See discussions, stats, and author profiles for this publication at: [Link]

net/publication/258837311

SEDIMENTATION AND CONSOLIDATION OF IN-LINE THICKENED FINE

TAILINGS

Conference Paper · December 2008

CITATIONS READS

9 690

5 authors, including:

Silawat Jeeravipoolvarn J. Don Scott

University of Alberta University of Alberta
20 PUBLICATIONS   174 CITATIONS    75 PUBLICATIONS   818 CITATIONS   

SEE PROFILE SEE PROFILE

Rick Chalaturnyk
University of Alberta
219 PUBLICATIONS   1,974 CITATIONS   

SEE PROFILE

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

IRC-FCMG View project

Analytical and Experimental Investiagtion of Shale Permeability View project

All content following this page was uploaded by Silawat Jeeravipoolvarn on 05 June 2014.

The user has requested enhancement of the downloaded file.

 SEDIMENTATION AND CONSOLIDATION OF IN-LINE THICKENED
FINE TAILINGS
S. Jeeravipoolvarn, J. D. Scott and R.J. Chalaturnyk
Department of Civil and Environmental Engineering – University of Alberta, Edmonton, AB, Canada
W. Shaw
WHS Engineering Ltd., Calgary, AB, Canada
N. Wang
Syncrude Canada Ltd., Fort McMurray, AB, Canada

ABSTRACT sedimentation ceases and the process of self-

weight compression fully governs. Field
An alternative technology to improve oil sands measurements show that the material does not
tailings settling behavior, an in-line thickening appear to be compressing further in the tailing
process (ILTT), was investigated. To achieve these ponds and the fine tailings at this point is called
new tailings, flocculants and coagulant are mixed mature fine tailings (MFT).
with cyclone overflow tailings in an in-line multi
stage fashion. Conceptually by binding fine The slow compression behavior of the MFT causes
particles at low solids content into flocs, the major problems for oil sand companies to manage
hydraulic conductivity is increased, tortuousity is containment ponds and to increase the amount of
decreased and the mass of the falling flocs is the recycle water for bitumen production. This led
increased. Field and laboratory studies were to considerable research on alternative solutions
carried out and preliminary findings indicate that for tailings deposition and dewatering. Major
the field in-line thickening process produces tailings developments include production of
tailings that undergo rapid hindered sedimentation composite tailings (CT), nonsegregating tailings
from fines void ratios of 67 to 5 followed by (NST), and the use of thickeners and centrifuges.
consolidation and segregation phenomena at lower
void ratios. Shearing during deposition is believed To search for an alternative to MFT formation, a
to cause the segregation. History matching of the new type of tailings called in-line thickened tailings
field pilot pond was performed to determine the (ILTT) is being developed. The process is aimed to
geotechnical consolidation properties of the field improve settling and strength behavior of thin fine
deposits and these properties were compared to tailings. This will result in a reduction of MFT
untreated cyclone overflow consolidation formation and storage by rapidly dewatering the
properties, which forms new MFT, and to cyclone overflow tailings stream which is a high
consolidation properties of laboratory remolded fines stream and the main source of new MFT. The
ILTT samples. Production of Composite Tailings schematic implementation of ILTT into the
using ILTT instead of MFT was examined and it is conventional oil sands operation is shown in Figure
postulated that ILTT CT would be far superior to 1. After extraction, the oil sands tailings are passed
MFT CT as ILTT’s permeability is more than 20 through cyclones which produce coarse and fine
times the permeability of MFT. tailing streams. Cyclone overflow tailings, the fines
tailings stream, is thickened via the in-line
thickening process. Rapid water release results in
INTRODUCTION hot water which can be returned to the extraction
plant reducing energy costs.
The oil sands mining operations in Northern
Alberta produce large volumes of tailings After deposition in a containment pond the ILTT
composed of sand, silt, clay and a small amount of reaches a solids content of about 30% in about 10
bitumen. Upon deposition, the tailings segregate days and a solids content up to 50% in a few
with the sand dropping out forming dykes and months. At this higher solids content, the ILTT can
beaches and about one-half of the fines and most be pumped out and mixed with the cyclone
of the bitumen flowing into the tailings pond as thin underflow to produce a superior Composite
fine tailings (TFT) which have about 8% solids and Tailings than that produced with cyclone underflow
90% fines. In the pond, sedimentation occurs and and MFT. This CT process is also shown in Figure
in about two years reaches approximately 30% 1.
solids. At this solids content the process of

1
 Make-up
ILTT Deposition Cells
Water Release
Water
In-Line Thickening
Process

Cyclone Over
Oil Sands Flow
Oil Sands 50-90% Fines
Tailings
Extraction Cyclones Flocculant Flocculant Flocculant Aggregate
Addition Distribution Adsorption Growth
Cyclone
Caustic Under Flow
5-10% Fines (a) Stage 1: First addition of flocculant
Bitumen Froth

CT Process
Thickened Fine
Tailings
Composite Tailings Disposal

Figure 1. Syncrude ILTT production

Coagulant Coagulant Coagulant Aggregate

To achieve these new tailings, chemical additives Addition Distribution Adsorption Growth
are used to increase size of clay flocs, settling rate, (b) Stage 2: Addition of coagulant
hydraulic conductivity and shear strength of the
cyclone overflow tailings. This new method uses a
multi-stage flocculation and coagulation mixing
process which is also known as tapered
flocculation (Yuan and Shaw 2007). The process is
illustrated in Figure 2. During the first stage (Figure
2(a)), a flocculant solution is added and distributed
into the cyclone overflow tailings. The long chain
molecules of the flocculant, which has various
charges along the chain, attach to coarse and fine
particles and brings these particles together to Flocculant Flocculant Flocculant Aggregate
form a floc. Some clays particles are not captured Addition Distribution Adsorption Growth
during this stage. In the second stage (Figure 2(b))
the coagulant solution is added into the tailings (c) Stage 3: Second addition of flocculant
from the first stage. The coagulant alters surface - Sand size particle - Clay particle
charges on the clay particles and results in - Flocculant
- Silt size particle - Coagulant
coagulation of the particles. The final stage (Figure
2(c)) is to add a flocculant solution into the product
Figure 2. The ILTT process (Modified from Yuan
of the second stage to bind the coagulated clay
and Shaw 2007)
with the flocs to form large aggregates.

