Part5 240408 032427
Part5 240408 032427
Part5 240408 032427
A. Commodity Summary
Phosphorus is the twelfth most abundant element, almost all of which occurs as salts of phosphoric acid.
Phosphate rock deposits occur as marine phosphorites, apatite-rich igneous rock, and modern and ancient guano.
Apatite minerals comprise the majority of phosphate constituents in phosphate rock. All domestic production is
from marine phosphorites. According to the U.S. Bureau of Mines, nearly 93 percent of the phosphate rock sold or
used by U.S. producers in 1991 was for the manufacture of wet-process phosphoric acid, single superphosphate, and
triple superphosphate; the balance (approximately 7 percent) was used to produce elemental phosphorus.1
Solid elemental phosphorus exists in several allotropic forms -- white, red, and black. The most
commercially important is white (elemental) phosphorus. Red phosphorus is also important commercially. Black
phosphorus has been prepared only in a few laboratories.2
Phosphorus-based materials are used mostly in fertilizers, detergents, foods and beverages, and metal
treatment coating. Elemental phosphorus is used as a process input to produce a wide array of phosphorus
chemicals. Most phosphorus is converted to derivatives, including phosphorus sulfides and halides, phosphorus
pentoxide, and phosphoric acid. Elemental phosphorus is used in the deoxidation and alloying of copper; and
elemental phosphorus is used with ferrophosphorus in ferrous metallurgy. White phosphorus is also used in roach
and rodent poisons, chemical warfare, and other military purposes. Generally, red phosphorus is made from white
phosphorus. Red phosphorus is used for wooden and paper safety matches and in the manufacture of fireworks.3
According to the largest U.S. producer of phosphorus, there are four domestic producers of elemental
phosphorus. FMC operates a facility in Pocatello, ID and Monsanto operates a facility in Soda Springs, ID. The
remaining two facilities are owned and operated by the Rhone Poulenc Basic Chemical Company and the Occidental
Chemical Company.4 These are located Silver Bow, MT and Columbia, TN, respectively.
Phosphate rock is mined using both surface and underground mining techniques. A modern electric furnace
process for the production of phosphorus consists of a sequence of four operations: preparing the furnace burden,
charging and operating the furnace, collecting the liquid products, and collecting the gaseous products.5
1
David Morse, "Phosphate Rock," from Minerals Yearbook Volume 1. Metals and Minerals, U.S. Bureau of Mines,
1992, pp. 977-980.
2
"Phosphorus and the Phosphides," Kirk-Othmer Encyclopedia of Chemical Technology, 3rd ed., Vol. XVII, 1982,
pp. 473-490.
3
Ibid.
4
FMC Corporation. Comment submitted in response to the Supplemental Proposed Rule Applying Phase IV Land
Disposal Restrictions to Newly Identified Mineral Processing Wastes. January 25, 1996.
5
Kirk-Othmer Encyclopedia of Chemical Technology. Op. Cit.
2. Generalized Process Flow Diagram
White Phosphorus
Exhibit 2 presents a process flow diagram for the production of white elemental phosphorus. The furnace
burden must be porous enough to allow gases to escape from the reaction zone near the bottom of the furnace.
Several agglomeration methods must be employed to prepare phosphate rock fines for the electric furnace. The
fines must be sintered and then crushed to size and screened. Another agglomeration method is nodulizing. In this
process, phosphate fines are heated in a rotary kiln to incipient fusion. The tumbling in the kiln causes the material
to cohere and form spheroidal agglomerates. A final method of agglomeration is formation of pellets by tumbling.
The pellets can then be calcined in a rotary kiln.6
The agglomerated phosphate rock is charged to an electric arc furnace with coke as a reductant and silica as
a flux. The reduction generates a calcium silicate slag and ferrophosphorus, which are tapped, and carbon monoxide
offgases that contain volatilized phosphorus. Dusts are removed from the offgases using dry separation techniques
such as electrostatic precipitation, and phosphorus is removed by condensation in the presence of recirculation water
above the melting point of phosphorus. The carbon monoxide gases are subsequently burned, and phosphorus is
decanted from the water and stored for sale. The recirculating water is neutralized, and a purge of mud and soluble
impurities is removed and disposed.7
Red Phosphorus
Although red phosphorus is usually manufactured by a batch process, continuous methods are now being
used. In the batch process, white phosphorous is converted to red phosphorus in a steel or cast-iron vessel. The
liquid phosphorus, which is protected by a layer of water, passes into the vessel, which then is closed. The vessel is
heated gradually so that unconverted liquid phosphorus does not boil violently and erupt. A reflux condenser is used
to retain the phosphorus. The mass remains fluid until almost half of the phosphorus has turned into red phosphorus.
As the process continues, the mass thickens and solidifies. The mass is cooled and red phosphorus is removed. The
material is then wet-ground and boiled with sodium carbonate solution to remove any traces of white phosphorus,
which is flammable in air. The red phosphorus is sieved, washed on a rotary filter, vacuum dried, and stabilized by
one of two methods. In the first method, red phosphorus is suspended in a sodium aluminate solution and then
aerated. In the second method, magnesium oxide is precipitated onto the red phosphorus.8
FMC Corporation is the world's largest producer of elemental phosphorus, producing about 240 million
pounds of elemental phosphorus per year. Under normal operating conditions, this process operates 24 hours per
day, 365 days per year.10 Because of the large quantities of elemental phosphorus produced here and the importance
of this facility in the market, FMC's process is described below and a process flow sheet is presented in Exhibit 3.
This information was provided by FMC. The Agency may not necessarily agree with FMC’s characterization of its
waste streams.
6
Ibid.
7
Ibid.
8
Ibid.
9
The processes discussed in this technical background document pertain to production at the FMC Pocatello, ID
plant, and may not be fully representative of industry practices employed by the other three domestic phosphorous
producers.
10
FMC Corporation. Op. Cit. January 25, 1996.
Storage and Preparation of Raw Materials
Storage and preparation of raw materials consists of stockpiling; screening and crushing; briquetting;
calcining; and proportioning of the shale ore, coke, and silica. Shale ore arrives at the plant by railcar and is
unloaded by a rotary car dumper. The ore is conveyed to the stacker for distribution on one of two storage piles.
Ore is then collected from the piles by a reclaimer which deposits it on a conveyer belt. The conveyor belt carries
the ore to crushing and screening. Ore is first screened to remove oversized material and then crushed to a uniform
size and sent to the briquetting process. Fugitive dust from screening and crushing is collected by three baghouses.
The briquetting process presses the crushed material into briquettes similar to the size and shape of charcoal
briquettes, which are fed to the calcining process. FMC Pocatello’s calciners are sintering operations in which the
briquetted shale is heated to form a coherent mass without melting. The calcining operation is therefore principally
an agglomeration method. The process occurs in grate calciners in which the briquettes are carried on moving grates
or pallets through the calcining zone where hot gases are pulled through the feed bed (the briquettes) and grate.
Because the briquettes ride on the moving pallets they undergo little or no tumbling or other motion during this
process. Water and carbon dioxide are given off during this process, but the temperatures are carefully controlled
below the fusion point of the phosphate shale. Temperatures above the fusion point result in a phenomenon known
as “fusing the bed” in which the briquettes on the pallets fuse into a single, rigid mass that cannot be handled in
FMC’s downstream material handling equipment. FMC calcines the ore for two reasons: (1) to harden the
briquettes, thereby limiting briquette breakage; and (2) to drive off moisture content and carboniferous matter. (The
carbon is driven off as carbon dioxide.) Calcining does not chemically alter the phosphate shale.
It is important to harden the briquettes during calcining to reduce breakage and allow proper flow of
gaseous reaction products in the furnaces. Hardened (calcined) phosphate shale briquettes (nodules) are choke-fed
into the furnace following mechanical conveyance and mixing with coke and silica. Significant breakage can occur
to improperly hardened nodules during drops from one conveyor to the next and as the nodules are forced into the
tops of the furnaces. Phosphate in the shale nodules, which is unchanged chemically during calcination from
phosphate in the as-received mined shale, is reduced in the furnaces with coke to form elemental phosphorus and
carbon monoxide. This occurs near the bottom of the furnaces in the plasma arcs at the tips of the carbon electrodes.
Gas-phase reaction products, primarily elemental phosphorus and carbon monoxide, percolate up from the reaction
zones through gas channels in the incoming bed of phosphate nodules, coke, and silica. Nodule fragments, formed
by breakage from broken or improperly calcined briquettes, can plug these gas channels, interfering with proper,
steady-state furnace operation.
It is also important to drive off water during calcining of the briquettes to prevent volatilization of steam in
the high temperature furnaces and to prevent an unwanted side reaction involving water and elemental carbon. In
the calciners, water content is reduced from about 11 percent in the incoming phosphate shale to less than 1 percent
in the hardened briquettes or nodules. The dried and hardened nodules are conveyed and fed to the phosphorus
furnaces, which operate at very high temperature. At the top, where the nodules, coke, and silica are being choke-
fed, the temperature is about 400oC, while near the bottom, temperatures can reach 1500oC in the plasma-arc zones.
If wet briquettes were allowed to enter the furnaces, steam would be uncontrollably and possibly explosively
volatilized. At these temperatures, and in the reducing conditions found inside the furnaces, water and elemental
carbon can undergo the watergas reaction to form hydrogen and carbon monoxide. This is a very destructive side
reaction that can consume both coke, a reagent necessary for the reduction of phosphorus shale to elemental
phosphorus, and the furnace electrodes and sidewall refractory bricks (both of which are solid carbon). Finally,
water generated in the furnaces can cause increased corrosion in downstream process equipment.
Although less important than water removal, it is also necessary to drive off carboniferous material during
calcination. Volatilization of these materials in the furnaces can create severe furnace pressure excursions. During
calcining of phosphorus shale, quantities of low boiling point metals may be volatilized.11
11
FMC Corporation. Op. Cit. January 25, 1996.
EXHIBIT 2
ELEMENTAL PHOSPHORUS PRODUCTION
(Adapted from: Phosphorus, A Mineral Processing Waste Generation Profile.)
CO to
Combustion
Dry
Offgases Phosphorus
Dust
Separation Condensation
H2O
Phosphate Rock or
Sintered/Agglomerated Fines
Ferrophosphorus to Storage
Shale
Shale Stacker- Screening Crushing Briqueting Calcining
Reclaimer
Calcined Nodule
Phosphate Storage &
Reclaim
Coke
Silica
Secondary Primary Electric
Precipitators
Condensers Condensers Furnace Proportioning
Lined
P4 Storage Ponds Crushing Slag
The burden is then fed to one of four electric arc furnaces through feed chutes located on top of the
furnaces. There are 10 feed chutes per furnace to distribute burden uniformly about the electrodes. Each furnace is
equipped with three electrodes that heat the furnaces to reduce the phosphate to gaseous elemental phosphorus.
Silica is used as a fluxing agent to bind with the calcium present in the phosphate ore and form slag. The coke reacts
with the phosphate ore to form carbon monoxide and ferrophosphorus. The furnace offgas, which contains
elemental phosphorus, carbon monoxide and particulates, passes through an electrostatic precipitator (ESP) for
particulate removal. The gas is then passed through a primary condenser where the phosphorus is cooled by water
sprays and condensed to a liquid. Each furnace is equipped with a precipitator and primary condenser. The majority
of the phosphorus is condensed in the primary condenser. The gas stream leaving the primary condenser is
combined with the gas streams from the other primary condensers at the carbon monoxide header. The combined
gas stream, which is primarily carbon monoxide, flows to a second condenser for additional phosphorus removal.
The carbon monoxide stream from the secondary condenser is used as fuel for the calciners. Excess carbon
monoxide goes to the roof flare and to the flare pit.
Elemental phosphorus is gravity fed from the condenser to sumps in the furnace building. Phosphorus in
the sumps is kept under water to prevent contact with air. Phosphorus is transferred from the sumps to one of seven
storage tanks at the phosphorus loading dock by pumping water into the top of the sumps to displace the phosphorus
and forced it out the bottom of the sump and into the bottom of the storage tanks. The phosphorus displaces water
from the top of the storage tank, and this water flows back to the sumps to form a closed-loop system. The
combined capacity of the seven storage tanks is 3,131,000 pounds of phosphorus. Phosphorus is pumped from the
storage tanks into railcars for off-site shipment. FMC also has 12 underground storage tanks for long term storage of
phosphorus. The phosphorus dock also processes sludge generated at the furnace building sumps, storage tanks, and
in the returning railcars. The sludge is dried in a centrifuge and stored before being pumped back to the furnaces for
phosphorus recovery.12
Air emissions from furnace operations result from normal furnace operation, furnace venting, and slag
tapping. The electrothermal process generates carbon monoxide which is used as fuel at the calciners or vented to
the flare pit. Furnace venting occurs during furnace maintenance/repair or process upsets. During maintenance/
repair, the furnace is vented to the roof flare. Emissions to the flare bypass the carbon monoxide header. These
emissions consist primarily of phosphorus pentoxide. Emissions from process upsets, such as furnace
overpressurization, are vented to a pressure relief valve. These emissions are also primarily phosphorus pentoxide.
Emissions from slag tapping are vented to the atmosphere through a Medusa scrubber followed by an Anderson
scrubber.13
None identified.
4. Beneficiation/Processing Boundaries
EPA established the criteria for determining which wastes arising from the various mineral production
sectors come from mineral processing operations and which are from beneficiation activities in the September 1989
final rule (see 54 Fed. Reg. 36592, 36616 codified at 261.4(b)(7)). In essence, beneficiation operations typically
serve to separate and concentrate the mineral values from waste material, remove impurities, or prepare the ore for
further refinement. Beneficiation activities generally do not change the mineral values themselves other than by
reducing (e.g., crushing or grinding), or enlarging (e.g., pelletizing or briquetting) particle size to facilitate
processing. A chemical change in the mineral value does not typically occur in beneficiation.
12
National Enforcement Investigations Center (NEIC), Multi-Media Compliance Investigation, FMC Corporation
- Phosphorus Chemicals Division, Pocatello, Idaho, January 1994.
13
Ibid.
Mineral processing operations, in contrast, generally follow beneficiation and serve to change the
concentrated mineral value into a more useful chemical form. This is often done by using heat (e.g., smelting) or
chemical reactions (e.g., acid digestion, chlorination) to change the chemical composition of the mineral. In contrast
to beneficiation operations, processing activities often destroy the physical and chemical structure of the incoming
ore or mineral feedstock such that the materials leaving the operation do not closely resemble those that entered the
operation. Typically, beneficiation wastes are earthen in character, whereas mineral processing wastes are derived
from melting or chemical changes.
EPA approached the problem of determining which operations are beneficiation and which (if any) are
processing in a step-wise fashion, beginning with relatively straightforward questions and proceeding into more
detailed examination of unit operations, as necessary. To locate the beneficiation/processing "line" at a given
facility within this mineral commodity sector, EPA reviewed the detailed process flow diagram(s), as well as
information on ore type(s), the functional importance of each step in the production sequence, and waste generation
points and quantities presented above.
EPA determined that for the production of elemental phosphorus, the beneficiation/processing line occurs
between agglomeration and electric arc furnace reduction because the agglomerated phosphate rock undergoes a
significant thermal reaction inside the electric arc furnace to yield gaseous phosphorus. Calcining is recognized as a
beneficiation activity in the production of phosphorous. It is one of the final activities performed prior to the
chemical change of phosphate ore into the elemental phosphoro\us product.14 Because EPA has determined that all
operations following the initial "processing" step (in this case, the introduction of proportioned ore into the electric
furnace) in the production sequence also are considered processing operations, irrespective of whether they involve
only techniques otherwise defined as beneficiation, all solid wastes arising from any such operation(s) after the
initial mineral processing operation are considered mineral processing wastes, rather than beneficiation wastes. EPA
presents below the mineral processing waste streams generated downstream of the beneficiation/processing line,
along with associated information on waste generation rates, characteristics, and management practices for each of
these waste streams.
1. Extraction/Beneficiation Wastes
Fugitive dust is generated from screening and crushing. FMC collects this dust in baghouses.15
Calcining offgas solids. FMC sends air emissions from the calciners to scrubbers for removal of
particulates and radionuclides.16 It should be ruled that if calcining at FMC drivers off more than water and carbon
dioxide, it is not a beneficiation activity (See 40 CFR 261.4(b)(7).
Surface impoundment waste solids are generated at a rate of 373 kg per kkg product.17 Existing data and
engineering judgment indicate that this waste does not exhibit characteristics of a hazardous waste. Therefore, the
Agency did not evaluate this material further. Waste characterization data are presented in Attachment 1.
14
FMC Corporation. Op. Cit. January 25, 1996.
15
NEIC. Op. Cit. 1994.
16
Ibid.
17
U.S. Environmental Protection Agency, Multi-Media Assessment of the Inorganic Chemicals Industry, Volume II,
Chapter 8, 1980.
Slag, a RCRA special waste, contains trace amounts of radioactive materials in a stable, calcium silicate
matrix. For every pound of white phosphorus produced, eight to ten pounds of slag are generated. In 1988, this
waste was generated at a rate of 2.6 million metric tons per year.18 At FMC, slag is tapped into the slag pit which is
located next to the furnace building. The slag is allowed to cool before it is loaded and hauled to the slag storage
piles. Some slag is screened and crushed for use in berm and road maintenance.19
Dust. Phosphatic dusts may contain slightly elevated levels of radioactivity as well as volatilized, reduced
heavy metals. Phosphatic dusts are normally sold for blending with fertilizer materials20 and formerly were
classified as byproducts. Past management practices have included storage in a waste pile and offsite landfill
disposal.21 Dust is generated at a rate of 4,400 metric tons per year (adjusted from a reported value to reflect changes
in the sector). Dusts may exhibit the characteristics of toxicity (for cadmium), ignitability, and reactivity.
