Professional Documents
Culture Documents
VICTORIA CHAPTER
PREFACE
The Victorian Chapter of the Australian Geomechanics Society (AGS) is pleased to
announce the 2023 AGS Victorian Symposium titled “Novel Solutions in Geotechnical
Engineering” to be held on 25 October 2023.
The symposium forms part of the continuing program of events organised by the Victoria
Chapter of the AGS. The event is designed to showcase recent novel solutions and
innovations to address geotechnical challenges in practice. This symposium will be held as
a face-to-face event in Melbourne.
For the past decade, Australia and particularly Victoria has seen an unprecedented boom
in infrastructure investment involving mega-projects, transport and rail developments.
These significant infrastructure developments have generated new geotechnical
challenges due to complex geological condition, spatial variability of the ground and urban
construction. Adding to these challenges are the need for higher efficiency, cost saving and
consideration for sustainable developments. All these challenges and requirements have
spurred novel and innovative solutions across all aspects of geotechnical engineering.
Fuelling the geotechnical creativity and innovations is the advancement in technology
and knowledge, which allows novel solutions and tools to be developed to tackle
these geotechnical challenges using a variety of approaches (e.g. advanced numerical
modelling, physical testing, data-driven solutions, field instrumentation and monitoring).
This symposium seeks to bring together the local geotechnical community to share
and discuss recent geotechnical advances, with a particular emphasis on tackling the
geotechnical challenges in the Melbourne region and regional areas of Victoria. To
promote engagement with young engineers, the Symposium also hosts an inaugural
Young Geoprofessionals Poster Exhibition. In addition, a Heritage Time Capsule (HTC)
session has been included to collect feedback from the geotechnical community
regarding Geo-Education (current state, needs and challenges) in an Australian context.
This valuable feedback will contribute towards the development of our next generation
geotechnical workforce.
All right reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form without the permission of the Australian Geomechanics Society.
© 2023 Australian Geomechanics Society.
TECHNICAL REVIEWERS
All papers have been peer reviewed in accordance
with the AGS Symposia review procedure.
SPONSORS
PLATINUM
Chadwick Geotechnics
GOLD
Insitu Geotech Services Pty Ltd
Global Synthetics Pty Ltd
SILVER
Black Insitu Testing
Insitutek
Rocscience
Cover image: Don Valley Brick Works Park, courtesy of Zhenyu Liu
SESSION 1 1 SESSION 3 41
NOVEL GEOTECHNICAL ANALYSIS ROCK BEHAVIOUR IN MINING
APPLICATIONS
Numerical Analysis of the Melbourne 3
Metro State Library Station Cavern Novel Solutions in Utilising Underground Coal 43
B. Coombes, D. Sainsbury and R. Storry Seam Gas Reservoirs for Gas Storage/Recovery
M. S. A. Perera
Interpretation of Shape Array Inclinometer 9
Data to Assess Trigger Action Levels Vein Effect on Mechanical Properties and 49
in Tailings Dams – A Case Study Fracture Behaviour of Porphyry Rock
M. Naeini and B. G. Nejad H. J. Wang and Q. B. Zhang
SESSION 2 21
CHALLENGING GEOTECHNICAL
CONDITIONS
Go to table of contents
SCHEDULE
8:30 am Welcome and opening remarks SESSION 3
ROCK BEHAVIOUR IN MINING
SESSION 1 APPLICATIONS
NOVEL GEOTECHNICAL ANALYSIS Session Chair: Yunlong Tang (Tonkin+Taylor)
Session Chair: Shiaohuey Chow (University of Melbourne) 1:35 pm KEYNOTE ADDRESS
8:40 am KEYNOTE ADDRESS Novel Solutions in Utilising Underground
Numerical Analysis of the Melbourne Metro Coal Seam Gas Reservoirs for Gas Storage/
State Library Station Cavern Recovery
Ben Coombes (NOMA Consulting) Samintha Perera (University of Melbourne)
Go to table of contents
YOUNG GEOPROFESSIONAL
POSTER EXHIBITION
Go to table of contents
SPONSORS
PLATINUM SPONSOR
CE
G
INSITU
S
re d u
cin
ge
ain
ty
g
ote
c h n i c a l u n c e rt
SILVER SPONSORS
Go to table of contents
OVER 25 YEARS
OF ENGINEERING
TESTING EXCELLENCE
chadwickgeotechnics.com.au
Australia’s growing investment in major new
infrastructure projects across a range of sectors
is driving a surge in demand for high-quality
geotechnical data.
TECH SERV
EO I
CE
G
INSITU
COMPREHENSIVE RANGE OF
redu
ote ai
c h nical u n c ert
mick@insitu.com.au mark@insitu.com.au
0407 467 025 0437 824 776 www.insitu.com.au
BRISBANE | MELBOURNE | SYDNEY | HOBART | PERTH | TOWNSVILLE | PNG
Applied Landslide
Risk Assessment
Course
The Australian Geomechanics Society is
running the Applied Landslide Risk Assessment
Course in Victoria in February 2024.
ABSTRACT
The Melbourne Metro Tunnel Project is delivering twin nine-kilometre rail tunnels in Melbourne, Australia. In addition
to the tunnels, five new underground stations are being constructed. Two of the new stations – State Library and
Town Hall – are complex cavern and adit excavations located in Melbourne’s City Centre. The State Library Station,
located predominantly underneath Swanston Street and a busy tram route, is surrounded by a mixture of modern,
educational and heritage developments requiring the excavation sequence and primary support to be designed to
ensure minimal surface impacts. To simulate the anisotropic rock mass response to the excavation of the State
Library Station, FLAC3D numerical analysis was undertaken. The analysis adopted the ubiquitous joint constitutive
model approach and was used to assess the performance of the primary lining design and to determine the impacts
the predicted ground displacements may have on the surrounding structures. Calibration and validation of the
ubiquitous joint modelling parameters for the Melbourne Formation have been completed against the measured
and observed responses throughout excavation. This was also completed against the structural elements, which
are often overlooked with respect to calibration and validation.
The Melbourne Metro Tunnel Project is a The Melbourne Formation (also known as the
metropolitan rail infrastructure project located in Melbourne Mudstone) is an anisotropic rock mass,
Melbourne, Australia. The project comprises the both in terms of its intact properties and in terms of
construction of twin nine-kilometre tunnels with five its structural geological features.
underground stations. The two stations under
Swanston Street, State Library and Town Hall, were The predominant rock type in the Melbourne Formation
constructed as mined trinocular form caverns with the is siltstone, which typically com-prises 70% to 80% of
central ‘station’ tunnel excavated prior to the outer the formation. Bedding in the siltstone is typically well-
‘platform’ tunnels using sequential excavation developed and varying in the range 10 mm to 300 mm
methods (Figure 1). (Storry et al. 2023), provides a thick, but more typically in the range 30 mm to 60 mm
broad description of the project and construction thick. The remainder of the formation comprises
techniques used. sandstone and rare (<2%) beds of con-glomerate and
claystone. The bedding defects are highly persistent,
limited only by fold hinge axes, major fault structures
and igneous dykes (Figure 2), which resulted in
anisotropic strength and deformation behaviour.
Go to table of contents
Numerical Analysis of the Melbourne Metro State Library Station Cavern B. Coombes, D. Sainsbury and R. Storry
colour change from light yellow and grey through yellow design packages prior to construction. The
browns to dark grey (Paul et al. 2017). At the State advantage of using this technique was the improved
Library Station, the weathering profile of the encountered understanding of the rock mass response with the
Melbourne Formation was complex because of the inclusion of structural elements, allowing for
geological structure (intrusive igneous dykes and folds) improved accuracy with respect to predicted
and varied significantly over short distances. structural element forces, allowing greater efficiency
with the design of these elements. This numerical
3 NUMERICAL ANALYSIS analysis approach also had the benefit of a more
realistic rock mass response for inclusion in the
3.1 Background ground movement and building impact assessments
given the dense, urbanised environment that the
Geotechnical numerical modelling techniques can be State Library Station was constructed.
broadly grouped into continuum, discontinuum, and
equivalent continuum methods. The most common 3.2 Constitutive Model
equivalent continuum method is the ubiquitous (or
subiquitous) joint constitutive model which has been The Ubiquitous Joint model corresponds to a Mohr-
implemented within the finite difference codes FLAC, Coulomb material that exhibits a well-defined strength
FLAC3D, UDEC and 3DEC (Itasca, 2018). Due to the anisotropy due to embedded planes of weakness. As
inherent nature of the ubiquitous joint formulation, it shown in Figure 3(a), the planes of weakness can be
has been found that direct use of discontinuum assigned different orientations for each zone in the
properties in a ubiquitous joint model without careful model. The criterion for failure on the plane of
calibration of the material response can provide weakness consists of a composite Mohr-Coulomb
misleading model results by not capturing the rock envelope with a tension cut-off. The propagation of
mass response accurately (Sainsbury et al. 2017; damage within a Ubiquitous-Joint model can be
Perras and Diederichs, 2009). observed through the progressive degradation of
matrix cohesion and ubiquitous joint-failure plots at
Various discontinuum modelling techniques are various stages of loading in a simulated unconfined
available that explicitly simulate joints and compressive strength test, illustrated in Figure 3(b).
discontinuities within an anisotropic rock mass. Due
to the computational intensity of these numerical The Ubiquitous-Joint model formulation assumes
techniques, it is not practical to explicitly simulate the infinitesimal spacing and no length scale to their
joint fabric of an entire rock mass for routine analyses implementation. As such, a ubiquitous-joint material
of large-scale excavations. To overcome this, the cannot account for the bending stiffness of the
continuum based ubiquitous joint constitutive model individual layers of rock. As demonstrated by (Leitner
is commonly used to represent the strength and et al. 2006; Sainsbury and Sainsbury 2017), the
deformation behaviour of anisotropic rock masses selection of matrix and joint properties based on direct
(Clark 2006; Leitner et al. 2006). input of the measured block and joint strength will
result in a simulated material response that does not
This approach to numerical analysis was completed represent the true rock mass strength or deformational
for all adits and cavern primary ground support profile and provide misleading model results.
