In recent years, the development of oil and gas from shale has proceeded quickly in the world due to the application of multi-stage fracturing technology in horizontal wells. It is imperative to study the poroelastic characteristics of the rock for modeling the performance of rock under in-situ conditions, thus ensuring the success of hydraulic fracturing. Biot's coefficient is one of the key poroelastic parameters for calculating the effective stress for creating artificial fractures in the shale formations. In this study, we propose anew method to measure the Biot's coefficient. Our method simplified the measuring procedures to obtain the Biot's coefficient by controlling the confining pressure, which isused to maintain the volume of the sample, while altering the pore pressure. Shales amples recovered form Bakken formation in Willistion Basin is tested using this method. The results of our experiments show that the Biot's coefficient of Bakken samples obtained from horizontal drilling and vertical drilling are significantly different from each other. This significant difference of Biot's coefficient with different drilling-direction provides scientists and engineers a solid base for in-situ stress analysis during multi stage hydraulic fracturing and reservoir depletion due to production.
Hydraulic fracturing technique has been widely applied in the enhanced geothermal systems, to increase injection rates for geologic sequestration of CO2, and most importantly for the stimulations of oil and gas reservoirs, especially the unconventional shale reservoirs. One of the key points for the success of hydraulic fracturing operations is to accurately estimate the redistribution of pore pressure and stresses around the induced fracture and predict the reactivations of pre-existing faults. The fracture extension as well as pore pressure and stress regime around it are affected by: poro- and thermoelastic phenomena as well as by fracture opening under the combined action of applied pressure and in-situ stress. A couple of numerical studies have been done for the on this for the purpose of analyzing the potential for fault reactivation resulting from pressurization of the hydraulic fracture. In this work, a comprehensive analytical model is constructed to estimate the stress and pore pressure distribution around an injection induced fracture from a single well in an infinite reservoir. The model allows the leak-off distribution in the formation to be three-dimensional with the pressure transient moving ellipsoidcally outward into the reservoir with respect to the fracture surface. The pore pressure and the stress changes in three dimensions at any point around the fracture caused by thermo- and poroelasticity and fracture compression are investigated. Then, the problem of constant water injection into a hydraulic fracture in Barnett shale is presented. In particular, with Mohr-Coulomb failure criterion, we calculate the fault reactivation potential around the fracture. This study is of interest in interpretation of micro-seismicity in hydraulic fracturing and in assessing permeability variation around a stimulation zone, as well as in estimation of the fracture spacing during hydraulic fracturing operations.
Luo, D. (University of Alaska Fairbanks) | Chen, G. (University of Alaska Fairbanks) | Patil, S. (University of Alaska Fairbanks) | Abhijit, D. (University of Alaska Fairbanks) | Santanu, K. (University of Alaska Fairbanks)
Wellbore stability is of critical importance in all drilling operations. Wellbore instability may cause stuck pipes, lost circulation, and/or collapse of the wellbore, resulting in high drilling cost and significant time loss. In this study, computer simulations using FLAC (Fast Lagrangian Analysis of Continua) were conducted on the stability of horizontal wells in shale formations. Laboratory-tested geomechanical properties of seven shale samples and in-situ stress conditions collected from the literature were used. Computer simulations were carried out to estimate minimum downhole pressures for maintaining wellbore integrity in each type of shale formation under different states of in-situ stresses. The results showed that the minimum downhole pressure to maintain wellbore stability is positively related to stress differentiation and pore pressure, and negatively related to internal frictional angle and cohesion of the surrounding rock. The determination of the minimum downhole pressure from the regression analyses may serve as a basis for engineers to quickly select proper mud density when drilling horizontal wells in potentially problematic rock formations, particularly shale formations.
