max
t/m
3
1.771.79 1.771.79
min
t/m
3
1.281.30 1.281.31
dStandard Proctor
t/m
3
1.651.69 1.601.65
Permeability cm/s 1x10
3
to
1x10
4
1x10
3
to
1x10
4
Fines(200mesh % 1418 1216
Classification USC
S
SM SM
It should be noticed that for a density of 1.60 t/m
3
the permeability of the sand varies from about 1x10
4
for 25% of
fines to about 5x10
4
cm/s for 15% of fines (Barrera and Lara,1998).
River alluvial
Granodiorite bedrock
Tailings
Figure 1: Typical Cross Section of Los Quillayes Dam
TAILING SAND
STARTER DAM
TAILINGS
Consolidated isotropicaly undrained (CIU) triaxial tests on 2 samples prepared to a density of 1,50 t/m
3
were carried
out on El Chinche tailings sand. The test were performed for effective confining stresses of 0.1; 0.2; 0.4; 0.8; 1.6 and
2.4 Mpa in order to cover most of the stress range that would be present in the dam during its operation, according to
the original design of a 175m high dam.
a) Stressstraincurve
Fig. 2: Stressstraincurveandstresspaths
of El Chinchesand.
0 5 10 15 20
1
(%)
0
5
10
15
20
25
q(0.1MPa)
0 5 10 15 20 25
p'(0.1MPa)
0
5
10
15
20
25
q(0.1MPa)
M=1.43
b) Stresspaths
q=
1

