- - -- - - --

,
7:!

PROCEEDINGS OF THE EIGHTH REGIONAL CONFERENCE FOR

AFRICA / HARARE / 1984

SOIL MECHANICS AND

FOUNDATION ENGINEERING
Technical editors:
l.R.BOYCE
WR.MACKECHNIE
K.SCHWARTZ

VOLUME ONE
.
1\ ,
I '
., ,

OFfPRINT

Published on behalf of the

Division of Soil Mechanics and Foundation Engineering
.of the Zimbabwe Institution of Engineers

A.A.BALKEMA/ ROTIERDAM / BOSTON

 Proceedings of the Eighth Regional Conference for Africa on Soil Mechanics and Foundation Engineering /Harare /1984

Geotechnical evaluations for tailings impoundments

J.A.CALDWELL
Steffen, Robertson & Kirsten (Vancouver) Inc., Canada

C.STEVENSON
Steffen, Robertson & Kirsten (Mining) Inc., Johannesburg, South Africa

SYOOFSIS

The design, operation and reclamation of tailings ill1);X)undnentsrequire a knowledge of the ge0-
technical characteristics of the site and imp:>unanent materials. 'Ibis piper describes the general
considerations and approaches involved in geotechnical investigations for tailings ill1);X)undnents.
Five case histories of geotechnical investigations for tailings impoundments are described to
illustrate the approaches and methods discussed. 'Ibe case histories vary fran studies of tailings
discharged into model p:>nds fran a pilot plant (to predict full scale tailings behaviour) to
site investigations in areas covered I: alluvial clays and collaISing sands.

IN'!ROOOCl'ION of the site, construction materials and

tailings, which will most likely be used
Tailings impoundments are argueably the in the design, construction, operation and
largest structures built by man. 'Ibey are closure of a tailings impoundment. This
primarily geotechnical structures: built on information is required in order to perform
soils and rocks, with soil and rock materials at least those geotechnical analyses and
to contain ground-up rock that behaves essen- evaluations listed in Table 2.
tially as a soil. The mining engineer, the
mill manager and the geotechnical engineer The tables referenced above are an attempt
all play a pirt in the safe design, econanic to define which geotechnical parameters
operation and effective reclamation of a are required and why they are used in tailirg;
tailings impoundment. For this reason a impoundment geotechnical evaluations. A
broad understanding of the geotechnical general philosophy of approach might be
nature and requirements of the foundations, that the designer or operater of a tailings
earthworks and tailings is an essential imp:>undnent must have a feel for the naterials
requirement for the proper functioning of on which, of which, and for wen the stru<xure
a tailings ill1);X)undnent. will be built. Quantifying the parameters
in the list is only a pirt of the engineering
'n1is piper sets out the elements of geotedmical synthesis that constitutes getting a feel
evaluations required to design, construct, for a tailings ill1);X)undnent.
operate and reclaim tailings ill1);X)undnents.
Case histories are described in order to The geotechnical characteristics must be
illustrate the princi~es discussed. quantified in order to carry out calculations.
The strength of the various materials must
be def ined in order to calculate the factor
GEX71'EQ{NIrAL REJ;JUIREI>IENTS of safety or probability of failure of the
embankment. Permeability must be measured
Geotechnical evaluations for tailings inp::ud- to calculate the quantity of seepage from
ments may be divided into the work required the impoundment, and hence to determine
to characterize: if this is acceptable. Consolidation pro-
the in situ foundation soils, rock perties must be known to allow an acceptable
rate of rise for the inpoundnent to be cal-
and grouna"ater. culated.
the construction materials for enI:m1kment
starter ckes and filter drains. SITE CHARACI'mISATIOO
the tailings.
In order to design a tailings iII1);X)undnent,
Table 1 lists the geotechnical characteristics the geotechnical characteristics of the

433
TABLE I : SITE & MATERIALS GEDTECHNICAL se
CHARAcr'ERIZAT ION Tt
PARAMETER .usE AI
tc
Site Material s ir,
me
I Soils iD
- DistribJtion Impoundnent Layout c.
Depth Layout, Stability Analyses of
In Situ Description (MOCSSO) Identification and Evaluation th
--
In Situ
In Situ
Density
Permeability
Definition of Strength
Seepage Analyses Fe
-- Moduli Strength
In Situ Stability and Settlement si
a
Seisnic Stability
- Grading Curve Identification and Evaluation en

