Digital Rock Research Articles

Exploring CO2 storage potential in Lithuanian deep saline aquifers using digital rock volumes: a machine learning guided approach

The increasing significance of carbon capture, utilization and storage (CCUS) as a climate mitigation strategy has underscored the importance of accurately evaluating subsurface reservoirs for CO2 sequestration [1]. In this context, digital rock volumes, obtained through advanced imaging techniques such as micro-Xray computed tomography (MXCT), offer intricate insights into the porous and permeable structures of geological formations [2]. This study presents a comprehensive methodology for assessing CO2 storage viability within Lithuanian deep saline aquifers, namely Syderiai and Vaskai, by utilizing petrophysical properties estimated from digital rock volumes [3, 4]. These petrophysical properties were derived from core samples collected from these formations. Utilizing machine learning algorithms, porosity was estimated while the Lattice Boltzmann method (LBM) was applied to determine permeability [5]. The methodology employed for estimating these petrophysical parameters was initially validated using samples from formations analogous to Lithuanian formations. Subsequently, it was applied to rock samples specifically obtained from Lithuanian formations. The estimated petrophysical properties were compared with peer-reviewed data from published literature. When fluids such as CO2 or H2 are injected into sub-surface reservoirs, they can alter pore and grain characteristics. Therefore, it is crucial to extract representative element volumes (REVs) from segmented volumes to study the impact of fluids on porosity and their distribution [6]. These mini models, representing small portions of the larger formation, assist in predicting fluid flow within the formation, which is vital for assessing the efficiency and safety of carbon capture and storage (CCS) operations. Subsequently, numerical modelling was conducted using the petrophysical parameters as inputs to assess the storage capacity of the Lithuanian formations using tNavigator software [7]. This research contributes to an enhanced understanding of pore space distribution and its role in various aspects of long-term CO2 storage. It also demonstrates the potential of integrating advanced imaging techniques, machine learning, and numerical modeling for accurate assessment and effective management of subsurface CO2 storage. This study shall aid in enhanced understanding of pore space distribution and their contribution towards various aspects of long-term storage. The results can be extended to study the geochemical reactions and geo-mechanical behaviour of the rocks. Such studies shall further facilitate identification of reservoir(s) wherein sequestration potential can be reliably explored.

Upscaling rocks mechanical properties to study Underground Hydrogen Storage feasibility

Underground Hydrogen Storage (UHS) is a feasible option for large-scale energy storage considering the advancements of the large-scale production of green hydrogen. One of the main engineering objective is to ensure the continuous safety of the storage, such that subsurface operations can be carried under safe stress regimes. Despite obvious similarities to Carbon Capture and Storage, a few differences make the task a new research challenge.
 The first one relates to the cyclic nature of UHS, which is expected to be carried at variable frequency and injection/production loads. Current models are not adequate for the lower frequency range considered in this application. In that case, the visco-plastic nature of rocks becomes non-negligible and needs to be taken into account.
 As a second observation, UHS revolves around a new gas, much lighter than CH4 and super critical CO2. Hydrogen’s atoms are so small they can diffuse even inside rock and this absorption causes rock matrix mechanical properties to weaken. This process is know as Hydrogen Embrittlement. When unaccounted for, such physical phenomenon could lead to catastrophic failure of the caprock, which is supposed to maintain stability to ensure safe storage. The caprock being responsible for the confinement of the hydrogen in the reservoir, development of cracks would enhance greatly permeability of an otherwise impermeable medium, resulting in an environmental disaster as the hydrogen suddenly leaks towards the subsurface and through groundwater aquifers.
 No empirical model is able to capture those two behaviours at the macro-scale since they are both phenomena principally related to grain-scale physics. As such, this contribution presents a Digital Rock Physics framework to upscale rock mechanical properties from the grain-scale. Rocks of interest are microCT-scanned to extract the digitized microstructure. Direct numerical simulations of elasto-plasticity are performed for different stress paths in order to compute the full yield surface (see Figure 1) instead of just the Uniaxial Compressive Strength. While most studies use Discrete Element Modelling to consider grain contacts explicitly, our simulator uses Finite Element Modelling which allows more flexibility in the approach to model multiphysics processes present during UHS. The contacts are modelled instead as an upscaled plastic law. Details of the numerical algorithms are presented in [1].
 As a first case study for this framework, we present a comprehensive parametric study on the impact of cementation on rock strength for real microstructures of granular materials. The framework is then coupled with a numerical erosion algorithm that simulates homogeneous precipitation of mineral matter to represent cementation. New results on the influence of cement property namely Young’s modulus, friction and cohesion on the rock’s yield surface are explored. This study contributes to preliminary results on Hydrogen Embrittlement which directly influences those same mechanical properties. However more work is needed to model realistically the Hydrogen Embrittlement, which is the aim of our new PhD project OCEAN. The process will be observed experimentally at the micro-scale in order to calibrate the simulator. MicroCT-scan images will determine the spatial distribution of the phenomenon. Visco-plasticity will be implemented from [2] to go one step further and determine the effect of Hydrogen Embrittlement during cyclic injection/production of hydrogen.

