You are on page 1of 10

SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN

Robert Bertuzzi
Pells Sullivan Meynink, Unit G3 56 Delhi Road, North Ryde NSW 2113, Australia

1 BACKGROUND
Inherent in any set of rock mass parameters are various assumptions regarding, amongst other things the depth of cover,
the proportion of materials encountered and the scale of the proposed excavations. The characteristics used to define
rock mass parameters of strength and stiffness comprise:
 rock type (lithology)
 strength of the rock substance (intact rock strength)
 fracturing of the rock mass by defects (bedding, joints, shears, etc.)
o persistence and spacings of defects
o number of defect sets
o infill material
o roughness of defects
 groundwater pressures
 reactivity of the rock substance to environmental change (shrink/swell, slaking)
The intention of this paper is to present typical geotechnical characteristics that can be used for tunnelling projects in
Sydney’s Hawkesbury Sandstone and Ashfield Shale based on the Sydney classification system (Pells et al 1998). It
provides the following information.
 A brief discussion about the Sydney rock mass classification system
 A table summarising suggested classification parameters for tunnelling projects for the 10 classes
 Validation of these suggested classification parameters from recent tunnelling projects
 Tables summarising suggested rock mass parameters for tunnelling projects. Two sets of material properties
are provided to cater for different scales as rock mass parameters are scale dependent.
o Overall tunnel scale – properties to be used in continuum analyses e.g. FLAC, Phase 2, PLAXIS, Abaqus.
Specific geological structures, one or two at most, can be included in these types of models.
o Approximately 1 to 2 m3 scale – properties to be used in discontinuum analyses where numerous
geological structures are explicitly modelled, e.g. UDEC and potentially jointed network finite elements
such as that available in Phase2
 Summary sheets showing examples of core photographs and typical Geological Strength Index (GSI) and Q values
for:
o Sandstone – Class I to V; Faults and Shears
o Shale – Class I to V; Fault and Shears.
This paper updates Bertuzzi and Pells (2002) with data from recent tunnelling projects. The database now includes
information from the Ocean Outfalls, Sydney Harbour Tunnel, M2, Eastern Distributor, M5 East, Cross City, cable
tunnels, Epping to Chatswood rail link, Lane Cove, CBD Metro, Wynyard Walk and the Northwest Rail link projects.
Detailed borehole logging from the last three projects has been especially used. This paper also brings into
consideration the rock mass behaviour types recommended by the Austrian Society for Geomechanics (2010). The
example of a nominally 8 m diameter TBM at depths of up to 50 m is used.
It is important for readers to appreciate that any set of design parameters also communicates the designer’s assumptions
regarding the ground conditions to those in the field responsible for implementing the designs. It is vital that those in
the field are vigilant in observing, documenting and interpreting geological structures that may dominate support
requirements at a particular location which would render the “average” rock mass class irrelevant and feed back to the
designers where conditions are different from those anticipated.

2 SYDNEY ROCK MASS CLASSIFICATION SYSTEM


The classification system for Sydney sandstone and shale - Pells et al. (1998) which updated Pells et al. (1978) - was
intended to assist in the design of foundations on rock in the Sydney area. The classification system is based on rock
strength, defect spacing and allowable seams as shown in Table 1. All three factors must be satisfied. Seams include
clay, fragmented or highly weathered zones.

Australian Geomechanics Vol 49 No 1 March 2014 1


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Table 1: Sydney Classification System (Pells et al., 1998)

DEFECT SPACING (a) ALLOWABLE


UCS
CLASS SEAMS
[MPa] [mm] Description %
I > 24 > 600 Widely spaced < 1.5
II > 12 > 600 Widely spaced <3
Sandstone

III >7 > 200 Moderately spaced <5


IV >2 > 60 Closely spaced < 10
V >1 NA NA
I > 16 > 600 Widely spaced <2
II >7 > 200 Moderately spaced <4
Shale

III >2 > 60 Closely spaced <8


IV >1 > 20 Very closely spaced < 25
V >1 NA NA
(a)
Defect spacing based on ISO/DIS 14689 & ISRM suggested methods replaced the degree of fracturing terms in the
Pells et al. (1978) paper
Pells et al. (1998) recommended that the zone of rock being classified be “over a length of core of similar
characteristics”. That is, “the classification system be applied to portions or units of rock mass having similar UCS,
defect spacing and seam characteristics” (Bertuzzi & Pells, 2002). An example profile of a mapped face was provided
in that paper to clarify the correct method of applying the classification system. This is reproduced in Figure 1.

