ASSESSMENT OF PROCESS OPERATIONS & WASTE CONTROL
HOMESTAKE GOLD MINE
LEAO • DEADWOOD, SOUTH DAKOTA
PREPARED BY
EDMUND STRUZESKI
FOR
D
I V—' ' N ^ L-/ C. I ^« V
-------
Assessment of Process Operations and Waste Control
Homestake Gold Mine
Lead-Deadwood, South Dakota
Prepared by
Edmund Struzeski
for
Division of Field Investigations - Denver Center
Office of Enforcement and Standards Compliance
Water Quality Office
Environmental Protection Agency
March 1971
-------
TABLE OF. CONTENTS
Introduction ;. • a ., . • . :. . . . •
Background ,. • - • . . . •- • a . • . . . - . • ., • • • ... . . • •
Plant Processes . • • • . . . • . • • . . .. . . . • . • -
• . Comminution—Grinding—Concentration • • ,• ..• • • •
Cyanidation of Sands and.Slimes . . • • • ..• . • • •.
Sands Treatment . 4. • • • • • • • .• .• .. , • . •• ..
‘Slimeg Treatment . • . . • • . •• • • • • .• •
Cyanide Circuits ‘. . • • . •. • .• •• .5 • • • •0
Previous Homestakelcyanidization • ,. . . .• .•
Waste Sources • • • . . . • • . . • . • • • •
Sand Plant . • . . . • . . .1 ,.• •
S limes Plant • • • • . . . .• • • •. .. . . . S.
Homestake Mine • . • * . . • •. •
Waste Treatment/Disposal • •. . • . . .,, . . . . . ,. •. . •
Installation of tailings pond(s) • • • • •. • •.• • •
Filtration of spent slimes • • • • • . •. •• .. . -•
Chlorination of mill waste flows • • • . -. . ..
On—line cyanide regeneration . . • . .. .••. . , • •
In—plant housekeeping
Greater recycling of cyanide—laden flows•
back into the process lines . . . • • • • .5 . 5. • • • • •
Creating certain chemical conditions . . .
• within tailings ponds . • • . . . . . • • . I.. •. • •
Separate storage àf liquids decanted from
tailings ponds. . . • • • .. • • -.:. - • • • •
Segregation of strong from weak waste8 • •. •. .. • . •. .• .•
Mixing acid wastes (if any) with alkaline
wastes • • . • . . . . . • . . . . . . . . • •• •. .
Greater’iise:.ofl condenser and cooliiig . . . S
- waters. :. • :,... . •:. • - . • • • • • S. f • • S • S S •
.Citatlons.4.... • . • a ..; • . ... .• .• . . • ... • .. • • 25
1 LLUSTRATIONS
Follows
Fig. page
1. Location Map Deadwood—Lead, South Dakota • • • • • . . • •
tocation of’ Hornestake Ni!ning Co.
Facilities at Deadwoodj—Lead, Southt Dakota ‘ ‘
2
• • I 5
• . a 5
• • • 7
• I S 7
• • 5 9
• • I 11
• . • 13
• S. • • •
• . I • •
• • • S •
S I • •
16
17
18
19
20
20
21
21
22
23
23
24
24
24
24
24
1
S
-------
1
ASSESSMENT OF PROCESS OPERATIONS AND WASTE CONTROL
IIOHESTAKE GOLD MINE
LEAD-DEADWOOD, SOUTH DAKOTA
INTRODUCTION
The following discussion gives a tentative assessment of conditions
believed now existing at the Homestake Mining Company operations at Lead—
Deadwood; changes that have probably taken place over the past few months;
and suggested alternative measures for waste handling/disposal practices.
This assessment is based upon past reports including 1959 river pollution
and industrial plant studies made by the U.S. Public Health Service and
the South Dakota Department of Health; a reconnaissance and in—plant
visit made by Environmental Protection Agency personnel on December 2,
1970; and a literature review of the status of technology in the gold
and silver milling industry.
