INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST
– 6TH
, 2014
Design Optimization of Bauxite Residue Dam
in Connection with Environment and Land Acquisition
in Mempawah SGA, Indonesia
A. Fitriyanto & W. Taruko PT. ANTAM (Persero) Tbk.
DESIGN OPTIMIZATION OF BAUXITE RESIDUE DAM
IN CONNECTION WITH ENVIRONMENT AND LAND ACQUISITION
IN MEMPAWAH SGA, INDONESIA
Mining and mineral processing activities cannot be separated from tailing. In its operation,
Smelter Grade Alumina processing plant will produce mud called bauxite residue. In order to
accommodate the mud output, we should build dam. The development of bauxite residue dam is
closely related to the environment and land acquisition. Because its form is mud, bauxite residue
dam must be designed in such a way that is environmentally friendly. In addition, the design of
bauxite residue dam sometime must be adjusted in the field because of the challenges in land
acquisition. This paper will explain the strategies that can be taken in the design optimization of
bauxite residue dam i.e. do pre-treatment by pressing and filtering the mud that would reduce the
volume and the toxicity of the mud. The dam also needs to be built by staging system to overcome
the challenges of land acquisition and to minimize catchment area of the dam. Because it is in a
dry form, the bauxite residue can be disposed forming bench so that it will reduce the large of dam
area required. The liner system also should be made to ensure there is no infiltration from the dam,
so it is safe for the environment. Based on research that has been done, bauxite residue can be used
as geopolymer brick with low compressive strength 53.5 kg/cm2 and high compressive strength
238.9 kg/cm2. It can also reduce the large of dam area required.
Keywords: bauxite residue dam, design optimization, environment, land acquisition
1. INTRODUCTION
Mining and mineral processing activities cannot be separated from tailing. In its operation,
Smelter Grade Alumina processing plant will produces mud called bauxite residue. In
order to accommodate the mud output, we should build dam. The development of bauxite
residue dam is closely related to the environment and land acquisition. Bauxite residue is
waste product of the Bayer Process. It is disposed as slurry with 10-30% of solid
concentration. Because Bayer Process uses dissolution of alumina in caustic soda for
extraction of the same, the waste also contains approximately 3wt.% sodium hydroxide
III- 1
that makes the mud highly alkaline (pH in the range of 12 to 13). The color of bauxite
residue is red. This color is caused by the oxidized iron, which can make up to 60% of the
mass of the bauxite residue. Because its form is mud, bauxite residue dam must be
designed in such way to meet environmental requirement.
Figure 1: The Sequence of Bauxite Mining and Processing Activities
2. OPTIMIZATION OF MATERIAL VOLUME AND TOXICITY
The processing of bauxite residue will be developed in stages in accordance with the
progress of plant operations Mempawah SGA. Bauxite residue mud from the mud line
through the clarification process is leading to pumping mud thickener. Inside the mud
thickener, its overflow which is a solution of caustic liquor will be recycled back to the
plant. While its underflow shaped high-density mud will be pumped directly to the disposal
area. For the next stage of development, the factory SGA Mempawah possible to install
vacuum filters in the red mud disposal area. The tool is a tool with an intensive and
operational special expertise needs, because it will installed accordance with the
development and needs of the plant. Vacuum Filters will filter the output becomes
oversized mud thickener and undersize. Undersize of vacuum filters will have a solid
output in the range of 150-300 kg/m2h. Furthermore, undersize will be pumped to the
system through a High Pressure Positive Displacement Pumping Mud Pump (PDP) headed
dam area. The oversize of vacuum filter will be transferred back to the mud thickener. The
simple diagram is as follows:
III- 2
FLOCCULENT
MUD THICKENER
Dilute Mud Slurry
Overflow Recycle
To Plant
Underflow High Density Mud
High Pressure Mud Pumping
To Mud Droppers Via Mud Line
Evaporation
CLoud
Sprinkler
P-30
E-13
DECANT
RUN-off
Recycle to
Process
Bauxite Residue Dam Area
Vacuum Filter
Oversize
Undersize
NEUTRALIZING AGENT
From Plant
Bauxite Residue Dam –
Stage 1
The Development of Bauxite
Residue Dam – Stage 2
Figure 2: Press & Filter of Bauxite Residue Diagram
Bauxite residue as slurry has a solid concentration in the range of 10-30%. Pre-treatment
by pressing and filtering the slurry will reduce the volume and the toxicity of the bauxite
residue. After pressing and filtering the slurry, the bauxite residue has a solid concentration
in the range of 60-70%. The volume of bauxite residue can be reduced. The toxicity of
bauxite residue also can be reduced.
The following are the differences between bauxite residue in the form of filter-cake and
slurry:
No Filter-Cake Slurry
1
Dam is relatively more simple in design
(thinner). It is because active earth pressure
is only from solid material.
Dam is relatively more complex in design
(thicker). It is because active earth
pressure is from solid material and water.
The volume requirement is relatively
smaller.
The volume requirement is relatively
bigger.
Liner specification for coping the water
seepage is relatively lower.
Liner specification for coping the water
seepage is relatively higher.
Transportation
Must prepare road and transport
equipment (truck) because it can not be
transported through pipelines.
Material transport is carried by pipelines.
Filtration Must set up facilities for filtration.
Facilities for filtration is not needed. It is
because the bauxite residue is passed
directly (slurry).
Filtration There is a cost for filtering. There is no cost for filtering.
TransportThe cost for transport equipment (fuel) is
relatively high.
Operation cost is relatively low (cost is
only from fuel for pump machine).
OPEX
Description
Financial
CAPEX
Dam
III- 3
Table 1: The Differences between Bauxite Residue in the Form of Filter-cake and Slurry
3. DESIGN OPTIMIZATION
3.1 Land Acquisition Limitation
After the Reformation Era in 1998, the land acquisition became the significant issue for all
projects in Indonesia. The government doesn’t have enough influence and power to
support the land acquisition. It causes many variations in price range and badly the land
lord usually offers much higher than the market price. The land lord decides what price is
appropriate for the land otherwise they don’t release the land. This situation affects to cost
and duration of project. The challenge for the project is how to manage the effective layout
plan in minimum area.
No Filter-Cake Slurry
2
Water Reclaim Water reclaim for process plant is higher. Water reclaim for process plant is lower.
Material
Transport
Using trucks. The possibility of redmud fall
is small.
Using pipelines. There is a potential
outbreak of slurry pipelines. It needs a
pipeline inspection.
VolumeVolume is relatively smaller (dam height is
relatively lower).
Volume is relatively bigger (dam Height is
relatively higher).
Stability Forming a more stable land. Relatively unstable.
ClosureRelatively easier. It is because the
embankment materials are solid.
Relatively more difficult. It is because the
embankment materials in the form of
mud.
3
Toxic liquid is filtered. Relatively more toxic.
Relatively lower. Relatively higher.
Piping potency is lower (potency of dam
failure is lower).
Piping potency is higher (potency of dam
failure is higher).
Noise and dust from vehicles (material
transport).
There are no noise and dust from vehicles
(material transport).
Lower. Higher.
Leakage potency (from material transport)
is smaller.
There is a potential outbreak of slurry
pipelines. It needs a pipeline inspection.
Possibility of overtopping is smaller (the
solid form reducing the quantity of the
water).
Possibility of overtopping is higher (the
solid form reducing the quantity of the
water).
Environment
Toxicity
Groundwater
Piping Potency
Air Pollution
Potency
Erosion Potency
Leakage Potency
(Material Transport)
Overtopping
Technical
Filtration
Transportation
Dam Design
Description
III- 4
Figure 3: Progress of Land Acquisition at Project Area
The figure 3 shows progress of land acquisition at project area that had been started from
2012. The hatch areas are the land that had been acquired by Antam. The total area of land
acquisition is 353 hectares. Still some areas haven’t been acquired (blank area) and the
total requirement is 580 hectares.
According to the land acquisition progress, the development of bauxite residue area should
be started from the north side of plan area then go to southwest.
3.2 Staging System
Based on the progress of land acquisition, the design is divided by several stages. The
design should cover the total bauxite residue from plant operation in 30 years plus extra
20%. The bauxite residue production is 1.15 million m3 per year, so the total bauxite
residue production is 41.4 m3 for 30 years. We have 6 stages on the design. Each stage in
design covers 5 year operations with total production of bauxite residue is 5.75 m3 per
stage. The staging system also minimize catchment area of the dam, so the water that have
to be treat is reduced. The staging system can be seen at figure 4.
III- 5
Figure 4: The Staging System of Bauxite Residue Dam
3.3 Benching System
Because the bauxite residue is in a dry form, it can be disposed forming bench. It will
reduce the large of dam area required. Based on the characteristic of the bauxite residue,
the slope can be designed with ratio 1:3. The following are the characteristic of bauxite
residue:
Figure 5: The Benching System of Bauxite Residue Dam
III- 6
Figure 6: Stability Analysis of Bauxite Residue Dam (without Earthquake)
Figure 7: Stability Analysis of Bauxite Residue Dam (with Earthquake)
3.4 Counterweight System
Embankment on the outer side of bauxite residue heaping has a function as a
counterweight and planned as a barrier so that the bauxite residue does not splatter in times
of rain.
Figure 8: The Counterweight System of Bauxite Residue Dam
3.5 Liner System
The liner system also should be made to ensure there is no infiltration from the dam, so it
is safe for the environment. Clay layer with thickness 1 meter and permeability 1 x 10 -7
cm/s needs to be made under bauxite residue layer to prevent seepage of water that may
still be contained in the bauxite residue that has been pressed and filtered. In addition to
release water from the dam subgrade as an effort to accelerate the consolidation process, it
required sand layer with 30 cm of thickness below the clay layer, where the sand layer at
the same time also functioned as a leak detection layer.
III- 7
Figure 9: The Liner System of Bauxite Residue Dam
3.6 Drainage System
Surface drainage channel is planned in the bauxite residue dam that drains the rain water
from dam area to the water storage pond. Bauxite residue which is carried by erosion will
settle on water pond and drained by opening the spillway. Some of the rain which fell on
dam area will be drained to the Waste Water Treatment Plant (WWTP) to be managed to
meet the environmental quality standards. Planned drainage system in the dam area is
separating the dam drainage and drainage of around the dam area. It minimizes the
environmental impact of the dam.
Figure 10: The Drainage System of Bauxite Residue Dam
3.7 Waste Water Treatment Plant
WWTP operation starts from anaerobic pond to maturation pond. Anaerobic pond is the
first collection tank for contaminated water, it functions to improve dissolved oxygen
(DO), solid precipitation, and stabilization in the influent. Facultative pond functions to
control biochemical oxygen demand (BOD). Maturation pond is to decrease waste
contaminant by evapotranspiration and bind metals.
Figure 11: Waste Water Treatment Plant
III- 8
3.8 Closure
The final planning is dam closure plan. Dam closure must be done to return the function of
the area as green land. The surfaces of bauxite residue dam should be reshaping,
landscaping, and covering with 50 cm of fertile topsoil. The planting of trees is preferably
using endemic plants grown in the area (acacia).
Figure 12: The Planning of Dam Closure
4. RED MUD UTILIZATION
Based on research that has been done, bauxite residue can be used as geopolymer brick
with low compressive strength 53.5 kg/cm2 and high compressive strength 238.9 kg/cm
2
(mix design = bauxite residue : fly ash : tailing : lime = 35% : 25% : 30% : 10%). It can
reduce the large of dam area required.
Adding of Sodium Silicate (%) Compressive Strength (kg/cm2)
0 53,5
0,5 75,8
1 85,8
2 91,5
4 135,6
8 238,9
Table 2: Increase of Compressive Strength by Adding of Sodium Silicate
Figure 13: Micro-photographs of Geopolymer Brick
Left: Low Compressive Strength 53.5 kg/cm2
Right: High Compressive Strength 238.9 kg/cm2
III- 9
5. CONCLUSION
The development of bauxite residue dam is closely related to the environment and land
acquisition. Bauxite residue is waste product of the Bayer process. Because its form is mud,
bauxite residue dam must be designed in such way to meet the environmental requirement
and land acquisition limitation. Pre-treatment by pressing and filtering the slurry will
reduce the volume and the toxicity of the bauxite residue.
The design should covers the total bauxite residue from plant operation in 30 years plus
extra 20%. The bauxite residue production is 1.15 million m3 per year, so the total bauxite
residue production is 41.4 m3 for 30 years. There are 6 stages on the design. Each stage in
design covers 5 year operations with total production of bauxite residue is 5.75 m3 per
stage.
Based on the characteristic of the bauxite residue, the slope can be designed with ratio 1:3.
The liner system made from clay layer with thickness 1 meter and permeability 1 x 10 -7
cm/s needs to be made under bauxite residue layer to prevent seepage of water that may
still be contained in the bauxite residue that has been pressed and filtered. Surface drainage
channel is planned in the bauxite residue dam that drains the rain water from dam area to
the water storage pond. Some of the rain which fell on dam area will be drained to the
Waste Water Treatment Plant (WWTP) to be managed to meet the environmental quality
standards.
Based on research that has been done, bauxite residue can be used as geopolymer brick
with low compressive strength 53.5 kg/cm2 and high compressive strength 238.9 kg/cm
2. It
can also reduce the large of dam area required.
REFERENCES
Tekmira and Antam (2011): Pencucian Bauksit, Pemanfaatan Tailing Skala
Pengembangan dan Pemanfaatan Residu Bauksit, Kerjasama Penelitian Antam
dengan Puslitbang Tekmira, pp. 12-35, Citatah, Indonesia.
Antam (2012): Basic Design for Bauxite Residue and Ash Disposal SGA Mempawah
Project, pp. various pages, Jakarta, Indonesia.
Davidovits, J. (2011): Geopolymer Chemistry and Applications, Institut Géopolymère,
Saint-Quentin, France.
Cablik, V. (2007): Characterization and Applications of Red Mud From Bauxite
Processing, pp. 2-9, Ostrava, Czech Republic.
Szépvölgyi, J. (2011): Properties, Disposal and Utilization of Red Mud, Institute of
Materials and Environmental Chemistry, CRC HAS, pp. 19-23, Budapest, Hungary.
III- 10
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST – 6TH , 2014
Design and construction of an exposed geomembrane sealing system
Paper Title Line 1 (14pt) hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf fffffjfjjfkkfjjj
for the Sar Cheshmeh tailings dam raising in Iran
2(14pt)
C. Noske & P. Sembenelli ATC Williams, Australia
SC Sembenelli Consulting, Italy
A. Scuero & G. Vaschetti Carpi Tech, Switzerland
ABSTRACT Sar Cheshmeh is a large copper mine in Iran. Its tailings dam comprises a 70m high Main
Embankment with inclined clay core and outer colluvial gravel shell. ATC Williams, engaged to
consider tailings management options for a production escalation involving almost 1 billion tonnes
of tailings over 31 years, designed a scheme comprising a 40m high downstream raise to the Main
Embankment in four stages. The critical design case included up to 20m of water ponding against
the upstream face. The design of the raised embankment was constrained by the existing inclined
core, as the raised core would become an upstream diaphragm with little rockfill cover to act as a
surcharge during seismic loading. Iran is a highly seismically active region, and the resultant seismic
factor of safety for this conventional raising approach was found to be unacceptable. Clay hence
needed to be eliminated from the raise, and a geomembrane sealing system on a rockfill embankment
was adopted. From a construction, performance and cost point of view, an exposed liner was
preferred. Due to its superior mechanical and durability properties, a 3mm thick PVC geocomposite
system, designed by Sembenelli and Carpi Tech, and supplied and installed by Carpi Tech,
represented the optimal solution. The geomembrane sealing system utilises a patented face
anchorage system fastening the geocomposite to a drainage layer of extruded porous concrete curbs.
This paper discusses the design issues and the installation of the first two stages of the raise,
completed in 2008.
Keywords: Geomembrane, geocomposite, PVC anchor strips.
1. INTRODUCTION
Designers of dams, particularly tailings dams, generally agree that the most cost effective
solution is to maximise the use of natural materials available on site. However, in some
situations a design cannot be made to work with the available materials. This was the
situation encountered at the Sar Cheshmeh mine tailings storage, located in a region of high
seismicity in central Iran.
Interest in the application of geomembrane sealing systems (GSS) in the design of
embankment dams for tailings storage facilities has been increasing in the last two decades
around the world. The reasons for this interest include favourable deformation
characteristics, environmental aspects (i.e. seepage minimisation), shortage of low
Abstract Number : 121
III- 11
permeability materials, installation efficiency and general constructability of geomembrane
liners in a wide range of climatic conditions.
This paper, which describes the design and installation of a GSS for the first two stages of
the Sar Cheshmeh dam raising, conceptually follows and amplifies two papers presented at
ICOLD Congresses: in 2006, the GSS with flexible face anchorage adopted at Sar Cheshmeh
dam raising was discussed merely as a new design concept for construction of embankment
dams with a geomembrane facing [Scuero & Vaschetti, 2006]; in 2009, the GSS was
presented as first field application on a project that at the time the paper was written was still
under construction [Scuero & Vaschetti, 2009]. In 2014, after six years of operation, the
whole construction process can be described, and field results and conclusions can be drawn
regarding the performance of the lining system.
2. BACKGROUND
2.1. The existing dam
The existing Sar Cheshmeh tailings dam, owned by National Iranian Copper Industries
Company and commissioned in 1980, comprises a 75 m high embankment consisting of an
inclined clay core and of an outer colluvial gravel shell. The site is located within an area of
complex geology, consisting of rocky, sparsely vegetated terrain well above 2,000 m
elevation, in a climate ranging in temperature from about 40°C to well below freezing point.