By following the above process, the hydraulic

To be able to engineer this new type of tailings, a
conductivity of the mixtures is significantly
study of the sedimentation and consolidation
increased because the aggregates are far apart,
behavior of the ILTT is essential. This paper
tortuousity is decreased and the mass of the falling
presents some early findings of field investigations
aggregates is increased. This results in tailings
of ILTT pilot scale ponds and laboratory research
that can sediment quickly to the end of
on remolded ILTT samples. Field sedimentation
sedimentation. Moreover, as some consolidation
and compression mechanisms are shown and
occurs, a suitable shear strength is obtained which
history matching of the field behavior by a finite
allows a large variety of reclamation methods.
strain consolidation theory is given.

2
FIELD INVESTIGATIONS RCW + Coagulant
COF 1st Flocculent nd
2 Flocculent
ILTT pilot ponds were started in late 2005 to
demonstrate the pilot scale behavior of tailings
produced from the in-line thickening process.
Geotechnical field investigations including pore
Half pipe chute
pressure measurements, solids and fines content
measurements and vane shear strength
measurements were performed to monitor the
compression behavior of the deposits.

In-Line Thickened Tailings Pilot Ponds

Figure 3. Half-pipe chute deposition in the west
The Syncrude ILTT pilot project was composed of pilot pond
two pilot ponds; the west pond and the east pond.
Although the tailings stream is produced from
extraction of ore bodies from many mining
locations; in the pilot operations the west pond
ILTT was mainly from marine oil sands ore and the
east pond ILTT was mainly from estuarine oil sand
ore. Marine ore contains a higher percentage of
highly active clay minerals which detrimentally
affect extraction and the resulting tailings contain
more active clay minerals and bitumen.
Flocculation and coagulation of these tailings may
Aggregate

also be less efficient because of these properties.

Size

Two distinct discharging methods were used for

each pond. The ILTT was discharged down a half- Shearing Effort
pipe chute (Figure 3) into the west pond (located at
the west side of the ILTT mixing plant). This type of
discharge results in high energy dissipation which Figure 4. Aggregate rupture (Modified from Shaw
causes shearing of the ILTT aggregates. The and Wang 2005)
relationship between shearing effort and aggregate
size is postulated in Figure 4. This shearing of the
RCW+ Coagulant
aggregates results in a smaller aggregate size COF 1st Flocculent nd
2 Flocculent
which is believed to have detrimental effects on the
ILTT settling, compression and segregation
behavior. To avoid aggregate breakage as much
as possible, another type of ILTT deposition, a
sub-TT discharge was used for the east pond Spill box
(Figure 5). This discharging method utilized an
inverted spill box which was located lower than the
mixing plant. The third ILTT stage (Figure 2(c))
was also moved to this box. This type of discharge
provides a lower energy dissipation technique.

The plan view of both ponds is illustrated in Figure Figure 5. Sub-TT deposition in the east pilot pond
6. The pond is equally divided into three sections
for sampling and pore pressure monitoring. Each
section contains a center station for monitoring and
there are also two extra stations at the inlet and
outlet of the pilot ponds.

3
 Flow direction An incongruity in the interface height
L measurements was observed at 70 days. This was
L/3 L/3 L/3
during the winter period where accurate
measurement was difficult. The last measurement
W/2
was performed at 235 days. At this time, the
Station 1 Station 2 Station 3 surface was desiccated by a winter freeze-thaw
W process and natural drying. The average interface
W/2
height was 1.1 m and an average crust thickness
was about 0.2 m.
Inlet station Outlet station

- Sampling location Solids and Fines Content Measurements

- Pore pressure location
Figure 6. Plan view of an ILTT pilot pond Sampling for the solids and fines contents were
done by using a thin walled sampler provided by
GeoForte Services Ltd. This sampler was
Field Investigation Results and Discussion designed to retrieve a core sample from a tailings
pond. The core sample was extruded out of the
This discussion is divided into four sections which sampler and divided into several sections for solids
are tailings-water interface measurements, solids and fines content profile determinations.
and fines content measurements, pore water
pressure measurements and vane shear strength The measurements of solids content profiles at
measurements. The results will be focused on the different elapsed times at the three sampling
east pilot pond as it better represents the ILTT stations from the east pilot pond are shown in
tailings behavior from a lower energy deposition Figures 8(a), 8 (b) and 8(c). The fines content
technique. profiles at the three stations are given in Figures
8(d), 8(e) and 8(f).
Tailings-Water Interface Measurements
The solids content profiles in Figure 8 indicate that
In November 2005, the ILTT east pilot pond was
there is a rapid increase in solids content from an
filled continuously for approximately 10 days at a
initial solids content of 3.7% to an average solids
solids content of about 3.7% (void ratio of 65.6)
content of 35.8% within 26 days. After filling, the
and a fines content of 89%. The average solids
pond was allowed to compress under self-weight
content at the end of the 10 day filling period was stress. The surface of the pond has gone through
32.6% (void ratio of 5.2). Then it was left to settle
freezing, thawing and drying during the quiescent
under its own weight for approximately 230 days.
period which developed a crust approximately 0.2
The tailings-water interface height of ILTT in this
m thick. Below the crust the average solids content
pond is shown in Figure 7.
was about 50% (void ratio of 2.5).

