Draft - University of Toronto T-Space · Draft 1 Shear and dewatering behaviour of densified gold tailings in a laboratory simulation of multi-layer deposition Farzad Daliri1 Paul

Draft

Shear and dewatering behaviour of high density gold

tailings in a laboratory simulation of multi-layer deposition

Journal: Canadian Geotechnical Journal

Manuscript ID cgj-2014-0411.R3

Manuscript Type: Article

Date Submitted by the Author: 25-Feb-2016

Complete List of Authors: Daliri, Farzad; Thurber Engineering Ltd., ; Simms, Paul; Carleton University, Civil and Environmental Engineering Sivathayalan, Siva; Carleton University, Civil and Environmental Engineering

Keyword: Tailings, Desiccation, Drying Box, Simple Shear, Cyclic

https://mc06.manuscriptcentral.com/cgj-pubs

Canadian Geotechnical Journal

Draft

1

Shear and dewatering behaviour of densified gold

tailings in a laboratory simulation of multi-layer

deposition

Farzad Daliri1 Paul Simms

2 and Siva Sivathayalan

3

1 Geotechnical Engineer, Ph.D., Thurber Engineering Ltd., 180, 7330 Fisher Street

SE, Calgary, Alberta, Canada, T2H 2H8, E-mail: [email protected] (Corresponding

author)

2 Associate Professor, P. Eng. , Department of Civil and Environmental Engineering,

Carleton University, 1125 Colonel By Drive, Ottawa, Canada, K1S 5B6, E-mail:

[email protected]

3 Associate Professor, P. Eng. , Department of Civil and Environmental Engineering,

Carleton University, 1125 Colonel By Drive, Ottawa, Canada, K1S 5B6, E-mail:

[email protected]

Page 1 of 41



Draft

2

ABSTRACT

Tailings may undergo desiccation stress history under varied climatic and depositional parameters.

While tailings substantially dewatered prior to deposition may experience desiccation under the

greatest range of climatic variation, even conventionally deposited tailings may desiccate in arid

climates at lower rates of rise. Bench-scale research has shown that the stress history imparted by

desiccation substantially improves strength in gold tailings. The present study further investigates

this phenomenon by simulating multilayer deposition of high density tailings using a modular

drying box, 0.7 m by 1 m in plan. The box is instrumented for directly measuring evaporation,

drainage, water content, vertical volume change, and matric suction, and additional measurements

included total suction at the surface, and observations of crack development. The dewatering

behaviour conforms to that predicted by previously published generic modelling, specifically that

the presence of partially desiccated tailings initially accelerates but then decelerates dewatering of

fresh tailings. The shear behaviour of samples obtained using buried tubes and by driving thin-wall

tubes into the multilayer simulation are compared with shear behaviour of samples from bench-

scale experiments. Shear strength is independent of the sampling method, and shows higher strength

than the bench-scale samples. The higher strength may be due to greater number of wet-dry cycles,

or other age-related processes.

Keywords: Tailings, Desiccation, Drying Box, Simple Shear, Cyclic

Page 2 of 41



Draft

3

1 INTRODUCTION

In the last few decades, more than fifty tailings impoundments have failed due to seismic

activity, liquefaction, foundation failure, etc. Many of these failures have been attributed to

liquefaction of the high water content tailings or tailings used to construct the retention dykes

(ICOLD 2001; James et al. 2011). Densified Tailings were first introduced by Robinsky (1975,

1999) as a way to not only reduce the risk of dam failure but also potentially reduce the cost of

the Tailings Storage Facility (TSF) for certain projects by both reducing dam construction and

increasing water recycling. In this paper densified tailings are defined as any tailings that are

dewatered prior to deposition to the extent that they are non-segregating and exhibit a yield

stress, and therefore will form a gentle slope. This could include thickened tailings, paste tailings

(which are sufficiently dense to allow pumping in the laminar range), or filter cake tailings

(prepared using filters and so dense as to require transport by conveyor belt or truck). The typical

minimum solids content (and maximum water content) for hard rock tailings to be classified as

thickened paste, and filtered tailings are respectively 65% (w <54%), 68% (w < 47%) and 77%

(w <29%). In this paper, we focus on thickened or paste tailings from hard rock mining. Any

general conclusions stated in the paper only applies to those kinds of tailings.

Cycling deposition between a number of spigots facilitates the development of thin lifts, and

allows for a fresh layer to densify due to desiccation and/or drainage before burial during

subsequent deposition. Cycling deposition has been employed at a number of sites (Cooper and

Smith 2011, Kam et al. 2011, Shuttleworth et al. 2005). As discussed in Daliri et al. (2014) and

Al-Tarhouni et al. (2011), a freshly deposited layer experiences a variable stress history (Figure

Page 3 of 41



Draft

4

1a). Figure 1b shows an example of measured void ratio – stress data for the tailings used in this

study. Such stress history will include initial settling (self-weight consolidation), potentially

followed by suction-driven densification due to desiccation and drainage, before being rewetted

and subsequently consolidated by burial under new tailings. The time required for desiccation in

this step (hereafter called “drying time”) has been shown to substantially influence the final

strength (Daliri et al. 2014). Drying time may also play a role in acid rock drainage generation,

and possibly the average slope angle as well (Daliri et al. 2011, 2012; Kim et al. 2011; Simms et

al. 2007).