Figure 1. Sar Cheshmeh dam prior to raising
Due to the changing political environment in Iran at the time, the documented design,
construction and performance history of the embankment were incomplete. It was however
evident from surveillance audits that repair works were carried out during the 1980s and
1990s to correct seepage, settlement and cracking issues relating to high pond levels.
2.2. The raising
ATC Williams were engaged in 2001 to design optimised tailings and water management
options for a proposed production escalation involving almost 1 billion tonnes of tailings
over 31 years. Integral to the project was the buttressing and raising the Main Embankment
and Saddle Dam No.1, to provide tailings storage for the remaining mine life.
The embankment design required a 40 m high, 1000 m long downstream raise constructed
in four stages, the first two of which, comprising 20 m of raising, are the subject of this
III- 12
paper. The critical design case included up to 20 m of water ponding against the upstream
face. In order to maintain the embankment as a water retaining structure, the conventional
downstream raising approach would have resulted in a raised core effectively being an
upstream diaphragm with little rockfill cover to act as a surcharge during seismic loading.
Iran is one of the more seismically active regions in the world, evident by the peak ground
acceleration for the Maximum Design Earthquake of 0.8g. Stability analyses showed that
the seismic factor of safety for this approach was unacceptable. Compounding the issue was
a shortage of suitable clay-based soils in what is a semi-arid area, and that the borrow sources
used in the original embankment construction had since been covered by tailings.
A detailed study of alternative water retaining elements was conducted. Options investigated
included asphaltic cores, and bituminous and polymeric membranes. A cross-section
consisting of an upstream GSS on a rockfill embankment was subsequently adopted, on the
basis of it being a more stable, efficient and buildable arrangement.
Figure 2. Inferred cross section of the existing dam, and scheme of the raising
3. GEOMEMBRANE SEALING SYSTEM
3.1. Design criteria and assessment of GSS options
In order to establish the criteria for the final selection of the GSS, a detailed study of the
published literature was conducted, giving preference to publications by independent bodies
such as ICOLD [2010]. A number of GSS options were identified, with the short-list
consisting of a) bituminous concrete/membrane, b) high density polyethylene (HDPE)
synthetic geomembrane, c) polyvinyl chloride (PVC) synthetic geomembrane, and d)
geocomposite (consisting of a PVC geomembrane heat-coupled during extrusion to a
polypropylene non-woven geotextile). Each alternative was analysed with respect to
hydraulic properties, mechanical properties, durability, practical requirements, and
established precedents.
It is not in the purpose of this paper to go into details regarding the analysis and design of
the embankment raising, or the detailed rationale and selection process for the GSS. Such
details are discussed by Noske [2010]. Upon completion of the assessment, it was concluded
that an exposed PVC geocomposite represented the GSS solution that would best meet the
necessary requirements for Sar Cheshmeh.
An exposed geomembrane was preferred to a covered geomembrane for several reasons. In
order to maintain the existing, relatively steep slope 1.5:1 (horizontal to vertical), the only
III- 13
viable membrane cover would be cast in situ concrete slabs. Flattening the slope to
accommodate a granular fill cover would result in a cost prohibitive increase in downstream
rockfill volume. Cast in situ concrete slabs, as with any type of cover, would require more
stringent Quality Control procedures, increase the risk of membrane damage, and increase
construction times and costs.
Over the last two decades, there have been considerable advances in the development of
geomembranes, additives, and anchorage methods. Exposed geomembranes are now a viable
alternative to a covered system, and in many ways they are now considered state of the art.
Exposed geomembranes have the advantages of minimised risk of membrane damage, easier
construction, access to the membrane for inspection or repair, and reduced overall
construction times and costs.
It was hence clear that from a construction, performance and cost perspective, an exposed
GSS should be preferred.
3.2. Upstream Exposed Geomembrane Sealing System
A tender process was initiated for the detailed design, manufacture, supply and installation
of an exposed PVC geocomposite GSS. Based on demonstrable experience with respect to
materials and installation expertise, the contract was awarded to Carpi Tech, a Swiss/Italian
specialist waterproofing company.
The selected waterproofing PVC geocomposite was SIBELON® CNT 4400 (manufactured
exclusively for Carpi Tech according to a proprietary formulation), consisting of a 3 mm
thick PVC geomembrane, heat-bonded during manufacturing to a 500 g/m2 non-woven
polypropylene geotextile. The SIBELON® manufacture and materials are specifically
formulated to impart to the geomembrane the tensile and weathering properties necessary to
survive in exposed conditions in the long term.
The design of the upstream face of the raise specified a relatively simple two-layer system
consisting of a GSS base/anchorage layer over a granular filter/transition/drainage layer.
This system was in turn supported by the rockfill body of the embankment raise.
The concept of the final design for the GSS, developed by SC Sembenelli Consulting of
Italy, was to provide imperviousness to the dam via a very deformable system that could
adapt to settlements and to seismic events. A key factor was the method of anchoring,
supporting and providing drainage for the GSS. An innovative method was devised, making
use of the so-called Ita method of support layer construction that was adopted for the dam.
The Ita method involves porous concrete extruded curbs 0.4 m high, installed on the
upstream face and providing a solid, uniform supporting layer for the placement, anchoring
and drainage of the GSS.
As the curbs are being extruded, 0.5 m wide anchor strips of geomembrane material are fixed
onto each curb; overlapping strips are heat-seamed, so that when the embankment is
completed continuous parallel bands of PVC geomembrane are embedded in the slope of the
embankment from crest to upstream toe. The horizontal spacing between adjacent PVC
anchor strips is 6 m, calculated as a function of the design wind uplift loads.
The waterproofing PVC geocomposite sheets are unrolled from the crest and anchored
against uplift by seaming them to the PVC anchor strips. Figure 3 shows the concept.
III- 14
Figure 3. Cross section at curbs, and scheme of PVC geocomposite placement (1. Curbs, 2. PVC
geocomposite, 3. PVC anchor bands, 4. Plinth, 5. Seaming to PVC anchor bands)
The transition between the existing embankment and the raising comprises an upstream
sloping, 10 m wide compacted clay zone seated in the upper part of the existing core. The
crest of the core transition forms the bottom peripheral anchorage for the exposed PVC
geocomposite, which was placed in a trench excavated in the core transition and then
backfilled with clay. At the rock abutments, the watertight seal is of the mechanical tie-down
type to reinforced concrete plinths.
The top seal of the first intermediate stage of the raise is mechanical and placed on a curb
formed with mass concrete at the crest. Top anchorage for the second stage is required to be
more permanent, and was made by positioning the PVC geocomposite in a trench excavated
at the crest of the stage, and by ballasting it with a mass concrete anchor beam.
TYPICAL SECTION ON ANCHOR STRIPS
0.15
40mm
Heat weld
Temporary ballast (concrete, gravel,Porous concrete curb
Anchor strips w=0.50
PVC geocomposite
Sibelon CNT 4400
0.15
Interlock keys between curbs
4 m long centered between two
anchor strips.
0.30
0.1
0
sand, steel bar, etc.)
1 2
3
4
5
5
6
III- 15
4. INSTALLATION
4.1. Curbs and anchor strips
The extruded porous concrete curbs upon which the GSS is anchored have the advantage of
very quick construction, provide a regular, non-erodible slope that can be quite steep, and
provide containment for the embankment fill during construction.
After each curb has been completed, a PVC geocomposite anchor strip is nailed to the curb
for temporary anchorage, permanent anchorage being provided by the compacted granular
fill placed against the curb as the embankment is raised. Each PVC anchor strip overlaps the
PVC strip installed on the curb underneath, and the two are heat-seamed at the overlap, thus
constructing parallel PVC anchor bands, 0.50 m wide, over the curbs. In this way, the
anchorage system for the PVC geocomposite is in place as soon as the embankment raise is
completed, as shown in Figure 4.
Figure 4. Heat-seaming of two overlapping PVC anchor strips, and the parallel PVC anchor bands
on the first stage of raising
This anchorage system is flexible and deformable, meeting the seismic design objectives for
the raise. PVC has an elongation at break greater than 230%, allowing the system to maintain
its integrity under the full range of predicted embankment deformations.
A
4.2. Placement and anchorage of PVC geocomposite
When the raising of the first, intermediate stage was completed, the waterproofing liner,
supplied in rolled sheets, was temporarily fastened at the crest and then deployed over the
curbs. The high friction at the geotextile/curb interface facilitates the positioning of the
geocomposite and increases the stability with respect to sliding.
The sheets of PVC geocomposite were placed to coincide with the PVC anchor strips, and
then permanently secured to the curbs by heat-seaming their edge to the PVC bands (Figure
5). Adjacent PVC geocomposite sheets were then watertight heat-seamed, forming one
continuous PVC liner over the curbs.
When the raising of the second stage was completed, the same procedure was followed. The
PVC geocomposite was deployed from the newly raised crest, down until it overlapped the
top anchorage of the first stage. A watertight, horizontal seal between the two sheets was
III- 16
then created by heat-seaming and the placement of a PVC geomembrane cover strip, which
was itself heat-seamed to the two adjacent geocomposite stages as shown in Figure 6.
Figure 5. The PVC geocomposite sheets are unrolled from the crest and heat-seamed to the
underlying PVC anchor strips.
Figure 6. The PVC geocomposite sheets of the two stages (Stage II B is covered by dust) are
seamed at the overlapping, and the seam is covered by a PVC geomembrane strip
4.3. Perimeter sealing
4.3.1. Bottom sealing
The bottom seal was made by placing the PVC sheets in a trench excavated in the compacted
clay core transition at top of the existing embankment. The anchor trench was then backfilled
with compacted clay, as shown in Figure 7.
At the concrete plinths constructed at the abutments, the bottom seal was of the mechanical
tie-down type as shown in Figure 7. The surface of the concrete was regularised with epoxy
resin, and the PVC geocomposite anchored to the plinth using a 60 x 6 mm stainless steel
batten strip placed on an EPDM gasket, and tied-down with stainless steel anchor rods
embedded in chemical phials placed into the concrete at 150 mm spacing.
4.3.2. Top sealing
The top seal of the first, intermediate stage was of the tie-down type, placed on a mass
concrete curb at the crest, as shown in Figure 7, while the top seal at Stage II C is made in a
trench excavated along the crest and backfilled with a mass concrete anchor beam.
III- 17
[Blank line 10 pt]
Figure 7. Perimeter sealing methods (first stage of raise)
Installation of the waterproofing system for the first stage of the raise started on 28 June
2008 and was completed on 14 August 2008. The second stage installation started on
12 October 2008 and was completed on 26 November 2008. Total installed geocomposite
was 20,500 m2 and 18,000 m2 for the first and second stages respectively.
5. SIX YEARS PERFORMANCE
After one year of service, the designer [Noske 2010] reported: “From the perspective of both
the designer and the owner, the selected GSS has resulted in an efficient and economic
waterproofing solution for the tailings storage raise. The extruded curbs have proven to be
an effective construction method, whilst the anchor strip installation became a simple,
routine process…. The installation was fast…. The overall construction period was also
significantly reduced…… Impoundment of water against the toe of raised embankment has
recently commenced, and seepage measurements downstream of the embankment have not
changed from their steady-state levels. It is concluded that the geocomposite faced rockfill
approach is a viable, effective means of tailings dam construction …. where natural materials
are either not available, or are unable to be used from a technical perspective”.
In 2014, after 6 years of service the raised embankment has continued to perform as intended,
although the water level at the toe of the raise has remained relatively static over the period
(refer to the photos in Figure 8). This has been due to the commissioning of the tailings
thickening system designed by ATC Williams, which has resulted in a steeper beach slope
depositing tailings further back up in the heads of the valleys. However, the decant pond
will continue migrating back towards the Main Embankment, bringing the PVC
geocomposite into full service in the near future.
III- 18
4. CONCLUSIONS
The selected geomembrane sealing system with anchorage strips embedded in the upstream
face of the embankment during construction has minimised the impact that traditional
geomembrane installations have on the construction schedule of a dam. Installation of the
GSS is quick and easy, so that just a few weeks are required to complete the whole facing
system after completion of the embankment body. The waterproofing system is “more
forgiving” in comparison to conventional systems as far as construction skills are concerned.
Companies with limited previous experience in geomembranes can carry out the
construction of a relatively straight-forward rockfill embankment section and install the
extruded porous concrete curbs, while a specialised contractor supplies and installs the
waterproofing system to efficiently achieve a watertight retaining structure capable of
satisfying all necessary design and serviceability criteria.
After the Sar Cheshmeh tailings dam raising, other field applications of this new design have
been constructed, are under construction, or have been included in the design of embankment
dams yet to be built. Behaviour up to date indicates that the system constitutes a viable and
effective alternative to traditional upstream concrete facings and impervious cores.
Figure 8. Sar Cheshmeh dam raising at the end of the second stage construction works, and in
service during 2013.
REFERENCES
ICOLD, International Commission on Large Dams (2010): Bulletin 135. Geomembrane
sealing systems for dams - Design principles and review of experience, ICOLD,
Paris, France.
Scuero, A. M. and Vaschetti, G.L. (2006): Unconventional cross sections and materials in
embankment dams. ICOLD 22nd Congress, Barcelona, Spain. Proceedings Vol. I -
Question 84: pp. 87-105, International Commission on Large Dams - Paris – France.
Scuero, A. M. and Vaschetti, G.L. (2009): Unconventional design in dam raising: Sar
Cheshmeh tailings dam. ICOLD 23rd Congress, Brasilia, Brazil. Proceedings Vol. I
– Question 90: pp. 18-20, International Commission on Large Dams - Paris – France.
Noske, C. (2010): Geocomposite faced rockfill - an innovative means of water-proofing
tailings storages. Mine Waste 2010, Proceedings Vol. 1: pp. 291-302, Australian
Centre For Geomechanics, Perth, Australia.
III- 19
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST
– 6TH
, 2014
Tailings Storage Risk Reduction by Integrated Waste Management Mine
hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf fffffjfjjfkkfjjj
at Didipio Mine
2(14pt)
D.M. Brett, R.J Longey & S.P. Edwards GHD Pty Ltd, Hobart, Tasmania, Australia
ABSTRACT OceanaGold’s Didipio Mine Project comprises an open pit operation in the Dinauyan River valley
of northern Luzon, Philippines. The site is characterised by steep topography, high rainfall and
deeply weathered tropical soils (saprolite) prone to landslips on the steep valley slopes. In general
these conditions create a challenging situation for safe disposal of tailings, particularly
considering closure after completion of mining. The solution was developed by considering the
synergies of total waste from the mine; the estimated 40 million m3 of tailings and 70 million m3 of
waste rock over the projected 17 year mine life, moving the tailings dam upstream from its original
planned location and allowing it to be integrated into a conservatively stable waste rock dump.
“Flow-through” dump technology is being used to manage significant stream flow within the waste
storage area, creating “real-estate” for waste disposal and also managing the impact of peak
storm flows on the open pit which will eventually span across the valley floor with resultant risk of
flooding.
The integrated waste facility design allows a high degree of confidence to meet the closure
planning aims of ICOLD Bulletin 153 (draft) “Sustainable Design and Post-Closure Performance
of Tailings Dams”.
This paper outlines the design development of the Didipio tailings storage facility with reference to
ICOLD bulletin 153, and describes experience in the first two years of construction, including
initial placement of the “flow-through” dump structure.
Keywords: Integrated Mine Waste Management; Tailings Dams, Flow-Through Waste Dump
1. INTRODUCTION
OceanaGold‟s Didipio Project is a gold/copper deposit located near the village of Didipio,
on Luzon Island, Philippines, approximately 270 kilometres north of Manila and 100
kilometres to the east of the country‟s largest gold-copper mining operations centred in the
Baguio area, as shown in Figure 1.
The ore body is located beneath the Dinauyan River close to its confluence with the
Surong. The site is characterised by steep topography, high rainfall and deeply weathered
tropical soils (saprolite) prone to landslip on the steep valley slopes. In general these
conditions create a challenging situation for safe disposal of tailings, particularly
considering closure after completion of mining.
III- 20
Figure 1. Didipio Project Location
Planning of the mine developed from a small open pit with underground mining, to a
significantly larger open pit, resulting in a mine life of 17 years generating an estimated 40
million m3 of tailings and 70 million m
3 of waste rock. This change led to the open pit
encroaching on the Dinauyan River, with consequent risk of pit flooding, but created the
potential for a much lower risk water and mine waste management methodology.
The original mine concept wrestled with the difficulties of creating a Tailings Storage
Facility (TSF) in a significant river, with relatively small volumes of waste rock materials
for construction. Water management was envisaged to involve diversion tunnels or major
pump installations and relied on rapid construction of a dam in a wet tropical climate. The
solution was to push the tailings dam further upstream to a tributary valley of the
Dinauyan, develop a “flow-through” rock drain in the main river channel and take
advantage of the much increased waste rock volume expected from the pit, to construct a
large waste rock dump (WRD) able to contain major flood flows and modulate outflows to
the river as it passed the pit area downstream.
