See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/265913704

Geotechnical Design of Underground Infrastructure Works for the Mine

Chuquicamata in Chile

Conference Paper · June 2013

DOI: 10.1201/b14769-134

CITATIONS READS

2 411

3 authors:

Alexandre R.A. Gomes Gustavo Reyes

SMEC (member of the Surbana Jurong Group) 2 PUBLICATIONS   2 CITATIONS   
25 PUBLICATIONS   18 CITATIONS   
SEE PROFILE
SEE PROFILE

Juan Carlos Ulloa

6 PUBLICATIONS   4 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Metro Santiago View project

Metro Bangkok View project

All content following this page was uploaded by Alexandre R.A. Gomes on 19 July 2016.

The user has requested enhancement of the downloaded file.

 World Tunnel Congress 2013 Geneva
Underground – the way to the future!
G. Anagnostou & H. Ehrbar (eds)

Geotechnical Design of Underground Infra-structure Works for the

Mine Chuquicamata in Chile
(1) (2) (1)
Alexandre RA Gomes , Gustavo Reyes , Juan C. Ulloa
(1)
GEOCONSULT Latinoamérica, Santiago, Chile
(2)
HATCH, Santiago, Chile

ABSTRACT: The paper presents the design philosophy and criteria applied for the geotechnical design of the
permanent underground infra-structure works of the future Chuquicamata underground mine, located in the
Atacama Desert, Chile. The design was carried out at an advanced basic engineering level, as part of the integral
feasibility study of the mine, which will constitute one of the largest underground mining operations in the world.
The design scope included the permanent traffic access, ore transport and ventilation tunnels with lengths varying
from 4.3km to 7.45km, with steep gradients (-9% to -15%) and low to high overburden conditions (up to 1250m),
in addition to various associated ancillary structures, such as shafts and connection drifts. Excavations will be
carried out in intrusive and meta-plutonic rocks with different degree of alteration with the use of the conventional
excavation method.

1 Introduction
The mining complex Chuquicamata is located in the Atacama Desert, 1.650 km north of Chile’s capital
city, Santiago, and 2.870 m above sea level. It corresponds to one of the main operations of the
National Copper Corporation of Chile (CODELCO), which counts currently with two productive open
pit mines, denominated “Chuquicamata” and “Mina Sur”.
Since the large copper ore reserve that lies below the “Chuquicamata” open pit will no longer be
economically feasible to mine from the end of this decade, the mine is currently planning to change
the operation from an open pit to an underground mining type, where the block caving method with
macro-blocks will be used. The future underground mine will correspond to one of the largest
underground mining operations in the world, with reserves equivalent to more than 60% of the ore
extracted over the last 90 years, and a production rate of about 140 thousand tons per day.
As part of this major CODELCO´s project, several permanent underground works are required for the
permanent mine access and operation. This paper presents the approach used for the geotechnical
design of these works, carried out at an advanced basic engineering level for the construction tender,
as part of the integral mine feasibility study commissioned to Hatch, with Geoconsult as specialist
consultant for tunnelling and geotechnical design. Construction of these infra-structure tunnels have
started in early 2012 and the underground mine operation is expected to initiate in 2018.

2 Project Characteristics – Permanent Underground Infra-structure Works

The permanent underground infra-structure works consist of the following:
• A main bidirectional traffic access tunnel for personnel, equipment, supply, waste and machinery,
2
7.45 km long, cross section areas of 49-72 m and a longitudinal gradient of -8.85%;
• A conveyor tunnel (housing the future mineral conveyor belt that transports the ore to the surface),
2
6.25 km long, cross section areas of 43-63 m and a longitudinal gradient of -15%;

-1-
• Five (5) fresh air injection tunnels and five (5) air extraction tunnels, all with a length of 4.3 km and
2
cross section areas of 73-102 m and a steep longitudinal gradient of -15%.
• Associated ancillary structures, such as shafts, cross passages, parking and emergency bays,
niches and others.

