Rock Engineering for an Underground

Storage
g Facility
y in Singapore
g p

Dr Zhou Yingxin
Senior Principal
p Engineer
g
Defence Science & Technology Agency
Outline

 Site characterisation and rock mass classification

 Tunnel design
g and construction
 Instrumentation and monitoring
 Co c us o s
Conclusions
Geology
gy Map
p of Singapore
g p
T05
Geology
gy and Site Investigation
g
 Where to find geology information?
 How much to do in site investigation

W can’t
We ’t ttake
k everything
thi ffor G
Granite!
it !
Planning
g for SI
 Dependent on objectives of SI work and many
other factors.
 Norwegian tunnelling recommendations: 2-10%
of excavation cost for road tunnels; 5-15% for
sub sea tunnels
sub-sea
 Caverns: about 1-2% of rock excavation cost

You pay for site investigations,

whether you do them or not!
- Paraphrasing Prof Kurt Klima
Factors Affecting
g SI Scope
p & Cost

 Purpose and scope of the investigation (feasibility, planning,

or design)
 Expected subsurface material and ground water
 Size and extent of facility (e
(e.g.
g road tunnels vs storage
facilities)
 Site conditions ((topography,
p g p y, access,, marine,, etc))
 Project specific requirements
 Environmental constraints and impacts
 Availability of equipment, technology and specialists
 Time, budget, and resources
What Is Site Characterisation

 Engineering
g gggeological
g investigation
g of the rock,,
rock discontinuities and rock mass at site and in
laboratory
 Integral part of the engineering design process for
any projects involving the ground
 I
Important
t t for
f layout
l t planning,
l i supportt design
d i and d
costing
 Also an important aspect of construction safety
and geological risk management
Phases of Site Characterisation

 Desk studies and site reconnaissance

 Acquire maps, papers, air photographs, imagery and satellite data
 Site visits and renaissance to confirm data and identify areas
where engineering difficulties may exist and areas for focused
investigation
 Site investigations
 Rock and rock mass properties
 Discontinuities and week zones
 Ground water,, permeability,
p y, in-situ stresses
 Data analysis and geological modelling
 Reporting
p g
Techniques
q for Site Investigation
g

 Geological
g mapping
pp g ((exposure
p and discontinuities))
 Geophysical surveys (detective work)
 Exploratory
p y drilling
g ((soil drilling
g and diamond core
drilling)
 Laboratory testing (rock material properties)
 In situ testing (rock mass properties)
 In situ stress measurements
Important
p To Remember

 No single
g technique
q is best
 Select techniques based on geology and data
requirements
 Complementary methods reduce uncertainty
 Different methods as cross check
 Allow for Phased SI during planning, design, and
construction
Methods of Investigation
g Used

Type Methods Objective

Drilling Soil boring; diamond core drilling Overburden, and rock cores
Surface Seismic refraction/reflection; electric Main geological structures;
geophysical resistivity tomography overburden depth
surveys
Borehole surveys Borehole logging; seismic logging; Ground temperature; Seismic
and testing borehole camera acoustic imaging; velocities; joints; and
p
impression p
packer;; borehole radar;; permeability;
p y; geological
g g
Lugeon tests; rising head/falling head structures
tests; cross-hole tomography
Laboratory tests Point load; uniaxial/triaxial compression; Mechanical properties of intact
Brazil tensile; 3-point flexural rock and rock joints
In situ stress Hydraulic fracturing; 3-D overcoring Hydraulic fracturing; 3-D
overcoring (during construction)
Cross--hole Tomography
Cross g p y
Cross hole Tomography
Vertical Seismic Profiling
g

Damaged qry wall

Vertical Seismic Profiling
g

Un-damaged wall
Composite
p Geological
g Profile

Weathered
Weathered Weathered EN1-ES1 trench – T12
trench – T11 trench – T12 F11 EN2-ES2 F11

