Rock mass response ahead of an advancing face in faulted shale

https://doi.org/10.1016/j.ijrmms.2013.01.002Get rights and content

Abstract

In this study, the rock mass response ahead of an advancing test tunnel in the Opalinus Clay at the Mont Terri Rock Laboratory (Switzerland) was investigated. Characterisation of the excavation-induced damage zone at Mont Terri is a challenging task due to the anisotropic and heterogeneous nature of the shale: pronounced bedding leads to intact rock anisotropy and prevalent small-scale tectonic shears lead to rock mass heterogeneity. Rock mass damage ahead of an experimental tunnel or niche was characterised through single-hole seismic wave velocity logging, borehole digital optical televiewer imaging, and geological drillcore mapping. Three-dimensional elastic stress analyses were completed and showed that rock mass degradation can be correlated to changes in the maximum to minimum principal stress ratio (i.e., spalling limit). Numerical results showed that close to the niche boundary, unloading lowers stress ratios, which correspond with decreasing seismic wave amplitudes and velocities; thus, indicating that strength degradation resulted from increasing crack-induced damage. Considerations of tectonic shears and distance from a previously stressed volume of rock were necessary in understanding both the damage state and extent ahead of the face. By integrating field and numerical data, the investigation showed that geological structures (i.e., bedding and bedding-parallel tectonic shears) were most influential near the entrance but played a lesser role as the niche deepened. Additionally, a portion of the niche is located in the perturbed zone of the intersecting Gallery04.

Highlights

► Rock mass response ahead of an advancing tunnel in shale was investigated. ► Limited macro-damage but seismic wave data identified micro-scale damage. ► Three-dimensional elastic stress modelling completed. ► Principal stress ratios lowest where seismic wave velocities and amplitudes lowest. ► Tectonic shears and proximity to past damage necessary for characterisation.

Introduction

Tunnel construction damages the surrounding rock mass, which can lead to the alteration of rock mass transport properties and/or tunnel instability. Safety assessment of geological nuclear waste repositories necessitates understanding the processes that lead to rock mass perturbations induced by tunnel excavation. While many previous studies have investigated perturbations around tunnels in shale, few have considered the development of perturbations ahead of an advancing tunnel. However, degradation of the rock mass ahead of the advancing face may influence development of the perturbed zone around the tunnel away from the face: e.g., possibly leading to asymmetric tunnel breakouts [1].

Abel and Lee [2] demonstrated that changes in the stress state can be detected several tunnel diameters ahead of the face in both laboratory and field studies. The laboratory studies involved tunnelling in models built from acrylic (ideally elastic), concrete (heterogeneous elastic), and granite (approximately elastic brittle). The onset of stress changes were detected two to four diameters ahead of tunnels drilled into the laboratory models. Compressive stress peaks, one to two diameters ahead of the tunnel, were also measured in these models. In the field study, two probes were installed about 15 m ahead of a proposed crosscut in jointed and closely foliated gneiss and gneissic granite. Changes in stress associated with the advancing tunnel were measured more than seven tunnel diameters ahead of the face with a compressive stress peak about six tunnel diameters away from the advancing face. This was followed by a much larger decrease in compressive stress. Stress-change trajectories were also determined from the field measurements and demonstrated that local structural variations in the foliated and faulted metamorphic rocks controlled the rock mass response. The crosscut was driven orthogonal to the strike of the major geological weakness in the rock mass, or the foliation in this case. Because the foliation and associated jointing provided a ready avenue for tensile strain relief, the rock mass was postulated to have expanded preferentially perpendicular to the foliation and towards the advancing face. In this case, strain relief was provided parallel with the crosscut axis and normal to the strike of foliation. Overall, the tunnel advance resulted in decompression of the rock mass ahead of the face and to the side of the tunnel.

