Search

Summary Accurate characterization of oilwell cement mechanical properties is key to establishing long-term wellbore integrity. The most widely used method is curing cement in an autoclave, demolding, cutting, and transferring it to a triaxial compression apparatus. The drawback of this traditional technique is that the mechanical properties are not measured under in-situ curing conditions. In this paper, we developed a high-pressure and high-temperature vessel to hydrate cement under downhole conditions and then directly measure cement Young’s modulus and Poisson’s ratio without cooling or depressurization. We validated the setup with water and obtained a reasonable bulk modulus of 2.37 GPa under elevated pressure. We proposed a poroelastic method to calculate cement elastic properties accounting for boundary stiffness and changing pore pressure. We compared the in-situ measurements with traditional triaxial compression tests conducted on the same specimen after retrieval from the vessel. The results show that in-situ measured Young’s modulus is more than double, and the Poisson’s ratio is 20 to 100% higher than that measured by the traditional triaxial method. One mechanism could be that the depressurization and repressurization process in those traditional tests may generate microdefects or induced stresses that weaken cement mechanical properties. Finally, we applied our mechanical properties measurements to cement wellbore integrity analysis by using a thermoporoelastic model. We found that the initial state of stress plays a significant role in maintaining wellbore integrity. With only mechanical properties differences considered, the estimation with traditional measured properties may mistakenly show cement is safe under some pressure and temperature perturbations.

Fluid flow in fractures has the potential to drastically change the economic, environmental, and safety risks associated with a subsurface operation. This work focuses on the analysis of experimental shear fracturing and subsequent shear displacement of a shale specimen that contains multiple pre-existing natural fractures. A triaxial direct shear apparatus with concurrent µCT imaging allowed continuous monitoring of sample permeability and changes in fracture geometry during fracture creation, displacement, and changes in stress. Steady-state, low Reynolds number fluid flow simulations were used to determine the changes of each individual fracture’s hydraulic aperture due to displacement and effective stress. The development of anisotropy in the hydraulic aperture of a generated shear fracture was determined for shear displacements of 0.38, 0.74, and 2.04 mm. At high confining stress (30 MPa), shear displacement significantly reduced the hydraulic aperture of the shear fracture to below flow detectable limits and X-ray imaging resolution (23 × 23 × 23 µm3). The shear generated fracture was found to be more sensitive to changes in effective stress compared to the pre-existing natural fractures. Existing models derived from fluid flow equations accounting for stochastic fracture roughness were able to partially fit the simulated flow properties of the pre-existing fractures, but no model was found to fit the flow properties of the newly formed shear fracture. These results are the first report of simultaneous changes in the geometry and flow properties of surrounding, parallel natural fractures as a shear fracture is generated with varying stress and shear displacement conditions.

We present a novel workflow for forecasting production in unconventional reservoirs using reduced-order models and machine-learning. Our physics-informed machine-learning workflow addresses the challenges to real-time reservoir management in unconventionals, namely the lack of data (i.e., the time-frame for which the wells have been producing), and the significant computational expense of high-fidelity modeling. We do this by applying the machine-learning paradigm of transfer learning, where we combine fast, but less accurate reduced-order models with slow, but accurate high-fidelity models. We use the Patzek model (Proc Natl Acad Sci 11:19731–19736, https://doi.org/10.1073/pnas.1313380110, 2013) as the reduced-order model to generate synthetic production data and supplement this data with synthetic production data obtained from high-fidelity discrete fracture network simulations of the site of interest. Our results demonstrate that training with low-fidelity models is not sufficient for accurate forecasting, but transfer learning is able to augment the knowledge and perform well once trained with the small set of results from the high-fidelity model. Such a physics-informed machine-learning (PIML) workflow, grounded in physics, is a viable candidate for real-time history matching and production forecasting in a fractured shale gas reservoir.