Integrated Formation Evaluation for Site-Specific Evaluation, Optimization, and Permitting of Carbon Storage Projects

Participation in over 80 carbon capture and sequestration (CCS) projects spanning 25 years has led to the evolution of a recommended well-based appraisal workflow for CO2 sequestration in saline aquifers. Interpretation methods are expressly adapted for CCS applications to resolve key reservoir parameters, constrain field-scale modeling, provide answers required for the permitting process, and de-risk unique CCS evaluation challenges, such as Storage capacity Injectivity Containment. A challenge complicating all of the above is the eventual impact of three-way interaction among rock matrix, brine, and (impure) CO2 streams. Most logging, sampling, and laboratory techniques are adapted from established domains such as enhanced oil recovery, underground gas storage, and unconventional reservoir evaluation, though some CCS-specific innovation is also needed. Storage evaluation begins with established methods for lithology, porosity, permeability, and pressure, while special core analysis (SCAL) determines CO2 storage efficiency and relative permeability. Containment evaluation spans multiple disciplines and methods: the petrophysicist’s task to quantify seal capacity relies heavily on laboratory analysis, while geologists leverage downhole imaging tools to verify caprock structural/tectonic integrity. Geomechanics engineers define safe injection pressure via mechanical earth models (MEMs) built on advanced acoustic logs calibrated by core geomechanics, wellbore failure observations, and in-situ stress tests. The impact of rock-brine-CO2 interactions is studied via custom SCAL experiments and/or pore-scale digital rock simulations that rigorously represent chemical and thermal processes. Wireline formation tester samples provide representative formation brine as feedstock for SCAL. Water samples also enable operators to prove injection within regulatory limits while establishing baselines for future monitoring programs. Examples applied to recent CCS projects in North America are presented. All of the above data need to be integrated into a CCS model predicting the CO2 plume behavior across the area of interest and within multiple horizons.

Modeling and petrophysical properties of digital rock models with various pore structure types: An improved workflow

Pore structure is a crucial factor affecting the physical properties of porous materials, and understanding the mechanisms and laws of these effects is of great significance in the fields of geosciences and petroleum engineering. However, it remains a challenge to accurately understand and quantify the relationship between pore structures and effective properties. This paper improves a workflow to focus on investigating the effect of pore structure on physical properties. First, a hybrid modeling approach combining process-based and morphology-based methods is proposed to reconstruct 3D models with diverse pore structure types. Then, the characteristics and differences in pore structure in these models are compared. Finally, the variation laws and pore-scale mechanisms of the influence of pore structure on physical properties (permeability and elasticity) are discussed based on the reconstructed models. The relationship models between pore structure parameters and permeability/elastic parameters in the grain packing model are established. The effect of pore structure evolution on permeability/elasticity and the microscopic mechanism in three types of morphology-based reconstruction models are explored. The influence degree of pore structure on elastic parameters (bulk modulus, shear modulus, P-wave velocity, and S-wave velocity) is quantified, reaching 29.54%, 51.40%, 18.94%, and 23.18%, respectively. This work forms a workflow for exploring the relationship between pore structures and petrophysical properties at the microscopic scale, providing more ideas and references for understanding the complex physical properties in porous media.