Figure 3. (a) Ubiquitous joint model: matrix and joint properties, (b) stress-strain response of simulated ubiquitous
joint rock mass (Sainsbury et al. 2008)
Go to table of contents
Numerical Analysis of the Melbourne Metro State Library Station Cavern B. Coombes, D. Sainsbury and R. Storry
To provide meaningful modelling results, careful During the tender and detailed design phases of the
calibration of the matrix and ubiquitous joint Metro Tunnel project, detailed UJRM parameter
parameters to the emergent behaviour from calibration was conducted based upon laboratory
discontinuum modelling techniques and in situ testing and field measurement data (including
monitoring and observation is required. A detailed bedding spacing and persistence measurements at
Ubiquitous Joint Rock Mass (UJRM) calibration notable exposures of Melbourne Formation).
procedure to account for rock mass anisotropy in
open pit rock slopes, block cave mines, deep mine Figure 4 illustrates the response of a large-scale,
access development and civil tunnelling has been discontinuum (3DEC), synthetic rock mass specimen
developed continuously since 2006 (Clark, 2006; on Melbourne Formation with horizontal bedding
Sainsbury and Sainsbury 2017; Sainsbury et al. (β=0). Figure 5 presents the calibrated continuum
2016; Johnson et al. 2016). This included back- (FLAC3D) strength response to the discontinuum
analysis of historical underground excavations in the (3DEC) model response at bedding orientations from
Melbourne Formation, notably the Melbourne horizontal to vertical. The emergent strength
Underground Rail Loop and City Link Domain Tunnel anisotropy in both discontinuum and continuum
were conducted and reported by Sainsbury and modelling approaches can be observed.
Amon (2017).
Figure 5. Calibrated UJRM anisotropic strength response compared to the discontinuum anisotropic response for
Melbourne Formation
Go to table of contents
Numerical Analysis of the Melbourne Metro State Library Station Cavern B. Coombes, D. Sainsbury and R. Storry
Go to table of contents
Numerical Analysis of the Melbourne Metro State Library Station Cavern B. Coombes, D. Sainsbury and R. Storry
CH9200
CH9335
nearly 300 SMPs plus a suite of utility monitoring CL CL
0.0
points and building monitoring points.
-5.0 Predicted
A’BECKETT ADIT
bound K0 simulations. -30.0 9220 FRANKLIN ADIT
-35.0
(a)
-40.0
FRANKLIN EAST
9190
9200
9210
9230
9240
9250
9260
9270
9280
9290
9300
9310
9320
9330
9340
9350
9360
SHAFT
CHAINAGE
CENTRAL TUNNEL
CH9335
DYKE(S)
6.3 Structural Element Settlement
PROFILE
EXTENT
Settlement
(mm) FRANKLIN ADIT Figure 11 shows the measured settlement of the
0.0 western steel columns compared to the settlement
-5.0
-10.0 predictions with both the best-estimate and lower-
-15.0
-20.0
bound K0 simulations.
-25.0
FRANKLIN EAST
-35.0 TUNNEL
CL
0
-40.0
-2
Predicted
(Design K0)
-4
Predicted
CH9200
CH9335
SETTLEMENT (mm)
(Low K0)
-6
Measured
CH9305
-8
DYKE(S)
PROFILE -10
EXTENT
A’BECKETT ADIT
FRANKLIN ADIT
-12
WESTBOUND PLATFORM
TUNNEL
(b) -14
-16
Figure 9. Estimated settlement for the (a) central
9220
9230
9240
9250
9260
9270
9280
9290
9300
9310
9320
9330
9340
9350
9360
Predicted settlement from the central tunnel (Figure Figure 11. Calibration of predicted versus measured
9(a)) above Franklin Adit was predicted to be 15mm settlement for western steel columns (structural
with a maximum predicted settlement at CH9295 of element calibration)
20mm. With continued excavation of the platform
tunnels, the predicted settlement from the excavation
Go to table of contents
Numerical Analysis of the Melbourne Metro State Library Station Cavern B. Coombes, D. Sainsbury and R. Storry
8 ACKNOWLEDGEMENTS
Go to table of contents
Interpretation of Shape Array Inclinometer Data to Assess Trigger Action
Levels in Tailings Dams – A Case Study
M. Naeini1 and B. G. Nejad2
1
Senior Geotechnical Engineer, ATC Williams, Melbourne Victoria 3195. Email: MahdiN@atcwilliams.com.au
2
Chief Technical Officer, ATC Williams, Melbourne Victoria 3195. Email: BehroozG@atcwilliams.com.au
ABSTRACT
This article investigates the displacement data from Shape Array Inclinometers in a tailings dam constructed on a
bed of clay and silty sand overlying the bedrock. During embankment construction, two inclinometers indicated
rapid movements at the interface of sand and clay layers, which triggered further action with the operation being
temporarily ceased. Consequently, a numerical simulation was conducted to properly simulate the development of
stresses within the foundation materials during construction and operation, and to better understand the extent of
foundation/embankment deformation and its pattern. Staged construction of the embankment was simulated by
employing FLAC software (Itasca group), adopting suitable constitutive models and associated input parameters
for each material. The model parameters were calibrated based on a series of field and laboratory tests to replicate
the observed movements in the inclinometers. It was found that the inclinometer's movements were localised and
caused by the adopted staged construction technique and not indicating a global sliding mechanism through the
foundation clay. It is concluded that despite the significant benefits of monitoring displacements by inclinometers,
adopting trigger action levels purely on this basis might be misleading.
Keywords: inclinometer data, trigger level assessment, numerical simulation, FLAC, tailings dam
Go to table of contents
Interpretation of Shape Array Inclinometer Data to Assess Trigger M. Naeini and B. G. Nejad
Action Levels in Tailings Dams – A Case Study
The foundation beneath the embankment consists of possibility of static liquefaction with the increase in
fresh and weathered bedrock, lower clay, sand the stresses induced by embankment construction
(upstream side and downstream side), and upper and tailings deposition. The MCC constitutive model
clay. The complexity of the foundation layering has was implemented for the clayey foundation material
resulted in several scenarios being analysed with to adequately simulate the deformation
changing material characteristics and constitutive characteristics of this layer throughout the staged
models throughout the development of embankment construction loading.
construction and tailings deposition stages.
For all analyses undertaken, an embankment model
Two-dimensional plane strain, stress-deformation was generated, material properties were assigned
analyses simulating the development of the Tailings and then the model was run so stress states within
Dam to final configuration were undertaken using the embankment and foundation materials could be
FLAC (ITASCA group, 2022) and its built-in and user- progressively established. Sensitivity analyses were
defined constitutive models. Staged construction undertaken to optimise the mesh size within the
simulations were undertaken, from the starter model. Square zone (mesh) sizes of 0.5 m and 1 m
embankment construction and tailings deposition to were assessed, and it was determined that a mesh
the final design height. Considering the design of all size of 1 m provided reasonably close results to the
raises as well as the available piezometer data during mesh size of 0.5 meters while maintaining
construction, three piezometric lines were defined at reasonable runtimes for the analyses. This zone size
the end of the construction of each stage, to simulate was adopted to also allow conformance to wave
the pore pressure development within the tailings, propagation criteria required for subsequent dynamic
sandy and clayey foundation zones and deformation analyses, which are not the subject of
embankment materials. this paper.
A range of constitutive models from Mohr-Coulomb Five static models are presented in this study based
(M-C), NorSand, and Modified Cam-Clay (MCC) was on the results of the sensitivity analyses for different
employed to simulate the behaviour of embankment constitutive models and the adopted material
zones and foundation materials. The M-C elastic- parameters. These models have been assessed in
plastic model has been adopted for embankment terms of total deformation, stress distributions, and
materials and was initially used for both sandy and shear strain development for the Embankment
clayey foundation zones (i.e., the starting point from Raise 7 (existing) and Raise 8 (future) design
which to improve). Several sensitivity analyses were arrangements. Most importantly, the Raise 7 models
completed to calibrate the results with the overall were calibrated against inclinometer monitoring data
deformations of the available inclinometer records. (Figure 1).
The constitutive model for the sandy foundation
materials was replaced with NorSand to assess the
Table 1 summarises the key differences between the The results were assessed in terms of the following:
model runs for different scenarios. Run 1 (preliminary • Overall deformations for the embankment and
analysis) was carried out to evaluate the functionality foundation materials;
of staged construction simulation, verify the definition • Comparison of predicted foundation
of piezometric lines, and assess the suitability of deformations with the inclinometer records with
material parameters. In Run 1, the M-C elastic-plastic the focus being on foundation deformations; and
constitutive model was used for all materials, except • Developed shear strains at the end of each
for bedrock layers with adopted elastic model. construction stage.
Go to table of contents
Interpretation of Shape Array Inclinometer Data to Assess Trigger M. Naeini and B. G. Nejad
Action Levels in Tailings Dams – A Case Study
The Run 1 results were considered the starting point, compared the results against monitoring data from
including the embankment Raise 7 geometry inclinometers S7 and SN17 (see Figure 1). The only
(existing) with subsequent model runs including difference between Runs 4 and 5 was the constitutive
changes in constitutive models, and adopted model adopted for the clay layers.
parameters for Embankment Raises 7 and 8.
2.2 Material parameters
Table 1: Staged construction runs
Foundation Table 2 presents the material properties adopted for
Constitutive Model the numerical modelling. Most parameters for
Upper Interface different zones were determined from extensive field
Run No. Sand
and Layer and laboratory investigations, including field CPTu
and
Lower with shear wave velocity measurement, triaxial,
Gravel
Clay monotonic and cyclic simple shear, oedometer and
1 Mohr- Mohr- CRS, direct shear, and ring shear testing. However,
No
(preliminary) Coulomb Coulomb if data were not available, reasonable assumptions
Mohr were made based on other data published in the
2 NorSand No literature.
Coulomb
Modified
3 Cam- NorSand No The large-strain shear moduli adopted in the staged
Clay construction simulations were factored from the small
Mohr strain shear modulus at approximately one percent
4 NorSand Yes shear strain. The maximum shear modulus, Gmax, is
Coulomb
Modified a measure of the stiffness of the material at very
5 Cam- NorSand Yes small strains. The Gmax values were formulated
Clay (based on the test data where applicable) using the
empirical relationship developed by Seed and Idriss,
Runs 2 to 5 were conducted to assess the foundation as described in Kramer (1996).
sand susceptibility to static liquefaction, predict upper
bound deformations of the foundation clay layers, The variations in Gmax with 𝜎𝜎𝜎𝜎′𝑚𝑚𝑚𝑚 (mean effective stress
and evaluate the performance of different constitutive or p′) for foundation sand and gravel materials were
models. Run 2 examined the functionality of the determined using the available field Seismic Cone
NorSand model, while Run 3 used the MCC model to Penetration Testing (SCPT) data. Similarly, the Gmax
predict overall deformation in the foundation clay values for the foundation clay layers were determined
layers. Runs 4 and 5 analysed the effects of from the shear wave velocities obtained during field
foundation deformation at the interface layer investigation as well as bender element tests on
between the sand and upper clay layers and undisturbed triaxial samples.
Go to table of contents
Interpretation of Shape Array Inclinometer Data to Assess Trigger M. Naeini and B. G. Nejad
Action Levels in Tailings Dams – A Case Study
The stiffness of a material reduces at larger strains, During construction, S7 inclinometer consistently
and this is described by the Modulus Reduction presented larger horizontal movements at the
function. To estimate the large strain shear modulus contact of the sand and upper clay layers. In model
(G), the Gmax values were converted using different Runs 4 and 5, an interface zone with a friction angle
modulus reduction curves for each material type. of 15° was adopted at this location across the model
This procedure was used only for elastic and M-C to assess if a global slippage at the contact can be
constitutive models, while other models such as justified. In these model runs, the stiffness and
NorSand and MCC can effectively simulate the strength properties of the foundation soils were also
non-linear stress-strain response of soil. adjusted to predict similar results to the Raise 7
displacement records from S7.
