Pore-scale imaging and modelling

Deep learning in pore scale imaging and modeling

The study of fundamental flow and transport processes at the pore scale is essential to understanding how the mechanisms affect larger, field-scale, processes that occur in oil and gas recovery, groundwater flow, contaminant transport, and CO2 sequestration. Pore-scale imaging and modeling is one of the techniques used to investigate these fundamental mechanisms. Although extensive development of pore-scale imaging and modeling has occurred recently, some areas still need further advances. In this work, we address two areas: (1) imaging of bulk proppants and proppant-filled fractures under varying loading stress and flow simulation in these systems and (2) nanoparticle (NP) transport modeling in porous media. These are briefly explained below. Rock fracturing, followed by proppant injection, has been used for years to improve oil and gas production rates in low permeability reservoirs and is now routinely used in low-permeability resources such as a shales and tight sands. While field data makes clear the effectiveness of this technique, there is still much room to improve on the science, including how the proppant-filled fracture system responds to changes in loading stress which affect permeability and conductivity. Here, we use high-resolution x-ray computed tomography (XCT) to image two unsaturated rock/fracture/proppant systems under a series of stress levels typical of producing reservoirs: one with shale, one with Berea sandstone. The resulting XCT images were segmented, analyzed for structural and porosity changes, and then used for image-based flow modeling of Stokes flow using both finite element (FEM) and Lattice Boltzmann methods. NPs have been widely used commercially and have the potential to be extensively used in petroleum engineering as stabilizers in enhanced oil recovery operations or as tracers or sensors to detect rock and fluid properties. %In spite of a wide range of applications, many NP transport details are still unknown. In this work, we describe a Lagrangian particle tracking algorithm to model NP transport that can be used to better understand the impact of pore-scale hydrodynamics and surface forces on NP transport. Two XCT images, a Berea sandstone and a 2.5D micromodel, were meshed and used for image-based flow modeling of FEM Stokes flow. The effects of particle size, surface forces, flow rate, particle density, surface capacity, and surface forces mapped to XCT-image based mineralogy were studied.

Assessment of global unconventional oil and gas resources

Important geoscience and environmental applications such as geologic carbon storage, environmental remediation, and unconventional oil and gas recovery are best understood in the context of reactive flow and multicomponent transport in the subsurface environment. The coupling of chemical and microbiological reactions with hydrological and mechanical processes can lead to complex behaviors across an enormous range of spatial and temporal scales. These coupled responses are also strongly influenced by the heterogeneity and anisotropy of the geologic formations. Reactive transport processes can change the pore morphology at the pore scale, thereby leading to nonlinear interactions with advective and diffusive transport, which can strongly influence larger-scale properties such as permeability and dispersion. Therefore, one of the greatest research challenges is to improve our ability to predict these processes across scales (DOE 2007). The development of pore-scale experimental and modeling methods to study reactive processes involving mineral precipitation and dissolution, and biofilm dynamics allows more fundamental investigation of physical behavior so that more accurate and robust upscaled constitutive models can be developed for the continuum scale. A pore-scale model provides fundamental mechanistic explanations of how biogeochemical processes and pore-scale interfacial reactions alter flow paths by pore plugging (and dissolving) under different geochemical compositions and pore configurations. For example, dissolved CO2 during geological CO2 storage may react with minerals in fractured rocks, confined aquifers, or faults, resulting in cementation (and/or dissolution) and altering hydrodynamics of reactive flow. This can be observed in a natural analogue where primary porosity in sandstone is cemented by carbonate precipitates, affecting dissolved CO2 flow paths at the Little Garde Wash Fault, Utah (e.g., Fig. 1a–b). Several other examples demonstrating macroscopic characteristics of calcium carbonate (CaCO3) precipitation in Figure 1 include an elongated concretion along the groundwater flow direction, CaCO3 precipitation along the vertical pathway sealed …

