University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
Colorado School of Mines
|
ER-4609 134
centerline methods. Good control of the phreatic surface in
the vicinity of the embankment face is achieved by either
arrangement, provided that the downstream shell materials
are sufficiently pervious with respect to the core.
Prerequisite to the use of cores as a phreatic surface
control method is, of course, that suitable low-permeability
natural soils be locally available. Cores of some type are
usually mandatory where water will stand directly against
the upstream face of a pervious embankment.
6.6 Drainage (after Soderberg and Bush, 1980)
From the designer's viewpoint, it is desirable to
promote drainage of water from the tailings pond in order to
keep the phreatic surface as low as possible and help the
consolidation and stability of the embankment. For this
reason, the relatively pervious tailings dam is the most
common design used. It is also the cheapest because it can
be built from the coarse fraction of tailings or from the
borrow material. The impervious tailings dam is the least
common type and is only used where it is necessary to retain
polluted water or low-density solids that are slow to
consolidate. In either type, the stability of the dam is of
paramount importance and necessary provisions must be made
to control seepage through and under the embankment and to
control surface runoff into the pond.
Water control in a tailings pond is probably the most
critical single design item affecting slope stability,
followed by (1) weak foundations, (2) slopes that are too
steep and too high, and (3) low density of the material in
the dike and on the beach. Failures are often caused by
|
Colorado School of Mines
|
ER-4609 135
overtopping of pond water, by piping of fine materials
because of seepage through the embankment or foundation, and
by piping along the outside of decants and culverts because
of the lack of or inadequate seep rings.
6.6.1 Drains (after Soderberg and Bush, 1980)
The position of the phreatic surface in an embankment
is very critical for its stability. Of all the factors
controlling stability, the control of water within the
embankment is by far the most important. For example, a
slope composed of cohesionless sand having a total unit
weight of 120 pounds per cubic foot and an angle of internal
friction of 35° will remain stable at an angle of 35°
regardless of the height of the slope, providing the
phreatic surface is below the toe of the slope. If the
phreatic surface is raised to exit the downstream face of
the slope, the steepest stable angle is reduced to
approximately 18°.
A relatively impervious starter dam very definitely
requires drainage on the upstream side, and a pervious
starter dam does not necessarily preclude the need for
drains. Should the pervious starter dam become ineffective
and there is no drain to fall back on, the phreatic surface
can rise and seriously reduce stability.
Two of the basic types of drains are (1) the blanket
drain, which is coarse gravel sandwiched between protective
filter material which carries all the water through and
beneath the dam, and (2) the perforated pipe drain, which is
surrounded by gravel and a protective filter which collects
the drainage and carries it through the dam. The choice of
|
Colorado School of Mines
|
ER-4609 136
which one to use depends on the availability of drainage
material, drainage capacity required, foundation conditions,
and construction cost. Variations and combinations of these
two methods can be used, but they must be carefully
constructed, and the material used for filters must have
proper grain-size graduation.
6.6.2 Blanket Drain (after Soderberg and Bush, 1977)
Blanket drains would typically be used in a cross
valley dam with either a pervious or impervious starter dam.
Bedrock or a relatively impervious base is close below the
natural ground level and the upstream method of dam building
utilizing the tailing sands is planned. The purpose of this
blanket drain is to intercept the water that moves downward
out of the tailings as well as any springs or artesian water
that may come up from below. If springs are found in site
investigation or artesian water in drill holes, either from
the rock or below a stratum of impervious clay, the blanket
drains should have capacity to remove all this water plus
additional capacity for that which may not have been
discovered. It is very important that this water be removed
because in mountainous areas it could have a high head and
if trapped below the slime layer in a tailings pond it could
exert tremendous upward pressure and greatly reduce the
factor of safety of the embankment.
The drain consists of a layer of clean gravel up to 18
inches thick extending from above the upstream toe to below
the downstream toe and wide enough to cover the main valley
bottom. This gravel drain is protected by a 9- to 12- inch
filter layer of clean sand and gravel both above and below.
|
Colorado School of Mines
|
ER-4609 137
An additional drain of unprocessed sand and gravel up to 3
feet deep is also placed upstream to extend the drain area
to catch all the seepage (Figure 43) .
These drains must have a catchment ditch filled with
cobbles to intercept the drainage and prevent erosion on the
downstream face. Where the downstream slope of an
embankment is composed of fine-grained materials, water
should not be allowed to flow out of this slope. Lowering
the phreatic surface increases the stability, permitting the
use of steeper slopes, and reduces the volume of
construction material needed. In cold climates it is
especially important that the drain water be directed
through a drainage blanket below the compacted soil so that
it will not freeze and raise the phreatic surface causing
the entire embankment to become saturated behind a frozen
blanket of soil on the downstream face of the starter dam.
6.6.3 Pipe Drain (after Soderberg and Bush, 1977)
Where drainage pipes are to be used, the pipes should
be designed to withstand the maximum anticipated load of the
overlying tailings. When perforated pipe is used, it should
be perforated on the bottom half only and laid with the
perforations down, with a bed of gravel both top and bottom
and graded filter surrounding the gravel (Figure 44). The
diameter of the perforations should not be larger than one-
half of the 85-percent size of the drainage material
surrounding the pipe. Pipe drains can be very satisfactory
with a good foundation and careful construction, but the
blanket or strip drains may be more fail-safe. Various
arrangements of pipe drains can be made. A perforated pipe
|
Colorado School of Mines
|
ER-4609 140
parallel to the upstream toe of the starter dam with one or
more solid pipes through the dam to the downstream toe is
the simplest. This same arrangement can be used as a
collection for drains up to 600 feet long running parallel
to the valley at right angles to the dam axis and spaced at
50- to 100- foot intervals along the valley floor and walls
(Figure 45) . Pipes through the starter dam should not be
perforated and should have at least three cutoff collars
that extend at least 2 feet to prevent "piping".
If the foundation beneath a tailings embankment is
compressible and differential settlement is possible, pipe
drains should be avoided. The stresses may result in
opening pipe joints or breaking the pipe, which might allow
internal erosion.
6.6.4 Strip Drain (after Soderberg and Bush, 1977)
Strip drains are the same as blanket drains in design
and construction except that they are narrow strips of drain
material laid in the foundation prior to dam construction.
They are laid out to carry drainage through the dam and to
outlets beyond the downstream toe of the embankments. The
drains are laid out in strategic locations to catch the
drainage and must be arranged according to the contours of
the foundation. Strip drains can be used upstream from the
starter dam in the same manner as blanket or pipe drains.
If blanket or pipe drains are placed as much as 500 to 600
feet upstream from the toe of the starter dam and spigoting
is started with tailings, the slimes begin settling on the
top of the drains in the first 100 feet from the upstream
toe of the starter dam. As the sand builds up, this fine
|
Colorado School of Mines
|
ER-4609 142
material continues to move uphill in an increasingly thicker
layer. The layer of fines or slimes is quite impervious
even in thin layers, and under consolidation and drainage it
can have a permeability as low as 10-6 or 10”7 centimeters
per second. This layer of slimes over a drain renders the
drain useless except for the first few feet upstream of the
starter dam. For this reason when drains are constructed
upstream any distance above the starter dam, the tailings
should be cycloned and the coarse underflow should be
repulped and spigoted on the dam to make a beach of very
pervious sand that will cover the drain. Since this coarse
sand will have a slope angle of 4+ percent, depending on
grind, etc., it will require a lot of material and come up
quite high on the starter dam. After the drain is well
covered, spigoting of unclassified tailings can be started;
the blanket of slime will not cover the drain, keeping it
free and operating for the life of the dam. This type of
blanket drain would be used only where bedrock is relatively
shallow and the natural soil is saturated or will become
saturated from the tailings pond.
In areas where the bedrock is 100 to 500+ feet deep and
the soil is very pervious (10'2 to 10~4 centimeters per
second) the blanket, strip, or pipe drains extending
upstream from the starter dam would not be used because the
seepage through the bottom would go down toward bedrock and
not follow the drain. Nearly every mine has a different set
of conditions and each tailings area must be designed
accordingly.
Because of the layering in a spigoted embankment, the
permeability in the horizontal direction may be as much as 5
to 10 times that in a vertical direction, especially if the
grind is coarse and the pulp density is low. To determine
|
Colorado School of Mines
|
ER-4609 143
the seepage from the pool and from the spigoting on the
beach, a flow net should be used to estimate the seepage
rate to the drains. The quantity of seepage will depend on
the permeability values, hydraulic gradient, and area of
flow. In some embankments and possibly all of them, the
piezometric head from the downstream toe up and under the
beach (on a large dam 500 to 600+ feet distance) is
determined more by the water flowing on the beach during
discharge than by the water escaping from the pond area.
The water in the pond is contained in a saucer of slime with
the permeability lowest at the center of the pond and
increasing toward the beach. Because of this, control of
the height and location of the water pool is important in
controlling the seepage. An increase of 1 foot in the pond
will flood 200 to 600 feet of beach, which has a higher
permeability than that area at a lower elevation. To insure
that the drain design is adequate, the upper range of the
coefficients of permeability of the embankment should be
used in these calculations. In the same way the lower range
of permeability of the drainage blanket should be used to be
sure that the blanket has a greater capacity than the
calculated seepage of the embankment. With a high water
table, water can enter the drainage system from the
foundation strata, thus increasing the required capacity of
the drainage system. Layers of impervious materials in the
foundation strata will restrict seepage into the foundation.
Calculating the thickness and width of blanket and
strip drains is probably worth the effort from a cost
standpoint because the difference of cost between 1 foot and
2 feet of gravel over a large area could be considerable.
The drains should be as large as practicable considering the
cost and availability of materials. They should be uniform
|
Colorado School of Mines
|
ER-4609 144
and continuous and constructed of the proper gradation of
materials, without which they could become useless.
Granular materials incorporated in underdrainage
systems should be compatible with the properties of the
seepage water they are designed to carry. . Drainage
materials composed of carbonate rocks are unsuitable if the
seepage collected by the system is acidic.
Blanket drains and strip drains should be designed to
be capable of passing full design flow when the phreatic
surface within the drain is at or below the upper surface of
the drainage material.
6.7 Filters and Transition Zones (after Soderberg and Bush,
1977)
A filter or "protective filter" is any porous material
that has openings small enough to prevent movement of soil
into the drain and yet is much more pervious than the soil
it is protecting. Transition zones are also filters between
a very fine and a very coarse material where several
different filter materials have to be used. In the
construction of a zoned embankment with the permeability
increasing in the downstream direction, filters and
transition zones should be placed between layers of
significantly different gradation to prevent piping and
subsurface erosion (Figure 46).
Each situation is different, and the filter design
should be governed by field conditions, particularly by the
gradation of the soil to be protected and the material
protecting it. Two different gradations of filters would
probably be used where coarse mine rock was to be placed on
the downstream toe of a tailings dam to prevent seepage.
|
Colorado School of Mines
|
ER-4609 146
Experiments have shown that filters need not screen out
all the particles of the soil but only the coarsest 15
percent, or the Dss, of the soil. These coarser particles
(Des and larger) will collect over the filter openings and
screen out the smaller particles. Therefore, the screen or
holes in a perforated pipe must be smaller than the Des size
of the soil. By the same reasoning, if soils are used as
filters, the effective diameters of the soil voids must be
less than Dss of the soil being filtered. Since effective
pore diameter is about 1/5 Die, then Die filter < Dss soil.
The filter must provide free drainage, and since the
permeability coefficient varies as the square of the grain
size, the ratio of permeabilities of over 20 to 1 can be
secured by Die filter >5Die soil.
To satisfy the criteria for filter materials,
thefollowing rules should be applied as illustrated in
Figure 47.
Rules 1, 2, and 3 assure that the protected soil does
. not pass through the filter. All filters allow small
amounts of fines to pass and collect in the outlet pipes, so
provisions should be made for flushing out the pipe if at
all possible. Rules 4 and 5 are to assure that the
permeability of the filter material is high enough to take
all the seepage from the protected soil. Attempts have been
made to develop a universal filter that will filter the
finest soil and yet have a Dss large enough that it will not
pass through the 5/16-inch perforations of commercial
drainage pipes. These filters have such a wide range of
size that the particles segregate during handling and
placement. A filter with a Cu of 20 or less is not as
susceptible to segregation (Rule 6).
Transition zones and filters are very important in the
|
Colorado School of Mines
|
ER-4609 148
design and construction of some tailings ponds, particularly
where drains are required either upstream from and through
the starter dam or in the starter dam construction itself.
They are especially important where a water-type dam is to
be constructed. Where remedial measures are necessary for
seepage emerging on the downstream face of a tailings dam, a
protective surcharge of any available material can be used
with protective filters and drains.
The thickness of filters required varies with the head
of water from a few inches with a low head up to 10 feet
with a high head in a water-type dam. The filter thickness
also varies depending on whether it is placed with
construction equipment or by hand. Steeply inclined filters
placed with construction equipment should be wide enough to
accommodate this equipment easily, while on level ground
these filters should be at least 3 feet thick. Filters
placed by hand should be at least 6 inches thick.
In summary, filters must be of proper gradation to
protect the soil, have a permeability greater than the soil
being protected, and have an increasing permeability in the
direction of flow.
6.8 Sand yield (after Soderberg and Bush, 1977)
The yield of suitable sand obtained in separating cthe
coarser fraction from the raw tailings affects the design
and construction of the embankments. A cross-valley dam has
a big advantage over a flat-country impoundment where three
to four sides have to be built using sand. A shortage of
sand makes it difficult to keep the embankment crest above
the finer tailings in the pond. With fine grind (55 to 60
|
Colorado School of Mines
|
ER-4609 149
percent minus 200 mesh and finer) and in flat country,
cyclones with the upstream method may be one way to keep the
embankment ahead of the tailings. In this case it may be
necessary to use mine waste as a supplement to the sand.
Operations with : a) 35 or more acres per 1,000 tons of
daily capacity; b) with this fine-grind material ; and c)
with tailings areas divided into two separate ponds that are
used alternatively; can operate with peripheral discharge if
the ultimate dams are kept low and the tailings are allowed
to drain and consolidate after each 10+/- foot lift.
The yield of acceptable sand from cyclones can be
calculated from the gradation of the raw tailings and the
characteristics of the cyclones. The rate of embankment
construction will depend on the amount of available sand,
the length of embankment being built, and the weather, or
the number of months a year that it is possible to construct
embankment. Using cyclones and the downstream method, each
foot of rise takes longer and requires more sand than the
previous foot. The use of cyclones with the centerline
method is nearly as bad.
Spigoting and building an upstream embankment with
beach sand requires the same amount of sand each lift except
for the increased length as the embankment gets higher.
In planning any tailings site, the active time for
embankment utilization is far below 100 percent. The time
required to build embankment and replace spigots or cyclones
and the time necessary to raise the entire line to a new
berm are times when the pond is not available for
discharging tailings unless they can be "dumped” at some
other spot in the pond. For this reason, it is better to
have two complete and separate dams. This is especially
important at the start of a new operation. With two dams
|
Colorado School of Mines
|
ER-4609 150
there can be a complete shutdown of an area so that the sand
beach can be drained, dike built, and pipes or cyclones
replaced. By alternating sites, a regular schedule of
maintenance and operation can be set up; also the annual
rise of the embankment is reduced, which improves slope
stability. Where the winters are severe, dike building can
be done only in the 6 to 8 warmer months to prevent the
formation of ice lenses in the beach area. Enough sand must
be available to build enough dike in the summer to last
through the winter months. Where tailings sand is used for
mine stope fill, the amount of sand available for embankment
construction is further reduced; of course, the total volume
to be impounded is also reduced by this amount.
When the grind is such that the proportion of minus
200-mesh tailings is more than 55 to 60 percent, the use of
cyclones is almost mandatory in order to save the entire
volume of sand for dam building. Under certain conditions,
a water-type dam should be considered even though the
capital cost is high. For these dams the operating cost is
very low. Conditions that may warrant water-type dams are
high percentage of slimes, harmful chemicals in the
tailings, and, for phosphate clay, slimes with no sand in
the tailings.