2. DESIGN OVERVIEW
The design basis for the project evolved to allow for the following:
Mine life of 17 years, processing 365 days/year, 24 hr/day
Mill production of 2.5 Mtpa by the end of Year 2 and increasing to 3.5 Mtpa
Tailings delivery of 51 Mt at 60% solids with settled density 1.3 t/m3
Waste rock storage 136 Mt at average density 2.0 t/m3
Tailings geochemically benign Non-Acid Forming (NAF)
Waste rock is classified NAF with minor Potentially Acid Forming (PAF)
Design criteria based on o Australian National Committee on Large Dams (ANCOLD) guidelines;
o ICOLD Guidelines including Bulletin 153, Draft, 2011;
Didipio Project Location
III- 21
o “Policy Guidelines and Standards for Mine Wastes and Mill Tailings
Management” Memorandum Order No. 99-32 (24th November 1999),
issued by the Philippine Department of Environment and Natural Resources
(DENR).
TSF Consequence Category (ANCOLD, 2012) „High C‟
Construction spillways to pass 1:5 ARI flow events
Post-construction emergency spillway to safely store/pass a Probable Maximum Flood (PMF);
TSF Decanting System to utilise pumping for removal of decant water and maintain minimum flood storage capacity for the 100 yr ARI 72 hr (1.5 Mm3) flood
TSF embankment utilising downstream construction methodology
Use waste rock from mining operations where economical to do so
Where waste rock is unsuitable, maximise use of locally won materials
Low permeability clay core to retain supernatant water when required and minimise
seepage.
Low permeability cut-off trench to minimise seepage
Filter protection for clay core piping failure protection
Seismic Loading for Maximum Design Earthquake (MDE)1:10,000 AEP, 0.50g;
WRD spillway to pass 100 yr ARI 72 hr event
Storage of PMF on WRD without overtopping following Year 5
Flow-through drain to pass wettest monthly flow (2.6 m3/s)
TSF/WRD Closure
o Partial wetland/water cover, land use available for revegetation or
cultivation
o Flooded pit with diversion of flow from TSF/WRD (Dinauyan River) as a
large water body
o Stable (negligible erosion, settlement, blockage risk) spillway for the long
term
The final landform proposed is shown in Figure 2.
Key features of the design as shown in Figure 2 are:
Tailings storage upstream of and buttressed by the WRD;
“Flow-through” drain in the river bed under the WRD;
“Retention dam” wall and flood-plain formed by the WRD;
Operational spillway formed from coarse rock at the WRD face;
Open Pit formed across the full extent of the Dinauyan Valley
III- 22
Figure 2. Didipio Integrated Tailings and Waste Rock Storage
The design envisages a closure scenario where the pit is flooded and provides a silt-trap for
the upstream catchment while stable revegetation is developed. The long term spillway is
provided over the ridgeline to the south of the TSF to prevent the risk of discharge over the
face of the WRD in the event that the “flow-through” becomes blocked.
This concept closure design is considered to form the basis for sustainable closure in
accordance with ICOLD Bulletin 153 (ICOLD, 2011), meeting the aims of the Bruntland
Report (UNWCED, 1987) adopted by the International Council on Mining and Metals
(ICMM), namely, “development that meets the needs of the present without compromising
the ability of future generations to meet their own needs”. The integrated waste storage
facility is conservatively stable from both geotechnical and hydraulic aspects in the long
term, while providing potential for agricultural development on the final landform and
passive water quality management through a large water body at the downstream end of
the site. The final closure design is dependent on monitoring of performance during
operations and ongoing risk review as the final structure evolves.
3. SITE GEOTECHNICAL CONDITIONS
The valley slopes comprise residual diorite soils to varying depths, typically 5 m to 22 m,
overlying highly fractured, slightly to moderately weathered (MW-SW) Diorite. The
strength of the residual silty clay/clayey silt soils generally varies from firm to very stiff
consistency but strength reduces with increasing moisture content and mechanical
working. When exposed after excavation and subjected to wet conditions the residual soils
lose strength and cut slopes can fail if an adequate batter slope is not provided. Similarly
spreading and compaction can disrupt the soil structure.
Final Pit filling Dinauyan Valley
WRD Operational Spillway
WRD “Dam” Wall
WRD Floodplain
Final Long Term Spillway
Final Tailings Beach –
rehabilitated and vegetated
Flow-Through Drain Under
III- 23
The valley floor in general consists of 2 to 4 m of silts, sands, gravels, cobbles and
boulders (up to 1.2 m in size) overlying residual diorite soil or in some areas moderately
weathered diorite.
4. TSF EMBANKMENT
The TSF comprises a zoned earth and rockfill dam located at the upstream end of the
WRD, in the catchment of Luminag Creek. The TSF is being built in stages from a starter
dam of 40m high, expected to eventually reach 100 m high and merging with the WRD to
become an integrated structure. The final TSF embankment volume will total nearly
20,000,000 m3, approximately 30% of total waste rock from the mine life. The internal and
external geometry of the TSF/WRD has taken into consideration the outcomes both of
stability analysis and a construction methodology. The major construction material to be
used in the TSF is fresh non sulphide waste rock (Zone 3) won from the open pit and
carted by the mine fleet consisting of Cat 777 off highway haul trucks and 40 tonne
articulated dump trucks.
A typical Cross-Section of the TSF is presented in Figure 3.
Figure 3. TSF Typical Cross-Section
The majority of the waste rock, particularly during the initial few years will be placed in
the TSF by paddock dumping, spreading in layers and compacted by truck/roller passes.
The downstream zones in later years may be placed by developing tip heads of 10m
maximum height.
The Zone 1 material comprises extremely weathered diorite rock (saprolite), comprising
medium to low plasticity silty clays and clayey silts with an average PI of 40 % and field
moisture contents (FMC) of +8 % wet of optimum moisture content (OMC). The wet
nature of the Zone 1 affected the maximum achievable compaction in the field as discussed
later in the paper.
The embankment features primary and secondary filters, which are produced by crushing
the fresh mine waste rock. A primary sand filter (Zone 2A) protects against piping through
the embankment and foundations. A horizontal width of 1.5 m has been specified for the
Zone 2A and 2B filters placed downstream of the clay core, considered the narrowest
practical width based on construction tolerances. The Zone 2A and 2B filter blanket along
the downstream foundation has a minimum thickness of 600mm, this again is considered
the practical minimum for construction. The Zone 2A and 2B filters in the TSF have been
III- 24
extended to the crest level in the vertical direction and 30m beyond the extent of the Zone
1 clay core on the foundations in the horizontal direction. The horizontal (blanket) filter is
used to protect the extremely weathered diorite foundation from piping where there are
high hydraulic gradients in the foundation materials. The Didipio TSF embankment
materials zones pictured within Figure 4 are shown in a right to left direction (upstream to
downstream) Zones 3C, 1, 2A, 2B and Zone 3.
Figure 4. TSF Embankment Material Zones
5. WASTE ROCK DUMP
The WRD will be integrated with the TSF to form a single mass of waste rock in the latter
stages of the project. The WRD location and concept will provide the following benefits:
Sufficient waste storage capacity for the life of mine;
Provide additional stability to the TSF;
Provide a safe disposal cell for any Potentially Acid Forming (PAF) waste
materials;
Providing a method for passing normal flows of the Dinauyan River through the TSF and WRD in a controlled manner via a “flow-through” drain constructed at the
base of the dump;
Provide a method of safely storing flood flows from the Dinauyan River in a controlled manner by providing a detention basin formed between the TSF and
WRD, until such a time when the flow-through drain can pass the retained flood;
and
Provide a high level access to the TSF from the pit for materials haulage and access routes for tailings mill return water pipelines.
A key component of the WRD is the “flow-through” drain, which is proposed to be
constructed by dumping fresh, coarse NAF waste rock from a minimum tip head of 20m.
Upstream Zone 3C
Saprolite Abutment
Zone 1
Zone 2A
Zone 2B
III- 25
This minimum tip head is designed to encourage segregation of the waste rock so the large
boulders fall first to the base of the drain forming a thick permeable zone for flow to pass
through before the general waste rock is dumped above the drain. The flow-through drain
will not be sufficient to pass the required design PMF, but is designed to cater for the
average wettest monthly average flow.
To cater for large flood flows, the WRD will be constructed in a way to allow for detention
of flood water once the flow-through drain has reached its full flow capacity. When the
drain reaches capacity, excess water will back up within the detention basin provided
between the WRD and TSF crests.
The WRD detention basin is sized to eventually cater for storage of a PMF event, however
there is potential for overtopping, particularly during the initial development of the dump.
Accordingly, the WRD features a coarse rockfill spillway designed to cater for a 1:100
ARI flood event (72 hr critical case).
6. FLOW-THROUGH DUMP EXPERIENCE & PERFORMANCE MONITORING
There are several notable examples of flow-through drains for waste rock dumps. The
authors have particular experience in the “flow-through” drain at the Savage River Mine in
Tasmania, Australia (Brett et al, 2003). This experience in the performance of existing
“flow-through” drains has been used in sizing of the Didipio “flow-through”.
As the future pit expands to cross the Dinauyan River, pipes are proposed to pass flows
around the pit on a bench, thus the “flow-through” outflow has been sized large enough to
pass the average wet month flows, but not so large that the flows become unmanageable to
deliver around the open pit by gravity pipes. Performance monitoring is currently being
installed upstream and downstream of the drain to assess its constructed flow capacity
against the design target. The instrumentation and flow monitoring performance
assessment will become a critical tool to allowing ongoing design calibration and
optimisation using the observational approach (ANCOLD 2012). Adopting this
observational approach to the design and construction of the “flow-through” will enable
informed decisions to be made regarding modifying the as-constructed drain capacity. If
the drain flow capacity is found to be undersized the drain can be expanded or duplicated
higher up in the dump, if the drain capacity is deemed too high the drain could be
„throttled‟ by placing finer rock over the drain inlet to achieve the desired design flow
capacity. The TSF and “flow-through” drain upstream intake are shown in Figure 6.
III- 26
Figure 5. Didipio TSF Embankment Construction And Flow Through Dam Intake
7. SITE CONSTRUCTION EXPERIENCE
Construction of the Didipio Project TSF, especially the starter dam was a challenge for the
design and construction teams. The challenges encountered and project solutions
successfully implemented during the construction phase are briefly discussed in the
followings section.
The major source of materials for the TSF is the Didipio open pit, therefore careful
integration between the mine planning and TSF construction teams is critical to the success
of the project, especially in the very early stages of construction. The starter dam design
was reliant on fresh rock to construct the downstream shell of the TSF and to also provide
a coarse boulder overtopping spillway zone on the embankment. Difficulties with the
development of the mine starter pit occurred due to underestimation of pit stripping and
approval limitations which meant the TSF bulk rockfill shell material was not able to be
produced in significant quantities in the early stage of TSF construction. A solution was to
modify the starter embankment zoning to accommodate a less permeable highly weathered
rock and alluvial zone (Zone 3C) sourced from within the TSF storage combined with a
coarse rock drainage layer, this design was continued until the mine was able to strip the
starter pit and produce adequate rockfill volumes.
The same issue in delays with the starter pit resulted in the boulder zone for the TSF
overtopping spillway not being available, this was overcome by using a finer rockfill zone
sourced from the TSF surrounds. The finer rock available would have eroded in the
construction design flood flows thus a temporary reinforced mesh spillway was adopted.
The mesh spillway design was only to last 3 months until the boulders could be produced
from the mine starter pit. The temporary spillway design shown in Figure 6, utilised readily
available site materials including 50mm aperture mesh security fencing fastened to the fine
rock spillway surface using horizontal and vertical reinforcing bars, the bars on the
spillway face were secured by additional bars anchored within the embankment rockfill.
Flow Through Dam Intake TSF Construction
III- 27
This mesh spillway design was far more cost effective than alternative concrete options
and suited the short design life. The 50m wide mesh spillway was designed to pass
velocities up to 1:5yr ARI flow events. In July 2012, the permanent coarse boulder
spillway had been constructed to 10m high on the downstream side of the 20m high mesh
spillway when a 180mm rainfall occurred within 2 hours over the TSF catchment,
equivalent to a 1:2yr ARI event. Both spillways performed adequately as designed, with
only minor movement observed of the finer rockill contained by mesh. The coarse
overtopping rock zone was not impacted and has been adopted for future spillway designs
as the dam is progressively raised in the downstream direction. The overtopping spillway
on the embankment uses flood retention to retain peak flood flows allowing the remainder
to safely pass through the coarse boulder zone in the downstream embankment at non
eroding velocities. The design has been adopted to prevent the need for cutting more
conventional type spillways into the naturally steep side slopes of the TSF abutments
preventing large costly excavations which would potentially destabilise the natural
topography.
Figure 6. Left Plate; TSF Temporary Mesh Spillway In Starter Dam. Right Plate; Stage 7
Overtopping Spillway
The use of saprolite soils which are classified as clayey Silts and silty Clays for the low
permeability Zone 1 upstream sloping core presented a challenge in using the site available
materials which are in excess of +8% wet of OMC. The construction rate and wet tropical
climate makes significant drying back of these materials to a more commonly specified -
1% to +2% OMC quite impractical. Therefore, the design specification allows for a lower
level of compaction with a minimum 95% maximum dry density (MDD) using standard
compaction. The design accounts for the reduced Zone 1 material strength due to the lower
compaction and high FMC. The Zone 1 design specification requires minimum undrained
shear strength (Su) of 60kpa, which was derived from lower bound laboratory triaxial test
results from samples remolded to suit achievable actual field compaction conditions, this
low strength material causes instability in the upstream direction if not addressed. The
upstream stability issue was overcome in the design by incorporating a 10m wide upstream
stability zone consisting of granular general fill (Zone 3C) which enables the embankment
to be raised up to 30m above the tailings level. The upstream Zone 3C material also
provides erosion protection should a tailings spigot blowout on the embankment face occur
or a flood event results in ponding and wave action on the upstream embankment wall. The
Zone 3C also provides for ease of constructability allowing a trafficable zone on both sides
of the Zone 1 material. The use of Zone 3C has also utilises waste materials from the pit
which do not meet either Zone 1 or Zone 3, this maximises the material usage from the pit
at all stages of mining.
III- 28
The climate at the Didipio project site has a relatively high annual average rainfall of 3.3m
with three distinct seasons. The „dry‟ season lasts January through June, which has an
average monthly rainfall of 148mm, the „typhoon‟ season July through September having
300mm rainfall average per month and the „wet‟ season having 500mm average per month.
The wet climate combined with the naturally high MC saprolite soils, limits practical and
economical placing of the Zone 1 material in the „dry‟ season only. This requires a staged
construction of the TSF to suit the seasons such that the bulk of the downstream rockfill
shell zone is placed during the typhoon and wet seasons ready for raising of the Zone 1 and
filters by campaign in the dry season.
The remoteness of the Didipio mine and size of the TSF construction project made it a
necessity to setup an onsite soils testing laboratory. Laboratory technicians have been
trained to working to Australian Standard testing procedures for the TSF construction
QA/QC program. Laboratory testing has shown oven drying at standard temperatures
(110C) can impact the natural saprolite soil structure and thus compaction tests are
undertaken using the AS 1289.5.7.1 Rapid Hilf testing method using air drying techniques,
which has proven effective and results in quick turnaround time for compaction tests. In
addition to the field density tests for compaction a field shear vane is also used for both
compliance and indicative testing on Zone 1 materials placement. Required field vane tests
are set a minimum 15% higher than the required minimum design strengths to account for
field-lab vane variability.
9. CONCLUSION
The current design and construction of the Didipio Mine Waste Management system has
proven that integrated waste management can result in practical, economical and safe
storage of mine waste in a tropical climate in keeping with the recommendations within
ICOLD Bulletin 153. With planning and innovation, integrated waste management has
enabled the mine operation to successfully develop mine waste storage within a major
floodway which has also enabled future risk mitigation for managing pit flooding.
10. ACKNOWLEDGEMENTS
The authors express their gratitude to the management of OceanaGold Corporation for
permission to present this paper and to staff, consultants and contractors who have
contributed to the success of the project.
11. REFERENCES
ANCOLD, 2012: Guidelines On Tailings Dams - Planning, Design, Construction,
Operation And Closure, Australian National Committee on Large Dams.
Brett, D.M. and Hutchison, B. (2003): Design and performance of a “flow – through”
spillway at Broderick Creek waste rock dump – Savage River Mine, Australian
Journal of Water Resources, Vol 6 No.2, 2003, Institution of Engineers, Australia
ICOLD, 2011: Sustainable Design and Post-Closure Performance Of Tailings Dams,
International Committee of Large Dams
III- 29
1
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST
– 6TH
, 2014
MANUAL FOR DESIGN, CONSTRUCTION AND OPERATION OF
TAILINGS DAMS IN IRAN
VAHID FARIDANI, HAMID REZA TAMANNAIE, REZA BAGHI Bandab Consulting Engineers, Tehran, Iran
SHAHROKH TAHOUNI National Iranian Committee of Tailings Dams Tehran, Iran
ABSTRACT: Comparing to regular water reservoir dams, tailings dams potentially have more environmental
impacts. After some incidents in tailings dams in recent years in the country, and considering the
insufficiency of technical documents and national regulations for these dams, the ministry of power
which is the governmental organization; responsible for water recourses, decided to prepare a
national technical document to promote knowledge in this field. The intention is to guide the
owners of the mines to a safer construction, operation and abandonment of tailings dams. This
document is titled “Manual on design, construction, and operation of tailings dams”, and takes
into consideration the current trends of design and construction of tailings dams in Iran. The
manual is prepared under supervision of the committee of tailings dams of Iranian national
committee on large dams (IRCOLD), and though concentrates mostly on the above ground
facilities; it is expected to be a helpful document for other kinds of tailings storage facilities as
well.
Major part of this guideline is based on published experiences of other countries and accepted
recommendations thereof, so in this paper, after briefly introducing the structure of the document,
only those parts of the manual which deal with the problems from a national viewpoint, will be
introduced with some details.