--- Permeability
Plasticity Identification and Evaluation se
Seepage Analyses
Strength: Undrained Stability Analyses - Short term As
-- Consolidation
Effective Strength: c" Stability Analyses - Long term fi
co
Settlement and Strength
- O1emical Contaminant Mitigation an
of
2
-Rocks
DistribJtion Layout Definition ma
an.
- Depth Layout Definition
- Core Descriptions Indentification and Evaluation an.
-- Jointing
In Situ Permeability
and Fracturing
Seepage
Seepage and Stability
ty
dil
- Bedrock Acceleration,
As
Veloci ty , Displacement Seisnic Stability th
3 GrounGiater
-DistribJtion Layout Definition ~
ro
-Depth Layout Definition ane
-Quality Identification and Evaluation
Th;
Construction Materials so;
If
af1
4 Soils
-- Grading, Plasticity
Permeability
Identification
Seepage
and Evaluation ref
OVI
- Compaction O1aracteristics Construction Specifications
taj
ar
-- DurabilStrength Stability att
- O1emicali tyO1aracteristics Long-Term Behavior
Contaminant Control A r
5 Tailings lir
- Grading Identification and Evaluation
-- Strength:
Permeability
Static
Seepage Analyses
Stability Analyses
ree
met
-- Strength:
Consolidation
Dynamic Stability Analyses
Settlement and Strength Increase
pit
ciE
-- Beaching Angle
Deposited Density
Impoundnent Geanetry
Impoundnent Volume
soj
ma~
-- O1emical
Deposited Moisture Content
O1aracteristics
Water Balance Studies
Contamination Potential
pit
ma1
Soi
6 Slurry
- Slurry Density Water Content, Depositional O1aracteristics
and
- Slurry Supply Rate Impoundment Size, Rates of Rise, Distri-
bution Facilities
In
Afr
7 Mill
-- SUpply
Rock Type
Rate Impoundnent Size
Tailings Definition
gai
tig
ri!
-- Grinding Process
O1emical Process
Containment Potential
Post Deposition Behaviour
med
to
of
the

Det

434
 soils and rocks at the site must be known. an extensive body of experience is not
To cb this one carries out a site investigatioo. available of successful
impoundments on
similar foundation soils.
Aspects unique to site investigation for
tailings i~undnents are diSOJSSedbelow. M::st it has not been possible to describe the
impoundments oover a oonsiderable area. '!his soil profile properly, pirticularly where
may include the hillsidessurroundingthe the oonsistency of individual soil horizons
impoundnent (if in a valley) and the associated has not been assessed.
catchment areas. The impoundmentareas
of the case histories describedlater in local variations occur along the perimeter
this piper vary from 1 Jan2 to 4 Jan2. of the impoundment, or along critical
structures such as penstock decant pipe-
For such large areas the first step in a lines.
site investigation is what may be termed
a preliminary or medium scale survey. '!he the design involves a short high artankment,
emphasis is on the definition in a broad eg in a narrow valley (see Case History 2).
sense of the geology and soils at the site.
the impoundment is to retain water and
As described by Brink et al (1982),the also serve as a water storage dam.
first step in a medium scale survey is to
consult existing topographic, geological groundwater oontamination from jmp:)undnent
and pecblogical map;, and arrj other pblicatiOrE seepige is of ooncern.
of interest for the area. After studying
maps and air photos the site is visited, special soil problens which rEqUire further
and an attenpt is made in the field to identify evaluation have been identified during
and delineate landforms, land-use, vegetation the preliminary investigations (see Case
type, and soil and rock types and their History 4).
distributions.
The preliminary survey will have defined
As described in Case Histories 1 and 5, the constraints imposed by the site. The
the next step in the medium scale survey detailed investigation is planned to provide
may be to have the area flown so as to define data to solve the problems so defined.
rock outcrops, soil zones, faults, dykes
and zonesof noisture using infra-red iJTagery. During a detailed site survey nore detailed
This technique is most efficient where the stratigraphic definition may be done with
soil cover is thin (Warwicket aI, 1979). closely spaced reflection or refraction
. .. If the soil cover is thick, and thus may seismic studies or electrical resistivity
affect the design of the imp:>undnent, seismic profiling (see Case History 2). Detailed
refraction can be used to define soil depths on-site profiling and recovery of soil and
over large areas at reasonable cost. As rock samples is carried out, with frequent
tailings impoundments normally cover large test pits or core rigs. In situ testing
areas the use of geophysics is usually an may be done. The SPl'test provides a cpod
attractive eoonanic solution. measure of sand consistenCY1 the Dutch Cone
and the piezocone are useful for defining
A medium scale site survey may include a soil layeringand groundwaterconditions1
limited number of trial holes or pits to in situ pressure meters measure the soil
record the soil profiles. Around the peri- strength and deformation characteristics.
meter of a tailings impoundment, a trial
pit spicing of 200 to 500 m is usually suffi- Packer testing and in si tu tests in piezo-
cient. Where rapid changes of topography, meters can be used to define the hydraulic
soil or vegetation occur closer spacing conductivity (permeability) of the soil
may be warranted, but closely spiced trial and rock layers.
pits normally constitute the approach which
may be termed detailed site evaluation. Soil samples for classification, strength,
Soil samples may be taken for identification consolidation and hydraulic conductivity
:s and testing. testing may be obtained from Shelby tubes,
cores, and block or bulk samples. '!he choice
In many cases, particularly in southern of equipnent to obtain the sample depends
Africa, there may be extensiveexperience, on the soil type.
gained from other tailings imp:>undnent inves-
tigations of the soil types and their characte- Field and laboratory data obtained from
ristics. In such cases a preliminaryor preliminary and, if required, ci!tai1ed i.rwesti-
medilJll scale survey may be all that is rEqUired gations, are used to compile geological
to characterise the site, to identify sources maps and cross-sections of the site, to
of constructionmaterial,and hence design record detailed soil profiles and quantify
the imp:>undnent. those geotechnical parameters to be used
in the design and analysis of the in1;Dundnent.
Detailed investigation is necessary when :