Daliri et al. (2014, 2011) simulated the effects of desiccation stress history in a small scale by

depositing two layers of tailings of thickness of about 10 cm in a cylindrical column with 25 cm

diameter. While Daliri et al. (2014) showed how drying time influences strength through stress

history in laboratory prepared specimens, for such knowledge to be useful, the drying time itself

must be determinable in the context of the large uncertainties existing in field deposition. To this

end, Simms and co-workers developed a numerical method to predict drying time in hard rock

tailings (Simms et al. 2007; Fisseha et al. 2010), and subsequently made generic predictions for a

range of climatic, material, and deposition (layer thickness) parameters (Simms et al. 2010),

essentially creating charts for assisting design. However, only a limited set of data for deep

deposits of thickened tailings were available to validate the method.

To bring greater confidence to the shear strength behaviour measured in laboratory samples

(Daliri et al. 2014), and to provide additional data to validate / improve the method of Simms et

al (2010) to predict drying time, a multilayer deposition experiment (1m by 0.7 m in plan

Page 4 of 41



Draft

5

“Drying box”) was constructed in the laboratory. This paper reports the details of the

construction of the drying box, and reports selected results used to characterize the dewatering

and shear strength behaviour of the deposit.

2. MATERIALS AND TESTING PROCEDURE

2.1 Materials

Tailings used in this study were collected from a gold mine located in Tanzania. The tailings

were collected at the end of a spigot, and hence were shipped at the pumping water content

(38%) used at the mine. However, due to settling during transport, the water content decreased to

around 22-25% and it was required to remix the tailings with the bleed water produced by

settling in order to re-produce the tailings with w (gravimetric water content) = 38%. The particle

size distribution of tailings was established by the combination of sieve (wet technique) and

hydrometer analyses based on ASTM D 422-63 (2002). Specific gravity, liquid limit, plastic

limit and shrinkage limit were determined as 2.89, 22.5%, 20% and 18% respectively, while the

D90, D60, D50, D30, and D10 are: 0.12 mm, 0.034 mm, 0.03 mm , 0.012 mm and 0.0015 mm.

The complete particle size distribution is available in (Al-Tarhouni et al. 2011).

Figure 2 shows Soil Water Characteristics Curve (SWCC) of the tailings, derived from axis-

translation measurements previously reported in Simms et al. (2007), and paired measurements

of total suction and water content, where total suction is measured using a dewpoint hygrometer

and water content by oven drying. Other properties of these tailings are reported in previous

work of the authors (Al-Tarhouni et al. 2011). The tailings generally fall into the range of typical

hard rock tailings reported in other studies (Bussiere 2007; Qiu and Sego 2001).

Page 5 of 41



Draft

6

2.2 Drying box characteristics and deposition plan

A modular drying box was designed and built in the civil engineering laboratory at Carleton

University to sequentially deposit tailings layers. The intent was to mimic the field deposition

and desiccation processes. The Plexiglas box was reinforced by removable steel bars. The box

was modular so as to maintain a consistent height from the top of the box to the tailings, to

maintain a more or less constant potential evaporation rate. The plan area of the box was 0.7 m ×

1 m, with a maximum depth of 1.5 m. The drying box was mounted on load cells to measure

water loss and evaporation rate. Four load cells were placed on the floor to carry each corner of

the box. Drainage was monitored using a tipping bucket placed underneath a drainage port in the

centre of the box. The bottom of the box was lined with a geotextile. 5TE volumetric water

content (VWC) sensors from Decagon were used to measure water content and temperature. 5TE

sensors are designed to measure VWC, electrical conductivity and temperature of soil. 5TE

determines VWC of the soil by measuring the dielectric constant of the media using

capacitance/frequency domain technology. The sensor uses a 70 MHz frequency, which reduces

salinity and textural effects making the 5TE accurate in most soils. The 5TE measures

temperature with an onboard thermistor, and electrical conductivity using a stainless steel

electrode array. The accuracy of the sensors for measuring VWC after calibration was ±1% - 3%

(Daliri 2013). Three 5TE sensors were placed in each layer at different locations (three different

heights and three different horizontal locations). UMS T5 tensiometers were used to measure the

matric suction. T5 tensiometers are designed to measure pore water pressure from +100 kPa

(water pressure) to -95 kPa (suction), though the range may be temporarily extended by proper

preparation of reservoir water (Guan and Fredlund 1997). After 100 kPa, the results may be

Page 6 of 41



Draft

7

reliable; but should be considered cautiously. The sensor body of T5 tensiometers is made of

acrylic glass. The tensiometer cup consists of porous aluminum oxide ceramic. The pressure

transducer reports the soil water tension as a linear output signal, with 1 mV corresponding to 1

kPa. The precision of the tensiometes are approximately ±2 kPa. Each layer of tailings in the

drying box had two or three tensiometers. Tensiometers were positioned near sampling ports, to

enable replacement after cavitation. A WP4-T Dewpoint Potentiometer was employed to

measure the total suction at the tailings surface from grab samples of the top 1cm of the tailings.

WP4-T uses the chilled-mirror dewpoint technique to measure the water potential of a sample. In

fact, the sample is equilibrated with the headspace of a sealed chamber that contains a mirror and

a means of detecting condensation on the mirror. At equilibrium, the water potential of the air in

the chamber is the same as the water potential of the sample. WP4-T measures water potential

from 0 to - 60 MPa, with an accuracy of ±0.1 MPa from 0 to - 10MPa and ±1% from -10 to - 60

MPa. Four TS-30S ultrasonic distance sensors were employed to measure the height of deposited

tailings, which was used to estimate the void ratio in the fresh tailings layer. TS-30S ultrasonic

distance sensors consist of a rugged transducer in a stainless steel case. The minimum and

maximum ranges of the sensor were 10 cm and 427 cm respectively. Accuracy of the sensor is

better than 0.5% of the target distance in stable environment. Figure 3 presents a picture and

schematic diagram of the drying box and employed sensors. Tailings were prepared in buckets

before deposition. Crack geometry was monitored by pictures and by hand, crack widths were

measured every 5 cm, and crack depth was estimated as best as possible using ruler or string.