Furnace offgas solids. This waste may contain cyanide. The generation rate for furnace offgas solids is
24,000 metric tons per year22 (adjusted from a reported value to reflect recent changes in the sector). Waste
characterization data are presented in Attachment 1. This waste may be recycled and formerly was classified as a
sludge.
Sludge is generated in the sumps and consists of a phosphorus/water emulsion and particulates not removed
by the ESPs. Generally, sludge is distilled in iron retorts of steam heated vessels to recover phosphorus. Recovered
phosphorus is added to the product and phosphorus free residues from the retorts are landfilled. Sludge is generated
at a rate of 25 kg per kkg product.23 At FMC, the sludge is sent to the phosphorus dock where it is processed with
sludge generated from furnace building sumps, storage tanks, and returning railcars. The sludge is dried in a
centrifuge and stored before being pumped back to the furnace for phosphorus recovery.24 Existing data and
engineering judgment suggest that this waste does not exhibit any characteristics of hazardous waste. Therefore, the
Agency did not evaluate it further.
Precipitator slurry scrubber water. FMC treats the scrubber water with lime and discharges it to calciner
ponds for settling. The water is recycled back to the scrubbers from the calciner ponds.25 Existing data and
engineering judgment suggest that this material does not exhibit any characteristics of hazardous waste. Therefore,
the Agency did not evaluate this material further. Waste characterization data are presented in Attachment 1.
18
U.S. Environmental Protection Agency, Report to Congress on Special Wastes from Mineral Processing, Volume
II: Methods and Analysis, Office of Solid Waste, July 1990, p. 7-3.
19
NEIC, 1994, Op. Cit.
20
.S. Environmental Protection Agency. "Phosphate Rock," from 1988 Final Draft Summary Report of Mineral
Industry Processing Waste, 1988, pp. 2-120 - 2-127.
21
U.S. Environmental Protection Agency, Technical Background Document, Development of Cost, Economic, and
Small Business Impacts Arising from the Reinterpretation of the Bevill Exclusion for Mineral Processing Wastes,
August 1989, pp. 3-4--3-6.
22
Ibid.
23
U.S. Environmental Protection Agency, Op. Cit., Volume II, Chapter 8, 1980.
24
NEIC, 1994, Op. Cit.
25
Ibid.
Cooling water is generated from cooling of furnace domes by direct water spray. At FMC, this water is
discharged via a permitted outfall.26 Existing data and engineering judgment suggest that this material does not
exhibit any characteristics of hazardous waste. Therefore, the Agency did not evaluate this material further.
Furnace scrubber blowdown. FMC generates this waste at the rate of 43.4 million gallons per year.27
Management for this waste may include treating in a tank and sending the sludge to disposal impoundments.28 This
waste may exhibit the characteristics of toxicity for cadmium and corrosivity prior to treatment. Waste
characterization data for raw furnace scrubber blowdown are presented in Attachment 1.
Furnace building washdown. This waste is generated from numerous sources in the furnace building. The
term "furnace building washdown" is a generic term used at the FMC facility to designate water collected in the V-
3600 tank from numerous sources throughout the furnace building. The water is eventually discharged into a RCRA
Interim Status MTR pond. Water is used in condensing elemental phosphorus from the furnace off-gas stream,
water seals on pressure relief devices and overfill protection systems, electrical transformer and furnace cooling,
furnace tapping fume scrubber systems, conveying and storage operations to keep phosphorus under water cover,
slurrying and transporting electrostatic precipitator solids, cleaning (rinsing) of process equipment to ensure
maximum performance, and other miscellaneous uses such as pump packings and steam condensate. The V-3600
tank is the back-up water collection point for some of these streams, which do not normally report to this tank. The
discharge from the V-3600 tank collectively has been known as "furnace building washdown" because waters from
numerous sources in the furnace building are the primary components of this stream. This stream is generated
continuously during process operations. Contaminants in this stream originate from the mineral feedstocks
(phosphate shale, silica, and coke) used at Pocatello to produce elemental phosphorus. This stream does not contain
outside contaminants such as solvents or acidic or caustic cleaning agents. The furnace building washdown
occasionally exhibits the Toxicity Characteristic for cadmium (D006). The TSS average is .08 percent, and it is also
considered a wastewater for LDR purposes. Furnace Building Washdown contains elemental phosphorus and
NORM. The FMC facility generates approximately 79 million gallons of this waste a year.29
WWTP sludge/solids. This waste is not expected to be hazardous. Waste characterization data are
presented in Attachment 1.
Surface impoundment waste liquids. This waste is not expected to be hazardous. Waste characterization
data are presented in Attachment 1. This waste is completely recycled at FMC.
Spent furnace brick. Existing data and engineering judgment suggest that this material does not exhibit
any characteristics of hazardous waste. Therefore, the Agency did not evaluate this material further.
Waste ferrophosphorus is tapped from the furnaces. It is tapped into chill molds inside the furnace
building and allowed to cool. After cooling, the ferrophosphorus is crushed and screened before being sold as a raw
material to the steel industry.30 Based on existing data and engineering judgment, this waste is not expected to
exhibit characteristics of a hazardous waste. Therefore, the Agency did not evaluate this material further.
26
Ibid.
27
FMC Corporation. Op. Cit. January 25, 1996.
28
U.S. Environmental Protection Agency, Technical Background Document, Development of Cost, Economic, and
Small Business Impacts Arising from the Reinterpretation of the Bevill Exclusion for Mineral Processing Wastes,
August 1989, pp. 3-4 to 3-6.
29
FMC Corporation. Op. Cit. January 25, 1996.
30
Ibid.
WWTP liquid effluent is not expected to be hazardous. This waste may be discharged under NPDES.
The generation rate for this waste stream is 1,520,000 metric tons per year31 (adjusted from a reported value to
reflect recent changes in the sector). Waste characterization data are presented in Attachment 1.
Andersen Filter Media (AFM). Andersen Filter Media (AFM) is a felt-like material that is used in the
Andersen scrubbers to filter fine particulate. Andersen Cleanable Media High Efficiency Air Filter (CHEAF)
scrubbers are used in series with the Medusa Scrubbers to further clean fumes from furnace tapping and phosphorus
dock loading and operations. Andersen Filter Media is used in these scrubbers to filter fine particulates. The AFM is
generated at a rate of 420 cubic yards per year. The material fails the TCLP for cadmium and sometimes arsenic,
chromium, lead and selenium and is a RCRA hazardous waste.32
AFM rinsate. This waste stream has been eliminated by FMC as a waste reduction process modification.
FMC generates three additional waste streams, as described below, which may or may not be typical of
other phosphorus producers.33
Precipitator slurry. The elemental phosphorus product exits the furnaces as a gas along with the carbon
monoxide produced in the furnace reaction. The furnace off-gas also includes entrained solids and solids that have
volatilized in the furnace and condensed as the off-gas cools. Electrostatic precipitators are used to remove these
furnace off-gas solids prior to the water spray condensers that remove the elemental phosphorus as a liquid. At
FMC, these solids collect in a vessel at the bottom of the precipitator, known as the slurry pot, where water is added
with a mixer to form what is termed precipitator slurry. The slurry pot acts as a gas seal on the precipitators to
prevent in-leakage of air. Some elemental phosphorus condenses in the slurry pot and the solids contain low
volatile metals such as cadmium and zinc in elevated levels. Historically, precipitator slurry has been sent to ponds
where the solids settle out and the water is recycled. FMC produces 43 million gallons of precipitator slurry (may be
combined with NOSAP slurry as discussed below) each year. Although there are elevated levels of metals in the
precipitator slurry, the solids typically do not fail a Toxicity Characteristic Leaching Procedure (TCLP) test, unless
there are extenuating circumstances, in which case it will fail for cadmium (D006).
Based on preliminary data, EPA has indicated that precipitator slurry is ignitable (D001) and reactive
(D003). The slurry also contains NORM and elemental phosphorus and should be managed accordingly. The TSS
in the precipitator slurry typically exceed 1 percent, and the TOC concentration present in the precipitator slurry
does not exceed 1 percent. Therefore, the slurry is considered an LDR non-wastewater.
NOSAP Slurry. In 1994 and 1995 as part of its waste minimization efforts, FMC developed and installed
the NOSAP, which is a modification to the electrostatic precipitator and the slurry pot. Lime slurry is added to the
slurry pot to control the pH of the slurry to a set point of pH 12. The lime reacts with the phosphorus to form
phosphites and phosphine gas, thus reducing the concentration of phosphorus to below 1000 ppm. The lime also
prevents the metals from becoming leachable and ensures the slurry will not fail the TCLP test. The resulting slurry
that has gone through this process is known as NOSAP slurry. Based on preliminary data, EPA has indicated that
NOSAP slurry is reactive (D003). If all precipitator slurry went through the NOSAP process, FMC would produce
only 22 million gallons per year since the NOSAP slurry has a higher solids content. The solids in NOSAP slurry
are the same as precipitator slurry with the exception of the effect of the lime. The NORM content is the same and
there is still some residual phosphorus content. NOSAP slurry that does not meet specifications is a component of
precipitator slurry. The TSS in the NOSAP slurry typically exceed 1 percent, and the TOC in the NOSAP slurry does
not exceed 1 percent. Therefore, the NOSAP slurry is a non-wastewater for LDR purposes.
31
U.S. Environmental Protection Agency, Op. Cit., Vol. II, pp. 14-45 - 14-59.
32
FMC Corporation. Op. Cit. January 25, 1996.
33
FMC Corporation. Comment submitted in response to the Second Supplemental Proposed Rule Applying Phase
IV Land Disposal Restrictions to Newly Identified Mineral Processing Wastes. May 12, 1997.
The maximum volume of precipitator slurry and NOSAP slurry that would be generated in one year would
be 43 million gallons.
Phossy water. This water waste stream is called phossy water by FMC. Phossy water may carry the D003
RCRA waste code for reactivity. It is possible that phossy water could exhibit the TC for cadmium (D006) upon
process upset. As a result of its contact with phosphorus, phossy water contains suspended phosphorus and other
dissolved solids. In addition, the phossy water contains NORM. The majority of phossy water is recovered for
reuse, but excess phossy water is generated from two locations within the facility. Due to the presence of elemental
phosphorus, the solids in phossy water may spontaneously oxidize and ignite if dewatered. The TSS in the phossy
water typically exceed 1 percent, and the TOC in the phossy water does not exceed 1 percent. Therefore, the phossy
water is a non-wastewater for LDR purposes. FMC generates 89 million gallons total of the phossy water per year.
Solids in phossy water settle out, and are considered to be mineral processing wastes, because the 1984 rule states
that a waste of a mineral process waste is a mineral process wastes. Based on available data, solids from phossy
water may be a hazardous waste.
Although other phosphorous manufacturers may generate the following waste streams, FMC does not do so
at its facility: phosphatic dust, condenser water discard, precipitator slurry scrubber water, WWTP liquid effluent,
and WWTP sludge/solids.
Non-contact cooling water is generated by cooling of the ore calciner grates that transport the briquettes.
At FMC, the noncontact cooling water is discharged via a permitted outfall.34 This waste is a non-uniquely
associated waste. Ancillary hazardous wastes may be generated at on-site laboratories, and may include used
chemicals and liquid samples. Other hazardous wastes may include spent solvents (e.g., petroleum naptha), and
acidic tank cleaning wastes. Non-hazardous wastes may include tires from trucks and large machinery, sanitary
sewage, waste oil (which may or may not be hazardous), and other lubricants.
Three commenters provided additional factual information about the elemental phosphorous production
process (COMM 42, 70, 78)). This information, where appropriate, has been included in sector report.
Sector-specific Issues
Two commenters disagreed with the Agency’s assertion that certain wastes generated during elemental
phosphorous production are RCRA ignitable and reactive (COMM 42, COMM 70). One of these commenters has
since agreed to treat certain wastes as ignitable and reactive (COMM 70).
One commenter stated that water recycled from ponds on-site should not be considered a hazardous waste,
because it does not fit any one criteria that would make it a hazardous waste under the RCRA statute definitions
(COMM 70).
34
NEIC, 1994, Op. Cit.
BIBLIOGRAPHY
Bartels, James J., and Theodore M. Gurr. "Phosphate Rock." From Industrial Mineral and Rocks. 6th ed. Society
for Mining, Metallurgy, and Exploration. 1994. pp. 751-763.
Elemental Phosphorus Processing Waste Characterization Report for FMC Corporation, Pocatello, Idaho. 1991.
pp. 1-4.
Morse, David. "Phosphate Rock." From Minerals Yearbook Volume 1. Metals and Minerals. U.S. Bureau of
Mines. 1992. pp. 977-980.
"Phosphorus and the Phosphides." Kirk-Othmer Encyclopedia of Chemical Technology. 3rd ed. Vol. XVII. 1982.
pp.473-490.
U.S. Environmental Protection Agency. Technical Background Document, Development of Cost, Economic, and
Small Business Impacts Arising from the Reinterpretation of the Bevill Exclusion for Mineral Processing
Wastes. August 1989. pp. 3-4--3-6.
U.S. Environmental Protection Agency. "Elemental Phosphorus Production." From Report to Congress on Special
Wastes from Mineral Processing. Vol. II. Office of Solid Waste. July 1990. pp. 7-1 - 7-24.
U.S. Environmental Protection Agency. Newly Identified Mineral Processing Waste Characterization Data Set.
Office of Solid Waste. August 1992. Vol. I. pp. I-2 - I-8.
U.S. Environmental Protection Agency. Newly Identified Mineral Processing Waste Characterization Data Set.
Office of Solid Waste. August 1992. Vol. II. pp. 14-45 - 14-59.
U.S. Environmental Protection Agency. "Phosphate Rock, Phosphoric Acid, and Phosphorus." From 1988 Final
Draft Summary Report of Mineral Industry Processing Waste. 1988. pp. 2-120 - 2-127.
U.S. Environmental Protection Agency. Multi-Media Assessment of the Inorganic Chemicals Industry. Volume II,
Chapter 8. 1980.
ATTACHMENT 1
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - SURFACE IMPOUNDMENT SOLIDS - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - SLAG QUENCHWATER - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - CONDENSER PHOSSY WATER DISCARD - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - FURNACE OFFGAS SOLIDS - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - PRECIPITATOR SLURRY - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - FURNACE SCRUBBER BLOWDOWN - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - WASTEWATER TREATMENT PLANT SLUDGE/SOLIDS - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - SURFACE IMPOUNDMENT LIQUIDS - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - WASTEWATER TREATMENT PLANT LIQUID EFFLUENT - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - AFM RINSATE - ELEMENTAL PHOSPHOROUS
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
FLUORSPAR AND HYDROFLUORIC ACID
A. Commodity Summary
In 1994, approximately 73 percent of the reported fluorspar (CaF2) consumed in the United States was used
in the production of hydrofluoric acid. About 10 percent of the fluorspar was used as a fluxing agent in steelmaking,
and in iron and steel foundries. The remaining 17 percent was consumed in aluminum fluoride manufacture, primary
aluminum production, glass manufacture, enamels, welding-rod coatings, and other miscellaneous end uses or
products.1 Fluorspar is sold in three grades: metallurgical (minimum 85 percent CaF2), ceramic (85-96 percent
CaF2), and acid (minimum 97 percent CaF2).2 While there are seven active domestic fluorspar mines, the majority of
fluorspar used in the U.S. is imported.3
Hydrofluoric acid is an aqueous solution of hydrogen fluoride produced by a reaction of fluorspar and
sulfuric acid. Hydrofluoric acid is the feedstock used to produce almost all of the organic and inorganic fluorine-
bearing chemicals. Hydrofluoric acid also is used in aluminum and uranium processing.4 As of 1989, three facilities
actively produced hydrofluoric acid. Although several other facilities produce hydrofluoric acid as an intermediate
product during the formulation of commercial chemicals or compounds, these facilities are not included as part of
the primary hydrofluoric acid industry. The names and locations of the three hydrofluoric acid production facilities
are shown in Exhibit 1.
EXHIBIT 1
Hydrofluoric acid is produced by reacting acid-grade fluorspar and sulfuric acid in a kiln, and cooling and
purifying the product. This process is described in detail below.
1
M.M. Miller, "Fluorspar," from Mineral Commodity Summaries, U.S. Bureau of Mines, January 1995, p. 58.
2
M.M. Miller, "Fluorspar," from Minerals Yearbook. Volume 1. Metals and Minerals, U.S. Bureau of Mines, 1992,
p. 487.
3
U.S. Bureau of Mines, Randol Mining Directory 1994/1995, Randol International Ltd., Golden, CO, 1994, p. 165.
4
M.M. Miller, 1994, Op. Cit., p. 58.
2. Generalized Flow Diagram
Before fluorspar can be used to make hydrofluoric acid, the raw ore must be physically concentrated and
purified. Ceramic and acid grades of fluorspar are concentrated (not shown) by crushing and grinding fluorspar, and
purified by froth flotation. First the fluorspar is crushed and ground. Then lead and zinc sulfides are preferentially
floated away from the fluorspar. The easily floating fluorspar is removed and sent to the cleaner circuit. The tailings
are discarded and the middling product is reground and passed through a cleaner circuit. The flotation process yields
acid grade concentrate, and sometimes lower grade concentrates, which are sold either as ceramic or metallurgical
grade fluorspar. Metallurgical grade fluorspar is produced by hand sorting, crushing and screening, and gravity
concentration.5
Hydrofluoric acid is produced from acid-grade fluorspar (CaF2) which reacts with sulfuric acid in a heated
retort kiln to produce hydrogen fluoride gas, as shown in Exhibit 2. The acid grade fluorspar typically contains at
least 97 percent calcium fluoride, as well as silica, calcium carbonate, carbon, sulfur, phosphorus pentoxide,
chloride, mixed metal oxides, and a trace amount of arsenic. The sulfuric acid generally is between 93 and 99
percent pure. Both sulfuric acid and oleum (SO3) are commonly used.6 The residue remaining after retorting is
calcium sulfate anhydrite, commonly known as fluorogypsum, which is a RCRA special waste. This solid is slurried
in process water as it exits the kiln and is transported either to the waste management units7 or, at the duPont plant,
to a production operation for further processing for sale as a byproduct.8 The process wastewater, the second RCRA
special waste generated by this sector, is stored/treated in on-site surface impoundments and then either reused in the
process operations or discharged.
