-------
gases coming from the furnace. Although furnace 15 has electrode seals,
it is classified as an open furnace because the electrodes are sealed to an
open canopy hood, which extends down to the top of the furnace and limits
the amount of combustion.
In Japan the most commonly used air pollution control equipment consists
of venturi scrubbers, pressurized fabric filters, electrostatic precipitators,
Theisen centrifugal scrubbers or various combinations of these devices.
Fabric filter collectors, in conjunction with furnace and/or ladle hooding,
are used in some installations to control furnace tapping fumes. When
observed by EPA personnel, the furnaces listed in Table VI-18 had no visible
emissions. Reported emission rates were obtained by use of the Japan
Industrial Standard test method. Concentrations of emissions to the atmos-
phere from air pollution control systems for cleaning the collected furnace
gases varied from 0.001 to 0.035 gr/scf while mass emissions varied from
0.002 to 1.5 'Ib/mw-hr.
VI-54
-------
VII. DESCRIPTION OF CONTROL SYSTEMS
For 40 years various types of participate emission control systems
have been used on a limited basis by the ferroalloy industry to reduce
the high levels of suspended particulate matter often existing near
plant sites and causing ambient air problems. During this time the
industry has sought an economical means for controlling particulates and
has tried systems using scrubbers, filters, and electrostatic precipitators,
Over-the years, four different control systems have been used by the
industry:
1. Scrubbers serving open furnaces,
2. Cloth filters serving open furnaces,
3. Scrubbers serving covered furnaces, and
4. Electrostatic precipitators serving open furnaces.
A. SCRUBBERS SERVING OPEN FURNACES
The most prevalent type of wet collector used for cleaning the large
gas volumes from open furnaces is the high-energy venturi scrubber. Of
the several designs of venturi scrubbers, the one now generally used in
the United States is the flooded-disc scrubber (see Figure VII-1). The
adjustable disc can be raised or lowered to decrease or increase scrubber
efficiency. A two-phase jet scrubber system has also been developed and
used to reduce emissions from several furnaces.
Because large quantities of submicron particulates are emitted from
ferroalloy furnaces, venturi scrubbers require pressure drops of about
VII-1
-------
VI [-2
-------
60 inches water gauge to obtain a removal efficiency of 96 to 99 percent.
The venturi scrubber is capable of absorbing gas temperature peaks
by evaporating more water. Multiple units may be required if water
distribution across the throat is inadequate. Improved performance can be
obtained by increasing either the scrubbing water or the gas velocities.
One of the greatest advantages of scrubbers over other types of collectors
is this operational flexibility.
However, scrubbers also have some disadvantages. Steam plumes are
usually noticeable in the exhaust gases entering the atmosphere because
the hot furnace gas evaporates large quantities of scrubber water and the
moisture is often condensed in the cooler atmosphere. Large quantities
of scrubber water are necessary, and to minimize water usage it may become
necessary to recirculate dust-laden water to the cooling-spray zones of the
scrubber system. When this happens, previously captured dust in the re-
cycled water is released into the gas stream, and scrubber efficiency is
27
thus reduced. Another disadvantage of scrubbers is the potential trans-
formation of an air pollution problem into a water pollution problem. The
power required to operate these high-energy scrubbers is equivalent to about
10 percent of the total power supplied to the furnace.
Because most venturi scrubbers recirculate the scrubbing liquor, water
usage is held to that exhausted in the gas stream plus that bled off with
the collected solids. The scrubbing liquor containing the collected solids
is clarified in thickeners or settling ponds.
A new type of dust-removal system uses waste heat from the furnace
to provide the energy for gas scrubbing without the use of exhaust fans.
This system is shown in Figure VII-2. At one plant, four such units have
VII-3
-------
VII-4
-------
recently been installed to serve four ferroalloy furnaces. 'In brief, the
combusted reaction gases pass through a heat exchanger, a nozzle, and a
separator before the cleaned gas is discharged to the atmosphere. To
describe the process in more detail, heat from the reaction gases is
transferred to the water in the heat exchanger, increasing the temperature
of the water to 350 or 400 F and increasing water pressure to about
300 pounds per square inch (psi). As the heated water is expanded through
the nozzle of the scrubber, partial flashing occurs, and the remaining
liquid is atomized. Thus, a two-phase mixture of steam and small droplets
leaves the nozzle at high velocity. At the same time, reaction gas from
the furnace is aspirated by this high-velocity two-phase mixture, and
in the subsequent mixing, the high-velocity water entrains the particulates
that had been contained in the reaction gas. The mixture of steam, gas,
and dust-laden water droplets then passes through a separator where the
water and dust are separated from the gas-steam mixture. Cleaned gas
leaves the separator through the stack and dirty water is discharged from
the separator to a waste-water treatment system. Chemicals and other
treatment are applied to settle solids and other contaminants from the
water, and fluid waste is discharged to settling ponds. The water is then
deionized, filtered, and returned to a pump for recycling to the heat
exchanger. Makeup water is added to replace any losses.
B. CLOTH FILTERS SERVING OPEN FURNACES
In a cloth filter system, dust and condensed fumes are retained on
the "dirty gas" side of the filter, and cleaned gas passes through the
VII-5
-------
filter to the "clean gas" side. Collected dust particles are then
removed from the fabric by pneumatic or mechanical devices. Fabrics
R R R
used are cotton, wool, nylon, Orion , Dacron , Nomex , acetates, fiber-glass,
etc. Synthetic fibers are either produced as a continuous filament, then spun
and woven into yarn in the usual manner, or they are cut into short
lengths or "staple," which may be spun, woven, or impacted into a felt.
Woven fabrics are identified by thread count and by weight of fabric
per unit area. Felts are identified in terms of thickness and weight
per unit area.
The ability of the filter medium to pass clean air is stated as
"permeability," which is the volume of clean air in cubic feet per minute
that is passed through 1 square foot of the filter at a pressure differential
of 0.5 inches water gauge. The amount of furnace gas that a fiber-glass cloth
filter can handle without blinding is a maximum of about 2 actual cubic feet
per minute (acfm) per square foot of filter area. This low air-to-cloth
ratio can result in the use of thousands of bags so that baghouses serving
ferroalloy furnaces require large areas. A baghouse also has many moving
parts, because the bags must be flexed or shaken to discharge the dust.
Conveying the collected dust from the baghouse hopper to a dust storage
bin requires several enclosed screw conveyors and an enclosed elevator.
The baghouse in Figure VII-3 is typical of those used in the ferroalloy
industry. The walkway access area around the bottom part of the compartment
is shown as being solid, but these walkways are normally grated to allow
outside cooling air to enter and mix with the hot filtered gases at
approximately a 1-to-l ratio. The resultant cooling effect within the
VII-6
-------
FILTERED GAS
FABRIC
CLEANING
MECHANiSM
SERVICE
WALKWAY
DUSTUBES
DIRTY GAS
REVERSE FLOW
Figure VII-3. Typical baghouse. (Courtesy of
Wheelabrator-Frye, Inc.)
VII-7
-------
baghouse compartment makes it easier for personnel to replace bags
during operation.
Unequal bag life necessitates frequent bag replacement. Baghouses
in the ferroalloy industry are designed with compartments so that one
compartment can be shut down for maintenance while other compartments
continue to operate. Means for easy access to the bags are also included
in the original design. In most cases, open, pressure-type baghouses
are used; the fan is located on the dirty gas side of this kind of
baghouse. Bag replacement is facilitated with this type of baghouse
because a leaky bag is easier to locate when the gas flows-from the inside
of the filter bags.
Gas temperatures are limited to about 500°F for treated fiberglass.
Gases from the furnace must often be cooled-by heat transfer surfaces or
by air dilution before entering the filter media. Figure VII-4 illustrates
one type of cooling system. In some cases, arresters are used to prevent
overheated particles from reaching the fabric. Cooling the gas by water
spray is possible, but requires a reliable spray control system to prevent
condensation of moisture on the fabric filters and subsequent blinding.
Cloth filters with air-to-cloth ratios ranging from 1.5 to 2 acfm
per square-foot of cloth have been installed on 12 large ferroalloy
furnaces in recent years. Because of the high percentage of submicron
particulates and the high electrostatic charge, pressure drops across the
filters are high, ranging from 10 to 18 inches of water. However, the
pressure drops for cloth filters are lower than those for venturi scrubbers
VII-8
-------
VII-9
-------
with equivalent efficiencies and gas flows; consequently, the power
required for the cloth filter exhaust system is less.
C. SCRUBBERS SERVING COVERED FURNACES
A covered ferroalloy furnace has a water-cooled cover that seals the
top of the furnace, including the electrodes, mix spouts, and access
openings. This seal prevents the induction of ambient air that would
otherwise burn the gases coming from the reduction process. The dust-laden
furnace gas is withdrawn from under the cover, cleaned, and either used
as fuel or flared above the furnace building. The quantity of gas that
needs cleaning from a covered furnace can be only 3 to 5 percent of that
from an open furnace.
Two types of covered ferroalloy furnaces are currently in operation.
Developed in the 1930's, the initial version of the covered ferroalloy
furnace has mix seals at the electrodes and is generally called a semi-covered
or semi-enclosed furnace (see Figure VII-5). A later version is essentially
the same as the earlier one except that tight or fixed seals are used in
place of mix seals at the electrodes. This configuration is called a totally
enclosed furnace (see Figure VII-6). However, mix seals are maintained
within the chutes at the cover of the totally enclosed furnace by choke-
feeding the material.
With a semi-enclosed furnace, the mix is charged to the furnace through
the annul us around each electrode, and an air gap is established between the
furnace cover and the mix chute to prevent an electrical current flow. If
enough mix is added to keep this space filled, it acts as a seal that
prevents or limits the gases under the cover from escaping through the mix
VII-10
-------
MIX SEAL
FORMED BY
RAW MATERIAL
FEED AROUND
ELECTRODES
COVER
CLEANED
GAS
TO FLARE
I SLURRY TO
^"THICKENER
Figure VI1-5. Covered furnace with mix seals.
ELECTRODES
FIXED
SEALS
MIX
FEED
CLEANED
GAS
TO FLARE
COVER
TAP i
HOLE ' \
SLURRY TO
'THICKENER
Figure VII-6. Covered furnace with fixed seals.
VII-11
-------
around the electrodes; hence the term, "mix seals." To minimize emissions
at the mix seals, the air pollution control system should be designed with
enough flexibility to maintain a slight, negative pressure under the cover.