Cracks were assumed to have a triangular cross section, when estimating crack volumes.

Five layers were sequentially deposited, each at an initial GWC of 38%. The first and second

layers were targeted to dry to w = 12% and w = 16% respectively. The three top layers were

Page 7 of 41



Draft

8

targeted to dry to the shrinkage limit (18%). Except for the fourth layer, which had an initial

thickness of 14 cm, other layers had a thickness of 18 cm. Two re-saturation processes, between

the fourth and fifth depositions, were performed in order to simulate light and heavy rainfalls.

The volumes of water with the rate of 40 mm/day and 500 mm/day were added by spray bottle

over the course of 9 and 2 hours.

2.3 Sample extraction methods

Two sample extraction methods were used to extract samples from different layers of the drying

box. In the first method, three open-ended tubes were placed vertically in the box on top of the

previous layer of tailings before deposition of the next layer. Buried tubes were extracted after all

layers were deposited. To extract the tubes, the steel parts of the box were detached and the tubes

were extracted manually. Each tube provided two or three samples for the simple shear test. To

verify that similar stress history occurred for samples inside the tubes, and outside, water

contents were measured both inside and outside the tubes (around 18%). In the second method,

two thin wall tubes with inner diameter of 69.95 mm and 70 cm depth were pushed into the

tailings using a hydraulic jack. As recommended for fine grained soils, a thin, sharp edged tube

with the length-diameter ratio of 1.4 and area ratio less than 15% was used (Wijewickreme and

Sanin 2004). The tubes were pushed in the middle of the box where there were no sensors.

Figure 4a shows three tubes prepared for depositing fresh top tailings layer and a buried tube that

have been buried in the bottom layer. Figure 4b shows the method of sampling using a thin wall

tube.

Page 8 of 41



Draft

9

2.4 Simple shear test

A simple shear apparatus well-described in recent publications (Al-Tarhouni et al. 2011, Daliri et

al. 2014) was used to conduct the strength measurement tests. The apparatus employed in this

study was a NGI type simple shear located in the geotechnical research laboratory at Carleton

University. The apparatus consists of a shear load frame, a vertical single acting air piston, a

horizontal double acting air piston, a constant speed motor drive, load cells, Electronic-

Pneumatic Transducer (EPT), and Linear Variable Displacement Transducers (LVDT). In the

apparatus, the sample is surrounded and fixed by a steel wire reinforced rubber membrane in

order to minimize the lateral deformation. The constant volume condition is obtained during

shear loading by keeping the height of the sample constant using a clamping mechanism. The

decrease or increase of vertical stress in a constant volume simple shear test is equivalent to the

increase (or decrease) of excess pore water pressures in an undrained test (Dyvik et al. 1987). As

noted by Sivathayalan and Ha (2011), the average stress and strain measurements yield a

representative response of the material in spite of the non-uniformities that occur in simple shear

specimens. Samples were consolidated and sheared in the device, monontonic shearing

continued up to 11% shear strain, while the National Research Council (NRC) liquefaction

criteria of 3.75% shear strain was adopted for the termination of the cyclic tests. Table 1 and

Table 2 show characteristics of monotonic and cyclic simple shear tests performed on drying box

samples. All the cyclic simple shear tests (Table 2) were performed on desiccated/rewetted

samples under 100 kPa consolidation pressure.

3.RESULTS

Page 9 of 41



Draft

10

3.1. Results of Multilayer deposition in the drying box

Deposition of five layers and two rewetting events were performed over 73 days. Figure 5

presents gravimetric water content (GWC) results of all layers. The average GWC was calculated

from the average value of VWC sensors in each layer, and by considering the void ratio

estimated from TS-30S senix distance sensors and crack measurements (Figure 6 shows an

example). The surface GWC was obtained by oven-drying grab samples (~ 30 g) taken from the

top 1 cm of each layer. The final water content of each layer was targeted to different values in

order to examine the effect of desiccation on the dewatering of subsequent layers. The targeted

water contents were 12, 16, and 18% for the first, second, and third through fifth layers

respectively.

The dewatering behaviour of each layer showed an evolution in time, and had two different

phases. The first phase characterized by settling and drainage of the fresh layer was shorter as the

volume of previously desiccated tailings increased. The presence of even marginally desiccated

underlying tailings increased the rate of settling of the fresh layer. For the fifth layer, following

two resaturations, the length of this phase increased, as the tailings were almost completely

resaturated prior to deposition of this layer. Quantitatively, it is shown that while the drying time

of the phase one of the fifth layer is between 1 ~ 2 days; other layers took less than one hour to

pass first phase of dewatering.

The second phase comprises drying from the post-settling water content (30%) or somewhat

lower. In contrast to the first phase, the duration/length of the second phase increased as the

volume of underlying tailings increased. The variation in dewatering is not due to variations in

evaporation, which is quite consistent from layer to layer (Figure 7). Rather, the increase in

Page 10 of 41



Draft

11

drying time is due to water transport up from the tailings underlying the fresh layer. This

behaviour is very well illustrated in Figure 5 after placement of the second layer, where the water

content of that layer sharply rises as water is absorbed from the rapidly dewatering fresh layer.

Subsequently on Day 11, the water content of the bottom layer begins to decrease – by this point

no drainage is being reported, so the water is flowing upwards. This is also illustrated by the

tensiometer data in Figures 8. Figure 8c shows the change in matric suction in Layers 1 and 2

along with the water content of Layer 1. Upon deposition, matric suctions in layer 1 are much

higher than in the fresh layer, but by Day 11, the gradient in matric suction (and effectively in

total head due to the small change in elevation, < 20 cm) reverses, showing that water is now

flowing from layer 1 to 2. This correlates strongly with the change from increasing to decreasing

water content in Layer 1.

