- neural networks – NN) will be used here to achieve state and input-output matching simultaneously. In Jin, Liu and Durlofsky (2020), the authors use a NN architecture where it is possible to predict the state variables evolution after training an autoencoder coupled with a control system approach (Embed to Control - E2C) and adding some physical components (Loss functions) to the neural network training procedure. In this paper, we extend this idea by adding the simulation model output, e.g., well bottom-hole pressure and well flowrates, as data to be used in the training procedure. Additionally, we added a new neural network to the E2C transition model to handle the connections between state variables and model outputs. By doing this, it is possible to estimate the evolution in time of both the state variables as well as the output variables simultaneously. The method proposed provides a fast and reliable proxy for the simulation output, which can be applied to a well-control optimization problem. Such a non-intrusive method, like data-driven models, does not need to have access to reservoir simulation internal structure. So it can be easily applied to commercial reservoir simulations. We view this as an analogous step to system identification whereby mappings related to state dynamics, inputs (controls), and measurements (output) are obtained. The proposed method is applied to an oil-water model with heterogeneous permeability, 4 injectors, and 5 producer wells. We used 300 sampled well control sets to train the autoencoder and another set to validate the obtained autoencoder parameters.

scholarly journals Application of Machine Learning Method of Data-Driven Deep Learning Model to Predict Well Production Rate in the Shale Gas Reservoirs

Energies ◽  
2021 ◽  
Vol 14 (12) ◽  
pp. 3629
Author(s):  
Dongkwon Han ◽  
Sunil Kwon
Keyword(s):  
Deep Learning ◽  
Hydraulic Fracturing ◽  
Physical Phenomenon ◽  
Principal Component ◽  
Learning Model ◽  
Gas Production ◽  
Data Driven ◽  
Percentage Error ◽  
Proxy Model ◽  
Deep Learning Model

Reservoir modeling to predict shale reservoir productivity is considerably uncertain and time consuming. Since we need to simulate the physical phenomenon of multi-stage hydraulic fracturing. To overcome these limitations, this paper presents an alternative proxy model based on data-driven deep learning model. Furthermore, this study not only proposes the development process of a proxy model, but also verifies using field data for 1239 horizontal wells from the Montney shale formation in Alberta, Canada. A deep neural network (DNN) based on multi-layer perceptron was applied to predict the cumulative gas production as the dependent variable. The independent variable is largely divided into four types: well information, completion and hydraulic fracturing and production data. It was found that the prediction performance was better when using a principal component with a cumulative contribution of 85% using principal component analysis that extracts important information from multivariate data, and when predicting with a DNN model using 6 variables calculated through variable importance analysis. Hence, to develop a reliable deep learning model, sensitivity analysis of hyperparameters was performed to determine one-hot encoding, dropout, activation function, learning rate, hidden layer number and neuron number. As a result, the best prediction of the mean absolute percentage error of the cumulative gas production improved to at least 0.2% and up to 9.1%. The novel approach of this study can also be applied to other shale formations. Furthermore, a useful guide for economic analysis and future development plans of nearby reservoirs.

scholarly journals Physics-Aware Deep-Learning-Based Proxy Reservoir Simulation Model Equipped With State and Well Output Prediction

2021 ◽  
Vol 7 ◽  
Author(s):  
Emilio Jose Rocha Coutinho ◽  
Marcelo Dall’Aqua ◽  
Eduardo Gildin
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Simulation Model ◽  
Reservoir Simulation ◽  
The State ◽  
Data Driven ◽  
Training Procedure ◽  
State Variables ◽  
Input Output ◽  
Modeling Framework

Data-driven methods have been revolutionizing the way physicists and engineers handle complex and challenging problems even when the physics is not fully understood. However, these models very often lack interpretability. Physics-aware machine learning (ML) techniques have been used to endow proxy models with features closely related to the ones encountered in nature; examples span from material balance to conservation laws. In this study, we proposed a hybrid-based approach that incorporates physical constraints (physics-based) and yet is driven by input/output data (data-driven), leading to fast, reliable, and interpretable reservoir simulation models. To this end, we built on a recently developed deep learning–based reduced-order modeling framework by adding a new step related to information on the input–output behavior (e.g., well rates) of the reservoir and not only the states (e.g., pressure and saturation) matching. A deep-neural network (DNN) architecture is used to predict the state variables evolution after training an autoencoder coupled with a control system approach (Embed to Control—E2C) along with the addition of some physical components (loss functions) to the neural network training procedure. Here, we extend this idea by adding the simulation model output, for example, well bottom-hole pressure and well flow rates, as data to be used in the training procedure. Additionally, we introduce a new architecture to the E2C transition model by adding a new neural network component to handle the connections between state variables and model outputs. By doing this, it is possible to estimate the evolution in time of both the state and output variables simultaneously. Such a non-intrusive data-driven method does not need to have access to the reservoir simulation internal structure, so it can be easily applied to commercial reservoir simulators. The proposed method is applied to an oil–water model with heterogeneous permeability, including four injectors and five producer wells. We used 300 sampled well control sets to train the autoencoder and another set to validate the obtained autoencoder parameters. We show our proxy’s accuracy and robustness by running two different neural network architectures (propositions 2 and 3), and we compare our results with the original E2C framework developed for reservoir simulation.