The crude product gas is handled differently by the various manufacturers, but cooling and scrubbing are
always involved. Exhibit 2 shows the gas being cooled, purified by scrubbing, and condensed. The crude product
may be diluted and sold as an approximately 70 percent hydrofluoric acid solution, or distilled to remove any
remaining water and impurities, and sold as anhydrous hydrogen fluoride, a colorless fuming liquid. The wastes
from the refrigerated condenser go to an acid scrubber. The sulfuric acid used in this process unit is then sent to the
acid feed, to react with the fresh fluorspar. The waste stream from the acid scrubber is sent to a water scrubber
which generates fluorosilicic acid and gases. The fluorosilicic acid may be recovered or disposed.
No new novel processes have been utilized, however, a possible process under investigation involves
extracting fluorine as fluorosilicic acid from phosphate rock during the production of phosphoric acid. Also under
investigation is the production of calcium fluoride from calcium silicon hexafluoride (CaSiF6) produced by the
reaction of fluorosilicic acid and phosphate rock.9
5
M.M. Miller, 1992, Op. Cit., pp. 488-89.
6
"Fluorspar," from Kirk Othmer Encyclopedia of Chemical Technology, 4th ed., Vol. XI, 1994, p. 364.
7
Allied Signal, Inc., 1989, Public comments from Allied Signal, Inc. addressing the 1989 Proposed Reinterpretation
of the Mining Waste Exclusion (Docket No. MW2P00020); November 8, 1989; pg. 1.
8
At the duPont facility, lime is added when the fluorogypsum is quenched in order to enhance the chemical
characteristics of the material for construction applications.
9
"Fluorspar," 1994, Op. Cit., pp. 367-68.
EXHIBIT 2
HYDROFLUORIC ACID PRODUCTION
(Adapted from: Kirk-Othmer Encyclopedia of Chemical Technology, 1994, p. 367
and Development Document, Section 12, Hydrofluoric Acid Industry.)
H2SO4 Water
CaSO4 to
Fluorspar Disposal or
Furnace Co-product
Water
Water
Condensate to Stoprage
Ejector Dilution Distillation (or Recycled to Kiln)
Packaging Packaging
LEGEND:
Common Practice
To Sales To Sales Intermittent Process
(or process at only
some plants)
4. Beneficiation/Processing Boundaries
EPA established the criteria for determining which wastes arising from the various mineral production
sectors come from mineral processing operations and which are from beneficiation activities in the September 1989
final rule (see 54 Fed. Reg. 36592, 36616 codified at 261.4(b)(7)). In essence, beneficiation operations typically
serve to separate and concentrate the mineral values from waste material, remove impurities, or prepare the ore for
further refinement. Beneficiation activities generally do not change the mineral values themselves other than by
reducing (e.g., crushing or grinding), or enlarging (e.g., pelletizing or briquetting) particle size to facilitate
processing. A chemical change in the mineral value does not typically occur in beneficiation.
Mineral processing operations, in contrast, generally follow beneficiation and serve to change the
concentrated mineral value into a more useful chemical form. This is often done by using heat (e.g., smelting) or
chemical reactions (e.g., acid digestion, chlorination) to change the chemical composition of the mineral. In contrast
to beneficiation operations, processing activities often destroy the physical and chemical structure of the incoming
ore or mineral feedstock such that the materials leaving the operation do not closely resemble those that entered the
operation. Typically, beneficiation wastes are earthen in character, whereas mineral processing wastes are derived
from melting or chemical changes.
EPA approached the problem of determining which operations are beneficiation and which (if any) are
processing in a step-wise fashion, beginning with relatively straightforward questions and proceeding into more
detailed examination of unit operations, as necessary. To locate the beneficiation/processing "line" at a given
facility within this mineral commodity sector, EPA reviewed the detailed process flow diagram(s), as well as
information on ore type(s), the functional importance of each step in the production sequence, and waste generation
points and quantities presented above.
EPA determined that for this specific mineral commodity sector, the beneficiation/processing line occurs
when the beneficiated fluorspar is mixed with concentrated acid in the furnace/kiln where an intense exothermic
chemical reaction occurs and signficantly alters the chemical structure of the fluorspar. Therefore, because EPA has
determined that all operations following the initial "processing" step in the production sequence are also considered
processing operations, irrespective of whether they involve only techniques otherwise defined as beneficiation, all
solid wastes arising from any such operation(s) after the initial mineral processing operation are considered mineral
processing wastes, rather than beneficiation wastes. EPA presents below the mineral processing waste streams
generated after the beneficiation/processing line, along with associated information on waste generation rates,
characteristics, and management practices for each of these waste streams.
Gangue, lead and zinc sulfides, spent flotation reagents, and tailings are likely to be generated by the
beneficiation of fluorspar. The lead and zinc sulfides may be processed further to recover the lead and zinc. No
other information on waste characteristics, waste generation, or waste management was available in the sources
listed in the bibliography.
The hydrofluoric acid production process generates several waste streams. Two of these waste streams,
fluorogypsum and process wastewater, were classified as RCRA special wastes, and were studied in the July 1990
Report to Congress on Special Wastes from Mineral Processing.
Fluorogypsum. This waste is a solid material consisting primarily of fine particles of calcium sulfate,
usually less than 0.02 mm in diameter, that is slurried for transport from the kilns to waste management units. Using
available data on the composition of fluorogypsum, EPA evaluated whether the waste exhibits any of the four
characteristics of hazardous waste: corrosivity, reactivity, ignitability, and extraction procedure (EP) toxicity.
Based on analyses of four samples from two facilities (Geismar and Calvert City) and professional judgment, the
Agency does not believe the fluorogypsum exhibits any of these characteristics. All eight of the inorganic
constituents with EP toxicity regulatory levels were measured in concentrations (using the EP leach test) that were at
least two orders of magnitude below the regulatory levels.10 According to the Newly Identified Mineral Processing
Waste Characterization Data Set, approximately 894,000 metric tons of fluorogypsum are produced annually in the
United States.11 The La Porte, TX plant sells its fluorogypsum for use as a filler for a road base, railroad subbase,
and paving material.12
Process wastewater. This waste is an aqueous liquid, the chemical constituents of which include fluoride,
calcium, and sulfate, with smaller amounts of iron and silicon, as well as many trace metals. Using available data on
the composition of hydrofluoric acid process wastewater, EPA evaluated whether the wastewater exhibits any of the
four characteristics of hazardous waste: corrosivity, reactivity, ignitability, and extraction procedure (EP) toxicity.
Based on available information and professional judgment, the Agency does not believe the wastewater is reactive,
ignitable, or EP toxic. All eight of the inorganic constituents with EP toxicity regulatory levels were measured in
concentrations (using the EP leach test) that were at most 0.6 times the regulatory levels. Some wastewater samples,
however, exhibit the characteristic of corrosivity. Analyses of the pH of hydrofluoric acid process wastewater at the
Geismar and Calvert City facilities indicate that the wastewater is corrosive in all of the nine samples analyzed,
sometimes with pH values as extreme as 1.00 (for comparison, pH levels below 2.0 are operationally defined as
corrosive wastes).13 According to the Newly Identified Mineral Processing Waste Characterization Data Set,
approximately 13.6 million metric tons of process wastewater are produced annually in the United States.14
Sludges. Existing data and engineering judgement suggest that this material does not exhibit any
characteristics of hazardous waste. Therefore, the Agency did not evaluate this material further.
Off-Spec Fluorosilicic acid (H2SiF6). This waste is recovered from the water scrubber, and can be used in
water fluoridation after it is recovered. Although no published information regarding waste generation rate or
characteristics was found, we used the methodology outlined in Appendix A of this report to estimate a low,
medium, and high annual waste generation rate of 0 metric tons/yr, 15,000 metric tons/yr, and 44,000 metric tons/yr,
respectively. We used best engineering judgement to determine that this waste may exhibit the characteristics of
corrosivity. This waste stream is partially recycled and classified as a by-product.
APC Dusts. Existing data and engineering judgement suggest that this material does not exhibit any
characteristics of hazardous waste. Therefore, the Agency did not evaluate this material further.
D. Non-uniquely AssociatedWastes
Non-uniquely associated and ancillary hazardous wastes may be generated at on-site laboratories, and may
include used chemicals and liquid samples. Other hazardous wastes may include spent solvents, and acidic tank
cleaning wastes. Non-hazardous wastes may include tires from trucks and large machinery, sanitary sewage, and
waste oil other lubricants.
10
From the response of Allied Signal, Inc. and Pennwalt Corp. to EPA's "National Survey of Solid Wastes from
Mineral Processing Facilities", conducted in 1989.
11
U.S. Environmental Protection Agency, Newly Identified Mineral Processing Waste Characterization Data Set,
Office of Solid Waste, August 1992, p. I-5.
12
From the response of E.I. duPont to EPA's "National Survey of Solid Wastes from Mineral Processing Facilities,"
conducted in 1989.
13
From the response of Allied Signal, Inc. and Pennwalt Corp. to EPA's "National Survey of Solid Wastes from
Mineral Processing Facilities", conducted in 1989.
14
U.S. Environmental Protection Agency, 1992, Op. Cit., p. I-5.
E. Summary of Comments Received by EPA
"Fluorspar." Kirk Othmer Encyclopedia of Chemical Technology. 4th ed. Vol. XI. 1994. pp. 364-68.
Fulton III, R.B., and G. Montgomery. "Fluorspar." From Industrial Minerals and Rocks. 6th ed. Society for
Mining, Metallurgy, and Exploration. 1994. pp. 509-522.
Miller, M. M.. "Fluorspar." From Mineral Commodity Summaries. U.S. Bureau of Mines. January 1995, pp. 58-
59.
Miller, M.M., "Fluorspar." From Minerals Yearbook. Volume 1. Metals and Minerals. U.S. Bureau of Mines. 1992.
pp. 487-500.
MRI Inc. Draft Site Visit Report on Pennwalt Corporation, Calvert City. July 14, 1989.
U.S. Bureau of Mines. Randol Mining Directory 1994/1995. Randol International Ltd. Golden, CO. 1994.
U.S. Environmental Protection Agency. "Fluorspar." From 1988 Final Draft Summary Report of Mineral Industry
Processing Wastes. Office of Solid Waste. 1988. pp. 2-104 - 2-108.
U.S. Environmental Protection Agency. Newly Identified Mineral Processing Waste Characterization Data Set.
Office of Solid Waste. August 1992.
U.S. Environmental Protection Agency. Report to Congress on Special Wastes from Mineral Processing. Volume
II. Office of Solid Waste. July 1990. pp. 9-1 - 9-29.
Weiss, N.L., Ed. SME Mineral Processing Handbook. Society of Mining Engineers. New York. 1985. pp. 23-1 -
23-9.
GEM STONES
A. Commodity Summary
The gem stone industry in the United States is extremely small and relies on foreign trade to meet most of
its source requirements. The United States has no known large resources of precious gem stones (i.e., diamonds,
rubies, sapphires, and emeralds) and reserves are generally limited to semiprecious stones. Several semiprecious
gem stone deposits are mined in the United States. In 1992, 91% of the total U.S. gem stone production (by value)
was made up by the following states: Arizona, Nevada, Oregon, Maine, California, and Montana. In 1994,
domestic natural gem stone production was approximately $51.6 million.1
Most gem stone mining appears to be done by hobbyists and amateurs in Mitchell County, NC; Judith
Basin County, MT; San Diego County, CA; Oxford County, ME; and Gila County, AZ where gems such as
turquoise, tourmaline, kunzite, emerald, and sapphire are found.
Gem stones are formed in nature in one of three ways: (1) from metamorphic processes, (2) by
precipitating from aqueous solutions, and (3) by crystallizing from magmas. There are three major compositional
groups of gem stones: silicate minerals comprise one-third; alumino-silicates comprise one-fifth; and oxides
comprise one-seventh of gem minerals. The remaining groups are sulfides, phosphates, borosilicates, and
carbonates.
Some semi-precious stones are produced as by-products of other mining operations. For example, beryl,
tourmaline, spodumene, and gem quartz may be coproducts of mica, feldspar, quartz, or other pegmatite minerals.
Diamonds may be recovered from gold dredges, turquoise from copper mines, agate and petrified wood from gravel
pits, and gem garnet from abrasive garnet mines and mills.
Gem stones are used primarily for decoration. There are, however, some industrial applications for gem
stone material. For instance, industrial processes requiring clean homogeneous stones use low-quality diamond.
Tourmaline is used in laboratories to demonstrate the polarization of light, to measure the compressibility of fluids,
and to measure high pressures. Agate is made into mortar and pestle sets, knife edges for balances, textile rollers,
and spatulas. Gem stones are used as jewel bearings in timing devices, gauges, meters, and other applications
requiring precision elements.
Gem stone production includes three steps, (1) mining, (2) processing, and (2) enhancement. These steps
are discussed in greater detail below.
Mining
Gem stone mining operations vary according to size and complexity. Small shallow deposits are generally
mined by a few people with prybars, picks, shovels, and buckets. Drilling, blasting, and timbering may or may not
be employed. Mechanized hauling and hoisting are done only at the largest mines.
1
Gordon Austin, "Gemstones," from Mineral Commodity Summaries, U.S. Bureau of Mines, January 1995, pp. 64-
65.
Processing
In small operations, gem stone ores are broken, crushed, and concentrated by hand picking, washing,
screening, or jigging. In larger operations, mechanized processes are employed. For instance, diamond processing
involves standard gravity methods, grease belts, electrostatic separation, skin-flotation, magnetic separation,
separation by x-ray luminescence, and separation by optical sorting.
Enhancement
Gem materials are cut in four main operations: sawing, grinding, sanding, and polishing. An initial cut is
made with a diamond saw or blade to obtain a slice of desired thickness. Grinding of the stone may be done with
impregnated-diamond, silicon carbide, aluminum oxide wheels, or coated abrasive disks. Multiple grinding steps
ranging from 80 mesh through 600 mesh abrasives are used. Disk and belt sanders use abrasives bonded to cloth or
waterproof reinforced paper. Leather laps or hard felt are used with a polishing agent to obtain the final polish.
Polishing agents such as fine diamond compound, tin oxide, tripoli, chromium oxide, cerium oxide,
alumina, and rouge are typically used. These polished irregular shapes can then be further polished by tumbling
them in a rubber lined drum and using a grinding and polishing medium with or without water.
Finally, many gem stones are further treated to enhance their appearance. Several different chemical and
physical processes may be used, including bleaching, oiling, waxing, staining, dyeing, plastic and color
impregnation by diffusion or dyeing, surface modification with color coating, lasering, glossing, heat treatment to
change color, and irradiation by electromagnetic spectrum and by energetic particles to change color. Interference
filters, foil backings, surface decoration, and inscribing are used to alter the surface of gems. The most common
method of gem enhancement is heat treatment which can change color, structure, and clarity. A newer method of
gem enhancement is diffusion treatment. This involves a chemical heat treatment in a bath of chemicals containing
iron and titanium.2
Synthesis of materials that can replace rare crystalline materials has been encouraged by industry.
Synthetic gem stones may be used in electronics and semiconductors or as frequency controllers, polarizers,
transducers, radiation detectors, infrared optics, bearings, strain gages, amplifiers, lasers, lenses, crucibles, and
more.3
4. Beneficiation/Processing Boundaries
Based on a review of the process, there are no mineral processing operations involved in the production of
gem stones.
Existing data and engineering judgement suggest that the wastes listed below from gem stone production
do not exhibit any characteristics of hazardous waste. Therefore, the Agency did not evaluate these materials
further.
1. Extraction/Beneficiation Wastes
The extraction of gem bearing material in mines creates overburden. However, land disturbance due to
2
Gordon T. Austin, "Gemstones," from Minerals Yearbook Volume 1. Metals and Minerals, U.S. Bureau of Mines,
1992, pp. 501-519.
3
Jean W. Pressler, "Gemstones," from Mineral Facts and Problems, U.S. Bureau of Mines, 1985, pp. 305-315.
gem stone extraction is minimal since the number of underground mines in operation is minimal.4 Additional
miscellaneous wastes include spent chemical agents used to color the gem stones, spent polishing media, and
waste minerals.
D. Non-uniquely AssociatedWastes
Non-uniquely associated and ancillary hazardous wastes may be generated at on-site laboratories, and may
include used chemicals and liquid samples. Other hazardous wastes may include spent solvents, and acidic tank
cleaning wastes. Non-hazardous wastes may include tires from trucks and large machinery, sanitary sewage, and
waste oil other lubricants.
4
Ibid.
BIBLIOGRAPHY
Austin, Gordon. "Gemstones." From Mineral Commodity Summaries. U.S. Bureau of Mines. January 1995. pp.
64-65.
Austin, Gordon T. "Garnet." From Industrial Minerals and Rocks. 6th ed. Society for Mining, Metallurgy, and
Exploration. 1994. pp. 523-528.
Austin, Gordon. "Gemstones." From Minerals Yearbook Volume 1. Metals and Minerals. U.S. Bureau of Mines.
1992. pp. 501-519.
Pressler, Jean. "Gemstones." From Mineral Facts and Problems. U.S. Bureau of Mines. 1985. pp. 305- 315.
GERMANIUM
A. Commodity Summary
Germanium is recovered either as a minor byproduct of the refining of base metal ores, or as a constituent
of coal deposits.1 Germanium has a metallic grayish appearance and is hard and brittle. It is a semiconductor with
electrical properties between those of metal and an insulator.2 The Bureau of Mines estimated domestic
consumption at 25,000 kilograms during 1994.3 The domestic germanium industry is composed of three germanium
refineries, one each in New York, Pennsylvania, and Oklahoma and a mining operation located in Tennessee.4
Additional information is provided on a recovery operation in Utah. Exhibit 1 presents the names and locations of
the facilities involved in the production of germanium.