Surface grab samples of about 1 cm in depth were analyzed for total suction using the WP4

device. Figure 9 presents total suction results of samples obtained from the surface and at crack

edges. Crack edges were particularly interesting as they showed the first visual signs of salt

precipitation (Figure 10). Evaporation appears to be locally higher at crack edges but then would

be suppressed by salt precipitation. There is a decreasing slope to the total suction increase with

each layer. This is correlated with the lower rate of dewatering of each fresh layer, shown in

Figure 5. In other words, total suction values are correlated with the water content of the fresh

layer, and dewatering of the fresh layer decreases with increasing number of layers.

3.2 Results of simple shear tests

3.2.1 Results of monotonic simple shear tests

Page 11 of 41



Draft

12

Figure 11 presents results of monotonic simple shear tests performed on samples obtained from

the buried tubes extracted from different layers under 50 kPa vertical effective stress. All

samples exhibited a strain hardening response with a distinct phase transformation (SPT) point in

the stress path plot. The first layer exhibited higher strength than other layers. Layers 2, 3 and 4

exhibited a fairly similar response at similar void ratios. The shear strength at phase

transformation SPT in layers 2, 3 and 4 was about 10 kPa and the SPT of the first layer was around

12.5 kPa. Samples obtained by driving in the thin-walled tubes yielded virtually identical results

(Figure 12).

3.2.2 Results of cyclic shear test

Cyclic simple shear tests were performed on samples obtained from different layers of the drying

box. All samples were obtained from buried tubes and consolidated to 100 kPa consolidation

pressure before the application of 0.1 Hz sinusoidal cyclic load. As noted earlier, a strain

criterion (NRC 1985) was used to define the triggering of liquefaction. Constant volume cyclic

tests were conducted over a range of CSR value from 0.075 to 0.20. Typical results of an

individual test are presented in Figure 13. The excess pore pressure ratio (ru) is defined as the

ratio between excess pore water pressure and the initial effective vertical stress.

4. DISCUSSION

4.1 Comparison of dewatering of the fifth layer with Simms et al. (2010)’s model

Page 12 of 41



Draft

13

This dewatering behaviour was anticipated by the generic modelling results of Simms et al.

(2010). This model is based upon unsaturated flow, uses a different SWCC for fresh and

desiccated tailings, and simulates the relatively short consolidation phase (where the pore-water

pressures are positive), through appropriate selection of initial conditions and the storage

parameter: details of this method are given in Fisseha et al. (2010) and Simms et al. (2007). The

generic modelling results presented in Figure 14 are show small variability for tailings with an

AEV between 50 kPa and 500 kPa, and saturated hydraulic conductivity between 10-6

m/s and

10-7

m/s at a void ratio of 1. The generic results are also relatively insensitive to the depth of the

fresh layer (for < 0.5 m) and to the depth of underlying tailings, which are assumed to be slightly

unsaturated. The fifth layer of the drying box is compared with those previously published

generic predictions in Figure 14. The generic modelling results are exactly as presented in

Simms et al. (2010). Field data in Figure 14 was previously reported in Simms et al. (2007),

Simms et al. (2012) for the Bulyanhulu mine, and Kam et al. (2011) and Simms et al. (2012) for

the Musselwhite mine. Tailings at the Musselwhite mine are also gold tailings of low plasticity.

The fifth layer dewaters faster than the generic predictions, as the latter are done assuming a

substantial depth of tailings (10 m+), which provide greater capacity to supply the fresh layer

with water. The modelling results, and the drying box results, illustrate the prominent influence

of the underlying tailings on dewatering: Namely, to initially accelerate dewatering of fresh

tailings even if the underlying tailings are only marginally desaturated, and thereafter to suppress

dewatering as underlying tailings supply water to the fresh tailings once the hydraulic gradient

reverses.

4.2 Comparison of shear behaviour to previous small scale tests

Page 13 of 41



Draft

14

Daliri et al. (2014, 2011) simulated the effects of desiccation stress history by depositing two

layers of tailings of thickness of about 10 cm in a cylindrical column with 25 cm diameter.

Samples were obtained from the bottom layer after varying degrees of desiccation and twelve

hours after the next layer was added. These samples were then consolidated and sheared in the

simple shear device. Figure 15 presents results of monotonic simple shear tests performed on

samples prepared by Daliri et al (2014)’s method under 50 kPa vertical effective stress.

Increasing desiccation stress history leads to increasing degrees of strain hardening, while

samples with no desiccation exhibited strain softening behaviour.

Figure 16 compares simple shear test results from Daliri et al. (2014) to the drying box results

using samples with close to the same stress history, stress state (vertical consolidation pressure of

50 kPa), and void ratio. Drying box tailings exhibit a somewhat stiffer and stronger response

than the bench-scale results for strains less than 6%. Drying box tests which have simulated the

field processes more closely would presumably be more representative of the true behaviour in-

situ. The bench-scale tests would then be somewhat conservative for purposes of design. The

increased stiffness of the drying box tests may be due to aging mechanisms, or the greater degree

of wet-dry cycles experienced in the drying box tests compared to the bench scale tests.