-------
Cheyenne Arm Oahe Reservoir
S PEAR FISH
0 E A DV! 00 D
LEAD
SlUR C IS
R
RAPID CITY
0 10
I
o HOT SPRIIICS MILES
Figure 1. Location nap Deadwood Lead, South Dakota
-------
DEAD WOO 0
Slime Plant
Sand Plant
0
LEAD ____________________________
MILES
Figure 2. Location of Homestake Filning Co. Facilities at Deadwood - Lead, South Dakota
0”
/
-------
2
BACKGROUND
The ilomestake m±ne, near Lead, South Dakota, has operated continu-
ously since about 1887. The first sand leaching plant was started
:around. 1901 at Leád. One of the oldest cyanide mills in the United
States was located at Deadwoo4, designed by the Gold and Silver Extraction
Company about 1894. The Homestake Mining Company was founded in the
very early 1900’s and is currently the largest gold producer in the
Un .ted States. In 1948, monthly production from the Homestake Mining
Company wasreported inexcess of $7 million in gold bullion. Over the
years to the present time, all the milling and sand—leach operations
came to be located at the town of Lead, whereas the slimes are gravity—
fed toa central plant atDea wood, several miles below Lead. Gold
• ores àontaining.as little as .2 oz. troy/ton can be mined and Homestake
ores contairiaround 0.4 oz./tbn or possibly slightly higher in recent
years. Ore roductión ‘at:the Homestake Mining Company has increased
steadily over the years from 4,800 tons/day in 1949, to around 5,800
tons/day in 1963, and 5,800—6 ,OO0 tons/day at present.
Amalgamation (mercürytreatment) of ôrés has been used in the
Lea4 area since before the turn of, the century. Up to three months
ago, Homeatakehad ernployedamalgamationas a cheap method of gold
recovery preceding cyanidation...’ Figures given for the late 1930’s
and early 1940’s show that 70 to 75 percent of incoming Ilomestake gold
ore was considered freet gold. The Clark—Todd amalgamators removed
60:percentof.’this’coarse:gol ’a,giving 43percent ’ overall gold recovery.
-------
3
Thèàe early figures. show that only 1/8 oz. troy mercury was used per
ton of. ore crushed (about halt of this mercury was consumed). The
Clark—Todd amalgamators, until recently,were installed in the closed
circuit with the rod..and bal1 tnills. In contrast with these figures,
the 1959 study by the South Dikota Department of Realth showed that
75 percent of total gold recoVered was achieved by amalgamation compared
to: 25 percent by cyanidation. Howevet, in 1959, reported use of
mercury was greater than 0.45 oz. mercury/ton ore. These figures
indicate the . iIghly important role of amalgamation in the previous
line. Amalgamation at the Lead plant is reported to
have been completly phased out as of January 1971.
From the July 1959 stream studies conducted by the USPHS and
South Dakota Department ofHeâlth, the data show that total cyanide
concentrations in tTh itáwood Creek ranged from 0.60 — 9.10 rng/l with
an averageof 2.6 mg/i. In this same study, cyanide levels in Belle
Fourche River waters were around 0.50 mg/i. Cyanide levels critical
to fishlife are generally conSjdered to be in the range of. 0.1 0.3
mg/i. We alsO note the USPUS Drinking Water Standards specify a maximum
limit of 0.2 mgflcyanides, -but.they strongly advise that 0.01 mg/I
not be exceeded. For body contact recreation, fisheries and farmstead
supplies, unpublished water quality criteria guidelines have specified
:an upper li mit of .O.b2 mg/i cyanides.
Nearly all gold milling operations currently employ the all—sliming
process in cyánidation of gol4 ores.: Homestake has retained separate
treatment of sand and slimes because of separate recovery and reuse
-------
4
of sands as backfill in the nine; because of amalgamation; and due
to the high metallurgical efficiency achieved by this separation on
an ore that is somewhat different to treat. The relatively high
pyrrhotite (Fe 5 5 5 to Fe 16 S 17 ) and ferrous iron content of this ore
would normally indicate difficulties in direct cyanide extraction of
gold.
During the 1940’s, Homestake was recovering silver in addition
to gold. At that time, gold bars were produced at about 997 fine,
and silver at 980 fine. Current prices for gold and silver are about
$35/troy oz. gold and $1.70/troy oz. silver (1 troy oz. — 1.097 avoir.
oz.).
-------
5
P4ANT PROCESSES.
A. Comminut ion-Grinding-donden tra tion
I. Baseá.uponbackg*oundir formation,theores enter the Lead mill
previously crushed to about one—half inch size and then subjected to
fine grindingià a rod mill—ball mill—Dorr classifier closed circuit.
Water enters the mill with tHe crushed ores, and wet grinding is
employed, no cyanide or barren solutions (cyanide solution from which
gold has been precipitated) et ter the grinding circuits, although
!cansiderab].e water containing some cyanide is returned from the sand
dam site. According to the 1 PA trip report of December 14, 1970,
the South Mill (at Lead) was utilizing four banks of rod and ball mills.
Previous to January 1971, thd Clark—Todd amalgamators were contained
in this circuit at the efflu nt of the ball mills. The plant flow
diagram for 1956 shows three banks of rod and ball mills, with a
fourth bank consisting of a rod mill, a screening bank for separating
coarse material, and Kreb cydlones for separation of fine particle
sizes. Apparently the latteij grinding circuit has been replaced by
a rod mill—ball mill circuit The precise status of the amalgamators
is not known as of February 1971 but presumably they have been cleaned
and reinaià :° . standby.