Study on the Diverting Characteristics of Diverting Acid Based on Digital Rock Physics

Study on the Diverting Characteristics of Diverting Acid Based on Digital Rock Physics

Core analysis in a changing world – how technology is radically benefiting the methodology to acquire, the ability to visualize and the ultimate value of core data

Abstract Core analysts principally study the storage, flow and saturation properties of porous rocks and sediments. Some of the derived parameters are specific to hydrocarbon production but many have commonality with other subsurface disciplines such as hydrology and soil science. Traditional core analysis involves direct physical experimentation on core plugs to derive a range of parameters used as calibration for conventional well logs, and to predict hydrocarbon reserves and recovery. The mechanisms and processes for obtaining such data have evolved significantly during the last century, from the manual instruments of the mid-twentieth century to the accredited digital data collection and recording of the 1990s onwards. X-ray micro- and nano-scale computed tomography (CT) imaging led to the development of the digital rock physics subdiscipline in the early 2000s. This has subsequently allowed direct visualization of fluid flow at the pore scale, imaging the wetting phase and multiphase fluid mobility. Multiscale imaging workflows are being developed to overcome issues around heterogeneous rock and the limited field of view associated with the highest resolution X-ray CT images. Hybrid workflows, which combine digital rock physics with traditional core analysis, are becoming increasingly common to meet the challenges associated with some of the most difficult to constrain properties, such as relative permeability. At a larger scale, the recent development of multisensor core logging (MSCL) tools has allowed the cost-effective acquisition of essentially continuous high-resolution 1D, 2D and 3D datasets from both slabbed and unslabbed whole core. Often aided by artificial intelligence to manage and interpret these large physical and chemical datasets, both new and legacy core can be rapidly screened to allow representative subsampling for detailed laboratory experimentation. The context and data provided by the MSCL then allows effective upscaling of these time- and cost-intensive point-source measurements. In the last decade, extended reality (XR) has resulted in a step change in the ability to visualize and integrate core and core-derived information with other subsurface datasets. A very wide range of scales can be managed effectively, from micrometre- to centimetre-scale petrographical and core analysis data, to metre-scale well logs and up to kilometre-scale 3D and 4D seismic. These tools allow stakeholders to work and meet from any location in a common workspace, and efficiently scale and interrogate data in a virtual 3D environment. The various advances in core analysis and associated technologies during the early twenty-first century mean that the study of porous media to help enable the energy transition looks assured. During the coming decades, applications as diverse as carbon capture, utilization and storage (CCUS), hydrogen storage, geothermal energy generation, mining for critical minerals, palaeoclimate studies, radioactive waste management, and site surveys for windfarms will all continue to benefit from the data and understanding derived from core analysis.

3D reconstruction of digital rock guided by petrophysical parameters with deep learning

Digital rock reconstruction is the basis of digital petrophysics, aiming to provide diverse samples for heterogeneity analysis on microscopic pore structures. Traditional physical experiments and numerical simulation methods tend to lead to a high experimental cost or a long reconstruction time, which is improved by generative adversarial networks (GAN) in recent studies. However, GAN's reconstructions are still too stochastic to satisfy the need for joint control over user-defined feature combinations. Therefore, we propose a multi-conditional GAN (MCGAN) model, using petrophysical parameters to, in turn, guide and drive the training process. Prior information, such as porosity, tortuosity, specific surface area, and fractal dimension, is concatenated with random noise to precisely control the pore distribution of dummy samples. In the loss function, we design a feature loss term to accelerate convergence and allow users the freedom to choose parameter combinations requiring tighter constraints. We compare MCGAN with three other GAN-based models on petrophysical properties, statistical features, and visual effects for performance evaluation. The results show that for applicable parameters, our method supports better mapping performance and higher degrees of freedom in multi-conditional reconstruction of digital rock. Therefore, MCGAN can quickly reconstruct digital cores with specific feature combinations according to users' needs. This work proposes a generative model for binary digital rock based on deep learning, promising to help more precisely analyze the microscopic heterogeneity of pore structures.

Relative permeability curve prediction from digital rocks with variable sizes using deep learning

Recent advancements in artificial intelligence (AI) technology have offered new ways to obtain the relative permeability curve that is crucial for subsurface engineering problems. The existing data-driven AI approaches for the relative permeability prediction are only suitable for a specific geometry rock size and lack the cross-size prediction ability. In this paper, we propose a new physics-informed prediction model called the three-dimensional (3D) spatial pyramid pooling convolutional network (3DSPPConvNet) model embedding three important parameters: computed tomography (CT) scanning resolution, interfacial tension, and contact angle distribution. Our model can deal with 3D digital rocks with variable sizes as it is based on a multi-size digital rock dataset containing 1512 couples of relative permeability curves. Their relative permeability curves for training and testing are obtained from pore network modeling (PNM) results with a linear postprocessing process. The results show that the physics-informed 3DSPPConvNet has a high prediction accuracy of 95.3%, with scale information being the most important physics feature accounting for 51%. In addition, compared with PNM, the computational time and memory of the intelligent model are saved by 63% and 80%, respectively.