Ventilation Extraction Shafts/Tunnels Ventilation Injection Tunnels

Access Tunnel

Open Pit

Conveyor Tunnel
Underground
Mine

Figure 1. Overall View of Permanent Infrastructure Works: Air Injection, Air Extraction, Traffic Access and
Ore Transport Tunnels
Tunnels’ geometric design results from the best compromise among relevant factors, such as
geotechnical-structural behaviour, constructive and operational aspects, as well as durability and
safety requirements. The access and ore transport tunnels’ alignment are mainly governed by the
need to maintain longitudinal gradients to acceptable limits, resulting in straight lines from the surface
(entrance portals) to the first production level. For the mine traffic access tunnel, due to its hybrid
mine/transportation character, project-specific safety and operational criteria had to be defined. For
the ventilation tunnels, alignment is governed mainly by the mining layout. Injection tunnels consist of
parallel tubes, starting at the surface and ending at the mining level. Due to the great depth,
alignments are subdivided in three sub-horizontal segments interconnected by vertical intermediate
shafts, with about 180 m depth. Extraction tunnels, on the other hand, are connected to the surface by
means of very deep shafts (with about 970m depth), but have also a layout similar to the injection
tunnels, with three intermediate vertical shafts with about 180 m depth, connecting the system to the
different mining levels.
Excavation will be carried out mostly from the surface, with some sections carried out from within the
open pit and the mine interior. The use of the conventional excavation method, with consideration of
high mechanization and automation levels, is envisaged for the tunnelling works. Since the use of
TBM for the tunnel excavation had already been discarded by the client in the previous design stage,
mainly due to the steep tunnel inclination and the rock abrasivity, this construction method was not
further considered in the scope of the basic engineering design.
Geological setting consists of intrusive (granites, diorites, granodiorites, tonalites, quartzifer diorites
and monzonites) and meta-plutonic rocks (amphibolites) of the Chuquicamata complex, as well as
their respective hydrothermal alterations (weakly leached and structurally leached rocks). Along the
alignment, tunnels are expected to encounter mostly fair to good rock quality. Nevertheless, also some
adverse geological conditions are expected, such as in minor faults zones and areas affected by
leaching and weathering with different degrees of alteration. Ground water conditions are mostly
favourable and no major water inflows are expected, even though the presence of sodium sulphate
and other metals in the water, probably derived from adjacent mining chemical processes carried out
at the surface, could negatively affect the tunnel lining durability. The area is characterized by seismic
events of low to medium magnitude; however the region as a whole may be considered as highly
seismic with more than 80 events per day. Another relevant factor for the tunnel design and
construction is the in-situ stress conditions, with stress anisotropy and low to high tunnel overburden,
which will reach up to 1250m below the ground surface.

-2-
3 Geotechnical Design

3.1 Design Purposes and Process

For the advanced basic design, the main purposes of the geotechnical design were the following:
1) To analyze the geotechnical data in terms of sufficiency, coherence and confidence levels and
prepare a Geotechnical Data Report (GDR);
2) To assess the expected rock mass behaviour and potential geotechnical risks, which could affect
the project safety and construction baseline, and prepare a comprehensive Geotechnical
Reference Report (GRR) to support the tender process, portraying the best client´s understanding
of the geotechnical subsurface conditions;
3) To determine suitable excavation methods and permanent support measures, including technical
specifications for the safe and economic tunnel construction and provide a geotechnical model
that can be further adjusted and improved during the construction;
As a result of that, a probabilistic estimation of construction time and costs and material take-off for the
project construction baseline could be carried out.
The design was carried out on the basis of a step-by-step comprehensive methodology, similar to that
recommended by the Austrian Society for Geomechanics (2010), and outlined in the chart below:

Figure 2. Geotechnical Design Process (left); Design development throughout Project Stages (right)

3.2 Geotechnical Investigation Campaign

The geotechnical model is based on the study of extensive information obtained from the mine
operation, previous local experience (e.g. construction of exploratory tunnels) and investigations
carried out specifically for the new underground works. The investigation campaign carried out along
the tunnels´ alignment included mapping of both surface and existing underground galleries;
geophysical profiles (Vs and Vp waves); drilling of boreholes with core sampling, geophysical
scanning (e.g. caliper, gamma-ray, sonic probe, resistivity and neutron logs) and in situ verification of
rock mass parameters (e.g. PLT and permeability). Additionally laboratory testing was made to
selected core samples (e.g. UCS, TS, TX) and water (chemical tests). In certain portions of the

-3-
tunnels´ alignment, the existence of massive gravel and sterile mine dumps hindered the execution of
drillings, so that conditions at these sections had to be inferred by extrapolation of geotechnical data.