Combined sections of seismic refraction (better vertical resolution)

and electrical resistivity profiling (better horizontal resolution)
Core Drilling
g - Fresh Granite
Intact Rock Properties
p
Properties Range Average
Densit (g/cm3)
Density 2 62 ~ 2
2.62 2.67
67 2 65
2.65
Uniaxial compressive strength 108.09 ~ 163.83
(MPa) 224.89
Young’s modulus (GPa) 37.10 ~ 111.25 65.87
Poisson’s ratio 0.14 ~ 0.35 0.24
Cohesion (MPa)  24.51
Internal friction angle (o)  59.02
P i t lload
Point d iindex
d 5 6  16.1
5.6 16 1 87
8.7
Brazil tensile strength (MPa) 8.46 ~ 14.30 11.71
Three-point tensile strength (MPa) 13 25 ~ 27.30
13.25 27 30 19 94
19.94
Core Drilling
g - Weathered Granite
Weathered Trenches

Trench Strike Extent, m Depth, m Weathering

Grade
T1 SN to 750 39 II III
II,
NE30o
T2 NNW to 950 80 II, III, IV
NE30
T3 NE25 900 47 II, III
Predominant Sub-vertical Joints
Geometries of Rock Joints
Joint Set Video Impression Acoustic Borehole Qry wall
logging packer imaging radar mapping

Sub-vertical 310/70 278/70 233/74 239/80

311/77 308/71 110/79 9/83

68/79 178/83

Sub- 98/6 0/0 23/10

horizontal

181/11

Medium dip 115/37 282/65

angle

292/55
Rock Joint Properties
p
Joint conditions Friction Angle,  Cohesion, C
(o) (Kpa)
Freshly fractured and dry 45 6
45.6 258

Freshly fractured and saturated 42.6 172

Freshly fractured and dry (weathered 36 8

36.8 183
rock)
Natural and dry 36.5 266

Natural and saturated 33 4

33.4 108

Mineral filled and dry 32.5 71

Mineral filled and saturated 27 3

27.3 52

Weathered and dry 27.6 200

Weathered and saturated 20 1

20.1 136
Permeabilityy

Soil 10-05 – 10-06 cm/s

Heavily weathered rock 10-06 cm/s

Jointed rock mass 10-08 – 10-09 cm/s

>> No major water inflow expected during construction

In situ Stress
 Stresses before excavation (virgin stress)
 V ti l stress
Vertical t v = H H
 Horizontal stress h = K v
 St
Strong influence
i fl off local
l l variations
i ti
 Estimate of average horizontal stress factor:

K = 3 - H/500 (for depth < 1000 m).

Example: H = 100,
100 K = 2.8
28

K = 9/8 - H/800 (for depth > 1000 m)

In Situ Stress
Test Method Hydraulic Fracturing 3-D Overcoring

Stress, Orientatio Stress, Orientatio

Mpa n Mpa n

Vertical stress 2.25 --- 3.0 ---

Maximum horizontal 7.3 13o 8.2 23 o

stress

Minimum horizontal 4.56 103 o 3.4 113 o

stress

Horizontal stress ratio: v: hmin: hmax = 1:2:3

Summaryy of Geological
g Properties
p

 No major
j tectonic faults
 Three-layer geological profile
 Deepp weathering g trenches
 Sub-vertical strips of densely jointed rock
 Favourable highg horizontal stresses
 Relatively low permeability
 Rock mass g generallyyg
good to veryyggood for
cavern construction
Geological
g Model

Three layer formation



• Residual Soil
• Weathered granite
• Fresh granite
Weathering trenches &
dense joint strips
Rock Mass Classification

 Rock Tunnelling
g Quality
y Index ((Q-system):
y )

Q = (RQD/Jn)(Jr/Ja)(Jw/SRF)

Relative block size: (RQD/Jn)

Inter-block shear strength: (Jr/Ja)
Active stress: (Jw/SRF)
Rock Mass Classification
Q Value Rock Mass Quality Percent, %