3.0 The fines content profiles are shown in Figure 8.

East Pond Station 1
The average fines content of the ILTT was found to
2.5 East Pond Station 2 be 89% and the profiles indicate that the ILTT is a
East Pond Station 3 nonsegregating mix showing minimal changes in
Interface Height (m)

2.0 East Pond Average the fines content profiles during the observation
period of 235 days. This finding is also confirmed
1.5 by the slurry properties diagram in Figure 9. The
diagram reveals that the ILTT samples obtained
1.0 from the east pilot pond show small segregation. In
contrast to this finding, samples obtained from the
0.5 west pilot pond indicate segregation behavior by
showing a reduction of fines content at a constant
0.0 fine/(fines+water) ratio. The difference between the
0 50 100 150 200 250 two pilot ponds was the discharge method and the
Elapsed Time (Days)
tailings ore type. The higher energy deposition
method in the west pond is believed to be the main
Figure 7. Interface height vs. time of the east pilot
reason for aggregate breakage and the
pond
segregation behavior of the ILTT in this pond.

4
 2.0 2.0 2.0
26 Days 26 Days 26 Days
62 Days 62 Days 62 Days
1.6 124 Days 1.6 124 Days 1.6 124 Days
235 Days 235 Days 235 Days
1.2 1.2 1.2
Height (m)

Height (m)
Height (m)
0.8 0.8 0.8

0.4 0.4 0.4

(a ) Station 1 (b ) Station 2 (c ) Station 3

0.0 0.0 0.0
0 25 50 75 100 0 25 50 75 100 0 25 50 75 100
Solids Content (%) Solids Content (%) Solids Content (%)

2.0 2.0 2.0

26 Days 26 Days 26 Days
62 Days 62 Days 62 Days
1.6 124 Days
1.6 1.6
124 Days 124 Days
235 Days 235 Days 235 Days
1.2 1.2 1.2
Height (m)

Height (m)

Height (m)
0.8 0.8 0.8

0.4 0.4 0.4

(d ) Station 1 (e ) Station 2 (f ) Station 3

0.0 0.0 0.0
0 25 50 75 100 0 25 50 75 100 0 25 50 75 100
Fines Content (%) Fines Content (%) Fines Content (%)

Figure 8. Solids and fines content profiles of the east pilot pond

Water ILTT West Pilot Pond 70

ILTT East Pilot Pond 3. In-line Thickened Tailings
10 10 COF
20 20
60
30 30 2. In-Line Thickening Process
Solids
Content (%) 40 40 Fines/( Fines + 50
Fines Void Ratio

50 50 Water), (%)
4. Rapid Sedimentation
60 60 40
70 70
80 80
30 1. Original cyclone overflow
Fines Matrix
90 90
Sand Matrix

Sand 10 20 30 40 50 60 70 80 90 Fine
20
Fines Content (%)
5. Consolidation and Segregation
Figure 9. Slurry properties diagram 6. Segregation
10

To summarize the processes that occured in the 0

east ILTT pilot project, a mechanistic diagram of 60 70 80 90 100
the ILTT process is illustrated in Figure 10. First
Fines Content (%)
the cyclone overflow is mixed with recycle water
and chemical additives via the in-line thickening Figure 10. The ILTT process
process. This produces an ILTT material at a solids

5
content of about 4% (void ratio of 66 and fines void the bottom was equal to zero. A possible
ratio of 67). The ILTT then undergoes rapid explanation for Station 3 was that the underlying
sedimentation releasing more than 90% of its sand foundation was fully drained in this particular
original water volume as it reaches a fines void location. This bottom pore pressure would create a
ratio of about 5. At this point both consolidation draw down condition on the tailings in the pond.
and segregation phenomena can occur The pore pressures being approximately equal to
simultaneously in the material causing both a the total stress above the pond bottom indicate that
reduction in the fines void ratio and the fines any downward drainage must have consolidated a
content. Generally the ILTT continues to undergo layer of tailings at the bottom which somewhat
consolidation to a lower void ratio while some sealed the pond bottom and prevented further
whose structure was broken during the deposition significant downward drainage.
process showed a small amount of segregation.
Vane Shear Strength Measurements
Pore Water Pressure Measurements
The vane shear strength of the ILTT materials was
Pore water pressure measurements in a pond are measured by pushing in a field vane apparatus
essential to monitor dissipation of excess pore directly into the deposits. The vane was pushed
pressure which is an indication of the consolidation down 25 cm at a time to obtain shear strength with
progress in a tailings deposit. The pore pressure depth. The measured vane shear strength values
measurements were performed by using Roctest, of the east pilot pond at the last field investigation
model PWP, push-in piezometers with a vibrating program at 235 days are shown in Figure 12.
wire readout unit, model MB-6TL. In the pilot
ponds, four piezometers were pushed in the pond The vane shear strength of the east pond varied
simultaneously to different elevations. The from 1.6 kPa (solids content of 44% and void ratio
piezometers were kept in place until the pore water of 3.2) under the crust to 5.2 kPa (solids content of
pressure measurements became stable when the 52% and void ratio of 2.3) near the bottom of the
pressure values were taken. pond. The dry crust of the east pond had a surface
vane shear strength of about 97 kPa at a solids
The last pore pressure measurements at 235 days content of about 72% (void ratio of 1.0). The vane
in the east pond are shown in Figures 11(a), 11(b) shear strength profiles of the east pond indicate
and 11(c) for Station 1, 2 and 3 respectively. The good shear strength consistency between all
total stresses shown were calculated from the stations. This is because the settling processes are
solids content profiles from Figure 8. The pore similar at all stations and the ILTT materials are
pressure measurements at Stations 1 and 2 quite homogeneous throughout the pond.
indicate that double drainage conditions prevailed
in these locations as the pore pressure was equal
to the hydrostatic pressure at the bottom of the
stations. While at Station 3 the pore pressure at

