Planetary exploration depends heavily on 3D image data to characterize the static and dynamic properties of the rock and environment. Analyzing 3D images requires many computations, causing efficiency to suffer lengthy processing time alongside large energy consumption. High-Performance Computing (HPC) provides apparent efficiency at the expense of energy consumption. However, for remote explorations, the conveyed surveillance and the robotized sensing need faster data analysis with ultimate accuracy to make real-time decisions. In such environments, access to HPC and energy is limited. Therefore, we realize that reducing the number of computations to optimal and maintaining the desired accuracy leads to higher efficiency. This paper demonstrates the semantic segmentation capability of a probabilistic decision tree algorithm, 3D Adapted Random Forest Vision (3DARFV), exceeding deep learning algorithm efficiency at the utmost accuracy.

Researchers have used NMR to measure multi-phase fluid saturation and distribution inside porousmedia of natural rock. However, the NMR signal amplitude suffers reduction with the increase of temperature. The main reason is the Transverse Overhauser Effect, where heating increases the freedom for ionic motion, affecting spinning behavior by having two spins go in two opposite directions to form the Dipolar Coupling. We approach solving NMR thermal effects correction by melding experimentation and numerical simulation method. We use NMR for Cretaceous carbonate rock multi-phase flow research. We conduct time step in-situ temperature measurement for four different sections of the flooding system at the inlet, center, and outlet along the flooding path. In addition, we conduct a temperature measurement at the NMR device radial axis, representing the permanent magnet temperature. We build a 3D cylindrical heat transfer model for the numerical simulator that simulates thermal effectdistribution on the NMR for optimally generating the correction model. The insight provided by the simulator improved the understanding of the thermal distribution at the natural rock core plug to produce a better thermal correction model that meld experimentation and simulation, a method we call Th-CENS.

Each grid block in a 3D geological model requires a rock type that represents all physical and chemical properties of that block. The properties that classify rock types are lithology, permeability, and capillary pressure. Scientists and engineers determined these properties using conventional laboratory measurements, which embedded destructive methods to the sample or altered some of its properties (i.e., wettability, permeability, and porosity) because the measurements process includes sample crushing, fluid flow, or fluid saturation. Lately, Digital Rock Physics (DRT) has emerged to quantify these properties from micro-Computerized Tomography (uCT) and Magnetic Resonance Imaging (MRI) images. However, the literature did not attempt rock typing in a wholly digital context. We propose performing Digital Rock Typing (DRT) by: (1) integrating the latest DRP advances in a novel process that honors digital rock properties determination, while; (2) digitalizing the latest rock typing approaches in carbonate, and (3) introducing a novel carbonate rock typing process that utilizes computer vision capabilities to provide more insight about the heterogeneous carbonate rock texture.

Although capillary and permeability are the two most important physical properties controlling fluid distribution and flow in nature, the interconnectivity function between them was a pressing challenge. Because knowing permeability leads to determining capillary pressure. Geodynamics (e.g., subsurface water, CO2 sequestration) and organs (e.g., plants, blood vessels) depend on capillary pressure and permeability. The first determines how far the fluid can reach, while the second determines how fast the fluid can flow in porous media. They are also vital to designing synthetic materials and micro-objects like membranes and micro-robotics. Here, we reveal the capillary and permeability intertwined behavior function. And demonstrate the unique physical connectors: pore throat size and network, linking capillary pressure and permeability. Our discovery quantifies the inverse relationship between capillary pressure and permeability for the first time, which we analytically derived and experimentally proved.

For defining the optimal machine learning algorithm, the decision was not easy for which we shall choose. To help future researchers, we describe in this paper the optimal among the best of the algorithms. We built a synthetic data set and performed the supervised machine learning runs for five different algorithms. For heterogeneous rock fabric, we identified Random Forest, among others, to be the appropriate algorithm.

Permeability has a dominant influence on the flow behavior of a natural fluid, and without proper quantification, biological fluids (Hydrocarbons) and water resources become waste. During the first decades of the 21st century, permeability quantification from nano-micro porous media images emerged, aided by 3D pore network flow simulation, primarily using the Lattice Boltzmann simulator. Earth scientists realized that the simulation process holds millions of flow dynamics calculations with accumulated errors and high computing power consumption. Therefore, accuracy and efficiency challenges obstruct planetary exploration. To effic­­­iently, consistently predict permeability with high quality, we propose the Morphology Decoder. It is a parallel and serial flow reconstruction of machine learning-driven semantically segmented heterogeneous rock texture images of 3D X-Ray Micro Computerized Tomography (μCT) and Nuclear Magnetic Resonance (MRI). For 3D vision, we introduce controllable-measurable-volume as new supervised semantic segmentation, in which a unique set of voxel intensity corresponds to grain and pore throat sizes. The morphology decoder demarks and aggregates the morphologies’ boundaries in a novel way to quantify permeability. The morphology decoder method consists of five novel processes, which we describe in this paper, these novel processes are (1) Geometrical: 3D Permeability Governing Equation, (2) Machine Learning: Guided 3D Properties Recognition of Rock Morphology, (3) Analytical: 3D Image Properties Integration Model for Permeability, (4) Experimental: MRI Permeability Imager, and (5) Morphology Decoder (the process that integrates the other four novel processes).

Automated image processing algorithms can improve the quality, efficiency, and consistency of classifying the morphology of heterogeneous carbonate rock and can deal with a massive amount of data and images seamlessly. Geoscientists and petroleum engineers face difficulties in setting the direction of the optimum method for determining petrophysical properties from core plug images of optical thin-sections, Micro-Computed Tomography (μCT), or Magnetic Resonance Imaging (MRI). Most of the successful work is from the homogeneous and clastic rocks focusing on 2D images with less focus on 3D and requiring numerical simulation. Currently, image analysis methods converge to three approaches: image processing, artificial intelligence, and combined image processing with artificial intelligence. In this work, we propose two methods to determine the porosity from 3D μCT and MRI images: an image processing method with Image Resolution Optimized Gaussian Algorithm (IROGA); advanced image recognition method enabled by Machine Learning Difference of Gaussian Random Forest (MLDGRF). Meanwhile, we have built reference 3D micro models and collected images for calibration of the IROGA and MLDGRF methods. To evaluate the predictive capability of these calibrated approaches, we ran them on 3D μCT and MRI images of natural heterogeneous carbonate rock. We also measured the porosity and lithology of the carbonate rock using three and two industry-standard ways, respectively, as reference values. Notably, IROGA and MLDGRF have produced porosity results with an accuracy of 96.2% and 97.1% on the training set and 91.7% and 94.4% on blind test validation, respectively, in comparison with the three experimental measurements. We measured limestone and pyrite reference values using two methods, X-ray powder diffraction, and grain density measurements. MLDGRF has produced lithology (limestone and pyrite) volume fractions with an accuracy of 97.7% in comparison to reference measurements.

For defining the optimal machine learning algorithm, the decision was not easy for which we shall choose. To help future researchers, we describe in this paper the optimal among the best of the algorithms. We built a synthetic data set and performed the supervised machine learning runs for five different algorithms. For heterogeneous rock fabric, we identified Random Forest, among others, to be the appropriate algorithm.