6.9 Spigots (after Soderberg and Bush, 1977)
The most common method of placing tailings on a pond in
upstream construction is by spigots spaced 16 to 60 feet
apart along a header pipe which is installed on the
embankment crest (Figure 48). The material immediately
adjacent to the spigots is used for dike constructions on
|
Colorado School of Mines
|
ER-4609 152
the next lift and should be the best tailings material
available. The gravitational separation of the sand and
slime as it flows into the pond ranges from very poor to
very good depending on the grind, specific gravity, and pulp
density. Figure 36 shows the gradation of the mill tails
from mine A, the gradation at the dike, and the change in
gradation 1,000 feet from the dike of a material that has
about 38 percent minus 200 mesh and is spigoted at 30
percent pulp density by weight. Figure 37 shows the
gradations for a material that has 58 percent minus 200 mesh
and is spigoted at 48 to 50 percent pulp density. Figure 49
shows the uniformity of the permeability change; the
vertical permeability is 8 x 10'3 centimeters per second at
the dike and decreases to 9 x 10-4 centimeters per second
7 00 feet from the dike. The horizontal permeability had the
same relative decrease in a 1,000-foot distance from the
spigot point but was approximately five times the vertical.
In mines B and C (Figures 50 and 51) , the horizontal and
vertical permeability in surface test pits is about the same
at the dike as it is 500 feet from the dike and also near
the decant area 3,000 feet from the dike. From very limited
testing there is also no apparent difference in permeability
between the horizontal and vertical samples from surface
test pits. The vertical permeability does definitely
decrease with depth and is very low just above the natural
soil. These two examples might represent the extremes of
what can be expected from spigoting. Mine B could certainly
improve the separations of the sand and slimes if it were
necessary by decreasing the pulp density 10 to 20 percent.
|
Colorado School of Mines
|
ER-4609 155
6.10 Cyclones (after Soderberg and Bush, 1977)
As the grind in metal mines becomes finer, 50 to 60
percent minus 200 mesh, cyclones can recover a larger
percentage of the sand for dike building than can spigoting,
and they are being used for this purpose. They are usually
mounted on the crest of the embankment but are sometimes
mounted on an abutment at one side of the embankment. From
the cyclones the fines are piped to the tailings pond, and
the coarser sand drops directly onto the embankment or is
repulped and discharged through spigots or launders onto the
embankment. In some cases, the fines can be piped upstream
many hundred of feet so that no slimes are beneath the dike
area. In some cases the slimes may be piped to another
tailings pond temporarily.
6.11 Compaction (after Vick, 1983)
Compaction of cycloned sand is a significant issue in
embankment design, and, as explained above, the various
cycloning methods vary in their ability to accommodate
compaction procedures. Compaction of embankment sands is
often desirable to reduce pore pressure buildup during shear
and the possibility of flow slides. But compaction becomes
of critical design importance in determining the
susceptibility of saturated embankment sands to seismic
liquefaction.
The site seismicity, internal drainage provisions in
the embankment design, and compaction requirements are all
interrelated. Liquefaction is most likely to occur for
loose, saturated sands under high levels of seismic shaking.
|
Colorado School of Mines
|
ER-4609 156
The main design measures against liquefaction are good
internal drainage to prevent saturation, and densification
of the cycloned tailings sands by compaction. In the
extreme, these design measures tend to be mutually
exclusive : when the sand is unsaturated, liquefaction is
unlikely regardless of density, while if the sand is
sufficiently dense, liquefaction is improbable even under
complete saturation. Since compaction requires mechanical
hauling and placement at considerable expense, it is often
less costly in areas of lower seismicity to provide for
internal drainage rather than high densities. In areas of
high seismicity, both compaction to high densities and
internal drainage is recommended.
Where high seismicity indicates the need for
compaction, a remaining design consideration is the relative
density requirement. In reviewing various liquefaction-
related relative density criteria, it is apparent that
compaction to relative densities of 50-60% is appropriate
for areas of moderate seismicity with expected accelerations
up to 0.10g. Higher relative densities, in the range of
75%, may be necessary for higher seismicity. Usually,
reasonably high relative densities can be obtained for clean
sand tailings with only modest compactive effort. The major
cost associated with compaction results from the need to
haul, spread, and place the sands in thin lifts rather than
from the compaction operation itself. Use of the hydraulic
cell method, however, allows for compaction without
mechanical fill handling and at relatively little cost.
Compaction of a soil is usually accomplished by
spreading the soil in layers of specified thickness and
compacting with a mechanical compactor. A number of
arbitrary standards for determining the optimum moisture and
|
Colorado School of Mines
|
ER-4609 157
maximum densities have been established to simulate
different amounts of effort as applied by the full-sized
equipment used in soil construction. The simplest and the
most widely used are the Standard Proctor and the Modified
Proctor. Compaction may be specified by procedure (type of
compactor, layer thickness, number of passes, and moisture
content to be used) or by end product (minimum inplace
density required). The purpose of compaction may be to
increase the fill shear strength or to decrease the fill
permeability or both.
Variables affecting compaction are:
The type of compactor : Pneumatic-tired roller,
steel wheel, vibratory steel wheel, sheepsfoot,
grid roller, vibratory plate compactors, and
tract-type tractors.
The weight and energy of the compactor.
The thickness of layers.
The placement water content.
Table 17 shows the compactors recommended to be used to
attain 95 to 100 percent of the Standard Proctor on various
soils, but during construction regular density testing is
necessary. For cleanc ohesionless tailings sand, compaction
in thin lifts with proper moisture might be attained with a
tractor for building dikes, but it should be checked with
inplace density tests. Seldom does the sand have less than
8 percent minus 200-mesh material, and the moisture content
is in the bulking range where compaction is difficult.
|
Colorado School of Mines
|
ER-4609 158
TABLE 17 Compaction equipment and methods (after
Soderberg and Bush, 1977)
Requirements for compaction to 95 to 100 percent Standard
Proctor maximum density
Compacted Possible variations in
lift Passes or Dimensions and weight of equipment
thickness , coverages equipment
inch
SHEEPSFOOT ROLLERS
For fine-grained soils 6 4 to 6 passes F oot For earth dam, highway, and
or dirty coarse-grained for fine Soil type contact airfield work, drum of 60-
soils with more Chan 20 grained area , inch diameter loaded to
percent passing the sq in psi I 5 to 3 tons per lineal
No. 200 sieve. Not 6 to 8 passes Fine-grained 5-12 250-500 foot of drum generally is .
suitable for clean for coarse soil, plas utilized. For smaller
coarse-grained soils. grained ticity projects, 40-inch-diameter
Particularly appropri index, >30. drum loaded to 0.75 Co
ate for compaction of Fine-grained 7-14 200-4 00 1.75 tons per lineal foot
impervious zone for soil, plas of drum is used. Foot con
earth dam or linings ticity tact pressure should be
where bonding of lifts index, <30. regulated to avoid shearing
is important. Coarse-grained 10-14 150-250 the soil on the 3d or 4th
soil.
Efficient compaction o soils
wetter than optimum requires
less contact pressure than for
the same soils at lower mois
ture contents
RUBBER-TIRE ROLLERS
For clean, coarse 10 3 to 5 Tire inflation pressures of 60 to Wide variety of rubber-tire
grained soils with 4 to coverages. 80 psi for clean granular mate compaction equipment is
8 percent passing the rial or base course and subgrade available. For cohesive
No• 200 sieve. compac tion. soils, light-wheel loads,
For fine-grained soils 6- 8 4 to 6 Wheel load 18.000 to 25,000 such as provided by wobble-
or well-graded, dirty coverages. pounds. Tire inflation pres wheel equipment, may be sub
coarse-grained soils sures in excess of 65 psi for stituted for heavy-wheel
with more than 8 per fine-grained soils of high load if lift thickness is
cent passing the plasticity. For uniform clean decreased. For cohesionless
No. 200 sieve. sands or silty fine sands, use soils , large-size tires are
large-size tires with pressures desirable to avoid shear
of 40 to 50 psi. and rutting
SMOOTH-WHEEL ROLLERS
Appropriate for subgrade 8-12 4 coverages. Tandem-type rollers for base 3-wheel rollers are obtain
or base course compac course or subgrade compaction. able in wide range of sizes.
tion of well-graded 10- to 15-ton weight, 300 to 2-wheel tandem rollers are
sand-graveI mixtures. 500 pounds per lineal inch of available from 1 to 20 tons .
width of rear roller. 3-axle tandem rollers are
May be used for fine 6- 8 6 coverages. 3-wheel roller for compaction of generally used in the range
grained soils other fine-grained soil ; weights from of 10 to 20 tons. Very
than in earth dams. 5 to 6 tons for materials of low heavy rollers are used for
Not suitable for clean plasticity to 10 tons for mate proof rol ling of subgrade or
well-graded sands or rials of high plasticity. base course.
silty uniform sands.
VIBRATING -BASEPUTE COMPACTORS
For coarse-grained soils 8-10 3 coverages Single pads or plates should Vibrating pads or plates are
with less than about 12 weigh no less than 200 pounds. available, hand -propel led
percent passing No. 200 May be used in tandem where or self-propelled, single or
sieve. Best suited for working space is available. For in gangs , with width of cov
materials with 4 to 8 clean coarse-grained soil, erage from 1-1/2 to 15 feet.
percent passing vibration frequency should be no Various types of vibratlng-
No. 200, placed thor less than 1,600 cycles per drum equipment should be
oughly wet. minute. considered for compaction
in large areas.
CRAWLER-TRACTOR
Best suited for coarse 10-12 3 to 4 No smaller than 08 tractor with (Tractor weights up to 60,000
grained soils with less coverages. blade, 34,500-pound weight, for 1 pounds,
than 4 to 8 percent high compaction.
passing No. 200 sieve,
placed thoroughly wet.
POWER TAMPER OR RAMMER
For difficult access, 4-6 for silt 2 coverages. 30-pound minimum weight. Con Weights Up to 250 pounds.
trench backfill. Suit siderable range is tolerable , Foot diameter 4 to 10
able for all inorganic 6 for coarse - depending on materials and
grained conditions.
|
Colorado School of Mines
|
ER-4609 159
The pneumatic tire has proven to be an excellent
compactor for cohesionless and low-cohesion soils, including
gravels, sands, clayey sands, silty sands, and even sandy
clay. It applies a moderate pressure to a relatively wide
area so that enough bearing capacity is developed to support
the pressure without failure. The light rollers are capable
of compacting soils in 4-inch layers to densities
approaching the Standard Proctor maximum at optimum moisture
with three or four passes. The heavy rollers can obtain
densities above Standard Proctor maximum in layers up to 18
inches thick with four to six passes.
Because of the small loaded area and high unit
pressure, the sheepsfoot roller is adapted best to cohesive
soils such as clays. Tailings embankments with more than 20
percent minus 200-mesh material in the dike could be
compacted with the sheepsfoot or in the modified sheepsfoot,
which has some of the feet removed and flat plates 8 to 10
inches in diameter welded on the remaining feet. This
modified version is best for silty soils of low cohesion,
which better describes some of the tailings sand. Moisture
control is very important to efficiency of compaction of
this type of material.
For cohesionless borrow and waste materials, such as
rockfills and gravels, compaction by track-type tractors and
haulage trucks might be adequate if properly controlled.
Where additional compaction is required, heavy vibratory
steel wheel compactors are very efficient and are not
moisture dependent.
Compaction is very critical on starter dams whether
pervious or impervious and on tailings embankments and dams
built of tailings sand to raise density, increase stability,
and prevent liquefaction.
|
Colorado School of Mines
|
ER-4609 160
A particular compactor has its own optimum water
content which may not be the same as that determined in the
Standard or Modified Proctor test in the laboratory. The
laboratory tests will be within a few percentage points of
those applicable to most compactors, and the heavier the
compactor, the lower will be the optimum moisture content.
Field density tests will help determine the efficiency of a
particular method. Use of compaction layers which are too
thick for a particular compactor will result in
undercompaction at the base of the layers, with increased
horizontal permeability of that zone.
No one method of compaction can be established as best
for the sand zones of tailings embankments because of the
wide variations in the gradation of the sand on the beach
from one property to another. Each mine must determine the
most efficient compactor for the materials it is using.
6.12 Seismic Considerations (after Klohn, 1980)
Liquefaction of a soil is a temporary state in which
the structure of the soil is disturbed, causing the
particles to lose contact, and to transfer the weight of the
soil and any superimposed loads onto the water in the voids.
This causes the pore pressure to rise and the soil behaves
like a dense fluid. It is unable to support loads or to
resist significant shear forces, and the resulting pore
pressures approach or equal the confining pressures.
Liquefaction can be caused by changes in static stress
conditions, particularly in very loose soils. However, the
more general cause of structural disturbance leading to
liquefaction is dynamic loading such as can occur during an
|
Colorado School of Mines
|
ER-4609 161
earthquake.
A form of liquefaction, in which air acts as a fluid
medium, can occur in extremely loose aeolian soils such as
loess. However, for soils of the types and densities used
in constructing tailings dams, liquefaction is assumed to
occur only in saturated zones, i.e. below the phreatic
surface.
The ease with which liquefaction can be initiated
varies. Under the least favorable conditions, a soil can be
liquefied by the first pulse of an earthquake, whereas under
other circumstances it may not fail even after hundreds of
cycles.
The factors affecting liquefaction potential are
summarized following :
Saturation
More-or-less complete saturation is essential for
liquefaction in sand of the type and density under
consideration. Partially saturated sand, such as
may exist where water is percolating through the
fill above the phreatic surface, will not liquefy
because the contained air is highly compressible,
thus preventing a significant rise in porewater
pressure. Saturation by capillarity is not
significant in cycloned tailings sand.
Permeability
For liquefaction to occur, the soil must be
sufficiently low in permeability that no
significant migration of water and consequent
pressure relief is possible within the time frame
of the earthquake. Most cycloned tailings sands
|
Colorado School of Mines
|
ER-4609 163
of cycles that may cause liquefaction under a
given cyclic shear. It may also affect strain
movements once failure has occurred.
Initial Shear Stress
If the initial shear stress has been acting long
enough to allow full drainage, the resistance to
liquefaction is increased. However, if drainage
has not had time to occur under the initial shear
stress, the resistance to liquefaction is
decreased.
Other Factors
There are several other factors which appear to
affect the liquefaction potential of a fine sand
under earthquake loading. These include such
things as: age of deposit, method of deposition,
exposure to previous earthquakes, particle shape,
overconsolidation, etc.
The liquefaction problem is normally assessed by
running dynamic shear tests on laboratory samples of the
tailings sand. These may be so-called "undisturbed" samples
or reconstituted samples. The most commonly used method is
the cyclic triaxial test which uses conventional triaxial
equipment adapted to allow cycling of the deviator stress.
The output includes continuous records of deviator stress,
axial strain and porewater pressure.
Using the data obtained from the dynamic laboratory
tests a dynamic analysis of the tailings dam may be carried
out. Such an analysis is involved and complex. Briefly, a
seismic analysis involves the following steps :
|
Colorado School of Mines
|
ER-4609 165
saturated.
In tailings dam designs for areas rated as low seismic
risks it is usually more economical to strive for complete
drainage rather than for high densities. For areas of high
seismicity both high densities and good drainage are
considered desirable.
6.13 Rate of Seepage (after Soderberg and Bush, 1977)
The rate of seepage through a tailings embankment is
governed by the permeability of the materials, the hydraulic
gradient, and the elevation difference between the pond and
the point of emergence of the seepage. With a coarse grind,
low pulp density, and a wide beach the seepage vertically
through the beach is tremendous and could equal or exceed
that which escapes through the embankment. With a given
screen analysis, mineralogy, tonnage, pulp density, area,
and construction method (upstream or downstream), there is
little that can be done to change the seepage once the
operation has started. The length of the flow path cannot
be increased after the pond gets to the extreme upstream
position, or to the decant, except by adding material to the
toe. The difference in elevation has to increase, and no
impervious barriers can be placed. Therefore, it is
imperative that seepage be anticipated and incorporated into
the design. The mathematical expression for the rate of
seepage is
|
Colorado School of Mines
|
ER-4609 166
nf
q = k — h,
nd
where q = the rate of seepage per unit length
perpendicular to the plane of the
flow net,
k = the coefficient of permeability of the
soil.
nf = the number of flow paths (determined
from the flow net) ,
nd = the number of equipotential drops
(determined from the flow net),
and h = the difference in piezometric head
between the point of seepage entry and
the point of seepage exit.
An approximate coefficient of permeability can be
determined by laboratory tests if field tests are not
possible.
6.14 Pond inflow and control methods (after Portfors, 1980)
Water inflow to the pond is made up of tailings slurry
transport water, precipitation directly onto the
impoundment, runoff from the pond catchment area, seepage
recovery, ground water seepage, and other miscellaneous
sources, such as pit dewatering, surface diversions, and
domestic wastes. If a deficit exists in the system, a
supplemental water inflow source may be required.
In most tailings ponds, the water required to transport
the tailings in a slurry form is the single largest input
water source. In a closed circuit operation, reclaim water
is withdrawn at a rate proportional to the slurry transport
|
Colorado School of Mines
|
ER-4609 167
water input, consequently, the rate of transport water input
usually is not critical.