Keywords: Tailings Dam, Manual, Design-Construction-Operation.
1. INTRODUCTION
Growing mining activities in country; increased concern about the safety of tailings dams.
After some incidents in these dams in recent years with negative impacts on society, it has
become evident to the governmental authorities that, in addition to the lack of regulations,
there is a lack of criteria and standards related to all aspects of tailings dams. As the first
step to address this issue, the ministry of power, the governmental organization responsible
for water recourses, decided to prepare a national technical document to promote
knowledge in this field. This document is titled “Manual on design, construction, and
operation of tailings dams”. This Manual is prepared under supervision of the sub-
committee of the tailings dams of Iranian national committee on large dams (IRCOLD).
III- 30
2
2. CONTENTS OF THE MANUAL
Considering the fact that, tailings dams potentially have more environmental impact in
comparison with regular water reservoir dams; the main road map in composing this
manual was to preserve the sustainability of tailing dams. This involves consideration of
environmental, economical and technical aspects throughout the life of the structure,
including its abandonment and post closure.
The manual concentrates mostly on the above the ground facilities and consists of nine
chapters. More than fifty percent of the contents (three chapters) deal with the design
problems, such as storage methods, storage design, and design of confining structures.
Since Iran is located on earth seismic belt, and most parts of the country have an arid to
semi-arid climate, which makes water a scarce commodity, more attention is given to
prevention of contamination of natural water resource, water recovery, and seismic design
of structures in these chapters. In other chapters some topics about risk analysis,
management, construction, and operation, surveillance, and post closure issues are covered
in brief.
Considering the fact that the back bone of this guideline is based on ICOLD publications,
and published experiences of other countries and accepted recommendations thereof, it did
not seem to be of much interest to present the whole manual completely, so only main
parts which deal with the problems from a national viewpoint, will be presented here.
3. DESIGN AND SAFETY EVALUATION FLOOD
The Ministry of Power of Iran is presently in the process of publishing a guideline for
classification of water reservoir dams in the country. This classification is used to
determine the design requirements for dams. To classify a dam, the following factors are
considered:
Dam height and reservoir capacity
Area of affected Irrigated land
Annual water supply for domestic and industrial needs
Installed capacity of hydropower stations
Population at risk
Based on this classification, appropriate design flood and safety evaluation flood for dams
can be determined from table 1.
Table1. Recommended Minimum Design, and Safety Evaluation Flood for Water Reservoir Dams
Dam Type Flood Class of The Dam
1 2 3 4
Earth Fill Design(AEP)
1:10000 1:5000 1:2000 1:1000
Safety Evaluation(AEP) PMF 1:10000 1:5000 1:2000
Concrete Design(AEP) 1:5000 1:2000 1:1000 1:500
Safety Evaluation(AEP) 1:10000 1:5000 1:2000 1:1000
Annual Exceedance Probability
III- 31
3
It is obvious that the consequences of failure, or even any small deficiency in a tailings
dam, are much more hazardous than failure of a normal water storage dam, so a more
conservative approach should be considered to determine the design requirements of
tailings dams. In this context a qualitative assessment of consequences of tailings dams’
failure and water spillage from its reservoir by considering factors such as impact on
environment, damage to infrastructure, public health, social affects, and population at risk
is recommended. As Iran is a developing country and considerable changes in population
at risk and affected areas should normally be anticipated, it is recommended to carry out
this assessment periodically and in all stages of the life of the tailings dams. Results of this
assessment should then be compared with the requirements of an earth fill water storage
dam having similar sizes. The requirements for the tailings dam should be more than or
equal to the requirements of such water storage dam.
4. TAILING CONTAINMENTS TYPES
There are many different alternatives to design and construct a tailings dam. Description of
these alternatives could easily be found in literature, and selection amongst them for a
specific project depends upon the tailings properties, natural topography, site conditions,
and obviously economic factors. To facilitate a systematic approach to implement a
strategy to manage the tailings, it seemed necessary to classify these alternatives in a
manner shown in table 2.
Table2. Types of Tailings Containment
Construction
Method
Confining Structure
Construction Material Reservoir Location
Single or Multiple
Stage With Barrow Material
In Channel
Above
Ground Tailings
Dams
Up Stream Starter Dam+ Self-Stacking
Tailings (Raising Embankment) Down Stream
Centerline
Single or Multiple
Stage With Barrow Material
Side
Hill
Off
Channel
Up Stream Starter Dam+ Self-Stacking
Tailings (Raising Embankment) Down Stream
Centerline
Single or Multiple
Stage With Barrow Material
Paddock Up Stream Starter Dam+ Self-Stacking
Tailings (Raising Embankment) Down Stream
Centerline
Open Pits
Filling Existing Voids Co-Disposal with Mine Wastes
Underground Mines
Submarine Tailings Disposal Deposition in
The Environment
Notes: 1- Due to scarcity of water in the country, off channel construction is more encouraged.
2- Considering the high seismicity of most regions of the country, upstream construction method is less
implemented in Iran.
3- As there are almost no significant mining activities close to deep water in Iran, deep sea tailing placement is not an experienced practice in Iran.
III- 32
4
5. ARRANGEMENT FOR TECHNICAL DIRECTION OF THE WORK
In compliance with general recommendations made in many references, the importance
and necessity of engagement of competent personnel in all aspects, related to tailings dams
is emphasized in this manual. However it should be noted that, a safe and technically
appropriate tailings dam can not be built by just relying on responsibilities of individuals
such as designers, construction companies, operators and even planners or managerial
bodies; engaged in the project. A system which organizes the relationship between
involved entities that carry out the work is also of major importance. In figure 1 below
recommended structures for execution of works by a contractor is shown. This structure for
an in-house construction could be seen in figure 2. Along with these structures following
principals are also stressed in the manual.
Regardless of the size of the project, the design works and duties of the supervision over construction should be entrusted to properly certified legal entities rather than an
individual- Real person.
As far as possible the designer and supervisory body should be of a single legal entity.
In cases where, two separate entities are employed for design and supervision work, every effort should be made to ensure an effective presence of the design body during
the construction period of the dam, and; preferably throughout the life of the structure.
Figure1. Proposed Framework for execution of project by a contractor
III- 33
5
Figure2. Proposed Framework for in-house execution of project
ACKNOWLEDGEMENT
The writers wish to thank the Australian National Committee on Large Dams (ANCOLD)
who generously provided a copy of the final draft of the GUIDELINES ON TAILINGS.
The preparation of the manual for tailings dams in Iran, and also this paper, is supported by
Iran Ministry of power’s Bureau of Developing Plan of Water and Wastewater Technical
Standards and Criteria.
III- 34
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST
– 6TH
, 2014
Optimization of Tailings and Water Management Schemes
hhdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkjf
fffffjfjjfkkfjjj
in Taft and Dareh Alou Copper Mines
2(14pt)
H. R. Seif National Iranian Copper Industries Company, Integrated Water Master Plan, Tehran, Iran
A. Roshdieh ATC Williams, Melbourne, Australia
H. Zaker Middle East Water & Environment Consulting Engineers, Member of the Iranian National Committee on Large Dams,
Tehran, Iran
ABSTRACT The National Iranian Copper Industries Company is placing significant investment inexpanding its
existing mining complexes, as well as developing new mines. "Tailings and Water Management" is
regarded as a crucial aspect of these development projects. Regarding the dry climate and severe
shortages in water resources in Iran, maximizing the water recovery from the tailings outflow from
the concentrator plant, has been a major objective in the related engineering designs. In this paper,
Tailings and Water Management of two of the more recent mine development projects in the very
dry region of Central Iran are reviewed. Three major components of the "Tailings and water
Management" studies including tailings "Dewatering", "Transport" and "Storage" and different
combinations of the available options are investigated. Technical and economic aspects of these
components and their role in the overall "Feasibility Studies" of these projects are discussed. In
view of the water scarcity in the regions, raw water resourcing and transfer costs are also taken
into account. Both mines have mineable ore capacities of about 140 Million tons. Concentrator
plants with nominal capacities of 7 Mtpa (900 ton per hour), are designed to operate for 20 years
of mine life. The study shows that despite similarities in the mine and concentrator plant capacities,
Capital Costs of the optimized Tailings and Water Management Schemes may differ significantly,
mostly due to the topographic and environmental restrictions in each region.
Keywords: Water, Tailings, Feasibility, Copper, Stacking, Filtration, Beach Slope
1. INTRODUCTION
Technical, economic, environmental and social issues associated with mine tailings
deposition and water management are the main reason why a holistic approach is required
for the selection of the tailings storage facilities (TSF) in mining projects. In this article,
two copper mines with similar throughputs and minable reserves are selected and it is
shown that despite their similarities, the selected tailings disposal schemes are different.
Both of these mines are located in Iran and they are both part of the Iranian National
III- 35
Copper Company (NICICO) expansion plan. The environmental conditions in the region
are the reason why water management and conservation is of crucial importance in both of
these projects. The mines covered in this paper are the Taft mining complex and the Dareh
Alou mine – both open cut copper mines in central Iran.
Due to water scarcity in the region and the cost of water, the client instructed that at least
paste thickening should be utilized for tailings dewatering. Also, due to the complexities
and problems associated with pumping of slurry, the client was reluctant to use slurry
pumps and requested the minimization of slurry pumping in both projects.
2. BASIS OF DESIGN
2.1. Tailings Dewatering and Transportation
Various dewatering methods are adopted in the mining industry for the main purpose of
water recovery. Another advantage of dewatering the tailings is the fact that if water is
removed from the tailings the potential for storing the tailings in a stack with steeper
angles would become more viable. In the mining industry, the most common dewatering
facility is the thickener, which covers a wide range of forms from “conventional”
thickeners with the lowest underflow solid concentration, to “high-rate”, “high density”and
finally “paste” thickeners with the highest underflow solid concentration. Generally the
capital cost and also the operation costs of thickeners increase as the underflow solid
concentration increases.
Another less common dewatering technique is filtration. In general, there are two main
types of filter systems, pressure or vacuum filters. Usually vacuum filters are more
mechanically demanding and require intensive maintenance. On the other hand, pressure
filters use hydraulic or mechanical forces to apply the required pressure. Due to the
adverse impact on vacuum filters of the elevation of the sites (1600~2400 m), pressure
filters have been assumed to be more viable in these studies.
Typical effluent solid concentrations that can be achieved from each of the dewatering
technologies are presented in Table 1.
Table 1.Typical Slurry Concentrations for Copper Tailings
Case Solid Concentration
(%)
Un-thickened tailings slurry ex concentrator 20-25% w/w
High rate thickener underflow 50-54% w/w
Paste thickener underflow 60-64% w/w
Cake from pressure filters 75-80% w/w
Typically, as the solid concentration of the tailings increases, the Rheological properties
change. A typical tailings slurry rheogram is presented in Figure 1.
III- 36
Figure 1. Typical Rheograms for Copper Tailings Slurry at Various Concentrations
As the solid concentration of slurry increases, the volumetric flow rate decreases, but at the
same time, the required energy for slurry transportation usually increases. This required
energy could be in the form of potential energy in the case of gravity transportation of
slurry or required pump head. Thus, the combination of tailings dewatering and tailings
transportation system should be optimized. Gravity slurry transportation could take place
in slurry channels or pipelines with various types and shapes.
In the case of filtered tailings, the product is no longer a liquid and typically is referred to
as “cake” which requires solid handling techniques. Such techniques are usually more
expensive than fluid transportation and requiremore sophisticated equipment.
2.2. Tailings Storage and Return Water
Tailings with lower solid concentration, deposit in the TSF with a flatter beach slope which
usually results in higher embankments. On the other hand, tailings slurry that is thickened
sufficiently will not segregate, i.e. there will be no hydraulic sorting of particle sizes when
it is deposited sub-aerially on a beach, and will form a planar slope for a given percent
solids, which is referred to as the “stacking” of thickened tailings. The form of the stack
can be as a cone on flat terrain and is called Central Thickened Discharge (CTD) or as
Down-Valley Discharge (DVD) if the discharge is down-slope to a retaining embankment.
Based on the methods proposed by Pirouz and Williams (2007) and Fitton (2006),
prediction of the beach slopes has been undertaken. The beach slope for sub-aerial
deposition (i.e. not underwater deposition) of non-segregating tailings slurry is a function
of particle size distribution, mineralogy – expressed by specific gravity, clay content,
particle angularity – rheology, and flow rate. Most of these parameters are fixed for a
given ore type and mineral recovery process, but rheology and flow rate are variable. A
typical tailings particle size distribution curve is presented in Figure 2.
III- 37
Figure 2. Typical particle distribution curve for Copper Tailings
If the tailings are even further dewatered and “cake’ is produced, then they can be
deposited with steeper slopes, which could decrease the overall footprint of the TSF.
However, filtered tailings require more sophisticated and more expensive deposition and
spreading methods.
It is obvious that as the solid concentration of tailings increases, less water remains in the
deposited tailings which would result in production of less bleed water and typically less
return water from the TSF. Bleed water together with the surface run off that is generated
in the TSF area are usually returned back to the concentrator. This is called “return water”.
3. TAFT MINING COMPLEX
The Taft mining complex comprises two copper mines; Aliabad and DarehZereshk,
respectively 58 and 62 km south-west of Yazd city. Both the Aliabad, and
DarehZereshkmine sites are located on sloping planes surrounded by hills, grading up to
steep mountainous terrain at RL 2470m. The location of the concentrator is proposed at
approximately 1.5 km north of DarehZereshk village, which is immediately upstream from
the pit in the same floodway valley. On the other hand, the preferred location for the TSF
is in lower, flatter terrain, approximately 35 km to the south of the concentrator location at
RL 1600m, this would give an overall average slope of 25 m/km. This is in desert area
with minimum environmental and social impacts.
The total throughput at the concentrator is 7 Mtpa which comprises of 4 Mtpa from
DarehZareshk mine and 3 Mtpa from Aliabad mine over the 20 years of mine operation.
This equates to generation of 140 Million tons of dry tailings.
The TSF location is in an arid region with an average of 100 mm/year of precipitation and
3,800 mm/year of pan evaporation with low humidity.
III- 38
The average required fresh water for the two cases of paste thickening and filtration is
estimated at 387,200 m3/month and 169,200 m
3/month respectively.
In this study,central thickened discharge (CTD) and tailings filtration (Dry Stacking)
schemes have been investigated at the TSF location. For the CTD option three tailings
dewatering and transportation options including paste thickening at the concentrator and
paste thickening at the TSF plus the option of high rate thickener at the concentrator and
paste thickener at the TSF have been investigated. In the case of the filtration option, the
best arrangement was proven to be a high rate thickener at the concentrator and filtration at
the TSF location.
In the CTD option an overall earthwork of approximately 2Mm3 would be required mostly
in low bunds around the perimeter of the TSF and a central tailings delivery pipe ramp.
Thickened tailings is predicted to be stacking in the TSF at maximum slope of 3%. The
overall plan of the TSF for the CTD is presented in Figures 3.
Figure 3.TSF plan for CTD case for Taft mine
In the filtration option it is assumed that the “cake” stacks in a steeper angle but in order to
prevent erosion, the face of the stack is planned to be covered by mine waste. For the
Transportation of “cake’ to the TSF location various options including trucks and conveyor
belt were examined and conveyor belt is proven to be more cost effective. Also the
distribution of tailings in the TSF is undertaken by stacking equipment instead of trucks.
The overall plan of the TSF for the filtration option is presented in Figures 4.
III- 39
Figure 4.TSF plan for filtration case for Taft mine
In the case of transportation of un-thickened tailings to the TSF location, the required
energy gradient for the tailings pipeline (NS 750mm steel pipe) is estimated at 6.7 m/km,
for the transportation of partially thickened tailings from high rate thickener underflow (NS
500 mm steel pipe) an energy gradient of 8.7 m/km is estimated to be required and in the
case of transportation of paste the required energy gradient is 42 m/km (NS 400 mm steel
pipe).
A cost estimate and analysis have been undertaken with prices at 2011 rates, and a rate of
return equal to 8%. The result of the costs estimate is presented in Table 2.
Table 2.Cost Estimation Results at Taft TSF
Option CAPEX
(USD)
OPEX
(USD/year)
NPV
(USD)
Paste thickener at concentrator 202M 10.9M 283M
High rate thickener at
concentrator and Paste at TSF
166M 4.3M 169M
Paste thickener at TSF 160M 5.6M 175M
Filtration 280M 11.2M 320M
On this basis, the equivalent cost of tailings deposition in the least expensive CTD option
and the filtration option are $2.4/ton and $4.6/ton respectively.
The net present value of the cost of fresh water supply in the case of CTD is estimated at
$135M while the cost of fresh water supply for the filtration option is estimated at $25M.
By including these costs in the total evaluation, the net present value of the total costs at
the CTD option is $304M,and$345M for the filtration option.
4. DAREH ALOU MINE
III- 40
Dareh Alou mine is located approximately 4 km north west of Gerdootan village, which
lies about 23 km north east of the town of Rabor in Kerman province. Dareh Alou mine
site is located on a sloping plane in a valley that is surrounded by hills, grading up to steep
mountainous terrain. The surface elevations of the proposed pit are generally between 2800
and 2900 m above sea level. Since this is a green field project the location of the
concentrator was selected in conjunction with the tailings deposition options study. As the
result five contractor locations with various ore transportation options have been
investigated together with the tailings and water management study. The result of the
study indicated that the best concentrator plant location was at RL 3050m at a distance of
approximately 2.5km from the preferred TSF location.