Figure 1: The wrong and right way to classify (Bertuzzi & Pells, 2002)
While the Sydney Classification System was not intended for tunnelling, it does represent a good method for
communicating rock mass quality in Sydney sandstone and siltstone. It is also useful for linking values put forward for
designs with measured and back-figured parameters from existing excavations (e.g. Bertuzzi & Pells, 2002; Clarke &
Pells, 2007). Hence, it is often used as the basis for tunnelling projects in Sydney. However, the defect spacings which
are appropriate for foundation problems have been found to cover a too narrow range for tunnelling, which needs to
consider spacing in three dimensions.
Typical spacing for bedding and for jointing is therefore suggested in Table 2. It is not intended that this replaces the
Sydney Classification System (Pells et al., 1998) but rather used as a guide as to what to expect in the tunnelling
environment.
Table 2: Suggested conditions for tunnelling projects.

2 Australian Geomechanics Vol 49 No 1 March 2014


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

DEFECT SPACING [m]


ALLOWABLE
CLASS UCS [MPa] TYPICAL SEAMS
DEFECTS [%]
BEDDING JOINTS
I > 24 > 0.6 > 1.5 >2 < 1.5
II > 12 > 0.6 >1 >1 <3
Sandstone

III >7 > 0.2 > 0.5 > 0.5 <5


IV >2 > 0.06 > 0.2 > 0.2 < 10
V >1 NA NA
I > 16 > 0.6 >1 >1 <2
II >7 > 0.2 > 0.5 > 0.5 <4
Shale

III >2 > 0.06 > 0.1 > 0.2 <8


IV >1 > 0.02 > 0.1 > 0.1 < 25
V >1 NA NA

3 FIELD DATA VALIDATION


Recent tunnelling projects have provided a large database of drill core logs, comprising 6 km of logs containing over
7200 records of defect characteristics as well as field estimated strengths, point load index tests and laboratory UCS
tests. The data has been collected by several different geotechnical organisations and various geotechnical engineers
and engineering geologists. Hence, any bias of particular geotechnical organisations or individuals should be balanced
out in the collated data.
The database has been accessed to assess the variability in intact strength and the spacing, characteristics and
orientation of defects for each particular class of sandstone and shale. The variability is presented in Figures 2 and 3
and in Tables 3 and 4. The terms used in these tables are from the following standards and guidelines.
 Field estimate strength terms are those of AS 1729-1993 Geotechnical site investigations
 Defect spacing based on ISO/DIS 14689 & ISRM suggested methods
 Defect characteristics follow AS 1729-1993 Geotechnical site investigations
It is submitted that this data validates the bedding and joint spacing for tunnelling projects suggested in Table 2.
In terms of orientation, the recorded data shows bedding is obviously sub-horizontal (to 10°) in both the sandstone and
shale whereas joint patterns differ (Figure 4). In the sandstone, two sub-vertical joint sets are apparent - striking north-
northeast and east-southeast. While these two orientations are also apparent in the shale, other joint sets do occur
dipping between 30 to 50° in various directions.

Australian Geomechanics Vol 49 No 2 June 2014 3


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Figure 2: Distribution of logged spacing and aperture for bedding planes and joints for Hawkesbury Sandstone

Figure 3: Distribution of logged spacing and aperture for bedding planes and joints for Ashfield Shale

4 Australian Geomechanics Vol 49 No 1 March 2014


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Table 3: Variability in strength and spacing as recorded in the drillhole database.


TYPICAL SPACING
STRENGTH TOTAL CORE
RANGE [m]
CLASS AXIAL Is(50)[MPa] SAMPLE
FIELD
BEDDING JOINT LENGTH [m]
ESTIMATE TESTS RANGE AVERAGE
I High 1575 1.0 – 3.0 1.7 0.6 - > 6.0 0.6 - > 6.0 2478
II High 827 0.3 – 3.0 1.5 0.6 - 6.0 0.2 - 6.0 890
Sandstone