A Dual-Grid, Implicit, and Sequentially Coupled Geomechanics-and-Composition Model for Fractured Reservoir Simulation

SPE Journal ◽  
2020 ◽  
Vol 25 (04) ◽  
pp. 2098-2118 ◽  
Author(s):  
Xin Li ◽  
Xiang Li ◽  
Dongxiao Zhang ◽  
Rongze Yu
Keyword(s):  
Reservoir Simulation ◽  
Large Scale ◽  
Flow Model ◽  
Fractured Reservoirs ◽  
Stress Sensitivity ◽  
Fractured Reservoir ◽  
Flow Equations ◽  
Reservoir Simulations ◽  
Solid Deformation ◽  
Coupled Geomechanics

Summary In the development of fractured reservoirs, geomechanics is crucial because of the stress sensitivity of fractures. However, the complexities of both fracture geometry and fracture mechanics make it challenging to consider geomechanical effects thoroughly and efficiently in reservoir simulations. In this work, we present a coupled geomechanics and multiphase-multicomponent flow model for fractured reservoir simulations. It models the solid deformation using a poroelastic equation, and the solid deformation effects are incorporated into the flow model rigorously. The noticeable features of the proposed model are it uses a pseudocontinuum equivalence method to model the mechanical characteristics of fractures; the coupled geomechanics and flow equations are split and sequentially solved using the fixed-stress splitting strategy, which retains implicitness and exhibits good stability; and it simulates geomechanics and compositional flow, respectively, using a dual-grid system (i.e., the geomechanics grid and the reservoir-flow grid). Because of the separation of the geomechanics part and the flow part, the model is not difficult to implement based on an existing reservoir simulator. We validated the accuracy and stability of this model through several benchmark cases and highlighted the practicability with two large-scale cases. The case studies demonstrate that this model is capable of considering the key effects of geomechanics in fractured-reservoir simulation, including matrix compaction, fracture normal deformation, and shear dilation, as well as hydrocarbon phase behavior. The flexibility, efficiency, and comprehensiveness of this model enable a more realistic geocoupled reservoir simulation.

A data-driven reservoir simulation for natural gas reservoirs

2021 ◽  
Author(s):  
Shahdad Ghassemzadeh ◽  
Maria Gonzalez Perdomo ◽  
Manouchehr Haghighi ◽  
Ehsan Abbasnejad
Keyword(s):  
Natural Gas ◽  
Reservoir Simulation ◽  
Data Driven ◽  
Gas Reservoirs ◽  
Natural Gas Reservoirs

Real Time Gender and Age Prediction using Deep Learning Techniques

2020 ◽  
Vol 9 (6) ◽  
pp. 797-801
Keyword(s):  
Deep Learning ◽  
Face Recognition ◽  
Real Time ◽  
Vital Role ◽  
Learning Techniques ◽  
Gender And Age ◽  
The Face ◽  
And Gender ◽  
Gender Detection ◽  
Age Prediction

Face recognition plays a vital role in security purpose. In recent years, the researchers have focused on the pose illumination, face recognition, etc,. The traditional methods of face recognition focus on Open CV’s fisher faces which results in analyzing the face expressions and attributes. Deep learning method used in this proposed system is Convolutional Neural Network (CNN). Proposed work includes the following modules: [1] Face Detection [2] Gender Recognition [3] Age Prediction. Thus the results obtained from this work prove that real time age and gender detection using CNN provides better accuracy results compared to other existing approaches.

Deep Learning based Thermal Crack Detection on Structural Concrete Exposed to Elevated Temperature

2021 ◽  
pp. 136943322098663
Author(s):  
Diana Andrushia A ◽  
Anand N ◽  
Eva Lubloy ◽  
Prince Arulraj G
Keyword(s):  
Deep Learning ◽  
Crack Detection ◽  
Detection Method ◽  
Concrete Structures ◽  
Vital Role ◽  
Subjective Knowledge ◽  
Thermal Cracks ◽  
Reinforced Cement ◽  
Novel Method ◽  
Human Inspection

Health monitoring of concrete including, detecting defects such as cracking, spalling on fire affected concrete structures plays a vital role in the maintenance of reinforced cement concrete structures. However, this process mostly uses human inspection and relies on subjective knowledge of the inspectors. To overcome this limitation, a deep learning based automatic crack detection method is proposed. Deep learning is a vibrant strategy under computer vision field. The proposed method consists of U-Net architecture with an encoder and decoder framework. It performs pixel wise classification to detect the thermal cracks accurately. Binary Cross Entropy (BCA) based loss function is selected as the evaluation function. Trained U-Net is capable of detecting major thermal cracks and minor thermal cracks under various heating durations. The proposed, U-Net crack detection is a novel method which can be used to detect the thermal cracks developed on fire exposed concrete structures. The proposed method is compared with the other state-of-the-art methods and found to be accurate with 78.12% Intersection over Union (IoU).

Sign in / Sign up

Export Citation Format

Share Document