Anisotropy of fracture and permeability in high-rank coal analyzed using digital rock physics

Rock microstructure analysis and physical property simulation by means of digital rock physics (DRP) can overcome the deficiencies such as poor quantification and visibility in traditional petrophysical experiments. The establishment of a complete set of DRP is the basis of analyzing the microstructure of unconventional oil and gas reservoirs. In this study, high-rank coal, a material with anisotropic fracture and permeability properties, was tested using x-ray computer tomography scanning to reconstruct the digital core, and the fracture direction and structural characteristics were quantified with respect to the main inertia axis and the Feret diameter, respectively. A method for calculating the fractal dimension and tortuosity based on DRP is presented, and the optimal interaction between the lattice Boltzmann method seepage simulation and DRP is identified. The results show that the average length, width, aperture, and volume of fractures in the direction of face cleat (DFC) are 1.13, 1.10, 1.11, and 1.09 times that in the direction of butt cleat (DBC), respectively, and their surface area, count, fracture porosity, fractal dimension, and tortuosity are 1.17, 1.16, 1.26, 1.04, and 1.10 times that of DBC, respectively. The permeability of DFC was found to be 3.46 times that of DBC. This study presents an effective method for determining the dominant direction of fracture structure and fluid migration that is not limited to pores and fractures in rock but can also be used to characterize the physical properties of skeletons or solid particles in other materials.

A detailed rock density model of the Hong Kong territories

We used the geological map and published rock density measurements to compile the digital rock density model for the Hong Kong territories. We then estimated the average density for the whole territory. According to our result, the rock density values in Hong Kong vary from 2101 to 2681 kg·m−3. These density values are typically smaller than the average density of 2670 kg·m−3, often adopted to represent the average density of the upper continental crust in physical geodesy and gravimetric geophysics applications. This finding reflects that the geological configuration in Hong Kong is mainly formed by light volcanic formations and lava flows with overlying sedimentary deposits at many locations, while the percentage of heavier metamorphic rocks is very low (less than 1%). This product will improve the accuracy of a detailed geoid model and orthometric heights.

A study of transfer learning in digital rock properties measurement

The measurement of physical parameters of porous rock, which constitute reservoirs, is an essential part of hydrocarbon exploration. Typically, the measurement of these physical parameters is carried out through core analysis in a laboratory, which requires considerable time and high costs. Another approach involves using digital rock models, where the physical parameters are calculated through image processing and numerical simulations. However, this method also requires a significant amount of time for estimating the physical parameters of each rock sample. Machine learning, specifically convolutional neural network (CNN) algorithms, has been developed as an alternative method for estimating the physical parameters of porous rock in a shorter time frame. The advancement of CNN, particularly through transfer learning using pre-trained models, has contributed to rapid prediction capabilities. However, not all pre-trained models are suitable for estimating the physical parameters of porous rock. In this study, transfer learning was applied to estimate parameters of sandstones such as porosity, specific surface area, average grain size, average coordination number, and average throat radius. Six types of pre-trained models were utilized: ResNet152, DenseNet201, Xception, InceptionV3, InceptionResNetV2, and MobileNetV2. The results of this study indicate that the DenseNet201 model achieved the best performance with an error rate of 2.11%. Overall, this study highlights the potential of transfer learning to ultimately lead to more efficient and effective computation.

Digital Rock Physics in Cuttings Using High-Resolution Thin Section Scan Images

Digital rock physics (DRP) has undergone significant advancements in the use of various imaging techniques to acquire three-dimensional volumes and images of rock samples for the computation of petrophysical properties. This study focuses on developing a DRP workflow using high-resolution thin section scans for computing porosity and permeability in cuttings samples. The workflow was tested on quarry sandstone plug samples and artificially generated pseudo-cuttings before applying it to real cuttings from oil and gas wells. The results show that the porosity and permeability values obtained through the DRP workflow are statistically equivalent to those obtained through conventional routine core analysis (RCAL). The workflow was also able to handle the presence of various lithologies in real cuttings samples. The study demonstrates the feasibility of obtaining porosity and permeability values in cutting samples using the DRP approach, offering a fast and cost-effective methodology that provides additional data and allows linking petrophysical properties to image data from the cuttings.