3.3 Geotechnical Model and Basic Geotechnical Units (BGUs)

The initial activity comprises an extensive review of the geological-geotechnical and hydro-geological
information, which is checked for sufficiency, coherence and confidence levels. Following that, ground
types with similar geotechnical characteristics, identified at the influence zone of the project, are
grouped in distinct “Basic Geotechnical Units (BGUs)”, namely the following: Granodiorite (GRD);
Anphibolites (ANF); Silificate Anphibolite (ANF-SIL); Granodiorite Elena Sur (GES); Weakly Leached
Rock (LXD) and Structurally Leached Rock (LXE).
Each BGU is characterized both in qualitative (descriptive) and quantitative terms, always considering
maximum, minimum and average values for each specific set of key parameters. The following
aspects are covered in the description of each BGU: lithology, alteration and mineralization, rock mass
description (sketches, photos, characterization) and behaviour, geo-structural characterization, hydro-
geological conditions (permeability and water inflow), presence of minerals, intact rock parameters,
rock mass parameters (based on the Hoek-Brown strength criteria), information confidence levels and
geo-mechanical classification Indexes such as RQD, IRS, GSI, Q’, RMR and RMR Laubscher.
Each tunnel sector is then characterised with the respective BGU, considering also a description of
eventual singular local characteristics, ground behaviour and key parameters variation. The figure
below shows a simplified example of a tunnel section with its respective geotechnical “sectorization”.

Sterile Dumps

LXD

GRD ANF ANF-SIL ANF GRD

Figure 3. Example of Tunnel Profile with Geotechnical Sectorization – Access Tunnel

For this project, the existing stress field around and below the open pit was determined by means of
stress measurements carried out with the “hollow inclusion” and hydraulic fracturing methods, both
inside and outside of the influence field of the open pits. These data was used to calibrate a
tridimensional numerical model (FLAC3D, Itasca Inc.) in regional scale, Geotechnical Update (2009),
where the average deviation error obtained with respect to individual measurements is about 44%.
High stress-field anisotropy was identified, with the principal stresses axis oriented at an East-West
direction. The maximum estimated stress ratios (horizontal/vertical) were about 2.0 and 0.81 for the
East-West and North-South direction, respectively.

3.4 Analysis of Tunnel Unsupported Sections – Rock Mass Behaviour

In this design step, the rock mass behaviour is assessed at each tunnel section, describing the
response of the ground to the full face excavation of openings, without the influence of supports,
division of face or auxiliary measures (Goricki et al, 2004). This assessment considers the respective
BGU characteristics and properties variation, as well as the specific influencing factors at each tunnel
sector, such as the tunnel geometry (shape/size) and orientation, geo-structural characteristics, stress
and water conditions as well as adjacent ground conditions. Generally, the expected ground behaviour
is categorized in one of the main general failure modes or as a combination of them. The Austrian
Society for Geomechanics (2010) proposes 11 general categories with possible sub-categories. For
further reference purposes, the table below shows a simplified list of major sources of opening
instability, in relation to the main controlling affecting effects, as proposed by Gomes, A. (2009):

-4-
Table 1. – Main Sources of Opening Instability – Gomes, A. (2009)
Sources of Instability Failure mechanisms mainly controlled by

(GD) Gravity and discontinuities (e.g. block and wedge failure, etc.)
(SI) Stress-induced (with/without association with discontinuities) due to stress redistribution
or dynamic loads (e.g. buckling, plastic or brittle fractures and creep).
(GW) Effect of groundwater (e.g. flowing ground, water pressure exerting, water inflow, etc.)

(MI) Presence of minerals (e.g. swelling, slaking, etc.), affected by water and environment
(CX) Presence of complex conditions, such as faults, mixed face, stratification, bim-rocks,
frequently changing ground, highly disturbed zones, low overburden, etc. (e.g. unstable
ground, ravelling, cave-in, etc.)

