Microtunneling in Mixed-Ground

Steve Hunt

Black &
Veatch 1
 Topics
1. Potential cobble and boulder impacts

2. Cobble and boulder concepts

3. Geologic setting

4. Subsurface investigation

5. GBRs and baselining

6. Risk mitigation measures

7. Conclusions
Black &
Veatch 2
Boulder Obstruction, Rescue Shaft
Two boulder
obstructions and
rescue shafts first
350 feet
(Minneapolis
MN320)
Boulder Obstruction, MTBM Abandoned

1115-ft drive
obstructed at
500 ft, MTBM
and 60” steel
pipe
abandoned
 Pushing cobbles, boulders aside
Peel Hwy 50 Trunk Sanitary Sewer – Tunnelling Complications with Rolling Rocks and
Damaged Pipes

Likely if soil matrix around cobbles and boulders is soft or lose

Over-ridden boulder, misalignment, cracked RCP

➢ Boulder not cut,

➢ Pushed aside,
➢ Rode over,
➢ Steering
difficulty,
➢ Misalignment,
➢ High contact
stresses,
➢ Cracked RCP,
➢ Rescue shaft.
Overbreak from plucked boulders resulting in
excess lost ground and severe settlement

Plucked perimeter boulders

leaves a large void
Jammed, stalled excavation chamber, stuck drive from
excess gravel, cobbles and boulders
Perhaps the most difficult ground condition for
microtunneling is a combination of cobbles and
boulders with a CVR + BVR greater than 10 percent in
a matrix of open, high permeability gravel.
Jammed, stalled excavation chamber, stuck drive
from excess gravel, cobbles and boulders
Overmining and sinkholes from excess flow of gravel
and cobbles into excavation chamber
Intake port abrasion, jammed excavation chamber
Severe cutter wear and breakage
Cutterhead wear
Broken pinion gears, gearbox from boulder impacts

Akkerman SL-74, Alameda Siphon No, 4

 Advance rate reduction as BVR increases

Columbia 58 ft per shift at 0 boulders per tunnel ft to

Slough, 22 ft per shift at 30 boulders per tunnel ft
(48% reduction)
Portland -
34,300
boulders
within,
8,300 ft of
15.2 ft
diameter
tunnel !
 Cobble and boulder concepts

1.Volume ratios and ranges of CVR, BVR, RVR

2.Energy consumption and clast strength

3.Abrasion, ‘tool’ wear, and breakage

Black &
Veatch 15
Cobble, Boulder Volume Ratio – Definitions
CobbleVolume
CVR(%) =
ExavationVolume
BoulderVolume
BVR(%) =
ExavationVolume
TotalRockVolume
RVR(%) =
ExavationVolume =CVR+BVR
 0
Trace (0-1%) Boulder Volume Ratio –
1
Descriptive Terms
Few (1-2%)
2

Many (2-5%)

5
Don’t use
Frequent (5-10%)
descriptive
10
terms
without
defining 20
Very frequent (10-50%)
them!

50

Extremely frequent (>50%)

100
Solid rock (100%)
Relative Cobble and Boulder Concentrations

(63%) of cases plotted

had a BVR <1%)
Relative Cobble and Boulder Concentrations

Microtunneling can be
significantly impacted at
BVR>0.5% and RVR values
>1%!
Relative Cobble and Boulder Concentrations – Allluvium

Granular Conditions

RVR 2-10% RVR 10-20% RVR 20-40% RVR 40-60%

Relative Cobble and Boulder Concentrations - Till

Cohesive Conditions
RVR 0.5-1% RVR 1-5% RVR 5-10% RVR 10-25%
 Cobble and boulder concepts

1.Volume ratios and ranges of CVR, BVR, RVR

2.Energy consumption and clast strength

3.Abrasion, ‘tool’ wear, and breakage

Black &
Veatch 22
Energy Consumption Matters
• Commutation (cutting and crushing rock clasts) requires
considerable energy.
• Higher CVR and BVR increases energy consumption.
• Higher UCS clasts require more energy to commutate.
• MTBM can stall and loose rotation from high energy
demand during commutation.
• Abrasion and wear increase and cutter life decreases as
energy consumption increases.
• Energy consumption is a cost contractors must bear.
 Commutation Energy

3”=7.6 cm

Commutating to gravel size for slurry mucking

system pumping requires more energy than to
larger size clasts for conveyor mucking
 Commutation Energy vs. Cutter Life

Higher UCS →
Lower cutter life → more
cutter change
interventions

Disc cutter life decreases as UCS and energy consumption increase.

