Evaluation of liquefaction in tailings and mine waste: an ... 2018...

Evaluation of liquefaction in

tailings and mine waste:

an update

P.K. Robertson

2018

Definitions of Liquefaction

• Cyclic (seismic) Liquefaction– Zero effective stress

(during cyclic loading)

• Flow (static)

Liquefaction– Strain softening

response

Moss Landing, CA

Fundao, Brazil

Level – gently sloping ground

Steeply sloping ground

Tailings & mine waste

Flow Liquefaction is the one of the main design

issues for most tailings and mine waste structures

• High static shear (“steeply” sloping ground)

• Various trigger mechanisms

– Cyclic loading only one form of trigger

• High risk if significant strength loss possible

Flow liquefaction - Case histories

• Common soil features:– Very young age

– Non-plastic or low-plastic

– Little or no stress history (Ko ~ 0.5)

– Very loose (contractive)

– Low effective stress (s’vo < 3atm)

• Common instability features:– Some triggered by very minor disturbance

– Failures tend to occur without warning

– Failures tend to be progressive & rapid

– Observation approach not valid

Fundao, Brazil, 2015 19 deaths

Stava, Italy, 1985; 268 deaths, 190,000m3

Flow liquefaction – Evaluation Steps

Evaluation Sequence:

1. Evaluate susceptibility for strength loss

2. Evaluate stability using post-earthquake shear

strengths

3. Evaluate trigger for strength loss

If soils are susceptible, and instability possible

(FS<1), it is often prudent to assume trigger

will occur

Outline• Trigger events

– Small events can trigger strength loss

– Critical stress paths

• Classify materials susceptible to strength loss

– Application of SCPTu

• Stability

– Influence of high stresses

– Unsaturated tailings/waste

– Progressive failure

Triggers - Flow Liquefaction

After Olson & Stark, 2003

Undrained loading e.g. rapid construction A-B-C

Undrained cyclic e.g. earthquake A’-E-C

Unloading e.g. increasing GWL A-D-C

Example - Fundao

Unloading stress

paths are the

most critical -

can be drained

or undrained

Higher static

shear stress ratio

- smaller the

trigger

Lab. testing

http://fundaoinvestigation.com/

Classify susceptibility to strength loss

• Geo-materials must be strain softening in

undrained shear

• Strain softening geo-materials are contractive

at large strains

CPT can be used to identify contractive soil

for young, uncemented soils - Robertson(2010)

Case Histories of Flow Liquefaction

Case Histories

Nerlerk (sand) – 19,20,21

Jamuna (sand) - 34

Fraser River (silty sand) - 27

Sullivan mines (silty tailings) - 35

Northern Canada (silty clay) – 36

L. San Fernado Dam (silt) – 15

CPT data in critical layers +/- 1 sd.

(Average stress level < 200 kPa)

All flow liq. case histories plot in ‘contractive’portion of CPT SBT chart

Good theoretical support via State Parameter

19,20,21

34

35

27

36

15


su(liq) / s ’vo = liq. undrained strength ratio (sand-like)

Robertson, 2010

19,20,21

34

35

27

36

15



Robertson, 2010

Fundao

Fundao

Soil structure

• Some natural soils and mine tailings/waste have

some form of ‘structure’ that make their behavior

different from ‘ideal’ soil

– Macrostructure (layering, fissuring, etc.)

– Microstructure (particle scale – aging, bonding, etc.)