A number of different TSF site locations were also investigated, and the results of initial
these investigations suggested that only one TSF site was worthy of further studies. The
tailings disposal management method at the selected site is a Central Thickened Discharge
(CTD) system. Two embankments are required for this site. Embankment 1 (South West
Embankment) is a 44 m high embankment. Embankment 2 (North East Embankment) is 80
m high. The total volume of required embankment construction is 7.2 Mm3 which will be
constructed in various stages. Tailings will be delivered to the center of the TSF on an
earthfill pipeline ramp that would be extended during the life of the mine. The thickened
tailings, similar to Taft CTD case, is predicted to be stacked at a maximum beach slope of
3%. As mentioned earlier this is the maximum beach slope at the top of the stack and
gradually reduces as it gets closer to the bottom of the stack and the embankment. There
are very few social issues associated with the location of the TSF.
The total throughput at the concentrator is 7 Mtpa with a life of mine of 20 years which
would result in production of 140 Million tons of dry tailings. An overall plan of the TSF
in this study is presented in Figures 5.
Figure 5. TSF plan for Dareh Alou mine
Tailings pipeline energy requirement at conceptual study stage was assumed to be similar
to those of the Taft project.
III- 41
Precipitation at the location of the TSF is not as low as for the Taft project with, an average
annual rainfall of 251mm/year. Also, the amount of pan evaporation is not has high as that
in Taft, and is recorded at 1,930 mm/year.
In the Dareh Alou project, water is not as scarce as at Taft, and based on previous
experience, it is anticipated that filtration would not be attractive, hence three tailings
dewatering and transportation options were investigated. These options included; paste
thickening at the concentrator location, paste thickening at the TSF location plus the option
of a paste thickener at the TSF and a high rate thickener at the concentrator.
Similar to the Taft project in cost estimate and analysis, 2011 rates are used with a rate of
return equal to 8%. The result of the cost estimateis presented in Table 3.
Table 3.Cost Estimation Results at Dareh Alou TSF
Option CAPEX
(USD)
OPEX
(USD/year)
NPV (USD
Paste thickener at concentrator 143M 3.6M 122M
High rate thickener at
concentrator and Paste at TSF
148M 5.3M 189M
Paste thickener at TSF 143M 3.9M 123M
On this basis, the lowest annual equivalent cost of tailings depositionis $1.8/ton.
As mentioned earlier, since the availability of water at Dareh Alou was not as problematic
as for the Taft project, and also, since all the proposed options present the same amount of
fresh water requirement, the cost of water supply and transportation have not been included
in the study.
5. DISCUSSIONS
In the Taft project the TSF, regardless of the type of deposition, is located in flat terrain in
an arid area. The distance between the concentrator location and the TSF is long enough
that the extra cost of additional high rate thickeners could be compensated by the savings
in the reduced size of tailings and return water transportation. On the other hand, the
elevation difference between the concentrator and the TSF was such that gravity flow of
the partially thickened tailings (underflow from a high rate thickener) was viable. On this
basis, and if the cost of water is excluded, the central thickened discharge would present
the lowest net present cost. However, water scarcity is severe, so much so that sufficient
water may simply not be available at all. In this case, the only remaining tailings
deposition option would be tailings filtration. In the case of tailings filtration, in order to
reduce the cost of tailings and water transportation, utilization of high rate thickeners at the
concentrator is suggested. If the environmental and social constraints were removed, some
saving in the filtration option could be achieved by moving the TSF location closer to the
concentrator.
In the case of the Dareh Alou project, the distance between the concentrator and the TSF is
such that transportation of un-thickened tailings to the paste thickener adjacent to the TSF
is the most effective option. Also, despite that fact that relatively large embankments
would be required in the Dareh Alou project, thickening the tailings could still provide
III- 42
benefits in reducing the construction cost by adopting a CTD scheme. This is made
possible due to the special topographical conditions at the Dareh Alou TSF site.
6. CONCLUSIONS
The conclusions that are derived from these studies are as follows:
There is no single answer for tailings and water management projects. Each project
is unique and thorough, specific study is required.
Factors that may have significant impact on the preferred tailings and water
management projects are; site topography, environmental constraints, climate
conditions, social sensitivities, distance from the concentrator and tailings
properties.
With increasing importance of water saving and scarcity, filtration gets more and
more attention, and there are cases in which filtration may become the viable option
if the cost of supply and transportation of fresh water is included in the study.
If the TSF is far enough away from the concentrator, two stages of dewatering
might be the most viable option.
With advancements in thickening technologies and based on the recent researches
and understanding of the beach slope, stacking of thickened tailings in the Central
Thickened Discharge and Down Valley Discharge options provide the most viable
options.
In some cases the clients prefer to select the option with lower up-front capital cots.
This is usually to be able to overcome the cash flow problems prior to the
commissioning of the plant.
ACKNOWLEDGEMENT
The authors wish to express their gratitude for the support and cooperation of the managers and
engineers at NICICO’s integrated water master plan technical office, ATC Williams and Middle
East Water and Environment.
REFERENCES
Pirouz, B and Williams, P. (2007):“Prediction of Non-Segregating Thickened Tailings
Beach Slope – A New Method”. Proceedings of the 10th International Seminar on Paste
and Thickened Tailings, Fremantle, Western Australia.
Fitton, T.G. and Bhattacharya, S.N. (2006):“Tailings Beach Slope Prediction. A New
Rheological Method”.International Journal of Surface Mining, Reclamation and
Environment. Vol. 20, 93, pp 181-202.
III- 43
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST – 6TH , 2014
Comparison of tailings dams dynamic response in case of central and
downstream method of construction
Ljupcho Petkovski, Professor, PhD, BSc. Civ. Eng. Ss. Cyril and Methodius University, Civil Engineering Faculty in Skopje, Republic of Macedonia
Stevcho Mitovski, MSc., BSc. Civ. Eng. Ss. Cyril and Methodius University, Civil Engineering Faculty in Skopje, Republic of Macedonia
ABSTRACT: The tailings dams are complex engineering structures, composed of: initial (starter) dam, sand dam, deposit pond, drainage system, water conveyors for cleared water conduction and structures for protection in case of incoming external water. The tailings dams along with the enormous volume of sediment’s lake are structures with highest potential hazard for the surrounding. A numerous tailings dams had a break or suffered enormous displacements during past earthquakes. Namely, the first main reason for tailings dam break is overflow, while the second is the action of earthquakes, causing tailings dams break at around 17% of the total number of breaks. The aim of this research is to contribute on the understanding of tailings dams behavior on action of strong earthquakes, by comparison analysis of the seismic response of the tailings dam constructed by different construction method. In this paper are presented results and conclusions from the comparison analysis (tailings dams alternatives with central and downstream method of construction) of the dynamic response of tailings dam no. 4 of lead and zinc mine Sasa, located in the north-east part of Republic of Macedonia. This region, as part of the Western Balkan, is seismic active area with maximal intensity of VIII-th degree for the expected earthquake for return period T = 1,000÷10,000 years, magnitude M ≈ 6.5, and peak ground acceleration PGA ≈ 0.35 g in case of Maximum Credible Earthquake. The analyzed tailings dam, currently at design stage, is planned with dam crest width of 5.0 m, downstream slope of 2.7 and height of 79.0 m measured from the tailings dam crest to the downstream toe.
Keywords: tailings dams, dynamic analysis.
1. INTRODUCTION
The tailings dam, imposing large volume of deposit pond, are structures with highest potential hazard on the environment. The similarities between tailings dams and convectional embankment dams (creating water reservoirs) have contributed many of the procedures and techniques at designing, construction and maintenance of the embankment dams to be applied also at tailings dams, thus improving their safety. But, numerous reports on accidents at tailings dams in the last three decades worldwide and also in Republic of Macedonia, are indicating on the ascertainment that structural (Petkovski L., Tančev L., Mitovski S., 2007;, Petkovski L., Tančev L., Mitovski S., 2013), dynamic (Daghigh Y., ..., 2005. Petkovski L., Paskalov T., 2003; Petkovski L., 2005), hydrologic
III- 54
and hydraulic safety is not secured by same strictness – as for embankment dams (Wieland M., Malla S., 2002; Seid-Karbisi M., Byrne P.M., 2004; Petkovski L., Tančev L., 2003). If we take in consideration following facts for the tailings dams: (1) there is no terminal control on the dam quality by applying the procedure of “first filling of the reservoir”, (2) there is no bottom outlet on eventual reservoir emptying and (3) in case of eventual dam break, beside the human victims and material damage, there is also a lasting degradation of the environment in the downstream valley. This requires the strictness for estimation of the tailings dams safety to be on higher level, compared with conventional dams.
The purpose of this study is to contribute to the understanding of the tailings dams behavior, constructed by different methods, caused by earthquake action. In this paper are presented the results and conclusions of the comparative analysis of the dynamic response in time domain (Salehi D., Mahin Roosta R., 2005; Matsumoto N., ..., 2005.) of the tailings dam Sasa no. 4, according to alternative construction methods – central and downstream method. In composition of the tailings dams of mine Sasa, M. Kamenica, located in north-eastern part of Republic of Macedonia, along the valley of Saska river, up to now are constructed 4 tailings dams, by application of the downstream method of construction (Petkovski L., Ilievska F., 2010.08). Currently, is active deposit pond no. 3-2, for which creation is constructed downstream tailings dam no. 3-2 with designed level of 975.0 masl. In order to use all available space along Saska river in the future service period of mine Sasa (after 2016), the following basic parameters of the tailings dam no. 4 are adopted: (а) crest elevation 952.0 masl, (b) deposit lake elevation 950.0 masl and (c) location of the dam site axis – most downstream from dam no. 3-2, limited in order all civil engineering structures to be placed in the zone of urban coverage of mine Sasa.
By such choice of parameters the downstream slope of tailings dams no. 4 is at vicinity of settlement. Due to the vicinity of the settlement (downstream from the zone of tailings dams no.4), in order to protect the inhabitants more efficiently from the air pollution, it is necessary to start the cultivation process of the tailings dams downstream slope as soon as possible. The initial thoughts for the alternative with downstream method of construction, as mostly acceptable solution from ecological point of view, are rejected, due to the very active earthquake region and this type of method obtains lowest seismic resistance of the structure. Therefore it is analyzed alternative differencing from the so far practice of downstream construction, apropos applying the method of central advancement of the sand dam. At such alternative, by embedding the rock material (from mine excavation) in the downstream shell, the final downstream slope can be formed practically in the same time with the raising of the dam crest elevation. It will enable in the initial stage of the dam construction to start the cultivation of the downstream slope, much more favorable from ecological point of view, compared with the alternative on downstream advancement of the sand dam. Namely, at downstream method of construction of the dam, by advancing in skew layers, the cultivation can start after the completion of the service period, apropos after the reaching of the final dam crest and last skew layer on the downstream slope. Therefore two alternatives are envisaged: (a) alternative no. 1 - downstream method of construction (proved as safe solution in the previous 4 tailings dams) and (b) alternative no. 2 – modified central method of construction. If the structural analysis confirms that alternative no. 2 poses the required safety and even if it is more expensive (at reasonable amount), then it will be adopted as most favorable solution, due to the possibility of faster cultivation of the downstream slope, apropos securing more acceptable level of inhabitants protection from air pollution.
III- 55
2. REPRESENTATIVE CROSS SECTION FOR STRUCUTRAL ANALYSIS AND AUTHORITATIVE MATERIAL PARAMETERS
The tailings dams cross section with maximal dimensions is adopted as representative section for dynamic analysis. For this section are foreseen several approximations, contributing on simplification of the numerical experiment, and not decreasing the analysis safety. The cross section simplification is done with the following:
i. The rock foundation is composed of gneiss at around 100 times higher deformable properties compared with the deposit (and also to the materials in the tailings dam), and in the analysis is treated as non-deformable and rigid boundary condition.
ii. The river bed deposit has variable depth, at around 27 m upstream of the dam (in near by of the upstream toe), 25 m at tailings dam site axis (dam crest) and 23 m in near by of the downstream toe of the sand dam.
iii. The river deposit will be cut in the initial dam axis by plug with depth of 8 m and bottom width of 3 m, thus securing the required casual seepage strength in the alluvium.
iv. The initial dam will be constructed as symmetrical homogeneous dam, made of graphite shale, with drainage blanket in the downstream toe, crest width of 3 m, crest elevation at 906.0 masl (due to securing of reserved volume in case of flood with return period of 20 years) and slope m = 1.5.
v. The sand dam will be created up to elevation 952.0 masl, (2.0 m above deposit lake), with crest width of 5 m, and slopes: upstream m1=1.5 (alternative no. 1) and downstream m2 = 2.7 (for both alternatives).
vi. The influence of the deposit lake (up to elevation of 950.0 masl) and the river deposit upstream of the dam site will be analyzed at section of 100.0 m upstream of the dam site, and the influence of the river bed deposit will be analyzed at section at around 40 m downstream of the downstream toe of the sand dam.
In such a way is prepared the idealized cross section for the dynamic analysis. The heterogeneous composition of the tailings dams is modeled by 6 different materials, fig. 1 and fig. 2.
Figure 1. Representative cross section for downstream method of construction (1) alluvium, (2) initial dam of shale, (3) tailings sand and (4) sludgein the deposit lake.
1
2
34
884.4
952.0950.0
906.0
882.0 873.0888.4879.4 874.6
Distance [m]
0 50 100 150 200 250 300 350 400
Ele
vatio
n [m
]
840
865
890
915
940
965
III- 56
Figure 2. Representative cross section for central method of construction (1) alluvium, (2) initial dam of shale, (3) tailings sand, (4) sludgein the deposit lake, (5) mix of tailings sludge and sand and (6) rock from mine excavation.
The adoption of the strength, deformable and water impermeable properties of the materials is based on number of terrain and especially laboratorial testing (three axial and oedometar testing). The adopted values for the geomechanical parameters, applied as input data in the analysis, are systemized in Tab. no. 1.
Table 1. Basic geomechanical parameters of the materials.
no. dim. 1 2 3 4 5 6
material
gravel shale tailings
sand tailings sludge
mix sludge-sand
mine rock
element foundation initial dam lake lake dam γspec kN/m3 26.5 27.0 32.0 31.0 31.5 32.0γdry kN/m3 19.0 19.2 18.0 15.0 16.5 20.0n 0.283 0.289 0.438 0.516 0.476 0.375
ω < ωsat % 8.0 8.0 10.0 15.0 12.0 8.0γsat kN/m3 21.8 22.0 22.3 20.1 21.2 23.7γ kN/m3 20.5 20.7 19.8 17.3 18.5 21.6φ o 37.0 31.0 32.0 15.0 20.0 33.0c kN/m2 0.0 15.0 0.0 5.0 2.0 0.0
k_s m/s 2.0E-05 1.0E-07 2.0E-06 2.0E-07 7.0E-07 1.0E-05Ko(φ) = 0.40 0.48 0.47 0.74 0.66 0.46ν(Ko) = 0.28 0.33 0.32 0.43 0.40 0.31
Mv compr kN/m2 37,000 35,000 30,000 22,000 25,000 40,000
3. INPUT DATA FOR THE DYNAMIC ANALYSIS
The analysis of the dynamic stability of the tailings dams Sasa is elaborated in accordance with the current standards on aseismic designing in Republic of Macedonia. In accordance with geographic coordinates of the location of the dam site of tailings dam Sasa – M. Kamenica, by seismological maps of Republic Macedonia, the dam is in moderate seismically active area. The maximal intensity of the expected earthquake at the location, according to MKS-64, for return period T = 1,000÷10,000 years is VIII degrees. In accordance with the scales on comparison of the earthquake intensity and magnitude, for
1
2
34 5
6
952.0950.0
906.0
878.0873.0
917.0
884.0 880.4 877.0
910.0902.5
895.0887.5
Distance [m]
0 50 100 150 200 250 300 350 400
Ele
vatio
n [m
]
840
865
890
915
940
965
III- 57
earthquake intensity of VIIIth degree corresponds magnitude M ≈ 6.5. By research of the dependence between the earthquake magnitude “M” and peak ground accelerations (PGA) for the wide region, where Republic of Macedonia belongs, if it is case of rock medium and attenuation of the seismic action on distance from the epicenter at around 5 km, it can be assumed Maximum Credible Earthquake (MCE) – strongest possible earthquake, to be PGA ≈ 0.35 g. In accordance with Euro Code 8, (Eurocode 8, 2003; Wieland M., 2003), the time lasting of the earthquake excitation is in dependence of the PGA, and is adopted rounded at ± 5 s, and the vertical component of the acceleration is adopted at 2/3 оf the horizontal. In this analysis, for MCE with PGAX = 0.35 g, the time lasting is adopted t=25 s, and the vertical component is PGAy = 0.23 g. In line with the advanced practice (ICOLD 1989), the dynamic analysis in time domain is analyzed with several accelerograms: (а) synthetic – according to the norms in Macedonia (Paskalov T., Zelenović V., 1986), and (b) realistic scaled earthquakes – El Centro, 1940, Ulcinj, 1979. In the following are presented results of the dynamic response of the tailings dams, at action of synthetic earthquake (SIMQKE 1997) – generated according to the norms in Macedonia, fig. 3.