1.4 1.4 1.4

Total Stress Total Stress Total Stress
Pore Pressure Pore Pressure 1.2 Pore Pressure
1.2 1.2
Hydrostatic Pressure Hydrostatic Pressure hydrostatic Pressure
1.0 1.0 1.0
(a ) Station 1 (b ) Station 2 (c ) Station 3
Height (m)
Height (m)

Height (m)

0.8 0.8 0.8

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0.0 0.0 0.0

0 5 10 15 20 0 5 10 15 20 0 5 10 15 20
Stress and Pressure (kPa) Stress and Pressure (kPa) Stress and Pressure (kPa)

Figure 11. Total stress and pore pressure profiles of ILTT east pond at 235 days

6
 tailings as the ILTT flocculated aggregates are
1.2 completely dispersed.
East Pond Station 1
1.0 Figure 14 shows a plasticity chart containing data
East Pond Station 2 from the east pond ILTT, west pond ILTT, cyclone
0.8 overflow and 1982 MFT. The east pond ILTT has
East Pond Station 3
an average liquid limit of 50 and an average
Height (m)

0.6 plasticity index of 29. These values are in the same

range as the cyclone overflow and 1982 MFT. The
0.4
west pond, however, has an average liquid limit of
71 and an average plasticity index of 47.
0.2
100
0.0
0 2 4 6 8 10
80
Vane Shear Strength (kPa)
Figure 12. Vane shear strength profiles of ILTT 60
east pond at 235 days

% Finer
40

LABORATORY EXPERIMENTS Cyclone Over Flow

20
ILTT West Pond
ILTT samples obtained from the pilot scale test
pond were delivered to the University of Alberta 0
and tested in the summer of 2006. The objectives 0.0001 0.001 0.01 0.1 1
of the laboratory test program were to obtain Particle Size (mm)
particle size distributions, index properties and Figure 13. Particle size distribution
consolidation characteristics of the ILTT materials.

As well as testing the ILTT materials; samples of 60

CL CH
cyclone overflow, taken at the same time as the 1982 Syncrude MFT

ILTT was made and deposited, were similarly 50 2005 ILTT West Pilot Pond
tested. The difference in material properties should 2005 ILTT East Pilot Pond
show the influence of the flocculation and
Plasticity Index, I p

40
2008 Syncrude Cyclone
coagulation processes. As well, the untreated Over Flow
cyclone overflow is the same as new MFT and 30
should indicate the differences between ILTT and
MFT. Samples of MFT from the MFT tailings pond 20
were not tested as ageing of MFT has a significant
effect on its thixotropic properties which affect its 10
consolidation properties (Jeeravipoolvarn et al. CL-ML ML-OL MH-OH
2008 (a) and (b)). Comparing ILTT and MFT of the 0
same age was preferred. 0 20 40 60 80 100
Liquid Limit, w L
Material Index Properties Figure 14. Plasticity chart of various oil sands
tailings
Figure 13 provides particle size distributions of
ILTT and cyclone overflow. The cyclone overflow
tailings consist of 8% sand, 92% fines and 52% The relatively higher values of wL and Ip of the west
clay size materials similar to that of ILTT. The pond ILTT could indicate two possibilities. One is a
particle size distributions were obtained by a variation in bitumen content and another is the
standard hydrometer-sieve test (ASTM D422-63). variations in the ore type of cyclone overflow used
The test is therefore a dispersed test and results in for the two ponds.
identical particle size distributions of the two

7
The segregation of the material in the west pond Table 1 Tailings properties
provided samples with higher fines content which Solids Water Fines
in turn also indicates a higher bitumen content. Tailings Content Content ei Gs Content
This will generally increase both wL and Ip of the (%) (%) (%)
tailings. The tailings with more bitumen (usually COF 30.9 223.6 5.66 2.53 93.9
finer tailings) tend to have a higher liquid limit
(Scott et al. 1985). Estuarine and marine ores may East pond
35.5 181.7 4.61 2.52 87.4
pose differences in both the extraction and the in- ILTT
line thickening process efficiency. Future research West
is essential to investigate the ILTT production from pond 29.7 236.7 5.80 2.45 95.3
the two different ores. ILTT