Table 3 presents the MCC model parameters used in
Runs 3 and 5. The parameters were selected based 3 RESULTS AND DISCUSSIONS
on the laboratory test results and sensitivity analyses
with further calibration undertaken using the field This section provides a summary of the results of the
displacement records. static stress-strain analyses (staged construction) for
the identified critical cross-section, with a particular
Table 3: Modified cam-clay constitutive model focus on comparing the results with the available
parameters deformation data for inclinometers S7 and SN17.
Run No.
Parameters
3 5 Figure 2 presents the foundation horizontal
λ 0.03 displacements at the end of Embankment Raise 7
κ 0.01 (existing conditions) for different model runs together
M with those recorded at S7 inclinometer (i.e., actual).
1.03 1.33 The first three runs were aimed at replicating the
overall foundation deformation profile recorded by S7
νλ 0.840
inclinometer. As can be seen from Figure 2, all three
Pre-consolidation runs provide a conservative (upper bound)
400 600
stress, p´0 (kPa) deformation profile, particularly in the sand layer.
p′0: preconsolidation pressure,
λ: slope of critical state line (CSL) or isotropic normal Runs 4 and 5 were to evaluate the effect of the
consolidation line (INCL) in Specific Volume (ν) vs
interface layer on the predicted deformation profile.
Mean Effective Stress (p´) plane,
κ: slope of swelling (unload/reload) line in ν-p´ plane, The stiffness of the sand and upper clay layers was
M: frictional constant, slope of the CSL in q-p´ plane also adjusted to closely simulate the overall
νλ: reference specific volume on the INCL deformation response in the first three runs. The
corresponding to p′ = 1 kPa results show that Runs 4 and 5 can predict similar
displacements to Runs 1-3 at S7 inclinometer.
Table 4 presents the NorSand model parameters
adopted for the foundation sand layer in Runs 2 to 5. Based on the preliminary analyses, Run 2 was adopted
The NorSand model parameters were calibrated and progressed as the seed model for prediction of
based on multiple triaxial tests on loose/dense embankment and foundation deformations due to
samples under both drained and undrained construction of subsequent embankment raises. Run 2
conditions. model predicted similar responses to Runs 1 and 3,
and can effectively assess the potential for static
Table 4: NorSand constitutive model parameters liquefaction of the foundation sand.
Run No.
Parameter
2 3 4 5
Gref 27,500
Elastic m 0.5
µ 0.2
Γ 0.53
Critical state λ 0.024
M 1.32
N 0.3
Plasticity
H 50 50 200 200
parameters
χtc 4.5
State
ψ -0.2
parameter
Gref: reference shear modulus,
m: exponent of mean effective stress ratio,
µ: Poisson’s ratio,
Γ: Reference void ratio on the CSL corresponding
to p´ = 1 kPa, Figure 2. Comparison of results with S7 inclinometer
N: Volumetric coupling coefficient, slope of stress data
dependency plot
H: Plastic hardening modulus, Figure 3 compares the recorded horizontal
χtc: Constant relating dilatancy to the state displacements of the foundation materials in Raises 5
parameter.
to 7 with those predicted by Run 2. As can be seen,
Go to table of contents
Interpretation of Shape Array Inclinometer Data to Assess Trigger M. Naeini and B. G. Nejad
Action Levels in Tailings Dams – A Case Study
the magnitudes of the predicted deformations at S7, Figure 5 illustrates the predicted shear strain
although relatively small, are in reasonable agreement increment contours for the final geometry (Stage 10)
with the recorded displacements at this location. of embankment construction. The vertical scale has
been exaggerated with the embankment boundary
The predicted deformations for Raise 8 are also schematically shown.
depicted in Figure 3, which are slightly higher than
those for Raise 7, as expected. During the The localised shear strain increments are set to
construction of Raise 8, the S7 inclinometer was develop in two main locations of the foundation, one
damaged, and as a result, the predicted deformation under the downstream side of the starter
for Raise 8 could not be verified against the actual embankment and the other near the S7 inclinometer
data at this location. location. It is evident that the developed shear strains
are localised and not continuous across the entire
foundation (i.e. a potential failure plane), as also
confirmed by the recorded displacements at SN17.
Go to table of contents
Interpretation of Shape Array Inclinometer Data to Assess Trigger M. Naeini and B. G. Nejad
Action Levels in Tailings Dams – A Case Study
are not susceptible to the static liquefaction during adopting trigger action levels purely based on a
construction and operation stages. limited number of inclinometer monitoring data
• The movements recorded by the inclinometers are might be misleading. Employing numerical
localised and caused by the adopted staged simulation methods has the potential to evaluate
construction technique and not indicating a global the degree of movements and its influence on the
sliding mechanism through the foundation clay. overall performance of the structure.
• It is concluded that despite the significant benefits Shear strain increment
of displacement monitoring by inclinometers,
Figure 5. Predicted shear strain increment contours at the end of construction (not in scale)
REFERENCES Konno, T., Hatanaka, M., Ishihara, K., Ibe, Y. and Izuka, S.
(1994), “Gravelly Soil Properties Evaluation by Large
Abuel-Naga, M. A., & Abdel-Gawad, F. M. (2018). Application Scale In-situ Cyclic Shear Tests, Ground Failures Under
of shape arrays in the evaluation of the deformation Seismic Conditions” Geotechnical Special Publication,
behavior of the foundation of a tailings dam. Journal of No. 44, ASCE Reston, Va., 177-200.
Rock Mechanics and Geotechnical Engineering, 10(3), Li, H., Chen, Y., Zhang, J., & Li, G. (2018). Numerical analysis
496-504. and field monitoring of the stability of a tailings dam
Hagan, P. C., Walton, G., Galvin, J., & Gu, Z. (2017). The use foundation. Arabian Journal of Geosciences, 11(10), 231.
of shape arrays in the monitoring of tailings dams. In Schnabel, P.B. (1973), “Effects of Local Geology and Distance
Proceedings of the 20th International Seminar on Paste from the Source on Earthquake Ground Motions”, PhD.
and Thickened Tailings (pp. 237-246). Thesis, University of California, Berkeley, California.
ITASCA Consulting Group, “FLAC Version 8.1 Manual”, Vucetic, M. and Dobry, R. (1991), “Effect of Soil Plasticity on
Minnesota, USA. Cyclic Response”, Journal of Geotechnical Engineering,
Jefferies, M. and K. Been, “Soil Liquefaction - A Critical State Vol. 117, No. 1. ASCE, pp. 89-107.
Approach (Second Edition)”, Taylor & Francis Group, Williams, J. R., Harrison, J. P., & Young, R. P. (2015). Seismic
2016: p. 712. deformation monitoring of a tailings dam foundation
Kachanov, S. T., Zhu, Y., & Yin, J. (2018). Application of shape using a shape array. Soil Dynamics and Earthquake
array calibration in the monitoring of tailings dam Engineering, 72, 90-101.
foundations. In Proceedings of the 21st International Zavodni, B., Maloney, S., Eitzenberger, A., & Hagan, P. C.
Seminar on Paste and Thickened Tailings (pp. 33-43). (2016). Monitoring of a tailings dam foundation using a
shape array. In Proceedings of the 19th International
Seminar on Paste and Thickened Tailings (pp. 101-108).
Go to table of contents
Back Analysis, Monitoring and Remediation of Instability in a Soil Nail Wall in
Fissured Basaltic Clay
A. E. M. Dupleix1, BEng GradIEAust, R. A. Day2, MEngSc FIEAust CPEng NER
and D. N. Hagues3, MSc DIC CGeol
1
Geotechnical Engineer, Mott MacDonald Pty Ltd, Melbourne austen.dupleix@mottmac.com
2
Associate Principal, Arup Australia Pty Ltd, Melbourne email: rob.day@arup.com
3
Principal Geotechnical Engineer, Mott MacDonald Pty Ltd, Melbourne email dominic.hagues@mottmac.com
ABSTRACT
Soil nail walls are increasingly used as one of the few feasible construction methodologies for constrained urban
environments. The top-down construction methodology used for soil nail walls is inherently unstable during
construction and there are documented instances of instability occurring during construction staging. Temporary
stability during excavation stages relies heavily on short-term undrained strength of clay, the adhesion of shotcrete
to the face, and the ability of soil nails to act in shear, none of which are relied on in the completed soil nail wall for
long term stability. Fissuring in basaltic clay further exacerbates this, as it can lead to over-break in the excavation
face and rapid onset of drained strength behaviour. For the design and construction of the soil nail wall at Ferguson
Street Level Crossing Removal, laboratory testing had identified the presence of low drained strength in the basaltic
clay indicative of fissuring. Confined space for construction and a limited occupation window for construction also
put tight constraints on construction staging. These factors were addressed with design measures to reduce reliance
on nail shear and shotcrete adhesion during construction, construction methodology, close monitoring and
supervision, and a toolbox of treatments developed to ensure construction could take place quickly before onset of
drained conditions. However, an unprecedented industry-wide two-week shut during the Covid epidemic, followed
by several weeks of reduced shift work occurred mid-way through construction of the soil nail wall, leading to
significant delays in the construction program. Significant overbreak and wall movements were subsequently
observed, and additional propping and remedial measures were implemented to prevent further movement and
allow completion. Back analysis was carried out on the movements, which was able to demonstrate the behaviour
was consistent with the development of drained conditions and that there was no long-term detrimental effect to the
completed wall.
Go to table of contents
Back Analysis, Monitoring and Remediation of Instability in a Soil Nail Wall in Fissured Basaltic Clay A. E. M. Dupleix, R. A. Day and D. N. Hagues
stages showed quite high excess pore pressures during was well below peak friction angle documented for
shearing, but removing the individual tests with high soils of this plasticity and was more representative of
excess pore pressures did not change the line of best fit critical state friction angle. This suggested that the clay
significantly. had undergone significant reworking through fissuring
and shrink/swell movements. It was recognised that
multi-stage testing may still have resulted in some
increase in apparent cohesion. As small increases in
cohesion can have a big impact on the stability of soil
nail walls, a slightly lower cohesion of c’=6kPa was
adopted with a corresponding slight increase in friction
to 22o adopted to maintain a line of fit with almost the
same correlation coefficient as the best fit line.