Canadian Geotechnical Journal, 2016

“CPT-based Soil Behavior Type (SBT) Classification System – an

update” P.K. Robertson

Seismic CPTu

After Mayne, 2014

SCPTu6 -7 measurements!

qt

fs

u2

Vs (Vp)

t50

uo

i

Diss.test

Identification of microstructure

• CPT penetration resistance, qt – controlled by

peak strength

• Shear wave velocity, Vs – controlled by small

strain stiffness

• Potential to identify ‘structured’ soils from

SCPT by measuring both peak strength and

small strain stiffness

New Go/qn Chart

1

10

100

1000

1 10 100 1000

Qtn

IG = Go/qt

Cementation/bonding

& aging

KG = (Go/qt)(Qtn)0.75

Small strain

rigidity index:

IG = Go/(qt-svo)

Normalized

rigidity index:

K*G = IG (Qtn)

0.75

Go / qn

Robertson, 2016

New Go/qn Chart

1

10

100

1000

1 10 100 1000

Qtn

IG = Go/qn

Soils with microstructure

(e.g. cementation/bonding

& aging)

K*G = (Go/qn)(Qtn)0.75

1

2

3

4

5

6

7

8

9 10

11

12

13

14

15

16 17 18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

Average

normalized

rigidity index for

young,

uncemented

silica-based

soils:

K*G = 215

Robertson, 2016

Updated CPT-based SBT Charts

Normalized SCPTu parameters: Qtn, Fr, U2 and IG

Microstructure

Ideal

Microstructure

Ideal soils –

no microstructure

Robertson, 2016


“CPT-based Soil Behavior Type (SBT) Classification System – an update” P.K. Robertson

Updated SBTn Charts

1

10

100

1000

0.1 1 10

Qtn

Fr (%)

IB = 32

IB = 22

CD = 70

CD = (Qtn - 11)(1 + 0.06Fr)17

IB = 100(Qtn + 10)/(70 + QtnFr)

SD

CD

CC

SC TD

TC

Soil Behaviour Type 1: CCS Clay-like - Contractive - Sensitive

2: CC Clay-like - Contractive

3: CD Clay-like - Dilative

4: TC Transitional - Contractive

5: TD Transitional - Dilative

6: SC Sand-like - Contractive

7: SD Sand-like - Dilative

CCS

Behavior Descriptions

Robertson, 2016

Soils with no microstructure

Updated SBTn Charts

1

10

100

1000

0.1 1 10

Qtn

Fr (%)

IB = 32

IB = 22

CD = 70

CD = (Qtn - 11)(1 + 0.06Fr)17

IB = 100(Qtn + 10)/(70 + QtnFr)

SD

CD

CC

SC TD

TC








CCS


Young-uncemented

NC to LOC soil

Aging, cementation

OCR

Sensitivity

Density

Robertson, 2016


Updated SBTn Charts

1

10

100

1000

0.1 1 10

Qtn

Fr (%)

IB = 32

IB = 22

CD = 70

CD = (Qtn - 11)(1 + 0.06Fr)17

IB = 100(Qtn + 10)/(70 + QtnFr)

SD

CD

CC

SC TD

TC








CCS


DRAINED CPT

UNDRAINED CPT

Transition -

Partial drainage


Saturated soils

Updated SBTn chart

Robertson, 2016

OCR > 4


Fundao

1536

Contours of Qtn,cs

70

6044

Contours of Qtn,cs on updated SBTn chart

OCR >4

4460

Ic

Kc

Robertson & Wride, 1998

Based on case histories of cyclic liquefaction

Contours of su(liq)/s’vo

0.25

0.100.05

Contours of su(liq) / s’vo based on Qtn,cs

Based on Qtn,cs(extended to Ic > 2.6)

OCR >4


0.25

0.100.05



OCR >4


Only applies to

Sand-like (SC)

drained CPT

Ic < 2.6 or IB > 32

Contours of su(liq/R)/s’vo

0.60

0.25

0.100.05

Robertson (2016)

OCR

Clay-like soils (undrained CPT)

remolded strength & sensitivity (St)

Modified for

clay-like soils(when Ic > 2.6)

St

St

OCR

NC Clay

St ~ 1

OCR >4

su(R) / s ’vo = remolded undrained strength ratio for clay-like soils

su(R) / s ‘ vo = fs / s ‘ vo = (Fr Qtn )/100

fs ~ su(R)

fs /s’vo = Fr Qtn /100

0.100.01 0.25


0.60

0.25

0.100.05

OCR

su(liq) / s ’vo = liq. undrained strength ratio for sand-like soils


Modified for

clay-like soils(when Ic > 2.6)