EXHIBIT 1
Germanium is available commercially as a tetrachloride and a high purity oxide, and is commonly found in
the form of zone-refined ingots, single crystal bars, castings, doped semiconductors, and optical materials.5 Some of
the major end uses for germanium include infrared optics, fiber-optics systems, detectors, and semiconductors.6
Germanium is recovered as a by-product of other metals, mostly copper, zinc, and lead. The process
described in detail below refers to the recovery of germanium from residues at zinc ore processing facilities.
Exhibits 2 through 4 present process flow diagrams for the production of germanium.
1
Thomas O. Llewellyn, "Germanium," from Minerals Yearbook Volume 1. Metals and Minerals, 1992, p. 531.
2
Ibid., p. 531.
3
Thomas O. Llewllyn, "Germanium," from Minerals Commodity Summaries, U.S. Bureau of Mines, 1995, p. 66.
4
Ibid.
5
Thomas O. Llewellyn, 1992, Op. Cit. p. 531.
6
Thomas O. Llewellyn, 1994, Op. Cit., p. 70.
EXHIBIT 2
RECOVERY OF GERMANIUM DURING ZINC ORE PROCESSING
(Adapted from: 1988 Final Draft Summary Report on Mineral Industry Processing Wastes, 1988, pp. 3-93 - 3-99.)
Leaching Leaching
Reagents Leaching
Waste
Precipitation
Reagents Precipitation Wastewater
Ge Scrap,
Concentrates
Water or Caustic
Solution
Either HCl or Cl2
Chlorination
Still Liquor
H 2O Hydrolysis Hydrolysis
Filtrate
GeO2
Purchased
GeO2 H2
Reduction
Ge
Melt and
Cast
Zone
Refining
High Purity
Ge Product
EXHIBIT 4
PROCESS FLOW SHEET FOR GALLIUM AND GERMANIUM PRODUCTION
(Adapted from: 1988 Final Draft Summary Report on M ineral Industry Processing Wastes, 1988, pp. 3-93 - 3-99.)
Ore
Sulfuric Acid
SO 2 Leach Wash Tailings to Pond
Fluorspar
Copper
Iron Cem entation Cement
(Product)
Sulfuric
Salt Acid
98% Pure
H 2S Germ anium
Leach Distillation Germanium Oxide
Precipitation
(Product)
Germanium Cominco Ltd.
Precipitation Still Residue to Pond Proprietary
Technology
Solvent Gallium
Gallium Electrolysis
Extraction Purification Metal
Precipitation
(Product)
Gallium
Hydroxide
As shown in Exhibit 2, germanium-bearing residues from zinc ore processing facilities are a main source
for germanium metal. In general, the ore is roasted and sintered. The sintering fumes, which contain oxidized
germanium are then collected and leached with sulfuric acid, yielding a solution that contains germanium.
Germanium can then be selectively precipitated from the solution through the addition of zinc dust. The remaining
solids can be recycled to maximize the recovery of germanium.7
Most germanium, regardless of the process by which it was recovered from ore, is refined using
chlorination. As shown in Exhibit 3, germanium concentrates are chlorinated with concentrated hydrochloric acid or
chlorine gas to produce germanium tetrachloride (GeCl4) in solution.8 Chlorine is usually added to the primary
distillation or subsequent fractionation, or both, to suppress the volatility of arsenic.9 Solid impurities and still liquor
are separated and discarded as waste or processed further, while the filtrates and other wash water are sent for further
germanium recovery.10
The resultant purified germanium tetrachloride is then hydrolyzed with deionized water to produce a solid
germanium dioxide (GeO2). The germanium dioxide is removed by filtration, dried, and reduced to germanium
metal with hydrogen at approximately 760 oC. The metal can then be melted and cast into first reduction or as-
reduced bars.11
An alternative process, shown in Exhibit 4, is used at the Musto Exploration site in Utah to recover
germanium. Fluorspar and ore are leached with sulfuric acid and sulfur dioxide. The fluorspar forms hydrofluoric
acid which then helps leach both germanium and gallium. Copper can be removed from the leachate by cementation
on iron and can then be sold as a byproduct. Hydrogen sulfide is used to precipitate the germanium. Following this
precipitation, the remaining liquid is sent for further gallium recovery. The recovered germanium is refined via
leaching and distillation. Any still residues are discarded.12
None Identified.
4. Beneficiation/Processing Boundary
Since germanium is recovered as a by-product of other metals, all of the wastes generated during
germanium recovery are mineral processing wastes. For a description of where the beneficiation/processing
boundary occurs for this mineral commodity, please see the reports for copper, zinc, and lead presented elsewhere in
this background document.
7
U.S. Environmental Protection Agency, Development Document for Effluent Limitations Guidelines and Standards
for Nonferrous Metals Manufacturing Point Source Category, Vol. X, Office of Water Regulations Standard, May
1989, p. 5256.
8
Ibid., p. 5256.
9
"Germanium," Kirk-Othmer Encyclopedia of Chemical Technology, 4th ed, Vol XI, 1994, p. 796.
10
U.S. Environmental Protection Agency, 1989, Op. Cit., p. 5256.
11
"Germanium," 1994, Op. Cit. p. 796.
12
U.S. Environmental Protection Agency, "Germanium," from 1988 Final Draft Summary Report on Mineral
Industry Processing Wastes, Office of Solid Waste, 1988, p. 3-96.
C. Process Waste Streams
1. Extraction/Beneficiation Wastes
Not applicable.
Exhibit 3 identifies the following wastes from the recovery of primary and secondary germanium and
gallium recovery. Although no published information regarding waste generation rate or characteristics was found,
we used the methodology outlined in Appendix A of this report to estimate low, medium, and high annual waste
generation rates.
Waste Still Liquor. As shown in Exhibit 3, plants which chlorinate germanium raw materials generate an
acidic still liquor containing arsenic, nickel, zinc and germanium, and suspended solids.13 Low, medium, and high
annual waste generation rates were estimated as 10 metric tons/yr, 210 metric tons/yr, and 400 metric tons/yr. We
used best engineering judgment to determine that this waste may exhibit the characteristics of toxicity (arsenic,
cadmium, chromium, lead, selenium, and silver) and ignitability. Waste characterization sampling data for this
waste stream is included as Attachment 1.
Chlorinator Wet Air Pollution Control Sludge. Plants chlorinating germanium use a wet scrubbing
system to control HCl and H2 fumes. Waste from the scrubbing system contains cadmium, lead, nickel, germanium,
suspended solids, and an alkaline pH.14 Low, medium, and high annual waste generation rates were estimated as 10
metric tons/yr, 210 metric tons/yr, and 400 metric tons/yr. We used best engineering judgment to determine that this
waste stream may be partially recycled and may exhibit the characteristic of toxicity (arsenic, cadmium, chromium,
lead, selenium, and silver). This waste is classified as a sludge. Waste characterization sampling data for this waste
stream is included as Attachment 1.
Waste Acid Wash and Rinse Water. Germanium ingots or bars are washed with a HF-HNO3 mixture
and then rinsed with water to remove residual acid from the bar. The discharged spent acid and rinse water contain
treatable concentrations of lead, germanium, and fluoride.16 Low, medium, and high annual waste generation rates
were estimated as 400 metric tons/yr, 2,200 metric tons/yr, and 4,000 metric tons/yr. We used best engineering
judgment to determine that this waste stream may be partially recycled and may exhibit the characteristics of toxicity
(arsenic, cadmium, chromium, lead, selenium, and silver) and corrosivity. This waste is classified as a spent
material. Waste characterization sampling data for this waste stream is included as Attachment 1.
Exhibits 2 through 4 also identify the following mineral processing wastes associated with the production of
germanium. No information on waste generation or management is available for these wastestreams.
13
U.S. Environmental Protection Agency, 1989, Op. Cit., p. 5273.
14
Ibid., p. 5273.
15
Ibid.
16
Ibid.
Roaster off-gases. Off-gases containing sulfur dioxide are generated during roasting and sent to an acid
plant for treatment.
Leach Residues. Ferrous residues are removed and sent to disposal. Residues containing copper and
cadmium are sent to further treatment and distribution. This waste stream has a reported annual waste generation
rate of 10 metric tons/yr. We used best engineering judgment to determine that this waste may exhibit the
characteristic of toxicity (cadmium and lead).
Germanium-Oxides Fumes. As shown in Exhibit 2, fumes containing germanium oxide are generated
during sintering. From these fumes, scrubbing water or solids as well as air pollution control dusts may also be
generated.
Spent Acid/Leachate. Spent acid is generated by leaching the germanium oxide fumes from the zinc
sintering. The spent acid likely contains arsenic and other heavy metals, including lead and zinc. Low, medium, and
high annual waste generation rates were estimated as 400 metric tons/yr, 2,200 metric tons/yr, and 4,000 metric
tons/yr. We used best engineering judgment to determine that this waste stream may be partially recycled and may
exhibit the characteristics of toxicity (arsenic and lead) and corrosivity. This waste is classified as a spent material.
Wastewater. Some wastewater will result from the precipitation step. Existing data and engineering
judgment suggest that this material does not exhibit any characteristics of hazardous waste. Therefore, the Agency
did not evaluate this material further.
As shown in Exhibit 4, tailings are generated as a result of the initial leaching at the Musto Exploration
process. Still residues are generated as a result of further refining operations at Musto Exploration and sent to ponds
for further treatment.17
D. Non-uniquely AssociatedWastes
Non-uniquely associated and ancillary hazardous wastes may be generated at on-site laboratories, and may
include used chemicals and liquid samples. Other hazardous wastes may include spent solvents, and acidic tank
cleaning wastes. Non-hazardous wastes may include tires from trucks and large machinery, sanitary sewage, and
waste oil other lubricants.
17
U.S. Environmental Protection Agency, 1988, Op. Cit., p. 3-96.
BIBLIOGRAPHY
Llewellyn, Thomas O. "Germanium." From Minerals Yearbook Volume 1. Metals and Minerals. 1992. pp. 531-
534.
U.S. Environmental Protection Agency. "Germanium." From 1988 Final Draft Summary Report on Mineral Industry
Processing Wastes. Office of Solid Waste. 1988. pp. 3-93-3-99.
Germanium Specialist. "Germanium." From Mineral Commodity Summaries. January 1995. pp. 66-67.
U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines and Standards
for Nonferrous Metals Manufacturing Point Source Category. May 1989. Vol. X. Office of Water
Regulations Standards. May 1989. pp. 5231-5352.
"Germanium." Kirk-Othmer Encyclopedia of Chemical Technology. 3rd ed. Vol XI. 1994. p. 796.
ATTACHMENT 1
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - WASTE ACID WASH AND RINSE WATER - GERMANIUM
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - HYDROLYSIS FILTRATE - GERMANIUM
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - WASTE STILL LIQOUR - GERMANIUM
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
SUMMARY OF EPA/ORD, 3007, AND RTI SAMPLING DATA - CHLORINATOR WET APC - GERMANIUM
Non-detects were assumed to be present at 1/2 the detection limit. TCLP data are currently unavailable; therefore, only EP data are presented.
GOLD AND SILVER
A. Commodity Summary
Gold and silver are discussed together in this report since most of the processes used to recover one will
also recover the other. In addition, both metals are often found together in nature. A mine is generally classified as
a gold or silver mine based on which metal recovered yields the greatest economic value to the operator. Exhibit 1
presents the names and locations of known gold and silver smelters and refineries. Exhibit 2 presents the names and
locations of the 25 leading gold-producing mines in the United States.
EXHIBIT 1
TWENTY-FIVE LEADING GOLD-PRODUCING MINES IN THE UNITED STATES (IN ORDER OF OUTPUT)
Silver occurs as native metal, but is usually found in sulfur-bearing minerals. About two-thirds of the world
silver reserves and resources are contained in copper, lead, and zinc deposits. Ores in which silver or gold is the
main component account for the remaining one-third of total world reserves and resources. The chief silver minerals
found in domestic reserves are native silver, argentite, ceragyrite, polybasite, proustite, pyrargyrite, and tetrahedrite.
Other ore minerals of silver are the tellurides, stromeyerite, and pearceite. Gold occurs mainly as native metal,
alloyed with silver and/or other metals, and as tellurides. A naturally occurring alloy of gold and silver is known as
electrum. Other gold minerals are rare. Gold is commonly associated with the sulfides of antimony, arsenic, copper,
iron, and silver.3
Precious metals may be recovered from the ore or from refining processes of base metals such as copper
and lead. Because these are distinct and separate recovery methods, they are discussed separately in this report.
Section 1 describes precious metal recovery from ore, and Section 2 describes precious metal recovery from refinery
slimes. Section 3 is a discussion of precious metal refining operations.
Most domestic gold comes from surface lode mines. Silver is mined using open pit and underground
methods. Several processes may be used to recover gold and silver from their ores. These include gravity
separation, amalgamation, froth flotation, and cyanidation. Several processes may be combined at any given plant.
These processes are discussed in more detail below.
Gravity Separation
Gravity separation relies on density differences to separate desired materials from host rock. Devices used
include gold pans, sluices, shaking tables, and jigs. Gravity separation is used at most placer mines and at some lode
or vein deposits.4
1
John Lucas, "Gold," from Mineral Commodity Summaries, U.S. Bureau of Mines, January 1994, pp. 72-73.
2
Robert Reese, "Silver," from Mineral Commodity Summaries, U.S. Bureau of Mines, January 1995, pp. 154-
155.
3
John M. Lucas, "Gold," from Minerals Yearbook Volume 1 Metals and Minerals, U.S. Bureau of Mines, 1992,
pp. 535-561.
4
U.S. Environmental Protection Agency, "Gold and Silver," from, 1988 Final Draft Summary Report of Mineral
Industry Processing Wastes, Office of Solid Waste, 1988, pp. 3-100- 3-115.
Amalgamation
Fine gold in placer deposits is often not separable from the ore minerals by density alone. The fine
concentrate stream from a gravity separator, called "black sand" because of its color, often contains several dense
minerals as well as fine gold. This fine gold may be recovered by amalgamation which involves the dissolution of
gold or silver in mercury. The resulting alloy, amalgam, is relatively soft and will adhere readily to other pieces of
amalgam or to mercury.5
Historically, amalgamation was widely used in the United States for recovery of gold and silver from their
ores. Although this method is still practiced in other parts of the world, amalgamation most likely occurs
domestically on a very limited scale.
Ore Preparation
Extracted ore must be milled to prepare it for further recovery activities. Uniformly sized particles may be
obtained by crushing, grinding, and wet or dry classification. The degree of milling performed on the ore depends
on the gold concentration of the ore, mineralogy and hardness of the ore, the mill's capacity, and the next planned
step for recovery. Milled ore is pumped to the next operation unit in the form of a slurry. Fugitive dust generated
during crushing and grinding activities is usually collected by air pollution control devices and recirculated into the
beneficiation circuit. Most mills use water sprays to control dust from milling activities.6
After milling, sulfide ores may be subjected to oxidation by chlorination, bio-oxidation, roasting, or
autoclaving. Chlorination is not commonly used to oxidize sulfide ores because of high equipment maintenance
costs caused by the corrosive nature of the oxidizing agent. Bio-oxidation of sulfide ores employs bacteria to
oxidize the sulfur-bearing minerals. Roasting of sulfide ores involves heating the ores in air to convert them to oxide
ores and break up their physical structure, allowing leaching solutions to penetrate and dissolve the gold. Roasting
oxidizes the sulfur in the ore, generating sulfur dioxide that can be captured and converted to sulfuric acid. Roasting
temperatures are dependent on the mineralogy of the ore, but range as high as several hundred degrees Celsius.
Roasting of carbonaceous ores oxidizes the carbon to prevent interference with leaching, which, in time, improves
gold recovery efficiency. Autoclaving (pressure oxidation) is a relatively new technique that operates at lower
temperatures than roasting. Autoclaving uses pressurized steam to start the reaction and oxygen to oxidize sulfur-
bearing minerals. Heat released from the oxidation of sulfur sustains the reaction. The Getchell and Barrick
Goldstrike Mines in Nevada, the McLaughlin Mine in California, and the Barrick Mercur Mine in Utah are currently
using pressure oxidation (autoclave) technology, totally or in part, to beneficiate sulfide or carbonaceous gold ores.7
Agglomeration
Because ores with a high proportion of small particles may retard the percolation of the lixiviate,
agglomeration is used to increase particle size. This operation includes mixing the crushed ore with portland cement
and/or lime, wetting the ore evenly with cyanide solution to start leaching before the heap is built, and mechanically
tumbling the ore mixture so fine particles adhere to larger particles.
Cyanidation - Leaching
Cyanidation leaching is the primary means of recovery of fine gold and silver. In this process, solutions of
sodium or potassium cyanide are brought into contact with an ore which may or may not require extensive
preparation prior to leaching. Gold and silver are dissolved by cyanide in high pH solutions in the presence of
5
Ibid.
6
U.S. Environmental Protection Agency, Technical Resource Document, Extraction and Beneficiation of Ores and
Minerals, Vol. II, July 1994.
7
Ibid.
oxygen. There are three general methods of contacting ores with leach solutions: (1) heap leaching, (2) vat
leaching, and (3) agitation leaching. Cyanidation heap leaching and agitation leaching account for most gold and
silver recovery.8,9 These leaching methods are discussed in detail below.
Heap leaching, shown in Exhibit 3, is the least expensive process and is used most often to treat low value
ores. In 1993, heap leaching accounted for 39 percent of gold production.10 In many cases, heaps are constructed on
lined pads and ore is sent directly from the mine with little or no preparation. However, at about half of the heap
leaching operations, ore is crushed and agglomerated prior to placement on the heap to increase permeability of the
heap and maintain the high pH (optimally 10.5) needed for leaching to occur.