0 01 – 0.1
0.01 01 Extremely poor 19
1.9

0.1 – 1.0 Very Poor 3.7

1–4 Poor 5.8

4 – 10 Fair 13.6

10 – 40 Good 51.8

40 – 100 Very Good 19.3

> 100 Extremely Good 3.8

Tunnel Design and Construction
 The Drill and Blast Cycle
y
Drilling Charging

Survey Blasting

Rockbolt
Ventilation

Shotcrete Mucking out

Scaling
 Cycle
y Time for a Single
g Face

16 15
14 12.7
1999 2001
12
Time ((hrs)

10
8
6 4 3.6 4.5
4 2.6 1.9
1 0.4 1.51.6 1 0.5 1 1.51.6
2 0.50.5
0

rt
n

l
ng
g

g
n

ou

ta
tio

tio

po
lin

in
io

To
gi

al
at

ga

ng
ila
ril

up
r

Sc
ha
ilis

nt
i

ki

lS
av

Ve

uc
ob

tia
N

M
M

g/

i
In
tin
as
Bl

Activity
 Typical
yp Blasting
g Patterns
Perimeter holes

Production holes
65 m2 101 125 m2 151
holes holes

Cut holes

Lifter holes

Slash Pilot

Chamber
section: 275 m2

Bench 3
Bench, 3.5-m
5 m high
Blasting
g and Safety
y

 Drill and blast most common method for rock

excavation
 Storage and transport of explosives
 Vibrations effects:
 Building damage
 Sensitive
S iti equipment/buildings
i t/b ildi
 Human annoyance
 Blasting
g Using
g Bulk Emulsion
Rock cavern for on-site storage of
• On-site Storage detonating devices
– Licensed magazine to store detonators &
booster charges in temp cavern on site
– Reduced transport hazards to public
• Bulk Emulsion
– Non-explosives until being charged
– On-site storage reduced cost for magazine
rental and long lead time to withdraw
explosive for daily consumption
– Less toxic fumes means less ventilation
– Mechanised charging minimises human
exposure at drilling face

Cost savings
savings, better safety,
safety
Mobile Charging Unit
and improved productivity!
Blasting
g Efficiency
y and Overbreak
 Powder factor reduction from 2.5 kg/m3 to about 2 kg/m3
 Underbreak is more expensive to remove than overbreak
 Overbreak related to look-out angles, geology, and drilling accuracy,
and blasting design
 Typical
yp overbreak in g
granite rock: 14-28% tunnel area
12
Kg/m 3

Data for 65 m2 tunnel

10
Powder Factorr ,

4 y = 0.0845x + 1.6906

0
0 10 20 30 40 50
% off O
Overbreak
b k
Blasting
g Vibrations
n
 R
V  H B 
V = Peak Particle Velocity; H = constant;

Q  B = scaling
li llaw; n = attenuation
tt ti coefficient
ffi i t

H value, Exponent n Exponent B Remarks

mm/s
1109 1.4 1/3 Fully coupled tests in granite
21 - 804 0.88 - 2.8 1/3 Cube Root: China & Japan
1200 1.6 1/2 Square Root: Mining (USBM), Civil
engineering
1130 1 77
1.77 1/2 Hong Kong granite tunnel blasting
700 1.5 0.467 Average Swedish bedrock
700 2.0 1/3 Granite rock
193-1930 1.6 1/2 Down hole bench blasting
50 - 220 1.10 1/2 Coyote (large chamber) blasting
Mandai Granite
1.44
 R 
V  1099 1/ 3 
Q 
1000 Rock free field data
Tunnel Wall-Adjusted
Water pipe on soil
m/s

100 Quarry wall adjusted

velocity, mm

Soil surface - horizontal ppv

Soil surface vertical ppv

10
eak particle v

1
Pe

0
1.0 10.0 100.0
Scaled horizontal distance, m/kg1/3
Blasting
g Vibration Criteria

Country PPV (mm/s) Remarks

Norway/Sweden 18-70 Specifically stated for vertical PPV for different
geological media. Corrections are made for other
factors.
USA 50 Mostly based on US Bureau of Mines studies relating
to surface mines
UK 50
Switzerland 30