Lower cutter life results in closer cutter change interventions.
Rock Clast Unconfined Compressive Strength
Most cobbles and boulders are the survivors of weathering and transport

22-45 ksi

14-49 ksi

Hunt & Del Nero, 2012, Mechanical Excavation in Bouldery Ground, Breakthroughs in Tunneling Short Course
 Cobble and boulder concepts

1.Volume ratios and ranges of CVR, BVR, RVR

2.Energy consumption and clast strength

3.Abrasion, ‘tool’ wear, and breakage

Black &
Veatch 29
 Soil Matrix
Cutter wear
Abrasivity
assessment Mixed Face Ground

involves three Cutter wear, breakage

types of ground
Rock Clasts
Cutter wear
Rock Abrasivity – Cerchar Abrasivity Index

Follow ASTM D7625-10, Laboratory Determination of

Abrasiveness of Rock Using the CERCHAR Method
 Rock Clast Abrasivity

Amoun et.al, 2015, W2015.352

Most cobbles and boulders have high UCS and high CAI.
 Abrasivity of the Soil Matrix - SAT
Over 12 Soil
Abrasivity
Testing (SAT)
methods have
been
developed –
none are
standards yet,
few directly
correlate to
cutter life
Most test 20
mm or less
None of the SAT methods account for cobbles and boulders
Abrasivity of the Soil Matrix - EQC
EQC Determination
1. Mineral content, Ai, quantified bya thin section analysis
or an X-ray-diffraction (XRD) analysis
2. Mohs hardness correlated to grinding hardness, Ri, for
each mineral after Rosiwal (1896, 1916), normalized to
quartz = 100

Soil Matrix EQC methods do not account for cobbles and boulders
Abrasivity – Typical EQC Ranges
2016, Moradizadeh et.al

Typical EQC Ranges

• Clays, silty clays: 0-15%
• Silts, sandy clays, clayey sand: 10-50%
• Sand, sand and gravel: 40-100%
• Cobbles and Boulders: 50-90%
• Igneous rocks: 40-65%
• Metamorphic rocks: 3-71%
• Sandstones rocks: 30-90%
• Limestone rocks: 3-6%

Total soil matrix + rock clast EQC correlates to cutter wear but not breakage
Cutter Life Considerations
• Cutter Life – cutter travel distance, Sc
• Cutter Life - average cutter ring life in m of tunnel driven, Hm
• Cutter Life – tunnel volume per cutters changed, Vc
• Intervention Interval in m, Li
• Abrasivity of the soil matrix – clay, silt, sand, gravel
• Extent of mixed-face ground or cobbles and boulders
• Cutter type – ripper, disc, scraper, pick
• Cutter steel and hardening inserts
• Cutter impact stresses – thrust force and speed
 Cutter Life - Sc, Vc, Li Concepts

Gage cutter
spiral tracking
distance
Sc = spiral cutter tracking distance until cutter worn
Vc - average cutter life per
excavated tunnel volume in De
m3/cutter = excavated
volume for Li / cutters
changed Li = Length bored until
intervention
 Ranges in Tracking Distance, Sc

From Kim et.al W2017.77

Cutter
Disc
Ripper
Scraper

Tracking distance Sc ranges from > 1000 km in clay to < 250 km in gravel
and till and is even less in mixed-face or gravel-cobble-boulder ground
Tool life in volume per cutter by EQC

EQC can be
used to
estimate
cutter life
for soil
matrix
abrasivity

Soil Matrix EQC methods do not account for cobbles and boulders
Tool life in length bored per cutter by EQC
EQC can be
used to
estimate
cutter life
for soil
matrix
abrasivity

Soil Matrix EQC methods do not account for cobbles and boulders
 Cutter Breakage From Impacts
TU2016.110

2017 Ren et.al

Impact breakage reduces

cutter life significantly!
Cutter Life and Intervention Interval - SAI
Koppl et.al 2013-2015
• Study of 18 slurry TBM drives
• Defined Soil Abrasivity Index - SAI
• Estimation of cutter life and intervention intervals