St

Contours of su(liq) / s ’vo & su(R) / s ’vo

Contours of su(liq)/s’vo

0.25

0.100.05



OCR >4



0.60

0.25

0.10

0.05

OCR



St

Contours of

su(liq) / s ’vo

su(R) / s ’vo

Based on case histories:

•Flow liquefaction for sand-

like soils

•Remolded shear strength for

clay-like soils

Both large strain

undrained strengths

IB


Stability

• Tailings dams are becoming increasingly

higher (>100m) with much higher overburden

stresses (> 8 atm).

– How do we extrapolate to higher stresses?

• Many tailings dams/waste pads have large

regions that are unsaturated

– How do we evaluate influence of lack of

saturation?

Questions?

Are all contractive soils strain softening?

Are all strain softening soils brittle?

How does stress level affect these?

PBDIII Earthquake Geot. Eng. , Vancouver - 2017

“Evaluation of Flow Liquefaction: Influence of high stresses” P.K. Robertson

State parameter in sands

State Parameter after Jefferies and Been, 1985Critical State Line

(CSL)

(+) Loose

CONTRACTIVE

(-) Dense

DILATIVE

Y = eo - ecsecs

eo

State Parameter after Been and Jefferies, 1985

p’op’csp’o /p’cs = 10(y/l)

Brittleness

Bishop (1967) defined brittleness:

IB = (tp – tr)/ tp

tp = peak strength

tr = residual (large strain) strength at same effective

normal stress

IB = 1.0 (100% strength loss)

IB = 0 (no strength loss)

tp

tr

t

g

Critical State Lines (CSL)Contractive

Dilative

Jefferies and Been, 2016

Limited

stress range

Linear approximation for CSL

CSL for range of sands

Critical State Lines (CSL)

Dilative

Contractive

CSL over wide stress range

Schnaid et al, 2013

Verdugo & Ishihara, 1996

CSL non-linear

over wide stress range

Schnaid et al, 2013


Boulanger 2003

Bolton (1986) Relative Dilatancy Index

Bolton (1986) Relative Dilatancy IndexIncreasing

compressibility/crushability

Schnaid et al, 2013


Boulanger 2003

Bolton (1986) Relative Dilatancy Index

Bolton (1986) Relative Dilatancy IndexIncreasing

compressibility/crushability

This format is very helpful to estimate CSL

Bolton’s empirical relationship is helpful guide

Y = 0.07

(y/l ~2.0)

Y = 0.20

Y = 0.25

(y/l ~0.9)

(y/l ~0.6)

p’o /p’cs = 10(y/l)

Mtc = 1.2

p’o /p’cs = 100

p’o /p’cs = 4.5

p’o /p’cs = 3.5

Example – Erksak Sand

Contractive

Dilative

Loosest state

Brittleness (IB) vs (p’o /p’cs)

po’/p’cs~NC Clay

A

B

IB = 0.9 – 0.84(p’o/p’cs)

C

Regardless of fabric and direction of loading

IB

Modified from

Sadrekarimi & Olson, 2011

(Ottawa)

(Illinois)

(Mississippi)

Brittleness (IB) vs (p’o /p’cs)

po’/p’cs~NC Clay

A

B

IB = 0.9 – 0.84(p’o/p’cs)

C

IB

Modified from


(Ottawa)

(Illinois)

(Mississippi)

Can use y but requires slope of CSL l

(y/l) which is changing

Brittleness (IB) vs su,cs/s’vo

IB

A

B

C

su,cs/s’vo

~NC Clay

Modified from


(Ottawa)

(Illinois)

(Mississippi)


IB

A

B

C

su,cs/s’vo

High Brittleness

Low Brittleness

Case histories

Modified from



IB

A

B

C

su,cs/s’vo

High Brittleness

Low Brittleness

Case histories

Modified from


su,cs/s’vo < 0.15

have higher brittleness

IB > 0.4

0.60

OCR



Contours of

su(liq) / s ’vo

su(R) / s ’vo

Based on case histories:

•Flow liquefaction for sand-

like soils

•Remolded shear strength for

clay-like soils

Both large strain

undrained strengths


0.15

High IBat low stress



0.60

0.25

0.10

0.05

OCR



St

IB

Sand-like and Dilative – SD

•Potential for cyclic liquefaction –

depends on size and duration of

cyclic loading

•CRR will increase with

increasing static shear

•Sampling difficult & in-situ

testing preferred

•Estimate CRR based on Qtn,cs

•No strength loss expected unless

state changes and/or some

microstructure (e.g. cementation)

Liquefaction Summary



0.60

0.25

0.10

0.05

OCR



St

IB

Sand-like and Contractive –

SC•Potential for cyclic & flow

liquefaction

•CRR will decrease with

increasing static shear

•Sampling difficult & in-situ

testing preferred

•Strength loss possible

•Estimate state (y) and su(liq)/s’vo

based on Qtn,cs

•Strain to trigger strength loss can

be small

Liquefaction SummaryContours in SC/TC

region based on

s‘vo < 2atm)


0.60

0.25

0.10

0.05

OCR



St

IB

Clay-like and Contractive –

CC/CCS•Potential for cyclic softening &

flow liquefaction

•CRR can decrease with increasing

static shear

•Sampling possible and

recommended

•Strength loss possible

•Estimate state (OCR) and su(R)/s’vo

based on CPT & FVT

•Strain to trigger strength loss can

be large – depends on plasticity

•Evaluate sensitivity using CPT fs

and FVT

•Check drainage CPTu dissipation

tests

•Measure wn and Atterberg




0.60

0.25

0.10

0.05

OCR



St

IB

Clay-like and Dilative – CD

•Liquefaction unlikely

•Sampling possible, if not too stiff

•No strength loss expected unless

state changes and/or some

microstructure

•Measure wn and Atterberg

•Drained strength valid




0.60

0.25

0.10

0.05

OCR



St

IB

Transitional– TD/TC

•Evaluate plasticity and drainage

(Atterberg Limits & CPTu

dissipation tests)

•Sampling possible depending on

plasticity and fines

•Evaluate assuming both sand-

like and clay-like and compare

• typically drained

strengths are more

conservative in SD, SC,

TD, TC and CD



Case History• Filtered mine waste placed via conveyor and radial arm

stacker with surface irrigation

• Placed in uncompacted lifts of about 20 to 30m thickness

• Overall slope of about 4H:1V

• Current max. height of 200m

• Approx. 100 million tons of waste

• Low precipitation & high evaporation region


“Characterization of unsaturated mine waste: case history”

Robertson, P.K., da Fonseca, A.V., Ulrich, B. & Coffin, J.

Typical index test results

• Uniform grain size distribution – sandy silt

• Fines content ~ 50%

• D50 ~ 0.065mm

• Specific gravity, Gs = 2.73

• Mostly non-plastic

• In-situ water content (w) varies from around

2% to 15% (depending on depth, location and

irrigation) with degree of saturation

5% < S < 50%

Typical CPTu

High stress

s’vo ~15atm

Typical Normalized CPTu

High stress

s’vo ~15atm

Typical Normalized CPTu

Updated SBTn Charts

Little or no microstructureMostly drained

CPT penetration

Mostly

Transitional Contractive

Updated CPT SBTn chart

Qtn-Fr chart indicates

that material is mostly

sand-like and

CONTRACTIVE at

large strains

(assumes no suction)

Robertson, 2016

0 to 15m depth range

s’vo < 300kPa


Qtn-Fr chart indicates

that material is mostly

sand-like/transitional

and CONTRACTIVE

at large strains

(assumes no suction)