Precipitation onto the pond becomes completely
available as input water to the pond. On the other hand,
runoff from the impoundment catchment and exposed beaches
will not equal the total precipitation, but will reflect the
basin losses to evapo-transpirâtion, groundwater recharge,
diversions, other withdrawals for use, etc.
Hydrological studies are required to estimate catchment
basin yields in average, wet and dry years. Average annual
precipitation and runoff values would be used to estimate
available water to balance other losses, and to estimate the
average quantity and cost of any make-up water that may be
required. Average conditions determine the overall pond
management strategy over the life of the mine.
Wet year inflow conditions are used in combination with
flood discharges to calculate the required dam freeboard or
spill capacity. Dry year flows, on the other hand, can be
critical to plant operation. If the precipitation and
runoff are required as water supply sources, the probability
of a dry year, or a series of dry years, must be established
to define water deficiencies.
Timing of inflows throughout the year will be important
as dam construction must be advanced sufficiently to store
inflow during the high runoff season. The freewater volume
is then depleted during the dry seasons of the year.
Available precipitation records are used to establish
amounts and timing of precipitation, and stream flow records
to estimate runoff. Mines are often located in remote
regions where hydrologie data are scarce or entirely absent.
The designer then must use information from neighboring
areas to estimate flows at the site.
|
Colorado School of Mines
|
ER-4609 168
Catchments above tailings impoundments are generally
small, and caution must be used when information from large
basins is applied on small ungauged basins. For example, in
mountainous regions, weather effects on local precipitation
can be larger than the regional average.
In the case of large upstream catchments, diversion
channels probably will be necessary to carry upstream runoff
around the pond, thereby limiting the input of surface
water. Design discharges adopted for these channels can
vary depending upon the consequences of overtopping.
Commonly adopted design floods range between the 100 year
and 1000 year recurrence interval events. With such a
design, only in rare cases will the channels be overtopped,
and flood waters enter the pond.
Offtakes can also be added to the diversion channels
enabling extra water to be added to the pond during dry
years.
Mine drainage water may be contaminated and
consequently cannot be released to the downstream. The
tailings pond then becomes a convenient storage location,
and the drainage water becomes another input water source.
Surface drainage water from around the pit and mill area
also may carry a high concentration of suspended sediments
and other pollutants. This water can be directed into the
tailings pond rather than into separate settling ponds.
Controls (after Vick, 1983). The preferred situation
is to limit water inflow by proper siting of the
impoundment. There are several methods available for
handling remaining flood inflows.
A primary method of inflow control is storage.
Adequate dam freeboard or surcharge is maintained at all
times by raising the embankment at a rate such that
|
Colorado School of Mines
|
ER-4609 169
sufficient volume is always available for storage of the
design inflow. If the design flood inflow has been
determined with an appropriate degree of conservatism, it is
unlikely to be experienced during the life of the
impoundment. In the improbable event that it should occur,
the impounded runoff could be stored for eventual
evaporation in regions of arid climate. In other areas, the
water, if contaminated by mixing with the mill effluent,
could be treated and released at a gradual rate. Control of
flood inflow by storage avoids the necessity of expensive
and sometimes difficult treatment of contaminated flood
inflow prior to release into surface watercourses.
In some areas, topographic constraints on practical
embankment height and impoundment volume, combined with high
precipitation or high rates of net mill water discharge, may
make storage of flood inflows unfeasible. In such cases,
the only option may be to provide treatment of all mill
effluent prior to its discharge into the impoundment. Flood
inflows entering the impoundment, presumably no longer at
risk of contamination by mixing with the treated mill
effluent, can be passed through the impoundment and
discharged through a spillway designed according to
conventional methods. In applicable areas, the thunderstorm
rather than general storm PMP may control the design of
spillways where peak flow rate rather than total in flow
quantity is of primary importance. The use of spillways
with raised embankments, however, is awkward at best. With
each embankment raise, a new spillway must be constructed at
the new crest elevation. This adds appreciably to cost and
difficulty of construction. In extreme cases, water
handling by spillways may preclude the use of raised
embankments, instead requiring water-retention type
|
Colorado School of Mines
|
ER-4609 170
structures built initially to full height.
Diversion channels are desirable in most cases and
essential in some for diversion of normal runoff flows.
Although diversion channels can also be used to divert flood
flows around the impoundment, major design floods often
produce flows that require large channels (channel widths
for PM F diversion in excess of 100ft are not uncommon) and
heavy riprap for protection of the channel banks against
high flow velocities. As a result, diversion channels
designed for peak flood flow are sometimes impractical,
depending on the channel length required. For raised
embankments, however, excavation of even relatively large
flood diversion channels may provide a convenient source of
starter dike fill material.
For open-pit mining operations, creative planning of
waste rock dumps and even the pits themselves can provide
useful water-control benefits for the tailings impoundment.
Process flow considerations in mine facility layout usually
dictate that the tailings impoundment and mill be located at
lower elevations than the pit and associated waste rock
dumps. If the pit is located within the drainage basin of
the tailings impoundment, the pit volume itself can be
credited with flood water storage for extreme PMF-type
floods. If the waste rock dumps can be extended across the
tailings impoundment drainage area without excessive haul
distance, diversion of extreme floods by the mass of the
waste rock can be realized at essentially no cost. This
requires that waste dump and tailings impoundment planning
be integrated and carried out together.
A related method of runoff diversion is to construct
non-impounding diversion dikes across the impoundment
drainage area. To be effective, such dikes should be
|
Colorado School of Mines
|
ER-4609 171
located as close as possible to the maximum upstream extent
of the impoundment. This method of surface-water control
may be of particular advantage if the impoundment is located
over relatively shallow bedrock where construction of
diversion channels would require expensive rock excavation.
Flow velocities against the diversion dike may be high,
however, possibly requiring the use of riprap if the dikes
are constructed of native soils. The use of open-pit waste
rock for diversion dike construction, usually relatively
coarse and erosion resistant, may be attractive.
In extreme cases where the tailings impoundment is
located in a narrow, constricted valley with a large
upstream drainage area, steep valley sidewalls may make it
impractical to divert flood runoff around the tailings
impoundment by either diversion channels or nonimpounding
dikes. Here it may be necessary to construct an entirely
separate flood-control dam upstream from the tailings
impoundment. The flood-control dam can provide full storage
of expected flood runoff from the upstream drainage area,
with gradual evacuation of the stored water by a culvert
passing through the water-control dam and around the
tailings deposit. This arrangement is to be avoided if
possible, since fill requirements for the water-control dam
may be large, in some cases even exceeding fill required for
construction of the tailings embankment itself. Moreover,
construction of the flood-control dam cannot be staged; it
must be completed prior to tailings impoundment operation in
order to serve its intended flood-protection function.
Finally, maintenance of buried culverts is problematic, and
the limited life of the culvert structure may make it
necessary to provide more permanent water-control measures
after impoundment abandonment and reclamation.
|
Colorado School of Mines
|
ER-4609 172
Figure 52 summarizes the various methods for handling
extreme flood inflows from external drainage areas. It is
important to note that even with complete diversion of flood
runoff from tributary drainage areas, that portion of the
extreme flood precipitation which falls directly on the
impoundment surface must still be accounted for.
6.15 Effects of floods on impoundments (after Vick,1983)
A completely separate consideration is control of
floods that may pass at the toe of the embankment. This
inflow into the impoundment may result in erosion,
undercutting, and eventual failure of the exterior face of
the embankment, and it is especially critical for
impoundments located in low-lying river floodplain areas or
narrow canyons. The same general flood criteria as for
impoundment inflow apply, only here the concern is with flow
velocities of the water at the embankment toe. Riprap of
affected portions of the embankment face provides one
solution. The method of choice is to avoid impoundment
siting in floodplains, particularly in areas of
predominantly soft, sedimentary rocks where riprap may be
unavailable locally and difficult or expensive to import.
For some Southeastern U.S. phosphate tailings impoundments
located in low-lying coastal areas, imported riprap for wave
protection during hurricanes constitutes the major expense
in impoundment construction.
|
Colorado School of Mines
|
ER-4609 174
6.16 Embankment freeboard and wave protection
(after Soderberg and Bush, 1977)
Tailings embankments built of borrow material need
protection from wave action if the water is allowed to come
against the borrow dike. Wave action is a very destructive
force and can erode and overtop an embankment unless ample
freeboard is provided. The height of the waves depends on
the wind velocity, the duration of the wind, the fetch (the
distance the wind can act on the water) , and the depth of
water. On steep upstream slopes, riprap will limit the
uprush of the waves to approximately 1.5 times the height of
the waves and will prevent erosion by wave action. Tailings
embankments should not have free water standing against the
dike except where a water-type dam is impounding the
tailings, and then they should have riprap on the upstream
face. Table 18 gives the approximate wave heights for
various wind velocities and fetch, and the freeboard and
riprap gradation for the 3:1 slope. For 2:1 slopes, the
thickness should be 6 inches greater. A layer of filter
gravel should be placed between the riprap and the
embankment.
Wave action is not a problem if even a small beach is
provided between the dike and the water, because the waves
dissipate harmlessly in the shallow water on the beach.
6.17 Crest Width (after Soderberg and Bush, 1977)
The crest width should be no less than 12 feet for easy
equipment operation in a situation where water is against
the embankment. The most suitable crest width will depend
on the allowable percolation distance through the embankment
|
Colorado School of Mines
|
ER-4609 177
CHAPTER 7
SEEPAGE AND SEEPAGE CONTROL MEASURES
The information presented here on seepage and seepage
control has been extracted from several sources. Section
7.1, 7.4, 7.5, were extracted from Control of Seepage in
Tailings Dams, written by Patrick M. Griffin. Section 7.2
and part of section 7.3 were compiled from Earl J. Klohn1 s
article Seepage Control for Tailings Dams. Sections 7.3 and
7.8.4 were extracted from the Bureau of Mines information
circular 8755, Design Guide for Metal and Nonmetal Tailings
Disposal, written by Roy L. Soderburg and Richard A Busch.
Sections 7.6, 7.8.1, 7.8.2, 7.8.3, 7.8.5, 7.8.6, and 7.8.7
were compiled from Seepage Control Measures in Tailings
Dams, written by Robin Fell. Sections 7.7 and 7.8.8 were
compiled from Steven G. Vick's book titled Planning, Design,
and Analysis of Tailings Dams. The information in section
7.9 through the end of the chapter was extracted from Mine
Waste Management, edited by Ian Hutchison and Richard D.
Ellison.
7.1 Introduction to Seepage (after Griffin, 1990)
When considering seepage, the designer of a tailings
dam is faced with three major concerns:
Resistance to piping and internal erosion.
Location of the phreatic surface and its effect on
|
Colorado School of Mines
|
ER-4609 179
water in the tailings pond can be estimated using
approximate flow nets based on the anticipated width of
beach and the permeabilities of the materials involved.
Estimating the quantity of seepage due to spigotting is much
more difficult, however, a reasonable approximation can be
made by assuming that the spigotting completely saturates
the beach so that in effect the free water in the pond
extends to near the top of the spigotted beach. A flow net
drawn for this condition should provide an estimate of the
combined seepage, due to both the free water in the pond and
the effects of spigotting.
Estimating the quantity of the construction water that
the drains must handle varies from a simple exercise for the
case of on-dam cycloning, where all the construction water
seeps into the downstream sand dam, to a complex exercise
for the case where large volumes of hydraulic fill transport
water flow across the dam and then exit either upstream into
the tailings pond or downstream behind the seepage recovery
dam. For the on-dam cycloning, all the water contained in
the cyclone underflow is assumed to reach the underlying
drain. This value can be estimated with reasonable accuracy
from the density of the underflow and the recorded tonnage
of sand placed each day. For the case of hydraulic fill
placement, where large volumes of water flow across the sand
dam, the seepage loss into the dam may be estimated by
assuming downward seepage under a hydraulic gradient of 1
and using the expression :
q = k i A
where q = rate of vertical seepage
A = total area over which the water is
flowing
|
Colorado School of Mines
|
ER-4609 180
i =hydraulic gradient (1 for this case)
k =effective coefficient of permeability =
Vk x K
h v
The fourth potential source of seepage, consolidation
water squeezed out of the slimes as they consolidate is more
difficult to quantify. In effect, this action adds an
additional increment of pore pressure to the normal
hydrostatic pore pressure that would exist in the slimes if
they were completely consolidated under their own weight.
In high rainfall climates precipitation can be a major
contribution and may cause the sand dam to be almost fully
saturated during long periods of heavy rainfall.
Of the above five outlined sources of seepage water,
construction water from cyclone underflow or hydraulic fill
placement operations is usually several times greater than
all other sources combined. In designing the drains, the
highest probable seepage flows that can enter the drains
should be used and the drains should be assigned their
lowest probable permeabilities and gradients. This is
essential, as once constructed and buried, the drains must
continue to perform satisfactorily throughout the life of
the structure and, in many cases, for many years after the
mining operation is completed. All drains should be sized
to handle flows several times the largest probable flows
computed using the above outlined methods.
7.3 Use of flow nets (after Soderberg and Bush, 1977)
The uppermost line of seepage that is at atmospheric
pressure is known as the phreatic surface and is the
|
Colorado School of Mines
|
ER-4609 181
uppermost flow line. An equipotential line is a line of
equal head; therefore, water rises to the same level in
piezometers installed along a given equipotential line. The
equipotential lines in a flow net must intersect the free
water surface at equal vertical intervals.
Flow lines and equipotential lines must intersect at
right angles to form areas that are basically squares when
the materials are isotropic. Adjacent equipotentials have
equal head losses. The same quantity of seepage flows
between adjacent pairs of flow lines.
Usually it is best to start with an integral number of
equal potential drops by dividing the total head by a whole
number and drawing flow lines to conform to these equal
potentials. The outer flow path will generally form a
distorted square figure, but the shape of these distorted
squares (the ratio B/L) must be a constant (Figure 53).
In a stratified soil profile where the ratio of
horizontal to vertical permeability exceeds the
( K h / K v ) 1 0 ,
flow in the more permeable layer controls. The flow net may
be drawn for the more permeable layer, assuming the less
permeable layer is impervious. The head on the interface
from this permeable layer is imposed on the less pervious
layer for the construction of the flow net within that
layer. This situation can occur when a starter dam is
impervious to the tailings it is retaining and has
insufficient, ineffective, or no drains, so that water at
the interface builds up to and goes over the top of the
starter dam. The flow net through the starter dam then
would be the same as if there was free water against it, as
illustrated in Figure 54.
In a stratified layer where the ratio of permeability
of the layers is less than 10, the flow net is deflected at
|
Colorado School of Mines
|
ER-4609 184
the interface in accordance with Figure 54.
When materials are anisotropic with respect to
permeability, the cross section should be transformed by
changing the scale. The horizontal dimension of the section
is reduced by the square root of The flow net is then
K v K h .
drawn as for isotropic materials and can be transposed to a
true section. If Kh > Kv the L dimension becomes elongated,
and the net is no longer a square. In computing the
quantity of seepage, the differential head is not altered
for the transformation. Where only the quantity of seepage
is to be determined, an approximate flow net suffices.
Where pore pressures are to be determined, the flow net must
be accurate.
Earl J. Klohn, in his article Seepage Control for
Tailings Dams states that seepage flows are predicted from
the flow net using the following relationship :
q = k . h . nf per unit of length
nd
Where : q = rate of seepage flow
k = coefficient of permeability of the
soil
h = the hydraulic head acting across
the structure
nf = number of flow paths in the flow
net
nd = number of equipotential drops in
the flow net
|
Colorado School of Mines
|
ER-4609 185
7.4 Piping protection (after Griffin, 1990)
The major issues which must be addressed to prevent
piping within the tailings embankment are control of
hydraulic flow gradients, care in the use of erodible
materials, and effective use of filters to separate
materials with large differences in gradation. Where a
beach is used, drainage paths will be lengthened, which will
have a positive effect in reducing the seepage flow gradient
through the upstream portion of the embankment. Without a
beach, it is necessary for designers to provide appropriate
control of the gradient in the upstream dam section.
If sand tailings are used as an integral part of the
dam, thed esigners must take into account the potential
erodibilities of these materials. Generally, non-plastic
sand and slime tailings can be expected to be extremely
erodible. Materials containing some clay minerals, such as
phosphate tailings, could have some inherent resistance to
erosion and dispersion.
Filter systems are used in tailings dams in much the
same manner as for conventional embankment structures.
Often, the sand fraction of the tailings can be used as a
filter or a component of a filter system. Care must be
taken, however, that the sand material has sufficient
durability and will be chemically stable when exposed to the
effluent seepage. Physiochemical decomposition of the
tailings sand could create problems over time if not
properly accounted for in the design.