Since amalgamation has been stopped, the Lead mill has likely
reverted to grinding its ore to finer particle size so as to enable
more intimate contact and greater extraction of gold in the cyanide
leaching circuits. To compensate for the removal of amalgamators
from the circuit, the plant may also be currently substituting some
-------
6
gravity—concentration unit operations. These could take the form of
hydraulic traps, unit cells and hydraulic cones such as the Denver
Sub—A flotation cell, jigs (t sually mechanically—pulsated), and the
possible use of corduroy blankets (alternate to gold tabling) These
:t1 t operatiOnè are rneiitioned as alternates rather than probable
substitutes since they have 1n the past been followed by amalgamation.
Hornestake may now be giving serious consideration to gravity concen—
Lore. likely èhey will attempt.to extract.
the large majority of gold by use of previously existing grinding
circuits and cyanide circuits 1 . If a high level of gold extraction
cannot be achieved by this means, a change to finer grinding (higher
percentage of slimes) and significantly increased cyanide consumption
may be made. Whether with th ese changes the Homestake operations can
maintain a 97 percent gold e traction (as cited in the 1960 reports)
.:iB not known.
Flotation processes have been employed extensively in milling
operations. However, Homesta 1 ke ore is not amenable to this procedure
Secondiry classificatiot of ground ore at Homestake is achieved by
aserieS of.large Dorr bowl classifiers (settlers) which serve to
:separate theoré intosands and slimes.: According to the 1967 size
distribution table provided by }Iomestaké Mining, ground particles
are roughly separated into sands and slimes, somewhere around the
.200—mesh.sizérânge. In. the metallurgical industry, slimes are
generall > considered as comprising those particles finer than 50
microns, i a., passing through 325—mesh screens. Homestake maintains
-------
7
Its cyanide circuit feeds sornewhat high on the sands side. In 1949,
the Homestake operation was separating its ore into 56 percent sands and
.44 percent slimes x amill throughput rate of 4,800 tons per day. In
1967, based upon a milling ra 1 te of 5,195 tons ore/day (note yearly
average rate),.the percentage of.eands was increased to 60.1 percent
and slirnes decreased to 39.9 percent. It:appears the Leadmill has
more flexibility for increased capacity compared to the Deadwood mill.
All slimes are transferred to the Deadwood mill for cyanidation, and
all sands are cyanidized at the East and West plants at the Lead mill.
. an1dationof SandsandSlimes
Sands Treatment . The sands fraction in the form of a sand—
water slurry is diverted to the East Sand Plant and the West Sand Plant
at the Lead mill. The sands are loaded into 35 sand vats (23 in the
Ea Sand Plant and 12 jn. th& West Sand Plant), through which cyanide
solution is allowed to slow1y percolate downward removing the gold
values in the sands. Sand leaching at the Homestake mill is entirely
a batch—type operation requiring in the order of seven to eight days
to complete the sand treatmen cycle. The vats are-filled with
water, charged with sands, an i the tank water content is drained.
This is followed by first aeration, a1kali ie wash and first strong
cyanide leach solution. The vat is again drained of water, aerated
a second time, subjected to a second strong cyanide leach solution,
drained and aerated a third tItus This J.s followed by a weak cyanide
leach; washing and finally sli icing the sands to the sand dam site.
Gold extraction efficiency InJl967 was reported in the order of 84.5
to 87.5 percent.
-------
8
: is”axiomatic that the finer thesand the higher the, gold extrac-
tion. Previous data from Homestake show that extraction varies from
better than 90 percent for t e minus 200 mesh particles, to less than
‘50 ,percent for particles ‘coar ser than 50—mesh. Homestake maintains
around 46 percent of it8 material as somewhat finer than 200—mesh and
‘S4percent’greater’than 200-mesh but nonecoarser than .48—mesh.