Figure 17 summarizes the drying box cyclic tests and compares them with the results of Daliri et

al. (2014). Figure 17 shows similar trends to the monotonic results and shows that the oldest and

bottom layer (Layer 1) is the strongest, and that all drying box samples show a somewhat

stronger response than the small-scale tests of Daliri et al. (2014). In the small-scale tests,

samples that had been desiccated to the shrinkage limit (18%) and below prior to re-saturation

Page 14 of 41



Draft

15

exhibited very similar cyclic resistance ratios. The variation in strain hardening at high strains

observed in the monotonic tests for different degrees of desiccation does not probably manifest

in the cyclic behaviour. The failure criteria in the cyclic tests was a shear strain of 3.75%, while

the maximum strain measured in the monotonic tests was between 10% and 11%. In general, the

bench-scale data showed somewhat lower strength than drying box data, but not excessively

weaker compared to measurements of the drying box samples. The higher strength of the drying

box samples might be due to increased number of wet-dry cycles, or some aging process.

For context, it is noted that Wijewickereme et al. (2005) and James et al. (2011) reported cyclic

resistance ratio (CRR) of gold tailings, gold-copper and copper gold-zinc tailings somewhat

higher than cyclic resistance ratio of the tailings of this study (0.02 ~ 0.1 higher in CRR10). The

difference could be attributed to different sample preparation methods, field sampling and aging

effects, different density, and differences in the materials properties.

5. SUMMARY AND CONCLUSIONS

Multilayer deposition of high density gold tailings was simulated by sequentially depositing five

layers, each at the pumping water content of 38% GWC, and with layer thicknesses between 0.14

and 0.18 m, in an instrumented drying box with a plan area of 0.7 m by 1 m.

The target water content of the last layer was varied to examine the influence of desiccation of

old layers on dewatering of fresh layers. Samples for shear strength testing were obtained by two

methods, one by excavating buried tubes, the second by driving thin-walled tubes from the

surface.

Page 15 of 41



Draft

16

Conclusions may be divided into two aspects, the dewatering behaviour, and the strength

behaviour. Regarding the dewatering behaviour:

• Dewatering behaviour of a given layer could be divided into two phases, distinguished by

a change in vertical head gradient between the fresh layer and underlying tailings from

downward to upwards.

• The rate of dewatering in Phase I was accelerated if underlying tailings were desaturated.

Capillary forces in the underlying layers would accelerate the self-weight consolidation

process.

• The rate of dewatering in Phase II, was decelerated by the presence of underlying

tailings. The flow of water in Phase II is reversed, and the underlying tailings, even if

partially desaturated, have the capacity to recharge the fresh layer and slow its

dewatering.

• This dewatering behaviour was anticipated by the generic modelling published in Simms

et al. (2010), based on the methodology developed in Simms et al. (2007) and Fisseha et

al. (2010). Comparisons of drying of the top layer show that the top layer dried faster

than the generic predictions. This is reasonable, as the depth of previously deposited

tailings in the generic predictions was much thicker than in the drying box.

• The potential influence of cracking in these tests seems to be blunted by the precipitation

of salts at crack edges. No signal in the evaporation data due to cracking was detected.

This is in remarkable contrast to similar work on fine grained tailings such as oil sand

fine tailings (Innocent-Bernard et al. 2014), where cracking is integral to evaporation.

Page 16 of 41



Draft

17

In terms of shear strength:

• Shear behaviour of samples obtained by the two types of sampling (extraction of buried

tubes, or driving thin-walled tubes into the tailings) was almost identical.

• Monotonic strength and stiffness, and cyclic resistance, were somewhat higher than those

obtained by Daliri et al. (2014), who simulated desiccation stress history in bench-scale

samples. This may be due to the increased number of wet-dry cycles experienced by the

tailings in the drying box, or some other age-related phenomenon. The first layer gave the

strongest response, which could not be explained by its desiccation stress history alone.

• The results of Daliri et al (2014) can be judged as somewhat conservative compared to

the tailings in the drying box. We expect this to be true for comparisons with field

samples as well. We hope to test this hypothesis in the near future.

ACKNOWLEDGEMENTS

This work was funded primarily by a collaborative research and development grant supported of

by Golder Associates and the Natural Science and Engineering Research Council of Canada.

Supply of tailings by Barrick Gold is gratefully acknowledged.

Page 17 of 41



Draft

18

REFERENCES

Al-Tarhouni, M., Simms, P., and Sivathayalan, S. 2011. Cyclic behaviour of reconstituted and

desiccated samples of thickened gold mine tailings, Canadian Geotechnical Journal, Vol.

48(7): pp. 1044-1060.

ASTM D422-63. 2002. Standard test method for particle-size analysis of soils, Annual Book of

ASTM Standards, American Society for Testing and Materials, Philadelphia, PA. Vol.

04.08.

Bussiere, B. 2007. Hydrogeotechnical properties of hard rock tailings from metalmines and

emerging geoenvironmental disposal approaches, Canadian Geotechnical Journal, Vol.

44(9): pp. 1019-1052.

Cooper, R.A., and Smith, M.E. 2011. Case study - operation of three paste disposal facilities.

Proceedings of the 14th International Conference on Paste and Thickened Tailings,

Perth, Australia, pp. 261-270.

Daliri, F. 2013. The influence of desiccation and over-consolidation on monotonic and cyclic

shear response of thickened gold tailings. Ph.D. Thesis, Carleton University, Ottawa,

Canada.

Daliri, F., Simms, P., and Sivathayalan, S. 2011. A comparison of different laboratory techniques

to simulate stress and moisture history of hard rock mine tailings. Proceedings of Tailings

and Mine Waste 2011 Conference, Vancouver, B.C., Canada. pp. 163-175.