TOP OF ROCK
Another known issue with multi-stage tests is that Figure 4. Undrained Strength Results
progressive softening of the sample can occur
between the stages of the test with the third stage in It is important to note that the triaxial testing
particular giving a low strength closer to residual consistently gave lower undrained strengths than
strength. Discounting the third stage of all tests to inferred from either shear vane or pocket
investigate this possibility also had negligible impact penetrometer tests. At the time this was attributed to
on the line of best fit. On this basis, it was considered sample disturbance, but with the benefit of hindsight
that the results were accurate. The low friction angle it seems likely that small scale fissuring in the clay
Go to table of contents
Back Analysis, Monitoring and Remediation of Instability in a Soil Nail Wall in Fissured Basaltic Clay A. E. M. Dupleix, R. A. Day and D. N. Hagues
contributed to this result as triaxial testing can shear shotcrete lip and facing shotcrete. Stage 2 was the
the sample along fissures, while the other tests excavation and strip drain placement for the subsequent
require shearing through the intact clay. drop, Stage 3 is installation of nails and shotcrete. Steps
2 and 3 were repeated until wall the wall was completed.
3 SOIL NAIL DESIGN AND DETAILS
Go to table of contents
Back Analysis, Monitoring and Remediation of Instability in a Soil Nail Wall in Fissured Basaltic Clay A. E. M. Dupleix, R. A. Day and D. N. Hagues
Go to table of contents
Back Analysis, Monitoring and Remediation of Instability in a Soil Nail Wall in Fissured Basaltic Clay A. E. M. Dupleix, R. A. Day and D. N. Hagues
cause of the movements and a remedial works plan 5.3 Impacts on Construction and
could be determined. At the point of intervention and Completed Design
backfilling, maximum movements of 20mm and a rate
of increasing movement of 1mm/hr were observed Having identified the failure mechanism, the key
(Figure 10). Over the following week, despite questions to be addressed were then:
backfilling, movements continued to increase before
stabilising at a maximum of approximately 50mm • Had the stability of the wall been compromised
lateral movement and 30mm of vertical settlement. by the ground movements?
• How would the remainder of construction be
5 BACK ANALYSIS achieved without causing collapse?
Go to table of contents
Back Analysis, Monitoring and Remediation of Instability in a Soil Nail Wall in Fissured Basaltic Clay A. E. M. Dupleix, R. A. Day and D. N. Hagues
6 REMEDIAL ACTIONS
8 ACKNOWLEDGEMENTS
Figure 13. Completed "hit" shotcrete panel supported The authors would like to acknowledge the member
on base of excavation organisations of the Western Program Alliance for
permission to present this paper, specifically our
owner and construction partners: LXRP, Metro
Trains and McConnell Dowell.
REFERENCES
Gniel, J., Lenthall, C., Paul, D. (2019). “Recent experiences
with soil nailing in deep cuts as part of level crossing
removal works in Melbourne.” Australian
Geomechanics Journal, vol. 54 (1), 41-51.
Go to table of contents
SESSION 2
CHALLENGING
GEOTECHNICAL
CONDITIONS
Photo courtesy of Narges Khajavi (ATC Williams)
Ground Improvement Works for Cutter Head Intervention and Cross Passages
for Klang Valley Mass Rapid Transit (KVMRT)
L. H. Ooi1, CPEng, PEng, MIEM and Y. S. Teo2, CPEng, PEng, MIEM.
1
Gamuda Engineering Australia, Victoria, Level 12 West, 303 Collins Street, Melbourne, VIC 3000, Australia;
email: ooi.leanhock@gamuda.com.au
2
Gamuda Engineering Australia, Victoria, Level 12 West, 303 Collins Street, Melbourne, VIC 3000, Australia;
email: teo.yeesoon@gamuda.com.au
ABSTRACT
Klang Valley Mass Rapid Transit (KVMRT) consists of 3 main lines, namely Sungai Buloh Kajang (SBK Line 1),
Sungai Buloh Serdang Putrajaya (SSP Line 2) and Circle Line to date. SBK Line 1 and SSP Line 2 have been
completed with Circle line at planning stage. A total of 23km tunnel length from both Line 1 and Line 2 were
constructed. The general geology of the underground section is in Kenny Hill formation and Kuala Lumpur (KL)
Limestone formation. This paper discusses the ground improvement works carried out for cross passages and
scheduled or planned intervention for cutter head inspection and maintenance. The performance of the treated
blocks for SBK line 1 was reviewed and the mixed successes will be presented. A case history of a cross passage
with highly complex karst in SSP Line 2 involving extensive investigation and ground treatment works is presented
as well.
Keywords: cutterhead intervention, ground improvement, cross passage, Kenny Hill, karstic limestone
1 INTRODUCTION
Go to table of contents
Ground Improvement Works for Cutter Head Intervention and Cross L. H. Ooi and Y. S. Teo
Passages for Klang Valley Mass Rapid Transit (KVMRT)
3 SELECTION OF INTERVENTION
LOCATIONS AND GROUND
IMPROVEMENT WORKS
The KL Limestone Formation along the alignment is In all tunnelling works, including SBK and SSP line,
composed of fine to coarse grained, white to grey, to achieve the best productivity and trouble-free
predominantly recrystallised limestone, with local mining with the TBM, maintenance of the TBM cutter
developments of dolomitic limestone. The highly head is a major consideration.
irregular bedrock is overlaid by alluvium. The
limestone/dolomite rock has steep-sided valleys and The selection of the planned intervention location
an intricate three-dimensional network solutions considered the land use, impact to the community
channels and voids. These features are consistent and stakeholders and feasibility/accessibility of the
with Extreme tropical Karst classification according to location if ground treatment from surface is required.
Waltham (2003). This Karstic feature creates various The desired intervention block should facilitate free
problems and considerations for design and air intervention, if possible. The spacing of the
construction of cross passages and maintenance or intervention blocks also considers the geology and its
intervention blocks. impact on the degree/rate of wear and tear due to
mixed face condition and the hardness/abrasiveness
Figure 3 provides an indicative nature and formation of the mined material.
of KL limestone for a better understanding.
There were several unplanned interventions for SBK
Line 1 due to:
Go to table of contents
Ground Improvement Works for Cutter Head Intervention and Cross L. H. Ooi and Y. S. Teo
Passages for Klang Valley Mass Rapid Transit (KVMRT)
no ground treatment works was expected however, the intervention blocks in mixed face condition is
scheduled inspection and maintenance around 340m presented in Figure 7. The sectional view represents
intervals was adopted as part of the Standard the approximate location of the TBM cutter head
Operating Procedure (SOP) for SBK Line 1. intervention. The longitudinal section reflects the
condition variation of the mixed face condition over
3.2 Intervention Locations in KL the shield section. The different shade demarcated
Limestone the grout sections, and the number is the grout take
in litres.
In the extreme karst conditions, it is anticipated that
the highly variable rock head, steep-sided valleys
and solution channels have a significant impact on
the wear and tear of the disc cutters. Intervention
blocks coinciding with cross passage locations are
treated with appropriate ground improvement
methods at approximately 240m intervals and are
often tweaked or adjusted to locations with least
impact to utilities, public and traffic.
Go to table of contents
Ground Improvement Works for Cutter Head Intervention and Cross L. H. Ooi and Y. S. Teo
Passages for Klang Valley Mass Rapid Transit (KVMRT)
Go to table of contents
Ground Improvement Works for Cutter Head Intervention and Cross L. H. Ooi and Y. S. Teo
Passages for Klang Valley Mass Rapid Transit (KVMRT)
REFERENCES
A.C. Waltham and P.G. Fookes (2003). “Engineering
classification of karst ground conditions”
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west
– A Crack in the Code?
A. McIntrye1, C. Bridgeman2 and C. Crowe3
1,2,3
Douglas Partners Pty Ltd, 231 Normanby Road, South Melbourne 3205, PO Box 5051
ABSTRACT
Design for earthquake actions in Australia is guided by the Australian Standard (AS 1170.4-2007) where the
design earthquake structural actions are a function of the size and nature of the structure being assessed, the
behaviour of the local ground profile during a seismic event and the probability of an earthquake occurring. To
assist designers, AS 1170.4-2007 outlines a simplified approach to model the behaviour of the ground profile,
where the ground profile is assigned to one of five subsoil classes. There are geographical areas west and south-
west of Melbourne where the local ground profile does not always align with one of the five ground profile (or
subsoil) classifications defined in AS 1170.4-2007. Strict adherence to the subsoil definitions may lead to either
a conservative or an optimistic assignment of subsoil class, both of which can have an impact on the design
earthquake loading and resulting seismic-force-resisting system. This paper discusses the nature of the ground
profile in the west and south-west of Melbourne and how a site subsoil class could be applied in this region,
including discussion on site natural period.
1 INTRODUCTION
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west – A Crack in the Code? A. McIntyre, C. Bridgeman and C. Crowe
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west – A Crack in the Code? A. McIntyre, C. Bridgeman and C. Crowe
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west – A Crack in the Code? A. McIntyre, C. Bridgeman and C. Crowe
magnitude for inadequately designed structures or basalt rock in the Werribee Plains Lava Field is
those in a deteriorated condition. variable. A review of 55 published boreholes put the
basalt thickness at between 1m to 125m. Figure 3
Even though the first seismograph in Melbourne was shows the depth to the base of the basalt.
installed in 1888, more comprehensive recording of
seismic activity only commenced in 1976. Analysis of
seismic data in the Melbourne area suggests that
earthquakes are considered shallow (less than
17km) and are due to horizontal compressive stress
from south-east to northwest with movement along >100m
on reverse faults (Gibson et al. 1981). Given the
limited time scale of seismic recordings in Victoria in
comparison to a notional return period of moderate or >50m >20m
large earthquakes, it is possible that larger
earthquakes may occur in the future. Noting that the
general area would still be considered a region of
lower seismicity.
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west – A Crack in the Code? A. McIntyre, C. Bridgeman and C. Crowe
Table 2: Summary of Measured Shear Wave Velocity for Inferred Geological Units
Geological Unit Approximate Material Measured Shear
Age of Unit Description Wave Velocity
(Mya) (m/s)
Newer Volcanic Group (Qvn) 1-5 Basalt 1800
Newport Formation (Tmn) 8-25 Silt, sand, clay 500
(currently referred to as Gellibrand Marl)
Altona Coal Seam (Tea) 23-60 Brown coal, clay 1300
(a subunit of the Werribee Formation)
Werribee Formation (Tew) 23-60 Sand, silt, clay 500
Note: The table does not include all known Tertiary age deposits e.g., Brighton Group / Sandringham Sandstone.
Figure 4. Shear Wave Velocity from Roberts et al at Altona with extract from GSV 1974
Class Be could be assigned based on assumption the Table 3: Estimate of the Permissible Depth to
Tertiary soils have a shear wave velocity >300m/s Bedrock to Achieve a Certain Site Period
and a UCS >0.8MPa (or approximate undrained Average Maximum depth to top of
shear strength of 400kPa); and there is sufficient Shear Wave bedrock
basalt thickness to achieve the average shear wave
Velocity of Site Period of Site Period of
velocity of 760m/s in the top 30m.