The degree to which furnace emissions escape from the mix seals may also be
influenced by the quality of charge materials and by the operating condition
of the furnace. Particulate losses from mix seals are reported to range from
2 to 12 percent of the dust and fumes generated in the furnace. Semi-enclosed
furnaces are used in the United States to produce calcium carbide, ferrosilicon
containing 50, 65, and 75 percent silicon, ferromanganese, silicomanganese,
HC ferrochrome, and ferrochrome-silicon.
In a totally enclosed furnace, seals are fixed insulators around the
electrodes and cover which allows the air pollution control system to collect
essentially all of the dust and fumes. These furnaces are used at several
foreign installations to produce calcium carbide, ferromanganese, silicomanganese.
ferrochrome, and several grades of ferrosilicon to a limited extent.
No silicon metal or ferrosilicon alloys with 80 to 90 percent silicon
are produced in either semi-enclosed or totally enclosed furnaces due to the
hot furnace conditions and complications in stoking the furnace.
The quality and size of the raw materials used in the feed mixture have
a major influence on the operation of a covered ferroalloy furnace which may
adversely affect the gas scrubbing system. Favorable operation requires
the use of charge materials with porsity and non-fusing properties that will
permit the uniform release of gases from the reduction process and free
flow of feed materials into the super-heated zone at the electrode tips.
Bridging of the mixture through fusion may cause gas pockets to form within
vn-12
-------
the furnace burden. Upon their collapse, heavy or violent hot-gas blows
may suddenly occur which cannot be handled by the gas removal system.
Stoking breaks up fused material and controls gas blows. Occasionally, the
power may be turned off a semi-enclosed furnace for a short period and the
cover doors opened to permit stoking. Totally enclosed furnaces in foreign
installations have been found to be rarely if ever stoked. The frequency
of stoking in a covered furnace is dependent upon the furnace operating
conditions, the type of product, the quality of material, feed preparation,
and furnace design.
Nine tests by EPA show that total particulate losses from two semi-
enclosed furnaces equipped with scrubbers are higher than the losses from
five well-controlled open furnaces and are-considerably higher than those
from two totally enclosed furnaces equipped with scrubbers. The particulates
in the fugitive fumes from the mix seals of the two semi-enclosed furnaces
tested averaged 342 Ibs/hr and 58 Ib/hr respectively. The weight of parti-
culates represented a computed loss of 8.7 percent and 6.7 percent of the
total particulates generated in open furnaces of the same size and products
based on the emission factors, Table VI-3. A reduction in the fugitive fumes
from the two semi-enclosed furnaces tested is required to meet the control
equipment performance measured in the other seven EPA tests. Emission reduction
may be accomplished by one or more of several methods; improved furnace
operation, more advance furnace design, or the use of secondary emission
control.
In a covered furnace design using a shaft kiln, 80-90 percent of the
r
dust containing 32.5 percent of MnO from a HC ferromanganese furnace is re-
ported to be retained in the furnace feed mixture and returned to the
oc
furnace. As illustrated in Figure VI1-7, dust is collected by impaction
VII-13
-------
DUST MOSTLY
RETAINED BY
FEED MATERIALS
FURNACE
FEED
MATERIALS
PARTIALLY
CLEANED
GAS
FEED CONTROL
ELECTRODE-
Figure VII-7. Shaft kiln on HC ferromanganese furnace.
VII-14
-------
&
7
1
o
o
CD
O)
O
cu
CD
"o
(0
0)
o
O
CO
-------
on the feed as the gas stream is exhausted from the furnace. The benefits
of this system have not been verified.
In the United States, gases from semi-enclosed furnaces are often
cleaned in a centrifugal scrubber consisting of a multi-stage centrifugal
fan equipped with water spray nozzles. The two pictures in Figure VII-8
show the exterior and interior views of a typical centrifugal scrubber
used in the ferroalloy industry on several semi-enclosed furnaces. Uncleaned
gases enter on the left and after passing through four cleaning states,
the cleaned gas is discharged from the "inlet" of the wheel on the right.
Although these and other scrubbers generally remove most of the particulates,
they do not remove all the tars. Usually, the cleaned gas is flared, and
as a result, the tarry material is burned. Absence of visible emissions
in the low volume of flared gas indicates that no significant discharge
of particulates occurs. However, centrifugal scrubbers are limited to a
capacity of about 2800 acfm, which is comparable to gas flowrates from a
medium-sized covered furnace. The volumes of gases from larger covered
furnaces can be controlled by the use of two or more centrifugal scrubbers.
Power and water requirements of a centrifugal scrubber are generally greater
than those of a venturi scrubber. A centrifugal gas scrubber when used on a
covered furnace that produces calcium carbide and other alloys compatible
with the scrubber's limitations, has a particulate removal efficiency of
up to 99 percent. Venturi scrubbers currently serving covered furnaces
and operating at pressure drops up to 80 inches water have been found to
be somewhat more efficient than centrifugal scrubbers.
Scrubbers used to clean the gas from a covered furnace collect the dust
in the water. Consequently, treatment facilities are required to remove
sludge from the waste water before it is discharged into public waterways.
VII-16
-------
Cleaned gas from covered ferroalloy or calcium carbide furnaces has
significant value as fuel if the gas can be used within a reasonable
distance of the furnace. Fuel value of the gas, based on the cost of an
equivalent amount of coal, may exceed $100,000 per year for a moderate-sized
ferroalloy furnace.
Retrofitting an open furnace to a covered furnace generally cannot be
done without completely rebuilding it to include additional head room, which
provides space for the cover itself. In addition, head room is required so
that the electrode suspension mechanism may be raised to provide working space
over the cover. The mixture supply system also requires increased head room
since the mixture flows by gravity into the furnace, and the chutes need to
be clear of the electrodes. Usually, self-baking electrodes are used for
covered furnaces, and ample head room is needed to provide a working area
for installing the metal electrode shells and an overhead crane for handling
electrode materials.
D. ELECTROSTATIC PRECIPITATORS SERVING OPEN FURNACES
Theoretically, the electrostatic precipitator has the lowest pressure
drop of any large-volume device capable of removing micron-size particles
from gas streams. As a result, precipitators usually have lower power and
operating costs than other devices of comparable efficiency. Precipitators
are also able to operate at higher temperatures than fabric filters. Figure
VII-9 shows a cutaway view of a typical high-voltage electrostatic precipitator.
In the precipitation process, particles suspended in the gas are
electrically charged and passed through an electric field where electric
forces move the particles toward the collection surface. The particles
are retained on the collection electrode and subsequently removed from the
VII-17
-------
o
<
O
o
O
o
h-
u
r)
L/>
Z
± _ a:
^00
LL
> > a
o
-
00
_
3 13
O) O
,70
VII-18
-------
predpitator. Various physical configurations are used to accomplish
these basic functions of charging, collection, and removal, depending
upon the type of application and properties of the dust and gas.
Electrical resistivity of the dust is an important factor in the
performance of electrostatic precipitators in the ferroalloy industry.
If the resistivity of the collected dust is high, excessive sparking or
reverse corona can occur, thereby limiting precipitator performance.
High conductivity of the dust is another property necessary for
satisfactory dust removal in a precipitator. Two distinct types of
electrical conduction occur. One type is conduction by free electrons
within the dust particles. This type of conduction depends upon the
electron activation energy, which is a material property varying with
temperature. Most of the particulates in the gas streams from the
ferroalloy furnaces are composed of metallic oxides and have low
activation energies. The electrical conductivity of these particulates
is low at temperatures of 300 to 400°F, but is improved substantially
when temperatures are between 450 and 500°F.
The second type of conduction occurs over the particle surfaces
because of the adsorption of moisture or of certain chemicals such as
ammonia. Adsorption increases with decreasing temperature; hence,
particle conductivity also increases with decreasing temperature. Moisture
is often referred to as the primary conditioning agent, and other chemical
adsorbates are called secondary conditioning agents.
Unfortunately, most ferroalloy furnace fumes at temperatures below
500"F have too high an electical resistivity for satisfactory precipitator
operation. Resistivity is in an acceptable range on-ly if the gas temperature
is maintained above 500"F. The alternative to operating at high temperatures
VII-19
-------
is to humidify the furnace gases along with adding a secondary conditioning
agent, like ammonia. Humidification of the furnace gas by water sprays
requires good atomization and sufficient residence time and heat to obtain
vaporization. Thus, a conditioning tower physically larger than the
precipitator may be required (see Figure VII-10). Stainless steel construc-
tion would be required for the conditioning tower and interior surfaces of
the precipitator in order to control corrosion. The conditioning tower
performs like a scrubber and in actuality removes 20 to 30 percent of the
particulates from the gas stream. This system thus requires a waste-water
disposal system as well.
Only two modern precipitators are in operation on ferroalloy furnaces
in the United States. These are on open furnaces. The dust-removal
efficiency of these precipitators on open furnaces in the manufacture of
chrome alloys under optimal conditions can be expected to be about 98 percent.
E. WASTE-WATER TREATMENT
Large quantities of water are used in the operation of both the
ferroalloy furnaces and the wet-type air pollution control devices
(scrubbers and electrostatic precipitators).
Furnace cooling services require by far the largest portion of'the
water used in ferroalloy manufacturing processes. From 700 to 5,000 gallons
per minute may be needed to cool the furnace and certain components of the
electrical conductors. Additional water is, of course, required for
wet-type air pollution control devices, and approximately one-third of the
furnaces in the ferroalloy industry use such devices. For the year 1968,
the U.S. Census Bureau tabulated water intake for the electrometallurgical
industry (20 out of 34 establishments reporting) as 298.2 billion gallons
VII-20
-------
4_
L
I
••-,
1 — =1
. =r -~ n
3 t.
o
c +->
o
O) oo
s o
o s_
+-> -i-j
o
>>
-------
with gross water usage of 320.3 billion gallons. Since some plants are
equipped with their own steam electric generators, 185.3 billion gallons
of this total were used for this purpose. Total water discharged was
296.1 billion gallons.
Water quantities needed for furnace cooling and for scrubbers range
widely from 3,000 to 10,000 gallons per megawatt-hour and from 500 to
3,500 gallons per megawatt-hour, respectively. Since each ferroalloy
plant may differ in its water needs, typical water requirements are
difficult to establish. Water use may range from 0.5 million gallons
per day for a small plant up to 100 million gallons per day for a large
plant with steam electric generators.
Treatment facilities for the scrubber water differ, depending on
the product being made and the type of scrubber system used. Water
pollutants from a ferroalloy plant may include one or more of ihe following:
suspended insoluble metal compounds, soluble metal compounds, cyanides, acid or
basic effluents, tars, and thermal pollution. Chemical and physical treatment of
the waste streams is usually sufficient; biological treatment methods are not
normally considered necessary. Scrubber water is always clarified to remove the
dust scrubbed from the ferroalloy furnace fumes. The solids consist of
burden fines ranging from less than 1 micron up to 10 microns. The slurry
bleed-off rate from a scrubber ranges from 200 to 500 gallons per minute.
