Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 191 (2017) 402 – 409

Symposium of the International Society for Rock Mechanics

Stability Analysis of an Over-Tilted Slope in a Granite Quarry:

the Role of Joint Spacing
L.R. Alejanoa,*, U. Castro-Filgueiraa, I. Pérez-Reya, J. Arzúab
a
Natural Resources & Environmental Engineering Department, University of Vigo, Spain
b
Department of Metallurgical and Mining Engineering, Universidad Católica del Norte, Antofagasta, Chile

Abstract

This paper presents the case study of a stability analysis of an over-tilted (inverse) slope in an ornamental granite quarry. Based
on traditional small quarrying practices and due to space constraints, the SW slope was carved following the occurrences of
highly persistent joints that were dipping around 80º counter-slope. Experience dictates that this sort of slopes tends to be
unstable, at least in case of average to low quality rock masses. In this way, a stability analysis was due to analyze slope stability.
The good quality granite rock mass in the slope was characterized, joint data was recorded and laboratory testing was done to
estimate the main significant parameters involved in the study and the geometry of the slope. A stability analysis of the slope was
performed by means of the calculations of safety factors against toppling. These calculations have been done contemplating
various possibilities regarding the occurrence of joints and its spacing, which was found to be the most relevant parameter
controlling stability.
©2017
© 2017TheTheAuthors.
Authors. Published
Published by Elsevier
by Elsevier Ltd. is an open access article under the CC BY-NC-ND license
Ltd. This
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of EUROCK 2017.
Peer-review under responsibility of the organizing committee of EUROCK 2017
Keywords: Rock slope; Stability analysis; Toppling; Rock joint spacing

1. Introduction

This paper presents the case study of a stability analysis of an over-tilted (inverse) slope in a granite quarry. Due
to space constraints of the quarry located in a hill slope, the SW end of the quarry happens to be parallel to a set of
highly persistent joints that were dipping around 80º counter-slope. Therefore, as shown in Fig. 1, large parts of

* Corresponding author. Tel.: +34-986-812-374.

E-mail address: alejano@uvigo.es

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of EUROCK 2017
doi:10.1016/j.proeng.2017.05.197
 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409 403

the slope are overt-tilted, that is, the slope is inverse –dips more than 90º– in these areas. In such situations, the local
mining regulations enforce the owner to carry out a slope stability study appropriately justifying the stability of
the slope. In this paper, the authors presents the most relevant topics of this geomechanical study, together with some
considerations regarding the actual occurrence of instability toppling phenomena in rock slopes.

Fig. 1. General view of the general quarry slope roughly 40 m high.

1.1. Brief description of the quarry and mined stone

The quarry is located in the municipality of Tui in NW Spain. Fig. 1 shows a general view of this exploitation.
The quarry activity is mining local slightly weathered granite, whose commercial name is ‘Silvestre Moreno’.
Quarry production tends to be around 100,000 m3 per year according to market evolution. Final products of
the quarry are granite blocks, semi-blocks and ashlar masonry; appropriate descriptions of these products are
provided below. The exploitation method is quarrying using black powder to cut the rock following the typical three
weakness directions and recently diamond wire has been introduced in the quarry to make precise vertical cuts in
the stone, improving recovery levels in the extraction of the ornamental rock.
The rock quarried can be defined as an alkaline two-mica granite moderately weathered. Grain size is average
5 mm and it is averagely composed by quartz (22 %), microcline (42 %), plagioclase albite (23%), muscovite (6 %)
and biotite (7%). From a structural scope, this is a post-kinematic (post-Variscan or post-Hercynian) granite
batholite, and therefore it has not suffered large tectonic stresses. A number of rock mechanics tests in this kind of
granites were performed in the past and their results presented elsewhere [1, 2].
The attempt to obtain larger, and therefore more profitable, ornamental granite blocks, means that the primary
goal is to extract primary blocks of the largest size possible, to be subsequently squared off to obtain parallelepiped
blocks that are transported to the gang-saw for sawing into slabs.
In general terms, a block suitable for sawing in the gang-saw has the following characteristics: length 1.90-
3.30 m, width 1.00-1.80 m and height 0.90-1.60 m. The slabs that are obtained after sawing will normally be 2-3 cm
thick and surface area will be as large as possible so as to permit the production of large pieces. Next down the scale
in terms of ideal quarry product size is the semi-block; although this may not have the ideal minimum block
measurements indicated above, after sawing it will be of a size that is both commercially acceptable and
economically viable. When the natural fracturing of the rock mass is such that it is impossible to obtain blocks or
semi-blocks, smaller sized particles - greater than 15 cm thick, 42 cm high and of variable length - can be extracted
for sale as ashlar and rubble masonry [3].
In the studied quarry due to its geographical situation far from large towns, the rest of material is carried to
a waste dump. In other quarries, having available crushing equipment and located close to consumption areas,
the remaining material is typically crushed into a number of suitable sizes, sold to the aggregate market.
Average annual rainfall for a return period of 100 year has been quantified in 165 l/m2 by regional records and
404 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409