Ground 0.2s 0.6s
6.3 Can Class Be apply if the natural Profile
site period is >0.6s? 1500m/s 75m 225m
760m/s 38m 114m
AS 1170.4-2007 is silent on maximum depth limits 300m/s 15m 45m
and natural site periods for the rock classes. Authors, Profile as per
including Poulos (1993), note that structural damage Roberts et al. 45m 103m
in buildings can be more prevalent when the natural (2004) Table 2
period of the building coincides with the natural period
of the ground. Designers may assume that by
assigning a Class Be a reasonably low (or fast) natural
site period would apply e.g., earthquake standards >50m
elsewhere in the world refer to “rock sites” as having >100m
natural site periods in the range of 0.1 to 0.2s.
>200m
AS 1170.4-2007 provides an upper limit on Class Ce
as a natural site period of 0.6s. If this site period of
0.6s is exceeded, then a Class De applies. A method >250m
for estimating site period is provided in AS 1170.4-
2007 as four times the shear wave velocity travel time
from surface to the underlying rock. Using this
methodology, Table 3 presents a summary of
estimated permissible depth to bedrock to achieve Figure 5. Depth to Top of Silurian and Devonian
certain site periods, e.g., where the ground profile Bedrock based on 13 published boreholes (geo.vic)
has an average shear wave velocity say of 760m/s,
then the estimated permissible depth to top of 6.4 The significance of a nominal 3m
bedrock is about 114m to remain within Class Ce surficial layer
definition of having a site period less than 0.6s. As
shown in Figure 5, the depth to Silurian and Devonian In AS 1170.4-2007, the difference between assigning
age bedrock is deep and typically exceeds the depths a subsoil class of Be or De could potentially rely on
in the table. the presence of a 3m thickness of soil strength
materials overlying the basalt rock, i.e., residual
clays, weathered materials or any alluvial or
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west – A Crack in the Code? A. McIntyre, C. Bridgeman and C. Crowe
windblown silts that were discussed previously. A 3m on a case-by-case basis. If earthquake load was a
thick layer of a surficial soil would not be expected to critical design case, then the building could be at risk
change the behaviour of the ground in an of damage from being under-designed, particularly if
earthquake. However, the impact from a design the building was located near an earthquake
perspective can be significant when considering epicentre.
force calculations presented in Section 4 of this
paper, where the resulting force estimates could Where Class Ce has been assigned and depending
double. on the size of the building and the Earthquake Design
Category, it is possible that earthquake loads were
With reference to the specific ground profile under-estimated by 10-20%. As a Class Ce seismic-
encountered by Roberts et al. (2004) in Altona force-resisting system would have been constructed,
(Table 2), the method of assigning a subsoil class this additional load may not induce as much damage
would be as follows: the basalt has a shear wave as the previous scenario where Class De site
velocity of approximately 1,800m/s over the top 30m, assigned as a Class Be.
so subclause (a) for both Class Ae and Be are
satisfied. For sub-clause (b), the strength of the 8 SUGGESTIONS MOVING FORWARD
Tertiary soils is not provided, but the shear wave
velocity profile indicates values between 500m/s and AS 1170.4-2007 could provide a definition of
1300m/s, so the clause (b) for Class Be is satisfied rock/bedrock so that an appropriate ground profile
but not Class Ae. Therefore, provided there is no can be determined. AS 1170.4-2007 could also
more than 3m of soil or highly/completely weathered provide a consistent criterion across all of the subsoil
material above the basalt (thickness not reported by classes, such as site period. A proposal for a site
Roberts et al. (2004)), then a Class Be applies. If the subsoil classification scheme has been prepared by
thickness of soil is exceeded, then the site period for the University of Melbourne for regions of lower
this profile did not encounter Silurian bedrock within seismicity (Tsang et al. 2015).
150m and is well above the 0.6s i.e., Class De
applies. The spatial extent of the issue can be defined by the
depth to the basement rock. At the north-eastern
6.5 Calculating Natural Site Period extent of the Werribee Plain Lava Field near
Footscray, where the development density and
It is acknowledged in AS 1170.4-2007 that the building height increases, the basement rock also
proposed method of estimating natural site period rises up and the calculated site period may be less
(four times shear wave velocity) is simplified. A deep than 0.6s. Where the site period is less than 0.6s,
sedimentary basin can either attenuate or amplify there is less uncertainty over the site subsoil class. It
seismic energy depending on the wave amplitude of may be advantageous to conduct investigations to
the nominated earthquake and the strength of the locate the top of the Silurian age bedrock as either
soils. A thick cap of basalt rock creates a shear wave part of the master planning for these areas or as part
velocity inversion that could potentially confine or of an AGS initiative. Drilling deep geotechnical
dampen the effects of seismic wave energy. Both of boreholes for an individual development on a case-
which could be considered in estimating the site by-case basis is unlikely to occur or be cost effective.
period, as suggested by Fabozzi et al. (2021).
However, it is not easy to predict seismic wave The properties of the underlying Tertiary age
energy in near-surface sediments without materials are not well understood. This is partly due
measurement. Unfortunately, buildings are not to most geotechnical investigations not penetrating
typically installed with seismographs to measure the basalt and the geological investigations that do
performance during an earthquake. penetrate the basalt are primarily interested in coal
and mineral deposits, so do not provide detailed
A separate issue with estimating natural site period strength data. As a result, properties of the Tertiary
using the method suggested in AS 1170.4-2007 is materials may need to be derived from non-intrusive
the basalt thickness is highly variable in the Werribee methods, such as seismic profiling (Roberts et
Plains Lava Field. The depth to sound rock can vary al. 2004).
significantly over short distances and the unit is
composed of multiple volcanic layers that each have To assign a subsoil of Class Be in critical areas within
experienced variable erosion and weathering. The Werribee Plains Lava Field, a borehole drilled to at
underlying base of the basalt is also variable, as it least 30m should be considered a minimum
reflects the relic topography of the Tertiary basin. requirement.
Go to table of contents
Earthquake Site Subsoil Classification in Melbourne’s South-west – A Crack in the Code? A. McIntyre, C. Bridgeman and C. Crowe
9 CONCLUSIONS
REFERENCES
AS 1170.4. (2007). Structural Design Actions, Part 4:
Earthquake Actions in Australia. Reconfirmed 2018.
Incorporating Amendments 1 & 2: Standards
Australia.
Boyce, Julie. (2012). Victoria erupts! The Newer Volcanics
Province of south-eastern Australia: a new database
of eruption points and an updated map.
Engineering geology of Melbourne : proceedings of the
Seminar on Engineering Geology of Melbourne,
Melbourne, Victoria, Australia, 16 September 1992 /
edited by W.A. Peck ... [et al.]
Eruption points of the newer volcanics province of Victoria :
an inventory and evaluation of scientific significance /
Neville Rosengren. Book. Bib ID, 2412822.
Fabozzi, S. Catalano, G. Falcone, G. Naso, A. Pagliaroli, E.
Peronace, A. Porchia, G. Romagnoli, M. Moscatelli,
Stochastic approach to study the site response in
presence of shear wave velocity inversion:
Application to seismic microzonation studies in Italy,
Engineering Geology, Volume 280,2021,
Geoscience Australia website https://www.ga.gov.au/
accessed September 2023
Gibson, G., Wesson, V., & Cuthbertson, R. (1981).
Seismicity of Victoria to 1980. Journal of the
Geological Society of Australia, 28(3–4), 341–356.
https://doi.org/10.1080/00167618108729173
GSV (1967) Geology of the Melbourne District, Victoria:
Explanatory Notes on the Stratigraphy, Structure and
Economic Geology to Accompany the Geological
Map of Melbourne and Suburbs (1959, Scale 1:31,
680) Mines Department, 1967
GSV (1974). Melbourne Sheet SJ55-1. Melbourne:
Geological Survey of Victoria and accompanying
notes
Geo.vic database website accessed September 2023
Leonard, John. (2022). GEOLOGY OF VICTORIA.
Poulos, H.G. (1993). Effect of earthquakes on settlements
and axial pile response in clays. Aust. Civil Eng.
Transactions, IEAust, Vol. CE35(1): 43-48.
Roberts, J., Asten, M., Tsang, H., Venkatesan, S., & Lam,
N. (2004). Shear Wave Velocity Profiling in Melbourne
Silurian Mudstone Using the SPAC Method. AEES
Conference. Mount Gambier, SA.
Tsang, H.H., Wilson, J.L., Lam. N.T.K., 2015.
Recommended Site Classification Scheme and
Design Spectrum Model for Regions of Lower
Seismicity. In: Proceedings of the Tenth Pacific
Conference on Earthquake Engineering, Sydney,
Australia, November 6-8, 2015.
Go to table of contents
Gas Permeability of Cement-Bentonite Cut-off Wall Materials with Powder and
Granular Bentonite
S. E. Al-Baiaty1, A. Bouazza2
1
Ph.D. Candidate, Department of Civil Engineering, 23 College Walk, Monash University, Vic. 3800, Australia;
PH +61 3 9905 0538; email: sarah.al-baiaty@monash.edu; Geotechnical Engineer, Tonkin + Taylor Pty Ltd,
Kings Technology Park, Level 3, 99 Coventry Street, Southbank, Vic 3006: SAl-Baiaty@tonkintaylor.com.au
2
Professor, Department of Civil Engineering, 23 College Walk, Monash University, Vic. 3800, Australia;
PH +61 3 9905 4956; email: malek.bouazza@monash.edu
ABSTRACT
A gas flow unified measurement system (UMS-G) was used to examine the gas permeability of cement-bentonite
(CB) mixtures prepared with powder and granular bentonite over various saturation states. The influence of
bentonite particle size was evident at volumetric water content > 50%; the CB mixture with granular bentonite
experienced less gas migration. This is due to the interfacial transition zone (ITZ) that could be present in the
microstructure of the CB mix with granular bentonite. Also, the granular bentonite contains a higher content of quartz
than the powder bentonite, which can provide a further pozzolanic reaction and denser microstructure to the CB
mixture. A reverse effect for the ITZ was found at 50% volumetric water content (VWC), resulting in severe fissure
cracks within the CB matrix compared to the CB mixture with powder bentonite that achieved a stable mixture
without cracks. At volumetric water content <50%, no significant difference was observed in gas migration results
between the CB mixes with powder and granular bentonite. The CB mixture with granular bentonite was disregarded
from the test at 40% VWC compared with 30% VWC for the CB mixture with powder bentonite due to the formation
of apparent cracks. Such findings can enhance the confidence in assessing the effect of bentonite particle size on
the gas permeability properties of the CB cut-off wall materials.