Page 18 of 41



Draft

19

Daliri, F., Kim, H., Simms, P., and Sivathayalan, S. 2012. Contribution of desiccation to

monotonic and cyclic strength of thickened gold tailings ‒ not the same as over-

consolidation. Proceedings of the 15th international seminar on paste and thickened

tailings, Sun City, South Africa, pp. 73-84.

Daliri, F., Kim, H., Simms, P., and Sivathayalan, S. 2014. Impact of desiccation on monotonic

and cyclic shear strength of thickened gold tailing, Journal of Geotechnical and

Geoenvironmental Engineering (ASCE), 10.1061/(ASCE)GT.1943-5606.0001147

Dyvik, R., Berre, T., Lacasse, S., and Raadim, B. 1987. Comparison of truly undrained and

constant volume direct simple shear tests, Geotechnique, Vol. 37, No. 1, pp.3-10.

Fisseha, B., Bryan, R., and Simms, P. 2010. Evaporation, unsaturated flow, and salt

accumulation in multilayer deposits of a paste gold tailings. Journal of Geotechnical and

Geoenvironmental Engineering (ASCE), Vol. 136, No. 12, pp. 1703-1712.

Guan, Y., and Fredlund, D.G. 1997. Use of tensile strength of water for the direct measurement

of high soil suction. Canadian Geotechnical Journal Vol. 34: pp. 604-614.

ICOLD and UNEP. 2001. Tailings dams, risk of dangerous occurrences. Bulletin 121: Lessons

learnt from practical experiences Paris, 144.

Innocent-Bernard, T., Simms, P., Xiaoli, Y., and Sedgwick, A. 2014. Multilayer Deposition of

Two Batches of Thickened Oil Sands Tailings: Experiments and modeling. International

Oil Sands Tailings Conference, December 7th

-10th

, 2014, Lake Louise, Alberta.

James, M., Aubertin, M., Wijewickreme, D., and Ward Wilson, G. 2011. A laboratory

investigation of the dynamic properties of tailings. Canadian Geotechnical Journal, Vol.

48(11): 1587–1600.

Page 19 of 41



Draft

20

Kam, S., Girard, J., Hmidi, N., Mao, Y. and Longo, S. 2011. Thickened tailings disposal at

Musselwhite Mine, Proceedings 14th International Seminar on Paste and Thickened

Tailings (Paste2011), Perth, Australia, Australian Centre for Geomechanics, Perth, pp.

225–236.

Kim, H., Daliri, F., Simms, P. and Sivathayalan S. 2011. The influence of desiccation and over-

consolidation on monotonic and cyclic shear response of thickened gold tailings.

Proceedings of the 64th Canadian Geotechnical Conference (Pan-Am CGS), Toronto,

Canada.

National Research Council (NRC) 1985. Liquefaction of soils during earthquakes, National

Academy Press, Washington, D.C.

Qiu, Y., and Sego, D.C. 2001. Laboratory properties of mine tailings, Canadian Geotechnical

Journal, Vol. 38, pp. 183- 109.

Robinsky, E.I. 1975. Thickened Discharge - A New Approach to Tailings Disposal, Canadian

Mining and Metallurgical Bulletin, Vol. 68, December 1975, pp. 47-53.

Robinsky, E.I. 1999. Thickened Tailings Disposal in the Mining Industry, Toronto, Ontario,

Canada, E. I. Robinsky Associates Ltd.

Shuttleworth, J.A., Thomson, B.J., and Wates, J.A. 2005. Surface disposal at Bulyanhulu:

practical lessons learned. Proceedings of the 6th International Conference on Paste and

Thickened Tailings, Santiago, Chile, pp. 20-22.

Simms, P., Grabinsky, M., and Zhan, G. 2007. Modeling evaporation of paste tailings from the

Bulyanhulu mine, Canadian Geotechnical Journal, Vol.44, No. 12.,pp.1417-1432.

Page 20 of 41



Draft

21

Simms, P, Dunmola, A., Fisseha, B, and Bryan, R. 2010. Generic modelling of desiccation for

cyclic deposition of thickened tailings to maximize density and to minimize oxidation. In

Proceedings of Paste 2010, 13th International Seminar on Paste and Thickened Tailings,

Eds Jewell, R, and Fourie, A. Toronto, Canada, May 3rd

to 6th

2010, pp. 293-302

Simms, P., Daliri, F., and Dunmola, A. 2012. Deposition sequencing or “drying time” for multi-

point deposition of high density tailings, Proceedings of Tailings and Mine Waste 2012

Conference, Colorado, B.C., U.S.A.

Sivathayalan, S., and Ha, D. 2011. Effect of static shear stress on the cyclic resistance of sands in

simple shear loading, Canadian Geotechnical Journal, Vol. 48(10): pp. 1471-1484,

10.1139/t11-056

Wijewickreme D., and Sanin, M. 2004. Cyclic shear loading response of Fraser river delta silt.

Proceedings of 13th World Conference on Earthquake Engineering, Vancouver, B.C.,

Canada, paper No. 499.

Wijewickreme, D., Sanin, M. V., and Greenaway, G. R. 2005. Cyclic shear response of fine-

grained mine tailings, Canadian Geotechnical Journal, Vol. 42, pp 1408-1421.