the local seism risk is low with recommended horizontal accelerations for design of 0.04 g.

1.2. Ornamental granite quarrying

Granite quarrying requires the three weakness planes of the rock to be taken into account in the extraction of
blocks. These three planes - where cut is easiest [3, 4] - are roughly orthogonal in relation to each other. In mining
terminology, they are called the rift, hardway and grain planes.
The rift plane is sub-horizontal, and it is here that the rock separates from the mass. As the plane of least
resistance, this governs how a block is sawed in the gang-saw. For this reason, the largest face of the rock should
ideally correspond to this plane, so as to enhance slab yield in the gang-saw. The rift plane is a result of
the orientation of the minerals in the rock, whose greatest exfoliation planes are parallel to the horizontal plane.
The grain plane is sub-vertical - i.e. normal to the rift - and produces a poorer break in the rock. The origin of
the grain plane possibly lies in structural tensile stresses [4, 5]. Finally, the hardway plane, which is again-sub-
vertical and roughly orthogonal to the rift and grain planes, is the minimum weakness plane, even if it is still
a weakness plane. This is due to micro-cracking, possibly resulting from tectonic compression.
The rift, grain and hardway rift planes guide the orientation of the exploitation operation. Since those are relevant
weakness planes, it is not surprising that discontinuities in this type of granites tend to occur following these
directions. For this reason, it seems logical to relate the three weakness planes to the fracturing of the rock mass,
since both concepts are directly related to historical tectonic processes in the mass.

2. Rock mass characterization

Rock mass characterization was carried out following ISRM Suggested methods [6]. Though the rock mass under
scrutiny presented a good geotechnical quality, it was still possible to recover data of a good number of
discontinuities. In this process, it was easily recognized that the three clearly identified joint sets orientate in parallel
to the three previously defined weakness planes occurring in granite rock and used to traditional ornamental granite
quarrying, namely rift, grain and hardway.
Orientation data of discontinuities recovered have been represented by means of stereographic projection
techniques (Fig. 2) [7] and the average orientations have been obtained. These orientations and other geomechanical
features including JRC, JCS, weathering (according to Moye [8]) and water presence of the observed discontinuities
are gathered in Table 1 in statistical terms (mean value and standard deviation) whenever possible.