Hydraulic fracturing re-distributes pore pressure and stresses inside rock and causing failure by fracture initiation and/or activation of discontinuities such as natural fractures or layering boundaries. The clear result of this process would be enhancement of the formation permeability. In this paper, poroelastic numerical method is employed to investigate interactions of hydraulic fractures and porous rock. Besides, evolution of potential failure (microseismic events) during hydraulic stimulation is studied. The model uses indirect boundary element method. Temporal variations and pressure-dependent leak-off, hydro mechanical response of porous matrix, fluid flow in matrix, couplings of matrix volumetric deformation and pore fluid dissipation, and hydraulic fractures interaction are taken into account. Results clearly show the modification/redirection of principal stresses around pressurized hydraulic fracture. It also shows that modified stresses cause failure around the fracture tip which generally covers a bigger area than the fracture itself and could results in an overestimation of the stimulated reservoir volume. Then, pressurization of multiple parallel fractures studied. As expected, it is found that fracture geometry and the distance between hydraulic fractures are the most important factors in modifying the stress state and pore pressure and consequently extent of failure region. It was also observed that the opening of a fracture induces shear stresses on adjacent fractures. The SIF for pressurized cracks was calculated for Mode I and Mode II, and it was shown that when the distance between hydraulic fractures increases, the Mode I SIF also increases and the Mode II SIF decreases.'ép.
Often, a key factor in the successful hydraulic fracture stimulation of unconventional reservoirs is the opening or shearing (and later extension) of natural fractures or weakness planes around a created hydraulic fracture. The behavior of natural fractures, or weakness planes, in response to hydraulic fracture stimulation can be complicated. Furthermore, the stimulation of these fractures and weakness planes is dependent on several critical, in-situ conditions that can increase (or decrease) the contribution of natural fractures and weakness planes to well production. The optimal economic completion, then, requires considering these factors during both stimulation design and post-stimulation evaluations.
The simplistic, and traditional, assumption that hydraulic fractures are bi-wing, planar and symmetric around the wellbore has tended to bias the interpretation of different aspects of the stimulation process. However, hydraulic fracture monitoring methods, such as microseismicity, pressure evaluations, and the coring through of hydraulic fractures, have confirmed the complex nature of fracture propagation in unconventional plays, often due to the presence of natural fractures and weakness planes. Therefore, an improved consideration of natural fracture and weakness plane behavior during hydraulic fracturing will result in a better understanding of fluid treating pressures and hydraulic fracture geometry, which will help lead to more accurate estimations of production for unconventional plays.
In this paper, the results of an extensive parametric study of in-situ stress conditions, in-situ pressure, natural fracture mechanical properties (cohesion and friction angle) and characteristics (joint orientation and initial aperture), and different operating conditions (single stage, simultaneous hydraulic fracture stages, and sequential hydraulic fracture stages) on injection (net) pressure behavior is presented. The results were generated using a 2-D distinct element model and capture the important role that, for example, initial natural fracture aperture and in-situ pressure play in the development of hydraulic fracture injection pressures in unconventional reservoirs.
In complex reservoirs with geological discontinuities such as faults, the risks of seal breach increased by injection and depletion. One of the phenomena which can cause serious production losses and environmental accidents is the reactivation of geological faults. This occurs in association with variations in stresses induced in the formation, which can be high enough to reactivate the faults and significantly modify reservoir behavior. In such cases, fluid can migrate to shallow depths compromising the integrity of the entire fault plane and generating one of the most critical situations in the oil industry. This work investigates through the finite element method the phenomenon of reactivation of geological faults. Four numerical formulations are used for this purpose: (1) explicit representation of the fault with interface elements, (2) continuum model with fault strength criterion applied to stresses on the fault plane, (3) equivalent anisotropic continuum with a fault guidance plane and (4) equivalent continuum with yield criterion using equivalent homogeneous properties for the reservoir-fault system. These methods are applied to a synthetic two-dimensional finite element model comprising a normal fault. Comparative results are shown.
The long-term thermal-hydraulic-mechanical response of a generic salt repository for high-level nuclear waste is investigated. In-drift emplacement of the waste packages and subsequent backfill of the drifts with crushed salt are assumed. The aim of this research is to evaluate the long-term integrity of the natural salt and consolidated backfill barriers. For this purpose, we use an updated version of the TOUGH-FLAC simulator, able to deal with large strains and creep. The simulator also includes state-of-the-art constitutive relationships and coupling functions. The Lux/Wolters constitutive model for natural salt is used. The simulations are two-way coupled and include the stages of excavation, waste emplacement, backfilling and a post-closure period of 100,000 years. The simulation results show that the excavation damaged zone is healed within the first years and that the backfill reconsolidation is complete within the first decades. Depending on the magnitude of the pore pressure relative to the minimum principal stress, hydraulic damage within the host rock may occur at a larger scale. The comparison of coupled simulation results with those issued from a case that disregards the mechanical processes shows the necessity to account for the mechanical effect in order to accurately predict the long-term evolution of the barriers.