Go to table of contents
Gas Permeability of Cement-Bentonite Cut-off Wall Materials with Powder and Granular Bentonite S. E. Al-Baiaty and A. Bouazza
bentonite than the Volclay bentonite, indicating the Jefferis 2013, Carreto 2014). After 48 hours of
presence of a high percentage of granular bentonite hydration, the bentonite suspension's rheological
in VSP. properties were measured to identify the minimum
viscosity achieved during hydration, as Jefferis
The cement used in the current investigation (2013) recommended. Two methods were used: the
complies with AS3972, LH/SR type, special purposes indirect method (Marshal funnel) and the direct
cement, and it is composed of cement clinker and method (Viscometer). Table 2 summarises the
ground granulated blast furnace slag (GGBSF) (40 viscosity results. The indirect method was found
cement/60 GGBSF) and gypsum. It is typically used unreliable for both mixtures (i.e. foam from bentonite
in applications where an aggressive sulphate-rich suspensions blocked the flow of the suspensions
environment and other aggressive salt and solution from the funnel). The apparent viscosity from the
are expected. direct method was used as the indicative parameter
for the suspension rheological properties.
Table 1: Mineralogical composition of the bentonites
Volclay VSP In stage two, cement was added to the bentonite
Chemical
(P) (G) suspension and mixed for a time not less than 5
Bulk Ca saturated minutes to form a homogenous CB mixture. The
Quartz % 12 11 fluidity and workability properties of the CB slurry
Cristobalite % 1 10 were measured using a viscometer (Table 2). It can
Montmorillonite % 78 66 be seen from Table 2 that the apparent viscosity was
Calcite % <1 - significantly increased for both CB-P and CB-G due
Albite/ Anorthite % 7 8 to the increase of solid content in the mixture after
Kaolin % 1 <1 adding the cement. Also, the viscosity results at both
Anatase % - <1 stages are higher in the CB mixture with powder
bentonite (CB-P) than the CB mixture with granular
Zeolite % 1 5
bentonite (CB-G) due to the highest montmorillonite
Particle Size >2µm 24.4 39.2
content in powder bentonite.
<2µm fraction
Quartz % 3 2 The set properties of the CB mixture were
Cristobalite % <1 9 characterised by measuring the mixture’s slurry
Montmorillonite % 95 86 density and filter loss. The density measurement was
Calcite % - - conducted using a known cylindrical volume, and the
Albite/ Anorthite % 1 <1 filter loss test was by using a cylinder graduated in 20
Kaolin % 1 1 ml up to 100ml as recommended by ICE (1999).
Anatase % <1 <1 Typically, the density of CB materials should be
Zeolite % - 1 around 1220kg/m3, and the filter loss not be more
Particle Size <2µm 75.1 55.7 than 2% within 24 hours. The CB mixes set
CEC (cmol/Kg) by Ba and properties were within the requirement (ICE 1999).
93 77
XRF
Mineralogical test was done by Mineralogical Services The CB mixtures were poured inside PVC tubes
(CSIRO), Land and Water Division, South Australia closed from both ends to prevent moisture losses.
The PVC tubes were kept inside a fog room
2.2 Experimental methods (RH=100%, T=20oC) for curing. The mechanical
properties of the CB materials were measured after
2.2.1 CB mixture preparation curing at 28 days. The hydraulic conductivity was
measured at 90 curing days as recommended by ICE
The CB slurry mixture was prepared to a standard (1999). ICE (1999) specifies that unconfined strength
specification according to the Institution of Civil at 28 days should be more than 100kPa, and
Engineers, UK guidelines (ICE 1999), aided by other hydraulic conductivity is recommended to be 10-8 to
studies (Daniel and Koerner 2007, Jefferis 2012, 10-9 m/s at 90 days. In the current study, the strength
Jefferis 2013, Lam and Jefferis 2018). The mixture for both CB mixtures was within the recommended
included 7% bentonite and 20% cement of the total value; the strength of the CB-P mixture was 792kPa
weight of water. The preparation was done in two and 1515 kPa for the CB-G mixture. The permeability
stages; the first stage aimed at forming a bentonite to water of CB-P was 4.1 x 10-10 m/s and 3.06 x 10-9
suspension (BS) by mixing the bentonite with water m/s for CB-G; both values were within the specified
for not less than 5 minutes. Once formed, the range recommended by ICE (1999).
suspension was kept inside closed containers for
about 24 hours for hydration purpose. It should be 2.3 Gas permeability measurements
noted that granular bentonite (VSP) was remixed
after 24 hours of hydration and rehydrated for The gas permeability test was conducted using the
another 24 hours, as the accumulated hydrated Monash unified measurement system (MUMS-G)
bentonite particles were observed due to the developed by Rouf et al. (2016). The set-up
bentonite particle size. Sufficient hydration time measures sequentially gas diffusion and
allows the bentonite particles to be dispersed, permeability.
supporting the cement grains later during the
cementitious system hydration (Jefferis 2012,
Go to table of contents
Gas Permeability of Cement-Bentonite Cut-off Wall Materials with Powder and Granular Bentonite S. E. Al-Baiaty and A. Bouazza
The set-up comprises three major parts made of The results were plotted against volumetric water
acrylic: a top cylinder of 100mm inner diameter and content (VWC). One can observe that there is a
70mm height, a middle cylinder to accommodate the marked difference between CB-P and CB-G at VWC
sample, and a bottom cylinder of 100mm inner range between 75% up to 50%; where the gas
diameter and 95mm height. The bottom chamber permeability of the CB-P mixture is higher than the
included two valves and the oxygen and pressure CB-G mixture; beyond 50% VWC, there is no marked
transducer sensors. difference between CB-P and CB-G. This is
attributed to factors influencing the gas permeability
One of the valves in the bottom chamber was used of CB mixtures such as porosity, bentonite type and
to pressurise the gas into the bottom chamber. In the mixture sensitivity to the external environment; for
contrast, the second valve was used to expel it example, drying can cause cracks that affect the
during the gas diffusion test (i.e. oxygen diffusion). permeability process. As a cement-based material,
The top cylinder contained a movable rod and six the porous system interconnectivity can affect its gas
ports. Only one port can be opened during the gas transport properties (Jennings 2008, Jennings et al.
permeability test, and an interchangeable gas flow 2015, Wu et al. 2017). The pores system in cement-
meter was attached to this port to measure the gas based materials is referred to as capillary pores (i.e.
flow through the sample. Different ranges of flow between 8nm to 10μm),
meters were connected to this port to measure the
gas flow through the sample, depending on the CB
1E-6
degree of saturation.
CB-P
CB-G
Nitrogen gas (N2) was used for the gas permeability
1E-7
measurement because it is inert and has low water
Gas Permeability, m/sec
The gas permeability (kg) for both CB-P and CB-G The granular bentonite and its mineralogical
mixtures exposed to drying is presented in Figure 1. properties in the CB-G mixture can provide high
Go to table of contents
Gas Permeability of Cement-Bentonite Cut-off Wall Materials with Powder and Granular Bentonite S. E. Al-Baiaty and A. Bouazza
resistance against gas transport compared to the accumulate high ettringite, which could cause
CB-P mixture. The bentonite of the CB-G mixture expansion and is considered the main reason for
contains a high percentage of cristobalite (i.e. quartz the cracking in ITZ. In the case of the CB materials,
family), which can facilitate the pozzolanic reaction ITZ was observed under the secondary electron
and develop a denser microstructure. In an microscopic (SEM) (Figure 3). This is evidence that
environment rich with pozzolanic, ettringite forms bentonite is not entirely dissolved during the
quickly; this can provide more filling and denser interaction with cement, which contradicts what was
internal structure to the porous system and observed by other researchers (Evan 2021, Jefferis
decelerate calcium leaching through a denser 2012). Granular bentonite can provide a better
structure network, developing a chemically and barrier against the gas flow at high saturation levels
physically stable barrier. In contrast, the bentonite of because it has high porosity between (10 and 50
CB-P contains a high percentage of montmorillonite um) that can facilitate moisture ingress through the
that, in turn, can increase the calcium percentage in formation of ITZ. Still, it can cause fissure cracks at
CB pore solution, causing high carbonation. a moderate saturation level.
Carbonation can cause dimensional changes to the
porosity (i.e. disordered structure) of the porous At 40% VWC, the CB-G mix was disregarded from
system through high dissolution/precipitation to the the test due to the apparent cracks (i.e. around 2 mm,
cementitious system hydration products and see Figure 4). In contrast, the CB-P mix experienced
influence the porous system gas transport properties fissure cracks at 40% VWC and was disregarded
(Jain and Neithalath 2009). Carbonation can also from the test at 30% VWC.
induce finer cracks in the porous system and
increase the porous system’s gas migration.
Go to table of contents
Gas Permeability of Cement-Bentonite Cut-off Wall Materials with Powder and Granular Bentonite S. E. Al-Baiaty and A. Bouazza
Go to table of contents
Gas Permeability of Cement-Bentonite Cut-off Wall Materials with Powder and Granular Bentonite S. E. Al-Baiaty and A. Bouazza
Jennings, H. M. (2008) Refinements to colloid model of Wu, Z., Wong, H. S. & Buenfeld, N. R. (2017) Transport
CSH in cement: CM-II. Cement and Concrete properties of concrete after drying-wetting regimes to
Research, 38(3), 275-289. elucidate the effects of moisture content, hysteresis
Jennings, H. M., Kumar, A. & Sant, G. (2015) Quantitative and microcracking. Cement and Concrete Research,
discrimination of the nano-pore-structure of cement 98, 136-154.
paste during drying: New insights from water sorption Xanthakos, P. P. (1979). Slurry walls, McGraw-Hill Book
isotherms. Cement and Concrete Research, 76, 27- Company, New York.
36. Zhang, Y., & Zhang, M. (2014). Transport properties in
John, E., Epping, J. D. & Stephan, D. (2019) The influence unsaturated cement-based materials–A
of the chemical and physical properties of CSH seeds review. Construction and Building Materials, 72, 367-
on their potential to accelerate cement hydration. 379.
Construction and Building Materials, 228, 116723.
Jain, J. and N. Neithalath (2009). “Analysis of calcium
leaching behavior of plain and modified cement
pastes in pure water.” Cement and Concrete
Composites 31(3): 176-185.
McBean, E. A., Rovers, F. A. & Farquahar, G. J. (1995)
Solid waste landfill; engineering and design. Prentice
Hall.
Opdyke, S. M. and J. C. Evans (2005). “Slag-Cement-
Bentonite Slurry Walls.” Journal of Geotechnical and
Geoenvironmental Engineering 131(6): 673-681.
Ratnam, S. (2002). Development of a novel self-boring
permeability measurement technique, University of
Cambridge.
Rouf, M. A., A. Bouazza, R. M. Singh, W. P. Gates and R.
K. Rowe (2016). “Gas flow unified measurement
system for sequential measurement of gas diffusion
and gas permeability of partially hydrated
geosynthetic clay liners.” Canadian Geotechnical
Journal 53(6): 1000-1012.
Ryan, C. R. and S. R. Day (2002). Soil-cement-bentonite
slurry walls. Deep Foundations Congress 2002 -
International Perspective on Theory, Design,
Construction, and Performance_, Orlando, FL.
