InterPore2024

Session

MS15

13 May 2024, 11:25

67. Machine learning accelerated molecular simulation: Implications for oil and gas problems

Dr Jie Liu (King Abdullah University of Science and Technology)

13/05/2024, 11:25

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

In the last few decades, deep learning (DL) has afforded solutions to macroscopic problems in petroleum engineering, but mechanistic problems at the microscale have not benefited from it. Mechanism studies have been the strong demands for the emerging projects, such as the gas storage and hydrate production, and for some problems encountered in the storage process, which are common found as...

187. CO2 Leakage Detection using Optimized Deep Learning

Dr Xupeng He (Saudi Aramco)

13/05/2024, 11:40

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

1. OBJECTIVE/SCOPE
   Geologic CO2 sequestration (GCS) has been considered as a promising engineering measure to reduce global greenhouse emission. Real-time monitoring of CO2 leakage is an essential aspect of large-scale GCS deployment. This work introduces a deep-learning-based algorithm using a hybrid neural network for detecting CO2 leakage based on bottom-hole pressure measurements....

592. Prediction of Upscaled Permeability of Digital Rock Cores Using Machine Learning Techniques

Yaotian Guo

13/05/2024, 11:55

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

A workflow has been proposed to directly predict the upscaled absolute permeability of a rock core from CT images whose resolution is insufficient to directly calculate the pore-scale permeability. The workflow employed the deep learning technique with the raw CT image data of rocks and their corresponding permeability values, which were obtained through high-resolution flow simulation on...

82. Large Scale Efficient 3D Domain Transfer for Digital Images of Porous Materials using Pseudo-3D Architectures

Kunning Tang (UNSW)

13/05/2024, 12:10

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Characterisation of the internal 3-dimensional (3D) structure of complex porous materials has been revolutionised with deep-learned image processing and segmentation, promising second-scale scan times with hour-scale quality, and beyond-human multi-label segmentation accuracy at a fraction of the time. However, these claims are currently only true for single-sample, single-domain cases using...

205. Deep learning-assisted technology transition in natural hydrogen development

tao zhang (King Abdullah University of Science and Technology)

13/05/2024, 13:25

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Although the global energy sector is shifting from the fossil-based energy systems to the renewable energy resources, the conventional energy development techniques has received increasing attentions with the mature development and the recharge by AI. With the help of AI techniques, drawing lessons from thousands of years of traditional energy development in the technology transition into the...

313. Multiparameter Inversion of Reservoirs Based on Deep Learning

Qian Wang

13/05/2024, 13:40

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Reservoir parameter inversion is an important technique in oil and gas exploration and development that can estimate the reservoir physical properties, such as skin factor and permeability, using observed data, such as well test data and production data. In this paper, we propose a physical accelerated neural network with multiple residual blocks (PRNN-Acc) for multiple parameter inversion of...

450. A Vision Transformer for Size-Agnostic Modelling of Two-Phase Drainage in Complex Porous Media Considering Wettability, Interfacial Tension, and Resolution

Ping Wu (School of Science, Southwest Petroleum University, Chengdu 610500, China)

13/05/2024, 13:55

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

The estimation of pore-scale multiphase flow fields in complex geometries using deep learning has proven challenging. This is partly because researchers have historically focused on model architecture and data quality, while the volume and variety of data may have been inadequate to capture the intricacies of multiphase flow. In this work, we introduce a novel deep learning methodology to...

674. A neural network model with physics constraints for simulating CO2 storage in deep saline aquifers during and after injection

Mengjie Zhao

13/05/2024, 14:10

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

The exploration of CO2 capture and storage has become a crucial element in strategies aimed at mitigating climate change, where deep saline aquifers are of particular interest due to their extensive storage capacity and widespread availability. The complexities involved in effectively monitoring and simulating CO2 behavior within these geological formations present significant challenges. To...

880. 2D to 3D deep learning reconstruction of CO2 electroconversion Gas Diffusion Electrode : a validation study

Georgy Borisochev (CHLOE)

13/05/2024, 14:25

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Direct 3D imaging of natural or synthetic porosity below ~1μm in diameter often requires the application of Focused Ion Beam Scanning Electron Microscopy (FIB-SEM). This technique has several limitations: high cost and time demands, instrument availability, complex sample preparation and low field-of-view (FoV), restricting its suitability for operational industrial studies. SliceGAN, a...