Digital Tools for the Promotion of Geological and Mining Heritage: Case Study from the Thames Goldfield, Aotearoa, New Zealand

The geoheritage value in the Thames District, North Island, New Zealand, is notable for its over 50 epithermal gold deposits associated with the Coromandel Volcanic Zone, significant to the region’s geological, cultural, and mining heritage. This case study was conducted in collaboration with the Thames School of Mines and Mineralogical Museum to develop a series of web-based applications for public outreach and an accessible museum experience through the utilization of specimens from the mineralogical museum. This research applies a conceptual framework of ex situ geoheritage to explore links between local geology with cultural and mining heritage. Minerals and rock specimens collected for the Thames School of Mines Mineralogical Museum were used to create 3D virtual models demonstrating the epithermal mineralization in the Thames Goldfield. Outputs of this project consist of two digital products, including a digital mineral and rock repository and the dissemination of the geological collection through integrating the photogrammetric models into a user-friendly outreach, ArcGIS Storymaps, to depict the geoheritage relationship of the specimens to regional gold mining, and at the same time, to be developed to be implemented in geoscience education and communication. The results of this paper are intended to promote the use of digital tools for enhancing and raising awareness of the geoheritage values of the Thames Goldfield. This approach has relevance for Papua New Guinea and the Pacific Islands in raising awareness of geological, mineral, and mining heritage within widely distributed and often isolated communities across island archipelago nations.

Reconstruction of 3D digital rocks with controllable porosity using CVAE-GAN

Digital rock technology provides an effective approach for analyzing the pore structure and physical properties of rocks in geophysics and petroleum science. Although deep learning techniques have significantly improved the efficiency of digital rock reconstruction, existing works within this field often lack necessary constraints. Therefore, our study proposes a new method for digital rock reconstruction that employs a conditional variational auto-encoder generative adversarial network (CVAE-GAN). This approach integrates the strengths of a conditional variational auto-encoder (CVAE) and a conditional generative adversarial network (CGAN) to produce high-quality digital rocks. To enable the controllable porosity reconstruction of digital rocks, the proposed method includes porosity information within the encoder, decoder, and discriminator. Additionally, we used the Wasserstein GAN with gradient penalty (WGAN-GP) during the training process to enhance the stability of the neural networks. We conducted experimental evaluations using different types of samples to validate the effectiveness of our reconstruction approach. Calculations of pore structure parameters and simulations of rock physical properties were performed on the reconstructed digital rocks, indicating a robust correspondence with both the real samples and theoretical models. These results provide compelling evidence for the accuracy of the proposed method in digital rock reconstruction and suggest its promising prospects for investigating petrophysical properties.

Estimating Permeability of Porous Media from 2D Digital Images

Digital rock physics (DRP) has been widely used as an effective approach for estimating the permeability of porous media. However, conventional implementation of DRP requires the reconstruction of three-dimensional (3D) pore networks, which suffer from intensive memory and underlying uncertainties. Therefore, it is highly significant to develop an approach only based on two-dimensional (2D) cross-sections of parent samples without 3D reconstruction. In this study, we present a novel approach that combines the Kozeny–Carman equation with fractal theory to derive a bridge function that links 2D cross-sectional images and 3D pore structures of parent samples in flow equivalence. Using this bridge function, we predicted the physical properties of the parent samples, including the permeability, bulk porosity, tortuosity, and pore fractal dimension. To validate our model, we performed Lattice Boltzmann (LB) simulations on nine carbonate samples and compared the LB simulation results with our model’s predictions. We also compared our predicted results with available data on various porous materials, such as sandstone, glass beads, and carbonate, in the literature. Our findings demonstrate that without reconstructing 3D pore networks, our method provides a reliable estimation of sample permeability using 2D cross-sectional images. This approach not only simplifies the determination of sample permeability in heterogeneous porous media but also sheds new light on the inherent correlations between 2D cross-sectional information and 3D pore structures of parent samples. Moreover, the derived model may be conducible to a better understanding of flow in reservoirs during the extraction of unconventional onshore and offshore oil/gas.

3D-porous-GAN: a high-performance 3D GAN for digital core reconstruction from a single 3D image

The 3D digital rock technology is extensively utilized in analyzing rock physical properties, reservoir modeling, and other related fields. This technology enables the visualization, quantification, and analysis of microstructures in rock cores, leading to precise predictions and optimized designs of reservoir properties. Although the accuracy of 3D digital rock reconstruction algorithms based on physical experiments is high, the associated acquisition costs and reconstruction processes are expensive and complex, respectively. On the other hand, the 3D digital rock random reconstruction method based on 2D slices is advantageous in terms of its low cost and easy implementation, but its reconstruction effect still requires significant improvement. This article draws inspiration from the Concurrent single-image generative adversarial network and proposes an innovative algorithm to reconstruct 3D digital rock by improving the generator, discriminator, and noise vector in the network structure. Compared to traditional numerical reconstruction methods and generative adversarial network algorithms, the method proposed in this paper is shown to achieve good agreement with real samples in terms of Dykstra-Parson coefficient, porosity, two-point correlation function, Minkowski functionals, and visual display.