ABSTRACT: Cement initial effective stress is defined as the state of stress after waiting on cement (WOC). Many factors can affect cement’s initial effective stress distribution, including formation environment, cement formulation, and hydrostatic pressure. Our recent experiments using a plug shape setup show that under high-pressure conditions, cement initial effective stress at the cement-casing interface usually drops below hydrostatic pressure but higher than zero, and cement permeability is in the range of nano-Darcy. In a typical cement annulus geometry, we care about initial stress at both the cement-casing interface and the cement-formation interface. The total stress and pore pressure distribution across the annulus may not be isotropic. Therefore, the initial effective stress in the radial direction may not be the same. This non-isotropic initial effective stress distribution places the cement-casing interface and cement-formation interface at different levels of risks under pressure and temperature oscillations caused by carbon injection into the subsurface through wells. 1. INTRODUCTION Cement’s initial state of stress has confused the petroleum industry for decades. The initial state of stress includes total stress, pore pressure, and effective stress. The total stress determines the carbon leakage threshold at the interface, while the effective stress controls the poromechanical response of cement during carbon injection through wells (Meng et al., 2021a). The total stress and effective stress may or may not show consistent behavior considering the effect of cement pore pressure. Accurately understanding initial stress is essential to well plugging & abandonment and maintaining well integrity in carbon sequestration wells. The initial stress has proven to strongly affect cement failure (Saint-Marc et al., 2008; Bois et al., 2012; Meng et al., 2021b). Working on well integrity simulations, most researchers either assume initial effective stress as zero (Zhang et al., 2017), or equal to hydrostatic pressure when cement is in a liquid state (Gray et al., 2009; Nygaard et al., 2014). Such discrepancy in assumption causes great uncertainty in analyzing safe operational windows for maintaining well integrity during carbon injection.

ABSTRACT: Hydraulic fracture branching has been observed in the laboratory and in the field. Our previous studies demonstrate that hydraulic fractures can branch into multiple directions for weak anisotropic in-situ stresses when intermedium injection rate and fluid viscosity are used. However, it is quantitatively unclear how the in-situ stress anisotropy affects hydraulic fracture branching. Understanding this question is important for the optimization of hydraulic fracture design for real-world applications because reservoirs are subject to varying degrees of in-situ stress anisotropy. To address this problem, we designed a transparent true triaxial (TTT) cell that allows us to investigate hydraulic fracture branching under different anisotropic stress conditions with distinct injection parameters and sample heterogeneity. This self-designed TTT cell has a transparent bottom window for us to record hydraulic fracture propagation in real-time. We completed a few benchmark experiments to verify the capability of the TTT cell for hydraulic fracturing tests. The most striking experimental observation is that hydraulic fractures can grow and curve in the direction inclined to the minimum principal stress. This crack growth can stop at some critical points, after which new hydraulic fractures are initiated in the direction normal to the minimum principal stress. The curvature-stop-reinitiation cycle repeats during hydraulic fracturing and generates crack strands. Our experimental finding agrees with the observations at the hydraulic fracturing test site. Therefore, the TTT cell offers unique capabilities for us to experimentally understand complex hydraulic fracture branching and stranding behavior in subsurface conditions. 1. INTRODUCTION Hydraulic fracturing is a key technology that enables economic hydrocarbon production from low-permeability reservoirs (Smith and Montgomery, 2015). The same technology is also used to develop the Enhanced Geothermal System (EGS) for carbon-free geothermal energy extraction from the subsurface (Frash et al., 2014; Fu et al., 2021). Significant efforts have been made to advance our understanding of hydraulic fracture initiation and propagation in rock formations with anisotropic in-situ stresses (Haimson and Fairhurst, 1967; Detournay 2004; Fisher and Warpinski, 2012; Frash et al., 2015 & 2019). However, it remains unclear how do we effectively produce complex hydraulic fracture patterns in the subsurface which is very often subject to anisotropic in-situ stresses.

SummaryIn conventional wellbore-integrity analysis, the cement sheath’s initial state of stress and transient thermoporoelastic effects are often neglected. However, the initial state of stress is prerequisite information for accurately predicting the safe operating conditions that prevent a cemented well from being damaged. In addition, transient thermoporoelastic effects can have a profound effect on when damage will occur. In this paper, we propose a model that includes these effects to predict the safe operating pressures and temperatures that will prevent cement-sheath failure. For the initial state of stress, we proposed an empirical model using measurements. Subsequent stress changes are evaluated by a fully coupled transient thermoporoelastic model to analyze the mechanical behavior of the cement sheath. We predict the safe operating envelope (SOE) for shear, tensile, and debonding cement-sheath failures caused by pressure and temperature perturbations after the cement sets. Our model predicts that pore pressure is a key factor for cement failure, especially for rapid temperature changes. If the formation is low permeability, the transient pore pressures are amplified, causing the risk of damage to increase. Compared with conventional thermoelastic models, the thermoporoelastic model predicts a smaller SOE when heating the internal casing fluid and a larger envelope when cooling the internal casing fluid. Finally, the heating rate was considered with respect to field applications. The heating rate was also considered, and slower heating/cooling rates can prevent damage to the cement sheath. Finally, the thermoporoelastic model was applied to explain several laboratory and field experiments and achieved good matches.