Phosphorous levels are rnaintai.ned ‘on ‘the’ high alkaline side by lime
additions and’ sodium cyanide olution strengths are reported as 0.02
to 0.06 percent. 1-Iowever, repeated mention of Aero—cyanidè’is made
which implies that calcium cyanide is apparently used rather than
sodium cyanide. Aero—cyanide 1 (having 50 percent NaCN equivalent) had
a reported use of 0.57 lbft’onj sand treated in 1967 together with 1.7
lb lime/ton sand. Various cy nide waste flows result from draining
alkaline wash, leach discharge and from sluicing the sands to the
8and dam. Some slime overflo zs from the vats enter the Dorr thickeners,
also located in the sand p1an t. Barren solution is reused back in the
cyanide circuit, In 196,3,’ t was reported that, of the 3,420 tons/day
exhausted sands and 8,945 tonr water discharged from the mill to the
sand dam, approximately 685 tons/day of sand and 6,730 tons/day of
“water t1 (77 percent of’ total), 1 was subsequently discharged from the
sand dam to Gold Run. Cyanides and cyanide complexes are present in
‘all of these,”dis harges. Two other wastesources previously reported,
in the 1959 study, are the ov rf1ow from the Dorr thickener and the
sand—vat filling—water overflow. Excess from the thickeners not reused
in the mill finds its way to pe city sanitary sewers and/or the mill
-------
9
and Gold Run. The vat verf low was also previously released
to the sanitary sewer. Some of the above waste streams may have been
• eliminated.over the past few years.
• 2. Slimes Treatment . Shines after concentration and dewatering
in the Dorr thickeners (at thie Sands Plant) to a relatively high solids
content,’areconveyéd’inslu ry form to the Deadwood plant”and directly
into a series of,rather larg Merrill filter presses. The Merrill
press Is essentially a plate and frame press and is practically
automnatic.infilhing nd.discharge. Each of the presses has 90 frames;
“6 feet x 4 feet x 4 Inches, and the press will hold 26 tons or more of
shines.’ The D dvood plant, in the 1940’s, had 31 Merrill presses.
Whereas other mill8’.may .use Merrill presses for washing, the Homestake
‘: ‘,rnjj.l” is believed to be the ‘or ly plant utilizing direct cyanide treatment
in ‘the Merrill presses..
The complete slime trea ment cycle in the Merrill presses requires
only about 8 hours, it is a batch—type operation, and 2—3 charges
;‘èan beniade in a siügle day.. . Gold extraction efficiency is reported
between 88 and 90.5 percent. I Slime pulp is loaded into the frames
with linie followed by’aëratio n; a strong cyanide solution leach; a
second aeration; asecondstr ong cyanide, solution leach; a weak leach
(using barren. soiution) ashing; and,, finally, sluicing of. the slirnes
from the frames. Leach solution concentrations vary from around .015
to .04’percent (as MaCN) and the useof Aero—cyanide and lime respectively
appears to be in the order of 1 0.62 and 2.0 lb/ton slitne treated.
Pregnant cyanide solution (co itaining gold) is withdrawn from the first
-------
10
;1s01t1t10h1 leach;the second solution leach, and the first half of the
weak leach period. The pregt ant liquors are presumably then deaerated
in separate facilities followed by immediate additionof zinc, and
precipitation of the gold, silver, etc., without exposing the solution
to the atmosphere. The rare etals precipitates are likely removed
via small filter presses and the barren cyanide solutions returned for
reuse in the Merrill presses.
In 1963, it was reported that 2,380 tons/day of slime were entering
the Deadwood mill and this solids (slime) material, after extraction,
together with 12,520 tons/day water and cOnsiderable metals impurities
was entirely discharged into Whitewood Creek.. Approximately three—
quartersof.the. waterneeds.àrerapresented by that required in sluicing.
the slimes to the creek Sluicing water originates as ditch water, part
coming from the sand dam, and part as recycle water from inside the
slimes. plant. In addition to cyanides and cyanide complexes contained
in the interstitial water of he exhausted slimes, all of the above
sources oE:slüice water excép ditch water contain some cyanides.
Under normal operating conditions, the mechanical loss of cyanide
discharged with the tailings (both slimea and sands) is a very important
factor in cyanide loss in the plant and consequently in the amount of
cyanides found in the plant waste effluents. The loss of cyanides
would appear higher.Ln the sllmes.than the sands at Homestake although
the latteri&. also probably.h .gh .. Another source of cyanide loss at
the Deadwood mill is the overflow from the press filling water reclaim
tank. Used compressor water also is discharged from the Deadwood mill.
-------
11
Its cyanide content would seem very, minimal, but depending upon plant
origin may eritfurtherchec . Entrainment of high strength liquids
aiid solids across .cOndenser and boilers may be possible. Some of
• these waste streams may have been eliminated in recent years.
3. Cyanide Circuits . •, (a) Homestake has. previously used relatively
: 10w amounts of cyanide inth4r leaching circu±ts owing inlarge part to
previously—employed amalgamation and a careful chemical balance in the
circuits. .;(b) Cyanide consum tion in direct sands and alimes. leaching
will probably now show significant increase which in turn will, require
Cc) greatly improved waste treatment and abatement measures.