Fig. 2. Stereographic pole representation and distribution of discontinuities measured in and around the quarry. A photographic scheme showing
how granite is mined according to joint set directions is also included.
 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409 405

Table 1. Average geomechanical discontinuity features.

n Set Dip dir. (º) Dip (º) JRC JCS (MPa) Weathering [8] Water
Mean St. dev. Mean St. dev. Mean St. dev. Mean St. dev.
1 Rift 49 13 19 2 6 2,7 44 8 II Wet (89 %)
3 Grain 304 21 80 12 8 2,6 41 10 II Wet (81 %)
2 Hardway 207 6 82 10 8 3,3 44 12 II Wet (66 %)

Particularly relevant, regarding to identification of possible failure mechanisms in good quality rock mass such as
this (RMR = 70-75), are the spacing of the discontinuities observed. This spacing of rock joints have been estimated
following two approaches. They have been measured in place together with the standard characterization process
and they have also been measured starting from photographs. The rift spacing have been estimated from
photographs of the quarry slope (Fig. 3.a), and that of the sub-vertical grain and hardway planes have been estimated
from aerial photographs (Fig. 3.b). Interestingly, both kind of approaches have provided similar statistical results as
presented in Table 2. The combination of both approaches is deemed convenient, in the light of the fact that these
parameters and particularly the spacing of the grain-oriented planes will result critical to assess the stability of
the slope. Continuity or persistence of rock joints was also estimated in place and it is presented in Table 2.

Table 2. Mean and standard deviation of spacing and continuity (based on in-situ and photographic estimates) and continuity.
n Set Spacing / In-situ (m) Spacing / Photos (m) Continuity / strike (m) Continuity / dip (m)
Mean St. dev. Mean St. dev.
1 Rift 2,66 1,28 2,80 1,26 > 20 > 20
3 Grain 6,61 6,96 7,63 5,38 10-20 10-20
2 Hardway 11,61 7,23 12,70 6,5 10-20 10-20

Fig. 3. Photographic quantification of the spacing of discontinuities a) estimate of spacing of the rift discontinuity planes; b) estimate of spacing
of the grain and hardway discontinuity planes. Remark the position of the quarry slope.

Other interesting observations regarding the characterization of the observed joints were performed thanks to
having available rock faces recently cut with diamond wire saw. In one of this faces (Fig. 4.a) it was possible to
observe how water flow through a rift discontinuity only occurs in a conspicuous manner. It means that flow only
occurs through particular preferential flow channels within the joint surface. In this way, water pressures in rock
406 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409

blocks only take place in particular areas of the joints but not in all their surfaces. Additionally in recently mobilized
rock block it was observed again in rift joint the occurrence of small rock bridges (Fig. 4.b), that is small areas of
the joints where both surfaces are in contact. Accordingly, considering continuous joints results a conservative
approach since these joints do have some non-frictional shear strength and as well as some tensile strength. These
relevant engineering geology observations are difficult to quantify in rock mechanics practice. They do indicate
however, that considering the joints fully continuous is a rather conservative approach [9].
Standard rock mechanics laboratory testing following ISRM suggested methods [6] was carried out on a small
number of samples (Fig. 5). The obtained results are presented in Table 3.

Fig. 4. a) Recently diamond wired cut surface showing a rift joint where flow occurs in a conspicuous manner (water flow locates in particular
areas of the joint); b) Recently moved rock block base associated to a rift joint where rock bridges (intact rock, no-jointed areas) are visible.

Fig. 5. Samples of granitic rock of the quarry to be submitted to compressive strength and Brazilian testing.

Table 3. Mean and standard deviation basic rock mechanics parameters as obtained in the lab.
Rock Density (g/cm3) UCS (MPa) UTS (MPa) Ib (º)
Mean St. dev. Mean St. dev. Mean St. dev. Mean St. dev.
Weathered granite 2,58 0,02 65 7,3 2,4 0,4 29,4 1,8

3. Geometry of the slope and identification of failure mechanisms.

Based on traditional surveying and topographic maps of the quarry, three cross sections of the main quarry slope
to be studied were represented as shown in Fig. 6. Since cross-cut section 3 is the higher and steeper section, this
will be taken as the base case to perform stability analyses. In the lower part of the represented profiles, statistical
representation of hardway and rift joints are presented (grain is parallel to the analyzed sections), which provide
an illustrative view of the fracture patterns actually occurring in the rock mass behind the slope. Based on these
 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409 407

fracture patterns and in the stereographic representation of the orientation of the main joint sets together with slope
orientation, it can be derived that (kinematic) possible slope failure mechanisms [10] comprehend planar failure
through rift planes and toppling of blocks formed by rift and (sub-vertical) hardway discontinuity planes.