III Medium to High 625 0.3 – 3.0 1.4 0.2 - 2.0 0.2 - 2.0 590
IV Low to High 260 0.1 – 3.0 1.0 0.06 - 2.0 0.06 - 2.0 241
V Very low to High 41 <0.03 – 3.0 0.9 0.06 - 2.0 0.06 - 2.0 42
I Medium to High 69 0.3 – 3.0 1.5 0.2 - 6.0 0.2 - 6.0 471
II Low to High 130 0.1 – 3.0 1.3 0.2 - 6.0 0.2 - 6.0 499
Shale

III Low to High 100 0.03 – 3.0 1.0 0.06 - 2.0 0.06 - 2.0 271
IV Very low to High 31 <0.03 – 3.0 0.5 0.06 - 2.0 0.06 - 2.0 130
V Very low to High 10 <0.03 – 3.0 0.08 0.06 - 2.0 0.06 - 2.0 96

Table 4: Variability in the defect characteristics as recorded in the drillhole database.

DEFECT DEFECT CHARACTERISTICS DEFECTS


CLASS
TYPE ROUGH SHAPE APERTURE (mm) INFILL RECORDED
BG Rough Planar – Undulating Clean to 1 - 5 Clay 1066
I
JN Rough Planar – Undulating Clean - 315
BG Rough Planar - Undulating Clean to 1 - 10 Clay 952
II
JN Rough Planar Clean to Veneer Clay - Fe 230
Sandstone

BG Rough Planar Clean to 1 - 50 Clay - Fe 1369


III
JN Rough Planar Clean to Veneer Clay - Fe 511
BG Rough Planar 1 – 50 Clay - Fe 897
IV
JN Smooth - Rough Planar Clean to 1 - 5 Clay - Fe 376
BG Rough Planar - Undulating 1 – 100 Clay - Fe 117
V
JN Smooth - Rough Planar - Undulating 1–5 Clay – Fe 73
BG Smooth - Rough Planar Clean to Veneer Clay 106
I
JN Smooth - Rough Planar Clean - 217
BG Smooth - Rough Planar Clean to 1 - 5 Clay - Fe 314
II
JN Smooth - Rough Planar Clean to Veneer Clay - Fe 467
BG Smooth Planar Clean to 1 - 50 Clay - Fe 420
Shale

III
JN Smooth - Rough Planar - Undulating Clean to Veneer Clay - Fe 411
BG Smooth Planar 1 – 50 Clay - Fe 205
IV
JN Smooth - Rough Planar - Undulating Clean to 1 – 5 Clay - Fe 108
BG Smooth Planar 1 – 100 Clay - Fe 124
V
JN Smooth - Rough Planar - Undulating Clean to 1 – 5 Clay - Fe 18

Australian Geomechanics Vol 49 No 2 June 2014 5


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Based on the field data presented, the typically observed defect characteristics and those characteristics which can occur
but represent adverse conditions are summarised in Tables 5 and 6 for Hawkesbury Sandstone and Ashfield Shale,
respectively.
Table 5: Typical and adverse conditions of defects in Hawkesbury Sandstone.

PARAMETER TYPICAL ADVERSE


Persistence [m] > 20 > 20
Roughness Rough Slightly rough
Bedding Shape Undulating Planar
Aperture [mm] <1 > 10
Infill None / limonite Sandy clay
Persistence [m] > 10 > 10
Class I to V

Roughness Rough Slightly rough


Cross-bedding Shape Undulating Planar
Aperture [mm] <1 <1
Infill None / limonite None
Persistence [m] 5 > 10
Roughness Rough Slightly rough
Jointing Shape Planar Planar
Aperture [mm] <1 > 10
Infill None / limonite Sandy clay
Table 6: Typical and adverse conditions of defects in Ashfield Shale.

PARAMETER TYPICAL ADVERSE


Persistence [m] > 20 > 20
Roughness Slightly rough Smooth
Bedding Shape Undulating Planar
Class I, II & III

Aperture [mm] <1 > 10


Infill None / limonite Silty clay
Persistence [m] 3 >5
Roughness Slightly rough Smooth
Jointing Shape Planar Planar
Aperture [mm] <1 > 10
Infill None / limonite Silty clay
Persistence [m] > 20 > 20
Roughness Slightly rough Smooth
Bedding Shape Undulating Planar
Class IV & V

Aperture [mm] <1 50


Infill Clay coating Silty clay
Persistence [m] 3 >5
Roughness Slightly rough Smooth
Jointing Shape Planar Planar
Aperture [mm] <1 50
Infill Clay coating Silty clay
Notes:
Persistence of cross-bedding is controlled by sandstone bed thickness
Geotechnical models should consider and include adverse conditions, particularly that of defect persistence