Durability is commonly evaluated by performing the same
types of tests normally required for concrete sand and
aggregate (i.e.-pétrographie analysis, specific gravity and
absorption, sodium/magnesium soundness, abrasion, freeze-
|
Colorado School of Mines
|
ER-4609 186
thaw, etc.), but the acceptance limits are usually more
flexible than that required in ASTM Method C-33. For
example, chert is deleterious to concrete, but may be
perfectly adequate in a filter or drain. Similarly,
requirements for resistance to mechanical and chemical
breakdown need not always be as stringent as for concrete
aggregate.
7.5 Location of the phreatic surface (after Griffin, 1990)
The stability of the downstream slope of the tailings
dam will be heavily influenced by the depth of the phreatic
surface through the embankment. This is not as critical an
issue for the upstream slope, because of the presence of
solids within the pond. On the downstream side of the
embankment, the strength of cohesionless materials is
influenced by both the buoyancy effect of water below the
phreatic surface, and also the effect of seepage pressures.
The technical issues are essentially similar to those for
conventional embankment dams. However, with the combination
of features such as beach zones and extended drainage paths,
it is often possible to reduce the downstream phreatic level
to a very low surface within the embankment, and therefore
steepen the downstream slope while still maintaining
acceptable slope stability. This issue can often have major
economic consequences on the overall cost of a tailings dam.
For borrow dams, the most desirable materials for the
embankment are free draining hard durable rock or gravelly
materials. However, less desirable materials, such as mixed
soils, and decomposed shales and sedimentary rocks, might be
more economical or more readily available locally. In these
|
Colorado School of Mines
|
ER-4609 187
cases, it may be desirable to introduce a chimney and/or
blanket drain arrangement to help lower the phreatic
surface. The added cost of providing the zoning detail in
the embankment, including provision for staged enlargement,
might well be justified if the provision allows for the use
of inexpensive materials and for steepening of the
downstream slope.
Another approach for dealing with the economies of
steepening a downstream slope would include full utilization
of the observational approach. The observational approach
is well established and well accepted within the
geotechnical engineering profession, and has definite
applicability to tailings dams. In a specific project, a
system can be designed with a reasonably conservative
estimation of the phreatic level. Then as the dam is
enlarged in stages, instrumentation can be used to determine
the actual performance of the structure, including the
actual phreatic levels, stresses and displacements. If it
can be shown from instrumentation data that the original
design assumptions were overly conservative, then there will
be ample justification for modifying the design to a more
economical, but still reasonably conservative, scheme.
For dams constructed with tailings material, it becomes
more difficult to introduce zoned elements, such as internal
filters and drains. However, it is possible to introduce a
blanket drain and appropriate filter along the foundation of
the dam which extends to thé eventual downstream embankment.
Here again, tailings sands could serve as a filter or a
significant component of a filter system. Combined with the
observational approach, this method could allow for
refinement of the steepness of the downstream slope over
time.
|
Colorado School of Mines
|
ER-4609 188
Blanket drains and similar underdrain systems have a
significant additional advantage where there is a necessity
to collect hazardous tailings effluent leachate. It is
possible to design an underdrain system which also
incorporates groundwater cutoffs and doubles as a wastewater
containment and collection system. Another advantage to
lowering the phreatic level through the downstream
embankment section of the dam is the beneficial effect on
long-term durability of the embankment materials.
Physiochemical interaction is minimized, and material
degradation is significantly retarded when materials do not
have tailings fluids flowing through them in a saturated
condition. It is often possible to justify use of very low
cost "random fill zone" materials in a downstream dam
section when the material will be completely protected from
saturation by a chimney and underdrain system.
7.6 Contaminants (after Fell, 1990)
Many mine and industrial tailings have accompanying
water or "liquor" which contains dissolved salts, heavy
metals and other residual chemicals from the mineralogical
processes. If these escape to the surrounding surface and
groundwater in sufficient quantities, they can lead to
unacceptable concentrations, making the water unusable for
drinking water and affecting aquatic life. Some common
problems are :
Gold tailings — cyanide
Coal — high salts content
Copper, lead, zinc — heavy metals, sulphides
|
Colorado School of Mines
|
ER-4609 190
contaminants in the seepage water willnot all join
the groundwater. Much will be absorbed in the
foundation soil and rock. Contaminant load does
not equal flow rate x contaminant concentration in
the storage
having joined the groundwater saturated zone,
dispersion takes place,
reduction of contaminant concentration may occur
in mixing with stream flows
it is contaminant concentration in ground and
surface water which is generally critical, not the
total quantity. Hence, absorption, dispersion and
dilution can yield acceptable water quality in
streams or well points, even though the original
contaminant levels may have been unacceptable.
It is important to realize that in many cases, a
partially saturated flow condition will exist in the
foundation, at least at the start of operations and possibly
on a permanent basis if the tailings permeability is low
compared to the foundation permeability. Figure 56 shows
the stages in the development of seepage.
It may take years for the seepage mound to rise to
connect to the tailings (stage 3) or it may never happen.
Note that flow in the tailings in stages 1 and 2 will
be vertical, will not emerge at the toe of the embankment,
and will be virtually unaffected by any foundation treatment
such as grouting. In stages 2 to 4, flow will be in all
directions away from the storage.
|
Colorado School of Mines
|
ER-4609 192
In many other cases, the tailings will be stored in a valley
in which a dam has been constructed. It is usual to site
such dams at the head of a catchment so as to limit the
water runoff from the area outside the storage. Catch
drains and other measures may be used to reduce this inflow.
Figure 57 shows an example of a tailings storage. In
this case the final development will consist of several
embankments. When estimating seepage from such a storage it
is important to remember that seepage will occur under each
of the embankments, and depending on the base groundwater
levels, into the hillsides adjacent the embankments. Hence
in Figure 57 seepage will occur to the west, north and east,
but not to the south where natural groundwater levels are
higher from the storage. Note that the groundwater does not
always mirror the topography and may be affected by local
variations in geology — e.g. permeable dikes.
Many inexperienced engineers and geologists will either
forget completely that seepage will occur in all directions
from the storage, or at least apply an excessive amount of
the site investigation effort and analysis to the seepage
which will flow through and beneath the main embankment.
From the examples shown in Figures 55, 56 and 57, it
will be apparent that the assessment of seepage flow rates
will involve:
knowledge of the permeability of the tailings, as
these are commonly part of the seepage path. In
many cases they may control the seepage rates; In
Figure 57 it would be necessary to be able to
model the whole of the area between the streams,
necessitating knowledge of rock permeabilities
well beyond the storage area
|
Colorado School of Mines
|
ER-4609 194
mmodeling of the seepage, usually by finite
element methods, which may involve several
sectional models and/or a plan model. This
modeling should account for the development of
flow as shown in Figure 55, not just model an
assumed steady state coupled flow situation, i.e.
the storage and groundwater coupled as in stage 3,
Figure 56. A common error is to do this when in
fact it does not apply, and worsen the accuracy
further by assuming seepage emerges at the toe of
the dam and at ground surface, when in fact flows
downstream of the embankment are inadequate to
raise the phreatic surface to ground level.
7.7 Objectives of seepage control (after Vick, 1983)
Some general principles related to seepage control:
Not all mill effluents contain toxic constituents.
Depending on ore type, mill process, and pH,
contaminants may range from toxic heavy metals
(that is, cadmium, selenium, arsenic) to such
relatively innocuous materials as sulfates or
suspended solids.
For mill effluent that does contain toxic
constituents, it is not necessarily the case that
seepage of this effluent will result in pervasive
groundwater contamination. Geochemical processes
may retard or inhibit movement of some
constituents, and these processes are often most
|
Colorado School of Mines
|
ER-4609 195
effective in reducing mobility of the most
troublesome metallic ions associated with low-pH
effluents.
If toxic constituents do enter the groundwater
regime, the ultimate effects on the groundwater
environment must be determined before deciding on
a seepage-control strategy intended to minimize
these effects. Essential to this determination
are hydrogeologic factors, baseline water quality,
and intended use of the groundwater resource both
present and future.
From these principles, it seems reasonable that the
type of seepage-control strategy should match the chemical
conditions of the effluent and the geochemical and
hydrogeologic conditions of the site. Three types of
systems used with regard to uranium tailings seepage control
can be defined and extended to tailings in general :
Type I System. Seepage from the impoundment is
essentially uncontrolled. Due either to a lack of
troublesome contaminants in the mill effluent or
to uptake of these contaminants by geochemical
processes, groundwater contamination potential is
not serious, irrespective of seepage quantity.
Type II System. In this case, impoundment
effluent is partially retained, but some seepage
loss is anticipated. Contamination potential is
greater than for Type I effluents, and a higher
level of seepage-contaminâtion analysis is
|
Colorado School of Mines
|
ER-4609 198
Liners
7.8.1 Controlled placement of tailings
(after Fell, 1990)
In many cases the most cost effective way of
controlling seepage will be to place the tailings so they
blanket the base of the storage. Figures 58 and 5 9 show
this effect.
The effectiveness of the tailings as a "liner" is
dependent on placement methods. If tailings are placed sub-
aerially and allowed to desiccate, lower permeabilities will
result from the drying. If placed subaqueously lower
densities and higher permeabilities are likely to result.
A difficulty with this approach is that the coarser
fraction of the tailings tend to settle out more quickly
than the fine (or slimes) fraction. Hence a "beach” of
sandy tailings often occurs near the discharge point. If
water is allowed to cover this area subsequently, it can
allow high local seepage rates.
This can be overcome by shifting the tailings discharge
points from one end of the storage to the other, placing
slimes under the beach area, and/or by using a liner or
seepage collector system under the sandy area. Seepage can
also occur down the contact between tailings and embankment
if rock riprap is used.
The actual permeability of the deposited tailings can
be determined by laboratory permeability or consolidation
tests, preferably on undisturbed samples from the tailings
storage (taking account of variations within the deposited
tailings, particularly in regard to distance from the
discharge point and the degree of consolidation of the
|
Colorado School of Mines
|
ER-4609 201
tailings. Almost invariably, the horizontal permeability
will be one or two orders of magnitude higher than the
vertical permeability due to a stratification on cycling of
deposition (particularly near the discharge point). This
often leads to thin sand or silt partings on the top or
bottom of each layer (depending if the coarse particles are
low or high specific gravity).
7.8.2 Foundation Grouting (after Fell, 1990)
It is common to grout the foundation of most large
water storage dams, at least those which are founded on
rock. The grouting is usually carried out by pumping a
slurry of cement into holes drilled into the foundation.
Chemical grouts may be used to grout soil (e.g. sand and
gravel) or lower permeability rock.
In such dams the grouting is carried out to
reduce leakage through the dam foundation
reduce seepage erosion potential
reduce uplift pressures under concrete dams
(when used with relief drains)
strengthen the dam foundation and reduce
settlement (for concrete gravity and arch dams.
Grouting has a limited effect on reducing seepage,
particularly in rock which is closely fractured. The
critical issues are:
Portland cement grout is a suspension of silt and
sand size particles in water. It cannot penetrate
|
Colorado School of Mines
|
ER-4609 202
fractures finer than about 0.15-0.2mm wide. Table
20 shows estimated minimum groutable rock lugeon
values. One Lugeon ia a flow of 1
litre/minute/meter of borehole under a pressure of
1000 KPa.
The grout penetrates a limited distance into the
fractures in the rock. This is dependent on the
fracture opening and roughness, grout pressure and
time, and grout viscosity. Table 21 shows
approximate penetration of grout into fractures
based on laboratory tests and comparison with
field behavior.
Because the grout only penetrates a limited
distance, and does not render the rock
"impervious", but only reduces the permeability,
the seepage is only significantly reduced if the
rock is high permeability. A simplified
permeability analysis gives the figures in Table
.
22
Since it is not uncommon to adopt a closure criteria
(i.e. the lugeon value at which no further grouting is
carried out) for grouting of say 3 to 5 lugeons, it is
apparent that unless the rock is very high in permeability
there will be little reduction in seepage yielding little
overall reduction in seepage except in high permeability
zones in the foundation.
Chemical grouts may achieve a lower permeability in the
grouted zone than cement grout, but the penetration
distances are small.
|
Colorado School of Mines
|
ER-4609 204
TABLE 22 Effectiveness of grouting in reducing seepage,
(after Fell, 1990)
Permeability Grout Seepage with Grouting x 100 %
Permeability Rock Seepage without Grouting
0.5 90
0.2 70
0.1 50
0.05 35
0.02 20
0.01 10
On one project, the grouting of a 5km long dam to a
depth of about 25m, on average would have reduced the
estimated seepage by 1%, nearly all of this in the small
portion of the foundation affected by faulting, and having a
permeability of the order of 100 lugeons.
It will be seen from the above discussion that it is
unlikely that grouting of tailings dam foundations can be
justified on the grounds of reducing seepage. It may be
justified on other grounds, such as reducing
potentialerosion in weathered rock, or where the high
permeability zones can be identified from geological
information, allowing grouting of only those parts of the
foundation.
7.8.3 Foundation cutoffs (after Fell, 1990)
For tailings storages constructed on soil foundations,
particularly sand or sand and gravel, a significant
reduction in seepage may be achieved by construction of an
earthfill cutoff or a slurry trench cutoff wall as shown in
Figures 62 (a) and (b) . There are several types of slurry
trench wall.
Table 23 summarizes these alternatives, and their
|
Colorado School of Mines
|
ER-4609 206
potential application.
When a relatively impervious layer is at a shallow
depth below the foundation, a core trench can be used to
limit seepage. A core trench is illustrated in Figure 61.
7.8.4 Hydraulic barriers (after Soderberg and Bush,
1977)
Where the tailings dam is constructed on a thick
pervious foundation, and pollution control requirements
preclude the escape of water from the tailings pond, seepage
losses may be controlled by developing a hydraulic barrier
downstream of the tailings dam.
The hydraulic barrier (illustrated in Figure 62) can be
produced by a line of pumping wells and a line of injection
wells downstream of the embankment, the injection wells
being located downstream of the pumping wells. Fresh water
is supplied to the injection wells, while ground water is
extracted from the pumping wells. If the piezometric water
levels along the line of the injection wells are maintained
at elevations higher than the piezometric water levels along
the line of the pumping wells, a hydraulic barrier will be
formed that will prevent the flow of seepage from the
tailings pond past the line of pumping wells. This should
be checked in the field by piezometric measurements. The
hydraulic barrier may be usable for up to 100 feet of
overburden, but would not be feasible where the depth was
much greater. This method does not eliminate contamination
of the ground water in the long run, but only as long as the
injection and pumping wells are operating.
|
Colorado School of Mines
|
ER-4609 210
7.8.6 Toe drains (after Fell, 1990)
A drain may be provided at the downstream toe of the
embankment with a view to collecting seepage which emerges
at that location.
These drains can be reasonably successful in
intercepting seepage, but only if the seepage naturally
emerges in this location. In many cases the flow rates will
be such that the phreatic surface stays below the level of
the drain. Even when the seepage is sufficient to raise the
phreatic surface to flow to the drain, much may still bypass
by flowing beneath the drain. Ideally the drain has to
penetrate to a low permeable stratus, and this is often not
practicable.
Such drains may also intercept surface runoff from the
downstream face of the dam and groundwater from downstream,
and if the seepage is to be returned to the dam or process
plant, may exacerbate water management problems if a no
release system is being operated.
7.8.7 Seepage collection dams (after Fell, 1990)
In many cases, a practical way of collecting seepage
from tailings storages will be to construct a seepage
collector dam or dams. Figure 57 shows such a system.
The seepage collector dams may be located close to the
storage, sufficiently close to collect (the bulk of) the
seepage but not too far away so as to limit the external
catchment— e.g. Dam A on Figure 57. In this case one would
be anticipating pumping the water back into the dam or
process system.
|
Colorado School of Mines
|
ER-4609 211
Alternatively one may deliberately locate the collector
dam sufficiently far downstream that the runoff from the
catchment to the dam is sufficient to dilute the seepage to
acceptable water quality— e.g. Dam B in Figure 57. This has
the advantage that the seepage which occurs after shutdown
of the tailings storage is also diluted to acceptable
quality.
Whether such an approach is acceptable will be
dependent on the particular circumstances for the tailings
storage. For example, it may be unacceptable to have
substandard water quality in the stream between Dam B and
the tailings storage. Seasonal effects can also be
important e.g. If there is a prolonged dry season, water may
pond in the stream and concentration of contaminants may
occur.