Robertson, 2016

15 to 90m depth range

s’vo > 300kPa


0.25

0.10

0.05

With increasing depth,

Qtn decreases and Fr

increases

Trends toward a more

clay-like contractive

behavior, but often less

brittle

Seismic Velocity (SCPTu)

Vs

Vp

Vs1 = Vs (pa/s’vo)0.25

Seismic Velocity (SCPTu)

Vs1 ~ 225m/s

Vp < 1500m/s

Vs

Vp

Vs1

DilativeContractive

Robertson, 1995

Unsaturated

In-situ test interpretation

• CPT (Qtn) based interpretation suggests that

soils are mostly CONTRACTIVE at large

strains

• Shear wave velocity (Vs) based interpretation

suggests that soils are DILATIVE at large

strains

Which one is correct?

Critical State line (saturated)

Contractive

Dilative

Reconstituted samples

Loosest state

Limiting compression

curve (saturated)

?

In-situ stress range

Approx.

e(max)

Approx.

e(min)

Vs contours and CSL

Vs(field) > 200m/s

Assuming materials are saturated

Contractive

Dilative

Vs measured in lab.

at end of consolidation

Vs1 as a function of suction (s)

Increasing suction

(decreasing S)

3 tests 50 < sc < 600 kPa


Normalize Vs1 based on modified s’vo (adjusted for suction)

Sandy silt tailings


Vs1 as a function of suction (s)

Increasing suction

(decreasing S)

?3 tests 50 < sc < 600 kPa


In-situ range

Sandy silt tailings

CSL as a function of suction (s)

s = 0 kPa

Moist samples prepared to ei ~ 1.1 to 1.2

Sandy silt tailings

~loosest compression line after saturation


s = 0 kPa

Moist samples prepared to ei ~ 1.1 to 1.2

Sandy silt tailings

~loosest compression line moist

saturated samples


s = 200 kPa

s = 20 kPa

s = 0 kPa

Moist samples prepared to ei ~ 1.20

CSL (unsat.)

Sandy silt tailings

~loosest compression line moist


s = 200 kPa

s = 20 kPa

s = 0 kPa

CSL (unsat.)

Project tailings - Strain hardening when unsaturated

Limited strain softening if saturated

Sandy silt tailings

Case History - Summary

• Shear wave velocity (Vs) appears to be sensitive

to suction hardening, since it is a small strain

measurement

– Vs better indicator of unsaturated behavior

• Cone resistance (qc) appears to destroy

beneficial effects of suction hardening, since it is

a large strain measurement

– qc may better indicate behavior if soil becomes

saturated

Critical regions to investigate?

Tailings/waste

Foundation soils

Simplified piezometric

surface

Critical regions to investigate?

Tailings/waste

Foundation soils

Simplified piezometric

surface

Toe region:

• Higher static shear stress ratio

• Low effective confining stress

• Higher possibility to be 100% saturated

• Risk of progressive failure

Conclusion (1)

• Flow liquefaction can be triggered by

relatively minor conditions and ’unloading’

stress paths are the most critical, e.g.

– Earthquakes (even small earthquakes)

– Rising piezometric surface

– Unloading (movements in foundations or tailings)

• Not all CONTRACTIVE soils are strain

softening in undrained shear

• Not all strain softening soils are brittle

– brittleness varies with state, stress level and PI

• Although soils tend to become more contractive

with increasing stress they also tend to become

more ductile

• Low stress regions maybe most critical

Conclusion (2)

• SCPTu an excellent tool to evaluate potential

risk of flow liquefaction

– 5 to 6 measurements

– Ability to identify microstructure

– Classify soil behavior

– Vs and Vp powerful additional measurements

• Lab. testing to helpful to support evaluation via

CSL

Conclusion (3)

Questions?

PDF Copy of slides:www.cpt-robertson.com

Free (recorded) Webinars:www.greggdrilling.com/webinars

http://www.cpt-robertson.com/

Evaluation of liquefaction in tailings and mine waste: an ... 2018...

Documents

Transcript of Evaluation of liquefaction in tailings and mine waste: an ... 2018...