SUMMARY: The paper introduces and presents instances of application of methodologies for analysis of the stress field obtained from the results of in situ stress measurements using the overcoring, hydraulic fracturing and flat jacks methods. These methodologies integrate all stress measurements, and use numerical models of the rock mass that represent the ground topography and the underground excavations, so that the most likely stress field in the zone of interest for the design of several large underground structures is obtained. On one hand, all testing methods have limitations, inherent to its nature, and assumptions such as verticality of one principal stress or the rock mass linear elastic behaviour. On the other hand, the stress field presents significant spatial variability, influenced by ground topography, existing excavations, rock mass heterogeneity, tectonic forces, or time dependent effects (Figueiredo et al., 2012). However, release of the in situ stresses is often the most relevant action for design of underground structures, and the magnitude and the direction of the stresses can influence the location and orientation of a cavern, the support design and the excavation method.
Important research has been developed in recent decades with the objective of the elastic characterization of shale formations, for their usefulness to optimize designs of hydraulic fracturing, seismic exploration, and estimates of horizontal stress and decreased wellbore stability problems. However, the investigation of shale is hampered by its mechanical and chemical instability, hindering the recovery of cores (cm scale) used in conventional mechanical testing. Thus, the development and implementation of experimental correlations to obtain the elastic modulus or Young’s modulus by indentation technique, exhibits a great potential for the elastic characterization of these formations. Currently there is an important advance in the technique of indentation allowing the use of small samples as drill cuttings obtained during drilling and forces significantly lower than required in laboratory tests such as the uniaxial test. The correlations obtained between dynamic Young’s modulus and the results of the indentation test are presented in this article. For this end, cylindrical samples of shale were considered that had both ultrasonic measurements (centimeter scale deformations) for obtaining dynamic Young’s modulus and force-displacements curves obtained with the indentation technique, which allow detection of displacements as small as 0.001 mm.
Permeability is an intrinsic property of porous materials, dependent of pore size and interconnection in microscopic level, as well as fissure width and orientation in macroscopic level. Mechanical loading of such a porous material leads to changes in the geometry of the pore space with impacts on the macroscopic measured permeability. Realistic reservoir simulation models should consider these permeability changes as these may affect the reservoir pressure distribution, saturation front shapes and also the drilling mudwindow of infill drilling wells. However, the petroleum industry still lack of reliable and comprehensive permeability variation models and parameters. The article presents results of a campaign of hydrostatic tests underwent on samples from limestone and dolomite outcrops. The main target of those tests is the strain-dependent permeability measurement and the validation of the model proposed by Petunin et al. (2011). This model is an extension to the Carman-Kozeny formulation and, although still simplistic by considering pore volume changes alone as the main driver for the permeability variations, it may be quite effortlessly introduced in the conventional numerical reservoir simulators as Carman-Kozeny exponents or even as cell permeability multipliers. The results presented here are the first step of an internal project which aims to determine the correlation between variation in permeability caused by variation of in situ stress and strain state due to reservoir depletion for limestone and dolomite oil reservoirs. The chosen commercial Silurian Dolomite and Indiana Limestone outcrops can be considered as quasi-homogeneous standard materials when compared to real reservoir samples, a key aspect for this testing methodology validation.
Firme, Pedro A.L.P. (Alis Soluç ões em Engenharia e Sistemas) | Quispe, Roberto Quevedo (Alis Soluç ões em Engenharia e Sistemas) | Roehl, Deane (Instituto Tecgraf) | Oliveira, Maria Fernanda (Instituto Tecgraf) | Parotidis, Miltiadis (Instituto Tecgraf) | Glassborow, Brent (BG Group plc.)
For underground reservoirs, the decrease of pore pressure during hydrocarbon production can trigger reservoir compaction and subsidence problems. In order to analyse and forecast those phenomena, the finite element method (FEM) has been employed in several applications because of its flexibility in the treatment of geometry and topological heterogeneity. Those FEM features can provide more realistic estimates in real fields. However, a linear elastic model is usually adopted to simulate the behaviour of rock materials, disregarding different deformation mechanisms. This paper takes into consideration that issue through the analysis of a real field adopting the linear-elastic and the elasto-plastic Mohr-Coulomb and Modified Cam-Clay models. The results suggest that a careful definition of constitutive models should be given since the compaction and subsidence predictions may be significantly different.
Napa-García, Gian Franco (Department of Geotechnical Engineering EESC-USP) | Beck, Andre Teófilo (Department of Geotechnical Engineering EESC-USP) | Celestino, Tarcisio Barreto (Department of Geotechnical Engineering EESC-USP)
Risks involved in engineering projects need to be quantified objectively. Therefore, a keystone to this process is the quantification of the probabilities of failure of the single and combined failure modes. In geotechnical engineering, the most common technique used to calculate the probability of failure of an engineering system is the Monte Carlo simulation MCS. This technique leads to the exact answer when the number of realizations tends to infinite. This condition can be a handicap. To overcome this handicap, the direct coupling surge as novel technique that can lead with complex scenarios in an efficient way. Thus, this paper presents an application of the direct coupling method to estimate the probabilities of failure of a rock slope analyzed by Hoek (2007). Hoek (2007) performed MCS to obtain the probability of failure of that slope. The direct coupling was implemented in a Mathematica 9.0 version of the StRAnD. Also, the analytical formulation of the problem was written in a Mathematica notebook and coupled to StRAnD. The probability of failure was approximated using FORM and SORM. The results obtained by using SORM are very close to those obtained by Hoek using MCS. MCS performed in this study showed that both FORM and SORM approximations are good estimators of the probability of failure. This similarity between FORM and SORM suggest that the limit state surface was reasonably linear. Our results show that precise estimates of the probability of failure can be obtained by using a small number of callings of the performance with FORM or SORM approximations. With this, the used of advanced reliability techniques can be used to quantify the risk involved in engineering project in a more efficient fashion by the use of the direct coupling approach.