Concentration of suspended solids in the slurry bleed-off varies considerably
not only from plant to plant but also from hour to hour from the same
ferroalloy furnace; a typical range is 3,000 to 17,000 ppm. A well-designed
clarifier can reduce the concentration of suspended solids to less than
OQ
50 ppm.
VII-22
-------
In some plants, thickened sludge from the clarifier is dewatered by
vacuum filters. In some foreign installations,
-------
CHLORINE
COMPOUNDS
POLYELECTROLYTES(FLOCCULENTS)
! CYANIDE" "!
REMOVAL
CYANIDE
|_ REMOVAL j
EFFLUENT ~\ ]
SCRUBBER WATER ! j
SUMP OR
LIFT
STATION
1
RAK
CLA
-LIME
•pH ADJUSTMENT
CLARIFIED
EFFLUENT
—••JBH-Mi^.
5% SOLIDS
PONDS OR DRYING BEDS
FURNACE COOLING WATER.
RECYCLE
STREAM
Figure VI1-11. Flow diagram of typical waste-water treatment facility.
VII-24
-------
of the Act requires the achievement by July 1, 1977, of effluent
limitations which require application of the "best practicable control
technology currently available," and the achievement by July 1, 1983, of
effluent limitations which require application of the "best available
technology economically achievable." In addition to setting effluent standards
for existing point sources, EPA also sets standards for new point sources.
F. SOLID WASTE DISPOSAL
The application of more highly efficient air pollution and waste-water
treatment facilities in the industry intensifies the solid waste problem.
This means that wet and dry air pollution control equipment that is only
95 percent efficient will potentially collect 342,000 tons per year of
solid waste. The industry also disposes of slag when it cannot be used
in other processes. The industry produces about 450,000 tons per year of slag.
Slag from some operations is crushed and used as road material. Also,
waste slag from ferroalloy manufacture has been used to construct docks
and reclaim land.
Presently very little use is made of collected ferroalloy particulates.
Most of the collected dust is deposited in refuse lagoons and landfill areas.
In the dry state the collected material, which contains considerable
quantities of submicron fumes, can easily become reentrained when transported
to open dumps. Therefore, dust collected in the dry state should be mixed
with water or pelletized before disposal.
VII-25
-------
VIII. EMISSION CONTROL GUIDELINES
A. FIELD SURVEILLANCE GUIDELINES FOR AIR POLLUTION CONTROL OFFICIALS
1. Typical Emission Control Regulations Pertaining to the
Ferroalloy Industry
Particulate matter is the principal air contaminant emitted by
ferroalloy plants. These particulates are primarily metallic oxides and
originate mainly from the electric smelting process. Unless adequately
controlled, ferroalloy plant emissions are usually noticeable as a cloud
mass formed from several individual plumes and, depending upon weather
conditions, can prevail downwind for several miles. Most of the partic-
ulates are submicron in size and produce extensive light scattering;
consequently, low particulate concentrations are necessary before the
furnace fumes become invisible.
Most state regulations pertaining to allowed particulate emissions
are based on the weight of materials introduced into a specific process.
Table VIII-1 summarizes the general regulations for the states where most
ferroalloy plants are located. The regulations express allowable emissions
in pounds of particulate matter that can be emitted per hour as a function
of the total pounds per hour of raw material process weight rate. Included
in this table is the percent opacity allowed by those states that have
opacity restrictions.
VIII-1
-------
I/}
QJ a
r- O
-Q r-
'* £
> 'i
OJ
QJ"
S-
.G
0)
* §•
2^
e
a.
^
<_>
re
a.
°
&$
i
o
o
o
8
1
O
o
o
8
c?
o
o
o
o
S
8
0
o
o
o
o
o
o
o
LO
o
o
o
o
o
cu
tJ
CO
CM
en
LO
£
LQ
m
ro
OO
ro
CO
OJ
1
°1
<*•
£
en
ro
n
•SI-
rO
CM
UP
LO
O
"i
.a
(O
o
CT*
LO
OJ
LO
ro
vo
O
O
LO
O
CO
0
CM
cn
i—
8
CM
CO
CO
OJ
LO
o
(0
1
o
o>
CM
LO
ro
«i
o
o
LO
S
o
CM
O1
i
O
O
CM
OO
CO
LO
OJ
LO
o
o
I
1
^ oi*
o •— o o o
<3- i— O~i C7i CTi
CM
r-^ CM oj oJ
Si i — i— i—
u~> ID in
r^ ro ro oo
00 ' LO V£> LO'
u
O CO O O O
0 r-^ 0 0 0
LO LO LO ro
01 ro ro CM
~o
CM CM CM < —
1 ...
§OOO
CO O O O
OJ O OJ CM O
co co co r^-.
OO CO CO O
LO CO LO LO CO
OJ OJ OJ OJ OJ
LO LO Lf) VO
O O O 0 O
-o
rO
a. c
O- ra
to "^ E ^~
i/i O o c: >,
r- S- .C O (rt
irt o to a> c:
t/> 2 -r- i— CU C
•r- OJ J= ^ t. QJ
s: sr o o o a.
QJ OJ OJ
OJ OJ OJ
O P»> CJi O
O% LO OO O"i
,_,
CM i — oj i—
r— VO LO to
LD m a> co
CO
LO CO i ro
C3 VO •* *JD
o en r-. a\
*3- CM LO OJ
LO o r*H
o ro 01 i
m CM LO
en
OJ O r— Ol
2 ^ S5 ^
O O CO
O r-x CM i^.
CM O\ Ln O%
LO oo r*-w ro
OO «*• LO
LO CO VO CM
CM OJ r- CM
LO LO LO
LO LO LO
O O O
C (4- C
•r- CU O
r—
girt CT)
Wl C
fO CU t/1 Tf~
C X (rt
CU (D (13
CO 1— h- 3
S
Cf>
.^
Ol
i-
-*->
1
.? e
" S
5 S
T, Su.
(U O
wi O) •*- II
rtJ S- t/i
j3 d) in x
§•§'£•§
cu cr
<, -r
-------
Since process weights are not easily obtained for electric smelting
furnaces, an indirect approach can be used with reasonable success. This
approach is based on the stoichiometry of ferroalloy smelting, which pro-
vides a relationship between process weight and electrical energy input
(obtainable from control room instruments) required to produce a unit
weight of product. These relationships are expressed as factors that vary
with the product made and are shown in Table VIII-2. This table provides
the range of raw material usage and electrical energy requirements
experienced in furnaces normally used to produce general classes of ferro-
alloys. Average factors are shown for convenience in applying them to
process weight information and furnace capacity in megawatts.
Process weight (in pounds per hour) for each individual product in
Table VIII-2 may be determined by multiplying the average integrated
furnace load in megawatt-hours (obtained from instruments at the furnace)
by the appropriate factor from the table (column 7). Likewise, net tons
of the ferroalloy produced may be determined by multiplying the average
integrated furnace load in megawatt-hours by the listed factor (column 6).
Uncontrolled mass-rate emissions may then be estimated by multiplying
the net tons of ferroalloy produced by the emission factor shown for the
appropriate ferroalloy in Table VI-3.
State emission control regulations for particulates vary to some degree
but are generally based on process weight for each operating furnace or
source. Because they do vary, however, a careful study of the emission
control regulations and a general knowledge of the plant operation,
including furnace loads and products, are needed to determine allowed
VIII-3
-------
Table VIII-2. FACTORS FOR PROCESS WEIGHTS AND FERROALLOY PRODUCTION
RELATED TO FURNACE KILOWATT CAPACITY
1
Product
Silvery
iron
50% FeSi
65-75% FeSi
Silicon
metal
SMZ
CaSi
HC FeMn
SiMn
FeMnSi
Mn ore/lime
melt
Chg Cr
HC FeCr
Cr ore/lime
melt
FeCrSi
Ca carbide
2
Charge
Ib/lb
Range
1.7-1.9
2.3-2.5
4.3-4.5
4.6-5.0
4.3-4.5
3.8-4.0
2.9-3.3
2.7-3.3
4.2-4.4
3.2-3.6
3.7-4.1
3.7-4.1
3.2-3.6
1.5-1.7
3
weight, l
alloy
Approximate
average
1.8
2.5
3.6
4.9
4.5
3.9
3.0
3.1
4.3
3.5
4.0
4.0
1.2
3.4
1.6
4 :
Furnace
KW-hr/lb
Range
1.2-1.4
2.4-2.5
4.2-4.5
6.0-8.0
4.2-4.5
5.7-6.1
1.0-1.2
2.0-2.3
2.4-3.0
0.6-1.0
2.0-2.2
2.0-2.2
0.5-0.7
3.6-3.8
1.3-1.4
5
load ,
alloy
Approximate
average
1.3
2.5
4.4
7.0
4.4
5.9
1.2
2.2
2.7
0.8
2.1
2.1
0.6
3.7
1.3
6
Product,
Ib/Mw-hr
770
400
227
144
227
170
834
454
370
1350
476
476
1670
270
770
7
Charge
weight,
Ib/Mw-hr
1380
1000
1020
700
1020
660
2500
1410
1590
4280
1900
1900
2000
920
1230
VIII-4
-------
losses from each individual plant.
2. Process Description and Sources of Emissions
A general description of the ferroalloy industry has been given in
Chapter V. The operations important to the air pollution control in-
spector are the submerged-arc furnaces, the raw material preparation and
handling system, and the product sizing and handling. Figure VI-1 in
Chapter VI shows a typical flow diagram for ferroalloy production.
Ferroalloys are alloys of iron and some other metal or metals,
such as manganese, chromium and silicon. They are produced by reducing
an ore of the alloying element with carbon in the presence of heat and
iron from scrap steel. A submerged-arc furnace provides the high-
temperature vessel for the carbon reduction process. Furnaces are rated
by electrical energy input and vary in size from about 7,000 to 50,000
kilowatts, depending upon furnace use and age.
Raw materials from open storage piles are conveyed to overhead bins
in the mix house, where charge material is mixed and weighed for each
individual furnace. The lump size of the raw materials varies from
approximately 0.25 to 4 inches. In some plants, part of the mix materials
are dried before they are conveyed to the mix house; in this case, up to
3 percent of the material charged may be lost as particulate unless proper
air pollution control equipment is used. However, under ordinary circum-
stances when no intermediate operations are performed, the dust generated
during conveyance of material to the mix house can be held to a low level
with proper handling.
