Image processing for fractal geometry-based discrete fracture network modeling input data: A methodological approach

Kovács, Dorottya; Dabi, Gergely; Vásárhelyi, Balázs

doi:10.1556/24.61.2018.09

Cited by 3 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Analysis of the geometric parameters from the prepared specimens should be a straightforward process with not much room for error, and the process remained systematic [13]. This means that fluorescent images were produced in fixed conditions (light, camera settings, possible post processing) before the segmentation to avoid biases.…”

Section: Discussionmentioning

confidence: 99%

“…Characterization of the microfracture geometry can be achieved by modelling the fracture system's characteristic geometric data: interconnectivity, openness, and density of the microfracture systems. There are three major approaches to modelling the hydraulic properties: An (1) equivalent continuum, (2) DFN, or (3) hybrid models combining an equivalent continuum and DFN [5,[11][12][13].…”

Section: Fracture Network Simulationmentioning

confidence: 99%

See 1 more Smart Citation

Comparison of DFN Modelled Microfracture Systems with Petrophysical Data in Excavation Damaged Zone

et al. 2021

Self Cite

View full text Add to dashboard Cite

Physical and petrographic properties of drill core specimens were determined as a part of investigations into excavation damage in the dedicated study area in the ONKALO® research facility in Olkiluoto, Western Finland. Microfractures in 16 specimens from two drillholes were analysed and used as a basis for fractal geometry-based discrete fracture network (DFN) modelling. It was concluded that the difference in resistivity between pegmatoid granite (PGR) and veined gneiss (VGN) specimens of similar porosity was likely due to differences in the types of microfractures. This hypothesis was confirmed from microfracture analysis and simulation: fractures in gneiss were short and mostly in one preferred orientation, whereas the fractures in granite were longer and had two preferred orientations. This may be due to microstructure differences of the rock types or could suggests that gneiss and granite may suffer different types of excavation damage. No dependencies on depth from the excavated surface were observed in the geometric parameters of the microfractures. This suggests that the excavation damaged zone cannot be identified based on the changes in the parameters of the microfracture networks, and that the disturbed layer observed by geophysical methods may be caused by macro-scale fractures.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Fracture Network Simulationmentioning

confidence: 99%

Comparison of DFN Modelled Microfracture Systems with Petrophysical Data in Excavation Damaged Zone

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…In recent years, interactive (supervised) machine learning algorithms like the RF classifier (Ho 1994;Breiman 2001; e.g., implemented in the Trainable Weka Segmentation, TWS: Arganda-Carreras et al 2017 or in the opensource software package Ilastik by Berg et al 2019) were primarily developed for biological cases (Babel et al 2020;Brady et al 2020;Eyolfson et al 2020) and offer an alternative to deep learning methods. Berg et al (2018) showed that depending on the training, the RF classifier performs better than the classical segmentation methods and does not require any prior filtering (Garfi et al 2020;Kovács et al 2019;Scanziani et al 2018). The drawbacks of the TWS implementation include high computational demands and the variety of available filters and classifiers, which makes it difficult to implement without sufficient resources and advanced knowledge of the methods.…”

Section: Introductionmentioning

confidence: 99%

Benchmarking conventional and machine learning segmentation techniques for digital rock physics analysis of fractured rocks

et al. 2022

View full text Add to dashboard Cite

Image segmentation remains the most critical step in Digital Rock Physics (DRP) workflows, affecting the analysis of physical rock properties. Conventional segmentation techniques struggle with numerous image artifacts and user bias, which lead to considerable uncertainty. This study evaluates the advantages of using the random forest (RF) algorithm for the segmentation of fractured rocks. The segmentation quality is discussed and compared with two conventional image processing methods (thresholding-based and watershed algorithm) and an encoder–decoder network in the form of convolutional neural networks (CNNs). The segmented images of the RF method were used as the ground truth for CNN training. The images of two fractured rock samples are acquired by X-ray computed tomography scanning (XCT). The skeletonized 3D images are calculated, providing information about the mean mechanical aperture and roughness. The porosity, permeability, flow fields, and preferred flow paths of segmented images are analyzed by the DRP approach. Moreover, the breakthrough curves obtained from tracer injection experiments are used as ground truth to evaluate the segmentation quality of each method. The results show that the conventional methods overestimate the fracture aperture. Both machine learning approaches show promising segmentation results and handle all artifacts and complexities without any prior CT-image filtering. However, the RF implementation has superior inherent advantages over CNN. This method is resource-saving (e.g., quickly trained), does not need an extensive training dataset, and can provide the segmentation uncertainty as a measure for evaluating the segmentation quality. The considerable variation in computed rock properties highlights the importance of choosing an appropriate segmentation method.

show abstract

Effects of Excavation Damage on the Physical Properties of Rock Matrix

Kiuru

Jacobsson

Király³

et al. 2021

IOP Conf. Ser.: Earth Environ. Sci.

View full text Add to dashboard Cite

Posiva Oy has conducted investigations into excavation damage, including comprehensive laboratory testing of physical properties of rock specimens from excavation damaged rock mass. Laboratory testing was conducted on drill core specimens extracted from the excavated surface of a tunnel located at approximately 345 m depth in Olkiluoto, Finland. A total of 141 drill core specimens of three main rock types, a structurally isotropic coarse-grained pegmatoid (PGR) and structurally anisotropic veined gneiss (VGN) and diatexitic gneiss (DGN), were subjected to petrophysical testing, rock mechanics testing and petrographic analyses. Results from the various tests were subjected to rigorous statistical analysis in order to reveal the effects excavation damage has on the physical properties of the rock mass. Results of the study revealed changes that are credited to excavation damage in resistivity, S-wave velocity and various elastic properties of the rock specimens. Effects of excavation damage and the depth of the excavation damaged zone seem to be different to gneiss compared to pegmatoid. On microscopic level, the extent of excavation damaged zone appears to be 0.2 – 0.4 m depending on the measured property. This means that the deeper excavation damaged layer observed by geophysical surveys may be caused by larger scale fractures.

show abstract

Image processing for fractal geometry-based discrete fracture network modeling input data: A methodological approach

Cited by 3 publications

References 17 publications

Comparison of DFN Modelled Microfracture Systems with Petrophysical Data in Excavation Damaged Zone

Comparison of DFN Modelled Microfracture Systems with Petrophysical Data in Excavation Damaged Zone

Benchmarking conventional and machine learning segmentation techniques for digital rock physics analysis of fractured rocks

Effects of Excavation Damage on the Physical Properties of Rock Matrix

Contact Info

Product

Resources

About