Where τc =shear strength of the soil in the excavation face, D60 grain size

SAI does not account for cobbles and boulders

Cutter Life - SAI [Also see Koppl et.al 2015]

Cutter Life, Sc, may be

used with SAI to Disc
estimate intervention Cutters
intervals
Scraper
s

SAI does not account for cobbles

and boulders
Example
cutter
change
percentage
in rock 33
%
Gauge and
11
perimeter
%
cutters
changed
most
frequently
Farrokh & Kim 2018
Cutter Life Vc for Mixed-Face Ground

Study of 30 mixed-ground cases

showed that cutter life dropped
rapidly as percent rock (RVR)
increased from <0.1% to ~5%

Hunt & Frank 2021

R2021.21
 Intervention Interval Li for Mixed-Ground

Study of 30 mixed-ground cases showed

5,000 ft
intervention interval, Li drop rapidly as
Mostly RVR increased from <0.1% to >~2%
Clay

Cobbly-
Bouldery Mixed-Face
Ground Rock
150 ft Ground

Hunt & Frank 2021

R2021.21
Intervention Interval Li for Bouldery Ground
Intervention interval Li increased rapidly as
BVR decreased from 5% to <0.5%

Cobbly-
Bouldery
Ground

Mostly
Clay
 Cutter Life Estimating in Mixed-Ground

EQC Use
• EQC can be determined for soil matrix and rock clasts
• EQC-mixed = EQC-soil x %soil + EQC-rock x %rock
• EQC-mixed does not account for cutter breakage from
impacts with rock, cobbles and boulders
• Cutter Life Vc using EQC-mixed should be decreased to
account for impacts – 20 % reduction suggested
Estimating Cutter Life, VC for Mixed-Ground

See Hunt & Frank 2021, RETC, R2021.27

 Topics
1. Potential cobble and boulder impacts

2. Cobble and boulder concepts

3. Geologic setting

4. Subsurface investigation

5. GBRs and baselining

6. Risk mitigation measures

7. Conclusions
Black &
Veatch 50
Geologic Setting Matters
 Glacial Deposits
Till Isolated boulder
Outwash
Ice Margin
Nested cobbles and
boulders

Nested boulders
Glacial Deposits at Soil Rock Interface

Concentrated angular cobbles and boulders

near soil-rock interface in Milwaukee

Mostly silty
clay till Glacial Advance

Bedrock
Glacial Deposits and Bedrock Ridges
Concentrated cobbles and boulders
past downstream side of bedrock ridge

Mostly silty Sand zones

clay till Glacial Advance

Bedrock Ridge
 Glacial Deposits, Erosional Lag Zones
Cobble-boulder lag zone in Outwash
Scattered cobbles and boulders
and small nested zones in Till

Till

Till Till

Outwash Till

Till
Till
Alluvial Deposits
Surface and buried
cobble zones

Bradshaw 8 Interceptor along American River, Sacramento

Colluvium - Talus
RVR>50%

Bedrock
Alluvial Fan Deposits
Alluvial Fan Deposits
Residual Soil, Weathered Granitic Deposits
Cobble and
boulder zone

Liscak et al, 2001 Thuro & Scholz, 2003

Weathered granitic corestones
Beach Erosional Deposits
Boulder Beach – Northwest Coast
Mixed-Ground, Wide Fault Zones
 Topics
1. Potential cobble and boulder impacts

2. Cobble and boulder concepts

3. Geologic setting

4. Subsurface investigation

5. GBRs and baselining

6. Risk mitigation measures

7. Conclusions
Black &
Veatch 65
Subsurface investigation for cobbles and boulders

Legget in his 1979 Terzaghi lecture reminded us that:

“There can never be any certainty about geological
conditions between adjacent boreholes, even 5 ft apart,
until the excavation has actually opened up the ground.”

Recognized uncertainties can be managed.