Two types of pads used in gold heap leaching are permanent heap construction on a pad from which the
leached ore is not removed, and on-off pads, which allow the spent ore to be removed from the pad following the
leach cycle and fresh ore to be placed on the pad. Permanent heaps are typically built in lifts. Each lift typically is
composed of a 5- to 30-foot layer of ore, though lifts may be higher at times.11 On-off pads are not commonly used
in the industry.
After the ore is piled on a leaching pad, the leaching solution is applied to the top of the pile by sprinklers.
The solution generally has a concentration of 0.5 to 1 pound of sodium cyanide per ton of solution, though one major
gold producer reports that the leaching solution is generally in the range of 0.25 pounds of sodium cyanide per ton of
solution.12,13 The precious metals are dissolved as the solution trickles through the pile, and the metal bearing
solution is collected on the impervious pad and pumped to the recovery circuit. Following rejuvenation, which
involves removing the metals, the solution is returned for reuse. The leaching process continues until no more
precious metal is extracted. Typical operations will involve leaching for several months on each heap. The process
is relatively inexpensive and can be operated for less than two dollars per ton of ore. However, as much as half of
the gold and silver may not be extracted either because the leach liquor never contacts the precious metal or because
the metal bearing solution is trapped in blind channels. At one facility, at least 60 percent (and often much higher14
percentages) of the gold contained in leach-grade ore is recovered through heap leaching.15 Waste streams from this
process include spent ore and leaching solutions as well as residual leach liquor in the pile.
8
Personal communication between ICF Incorporated and Robert G. Reese, U.S. Bureau of Mines, September 23,
1994.
9
Newmont Gold Company. Comments submitted in response to the Supplemental Proposed Rule Applying
Phase IV Land Disposal Restrictions to Newly Identified Mineral Processing Wastes. January 25, 1996.
10
Personal communication between ICF Incorporated and John M. Lucas, U.S. Bureau of Mines, September 15,
1994.
11
Newmont Gold Company. Op. Cit.
12
U.S. Environmental Protection Agency, Technical Resource Document, Treatment of Cyanide Heap Leaches
and Tailings, Office of Solid Waste Special Waste Branch, 1994, pp. 2-4.
13
Newmont Gold Company. Op. Cit.
14
U.S. Environmental Protection Agency, 1988, Op. Cit., pp. 3-100 - 30-115.
15
Ibid.
E X H IB IT 3
G O L D -S I L V E R L E A C H IN G
(A d a p t e d f r o m : 1 9 8 8 F in a l D r a ft S u m m a r y R e p o r t o f M in e r a l I n d u s t r y P r o c e s s in g W a s te s , 1 9 8 8 , p p . 3 - 1 0 0 - 3 -1 1 5 .)
H eap or V at
O re
H eap or S p en t O re
V at
B a r re n
liq u o r
R e c y c le
C a rb o n
C o lu m n
C a rb o n
S t r i p p in g
G o ld
E le c tr o m in i n g
D o re ’ R e fin in g
(2) Cyanidation - Vat Leaching
Vat leaching, shown in Exhibit 3, is used when greater solution control than that afforded by heap leaching
is necessary. In this system, prepared ore is placed in a vat or tank and flooded with leach liquor. The solution is
continuously cycled through, draining from the bottom of the vat, proceeding to gold recovery, rejuvenation, and
returning to the top of the vat. The process is more expensive than heap leaching because the material must be
removed from the vat at the end of the leaching process. While the primary advantage of vat leaching is better
solution contact, channelization and stagnant pockets of solution still occur (almost as severely as in heap leaching)
when solution is drained from the vat. However, some of the trapped solution is recovered when the solids are
removed from the vat. Wastes from this process include spent ore and leaching solutions.16
Agitation leaching is the most commonly used leaching process in gold beneficiation operations in the
United States.17 High value ores are treated by agitation leaching, shown in Exhibit 4, to maximize the recovery of
metal values. The ore is crushed and ground in water to form a slurry. Cyanide is usually added at the grinding mill
to begin the leaching process, and more cyanide may be added to the leaching tanks. Ores may be leached anywhere
from 24 to 72 or more hours. Silver ores tend to require longer leaching times. The method of recovering the
precious metal from solution determines how the solution is separated from the solids. If the Merrill-Crowe or
carbon-in-column metal recovery process is used, the leach liquor will be washed out of the solids, usually by a
combination of counter-current decantation and filtration washing with water. This produces a concentrated wash
solution and recovers the maximum pregnant liquor from the solids. The resultant slurry will contain very little
cyanide or gold and would not be expected to exhibit any hazardous characteristics. The carbon-in-leach and
carbon-in-pulp beneficiation processes are the most commonly used metal recovery processes used in gold
beneficiation operations.18 If carbon-in-leach or carbon-in-pulp metal recovery is practiced, the slurry may be
discarded without washing. The carbon should remove all of the precious metals, and the solution is recovered from
the tailings treatment and recycled back to the process.19
In leaching operations, after dissolving the metal, the leach solution is separated from the ore, and the gold
and silver are removed from solution in one of two ways: (1) the Merrill-Crowe process, or (2) activated carbon
loading followed by activated carbon stripping.20 The primary difference between recovery methods is whether the
metal is removed by precipitation with zinc or by adsorption on activated carbon. Zinc cyanide is more soluble than
gold or silver cyanide and if pregnant liquor is contacted with metallic zinc the zinc will go into solution and the
gold and silver will precipitate.21 The two different recovery methods are described below.
16
U.S. Environmental Protection Agency, 1988, Op. Cit., 3-100 - 3-115.
17
Newmont Gold Company and National Mining Association. Comments submitted in response to the
Supplemental Proposed Rule Applying Phase IV Land Disposal Restrictions to Newly Identified Mineral Processing
Wastes. January 25, 1996.
18
Ibid.
19
U.S. Environmental Protection Agency, 1988, Op. Cit., 3-100-3-115.
20
Newmont Gold Company and National Mining Association. Op. Cit.
21
U.S. Environmental Protection Agency, 1988, Op. Cit., pp. 3-100 - 3-115.
EXHIBIT 4
AGITATION LEACHING WITH MERRILL-CROWE RECOVERY
(Adapted from: 1988 Final Draft Summary Report of Mineral Industry Processing Wastes, 1988, pp. 3-100 - 3-115.)
Ore
Size
Reduction
Leaching
Washing Tailings
Filtration Solids
Deaeration
Zinc Dust
Precipitation
Solid/Liquid
Zn CN Solution
Separation
Refining
Dore’
Casting
(1) Cyanidation - Metal Recovery - Merrill-Crowe
In the Merrill-Crowe process, the pregnant leaching solution is filtered for clarity, then vacuum deaerated to
remove oxygen and decrease precious metal solubility. The deaerated solution is then mixed with fine zinc powder
to precipitate the precious metals. The solids, including the precious metals, are removed from the solution by
filtration, and the solution is sent back to the leaching circuit. The solids are melted and cast into bars. If silver and
gold are present, the bars are called doré. In most cases, the metal is then sent to an off-site refinery. Most
operations using zinc precipitation in the United States use some variation of the Merrill-Crowe process.22
Precious metal leach solutions can be brought into contact with activated carbon by carbon-in-column,
carbon-in-pulp, and carbon-in-leach processes.
Carbon-in-column systems are used at heap and vat leach operations and in other situations where the
leaching solution is separated from the solids being leached prior to precious metal recovery. The leaching solution
is passed through a series of columns containing beds of activated carbon. The gold and silver are adsorbed as
cyanide complexes on the surface of the carbon. After passing through the columns, the solution is returned to the
leaching circuit. When the carbon in a column is loaded with precious metals, the column is switched to a stripping
circuit.23
In many agitation plants, the gold is recovered from the leached material before the solution is separated
from the solids. In the carbon-in-pulp system, the leached pulp passes from the last stage of the leaching circuit into
another series of agitation tanks. Each tank contains activated carbon granules. The slurry flows from tank to tank
in series while the carbon is retained by screens. When the carbon in the first tank is fully loaded with precious
metals, it is removed and sent to the stripping and reactivation circuit; the carbon in the other tanks is moved ahead
one stage, and new carbon is added to the last stage. The carbon moves counter-current to the leached slurry and the
leached slurry is finally sent to the tailings area for dewatering.24 A process flow diagram of carbon-in-pulp metal
recovery is shown in Exhibit 5.
Carbon-in-leach is similar to carbon-in-pulp except that the carbon is in the leaching tanks instead of in a
separate recovery circuit. One advantage of carbon-in-leach over carbon-in-pulp is that some cyanide is released
when gold adsorbs on carbon, making it available for more leaching. Another advantage is that fewer agitation tanks
are necessary since the separate recovery circuit is eliminated. However, the agitation is more aggressive in the
leach circuit causing more attrition of the carbon than in the carbon-in-pulp. Thus, the finely abraded carbon and its
load of precious metals may be lost, reducing recovery and increasing costs due to increased carbon replacement.25
A process flow diagram of carbon-in-leach metal recovery is presented in Exhibit 5.
Gold stripping from loaded activated carbon is usually done with a hot, concentrated alkaline cyanide
solution, sometimes including alcohol. These conditions favor the desorbtion of the precious metals into the
stripping solution. The solution then goes into an electrowinning cell where the precious metals are plated out,
generally onto a steel wool cathode. The solution is recycled to the stripping stage and the cathode is sent on to
22
Ibid.
23
Ibid.
24
Ibid.
25
Ibid.
EXHIBIT 5
CARBON-IN-PULP AND CARBON-IN-LEACH METAL RECOVERY
(Adapted from: 1988 Final Draft Summary Report of Mineral Industry Processing Wastes, 1988, pp. 3-100 - 3-115.)
CARBON-IN-PULP
Cyanide
Ground
Ore To
Slurry Waste
Loaded Carbon
to Stripping
CARBON-IN-LEACH
Cyanide
Ground
Ore To
Slurry Waste
Carbon
Loaded Carbon
to Stripping
refining. Some operations refine the steel wool on site to make doré while others ship it directly to commercial
refineries. The primary waste from carbon stripping is the spent stripping solution.26
The Anglo-American Research Laboratory (AARL) elution method is an alternative stripping process being
used by at least one facility. In the first step of the AARL method, the loaded carbon is rinsed with a dilute (i.e.,
three percent) hydrochloric acid solution and then flushed with water to render the gold more amenable to separation
from the carbon. The rinse waters resulting from this process are recycled by pipe to the carbon-in-leach circuit to
recover any loaded carbon that is flushed out by the acid wash (loaded carbon typically contains 150 ounces of gold
per ton). Following the acid wash/water rinse stage, desorption occurs. The loaded carbon is soaked in a
concentrated solution composed of six percent sodium hydroxide and three percent sodium cyanide. This soaking
loosens the bond between the carbon and gold. A fresh water rinse then picks up the gold. The resulting pregnant
gold-bearing solution is pumped to the electrowinning circuit.27
Carbon Regeneration
After stripping, the carbon is reactivated on- or off-site and recirculated to the adsorption circuit. Carbon
used in adsorption/desorbtion can be reactivated numerous times. The regeneration technique varies with mining
operations, but generally involves an acid wash before or after extraction of the gold-cyanide complex, followed by
reactivation in a kiln. The activated carbon is washed with dilute acid solution (pH of 1 or 2) to dissolve carbonate
impurities and metal-cyanide complexes that adhere to the carbon along with the gold. This technique may be
employed either immediately before or after the gold-cyanide complex is removed. Acid washing before the gold is
removed enhances gold recovery. The Barrick Mercur Mine in Utah, the Barrick Goldstrike Mine in Nevada, and
the Ridgeway Gold Mine in South Carolina are examples of facilities using acid prewash techniques. The Golden
Sunlight Mine in Montana and the Battle Mountain Mine in Nevada use acid postwash techniques.28
The specific acid used for carbon washing is determined by the types of impurities need to be removed.
Usually, a hydrochloric acid solution is circulated through 3.6 metric tons of carbon for approximately 16 to 20
hours. Nitric acid also is used in these types of operations, but is thought to be less efficient than hydrochloric acid
in removing impurities. The resulting spent acid wash solutions may be neutralized with a high pH tailings slurry,
dilute sodium hydroxide solution, or water rinse. When the spent acid wash solution reaches a stable pH of 10, it is
sent to a tailing impoundment. Metallic elements may also be precipitated with sodium sulfide to remove them from
the carbon.29
The carbon is screened to remove fines and thermally reactivated in a rotary kiln at about 730 /C for 20
minutes. The reactivated carbon is subsequently rescreened and reintroduced into the recovery system. Generally,
less than 10 percent of the carbon is lost during the process because of particle abrasion.30 Recirculating the carbon
material gradually decreases performance in subsequent absorption and reactivation series. Carbon adsorption
efficiency is closely monitored, and fresh carbon is added to maintain efficiency at design levels.31
26
Ibid.
27
Newmont Gold Company. Op. Cit.
28
U.S. Environmental Protection Agency, July 1994, Op. Cit., pp. 1-12.
29
Ibid.
30
Newmont Gold Company. Op. Cit.
31
U.S. Environmental Protection Agency, July 1994, Op. Cit., pp. 1-12.
3. Identification/Discussion of Novel (or otherwise distinct) Processes
None identified.
4. Beneficiation/Processing Boundaries
EPA has determined that all of the activities discussed in Section 1 are classified as beneficiation activities.
Gold and silver also are recovered from the refining processes for base metals, primarily lead and copper.
Smelting operations remove iron, sulfur, and other impurities from the ore and produce copper anodes for
electrolytic refining. In refining operations, the anodes produced from smelting are purified electrolytically to
produce copper cathodes. The refinery slimes from these operations are processed for precious metals recovery, as
portrayed in Exhibit 6. The recovery of precious metals in lead refineries is a normal part of the operation called
"desilverizing."
A major source of precious metals from the copper industry is electrolytic cell slimes. The slimes are
periodically removed from the cells in the refinery for treatment. The first stage of treatment removes the copper in
the slimes by acid leaching, either as is or after roasting. The decopperized slimes are then placed in a furnace and
melted with a soda-silica flux. The siliceous slag formed in this melting is removed, and air is blown through the
molten material. Lime is added, and a high lead content slag is formed which is combined with the siliceous slag
and returned to the copper anode casting furnace. Next, fused soda ash is added to the furnace and air is again blown
through the melt, forming a soda slag which is removed and treated to recover selenium and tellurium. The
remaining doré in the furnace is removed and sent to refining to recover the precious metals.32 See the selenium and
tellurium commodity reviews for a more detailed discussion of product recovery.
The desilverizing process takes advantage of the solubility of precious metals in molten zinc which is
greater than their solubility in molten lead. Lead from previous stages of refining is brought in contact with a zinc
bath, either in a continuous operation or in batches. The zinc absorbs the precious metals from the lead, and the lead
is then passed onto a dezincing operation. The zinc bath is used until it contains 5,000 to 6,000 troy ounces of
precious metal per ton of zinc. The zinc bath is then retorted to recover zinc by distillation. The zinc is returned to
the desilverizing process, and the "retort metal" is treated by cupellation to produce doré bullion. In the
cupellation,step, the base metals in the retort metal are oxidized with air and removed from the precious metals. The
oxides are all treated for the recovery of their various precious metals. The doré is then sent to refining.33
None identified.
4. Beneficiation/Processing Boundaries
Because the slimes from which gold is recovered are mineral processing wastes generated in the recovery of
other metals, all of the wastes generated during gold recovery from refinery slimes are, therefore, mineral processing
wastes as well.
32
U.S. Environmental Protection Agency, 1988, Op. Cit., pp. 3-100 - 3-115.
33
Ibid.
EXHIBIT 6
Crushing and
Grinding
Flotation
Base Metal
Smelting
Gold Slimes to
Precious Metal
Recovery
Doré
Gold Refining
SECTION 3: PRECIOUS METAL REFINING
The refining process used for gold and silver depends on the composition of the material in the feed. The
most basic operation is "parting" which is the separation of gold and silver. Parting can be done electrolytically or
by acid leaching. In either case, the silver is removed from the gold. Further treatments may be necessary to remove
other contaminants. These treatments have the potential to produce wastes with hazardous characteristics, primarily
corrosivity, since strong acids are used.34
Like several other gold refineries, the Newmont facility in Nevada electrowins its gold cyanide solution
onto steel wool cathodes after carbon stripping. The barren cyanide solution is returned to the leach circuit for gold
recovery. Sludge from the bottom of the electrowinning cell is filtered and sent to the retort for mercury recovery.
The gold/steel wool cathode is placed in a vat containing a sulfuric acid solution. The solution dissolves the steel
wool from the gold and silver, leaving a solid gold and silver residue. The waste sulfuric acid and steel wool
solution is discharged to the tailings slurry. The gold and sislver solids are filtered under vacuum through
diatomaceous earth. The gold and silver filter cake is then sent to the retort furnace where it is subjected to 1,200 °F
for 14 hours. After retorting, a flux of silica and borax is added, and the gold and silver mixture is smelted in an
induction furnace. It is from this induction furnace that gold doré bars are poured. Within days of generation, the
slag generated from this smelting is sent to a ball mill for crushing and grinding and is then leached in tanks with
sodium cyanide. The resulting gold rich slurry is conveyed to the cyanidation/leaching circuit where it is processed
with primary gold-bearing slurries. In some cases, portions of the slag are recycled directly back into the induction
furnaces for gold recovery. The gold slag may have between 3 and 4 ounces per ton of recoverable gold.35 The slag
from one facility reportedly averages 150 ounces of gold per ton.36
Silver metal is produced from silver chloride by a dissolution and cementation process. The silver chloride
is dissolved in a dilute solution of ammonium hydroxide and recovered by cementation. The silver is replaced in
solution, causing the silver ions to be reduced and precipitated from solution as silver metal.
