Project Topics
Get Complete Project Material File(s) Now! »
Determination of the stress field in rock masses
Methods for measuring in situ stresses
Depending on the domain of application, the most commonly used stress determination techniques include hydraulic methods, relief methods and jacking methods. More techniques exist, and comprehensive reviews of stress determination methods may be found for example in Amadei and Stephansson (1997) and Ljunggreen et al. (2003).
The main objective of hydraulic methods (Cornet and Valette, 1984) is to measure in situ stresses by isolating a section of a borehole and applying a hydraulic pressure on its wall. The applied pressure is increased until existing fractures open or new fractures are created. The fluid pressure required to open, generate, propagate, sustain, and reopen fractures in rock at a given depth is measured and is related to the existing stress field. The directions of the measured stresses are inferred from the measurement of the orientation of the fractures. Explicit uncertainties associated with stress determination through hydraulic tests depend on the accuracy on the test location and on uncertainties on both normal stress measurements and fracture orientations. However, no hypothesis is formulated with respect to the stress-strain relationship for the rock mass and the only implicit uncertainties to be considered concern the a priori assumption proposed for describing spatial stress variation when large volumes of rock are considered.
Relief methods (Sjoberg et al. 2003) consist of the isolation of a rock sample from the stress field in the surrounding rock mass and the monitoring of its response. This can be achieved by different methods, such as overcoring or undercoring holes and cutting slots. However, the stresses are not related to applied pressures as in hydraulic methods. Instead, the stresses are inferred from strains or displacements created by the relief process and measured on the isolated rock samples, in boreholes, or on the surrounding rock associated with the relief process. The interpretation of the stress relief tests depends on a stress-strain (displacement) relationship for the rock. Uncertainties associated with such stress determination depend explicitly on the precision on measurements of both the stress relief induced strains or displacements and the rock elastic parameters. But they depend also implicitly on the validity of the hypotheses implied by the corresponding technique (e.g., homogeneity, linear elasticity, uniformity of the stress components within the volume where the data have been gathered).
In jacking methods (Habib and Marchand, 1951), the equilibrium of a rock mass is disturbed by cutting slots on the surfaces of rock excavations. This creates deformations that are measured by means of reference pins or strain gauges that are placed on either side of the slots. Then, a jack is inserted into the slots and pressurised until the equilibrium is restored. The stresses normal to the slots are directly related to the pressure necessary to restore the equilibrium. Explicit uncertainties associated with stress determination through jacking methods depend on the uncertainties on both normal stress measurements and slot plane orientations. Implicit uncertainties depend on the knowledge of the rock’s constitutive behaviour, the concentration factors along the tested rock mass surface required to relate the measured normal stresses to the stresses at the far-field, the geometry of the excavation and the uniformity of the stress component perpendicular to the jacks.
Integration of in situ tests results using numerical modelling
In situ tests only enable the characterisation of the rock stresses at the location where the tests are performed due to the various factors that influence rock stresses, such as the heterogeneities, the existing geological structures and the correlated spatial variability of rock mass properties. Thus, it is essential to consider a model that integrates the results of the various in situ measurements to extrapolate the results to a rock mass volume of interest for the design of the underground structures of concern.
When different types of measurement techniques are used at the same location, the results must be combined into a single inversion scheme to determine the stress field that best fits all the data. Such integration schemes have already been presented by several researchers: hydraulic fracturing (HF) tests and hydraulic tests on pre-existing fractures (HTPF) (Cornet and Valette, 1984), hydraulic tests (HF and HTPF) and flat jacks (Cornet, 1996), hydraulic tests and focal mechanisms of induced seismicity (Cornet and Yin, 1995), hydraulic tests and overcoring tests (Ask, 2004) and overcoring tests and flat jacks (Lamas et al., 2010).
Numerical models are then sometimes developed to integrate all the tests results within a single model that is used to determine the most probable stress state in the volume of interest (Sousa et al. 1986, Hart 2003, Lamas et al. 2010, Matsuki et al. 2009, Muralha et al. 2009). In mountainous regions, with irregular terrain surfaces, these models are an essential tool to evaluate the influence of topography on the stress field, because analytical solutions, such as those provided by Savage et al. (1985) are not applicable.
The discrepancy found between the stresses provided by in situ tests and the results of the numerical models obtained due to gravity loading has been justified with the existence of tectonic stresses. By assuming this hypothesis, a linear elastic analysis is usually conducted to determine the horizontal stresses that must be applied to the model boundaries to achieve a good fit with the data (Li et al. 2009, McKinnon 2001, Tonon et al.2001). In this analysis, rheological effects are commonly neglected.