In many cases the catch dam with pump-back or dilution
may be far more appropriate than expensive measures to
control the seepage. Many engineers have an over-optimistic
view of the efficacy of these seepage control measures, or
an unrealistic view of what costs a mining operation can
reasonably bear to construct such measures.
7.8.8 Liners (after Vick, 1983)
Liners constitute the final category of seepage-control
measure and are usually reserved for conditions where a Type
III system is called for because of stringent groundwater
protection requirements and relatively high concentrations
of toxic constituents in the mill effluent. Liners of any
type are inherently high in cost, but their effectiveness is
at least somewhat less in doubt than high-cost barrier
|
Colorado School of Mines
|
ER-4609 212
systems, principally because a liner is a surface
installation that can be constructed under controlled
conditions and inspected. This does not guarantee, however,
that even a properly constructed liner will function as
intended during actual operation or that any lined
impoundment will be a "zero discharge" facility. Liners do
have a major advantage over seepage barriers or seepage
return systems insofar as they are completely independent of
subsurface conditions. Whereas the effectiveness and
construction feasibility of grout curtains, slurry trenches,
and collector ditches or wells depend on the presence of a
lower impervious layer as well as the nature of the material
to be penetrated, liners suffer no such limitations and can
be constructed on any surface sufficiently dry and competent
to allow for normal earthwork operation, without regard for
the nature of subsurface soil, rock, or groundwater
conditions. Liners, however, must be resistant to both
chemical attack by the retained effluent and a variety of
types of physical disruption.
Because of the critical nature of seepage problems in
those cases where liners are required, there is considerable
debate over the relative merits and effectiveness of the
various kinds of liners and liner materials.
7.9 Liner systems and properties (after Hutchison, 1992)
Liners and liner system technologies for application to
tailings dams have made great strides in recent years, with
environmental legislation and restrictions continuing to
increase, liner systems installed in closed circuit tailings
dams will become more common. This section will identify
|
Colorado School of Mines
|
ER-4609
215
TABLE 24 Available materials or procedures for liner system
components. (after Hutchison, 1992)
A. LOW HYDRAULIC CONDUCTIVITY LINERS
A.l Low Hydraulic Conductivity Natural Soil or Rock - Natural soils or rock may be used as a low hydraulic
conductivity liner so long as it is possible to demonstrate by field investigations that the material is uniform over the entire area
requiring the liner. This demonstration may be particularly difficult for rock that is extensively fractured.
A.2 Constructed Low Hydraulic Conductivity Liners - Low hydraulic conductivity liners which are constructed
beneath a mine waste management unit may consist of any of the following materials:
• Compacted, low hydraulic conductivity soils (e g., clayey-silt to clay depending upon required hydraulic
conductivity ).
• Soil and bentonite or cement mixtures.
• Pre-formed flexible membrane liners, made from a variety of available polymer material, generally called
geomembranes; varying in thickness from about 20 to 100 mils.
• Field-applied liners, varying from about 80 mils of spray-on asphaltic materials to 6 inches of
conventionally placed asphaltic materials.
• Composite liners, consisting of combinations of soil and geomembrane low hydraulic conductivity layers.
A3 Waste Material - Settled or mechanically placed tailings often have a low hydraulic conductivity and can be used as pan
of the long-term liner system, provided the tailings serve one of the low hydraulic conductivity liner functions illustrated in
Figure 7.1.
B. CUSHION OR LINER PROTECTION MATERIALS
B.l Geo textile - Synthetic geotextile materials varying in weight from 4 to 20 ounces per square yard may be used above or
below geomembranes to protect against penetrations from rock particles due to loads from construction activities or the weight of
the waste material. The suitability of a geotextile material varies with its method (density) of fabrication.
B3 Fine-Grained Soil for Geo membra ne Protection - Soils varying from clay to sand can also be used to protect
most geomembranes from equipment traffic or static loading of the waste material. Small gravel size material has also been used
to protect thick geomembranes. The protective soil must be relatively free of large rock panicles which could cause stress
concentrations on the liner.
B3 Coarse-Grained Material for Clay Liner Protection - Gravel protection may also be required for a compacted soil
liner, if the liner may be subjected to extreme loads such as impacts from large boulders placed by the end dump method.
C. HYDRAULIC HEAD CONTROL COMPONENTS O'O'O-O
C.l Free-Draining Gravel Layer - Several inches of free-draining gravel (including coarse sand) are usually adequate
to rapidly remove small volumes of leakage. However, thicker layers (8 to 18 inches) are usually placed to facilitate construction.
The waste material itself may function for this purpose if the material is granular and relatively free draining.
Cut Perforated Pipes - Closely spaced perforated pipes can be used to control hydraulic head above a liner. The required
spacing is calculated considering the maximum desired head, and the flow rate and hydraulic conductivity of the waste material
between the pipes.
CJ Geocomposite Systems - Composite systems consisting of synthetic drainage associated with geotextile filters have
recently been developed for a wide range of drainage control functions. Presently these systems have limited load-carrying
capacity, but are routinely being improved for wider use.
D. LEACHATE COLLECTION AND REMOVAL SYSTEMS (LCRS) O O O O O
D.l Synthetic Geonet Materials - Geonets arc net-like polymer products designed to allow high rates of transverse flow
Typical thicknesses of these materials vary from 0.16 to 0.30 inches.
D.2 Free-Draining Gravel Layer (See Item C.l)
|
Colorado School of Mines
|
ER-4609 217
pervious layer above the low hydraulic conductivity liner.
To be effective, the pervious layer must be free-draining
and direct the liquids away from the liner system. Often,
the mine waste material itself is porous so that it alone
can act as the free-draining over-liner material. For
finer-grained wastes such as tailings, the free-draining
layer must be constructed from imported materials.
In design, control of the hydraulic head can be an
important consideration in determining the potential for
leakage to occur. Often it is necessary to establish a
trade-off between head control and liner hydraulic
conductivity, especially at sites where low hydraulic
conductivity materials are not readily available.
Figure 63 illustrates that cushions or liner protection
layers may also be appropriate for this type of system, if
required to avoid liner penetrations from underlying
conditions or from the free-draining material.
7.9.4 Composite liners with and without overlying head
control (after Hutchison, 1992)
This type of liner system usually involves the
placement of a geomembrane liner directly on top of a low
hydraulic conductivity soil or clay layer. To be effective,
the two layers must be in close contact so that any leakage
through imperfections in the geomembrane liner must also
pass through the soil or clay layer as a limited point
source. Without this contact, leakage through the
geomembrane can spread and hence leak through a large soil
or clay area, thus increasing the total leakage through the
liner system. Usually in mine waste application, there is
sufficient load placed on the upper geomembrane to assure
|
Colorado School of Mines
|
ER-4609 218
good contact between the two layers.
As for single liner systems, liner protection and/or
hydraulic head control layers can be located above the
composite liner system (see Figure 63).
7.9.5 Double liners with an intervening leachate
collection and removal system (after Hutchison,
1992)
This type of system assures a very low hydraulic head
on the bottom liner. This is provided by a free-draining
LCRS medium above the bottom, low hydraulic conductivity
liner, plus a low hydraulic conductivity top liner.
Acceptable quantities of leakage through the top liner
and into the LCRS should avoid the potential for
unacceptable leakage through the bottom liner. Leakage
through the bottom liner depends upon the rate at which the
LCRS material can remove any leakage which does occur and
the hydraulic conductivity of the bottom liner. In many
cases, it is possible to remove this leakage so that the
head on the bottom liner is just a fraction of an inch in
those areas exposed to top liner leakage, and negligible or
zero elsewhere. It is important that the LCRS removal rate
be adequate to avoid the buildup of a continuous column of
water between the top and bottom impervious liners.
Otherwise, the hydraulic pressure above the top liner would
be transmitted directly through to the bottom liner, and the
double-lined system would have essentially no benefit toward
reducing the potential for leakage through as a result of
the hydraulic head.
|
Colorado School of Mines
|
ER-4609 219
7.9.6 Drainage system above double liners, with an
intervening collection system (after Hutchison,
1992)
This system provides an additional level of assurance
that the liner system will not be exposed to excessive
hydraulic head. The top, free-draining layer, which can be
the waste itself, reduces the hydraulic pressure on the top
liner and therefore reduces the potential for leakage to
occur into the LCRS between the liners. The reduced leakage
into the LCRS, in turn, reduces both the thickness of the
film of water which may flow above portions of the bottom
liner, and the resulting potential for leakage to occur into
the underlying unsaturated zone.
7.9.7 Low conductivity liners (after Hutchison, 1992)
Tables 25, 26, and 27 provide summary data for a
variety of low hydraulic conductivity liner materials which
can be applied to mine waste management units. The liner
types are divided into categories of low hydraulic
conductivity soils, geomembranes, field-applied liners, and
composite systems.
7.9.8 Basic liner properties (after Hutchison, 1992)
Table 25 summarizes typical hydraulic conductivity
values for a variety of soil and synthetic materials. Table
26 indicates typical physical properties for commonly-used
low hydraulic conductivity synthetic liner materials. Table
27 qualitatively indicates problems which can occur for each
liner type and indicates the potential degree of difficulty
|
Colorado School of Mines
|
ER-4609 222
uonipiitiQ praifofjig « 6 ° o O ° ° o o 0 «
uonipuSaQ O O O o O o o O o o O o o o o O o O o o O
s
Iui^wq uoanraaQ ® © ® © © © o
Xupqnsui adoy • • • • • • • • •
o
o o o o o
o
lunia*s Xq i$raiiu| Xiqiqtimty o o © ® o
uoaipulaa |rjnu*o • • • • e • o o o o • 0 o o o o o o o o o
luiptap) may uoninnzy © # # o © o ® @ ® © s « ©
iiniudmbg uioy nentrouy # # ® © © © o # ® © ® ® ©
uoniuusjaO oi.ana •UT13«'D o • © o © • o O O ° ° 0 o o o o o o o ° °
qsniyiwquv o • o o o ° o o o o o o o o o
uon»put>a wjowmin © © o o @ o o © 0 o o © © o
!
Xiipqmui nfcis o O o o o o o
«ny o ° o o o o o
i
t Itnqjuj «tans jtuioqi • o o 0 o o o o o o o o o o O
Sunpuo «irqyœay # # © @ © o
Ï
luiptO uontaroaQm # © @ # © o
o
Xiiuuojiu^ |>uanH UD OCVYÛ o # o o o o o o o o o o o 0 00 0 O 0 o
ipq or yu ua r« oK y t uu om nt up pV u < y" 0 @ © © © © 6 e # © o •
iiomdmbg may uontnouy # @ ©© O © o # e © © # « o
uontuuopa m joq tuppuj o ° o o o o o O ° ° o 0 o 0 o o o o o ° °
limpuo *eiu/*oay # ©# ® 0 o o
‘0
tuiqnr) uoanWBQ©# @®© o o © •
o
I © •
1 Î! 1 © © •
1 Î
1 SI 1 ! I 11 o o *
1 i
JG 5111
j
9 i1 II »1I
1
!! 1 6 I ° ° •
1 i i i 1
s & B s I Ï u »
Qpaatapiqaai) (pumpmtuooQ)
Al olA nU nS D vT nI wC OK iAN HOO siQN nn o aN xw ruawaw smm 03nddv<mu US JW SO1L iS KÀ OS D
OHWUOdTUd
ytivitcudnoc
ciluardyh
wol
rof
smelborp
laitnetoP
72
ELBAT
)2991
,nosihctuH
retfa(
srenil
DETALER
NOITAREPO
UUTAIER
NOITCURTSNOC
METSYS
RENIL
|
Colorado School of Mines
|
ER-4609 223
that may be realized in attempting to solve each problem.
7.9.9 Potential problems with low hydraulic
conductivity liners (after Hutchison, 1992)
Table 27 qualitatively illustrates potential problems
which may be realized with various low hydraulic
conductivity liners. The size of the circle in the matrix
indicates the relative potential for the identified problem
to adversely affect liner performance. The darkness of
shading in each circle indicates the degree to which special
efforts may be required to resolve the difficulty.
The most consistent potential problem with low
hydraulic conductivity soils is associated cracking, from
the effects of either desiccation (drying) or freeze/thaw
cycles. Desiccation cracking can occur up to several feet
deep with a substantial increase in hydraulic conductivity.
Freeze/thaw cracking is more difficult (costly) to
protect against because a greater cover thickness is
required. For this reason, compacted soil liners may not be
appropriate when large portions of liner will be leftexposed
throughout freezing winter conditions.
Additional important potential problems with low
hydraulic conductivity soil/rock liners are :
QA/QC verification of liner continuity for
natural, in situ soil and rock liner conditions.
This is particularly important for rock formations
that may be fractured due to factors such as
stress-releasing tectonic folding, differential
weathering, etc.
|
Colorado School of Mines
|
ER-4609 224
Soil and rock hydraulic conductivity can be
unaffected or greatly increased due to the
chemistry of the liquid to be contained. High pH
liquids or liquids which contain high levels of
dissolved salts which could exchange with ions in
the soil, can cause hydraulic conductivity
increases in soil liners. The most severe
potential impacts are associated with high
concentrations of organic solvents. However, this
is not generally important, as most mine waste
management units are not exposed to organics.
The most important potential problems for most of the
geomembrane liners are associated with the possibility for
penetration of the liner as a result of construction
activities or waste material loading. These types of
problems are most easily mitigated by the provision of
cushioning (or protection layers) below and/or above the
synthetic material to reduce the effects of concentrated
loading conditions.
Polyvinyl chloride (PVC) liners also can be subjected
to accelerated deterioration if exposed for long periods to
sunlight or to certain organic chemicals.
Composite systems, which consist of a combination of
compacted soil and a geomembrane, have less potential for
either penetration or cracking problems than either material
alone because the dissimilarities in the two provide
redundancy. The system least likely to suffer from
potential penetrations is one in which a geomembrane is
sandwiched between two compacted soil layers. This type of
system can only be economically feasible, however, when the
low hydraulic conductivity soils are readily available and a
|
Colorado School of Mines
|
ER-4609 226
7.9.11 Hydraulic head control layers (after Hutchison,
1992)
Provisions to control hydraulic head above a low
hydraulic conductivity liner are important in minimizing the
potential for excessive leakage. This is accomplished by
providing a free draining system above the liner, with the
primary function of removing liquids without the need to
establish a large hydraulic gradient. Table 23 provides
descriptions of three basic types of hydraulic head control
systems commonly used in mine waste management practice :
free-draining gravel, perforated pipes, and geocomposite
systems. The rate at which liquids are removed from any of
these systems depends on: (1) the system's hydraulic
conductivity, (2) the slope of the waste management unit
base, and (3) the volume of liquid involved.
7.9.12 Leachate collection and removal systems (LCRS)
(after Hutchison, 1992)
Instead of controlling the liquids above the low
hydraulic conductivity liner systems, the primary function
of an LCRS is to detect and collect liquids that may pass
through a top, low hydraulic conductivity liner. As shown
in Table 23, an LCRS is. placed immediately between two low
hydraulic conductivity liners. Therefore, liquids that pass
through the top liner can be removed with only a fraction of
an inch of hydraulic head buildup on the lower liner.
An LCRS usually consists of the following components:
The drainage layer itself
|
Colorado School of Mines
|
ER-4609 228
of water, especially for ground water at depth.
For a given hydraulic head and unsaturated zone
characteristic, a geomembrane or composite liner
will perform better than a single clay (compacted
low hydraulic conductivity soil) liner. Composite
liners, consisting of both geomembrane and clay,
provide greater protection than either material
alone.
A liner system performance can be predicted best
by its overall design and construction QA/QC
procedures, and not simply by its component
specifications (e.g., thickness). For example, if
a liner will be subjected to constructed or
operational loads, geomembrane components may
require protective layers below and/or above the
liner, to avoid penetrations. In those cases, the
properties of the protective layers and not the
geomembrane thickness often will be the most
important design criteria.
Clay liner performance does not increase
proportionately to its thickness, but potential
leakage does decrease as the thickness is
increased. The thickness of the clay liner should
be designed to be sufficient to avoid leakage
breakthrough which poses a threat to the
beneficial uses of ground water during operation
of the waste management unit. At sites with
limited availability of clay, the minimum
practical thickness of compacted soil is
warranted. Based on experience, this is
|
Colorado School of Mines
|
ER-4609 229
considered to be 1 foot, although with special
construction, considerations 8 to 9 feet can be
utilized.
The potential threat to beneficial uses of water
can vary greatly, depending on operational factors
such as the duration of waste unit operation and
controls employed to reduce the hydraulic head
pressure exposed to the liner. Low hydraulic
heads can be maintained by providing a control
component above the low hydraulic conductivity
layer, or a double liner with an intervening LCRS.