3
p'=('
1
+2'
3
)/3
3
'=2.4MPa
3
'=1.6MPa
3
'=0.8MPa
3
'=0.2MPa
3
'=0.1MPa
Figures 2a and 2b show the deformation and stress path characteristics for El Chinche sand. From these tests an
average frictional angle of 35 with cero cohesion was assumed for stability analysis purpose.
1.3 Cyclic shear strength of El Chinche tailings sand
Cyclic shear triaxial tests with controlled stresses were carried out to evaluate the liquefaction potential of the tailings
sands, using 2 samples of sand from El Chinche dam, prepared by wet compaction (7 to 8%) to a density of 1,50
t/m
3
. Confining stresses of 0.2; 0.4 and 0.8 Mpa were applied as well as a cyclic stress ratio of 0.05 to 0.23 at a
frecuency of 0.1 to 2 Hz.
1 10 100 1000 10000
Number of CyclesN
0
0.1
0.2
0.3
0.4
0.5
C
y
c
l
i
c
S
t
r
e
n
g
t
h
d
/
2
o
'
(A)El Chinchesand(1997)
19%F,Kc=1.0
(B)Verdugo30%F,Kc=1.0
(C)Verdugo22%F,Kc=1.0
(D)Verdugo15%F,Kc=1.0
(E)Verdugo0%F,Kc=1.0
(F)Obilinovic&Barrera(1982),
1015%F,Kc=1.5,
3C
'=0.2Mpa
(G)Quillayessand(2001),1418%F
Fig. 3: Cyclic strengthvsNumber of Cycles.
Curve A in figure 3 shows the cyclic strength of El Chinche sand for different number of cycles of applied load. Curves
B to E represent the same type of relationship for other hard rock tailings sands as reported by Verdugo (1983).
Curve F shows a reinterpretation of this relationship for Talabre tailings sand as reported by Obilinovic and Barrera
(1982). Curve G is the relationship obtained for the Los Quillayes tailings sand as it will be discussed later on in this
paper.
In general it can be concluded that all the tailings sands present similar pattern, showing a reduction of cyclic shear
strength with the number of cycles. Also the cyclic shear strength increases significantly with the reduction of the
fines content.
1.4 Seismic risk analysis
From the seismic risk analysis that was carried out specifically for this project the main earthquake source identified
was the subduction of Nazca plate, where most of the high magnitud earthquakes are produced ( Ms>7.5), being the
best example one earthquake ocurred in April 1943 (Ms=8.3) with a focal depth of 23 km and only 123 km away from
the dam site.
Main characteristics of the design earthquake are indicated in Table 2.
Table 2. Design Earthquake Characteristics
Characteristics
EQ
Horizontal
EQ
Vertical
Max Acceleration (g) 0.37 0.25
Duration of Strong Movement (s) 27 27
Total Duration (s) 80 80
Dominant Frecuency (Hz) 3 3
1.5 Seismic stability analysis at design stage
The basis for the seismic design of Los Quillayes tailings dam relied mainly in the specification of the most convenient
characteristics of the sand to be obtained economically through the cycloning of the tailings and also in the
specification of relatively conservative construction procedures including some compaction effort, all this in order to
guarantee an adequate behaviour of sands under extreme seismic loads. But the main feature of the design is related
to the relatively high permeability of sands due to a low percentage of fines (less than 18%), allowing a rapid drainage
of the sand dam body during construction and operation, limiting in this way the potential liquefaction of sands under
the action of very strong earthquakes to very minor and isolated sectors within the dam thus not affecting its overall
stability. Great attention was given then to the design of drains under the downstream part of the dam to adequately
take care of the seeping water.
Static stability analysis using limit equilibrium methods and drained shear strength parameters for static analysis and
undrained parameters for postseismic analysis resulted in factors of safety of 2.57 and 1.0 respectively. Pseudo
static analysis with undrained shear strength parameters and a seismic coefficient of 0.15 resulted in a factor of
safety 1.34 for abandonment stage. The pseudodynamic analysis proposed by Makdisi and Seed (1978), together
with Pelambres seismic response spectrum were used to estimate a maximum horizontal deformation of 2.60 m,
quite acceptable for a dam of this type and height.
With the described design features there was no practical concern on either static or seismic stability of such a dam,
consideration that was confirmed by the results of the stability analysis just commented. Nevertheless there were
some factors considered to be unique for this dam, mainly: the high raising rate of more than 20 m per year in the first
two years; the geometry of the relatively narrow valley where the dam had to be build and the height of the dam, the
highest of its type in the country at that time. These factors supported the decision to carry out a dynamic stability
analysis of the tailings dam, using the finite differences FLAC
3D
code.
The model used for this dynamic stability analysis considered a MohrCoulomb failure criteria including the estimate
of increments in pore water pressure according to Martin and Finn (1975). Although seepage calculations through the
dam as well as previous experiences with similar dams showed that the free pore water pressure line would not be
over 4 m above foundation level, a conservative assumption with the free pore water pressure line located up to 10 m
above basal drains level all over the downsream body of the dam was adopted as the initial situation before the
earthquake load was applied.
In order to consider the effect of the geometry of the narrow valley a 3D model was used, to complement the results
of the 2D analysis. The analysis were carried out for a total dam height of 180 m.
The main results of this dynamic stability analysis were:
Maximum deformation of the order of 2.85 m near the toe of the dam, in the slope direction.
Potential liquefaction within the dam is restricted to small zones near the toe of the dam. Excess pore water
pressure in this area does not exceed 50% of effective confining pressure.
The impact on the dam behaviour of the geometry of the narrow valley is only reflected in the slightly bigger
deformation at the toe of the dam when comparing 2D and 3D results.
The higher accelerations occur at the crest of the dam: 0.6 g, larger by a factor of 1.62 than the 0.37 g considered
at the base of the dam at foundation level.
2 CONSTRUCTION AND OPERATION OF LOS QUILLAYES DAM
The construction of the 70 m high starter embankment dam and of the basal drains of the main sand dam were
completed in 1999 (see Fig. 4 and Photo 1). Main findings during the construction of these works were:
Open long run fractures in the rock outcrop that formed part of the left abutment of the dam were identified. These
fractures were treated with extensive concrete infilling and dental concrete. In the starter dam area where water
could be in direct contact with the abutment before the tailings slimes would push it back in the deposit, some
concrete grouting was also carried out.
During the construction of the 70 m starter dam most of the alluvial foundation soils were removed, resting only
some minor sectors with 3 to 5 m of very dense alluvial clayey gravels. In the foundation of the sand dam these
alluvial clayey gravel soils less than 5 m in thickness are also quite dense.
Fig. 4: Starter dam cross section
At present about 1,000,000 m
3
of sand are deposited and compacted monthly in the dam. Figure 5 shows the raising
curve for the dam that in J une 2003 would have a height of about 150 m. During the first two years the dam grew
more than 30 and 20 m respectively. This is about 60 and 40 cm per week, requiring monthly relocations of the sand
slurry distribution system.
After the sand slurry is placed, with a solids content in weight of 70 %, on the downstream slope of the dam in lifts of
variable thickness, intensive spreading, reshaping and compaction of the sand is carried out with a D8 bulldozer and
Fig. 5: Raising Curve of Los Quillayes Tailings Dam
1320
1330
1340
1350
1360
1370
1380
1390
1400
1410
1420
0
1