Mercury Recovery
Many gold-bearing ores from the western United States contain small quantities of mercury. The presence
of mercury decreases the gold-loading capacity of the activated carbon. During cyanidation of mercury-bearing
gold-silver ores, significant amounts of mercury are extracted. Addition of calcium sulfide to the cyanide leach
slurry precipitates the solubilized mercury and also some silver.37 Primary mercury is also produced from gold-
bearing ores by roasting or calcining. These processes are described in more detail in the chapter on mercury.
Exhibit 7 presents an overall process flow sheet for gold production from ores.
34
Ibid.
35
U.S. Environmental Protection Agency, Trip report for Newmont Gold Corporation, South Operations
Facilities, Carlin Nevada, May 17, 1995.
36
Newmont Gold Company. Op. Cit.
37
Simpson, W.W., W.L. Staker, and R.G. Sandberg, Calcium Sulfide Precipitation of Mercury From Gold-Silver
Cyanide-Leach Slurries, U.S. Department of Interior, 1986.
3. Identification/Discussion of Novel (or otherwise distinct) Processes
None identified.
4. Beneficiation/Processing Boundaries
EPA established the criteria for determining which wastes arising from the various mineral production
sectors come from mineral processing operations and which are from beneficiation activities in the September 1989
final rule (see 54 Fed. Reg. 36592, 36616 codified at 261.4(b)(7)). In essence, beneficiation operations typically
serve to separate and concentrate the mineral values from waste material, remove impurities, or prepare the ore for
further refinement. Beneficiation activities generally do not change the mineral values themselves other than by
reducing (e.g., crushing or grinding), or enlarging (e.g., pelletizing or briquetting) particle size to facilitate
processing. A chemical change in the mineral value does not typically occur in beneficiation.
Mineral processing operations, in contrast, generally follow beneficiation and serve to change the
concentrated mineral value into a more useful chemical form. This is often done by using heat (e.g., smelting) or
chemical reactions (e.g., acid digestion, chlorination) to change the chemical composition of the mineral. In contrast
to beneficiation operations, processing activities often destroy the physical and chemical structure of the incoming
ore or mineral feedstock such that the materials leaving the operation do not closely resemble those that entered the
operation. Typically, beneficiation wastes are earthen in character, whereas mineral processing wastes are derived
from melting or chemical changes.
EPA approached the problem of determining which operations are beneficiation and which (if any) are
processing in a step-wise fashion, beginning with relatively straightforward questions and proceeding into more
detailed examination of unit operations, as necessary. To locate the beneficiation/processing "line" at a given
facility within this mineral commodity sector, EPA reviewed the detailed process flow diagram(s), as well as
information on ore type(s), the functional importance of each step in the production sequence, and waste generation
points and quantities presented above.
EPA determined that for recovering gold and silver from precious metal refining, the beneficiation/
processing line occurs between electrowinning and retorting because this is where a significant chemical change
occurs. Therefore, because EPA has determined that all operations following the initial "processing" step in the
production sequence are also considered processing operations, irrespective of whether they involve only techniques
otherwise defined as beneficiation, all solid wastes arising from any such operation(s) after the initial mineral
processing operation are considered mineral processing wastes, rather than beneficiation wastes. EPA presents
below the mineral processing waste streams generated during the production of gold and silver, along with
associated information on waste generation rates, characteristics, and management practices for each of these waste
streams.
1. Extraction/Beneficiation Wastes
Mining
Mine water is a waste stream generated from gold and silver production. This waste consists of all water
that collects in mine workings, both surface and underground, as a result of inflow from rain or surface water and
ground water seepage. If necessary, the water is pumped to allow access to the ore body or to keep the mine dry.
This water may be pumped from sumps within the mine pit or from interceptor wells. Mine water may be used and
recycled to the beneficiation circuit, pumped to tailings ponds, or discharged to surface water. Quantity and
chemical composition of mine water varies from site to site.38
38
U.S. Environmental Protection Agency, Mining Industry Profile, Gold, Office of Solid Waste, Special Waste
Branch, 1993, pp. 41-45.
EXHIBIT 7
OVERVIEW OF GOLD PRODUCTION FROM ORES
(Adapted from: Technical Resource Document, Extraction and Beneficiation of Ores and Minerals, July 1994, pp. 1-12.)
Extraction
(open pit or
underground)
Ore
Preparation
Cyanidation-Leaching
Crushing
Heap or Vat
Tank Leaching
Cyanidation-Metal Recovery
Elution
(stripping gold from
activated carbon)
Zinc
Electrowinning
Precipitation
Retorting
Smelting
Refining
Waste Rock. Overburden and mine development is referred to by the industry as waste rock. This waste is
generally disposed of in waste rock piles or dumps. An estimated 25 million metric tons of overburden and mine
development rock was generated in 1980 and 39 million metric tons in 1982. At surface mines, 71 percent of all
material handled is discarded as waste. At underground mines, 20 percent is discarded as waste. The quantity and
composition of the waste rock varies by site. Depending on the composition of the ore body, this waste may contain
sulfides or oxides.
Amalgamation
Black sand may contain residual mercury and be disposed of in a tailings pond.
Ore Preparation
Sulfur dioxide may be routed to an acid plant and converted to sulfuric acid. This may be sold to other
mines or used on-site for carbon washing and regeneration. At least one facility, Newmont's operation in Nevada,
generates sulfuric acid,.39
Cyanidation
Spent Ore. The ore from leaching may contain residual cyanide. The ore in continuous or valley fill heaps
is stacked in lifts and left in place for subsequent leaching, detoxification, and closure. Ore removed from on-off
heap leaching pads is permanently disposed at waste or spent ore disposal sites. Typically, detoxification of the
spent ore involves rinsing with water until the cyanide concentration in the effluent is below a specific standard set
by the State regulatory agency. The heap may then be reclaimed with wastes in place. Spent ore from vat leaching
exists in the form of a slurry composed of gangue and process water bearing cyanide and cyanide-metal complexes.
The spent ore may be treated to neutralize cyanide prior to disposal. The slurry is typically disposed of in a tailings
impoundment with some of the liquid component being recirculated to the tank leach as make-up water.40
Spent Leaching Solution. During the leaching operations, most of the barren cyanide solution is recycled
to leaching activitie. On rare occasions, however, the build-up of metal impurities may interfere with the dissolution
and precipitation of gold and, therefore, require a portion of the solution volume to be bled off and disposed. These
solutions may contain free cyanide and metallo-cyanide complexes of copper, iron, nickel, and zinc, as well as other
impurities, such as arsenic and antimony, mobilized during the leaching. Management practices for these solutions
are unclear; however, they have been discharged to tailings impoundments.41
Merrill-Crowe
Filter cake resulting from zinc precipitation consists primarily of fine gangue material and may contain
gold-cyanide complex, zinc, free cyanide, and lime. The filter may be washed with water, which is disposed of as
part of the waste. The waste is typically sent to tailings impoundments or piles.
Spent leaching solution from zinc precipitation is often returned to the leaching process.
39
Newmont Gold Company. Op. Cit.
40
U.S. Environmental Protection Agency, 1994, Op. Cit., pp. 1-12.
41
Ibid.
Tailings in slurry form, composed of gangue (including sulfide materials and dissolved base metals) and
process water bearing cyanide and cyanide-metal complexes, are generated from carbon-in-pulp and carbon-in-leach
processes. The characteristics of this waste vary depending on the ore, cyanide concentration, and water source
(fresh or recycled). The characteristics of the gangue are dependent on the ore source. The slurry is typically
disposed of in a tailings impoundment with some of the liquid component being recirculated to the tank leach or
other water consumptive system.42
Waste sulfuric acid from elution is exempt under the Bevill Amendment because it is generated in a
beneficiation activity that is uniquely associated with mineral processing. This waste may be corrosive.
Slag. Slag is typically generated at gold mining and milling operations.43 At one facility, metal-bearing
slag is broken off the molten dore and then placed into barrels inside the refinery building. The slag is then
processed for gold recovery, normally within several days of its generation. Specifically, the slag is ground and then
leached in tanks with sodium cyanide. The gold-rich slurry that results is then conveyed, by pipe, to the primary
gold-bearing slurries in the mill for mixing. The facility also reports that in the past, slag also was processed by
placing it directly back into the induction furnaces for gold recovery. Assays performed by Newmont Gold during
1995 and 1997 show that slag generated at its Nevada Mines Complex typically contains between 100 and 700
ounces of gold per ton. Tests also indicate that the slag may exhibit the characteristic of toxicity for cadmium.44 No
published information regarding waste generation rate or characteristics was found, though one facility reportedly
generates approximately 38 to 57 tons of slag per year.45, 46 This facility also indicated that it takes weeks to
accumulate enough slag to constitute a large enough batch for cost effective metals recovery. The total industry
generation rate for slag is thus probably less than 500 metric tons per year. One facility also indicated that slag may
possess the characteristic of toxicity for cadmium.47 The slag is not stored or processed on the land, nor does it enter
the outside environment. Slag is believed to be fully recycled and was formerly classified as a byproduct.
WWTP Sludge. WWTP sludge is typically generated at gold refineries.48 Although no published
information regarding waste generation rate or characteristics was found, we used the methodology outlined in
Appendix A of this report to estimate a low, medium, and high annual waste generation rate of 100 metric tons/yr,
360,000 metric tons/yr, and 720,000 metric tons/yr, respectively. We used best engineering judgment to determine
that this waste may exhibit the characteristic of toxicity for silver. This waste may be recycled and was formerly
classified as a sludge.
42
Ibid.
43
Precious Metals Producers. January 25, 1996. Op. Cit.
44
Newmont Gold Company. May 12, 1997. Op. Cit.
45
Newmont Gold Company. January 25, 1996. Op. Cit.
46
Newmont Gold Company. May 12, 1997. Op. Cit.
47
Ibid.
48
Precious Metals Producers. Op. Cit.
Spent Furnace Dust. Spent furnace dust is typically generated at gold mining and milling operations.49 As
part of the smelting process, gold-bearing dust is generated in the induction furnaces. This dust is conveyed by pipe
to a baghouse located in the refinery building and collected in barrels. The barrels of baghouse dust are never stored
outdoors. At the Nevada Mines Complex, sealed barrels of baghouse dust are accumulated for up to four months
before being shipped off-site for smelting. At other facilities, smelting of baghouse dust may be done in the on-site
induction furnace. Assays performed by Newmont Gold in 1995 and 1996 show that the baghouse dust from its
Nevada Mines Complex contains approximately 2,200 ounces of gold per ton and that the dust may exhibit the
characteristic of toxicity for cadmium and selenium. This facility generated approximately 1,550 ounces of gold in
1996 by smelting baghouse dust, generating revenues of over $600,000.50 Thus, the facility generated less than one
ton of baghouse dust in 1996, suggesting an industry-wide generation rate of less than 9 tons per year. The dust is
entirely recycled and was formerly classified as a byproduct. At one facility, several months are required to
accumulate sufficient baghouse dust to constitute a large enough batch to ship off-site for smelting.51 We used best
engineering judgment to determine that this waste may exhibit the characteristic of toxicity for silver. At one
facility, the dust exhibits the characteristic of toxicity for cadmium and selenium. The dust is not land stored and
never enters the outside environment.52,53 This waste is recycled and was formerly classified as a byproduct.
Retort Cooling Water. The retorting process at Newmont Gold’s Nevada Mines Complex generates
mercury-, silver- and gold-bearing gases. Water is used to cool and condense those gases. Through this process, the
cooling water becomes contaminated with gold, silver, and mercury. This water is conveyed by pipe to the main
beneficiation circuit to allow recovery of the metals and reuse of the water.54
Wastewater is typically generated at gold refineries and is generated from numerous sources, including the
smelter air pollution control (APC), silver chloride reduction, electrolytic cell wet APC, and electrolyte preparation
wet APC.55 Wastewater from electrolyte preparation wet APC, electrolytic cell wet APC, and smelter wet APC may
contain toxic metals, suspended solids, oil, and grease. This waste may be recycled.56 Although no published
information regarding waste generation rate or characteristics was found, we used the methodology outlined in
Appendix A of this report to estimate a low, medium, and high annual waste generation rate of 440,000 metric
tons/yr, 870,000 metric tons/yr, and 1,700,000 metric tons/yr, respectively. We used best engineering judgment to
determine that this waste may exhibit the characteristic of toxicity for arsenic, silver, cadmium, chromium, and lead.
This waste was formerly classified as a sludge.
Refining Wastes. The most basic refining operation for the separation of gold and silver is "parting" which
can be done electrolytically or by acid leaching. Further treatments are sometimes necessary to remove additional
contaminants. Although no published information regarding waste generation rate or characteristics was found, we
used the methodology outlined in Appendix A of this report to estimate a low, medium, and high annual waste
generation rate of 100 metric tons/yr, 360,000 metric tons/yr, and 720,000 metric tons/yr, respectively. We used best
49
Ibid.
50
Newmont Gold Company. May 12, 1997. Op. Cit.
51
Ibid.
52
Newmont Gold Company. January 25, 1996. Op. Cit.
53
Newmont Gold Company. May 12, 1997. Op. Cit.
54
Ibid.
55
Precious Metals Producers. Op. Cit.
56
U.S. Environmental Protection Agency, Development Document for Effluent Limitations Guidelines and
Standards for the Nonferrous Metals Manufacturing Point Source Category, Vol. V, 1989, pp. 2185-2186.
engineering judgment to determine that this waste may exhibit the characteristics of toxicity for silver and
corrosivity. This waste is recycled to extraction/beneficiation units.
Ancilary hazardous wastes also may be generated at on-site laboratories, and may include used chemicals
and liquid samples. Other hazardous wastes may include spent solvents (e.g., petroleum naptha), acidic tank
cleaning wastes, and polychlorinated biphenyls from electrical transformers and capacitors. Non-hazardous wastes
may include tires from trucks and large machinery, sanitary sewage, waste oil (which may or may not be hazardous),
and other lubricants.
Three commenters submitted comments in response to the January 25, 1996 Supplemental Proposed Rule
(COMM43, COMM57, COMM58), and one commenter submitted comments in response to the May 12, 1997
Second Supplemental Proposed Rule (Newmont Gold Company). All four commenters provided new factual
information that has been incorporated into the gold and silver sector report.
Sector-specific Issues
C Three commenters stated that the retorting step is, in fact, a beneficiation activity (COMM43,
COMM57, COMM58). EPA has clarified in the report that the beneficiation/mineral processing
boundary for gold recovery from ores is between electrowinning and retorting. Therefore, retorting is a
mineral processing activity.
C One commenter stated that slag and spent furnace dust are not wastes because they are destined for
reclamation (COMM57). EPA recognizes that these materials contain high concentrations of precious
metals and that they are reclaimed. However, they are still considered to be wastes, and no change was
made to the report.
C One commenter expressed confusion over the status of acid washing solution (COMM43), while
another commenter stated that EPA incorrectly classified acid washing during the elution process as
non-uniquely associated with mineral processing (COMM57). The Agency clarified in the report that
the use of an acid solution to dissolve the steel wool from the gold/steel wool cathode is a beneficiation
activity that is uniquely associated with mining or mineral processing. Acid wash solution from carbon
regeneration activities after the gold is stripped, however, is not uniquely associated and, therefore, is
not exempt under the Bevill Amendment.
C One commenter requested that EPA clarify the status of sulfuric acid (COMM43). The report was
modified to indicate that waste sulfuric acid from elution is exempt under the Bevill Amendment
because it is generated in a beneficiation activity that is uniquely associated with mining or mineral
processing.
C One commenter noted an inconsistency in that spent carbon is identified as a beneficiation waste
whereas carbon fines are not uniquely associated with mineral processing (COMM43). One commenter
stated that spent carbon is routinely regenerated and reused and, therefore, is not a waste (COMM43).
Another commenter stated that carbon fines generated at on-site carbon regeneration kilns as well as
other secondary materials generated during carbon regeneration are uniquely associated with mineral
processing (COMM57). Spent carbon after the gold has been stripped is, in fact, a waste. The Agency
clarified that spent carbon from cyanidation is a non-uniquely associated waste because spent carbon
results from carbon regeneration which is a non-uniquely associated operation because many industries
routinely regenerate carbon and generate wastes such as carbon fines. Therefore, carbon regeneration
and the resulting wastes are non-uniquely associated with mining or mineral processing.
BIBLIOGRAPHY
Lucas, John. "Gold." From Mineral Commodity Summaries. U.S. Bureau of Mines. January 1995. pp. 68-69.
Lucas, John. "Gold." From Minerals Yearbook Volume 1. Metals and Minerals. U.S. Bureau of Mines. 1992. pp.
535-561.
Personal communication between ICF Incorporated and Robert G. Reese, U.S. Bureau of Mines. September 23,
1994.
Personal communication between ICF Incorporated and John M. Lucas, U.S. Bureau of Mines. September 15,
1994.
Reese, Robert. "Silver." From Mineral Commodity Summaries. U.S. Bureau of Mines. January 1995. pp. 154-
155.
Reese, Robert. "Silver." From Minerals Yearbook Volume 1. Metals and Minerals. U.S. Bureau of Mines. 1992.
pp. 1199-1211.
Simpson, W.W., W.L. Staker, and R.G. Sandberg. Calcium Sulfide Precipitation of Mercury From Gold-Silver
Cyanide-Leach Slurries. U.S. Department of Interior. 1986.
U.S. Environmental Protection Agency, Trip report for Newmont Gold Corporation, South Operations
Facilities, Carlin Nevada, May 17, 1995.
U.S. Environmental Protection Agency. Technical Resource Document, Treatment of Cyanide Heap Leaches and
Tailings. Office of Solid Waste, Special Waste Branch. 1994. pp. 2-4.
U.S. Environmental Protection Agency. Technical Resource Document, Extraction and Beneficiation of Ores and
Minerals. Office of Solid Waste, Special Waste Branch. Vol. 2. 1994. pp. 1-12.
U.S. Environmental Protection Agency. Mining Industry Profile, Gold. Office of Solid Waste, Special Waste
Branch. 1993. pp.41-45.
