Multimodal imaging and machine learning to enhance microscope images of shale

Anderson, T. W.; Vega, Bolivia; Kovscek, Anthony R.

doi:10.1016/j.cageo.2020.104593

Cited by 38 publications

(31 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An additional advantage of multimodal imaging is to overcome the relatively poorer contrast in CT as compared to SEM images. 151 Even if acquired at nominally the same spatial resolution, a CT image tends to have lesser contrast in grayscale values and will experience some washing out of fine details. Anderson et al (2020) 151,152 show how to use a few spatially correlated nanoCT and SEM images to train a machine learning algorithm to improve contrast in nanoCT images, thereby reducing the need for destructive SEM imaging to obtain highcontrast images.…”

Section: X-ray Computed Tomographymentioning

confidence: 99%

Chemical and Reactive Transport Processes Associated with Hydraulic Fracturing of Unconventional Oil/Gas Shales

et al. 2022

Self Cite

View full text Add to dashboard Cite

Hydraulic fracturing of unconventional oil/gas shales has changed the energy landscape of the U.S. Recovery of hydrocarbons from tight, hydraulically fractured shales is a highly inefficient process, with estimated recoveries of <25% for natural gas and <5% for oil. This review focuses on the complex chemical interactions of additives in hydraulic fracturing fluid (HFF) with minerals and organic matter in oil/gas shales. These interactions are intended to increase hydrocarbon recovery by increasing porosities and permeabilities of tight shales. However, fluid–shale interactions result in the dissolution of shale minerals and the release and transport of chemical components. They also result in mineral precipitation in the shale matrix, which can reduce permeability, porosity, and hydrocarbon recovery. Competition between mineral dissolution and mineral precipitation processes influences the amounts of oil and gas recovered. We review the temporal/spatial origins and distribution of unconventional oil/gas shales from mudstones and shales, followed by discussion of their global and U.S. distributions and compositional differences from different U.S. sedimentary basins. We discuss the major types of chemical additives in HFF with their intended purposes, including drilling muds. Fracture distribution, porosity, permeability, and the identity and molecular-level speciation of minerals and organic matter in oil/gas shales throughout the hydraulic fracturing process are discussed. Also discussed are analysis methods used in characterizing oil/gas shales before and after hydraulic fracturing, including permeametry and porosimetry measurements, X-ray diffraction/Rietveld refinement, X-ray computed tomography, scanning/transmission electron microscopy, and laboratory- and synchrotron-based imaging/spectroscopic methods. Reactive transport and spatial scaling are discussed in some detail in order to relate fundamental molecular-scale processes to fluid transport. Our review concludes with a discussion of potential environmental impacts of hydraulic fracturing and important knowledge gaps that must be bridged to achieve improved mechanistic understanding of fluid transport in oil/gas shales.

show abstract

Section: X-ray Computed Tomographymentioning

confidence: 99%

Chemical and Reactive Transport Processes Associated with Hydraulic Fracturing of Unconventional Oil/Gas Shales

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…Juyal et al (2019Juyal et al ( , 2020 2D (SEM-EDX) and 3D (X-ray CT) approaches may be combined to build 3D chemical maps of soil samples based on a statistical approach such as that proposed by Hapca et al (2015). Similarly, Anderson et al (2020) combined 2D (FIB-SEM) with 3D images (TXM, transmission X-ray microscopy) using machine learning on sediment rock materials. Another approach is to merge 2D images acquired at 3 different scales to create one final image with a better resolution and multiscale porosity information (Karsanina et al, 2018).…”

Section: Upscaling Issuesmentioning

confidence: 99%

“…New methods are now able to provide much more information than before: not only are the resolutions of images higher but it is also possible to obtain spatial information on soil physical, chemical and biological characteristics in three dimensions. Furthermore, image processing and analysis tools have become more efficient, allowing for better correction, segmentation of areas of interest or even predictions of chemical composition (Hapca et al, 2015;Anderson et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

Opportunities and limits in imaging microorganisms and their activities in soil microhabitats

Védère¹,

Laure²,

Nunan³

et al. 2022

Soil Biology and Biochemistry

View full text Add to dashboard Cite

“…This limits the number of applications, especially in shale fabric reconstruction. Other related work on shale image reconstruction has focused on reconstructing fine-scale features from coarse-scale images of shales [34] or on assimilating multimodal imaging data [35]. However, beyond these studies, there is little work on synthesizing or reconstructing source rock images using deep generative models.…”

Section: Synthesis Of Geologic Samplesmentioning

confidence: 99%

“…The dataset used is 2D aligned transmission X-ray microscopy (TXM) and FIB-SEM nanoscale images acquired for a Vaca Muerta shale sample. Such a dataset is usable for training image translation models to predict high-contrast, sample-destructive FIB-SEM data from low-contrast nondestructive TXM data [35]. We synthesized the multimodal data by training the multimodal image as a 2-channel input image to the flow model.…”

Section: Multimodal Image Synthesismentioning

confidence: 99%

RockFlow: Fast Generation of Synthetic Source Rock Images Using Generative Flow Models

et al. 2020

Self Cite

View full text Add to dashboard Cite

Image-based evaluation methods are a valuable tool for source rock characterization. The time and resources needed to obtain images has spurred development of machine-learning generative models to create synthetic images of pore structure and rock fabric from limited image data. While generative models have shown success, existing methods for generating 3D volumes from 2D training images are restricted to binary images and grayscale volume generation requires 3D training data. Shale characterization relies on 2D imaging techniques such as scanning electron microscopy (SEM), and grayscale values carry important information about porosity, kerogen content, and mineral composition of the shale. Here, we introduce RockFlow, a method based on generative flow models that creates grayscale volumes from 2D training data. We apply RockFlow to baseline binary micro-CT image volumes and compare performance to a previously proposed model. We also show the extension of our model to 2D grayscale data by generating grayscale image volumes from 2D SEM and dual modality nanoscale shale images. The results show that our method underestimates the porosity and surface area on the binary baseline datasets but is able to generate realistic grayscale image volumes for shales. With improved binary data preprocessing, we believe that our model is capable of generating synthetic porous media volumes for a very broad class of rocks from shale to carbonates to sandstone.

show abstract

Multimodal imaging and machine learning to enhance microscope images of shale

Cited by 38 publications

References 25 publications

Chemical and Reactive Transport Processes Associated with Hydraulic Fracturing of Unconventional Oil/Gas Shales

Chemical and Reactive Transport Processes Associated with Hydraulic Fracturing of Unconventional Oil/Gas Shales

Opportunities and limits in imaging microorganisms and their activities in soil microhabitats

RockFlow: Fast Generation of Synthetic Source Rock Images Using Generative Flow Models

Contact Info

Product

Resources

About