INTERNATIONAL SYMPOSIUM ON Bali, Indonesia, June 1 ST – 6 TH , 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. [email protected]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/cm 2 and high compressive strength 238.9 kg/cm 2 . 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
71
Embed
Design Optimization of Bauxite Residue Dam in Connection with ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
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.
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
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
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
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
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
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
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