At present, Portugal faces an important movement regarding underground construction in rock masses that is the return to the construction of new large hydroelectric power schemes, which had been stopped for many years mainly due to environmental constraints. Not only several large dams but also powerhouse reinforcements of existing schemes are in the design phase or already under construction. The powerhouse reinforcements have two main purposes: in some cases, they simply increase the production of electrical power through the installation of new turbines in new powerhouses, and in other cases, turbine-pump systems are installed to store energy during periods in which the production of electricity due to other sources (especially wind farms) exceeds the demand. Most of these schemes include large underground caverns and long hydraulic pressure tunnels with significant cross sections. The National Programme for Dams with High Hydroelectric Energy Potential includes the construction of several new dams that will involve the excavation of new shafts, powerhouses and long hydraulic circuits.
This thesis analysed a case study that included data that were obtained through different techniques (hydraulic, overcoring and flat jack testing) in various locations within a rock mass, in a mountainous region. The measurements were conducted for the design of a re-powering scheme of an existing hydroelectric system that includes a large underground cavern and a long hydraulic pressure tunnel. The aim of the present study is to develop a data integration approach to extrapolate the results of the various in situ tests to the rock mass volume of interest for the design of the underground hydroelectric power scheme. The specific objectives of the study are the following:
• to complement an existing in situ testing program, by performing additional in situ tests exclusively for this thesis;
• to develop an inverse model for the determination of the regional stress field that best fits the data provided by the various testing techniques;
• to evaluate the influence of the topography effects on the stress field obtained due to gravity and tectonic loadings;
• to estimate the role of tectonic and gravitational stresses in the global stress field;
• to compare in situ measurements obtained by overcoring and flat jack techniques in the vicinity of an existing adit;
• to study the influence of the compliance of existing large-scale fractured zones on the stress field;
• to investigate the influence of the rheological properties of the rock mass on the interpretation of the stress measurement results;
• to use stress field information to ascertain the long-term mechanical properties of an equivalent geomaterial.
Thesis outline
In face of the objectives presented in the previous section, the thesis is structured in four chapters, of which this introduction is the first one.
Chapter 2, In situ stress measurements, discusses the collection and preliminary interpretation of the data obtained for this thesis. First, is described how hydraulic tests help determine the vertical stress profile. The overcoring tests are then described, and the results are combined in an inversion scheme to constrain the stress tensor at their location. Finally, the small flat jack tests are discussed, and their results are compared to those of the overcoring tests that were conducted at almost the same location.
Chapter 3, Determination of the regional stress field, presents an equivalent and continuum mechanics model to integrate the various data to assess the regional stress field. The model is used to analyse the role of topography on in situ stresses distribution due to gravity and tectonic loadings, to investigate the influence of the compliance of existing large-scale fractured zones and potential tectonic stresses on the stress field and to ascertain the long-term rheological characteristics of the equivalent geomaterial.
In the last chapter the more relevant and innovate conclusions are summarized and suggestions for future developments of the work are suggested.
The case study considered in this thesis refers to the re-powering scheme of the Paradela hydroelectric scheme located on the Cávado River in the North of Portugal (Figures 2.1 and 2.2). The Paradela II scheme includes a new 10 km hydraulic conduit and a powerhouse complex placed about halfway in the conduit (Figure 2.3) and will primarily be excavated in granite. It includes a new powerhouse cavern, a valves chamber and a large surge chamber with several adits located 500 m below ground level.
The water intake is placed at Paradela dam (Figure 2.4), which is a concrete face rockfill dam built in the 50’s. The circuit will run approximately along the Cávado River exiting near the confluence of the Rabagão River (a left bank tributary). It will cross the flank of Serra do Gerês that corresponds essentially to an outcrop of the Gerês granite, which is a post-tectonic biotite granite with calcite plagioclase of medium size with porphyritic trends.
In situ stress measurements were performed for the design of the re-powering scheme. Hydraulic tests provided the means to determine the natural stress profiles in two vertical 500 m deep boreholes (PD19 and PD23; Figure 2.5a). These two boreholes are separated by 100 m and reach the location of the future powerhouse cavern. The tests were conducted between 200 m and 500 m below the surface, i.e., above and below the ground surface, i.e., above and below the altitude of the nearby river bed.
In another location, approximately 1.7 km away from the hydraulic test site, OverCoring (OC) and Small Flat Jack (SFJ) tests were performed. Overcoring tests were conducted to determine the complete stress tensor at six different depths in two 60 m deep vertical boreholes (PD1 and PD2; Figure 2.5b), located approximately 4550 m downstream from the water intake. These boreholes are separated by 150 m and were drilled from an existing adit located 160 m below ground level and 170 m above the future hydraulic circuit. In borehole PD1, tests were conducted at depths between 205 m and 251 m below ground level. In borehole PD2, tests were conducted at depths between 163 m and 202 m below ground surface. Two tests were performed at approximately the same depth in each borehole. All of the tests were conducted above the river bed.