We present hydro-mechanical measurements that characterize shear on natural fractures in schist, amphibolite, and rhyolite specimens from the enhanced geothermal system (EGS) Collab Project's Experiment 1 and 2 sites (E1 and E2) at the Sanford Underground Research Facility. We employed a triaxial direct shear method augmented with X-ray imaging to perform hydroshearing (injection-induced shearing) and mechanical shearing on naturally fractured specimens at in situ stress conditions. Measurements included fracture permeability, strength, stress-dependent aperture, shear dilation, and frictional strength. Results reveal that in situ natural fractures must be permeable, weak, and shear-oriented to be hydrosheared, and only a subset of the observable in situ fractures were suitable. When sheared, the fracture permeability typically increased by a factor of 10 or more and this increase was retained over time. However, shear slip did not always result in permeability increase. High phyllosilicate content associated with exceptionally weak fractures exhibited poor or even decreased permeability after stimulation. These measurements in combination with site data were used to conduct a slip-tendency analysis for different fracture sets, and we selected the top candidate natural fractures for hydroshearing at the EGS Collab sites. We also found that the lower in situ shear stress and stronger fractures at the E2 site make hydroshearing more challenging than at the E1 site. Overall, the methods and analysis used in our work can be applied to any geothermal project to identify in situ joint sets that are best suited for hydroshearing, which in turn can help to optimize well placement and energy production.

Rocks are host to complex fracture networks that are difficult to locate in situ, and yet characterization of these fractures is crucial to predicting the effects of hydraulic stimulation. We analyze three-dimensional hydraulic fracture patterns among varied laboratory experiments to identify recurring geometries. Building on the constitutive tensile and shear fracture modes, we observe examples of offset fracture branching, traversing fracture coalescence, and smooth fracture reorientation as relatively simple structures within complex fracture networks. The evolution of fracture branching, also referred to as stranding, is revealed to be a fundamentally three-dimensional process, in which continued propagation can result in traversing fracture coalescence. Fracture branching, therefore, can create an illusion of unconnected, staggered, and offset hydraulic fracture growth when viewed from a single cross section; meanwhile, these fractures are likely connected through a common fracture surface elsewhere. The fractures are also investigated at a smaller scale, where similar fracture patterns are observed. In the field, these complex patterns are likely to hinder proppant settling, reduce open fracture permeability, create larger fracture surface areas, and lead to increased stimulated reservoir volumes. A balance of stimulation methods to prevent this complexity in some areas and exploit it in others could be key to improving recovery from oil and gas resources, improving geothermal energy efficiency, and optimizing disposal of waste water or CO2 via injection wells. Representation of these complex structures is needed for accurate modeling predictions for reservoir management.

Recent advances in hydraulic fracturing, in conjunction with horizontal drilling, have enabled large-scale extraction of natural gas and oil from shale formations. Despite its advances and enormous economic benefits, opportunities remain to increase hydraulic fracturing efficiency and minimize potential environmental impacts. This review specifically examines three key themes associated with development and utilization of hydraulic fracturing fluids: 1) characteristics and behavior of fracturing fluids, 2) understanding and predicting migration and fate of fracturing fluids, 3) technologies to reduce environmental impact of fracturing fluids. The paper discusses key and new techniques and findings on rheology of hydrogel-based fluids, high fidelity simulation of propagation transport, potential environmental impacts, geosynthetics in mitigating contamination, and greener fracturing fluids. It is indicated that future development relies on advances in understanding of physical processes, modeling capabilities, and monitoring techniques.