U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines and
Standards for the Nonferrous Metals Manufacturing Point Source Category. Volume V. Office of Water
Regulations Standards. May 1989. pp. 2185-2186.
U.S. Environmental Protection Agency. "Gold and Silver." From 1988 Final Draft Summary Report of Mineral
Industry Processing Wastes. 1988. 3-100 - 3-115.
IODINE
A. Commodity Summary
Iodine compounds are found in seawater, seaweed, marine organisms, and brines. Iodine and its
compounds are generally marketed in the form of crude iodine, resublimed iodine, calcium iodate, calcium iodide,
potassium iodide, sodium iodide, and numerous organic compounds.1 Final uses of iodine include animal feed
supplements, catalysts, inks and colorants, pharmaceuticals, photographic chemicals and film, sanitary and industrial
disinfectants, and stabilizers.
Japan and Chile are the largest producers of iodine in the world and account for 99% of the U.S. iodine
imports. All domestic iodine production is from iodine-rich natural brines in the deep subsurface of the Anadarko
basin of northwestern Oklahoma. Oklahoma production began in 1977 and at present, three companies operate a
total of four facilities (three major plants and one miniplant) for the recovery of iodine. The U.S. Bureau of Mines
estimates that domestic production was 2,000,000 kilograms in 1994.2 Exhibit 1 presents the names and locations of
the facilities involved in the production of iodine.
EXHIBIT 1
All three facilities (Asahi Glass Company of Japan, Iochem Corporation of Japan, and North America Brine
Resources) obtain iodine-rich brines from the Morrowan sandstones. Asahi Glass Company operates 22 production
wells and 10 injection wells ranging in depth from 2,130 to 2,290 meters. The Iochem facility has nine production
wells and four injection wells ranging in depth from 3,000 to 3,183 meters. The North American Brine Resources
facility operates two production wells and three injection wells drilled to about 1,800 meters.
North American Brine Resources also operates a mini facility near Dover, OK. At the Dover facility, North
American Brine Resources recovers iodine from oil-field brines collected from a number of oil and gas wells in
nearby parts of northwestern Oklahoma.
1
Kenneth S. Johnson, "Iodine," from Industrial Minerals and Rocks, 6th edition, Society for Mining, Metallurgy,
and Exploration, 1994, pp. 583-587.
2
Phyllis Lyday, "Iodine," from Mineral Commodity Summaries, U.S. Bureau of Mines, January 1995, pp. 82-83.
Brines are separated from hydrocarbons by using the blowing-out process. Iochem Corporation and North
American's Woodward facility both use this process.3 Exhibit 2 presents a typical process flow diagram for the
production of iodine from brines by the blowing-out process.
Exhibit 2 displays the blowing-out process. In the first stage of this process, hydrogen sulfide gas
(contained in the brine) is removed. This gas is reacted to form sulfur compounds which are sent to a hazardous
waste disposal facility. The second stage of processing is chlorine oxidation to convert iodide to iodine. The iodine
is then removed from the brine by air vapor stripping (air-blowout). The waste brine is treated with lime to adjust
pH and is reinjected into Class IV disposal wells. The iodine vapor is absorbed by a solution of hydriodic and
sulfuric acids. Sulfur dioxide is added to reduce the absorbed iodine to hydriodic acid. Most of the solution is
recirculated to the absorption tower, but a bleed stream is sent to a reactor for iodine recovery. In the reactor,
chlorine is added to oxidize and liberate the iodine which precipitates and settles out of solution. The settled iodine
is filtered to remove waste liquor and melted under a layer of concentrated sulfuric acid. The melted iodine is then
solidified either as flakes or ingots.4
Iodine is also recovered from oil well brines. In a settling tank, the iodine containing brine settles to the
bottom and the oil rises to the top. The oil is skimmed off and processed with other oil from nearby wells. The brine
is sent through a chlorinator which frees the iodine. It is then absorbed onto charcoal which is back-flushed with
potassium or sodium hydroxide when full. This solution is treated with hydrochloric acid which results in a 90%
crude iodine product. The spent brine is reinjected and the potassium/sodium hydroxide is recycled.5
While domestic iodine production employs the chlorine-oxidation air-blowout method for recovery of
iodine, three other brine clarification processes exist. In one process, silver iodide is precipitated by the addition of a
silver nitrate solution. The silver iodide is filtered and treated with scrap iron to form metallic silver and a solution
of ferrous iodide. The silver is redissolved in nitric acid and recycled, and the solution is treated with chlorine to
liberate the iodine. In a second process, chlorine is added after clarification to liberate the iodine as a free element in
solution. Passing the solution over bales of copper wire precipitates insoluble cuprous iodides. At intervals, the
bales are agitated with water to separate the adhering iodide; the bales are then recycled. The cuprous iodide
suspended in the water is filtered, dried, and sold. The third process uses ion-exchange resins on brines which have
been oxidized to liberate iodine. The liberated iodine, which is in the form of polyiodide, is absorbed on an anion-
exchange resin. When the ion-exchange resin is saturated, it is discharged from the bottom of the column and then
transferred to the elutriation column. Iodine is elutriated with a caustic solution followed by sodium chloride. The
regenerated resin is returned to the absorption column. The iodine-rich elutriant is acidified and oxidized to
precipitate iodine. The crude iodine is then separated in a centrifuge and purified with hot sulfuric acid or refined by
sublimation.6
3
Phyllis A. Lyday, "Iodine," from Minerals Yearbook Volume 1. Metals and Minerals, U.S. Bureau of Mines,
1992, pp. 609-612.
4
U.S. Environmental Protection Agency, "Iodine," from 1988 Final Draft Summary Report of Mineral Industry
Processing Wastes, 1988, pp. 2-109 - 2-112.
5
Personal Communication between ICF Incorporated and Phyllis Lyday, U.S. Bureau of Mines. October 11, 1994.
6
"Iodine and Iodine Compounds," Kirk-Othmer Encyclopedia of Chemical Technology, 3rd ed., 1981, Vol. XIII,
pp. 655-656.
EXHIBIT 2
THE BLOWING-OUT PROCESS
(Adapted from; 1988 Final Draft Summary Report of Mineral Industry Processing Wastes, 1988, pp. 2-109 - 2-112.)
Air
Waste Brine
Stripping
Sulfur Dioxide
Treatment
Absorbtion
Tower
Bleed
Precipitation
Waste Bleed Liquor
With Chlorine
Melting Under
Iodine Product
Sulfuric Acid
The Chilean nitrate industry employs another method for iodine recovery. Iodine is extracted from caliche
as sodium iodate, along with sodium nitrate. The iodate accumulates in mother liquors during crystallization of the
nitrate. Part is drawn off and treated with sodium bisulfite solution. Fresh mother liquor is added to the solution to
liberate the iodine. The precipitated iodine is filtered in bag filters and the iodine-free mother liquor is returned to
the nitrate leaching cycle after neutralization with soda ash. The iodine cake is washed, pressed, broken up, and
sublimed in retorts. The product is then crushed and packaged.7
4. Beneficiation/Processing Boundaries
Based on a review of the process, there are no mineral processing operations involved in the production of
iodine.
Existing data and engineering judgement suggest that the wastes listed below from iodine production do not
exhibit any characteristics of hazardous waste. Therefore, the Agency did not evaluate these materials further.
1. Extraction/Beneficiation Wastes
Sulfur compounds from hydrogen sulfide removal are sent to a hazardous waste disposal facility.
Waste brine. Waste brine contains 6,800 kkg of spent brine solids per kkg of product iodine. Waste brine
is processed for other solids recovery and then either used in chlor-alkali manufacture or returned to the source.
Bromine, calcium chloride and magnesium hydroxide may be recovered from these spent brines.8
Filtration
Sludge.
Waste Acid.
D. Non-uniquely AssociatedWastes
Non-uniquely associated and ancillary hazardous wastes may be generated at on-site laboratories, and may
include used chemicals and liquid samples. Other hazardous wastes may include spent solvents, and acidic tank
cleaning wastes. Non-hazardous wastes may include tires from trucks and large machinery, sanitary sewage, and
waste oil other lubricants.
7
Ibid.
8
U.S. Environmental Protection Agency, Multi-Media Assessment of the Inorganic Chemicals Industry, Volume II,
1980, Chapter 9.
E. Summary of Comments Received by EPA
"Iodine and Iodine Compounds." Kirk-Othmer Encyclopedia of Chemical Technology. 3rd ed. 1981. Vol. XIII.
pp. 655-656.
Johnson, Kenneth S. "Iodine." From Industrial Minerals and Rocks. 6th ed. Society for Mining, Metallurgy, and
Exploration. 1994. pp. 583-587.
Lyday, Phyllis. "Iodine." From Mineral Commodity Summaries. January 1995. pp. 82-83.
Lyday, Phyllis. "Iodine." From Minerals Yearbook Volume 1. Metals and Minerals. U.S. Bureau of Mines. 1992.
pp. 609-612.
Personal Communication between ICF Incorporated and Phyllis Lyday, Bureau of Mines. October 11, 1994.
U.S. Environmental Protection Agency. "Iodine." From 1988 Draft Report Summary Report of Mineral Industry
Processing Wastes. 1988. pp. 2-109 - 2-112.
U.S. Environmental Protection Agency. Multi-Media Assessment of the Inorganic Chemicals Industry. Vol. 2.
1980. Chapter 9.
IRON AND STEEL
A. Commodity Summary
The iron and steel industry, including primary and secondary producers, is composed of 79 companies
that produce raw steel at 116 locations. Iron is generally produced from iron ore (taconite) in a primary mineral
production process, while steel is produced using both primary and secondary processes. Primary production refers
to those operations where the feedstock is composed of at least 50 percent ore (or ore that has been beneficiated).
Electric arc furnaces use a high percentage of scrap steel as the feedstock in their operations and are therefore
classified as secondary production and not considered primary minerals processing. Although the electric arc
furnace process is described in this section, some of the wastes generated from this operation are currently regulated
under RCRA Subtitle C. Specifically, electric arc furnace dust (K061) is a listed hazardous waste.
The annual aggregate raw steel production capacity is 99 million metric tons; 1993 production is reported to
be 87 million metric tons. According to the U.S. Bureau of Mines, the iron and steel producers and ferrous foundries
produced goods valued at $55 billion. Currently, pig iron (i.e., molten iron from iron blast furnaces) is produced at
15 companies operating integrated steel mills, with approximately 58 blast furnaces (of which 41 or 42 are in
continuous operation). Integrated companies accounted for approximately 67% of steel production, including output
of their electric arc furnaces (which are classified as secondary production).1
Pig iron production in 1994 is estimated at 49 million metric tons. Pig iron is sent to either basic oxygen
furnaces or electric arc furnaces for further processing at steel facilities. Basic oxygen furnaces (BOFs) and electric
arc furnaces (EAFs) account for 61 percent and 39 percent of steel production, respectively. Continuously cast steel
accounted for 89 percent of steel production. Lastly, open hearth furnaces (OHFs) have been phased out and were
not used domestically to produce steel in 1993.2 Exhibit 1 presents the names and locations of facilities involved in
the primary production of iron and steel.
EXHIBIT 1
1
Gerald Houck, "Iron and Steel," from Mineral Commodity Summaries, U.S. Bureau of Mines, January 1995, p. 86.
2
Ibid.
EXHIBIT 1 (continued)
a
Gerald Houck, "Iron and Steel." from Minerals Yearbook Volume 1. Metals and Minerals, U.S. Bureau of Mines, 1992, p. 649.
On a tonnage basis, about nine-tenths of all metal consumed in the United States is iron or steel. Iron and
steel are used in the manufacture of transportation vehicles, machinery, pipes and tanks, cans and containers, and the
construction of large buildings, roadway superstructures, and bridges.3 According to the U.S. Bureau of Mines in
1993, steel consumption was divided amongst the following uses: warehouse and steel service centers, 26%;
3
Gerald Houck, "Iron and Steel," from Mineral Facts and Problems, U.S. Bureau of Mines, 1985, p. 412.
transportation (mainly for automotive production), 16%; construction 15%, cans and containers, 5%; and other uses,
38%.4
The production of steel products from iron ore involves two separate steps: ironmaking and steelmaking.
Each of these is described in detail below. Iron blast furnaces produce molten iron (pig iron) that can be cast
(molded) into products; however, the majority of pig iron is used as the mineral feedstock for steel production. Steel
furnaces produce a molten steel that can be cast, forged, rolled, or alloyed in the production of a variety of materials.
Ironmaking
Iron is produced either by blast furnaces or by one of several direct reduction processes; blast furnaces,
however, account for over 98 percent of total domestic iron production.5 The modern blast furnace consists of a
refractory-lined steel shaft in which a charge is continuously added to the top through a gas seal. The charge
consists primarily of iron ore, sinter, or pellets; coke; and limestone or dolomite. Iron and steel scrap may be added
in small amounts. Near the bottom of the furnace, preheated air is blown in. Coke is combusted in the furnace to
produce carbon monoxide which reduces the iron ore to iron. Silica and alumina in the ore and coke ash are fluxed
with limestone to form a slag that absorbs much of the sulfur from the charge. Molten iron and slag are
intermittently tapped from the hearth at the bottom. The slag is drawn off and processed. The product, pig iron, is
removed and typically cooled, then transported to a steel mill operation for further processing in either an electric arc
furnace or a basic oxygen furnace, as depicted in Exhibit 2. As shown in Exhibit 2, the iron can also be directly
reduced before it is sent for further processing.
Recent changes in the process include modifications in the fluxing practices. Flux is often introduced
through fluxed sinter or fluxed pellets rather than by direct charging. The use of external desulfurization of hot
metals prior to steel making has also increased.6
Steelmaking
All contemporary steelmaking processes convert pig iron, scrap, or direct-reduced iron, or mixtures of
these, into steel by a refining process that lowers the carbon and silicon content and removes impurities (mainly
phosphorus and sulfur). Three major furnace types can be used for making steel:
The latter predominantly uses scrap (i.e., non-mineral material) as feedstock and is classified as a secondary
process. The open-hearth process was prevalent in the United States between 1908 and 1969, but it is no
4
Gerald Houck, 1994, Op. Cit., p. 90.
5
American Iron and Steel Institute, "Annual Statistical Report," 1984, p. 78.
6
Harold R. Kokal and Madhu G. Ranade, "Fluxes for Metallurgy," from Industrial Minerals and Rocks, 1994, pp.
668-669.
EXHIBIT 2
IRONMAKING
Coal
Coke
Ore briquetting)
Blast Furnace
APC Gas Process Water
Cooling Tower
Blowdown
Steel Scrap
Electric Arc Furnace Open Hearth Furnace
Basic Oxygen Furnace
(secondary production) (no longer used)
APC Dust
(on-site disposal)
Alternate Process
longer in use domestically. The basic oxygen process has supplanted it as the predominant primary steel-making
process, making up approximately 95 percent of domestic primary steel production in 1987.7
Modern steelmaking also includes treatment of steel in ladles. This use of ladles (1) improves the
cleanliness of the steelmaking process, (2) increases throughput in steel vessels, and (3) allows for shape control of
inclusions in continuous casting operations.8
A general flow diagram for the production of raw steel from iron ore is presented in Exhibit 2. In general,
the process involves (1) beneficiation of the iron ore, (2) either direct-reduction or reduction in an iron blast furnace,
(3) processing in steelmaking furnaces, and (4) casting.
Ironmaking
There are a variety of beneficiation methods that can be used to prepare iron ores, depending on the iron
content in the ores. Some ores contain greater than 60 percent iron and require only crushing and blending to
prepare them for further processing. In other cases, operations including screening and concentrating are necessary
to prepare the raw materials. The characteristics of the iron-bearing ores vary geographically. Specifically,
magnetite is the main iron-bearing ore in the Lake Superior district and in the northeastern United States, while
hematite and hematite magnetite mixtures tend to be found in ores in Alabama and the Southwest.
When magnetite occurs in lower grade deposits, the ore is ground, and the concentrate is separated
magnetically from the gangue with the ore in a water suspension. Ore containing hematite can be high in clay
content and requires washing to remove the clay and concentrate the iron. Low grade ores that can not be separated
magnetically may also need to be concentrated via washing, jigging, heavy media separation, or flotation.9
Ores that will be sent to blast furnaces for ironmaking need to be permeable to allow for an adequate flow
of gas through the system. Additionally, concentrates in raw ores that are very fine need to be agglomerated before
they can be used as feed stock for the blast furnaces. The three major processes used for agglomeration include:
C sintering;
C pelletizing; and
C briquetting.
Sintering. Sintering involves mixing the iron-bearing material such as ore fines, flue dust, or concentrate
with fuel (e.g., coke breeze or anthracite). 10 The mixture is then spread on surface beds which are ignited by gas
burners. The heating process fuses the fine particles, and the resulting product is lumpy material known as sinter.
The sinter is sized and the fines are recycled. Sintering operations are used to recycle wastes from other iron and
steel manufacturing processes.
7
Frederick J. Schottman, "Iron an Steel," from Minerals Yearbook Volume I. Metals and Minerals, U.S. Bureau of
Mines, 1989, p. 511.
8
Harold R. Kokal and Madhu G. Ranade, 1994, Op. Cit., pp. 668-9.
9
U.S. Environmental Protection Agency, "Iron and Steel," from 1988 Final Draft Summary Report of Mineral
Industrial Processing Wastes, Office of Solid Waste, 1988, p. 3-128.
10
Ibid.
Pelletizing. Pelletizing involves forming pellets from the raw ore or concentrates, then hardening the
pellets by heating. Solid fuel can be combined with the concentrate to promote the heating necessary to harden the
pellet. Common binders added to strengthen the pellets include limestone, dolomite, soda ash, bentonite, and
organic compounds. After the pellets are sized, any remaining fraction of materials are recycled back through the
sintering process.