Raw materials in the mix house are weighed into larry cars or conveyors
according to the mix required for product specifications. The small amount of
VIII-5
-------
dust generated during mix preparation 1s generally confined to the mix
house. The Tarry car next dumps the mix into a skip hoist, which in turn
lifts the contents to an overhead belt conveyor within the furnace build-
ing. If there are several furnaces within the building, the conveying
system is designed to transport the mix to the correct furnace mix bins.
Charge material from the overhead mix bins is normally gravity-fed into
the furnace through several enclosed chutes, although a few open furnaces
charge mix materials with manually operated skip loaders.
The principal source of emissions is the submerged-arc furnace.
Emissions vary widely 1n type and quantity, depending upon the product
being made, the type of furnace used, and the power input to the furnace.
Gas containing large quantities of carbon monoxide, metallic oxide fumes, and
dust is continuously generated; gas flow will vary with furnace operation.
In an open furnace, large gas volumes result when the carbon monoxide
is burned to carbon dioxide at the surface of the charge. In a covered
furnace, the unburned carbon monoxide is withdrawn and may be used as
fuel or flared.
Another source of fumes associated with ferroalloy furnaces Is
the tapping operation, which occurs every 1 to 5 hours, depending on
the product, and lasts for 10 to 15 minutes. After being tapped into a
ladle, the molten material may be transferred by overhead crane to
another ladle for repouring, which results in additional fume emissions.
In some processes, still additional ladle operations involve slag-
metal reactions, chlorination, and oxidation, all of which produce
emissions.
VIII-6
-------
The molten product is poured into chills or molds, during which
time fumes are generated. After solidifying in chills, the product
is either manually sized or mechanically crushed and screened to size
before shipment. The crushing and sizing system (jaw crushers, cone crushers
and screens) generates sufficient dust to require an air pollution control system.
3. Emission Control Systems
Emission control systems are described in Chapters VI and VII. The
major emission problem of the ferroalloy industry concerns the capture
and collection of fumes from electric furnaces. Control sytems are in
use on both open and covered furnaces. Fumes from an open furnace are
collected by a canopy hood located 5 to 6 feet above the furnace and are
ducted through stacks to an emission control device. The combustion
taking place at the surface of the unreacted charge material results in
large quantities of gases.
Most conventional control devices have been used on open furnaces.
Several push-through baghouses are used on open furnaces producing
different products and, in the absence of other suitable devices, have
been found to be particularly adaptable for silicon metal and high-
silicon alloys. High-energy scrubbers are used on furnaces producing
silicomanganese, HC ferrochrome, and ferrochrome-silicon. Two electro-
static precipitators are in service on domestic furnaces producing
HC ferrochrome and ferrochrome-silicon.
Particulate emission rates from the better controlled open furnaces in
the United States range from 1.0 to 1.5 pounds per megawatt-hour. Particulate
emission rates from the most efficient air pollution control systems cleaning
collected gases from semi-enclosed furnaces are much lower, ranging up to 0.1
VIII-7
-------
pound per megawatt-hour. Unless losses from the mix seals are controlled
however, semi-enclosed furnaces generally are not as well controlled as
the better controlled open furnaces.
The covered furnace restricts air ingestion so that the reaction gases,
consisting of a high percentage of carbon monoxide, are mostly unburned. Gas
quantities from a covered furnace are from 2 to 5 percent of those from an
equivalent-size open furnace. Consequently, a control device serving a closed
furnace is small compared to one on an open furnace. High-energy venturi
scrubbers and centrifugal scrubbers are generally used to control emissions
from covered furnaces.
All ferroalloy products cannot presently be produced in covered furnaces.
The semi-enclosed furnace (with mix seals) is used in the United States for
the production of calcium carbide, HC ferromanganese, silicomanganese, 50-
percent ferrosilicon, and several other ferroalloys with relatively low gas
evolution. In general, however, the high-silicon alloys (75 percent silicon and
higher) and silicon metal are not produced in covered furnaces either in the
United States or abroad.
4. Maintenance and Operating Problems
Even the most effective emission control systems in the ferroalloy
industry, whether serving open or covered furnaces, occasionally will
have operating problems during which the allowed losses may be exceeded.
These operating problems may be caused by equipment failures, plugged
ducts, electrical difficulties, or some other type of problem. In some
instances, the proper performance of the control equipment bears a
direct relationship to the manner in which the furnace is operated. For
example, if rough furnace operation suddenly increases the normal
exhaust gas temperature, the performance of an electrostatic precipitator
VIII-8
-------
may be adversely affected.
Ferroalloy furnaces are costly and complicated pieces of heavy equip-
ment that require a well-planned maintenance program. Their operation requires
large quantities of selected feed materials, considerable amounts of electrical
power that is converted into heat, and rapidly circulated cooling water to
prevent heat damage to the equipment. The complexity of the furnaces causes
some operating problems even when the equipment is operated by qualified
personnel and preventive maintenance is performed.
Principal furnace operating problems may involve equipment failures,
furnace tapping problems that may cause the electrodes to be higher than normal,
water leaks, furnace interruptions for equipment repairs followed by startup
periods, mixture feed problems, electrode failures, electrical difficulties,
and electrical power shortages. Although the majority of furnace operating
problems cause interruptions of less than an hour, major furnace operating
difficulties do occur and production may then be interrupted for several
days. When a furnace with a cold hearth is starting up after a long
shutdown, emissions will be heavier than normal for up to a week. In
general, ferroalloy furnaces operate from 90 to 95 percent of the time
on an annual basis.
Baghouses usually contain several thousand cloth bags with a usable
life of about 2 years. Bags are normally replaced as they wear out; replace-
ment may be necessary daily. Sometimes locating the torn bags requires a
prolonged search, frequently under limited light conditions.
Scrubber systems use large exhaust fan motors to drive high-speed
fans that require continuous monitoring because imbalances may occur.
Close control of the scrubber system's water supply is also necessary
to ensure desired performance. Both recirculated water and waste water
VIII-9
-------
require treatment before discharge.
When electrostatic precipitators are used, the gas usually requires
preconditioning before passing through the electrical fields of the pre-
cipitator in order to reduce electrical resistivity of the particulate
matter. Monitoring is necessary to ensure that the gas is correctly
preconditioned. Other reasons the precipitators may not achieve design
efficiency are improper gas flow, inadequate rapping, electrode mis-
alignment, adhesion of collected material to interior surfaces, blocked
hoppers, and failure of the high-voltage electrical supply. Most ferro-
alloy furnaces are capable of producing different products, but the
high resistivity of some fumes may prevent a precipitator from attaining
required collection efficiency.
5. Monitoring Instruments
Because emission monitoring instruments are complex and difficult
to maintain, they are used on only a limited basis in the ferroalloy
industry. Opacity meters, which measure the attenuation of light beams
caused by particulate matter suspended in stack gases, are used to a
limited extent. To have reliable readings these instruments must be
cleaned frequently. Although continuous monitoring instruments
indicating the mass emission rate of particulates are available on the
market, the adequacy of these instruments for use in the ferroalloy
industry has not been fully demonstrated.
Newer furnaces usually have emission control equipment, sample ports
in the exhaust ducts, and platforms and ladders to permit stack sampling.
The furnace kilowatt-hour meter or kilowatt-load chart is used to
determine quantities of electrical energy consumed by furnace operation.
VIII-10
-------
Usually, no other process control instruments can be related to the
furnace production rates or process charge weights. Furnace pro-
duction rates and process weight rates can be computed from the
furnace power consumption, as shown in Table VI-4.
B. PROCEDURES FOR REDUCING EMISSIONS DURING EMERGENCY AIR POLLUTION
EPISODES
Air pollution control measures are promulgated to meet national
ambient air quality standards during normal meteorological conditions.
However, adverse meteorological conditions may cause a buildup of air
pollutants. To avoid a catastrophe in this event, each State is
responsible for establishing emergency episode procedures. These State
procedures will be necessary until emergency provisions of the Clean
Air Act Amendments of 1970 have been fully implemented.
The objective of an emergency episode plan is the immediate reduc-
tion of emissions. Control strategies specify the control actions and
the degree of control required for each source. These measures are
necessarily selective, requiring emergency curtailment of nonessential,
easily controlled sources first and postponing drastic measures until
initial curtailments are obviously insufficient.
Following is an example of an episode criteria plan used as a
guide for conditions justifying the proclamation of an air pollution
30
alert, air pollution warning, or air pollution emergency:
(a) "Air Pollution Forecast": An internal watch by the
Department of Air Pollution Control shall be actuated
by a National Weather Service advisory that Atmospheric
Stagnation Advisory is in effect or the equivalent
local forecast of stagnant atmospheric condition.
VIII-11
-------
(b) "Alert": The Alert level is that concentration of
pollutants at which first stage control actions are
to begin. An Alert will be declared when any one
of the following levels is reached at any monitoring
site:
SOp -- 800 yg/m (0.3 p.p.m.), 24-hour average.
Particulate -- 3.0 COHs or 375 yg/m3, 24-hour average.
SOp and particulate combined -- product of SO, p.p.m.,
24-hour average, and COHs equal to 0.2 or proauct of,
SOp -- yg/m , 24-hour average, and particulate yg/m ,
24-hour average equal to 65 X 10 .
CO -- 17 mg/m (15 p.p.m.), 8-hour average.
Oxidant (0,) -- 200 yg/m (0.1 p.p.m.) — 1-hour average.
NOp — 1130 yg/m3 (0.6 p.p.m.), 1-hour average, 282 yg/m3
(0.15 p.p.m.)) 24-hour average.
and meteorological conditions are such the pollutant con-
centration can be expected to remain at the above levels
for twelve (12) or more hours or increase unless control
actions are taken.
(c) "Warning": The warning level indicates that air quality
is continuing to degrade and that additional control
actions are necessary. A warning will be declared when
any one of the following levels is reached at any
monitoring site:
S02 -- 1,600 yg/m3 (0.6 p.p.m.) 24-hour average.
3
Particulate -- 5.0 COHs or 625 yg/m , 24-hour average.
SOp and particulate combined -- product of SOp p.p.m.,
24-hour average and COHs equal to 0.8 or product of
SOp yg/m , 24-hour average and,particulate yg/m , 24-
hoar average equal to 261 X 10 .
CO — 34 mg/m (30 p.p.m.),8-hour average.
Oxidant (03) -- 800 yg/m (0.4 p.p.m.), 1-hour average.