S
Singapore: 15 mm/s
/ is commonly
l imposed.
d
Good to be safe, but there is a price to pay! So excessive
conservatism is a waste of money!
Relative Costs for Vibration Control
Allowable Peak Particle Velocity, mm/s
Distance to
structure 20 mm/s 40 mm/s 60 mm/s

10 m 750 300 240

20 m 350 220 175

50 m 200 175 120

100 m 150 125 100

* Relative cost based on

study
t d for
f cavern excavation
ti
in Hong Kong
Ref: Berthelsen, 1992
Observed Threshold Values For RC
Structures
Material Building PPV Remarks
Type (mm/s)
Light concrete Residential 110
Old concrete Industrial 254 Structures expected to crack at 5-18 cm/s
in predictions
Concrete with Industrial 150 250
150-250 Initial concrete block cracks
masonry
foundations
Concrete Industrial 300 Tests showing lowest level corresponding
to cracking
Native stone with 1 1/2-storey 180-510 Subjected to progressively more intense
mortar joints & residential blast vibrations until damage was
rubble foundation observed.
Walls Residential 12.7 Door slams, Converted from strain
Walls Residential 22.4 Pounding nails. Converted from strain.
Walls Residential 76 Daily environmental changes

Singapore: 15 mm/s is commonly imposed.

 Rock Support
((Reallyy Reinforcement))
Rock Support
pp

 Rock bolts and shotcrete – still veryy much

empirical (usually for hard rock)
 No concrete lining used
 Water control or drainage a major consideration
 Spiling for some week zones
Empirical
p Design
g Process

 Site investigation
g and rock mass classification
 Preliminary design based on precedence and
rock mass quality
 Tunnel mapping by engineering geologist after
excavation
 Final support design prescribed based on
mapped conditions and actual rock mass quality
 S
Supported
t d by
b numerical
i l modelling
d lli (if necessary))
and instrumentation
Q-chart for Support
pp Design
g
Rock Bolt Length
g
Ref: E Hoek, 2001

General recommendations:
Roof: L = 2 + 0.15xSpan/ESR. Walls: L = 2 + 0.15xHeight/ESR
L = meters; B = Span; H = wall height; ESR = Excavation Support Ratio
Quantifying
y g Shotcrete Design
g

 Shotcrete not normallyy modelled as structural

element in numerical analysis
 Nominal shear strength = 2 Mpa
 Nominal bonding strength = 0.5 Mpa
 Estimate rock wedge volume
 Compare block weight to shear strength and
block weight to bonding strength
 Shotcrete increasingly accepted as final lining
for tunnels
Typical
yp Rock Support
pp
Class Q Type I Type II Type III
A >40 Spot Spot Spot
40 mm 40 mm 40 mm
B 10-40
10 40 L3(2.4) L4(2.4) L5(2.4)
40 mm 40 mm 50 mm
C 4-10 L3(2.2) L4(2.2) L5(2.2)
40 mm 40 mm 50 mm
D 1-4 L3(1.9) L4(1.9) L5(1.9)
50 mm 50 mm 75 mm
E <1 L3(1.5) L4(1.5) L5(1.5)
75 mm 75 mm 100 mm
Note: L3(2.4) = rock bolt length of 3 m at 2.4m center-to-center spacing
 Rock Bolts
 22 mm diameter Combination Tube rock bolts (CT-bolt) with end
anchoring and full grouting
 Corrosion
C i protection
t ti ((cementt grout,
t galvanising,
l i i d
double
bl coating)
ti )
Steel Fibre Reinforced Shotcrete
 Wet mix. Steel fire: 45 kg /m3. Alkaline-free accelerators. Water-cement ratio of
about 0.45. Rebound: 9 – 13%%, avg = 10%.
 Energy capacity test. (EFNARC, panel: 600 mm x 600 mm @100 mm thick loaded
at centre point. Joules)