435

. -----
 Due to the large areas normally covered The behaviour of the tailings as-deposited
-
by tailings impoundments, a fence diagram may be quantified simply by measuring sud!
is a useful tool at this stage as it displays parameters as the beach angle, moisture
the lateral extent of the soil profiles content, drying time, and sOlids-settling
along with depth. rate within the px>l. Piezometers can measure
the water pressures that develop.
~STRUCl'ION fo!ATERIAU)
In situ testingwith piezocones(Jonesand
Construction of a tailings inq;:oundnent usually Van Zyl, 1981), pressure meters, SPT or
rEquires starter dykes, E!I11:anI<IIentor tailin:Js shear vanes may be used to determine the in
guide walls. These may be constructedof situ geotechnical properties of the deposited
soils or quarried rocks that are availab[e tailings.
within reasonable proximity to the inp)lninent.
To the extent that these materials are soils, If there is no tailin;Js din of similar naterials
conventional geotechniques may be used. locally available,then the designer (in
Essential tests includemoisture-density order to characterizethe tailings)must
r determinations to assess aJlllpiction d!aract~ acbpt a different approach.
1 ristics,triaxial tests on aJlllpicted material
t for shear strength, and permeability tests First, the literatureshould be reviewed
to measure the water retention or drainage
- d!aracteristics of the aJlllpicted material.
to ascertain previouslymeasured values.
Next, the designer may test bench-scale
samples of tailings. Such bench samples
Many mines usually have large quantities are not always representative of the proto-
of waste dump rock available. This dump type mill grind, and caution when using
rock may be d!eaply used to cxmstructstarter the results is therefore espacially warrantEd.
embankments. The characterization of the Again, standardgeotechnicaltests may be
3
geotechnical propertiesof dumprockis used to measure grading, strEngth, ~creability Tt.
t certainly not a matter of CDnventiOnU~actice and consolidation d!aracteristics. s~
N
(Barton and Kjaerusli, 1981). If only rock wi
l'
core is available this may be examined in If possible, pilot plant tailings should be de
E
an attempt to determine the characteristics discharged into a small model pond and the at
of the rock as-mined. Joint spacings may properties of the deposited tailings can be t~
rE
give an indication of the grading of the measured. Case History 3 describes such of
rock, although blasting is likely to break a pond and the studies performed. While an
the rock into smaller fragments than those not traditionally viewed as a geoted!nical
estimated fran joint spicing only. If there parameter, the chemical composition of the In
is an existingwaste rock dump, thismay tailingsmust be determinedso as to study af
n be examined in order to obtain the grading, the pollution potential of the inq;:oundnent.
r ta
particle shape, .and uniaxial strength of ar
the rock. OPERATIONRID a.oouRE
c to
;) TAILOOS GEDImlNICAL OJARACrERIZATION The process of geotechnical evaluation ooes a
3 in
not cease once the tailings impoundment
t shJ
'!be methods used to characterize the geotech- is designed, constructed and commissioned.
re
r nical propertiesof tailings depends on Geotechnical instrumentation and monitoring
an.
whether an existing impoundment of similar are inq;:ortant aspects of the proper operation
tailings exists locally. Table 3 lists of a tailings inq;:oundnent. '!be data obtained ~
the methods used to establish the geote<:hni.cal fran an instrUTIentation and JOOnitceingprograme
characteristicsof tailings. A different determines the actual geotechnical perfocnanoe Th
level of knowledge of the tailings properties of the tailings impoundnent and allows this mi
J
may be required depending on whether the to be compared with the predicted or design SUi
impoundment containment dykes are to be performance. ro
constructed of tailings or borrow material. th
e Monitoring is an essential pitt c the ~actice
often referred to as the "Observational pr
If an imp;>undnent of similar tailings exists,
this should be examined in detail and a Method" (Peck, 1969). '1his aRXoad! essentially '1h
C careful note made of the successand problems consists of the following step;:
J ti
associatedwith the imp;>undnent, especially th
:) deposition techniques employed. Table 4 exploration sufficientto establish at
e ta
is a checklist for examining existing tailin:Js least the general nature, );:8ttern and pro- th
:t impoundments. Samples may be taken from perties of the deposits, rot not necessarily
i on
the existing impoundment and tested in the in detail. an
j laboratory. These parameters, previously
E wi
listed in Tab[e 1, nay be neasurEd fce disturl:ed assessment of the IOOStprobable conditions
or undisturbed samples. '!bus, the strength, and the most unfavourable conceivable
permeability or consolidation behaviour deviations fran these conditions.
may be determined using standard testing
methods.