Sicat, E., Gong, F., Ueda, T. and Zhang, D., 2014.
Experimental investigation of the deformational
behavior of the interfacial transition zone (ITZ) in
concrete during freezing and thawing
cycles. Construction and Building Materials, 65,
pp.122-131.
Soga, K. and K. Joshi (2015). “Cement bentonite cut-off
walls for polluted sites.” Multilevel Modeling of Secure
Systems in QoP-ML: 149.
Sun, D., Wu, K., Shi, H., Zhang, L. and Zhang, L., 2018.
Effect of interfacial transition zone on the transport of
sulfate ions in concrete. Construction and Building
Materials, 192, pp.28-37.
Rouf, M. A., Bouazza, A., Singh, R. M., Gates, W. P. &
Rowe, R. K. (2016a) Gas flow unified measurement
system for sequential measurement of gas diffusion
and gas permeability of partially hydrated
geosynthetic clay liners. Canadian Geotechnical
Journal, 53(6), 1000-1012.
Tedd, P. and S. A. Jefferis (2000). The use of slurry trench
cut-off walls to repair embankment dams in the UK.
Proceedings of the biennial conference of the BDS
held at the University of Bath on 14-17 June 2000:
425-438.
Vangpaisal, T. and A. Bouazza (2004). “Gas permeability of
partially hydrated geosynthetic clay liners.” Journal of
Geotechnical and Geoenvironmental Engineering
130(1): 93-102.
Wu, Z., Wong, H. S., & Buenfeld, N. R. (2015). Influence of
drying-induced microcracking and related size effects
on mass transport properties of concrete. Cement
and Concrete Research, 68, 35-48.
Go to table of contents
SESSION 3
ROCK BEHAVIOUR IN
MINING APPLICATIONS
Novel Solutions in Utilising Underground Coal Seam Gas Reservoirs
for Gas Storage/Recovery
M. S. A. Perera1
Department of Infrastructure Engineering, Building 175, Grattan Street, Parkville, The University of Melbourne, Victoria 3010
1
ABSTRACT
Today, the world is moving to a green energy future by targeting net zero carbon emissions, which wouldn't be
successful without the support of underground reservoirs. Geological reservoirs, such as saline aquifers and coal
seam gas reservoirs, play an essential role in decarbonisation by providing ideal candidates to store CO2. On the
other hand, during the transition period of moving to renewable energy from fossil fuels, natural gas extracted
from underground can play a major role as a low-emission fuel. However, when underground reservoirs,
particularly low-stiffness geological reservoirs such as coal seams, are used for gas storage and extractions,
numerous issues may arise due to extreme pressures, temperatures, and geological and environmental
complexities. This paper focuses on potential novel solutions to overcome such challenges. Our research focuses
on minimising reservoir stimulation-related issues, such as effective proppants usage with minimising embedment
and poor distribution, focusing on damage mechanisms, fluid dynamics (e.g., flow complications caused by
inertia) and the geological reservoirs' characteristics. Also, according to our research, some CO2 storage issues in
coal seam gas reservoirs can be minimised by reducing the swelling effect by controlling injection scenarios and
introducing a stream of inert gas (e.g. nitrogen) after each CO2 cycle. This paper thus provides an excellent
summary of the issues and opportunities to overcome those issues during underground coal seam gas reservoirs
utilisation during the gas storage and recovery.
Keywords: underground utilisation, gas storage, gas recovery, challenges, novel solutions
Go to table of contents 1
Novel Solutions in Utilising Underground Coal Seam Gas Reservoirs for Gas Storage/Recovery M. S. A. Perera
fracture network in the targeted reservoir to According to our research, it is not worth using
enhance permeability. However, this fracture proppants at shallow depths as it reduces the
network is rapidly closed due to the extreme conductivity through proppants clogging in fractures.
pressure at deep underground reservoirs. To For example, proppants should only be used after
overcome this issue, tiny particles known as around 300m depth for bituminous coal. Likewise,
proppants are mixed with the fracturing fluid. These conducting a preliminary investigation on the
proppants can keep open the fracture after effectiveness of proppants for any given reservoir is
releasing the fracking fluid upon completion of the worthwhile to do. Further investigation showed that
fracking process, offering permeable pathways for proppants with irregular shapes, like sand, are
the CSG to flow to the wellbore. effective in shallow depths, and regular, spherical-
shaped ceramics proppants are effective at deeper
2.1.1 Improper proppants usage issues depths (Figure 3).
Go to table of contents 2
Novel Solutions in Utilising Underground Coal Seam Gas Reservoirs for Gas Storage/Recovery M. S. A. Perera
As can be seen from Figure 4, under low As shown in Figure 7a, slowdown or minimised flow
confinement, sand proppants can rotate and re- through the branching fractures minimises the
arrange to provide spacing for fluid flow and under proppants entering them, resulting in the distribution
high confinement, in addition to proppants of proppants only along the major fractures, which
embedment, sharp contact points cause stress explains the reason for poor proppants distribution.
concentration and breaking the sand particle,
generating fine and blocking the flow paths. Due to The natural gas industry spends millions of dollars
its high strength and regular shape, this does not on proppants and injection schemes with
happen in ceramic proppants (Figure 5), so it works hydrodynamics that might not be favourable. For
better under high confinements. example, fracture stimulation of a typical
unconventional reservoir well may consume over 4
2.1.2 Inertia effect on fracture flow behaviour million pounds of proppants, costing over 1 million
dollars (Earthworks 2023). Thus, understanding the
The fluid inertia caused non-Darcy flow or fluid flow real fluid dynamic in a fracture network is essential.
transition from regular linear flow to nonlinear flow Opening this black box will provide the crucial
behaviour, significantly impacting the flow behaviour scientific foundations to rigorously engineer effective
through fractured geologically reservoirs (naturally fracture stimulation and proppant usage to direct
fractured or hydraulically fractured). The inertia proppants into secondary/tertiary fractures, reducing
causing flow irregularities (Figure 6) creates the wasteful disposition of proppants.
complicated flow behaviours in underground
reservoirs, making it hard to describe the fluid flow Further, formation damage and pore plugging in
through rock fractures through simple theories. reservoirs due to fines migration are commonly
associated with oil/gas wells. Several studies have
shown the contribution of fluid inertia effect on fines
migration mechanism with high flow rates employed
in reservoirs (Zeinijahromi et al. 2012, Bai et al.
2015, Miri et al. 2021).
Go to table of contents 3
Novel Solutions in Utilising Underground Coal Seam Gas Reservoirs for Gas Storage/Recovery M. S. A. Perera
Go to table of contents 4
Novel Solutions in Utilising Underground Coal Seam Gas Reservoirs for Gas Storage/Recovery M. S. A. Perera
a)
Go to table of contents 5
Novel Solutions in Utilising Underground Coal Seam Gas Reservoirs for Gas Storage/Recovery M. S. A. Perera
This has been discovered by conducting a series of the CO2 storage process in deep coal seam gas
tri-axial permeability tests. After the first cycle of reservoirs causes a coal matrix swelling,
CO2 injection, N2 gas was injected into coal samples significantly reducing the seam's storage capacity
for 24h, and permeability tests for CO2 were then and mechanical stability, where the damage rate
repeated to investigate changes in CO2 flow ability depends on the coal maturity and reservoir location.
after the N2 injection. As shown in Figure 15, CO2 According to our studies, the impact is much more
permeability showed an increment in permeability. significant in the super-critical CO2., a highly
reactive physical state, than in gas CO2. Thus,
Around 20% of permeability recovery could be seen selecting a suitable type of reservoir located at
at 20MPa confining pressure. Thus, it is clear that desired depth plays an important role here. Also, the
there is a significant impact coming from N2 flooding swelling area is found to be partially recovered by
after the first cycle of CO2 storage on CO2 injectivity injecting a stream of N2 after the first injection cycle
into the seam and can be used in commercial CO2 of CO2.
storage sites.
5 ACKNOWLEDGEMENTS
REFERENCES
Ahamed, M.A.A. (2020). " Efficient ways of proppant usage
in hydraulic fracturing of coal seam gas reservoirs."
PhD Thesis, The University of Melbourne, Australia.
Bai, T., Z. Chen, S. M. Aminossadati, Z. Pan, et., (2015).
"Characterisation of coal fines generation: A micro-
scale investigation." Journal of Natural Gas Science
and Engineering, 27, 862-875.
IEA, Gas-Fuels and Technologies,
http://www.iea.org/fuels-and-technologies/gas (Jul.
10, 2023)
Figure 15. Effect of N2 flood on CO2 permeability at Day, S., L. Connell, D. Etheridge, et al. (2000). Fugitive
20 MPa confining pressure (Perera et al. 2011) greenhouse gas emissions from coal seam gas
production in Australia, Energy Technology, CSIRO,
4 CONCLUSION Australia.
Earthworks. "Hydraulic fracturing 101."
https://earthworks.org/issues/hydraulic-fracturing-
The study focuses on novel solutions discovered to
101/ (July. 10, 2023)
optimise the usage of underground geological Hansika H. (2023). Influence of inertial and centrifugal
reservoirs for gas storage and recovery. forces on the flow rate and patterns of flow in
fracture networks, MPhil Thesis, The University of
In the gas storage and recovery processes, Melbourne, Australia.
reservoir stimulation techniques, such as hydraulic Ltd, U.A.P. (2014). Background review: Hydraulic
fracturing, play a significant role. The effectiveness fracturing ('fraccing') techniques, including reporting
of proppants used in the fracking process depends requirements and governance arrangements,
Department of the Environment of Australian
on the type and location of the reservoir rock and Government.
the fluid dynamic there. Thus, the optimum usage of Miri, R., M. Haftani and A. Nouri (2021). "A review of fines
proppants in underground should clearly understand migration around Steam Assisted Gravity Drainage
these aspects for maximum storage capacity and wellbores." Journal of Petroleum Science and
gas recovery. In particular, low-stiffness rocks like Engineering, 205,108868.
coal causes proppants to embed while shrinking the Perera, M. S. A. (2011). "Investigation of the effect of CO2
flow paths and tend to settle at junctions and avoid sequestration on hydro-mechanical properties of
branching fractures due to the inertia causing flow coal." PhD Thesis, Monash University, Australia.
Sampath K. H. S. M. (2020). " A comprehensive analysis
complexities.
of the structural alterations occur in CO₂- interacted
coal: An insight into CO₂ sequestration in coal." PhD
According to our studies, identifying the desired Thesis, The University of Melbourne, Australia.
depths and the type of proppants through proper Zeinijahromi, A., A. Vaz and P. Bedrikovetsky (2012).
preliminary investigation can significantly enhance "Well impairment by fines migration in gas fields".
the effectiveness. On the other hand, identifying Journal of Petroleum Science and Engineering, 88–
potential inertia causing flow instability and 89, 125–135.
controlling the gas injection rates accordingly helps
to improve the proppants distribution while avoiding
unnecessary proppants clogging at fracture
intersections.