Figure 3. Time history (accelerogram) of the horizontal acceleration component for MCE earthquake with t2 = 25 s, PGAx2 = 0.35 g, Z2_MAC (synthetic, MK-norms)
In the analysis is applied equivalent linear model (Geo-Slope QUAKE/W, 2007), where shear modulus Gmax [kPa] is in dependence of mean effectives stress σ'm [kPa], according to the equation no.1:
= ( ) (1)
Exponent "n" in this relation increases by deformation increscent, and for materials in composition of the tailings dams is adopted n = 0.5. Modulus "K" for calculation of the maximal value of the shear modulus "Gmax", is elastic parameter, and therefore is not directly dependable of the strength parameters of the local materials (cohesions and internal friction), is estimated according to the following two procedures: First procedure for estimation of "K" regards the reference data (Kramer S.L., 1996.) according to the geomechanical description of the materials and numerous empiric relations (Salehi D., Mahin Roosta R., 2005) in dependence of the material type and compaction – expressed by coefficient of porosity, after authors: Kokusho and Esashi, 1981; Hardin and Black, 1968; Seed and Idriss, 1970. By the second procedure for calibration of "K" are used dependences between: (a) distribution of the mean effective stresses in internal part of the tailings dams from initial stress state, (b) registered values of the propagation velocities of the transversal seismic waves in the internal part of the tailings dam, obtained with refraction geophysical measurements (Aleksovski D, 2003.), and (c) empirical dependence between velocities of the transversal waves and elastic parameters of the materials, in accordance with the equation no. 2:
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
0 5 10 15 20 25
t [s]
Acc
[g]
PGA z2_mac
III- 58
= ( / ) (2)
The approximate non-linearity and non-elasticity of the dynamic parameters of the local materials (fig. 4) is assumed according to the properties of the materials and experience dependences taken over from technical literature in domain of geotechnical earthquake engineering – by dynamic testing of materials: clay, sand, gravel, alluvium and crushed stone by the authors: Seed (1986), Idriss (1990), Vucetic and Dorby (1991), Kokusho (1980), and Tanaka (1987), as well and with comparison with the material parameters from dynamic testing of the tailings dams (Seid-Karbisi M., 2005).
Figure 4. Reduction of shear modulus G/Gmax and increase of damping ratio DR in dependence of the cyclic shear strains CSS
4. RESULTS OF THE DYNAMIC ANALYSIS
The dynamic response of the tailings dams, by application of equivalent linear analysis, is displayed by acceleration accelerograms (fig. 5 and 6) and response spectra of the accelerations (fig. 7 and 8) in the upstream edge of the dam crest, node with coordinates X=97.5m, Y=952 m.
Figure 5. Absolute horizontal acceleration, alternative no. 1 a[g]÷t[s], Z2_МАC, with PCA=0.447g
G /
Gm
ax R
atio
Cyclic Shear Strain (%)
0
0.2
0.4
0.6
0.8
1
0.001 1000.01 0.1 1 10
1
2
3
4
5
6D
ampi
ng R
atio
Cyclic Shear Strain (%)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.001 1000.01 0.1 1 10
X-A
ccel
erat
ion
(g)
Time (sec)
-0.1
-0.2
-0.3
-0.4
0
0.1
0.2
0.3
0.4
0.5
0 10 20 30
III- 59
Figure 6. Absolute horizontal acceleration, alternative no. 2 a[g]÷t[s], Z2_МАC, with PCA=0.574g
Figure 7. Response spectra for alternative no. 1, Sa[g]÷t[s], for DR=0.05, for acceleration in horizontal direction, Z2_MAC, in the foundation and in the crest
Figure 8. Response spectra for alternative no. 2, Sa[g]÷t[s], for DR=0.05, for acceleration in horizontal direction, Z2_MAC, in the foundation and in the crest
The estimation of the permanent deformations during excitation, caused by the internal dynamic forces, is done by application of Newmarks’ analysis (Geo-Slope SLOPE/W 2007; Paskalov T., 1985; Petkovski L., 2007.). In this analysis, based on the stability factor and dynamic internal forces, for each time period of the dynamic excitation, is calculated the mean acceleration for the total potential sliding body and dependence of the stability factor in function of the acceleration is determined. In case where the stability factor equals F=1.0, acceleration of fracture or creep is obtained. In case when the mean acceleration exceeds the creep acceleration, the mass would slide (fig. 9 and fig. 10).
X-A
ccel
era
tion
(g)
Time (sec)
-0.2
-0.4
-0.6
0
0.2
0.4
0.6
0 10 20 30
X-S
pect
ral A
ccel
erat
ion
(g)
Period (sec)
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 0.5 1 1.5 2
X-S
pect
ral A
ccel
erat
ion
(g)
Period (sec)
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0 0.5 1 1.5 2
III- 60
Figure 9. Time history of permanent displacements for critical sliding surface for alternative no .1, Z2_MAC, Xdisp = 0.381 m
Figure 10. Time history of permanent displacements for critical sliding surface for alternative no .2, Z2_MAC, Xdisp = 0.183 m
5. CONCLUSION
The engineering estimation on the degree of endangerment of the tailings dams in dependence of the value and geometry of the permanent displacement is based on the value of deformations at sand dam crest. Finally, at tailings dams no. 4 of mine Sasa, an earth sand dam, an uncontrolled emptying of the deposit lake would appear when: (1) the critical sliding surface of the slope (area with highest permanent displacements caused by earthquake), passes at least 2 m below the crest and if the value of the displacement is bigger than the crest width (b = 5 m) and/or (2) the settlements are bigger than the height from crest (952.0 masl) to the highest elevation of sludge in the lake (950.0 masl), apropos bigger then h=2.0 m.
The mechanism conditioning sliding of the downstream slope of the sand dam results from increase of the shear stresses at action of the dynamic inertial forces during earthquake action. If these stresses, in some zones, exceed the material shear resistance (most often in relatively shallow sliding surfaces), cause deformations, so the cumulative effect of these deformations is permanent deformation during the seismic excitation caused by increase of the tangential stresses. From the analysis of the dynamic stability of the downstream slope of the dam, at action of MCE, Z2-MK, it can be noticed that minimal value of the stability factor only in few cases is less then F=1.0. So at action of catastrophic earthquake small permanent displacements will appear (caused by increase of the shear stresses), that would be manifested as longitudinal cracks along the crest (or the slope) of the sand dam. These displacements, for action of earthquake Z2-MK, for alternative no. 2 (with downstream
Def
orm
atio
n
Time
0
0.1
0.2
0.3
0.4
0 10 20 30
Def
orm
atio
n
Time
0
0.05
0.1
0.15
0.2
0 10 20 30
III- 61
construction method) are 38.1 cm, and for alternative no. 2 (with central construction method) has a value of 18.3 cm.
In the dynamic analysis is done approximate calculation of the displacements in the dam due to additional compaction of the stresses redistribution for the local materials – exposed on cyclic action. The maximal settlement from these displacements, for action of earthquake Z2-MK, for alternative no. 1 (with downstream construction method) is 66.4 cm, and for alternative no. 2 (with central construction method) has a value of 24.2 cm.
At earthquake action is possible appearance of liquefaction in the water saturated zones of the tailings sand in the dam (at eventual plugging of the drainage structure), by increase of the pore pressure up to dPw = 94.4 kPa (for alternative no. 1), and dPw = 12.1 kPa (for alternative no. 2) for action of earthquake Z2-MK. But, composition of the cross section of the tailings dams, conditioned by the adopted technology of advancement of the sand dam (in downstream direction) enables to attain the stability of the downstream slope, at action of the excess pressure at the liquefiable material – directly after the earthquake action. By the slow hydrodynamic process of dissipation of the excess pressure in the pores of the tailings sand (till the once again establishment of the initial steady regime), some displacements in the dam body can be expected, The maximal settlement by these displacements, at action of earthquake Z2-MK, for alternative no. 1 (with downstream construction method) is 37.4 cm, and for alternative no. 2 (with central construction method) has a value of 42.4 cm.
The general conclusion from the dynamic analysis of tailings dams Sasa no. 4 with crest elevation 952.0 masl would be that heterogeneous medium, with adopted geometry and distribution of the materials, possess the required seismic resistance, and there is no disruption of the dynamic stability of the sand dam – not during the earthquake excitation, nor immediately after its termination. By the displacements caused from catastrophic earthquake (value of 141.9 cm for alternative no. 1 or 84.9 cm for alternative no. 2) the height of 2 m is not exceeded (from dam crest 952.0 masl till the highest elevation of the tailings sludge in the lake at 950.0 masl), so there is no risk for rapid leakage of the sludge from the deposit lake.
REFERENCES
Aleksovski D., (2003). „Defining of dynamic properties of the materials for embankment dams by geophysical testing in natural scale“, PhD dissertation, IZIIS, Skopje
Daghigh Y., Davoudi M.H., Shokri A.(2005) "Nonlinear dynamic analysis of earth dams using Diana code: (a case study in Alavian dam)", 73rd Annual Meeting of ICOLD, Tehran, IRAN, Paper No.:160-W4,
Eurocode 8, (2003), Design of structures for earthquake resistance, Doc CEN/TC250/SC8/N335, DRAFT No 6, Brussels
Geo-Slope QUAKE/W, 2007."Dynamic modeling", Geo-Slope SLOPE/W 2007. "Stability analysis", GEO-SLOPE International Ltd., Calgary, Alberta, Canada
ICOLD (1989), Selecting Seismic Parameters for Large Dams, Guidelines, Bulletin 72, Committee on seismic aspects of dam design,
Kramer S.L., (1996). "Geotechnical Earthquake Engineering", Prentice Hall, New Jersey, USA
Matsumoto N., ..., (2005). "Analysis of strong motions recorded at dams during earthquakes", 73rd Annual Meeting of ICOLD, Tehran, IRAN, Paper No.: 094-W3
III- 62
Paskalov T., (1985). "Earthquake induced deformations on earth-fill and rock-fill dams", International Journal "Soil Dynamics and Earthquake Engineering", Vol.4m No.1, CML Publications, UK, p35-42
Paskalov T., Zelenović V., 1986., "Normative of technical standards for design and computation of engineering structure in seismic areas", Belgrade
Petkovski L., (2005). “Dynamic Analysis of a Rock-filled Dam with Clay Core“, International Conference IZIIS 40 EE-21C, Skopje/Ohrid Macedonia, Proceedings, CD-ROM
Petkovski L., (2007). “Seismic Analysis of a Rock-filled Dam with Asphaltic Concrete Diaphragm“, 4th International Conference on Earthquake Geotechnical Engineering, Thessaloniki, Greece, CD-ROM;
Petkovski L., Ilievska F., (2010.08) “Comparison of Different Advanced Methods for Determination of Permanent Displacements of Tailings Dams in Earthquake Condition“, 14th Europian Conference on Earthquake Engineering, Ohrid, R.Macedonia, paper #1511, CD-ROM;
Petkovski L., Paskalov T., 2003. “Comparison of Dynamic Analyses of Embankment Dams by Using Lumped Mass Method and Finite Element Method“, International Conference in Earthquake Engineering - Skopje Earthquake - 40 Years of European Earthquake Engineering, Skopje, Ohrid, R.Macedonia, Proceedings, CD-ROM
Petkovski L., Tančev L., 2003. “Dynamic Analysis of a Rock-filled Dam with Geosynthetic Screen”, International Conference in Earthquake Engineering - Skopje Earthquake - 40 Years of European Earthquake Engineering, Skopje, Ohrid, Republic of Macedonia, Proceedings, CD-ROM,
Petkovski L., Tančev L., Mitovski S., (2007). “A contribution to the standardization of the modern approach to assessment of structural safety of embankment dams", 75th ICOLD Annual Meeting, International Symposium “Dam Safety Management, Role of State, Private Companies and Public in Designing, Constructing and Operation of Large Dams”, St.Petersbourg, Russia, Proceedings p.66, CD-ROM
Petkovski L., Tančev L., Mitovski S., (2013) "Comparison of numerical models on research of state at first impounding of rockfill dams with an asphalt core", International symposium, Dam engineering in Southeast and Middle Europe - Recent experience and future outlooks, SLOCOLD, Ljubljana, R.Slovenia, ISBN 978-961-90207-9-1, Proceedings, 106-115
Salehi D., Mahin Roosta R., (2005). "Evaluation of different methods for dynamic stability analysis of embankment dams", 73rd Annual Meeting of ICOLD, Tehran, IRAN, Paper No.:063-W4,
Seid-Karbisi M., (2005). "Seismic stability of embankment dams", 73rd Annual Meeting of ICOLD, Tehran, IRAN, Paper No.: 113-W4
Seid-Karbisi M., Byrne P.M., (2004). “Embankment dams and earthquakes”, Hydropower & Dams, Issue Two
SIMQKE 1997, manual of computer program for simulation of acceleration time-history of synthetic earthquakes, from target response spectrum
Wieland M., (2003). "Seismic Aspects of Dams", General Report of Question 83, ICOLD, Montreal, Canada
Wieland M., Malla S., (2002), “Seismic Safety Evaluation of a 117 m High Embankment Dam Resting on a Thick Soil Layer”, 12th European Conference on Earthquake Engineering, London, Paper Reference 128
III- 63
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST
– 6TH
, 2014
Assessment of static and seismic stability of
Kumtor’shdTTjjhkljdjjsgshjhfsdkjhskslsl;s;s;;s;;s;;sjsjkjffffrtttttttfggjfgjgkfkjkj
f fffffjfjjfkkfjjj
Kumtor’s gold mine tailings dam
2(14pt)
in Kyrgyz Republic
B.A. Chukin & R.B. Chukin Institute of Geomechanics and Development of Mineral Resources (IGDMR), Bishkek, Kyrgyz Republic
ABSTRACT: Kumtor gold mine is situated in the Kyrgyz Republic in Central Tien Shan Mountains at an altitude 4000 meters in permafrost area. Construction and exploitation of the tailings dam was started in
1995. In 1999 the displacement of the dam to downstream side was detected. The dam height was
20 meters. Analysis of monitoring data showed that displacement took place in ice rich loamy layer in the foundation on 4 meters depth. To stop the displacement the decision was made to excavate
loamy layer beyond downstream and change it by construction shear key made of macro
fragmental soil. The depth of shear key was 5 meters. In the following the monitoring data showed
that tailings dam continue to move on underlying soils. The additional geological investigation was done. It showed that more solid soils were located on the depth from 10 to 12 meters. Numerical
modeling of the dam was made in FLAC codes. The methodology of displacement stoppage was the
same. Rheological parameters of the soils in numerical model calibrated on the basis of back analysis. Forecast calculations were made to 2016 when the dam height would be 42.7 meters.
Also the assessment of seismic stability was made with consideration of layered foundation.
Calculation in FLAC codes showed the influence of soils condition. Peak ground acceleration was increased. Worked out measures to stop the tailings dam displacement was accomplished in the
period from 2006 to 2010. The new monitoring data of displacement confirm the efficient of the
taken actions.
1. INTRODUCTION
1.1. Overview
The tailings dam is located in the bed of Ara-Bel River. Water of the river was redirected
around the tailings dam through upper bypass canal with returning to the former bed below
the dam. The tailings dam was designed with filling method of wastes lying. Dam filling
was started in 1995. The dam is raised by stages to the downstream side. The dam body is
filled with macro fragmental soil. The upstream and downstream sides are formed with
gradient 1:3. There is impervious screen with the length 100 meters which is lying along
the upstream and the bottom of reservoir. It’s made of polyethylene film of high density
with thickness 1.5 millimeters. General view of the dam is shown in Figure 1.
III- 64
Figure 1. Kumtor tailings dam
1.2. Description of the Problem
In 1998, holes for installation inclinometers were drilled. At that time maximum height of
the dam was 20 meters. Data of field observations showed that dam was moving to
downstream side. Allocation of horizontal offset showed that displacement was caused by
ice rich loamy layer in the foundation. In order to stop displacement the decision was made
to remove loamy layer beyond downstream side and change it by construction of shear key
from macro fragmental soil. In 2003, works for arrangement of shear key were performed.
At that time dam height was 24.7 meters. Trench for shear key with depth 5 meters and
slope ratio 1:1 had a length about 20.5 meters. Shear key depth was determined on the
basis of analysis of horizontal offsets with consideration for meter long incut in soil which
had no any displacement. Cantledge with 5 meters height was dumped on the shear key.
Soil for organization of shear key, cantledge and dam body was selected from one pit. In
the following the data of field observations showed that the measures which had been done
did not lead to stoppage of the dam displacement. Temporary group of experts from
Canadian consulting firm BGC Engineering INC and IGDMR was created to solve this
problem. The following are results of IGDMR. Tailings dam numerical model was
performed in FLAC codes as a creep model. Regulatory requirements acting in Kyrgyz
Republic about assessment of stability based on the value of factor of safety. In this case
the factor of safety was calculated after creep modeling had been finished.
2. NUMERICAL MODELING
2.1. Design and Calibration of Numerical Model
Numerical modeling was performed using the program FLAC (Itasca 2011). The important
stage of numerical modeling is selection of soil model, describing the relation between
creep deformation and relaxation of stress. In the result of analysis of different models for
description of rheological processes in soil, Norton’s power law dependence was selected.
Stress-strain analysis with consideration of rheological processes was based on the
comparison of calculated displacements and monitoring data. Monitoring of displacement
was made on the basis of inclinometers. Figure 2 (a) describes observed displacements
according to inclinometer INC98-1 in period from 21.12.01 to 13.08.05.
III- 65
Figure 2. (a) INC98-1 inclinometer data; (b) allocation of deformation shift in loamy layer
The depth in meters from the dam crest is on vertical axis and total displacements in meters
are on horizontal axis. The figure shows that the most intense displacements occur between
22 meters and 24 meters marks. Original ground level corresponds to 20 meters mark.