436

_ _.u. ___+__ ___

 establishment of the design based on a The fieldwork for the investigationwas
working hypothesis of behaviour anticipated done in two Ihases, the first of which was
under the most probable conditions. the excavationof test pits in positions
determinedfrom air Ihoto interpretation.
~ selection of quantities to be observed This was found to be unsatisfactory as the
as construction proceeds and calculation air Ihotos showed insufficientdetail and
of their anticipated values on the basis siting of test pits could only be done on
of the working hypothesis. a .wild cat" basis. '!he conditions on site
were considered ideal for obtaining optinun
1 calculation of values of the same quantities results from an airborne thermal infra-red
3 under the most unfavourable conditions axn- linescan survey. The infra-red imagery
patible with the available data concerning gave extremelygood resolution of the rock
the subsurface conditions. structure and showed zones of probable deeper
soil cover.
selection in advance of a course of action
or modification of design for every fore- The second phase of fieldwork consisted
seeable significant deviation of the obser- of further test pits excavated in positions
vational findings from those predicted determined from the interpretation of the
:I on the basis of the working hypothesis. thermal infra-red imagery.

~ mesurement of quantities to be observed A geological and pedological map was prepared

~ and evaluation of actual conditions. to show areas of potential borrow material.
The north-western and south-eastern corners
;J modification of design to suit actual of the area mapped were considered to be
conditions. the best sources, tut were rot in the iJmediate
~ area of construction. Other scattered areas
( The advantage of a comprehensivemonitoring were more suitable in this respect.
system is that it provides the operators
with an advance warning of potential and However, during construction it was found
:! developing problems. The assumptions made that in the latter areas the soil contained
e at the design stage can be reassessedduring disseminated large dolomite boulders which
:! the operation of the impoundment. Such a make workability difficult. It was therefore
:1 reassessment can allow the introduction decided to use the deepersandy soil in >
e of more economic operating techniques or the south-eastern part of the site for con-
L an extension of the life of the illiX>unanent. struction.
:.
e l'
I Instnlllentation and monitoring should conti.rVJe Laboratory tests of samples taken in the
after reclamation and abandonment of the potential borrow pits showed the soil to
tailings impounanent. Tailings impounanents have acceptableshear strength and permea-
are large structures which ha\le great IDtential bility for use as construction materials
to adversely affect their surroundings over and as foundation materials below the inp:ud- C'
~ a long period. Consequently, a long term ment walls.
t instrumentation and monitoring programme
should be designed to provide the information The imagery obtained fran the thermal infra- i
;J required to minimize long term problems red linescan survey clearly showed areas c
11 and impact. of potential sinkhole development, such
:I as along soil-filled fracture zones in the m
e CASE HIS'IDRY 1 dolomite. Recommendations were made to
c
e prevent drainage of water into these zones r
~ The new tailings impoundment for a gold as downr.>ardpercolation of water will accelerate 'I
, mine is situated on Cblanite on the Malmani sinkhole development.
Subgroup, Transvaal Supergroup. '!he dolanite
rock has a thin soil cover over most of The spectacular detail of the rock structure
the site, but some thicker soil zones are obtained from the thermal infra-red imagery 1-
:! ),
L present. was the most illiX>rtant aspect of this inves- a
'I tigation. It provided the basis fran which
much of the information rEquired to satisfy .t
'!hemain objective of the geotechnical inves-
tigation was to locate borrow material for the main objectives of the investigation
t the construction of the toe walls for the was obtained. The case history illustrates
that the use of less conventional methods :,
tailings impoundment and for the walls of I:
the return water d:un. In addition infomatioo may be very informative in the ricjlt mnii.tias
on the permeability of the soils and rocks and should be considered when planni.n;Jge0tech-
and on the possibility of sink hole cEveloprent nical investigations.
:It
s within the impoundnent area was rEquired. .c
e \<.-