NO
56
2 — 2,260 yg/m (1.2 p.p.m.) — 1-hour average;
5 yg/m (0.3 p.p.m.), 24-hour average.
and meteorological conditions are such that pollutant
concentrations can be expected to remain at the above
levels for twelve (12) or more hours or increase
unless control actions are taken.
VIII-12
-------
(d) "Emergency": The emergency level indicates that
air quality is continuing to degrade toward a level
of significant harm to the health of persons and that
the most stringent control actions are necessary. An
emergency will be declared when any one of the fol-
lowing levels is reached at any monitoring site:
S02 -- 2,100 yg/m3 (0.8 p.p.m.), 24-hour average.
•3
Particulate -- 7.0 COHs or 875 yg/m , 24-hour average.
S02 and particulate combined -- product of S02 p.p.m.,
24-hour average and COHs equal to 1.2 or product of
S0? vg/m . 24-hour average and pacticulate yg/m ,
24-hour average equal to 393 X 10 .
CO -- 46 mg/m (40 p.p.m.), 8-hour average.
Oxidant (03) — 1,200 yg/m (0.6 p.p.m.), 1-hour average.
2
NO, -- 3,000 yg/m (1.6 p.p.m.), 1-hour average;
750 yg/m (0.4 p.p.m.), 24-hour average.
and meteorological conditions are such that this condition
can be expected to remain at the above levels for twelve
(12) or more hours.
(e) "Termination": Once declared, any status reached by
application of these criteria will remain in effect
until the criteria for that level are no longer met.
At such time, the next lower status will be assumed.
Additional information relative to air pollution emergency episodes
is reported in the Federal Register, August 14, 1971, and the amendment
of October 23, 1971.
A ferroalloy plant can curtail atmospheric pollution during an
episode in a number of ways:
1. A plant with both well-controlled and uncontrolled furnaces
should shut down the uncontrolled furnace(s) first.
2. A plant with no controlled furnaces should consider shutting
down one or more furnaces in preference to an overall reduction
of loads. The worst polluting furnaces should be the first to
VIII-13
-------
be shut down. For example, if there are four furnaces
and a 25-percent reduction in emissions is required,
better emission reduction can be attained by shutting down
one of the furnaces than by cutting the operating load 25
percent on all four furnaces. Also, the furnace with the highest
emission levels to the atmosphere should be the first to shut
down.
3. Curtail or stop the material handling system as much as possible.
4. Curtail or stop the alloy sizing operations.
Furnaces can shut down almost immediately by stopping the electrical
power input. Once this is done, emissions are immediately lowered
substantially and will gradually disappear over a few hours.
VIII-14
-------
IX. ECONOMICS OF EMISSION CONTROL
A. INTRODUCTION
Under the Clean Air Act (1970) each State has the primary responsibility
for assuring air quality within the geographic confines of that State by for-
mulating and adopting, after formal approval by the Administrator of the
Environmental Protection Agency, implementation plans for achievement of
primary and secondary ambient air quality standards. These plans contain
emission regulations and timetables of compliance for industrial sources,
among other provisions, such as monitoring air quality and enforcement, for
the attainment of the ambient air quality standards.
Ferroalloy plants are subject to implementation plans as required under
the Clean Air Act. Air pollution control equipment has been installed in a
few plants for the past several years. This chapter is a study of the better
controlled facilities in the industry. It provides that basis for estimating
the investments and annual expenditures that will be incurred over the next
two to three years to meet emission regulations that will be adopted by
state implementation plans.
To provide a general economic profile of the industry, present ownership,
type, and location of plants will be given. Three major products—ferro-
manganese, ferrochromium, and ferrosilicon--will be analyzed for supply-demand
characteristics, growth, and price movements. Model plants will then be
presented along with their control requirements, associated investment costs,
and annualized costs. Finally, economic repercussions will be discussed in
terms of impact on income and the ability of the firm to pass along the costs.
In order to reduce labor costs and to remain competitive, the ferroalloy
industry in the United States in recent years has constructed very large sub-
merged-arc furnaces with well-developed mechanical equipment for handling
IX-1
-------
raw materials and the finished product. The more efficient furnaces are
expected to be retrofitted with control devices such as baghouses and
high-energy scrubbers. A number of smaller, inefficient furnaces may be
shut down and replaced with larger furnaces (e.g., a single new furnace of
30,000 to 50,000 kw size replacing several smaller existing furnaces).
B. FCONOMIC PROFILE
1. Introduction
The economic scope of the ferroalloy industry is international and
highly complex. Probably no other activity linked to steel making is so
completely subject to the forces of world trade. Normally the ores and
other raw materials, the finished ferroalloys, and even the final steel
products are all international commodities. More than 50 different
alloys and metals, in hundreds of various compositions and sizes, are
produced for use in the manufacture of iron, steel, and nonferrous metals.
Most of the products are classified into three aeneral groups—ferromanganese,
ferrochromium, and ferrosilicon. This sector of the chapter will present
a picture of the industry's structure followed by an analysis of the current
economic situation for each of the three ma.ior product groups. An estimation
of the frequency of new installations will also be made.
2. Industry Structure
The ferroalloy industry is comprised of establishments that reduce
oxidic ores with carbon, for the most part, to obtain the various ferroalloy
products. The sources of carbon are most commonly byproduct coke or
low-volatile coals. The ores are either imported, like manganese or
chromium ores used to produce ferromanganese and ferrochromium, or mined
domestically, like quartz used to produce ferrosilicon. After reduction
IX-2
-------
1n the electric furnace, the product is cast and crushed to meet
consumer specifications.
As of January 1, 1972, there were approximately 44 known plants in
the United States, owned by 24 firms. Table IV-1 outlines plant ownership,
geographic location, general run of products, and types and numbers of
furnaces producing both ferroalloys and calcium carbide. As shown, there
are 145 ferroalloy furnaces and 13 calcium carbide furnaces. This
tabulation should be reasonably accurate, although the companies involved
frequently change their product lines or take furnaces in and out of
production according to demand.
The smaller ferroalloy companies are not diversified and rely
principally on sales of ferroalloys for their income. The larger companies,
however, have other interests, such as the manufacture of industrial gases,
chemicals, and steel, and their ferroalloy sales comprise only a part
of their income. In any case, the cost of controlling emissions to
atmosphere to meet new air pollution standards will have a major effect
on ferroalloy production costs and may force closure of marginal ferroalloy
plants, especially those that have open furnaces with no dust hoods and
ductwork for handling the furnace's gaseous emissions. Furnaces of this
type cannot be effectively equipped with emission control equipment without
completely rebuilding the furnace and extending the building height to
provide the head room required for the dust hood, ducts, and more
mechanized electrode columns. Rebuilding a ferroalloy furnace requires
substantial new capital investment, and currently weak markets for
ferroalloys do not justify such an investment.
IX-3
-------
a. Ferromanganese - Silicomanganese
The Ore - Except for a small quantity of metallurgical oxide nodules
shipped from stocks by The Anaconda Company and made several years ago from
Montana carbonate ore, no manganese ore, concentrates, or nodules have
been produced or shipped in the United States since 1971. Thus the U.S. is
now totally dependent on foreign sources of manganese ore. The effect of
recent monetary fluctuations on ore prices is not known. Principal suppliers
are Africa, 50 percent; Brazil, 30 percent; and India, 5 percent. None
of the countries is noted for its hard currency. Table IX-1 traces
the recent volume and price history of manganese ore supplies.
Under the "Kennedy round" of General Agreement on Tariffs and Trade
(GATT), tariffs were reduced on January 1, 1972, from 0.22 cent per pound
of contained manganese to 0.12 cent per pound. Actually, with only one
exception, no tariffs have been imposed since June 1964 because of con-
gressionally approved suspensions. The exception is a special tariff
of 1.0 cent per pound of contained manganese on ores from the U.S.S.R.
and mainland China, deterring their import.
U.S. consumption of metallurgical-grade ore has been cyclical, depending
upon the domestic market situation for ferromanganese and steel. The
market for ores became progressively weaker during the latter half of the
sixties, and imports generally decreased. Prices followed the downward
trend, with the average value of imports at the foreign port falling from
$34 per gross-weight ton in 1962 to $22 in 1971, a drop of 35 percent.
All manganese ore prices are negotiated because they are dependent, in part,
on the characteristics and quantity of ore offered, delivery terms, and
IX-4
-------
Table IX-1. SOURCES AND VALUES OF ORES CONTAINING .
35 OR MORE PERCENT MANGANESE, 1962-197V
Year
1971
1970
1969
1968
1967
1966
1965
1964
1963
1962
U.S.
production, ,
short tons x 10
b
5
6
11
13
14
29
26
11
25
Imports
(gross weight) ,o
short tons x 10
1914
1735
1960
1828
2059
2554
2575
2064
2093
1970
Value,3
$/ton
22
20
20
25
27
30
43
37
32
34
At foreign port.
1971 production 142 tons.
IX-5
-------
fluctuating shipping rates. Transportation is a major cost item, with
ocean freight rates comprising roughly one-third of the price of imported
ore at eastern seaports. Including domestic rail transportation, the
price of ore now reaching the domestic processing plant is probably about
$30 per gross ton.
The outlook is for price stability in the world markets. At present,
supply about equals demand and prices should halt their decline. Current
and near-term prices are showing some strength. The Bureau of Mines
estimates that nearly stable prices are expected to prevail through the
end of the century. Given current technology and projected growth rates
for steel, by the year 2000 the U.S. may need over 3.5 million tons of
ore per year.
The Product - Over 90 percent of the imported manganese ore is used to
make ferromanganese (75 percent) and other alloys (15 percent). The
consumption of ferromanganese and silicomanganese is tied closely to
the steel industry, which consumes 95 percent of the output. These alloys
are needed principally to counteract the effects of sulfur in cast iron
and steel. They also improve the characteristics of steel during rolling
and add strength and toughness to the finished product. The remaining
5 percent of the output is used in electric dry batteries and in chemicals.
Approximately one-half of the total U.S. ferromanganese output is from
electric submerged-arc furnaces, and ferromanganese production is usually
integrated with silicomanganese production. Annual production of
ferromanganese has fallen every year since 1965. Silicomanganese production
IX-6
-------
peaked in 1968 and has since dropped 42 percent. Table IX-2 presents
these recent trends. However, with the already evident recovery of the
steel industry, the two ferroalloys should display recovery trends in
1972 and 1973.