Small flat jack tests were performed to determine the stresses normal to various planes perpendicular to the walls of the existing adit near the boreholes where the overcoring tests were performed. Four tests were performed at a site (SFJ1) located within 3 m of borehole PD1, and eight tests were performed at sites (SFJ2 and SFJ3) located within 50 m of borehole PD2. Figure 2.6 presents a vertical cross section along the axis of the adit, along with the location of the overcoring and small flat jack tests.
In this chapter, is discussed the collection and preliminary interpretation of the data. First, is described how hydraulic tests help determine the vertical stress profile. The overcoring tests are then described, and the results are combined in an inversion scheme to constrain the stress tensor at their location. Finally, the small flat jack tests are discussed, and their results are compared to those of the overcoring tests that were conducted at almost the same location.
In situ tests
Hydraulic tests
Hydraulic testing has involved two different techniques: Hydraulic Fracturing (HF) and Hydraulic Testing of Pre-existing fractures (HTPF), according to the procedures described by Haimson and Cornet (2003).
The hydraulic testing equipment that was used in these measurements includes a geophysical logging cable, an HTPF testing system, and all of the necessary recording equipment. The cable is moved within the borehole using a winch system that is driven by an electrical motor. A schematic view of the system is presented in Figure 2.7.
The HTPF testing system includes two steel-reinforced packer elements, which are used to create or to reopen pre-existing fractures, and an electrical imaging device (HTPF tool), which provides a characterisation of the orientations of the fractures. The packers and the HTPF tool are shown in Figure 2.8.
In Figure 2.8, the top of the downhole assembly is shown in the lower-left corner and terminates with the upper packer element. The second and third parts on the figure include the lower packer element and a weight that is used to limit the associated friction problems (mounted below the HTPF tool). The lower-right corner on this figure shows the HTPF tool. The system is hydraulically connected to the surface by two coiled tubings that are attached to the logging cable with clamps that are sets at regular intervals. The coiled tubings (one is used for the packers, and the other is used for the test interval) are composed of stainless steel.
The HTPF tool (Mosnier and Cornet, 1989) is composed of a ring with 24 electrodes, which are set at various azimuths on a ring that is placed at the centre of the tool, 3 magnetometers and 2 inclinometers for measuring the position of the tool in three-dimensional space, and the electronic components required to transmit the signal to the surface. The electrical imaging technique has been adapted from the azimuthal laterolog described by Mosnier (Mosnier, 1982). During the measurement, an alternating electric voltage is applied between a distant electrode (the armour of the logging cable) and the 24 electrodes. The electric current emitted (or received) by each of the electrodes on the central ring is proportional to the conductance of that part of the borehole wall facing the electrode. The focusing electrodes located on both sides of the electrode ring ensure that the electric current lines are normal to the borehole wall. The results can be displayed either as polar diagrams or graphically as horizontal bands composed of juxtaposed squares (one square per electrode). Because the intersection of a plane with a cylinder is an ellipse, planar fractures are easily detected in the horizontal bands diagrams by their characteristic sinusoidal shape.
Before the tests are conducted, the tripod and all of the equipment need to have been placed at the location of the boreholes (Figure 2.9a). The system, which includes the two packers and the HTPF tool, is then descended (Figure 2.9b), and the first reading (reading R1) for the fracture depths and orientations is obtained through the HTPF tool.
Table of contents :
CHAPTER 1. INTRODUCTION
1.1. General
1.2. Determination of the stress field in rock masses
1.2.1. Methods for measuring in situ stresses
1.2.2. Integration of in situ tests results using numerical modelling
1.3. Objectives
1.4. Thesis outline
CHAPTER 2. IN SITU STRESS MEASUREMENTS
2.1. Introduction
2.2. In situ tests
2.2.1. Hydraulic tests
2.2.2. Overcoring tests
2.2.3. Small flat jack tests
2.3. Laboratory test results
2.4. Stress measurement results
2.4.1. Hydraulic test results
2.4.1.1. Calibrations
2.4.1.2. Description of the pressure and fracture orientations data37
2.4.1.3. Normal stress component determination
2.4.1.4. Fracture orientations determination
2.4.1.5. Summary of the pressure and fracture orientations data
2.4.1.6. Determination of the stress field
2.4.2. Overcoring test results
2.4.3. Small flat jack test results
2.5. Conclusion
CHAPTER 3. DETERMINATION OF THE REGIONAL STRESS FIELD
3.1. Introduction
3.2. Inverse model
3.3. The role of the topography on the in situ stresses distribution
3.3.1. Gravity loading
3.3.2. Tectonic loading
3.3.3. Gravity and tectonic loadings
3.4. Comparison between in situ tests and results from the FLAC3D model when considering gravity loading alone
3.5. Influence of faults and fractured zones
3.6. Testing the existence of tectonic stresses
3.7. Examining rheological effects
3.8. Design of the hydraulic pressure tunnel
3.9. Conclusion
CHAPTER 4. CONCLUSIONS
4.1. Summary and conclusions
4.2. New contributions
4.3. Recommendations