Geologic CO2 sequestration (GCS) has received high-level attention from the global scientific community as a response to climate change due to higher concentrations of CO2 in the atmosphere. However, GCS in saline aquifers poses certain risks including CO2/brine leakage through wells or non-sealing faults into groundwater or to the earth’s surface. Understanding crucial reservoir parameters and other geologic features affecting the likelihood of these leakage occurrences will aid the decision-making process regarding GCS operations. In this study, we develop a science-based methodology for quantifying risk profiles at geologic CO2 sequestration sites as part of US DOE’s National Risk Assessment Partnership (NRAP). We apply NRAP tools to a field scale project in a fractured saline aquifer located at Kevin Dome, Montana, which is part of DOE’s Big Sky Carbon Sequestration Partnership project. Risks associated with GCS injection and monitoring are difficult to quantify due to a dearth of data and uncertainties. One solution is running a large number of numerical simulations of the primary CO2 injection reservoir, shallow reservoirs/aquifers, faults, and wells to address leakage risks and uncertainties. However, a full-physics simulation is not computationally feasible because the model is too large and requires fine spatial and temporal discretization to accurately reproduce complex multiphase flow processes. We employ the NRAP Integrated Assessment Model (NRAP-IAM), a hybrid system model developed by the US-DOE for use in performance and quantitative risk assessment of CO2 sequestration. The IAM model requires reduced order models (ROMs) developed from numerical reservoir simulations of a primary CO2 injection reservoir. The ROMs are linked with discrete components of the NRAP-IAM including shallow reservoirs/aquifers and the atmosphere through potential leakage pathways. A powerful stochastic framework allows NRAP-IAM to be used to explore complex interactions among a large number of uncertain variables and to help evaluate the likely performance of potential sequestration sites. Using the NRAP-IAM, we find that the potential amount of CO2 leakage is most sensitive to values of permeability, end-point CO2 relative permeability, hysteresis of CO2 relative permeability, capillary pressure, and permeability of confining rocks. In addition to demonstrating the application of the NRAP risk assessment tools, this work shows that GCS in the Kevin Dome has a higher probability of encountering injectivity limitations during injection of CO2 into the Middle Duperow formation than previous studies have calculated. Finally, we estimate very low risk of CO2 leakage to the atmosphere unless the quality of the legacy well completions is extremely poor.

Widespread seafloor methane venting has been reported in many regions of the world oceans in the past decade. Identifying and quantifying where and how much methane is being released into the ocean remains a major challenge and a critical gap in assessing the global carbon budget and predicting future climate [C. Ruppel, J. D. Kessler. Rev. Geophys. 55, 126–168 (2017)]. Methane hydrate ([Formula: see text]) is an ice-like solid that forms from methane–water mixture under elevated-pressure and low-temperature conditions typical of the deep marine settings ([Formula: see text] 600-m depth), often referred to as the hydrate stability zone (HSZ). Wide-ranging field evidence indicates that methane seepage often coexists with hydrate-bearing sediments within the HSZ, suggesting that hydrate formation may play an important role during the gas-migration process. At a depth that is too shallow for hydrate formation, existing theories suggest that gas migration occurs via capillary invasion and/or initiation and propagation of fractures (Fig. 1). Within the HSZ, however, a theoretical mechanism that addresses the way in which hydrate formation participates in the gas-percolation process is missing. Here, we study, experimentally and computationally, the mechanics of gas percolation under hydrate-forming conditions. We uncover a phenomenon—crustal fingering—and demonstrate how it may control methane-gas migration in ocean sediments within the HSZ.

Leakage in carbon storage and hydrocarbon wells continues to be an area of concern in the development and abandonment of reservoirs. Industry need for a leakage remediation sealant that can perform in systems beyond the capability of cement squeezes has driven the development of a CO2/pH activated "smart" gel. Exploratory laboratory tests and a mock field scale well test were performed to determine the effectiveness of the smart gel. Control of the smart gel particle size distribution was demonstrated through batch synthesis experiments. Microfluidic experiments show some of the mechanisms leading to the successful sealing of an engineered fracture system. Initial and subsequent testing of the deployed smart gel in a leaky mock well completion proves the effective scale up of the smart gel sealing capability and can further drive wider adoption of this unique technology.

The transport of particles and fluids through multichannel microfluidic networks is influenced by details of the channels. Because channels have micro-scale textures and macro-scale geometries, this transport can differ from the case of ideally smooth channels. Surfaces of real channels have irregular boundary conditions to which streamlines adapt and with which particle interact. In low-Reynolds number flows, particles may experience inertial forces that result in trans-streamline movement and the reorganization of particle distributions. Such transport is intrinsically 3D and an accurate measurement must capture movement in all directions. To measure the effects of non-ideal surface textures on particle transport through complex networks, we developed an extended field-of-view 3D macroscope for high-resolution tracking across large volumes ($$25\,\hbox {mm} \times 25\,\hbox {mm} \times 2\,\hbox {mm}$$ 25 mm × 25 mm × 2 mm ) and investigated a model multichannel microfluidic network. A topographical profile of the microfluidic surfaces provided lattice Boltzmann simulations with a detailed feature map to precisely reconstruct the experimental environment. Particle distributions from simulations closely reproduced those observed experimentally and both measurements were sensitive to the effects of surface roughness. Under the conditions studied, inertial focusing organized large particles into an annular distribution that limited their transport throughout the network while small particles were transported uniformly to all regions.