98. Deep Learning enhanced multiscale rock typing for digital core modeling

Denis Orlov (Skolkovo Institute of Science and Technology)

13/05/2024, 14:40

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Currently, there is rapid development in the approaches for constructing and utilizing digital cores. Digital Rock Physics (DRP) methods allow for quick and non-destructive acquisition of rock properties. The process of digital rock physics involves two primary stages: model construction and simulation of physical processes on the created models.
For heterogeneous reservoir rocks, the usage...

59. Application of Diffusion Models to Generate Multiphase Fluid Pore-Scale Images

Linqi Zhu (Imperial College London/ Chinese Academy of Sciences)

13/05/2024, 17:00

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Generative Adversarial Networks (GANs) have been a typical example of how machine learning has been successfully applied, using three-dimensional images as training datasets, to generate realizations of the pore space, as well as to produce super-resolution images. We further this work with a new generative model: diffusion models (DMs), to generate images of both the pore space and two fluid...

64. Predicting ultimate hydrogen production and residual volume during cyclic underground hydrogen storage in porous media using machine learning

Raymond Mushabe (University of Bergen)

13/05/2024, 17:15

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

In the context of climate change mitigation, underground/subsurface hydrogen storage (UHS) is regarded as a solution that could help tackle the imbalance in renewable energy supply. Excess energy can be stored as molecular hydrogen (H2) and re-used when it is needed. To enable large-scale storage in underground geologic formations, reservoir simulation of cyclic loading scenarios will be used...

259. Efficient 3D Digital Rock Detail Reconstruction and Quality Enhancement with Super-Resolution Transformer

Mr Zhihao Xing (China University of Petroleum (East China))

13/05/2024, 17:30

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

High-quality digital rocks are essential for high-precision pore-scale modeling. However, limited by the imaging hardware, meeting the requirements of high resolution (HR) and a wide field of view (FOV) simultaneously is challenging. In this study, we propose a novel Efficient Attention Super-Resolution Transformer (EAST) to boost digital rock quality, which reconstructs HR details from low...

1050. Integrating deterministic geological model with multimodal machine learning to predict shale productivity

Gang Hui (China University of Petroleum Beijing)

13/05/2024, 17:45

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Over the past ten years, diverse machine learning techniques have been extensively employed in forecasting output for non-traditional reservoirs. Nevertheless, these techniques primarily utilized discrete point data obtained from field databases, such as well drilling, completion, monitoring, experiments, and production data of horizontal wells. However, this data fails to capture the spatial...

405. Stable Diffusion in Digital Rock Analysis: Applications, Challenges, and Future Prospects

Yutian Ma (China University of Petroleum (Beijing) Karamay Campus)

14/05/2024, 10:55

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Digital rock analysis has shown promise in visualizing geological microstructures and elucidating transport mechanisms in subsurface rocks, particularly in unconventional reservoirs such as tight sandstone and shale. Accurate image reconstruction techniques, which provide valuable insights into the pore network, grain distribution and connectivity, are essential to capture the intricate...

455. Optimizing Underground Hydrogen Storage through Surrogate Modeling: A CNN-LSTM-Attention Network Approach

Mr Zhilei Han (KAUST)

14/05/2024, 11:10

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Underground hydrogen storage (UHS) presents a viable solution for storing excess energy in suitable geological sites, ensuring a stable and scalable energy supply [1]. While extensive experience exists in underground natural gas storage [2], the significant differences in the properties of hydrogen pose unique challenges [3]. To deepen insights into the hydrogen recovery in UHS projects,...

319. Feasibility study of the inversion method for non-uniform hydrate saturation distribution based on ensemble Kalman filter algorithm

Yongge Liu (China University of Petroleum (East China))

14/05/2024, 11:25

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Field tests and laboratory experiments indicate that the spatial distribution of hydrate saturation in hydrate reservoirs is non-uniform. This non-uniform distribution significantly impacts the reservoir’s temperature changes, and gas and water production rates during reservoir development. Currently, the primary methods for determining hydrate saturation distribution in porous media are...