437
CASEHIS'roRY2 normally consolidated with an undrained
shear strength falling between 10 m and
'!his investigation was performed for a Irop:>sed 15 m to 40 kPa to 30 kPa respectively; ther~
valley-type impoundment, the bedrock of after, the strength increased to a maximum
which is hard banded schist. The valley of 70 kPa. a
is filled with glacial sandy gravels, overlain
by silts and clays, and covered by peat The coefficient of consolidation varies Tt
and an extensive muskeg aep,sit. Site investi- between 10-5 ~/s and 10-8 ~/s. Soil l<xar of
gation involved seismic profiling, drilling lic conductivi~ calculated fram these tests mE
for soil and rock samples, and in situ testiIj. varies fram 10-9 mfs to 10-10 mfs. iIT
me
2 000 m of refraction seismic lines were The sandy gravels beneath the silty clay li
performed. '!hese showed that bedrock beneath have SPT blow counts of about 14 and an
the imlDunanent embankment forms two traxhes, angle of friction of 300. Bedrock core '1h
one 30 m deep, the other 17 m deep. Bedrock recovery was good and hvdraulic conductivity I::!i
velocities are about 4 500 mfs. from 10-8 m/s to 10-10 mVs was measured su
in the intact rock.
Five holes were drilled with a tripod dri1~ Fh
rig. Standard penetration tests were done CASE HIS'lURY 3 in
and they showed very low blow counts. Intact wi
samples of silts and clays were recovered Pilot plant studies were perfa:med to establifh sp
with Shell::!i tubes and split SlX>Ons. A pitd1er the extraction process of copper and zinc an
sampler was used to recover sandy gravels. f rom a massi ve sul{:hide ore body. About 11
Rock cores were taken. 100 T were milled and the tailings disd!arged pa:
into a small IDnd. '!he floor was an existing the
Plastic standpipe piezometers, slotted at concrete pad and the walls were pre-cast COI
the bottom, were set in all holes. The concrete sections of about 1 m height. su:
bedrock permeability was measured with F8cker The pond was approximately triangular in weJ
tests. plan, the hypotenuse about 22 m, and the Th,
equal shorter sides of. about 15 m length. 56
Pressuremeter testing was performed in two Tailings were discharged into the pond at fr(
holes. The pressuremeter was inserted into one corner I::!iopen-ended disd!arge. sar
the ground and inflated to expmd rorizontally. sd!
The inflation pressure and corresponding The purpose of the study was to obtain data
radial strain were measured. Tests were on : Sel
performed at 1,5 m depth intervals down the
the hole, and thus an accurate profile of the bead! angle of the sedimented tailings. pre
the in situ soil strength was measured. Deposition was sub-aerial, slurry oozed to
down the beach to form an average bead! cIa
At one hole, a piezocone was driven into the angle of 4,70. of
valley fill. Hence, the layering of silts,
clays and sands was measured. Standard soil the tail ings densi ty. The known volt.me The
classification tests and mineralogical analyses of the pond and the tonnes deposited san,
were done on the silty clays. were used to calculate an average dry and
densi ty of 2 018 kN/m3. The specific tur
Undrained triaxial tests with IDre Iressure gravity of the tailings is about 4,5. and
measurements were performed on the silts bed
and clays to measure friction angles. Con- the grain size distribution along the mt
solidation tests were performed both in beach. The tailings gradation, 95\ less the
an oedometer and in a triaxial machine so than 0,075 mm, 0\ less than 0,01 mm,
as to measure soil response to earthquake did not vary down the beam. Fha,
induced stresses. con
the consolidation characteristics of to
The sil ty clay, axtIlDsed of quartz and feld- the tailings. Undisturbed samples were pro
spar with illite and chlorite, had 40 to taken with a Shelby tube and with a con- one
50\ silt size particles and 20 to 40\ clay. solidometer ring. An average value of for
The material has a low plasticity index Cv = 4 x 10-8 ~/s was measured. taiJ
of about 10. The natural moisture content
is between 15 and 30\, and in some cases the effectiveness of different drain Lab
the natural moisture content exceeds the designs. Drains of three different cEsigns was
liquid limit. Drained friction angles are were installed at the base of the IDnd. of .
between 300 and 350. High pore pressures The drain designs were nonitcxed I::!ineasurirg taiJ
were generated on shearing. The undrained both short and lorq-teIJn disdJarge; neasurirg
strength measured with the pressuremeter fines discharged throu<j1 the underdrains; Moil
shows that wi thin the top 6 m, the soil monitoring pore pressure between drain cart
is slightly overconsolidated with a strength lines to quantify in situ drainage d!arac-
of 40 kPa. Below this the silty clay is