3.5 Analysis of the Supported Tunnel – System Behaviour

Excavation and Support Classes (ESC) must be established for each tunnel, including a complete set
of specifications of support, auxiliary measures, excavation and construction sequence. These classes
are designed to cover the complete range of expected geotechnical conditions, as described in the
project construction baseline.
For the system behaviour analysis, a preliminary excavation and support classes (ESCs) concept was
established on the basis of engineering judgment, experience in similar conditions, ground behaviour
analyses and recommendations based on the Q classification system (NGI, 1999). These ESC
concepts were considered for the analyses of the supported tunnel behaviour and its interaction with
the ground (system behaviour) for different combinations of ground quality (range of expected
variation) and applicable ESCs. In each analysis, results are compared to pre-specified acceptance
criteria for the rock mass (e.g. stress concentration/relaxation, disturbance/damage zones around the
excavation and magnitude of deformations and strains); and for the support elements (structural safety
factors). The process is interactive, allowing an optimization of the initially proposed ESCs, as to best
fit the range of stress-strain conditions and prognosticated failure modes. Additionally, the process of
ESCs optimization also takes into account the analysis of construction cycles, as explained in section
3.8, incorporating further efficiency in the construction logistic and maximization of excavation rates.
The following table shows the proposed ESCs in association to the ground behaviour and associated
failure modes, as considered for the system behaviour analyses.
Table 2. Relationship between ESCs and Possible Behaviour/Failure Modes (Gomes, A., 2009)

Exc. Supp. Possible Ground Behaviour / Failure Mechanisms

Classes (ESCs) (note: GW can theoretically occur in any ESC; MI is not expected)
ESC1 (GD): Stable, with elastic behaviour and negligible deformations. Eventual local block and
wedge failures, sliding or rotation along discontinuities;
ESC2 (GD): Mostly stable, with systematic minor rock falls which stabilizes quickly. Eventual
local block and wedge failures, sliding or rotation; (SI): minor brittle failure.
ESC3 (GD): Mostly stable with some local instability due to discontinuities; (SI): Stress-induced
shallow depth brittle or plastic failures. Minor tunnel deformation.
ESC4 (GD): Very disturbed/fractured ground with local instability due to discontinuities; (SI):
Stress-induced deep-seated brittle (spalling/bursting) or plastic failure. Tunnel
deformation expected in weak rocks. (CX): unstable ground condition due to weak zones
intersecting excavation or other complex conditions.
ESC5/
(GD): Highly disturbed/fractured and unstable ground; (SI) Pressure exerting with large
ESC6 / ESC- deformations in weak zones or spalling/bursting in massive rock; (CX) similar to ESC4
Portal (requiring advanced support and face subdivision).

3.6 Excavation and Support Classes (ESCs)

As a result of the previous analyses, the definitive ESCs are established and a probabilistic forecast of
ESCs distribution along the tunnels´ alignment is made, forming the basis for compensation clauses in
the tender documents, together with the respective technical and administrative specifications and
material take-offs.

-5-
The next table provides the general concept applied for the ESCs regarding excavation and support
measures for the principal tunnels (not applicable to singular sections, such as intersections, enlarged
profiles or the other ancillary structures).
2
Table 3. Excavation and Support Classes (ESCs) Concept - Principal Tunnels - Typical Sections (43-49m )

ESCs Max. Excav. Suitable Method of Support

Round (drilling length)
ESC1 ≤5-6m Fibre reinforced shotcrete and bolt pattern at crown
ESC2 ≤5-6m/ Fibre reinforced shotcrete and bolt pattern at crown-bench with
/ESC3 ≤4.5m increasing thickness/density
ESC4 ≤3.5m wire-mesh reinforced shotcrete and bolt pattern at crown-bench and
eventual forepoling (with reduction of exc. length)
ESC5/ESC6/ ≤1,5m wire-mesh reinforced shotcrete and bolt pattern at crown-bench,
ESC-Portal Eventual Mechanical steel girders and forepoling/pipe roof. May require top-
excavation heading/bench excavation. Temporary and/or permanent invert and
longer bolts for ESC6. ESC-Portal similar to ESC5 but with thicker
lining and less dense rock bolt pattern (higher structural capacity in
low confinement zones).

It shall be noticed that, only in case of unexpected conditions, such as occurrence of zones with large
water inflow, inner concrete lining installation is foreseen. Therefore, installed support measures are
generally proposed as structures of permanent character.