H, Hornestàke in1967 wasusing approximately 0.57 lb. Aero—cyanide/ton
H
sands treated and 0.62 ib.Aero—cyanide/ton slimes treated. Since Aero
brand cyanide contains about 50 percent NaC1 and lime, it has an NaCN
equivalency of only 0.5. NaCN equivalent use in 1967 was, therefore,
about 0.3 lb/ton ore treated which is low in gold milling. In 1959,
plant records show augmented onsumption of 0.90 lbs Aero—cyanide/ton
ore equals :0.45 lb. NaCN equivalent per ton. Newton (1942) mentions
that slimes ’extractionnorinally requires 0.1 to l.0’ lb. of cyanide
per ton of ores. He also stares that sand leaching usually requires
4—6 lb. of cyanide/ton sands. It is noted that cyanidization of sands,
although generally consuming higher amounts of cyanide is considered
an inexpensive process when fine grinding is not necessary for good
extraction and when the chemical compoàltion of the ore permits this
process Homestake not only needs sands for backfill in the mine, but
also, the prior use of atnalgai ation had taken the burden of high extrac-
tion from the cyanide circuit . It is believed that small amounts of
-------
12
mercury in the cyanide circuits have served as a catalyst or active
agent toward increasing subsequent cyanide extraction efficiency. Dorr
and’3osqu (1950) obsarved.that mechanical loss of cyanide will depend
upon the type of. treatment used, but for gold ores, the total cyanide
consumption will average about 1.5 lb. MaCN per ton of ore. Although
this cyanide use would•appear high, it does illustrate the very large
amáunts of :.cyartide that may b necessary in straight alkaline cyanide
leaching.. From past information, Homestake ores are moderately difficult
to treat by straight cyanidization.
Understanding the cyanide circuit is in large part understanding
the impurities and cyanicides (CN consumers) present in the cyanide
-solutions.. Impurities are’kept below tolerable levels by “bleedout’,
usually replacing the moisture leaving the plant in the solids residues
by freshwater, sometimes by dumping part of unworkable CN solutions, or
• by regenerating CN sO1utions. Without doubt, the very large quantities
of cyanide—laden waters leaving the Homestake mills with the exhausted
sands and slimes serve as a mc st effective means of bleeding the cyanide
circuits of-impurities.; If these waste streams were-contained in the
plant, cyanide. and alkaline levels would ise substantially, and
additionally, other means of relieving the circuits would need to be
found. Cyanicides-most frequently comprise copper and capper compounds,
ferrous hydroxIde, antimony arBeniC and bismuth materials, pyrrhotite,,
native iron and metallic iron rom grinding, graphite, organic material,
zinc, reduced sulfur, etc. Oxygen and lime content are maintained
at high levels in the cyanide so1utions to minimize reaction with
-------
13
potential cyánicides. ‘Barren solutions will usually contain appreciable
amounts of zinc, copper, thiQcyonates and ferrocyanidee. Many complex
forms of cyanide will exist in the circuit, i e., Na 2 Zn(CN) 4 , Na 2 ZnFe(CN) 6 ,
Na 2 Zn 3 Fe(CN) 12 , KCu(CN) 2 , Na 4 Fe(CN) 6 , to mention only a few The complex
,cyanides,’ once ‘formed, do no appear ‘to harm ,the circuit but neither
do they have any attraction for gold. Their accumulation results ‘in
their eventual discharge inplan’t effluents.
It i’s’ : commonly thought that the complex cyanides have very low
toxicityto’aquatic’fauna compared to free cyanide. Yet’past studies
have shown that 15.5 mg/I sodium ferrocyanides, i.e., Na 4 Fe(Ct’1) 6 yields
3.8 mg/I. cyanide in 30 minüt s. Likewise 2 mg/i potassium ferrocyanide
was found to produce 0.36 — 0.48 mg/i CN which in the particular study
‘cited ‘in the. literature, .was’ suffic1ent to kill all test species of
fish’within.60 —90 , minutes.:
4. . :P i u Hoine’stake Cyanidizàtion. ” ilomestake ores appear to
have various minerals ‘containing iron in the ferrous state such as
pyrrhotite (Fe 5 S 6 to Fe 16 S 17 ), chlorite, cummingtonite (iron—magnesium
amphibole), etc. Arsenopyrit e and pyrite (FeS 2 ) are also constituents.
The suif ides oxidize rather rapidly and steadily through the cyanide
circuits, taking up CN and p oducing thiocyanates. With this material,
past experience shows that c anidization should be started promptly
following grinding, and high 1eve1s of oxygen are necessary to prevent
gold mixtures from settling qut in the circuit. Ores containing
pyrrhotite have always been’ 4ifficult to treat, ‘properly’ by cyanide
owing to their rapid decompo ition in moat atmospheres, particularly
-------
14
in aqueous solutions. Large ‘quantities of ferrous compounds are formed
in solution, compared to nonpyrrhotitic ores, and special care must
be taken in grinding and in preaerating such puips before cyanidization.