214. A Comprehensive Approach to In-Situ Stress Estimation in Subsurface Energy Structures using Numerical Simulation and Machine Learning

Dr Erfan Mohammadian

14/05/2024, 11:40

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

The objective of this research is to establish a consistent relationship between nonlinear numerical simulations and the obtained results for use in inverse analysis. We simulate the shape of breakouts, taking into account inelastic deformation of high-porosity limestone, using developed finite element methods under various in-situ conditions. Subsequently, the dataset is employed to train...

5. A deep learning enabled massive parallel simulator for porous media flow

Prof. Chensong Zhang (Academy of Mathematics and Systems Science)

14/05/2024, 12:00

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Due to the complex composition of oil and gas resources, reservoir engineers usually switch between different mathematical models when describing the properties of petroleum reservoirs. In addition to the commonly used black oil model, various compositional models have been proposed. Some EOR techniques, such as polymer flooding, must be simulated based on the framework of compositional...

57. Application of Machine Learning and Deep Learning Methods in Reservoir Development

Prof. Kai Zhang (China University of Petroleum (East China);Qingdao University of Technology)

14/05/2024, 12:15

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Machine learning (ML) has revolutionized various aspects of underground seepage, geological modeling, reservoir numerical simulation, production optimization, and big data analysis in the oil and gas industry. In particular, when it comes to reservoir development, ML methods, e.g., deep learning (DL) and intelligent computing, have proven to be superior to traditional methods in terms of...

434. 3D Pore Segmentation and Pore-Scale Simulation by Deep Learning

Haotian Li

14/05/2024, 12:30

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Objectives
This study aims to optimize the characterization and prediction of permeability and relative permeability in porous media through a multi-faceted approach. The primary objectives include achieving accurate 3D reconstruction of rock core images, implementing advanced deep learning models for segmentation, and addressing computational challenges associated with the Lattice...

906. A multi-well deep learning model considering geological and engineering parameters for the long-term forecasting of shale gas production

Yilun Dong (University of Science and Technology of China)

14/05/2024, 12:45

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

The production forecasting of shale gas wells is an important research topic in the natural gas industry. The underground pore structure is extremely complicated after hydraulic fracturing. Conventionally, researchers try to construct forecasting models via theoretical or experimental approaches. However, both theoretical and experimental approaches are faced with difficulties. Firstly, many...

87. Unsupervised resolution boosting of µCT scans integrated into a supervised convolutional network to predict 3D rock properties

Prof. Saeid Sadeghnejad (Applied Geology, Institute for Geosciences, Friedrich-Schiller-University Jena, 07749 Jena, Germany)

15/05/2024, 10:55

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

The anticipation of fluid transport behavior within porous media holds significant importance in a diverse array of applications, encompassing subsurface hydrology (Hu and Pfingsten 2023); petroleum industry (Moslemipour and Sadeghnejad 2021), geothermal energy utilization (Meller et al. 2017), and secure subsurface storage of hydrogen or CO2 (Esfandi et al. 2023, Kanaani et al. 2023)....

103. Deep learning for microstructure analysis of porous media from image augmentation, and multiscale fusion to image auto-segmentation

Dr Fugui Liu

15/05/2024, 11:10

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Based on deep generative adversarial networks, we present a comprehensive framework for image multiscale fusion and multi-component auto-segmentation, which aims to create a precise digital rock − a key part of Digital Rock Physics to predict the petrophysical properties of porous media. Compared to low-resolution images with a large field of view (FoV), high-resolution (HR) rock images are...

298. Machine Learning Assisted Numerical simulation of Propylene Glycol-mixed Steam Enhanced Extraction in Unsaturated soils

Zhixin Chen (Tsinghua university)

15/05/2024, 11:25

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Contaminated soils with hazardous persistent organic pollutants, such as Polycyclic Aromatic Hydrocarbons (PAHs), present significant challenges for efficient remediation, attributable to their low mobility and bioavailability. The technique of Propylene Glycol (PG)-mixed steam enhanced extraction has emerged as a promising remediation technology, markedly increasing the solubilization and...