Read et al. [1] and Martin [3] investigated the development of v-shaped notches around a test tunnel that was excavated in the Lac du Bonnet granite, and concluded that notch development depended on changes in the stress state ahead of the face. The Mine-by Experiment tunnel was aligned roughly with the intermediate principal stress (σ2) axis while the minimum principal stress (σ3) axis was sub-vertical. The maximum principal stress (σ1) was nearly orthogonal to the tunnel axis. V-shaped notches formed in the roof and floor of the tunnel about 0.6 m behind the face. The notches were not diametrically opposed due to non-symmetric stress concentrations caused by a 10° offset of the tunnel axis with the intermediate principal stress direction [1]. Stress path analyses indicated that well ahead of the advancing face in regions where the notches formed, the crack initiation threshold [4] was exceeded and thus, damage was in the form of micro-fracturing. Additionally, principal stress rotations, when stress levels exceeded crack initiation, also initiated ahead of the face with the maximum rotation in σ3 occurring in the roof. Consequently, rock mass degradation near the tunnel perimeter may be further exacerbated. Through a numerical study, Eberhardt [5] found magnitude and directional changes in the redistributed stress field near the face differed depending on the tunnel alignment with the far-field principal stress axes. It was postulated that magnitude and directional changes in the redistributed stress field would lead to progressive accumulation of damage.

Investigations relating to nuclear waste storage in argillaceous media have also shown rock mass perturbations initiating ahead of the face. Induced fracturing has been mapped in tunnel faces at the Meuse/Haute-Marne Underground Research Laboratory in France [6] and at the HADES Underground Research Facility in Belgium [7]. In both cases, fracturing formed an open “v” with a horizontal axis of symmetry near the tunnel springline. Fracturing was also found to be more pronounced when the tunnel axis was aligned parallel with the maximum horizontal stress at Meuse/Haute-Marne.

The Opalinus Clay in Switzerland is under consideration as a potential host rock for the storage of nuclear waste [8]. At the Mont Terri Rock Laboratory, the typical zone of perturbation around the tunnel cross-section consists of sub-vertical extension fracturing in the sidewalls and bedding-parallel fracturing above the crown and below the invert [8]. The rock mass response several metres ahead of the face has been examined in the EDB section of Gallery98 [9], HG-A niche in Gallery04 [10], and the MB section of Gallery08 (unpublished reports are currently under review, CD Martin, pers. comm.). This paper examines the rock mass response immediately ahead of an advancing face before and during the excavation of a short test tunnel, the EZ-B niche (Fig. 1), at the Mont Terri research facility.

Section snippets

The Mont Terri Rock Laboratory

The EZ-B niche is located in Gallery04 (Fig. 2) at the Mont Terri Rock Laboratory in northern Switzerland. Mont Terri is the northernmost in a series of anticlines in the Jura Mountains and the research facility is located in the southern limb, which is weakly deformed and less tectonically disturbed [11]. The anticline was formed by fault-bend and fault-propagation folding [12]. At the laboratory scale, tectonic features consist of networks of thin (i.e., in the order of millimetres) shear

The EZ-B niche

The EZ-B niche has a diameter of 3.8 m and length of 6–7 m (Fig. 2). Construction of the niche and associated borehole drilling campaigns spanned a period of 5 months from December 2004 to April 2005. In December, an entrance to the niche with a length of 1–2 m was excavated. The rock surface was lined with 150-millimetre-thick fibre-reinforced shotcrete and a 300-millimetre-thick concrete floor slab. Three 100-millimetre-diameter horizontal observation boreholes (BEZ-B1 to B3) with lengths of

Data integration

The rock mass response ahead of the advancing niche face was evaluated by integrating data collected from the central borehole, BEZ-B3 (Fig. 1, Fig. 3), as it was shortened to its final length of roughly 3 m. Borehole field data included drillcore mapping, digital optical televiewer (DOPTV) imaging, and single-hole seismic wave measurements. Other field data included laser scans and geological mapping of the niche surfaces. Numerical data from elastic continuum simulations were also incorporated.

Damage observations

In the pre-excavation stage, 38 fractures were mapped in the BEZ-B3 drillcore. None of the fractures were identified as excavation-induced fractures (or “unloading joints”). Five bedding-parallel shears were mapped while the remainder were identified as “artificial discontinuities” [13]. The lack of “unloading joints” is not surprising as excavation of the entrance removed the first one to two metres of rock mass where the induced macroscopic fracture frequency would be at its highest. However,

Assessing damage

Damage ahead of the niche face was assessed by integrating the data and observations collected in the field with the stress analyses carried out via the numerical simulations. While field data and observations provide phenomenological evidence of excavation-induced damage, the stress analysis provides a framework for its interpretation.