The potential threat to beneficial uses also will
vary depending on liner system design and the
condition of the unsaturated zone below the waste
management unit. Potential migration periods
through the unsaturated zone can be tens or
hundreds of years with a properly designed and
constructed geomembrane liner and verification by
state-of-the-art QA/QC procedures, especially when
ground water is deep and the unsaturated zone
material has relatively low hydraulic
conductivity. The rate of migration and
probability of seepage reaching ground water
depend upon the hydraulic conductivity of the
unsaturated zone and the depth to water.
The need to provide redundancy within the liner
system or to rely upon natural subsurface
conditions will depend on waste and site-specific
factors. At some sites, liner system construction
|
Colorado School of Mines
|
ER-4609 230
and operating procedures may provide basic
containment so that the unsaturated zone can be
relied upon top rovide redundancy, with an
adequate factor of safety. For example :
Double liners may not be necessary at
many sites if hydraulic heads are
maintained sufficiently low.
Small rates of leakage into the LCRS for
double liners can be acceptable so long
as it is collected and removed to avoid
significant head buildup on the lower
liner.
Uniform design specifications either are likely to be
inappropriate, with insufficient containment for sites with
unfavorable conditions, or too restrictive, resulting in
over-regulation of sites with favorable conditions. This
need for flexibility is consistent with the conclusion
reached by the ERA in the development of draft criteria for
designing Municipal Solid Waste Landfill(MSWLF) containment
systems. The ERA studies showed that a site-specific, risk-
based approach would be most appropriate. In reaching its
conclusion, the ERA specifically recognized the importance
of climate and geologic site factors, including the effects
of these factors on travel time from the waste management
unit to ground water. Table 28 summarizes factors that
should be considered for site-specific liner system designs.
|
Colorado School of Mines
|
ER-4609 231
TABLE 28 Site-specific factors to consider in liner system
design (after Hutchison, 1992)
Potential Waste Material Toxicity
Chemical Properties of the Waste
Physical or Chemical Changes REsulting from Mining or Ore Extraction
Net Acid Generation Potential
Soluble Constitutents for Anticipated Environmental Conditions
Special Treatment or Neutralization Procedures Utilized
Total Resulting Mass of Soluble Constituents Which Could be Mobilized Under Site Conditions
General Water Resource Values at the Site
Adequate Quality for Beneficial Use
Sufficient Quantity for Beneficial Use
Existing or Identified Beneficial Uses
Probable Locations of Future Beneficial Uses
Leachate Availability to the Environment
Waste Material Characteristics
Hydraulic Conductivity Based on Direct Measurement, Laboratory Tests, or Grain Size and
Density
Moisture Retention Capacity
Thickness of the Waste
Site Climatic Conditions
Provisions at Closure to Restrict Infiltration
Site Factors
Topography
Geology, Including Predictability of Uniformity and/or the Potential for Discontinuities
Unsaturated Zone Thickness, Contiinuity, Hydraulic Conductivity and Natural Water Content
Potential Migration Time for Seepage to Ground Water
Effects of Climatic Conditioins on Long Term Unsaturated Zone Migration Characteristics
Constituent Attenuation Potential
Waste Unit Management Practices
Facility Type
Waste Placement Method
Protection of Liner Systems From Environmental or Physical Damage
Controls on the Hydraulic Head
Risk Reduction Practices, Such as Placement of Underdrains, Sub-Aerial Depositions, Limited
Time of Operations
Non-Liner Barriers, Such as Cutoff Walls
Installation of Special Early Warning Monitoring Systems
|
Colorado School of Mines
|
ER-4609 232
CHAPTER 8
SLOPE STABILITY
In chapter eight, sections 8.1, 8.3, and 8.6 were
compiled from the Bureau of Mines information circular 87 55,
Design Guide for Metal and Nonmetal Tailings Disposal,
written by Roy L. Soderburg and Richard A Busch. Section
8.2 was extracted from Steven G. Vick's book titled
Planning, Design, and Analysis of Tailings Dams. Sections
8.4, 8.5, and 8.7 were compiled from the book Geotechnical
Engineering and Soil Testing, written by Amir Wadi Al-
khafaji and Orlando B. Andersland. Section 8.8 was
extracted from an article titled Seismic Assessment of
Tailings Dams, written by Thomas G. Harper, Harvey N.
McLeod, and Michael P. Davies.
8.1 Introduction to Slope Stability (after Soderberg and
Bush, 1977)
Since no slope can be regarded as permanently stable,
slope stability is a relative matter. However, in soils
engineering practice, the term is used in reference to the
possibility of a sudden relatively deep-seated slide.
Soil and rock materials fail in shear if the applied
shearing stresses on any surface exceed the shear strength
of the materials along that surface. The resisting forces
are the shear strength of the materials, both frictional and
cohesive. The cohesive strength is minimal or negligible in
most cases. Pore water pressure at the failure surface
lowers the resistance to sliding because it reduces the
ARTHUR LAKES LIBRARY ^
COLORADO SCHOOL OF MINES
GOLDEN, CO 80401 x
|
Colorado School of Mines
|
ER-4609 233
effective stress. Artesian water from the substrata into
the embankment will also reduce the stability of the
embankment owing to hydraulic uplift. The shear strength of
the materials can be further reduced by weathering and
softening by water. The shear strength may be increased by
compaction or by chemical cementing of the waste materials.
Cracking of the embankment caused by differential
settlement can reduce the shearing resistance along
potential failure surfaces. This cracking may lead to slide
failures or piping. Dense tills are usually strong, their
shear strengths can have both frictional and cohesive
components, and they may be relatively impermeable and
incompressible. Drilling and sampling are necessary to seek
out inconsistencies in the materials.
Sands and gravels are relatively incompressible, and
their shear strength is primarily frictional with no
cohesion. Here again, the density, gradation, and particle
shape determine their behavior. Loose, fine sand acts the
same as the same gradation material in mine tailings. If it
is saturated and below the critical density, it is subject
to liquefaction under shock load. Silts develop strength
from either friction or cohesion, depending on density,
gradation, and moisture content.
Clay in the foundation may cause embankment settlement
and instability. As the embankment rises, the clay may
consolidate and gain shear strength. Uncompacted clays in
waste piles saturate and swell, reducing their shear
strength to almost zero. The finer portions of tailing from
metal mines act as clays. The entire output of some mines
is mudstone or clay and requires specially designed dams to
contain it. Phosphate slimes are also a special case
because of the fineness of the clay and the pulp density of
|
Colorado School of Mines
|
ER-4609 234
the material being impounded.
8.2 Phreatic surface determination (after Vick, 1983)
Seepage conditions within the embankment exert a
controlling influence on its stability, and a primary
purpose of seepage evaluation is to assess pore pressures
for input to stability analyses. Seepage in tailings
embankments is commonly assumed to occur under gravity flow
and, for purposes of pore pressure evaluation, is usually
determined for steady-state conditions.
It is important to note that the Darcy assumptions of
steady seepage conditions and gravity flow are useful in the
context of stability analysis because theyu sually yield
conservative estimates of pore pressures.
The most common procedure for determining phreatic
surface location is by flow net analyses.
8.3 Rotational Slides (after Soderberg and Bush, 1977)
A somewhat idealized concept of a rotational slide is
shown in Figure 64. The crosshatched areas of the figure
represent a cross section of the sliding mass. The surface
on which sliding occurs is curved and may often be
approximated in cross section by a circular arc. The
sliding tendency is created by the moment of the mass about
the center of the arc. This moment is opposed by shearing
resistance developed along the sliding surface. When all
available resistance is overcome, failure occurs as shown in
the bottom panel of Figure 64. Two pictures of a typical
rotational slide are shown in Figure 65.
|
Colorado School of Mines
|
ER-4609 237
See Table 29 for procedures for stability analyses
TABLE 29 Generalized procedures for performing tailings
embankment stability analyses. (after Soderberg
and Bush, 1977
Step Method
(1) Select trial embankment slope configuration Experience and judgment
(2) Determine phreatic surface location based Flow nets
on internal zoning, material permeabilities, Numerical models
and boundary conditions Published solutions
(3) Establish whether or not initial excess Compare raising rate to pore pressure
pore pressures will result from embank dissipation rate for tailings or soft
ment raising foundation soils
(4) Perform stability computations for Use any of several available computational
applicable conditions methods after defining loading ^
cases for analysis, and appropriate
strength behavior under drained and
undrained conditions
(5) Return to Step 1 and revise trial configuration
if factors of safety are not adequate
Hand calculations for the factor of safety of a given
embankment by the "method of slices", utilizing trial and
error, is a long and tedious process.
8.4 Stability of homogeneous slopes (after Al-khafaji,
1992)
When dealing with slopes in homogeneous soil deposits,
it is possible to derive a general expression similar to
those developed for infinite slopes. All other cases
require use of approximate numerical or graphical
techniques. The factor of safety in all approximate methods
of analysis for finite slopes is defined in terms of moments
about the center of an assumed circular failure arc. This
concept is illustrated graphically in Figure 66. The
analysis of a finite slope in any soil can be made by first
|
Colorado School of Mines
|
ER-4609 239
Thus far, the factor of safety has been calculated for
only one failure surface. To determine the minimum factor
of safety we must calculate the FS for many trial failure
surfaces by assuming other radii and centers. The proper
factor of safety corresponds to the lowest value computed
for all of the assumed failure surfaces. This concept is
depicted in Figure 67. Here, a grid representing the
centers of all circles to be investigated is first
specified. The radius at each of the nodes is varied so
that the lowest factor of safety for circles with radii
ranging from bedrock to the top of the slope is normally
investigated. The lowest value is then placed at the node j
(j=l,2,... ,m) , where m is the number of nodes in the grid.
This procedure is repeated for each node, after which
contours of equal factors of safety are drawn and the lowest
factor of safety is determined. For the case shown in
Figure 67, the soil cross section involves stratified soil
layers. The individual soil layer i is homogeneous within
its thickness Li. The implication is that the weight of
each stratus within the assumed failure surface must be
determined separately. The factor of safety for a given
center j is then given by:
n
T TiLi
(FS)J = H = J=1.....m <3)
The stability of homogeneous finite slopes can be
determined using stability charts. Typical stability charts
for homogeneous clay slopes are shown in Figures 68 and 69.
These charts have been extended to include the influence of
surcharge loadings, submergence, and tension cracks as shown
|
Colorado School of Mines
|
ER-4609 243
in Figure 70. These charts have one thing in common, they
apply to finite slopes in homogeneous soils. Nonhomogeneous
soils require a more elaborate analysis.
8.5 Total stress analysis (after Al-khafaji, 1992)
Not all materials are going to be a clay with T = c =
constant along the failure surface. Therefore, we must look
at materials with s = c + a tan 0 . Note that it is also
possible to use effective stress strength parameters.
Stability chart solutions are available for soils with these
strength characteristics as shown in Figure 71. This chart
is extended to include the influence of surcharge loadings,
submergence, and tension cracks in Figure 70. The method of
slices deals with slopes in soils with variable shear
strength along the failure surface.
The method is extremely versatile and is based on
dividing a failure mass into several vertical slices.
Consider the problem using total stress strength parameters
(strength = i = c + a tan 0) . This assumption generally
implies that the pore water pressure is not known. Consider
the finite slope shown in Figure 72. The factor of safety
against stability failure is defined as before using
Eg.(2).Assuming the failure mass consists of vertical slices
only, the free body diagram for the ith slice can be
determined as shown in figure 73.
Since the slice is statically indeterminate, assume
that the resultant of EL and SL is equal to the resultant of
Er and SR and that their lines of action coincide. This
assumption is not a bad one, especially if the failure mass
is subdivided into several slices.
|
Colorado School of Mines
|
ER-4609 252
Substituting Eqs.(ll) and (12) into (2) yields the desired
factor of safety
n
-U1X1-CT2X2 + r Y / W^sinocj
2=1
When hand calculations are necessary, tabular summary
sheets are generally used. Otherwise, computer programs are
used for analyzing the majority of practical problems.
8.6 Computer input for factor of safety calculations
(after Soderberg and Bush, 1977)
In most cases it is extremely difficult, if not
impossible, to obtain soil samples that are truly
representative of the zone being studied. Consequently, the
soil properties developed from these samples must be
interpreted and applied with great care. Assuming that
input values developed are representative of the actual case
being studied, the computer factors of safety are general
guidelines and are meaningful only if used in conjunction
with all of the other design considerations.
The engineer must anticipate and design for the worst
possible situation; that is, ultimate height, maximum
phreatic line, saturated soils, and seismic activity. The
safety factor will only be meaningful if such values are
used for the computer program.
Grain-size distribution, the area of the tailings pond,
and the rate of discharge have much to do with the stability
|
Colorado School of Mines
|
ER-4609 253
of the embankment. A finer grind in the mill with the
coarse fraction being taken out for use as underground fill
will make dike building more difficult. Combine these
factors with the small impoundment area compared with the
tons of waste per day, and the situation becomes worse. The
tailings used underground at some properties do reduce the
total that must be impounded on the surface by 40 percent or
more, but they also remove the coarse sand (the best
material for building the dike) and the coarse sand beach,
which provides added safety.
An example of rapid building is a situation in which a
500-ton-per-day operation is impounded in a 5-acre site with
a pond rise of 1 foot in 33 days. Even a 10-acre site with
a 1-foot rise in 66 days is much too fast. Such situations
could cause a rapid increase of pore pressure because the
water does not have time to percolate through the fine
material. Even with the best of conditions, a rapid
building rate is not good, and every effort should be made
to keep the annual rise compatible with the seepage ability
of the soil or drains.
Piezometers installed in proper places in the
embankment will allow monitoring of the pore-water pressure
in the dam, which can be related directly to the safety
factor as shown in Figure 77, a typical graph showing the
variance of safety factor with phreatic water height in the
dam. This type of chart can be developed for any dam and
used by the operators to predict the safety of the
embankments. The phreatic surface is related to the rate
that material is placed around the periphery of the dike.
Safety factors cannot be accepted at face value if there is
the possibility of liquefaction from either earthquake,
sonic blasts, or sudden load. Soils in the 80- to 280-mesh
|
Colorado School of Mines
|
ER-4609 255
sizes, saturated and above the critical void ratio, are very
sensitive to liquefaction. Soils finer than that would be
too sluggish in their reaction to shock because of lower
permeability, and those coarser would dissipate the water
fast enough to make failure from shock unlikely.
All soil testing requires stringent testing conditions
and experienced personnel. Since Coulomb's theory and
stability equations are approximations, these samples have
to be studied thoroughly by experienced soils engineers to
insure that true values are obtained for any stability
analysis. Once these values have been interpreted, it is
simple to predict the stability using the computer.
Ultimate height, slope, and water movement can be determined
by soil engineering analyses. Before any major construction
of this type, these analyses should be made either by the
operating firm or by a consulting agency. The cost is only
a few percent of the total investment.
8.7 Recommended factors of safety (after Al-khafaji, 1992)
Table 30 gives factors of safety suggested by various
sources for mining operations. All of these factors are
based on the assumptions that the most critical failure
surface is used in the analysis, that strength parameters
are reasonably representative of the actual case, and that
sufficient construction control is ensured. There is no
substitute for a sound sub-surface investigation and for a
credible laboratory program for soil property determination.
The lower the uncertainty, the lower the factor of safety
required to ensure safety.
|
Colorado School of Mines
|
ER-4609 256
TABLE 30 Factors of Safety Suggested for Mining Operations
(after Al-khafaji, 1992)
United States (Federal Register, 1977) Minimum Safety Factor
I End of construction 1.3
II Partial pool with steady seepage saturation 1.5
III Steady seepage from spillway or decant crest 1.5
IV Earthquake (cases II and III with seismic loading) 1.0
Design is based on peak shear strength parameters 1.5° 1.3*
Design is based on residual shear strength parameters 1.3° 1.2*
Analyses that include the predicted 100-year return period
accelerations applied to the potential failure mass 1.2° 1.1*
For horizontal sliding on base of dike in seismic areas assuming
shear strength of fine refuse in impoundment reduced to zero 1.3° 1.3*
°Where there is a risk of danger to persons or property.
*Where no risk of danger to persons or property is anticipated.
8.8 Seismic Assessment, (after Harper, 1992)
A significant portion of the operating and closed large
dams in North America are tailings dams used to impound
wastes from mining operations. Often these dams were built
by mine personnel, and may not have been subjected to the
rigorous quality control measures required for conventional
water storage dams. Construction methods and materials vary
considerably.
Tailings dams’ stability in earthquakes has become an
increasing concern as seismic knowledge has advanced. In
Chile, failure of the new and old El Cobre tailings dams
following the La Ligua earthquake in 1965 killed 200 people.