Page 21 of 41



Draft

22

Figure Captions

Figure 1. a) Possible stress histories and b) a measured void ratio- stress history for a thickened

gold tailings sample

Figure 2. SWCC of gold tailings used in this study

Figure 3. a) Picture of the drying box (1 m × 0.7m × 1m) and b) schematic diagram of the drying

box with the employed sensors

Figure 4. Methods of sampling: a) buried tubes b) thin wall tubes using a hydraulic jack

Figure 5. Average GWC and surface GWC of all layers after deposition

Figure 6. Void ratio of second layer, estimated using vertical height only (from Senix sensors),

and vertical height plus estimates of crack volume

Figure 7. Actual evaporation (AE), measured by mass change and subtracting drainage measured

by tipping bucket

Figure 8. a) Matric suction evolution measured at the first/bottom layer, b) matric suction after

each freshly deposited layer, c) matric suctions and water content after placement of second layer

Figure 9. Total suction measurements on samples obtained from top 1 cm of tailings

Figure 10. Evolution of cracks and salt precipitates, a) Day 22, b) Day 24, c) Day 27

Figure 11. Results of simple shear tests on samples obtained from buried tubes under 50 kPa

vertical effective stress

Figure 12. Difference between the responses of samples obtained from buried tubes and thin wall

tubes extracted by sampling method

Figure 13. Cyclic simple shear response of the sample obtained from Layer 1 at CSR = 0.075

Figure 14. Generic predictions of Simms et al. (2010) compared to fifth layer dewatering (red),

field data from Bulyanhulu (Simms et al. 2007), and Musselwhite (Kam et al. 2011). PE is

potential evaporation

Figure 15. Monotonic results of simple shear tests performed on samples prepared by Daliri et al.

(2014)’s small scale method of simulation under 50 kPa vertical effective stress

Figure 16. Comparison of small-scale tests from Daliri et al. (2014) with drying box samples

Figure 17. Cyclic stress ratio (CSR) versus numbers of cycles to reach liquefaction of drying box

samples, compared with small-scale samples from Daliri et al. (2014)

Page 22 of 41



Draft

a)

b)

Figure 1. a) Possible stress histories and b) a measured void ratio- stress history for a

thickened gold tailings sample

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

0.1 1 10 100 1000

Void

Ratio

Matric suction or Net normal stress (kPa)

Desiccation

Reweting

Consolidation after rewetting

Page 23 of 41



Draft

Figure 2. SWCC of gold tailings used in this study

Page 24 of 41



Draft

a)

b)

Figure 3. a) Picture of the drying box (1 m × 0.7m × 1m) and b) schematic diagram of the drying

box with the employed sensors

Page 25 of 41



Draft

(a) (b)

Figure 4. Methods of sampling: a) buried tubes b) thin wall tubes using a hydraulic jack

Buried Tube

Page 26 of 41



Draft

Figure 5. Average GWC and surface GWC of all layers after deposition

Page 27 of 41



Draft

Figure 6. Void ratio of second layer, estimated using vertical height only (from Senix

sensors), and vertical height plus estimates of crack volume

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

8 10 12 14 16 18

Void Ratio

Time (Day)

Void Ratio based on Vertical Height

Page 28 of 41



Draft

Figure 7. Actual evaporation (AE), measured by mass change and subtracting drainage measured

by tipping bucket

0

1

2

3

4

5

6

7

8

9

10

0 10 20 30 40 50 60 70

AE (mm/day)

Time (Day)

First Layer Second Layer Third Layer Fourth LayerFirst Resaturation Second Resaturation

Fifth Layer

Page 29 of 41



Draft

Figure 8. a) Matric suction evolution measured at the first/bottom layer, b) matric suction after

each freshly deposited layer, c) matric suctions and water content after placement of second layer

-3

17

37

57

77

97

117

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00

Mat

ric S

uctio

n (k

Pa)

Time (Day)

Tensiometer 1

Tensiometer 2

Tensiometer 3

First Layer

Second Layer

Third Layer

Fourth Layer

First Resaturation

Second Resaturation Fifth Layer

Deposition

-3

17

37

57

77

97

117

0.00 20.00 40.00 60.00

Mat

ric S

uctio

n (k

Pa)

Time (Day)

Tensiometer 1Tensiometer 2Tensiometer 3Tensiometer 4Teniometer 5Tensiometer 6Tensiometer 7Tensiometer 9Tensiometer 10Tensiometer 11Tensiometer 12Tensiometer 13

First Layer Second Layer

Third Layer

Fourth Layer Fifth Layer

Re-Saturations

b)

a)

c)

Page 30 of 41



Draft

Figure 9. Total suction measurements on samples obtained from top 1 cm of tailings

Page 31 of 41



Draft

a) b) c)

Figure 10. Evolution of cracks and salt precipitates, a) Day 22, b) Day 24, c) Day 27

Page 32 of 41



Draft

Figure 11. Results of simple shear tests on samples obtained from buried tubes under 50 kPa

vertical effective stress a) stress strain, b) stress path, c) excess pore pressure-shear strain

0

5

10

15

20

25

30

0 2 4 6 8 10 12

Shea

r Stres

s (k

Pa)

Shear Strain (%)

Buried Tube, Layer 1, Wd = 12.55%, ec= 0.625

Buried Tube,Layer 2, Wd=16.88%, ec=0.633

Buried Tube, Layer 3, Wd = 18.18%, ec =0.623

Buried Tube,Layer 4, Wd = 17.79%, ec = 0.619

a)

0

5

10

15

20

25

30

0 10 20 30 40 50 60

Shea

r Stres

s (k

Pa)

Effective Normal Stress (kPa)

Buried Tube, Layer 1, Wd =

12.66%, ec = 0.625


16.88%, ec=0.633

Buried Tube, Layer 3,

Wd=18.18%, ec =0.623


17.79%, ec = 0.619

�′�� = 50 kPa

b)

0

5

10

15

20

25

30

0 2 4 6 8 10 12

Exce

ss P

ore

Pre

ssure

(kPa)

Shear Strain (%)





c)