Briquetting. Briquetting, another form of agglomeration, involves heating the ore and pressing it into
briquettes while the materials are still hot. Once the briquettes are cooled, they are sent directly to the blast furnaces.
Direct Reduction. Direct reduction involves the reduction of iron ore that is in the solid state - at less than
1000 oC.11 The solid primary metal produced by direct reduction of iron ores (DRI) can be used to supply electric
arc furnaces.
Blast Furnace. During ironmaking, agglomerated iron ore is combined with prepared limestone, silica, and
coke and placed into a blast furnace. Heated air is blown into the furnace and causes the limestone and silica to form
a fluid slag which combines with other impurities. The slag can be separated from the molten iron and sent to a slag
reprocessing unit. Generally, the molten iron from the blast furnace is transferred directly to the steelmaking
furnaces.
A number of integrated steelwork facilities in the United States have increased their use of fluxed pellets,
which are more easily reducible. The fluxed pellets are produced by adding limestone (CaCO3) and/or dolomite
[(Ca,Mg)CO3] to the iron ore concentrate during the balling stage. Flux is added until the ratio of calcium and
magnesium oxide to silicon dioxide and aluminum oxide ((CaO+MgO)/(SiO2+Al2O3)) in the pellet is above 0.6.
The most common ratio documented is approximately 1.0.12
Steelmaking
Open Hearth Furnace (no longer used). During the open-hearth process, a relatively shallow bath of
metal was heated by a flame that passed over the bath from the burners at one end of the furnace while the hot gases
resulting from combustion were pulled out the other end. The heat from the exhaust gas was retained in the exhaust
system's brick liners, which were known as checker-brick regenerators. Periodically the direction of the flame was
reversed and air was drawn through what had been the exhaust system; the hot checker-bricks preheated the air
before it was used for combustion in the furnace. Impurities were oxidized during the process and fluxes formed a
slag; this slag was drawn off and processed or discarded.
11
J. Astier, "Present Status of Direct Reduction and Smelting Reduction," from Steel Times, October 1992, pp. 453-
458.
12
William S. Kirk, "Iron Ore," from Minerals Yearbook Volume 1. Metals and Minerals, 1992, p. 618.
Basic Oxygen Furnace. The basic oxygen process uses a jet of pure oxygen that is injected into the molten
metal by a lance of regulated height in a basic refractory-lined converter. Excess carbon, silicon, and other reactive
elements are oxidized during the controlled blows, and fluxes are added to form a slag. This slag, one of the RCRA
special wastes, is drawn off and processed or discarded.
The first step in the BOF process is charging the furnace. Hot metal (molten iron from the blast furnace)
which accounts for most of the metallic charge is added to the furnace by ladles. Once the furnace has been charged,
a water-cooled oxygen lance is lowered into the furnace and high purity oxygen is blown in the top of the furnace.
One modification to this process is the Q-BOP in which the oxygen and other gases are blown in from the bottom of
the furnace instead of the top. In the bottom blown process, oxygen is introduced through a number of tuyeres,
consisting of two concentric pipes in the bottom of the converter.13
In the furnace, oxygen combines with the carbon and other unwanted elements to oxidize the impurities in
the molten charge, and thereby converting the molten charge to steel. The lime and other fluxes help remove the
oxidized impurities as a layer of slag. The refined steel is then poured into ladles. At this point, any alloys can be
added to the steel to obtain the desired strength and characteristics required in the final product.
Electric Arc Furnace (secondary production). Electric arc furnaces are generally used for scrap
processing and have traditionally been used to produce alloy, stainless, tool, and specialty steels. Scrap steel is the
principal metallic charge to electric furnaces. Direct reduction of iron ore also produces pellets with high enough
iron content to be used. Limestone and other fluxes are charged after the scrap becomes molten. As in the blast
furnace operation, the impurities in the steel form a floating layer of slag that can be poured off. The molten steel is
then poured into ladles and sent to be cast.
In all steelmaking operations, gases from the furnace must be cleaned in order to meet air pollution control
requirements. Facilities may use dry collection (e.g., bag houses, filters, or electrostatic precipitators) or wet
scrubbers or, as is most often practiced, both types of controls. Large volumes of dust and scrubber sludge are
collected for either further processing or disposal. Some of these air pollution control residuals are RCRA special
wastes.
The molten steel, from whichever type of furnace is used, flows into ladles and is sent for further
processing at rolling mills to form the finished products.
C Dezincing and Detoxification of Electric Arc Furnace Steelmaking Dust via Ammonium
Carbonate Leaching. The use of ammoniacal ammonium carbonate (AAC) leaching for the treatment
of carbon steel making EAF dust has been investigated on a laboratory scale. The tests were performed
using dust samples from three European steel companies. The dusts were found to be toxic due to the
leachability of silver, mercury, lead, and cadmium. After treatment, the toxicity tests indicated
leachates below past and current EPA toxicity threshold limits.14
C Recovery of Manganese from Steel Plant Slag by Carbamate Leaching. The U.S. Bureau of Mines
investigated the feasibility of using ammonium carbamate leaching to recover manganese from steel
plant slag. It was found that treatment of the slag with hydrogen prior to the leaching increased the
13
Association of Iron and Steel Engineers, The Making, Shaping and Treating of Steel, 1985, pp. 539-652.
14
R.L. Nyirenda et al, "Dezincing and Detoxification of Electric Arc Furnace Steelmaking Dust via Ammonium
Carbonate Leaching," The Minerals, Metals, & Mining Society, 1993, pp. 894-906.
amount of manganese recovered. Results indicated that the method cannot be applied satisfactorily to
all steelmaking slags.15
C Glassification16 of Electric Arc Furnace Dust. A new process has been developed to treat hazardous
materials, including electric arc furnace dust, slag, and spent refractories. The process, known as
Glassification, utilizes electric arc furnace dust from both the steel and nonferrous metals industries to
produce glass products.17
C Treatment of Steel Plant Wastes by Magnetic Cyclones. Steel plants generate sludges containing
high concentrations of iron which display ferromagnetic properties. Methods of treating these wastes
to take advantage of these properties using magnetic cyclones have been evaluated. The results
indicated that the cycloning process creates an underflow with a high solids content and a clean water
overflow.18
4. Beneficiation/Processing Boundary
EPA established the criteria for determining which wastes arising from the various mineral production
sectors come from mineral processing operations and which are from beneficiation activities in the September 1989
final rule (see 54 Fed. Reg. 36592, 36616 codified at 261.4(b)(7)). In essence, beneficiation operations typically
serve to separate and concentrate the mineral values from waste material, remove impurities, or prepare the ore for
further refinement. Beneficiation activities generally do not change the mineral values themselves other than by
reducing (e.g., crushing or grinding), or enlarging (e.g., pelletizing or briquetting) particle size to facilitate
processing. A chemical change in the mineral value does not typically occur in beneficiation.
Mineral processing operations, in contrast, generally follow beneficiation and serve to change the
concentrated mineral value into a more useful chemical form. This is often done by using heat (e.g., smelting) or
chemical reactions (e.g., acid digestion, chlorination) to change the chemical composition of the mineral. In contrast
to beneficiation operations, processing activities often destroy the physical and chemical structure of the incoming
ore or mineral feedstock such that the materials leaving the operation do not closely resemble those that entered the
operation. Typically, beneficiation wastes are earthen in character, whereas mineral processing wastes are derived
from melting or chemical changes.
EPA approached the problem of determining which operations are beneficiation and which (if any) are
processing in a step-wise fashion, beginning with relatively straightforward questions and proceeding into more
detailed examination of unit operations, as necessary. To locate the beneficiation/processing "line" at a given
facility within this mineral commodity sector, EPA reviewed the detailed process flow diagram(s), as well as
information on ore type(s), the functional importance of each step in the production sequence, and waste generation
points and quantities presented above.
EPA determined that for this specific mineral commodity sector, the beneficiation/processing line occurs
between agglomeration (sintering, pelletizing, and briquetting) and reduction of iron ore in a blast furnace. EPA
identified this point in the process sequence as where beneficiation ends and mineral processing begins because it is
15
S.N. McIntosh and E.G. Baglin, "Recovery of Manganese from Steel Plant Slag by Carbamate Leaching," U.S.
Bureau of Mines, 1992.
16
Glassification is a registered trademark.
17
R.B. Ek and J.E. Schlobohm, "Glassification of Electric Arc Furnace Dust," from Iron and Steel Engineer,
April 1993, pp. 82-84.
18
John L. Watson and Suren Mishra, "The Treatment of Steel Plant Wastes by Magnetic Cyclones," Conference
Paper from Symposium on Emerging Process Technologies for a Cleaner Environment, Phoenix, AZ, February 24-
27 1992.
here where a significant chemical change to the iron ore occurs. Therefore, because EPA has determined that all
operations following the initial "processing" step in the production sequence are also considered processing
operations, irrespective of whether they involve only techniques otherwise defined as beneficiation, all solid wastes
arising from any such operation(s) after the initial mineral processing operation are considered mineral processing
wastes, rather than beneficiation wastes. EPA presents the mineral processing waste streams generated after the
beneficiation/processing line in section C.2, along with associated information on waste generation rates,
characteristics, and management practices for each of these waste streams.
1. Extraction/Beneficiation Wastes
Waste characterization data, waste generation data, and waste management data are not available for all of
the wastes identified as generated from the production of iron and steel.
Tailings. Wastes from magnetic separation include tailings consisting mostly of silicate rock. The
magnetite ore from lower grade deposits is ground, and the concentrate is separated magnetically from the gangue
with the ore in a water suspension. These wastes are typically managed in tailing impoundments.
Wastewater and Waste Solids. Ore containing hematite can be high in clay content and require washing
to remove the clay and concentrate the iron. The wastewater and waste solids generated from washing ores
containing clay are not expected to be hazardous. No information is available on management practices for these
wastes.19
Ferrous metal production operations generate four RCRA special mineral processing wastes that are exempt
from RCRA Subtitle C: iron blast furnace slag, iron blast furnace air pollution control dust/sludge, steel furnace
slag, and steel furnace air pollution control dust/sludge. The Agency did not evaluate the four RCRA special
mineral processing wastes further. Besides these RCRA special wastes, the only other types of wastes generated
appear to be various types of wastewater, including cooling water, wash water, and scrubber water.
Iron Blast Furnace Slag. In 1988, iron blast furnace slag was reported as generated at 26 of the 28 ferrous
metal production facilities in the United States surveyed by the U.S. Environmental Protection Agency in 1989 -- all
24 integrated iron/steel facilities and two additional blast furnace operations.
Blast furnace slag contains oxides of silicon, aluminum, calcium, and magnesium, along with other trace
elements. There are three types of blast furnace slag: air-cooled, granulated, and expanded. Air cooled slag
comprises approximately ninety percent of all blast furnace slag produced. The physical characteristics of the slags
are in large part determined by the methods used to cool the molten slag. In the surveys, all facilities characterized
their slags as solid, though slag is molten at the point of generation.20
The primary management practice for iron blast furnace slag is processing (e.g., crushing, sizing) and sale
for use as aggregate. In 1990, only one facility disposed its slag in an adjacent water body in order to build up a land
area that was intended for use managing other waste materials as part of an Army Corp of Engineers approved fill
project.21
19
U.S. Environmental Protection Agency, 1988, Op. Cit., p. 3-128.
20
U.S. Environmental Protection Agency, "Chapter 8," from Report to Congress on Special Wastes from Mineral
Processing, Vol II, Office of Solid Waste, July 1990.
21
Ibid.
Iron Blast Furnace Air Pollution Control (APC) Dust/Sludge. In 1988, iron blast furnace APC
dust/sludge was generated at 26 of the 28 ferrous metal facilities in the United States submitting surveys, including
all 24 integrated iron/steel facilities and the two additional blast furnace operations.
Air pollution control (APC) devices treat the top gases emitted from iron blast furnaces. The air pollution
control devices generate either dusts or sludges. APC dust/sludge is composed primarily of iron, calcium, silicon,
magnesium, manganese, and aluminum.22
The two primary waste management practices at the iron facilities regarding APC dust/sludge are disposal
in on-site units and the return of the material to the production process via the sinter plant operation or blast
furnace.23
Steel Furnace Slag. In 1988, steel furnace slag was generated at 26 of the 28 ferrous facilities in the
United States that submitted surveys, including all 24 integrated iron/steel facilities and the two additional steel
production operations. Steel slag is composed primarily of calcium silicates and ferrites combined with fused oxides
of iron, aluminum, manganese, calcium, and magnesium. At the point of generation, the slag is in a molten form.
The molten slag is air cooled and is broken into varying sizes once processing (e.g., crushing) begins.24
The primary management practice for steel slag is processing (e.g., granulating, crushing, sizing) and sale
for use as aggregate, though several facilities dispose or stockpile their steel slag.
Steel Furnace Air Pollution Control (APC) Dust/Sludge. Steel furnace APC dust/sludge was generated
at 26 of the 28 domestic ferrous metal production facilities surveyed in 1989, including all 24 integrated iron/steel
facilities and the two additional steel production facilities. Steel APC dust/sludge consists mostly of iron, with
smaller amounts of silicon, calcium, and other metals.
Waste management practices were reported for only ten of the 26 facilities in 1989. Eight of the ten
reportedly dispose the APC dust/sludge on-site; the remaining two return the material to the production process via
the sinter plant operation.
Wastewater. Wastewater is generated from a number of sources during both the ironmaking and the
steelmaking processes. In addition to process wastewaters, wastewater streams also are generated from non-contact
operations (i.e., cooling tower water, cooling tower blowdown) and from non-process operations including
maintenance and utility requirements. However, the primary source of wastewater from ironmaking is water used
for the cleaning and cooling of gases. Most plants either recirculate or recycle their cooling process wastewater to
reduce the total pollutant load discharged from their facilities. The wastewaters from the blast furnace process
contain suspended particulate matter and cyanide, phenol, and ammonia. All of these pollutants are limited by
NPDES permit requirements. Other wastewaters contain toxic metals (predominantly zinc) and organic pollutants
which come from the raw materials or form during the reduction process.
Many of the pollutants in the process wastewaters are the result of compounds found in the charges and
fluxes added to the furnace. In both iron and ferromanganese blast furnaces operations, ammonia is present in the
exit gases and as a result is also present in the process wastewater. The ammonia is formed from the various
nitrogen compounds that are removed from the coke charge during blast furnace operations. Fluoride is also present
in the wastewater as a result of fluoride compounds, primarily calcium chloride from the limestone flux. Manganese
is present in wastewaters from ferromanganese production and other elements may be present depending on the
various ores and alloys used in production. Lastly, cyanide is generated as a result of the reaction of nitrogen, in the
blast air, with carbon from the coke charge in the reducing atmosphere of the blast furnace.
22
Ibid.
23
Ibid.
24
Ibid.
Existing data and engineering judgment suggest that this material does not exhibit any characteristics of
hazardous waste. Therefore, the Agency did not evaluate this material further.
Wastes associated with the coke making process, stainless steel production, and steel finishings are
considered to be non-uniquely associated. In addition, ancillary hazardous wastes may be generated at on-site
laboratories, and may include used chemicals and liquid samples. Other hazardous wastes may include spent
solvents (e.g., petroleum naphtha), acidic tank cleaning wastes, and polychlorinated biphenyls from electrical
transformers and capacitors. Non-hazardous wastes may include tires from trucks and large machinery, sanitary
sewage, and some waste oil and other lubricants. Other ancillary wastes associated with the coke making process,
stainless steel production, and the spent pickling liquors resulting from steel finishing at some integrated steel mills
are currently classified as listed and/or characteristic wastes and regulated under RCRA Subtitle C requirements.
American Iron and Steel Institute, "Annual Statistical Report," 1984, p. 78.
Association of Iron and Steel Engineers. The Making, Shaping and Treating of Steel. 1985. pp. 539-652.
Astier, J. "Present Status of Direct Reduction and Smelting Reduction." From Steel Times. October 1992. pp. 453-
458.
Ek, R.B. and Schlobohm, J.E. "Glassification of Electric Arc Furnace Dust." From Iron and Steel Engineer. April
1993. pp. 82-84.
Houck, Gerald. "Iron and Steel." From Mineral Commodity Summaries. U.S. Bureau of Mines. January 1995. pp.
86-87
Houck, Gerald. "Iron and Steel." From Mineral Facts and Problems. U.S. Bureau of Mines. 1985. p. 412.
U.S. Environmental Protection Agency. "Iron and Steel." From 1988 Final Draft Summary Report of Mineral
Industry Processing Wastes. Office of Solid Waste. 1988. pp. 3-125-3-145.
Houck, Gerald. "Iron and Steel." From Minerals Yearbook Volume 1. Metals and Minerals. U.S. Bureau of Mines.
1992. pp. 642-659.
Kirk, William S. "Iron Ore." From Minerals Yearbook Volume 1. Metals and Minerals, 1992, p. 618-641.
Kokal Harold R., and Ranade, Madhu G. "Fluxes for Metallurgy." From Industrial Minerals and Rocks. 1994. pp.
668-669.
McIntosh, S.N. and Baglin, E.G. "Recovery of Manganese from Steel Plant Slag by Carbamate Leaching." U.S.
Bureau of Mines. 1992.
Nyirenda R.L., et al. "Dezincing and Detoxification of Electric Arc Furnace Steelmaking Dust via Ammonium
Carbonate Leaching." The Minerals, Metals, & Mining Society. 1993. pp. 894-906.
Schottman, Frederick J. "Iron an Steel." From Minerals Yearbook Volume I. Metals and Minerals. U.S. Bureau of
Mines. 1989. p. 511.
U.S. Environmental Protection Agency. "Chapter 8." Report to Congress on Special Wastes from Mineral
Processing. Vol. II. July 1990.
Watson John L. and Mishra, Suren. "The Treatment of Steel Plant Wastes by Magnetic Cyclones." Conference
Paper from Symposium on Emerging Process Technologies for a Cleaner Environment. Phoenix, AZ.
February 24-27 1992.