The toppling failure mode is a mechanical process in rock slopes with regularly spaced layers or foliation, which involves block overturning and bending of columnar structures. As those processes develop, the failing of internal structures in the slope may occur, leading the whole system to its collapse. The Discrete Element Method based on circular discrete elements was used to analyze two modes of toppling: the block and flexure toppling modes. The method of the study was structured by means of a progressive modelling of the phenomena, and the subsequent validation of the results through a comparison with analytical and semi-analytical approaches. The strain and strength of rock structures subjected to bending process were modeled using a new contact model between every discrete element, which resulted in a better response than other conventional contact models. Subsequently, in the numerical simulation of a physical model, the strains and strength behavior of the experimental slope are reproduced by the numerical calibration of the rock and the joint mechanical properties. The result of this process shows a significant dependence on the stiffness and frictional joint components rather than the properties of the rock itself. That was a different but an acceptable conclusion among other similar works, which aim the rock properties and the joint frictional angle the main factors that control the slope stability in toppling process.
A numerical approach to cave assessment has been developed by Itasca over the past 15 years through the industry-funded International Caving Study (ICS I & II) and Mass Mining Technology (MMT) projects. The procedure has been implemented successfully in the calibration and simulation of cave responses at a number of existing and planned caving projects. In this paper, the procedure is used to simulate block caving for an ongoing underground mining operation, which has been active for more than 40 years. Due to limitations in accessing available information, the problem was posed to reproduce caving and cratering for the last 10 years, starting with the presence of a known cave representing mining in previous years. The algorithm to simulate caving has been implemented within the three-dimensional, continuum based program, FLAC3D, following a rigorous mass-balance routine to ensure that the design production schedule is represented accurately. A constitutive model specially designed to represent caving behavior has been recently developed by Itasca applying the concept of strain-softening (CaveHoek constitutive model), whereby strain-dependent properties are adjusted to represent the effects of dilation and bulking that accompany caving. The main outcome of the study is to compare the simulated cratering at the surface against the known topography at the end of the simulation period. The validation of the field observations provides increased confidence about the model predictions for future mining operations.
The main reasons for rock masses to frequently exhibit impressive degrees of anisotropy, with properties varying with direction of observation and measurement, are clearly their varied geological origins. Origins may provide distinctive bedding cycles in sedimentary rocks, distinctive flows and flow-tops in basalts, foliation in gneisses, schistosity in schists and cleavage in slates, and faults through all the above. We can add igneous dykes, weathered horizons, and dominant joint sets. Each of the latter are rich potential or inevitable sources of velocity, modulus, strength and permeability anisotropy. The historic and present-day stress anisotropy provides a wealth of additional examples, which further reinforce disbelief in the elastic-isotropic-continuum intact-rock-based assumptions used and promoted by so many of us for modelling rock masses. RQD and Q are frequently anisotropic as well. Has the a priori assumption of homogeneous-isotropic-elastic behaviour any significant place in the scientific practice of rock mechanics?
Karekal, Shivakumar (CSIRO Earth Science and Resource Engineering) | Subramanian, Srikrishnan Siva (Central Institute of Mining and Fuel Research) | Porathur, John Loui (Central Institute of Mining and Fuel Research)
Highwall mining operation involves driving a series of parallel unsupported, unmanned and unventilated excavations into a coal seam exposed at the open pit Highwall using a remotely operated continuous miner with attached conveying system. These parallel excavations are separated by web pillars of pre-designed width which are critical to the Highwall mining operations. The Highwall slope must remain stable during Highwall mining operation to ensure safety of workers and machinery. In this paper, Highwall slope stability is investigated with respect to different Highwall mining parameters using FLAC3D numerical modeling software. The parameters included in the study are: (i) single seam and multiple seams Highwall mining excavations with different width to height ratios; (ii) different Slope angles; (iii) different excavation heights; and (iv) different cover depths. A narrow strip of rock mass is considered by taking a plane of symmetry. The modeling results reveal that stability of open pit slopes have profound influence on the Highwall mining parameters, and the web pillar design can affect the stability of Highwall slopes. In designing Highwall slopes for an open pit, the design must include Highwall mining excavations, otherwise, near critical failure slopes could become critical and fail with Highwall excavations. In authors’ knowledge, this work is the first attempt at exploring the effect of Highwall mining parameters on overall slope stability.
Carrapatoso, Carla (Computational Geomechanics Group / GTE) | da Fontoura, Sergio Augusto Barreto (Computational Geomechanics Group / GTE) | Inoue, Nelson (Computational Geomechanics Group / GTE) | Lourenço, Affonso (Baker Hughes) | Curry, David (Baker Hughes)
This paper deals with the drilling of rocks by fixed polycrystalline diamond cutters. Single-cutter experiments were carried out on evaporites and results are presented in terms of forces acting upon the cutter and mechanical specific energy spent during cutting action. The discrete element numerical method was used to simulate the single-cutter experiments based on rock properties obtained from high strain rate compression triaxial tests carried out on evaporites. The effects of depth of cut, cutter friction and geometry on drilling efficiency were studied. Drilling efficiency increases with DOC up to a certain level where it stabilizes. The decrease of drilling efficiency with increase of cutter friction and use of blunt cutters also indicates that the methodology is satisfactorily robust and accurate.