438

_.- ..- .---

 ~

d teristics of the tailings; and visual of near surface clayey sands which amfirmed
d inspection of the drainage material when their suitability as cnnpacted fill in starter
~ the tailings were ranoved. walls. These tests a:mfirmed the suitability
m of the sands. The sandy clays were also
CASE HIS'IDRY 4 tested but were not recommended for use
provided sufficient clayey sand is available.
's This investigation was performed at the site
I- of a proposed ring dyke tailings impound- The resul ts of indicator testing performed
.8 ment for power station ash. The proposed on samples taken over the entire soil pcofile
im~undment is to replace an existing ~ confirmed the medium to very high potential
ment which is nearing the end of its useable expansiveness of the soils observed during
y life. visual pcofiling.
n
e The site is essentially level and is underlain Fran the results of the p:eliminaty investigation
y by clay. An air Ptoto mosaic revealed only it was concluded that the site was suitable
d surface drainage features. for the oonstruction of the pcoposed tailings
dam.
Phase one of the preliminary investigation
included drilling large diameter trial holes Subsequent to completion of the oonceptual
with a bucket auger machine. Holes were design, more localised, detailed investigatiom
h spaced between 300 and 500 m over the site were carried out by means of test pits and
c and taken to depths of between 5,5 m and further laboratory testing.
t 11,3 m. The trial holes were pcofiled with
d particulcr em~hasis on soil characteristics Before setting out the borrow pit area,
g that pcovide an indication of shear strength, further soil samples were taken from test
t compressibility and permeabilityof the pits for moisture-density and indicator
subsoil. In situ hand vane shear tests testing. These enabled the extent of the
n were carried out in two of the trial holes. borrow area and the axnpaction d1aracteristics
e The peak vane shear strength varies from to be accurately defined.
56 kPa to 110 kPa; residual values varies
t from 15 kPa to 24 kPa. Undisturbed block The potentially expansive nature of the
samples were cut for laboratory testing soils poses no pc obI em to the tailings iJliX>u-
scheduled to determine foundation amditions. ndnent performance. It does, however, affect
a the design of the foundations of various
Selcted disturbed samples were taken from service structures. Consequently detailed
the soil horizons identified during visual foundation investigations were carried out
profiling. Lalx>ratory testing was performed for the return water PJIIP station, the elec-
d to provide index properties, detailed soil tricity switch house, the main penstock
h classifications and compaction properties towers and an ablution block.
of ~tentially suitable borrow material.
A detailed investigation of near surface
e The tr ial holes revealed near surface clayey soils was required for the return water
d sands over soft to stiff, highly slickensided dam, which is to be oonstructed on a slight
'I and micro-shattered sandy clays, which in slope. Laboratory permeability testing
c turn overlie medium dense and dense clayey of undisturbed samples was carried out to
and sil ty sands. These soils overlie sarx:st<re confirm the suitabilityof the in situ soils
bedrock which occurs at depths from 5,5 as side walls to the dam in areas of cut.
e m to 10,5 m. Extensive water seepage into
s the trial holes occured below 4 m. MOisture-density and indicator testing was
carried out to assess the suitability of
Phase two of the preliminay investigation the soils from the areas of cut for use
consisted of a laboratory testing programme in oonstructingcanpacted side walls.
to establish the nature and engineering
! properties of the suoooil profiled in Iitase The chemical effect of return water on the
one. The results of this work were used soils was assessed. The clay minerals in
for the conceptual design of the proposed the soil are sodium-based and highly prone
tailings iJliX>undnent. to rapid dispersion, which would lead to
erosion and piping when subject to hydraulic
'1
Laboratory testing of undisturbed samples gr adi ents. However, on analysis the return
s was carried out to assess the suitability water was foundto be high in calciun ions
of the sandy clays as a foundation for the and low in sodium ions. Therefore cation
} tailings iJliX>undnent. exchange between the water and the soil
} resultingin dispersionof the soil would
Moisture-density and indicator testing was not occur and the return water dam walls
'1 will be stable.
carried out on disturbed samples fran areas