3.7 Engineering Tools for the Geotechnical Assessment and Analysis

3.7.1 General
The assessment of the rock mass/system behaviour is carried out with the support of engineering
tools (e.g. empirical, analytical or numerical methods), in the light of savvy engineering judgment.
Since geotechnical analyses are intrinsically affected by uncertainty - both in terms of input data and
inherently limitations of engineering tools – results must be treated probabilistically. This condition
must be properly expressed in the tender documents to allow for flexibility and an adequate risk
management during construction. In all cases, the actual tunnel behaviour must be monitored, verified
and corroborated on site (observational approach).

3.7.2 Empirical Methods

These methods were used to assess potential stress-induced (SI) failure modes in the two limits of the
rock competency scale, as well as failure modes controlled by gravity and discontinuities (GD). This
included estimation of the extent of squeezing (Hoek and Marinos, 2000) in rocks with GSI≤40; and
verifications of brittle failure (and potential of rock burst) in massive to moderately jointed rock with
GSI≥70 (Martin et al, 1999, 2009 and Hoek and Brown, 1980), among others. Additionally, Q’, RMR
and GSI indexes were also considered to assess the quality of the rock fabric of jointed rock in the
range of GSI=40-70, to support engineering judgment in the definition of required support measures.

3.7.3 Analytical Methods

Analytical methods were used to carry out assessments of ground characteristic lines (Carranza–
Torres & Fairhurst, 2000), stress distribution around the cavity (Feder, G. 1978), face stability and
others. Also, the structural stability of sectors where failure modes are controlled by discontinuities and
gravity (GD), with or without association with stress conditions (SI), were analyzed with simplified
structural checks (calculation sheets) and the use of stereographic projection and kinematics analysis
of wedges (e.g. software "Unwedge" of Rocscience Inc., based on Goodman and Shi, 1989).

3.7.4 Numerical Analysis Methods

To estimate the stress-strain (SI) behaviour of the ground and for the dimensioning the support
elements (system behaviour) under anisotropic loading conditions (e.g. ground and water pressure
and seismic loads), finite element models were used. The models were carried out with the software

-6-
Phase2D V7.0 of Rocscience, with the consideration of multiple stage analyses, allowing the
simulation of the tridimensional excavation process and the estimation of excavation longitudinal
displacement profiles (Vlachopoulus, 2009). Typically, the ground was modelled as a continuum,
applying the Hoek-Brown failure criteria based on the GSI (Hoek et al 2002). For rock masses with
GSI≤40, the post-failure behaviour was simulated with an elasto-plastic constitutive model, whereas
for GSI in the range 40 to 70, a simplified strain-softening constitutive model (with peak and residual
rock mass parameters as a function of ground strain) was applied, as shown in item “c)” of the
illustration below. The reduced GSI (GSIr) shall roughly cater for the ‘post’ excavation rock mass
degradation and the effects of anisotropic behaviour due to the orientation of weakness planes and
structural features.

GSIr=fr*GSI

fr=0.9 fr=0.7 fr=0.5

εp εr
Figure 5. a) FEM model b) BEM Model and c) Residual Parameters Criteria (GSI=40-70)

For competent rock masses (GSI≥70), failure criteria was mostly based on the analysis of stress
distribution in the rock mass (e.g. degree of concentration and relaxation), supporting the analysis of
tensile/brittle failure modes by means of empirically based methods (see 3.7.2).

Also MAP3D (Mine Modelling Pty Ltd) and FLAC3D (Itasca Inc.) modelling, based on the boundary
element and finite-difference methods, respectively, were applied to assess tunnel sections where 3D
stress conditions (overstress or loss of confinement) were relevant, such as in fault zones, tunnel
crossings and enlargements.