Large Strain Consolidation Tests with

Hydraulic Conductivity and Vane Shear Figures 15, 16 and 17 show comparisons of the
Strength Measurements compressibility, hydraulic conductivity and vane
shear strength of all the tested materials
The large strain consolidation test performed at the respectively. From Figure 15, the west pond ILTT
University of Alberta is a multi-step loading test. has essentially the same compressibility as the
The test apparatus and procedures have been cyclone overflow. In Figure 16, the hydraulic
slightly modified for the in-line thickened tailings to conductivity of the west pond ILTT is lower than
allow vane shear tests to be performed at different the cyclone overflow at the same void ratios while
void ratios. the east pond ILTT has the highest hydraulic
conductivity. As fines void ratio is used in Figure 16
The consolidation cell is 140 mm inside diameter to eliminate an influence of different fines contents,
and can accommodate samples up to 200 mm the higher hydraulic conductivity of the east pond
high. The wall friction is minimized by choosing the ILTT than the cyclone overflow indicates the
appropriate initial height of the samples so that the influence of the in-line thickening process on the
sample at the high stresses (10 kPa and over) has tailings permeability.
a diameter to height ratio of approximately 2.5 to 3.
Also the height of the sample should be kept as As mentioned in the introduction to the laboratory
high as possible to minimize errors in measuring experiments, cyclone overflow materials were
the thickness of the sample and in hydraulic tested as they, when deposited, will settle and form
conductivity tests. The ILTT samples are tested at new MFT. This new MFT will have the same
effective stress ranges from about 0.5 kPa to 300 mineralogy and initial water chemistry as the ILTT
kPa to cover the stress range that the ILTT might as both the cyclone overflow and ILTT samples
experience in the field. were obtained at the same time. The comparisons
of ILTT and cyclone overflow geotechnical
The hydraulic conductivity is measured at the end properties will then not be influenced by different
of consolidation for each loading step. The upward initial material properties. Figure 16 also shows the
gradient in the test is controlled to be as small as hydraulic conductivity of typical old MFT. As found
possible so the seepage force will not exceed the in other testing programs, the hydraulic
applied stress. The shear strength is then conductivity of old MFT is considerably smaller
measured after the hydraulic conductivity test by a than that of new MFT (new cyclone overflow) and
laboratory vane apparatus before the next load is ILTT. This observation is important when
applied. considering the properties of composite tailings
made with the high solids, high permeability ILTT.
Three materials were tested; cyclone overflow,
east pond ILTT and west pond ILTT. Sample For vane shear strength, the west pond ILTT has
preparations required the samples to be mixed higher strength compared to the cyclone overflow
thoroughly therefore all the samples are remolded at effective stresses above 10 kPa (Figure 17). At
samples. The properties of each tailings are listed the higher effective stresses, there appears to be
in Table 1. The lower value for specific gravity (Gs) no difference in the vane shear strength between
of the west pond ILTT results from the higher the two tailings. It is believed that the physical
bitumen content. forces dominate the behavior of the soils at
effective stress above 10 kPa and the chemical
effects are minor. As mentioned earlier, the ILTT

8
pilot samples have gone through sample
preparation processes. The structure of these 6
samples may not be representative of the original e = 3.298 σ '-0.204
5
field material. Cyclone Over Flow

Fines Void Ratio, ef

4 ILTT East Pond
For the east pond ILTT, the laboratory vane shear ILTT West Pond
strength is much smaller compared to the field 3
vane strengths at effective stresses lower than 10 e = 3.179 σ '-0.196
2
kPa. This could be caused by structural changes
from the laboratory sample preparation. At higher 1
effective stress than 10 kPa, shear strengths of e = 2.676 σ '-0.201
both field and laboratory east pond ILTT are 0
similar. 0.01 0.1 1 10 100 1000
Effective Stress (kPa)
In natural sedimentary clay deposits the undrained
shear strength has been found to increase with Figure 15. Compressibility of COF, east pond ILTT
depth and is proportional to the increase in and west pond ILTT
effective stress with depth (Holtz and Kovacs
1981). A relationship between plasticity index and
a ratio between undrained shear strength and
10-4
effective stress, τu/σ′vo, for normally consolidated k = 1.072×10-8 e 4.840
clays was established by Bjerrum and Simons
10-5
Hydraulic Conductivity (cm/s)
(1960). This relationship provides a τu/σ′vo ratio of k = 4.070×10-9 e 4.232
about 0.25 for oil sands fine tailings with plasticity k = 2.575×10-9 e 4.212
index of about 30. However, the large strain 10-6
consolidation test results (Figures 15 and 17) show
that this ratio is not constant at all fines void ratio 10-7
but increasing linearly with fines void ratio (Figure Cyclone Over Flow
18). The large values at high void ratios are -8 ILTT East Pond
indicative of the strong bonds in the flocs. As the 10
ILTT West Pond
flocs are compressed during consolidation the flocs Typical MFT
break and lose strength. 10-9
0 1 2 3 4 5
The research hypothesis of comparing ILTT and Fines Void Ratio, ef
cyclone overflow properties was that the difference Figure 16. Hydraulic conductivity of COF, east
in the behavior of both materials would be directly pond ILTT and west pond ILTT
related to the effect of the flocculants and
coagulant added in the in-line thickening process.
The function of the additives is to bind soil particles
together to form large aggregates of soil particles 6
that will settle quickly during deposition at low 5
Fines Void Rati o, ef

solids content. The hypothesis is that as the

additives flocculate the soil particles together, the 4
shear strength of the soil will increase as breaking
the structure requires additional energy. The soil 3
structure therefore becomes stronger. Compared 2
to the parent materials, the treated tailings would Cyclone Over Flow
not compress as much when subjected to the 1 ILTT East Pond
ILTT West Pond
same applied stress. This appears to be Field Measurement ILTT East Pond
insignificant in our preliminary test results. The 0
hydraulic conductivity should also be affected by 0.01 0.1 1 10 100
the process because the additives change the Vane Shear Strength (kPa)
structure of the soil reducing the tortuous channels
of the soil resulting in greater hydraulic Figure 17. Vane shear strength of COF, east pond
conductivity. ILTT and west pond ILTT and field ILTT

9
 1.6 amount of solids filled in a pond and the interface
height is also essentially linear during the filling
period. This is true when the filling rate is constant
1.2 τ u / σ ' vo = 0.239e f + 0.012 and the one dimensional pond condition is
2 satisfied. These conditions are justified for the
R = 0.660
current simulation and the interface during the
τu/σ'vo