Current oil recovery methods in hydraulically fractured shale reservoirs have a low recovery efficiency of about 10%. The objective of this work is to investigate the effectiveness of nitrogen and supercritical carbon dioxide in the huff-and-puff method for enhanced oil recovery as a means to re-energize reservoirs and improve recovery rates. We conduct direct visualization experiments with a microfluidic system to reveal the mechanisms and to quantify the recovery rates of oil from fracture networks. We compared the effectiveness of water, nitrogen and supercritical carbon dioxide at reservoir conditions in a process mimicking the huff-and-puff method in both dead-end and connected fracture systems. The microfluidic chips were made of glass and placed in a confining pressure system pressurized to 10 MPa, 50 °C. The system was allowed to equilibrate, and then depressurized to simulate huff-and-puff oil recovery. Fluorescence microscope images were continuously taken to visualize and calculate residual oil saturation as a function of pressure drawdown. As the system was depressurized from 10 MPa, gas exsolution from the oil liquid phase, including bubble nucleation, growth, and coalescence, appeared to be the main driver for mobilizing oil from the fracture networks. Injection of supercritical CO2 resulted in the highest recovery rate with an average end-point recovery of about 90% in the connected fracture network and 60% in the dead-end fracture network. N2 has lower solubility in oil and hence showed a lower recovery rate of 40% in the connected fracture network and 25% in the dead-end fracture network. Injection of water had no effect on oil mobilization since water is insoluble, immiscible and incompressible. The main mechanism of enhanced recovery was gas exsolution from the liquid phase as pressure was decreased below the bubble point pressure. Because the gas was distributed throughout the oil phase, bubble nucleation, growth, coalescence, and elongation occurred throughout the fracture network. Expansion of the gas forced oil out of the network through piston displacement in continuous oil areas and film flow in the dispersed oil areas. Bubbles began as spheres and grew until they touched the fracture walls where they elongated along the fracture length. The bubble growth rate depended on local mass transfer from liquid to gas phase and gas volume expansion due to pressure drop. The efficiency of the huff-and-puff process is dependent on the solubility and miscibility of the injection fluid with oil. High gas solubility allows for more bubble nucleation, growth and expansion during the depressurization cycle.

A three-dimensional multiphysics lattice discrete particle model (M-LDPM) framework is formulated to investigate the fracture permeability behavior of shale. The framework features a dual lattice system mimicking the mesostructure of the material and simulates coupled mechanical and flow behavior. The mechanical lattice model simulates the granular internal structure of shale, and describes heterogeneous deformation by means of discrete compatibility and equilibrium equations. The network of flow lattice elements constitutes a dual graph of the mechanical lattice system. A discrete formulation of mass balance for the flow elements is presented to model fluid flow along cracks and intact materials. The overall computational framework is implemented with a mixed explicit–implicit integration scheme and a staggered coupling method that makes use of the dual lattice topology enabling the seamless two-way coupling of the mechanical and flow behaviors. The proposed model is used for the computational analysis of shale fracture permeability behavior by simulating triaxial direct shear tests on Marcellus shale specimens under various confining pressures. The simulated mechanical response is calibrated against the experimental data, and the predicted permeability values are also compared with the experimental measurements. Furthermore, the paper presents the scaling analysis of both the mechanical response and permeability measurements based on simulations performed on geometrically similar specimens with increasing size. The simulated stress strain curves show a significant size effect in the post-peak due to the presence of localized fractures. The scaling analysis of permeability measurements enables prediction of permeability for large specimens by extrapolating the numerical results of small ones.

This study analyzes the dynamics and mechanisms of the interactions of carbonated brine with hydrated-Portland-cement; in particular, the study focuses on self-sealing, a process whereby hydrated-Portland cement reacts with carbonated brine to for silica and calcium carbonate in sufficient quantities to seal the flow pathway. The analysis is based on a comprehensive set of reactive-transport simulations that explore the complex coupled dynamics between the fluid flow and mineral reactions that underlie self-sealing, and it relies heavily on the synthesis of the extensive body of work on wellbore integrity that has been conducted over the past decade. The analysis explores a large chemical and mineralogical diversity and a wide range in physical conditions and flow regimes, attempting to assess the robustness of the analysis. Self-sealing conditions arise over a wide range in cement properties and reservoir conditions. Although some properties and conditions promote a stronger self-sealing response, self-sealing occurs for a wide range of Ca:Si ratios in cement and for various reservoir fluid compositions. Self-sealing conditions move along a wellbore proportional to the flux of the leaking carbonated brine, and the reaction zone spreads out proportional to the fluid velocity, where volumetric flux and velocity are related by porosity (flux = velocity * porosity). However, self-sealing conditions can be maintained in a specific section of a wellbore by controlling the pressure drive and/or effective wellbore permeability, which in turn can limit the flux and velocity of any leaking fluid. Finally, the phases produced by hydrating Portland cement represent a carbonic cement that will react with a carbonated brine to produce end products (calcium carbonate and silica) that can maintain integrity in the presence of carbonic acid. Moreover, the attributes that make hydrated Portland cement phases a carbonic cement are required for self-sealing.