Stress-driven failure depends on the rock mass strength, the in situ stress magnitudes, and principal stress orientations. Brittle rock fails in tension by

Conclusions

Field evidence for macro-damage was limited in the EZ-B investigation but seismic wave data provided a means for identifying micro-scale damage. By integrating field and numerical data, the investigation showed that geological structures (i.e., bedding and bedding-parallel tectonic shears) were most influential near the entrance but played a lesser role as the niche deepened. Additionally, a portion of the niche is located in the perturbed zone of the intersecting Gallery04.

Damage ahead of the

Acknowledgements

This intensely collaborative research project was funded by the Swiss Federal Nuclear Safety Inspectorate (ENSI) and in particular, Erik Frank is thanked for day-to-day support. Seismic data was provided by Kristof Schuster with field support from Torsten Tietz, Dieter Boeddener, Friedhelm Schulte, and Wilfried Stille. Geological mapping was completed by Christophe Nussbaum, Nicolas Basdertscher, and Olivier Meier. Thorsten Schulz and Hans-Martin Zogg with field support from Frank Lemy, Corrado

References (31)

  • Read R, Martin CD, Dzik EJ. Asymmetric borehole breakouts at the URL. In: Daemen JJK, Schultz RA, editors. Rock...
  • C.D. Martin

    Seventeenth Canadian geotechnical colloqium: the effect of cohesion loss and stress path on brittle rock strength

    Can Geotech J

    (1997)
  • Z.T. Bieniawski

    Mechanism of brittle fracture of rock, parts I, II, and III

    Int J Rock Mech Min Sci

    (1967)
  • Armand G, Wileveau Y, Morel J, Cruchaudet M, Rebours H. Excavation damaged zone (EDZ) in the Meute Haute Marne...
  • Martin CD, Lanyon GW. Excavation disturbed zone (EDZ) in clay shale: Mont Terri. Unpublished Mont Terri Technical...
  • Cited by (30)

    • Effect of principal stress direction interchange on the failure characteristics of hard rock

      2023, International Journal of Rock Mechanics and Mining Sciences
    • Influence of coal seam gas pressure on the propagation mechanism of outburst two-phase flow in visual roadway

      2022, Fuel
      Citation Excerpt :

      The evolution of the shock wave overpressure is shown in Fig. 6. It can be observed that the shock wave overpressure near the front of the outburst cavern alternated between positive and negative values[36]. After the first wave reached its peak value, the shock wave overpressure rapidly decreased and became negative.

    • Evolution of excavation damaged zones in Opalinus Clay shale inferred from seismic investigations

      2022, Engineering Geology
      Citation Excerpt :

      The extent and geometry of EDZs depends upon the gallery geometry and size, excavation direction with respect to rock bedding, rock mass fractures, strength, and stress, as well as the excavation technique. In general, EDZs around galleries at the MT URL excavated roughly normal to bedding strike are composed of new extensile fractures opening about normal to bedding at gallery sidewalls and bedding-plane breaks and tectonic fault reactivation in the invert and roof of the galleries (Bossart et al., 2002; Martin et al., 2004; Yong et al., 2013, 2017; Thöny, 2014). Tensile fractures are often associated with smaller shear fractures, which can result in an interconnected EDZ network within the first approximately 1.0 m of the rock mass (Nussbaum et al., 2011; Thöny, 2014).

    • Damage evolution mechanisms of rock in deep tunnels induced by cut blasting

      2016, Tunnelling and Underground Space Technology
      Citation Excerpt :

      In addition to using damage constitutive models, the factors affecting the damage of rock mass need to be analysed. Although the factors influencing excavation damage zone (EDZ), such as transient unloading (Cook et al., 1966; Abuov et al., 1988; Carter and Booker, 1990; Cai, 2008, 2013; Lu et al., 2008, 2012; Zhu et al., 2014; Fan et al., 2015), stress redistribution (Bäckblom and Martin, 1999; Kwon et al., 2009; Yong et al., 2013), and wavelength-to-excavation span ratio (Wang and Cai, 2015), have been studied, there are still lacking quantitative results. Since rock mass in deep tunnels is in a three-dimensional (3D) stress state, rock mass is indispensable under the action of in-situ stress.

    View all citing articles on Scopus
    View full text