Page 33 of 41



Draft

Figure 12. Difference between the responses of samples obtained from buried tubes and thin wall

tubes extracted by sampling method

0

5

10

15

20

25

30

0 2 4 6 8 10 12

Shear Stress (kPa)

Shear Strain (%)





Thin wall Tube, Layer 1, Wd = 12.55%, ec = 0.628

Thin Wall Tube, Layer 2, Wd= 16.88%, ec = 0.617

Thin Wall Tube, Layer 3, Wd = 18.18%, ec = 0.616

Thin Wall Tube, Layer 4, Wd = 17.79%, ec = 0.626

Page 34 of 41



Draft

Figure 13. Cyclic simple shear response of the sample obtained from Layer 1 at CSR = 0.075

-10

-8

-6

-4

-2

0

2

4

6

8

10

-5 -4 -3 -2 -1 0 1 2 3 4 5

Shear Stress (kPa)

Shear Strain %

CSR = 0.075, Layer 1 (Wd = 12.7%)

ec = 0.595 , NL = 102a)

-10

-8

-6

-4

-2

0

2

4

6

8

10

0 20 40 60 80 100 120

Shear Stress (kPa)


CSR = 0.075, Layer 1 (Wd = 12.7%) b)

0

0.2

0.4

0.6

0.8

1

1.2

0 20 40 60 80 100 120

Excess Pore Pressure Ratio (ru)

Number of Cycles

CSR = 0.075, Layer 1 (Wd = 12.7%)

ec = 0.595, NL =102

c)

Page 35 of 41



Draft

Figure 14. Generic predictions of Simms et al. (2010) compared to fifth layer dewatering (red),

field data from Bulyanhulu (Simms et al. 2007), and Musselwhite (Kam et al. 2011). PE is

potential evaporation.

Page 36 of 41



Draft

Figure 15. Monotonic results of simple shear tests performed on samples prepared by Daliri et

al. (2014)’s small scale method of simulation under 50 kPa vertical effective stress

0

5

10

15

20

25

30

0 10 20 30 40 50 60

Sh

ea

r S

tre

ss (

kP

a)


W=30%, ec=0.520

W=28%, ec=0.580

Wd=23%, ec=0.604

Wd=19%,ec=0.619

Wd=17%, ec=0.623

Wd=12%, ec 0.628

Wd= 4%, ec=0.636

��vc = 50 kPa

Page 37 of 41



Draft

Figure 16. Comparison of small-scale tests from Daliri et al. (2014) with drying box samples

0

5

10

15

20

25

30

0 2 4 6 8 10 12

Shear Stress (kPa)

Shear Strain (%)

Buried Tube, Layer 1, wd = 12.55%, ec = 0.625Buried Tube, Layer 2, wd =16.88%, ec = 0.633Small Scale, wd = 12%, ec = 0.628Small Scale, wd = 17%, ec = 0.64

Page 38 of 41



Draft

Figure 17. Cyclic stress ratio (CSR) versus numbers of cycles to reach liquefaction of drying box

samples, compared with small-scale samples from Daliri et al. (2014)

Page 39 of 41



Draft

Table 1. Monotonic simple shear tests - sample details

Test

No.

Drying

Box Layer

Water content

before rewetting

Wd (%)

Method of

Sampling

Consolidation

Pressure (�� )

kPa

Void ratio

after

consolidation

(ec)

1

2

3

4

5

6

7

8

9

10

11

12

One

Two

Three

Four

One

Two

Three

Four

One

Two

Three

Four

12.5

16.9

18.2

17.8

12.5

16.9

18.2

17.8

12.5

16.9

18.2

17.8

Buried

Buried

Buried

Buried

Thin Wall

Thin Wall

Thin Wall

Thin Wall

Buried

Buried

Buried

Buried

50

50

50

50

50

50

50

50

100

100

100

100

0.625

0.633

0.623

0.619

0.628

0.617

0.616

0.626

0.616

0.622

0.624

0.618

Page 40 of 41



Draft

Table 2. Cyclic simple shear tests - sample details

Test No.

Drying

Box Layer Wd

(%)

Void Ratio after

Consolidation (ec)

Cyclic Stress

Ratio (CSR)

Number of Cycles

to reach

Liquefaction (NL)

1 Layer 1 12.7 0.595 0.075 102

2 Layer 1 12.7 0.633 0.1 28

3 Layer 1 12.7 0.607

0.125 12

4 Layer 1 12.7 0.608 0.15 7

5 Layer 1 12.7 0.603 0.175 4

6 Layer 1 12.7 0.607 0.2 4

7 Layer 2 16.9 0.598 0.075 86

8 Layer 2 16.9 0.606 0.1 20

9 Layer 2 16.9 0.638 0.125 9

10 Layer 2 16.9 0.619 0.15 6

11 Layer 2 16.9 0.611 0.175 3

12 Layer 2 16.9 0.609 0.2 2

13 Layer 3 18.2 0.586 0.075 82

14 Layer 3 18.2 0.589 0.1 19

15 Layer 3 18.2 0.593 0.125 9

16 Layer 3 18.2 0.619 0.15 6

17 Layer 3 18.2 0.614 0.175 2

18 Layer 3 18.2 0.604 0.2 2

19 Layer 4 17.8 0.586 0.075 77

20 Layer 4 17.8 0.623 0.1 16

21 Layer 4 17.8 0.606 0.125 7

22 Layer 4 17.8 0.616 0.15 4

23 Layer 4 17.8 0.606 0.175 3

24 Layer 4 17.8 0.614 0.2 2

Page 41 of 41



Documents

Draft - University of Toronto T-Space · Draft 1 Shear and dewatering behaviour of densified gold tailings in a laboratory simulation of multi-layer deposition Farzad Daliri1 Paul