CO2 injection into saline aquifers for sequestration will initially result in buoyant supercritical (sc)CO2 trapped beneath the caprock seal. During this period, there is risk of CO2 migration out of the reservoir along wellbore defects or fracture zones. Dissolution of the scCO2 plume into brine results in solubility trapping and reduces this risk, but based on diffusion alone, this mechanism could take thousands of years. Gravitational (density-induced) mixing of CO2-saturated brine is shown to significantly accelerate this process in computational studies, but few experimental efforts have confirmed the phenomenon. Here, constant-pressure, 3-dimensional bench-scale experiments used the mass of added water to quantify the mass transfer of scCO2 into water-saturated porous media at 40–90 °C and 20 MPa, with Rayleigh numbers from 2093 to 16256. Experiments exhibit a period of 7–35X enhancement in mass transfer rates over diffusion, interpreted as gravitational mixing. Convective CO2 flux ranges from 1.6 × 10−2 to 4.8 × 10−3 mol s−1 m−2 in the experiments. Results are used to benchmark a computational model using PFLOTRAN. Experiments show an early diffusive onset period that is shorter with rates much higher than predicted by models and observed in analog experiments. Both experiments and models show convective mixing periods and similar overall rates of CO2 mass transfer.

Motivated by a triaxial coreflood experiment with a sample of Utica shale where an abrupt jump in permeability was observed, possibly due to the creation of a percolating fracture network through the sample, we perform numerical simulations based on the experiment to characterize how the effective permeability of otherwise low-permeability porous media depends on fracture formation, connectivity, and the contrast between the fracture and matrix permeabilities. While a change in effective permeability due to fracture formation is expected, the dependence of its magnitude upon the contrast between the matrix permeability and fracture permeability and the fracture network structure is poorly characterized. We use two different high-fidelity fracture network models to characterize how effective permeability changes as percolation occurs. The first is a dynamic two-dimensional fracture propagation model designed to mimic the laboratory settings of the experiment. The second is a static three-dimensional discrete fracture network (DFN) model, whose fracture and network statistics are based on the fractured sample of Utica shale. Once the network connects the inflow and outflow boundaries, the effective permeability increases non-linearly with network density. In most networks considered, a jump in the effective permeability was observed when the embedded fracture network percolated. We characterize how the magnitude of the jump, should it occur, depends on the contrast between the fracture and matrix permeabilities. For small contrasts between the matrix and fracture permeabilities the change is insignificant. However, for larger contrasts, there is a substantial jump whose magnitude depends non-linearly on the difference between matrix and fracture permeabilities. A power-law relationship between the size of the jump and the difference between the matrix and fracture permeabilities is observed. The presented results underscore the importance of fracture network topology on the upscaled properties of the porous medium in which it is embedded.

The process of cement slurry curing into a solid has a direct effect on the initial state of stress in a composite concrete and steel structure. Failure models for critical infrastructure and wellbores could benefit from measurements of this initial stress because it greatly influences structural performance. Here, we present measurements of the initial stress state (total stress, effective stress, and pore pressure) of multiple cement formulations cured under a high-pressure of 40 MPa within a semi-rigid steel pipe. Our results show that cements cured in this restrained environment exhibit dense microstructures with low permeability (nano-Darcy), develop an anisotropic stress state greater than zero and less than the curing pressure, and undergo a staged curing process where hydration and mechanical behaviors evolve from a slurry into a pre-stressed porous medium. Both drained and undrained scenarios were investigated, revealing a feedback from the semi-rigid boundary towards improved cement performance.