0.8 filling period was therefore obtained from this

approach. Boundary conditions for the simulation
are depicted by Figure 11. The pilot pond clearly
0.4 West Pond ILTT
shows a double drainage condition. Therefore a
Cyclone Overflow
Dirichlet boundary condition of u = 0 is assigned for
East Pond ILTT
both top and bottom boundaries.
0.0
0 1 2 3 4 5 The history match was done by searching
Fines Void Ratio, ef
compressibility and hydraulic conductivity functions
Figure 18. Laboratory determined shear strength- based on the power law relationships given in
effective stress relationship Figures 15 and 16. The mathematical formula of
the relationships remain the same, however, the
compressibility function was assigned to have a bi-
COMPARISON OF THE FIELD AND power form as at effective stress lower than 10
LABORATORY RESULTS kPa, large amounts of oil sands tailings
compressibility data suggest little to no variations
(Jeeravipoolvarn 2005). Therefore the form of the
In order to compare field performance and
compressibility is chosen as Equation 2 while a
laboratory consolidation characteristics of the ILTT
hydraulic conductivity-void ratio function remains a
from the east pilot pond, a finite strain
conventional power law functions expressed as
consolidation theory (Gibson et al. 1967) was
Equation 3.
utilized to history match the field interface height
measurements. A one dimensional condition is
assumed due to the much larger width and length
of the pond compared to the height of the slurry. A ⎧ A σ ′ B1 for σ ′ < 10kPa
e=⎨ 1 B [2]
one dimensional finite strain consolidation
⎩ A2σ ′ for σ ′ ≥ 10kPa
2

governing equation in terms of excess pore

pressure is given by Somogyi (1980) as in
Equation 1. k = Ce D [3]

It is noted that the A1 and B1 are laboratory

∂ ⎡ k (e) ⎤ ∂u k ( e) ∂ 2 u determined parameters while the A2, B2, C and D
⎢ ⎥ +
∂z ⎣ γ w (1 + e) ⎦ ∂z γ w (1 + e) ∂z 2 [1] are history matching parameters. An initial height
of the pond was back calculated from the final
de ∂u de ⎡ d (ΔZ )⎤
+ − ⎢ (Gs − 1).γ w =0 solids content profiles in Figure 8 and the initial
dσ ′ ∂t dσ ′ ⎣ dt ⎥⎦ solids content of 3.7%. The initial height was found
to be 20.4 m. The simulation period was allowed to
go up only to 70 days. This period was chosen due
Where u is excess pore pressure, z is a material to the fact that the pilot pond experienced freezing,
coordinate, k is hydraulic conductivity, σ' is effective thawing and drying after this period.
stress, e is void ratio, γw is unit weight of water, Gs
is specific gravity and t is time. A consolidation computer program was coded in
Visual BASIC. A fully implicit finite difference
To demonstrate an effect of the in-line thickening scheme with upwind in a convection term is used
process by increasing solids content with time to solve the governing equation. The scheme was
especially during a filling period, a quiescent selected due to its stability and numerical
condition was assigned in the model. For a filling parameters n = 40 and Δt = 10-4 are used in the
period, according to Carrier et al. (1983), a model (n and Δt denote a total node number and
prediction of an interface settlement is essentially incremental time respectively).
independent on filling history but follows the total

10
A result of an interface height simulation with time The good agreement between the history match
is shown with the average field measurements in simulation and the field data shown above was
Figure 19 while Figure 20 shows the increase in performed by the incorporation of void ratio-
solids content with time due to the in-line effective stress and hydraulic conductivity-void
thickening process in the east pilot pond. ratio relationships shown in Figures 21 and 22
respectively. The laboratory experimental data
from the large strain consolidation tests in Section
3.0 3.2 is also included for relationship comparisons.
Filling Period
2.5 For the compressibility, it requires compressibility
Quiescent Period
Interface Height (m)

of the field ILTT to be more compressible by

2.0
having a higher value of compression index. The
1.5
history matching compressibility function also
indicates that it would take a higher effective stress
1.0 to compress the ILTT material to the same void
Pilot Pond Average Interface Height Values
ratios in the pond compared to the laboratory
0.5 remolded ILTT sample. The field ILTT was more
Finite Strain Consolidation History Matching
difficult to compress than the remolded laboratory
0.0 samples because of the fairly strong large flocs in
0 10 20 30 40 50 60 70 the field. Mixing the field ILTT in the laboratory
Elapsed Time (Days)
would break these flocs making the material easier
to compress.
Figure 19. History matching of interface height vs.
time The hydraulic conductivity-void ratio relationship of
low solids or high void ratio materials is the most
important property governing the rate and
50
magnitude of settlement and consolidation
Filling Period (Suthaker and Scott 1994). For this reason it is
40 important to examine and compare the hydraulic
Solids Content (%)

conductivities in Figure 22. At 50% solids content

30 Quiescent Period (void ratio of 2.5) the field hydraulic conductivity is
2.5 times larger than the laboratory hydraulic
20 conductivity. This is a result of the breakdown of
Initial Solids Content of 3.7% the floc structure in the laboratory samples as they
were remolded. Of even more importance, at 50%
10 Pilot Pond Average Interface Height Values
solids the field ILTT is 12 times more permeable
Finite Strain Consolidation History Matching than the cyclone overflow (or new MFT) and even
0 larger than that of typical old MFT (Figure 16). So
0 10 20 30 40 50 60 70 even after compressing to 50% solids, the field
Elapsed Time (Days) ILTT still processes sufficient floc structure to have
a relatively large permeability.
Figure 20. History matching of increase in solids
content with time
At 30% solids (void ratio of about 5.9), a typical
MFT solids content, the field hydraulic conductivity
is 22 times more permeable than that of the
In Figure 19, the ILTT material was filled into a
cyclone overflow (new MFT). This great difference
pond in a period of 10 days. During this period,
in permeability at the same solids content is
Figure 20 provides that ILTT has changed its solids
caused by the difference in structure of the
content from the initial value of 3.7% to 33.2%
flocculated ILTT vs. the dispersed MFT.
(void ratios of 65.6 and 5.1 respectively). After it
released more than 90% of its water during this
Future research on hydraulic conductivity and
period, the ILTT was allowed to undergo self-
compressibility behavior of undisturbed in-line
weight compression up to 70 days when the
thickened tailings is required to further investigate
simulation period was over. Further increase in
these properties.
solids content slows dramatically as the hydraulic
conductivity reduces and excess pore pressures
dissipate slowly.