Loamy layer has a various width from 6 to 10 meters. The largest deformation shift is
observed in upper part of it within the limits of 2 meters. The roof of loamy layer is
situated at a depth of 2 meters from ground surface. All shift indexes above this mark are
virtually the same. That means that the dam body and two meters of natural soil above the
loamy layer have lesser deformation shift. In order to find out on which depth and with
what intensity deformation happens, schedule of inclinometer knees was built relative to
each other and provided on Figure 2 (b). This figure shows that the largest deformation
shift is registered at a depth from 2.5 to 4 meters (mark 22.5 - 24 meters). The most
deformable layer has width about 1.5 meters. Below 4 meters (mark 24.0 meters)
deformation shift drops evenly. Formation of rheological model is based on separation of
calculated rheological layers within the limits of whole loamy layer with a capacity of 9
meters. In view of the fact that shifts in loamy layer occurred before installation of INC98-
1, it is necessary to try to restore accumulated shifts from the beginning of the dam
construction. Difference in shifts between layers defines by the value of deformation and
this is the most important for assessment of resistance. Figure 3 provides restored shifts in
Figure 3. Horizontal displacement of loamy layers at depth from 2.0 to 4.5 meters based on INC98-1
(a) (b)
III- 66
separated layers of loamy layer on the basis of monitoring data. The results of layer shift
approximation with coefficient of determination R2 = 0.995 are also showed at Figure 3.
Layer shift at the stage of the dam upbuilding during the period from 1999 to 2005 occur in
a linear fashion. Model calibration was carried out by variation of rheological parameters
А and n according to Norton’s method. In this case n parameter is considered as constant
and equal 3. The most exact approximation of the results was obtained by rheological
model consists of six calculated layers which are provided on Figure 4. Figure 5 displays
comparison of calculated displacement and observation data according to INC98-1 to 2004
inclusively. The calculation data matched well with monitoring data obtained on basic
shifting layers, situated at a depth 22.5, 23.0 and 23.5 meters. Calibration test of 2004 was
associated with the fact that in 2008 shifts have some departure from linear fashion.
Figure 4. Calculated layers of dam foundation for 2004
Figure 5. Comparison of numerical results with monitoring data according to INC98-1 for 2004
The correction of А index in the middle layer in 2004 allowed reaching the optimal
approximation. In view of the fact that the last model hadn’t correction of A index, it can
believed that all the following calculations have prognostic character.
2.2. Forecast of Displacement and Stability Assessment
The construction of the dam will be finished in 2016. Figure 6 displays the results of
horizontal displacement forecast by layers according to INC98-1 for 2016 inclusively.
Loamy layer, outspreading under the dam body and shear key of 2003 cannot be removed.
FLAC (Version 6.00)
LEGEND
19-Jan-13 13:49
step 33360
Creep Time 2.8383E+08
Table Plot
21-22 m inc98-1
22.5 m inc98-1
23.0 m inc98-1
23.5 m inc98-1
24.0 m inc98-1
24.5 m inc98-1
21.0 m * #3
22.0 m * #4
22.5 m * #5
23.0 m * #6
23.5 m * #7
24.0 m * #8
24.5 m * #9
5 10 15 20 25 30
(10 ) 07
0.500
1.000
1.500
2.000
2.500
3.000
3.500
4.000
(10 )-01
JOB TITLE : Correlation X-disp Inc98-1 2004
IGDMR
Kyrgyz Republic
III- 67
The key feature in character of horizontal displacement allocation is that the shifts in
loamy layer under the dam areas continue, and shifts in areas where it was removed stop.
Figure 6. Forecast results of horizontal displacements of the layers for 2016
Figure 7 provides total horizontal displacement for the end of 2016. The largest horizontal
offsets concentrated in the area where loamy layer cannot be removed. In this area total
horizontal offsets to the end of 2016 are less than 80 centimeters. In the area where loamy
layer was removed and shear key was built, horizontal offsets are less than 20 centimeters.
It points to the fact that stopping measures for dam shifting caused by rheological
processes in loamy layer are effective. The question how to calculate factor of safety for a
Figure 7. Horizontal displacement for the end of 2016
a model of geotechnical object with rheological processes remains open. For evaluation of
overall stability we use the following method. The main differential characteristic from
early performed calculation of dam stability is that now we take into account alteration of
soils strength properties for loamy layer from deformation shift value, obtained as a result
of rheological process modeling. In other words, stability of the dam is evaluated with
consideration of time effect. The basis of coefficient of stability calculation has two
principal points:
If a structure is a subject to continuous rheological processes, then the moment of
beginning of stability loss will be defined not by time rheological processes
duration, but formation of surface of failure, along which soil reached critical
deformation;
FLAC (Version 6.00)
LEGEND
19-Jan-13 14:12
step 72156
Creep Time 6.6226E+08
Table Plot
21-22 m inc98-1
22.5 m inc98-1
23.0 m inc98-1
23.5 m inc98-1
24.0 m inc98-1
24.5 m inc98-1
21.0 m * #3
22.0 m * #4
22.5 m * #5
23.0 m * #6
23.5 m * #7
24.0 m * #8
24.5 m * #9
10 20 30 40 50 60
(10 ) 07
1.000
2.000
3.000
4.000
5.000
6.000
7.000
(10 )-01
JOB TITLE : Correlation X-disp Inc98-1 2016
IGDMR
Kyrgyz Republic
FLAC (Version 6.00)
LEGEND
19-Jan-13 14:12
step 72156
Creep Time 6.6226E+08
-2.103E+02 <x< 2.843E+02
3.683E+02 <y< 8.630E+02
X-displacement contours
0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
6.00E-01
7.00E-01
8.00E-01
Contour interval= 1.00E-01
Boundary plot
0 1E 2
3.750
4.250
4.750
5.250
5.750
6.250
6.750
7.250
7.750
8.250
(*10 2̂)
-1.750 -1.250 -0.750 -0.250 0.250 0.750 1.250 1.750 2.250 2.750
(*10 2̂)
JOB TITLE : X-disp 2016
IGDMR
Kyrgyz Republic
III- 68
As rheological processes of deformation shift in different points of structure are
various, overall stability of structure shall be calculated on the basis of deformation
shift distribution with determination of areas where deformation reached limit
values.
Zaretskiy points out that for soils there is a critical strain of form alteration. On the basis of
experiments with sand clay for Nurekskaya dam strain intensity was about 14%. This
circumstance for assessment of resistance upon calculation of creeping was realized as
follows. If it’s necessary to evaluate stability of the dam in certain period of time, then we
define the area in loamy layer, in which the amount of deformation reached limit value.
The limit value of deformation shift is 13%. This value was selected as the smallest among
all values provided in literary sources. It is accepted that this area doesn’t resist to shifting.
In this area strength properties cohesion and friction angle are set to zero. Special function
was developed in FISH programming language for exact area separation where soil doesn’t
resist shifting and maximum deformation reached value of 0.13. Figure 8 (a) provides
picture of maximum deformation shifts for 2016. The largest deformation shifts
concentrated in loamy layer. Figure 8 (b) as an example, describes friction angle
distribution for the same geometry after the function was used. After that we used standard
procedure for factor of safety calculation proposed by Dawson and Roth 1999. The
minimum factor of safety value is 1.36 for 2016 dam body. Note that shear-strain rate
contours doesn’t have exit to free surface and doesn’t form closed region.
Figure 8. (a) Maximum deformation shifts; (b) friction angle distribution
Kyrgyz Republic is a seismic active country. Seismic stability assessment of the dam was
based on dynamic theory. It was made on the basis of calculated deformations analysis.
The limit state of the dam is characterized by the presence of damage (vertical crest
displacement, cracks etc.) which can lead to disturbance of the dam with following
overflow of the pulp. We selected two criteria of seismic stability. The first criterion is
vertical dam crest displacement. Vertical crest displacement during earthquake must not
exceed reserve between the crest and the level of the water in upstream. The second
criterion is connected with shear-strains. This criterion is based on comparison of limit
shear-strains and shear-strains which evolving in dam body during earthquake. The
formation of surface of failure is related with the increase of plastic shear-strain increments
to value 2 – 5 %. Slope failure occurs when surface of failure cross the dam body and has
an exit to slope face. Thus, for overflow of the pulp this exit should be on upstream under
the water level. The earthquake motion is considered to occur when the reservoir level is at
fool pool. Pore-pressure calculation is important part, it determine pore-pressure
distribution on the upstream side of the dam and in the soils corresponding to the reservoir
elevation. Figure 9 displays the pore-pressure distribution and phreatic surface location
FLAC (Version 6.00)
LEGEND
19-Jan-13 14:14
step 72156
Creep Time 6.6226E+08
-6.927E+01 <x< 2.077E+02
4.772E+02 <y< 7.541E+02
Max. shear strain increment
0.00E+00
5.00E-02
1.00E-01
1.50E-01
2.00E-01
2.50E-01
Contour interval= 5.00E-02
Extrap. by averaging
Boundary plot
0 5E 1
5.000
5.500
6.000
6.500
7.000
7.500
(*10 2̂)
-0.250 0.250 0.750 1.250 1.750
(*10 2̂)
JOB TITLE : ssi 2016
IGDMR
Kyrgyz Republic
FLAC (Version 6.00)
LEGEND
19-Jan-13 16:29
step 72156
Creep Time 6.6226E+08
-8.630E+01 <x< 2.188E+02
4.712E+02 <y< 7.763E+02
friction
0.000E+00
1.500E+00
2.400E+01
3.500E+01
3.800E+01
4.400E+01
4.500E+01
Boundary plot
0 5E 1
5.000
5.500
6.000
6.500
7.000
7.500
(*10 2̂)
-0.500 0.000 0.500 1.000 1.500 2.000
(*10 2̂)
JOB TITLE : Friction 2016
IGDMR
Kyrgyz Republic
(a) (b)
III- 69
through the dam and foundation at steady state. The area where the dam was built
according to the confirmed “Map of seismic zoning of Kyrgyz Republic” classified as zone
Figure 9. Pore-pressure distribution and phreatic surface
with 8 points of intensity and repeatability 1 earthquake per 500 years. In the reports on
feasibility study of the project there is an analysis of Kumtor mine seismic ground motion.
This analysis based on methodology of evaluation of such factors as location of active
faults and their parameters and distance from these faults and construction site. As a result
peak ground acceleration for seismic analysis has a value 0.3g. For understanding the
influence of weak ice rich loamy layer on seismic stability the modeling of the foundation
was made. On the first stage the deconvolution analysis that is performed to obtain the
appropriate input motion was made without weak layer. On the second stage weak layer
was set. The dynamic simulation was made with three strong motions, one local
(Suusamyr) and two global (Kobe, CapeRio). The results of simulation with Kobe strong
motion earthquake are described further. It has duration 40.88 seconds and several high
peaks. The soil conditions of the dam foundation lead to dynamic amplification. Peak
ground acceleration increased from 0.3g to 0.34g. Also the character of frequencies was
changed. The Figure 10 displays acceleration time histories on the surface with (black line)
and without (red line) weak loamy layer. The Figure 11 displays power spectrum of
accelerations in three different locations of the model. For the better reading power
spectrum was converted from FLAC to Excel. Red line indicates power spectrum of input
Figure 10. Acceleration time histories on the surface of the foundation
Kobe acceleration. The dominant frequency is approximately 0.6 Hz, the highest frequency
component is less than 8 Hz, and the majority of the frequencies are less than 6 Hz. The
blue line is a power spectrum of acceleration on foundation surface. The highest frequency
FLAC (Version 7.00)
LEGEND
20-Feb-13 22:21
step 359600
Flow Time 7.0497E+07
3.155E+02 <x< 8.489E+02
3.357E+03 <y< 3.890E+03
Pore pressure contours
0.00E+00
1.00E+05
2.00E+05
3.00E+05
4.00E+05
5.00E+05
6.00E+05
7.00E+05
8.00E+05
Contour interval= 1.00E+05
Phreatic surface
Contour interval= 5.00E-01
Minimum: 0.00E+00
Maximum: 5.00E-01
Boundary plot
0 1E 2
3.400
3.500
3.600
3.700
3.800
(*10 3̂)
3.500 4.500 5.500 6.500 7.500
(*10 2̂)
JOB TITLE : PP and PS S2 2013
LSGO
FLAC (Version 7.00)
LEGEND
26-Feb-14 19:45
step 706397
Flow Time 7.0497E+07
Dynamic Time 4.0880E+01
3.152E+02 <x< 8.489E+02
3.357E+03 <y< 3.890E+03
Original Surface
Y-axis :
31 X acceleration( 358, 44)
X-axis :
100 Dynamic time
5 10 15 20 25 30 35 40
-2.000
-1.000
0.000
1.000
2.000
3.000
JOB TITLE : Acc_WI
LSGO
Kyrgyz Republic
FLAC (Version 7.00)
LEGEND
26-Feb-14 19:44
step 326299
Dynamic Time 4.0880E+01
3.155E+02 <x< 8.489E+02
3.340E+03 <y< 3.874E+03
HISTORY PLOT
Y-axis :
5 X acceleration( 151, 44)
X-axis :
1 Dynamic time
5 10 15 20 25 30 35 40
-2.000
-1.000
0.000
1.000
2.000
JOB TITLE : Acc_WO
LSGO
Kyrgyz Republic
III- 70
is 0.6 Hz and the second peak is 3.6 Hz. The black line is power spectrum of acceleration
on the crest of the dam. The highest frequency is 0.6 Hz and the second peak is 2.8 Hz.
Displacement of the dam after 40.88 seconds is primarily concentrated along upstream
slope. This is shown in the Figure 11. The maximum displacement is 50 centimeters.
Figure 10. Power spectrum of accelerations
Downstream side has a less value of the displacement due to the shear key efficiency.
Figure 11. Horizontal displacement contours
Figure 12 shows the plot of horizontal (green line) and vertical (blue line) displacements of
the crest. The maximum value of horizontal crest displacement is 42 centimeters and
vertical 3.3 centimeters. The value of maximum shear-strain occurred during seismic
loading is stored with special FISH function. The Figure 13 displays the maximum shear-
strain increment contours after 40.88 seconds. The maximum value of shear-strain is 50%
in weak loamy layer. This level of deformations shows that loamy layer reached the limits
and it doesn’t resist shifting anymore. The result of calculation showed that the condition
of slope failure, when the formed surface of failure crossed the dam body and had an exit
on slope face did not happen. None of the selected criteria of dam failure took place.
Seismic stability of the dam was confirmed.
FLAC (Version 7.00)
LEGEND
20-Feb-14 13:03
step 565409
Flow Time 7.0497E+07
Dynamic Time 2.3950E+01
3.152E+02 <x< 8.489E+02
3.357E+03 <y< 3.890E+03
X-displacement contours
-4.00E-01
-3.00E-01
-2.00E-01
-1.00E-01
0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
Contour interval= 1.00E-01
Boundary plot
0 1E 2 3.400
3.500
3.600
3.700
3.800
(*10 3̂)
3.500 4.500 5.500 6.500 7.500
(*10 2̂)
JOB TITLE : X-Disp
LSGO
Kyrgyz Republic
III- 71
Figure 12. Crest dam displacement
Figure 13. Maximum shear-strain increment contours after 40.88 seconds of seismic loading
REFERENCES
Ye.N. Bellendir, D.A. Ivashintsov, D.V. Stefanishin, O.M. Finagenov, S.G. Shulman
(2003): Probabilistic Methods of Assessment of Reliability of Soil Hydro-
Engineering Structures, St. Petersburg, Russia Federation.
C. Lomnitz and E. Rosenblueth (1976): Seismic Risk and Engineering Decisions, New
York, USA.
L. Suklje (1976): Rheological Aspects of Soil Mechanics, London, UK.
J.K. Zaretskiy and V.N. Lombardo (1983): Static and Dynamic of Soil Dams, Moscow,
USSR.
Seed, H. Bolton, and I. M. Idriss (1970): Soil Moduli and Damping Factors for Dynamic
Response Analysis, California, USA.
Seed, H. B., and I. Idriss (1969): Influence of Soil Conditions on Ground Motion During
Earthquakes, Journal of Geotechnical Engineering Division, ASCE, Vol. 95, pp. 99-
137, California, USA.
K.Ishihara (2006): Soil Behaviour in Earthquake Geotechnics, Tokyo, Japan.
L.H. Mejia and E.M. Dawson (2006): Earthquake deconvolution for FLAC, 4th
International FLAC Symposium on Numerical Modeling in Geomechanics, Itasca
Consulting Group, Inc., Minneapolis, USA.
International Commission on Large Dams (ICOLD) (1995): Tailings Dams and Seismicity,
Bulletin 98, 21-43 pp.