5 mm 10 mm 15 mm 20 mm 25 mm

354 641 866 1042 1180

354 637 862 1047 1198

Rock Class F E D C, B A
Energy Absorption, Joules, 560 400 280 200 NA
RDP (40 mm)
Energy Absorption, Joules, 1400 1000 700 500 NA
EFNARC (25 mm)
Grouting
g

 Used to minimise water seepage,

p g , not to stop
p
water completely
 Probe holes of 25 m drilled into face when
necessary
 Pre-grouting when flow in three boreholes exceed
15 l/
l/min
i
 Grouting pressure usually at 30-50 bar
(equivalent to 30-50 m water pressure)
 Grouting
g Principles
p
• Pre-grouting much more effective than
post-grouting
• Good penetration (joints aperture and
cement type)
• P
Pressure important
i t t (high
(hi h pressure iis
better)
• Overlapping of grouting zones ((> 5m)
• Thickness must be larger than rock bolts
(look out angles).
W t bearing
Water b i zone
working
face

TUNNEL Grout holes

w

Probe drillhole
Rock Mass Classification Method

 General, conservative and may be inaccurate

 Does not consider failure mode, deformation or
support interaction
 C
Cannot t consider
id complexl properties
ti off rock
k mass
 Same rock mass rating with various
combinations of rock parameters
 No information on safety margin
 Difficult to account for favourable stress
conditions
 Q-system: does not consider joint orientation
Tunnel Geometryy Design
g

 Minimum competent
p rock cover of 20 m or 1.5
times span
 Empirical rule of 0.2 x tunnel span for arch height
results in significant unused tunnel space
 Does not take into consideration of favourable
h i
horizontal
t l stresses
t
Optimisation
p of Rock Support
pp

 Numerical modelling
g
 Geological representation
 Realistic rock mass properties
 Tunnel stability criteria (is deformation a good criterion to use in
hard rock?)
 Instrumentation and monitoring
 Use of results
 Absolute deformation vs measured deformation vs supported
deformation
A Tale of Two Tunnels

Minimum rock separation?

 Rule of thumb: 1.5 times the larger tunnel

=> 15 m
 RC slab of 2 m thick suggested
 Optimised design => 8 m
 Rock mass properties
 S
Stress conditions
 Effects of rock reinforcement
 Construction method/sequence
Tale of two tunnels

Numerical modelling for

optimisation

Principal Stress Distribution Around Two Tunnels

Effects of Joint Orientation
Tunnel Orientations vs Wedge Sizes
Based on Joint Data from Acoustic Imaging BH12: (136/57) (66/73) (280/44)
Volume of Wedges for Different Tunnel Orientations
Rectangle Tunnel (10m Width, 0 o Plunge)
180

160

140

Results of UNWEDGE calculations

120
Volume (m 3)

100

80

60

40

20

0
0 20 40 60 80 100 120 140 160 180
Tunnel Orientation (Degrees)
Instrumentation and
g
Monitoring
Instrumented Cavern Sections

• Instrumentations &
monitoring performed during
& after construction:
– Borehole extensometer
– Convergence (tape)
– Bolt
B l load
l d (strain
( i gauges))
Cavern Excavation Sequence
q
CL

Slash

5 metres

Benching
Combined Instrumentation Section
Surface MPBX

Notes:
The locations of all anchors of
multi-point
p extensometers all
displayed with reference to the
crown surface of the cavern.

25m
MPBX-1(BH 1)

12m
MPBX-2 (BH 2) 6m MPBX-3 (BH 3)

2m
Diameter=76mm
Di t 76
Diameter=76mm Length=25 m
Length=25.m
0m

CT-1 CT-3
CT-4 Rebar-1
CT-2 Rebar-3
MPBX-4
MPBX 4 (BH4) Rebar-2
Rebar 2
MPBX-5 (BH 5)
Diameter=76mm
Length=12.m Rebar-5
Rebar-4
3.2m 3.2m
CT-5
Measured Deformation in Surface
MPBX
Displacement and Progress
(S f
(Surface MPBX M Main
i Phase
Ph Chamber
Ch b 44--CH50)
CH50)
4.0 100
25 m 12 m
3.0
6m 2m
rock; +:intto cavern mm)