11
A comparison of vane shear strengths from the
6
field history matching, field measurements and
laboratory measurements on ILTT and cyclone 5
overflow is given in Figure 23. History matching
vane shear strengths were obtained from the 4

Void Ratio
relationship between τu/σ′vo and fines void ratio in
Figure 18 with the compressibility data in Figure 3
21. The results show that the predicted undrained
2
shear strengths are closer to those data measured
Cyclone Over Flow
in the field than to the laboratory determined shear 1 Laboratory ILTT
Field Measurement ILTT East Pond
strengths. At higher void ratios above 2.5 the field Field ILTT (History Matching)
ILTT has a higher shear strength than the cyclone 0
overflow. As discussed above, this is a result of the 0.01 0.1 1 10 100
strong bonds in the ILTT flocs. Future research on Vane Shear Strength (kPa)
a relationship between undrained shear strength
and effective stress of the undisturbed ILTT is Figure 23. Comparison of the field history
required. matching, field measurements and laboratory
shear strengths

6
COMPOSITE TAILINGS USING ILTT
5

4 Composite tailings, CT, is presently produced by

e = 4.000 σ '-0.375
Void Ratio

mixing and depositing a blend of cyclone underflow

3 and MFT with a coagulant, generally gypsum,
e = 2.676 σ '-0.201 added to make the mix nonsegregating. The MFT
2 is pumped from the existing tailings ponds where it
Cyclone Over Flow
has accumulated over the years. This procedure is
1
Laboratory ILTT consuming some of the current inventories of MFT
Field ILTT (History Matching)
0 and it releases water fairly rapidly for make-up
water for the extraction plant (Matthew et al. 2002).
0.01 0.1 1 10 100 1000
Dewatering rates, however, are lower than
Effective Stress (kPa) expected and careful engineering of the CT
process is required to prevent segregation during
Figure 21. Comparison of the field history matching
deposition. Another drawback is that the amount of
and laboratory compressibilities
MFT which can be added to the cyclone underflow
is limited as too much fines in the CT reduces its
permeability and the dewatering rate will be too
10
-4
slow. Generally the sand-fines ratio, SFR, of CT is
k = 2.315×10 e-8 5.000 only 4 or 5, that is, the fines content is from 17% to
-5 20%.
Hydraulic Conductivity (cm/s)

10

As the oil sands tailings on average contains about

-6
10 17% fines and the cyclone overflow which contains
most of the fines is deposited in a tailings pond and
10
-7
forms more MFT, the CT process can not keep up
with the formation of more MFT and MFT volumes
-8 Cyclone Over Flow continue to increase. Another problem with the
10
Laboratory ILTT present CT process is that the dosages of gypsum
Field ILTT (History Matching)
to prevent segregation are quite large resulting in
-9 2+ 2-
10 the build up of Ca and SO4 ions in the recycle
0 1 2 3 4 5 water which in time will negatively affect bitumen
Void Ratio
extraction.
Figure 22. Comparison of the field history matching
and laboratory hydraulic conductivity Figure 24 shows the present CT process. All mixes
of MFT and cyclone underflow, CU, will lie on the

12
straight line joining the MFT and cyclone underflow water and produce thickened tailings without the
materials. The segregation boundary for untreated aid of a thickener. Field pilot programs were
tailings blends is also shown. All tailings above this conducted and were successful.
segregation boundary will segregate so all
untreated mixes of MFT and CU will segregate. Two ILTT pilot ponds were filled. The major
The segregation boundary for gypsum treated difference between the ponds was the discharge
tailings is also shown. All tailings below the method. The west pond had a half-pipe chute
gypsum segregation boundary will be discharge while the east pond had a sub-TT
nonsegregating so the CT mixes of MFT and CU discharge. The latter discharging method was a
will become nonsegregating with the addition of a lower energy deposition technique which did not
coagulant such as gypsum. Figure 24 also shows cause segregation of the ILTT.
all possible mixes of ILTT and CU which lie on the
line that joins these two materials. Possible CT In the east pilot pond, it was found that the tailings
mixes of ILTT and CU lie below the untreated compressed from an initial solids content of 3.7%
segregation boundary which indicates that these to 32.6% (void ratios of 65.6 and 5.2 respectively)
CT mixes would require no or little coagulant in 10 days and to about 50% (void ratio of 2.5) at a
addition with a resulting increase in the quality of depth of 1 m in 4 months. The fines content
the recycle water for bitumen extraction. profiles were consistent with depth and have an
average value of 89%. The pore pressure
measurements in this pond showed some
Water consolidation at the bottom at all stations as the
Segregation Boundary for Mixtures of MFT and
Gypsum Treated Tailings Mixture Cyclone Underflow pond had a permeable foundation. It appeared that
10 10
Segregation Boundary for Mature Fine Tailings (MFT)
a layer of tailings consolidated at the bottom and
20 20
Untreated Tailings
Mixtures of ILTT and
sealed it preventing further significant downward
30 30
Cyclone Underflow drainage.
40 40
Solids Fines/( Fines +
50 50
Content (%) Water), (%) Evident segregation was found in the west pilot
60 60
Cyclone
pond. A hypothesis on the cause of segregation
70 70
Underflow ILTT was that the high discharging energy to this pond
80 80
Fines Matrix broke floc structure allowing the sand to settle out.
90 90
Sand Matrix
Further research is required to confirm the
Sand 10 20 30 40 50 60 70 80 90 Fine influence of shearing on ILTT segregation
Fines Content (%) behavior.
Figure 24. Tailings properties diagram
Material index properties show that the cyclone
overflow and ILTT similarly consist of 8% sand,
The greatest advantage of CT made from ILTT and 92% fines and 52% clay size materials. The east
CU is the large hydraulic conductivity of the ILTT pond ILTT has an average liquid limit of 50 and an
fines. As shown in Figures 16 and 22, at solids average plasticity index of 29. These values are in
content of 30% the field hydraulic conductivity of the same range as the cyclone overflow and typical
ILTT is more than 20 times the hydraulic MFT.
conductivity of new MFT (cyclone overflow) and
even greater than that of old MFT. This means the The field pilot ponds were history matched by
ILTT CT will release water and consolidate more utilizing a finite strain consolidation theory and by
than 20 times faster than MFT CT allowing more searching compressibility and hydraulic
rapid CT disposal. The greater ILTT hydraulic conductivity functions based on power law
conductivity will also allow a larger amount of fines relationships. Good agreement was obtained with
in the ILTT CT than the 20% maximum in the MFT matching the field interface height with time and
CT. This would assist in the reduction of fine tails the increase in solids content with time. Field
volume stored in tailings ponds. compressibility and hydraulic conductivity functions
were obtained.