Subsurface investigation for cobbles and boulders

Medley 2002

Conventional borings alone are generally inadequate

Use conventional boring enhancement measures:
o RDR (Relative Drilling Resistance) logging
o SPT N-Value interpretations
Combine conventional with other SI methods: sonic
coring, large auger borings, and/or test pits
 Subsurface Investigation Options
Method Value for boulder invest.
Conventional borings Marginal with careful geotech logging,
evaluation

Cone penetrometer Will be obstructed in cobbles and boulders

Rotosonic coring Very useful and cost effective -

recommended

24 to 48-inch f auger bores Very useful, but costly and potentially

damaging

Test pits, quarries, outcrop Very useful, but may not penetrate tunnel
mapping zone soil

Geophysical (seismic refraction, Limited value for tunnel depths >20 ft

GPR, etc.) unless crosshole seismic

See: Hunt & Del Nero 2010, Hunt 2017

 Large Diameter Auger Boring

Sacramento, March 2003

30-inch
may
work –
36-48
inch
better
 Sonic Coring
Sonic coring is generally, the best method to
supplement conventional borings.
Sonic Coring Obvious cored
boulders
removed,
Cored sample measured and
dumped on weighed
Gilson sieve
machine

3” and
1-1/2” and
larger clasts
larger clasts
retained →
retained →
All retained
samples
weighed.
Particle Size Curves Typical misleading curve per ASTM with
100% missing cobbles and boulders.
Curve including cobbles an boulders
Instead of
poorly graded
gravel, GP, the
actual ground
may have been
Cobbles with
some gravel,
trace silt, sand
 Topics
1. Concepts and Objectives

2. Potential cobble and boulder hazards and risks

3. Subsurface investigation

4. Abrasion, ‘tool’ wear, and breakage

5. GBRs and baselining

6. Mitigation measures and specs

7. Conclusions

Black &
Veatch 73
Cobble and Boulder Baselining
GBR Items to Discuss
(Hunt and Del Nero, 2012+, Hunt 2002, Hunt &
Mazhar 2004, Hunt and Angulo, 1999)

Cobble and boulder volume ratios Baseline

Size distributions Baseline
Shapes, angularity
Unconfined compressive strength Baseline
Abrasivity Baseline
Quartz content Baseline
Matrix soil composition, strength Baseline
Geologic origin
Distribution in ground
 Cobble and boulder quantity baselining

Options: Avoid
potentially
1. Don’t baseline costly DSC
claims
2. Guess
3. CVR+BVR [Hunt: 1999, 2002; Boone et al, 1998]
4.Probabilistic: Frank & Chapman 2005; Medley,
2002; Felletti & Beretta 2009]
 Estimating Quantity Size Distribution
Enter
estimated
BVR, clast
size range,
mean and SD
for each
geologic unit
or tunnel zone
Example Boulder Baseline
 Clast Size Measurement and Baselining
ASTM D2487 or D2488 are often cited but are not clear:
“NOTE 2—For particles retained on a 3-in. (75-mm) US standard sieve,
the following definitions are suggested:
Cobbles—particles of rock that will pass a 12-in. (300-mm) square opening and be
retained on a 3-in. (75-mm) sieve, and
Boulders—particles of rock that will not pass a 12-in. (300-mm) square opening.”
Most labs don’t have a 12” square frame or sieve.

Orientation matters – is a 24” x 10” x 10” rock clast a

cobble or boulder? It would pass a 12” frame.
Clast Size Measurement and Baselining

Orientation
matters – is a
27” x 10” x 8”
rock clast a
cobble or
boulder? It
would pass a
12” square
frame.
 Clast Size Measurement and Baselining

Case for 27” x 10” x 8” clast = a cobble per ASTM D2488:

Dn = (2.0’∙ 0.833’∙ 0.667’)1/3 = 1.04 ft
VD = π/6(2.0’∙ 0.833’∙ 0.667’) = 0.58 ft3
De = (6VD/π)1/3 = (1.108)1/3 = 1.36 ft
Dmax = 2.25 ft, Davg = 1.25 ft, Dn = 1.04 ft, De = 1.36 ft
For a 12-inch = 1.0 ft f sphere, Vs = π/6(1.0)3 = 0.52 ft3
Vc / Vs = 0.58 / 0.52 = 1.12 or 12% more than volume for ASTM boulder size.
Conclusion: This clast should be considered a boulder. Define Dn or De in spec/GBR definitions.
Clast Size Measurement and Baselining Alts