439
 Tl\813-L..:-G.EaI.f~m'AU.'ATIrNS ffiR TAIr,n.r.S ll'l'O:'t-O'..EW.S

MATEP.IJ\L
I
DESIGN ! CXJNS'lWCrrON OPERATIOO RX:LAI'ATION
1
Site or Foundation - Defir:e Site
Geology

Soils and Reeks Layout

- Irr,pcunaCEont - Construction Rates - Contamination Potential - Restoration
Stacil ity
- Er.,barn.:ent - Sources of Materials or Migration Route Requirements

- Er.banklr.ent Flow Nets - or

- SeeJ;age
Foundation Mov.ents
Settlements
- long-Term
Response

Construction
".aerials (Soils
and Rocks) - Ieentificat,on and - Construction Tech- - Definition of - Covers
Availability niques Exp!cted RestOf'.se - Vegetation
- Er..t..'f'.IcT.e,nt DP.sign - Construction Control and Monitorir.g Potential 2
Requi r,=".er.ts Specif ications - Progr Evaluation - shapir.<J
Requirer:'ents

Tailir.<;s - Defir.ition of - Rate of Pl3cerner.t - Cover

In>tOun6-ent Type - Operating Requirer:'ents Requirer,ents
- Layout Requirents - "nitoring Program - long-Term
Specification Behaviour
- Cont<rninantPotential
- Cyclone Operation
--- P.atesof Consolidationl -.
3

The investigation was carried out in close of mining below the new dam is considerably
collaboration with the impoundment designers. deeper than where the ground surface displace-
Consequently, rather than providing a general ment took place, it was agreed in discussions
comprehensive site investigation from the with mine management that a IOOdified mining
outset of the project (which could have method wOt4d be used beneath the new tailings
been unnecessarily costly), the work was dam. This would ensure more stable oormtions
programmed to provide specific information at ground surface. 4
as required at each stage in the design
process of the tailings im[X)undment. ~CLUSIOOS
OOE HISTORY 5 This IBper has discussed the geotechnical
considerations affecting tailings impoundment
The new tailings impoundment was planned
as the main tailings disposal site for the
remaining life of a platinum mine. The
geotechnical investigation of the site entailed TABLE 3: ME'IH0C6'ID ESl'J\BLISH TAILIOOS
the excavation and profiling of test pits, GEDTE(}INICALrnARAcrERISTICS
air P'toto interpretation and the inteqretation
of imagery obtained from thermal infra-red Existing Dams
linescanning. In addition laboratory testing 5
of the soils to determine their engineering Literature Study
properties was undertaken. Examination of Design Re[X)rts
Sample Testing in the Laboratory
The site is on the Bushveld Complex which Field Measurement
in this area is composed of gabbro rocks beadt angles shear vane
with pec;JMtoid zones (coarse grained feldsIBr drying rates piezooone
rich zones). The rocks have been completely settling rates pressuremeter
decomposed in the upper 0,6 m to 1,5 m to piezaneters
form a black "turf" clay which is highly
expansive. The clay is relatively ~rmeab1e, No Existing Dams
but in the upper 0,5 m is highly fissured.
This makes it possible for water to flow 6
into the profile, causing expansion and Literature Study
in turn closing the fissures. Test Samples fran Similar tams of Similar 6.
Ores and Milling Processes
A number of dykes and fractures crossing Test Bendt Scale Samples in Laboratory
the site were identified on the aerial P'toto- Test Pilot Plant Samples in Laboratory
Tests of Interest
graph and thermal infra-red imagery. In
another area of the property undermining has standard classification (specific gravity
caused surface displacement along a fault. grading, permeability, strength,
Such an occurrence below a tailings dam consolidation)
could be catastrophic. Although the depth settling times