OBSERVATIONS AND CONCLUSIONS The field ILTT was more difficult to compress than
the remolded laboratory samples because of the
An in-line thickening process was developed to fairly strong large flocs in the field.
provide a method that can rapidly recover hot

13
 The field ILTT was more permeable than the REFERENCES
laboratory ILTT samples at the same solids
contents because of the breakdown of the floc Bjerrum, L. and Simons, N.E., 1960. Comparison
structure in the laboratory samples as they were of Shear Strength Characteristics of Normally
remolded. Consolidated Clays, 1st PSC, ASCE, 711-726.
Carrier, W.D., III, Bromwell, L.G., and Somogyi, F.,
At 30% solids content the field hydraulic 1983. Design Capacity of Slurried Mineral
conductivity was 22 times more permeable than Waste Ponds. Journal of Geotechnical
that of the new MFT (cyclone overflow). This great Engineering, 109(5): 699-716.
difference in permeability was caused by the Gibson, R.E., England, G.L. and Hussey M.J.L.
difference in structure of the flocculated ILTT vs. 1967. The Theory of One-dimensional
the dispersed MFT. Consolidation of Saturated Clays. I. Finite Non-
linear Consolidation of Thin Homogeneous
Possible Composite Tailings mixes of ILTT and Layers. Gėotechnique, 17(3): 261–273.
Cyclone Underflow lie below the untreated Holtz, R.D. and Kovacs, W.D., 1981. An
segregation boundary which indicates that these Introduction to Geotechnical Engineering.
CT mixes would require no or little coagulant Prentice-Hall, Englewood Cliffs, NJ, 733 p.
addition with a resulting decrease in cost to make Jeeravipoolvarn, S. 2005. Compression behavior
CT and an increase in the quality of the recycle of thixotropic oil sands tailings. MSc thesis,
water. University of Alberta, Edmonton, AB, 223 p.
Jeeravipoolvarn, S., Chalaturnyk, R.J. and Scott,
At solids contents of 30% the field hydraulic J.D., 2008 (a). Consolidation Modeling of Oil
conductivity of ILTT is more than 20 times the Sands Fine Tailings: History Matching,
hydraulic conductivity of new MFT and even Proceedings of 61st Canadian Geotechnical
greater than that of old MFT. The ILTT CT will Conference, Edmonton, AB, September 22-24,
release water and consolidate more than 20 times 190-197.
faster than MFT CT. It will also allow a larger Jeeravipoolvarn, S., Chalaturnyk, R.J. and Scott,
amount of fines in the ILTT CT than is presently J.D., 2008 (b). Three ten meter standpipes:
used in the MFT CT. experimental and prediction, submitted to
Canadian Geotechnical Journal.
Shearing of the field ILTT by thorough mixing in the Matthews, J.G., Shaw, W.H., MacKinnon, M.D.,
laboratory changed its compressibility and and Cuddy, R.G., 2002. Development of
permeability properties as shearing tends to composite tailings technology at Syncrude.
breakdown the floc structure. The ILTT became International Journal of Surface Mining,
more compressible under the same effective stress Reclamation, and Environment, 16(1): 24-39.
which had the advantage that it compressed more Scott, J. D., and Dusseault, M.B. and Carrier III,
and released more water. A major disadvantage, W.D., 1985, Behaviour of the
however, was that its permeability decreased. A clay/bitumen/water sludge system from oil
challenge exists, when making CT with ILTT, to sands extraction plants, Journal of Applied
develop procedures to pump ILTT from Clay Science, 1: 207-218.
depositional ponds and mix it with cyclone Yuan, S. and Shaw, W. 2007. Novel Processes for
underflow without a significant breakdown of the Treatment of Syncrude Fine Transition and
floc structure. Marine Ore Tailings. Canadian Metallurgical
Quarterly, 46(3): 265-272.
Shaw, W. and Wang, N., 2006. In-Line Thickened
ACKNOWLEDGEMENTS Tailings Pilot Program. CIM Fort McMurray, Oil
Sands Discovery Center Auditorium, Fort
The authors are grateful for the financial support McMurray, Alberta, February 21, 2006.
from the University of Alberta and Syncrude Somogyi, F. 1980. Large Strain Consolidation of
Canada Ltd. The authors also appreciate the Fine Grained Slurries. Presented at the
technical support from Gerry Cyre, Geoforte Canadian Society for Civil Engineering,
Services Ltd., Edmonton and Steve Gamble, Winnipeg, Manitoba, Canada, May 29-30.
Department of Civil and Environmental Suthaker, N.N., and Scott, J.D. 1994. Large Strain
Engineering, University of Alberta. Consolidation of Oil Sand Fine Tails in a Wet
Landscape. 47th Canadian Geotechnical
Conference, Halifax, Nova Scotia, 514-523.

14

View publication stats