Go to table of contents 6
Vein Effect on Mechanical Properties and Fracture Behaviour of Porphyry Rock
H. J. Wang1 and Q. B. Zhang2
1
Department of Civil Engineering, Monash University, VIC 3800, Australia; email: haojun.wang@monash.edu
Department of Civil Engineering, Monash University, VIC 3800, Australia; email: qianbing.zhang@monash.edu
2
ABSTRACT
Veins are commonly found in porphyry copper deposits and can greatly impact rock mass stability during block caving
processes. Understanding the mechanical properties and fracture behaviours of veined rocks under various
compression tests can help reduce potential hazards in rock engineering. In this study, laboratory tests were conducted
on veined porphyry from mining operations in Australia under uniaxial compression and triaxial compression with
different confinements. The results show that the orientation of calcite veins is the main inherent factor in reducing rock
strength, with the least favourable angle being between 15° and 45° relative to the normal loading direction. Among
these values, the weakest specimen was found at 30° in uniaxial compression tests. The real-time cracking process
of compression tests was monitored by an acoustic emission (AE) system, and surface cracking was captured by high-
speed cameras for uniaxial tests. AE events were concentrated at the final failure stage for veined samples, while
noticeable signals were detected during the early stages for intact samples. Furthermore, early-stage event energy
increased as confining pressure increased. X-ray computed microtomography (CT) was also used on tested samples
to obtain and reconstruct fracture patterns after peak load. The results show that veins dominate failure planes at low
confinement, while brittle fractures in the rock matrix increase at high confinement.
1 INTRODUCTION falling within the range of 2.0 to 2.5. Prior to testing, the
cylindrical surfaces of the samples were meticulously
Veins form through the deposition of minerals carried prepared to ensure they were flat and smooth. In this
by hydrothermal fluids circulating through paper, we focused on investigating the influence of two
mechanical discontinuities within rock masses. This key geometric factors: vein number and vein
natural formation process underscores the orientation shown in Figure 1. Vein number was
complexity of vein geometry, mineralogy, mechanical defined as the presence of continuous veins that were
properties (Day et al. 2019, Shang 2022). Veins play visible to the naked eye. Vein orientation was
a fundamental role in modulating the mechanical, measured as the angle between the veins and the
geophysical, and geochemical behaviour of rocks. loading direction (Turichshev and Hadjigeorgiou 2015).
Understanding the influence of veins on the
deformation and failure of veined rock is of
paramount importance for various applications in the
field of underground engineering. One such industry ①
is copper mining, where veined porphyry has proven
θ
to be a common and valuable deposit (Robbiano et
al. 2022, 2023). ②
Go to table of contents
Vein Effect on Mechanical Properties and Fracture Behaviour of Porphyry Rock H. J. Wang and Q. B. Zhang
Go to table of contents
Vein Effect on Mechanical Properties and Fracture Behaviour of Porphyry Rock H. J. Wang and Q. B. Zhang
captured by high-speed cameras, with typical cases spheres represent AE event locations. Larger sphere
shown in Figures 3 and 4. For intact case, surface radii correspond to higher energy magnitudes.
spalling and rock pillar splitting up can be observed
around peak load. For veined case, failure mode is In alignment with high-speed images, the behaviour
sliding shear along veins then splitting follows. of the intact sample unfolds with the initial
observation of spalling near the top, followed by
160 2500 crack propagation towards the bottom. Internal
(ii)(iii) cracks are also detected, contributing to the eventual
(i) (iv) splitting of the sample. On the other hand, in the
2000
veined sample, initial observations reveal bottom
120 ③
cracks, succeeded by shear failure on the right side,
along with subsequent cracking along the plane of
1500
the vein.
σ1 (MPa)
AE hit
80
160 2500
1000
① ④ ④
40 2000
500 120
② (ii)
(iii) 1500
(iv)
σ1 (MPa)
0 0
0.00 0.05 0.10 0.15 0.20 0.25
AE hit
80
ε (%)
1000
40
①② 500
③
(i) (ii) (iii) (iv)
0 0
0.00 0.05 0.10 0.15 0.20 0.25
ε (%)
Go to table of contents
Vein Effect on Mechanical Properties and Fracture Behaviour of Porphyry Rock H. J. Wang and Q. B. Zhang
orientation and strength displays a U-shaped predominantly concentrate around the point of final
distribution. Strength tends to decrease within the failure, other information is hard to collect. However,
0-30° range, reach its minimum value at around 30°, in the veined sample, the presence of veins
and then increase from 30 to 60°. Therefore, the substantially enhances AE response during the early
orientation of veins is a critical factor with significant stages of testing, suggesting a progressive fracturing
effects on the mechanical behaviour of the rock. process taking place.
160 350
Intact
240
210
300
120 250
180
Before After CT
150
σ1 (MPa)
σ3-σ1(MPa)
200
AE hits
80 120
150
90
100
40 60
50
30
0 0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0
0 1 2 3 4 ε(%)
Veins number
350 240
Veined
160
210
300
180
250
120
150
σ3-σ1(MPa)
200
AE hits
σ1 (MPa)
120
80 150
90
0 30 60 ε(%)
3.2 Veins effect on triaxial test Triaxial strengths are graphically represented in
Figure 7. An intriguing observation is that intact
Figure 6 presents the typical triaxial test results samples normally follow Mohr-Coulomb failure
conducted under a confinement of 10MPa, and it criterion, giving a linear positive correlation between
reveals a significant observation: the intact sample σ3 and σ1, whereas the veined samples present
demonstrates a strength and peak strain approximately discrete strength distribution. The difference more
double that of the veined sample, yet the elastic evident is under low confinement range (<10MPa)
modulus is only slightly weaken. This performance indicating the importance of vein study during
mirrors the trends observed in the UCS tests. In the excavation activities.
case of the intact sample, typical brittle failure manifests
along an oblique direction. The computed tomography
(CT) extraction reveals a fracture surface that is notably
thin, indicating a relatively smooth failure surface.
Conversely, in the veined sample, the presence of
weaker surfaces influences the direction of shear
damage, resulting in a CT-reconstructed fracture that
appears wide and rough.
Go to table of contents
Vein Effect on Mechanical Properties and Fracture Behaviour of Porphyry Rock H. J. Wang and Q. B. Zhang
5 ACKNOWLEDGEMENTS
REFERENCES
Figure 7. TCS results of intact and veined porphyry
ASTM, A. (2010). Standard test method for compressive
Upon comparing the stress results derived from the strength and elastic moduli of intact rock core
loading test with the observed failure patterns, it specimens under varying states of stress and
becomes evident that vein distribution wields temperatures. In.
significant influence over triaxial strength. This Bieniawski, Z. T., & Bernede, M. J. (1979). Suggested
methods for determining the uniaxial compressive
observation aligns harmoniously with the findings
strength and deformability of rock materials: Part 1.
from UCS tests, where it was consistently observed Suggested method for determining deformability of
that the majority of shear failures occurred along the rock materials in uniaxial compression. International
veins. These observations underscore the dominant Journal of Rock Mechanics and Mining Sciences &
role of structural features, such as veins, in dictating Geomechanics Abstracts, 16(2), 138-140.
the mechanical behaviour of the rock in triaxial Clark, M. D., & Day, J. J. (2021). Mineralogical and sample
testing conditions. selection implications for geomechanical properties of
intact heterogeneous and veined rocks from the
Legacy skarn deposit. Engineering Geology, 285,
4 CONCLUSION 106067.
Day, J. J., Diederichs, M. S., & Hutchinson, D. J. (2019).
In conclusion, our research has highlighted the critical Composite Geological Strength Index Approach with
role of veins in the mechanical behaviour of rocks, Application to Hydrothermal Vein Networks and Other
particularly in the context of uniaxial and triaxial tests. Intrablock Structures in Complex Rockmasses.
We observed that veins contribute to a wide range of Geotechnical and Geological Engineering, 37(6),
results in terms of compressive strength, and their 5285-5314.
Li, J., Zhao, J., Gong, S. Y., Wang, H. C., Ju, M. H., Du, K.,
presence significantly influences failure modes. In
& Zhang, Q. B. (2021). Mechanical anisotropy of coal
uniaxial compression tests, we noted distinct patterns under coupled biaxial static and dynamic loads.
of failure, with shear failures occurring primarily in International Journal of Rock Mechanics and Mining
veined samples and tensile cracks developing in intact Sciences, 143, 104807.
samples with higher UCS values. Liu, K., Zhang, Q., Wu, G., Li, J., & Zhao, J. (2019). Dynamic
mechanical and fracture behaviour of sandstone
Acoustic Emission monitoring revealed that veined under multiaxial loads using a triaxial Hopkinson bar.
samples exhibited cleaner AE signals during Rock Mechanics and Rock Engineering, 52, 2175-
2195.
breakage, and strong pre-signals provided insight into
Robbiano, F., Liu, K., Zhang, Q.-B., & Orellana, L. F. (2022).
impending failure. While traditional AE location Dynamic uniaxial compression testing of veined rocks
calculations assume material homogeneity, our study under high strain rates. International Journal of Rock
demonstrated that deviations caused by veins still Mechanics and Mining Sciences, 153, 105085.
allow for reasonably accurate crack position Robbiano, F., Liu, K., Zhang, Q.-B., & Orellana, L. F. (2023).
determination, although with slightly reduced Dynamic mechanical properties of veined rocks under
precision. biaxial confinement. International Journal of Rock
Mechanics and Mining Sciences, 170, 105538.
Shang, J. (2022). Stress path constraints on veined rock
Triaxial tests further underscored the impact of veins,
deformation. Rock Mechanics Bulletin, 1(1), 100001.
with the Hoek cell setup reducing AE sensitivity but Turichshev, A., & Hadjigeorgiou, J. (2015). Experimental
veins increasing AE response, signifying gradual and numerical investigations into the strength of intact
fracturing. The relationship between triaxial strength veined rock. Rock Mechanics and Rock Engineering,
and confining pressure was not always 48, 1897-1912.
straightforward, emphasizing the dominance of Turichshev, A., & Hadjigeorgiou, J. (2017). Quantifying the
structural factors. This aligns with UCS test results, effects of vein mineralogy, thickness, and orientation
where most shear failures were observed along veins. on the strength of intact veined rock. Engineering
Geology, 226, 199-207.
Wang, H. C., Zhao, J., Li, J., Braithwaite, C. H., & Zhang,
In summary, veins are a crucial factor in understanding Q. B. (2022). Progressive fracturing of concrete under
and predicting the mechanical behaviour of rock biaxial confinement and repetitive dynamic loadings:
materials. They introduce variability in strength, From damage to catastrophic failure. International
influence failure modes, and affect AE monitoring. Journal of Impact Engineering, 165, 104232.
Go to table of contents
NOTES
VICTORIA CHAPTER