Over the last few years there has been a slow attrition in the number
of U.S. companies and plants. As steel consumption recovers, it is
difficult to assess how much of the required ferromanganese and silicomanganese
will be provided by domestic production and how much by imports. Several
factors account for increasing foreign competition. Low transportation
costs are available to foreign producers as many of their plants are located
near seaports. A few countries, particularly in northern Europe, have cheap
hydroelectric power. As power costs vary from 15 to 40 percent of the
manufacturing cost (depending on, the product), a cost difference of 4 to
5 mils per kilowatt-hour can amount to $10 to $20 per ton of product, which
offers a considerable advantage to the foreign producer.
Tables IX-3 and IX-4 present recent values of both domestic production
and imports of ferromanganese and silicomanganese over the last several
years. As these tables show, both silicomanganese and ferromanganese
are under considerable price pressure.
b. Calcium Carbide
Calcium carbide, while not related chemically to the ferroalloys, is
considered in this study because it is also manufactured in the electric
submerged-arc furnace. Over 90 percent of the calcium carbide produced
IX-7
-------
Table IX-2. ANNUAL PRODUCTION OF FERROMANGANESE
AND SILICOMANGANESE
(short tons)
Year
1971
1970
1969
1968
1967
1966
1965
Growth
rate
Ferromanganese
759,896
835,463
852,019
879,962
940,927
946,210
1,148,011
-5.5%
Silicomanganese
164,682
193,219
222,877
284,499
245,798
253,134
240,667
-6.1%
IX-8
-------
Table IX-3. SOURCES AND VALUES OF FERROMANGANESE
Year
U.S.
production,3
short tons
Value, b
$/ton
Imports,9
short tons
Val uec
$/ton
1971
1970
1969
1968
1967
1966
1965
759,896
835,463
852,019
879,962
940,927
946,210
1,148,011
168
167
143
159
147
148
146
242,778
290,946
301,956
203,212
216,279
251 ,972
257,339
133
108
105
104
122
117
122
aBureau of Mines, Mineral Yearbook (includes both electric and blast
furnace output and all carbon content grades).
Value of shipments without freight or container cost.
°Value in foreign port of origin.
IX-9
-------
Table IX-4. SOURCES AND VALUES OF SILICOMANGANESE
Year
U.S.
production,3
short tons
Valueb
$/ton
Imports9
short tons
Value0
$/ton
1971
1970
1969
1968
1967
1966
1965
164,682
193,219
222,877
284,499
245,798
253,134
240,667
195
185
162
159
159
146
148
29,928
14,539
32,040
25,412
34,936
35,771
17,491
132
122
no
105
118
117
109
aBureau of Mines, Mineral Yearbook.
Value of shipments without freight or container cost.
cValue in foreign port of origin.
IX-10
-------
domestically comes from Air Reduction Company and Union Carbide Corporation.
Other companies producing this chemical are Midwest Carbide (a subsidiary
of Chemetron Corporation) in Keokuk, Iowa, and Pryor, Oklahoma, and
Pacific Carbide and Alloys in Portland, Oregon.
Table IX-5 presents the recent domestic production, import, and price
history of calcium carbide. The import tariff for non-Communist nations
is $4.20 per short ton. U.S. production has dipped 43 percent from
1,098,000 tons in 1965 to 625,000 tons in 1971, an annual growth rate of
-8.0 percent. Production capacity dropped from 1,195,000 tons in 1955
to 963,000 tons in 1970?1
The greatest use for calcium carbide is in the manufacture of acetylene,
a major chemical building block. However, it is fast losing ground to
acetylene made from petrochemicals.
c. Ferrochromium
The Ore - Commercial grades of chrome ore, a strategic and critical
commodity, are found only in limited areas of the world. North American
deposits are of poor quality and cannot compete economically with foreign
ores. No chromite ore has been mined in the United States since 1961,
when a small amount was produced under the Government Defense Production
Act. The world's largest deposits are found in the Transvaal area in
the Republic of South Africa. Other major ore deposits are located in
Rhodesia, the U.S.S.R., and Turkey. As detailed in Table IX-6, the
United States must import all of its chromite needs. As specified in
Tariff Classification 601.15* no rate of duty is placed upon chrome ore
imports.
IX-11
-------
Table IX-5. SOURCES AND VALUES OF CALCIUM CARBIDE
Year
1971
1970
1969
1968
1967
1966
1965
U.S.
production,
short tons
625,000
791 ,000
856,000
942,000
912,000
1,063,000
1,098,000
Value,
$/ton
90
81
78
94
94
87
89
Imports,
short tons
20,000
18,600
17,900
6,900
8,300
20,200
10,500
Value,
$/ton
75
70
68
70
69
65
70
Current Industrial Reports, Department of Commerce, Bureau of the
Census.
3U.S. Imports for Consumption and General Imports, U.S. Department
of Commerce, Bureau of the Census, FT 246 (Calcium Carbide TSUSA
4181400).
IX-12
-------
CO
LLJ
a:
o
o
3:
CJ
o
CO
(—t
CO
I—I
I-
co
o
to
I
X
.a
re
CO
.^
o
c
to
a
5
c.
o
o
CO
— J
c
o
«r—
a
3
to
O
O
•
CO
_^
LO
O
a.
••^
HH
" C
>> 0 0
to S- •!-> O
r— 4-> O
ttJ CO +J r—
-(-> 3 i-
cu -a o x
•1 — tO
n
CO CO
at c
•i- O O
s- +j o
i— -t-> O
i— to 4-> i—
«=t 3 S-
C -C
•i— CO
to
" C
>, 0 0
to S- +J CD
i— +-> O
ro to 4-> i—
•!-> Z5 U
CU T3 O X
s: c jr
•I- tO
to co
CD C
••- o o
S- -1-5 O
i— -t-> O
i— CO +-> i—
+^
CO O
f~*i C~
CM to
S- \
0 -t^
o
-o
CO CO O
CD 3 i—
C\J I"™
S-
CO
+-> o o
CO C +-> O
O CU O
CM +->+-> i—
S- C S-
0 0 0 X
o -c:
CO
CO
c
« o o
r— +J O
(t3 O
+->+-> I—
0 i~
1 — OX
-C
CO
s_
ro
CU
CO to to
co en en
CO CM CO
en o CM
o ^r •—
co t-^. en
-—"^-r -IJ^- — ~- — •- • —
CM 00 ^J"
i— en o
en oo oo
CO r- tO
O •— i—
^t" ^* CO
to oo in
i— in •<*•
en en to
«*• co co
in o en
o co co
oo o .—
i— o oo
CO CM i—
r^ to en
•=l- O en
vo LO <•
in to ^3-
o o oo
^1- r— O
o en co
r^. to to
cr» en en
en
in
^~
p^.
en
to
to
00
in
in
CO
00
co
CO
^«
oo
,
CM
CO
in
o
CM
to
en
CO
to
to
o
CO
oo
CM
oo
f—
to
*
CM
i —
to
CO
en
CO
o
CO
r—
oo
to
00
to
to
en
CO
^~
r-~
r—
fv.
O
en
^~
00
in
en
h>.
to
CO
en
CO
CM
in
CM
vo
00
oo
in
in
to
en
en
O
in
r-.
00
CM
CM
CO
CO
r«
in
to
m
CO
CO
CM
CM
to
to
oo
CM
to
en
to
00
to
CO
oo
in
~ -
CM
CO
to
p*^
co
oo
CM
CO
CO
in
CO
o
CM
in
o
to
en
CO
CO
to
en
IX-13
-------
The chief market for chrome ore is in the manufacture of chrome
ferroalloys, with this sector consuming over 60 percent. Other uses include
the manufacture of refractory products and the production of bichromates
within the chemical industry. Chromite ore has been traditionally classified
into three general grades—metallurgical, refractory, and chemical—depending
to a large extent upon chromium content, impurities, and the chromium/iron
ratio of the deposit. Although the Republic of South Africa accounts
for the greatest percentage of U.S. imports, deposits in the Transvaal
region are mostly of chemical grade due to the low chrome-iron ratio.
Consequently, lesser amounts of ore from this source are used by the
ferrochromium industry. Metallurgical-grade ore is normally imported from
the U.S.S.R., Rhodesia, and Turkey.
United States consumption of chromite ore has been cyclical, depending
upon the domestic market for ferrochromium and the international political
situation surrounding ore-exporting nations. From 1960 to 1966 chromite
ore imports and ore consumption generally increased at yearly rates of
about 5.0 percent and 3.5 percent, respectively. Average ore prices
remained relatively stable throughout this period, ranging between $33 and
$36 per short ton of chromium content (see Table IX-7). In December 1966,
however, the United Nations Security Council passed a resolution calling
for an economic boycott of Rhodesia by member nations, declaring that the
apartheid policies of that country's government constituted a threat to
peace. Pursuant to the U. N. resolution, a Presidential Executive Order
was issued in early 1967, imposing sanctions upon trade with Rhodesia.
As shown in Table IX-8, which presents import data for metallurgical-grade
IX-14
-------
to
Ul
o
o
a:
o
00
Di
D-
X
I—I
a>
[n
(O
r,
cu
o
•r- C
Q--8
S- V
CU C
-Q O
§r—
^^
O *tfl
Ol
d
OJ
1 1
— .
s-
CJ
»
cu
S-
o
E
3 t-
•i- 0
I+J
S- C
JC CU
CJ -P
C
o
o
>
S-
*t_J
C
3
o
0
o
PX*
CTl
r—
CTl
ID
CT1
r—
oc
CT
t—
vo
en
^~
U3
V£
CTi
|—
LO
VO
CT>
r^
^.
u:
P—
CO
VO
o^
r—
0
~"nj
S-
n
' Ol
•r-
S-
o
H-
O
o-
cr
z.
ex
•z.
o-
-Q
o-
^^
o
o
*
CO
CO
o
in
•
CO
CO
o
0
to
CO
CO
co
in
t
C£
•
oo
CO
zr>
o
in
in
in
o
0
00
o
o
oo
CO
o
o
in
CO
o
o
CO
CO
o
in
.
o
CO
o
o
r_
CO
o
o
r-.
co
CO
oo
>^
cu
\s
t_
13
I—
O
o
IjO
CM
in
CM
.
o
CM
in
CM
o
CM
in
CM
o
CM
in
p^
•
CTl
in
CM
f~~l
CM
o
in
oo
in
CM
o
CM
CM
^*
^~
u
*~
,
CO
,
ex
0)
cc.
O)
(U
01
•a
cu
o
n
cr
oo
03
S.
o
Q.
in
res
o
CO
O
0)
o
(U
S-
cu
cu
£
D)
CU
cu
a.