FLAC (Version 7.00)
LEGEND
27-Feb-14 13:03
step 706397
Flow Time 7.0497E+07
Dynamic Time 4.0880E+01
3.152E+02 <x< 8.489E+02
3.357E+03 <y< 3.890E+03
HISTORY PLOT
Y-axis :
52 reldispx (FISH)
53 reldispy (FISH)
X-axis :
100 Dynamic time
5 10 15 20 25 30 35 40
0.000
0.500
1.000
1.500
2.000
2.500
3.000
3.500
4.000
(10 )-01
JOB TITLE : Crest_Disp
LSGO
Kyrgyz Republic
FLAC (Version 7.00)
LEGEND
27-Feb-14 13:00
step 706397
Flow Time 7.0497E+07
Dynamic Time 4.0880E+01
3.152E+02 <x< 8.489E+02
3.357E+03 <y< 3.890E+03
EX_10 Contours
0.00E+00
1.00E-01
2.00E-01
3.00E-01
4.00E-01
5.00E-01
Contour interval= 1.00E-01
Boundary plot
0 1E 2
3.400
3.500
3.600
3.700
3.800
(*10 3̂)
3.500 4.500 5.500 6.500 7.500
(*10 2̂)
JOB TITLE : Max_ssi
LSGO
Kyrgyz Republic
III- 72
INTERNATIONAL SYMPOSIUM ON
Bali, Indonesia, June 1ST – 6
TH , 2014
Geotechnical Performance Evaluation of Sediment Dam
A Case Study on Fiona Dam at PT Vale Indonesia Tbk
Wiyatno Haryanto PT Vale Indonesia Tbk, Sorowako, Indonesia
Anom Prasetyo PT Vale Indonesia Tbk, Sorowako, Indonesia
ABSTRACT: This paper is intended to evaluate geotechnical performance of Fiona Dam as a sediment dam in terms of stability against seismicity since water filling stage until operational period. The dam is
located in the area of the mining operations of PT Vale Indonesia Tbk at Sorowako, East Luwu
Regency, South Sulawesi Province, Indonesia, constructed to impound solids of waste material (disposal) and also act as filter function before the mine effluent released to the downstream
waterbody. In this review, the analyses are referring to the detailed engineering design, survey &
monitoring serial data, seismicity calculation, and previous geotechnical site investigation report
through simulation and data analysis approaches.
Keywords: Sediment dam, solids of waste material, seismicity, peak ground acceleration (PGA),
geotechnical slope stability.
1. INTRODUCTION
The main purpose of the Fiona Dam construction as a sediment impoundment is to
containment of solids of fine-grained sediments or waste materials (disposal) by allowing
surface water passing the dam body (overtopping and seepages) and released to the
waterbody as mine effluent that must be complying with an environment quality standard
as regulated by Government of Republic of Indonesia, Ministry of Environment in
PERMEN-LH No.9/2006 regarding effluent quality standard of nickel ore mining industry
at eleven parameters such as TSS (total suspended solid), Cr+6
(chromium hexavalent),
Crtotal (total chromium), and some other metal minerals. [Blank line 10 pt]
Solid particles is derived from the operation of a nickel processing plant, mining face,
disposal, slag dump, revegetation and some exposed area in Anoa Hill in which transported
by surface runoff. Through the dam structure, sediment load transported by runoff contains
of suspended solid and floating particles shall be settling in the reservoir, thus clean water
shall be passing the dam structure with the whole parameters comply with the environment
quality standard. [Blank line 10 pt]
III- 73
Fiona Dam had been constructed in End of 2001, located on the position of 2°35’13” South
Latitude and 121°25’30” East Longitude with maximum elevation head about 22 m height,
design capacity 13,000,000 m3 with probable maximum flood 632 m
3/s, flood level on elv.
+396 m amsl and catchment area about 18.6 km2, dedicated to serve mining activities in
life of mine until 2035. The dam structure is categorized as rock-fill dam with drainage
core ripped slag and geotextile filter in front of core face. The dam crest is on elv. +397 m
amsl, 12 m width, 185 m length, upstream slope 40%, downstream slope 10%, and
spillway level on elv. +395 m amsl and 100 m width with flow capacity about 250 m3/s.
[Blank line 10 pt]
Layout of Fiona Dam is shown in the figure below; all of water effluents from mining
activities are trapped by some sediment ponds such as Diana Pond, Delaney Pond, and
Anoa Pond to settling solids before released to the reservoir. [Blank line 10 pt]
Figure 1. Aerial Photograph of Fiona Dam and Vicinity
[Blank line 10 pt]
Figure 2. Crest and Spillway of Fiona Dam
[Blank line 10 pt]
III- 74
Figure 3. Downstream Slope of Fiona Dam
[Blank line 10 pt]
Figure 4. Water Reservoir of Fiona Dam
[Blank line 10 pt]
[Blank line 10 pt]
2. DETAILED ENGINEERING DESIGN [Blank line 10 pt]
To evaluate technical performance of Fiona Dam referred to as a sediments dam; material
specification, detailed drawings, and standards are referring to technical report explained
below. [Blank line 10 pt]
2.1. Material Specification [Blank line 10 pt]
Fiona Dam is a rock-fill dam without impermeable core layer, constructed with rocky
material taken from quarry area consisting of peridotite rocks. Borrow material location is
relatively close to the dam site. Due to the dam is not equipped with impermeable core
layer, geotextile drainage is acting to replace it and as filter of granular particles passing
into the dam body. Material specification of the dam body is described below:
III- 75
� Riprap rock material is specified as selected boulder hard rock (size 450 mm, 600 mm
up to 1,000 mm), placed at upstream and downstream slope face as protection against
wave, abrasion, and scouring.
� Filter layer consists of three (3) vertical filter layers located in centerline of the dam
made of graded rock passing 40 mm, 300 mm and 150-450 mm. The filter layer surface
is coated with a fabricated geotextile filter layer to avoid fine materials passing into the
drainage layer that can decreases drainage capacity (clogging), thus clean water will
flow out of the dam body.
� Drainage core ripped slag material placed both on centerline and downstream shoulders
of the dam body, has function to drain water seepage from the dam body.
[Blank line 10 pt]
2.2. Detailed Drawing [Blank line 10 pt]
Detailed drawing of Fiona Dam is shown in the following cross section below.
Horizontal Drainage Ripped Slag
Fi lter Geotextil e
Maxi mum Water Level
Drainage Core Ripped Slag
Rockfill Ri prap
Jointed Peridotite
Horizontal Distance
0 20 40 60 80 100 120 140 160 180 200 220 240
350
360
370
380
390
400
410
420
430
0 20 40 60 80 100 120 140 160 180 200 220 240
Act
ua
l E
lev
ati
on
350
360
370
380
390
400
410
420
430
Figure 5. Detailed Drawing of Fiona Dam
[Blank line 10 pt]
The original base floor as bearing layer is a slightly weathered and jointed peridotite rock
with consistency hard to very hard. The dam body structure is laid on floor level from elv.
+375 m until reaching top crest on elv. +397 m. At centerline of the dam body, a drainage
core ripped slag covered by filter geotextile are constructed layer by layer reaching elv.
+395 m. A horizontal drainage ripped slag of 2 m thickness are constructed sloping down
parallel to final downstream slope. [Blank line 10 pt]
2.3. Geotechnical Parameter [Blank line 10 pt]
Geotechnical parameters are modeled in drained condition as critical state since water level
at front face higher and influencing stability of the dam body. The geotechnical parameters
are shown in the following Table 1. [Blank line 10 pt]
Table 1. Geotechnical Parameter of Respective Materials
Material Specification
Bulk Unit
Weight
Friction
Angle Cohesion
(kN/m3) (degree) (kPa)
Jointed Peridotite Rock 27 33 10
Rock-fill Riprap 26 38 20
Drainage Core Ripped Slag 25 33 0
Horizontal Drain Ripped Slag 25 33 0
III- 76
3. SEISMICITY [Blank line 10 pt]
Seismicity in the Indonesian Archipelago is quite high and about 10% of the whole
seismicity in the world. The epicenters were abundant in the young tectonic belt between
the continental shelf of South Asia and Australia. The belt starts from Sumatra-Java-Banda
turned to the northward through Sulawesi and Philippines. Generally, shallow epicenters
and its position are concentrated in western part of Sumatra, southern and eastern part of
Java, Banda Trench, western part of Papua and some territory between North Sulawesi,
Halmahera and Philippines. [Blank line 10 pt]
As we know Indonesia is prone to earthquakes, as it is the convergence of the Euroasia, the
Indo-Australia, the Pacific, and the Philippine Tectonic Plates. Fiona Dam that located in
Sulawesi Island is part of Pacific Tectonic Plate lines and Philippine Plate that's why
identification and evaluation of seismic hazard is very important for the dam design. [Blank line 10 pt]
There are several active fault systems on the Sulawesi Island including Matano Fault that
leads towards the bottom of Lake Matano adjacent to Fiona Dam. Sorawako that located
on the shores of the Lake Matano is a place where some frequent small earthquakes
emerge every few months. Thus, we have to careful in determining seismic coefficient for
the dam design. [Blank line 10 pt]
3.1. Determination of Peak Ground Acceleration [Blank line 10 pt]
In pre-design stage, determination of peak ground acceleration (PGA) for Fiona Dam on
horizontal direction is using some sources such as Indonesian Earthquake Study (1981),
NGDC/USGS Earthquake Catalogue (1999), and Golder Associates (1996). Results of the
peak ground acceleration (PGA) probability of Fiona Dam which is based on some sources
are mentioned above can be seen in Table 2. [Blank line 10 pt]
Table 2. Peak Ground Acceleration, PGA (g)
Source Return Periods (Years)
20 100 200 475 1000 2000
IES 1981 0.07 g 0.18 g 0.23 g 0.28 g 0.33 g 0.40 g
NGDC/USGS 1999 0.08 g 0.16 g 0.21 g 0.28 g 0.36 g 0.44 g
Golder 1996 0.03 g 0.18 g 0.28 g 0.39 g 0.48 g 0.58 g
Vale 2013 0.47 g 0.59 g 0.78 g
[Blank line 10 pt]
Calculation of maximum credible earthquake (MCE) is also based on Indonesian
Earthquake Study (1981), NGDC/USGS Earthquake Catalogue (1999), Golder Associates
(1996), which can be seen in Table 3 below. [Blank line 10 pt]
Tabel 3. PGA of Maximum Credible Earthquake
Source MCE PGA
Mean Mean + 1 T
IES 1981 M 7.9 @ 60-70 km
NGDC/USGS 1999 M 7.5 @ 60 km 0.12 g 0.20 g
Golder 1996 M 7.5 @ 10 km 0.43 g 0.73 g
Vale 2013 M 7.1 @ 10 km 0.37 g
[Blank line 10 pt]
III- 77
The probability calculation result shows that among sources are relatively similar. In line
with the earthquake calculation results on Balambano and Karebbe Dam in which using
sources from Golder (1996), then seismic design for Fiona Dam is also using Golder
(1996) as well. Earthquake design for Fiona Dam is recommended for the earthquake
periods of 400 years related to design basis earthquake (DBE) refer to Golder (1996) in
which the PGA is 0.36 g. [Blank line 10 pt]
3.2. Seismic Hazard Analysis [Blank line 10 pt]
PT Vale Indonesia Tbk in cooperation with Bureau of Meteorology, Climatology, and
Geophysics (BMKG) headquartered in Jakarta had accomplished Sorowako Seismic
Hazard Analyses in 2013. Sixteen in-situ micro-tremor tests were conducted in Fiona Dam
to obtain resonance frequency, amplification factor, and seismic susceptibility index. The
test results show that Fiona Dam dominantly has low seismic susceptibility index ranging
from 0.29 – 9.70 with resonance frequency between 1.9 – 9.3 Hz and amplification factor
low to medium (A < 6). These values indicate that the ground profile of Fiona Dam and
vicinity is relatively stable. [Blank line 10 pt]
Response spectra at the bedrock around Fiona Dam with return period of 475 years (equal
with 10% probability) and probability of exceedance 50 years (T = 0 second), peak ground
acceleration (PGA) is ranging from 0.44 – 0.49 g; response spectra at the bedrock with
return period 950 years (equal with 5% probability) and probability of exceedance 50 years
(T = 0 second), peak ground acceleration (PGA) is ranging from 0.55 – 0.62 g; and
response spectra at the bedrock with return period 2,475 years (equal with 2% probability)
and probability of exceedance 50 years (T = 0 second), peak ground acceleration (PGA) is
ranging from 0.74 – 0.83 g. [Blank line 10 pt]
[Blank line 10 pt]
4. DAM STABILITY ASSESSME NT [Blank line 10 pt]
Slope stability analyses both at upstream and downstream dam slopes and risk assessment
are conducted to obtain safety factor and risk matrix of Fiona Dam in case of failure in
which applied seismic loads modeled as pseudo static loads that represented with the peak
ground acceleration (PGA). [Blank line 10 pt]
4.1. Stability at Upstream Slope [Blank line 10 pt]
Pattern of critical slip failure for various seismic loads at upstream slope of the dam is
shown in the Figure 6 below. The slip failure is started from centerline of the dam crest
crossing a drainage core ripped slag until toe of rock-fill riprap with safety factor 2.231
(0.0 g), 1.859 (0.1 g), 1.602 (0.2 g), 1.401 (0.3 g), 1.250 (0.4 g), 1.129 (0.5 g) and 1.031
(0.6 g). [Blank line 10 pt]
Safety factor of upstream slope is lower than downstream slope due to the slope steeper. [Blank line 10 pt]
III- 78
Horizontal Distance
0 20 40 60 80 100 120 140 160 180 200 220 240
350
360
370
380
390
400
410
420
430
0 20 40 60 80 100 120 140 160 180 200 220 240
Act
ua
l E
lev
ati
on
350
360
370
380
390
400
410
420
430
Figure 6. Pattern of critical slip failure at upstream slope
[Blank line 10 pt]
4.2. Stability at Downstream Slope [Blank line 10 pt]
Critical slip failure pattern at downstream slope of Fiona Dam can be seen in the following
Figure 7. The slip failure is started from centerline of the dam crest as crossing a drainage
core ripped slag and rock-fill riprap then penetrating a jointed peridotite rock as original
base floor with respective safety factor various 4.513 (0.0 g), 3.153 (0.1 g), 2.533 (0.2 g),
2.120 (0.3 g), 1.834 (0.4 g), 1.624 (0.5 g) and 1.484 (0.6 g). [Blank line 10 pt]
Horizontal Distance
0 20 40 60 80 100 120 140 160 180 200 220 240
350
360
370
380
390
400
410
420
430
0 20 40 60 80 100 120 140 160 180 200 220 240
Act
ua
l Ele
va
tio
n
350
360
370
380
390
400
410
420
430
Figure 7. Pattern of critical slip failure at downstream slope
[Blank line 10 pt]
[Blank line 10 pt]
Table 3. Resume of Safety Factor based on Seismic Load
Geotechnical Stability at
Factor of Safety based on Seismic Load (PGA)
0.0 g 0.1 g 0.2 g 0.3 g 0.4 g 0.5 g 0.6 g
Upstream Slope 2.231 1.859 1.602 1.401 1.250 1.129 1.031
Downstream Slope 4.513 3.153 2.533 2.120 1.834 1.624 1.484 [Blank line 10 pt]
III- 79
Geotechnical Stability at the Dam Body based on Seismic Load
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
0.0 g 0.1 g 0.2 g 0.3 g 0.4 g 0.5 g 0.6 g
Seismic Load (Peak Ground Acceleration, PGA)
Fact
or
of
Safe
ty
FoS of the Upstream Slope FoS of the Downstream Slope Minimum FoS
Figure 8. Geotechnical Stability at the Dam Body with various Seismic Loads (PGA)
[Blank line 10 pt]
[Blank line 10 pt]
5. SUMMARY AND RECOMMENDATION [Blank line 10 pt]
Fiona Dam is located on low seismic susceptibility index zone that means relatively stable
in geotectonic terms. Based on slope stability analyses as summarized in Figure 8 above,
upstream and downstream slope of Fiona Dam with seismic loads for peak ground
acceleration (PGA) 0.45 g indicates that the dam body structure is in stable condition in
which the calculation results of respective safety factors more than 1.2 or meets
requirement. It means that Fiona Dam structure is expected having service life until 50
years with probability of exceedance 10%. [Blank line 10 pt]
Especially for upstream slope with peak ground acceleration (PGA) 0.49 g, safety factor is
ranging 1.0 – 1.2, it means the upstream slope is still in critical condition. Intensive
monitoring mainly soon after earthquake incident should be done to observe prior signs
before failure. [Blank line 10 pt]
At the dam crest and downstream slope, land clearing should be routinely conducted to
clean wild vegetation and bushes growing on it as well as trash in front of the spillway to
avoid any clogging and blocking in the spillway structure. [Blank line 10 pt]
[Blank line 10 pt]
ACKNOWLEDGEMENT We would like to express our grateful to Mr. H Basri Kambatu as Senior General Manager of
Mines & Exploration Department who has given encouragement during writing this paper, Mr.
Yusram Rantesalu as General Manager of Mine Engineering who has given direction and positive
feedback, and also colleagues BMKG who had conducted seismic study in Sorowako area. [Blank
line 9 pt]
[Blank line 9 pt]
III- 80
REFERENCES [Blank line 9 pt]
Bemmelen, R.W. Van. (1949): The Geology of Indonesia, Martinus Nijhoff The Hague,
Batavia, Indonesia.
Hamilton, W. (1979): Tectonic of the Indonesian Region, Geological Survey Professional
Paper 1078, Washington, United States of America.
Irsyam, Masyhur, et all. (2010): Summary of Study Team Indonesia Earthquake Map
Revised 2010, Ministry of Public Work, Jakarta, Indonesia.
Simandjuntak, T.O., et all. (1991): Geology of Malili, Ministry of Energy and Mineral
Resources, Geological Research and Development Center, Bandung, Indonesia.
____________________ (2009): Final Report of Dam Evaluation and Stability Analysis,
PT. Indra Karya (Persero) Wilayah – I Jawa Timur, Malang, Indonesia.
____________________ (2013): Seismic Hazard Analysis at Sorowako and Vicinity Area,
Bureau of Meteorology, Climatology, and Geophysics (BMKG), Jakarta, Indonesia.
III- 81