This research utilizes three-dimensional pore-scale imaging of cores from the Paaratte Sandstone in the Otway Basin, obtained through micro-CT, to investigate CO2 storage potential and multiphase flow characteristics. Capillary trapping, a critical mechanism for CO2 storage, was thoroughly examined in the context of CO2-brine displacement in a water-wet system. Using pore-scale modeling, drainage and imbibition curves were generated to determine key parameters, including initial and residual saturations, drainage and imbibition capillary pressures, and relative permeability for CRC-8 rock types. A comprehensive assessment of uncertainties and sensitivities is also included in the analysis. Model-generated curves enabled an exploration of the impact of variables such as wettability and initial water saturation, providing valuable insight into the sensitivity of CO2 trapping mechanisms. Notably, our model allowed for an exploration of sensitivity to contact angle, showing that increased contact angles correspond to a reduction in residual trapping. This study’s results underscore the potential of digital rock technology for enhancing understanding of geological CO2 storage. Through high-resolution imaging, experimental measurements, and pore network modeling, insights into the key mechanisms influencing CO2 trapping and multiphase flow in reservoir rocks were gained, providing a foundation for improved reservoir characterization and modeling.

The work demonstrates results of reservoir properties evaluation using a complex of laboratory and multiscale digital core or digital rock analysis. Rock properties (including relative phase permeabilities) were studied at different scales: from nanometers to meter (whole core). For the first time, cores from Turonian formation were characterized with digital rock analysis, which provided stationary relative permeabilities for gas-water under reservoir conditions. Lab determination of relative permeabilities was rather challenging for some low-permeability samples (<0.02 md), while digital analysis was successful even for them. Gas recovery in a depletion mode from different rock types was studied on a whole core model for different capillary pressures. Such studies are not conducted in the lab.

Geological carbon storage is a key strategy for mitigating climate change, but the long-term stability of trapped CO2 remains uncertain. Transport of dissolved CO2 in the aqueous phase can cause the rearrangement of capillary-trapped CO2 in the pore space, which is called Ostwald ripening. Using high-resolution three-dimensional X-ray imaging, we visualized the in situ evolution of CO2 ganglia in reservoir sandstone during storage and quantified its impact on trapped CO2 saturation. Pore-scale imaging showed the concurrent shrinkage and growth of CO2 ganglia, reduced morphological complexity, and enhanced connectivity, resulting from Ostwald ripening. Ganglia exhibited a size-dependent response: small ganglia dissolved and disappeared, intermediate ones shrank or grew, and large ganglia stabilized with occasional fragmentation. After waiting for 58 h with no flow, originally residual CO2 reconnected, and subsequent brine injection led to a decrease in saturation from 22.8% to 15.6%, consistent with previous estimates based on pore-scale modeling. This work suggests that measurements that ignore the effect of Ostwald ripening overestimate the residual saturation by a factor of approximately a third.

Pore-scale modeling is becoming widely applied in the oil industry. The objective of pore-scale imaging and modeling is to predict the properties of multiphase flow in porous media. Pore-network model construction of geologically realistic samples and determination of rock and two-phase fluid flow properties are described and discussed in this paper. Pore-network models were constructed using published micro-computed tomography images of different rock samples. Effective porosity, average absolute permeability, capillary pressure, tortuosity (in 3 directions), and relative permeability were calculated for 2 types of displacement (drainage and imbibition). Avizo® Software and two-phase code were used for volume rendering of rock samples and calculations of rock and two-phase fluid flow properties, since it is a valuable and reliable tool for prediction of petrophysical properties and has advances in visualization of the pore space. Created 3D models of core samples based on micro-computed tomography scan data will allow oil and oilfield service companies to create a digital core database on a computer instead of storing physical cores in warehouses, which in turn greatly facilitates access to cores for further work with them.