Assessment of the theoretical models of effective thermal conductivity based on digital rock physics for clean sandstones

Knowledge of the effective thermal conductivity (ETC) of porous media provides valuable insights into the heat flow and profoundly affects the feasibility analysis of thermal enhanced oil recovery methods and geothermal field development. Theoretical assessment of ETC based on structural properties of the porous medium and various simplification assumptions by prediction models have shown acceptable results towards ETC quantification; however, their applicability must be investigated further due to the availability of various types of such models. In this study, we first employed a digital rock approach to simulate the heat conductivity of micro-CT images of 10 clean sandstone samples using Avizo software in different directions for evacuated, air-saturated, and water-saturated conditions. The filtered and segmented images were applied to perform voxel-based ETC simulations under steady state condition. Then, the simulation results were compared to the computed ETCs for each sample by 11 theoretical models. Two well-known thermal conductivity bounds, i.e., Hashin and Shtrikman bounds and Wiener bounds were used to confirm the validity of models and simulation outputs. It was shown that the Haung model generates the best outputs for vacuum (MAPE = 0.08 and RMSE = 0.37) and air-saturated (MAPE = 0.07 and RMSE = 0.32) states. While the water-saturated condition was competently predicted by the Ribaud model (MAPE = 0.05 and RMSE = 0.31). The overall analysis indicated that the Haung model is the best model for the prediction of ETC in all three pore-filling states for clean sandstones.

Digital Rock Analysis Based on X-ray Computed Tomography of a Complex Pre-salt Carbonate Reservoir from the Santos Basin, SE Brazil

Digital Rock Analysis Based on X-ray Computed Tomography of a Complex Pre-salt Carbonate Reservoir from the Santos Basin, SE Brazil

Machine learning prediction of tortuosity in digital rock

Physical rock property measurement is an important stage in energy exploration, both for hydrocarbons and geothermal sources. The value of physical rock properties can provide information about reservoir quality, and one of these properties is tortuosity. Tortuosity is an intrinsic property of porous materials that describes the level of complexity of the porous arrangement when a fluid passes through it. Conventionally, tortuosity values are measured through laboratory analysis and numerical simulation, but these measurements can take a long time. An alternative method for measuring tortuosity is using machine learning with a convolutional neural network (CNN). A CNN is a type of deep neural network designed to analyze multi-channel images and has been applied successfully to classification and non-linear regression problems. By training a CNN on a dataset of digital rock samples that have been simulated using numerical computation to obtain their tortuosity values, it is possible to demonstrate that CNNs can accurately predict the tortuosity of digital rock. The result is that the CNN model can predict tortuosity values with the Xception model being the most accurate with the lowest RMSE value of 0.90962.

Applicability of 2D algorithms for 3D characterization in digital rocks physics: an example of a machine learning-based super resolution image generation

Digital rock physics is based on imaging, segmentation and numerical computations of rock samples. Due to challenges regarding the handling of a large 3-dimensional (3D) sample, 2D algorithms have always been attractive. However, in 2D algorithms, the efficiency of the pore structures in the third direction of the generated 3D sample is always questionable. We used four individually captured µCT-images of a given Berea sandstone with different resolutions (12.922, 9.499, 5.775, and 3.436 µm) to evaluate the super-resolution 3D images generated by multistep Super Resolution Double-U-Net (SRDUN), a 2D algorithm. Results show that unrealistic features form in the third direction due to section-wise reconstruction of 2D images. To overcome this issue, we suggest to generate three 3D samples using SRDUN in different directions and then to use one of two strategies: compute the average sample (reconstruction by averaging) or segment one-directional samples and combine them together (binary combination). We numerically compute rock physical properties (porosity, connected porosity, P- and S-wave velocity, permeability and formation factor) to evaluate these models. Results reveal that compared to one-directional samples, harmonic averaging leads to a sample with more similar properties to the original sample. On the other hand, rock physics trends can be calculated using a binary combination strategy by generating low, medium and high porosity samples. These trends are compatible with the properties obtained from one-directional and averaged samples as long as the scale difference between the input and output images of SRDUN is small enough (less than about 3 in our case). By increasing the scale difference, more dispersed results are obtained.