0
3

0
0
0
1

0
5

0
0
0
1

0
7

0
0
3
1

0
8

0
0
3
1

1
0

0
0
0
1

0
1

0
1
0
3

0
3

0
1
0
3

0
5

0
1
0
3

0
7

0
1
0
2

0
9

0
1
0
3

1
1

0
1
0
3

0
1

0
2
0
5

0
3

0
2
0
5

0
5

0
2
0
5

0
7

0
2
0
5

0
9

0
2
0
5

1
1

0
2
0
5

0
1

0
3
Time
E
l
e
v
a
t
i
o
n
(
m
.
a
.
s
.
l
.
)
Dam
Tailings
Photo 1: Starter dam just buit
a 10 ton smooth vibratory roller travelling along the slope (see Photo 2). Densities of more than 95% of Standard
Proctor density are obtained. These densities are checked at depths of 30, 60 and 90 cm.
During the first year of operation the main challenge was the placement of the sand according to the design geometry
of the downstream slope i.e. 1 to 4 (H:V). The sand was discharged from the crest of the dam and this was
particularly difficult due to the high raising rate and the initially reduced area of the sand dam footprint. In order to
overcome this, aditional discharge points were implemented in order to increase the placement surface (dam
footprint) and to make more efficient the drainage of the wet sand.
This problem was reduced at the same time the dam was heightened and it doesnt constitute a concern any more. At
present, with almost 30 millions of m
3
of dried sand already in place a new modification to the construction procedure
will be necessary: to build an intermediate berm in the dowmstream slope, in order to allow the discharged sand to
reach the toe of the dam without excesive use of earth moving equipment. Actually from the stability point of view it
could be perfectly possible to have a much steeper downstream slope of the order of 1 to 3 or 3.5 (H:V).
The unit cost of the sand placement (including tailings pipes relocation) is about 0.25 US$/m
3
.
3 RECENT GEOTECHNICAL INVESTIGATIONS IN THE SAND DAM
During 2001 the mining company retained SRK Consultores to carry out a soil investigation in the sand tailings dam
in order to evaluate insitu geotechnical properties of the sand as placed, to determine if conditions meet pre
Photo 2: Sand tailings dam
construction assumptions and to reassess the stability of Quillayes dam and to define parameters for use in the
design of a future tailings dam with the same sand.
The investigation considered the following elements:
Two rotatory drill holes 52 and 54 m deep, SPT and shelby sampling of sands
Four seismic cone penetrometer (SCPT) boreholes 5366 m deep
Down the hole shear wave surveys
Seismic refraction surveys, two profiles
Three sampling test pits 2 m deep
The results of this investigation has been presented by Swaisgood (2002) in a recent international seminar on
tailings. The basic sand properties obtained are also shown in Table 1, as those corresponding to Los Quillayes sand.
The results of the cyclic shear strength tests are included in Figure 3.
The main conclusions of this investigation confirmed a layered embankment without weak zones, with absence of
water.The investigation also concluded that the existing geotechnical properties meet predesign assumptions, that
the dam is stable and that the same construction method and general design can be applied in the new dam to be
designed.
Table 3 shows a comparison between the basic design parameters of the original design and those recomended for
the new tailings dam to be built with the same tailings sand.
Table 3. Comparison of Key Geotechnical Properties
Parameter Original
Design
Revision
Relative Density (%) 59 60
Friction Angle () 3535.7 35
CRR (*) 0.21 0.24
(*) Cyclic Resistance Ratio
4 HEIGHTENING OF LOS QUILLAYES TAILINGS DAM
The original design considered a final maximum height of the dam of 175 m corresponding to a capacity of 257
millions tons of dry tailings and approximately 7 years of operation at an average production of 114,000 tons of
mineral per day. Considering the satisfactory results of the investigation of Los Quillayes and adequate performance
of the dam and in order to gain time to optimize the location and design of the second tailings deposit, Minera los
Pelambres decided to study the feasibility of increasing the capacity of Los Quillayes tailings deposit to 360 millions
tons i.e. heightening the dam to a maximum height of 198 m, thus deferring the change to a new deposit in
approximate 3 years.
For this heightening in general the same design has been considered but adapting the basal drains to the new