80

m)
20
2.0

Progress (m
ment (-:intoo

Slash Pilot
1.0 Benching 60
0.0
Displacem

Excavation P
-1.0 40

-2.0
20

E
30
-3.0
-4.0 0
0 50 100 150 200 250 300 350 400
Days after 03 December 2001 (Installation date)
Measured Deformation in Internal
MPBX after Installation
Displacement and Progress
(MPBX-BH1 Main Phase Chamber 4--CH50)
4.0 100
90
3.0
80
ment (-:into rrock;

on Progress (m)
2.0
m)

70
o cavern mm

1.0 60
0.0 50
Displacem

Excavatio
+:into

-1.0 40
0m 2m 30
-2.0 6m 12m
20
Pilot Benching
-3.0 Slash 10
-4.0 0
0 50 100 150 200 250 300
Days after 19 March 2002
Summary of Deformation
Measurements

 Upwards movement of chamber crown observed,

peaking when top heading passed measurement
position and eventually stabilising at less
downward movements
 Inward movement of chamber walls consistently
observed
Measured Bolt Loads
Bolt Load vs Time
60.0
C
Compression:
i -, Tension:
T i +
40.0
Bolt Load (kN)

20.0

0.0
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
-20.0
Point 1 Point 2 1
2
3
-40.0 Point 3 Point 4 4

-60.0
Days after 15 March 2001
Comparison with Gjovik Stadium in
Norway (after Broch et al.
al 1996)
Rock Conditions & Bolt GjØvik Stadium, Norway Singapore Site
Parameters (Based on Broch at el. 1996)
Typical Rock Mass Quality 1 - 30 4 - 36
Vertical Stress, MPa 1 2-3
Max Horizontal Stress 3.5 8.2
Minimum Horizontal Stress 2 4.6
Ratio of Hori. to Vertical Stress 2-3.5 2-3
Tunnel/cavern span, meters 61 10 – 30
Type of Rock Bolts Fully grouted rebars Fully grouted CT-bolts
L th meters
Lengths, t 6 m (with
( ith alternating
lt ti 12-m
12 36m
3-6
long cables
Spacing, meters 2.5 m x 2.5 m 1.5 – 2.4
p y, KN
Bolt Capacity, 220 250
Minimum Measured Loads, KN 1 – 1.5 3 - 12
Typical Measured Loads, KN 30 - 60 20- 60
Typical Load Percentage 13 – 27% 8 – 24%
Ma im m Meas
Maximum Measured
red Load,
Load KN 87 70
Max Load Percentage 40% 28%
Comparison
p of World-wide Cost
45000
ost, $/m
World wide tunnelling cost (Hoek, 2001)
40000
upport Co

35000
30000
25000
Excavation and Su

20000
15000
10000
5000
0
6 8 10 12 14 16
Tunnel Span, meters

 Competitive
p cost in Singapore
g p due p
primarily
y to a)) large
g tunnel
sections; b) good rock; and c) low labour cost
Conclusions
 No surprises in construction due to good site
investigations
 Bulk emulsion explosives proven very beneficial
for safety and productivity
 High horizontal stress key factor for stability of
large-span
g rock caverns
 For favourable horizontal stresses, support
design using Q-system could be further
optimized
ti i d
Useful References

 “Practical rock engineering.” E. Hoek,

ww.rocscience.com

 “The Complete ISRM Suggested Methods for

R k Ch
Rock Characterization,
t i ti T
Testing
ti and
dMMonitoring:
it i
1974-2006" known as the blue book.
www.isrm.net
International NATM Workshop
Title: “Tunnelling in Difficult geological conditions – An
International NATM Workshop”
Organiser: Society for Rock Mechanics & Engineering Geology
Austrian Society for Geomechanics
Date: 11 12 November 2010
11-12
Venue: Traders Hotel
Further info: www.srmeg.org.sg
Email: srmeg@cma.sg
Thank you!