Fig. 6. a) Aerial view of the quarry showing three cross-cut sections; b) Representation of cross-cut sections profiles (upper-right) and estimated
fracture patterns in these cross-cut sections (lower-right) as obtained from field data and representation with the help of code UDEC [11].

4. Stability analysis.

Stability analysis of planar failure against sliding through a rift plane was performed for the higher slope of
the quarry (cross-cut section 3) and considering a tension crack 19 m behind the slope head as observed in place.
A water level was included to represent worst possible conditions, an horizontal seismic acceleration of 0.04·g was
given and rock and joint parameters have been input as derived from average field and laboratory estimates
(Fig. 7.a). With the help of Roc-plane code [6] an average factor of safety of 1.37 was estimated. Additionally
including measured or estimated variability a Monte Carlo approach was performed (Fig. 7.b) to estimate
a probability of failure of 1.64%.

Fig. 7. a) Input parameters, basic geometry and stability analysis by means of Roc Plane of the most likely planar failure contemplated in the
quarry slope; b) Monte Carlo analysis of the probability of failure.

Since the approach was carried out considering full continuity, no rock-bridges, no friction in the lateral surfaces
and full water filled joints, this is a conservative approach, so this type of failure is certainly highly unlikely to
occur.
Toppling failure of a rock block formed by sub-horizontal rift joints and sub-vertical hardway joints was
a possible kinematic failure mode, so it was also addressed. The section of this possible failure mode is sketched in
Fig. 8.
408 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409

Fig. 8. Sketch of the geometry of toppling failure of the largest possible overturning block depending on hardway spacing.

Factors of safety of a single rock block against toppling can be estimated as the ratio of the stabilizing moments
to the overturning moments. By performing the corresponding calculations [12], the formulation of the stabilizing
and overturning moments (basic, due to water presence and due to seismic considerations) for the toppling block
model can be obtained as in Eq. 1 in relation with Fig. 8:

M Stab. WT 1 sen 12  b T1-N + WR sen 12  b R-N  WT 2 sen 12  b T2-N

M Topl . basic WT 1 sen 12  b T1-N + WR sen 12  b R-N  WT 2 sen 12  bT2-N
M Topl . water U T bUt  U b bUb
M Topl . seism. DWT 1 sen 12  bT 1-P  DWR sen 12  bR-P  DWT 2 sen 12  bT 2-P 
 DWT 1 cos 12  bT 1- N  DWR cos 12  bR- N  DWT 2 cos 12  bT 2- N
(1)

For the case of hardway spacing around 19 m a large value of FoS well over 3 is obtained. So we have estimated
the spacing of the hardway that will make the slope unstable under different conditions. This spacing should be of
less than 3.5 m for dry and aseismic conditions, less than 5 m in the case of water table as in Fig. 8 and of less of 5.7
m for the case of seismic input and water presence (Fig. 9). Such small spacing is not observed behind the slope
face. Moreover, based on in-situ and aerial photography spacing observations, they are rather unlikely to occur in
the granite rock mass of the quarry. Accounting for the conservative assumptions associated to full continuity of rift
and hardway (estimated in the range of 10–20 m) planes, disregarding rock-bridges and conspicuous water flow and
disregarding stabilizing effect of lateral surfaces associated to grain planes; it can be considered that the slope is
very unlikely to present toppling failure modes.
Based on the results of the stability analysis against the two potential kinematic instability mechanisms, the slope
of the studied quarry will be perfectly stable in the coming years. Since very low probability of instability
phenomena were detected, a simple monitoring system based on the control of 5 blanks in the slope head every time
the topographic map of the quarry is updated (every 6 months) was recommended and implemented.
 L.R. Alejano et al. / Procedia Engineering 191 (2017) 402 – 409 409

Fig. 9. Histogram of spacing of the hardway joint set as measured in-situ and derived of aerial photographs in relation to the spacing producing
instability under various conditions.