production level. The axis of the dam has been slightly shifted near the right abutment in the downstream direction in
order to avoid a small depression inside the deposit area.
Preliminary pseudodynamic stability analysis have confirmed the feasibility of this heightening but nevertheless
cyclic shear strength tests now on Los Quillayes tailings sands are currently under way at higher confining stresses to
check sand behaviour at those stresses although no significant differences with the existing data is expected. On the
other hand (if necessary) there is room for many adaptations considering the conservative present design of Los
Quillayes dam.
From the environmental point of view this heightening of the existing dam is quite positive in the sense that there is
not significant new impact on the area of the deposit, the water diversion scheme is mantained but the regulatory
capacity of the impoundment to control extreme floods has been increased significantly as well as the capacity of the
structures for seepage control (aditional cut off and higher capacity of the seepage water collection pond).
5 CONCLUSIONS
The perfomance of Los Quillayes tailings dam after 3 years of operation has been quite satisfactory with geotechnical
properties of the placed sand meeting quite closely predesign assumptions and without showing weak zones in the
layered embankment for a high volume of 1,000,000 m
3
of tailings sand deposited and compacted monthly,. The very
low levels of pore water pressure registered up to now, practically at the level of the alluvial soil foundation, complete
this positive picture.
Despite the unprecedent high dam raising rate, the difficulties in the sand placement at the beginning of the operation
were overcome at very low cost through additional discharge points and the enlargement of the area of sand
placement complying with the geotechnical properties of the design.
The experience with Los Quillayes tailings dam shows that hydraulic sand dam adequately designed following the
general concepts associated to the practice of the large mines in the last two decades and properly operated could be
easily adapted to accommodate constraints or height increase.
ACKNOWLEDGMENTS
The authors thanks Compaa Minera los Pelambres for their support and for permitting the publication of the
information on Los Quillayes tailings deposit of their property.
The authors are very gratefull to engineers J os Luis Lara and Edgar Bard of Arcadis Geotecnica who helped with
the collection of information and review of the paper.
REFERENCES
Barrera, S., Lara, J .L. (1998) Geotechnical Characterization of Cycloned Sands for the Seismic Design of Tailings
Deposit, 3
rd
International Congress on Environmental Geotechnics, Lisboa, Portugal
Garga, V. K., CastillaCamacho, J . E. (1995) Embankments Dams, X Panamerican Conference of Soil Mechanics
and Foundation Engineering, General Report Theme V, Guadalajara, Mexico
Lara, J . L.; Gmez, P. y Barrera, S., (1999) Anlisis Dinmico Tridimensional de la Presa de Arena de Relaves
Quillayes PelambresChile, XI Conferencia Panamericana sobre Mecnica de los Suelos e Ingeniera Geotcnica
(XI CPMSIG), Foz do Iguau, Brasil
Makdisi, F.I., Seed, H.B.(1978) Simplified Procedure for Estimating Dam and Embankment EarthquakeInduced
Deformations, J ournal of the Geotechnical Engineering Division, ASCE, Vol. 104
Martin, G.R., Finn, L. And Seed, H.B. (1975) Fundamentals of Liquefaction under Cyclic Loading, J ournal of the
Geotechnical Engineering Division, May, pp. 423438.
Obilinovic, H., Barrera, S. (1982) Dynamic Analysis of the North Talabre Tailings Dam, First Chilean Conference of
Geotechnical Engineering, Santiago, Chile
Swaisgood, J . (2002) The Investigation and Evaluation of Quillayes Dam, Technical Visit, High Density & Paste
2002, Santiago, Chile
Valenzuela, L. (1996) Characteristics of Chilean Tailings Dams, Large Dams in Chile. Edited by ICOLD.
Valenzuela, L., Barrera, S. (1995) Large Tailings Dams Current Practice in Chile, X Panamerican Conference of
Soil Mechanics and Foundation Engineering, Guadalajara, Mexico
Verdugo, R., (1983) Influencia del Porcentaje de Finos en la Resistencia Cclica de Arenas de Relaves, Civil
Engineer Thesis, Catholic University, Chile.