6 Australian Geomechanics Vol 49 No 1 March 2014


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Figure 4: Equal area projections for Hawkesbury Sandstone on the left (4675 defects) and Ashfield Shale on the right
(1338 defects)

4 ROCK MASS PROPERTIES


The rock mass properties presented in Tables 7 and 8 are derived following the GSI approach (Hoek & Brown, 1997;
Hoek et al., 2002; Hoek & Diederichs, 2006) tempered with field measurements (e.g. Pells, 1990; Pells, 2004; Clarke &
Pells, 2007). Values of GSI and Q for the various classes are presented in the accompanying summary sheets. Table 9
lists suggested Mohr-Coulomb strength and stiffness parameters based on defect shear tests carried out for the
tunnelling projects.
In order to calculate the Q value an estimate of the in situ stress is required, or more correctly, the ratio of intact rock
strength to stress. Enever (1999) collated in situ stress data that suggested the following stepped profile for the upper
bound major horizontal stress (H):
 Approximately 2.5 MPa above the vertical stress (V) for depths less than 20 m
 Approximately 6.5 MPa above V for 20 – 200 m depths
 Approximately 15 MPa above V for 200 – 1200 m depths
Rather than adopting this profile with seemingly arbitrary steps, Pells (2004) suggested H can be related to V to
simplify the design process as: H = (1.2 to 2.0)V + 1.5. Adopting this same approach to the data presented in
McQueen (2004), the relationship, H = 2.0V + 2.5, is proposed as shown in Figure 5.
This means that the typical in situ principal stresses for depths less than 50 m (stresses at specific locations may be
different) can be assumed to be:
 σv up to 1.2 MPa
 σH ≈ 2.5 + 2.0 σv MPa, i.e. up to 5 MPa.

Australian Geomechanics Vol 49 No 2 June 2014 7


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Figure 5: Major horizontal stress versus depth with the design line H = 2.0V + 2.5 shown (after McQueen, 2004).

8 Australian Geomechanics Vol 49 No 1 March 2014


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Table 7: Hawkesbury sandstone

PARAMETER Class I Class II Class III Class IV Class V


Uniaxial compressive strength, UCS
30 25 15 10 5
(MPa)
Substance

Young’s modulus, E (MPa) 8000 6000 4000 3000 1000


Unit weight,  (kN/m3) 24
Poisson’s ratio,  0.25 0.3
mi 12

Mass Young’s modulus, Emass (MPa) 6000 4000 2500 1500


GSI 85 80 65 55

See tunnel scale


Not applicable
1-2 m3 scale

mb 7.023 5.874 3.438 2.406

parameters
Hoek-Brown (1) s 0.1889 0.1084 0.0205 0.0067
a 0.5 0.501 0.502 0.504
cꞋ (kPa) 1900 1200 400 250
Mohr-Coulomb (2) Ꞌ (°) 55 55 50 45
t (kPa) 500 250 100 25

Mass Young’s modulus, Emass (MPa) 3000 2000 1000 500 100
GSI 75 65 55 45 35
Tunnel scale

mb 4.914 3.438 2.406 1.683 1.178


Hoek-Brown (1) s 0.0622 0.0205 0.0067 0.0022 0.0007
a 0.501 0.502 0.504 0.508 0.516
cꞋ (kPa) 1000 500 300 200 150
Mohr-Coulomb (2) Ꞌ (°) 55 50 50 40 35
t (kPa) < 300 < 100 < 40 < 10 0

Table 8: Ashfield shale


PARAMETER Class I Class II Class III Class IV Class V
Uniaxial compressive strength, UCS
25 15 8 4 1
(MPa)
Substance

Young’s modulus, E (MPa) 6000 4500 2000 1000 250


Unit weight,  (kN/m3) 24
Poisson’s ratio,  0.25 0.3
mi 8

Mass Young’s modulus, Emass (MPa) 3500 2000 500


GSI 65 60 50
See tunnel scale
Not applicable
1-2 m3 scale

mb 2.292 1.917 1.341


parameters

Hoek-Brown (1) s 0.0205 0.0117 0.0039


a 0.502 0.503 0.506
cꞋ (kPa) 600 350 180
Mohr-Coulomb (2) Ꞌ (°) 50 45 38
t (kPa) 200 100 20