After amalgamation at Homestake, the sulfide minerals have been
found to yield their gold easily, provided careful attention is given
•to the chemical and mechanical preparation of the pulp for cyanidization.
For the above reasons,, Homestake crushes its ores in water rather than
recycling barren cyani es”olutions back to the grinding cycle, as is
the practice in many other gold mills. Aeration of the pulp prior to
cyanide leaching is carried out in an alkaline solution, The alkalinity
is kept as low’ as possible since excess alkalinity interferes with gold
dissolution. Homestake has apparently found that direct cyanidization
by grinding in solution is unsatisfactory with poor gold recovery and
high cyanide consumption. However, the addition of lead, compounds can
serve to largely overcome the 1 se unsatisfactory results. These lead
additions may cause a film to be produced on the pyrrhotitic which
acts as an’oxid4tion inhibitor. Past observations, however unverified,
suggest that very 8mal]. amounts of mercury, if present in cyanidization,
are of some benefit.in dissolving the gold into solution.
Counter—Current Decantat on (CCD in’ a series of 3 to 5 thickeners)
has been mentioned as a possibility at Homestake. Besides heavy initial
‘investment,. CCD. appears useful only for all—sliming cyanidation, which
is not., the case at Homestake. Although CCD would probably provide
a higher degree of reuse of cN solution, this generally implies that
partially—déw,atered slimes will still be discharged from’ the last
-------
15
thickener in the series. In essence, this is a form of bleedout from
the circuit. It is, however, recognized that CCD could have inherent
merits. The higher degree of automatic control in CCD would likely
reduce errors of judgment and mishandling associated with batch—type
operations. Although one of the main purposes of CCD is to recover
dissolved values from the finely ground solids without the need for
filtration, a filter can be installed on the effluents from the final
thickener; the filtrates are returned to CCD. It is believed that a
number of gold mills incorporate filters at the tailend of CCD when
the pulp will not settle to at least 50 percent solids, or where a
dewatered filter cake is desirable or necessary for tailings disposal.
-------
16
WASTE SOURCES
The waste effluents from the Lead and Hoinestake gold milling operations
represent very heavy waste loads placed upon Gold Run, Whitewood Creek,
the Belle Fourche, and the Cheyenne River. The mill effluents contain
very substantial and unacceptable amounts of dissolved, suspended and
total àolids together with high levels of various metals,. many of which
are toxic to fish ‘and aquatic life and impair receiving waters f or
practically all reasonable uses except waste disposal. Tailings solids
‘being discharged to the receiving streams are in excess of 3,000 tons/day.
Field surveys during the spring of 1971 by DFI—DC call for heavy metals
analyses on cyanide, mercury, zinc, arsenic and copper. esides
unknowns, additional metals that may merit attention include iron,
magnesium, sulfur (nonmetal), antimony, tin, chromium, cadmium, selenium
and tellurium. An emission spectrophotonieter scan of all possible
‘metals is probably indicated.
In discussing various waste sources in this report, smelting opera-
tions have not been included since Honiestake has not operated a smelter
‘for many years. Residue stock piles from past smelting operations
should possibly be checked. The spread of airborne contamination from
previous smelting operations over the surrounding countryside is not
known. Precipitation and recovery of gold and the fluxing of gold and
silver will not cause problems since these solutions will be recycled
and strict plant control is undoubtedly provided at this point in the
operations to minimize gold loss and theft. However, another check may
• be in order since, plant flow diagrams for these processes are lacking.
-------
17
Waste sources are summa ized as follows:
A) Sand Plant
1. Sand vat filling water overflows (previously routed to Lead
sanitary sewers). Cyanide will originate from the water obtained
from behind the sand dam.
2. Excess overflows from Dorr thickeners not reused in the mill
escape to the mill pond and/or Lead sewers or Gold Run. This
amount of water is considerable and contains some slimes,
cyanide etc. originating from sand vats.
3. First water drain from sand vats: after filling (approximates
characteristics of and vat filling overflows but likely
contains additional sulfates and thiosulfates).
4. Effluent from alkaline wash after first drain.
5. First: strong solution leach drain. Water is displaced from
vat, and saud’ interst ices until cyanide and thiocyanates are
detected. These wastes are high ‘in alkalinity, iron, sulfates,
various minerals, tI&ocyanates, etc. This effluent shut—off
. is’ presumed to be manual, and without close attention, appreci—
able’cyanide could be lost.
6 ’.Seconddrain(presumably thi8 is the same as 5 above).