Blanco, Martín L. (Lawrence Berkeley National Laboratory) | Rutqvist, J. (Lawrence Berkeley National Laboratory) | Birkholzer, J.T. (Lawrence Berkeley National Laboratory) | Wolters, R. (Clausthal University of Technology) | Rutenberg, M. (Clausthal University of Technology) | Zhao, J. (Clausthal University of Technology) | Lux, K.-H. (Clausthal University of Technology)
The long-term thermal-hydraulic-mechanical response of a generic salt repository for high-level nuclear waste is investigated using the TOUGH-FLAC simulator, developed at Lawrence Berkeley National Laboratory, and the FLAC-TOUGH simulator, developed at Clausthal University of Technology. Although these sequential simulators rely on the same flow and geomechanics software, they are based on different numerical schemes. One of the aims of using two different approaches to model the same scenario is to gain reliability on the results obtained. The two simulators include state-of-the-art constitutive relationships and coupling functions. The generic scenario studied assumes in-drift emplacement of the waste packages and subsequent backfill of the drifts with crushed salt. The Lux/Wolters constitutive model for natural salt is used. The simulations are two-way coupled and include the stages of excavation, waste emplacement, backfilling and post-closure. This work has been performed within the framework of a collaboration effort between Lawrence Berkeley National Laboratory and Clausthal University of Technology. Although the predictions presented in this paper cover a post-closure period of 100 years, it is intended to continue the benchmark until 100,000 years. The results obtained so far provide confidence in the capabilities of the two simulators to evaluate the barriers integrity over the long-term.
Deep saline aquifers have a great potential for geologic carbon dioxide (CO2) sequestration and proper assessment of host and cap rock is needed to guarantee that the procedure is safe. Temperatures and pressures at which most of the possible host rocks exist dictate that CO2 is present in a supercritical condition, having both gas and liquid properties. Hence, rock-fluid interaction has to be studied and measurements of poroelastic parameters are necessary. Sandstone formations are mostly considered as the possible host rock. However, in some countries only calcite-rich formations can satisfy the requirements for safe geologic CO2 sequestration.
This paper deals with measurements of poroelastic parameters of calcarenite (or Apulian limestone), which is 95-98% calcite. Jacketed and unjacketed hydrostatic compression experiments and undrained plane strain compression tests provided the full set of poroelastic parameters. Additionally, the specific storage coefficient was calculated. Inability to obtain constant values of Skempton B coefficient even at high pore pressures (~ 4 MPa) and the decrease in P-wave velocity with water injection revealed partial dissolution of calcarenite in water at high pressures. This phenomenon, as well as the mechanical behavior of rock in contact with supercritical CO2, are currently under consideration.
Hydraulic fracturing (HF) may lead to practical stress management possibilities by creating opportunities to control stress redistribution, or protecting locations where high stresses pose a threat to operations. These possibilities have application in petroleum engineering as well as mining. Understanding naturally fractured rock (NFR) behavior leads to better predictions of rock mass response to HF treatment and induced fracture initiation and propagation. Natural fractures exist in many different states, are reactivated by stress and pressure changes, and have alterable mechanical properties (e.g. stiffness, shear strength), leading to complex behavior in shear, opening, and closing reactions to stress changes. This article presents some attempts to understand and address simulation of HF interacting with a NFR using the Distinct Element Method (DEM) to represent the NFR. To this end, a coupled hydro-mechanical analysis is applied via the Universal Distinct Element Code (UDEC™) software to model both rock and fracture behavior in the HF/NFR system. In the current study, a Voronoi tessellated continuum has been generated to evaluate the effect of the stress ratio on flow into the joints by changing the differential principal compressive stresses. Given the difference in in situ stresses, pore pressure distribution is monitored and the distribution of slip and opening of fractures at different stress field anisotropy is investigated during pressurized hydraulic injection. Based on simulations, pore pressure decreases in a uniform pattern around the injection point in the isotropic stress state; however, the pressure distribution tends to become strongly anisotropic under a stronger differential stress. In addition, both normal and shear displacements show an increasing trend toward anisotropy under stronger stress differences. Applications may ensue with better understanding, such that HF strategies for strongly differential stress fields may evolve to be substantially different than for near-isotropic stress fields, and similar conclusions may ensue for random NFR fabrics, compared to cases with strongly oriented natural fracture fabric.