Caprock leakage is of crucial concern for environmentally and economically sustainable development of carbon dioxide sequestration and utilization operations. One potential leakage pathway is through fractures or faults that penetrate the caprock. In this study, we investigate the permeability induced by fracturing initially intact Marcellus shale outcrop specimens at stressed conditions using a triaxial direct-shear method. Measurements of induced permeability, fracture geometry, displacement, and applied stresses were all obtained at stressed conditions to investigate the coupled processes of fracturing and fluid flow as may occur in the subsurface. Fracture geometry was directly observed at stressed conditions using X-ray radiography video. Numerical simulation was performed to evaluate the stress distribution developed in the experiments. Our experiments show that permeability induced by fracturing is strongly dependent on the stresses at which the fractures are created, the magnitude of shearing displacement, and the duration of flow. The strongest permeability contrast was observed when comparing specimens fractured at low stress to others fractured at higher stress. Measureable fracture permeability decreased by up to 7 orders of magnitude over a corresponding triaxial confining stress range of 3.5 MPa–30 MPa. These results show that increasing stress, depth, and time are all significant permeability inhibitors that may limit potential leakage through fractured caprock.

Leakage from damaged caprock will depend on fracture properties in which we expect a difference between brittle and ductile deformations modes. We conducted experiments using integrated x-ray tomography and triaxial coreflood methods to characterize fracture-permeability behavior of shale as a function of the confining pressure at which fractures are created as a means of exploring the brittle-ductile transition. The experiments were conducted using a triaxial direct-shear coreflood system that creates through-going fractures for permeability measurements. Fracture characteristics were examined using 25-μm resolution x-ray computed tomography in both ex situ and in situ modes. At low confining pressure (3.5 MPa), permeability of fractured Utica shale was strongly dependent on fracture interactions with bedding planes and varied from 10s to 100s of mD (calculated with respect to the total sample area). Fracture systems propagating perpendicular to bedding planes resulted in tortuous paths with lower permeability. Fracture systems propagating parallel to bedding had substantially elevated permeability with some ex situ apertures greater than 500 μm. This behavior contrasted with fractures developed at high confining pressure (22 MPa). In situ x-ray tomography showed that localized deformation occurred on sub-resolvable fractures (less than 25-μm aperture) much smaller than apertures formed at low confining pressure. Permeability of the fractures formed at high confining pressure was less than 0.1 mD and about 3 orders of magnitude lower than the low confining pressure case.

An import factor for long-term storage is the integrity of wells penetrating the carbon storage reservoir. Well integrity will play a crucial role in establishing leakage risk in areas where there is a high density of existing (legacy) wells that could intersect operations including depleted petroleum fields where enhanced oil recovery or carbon storage will occur. The objective of this study was to develop methods that establish the baseline flow parameters from individual measurements in five wells in Wyoming. This paper describes the methods used to collect and analyze the data. The methods included isolation logging tools, in situ point and average permeability measurements, and laboratory tests on samples collected in the wells. The log results collected in the wells, considered in conjunction with the core measurements, indicate that interfaces and/or problems with cement placement related to eccentering provide preferential pathways for fluids, which increase the effective permeability of the barrier several orders of magnitude above that of intact cement. © 2017 Society of Chemical Industry and John Wiley & Sons, Ltd.

Mineral reactions during CO2 sequestration will change the pore-size distribution and pore surface characteristics, complicating permeability and storage security predictions. In this paper, we report a small/wide angle scattering study of wellbore cement that has been exposed to carbon dioxide for three decades. We have constructed detailed contour maps that describe local porosity distributions and the mineralogy of the sample and relate these quantities to the carbon dioxide reaction front on the cement. We find that the initial bimodal distribution of pores in the cement, 1–2 and 10–20 nm, is affected differently during the course of carbonation reactions. Initial dissolution of cement phases occurs in the 10–20 nm pores and leads to the development of new pore spaces that are eventually sealed by CaCO3 precipitation, leading to a loss of gel and capillary nanopores, smoother pore surfaces, and reduced porosity. This suggests that during extensive carbonation of wellbore cement, the cement becomes les...

Stress-dependent fracture permeability could significantly affect oil and gas production, underground CO2 storage, and deep waste disposal containment. Yet, measurements to characterize these effects are lacking. To be most relevant to field conditions, measurements of stress-dependent fracture permeability should target pre-existing fracture features, be completed at high-stress conditions, and enable visualization of the fracture geometry to characterize the in-situ aperture profiles. Here, we employ triaxial direct-shear (TDS) tests, integrated with real-time X-ray imaging, to measure the stress-dependent fracture permeability of Marcellus shale and ‘Western Texas’ shale at subsurface conditions. Our results indicate that the studied shale fractures are characterized by large variation of compressibility factor. We also found that surface roughness can reduce fracture permeability more severely than what literature values would suggest. We then implemented our stress-dependent fracture permeability measurements in a formation-linear flow model to investigate the significance for shale gas production. The model uses the MIP-3H well at the Marcellus Shale Energy and Environment Laboratory (MSEEL) as a base case. The model results reveal a narrow stress-dependent permeability range over which fracture closure can be critical for production and that MSEEL appears to be close to this critical condition, which could be important for reservoir pressure management.