O
(1)
1
l/l
0)
•p
o
JD
o
c
IX-15
-------
CU
o
0)
o.
tn
O
O
o
a: o
o
a: cu
o i—
<=t- o
o
•> o
CO i—
o: x
o
Q. (/>
•-i S_
s: o
o-c
D; w
co
i
X
CU
re
re
CU
o
r-.
en
en
to
en
co
to
en
*«Q
CTt
to
to
en
r—
LO
to
en
to
en
co
en
£
o
o
O 1^ LO 1^ CXI
O 00 01 UD
o en <* co oo
en r-^ *3~ i —
CXI i —
o LO en cxi o
oo LO en
i— en <3- LO o
o en "^j- en
i— cxi
«3- oo oo oo cxi
i — CXI i —
CM CXI i —
en LO co ^ LO
^^ r*^ co to cxi
CM CM
«3- CM O CO O
re
u
•i —
ti-
re
•i- C£.
10 • >> 00
CU CO CU S-
*O • *^ • CU
O 00 S- Q. .C
c- ZS CU +->
o: ^> i— cc o
o
OO •=*- IO LO
o to r^ LO
r^ co CM
0
CM
CO
en
en to to o
cxi t-^ en *j-
LO CM CM
co
to co r-. r-^
CO LO tD CO
^- exi r~
en
oo
en
CTl CM LO O
OO CO LO 53-
LO CM LO
O
LO
en co co oo
to «d- en co
IO CO CO
CO
co
to
r^ CM oo to
1^ LO r— OO
to oo en
CM
LO
i— "=d- LO to
LO CO CT> CO
to co i —
•*
CM
CT>
•=j- co •=!• cn
en en to co
CO i — LO
*\
O
o +->
•(-> O SL
C i— O
CU £1
-(-> x in
Q f/y Lf\
0
r— CO QJ CO
re o zs o c:
-(-> CM r— CVO
O S- to S- +->
1— 0 => C_>
IX-16
-------
ore, chromite shipments from Rhodesia ceased entirely in 1968. The boycott
caused a general decrease in the supply of ore and severe price increases
for metallurgical-grade chromite between 1967 and 1971. Although U.S. ore
imports diminished during this period, domestic consumption remained
relatively constant, causing a general decrease in consumer inventories.
This situation may be alleviated in the near future because in late
1971 the United States Congress passed a military procurement bill
containing an amendment that removed Presidential authority to ban imports
of Rhodesian chromite after January 1, 1972. The amendment, which forbids
embargoes on any strategic material that is also imported from a communist-
dominated country, could eventually lead to an improvement in the raw
material cost/product price relationships for ferrochromium producers.
A small amount of Rhodesian chrome ore entered the United States in 1972;
however, political factors have so far prevented normal shipments.
The Product - Ferrochromium is used in various percentages for producing
iron castings and all types of steel, with about 70 percent going into
stainless steel. Substantial quantities are consumed in the production
of superalloys, and small amounts are used in nonferrous alloy production.
Output is normally in the forms of HC ferrochromium, LC ferrochromium,
or ferrochromium silicon, depending upon the alloy usage and the undesirability
of excess carbon as an impurity.
Table IX-9 shows production, shipments, and price statistics for
ferrochromium from 1963 to 1970. Although the industry encountered several
problems during this period, domestic output of all ferrochromium products
IX-17
-------
^i
§
o
o
i—i
to
i—i
l—
o
LU
<_>
I
x:
-O
ro
E
B
O
i.
(J
O
0)
• o
01 w—
in X
in in
in E
O> o
E
•i- +J
4-> O
tO -E
O
O
O
O i—
in x
u in
3 E
0 4J
O.+J
s_
o
in
E
.E
(J
o
i.
i.
01
l*-
0
_J
iromium
o
o
01
o
o
4J
in
01
§
Domestic
u
+3
in
<1>
s
Domestic
in
+J
O
i — i
E
0 O
•(-> +->
w o
01 3
O O
Q S-
CL
in
4J
S-
1
Domestic
shipments
Domestic
production
M- 01
0 C
••-O J3
OJ id -i—
U -t-> S- -^
•r- E O-faO-
S- O
0. 0
•» in
in E
•(-> o o
E +JO
01 o
E 4->r-
•r- O X
.E .E
in in
n
C in
O E
•r- OO
+J +J O
0 O
3 •!-> i—
-a s-
0 O X
i. .E
Q. in
•O
H_ a;
O E
•r-.a
01 ID -XI
O •<-> S- i—
•i- E O ~~^
S- O *<>
Q. O
« in
in E
•M O O
E 4->0
01 O
•i- O X
_c jr
in in
production,
short tons
x 1000
S-
ro
01
j^
CM
CO
CM
-
co
S-
r—
CO
in
CO
CD
o
o
CO
CM
CD
CTi
CM
r-.
O
CTi
CM
00
CTl
IO
in
CM
IO
in
CO
co
IO
CM
o
CO
o
in
CTI
o
CM
o
o
CM
to
o
CNi
CTl
IO
CTl
CM
CM
CO
O
IO
in
CO
00
to
co
in
in
CM
0
CO
CTl
CD
CTi
O
CTi
0
JO
00
r—
CO
IO
CTl
CTl
•ft
E
CO
m
5
10
CM
in
in
CM
o
0
CO
o
CTl
o
to
CTl
o
CM
IO
CTl
CO
in
10
CO
CTI
CM
s
in
in
CM
O
in
CM
CM
CD
CTi
O
OO
B
to
CTl
CO
co
CTl
in
in
CO
CTl
CO
CO
in
in
CM
o
to
CM
CO
CD
CTl
0
in
to
£
in
to
CTI
CTl
CO
•sr
to
CM
CM
CTI
CO
01
CO
in
in
CM
o
CM
CM
O
CTi
CD
in
r-.
CTl
in
to
CTI
-;
o
CM
CTl
CM
0
o
CO
o
o
CO
0
CO
o
CM
o
CO
CTI
o
CM
0
in
to
co
to
CTl
O
O)
s.
(U
-a
c:
O)
•a
ai
+->
o
3
CT
£ E
ro O
3 i-
C J=
ro O
M-
O
ai
in
01
cr c
£ "~
-Q XI
01
E •!->
3 O
1 ^
U 3
C O
O i-
O -E
•r- U
c
u
i.
0}
a.
IO
O-
ro
"a.
I
•r- r-~
E a.
3 E
•r- 3
I I
.E •—
U 3
<1J 4J
l|- E
ro
O r-
_J Q.
-o
(U
-a
3
-a
ro
O
in
•o
ro
O
(U
U
s-
d)
ra-
in
fM
O
O t
O ro
3
ro <4-
O O
in
ro
re -Q
X 01
ro 0)
•SL-3.
-------
showed rather steady growth. Yearly Increases were attained through 1967,
after which a cyclical pattern set in for the remainder of the decade.
Linear regression shows a growth rate from 1963 to 1970 of approximately
3.2 percent per year, with total domestic production presently about
400,000 short tons per year.
Because of its close relationship with stainless steel (for which
market researchers predict long-term growth at a greater rate than for
raw steel), ferrochromium should have the most favorable outlook of any
of the major ferroalloys. However, three interrelated problems may plague
the domestic producers in the near future: (1) Higher chrome ore prices
and tightening supplies of metallurgical-grade lump, caused by the
Rhodesian boycott, have forced the domestic industry to raise alloy prices.
As measured by quotations for LC ferrochromium (maximum carbon content
0.025 percent), prices have increased from 25.5 cents per pound of contained
chromium in early 1968 to 39.5 cents per pound in late 1971, a jump of
some 55 percent. (2) Such price increases may stimulate higher levels of
alloy imports since foreign producers have a cost advantage going into the
marketplace. Ferrochromium imports have in fact grown at a rate of about
4.0 percent per year from 1963 to 1970, although the statistics show large
fluctuations. Norway, Sweden, and the Republic of South Africa are major
suppliers, having easy access to low-cost electric power and/or indigenous
raw materials. West Germany, France, and Japan also export substantial
quantities to the United States. (3) The rising prices for ferrochromium
and nickel have forced parallel increases in quotations for stainless
steel, estimated to be costing U.S. consumers an extra $100 million per
IX-19
-------
annum. The domestic steel industry, with mounting competition from
abroad, has also lost a substantial part of its stainless steel markets.
As shown in Table IX-9, stainless steel imports have risen dramatically
from the 1963 level, even with allowance for the programs of voluntary
steel import quotas underway since 1969. For 1971, stainless steel
imports stood at 191,000 tons, up about 8 percent from the 1970 level.
If international political problems can be settled, the prospects
for ferrochromium appear to be favorable. If Rhodesian ore supplies
become available once again, the prices for chrome ore should drop and
the ore markets should stabilize. Ferrochromium prices should follow a
similar downward pattern, which would benefit both alloy producers and the
domestic steel industry. Over the long run, growth in demand for ferro-
chromium should exceed the rate of expansion in steel output, reflecting
continued efforts to upgrade certain qualities of steel, particularly
corrosion resistance and increased strength. Between 1968 and 2000,
the U.S. Bureau of Mines predicts a growth of demand for ferrochromium
of 2.0 to 3.3 percent yearly. However, the United States will remain
dependent upon foreign sources of chrome ore, a situation that will
continue to be a potential problem.
d. Ferrosilicon
The Ore - Silica raw materials are widely distributed throughout the world,
and the processing required to retrieve the ore is relatively simple.
United States domestic supplies are plentiful, and the quarrying of quartz,
quartzite, and sandstone is essentially a domestic industry. Consequently,
ferrosilicon is one of the few ferroalloys made from an ore that is not
IX-20
-------
subject to the fluctuations of international political and economic
forces. Conventional processing consists of removing the material with
power equipment, followed by crushing, sizing, and washing.
United States production and consumption statistics for silica gravel
or crushed rock are not regularly collected. However, with an assumed
production ratio of 3 to 1 for guartzite converted to silicon contained
in all ferroalloys, the U.S. Bureau of Mines estimated that approximately
1.5 million tons of silica raw materials were guarried in !QfiB. Because
these raw materials are commodities having a low value-to-bulk ratio,
transportation costs are a major item and in most cases can determine
the source and the distance the material can be hauled. Prices for
quartz or quartzite are dependent to a large degree upon chemical
analysis, sizinq, quantity, and negotiated contracts. For 1%8, guoted
prices including transportation ranged from $R to $12 per ton in various
locations throughout the nation.
The outlook for silica rock is continued stability. Domestic raw
materials will be in ample supply well beyond the year 2000. The United
States will not have to depend upon imports for any part of its supply.