3.8 Construction Cycles and Probabilistic Time and Cost Estimation

As mentioned in section 3.5, extensive analyses of construction methodologies and working cycles,
together with the assessment of methods for muck handling and use of state-of-the-art construction
equipment were carried out, allowing further optimization of the ESCs. Probabilistic analyses of
construction cycles were carried out with the software @Risk (Palisade Corporation), based on the
Monte Carlo Method, which delivered a comprehensive and traceable estimation of construction time
and cost with the consideration of relevant influencing factors.
Estimated Overall Excavation Time (in months) – Estimated Advance Rate (in meters/month) –
Distribución
Access Tunnelpara Total Túnel
- Pessimistic Esp/BM193
scenario Access Tunnel - ESC3
0,250
Media=36,95302

0,200

0,150

0,100

0,050

0,000
32 36 40 44

5% 90% 5%
34,3136 39,9009

Figure 6. Example of Probabilistic Construction Time Charts - @Risk- Access Tunnel

-7-
 3.9 Conclusions
It is intrinsic that geotechnical information is hardly totally accurate and the most relevant tunnel
design and construction decisions have to be made in an earlier stage of the project, where actual
ground behaviour can only be inferred. Hence, the geotechnical design philosophy herewith outlined
focus on the definition of suitable and flexible excavation methods and permanent support measures,
based on the expected ground and system behaviours and the identification of geotechnical risks,
which may affect tunnel construction. The geotechnical design is aimed at providing a comprehensive
baseline and a traceable framework for both the tender process and the actual construction stage,
where tunnelling documentation and monitoring shall build on the proposed geological model and
enhance the understanding of the effective ground/system behaviour, i.e., requirements of excavation
and support, enabling the actual completion of the geotechnical design process.

3.10 Acknowledgments
The authors would like to express their gratitude to the National Copper Corporation of Chile
(CODELCO), and the companies Hatch and Geoconsult Latinoamerica, for the permission to publish
this article.

3.11 References
Austrian Society for Geomechanics (2010). Guideline for the Geomechanical Design of Underground Structures
with Conventional Excavation.
Bieniawski, Z. T.: Engineering rock mass classifica-tions. Wiley: New York, 1989.
Cai, M., Kaiser, P.K., Tasaba, M. (2007). Determination of residual strength parameters of jointed rock masses
using the GSI system. International Journal of Rock Mechanics & Mining Sciences 44 (2007) 247-265.
Carranza–torres, C. & Fairhurst, C. (2000). Application of the Convergence-Confinement Method of Tunnel
Design to Rock Masses that Satisfy The Hoek-Brown Failure Criterion. Tunnelling and Underground Space
Technology.
Chuquicamata Underground Project, 2009 Geotechnical Update, 2009, Itasca Denver, Inc.
Diederichs, M. S. 2007. Mechanistic interpretation and practical application of damage and spalling prediction
criteria for deep tunnelling. Can. Geotech. J.,
Feder, G. 1978. Versuchsergebnisse und analytische Ansätze zum Scherbruchmechanismus im Bereich
tiefliegender Tunnel. Rock Mechanics (6). 71-102
Hoek, E. and Brown, E.T. 1980. Underground excavations in rock. London: Inst. Mining and Metallurgy
Hoek, E and Marinos, P. 2000, Predicting tunnel squeezing problems in weak heterogeneous rock masses.
Hoek E., P. Marinos, EUROCK 2009, Tunnelling in Overstressed Rock.
Hoek E., C. Carranza-Torres, B. Corkum, 2002, “Hoek – Brown Failure Criterion – 2002 Edition”.
Gomes, Alexandre R.A., “Geotechnical Design of Tunnels”, Keynote Speaker “Conferencia Internacional de
Túneles y Construcción de obras subterráneas”, Elite Training; Lima, Perú, 2009
Goodman, Richard E., Introduction to Rock Mechanics, second edition, 1989.
Goricki A., Schubert W., Riedmueller G., 2004, New developments for the design and construction of tunnels in
complex rock masses, International Journal of Rock Mechanics and Mining Sciences
Martin, C.D., Kaiser,P.K., & McCreath,D.R. 1999. Hoek-Brown parameters for predicting the depth of brittle failure
around tunnels. Can. Geotech. Jour.
Martin, C.D. and Christiansson, R. 2009, Estimating the potential spalling around a deep nuclear waste repository
in crys-talline rock. Accepted for publication in Int. J. Rock Mech. Min. Sci. 46(2), 219-228.
Norwegian Geotechnical Institute, NGI (1999). Use of the Q-system in weak rock masses - Report nr: 592048-1;
Vlachopoulus N., Diederichs M.S., Improved Longitudinal Displacement Profiles for Convergence Confinement
Analysis of Deep Tunnels. Rock Mechanics and Rock Engineering (2009) 42.

-8-

View publication stats