Growing awareness of seismicity and the potential liability
associated with tailings dams are raising the requirements
for assessing the stability of both closed and operating
structures.
|
Colorado School of Mines
|
ER-4609 257
The main seismic design factors for tailings dams are
the governing site seismicity and the properties of the
tailings. Site-specific earthquake-induced motions can be
calculated either deterministically (based on the largest
earthquake thought possible given the geological evidence
within the region) or probabilistically (by estimating
ground motions and their annual probability of being
exceeded) .
Tailings behavior under seismic loading depends on the
soil characteristics and the size and duration of the
seismic event. The tailings and foundation materials may
behave in either a brittle or a non-brittle fashion.
Several soil characteristics influence the dynamic behavior
of hydraulically placed tailings and foundation materials,
including relative density, modulus, fabric of soil, and
aging.
When planning new impoundments or raising existing
ones, designers should look for a deposition method or
construction technique that optimizes the impoundment’s
seismic resistance. These include using a rockfill toe
berm; constructing a dam with cycloned tailings to use the
coarser, more liquefaction-resistant fraction in the dam;
depositing tailings by methods that ensure that coarse,
drained tailings form the dam; or compacting tailings to
increase liquefaction resistance.
However, selecting the method most appropriate for the
anticipated seismic conditions at a given site requires
broad experience. Deposition planning that satisfies
seismic requirements while not ignoring space restrictions
and environmental and economic considerations is very
challenging, and must take into account impoundment
geometry, seepage control measures, tailings deposition
|
Colorado School of Mines
|
ER-4609 258
mode, hydraulic fill composition and tailings line density.
8.8.1 Methods of analysis (after Harper, 1992)
Once site seismicity has been evaluated, the next step
is to assess the significance of the design seismic loading
on the tailings structure.
If large-scale brittle behavior is not considered
likely, a deformation analysis is recommended. It may be
argued that rigorous deformation analyses are more suited to
formal water-retention dams, but tailings impoundments are
often just as susceptible to catastrophe if significant
crest settlement occurs. Many have free water or slimes
contained with minimal freeboard and require permanent
deformation analyses to ensure that the freeboard is
sufficient.
If tailings behavior is potentially brittle, some form
of seismic stability analysis is required. The level should
depend on the consequences of failure in loss of life,
property damage or mill downtime. Site seismicity can be
evaluated empirically by examining relationships between
earthquake magnitude and the furthest liquefied site or
through one of a number of analytical procedures requiring
an ascending degree of effort :
Residual-strength, or steady-state, analyses use
the concept of minimum post-earthquake strength
with the well-accepted methods of conventional
limit equilibrium analyses. This type of analysis
is typically conservative, and if an adequate
post-earthquake stability is determined using it.
|
Colorado School of Mines
|
ER-4609 260
now, this level of analysis is typically reserved
for major tailings impoundments whose failure may
cause extreme economic loss or potential loss of
life; as the body of experience grows effective-
stress finite-element analyses will become a
commonplace design tool.
A companion process to analyzing the stability of
tailings structures is the quantification of the risk of
failure. In cases where a deterministic assessment
indicates that the dam will fail, it is important to
quantify the risk. This can be done on a probabilistic
basis, in which the probability of seismic events is coupled
with the behavior of the dam to calculate the annual
probability of failure.
One method uses the program PROLIQ (from the University
of British Columbia), which calculates the probability of
seismically induced liquefaction for level ground sites.
Modifications can be developed for sloping ground. Program
input involves field data and the results of a regional
seismicity evaluation. A slight modification of the PROLIQ
model has been used successfully on major tailings projects.
Other methods are available but are all essentially
similar. Regardless of the methodology, the results of a
probabilistic event should be carefully rationalized and
used with extreme engineering judgment.
8.8.2 Seismic upgrading (after Harper, 1992)
The requirements for seismic upgrading must be assessed
according to: (1) impact of the failure on people and the
|
Colorado School of Mines
|
ER-4609 261
environment; (2) cost and feasibility of upgrading works;
and (3) the risk or probability of failure. To decide
whether the risk of failure is acceptable requires
determining the risk criteria.
For a deterministic assessment, it is commonly accepted
that the dam must have a factor of safety of over 1.1 for
the maximum credible earthquake (MCE) if its failure could
cause loss of life or extensive environmental damage. The
MCE criterion normally uses the limit equilibrium method for
stability analysis, and it is commonly applied to water-
storage dams. A main concern with tailings dams is the
potential for liquefaction, and this requires closer
integration with the risk of liquefaction. Criteria for a
probabilistic assessment has not been rigorously set. An
annual probability of less than 0.0001 is currently
operative for numerous water-storage dams in the U.S. This
could be regarded as a minimum target criterion for tailings
dams.
The impact of a failure depends on the runout distance
for the failed section. Runout assessments have been
carried out for numerous failed structures worldwide and
form the basis for current estimations. The basic options
available for seismic upgrading are : (1) drainage; (2)
densification; or (3) construction of buttress/containment
structures.
Drainage is a very important parameter and can often be
improved at a relatively lower cost than other options. It
can, however, be very difficult because of the typically
fine grain size of the tailings and the fact that loose
deposits at the base of the pond could require pumping if
gravity drainage is not feasible. Densification methods
include dynamic compaction, vibro-wing and vibro-
|
Colorado School of Mines
|
ER-4609 264
Pressure cells.
9.2 Piezometers (after Kealy, 1972)
Incorrect choice of the piezometer has caused
unsuitable monitoring in the past because of mechanical
problems or incorrect pressure ranges. A laboratory testing
program was initiated by USBM to study the characteristics
of various types of piezometers. The knowledge gained
should be useful in selecting a suitable piezometer for
various water-monitoring applications.
In active tailings ponds the slime zone is not readily
accessible; therefore, when piezometers are placed in these
areas, lines have to be run horizontally to a reading
station on the bank or some other solid footing. In these
installations only hydraulic, pneumatic, or electric
piezometers can readily be used. Standpipes and other
Casagrade-type piezometers cannot be used because of the
lack of access to the piezometer and air-locking problems.
Four pneumatic piezometers, one electric vibrating-wire
piezometer, and one electric strain-gage piezometer were
obtained. All were tested under laboratory conditions and
monitored by instruments more sensitive than the piezometers
themselves. Table 31 was compares the six piezometers.
Figure 7 8 shows piezometers installed to sample pore
water pressures in tailings. Figure 7 9 shows piezometers
monitoring changes in water quality in tailings dam seepage.
|
Colorado School of Mines
|
ER-4609 265
TABLE 31 Comparison of piezometer devices by weighting
system (after Kealy, 1972)
— E
UH
5 1 E u
ill 3 JZ < c
2 S1 ?
r
Cost of piezometer j 3 ! 5 i 2 * | 6 1
Cose of readout 1 ! 5 ; 3 j A j 6 2
Total cose 2 i1 5 1i3 1
T
Ease of reading (hookup and 1 reading) 3 Î 4 ! 1 6 ! 5 2
Repeatability (total average standard deviation) 1 4 5 3 !
2
Accuracy at maximum reading 1 ! 3 5 ‘ i 2
. 1
Accuracy at maximum reading as percentage of \ ? ! , j ,
total range tested j ~ ! , 1
i i 6 !
Accuracy at mid-range reading 4 ! 2 j 1 3
! 5 i
Accuracy at mid-range reading as percentage of j , ! j
total range tested | 'l j V ^ 3 5
• i 1
1
Ease of calibration ! i 2 3 1
* h
Ease of determining head from reading 1 1 3
2 1
Slop lie ity of readout j 1 2 3 ! 1
i i
Water sample possible 5 1 ; 5 5 ! 5 j 3
1
Time lag 1 l | 2 I 3 | 2 | 5
Volume water required for head measurement 1 2 j ? ; ?! 6
Se1f-conrained 1 2 I
2 1
Diameter hole required (1 = rain hole) 3 1 4 j 2 j 3
-------------------------------- - 3 1
1 = roost favorable
6 = least favorable
cirtcelE
___)ecnatsiseR(__
cirtcelE
)eriW
gnitarbIV(
|
Colorado School of Mines
|
ER-4609 267
9.3 Slope movement (after Soderberg and Bush, 1977)
There are inherent hazards in constructing and
maintaining mine tailings embankments. These hazards are
difficult to recognize and quantify. Inspectors are trained
to look for certain "telltale" signs of hazardous
conditions, but often factors such as creep or slow local
failures are not noticeable without the ability to relate to
prior embankment conditions.
Numerous systems are available for observation of
ground movement and also for mapping areas of interest.
Slope indicator devices can monitor small movements of a
slope, and standard ground surveying techniques can be
applied to mapping. These standard surveying techniques,
which enable mapping and locating of specific points on a
structure, are time consuming.
One of the simplest but most effective ways to measure
the movement of embankment is to drive pipe or rebar
vertically into the berm of an embankment in a straight line
of sight so that any movement can be measured by simply
measuring the offset from the line of sight.
On cross-valley dams the permanent stations can be
placed on the natural ground or rock off the embankment with
the recording stations set on an abandoned berm where they
can be protected from the spigoting or dike building
operation. These should be measured for elevation changes
also. In flat country the survey line must be brought up to
the embankment from the natural ground at each end of the
downstream slope, and the movement markers placed the same
as described for the cross-valley dams. This is important
because the line of site must be made from points on top of
the sand berm, and these points themselves could have
|
Colorado School of Mines
|
ER-4609 268
considerable movement which then would not show the true
movement of the other points. The permanent points on the
natural ground at each end should be placed at a
considerable distance from the embankment so as to be
unaffected by the foundation deformation.
Another device for measuring both horizontal and
vertical movement is the slope indicator, which is lowered
into a casing with grooves which guide it and measures the
movement in two directions at right angles. For this
instrument the casing must be set into bedrock, or if the
alluvium is deep, the bottom of the casing must be set into
the foundation and the casing installed in the embankment as
it rises. Another way to install the casing is to drill a
hole and install the casing after an embankment is
constructed.
From the readings of the slope indicator a profile of
the hole can be drawn showing where the movement is taking
place from top to bottom (Figure 80).
9.3.1 Other detection methods (after Kealy, 1972)
As an aid to inspection personnel, a faster and more
reliable reconnaissance system must be integrated with a
suitable method of cataloging and evaluating the gathered
data.
In August 1975, CH2M Hill was awarded a contract to
identify, develop, and test a rapid system for monitoring
coal refuse embankments to aid in inspections. The rapid-
monitoring system has been shown to be an effective,
economic, and powerful monitoring tool. The system obtains
high accuracy by using convergent and vertical photography
|
Colorado School of Mines
|
ER-4609 270
from conventional fixed-wing aircraft. Either black and
white (B&W) or color-infrared (CIR) film may be used; B&W
film is superior from the standpoint of accuracy, while the
CIR film offers better evaluation of qualitative stability
indicators through photointerpretation techniques. The
rapid-monitoring system also provides a method for
maintaining a permanent, visual record of site conditions.
The rapid-monitoring system of aerial photography
proved to be practical for detecting and recording movement
of targets on slopes and embankments. The system uses
existing technology and conventional aerial photography with
fixed-wing photo aircraft* Targets for the rapid-monitoring
system are readily available, inexpensive, and easy to
install. A destroyed target can be replaced easily, and
base coordinates for detection of movement can be determined
at the next monitoring. The rapid-monitoring system could
be readily implemented for embankment inspection. Economics
are enhanced when several areas can be photographed in a
single photo mission.
Cost is often the most decisive factor in planning the
scope of monitoring programs. Some of the factors
influencing the cost are site location and size, vertical
relief, accessibility, vegetation, weather, and frequency of
monitoring. For the landslide and the coal refuse
embankments studied, costs of monitoring by the rapid-
monitoring system were estimated to be 15 to 40 percent of
the costs of monitoring by conventional ground survey
methods.
Besides being a useful inspection tool for coal refuse
embankments, the rapid-monitoring system can also be used
for a wide variety of other types of mining problems such as
the stability of mine spoil piles, open pit mines, and
|
Colorado School of Mines
|
ER-4609 271
tailings dams. The system's accuracy is such that major
earth dams and the stability of natural and manmade slopes
can also be monitored to detect and record movement.
9.4 Pressure Cells (after Soderberg and Bush, 1977 )
Pressure cells should be placed on the top of the
decant lines or other culverts that will be under the
embankment to measure the pressure of the soil and water
against the plane surface. The reinforced-concrete conduit
should be designed for the combined weight of material at
maximum height, but the pressure cells placed at strategic
spots along the line could verify the actual pressures.
This information can be valuable in designing future
conduits.
9.5 Records (after Soderberg and Bush, 1977 )
All instruments installed in an embankment should be
read and recorded on a regular basis, which at startup
should be quite frequently. Later, as patterns are
established, the time between readings can be altered to
suit the situation. These records are very important and
should be reviewed periodically to see if additional records
are needed and to note the trends. The piezometric level
measurements are probably the most important and have a
direct tie-in with the seepage or flow from the drains.
Precipitation measurements should be made, and positions of
the spigots or cyclones that are operating should be noted.
Often the length of time a group of spigots can be left in
operation depends on the rise of the piezometric level in
|
Colorado School of Mines
|
ER-4609 273
CHAPTER 10
MAINTENANCE AND INSPECTION
In chapter ten, sections 10.1 and 10.2 were extracted
from the Bureau of Mines information circular 8755, Design
Guide for Metal and Nonmetal Tailings Disposal, written by
Roy L. Soderburg and Richard A Busch. Sections 10.3 through
10.3.6 were compiled from the document Tentative Design
Guide for Mine Waste Embankments in Canada, D.F. Coates
being the project officer. Section 10.3.7 was compiled from
Jan Eurenius’ article Long Term Studies and Design of
Tailings Dams. Section 10.4 through 10.4.2 were extracted
from an article titled Reclamation of Mineral Milling
Wastes, written by K.C. Dean and R. Havens. Section 10.4.3
was compiled from an article titled Comparative Costs and
Methods for Stabilization of Tailings, written by Karl C.
Dean and Richard Havens.
10.1 Overview (after Soderberg and Bush, 1977)
The active life of mine tailings embankments may be
from a few years to as much as 100 years ; during this time
many changes can take place that affect the stability of the
embankment. This type of construction is radically
different from a water-type dam where the construction is
done in a relatively short time under close quality-control
of material and methods.
The physical properties of the tailings used in pond
construction may change over the years for many reasons.
|
Colorado School of Mines
|
ER-4609 274
Such changes can alter the stability of an embankment,
resulting in variations in the factor of safety. One of the
most common changes is the increase in tonnage to the mill
without a compensating change in tailings area; this will
mean an increase in the annual rise of the dam, reducing the
factor of safety. A change in grind with an increase in the
minus 200-mesh material can cause a higher phreatic line, a
possible decrease in the efficiency of the drains, and
increased seepage through the starter dam. Any one or
combination of these changes can mean a decrease in the
factor of safety of the embankment. It is therefore
important that a continuous program of inspection and
maintenance of the embankment be started at the beginning
and maintained throughout the life and even after the
abandonment of the embankment. The records of the
instrumentation as described previously are one of the most
important aids in determining the safety of the tailings
dam, and in a high dam are an absolute necessity. High
embankments should be thoroughly inspected by a competent
engineer at least twice a year during the active life of the
pond. A review of the records of the instrumentation in the
embankment should be included in this inspection.
Daily inspection should be made of the spigots or
cyclones, the decant lines, and position of the water pool
in relation to the decant or the tailings area boundary.
The drain lines should be checked for quantity of water and
sediment.
|
Colorado School of Mines
|
ER-4609 276
Whether the drains are operating, free of
sediment, and flowing at a regular rate.
Whether there is seepage on the downstream face of
the starter dam indicated by weeds growing along
the face, or worse yet, seepage on the downstream
face of the sand dam above the starter dam.
Whether the decant lines are intact and free of
cracks that could allow sand to pipe into the
lines and cause a total failure. These can be
visually inspected.
If the phreatic surface is as planned; or is there
excess pore water pressure from within the
foundation or perched water tables?
Whether there have been variations in the water
levels or a sudden rise in the water level, the
appearance of any new springs, or new seepage on
the face of the embankments, foundation, or
abutments. Conditions at the seepage exit points,
decant and drain pipe outlets should be reviewed:
Is the water clear or does it contain sediments;
is there sloughing in the area; is water coming
along the outside of these pipes; are there
sinkholes in the beach or slime zone which would
indicate piping?
Whether there has been an increase in embankment
movement as indicated by the surface control
points or slope indicator.
|
Colorado School of Mines
|
ER-4609 277
Any evidence of borrow from the toe or any other
area of the embankments that might affect
stability.