7. ‘.Sands sluicing to sands dam. This is likely the major source
‘of cyanide, etc. from the ‘sands plant. The very large majority
‘of this water’, is eventually discharged into Gold Run.
8.” Condenser and cooling waters at the Lead Plant.
9. Runoff from residual materials piles at the Lead site.
-------
18
10,. Mlii Pond Overflows’. This waste source is mentioned in the
196.0 report. 1t is’ not identified in the 1963 and 1967 flow
. “di grams and information, poBsibly this source does not exist
today.,
11. Sp 1ls, operator error, accidents, pump failure, etc.
12. Sanitary wastes.
B) Slimes Plant
1..’ Press filling reclaIm tank water overflows. Part of this
water originate from the slime leach circuit containing
cyanide, etc. Thislwaste stream is discharged to Wnitewood
Creek.
2. ‘:Firs .t :dràin or displacement water. This waste shows up in the
1960: State report but not in the 1967 plant information on
the slime plant treatment cycle, Flow diagram is insufficient
and this waste.should bechecked further.
3. .‘.Slimes ’siuicing direct to Whitewood Creek. This Is without
doubt the major source of cyanide in wastes leaving the
Deadwood: plant (likely the major source of. solids, cyanide
andthe various pollutants at Uómestake).
4. Condenser and cooling waters at Deadwood plant. Part of this
Is evidently recycled back into the plant and part wasted.
Origin. of this water should be verified.
5... Runoff from residual materials piles at ’plant site.
6. Spills, operator error, accidents, pump failure, etc.
7. Sanitary sewage.
-------
19
C) Homes take Mine
According to the 1960 State report, some water is returned from the
mine which originates mostly from drilling and sand filling operations.
Its characteristics and point(s) of discharge should be verified.
-------
20
WASTE TREATMENT/DISPOSAL
• Non—treatment alternatives, i.e.., process modifications or additions,
have been previously, mentioned in this report as possible substitutes for
amalgamation, which has now been phased out at liomestake. However,
serious consideration should also be given to waste treatment/disposal
practices and facilities which are indicated as absolutely necessary for
the Lead—Deadwood mills. Some of the possible waste abatement/treatment
methods believed to be tentatively feasible are described below.
‘1.’ Installation of tailings pond(s ) for detention and treatment
of waste flows from. both plants. As a bare minimum, the
.slimes and sands sluice waste streams should be directed into
tailings ponds. Th common tailings pond has been widely
used In the metal mining and milling industry ‘for separating
waterfroin solids; as a primary settler for gangue; as a
storage basin for equalization and chemical precipitation;
as a: surge basin for controlling waste discharge, and as a
water storage area.; Too often, however, these tailings ponds
not been properly designed and engineered to give the
require,d degree of treatment. Additionally, ‘a high level of
safety’, and stability must be built into these ponds. liomestake
is one of the very few metal milling operations today (if not
the only one) that does not employ tailings ponds for waste
‘d sposa1. ‘ Previous data’ from the 1959 USPHS—State field
studies indicated that Whitewood Creek waters when settled
for 3 days experienced a decrease in cyanide content from
-------
21
0.80 to 0.18 mg/i. Although conditions would be vastly
different in a tailings pond, we believe that anywhere from
0—3 days’ detention of plant flows will not be sufficient to
adequately reduce cyanide content. However, regardices of
detention time, the tailings pond(s), once constructed, could
serve to bring all effluents together into a single waste
stream, thereby expediting further treatment of total plant
flows.
2. Filtration of spent slimes . From the literature, it appears
that filtration of slimes is a process that has been incor—
• porated into a numb r of gold milling operations throughout
the world. Filtration of slimes Is indicated as being used
prior to discharge to tailings ponds. Filtration of slimes
before placement into the ponds would save storage space and
provide greatly increased stability of pond embankments. The
first drum filters used for dewatering tailings from counter-
current decantation of slimes were incorporated into the
Hollinger Mill, in Canada, in the 192O s;- Slimes from the
last thickener will generally contain from 50—75 percent
moisture. The cake from a filter, on the other hand, may
run from 15—20 percent moisture and may be as low as 7—8
percent. Filtrates 1 are recycled back to the plant. Although
filters are relatively slow and expensive compared to thicken-
ing, they are highly desirable for waste handling purposes.