Unplugged, abandoned wells are the greatest concern for potential open leakage pathways to groundwater and the atmosphere from geologic reservoirs storing CO2. Such wells will blow out when encountered by a buoyant fluid. Historical data on well blowouts during steam-enhanced oil recovery in California’s Oil and Gas District 4 provides perspective on blowout frequency during immiscible fluid injection. Well blowout rates are often characterized as events-per-well years. This implies that well integrity decreases with well age. This is at odds with an alternative conceptual model that suggests well leakage is predominantly due to initial well integrity defects. The rate of inactive well blowouts releasing injected steam in California Oil and Gas District 4 on a well-years basis declines contrary to the aging model. In contrast, the rate on the basis of wells intersected by steam is relatively constant. This is more consistent with initial defects rather than subsequent degradation creating the integrity defects that allow a blowout to occur. The initial defect hypothesis is further supported by the observation that blowouts typically occur one year or less after well shut in, plugging or the start of steam injection nearby. The well defect perspective implies that risk assessments for buoyant injected fluids, such as steam and CO2, should use a blowout rate per wells encountered by a plume rather than per well years to characterize acute well leakage risk. The predominance of inactive well blowouts due to initial defects rather than aging implies that blowout events are likely to occur earlier at the plume edge rather than later toward the plume center.

The challenge of characterizing subsurface fluid flow has motivated extensive laboratory studies, yet fluid-flow through rock specimens in which fractures are created and maintained at high-stress conditions remains under-investigated at this time. The studies of this type that do exist do not include in situ fracture geometry measurements acquired at stressed conditions, which would be beneficial for interpreting the flow behavior. Therefore, this study investigates the apparent permeability induced by direct-shear fracture stimulation through Utica shale (a shale gas resource and potential caprock material) at high triaxial-stress confinement and for the first time relates these values to simultaneously acquired in situ X-ray radiography and microtomography images. Change in fracture geometry and apparent permeability was also investigated at additional reduced stress states. Finite element and combined finite discrete element modeling were used to evaluate the in situ observed fracturing process. Results from this study indicate that the increase in apparent permeability through fractures created at high-stress (22.2 MPa) was minimal relative to the intact rock (less than 1 order of magnitude increase) while fractures created at low stress (3.4 MPa) were significantly more permeable (2 to 4 orders of magnitude increase). This study demonstrates the benefit of in situ X-ray observation coupled with apparent permeability measurement to analyze fracture creation in the subsurface. Our results show that the permeability induced by fractures through shale at high stress can be minor and therefore favorable in application to CO2 sequestration caprock integrity but unfavorable for hydrocarbon recovery from unconventional reservoirs.

•Reduced-order model development for CO2 and brine leakage along abandoned wellbores.•Coupled leakage effects are incorporated into ROMs intended for decoupled systems models.•Octree mesh refinement and sculpting reduces nodes and minimizes mesh effects.

We present constant-flow experiments of high-pressure CO2-saturated fluid through fracture channels in Portland cement with three different fracture geometries and with three different flow rates. The experiments were conducted in etched cement wafers within a microfluidics device. The evolution of pH was observed optically using phenolphthalein dye and changes in channel volume due to dissolution and/or precipitation were characterized with profilometry and scanning electron microscopy. Abundant precipitation for all three geometries was observed in the low flow-rate experiments, while only dissolution was observed in the high-flow rate experiments. The results bracket self-sealing versus self-opening behavior. The observed functional relationship among flow rate, fracture geometry and aperture in relation to self-sealing is consistent with a diverse set of experiments in the literature conducted under widely varying conditions and with behavior predicted in recent numerical models (Brunet et al., 2016; Cao et al., 2015; Guthrie et al., 2018; Iyer et al., 2017). The results confirm that self-sealing is a quantifiable behavior in ordinary Portland cement systems that is favored at low flow rates and in small-aperture or small hydraulic diameter fracture systems with clear limits imposed by dissolution-dominated conditions of high flow rates.