Acid treatment is a common well stimulation technique widely used for both oil and gas wells. However, there is a challenge, that for any given acid solution, rock lithology and permeability, and reservoir conditions, there are optimum values of acid injection volume and rate. Deviation from these optimal values during stimulation treatments reduces the acid job efficiency. We developed a robust and efficient method of identification of the optimum parameters based on digital core approach. The developed workflow includes: a) construction of a digital avatar of a core sample using 3D microCT tomography; b) pore-scale direct reactive flow modeling using a combination of the chemical kinetics/thermodynamics (in assumption of partial local equilibrium) with the method of density functional theory in hydrodynamics (an efficient tool for pore-scale modeling of multiphase flow able to handle different complex physical phenomena); c) core scale simulations in the framework of the Darcy based approach using upscaling from the results of the direct pore-scale simulations; d) input of the obtained parameters into the acidizing simulator to determine optimum acid type, rates, and volumes. We illustrate the developed workflow on example of an optimum injection rate determination in the case of Silurian dolomite dissolution by hydrochloric acid. The pore-scale simulations were performed using 3D microCT models with 2.5 μm/voxel resolution. These simulations allowed to determine the dependence of dolomite dissolution rate on the fluid injection rate and predict the transport properties of damaged rock. The correlations obtained from high resolution simulations were then applied in core-scale modeling of dissolution process using continuous Darcy based model (with 100 μm/voxel resolution). The transport properties of a core were populated using the results of pore scale simulations. Then several core scale simulations of dolomite dissolution with different acid injection rates were performed to obtain numerically the dependence of its influence on the number of pore volumes injected until the breakthrough (PVBT). PVBT dependence on the injection rate in a form of characteristic curve was incorporated into the advanced acidizing simulator. Being calibrated this way, the simulator was then used to model acidizing treatment in a dolomite reservoir with the similar properties as digitally acidized core. Modeling showed that post stimulation skin values are lower and expected wormhole length is bigger when digitally calibrated pore volume to breakthrough (PVBT) curve is used, if compared with modelling of the same treatment with non calibrated acid-rock interaction curves. Consequently, outcomes of this well scale modeling suggest that the use of digitally calibrated PVBT curves results, for this case, in optimization of required acid volume and associated operational footprint. The suggested approach improves the process of obtaining PVBT characteristic curve by application of digital core analysis technique. It allows to test numerous "what if’ scenarios and to evaluate the effect of different factors on mineral dissolution rate at pore scale. This paves the way for improvements in acidizing job design by increasing the consistency between the models used for reactive flow modelling and pore scale heterogeneity of real rocks.

Multiphase flow in porous media is relevant to amount of engineering processes, such as hydrocarbon extraction from reservoir rock, water contamination, CO2 geological storage and sequestration. Pore scale modeling, as an alternative approach to lab measurement, firstly serves as an effective bridge to link the pore scale properties (pore geometry and wettability) and displacement mechanisms to continuous scale multiphase flow in porous media; and secondly allows us to determine essential flow functions, such as capillary pressure and relative permeability curves, which are required for continuous scale modeling. In the literature, three methodologies, Bundle of Capillary Tube Modeling (BCTM), Direct Pore Scale Modeling (DPSM) and Pore Network Modeling (PNM), have appeared to be mostly widely adopted in the investigation of the pore-scale mechanics of fluid-fluid and fluid-solid interactions in porous media by numerical simulation. In this review article, a comprehensive review is provided to show their strengths and weaknesses and to highlight challenges that are faced in modelling of multiphase flow, key challenges include: are contact angle characterization, validation and upscale pore scale findings to core, or even field scale.

For the planning of an EOR project in a major mature oil field in East Malaysia, an extensive Routine and Special Core Analysis (RCA/SCAL) programme has been performed on unconsolidated clastic reservoir rocks. In view of the limited availability of homogeneous core plugs of suitable size for core flooding experiments and for "conventional" SCAL laboratory investigations, a complementary analysis of petrophysical properties was performed based on the acquisition of high resolution 3D Micro-CT (MCT) images, that are used to identify homogeneous sub regions of plugs, and to exclude zones that were damaged during coring. From these regions of undisturbed zones, reliable static reservoir parameters are derived by application of Pore Network Modeling (PNM) techniques. In addition to the generation of "static" parameters (e.g. porosity, permeability, grain size and pore size distributions), PNM simulations of primary drainage and imbibition, and the resulting pc and Krel curves were undertaken. The 3D MCT images were complemented by Fluorescense Microscopy and Field Emission Scanning Electron Microscopy (FESEM) investigations to visualise at a higher resolution possible effects of wettability changes due to cleaning, restoration and flushing of the cores. Variation of the residual oil phase distribution (ROS) due to wettability changes during the cleaning and core handling processes were observed. In view of the planned immiscible Water Alternating Gas (iWAG) process to increase the recovery, water injectivity tests and formation damage (FD) studies were performed in the lab. Potential causes for formation damage (e.g. changes in pore morphology or blocking of pore throats due to fines migration) were visualised by comparing registrations of MCTs at different stages of the flooding and FD experiments. The application of 3D MCTs and PNM proved to be a unique new option to visualise and understand the sensitivities during the handling of cores, and to quantify potential effects of the experimental procedures on the multiphase flow in conventional flooding experiments. As such the Digital Core Analysis and PNM technology is a very quick and robust complementary alternative for the optimised investigation of EOR options, especially in cases where conventional laboratory investigations are limited due to time constraints or due to the status of the core material and resulting potential experimental artefacts. Introduction As part of an extensive core study for a re-development of the EOR programme of a mature oil field in East Malaysia, the "conventional" core analysis program was complemented by acquiring 3D images and applying "Digital Core Analysis" (DCA) /Pore Network Modeling (PNM) methods at different scales with the following main objectives:Visualise the status and monitor potential changes of pore morphology, fluid distributions and wettability during the handling and execution of lab experiments on unconsolidated/friable rock samples.Evaluate and model the impact of sample heterogeneity and wettability changes on petrophysical properties, with special attention to primary drainage and imbibition effects.Provide petrophysical properties for individual distinct rock types at pore scale for heterogeneous or irregular shaped samples that are not suitable or available for RCA/SCAL and flooding experiments.