Mass Young’s modulus, Emass (MPa) 2000 1000 300 110 15


GSI 55 50 40 30 20
Tunnel scale

mb 1.604 1.341 0.939 0.657 0.459


Hoek-Brown (1) s 0.0067 0.0039 0.0013 0.0004 0.0001
a 0.504 0.506 0.511 0.522 0.544
cꞋ (kPa) 400 250 150 90 40
Mohr-Coulomb (2)
Ꞌ (°) 45 40 35 25 15
t (kPa) < 100 < 60 < 10 0 0
1. Damage parameter D of 0, assumes minimal disturbance to rock mass surrounding tunnel
2. Normal stress range of 0 to 2 MPa assumed

Australian Geomechanics Vol 49 No 2 June 2014 9


SYDNEY SANDSTONE AND SHALE PARAMETERS FOR TUNNEL DESIGN ROBERT BERTUZZI

Table 9: Defect properties.


SHEAR
STRENGTH STIFFNESS
INFILL
THICKNESS (MPa/m)
DEFECT TYPE INFILL TYPE MODULUS
(mm) cꞋ Ꞌ (MPa)
(kPa) (°) NORMAL SHEAR
kN kS
Tight Clean or hard mineral 0 40 NA >10000 >1000
Sandstone
1-5 Firm clay 50 30 20 6000 600
Bedding, joints
5-10 Soft clay 10 20 5 800 80
Tight Clean 0 30 N/A >10000 >1000
Shale
1-5 Firm clay 50 25 20 6000 600
Bedding, joints
5-10 Soft clay 10 20 5 800 80
Faults, shears, Equivalent to
Soft clay 10 20 5 150 15
erosional contacts 50 mm

5 ACKNOWLEDGEMENTS
The author is indebted to his colleagues Andrew de Ambrosis, Derek Anderson, Mark Eggers and Ben Rouvray, for
their constructive comments and opinions; and Dan Sandilands and James Smith for their manipulation of spreadsheets.
It is hoped that practitioners will find this paper useful in their work in Sydney.

6 REFERENCES
AUSTRIAN SOCIETY FOR GEOMECHANICS, 2010. Guideline for the geotechnical design of underground
structures with conventional excavation. Salzburg, Austria.
BERTUZZI, R. & PELLS, P. J. N. 2002. Geotechnical parameters of Sydney sandstone and shale. Australian
Geomechanics Journal, 37, 41-54
CLARKE, S. & PELLS, P. J. N. 2007. A large scale cable jacking test for rock mass modulus measurement, Lucas
Heights, Sydney
ENEVER, J. R. 1999. Near surface in-situ stress and its counterpart at depth in the Sydney metropolitan area.
Australian Geomechanics, 65-76
HOEK, E. & BROWN, E. T. 1997. Practical estimates of rock mass strength. International Journal of Rock Mechanics
& Mining Sciences, 34, 1165-1186
HOEK, E., CARRANZA-TORRES, C. & CORKUM, B. Hoek-Brown failure criterion - 2002 edition. North American
Rock Mechanics Symposium, 2002 Toronto
HOEK, E. & DIEDERICHS, M. S. 2006. Empirical estimation of rock mass modulus. International Journal of Rock
Mechanics and Mining Sciences, 43, 203-215
INTERNATIONAL SOCIETY FOR ROCK MECHANICS, 1978. Suggested Method for Quantitative Description of
Discontinuities in Rock Masses
ISO 14689-1:2003. Geotechnical investigation and testing -- Identification and classification of rock -- Part 1:
Identification and description
MCQUEEN, L. B. 2004. In situ rock stress and its effect in tunnels and deep excavations in Sydney. Australian
Geomechanics, 39, 43-58
PELLS, P. J. N. 1990. Stresses and displacements around deep basements in the Sydney area. 7th Australian Tunnelling
Conference. Sydney, Australia
PELLS, P. J. N. 2004. Substance and mass properties for the design of engineering structures in the Hawkesbury
Sandstone. Australian Geomechanics, 39, 1-21
PELLS, P.J.N., MOSTYN, G. & WALKER, B.F. 1998. Foundations on Sandstone and Shale in the Sydney Region.
Australian Geomechanics Journal , 33 Part 3

10 Australian Geomechanics Vol 49 No 1 March 2014

You might also like