3. Chlorination of mill waste flows at pH 8.5 and above.
Theoretically, it requires 2.73 partsof chlorine/part of
-------
22
cyanide and 3.08 parts of caustic/part cyanide to oxidize
cyanide to the cyan te (CNO) form. To convert the cyanate
tocarbondioxide and: nitrogen will, theoretically require
ánother4.09 parts of chlorine/part cyanide and 3.08 parts
caustic/part cyanide. Total theoretical requirements will
be 6.82 parts of chlorine and 6.16 parts of sodium hydroxide
part cyanate although mill wastes will already contain sub—
• stantial amounts of excess alkalinity. Practical require—
ments for complete oxidation wili.probably be in the order
Of 9 parts. chlorine/part cyanide. The first reaction is
almost instantaneous whereas the second reaction appears in
‘the order of 30—40 minutes. The amount and type of solids
present in the wast streams will undoubtedly affect efficiency
:0f b0n1 ti0 Sludges wIll also be produced, the quantity
of which will vary with the particular alkali used.
• . On—line c de regeneration . Cyanide regeneration has in
the past offered.a practical means of overcoming heavy cyanide
• consumption frequently encountered in treatment of gold and
• silver ores where.c anicides are. present. The comon method
for regeneration óf cyanide in the leach circuits is by
• . acidification of the solutions, although other methods of re—
generation such as he carbon—cyanidization, the bromocyanicide,
the. ammonia—cyanide, and chlorination processes, have been
previously used. •
In the acidification process., i.e., the Mills—Crowe process,
all or part of the yanogen is converted into hydrogen cyanide
-------
23
‘gas, which is in turn fixed by an alkali (generally lime) and
returned to the cyanide leach circuits. A weakor foul cyanide
solution is acidified generally by brining it into contact
with sulfur dioxide. The acidified solution is transferred to
a closed tank in which air and solution are mixed, The air
“leaving the tank charged with hydrogen cyanide Is then passed
• to another tank and mixed into an alkaline solution which
absorbs the hydrogen cyanide leaving clean air available
for’ more pickup of gas. .Large volumes of forced air are
nedessary in this process, and the higher the velocity of
circulated air, the higher the efficiency. On—line cyanide
‘regeneration is believed to be employed in many gold mills
in. Canada, overseas, and possibly In the United States.
5’ In—plant housekeeping . Whereas good housekeeping may afford
• a sizéable reductI n In waste, it probably cannot be substituted
(In any’ large degree) for the significant investment in waste
treatment facilities Indicated as necessary for the Romestake
‘operations. Errant; spills, pump or line-failures, careless—
ness, etc., should,; of course, be minimized to the maximum
I ex ent po.ssibl .
6. Greater recycling of cyanide—laden flows back into the process
lines . :The ‘1960. St te report indicates that, at that time,
only about one—third of total p ant(s) water requirement was
comprised of recycl water.’ Many more opportunities exist
at the Lead and Deadwood plants for greater reuse of water.
-------
24
These steps are probably essential for reducing wastewater
volumes from the plants in order that the most economical
treatment of wastes can be obtained.
7. Creating certain chemical conditions within tailings ponds
conducive to maximum precipitation and possible recovery of
metals from the waste flows.
8. Separate storage of liquids decanted from tailing(s) ponds .
It may be desirable to transfer liquids from tailings piles
to individual storage ponds for separate treatment, and/or
disposal by evaporation or other means.
9. Segregation of strong from weak wastes for separate handling
and/or treatment.
10. Mixing, acid wastes (if any) with alkaline wastes to effect
neutralization and coagulation.
11. Greater use of condenser and cooling waters in plant processes
(if not already practiced).
-------
25
CITATIONS
1. Environmental Protection Agency, Storm and Combined Sewer Pollution
Control Branch, Edison Water Quality Laboratory, “Impact of Highway
Salts on the Environment”, February 1971. Report awaiting publication.
2. Environmental Protection Agency, Personnel report on trip to
Honiestake Gold Mill at Lead—Deadwood, South Dakota, December 2,
1970, Files of Division of Field Investigations, Denver Center,
1970.
3. Dobson, 3. C., “The Treatnient of Cyanide Wastes by Chlorination”,
Sewage Works Journal , pp. 1007—1020, November 1947.
4. Dorr, 3. V. N. and F. L. Bosqul, “Cyanidization and Concentration
of Gold and Silver Ores”, McGraw—Hill Book Co., Inc., New York,
1950.
5. Curnham, F. C.., “Industrial. Waste Water Control”, date unknown.
6. Newton, J., “An Introduction to Metallurgy”, John Wiley and Sons,
inc., New York and London, February 1949.
7. South Dakota State Department of Health, Division of Sanitary
Engineering, “Report on Gold Recovery Wastes, Homestake Mining
Co., Lead, South Dakota, 1960”.
8. South Dakota State Department of Health, Division of Sanitary
Engineering and USPHS, “Report on Water Pollution Studies, Gold
• Run Creek—Whitewood Creek—Belle Fourche River—Cheyenne River,
• 1960”, Cincinnati, Ohio, 1960.
------- |