A variety of methods can be

used to determine – estimate
clast size and volume by:
• Measuring lengths: a, b, c
• Measuring perimeters P1
and P2 (and P3)
• Weighing clasts
• Volume displacement by
dunking
• Photo analyses
 Topics
1. Concepts and Objectives

2. Potential cobble and boulder hazards and risks

3. Subsurface investigation

4. Abrasion, ‘tool’ wear, and breakage

5. GBRs and baselining

6. Mitigation measures and specs

7. Conclusions

Black &
Veatch 82
 Geophysical Prospecting
1. Borings: conventional, auger, sonic –
not specific
2. Surface seismic refraction – limited
resolution with depth
3. TBM vibration monitoring –
encouraging success
4. TBM acoustic monitoring w
geophones – limited use
5. TBM seismic reflection – limited to
slurry TBMs

See 2018, ITA Tech, Geophysical Ahead Investigation Methods, Seismic Methods, 36p
Black &
Veatch 83
 Penetration Rate Mitigation
Modified from TU2018.316 Farrokh & Kim

Typical pressure balance tunnel boring machines are limited to a cutterhead rotational speed of less than 4 RPM due to
the limitations imposed by the main bearing protection sealing system. Typical optimal penetrations for disc cutters
into hard rock is less than 0.25 inches (6 mm) per revolution, which results in advance rates of well under 1 inch (25
mm) per minute. This is less than 25% of what penetration/rev would typically be expected in ground not containing
boulders.
To minimize premature plucking and maximize disc cutter chipping of boulders or
rock, the penetration rate should be reduced to less than 10 mm/rev.
 Cutterhead Opening Ratio
High torque, Manageble torque,
stalled at COR = success at COR =
35% 13%

Oregon - Hickey & Staheli, No-Dig 2007

 Cutterhead Opening Ratio

High torque, Manageble torque,

stalled at COR = success at COR =
20% 13%

Hickey & Staheli,

Sunol,
No-Dig
California
2007
Cutterhead Opening Ratio

Suggested COR limits for cobble and boulder

volumes in a sand or gravel matrix

Total RVR, percent Suggested maximum

COR, percent
10 - 19 25
20 - 29 20
30 - 39 18
≥ 40 15
Cutterhead Torque – MTBM Power
 Potential Obstruction Specification
An obstruction occurs when an unexpected ground condition ( large
boulder or cluster of cobbles and boulders, manmade object) is
encountered at the heading of a tunnel that stops or significantly
inhibits forward progress to less than 10 percent of normal progress
for at least 30 minutes under normal thrust and torque with
properly functioning cutters. In addition, the obstruction must be
removed by supplementary means such as from the MTBM
excavation chamber during an intervention, removal by an
excavation made from outside of the tunnel or a rescue shaft.
 Topics
1. Concepts and Objectives

2. Potential cobble and boulder hazards and risks

3. Subsurface investigation

4. Abrasion, ‘tool’ wear, and breakage

5. GBRs and baselining

6. Mitigation measures and specs

7. Conclusions

Black &
Veatch 90
 Conclusions
1. Tunneling in cobbly-bouldery ground generally requires time
and expense – it is not incidental.
2. Cobble and boulder conditions should be baselined even if
measurement is difficult.
3. Ballpark baselines are ok – uncertainties will remain.
4. Specify prescriptive tunneling methods depending on risks and
potential consequences of obstruction.
5. Require excavation chamber access and backloading cutters.
6. Fairly define conditions for an compensated obstruction.
7. Compensation should depend on anticipated cobble-boulder
quantities, risks, uncertainties, and tunneling methods
specified.
 Conclusions
8. Complete a thorough SI and baseline gravel, cobble and
boulder conditions, including CVR and BVR = RVR.

9. Complete abrasivity testing and baseline results.

10.Prohibit or avoid an upsized (skinned up) MTBM or specify or

use maximum available torque for MTBM size.

11.If the potential consequences of a jammed MTBM or boulder

obstruction occurrence are serious (high risk), require or utilize
a robust MTBM with special features.

12.Consider requiring or using a COR < 20 % when K>10-2

cm/sec and clast volume ratio > 10% - thick slurry may not be
enough.
Microtunneling in Mixed-Ground

Contact Steve Hunt

HuntSW@BV.Com