440
 TABLE 4 : EXISTING TAILINGS IMPOUNDMENI' DESCRIPllOO

1 GENERAL Freeboard
Current and Planned Heights
Mine Slope Angles
Owner Beach Angles
Location Rate of Rise
Personnel: Mine Manager Estimated Ranaining Life
Mill Superintendent
Impoundnent Supervisor
Designers
Others 6.2 Tailings Delivery and Distribution

Pipes (distance fran mill, type, slopes,

2 OPERATIOO operation, wear, valves, etc)
Other Delivery Features (launders,
Ore Body drop ooxes, spigots, cyclones)
'IYPe of Mine Behaviour of Tailings on Discharge
Rock 'IYPeS
Size of Ore Body
Production History 6.3 Water Control

Decant Facility (type, size, ca~city,

3 MILL~ PI~N'F-C:S operation)
Return Water Dam (type, area, ca~city,
Grind pumps, spillway, operation)
Olanicals Used Diversion Facilities (type, length,
Slurry Treatment (thickness, pi, other) size, ca~city, slope, operation)
Slurry Density Water Recovery
Slurry Tanperature
Water Intake to Mill
6.4 See~ge Control and Instrunentation

4 SITE rnARAcrmLC:TICS Drainage Blankets (location, size,

effectiveness, outlet, seepage quan-
Precipi tation tities)
Evaporation Impermeable Linings
Other Others (filters, slurry trenches, wells,
etc)
Topogra(hy
Geology Piezometers (number, location, type,
Grounciiater ronitoring, water table)
Seisnicity Other Instruments (inclinaneters, flow
Geotedmical meters, water quality wells)

7 EXTmSIOOS
5 TAILJH;S
Time Planned or Required
Description Location
Grading Curve Area
Geotechnical Parameters (PI, c, ~ , Ownership
k, c..,; Cc' SG) Current Use
Olanical, etc (including .radioactivity Description (topogra(hy, geology, soil,
and toxicity) etc)
Delivery SG and wls Ratio

8 1IBANIXD'IENT PLANS
6 TAn.Thr.S IMmJNIJoIENT
Sides
6.1 General Top
Water Diversion
Construction Method Vegetation
Area of Impoundnent Monitoring
Area of R>nd, Distance fran Crest,and Current Test Programs

441
 --- --~ ..

design, oonstruction, q:eration am reclanatioo.

Methods of measuring the geotechnical charac-
teristics of the soils and rocks on which,
and of which, the inq;x>unanents are oonstructed
have been described. Similarly, the ll'eCGuranent
of tailings geotechnical properties has been
considered. Some relatively new techniques
have been described, and reference made to
methods that are standard practice.
Actual case histories are described with
details of testing a};{Bratus, and the ci!finition
of material parameters.
This paper has attempted to establish the
geotechnical nature of tailings inq;x>unanents,
provide checklists for those who will design
or review the design of impoundnents, and
provide a framework for further discussion
of ways of ensur ing that tailings inp:xJ1dnents
are treated as, what indeed they are, the
largest geotechnical structures built by
man.

REFERENCES

Barton, N. and Kjaerusli, B. (1981) Shear

Strength of Rockfill. Journal of the Geot:ech-
nical Engineering Division, ASCE, 107(~)
873-890.

Brink, A.B.A., Partridge, T.C and Williams,

. A.A.B. (1982) Soil Survey for Engineering.
,Claredon Press, Oxford.
Jones, G.A. and Van Zyl, D.J.A. (1981) The
Piezometric
Tool.
Probe
Proceedings,
-
A Useful Investigation
Tenth International
Conference on Soil Mechanics and Foundation
Engineering, Stockholm, VollJ\le 2, 489-496.
Peck, R.B. (1969) Advantages and Limitations
of the observational Method in Applied Soil
Mechanics. Geoted1nique, 19 (2) 171-187 .
Warwick, D.W., HartoW, P.G. and Viljoen, RoP.
(1979) Application of the Thermal Infra-Red
Linescanning Technique to Engineering Ge0-
logical Mapping in South Africa. Quarterly
Journal of Engineering Geology, 12(3)159-179.

442