A multi-scale computational model for multiphase flow in porous media

Subsurface multiphase reactive flow in geologic CO2 storage: Key impact factors and characterization approaches

Fluid–mineral and fluid–rock interfaces are key parameters controlling the reactivity and fate of fluids in reservoir rocks and aquifers. The interface dynamics through space and time results from complex processes involving a tight coupling between chemical reactions and transport of species as well as a strong dependence on the physical, chemical, mineralogical and structural properties of the reacting solid phases. In this article, we review the recent advances in pore-scale imaging and reactive flow modelling applied to interface dynamics. Digital rocks derived from time-lapse X-ray micro-tomography imaging gives unprecedented opportunity to track the interface evolution during reactive flow experiments in porous or fractured media, and evaluate locally mineral reactivity. The recent improvements in pore-scale reactive transport modelling allow for a fine description of flow and transport that integrates moving fluid–mineral interfaces inherent to chemical reactions. Combined with three-dimensional digital images, pore-scale reactive transport modelling complements and augments laboratory experiments. The most advanced multi-scale models integrate sub-voxel porosity and processes which relate to imaging instrument resolution and improve upscaling possibilities. Two example applications based on the solver porousMedia4Foam illustrate the dynamics of the interface for different transport regimes (i.e., diffusive- to advective-dominant) and rock matrix properties (i.e., permeable vs. impermeable, and homogeneous vs. polymineralic). These parameters affect both the interface roughness and its geometry evolution, from sharp front to smeared (i.e., diffuse) interface. The paper concludes by discussing the challenges associated with precipitation processes in porous media, rock texture and composition (i.e., physical and mineralogical heterogeneity), and upscaling to larger scales.

The demand for energy is increasing along with the rise in population. Indonesian people rely on conventional resources such as coal, oil, and natural gas to meet their energy needs. It is estimated that coal can only be exploited for up to 61 years, natural gas for 34 years, and oil for 19 years. Meanwhile, Indonesia possesses unconventional oil and gas resources (e.g., coal bed methane (CBM), tight gas, shale gas and oil, and methane hydrate), estimated to reach 1,800 trillion cubic feet (TCF). These resources are in the exploration stage and have yet to be fully exploited due to technological limitations. Nevertheless, the Indonesian government continues to emphasize the development of this type of energy resource. Therefore, this study conducts a review of the potential of unconventional oil and gas resources in Indonesia, covering characteristics, potential occurrences in Indonesia, exploitation methods, utilization as a source of energy, and opportunities and challenges in their application. The method used is a narrative review based on secondary data by examining papers published in reputable national and international journals in the last ten years. Results show that unconventional oil and gas resources have different characteristics, including permeability, porosity, and depth. CBM can be found at the shallowest depth, followed by tight gas, methane hydrate, and the deepest is shale gas. Potential occurrences of these resources in Indonesia include gas hydrate (858.2 TCF), then shale gas (574.07 TCF), coal-bed methane (453.3 TCF), and shale oil 11.24 million tons. Exploitation can be done in various ways, such as dewatering for CBM, hydraulic fracking for tight and shale, and depressurization for methane hydrate. Once exploited, methane gas can be used for power plants, vehicle fuel, and industrial and household needs. Opportunities and challenges from various aspects, as well as applicable laws in Indonesia, are also discussed. In this light, the contribution of our study is to provide a comprehensive review of the characteristics, location, exploitation methods, opportunities, and challenges of utilizing unconventional oil and gas resources in Indonesia.