5. Discussion and conclusions.

The authors have presented the stability analysis of an over-tilted slope in a granite quarry. In order to do that,
an appropriate characterization of the rock mass object of exploitation was carried out. It is relevant to note that
standard geotechnical characterization may not be sufficient in this kind of good geotechnical quality rock masses.
The ranges of spacing and continuity of discontinuities provided by ISRM Suggested methods are too broad, so
more accurateness is convenient to define possible failure mechanisms in a sufficiently approximated manner.
Additionally, other engineering geology issues, such as the presence of rock bridges or conspicuous flow may play
a stabilizing role, which is often disregarded in standard rock mass characterization practice. Once characterized
the rock mass, a stability study was carried out regarding the two most prone failure mechanisms identified both
theoretically and in place, namely planar and toppling failure. For both types of mechanisms average stable factors
of safety have been computed and very low probabilities of failure were estimated. Starting from these results,
the main quarry slope can be considered stable. Moreover, no failure was observed for the last three years.
Nevertheless, a simple monitoring program was proposed in order to be aware of possible movements of the slope.

Acknowledgements

The authors thank the Spanish Ministry of the Economy and Competitiveness for funding this research, awarded
under Contract Reference No. BIA2014-53368P, partially financed by means of ERDF funds from the EU.

References

[1] L.R. Alejano, D. Posada, A. Rodriguez-Dono, Servo-controlled strength tests on moderately weathered granite, in: I. Vrkjan (Ed.),
Symposium, Rock eng. in difficult ground cond, Soft Rock & Karst, EUROCK 2009, CRC. Taylor & Francis Gr, London, 2009, pp. 181–186
[2] L.R. Alejano, J. González, J. Muralha, Comparison of different techniques of tilt testing and basic friction angle variability assessment, Rock
Mech. & Rock Eng. 45 (2012) 1023–1035.
[3] J. Taboada, A. Vaamonde, A. Saavedra, Evaluation of the quality of a granite quarry, Eng. Geol. 53 (1999) 1–11.
[4] J. Taboada, L.R. Alejano, F. García-Bastante, C. Ordóñez, Total exploitation of an ornamental granite qarry (in Spanish), Mat. Const. 42/4
(2005) 481–507.
[5] Y. Chen, T. Nishiyama, H. Kusuda, H. Kita, T. Sato, Correlation between microcrack distribution patterns and granitic rock-splitting planes.
Int. J. Rock Mech. & Min. Sci. 36 (1999) 535–541.
[6] ISRM, The complete ISRM suggested methods for rock characterization, testing and monitoring: 1974-2006, R. Ulusay & J.A. Hudson
(Eds.), Ankara, Turkey, 2007.
[7] Rocscience software products, DIPS, Roc-Plane. Rocscience Inc., Toronto, Canada, 2012.
[8] D.G. Moye. Engineering geology for the Snowy Mountains scheme. J. Inst. Eng. Aust. 27 (1955) 281–299.
[9] J. Shang, S.R. Hencher, L.J. West, Tensile strength of geological discontinuities including incipient bedding, rock joints and mineral veins.
Rock Mech. & Rock Eng. 2016.
[10] E. Hoek, J.W. Bray. Rock slope engineering, Second ed., London, The Institution of Mining and Metallurgy, 1977.
[11] Itasca. UDEC - User manual for UDEC, Universal Distinct Element Method. Version 5, Itasca Consulting Group Inc. Minnesota, 2012.
[12] R.E. Goodman, J.W. Bray. Toppling of rock slopes. Proc. Sp. Conf. Rock Eng. for Found. and Slopes, ASCE. Boulder, Colorado, 1976.