785. Data Quality Assurance Metrics for Federated Machine Learning

Bernard Chang (The University of Texas at Austin)

15/05/2024, 11:40

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Machine/deep learning (ML/DL) have emerged as powerful tools for driving science innovation. These methods tend to be data hungry, with large volumes that are not likely to be collocated. Further, industry or professional companies typically have a vested interest in keeping their data private. Recent work has shown the viability of federated learning (FL), a ML /DL framework where multiple...

227. Solving seepage equation using physics-informed residual network without labeled data

Shuaijun Lv (Hefei University of Technology)

16/05/2024, 09:20

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Physics-informed neural network (PINN) is an innovative universal function approximator which adds physical constraints to neural network to make the fitting results satisfy the physical laws better. In this paper, a physics-informed residual network (PIResNet) is proposed to solve the single-phase seepage equation without labeled data. The loss function is constructed by summarizing the...

574. Efficient Surrogate Modeling of Subsurface Flow in Porous Media Using Transfer Learning with Multifidelity Data

Mr Wenyue Sun (China University of Petroleum East China)

16/05/2024, 09:35

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

ABSTRACT
In subsurface flow settings, deep-learning-based surrogate modeling is shown to be an effective approach to deal with cases that require a substantial amount of model simulations. However, a large number of high-fidelity training simulations are usually required to construct these deep-learning-based surrogate models. For large-scale models, it can be computationally prohibitive to...

1053. Super-resolution imaging of multiphase fluid distributions in porous media using deep learning

Zhuangzhuang Ma

16/05/2024, 09:50

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

X-ray imaging has become an indispensable tool in the study of porous media, significantly enhancing our understanding of multiphase flow within these pore structures. High-resolution X-ray images enable researchers to accurately measure or calculate critical rock properties such as porosity, interfacial surface area, curvature, and contact angle distributions. These images are also pivotal in...

425. Anchored Physics-Informed Neural Network for Fluid Flow Simulation in Heterogeneous Porous Media

Jingqi Lin (China University of Petroleum(East China))

16/05/2024, 11:50

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Numerical simulation is a vital tool for analyzing and predicting fluid flow in porous media. Physics-Informed Neural Networks (PINNs) can work out systems of partial differential equations (PDEs) by leveraging the universal approximation ability of Neural Network (NN), offering a novel approach for numerical model solving. However, current PINN-based methods are rarely used to simulate...

631. A Transformer-based framework for brine-gas interfacial tension prediction: Implications for H2, CH4 and CO2 geo-storage

Tianru Song (University of Science and Technology Beijing)

16/05/2024, 12:05

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Brine-gas interfacial tension (γ) is an essential parameter to determine fluid dynamics, trapping and distributions at pore-scale, thus influencing gas storage capacities and securities at reservoir-scale. However, γ is a complex function of pressure, temperature, ionic strength and gas composition, thus very time-consuming and costly to cover all these influencing factors by experiment....

901. Machine learning algorithms for predicting breakthrough curves for pore scale reactive flow in porous media and application to parameter identification

Prof. Oleg Iliev (Fraunhofer ITWM)

16/05/2024, 12:20

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

Reactive flows in porous media play an important role in our life and are crucial for many industrial, environmental and biomedical applications. Very often the concentration of the species at the inlet is known, and the so-called breakthrough curves, namely the time dependent concentration at the outlet, are the quantities which could be measured or computed numerically. The measurements and...

981. Effect discussion and workflow establishment of machine learning algorithm in logging classification: A case study from the classification of sepiolite bearing strata in the first member of Maokou Formation

杉 任

16/05/2024, 12:35

(MS15) Machine Learning and Big Data in Porous Media

Oral Presentation

The carbonate reservoir in the first member of Maokou Formation (Maokou-1 Member) of Middle Permian in Sichuan Basin have the characteristics of self-generating and, self-storage. Maokou-1 Member is expected to become a new field of unconventional gas reservoir exploration in carbonate rocks. The organic matter and pore development of Maokou-1 Member carbonate rocks are closely related to...