Whether the embankment geometry is no steeper than
planned.
Whether diversion channels and pipes have
withstood spring runoff or storms. Are they
adequate and in good repair?
All of these points must be watched closely to check
stability, but if a tailings dam has been designed for a
total height of 500 feet and all seems to be going well at a
height of 200 feet, a thorough investigation should be made
by drilling holes into the fill material. Undisturbed
samples can be checked for inplace density, screen analysis,
0 angle, and cohesion. With this information and the
phreatic surface and geometry of the embankment, a static
and dynamic stability analysis can be run to get the FS at
that time. From this information the FS can be projected
for an embankment 500 feet high to determine if the slope
can be steeper or must be flatter, or if the construction
must be stopped short of the design height.
10.3 Remedial Measures (after Coates, 1972)
The extent and nature of remedial measures required to
maintain or improve the stability of mine waste embankments
will vary with circumstances. Some developments may require
|
Colorado School of Mines
|
ER-4609 278
extensive work in addition to that anticipated in design.
For others, minor repairs may be adequate. Descriptions of
the principal types of measures effective in improving
stability and in combating erosion follow.
10.3.1 Slope Flattening (after Coates, 1972)
This is illustrated on Figure 81. By removing the
weight of material near the crest of the slope, the driving
force tending to produce a slide is reduced. If the
material removed form the crest is dumped over the toe of
the slope, measures should be taken to ensure that it does
not impede drainage form the embankment. (A drain could be
required if the material has a permeability low in
comparison to that of the material at the base.)
10.3.2 Berms (after Coates, 1972)
This is a special case of slope flattening. It is also
illustrated on Figure 81. Generally, a berm can improve
foundation stability but may not be very effective in
increasing the factor of safety against slope failure unless
it is at least one third to one half the height of the
embankment. Its effectiveness will also depend on the shear
strength in the berm. Compaction may be necessary.
10.3.3 Height reduction (after Coates, 1972)
This could be in the form of excavation to forma berm,
as shown on Figure 82, thus flattening the overall slope.
|
Colorado School of Mines
|
ER-4609 281
10.3.4 Seepage control (after Coates, 1972)
Inverted filters can be used to prevent erosion and
sloughing caused by seepage from embankment foundation
surfaces, as shown on Figure 82. Such filters can prevent
the development of piping. However, if piping has developed
to the point where sink-holes are forming on the upstream
face of the embankment, it may be necessary to blanket this
upstream face with impervious fill or to seal the leaks by
grouting. Where sub-surface erosion is occurring into
culverts or pipes buried under the embankment, these methods
may be the only ones possible.
Relief wells can sometimes be used to lower the water
table under embankment slopes, as shown on figure 82. Such
systems can be expanded as the need arises. They are not
very effective where the holes are located outside of the
embankment fill. Inclined holes can be drilled under the
embankment; however, drilling at angles more than about 30°
off vertical usually involves additional costs. Relief well
systems are effective when installed under the embankment
before the start of waste disposal.
10.3.5 Surface Drainage (after Coates, 1972)
Generally, less erosion of waste pile slopes will occur
when the upper surface of the embankment is graded down
towards the hillside or towards the center, as shown on
Figure 83, and the runoff led away through drainage ditches
or pipes. A drainage system of this type is preferable to
one where drains are located close to the top of the slope,
as seepage from such drains can affect the stability of the
|
Colorado School of Mines
|
ER-4609 283
slope. Erosion of long slopes can be reduced by breaking
the length of the slope by berms and leading the drainage
water to drop pipes. Broken rock, coarse gravel or grass
can also be used to limit slope erosion.
Maintenance of embankment drainage systems shouldbe
directed to preventing unnecessary entry of water into the
embankments. The drainage system should be inspected at
intervals, particularly after heavy rain, with a view to
keeping the system free of obstruction. Arrangements should
be made for any work required to be carried out such as:
clearance of vegetation, sediments and refuse from
trash screens and drainage ditches,
cleaning pipe drains to clearsediments or salt
deposits; cleaning of silt traps,
attention to the outlets of any drainage zones, or
other seepage outlets.
repairs as required,
the diversion of any flow or accumulation of water
into the permanent drainage system,
attention to filters.
Tailings settlement ponds may require adjustment of
inlet and decant arrangements, the clearance of any
blockages, and rectification of any undercutting of the
embankment slopes brought about by wave action.
For completed tailings embankments it may be necessary
|
Colorado School of Mines
|
ER-4609 284
to maintain the decant or overflow systems to prevent the
accumulation of rainwater or to control surface run-off; the
clearance of silt from these systems is likely to be
necessary from time to time. Breaching a tailings
embankment is not generally an acceptable means of
preventing an accumulation of water; this method should only
be used if erosion caused by the outflow will not endanger
the embankment, block the drainage system or cause nuisance
in other ways.
10.3.6 Toe Embankments (after Coates, 1972)
Where waste piles are located on relatively steep
hillsides they may start to move downhill, particularly if
the materials are fine and become saturated, or they are
affected by weathering. Such movement can sometimes be
stopped by constructing an embankment of compacted material
at the toe of the pile.
10.3.7 Surface Erosion (after Eurenius, 1990)
The surface of the dam must be protected against water
and wind erosion. The major factors affecting water erosion
are amount of rainfall, catchment area, runoff, steepness
and length of slopes. The stability of natural soils
against water erosion is predicted by studying the influence
of these factors upon formations similar to a dam
construction. Such studies indicate that gentle slopes are
essential for the long-term stability. Flattening of the
slopes is considered being effective in preventing surface
erosion. For high dams (higher than 15 m) the slopes are
|
Colorado School of Mines
|
ER-4609 285
suggested to have an inclination flatter than 2:1 and 3:1
(horizontal to vertical). When non-toxic tailings are used
as damfill, the surface is mostly stabilized with an
appropriate vegetation. If the tailings contain toxic
material, the slopes are covered by suitable sealing
material.
10.3.8 Dust Control
Dust control has become a subject of importance in the
regulatory climate faced today. Just a few years ago,
measures taken against water erosion would generally serve
the purpose of controlling wind erosion. This is no longer
the case.
Active tailing ponds must be stabilized to reduce dust
and wind erosion. Sucessful techniques are simple, and
relitively easy to implement.
Tailings areas on the dam face and slope can be
stabilized by annual treatment with a water soluble organic
polymer. These areas see little traffic. Polymers that
have been proven successful include citrus based oils and
water soluble organic resin. Typically the Polymer is
applied mixed with water and sprayed from a water truck.
The polymer or resin bonds the fine grained particles
together and coats the surface of the tailing with a glaze.
This effectively stops dusting and wind erosion.
Active areas such as roads and pipe corridors can be
stabilized by similar methods. Polymers and resins can be
applied with success, but vehicle traffic will crush the
glazed surface making repeated applications necessary. Milo
hay and straw has been used to good effect by crimping the
straw into the active roads, and then applying polymers and
|
Colorado School of Mines
|
ER-4609 286
resins to the surface. Water trucks are then used to hit
areas that start dusting when the wind picks up.
10.4 Stabilization Procedures (after Dean, 1972)
The principal methods for stabilization of milling
wastes include :
Physical— the covering of the tailings with soil
or other restraining materials.
Chemical— the use of a material to interact with
fine-sized minerals to form a crust.
Vegetative— the growth of plants in the tailings.
The vegetative procedure is preferred in that esthetics
of the area are improved while obtaining stabilization.
Also, if a mineralized waste is to be conserved for possible
later retreatment or if the area is to be used later for
residential construction or recreational purposes, it is
beneficial to stabilize the area with vegetation.
Vegetation does not hinder retreatment procedures as much as
covering the tailings with other foreign materials.
Methods have been developed and applied in many areas
of the country using physical, chemical, vegetative, and
combined procedures. Several milling companies, either
independently or in cooperation with the Bureau of Mines,
have applied various stabilization techniques to differing
types of wastes and environments. The principal aim of
these companies has been to achieve effective low-cost
|
Colorado School of Mines
|
ER-4609 287
stabilization requiring minimal maintenance.
A summary of various procedures tested is given in the
following subsections on physical, chemical, and vegetative
stabilization.
10.4.1 Physical Stabilization (after Dean, 1972)
Many materials have been tried for physical
stabilization of fine tailings to prevent air pollution.
Other than water for sprinkling, perhaps the most used
material is rock and soil obtained from areas adjacent to
the wastes to be covered. The use of soil often has a dual
advantage in that effective cover is obtained and a habitat
is provided for local vegetation to encroach.
Crushed or granulated smelter slag has been used by
many companies to stabilize a variety of fine wastes,
notably inactive tailings ponds. On active tailings ponds,
however, the slag-covered portions are subject to burial
from shifting sands. Slag has the drawback, unlike soils or
country rock, of not providing a favorable habitat for
vegetation. Furthermore, suitable slag, like soil and rock,
must be locally available.
Other physical methods of stabilization include (1) the
use of bark covering and (2) the harrowing of straw into the
top few inches of tailings.
10.4.2 Chemical Stabilization (after Dean, 1972)
Chemical stabilization involves applying a material
which reacts with mineral wastes to form an air and water-
resistant crust or layer which will effectively stop dusts
|
Colorado School of Mines
|
ER-4609 288
from blowing and inhibit water erosion. Chemicals have the
drawback of not being as permanent a stabilizing means as
soil covering or vegetation. They, however, can be used on
sites unsuited to the growth of vegetation because of harsh
climatic conditions or the presence of vegetable poisons in
the tailings or in areas that lack access to a soil-covering
material. Chemical stabilization is also applicable for
erosion control on active tailings ponds. Chemicals can be
effectively used on portions of these ponds to restrict air
pollution while other portions continue to be active.
10.4.3 Vegetative Stabilization (after Dean, 1972)
The successful initiation and perpetuation of
vegetation on fine wastes involves ameliorating a number of
adverse factors. Mill wastes usually (1) are deficient in
plant nutrients, (2) contain excessive salts and heavy metal
phytotoxicants, (3) consist of unconsolidated sands that,
when wind-blown, destroy young plants by sandblasting and/or
burial, and (4) lack normal microbial populations. Other
less easily defined problems also complicate vegetative
procedures. The sloping sides of waste piles receive
greatly varying amounts of solar radiation depending on
direction of exposure. Studies have indicated that,
contrary to popular belief, photosynthesis of plants is not
continuous while the sun is shining; under high-temperature
conditions, photosynthesis may almost stop. Furthermore,
most accumulations of mill tailings are light in color and
may reflect excessive radiation to plant surfaces, thus
intensifying physiological stress. For these reasons,
vegetation that may be effective on northern and eastern
|
Colorado School of Mines
|
ER-4609 291
Ingredients : Lessons learned from experience and
the consideration of other factors will permit the
selection of plants of specific qualities and
proven usefulness. They will be compatible with
the environment at large since there appears to be
very few that will not grow on a prepared tailings
surface. New species will be introduced where it
is justified and profitable.
Economics : Study of the physical, economic, and
social framework will help determine the economic
aspects of the reclamation scheme. The best use
will be made of the tailings with regard to the
creation and maintaining of secondary industries
where applicable.
11.2 From Waste to Soil (after Leroy, 1972)
In nature, deleterious elements and substances as occur
from the presence of a base metal deposit, for example, are
not necessarily damaging to the, vegetation because their
release into the environment generally takes place at a
naturally controlled rate.
Conversely, the establishment of a vegetation cover on
tailings will result in the deleterious substances, if any,
being in fact sequestered or neutralized, precipitated by
the developing humus layer and the various natural colloids
present. Hence, the vegetation cover becomes a protective,
neutralizing, pollution-controlling natural device of
unsurpassed effectiveness. A fully vegetated acre will
transpire from 5,000 to 10,000 gallons of water daily, and
|
Colorado School of Mines
|
ER-4609 292
eliminate all previous erosion brought about with every
precipitation.
An area of tailings, unprotected and exposed to summer
heat, will quickly lose all available moisture in its
topmost layer and will readily reach temperatures of 100-120
F. Even a moderate wind, blowing unhindered, will cause
sand storm conditions; and at over 15 mph a sand blasting
effect will quickly take its toll of any freshly established
seedlings if there is an exposed partially reclaimed area.
Tailing waste will remain sterile until properly
conditioned for plant growth. But since uncontaminated,
clean and salt-free sand is the ideal medium for plant
growth, conditions very near to optimum will be achieved
provided one has :
moisture, of which there is generally no lack in
tailings;
fertilizing chemical elements, of which 16 are
necessary for plant growth, from N,P,K, down to
the lesser and the micro-nutrients;
an adequate bacterial population to promote
germination and continued growth.
Once the specific needs of the particular waste area to
be reclaimed are determined, it will be a mere matter of
weeks, with proper conditioning before vegetation blankets
the prepared area.
The basic combination is one of fast growing grasses to
act as nurse crop, and slower, nitrogen-fixing legumes to
provide the long-range permanent and maintenance free
|
Colorado School of Mines
|
ER-4609 293
vegetation cover.
A detail of considerable importance, is the homogeneity
of the tailings throughout the mass. It means that it can
be shaped to any desired form very easily, and the plants
established in it will be able to develop a full, coherent
root system, a factor of prime importance in the growth of
healthy plants. Therefore, this material/waste, can be made
more versatile than the common soil, particularly in poorer
podzolic soils with their thin fertile horizon.
11.3 Tailings and Slime Ponds (after BLM, 1992)
Tailings and slime ponds consist of impounded mill
wastes. Slime ponds are tailings ponds with high
percentages of silts and clays, which cause very slow
sediment drying conditions. Slime ponds are commonly
associated with phosphate and bauxite processing, and
reclamation is complicated by the slow dewatering.
Tailings impoundments are typically placed behind dams.
Dams and the impounded wastes may require sealing on a case-
by-case basis to avoid seepage below the dam or
contamination of the groundwater. This measure only may be
done before emplacement of the wastes. Long-term stability
of the structure must be assured in order to guarantee
ultimate reclamation success.
11.3.1 Tailings Characterization (after BLM, 1992)
The nature of the tailings to be impounded should be
determined as early as possible during the development of
' any plan. Tailings exhibiting phytotoxic or other
|
Colorado School of Mines
|
ER-4609 294
undesirable physical or chemical properties will require a
more complex reclamation plan. Analysis should include a
thorough review of groundwater flow patterns in the area and
a discussion of potential groundwater impacts. An
impermeable liner or clay layer may be required to avoid
contamination of groundwater. Where tailings include
cyanide, final reclamation may include either extensive
groundwater monitoring or pumpback wells and water treatment
facilities to assure (ensure) groundwater quality is
protected. The presence of cyanide in the tailings will not
normally complicate reclamation of the surface.
11.3.2 Dewatering (after BLM, 1992)
The first phase of actual reclamation will normally be
the dewatering or drying of the impoundment so that
equipment can gain access to the surface. This can range
from simply letting the tailing material dry naturally to
more complicated methods of trenching to allow water to
escape from the tailings. This phase of reclamation is
often complicated by surface crusting of the tailings. This
phase of reclamation can take up to several years.
Reclamation of slimes will typically require some form
of trenching using either balloon-tired vehicles or cable
trenching tools. Slimes reclamation can be greatly
accelerated by creating surface drainage for initial
stabilization using peripheral and feeder trenches. Feeder
trenches which drain into the peripheral trench are
typically 251 to 40 ' apart, and up to 2 ' deep. Once the
surface of the tailings has dried, heavier equipment can be
used. Farm equipment can usually operate when the solid
|
Colorado School of Mines
|
ER-4609 295
content exceeds 60% in the top 6 feet.
Revegetation of the tailings after trenching can
accelerate the drying process through evapo-transpiration.
Dust abatement may be required at this stage in order to
avoid airborne particulates which may constitute a
substantial environmental problem.
11.3.3 Reshaping (after BLM, 1992)
Depending upon the nature of the tailings, it may be
necessary to modify the overall shape of the top of the
impoundment to avoid the concentration of water and to
improve visual quality. This can involve the addition of
material to develop a "crown” on the impoundment and the
construction of artificial drainages.
11.3.4 Surface Treatment (after BLM, 1992)
Depending upon the nature of the tailings materials, it
may be necessary to construct a cover system to isolate the
waste. Where the tailing itself is a suitable growth
medium, or can be amended to provide a suitable growth
medium, this will not be needed. Cover systems typically
include: water exclusion layer, capillary break, and growth
medium. In some cases, it may be impractical to revegetate
the impoundment. Because dry tailings material is highly
susceptible to wind erosion and subsequent dust problems, it
is important to cap the tailings material with coarse
durable rock.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.