The Product - It is estimated that about QO percent of all silicon is
consumed by the iron and steel industry in the form of ferrosilicon alloys,
Ferrosilicon is regularly used to deoxidize the molten metal and remove
dissolved gases. It is also used to produce hiqh-silicon "alloy" steels
IX-21
-------
with greater corrosion resistance and improved strength, and low-iron-loss
steels for electrical transformers and motors. Ferrosilicon is also
used in gray iron foundries to increase the amount of silicon in the iron,
as it is necessary to add silicon when using scrap steel charge in the
cupola. Products are normally classified as silvery pig iron (15 to 20
percent silicon), ferrosilicon (21 to 95 percent silicon), and silicon
metal (96 to 99 percent silicon), with several percentage grades made
in each class. No silvery pig iron is now made in blast furnaces, but
it is produced in the submerged-arc furnace. The major market for silicon
metal is the aluminum casting industry. Because iron is unacceptable
in aluminum alloys, silicon metal is added to aluminum instead of the
normal grades of ferrosilicon to improve corrosion resistance, weldability,
and casting and machining properties. Secondary aluminum producers are
the largest consumers, accounting for some 65 percent of the demand for
silicon metal in 1969.
Table IX-10 gives production, shipments, imports, and price statistics
for two classes of ferrosilicon from 1965 to 1971. Based upon simple
linear regression, the growth rate for domestic production of silvery
pig iron dropped 4.4 percent while that for ferrosilicon rose 3.3 percent
per year from 1965 to 1971. Growth in shipments of silvery pig iron,
some 85 percent of which is consumed by the gray iron foundry industry,
followed the trend in cast iron output from foundries, where production
has been slowly receding since 1955. Domestic shipments of 21 to 95 percent
ferrosilicon increased about 3.3 percent per year, while prices ranged
IX-22
-------
O
CJ
CO
o
CO
O
Q
LU
O
X
I—I
0)
JD
CO
•r—
CO
.O
O
CM
"-'
C
o
•r-
co
o
S-
S-
[2?
• — *
•I —
CO
C3
CM
O
4J
LO
*^s
C.
2
•r—
•r—
CL
>^
S-
O)
^_
it-
IS)
(O O
O CO i —
+•> ra
0) S_
3 O J3
t— Q-i—
<0 E \
CO
c
« o o
CO 4-J O
•(-> 0
0 S- "~
+-> 0. O X
CU E -C
Z -i- CO
re
CO
14_ 4J
O C ••-
CO CO
CO E
-3 JS o
ro J^ ^
>• 10 •**
CO
(SI C
O -l-> O O
r- C 4-> O
•4-> CO O
(/>£•<->•—
O) CX S-
E -i- 0 X
0 .c x:
Q co co
C CO
O c
O -1- O O
•r- +J +J o
4-> 0 O
CO 3 +-> i—
cu -a s-
E 0 O X
o s- -c
Q Q- CO
re
CO
O C -r-
OJ CO
0) E
3 Q-JD
re -C *• —
> CO -<=>•
" CO
CO C
0 -M 0 O
•- C -P O
-t-> O) O
co E -l-1 <—
QJ CL S_
E •!- 0 X
O co co
^
C CO
o c
O T- O O
i- +J +-> O
•4-> O O
CO 3 +•> i—
CO T3 J-
E O O X
O s- x:
p Q- CO
s-
0)
o n oo ^-
• • * I
•— CO VO VO
^— ^~
o o
i— CM r~- r^
CvJ CM
cn LO «3- cn
• • • •
l-» f>. ** ^~
oo cn LO cn
r** LO ? — . o
vo vo vo vo
i — cn LO LO
oo o i— vo
vo i — r— vo
oo co ^- *3-
LO CM CO CO
CM CJ CM CM
^f CO " **• VO
f- O O r—
i— CM CM CM
CM VO -d- LO
r^ cn co cn
r— i— CM .—
r— o cn oo
r-« h- vo vo
cn cn cn cn
CO
•
p*i.
cn
r^
cn
•
*f
CO
o
vo
«^-
f%^.
vo
cn
•
LO
CM
O
csj
o
CSJ
CM
r--.
vo
cn
vo
•
p^,
LO
CM
CO
VO
00
LO
VO
fx.
LO
VO
•
VO
CM
VO
^~
CVI
LO
LO
CM
VO
VO
cn
en
•
^-
CM
r—
VO
•
LO
oo
oo
LO
LO
en
LO
o
•
r-
CM
^j-
CM
CM
f^
r—
CM
LO
VO
cn
-o
O)
o
13
O
CL
i.
0)
O
CJ
CD
O
Ol
CL
QJ
CT
ro
0)
C
QJ
CO
re
JD
CO
CO
o
o
i-
OJ
•r—
re
*->
c
o
o
S-
o
+J
D5
•f"~
01
s-
4_
•M *
H3 ^.
o o
1 * ,,_-!!-
•r— fO
s
CO I—
•4-> -^
c -<=>•
.
CO
4->
i-
O
CL
X
a>
a>
i \
c
ai
CO
OJ
i.
CL
ai
i.
CO
QJ
s-
3
Ol
•r-
O)
3
(O
•i- O
c cn
0) i—
>
•5, 1
re
O)
cn
r-
O
IX-23
-------
from 15 to 18 cents per pound of contained silicon. Markets for
ferrosilicon have been growing chiefly because steelmakers have increasingly
demanded a wider variety of specialized ferrosilicon grades to be used
for high-silicon alloy steels.
As reflected in Table IX-10, imports have not been a major influence
upon ferrosilicon markets.. Overseas shipments were normally less than
inflows until 1970, when exports exceeded imports. Most of the trading
is with Canada and the United Kingdom. The Norwegian Ferrosilicon
Producers Association announced in 1972 that it has adopted a major
marketing program to penetrate the U.S. market.
The outlook for ferrosilicon products is somewhat favorable. Future
demands for silvery pig iron and 21 to 95 percent ferrosilicon are expected
to follow the trends in iron and steel growth, estimated at 4 percent per
year. Because raw materials are plentiful and silica is relatively
inexpensive, there is little likelihood that silicon alloys will be
replaced by substitute products. Furthermore, technological advances
may even increase the use of ferrosilicons for replacement of the more
expensive corrosion-resisting additives such as chromium.
3. New Units
Concrete projections of the number of new units to be installed by
the industry are not readily available in the literature. However,
predictions can be made based upon the growth rates of the various product
classes and by a consideration of the replacements needed for older
furnaces. Although the growth rate for ferromanganese and silicomanganese
products has been negative in recent years, consumption should keep pace
IX-24
-------
with the 4 percent annual long-term trend of steel industry growth.
However, only 50 percent is currently supplied by electric furnace.
If this ratio holds, it will mean a net growth of one 30-mw furnace
every 2 years if all the increase can be obtained by domestic producers.
However, in the recent past domestic producers have only supplied about
80 percent of the domestic market for ferromanganese and 90 percent
of the silicomanganese market.
Generally it is made in the same type of equipment as ferromanganese
and silicomanganese. However, no new units are anticipated for the express
purpose of producing this material. One source quoted by the Chemical
Economics Handbook believes that acetylene derived from calcium carbide
may cease to be used for the production of chemicals. The only thing
that might reverse this trend would be a jump in the price of the competing
light hydrocarbons due to the energy shortage.
Growth rates of approximately 3 percent for ferrochromium and ferro-
silicon products seem to dictate that this segment of the industry will need
approximately one large furnace (30,000 to 40,000 kw) in alternate years.
There are a total of about 150 existing furnaces in the industry;
assuming an average furnace life of 30 years, about five furnaces per
year should have to be replaced. The trend in the industry, however, is
to replace smaller furnaces (average size estimated to be about 10,000 kw)
with much larger units (probably around 40,000 kw). Given a size ratio
IX-25
-------
of 4 to 1 for old-to-new furnaces, it is expected that approximately one unit
per year will be needed for replacement purposes. In total, about 5 to 8
new ferroalloy furnaces are estimated to be installed in the next 5 years.
C. CONTROL COSTS
1. Introduction
Capture of pollutants is the critical factor in designing-atmos-
pheric emission control systems for ferroalloy plants. The major
source of emissions is the carbon reduction of metallic oxides in the
submerged arc furnace. Carbon monoxide is generated continuously
along with other reaction gases and fumes. The carbon monoxide from
the furnace reaction zone may be withdrawn by an exhaust system with-
out combustion provided a furnace has a closed water-cooled cover
and seals around the electrodes. The covered ferroalloy furnace may
only be used for a limited number of products but offers the advantage
of producting smaller gas volumes to clean than an open, hooded furnace.
The small volume of dirty gases from a covered system is typically
cleaned by high-energy scrubbers.
The open-furnace system allows induced air to mix with and burn the
carbon monoxide above the charge. Depending on design of this particular
furnace type, evolution of gases may result in flows of 20 to 50
times those generated by the covered system. The volume of gas flow
depends on the hood design, the vertical opening required for stoking
the charge, and the diameter of the furnace. In addition, open furnaces
with provisions for adding electrode sections under electrical load,
require a protection area for electrode installation. Some older open
furnaces add electrodes under no load conditions at which time venting
occurs directly through roof monitors. To control such a system,
IX-26
-------
hooding requirements call for a much greater volume of gas than would be
required if a hood were placed at a lower elevation. Collection devices
for open furnaces include fabric filters, scrubbers, and electrostatic
precipitators.
In the following sections, costs will be discussed for scrubbers on
covered furnaces, and scrubbers and fabric filters on open systems. Not
enough data are available to discuss electrostatic precipitator costs.
It is important to point out that covered systems work only for a limited
number of products—ferromanganese, silicomanganese, 50 percent ferro-
silicon, some grades of HC ferrochrome and calcium carbide.
2. Model Plants
Model furnaces were developed to evaluate the control cost. Because
the trend in the industry is toward larger furnaces than in the past,
the size chosen for the models is large - 30 megawatts. Table IX-11
shows the pertinent design parameters associated with the model furnaces.
Since silicomanganese (SiMn) can be made in the same furnace inter-
changeably with high-carbon ferromanganese (HC FeMn), we have assumed that
the control equipment for the SiMn furnace will be the same as for the
HC FeMn furnace.
IX-27
-------
Table IX-11. MODEL FURNACE PARAMETERS
Product
Parameter
Power rating, mw
Product rate,3 tons/yr.
Gas volume from
HC FeMn
30
99,000
5,non
SiMn
30
44,000
5,noo
50% FeSi
30
47,500
6,000
HC FeCr
30
51,000
5,000
CaC?
30
91 ,000
4,000
totally enclosed
furnace,& scfm
Gas volume from open 350,000 350,000° 450,000 250,000 200,000
furnace,b acfm
O 400°F
Tapping fume gas 60,000 60,000 60,000 60,000 60,000
volume from all
furnace types, acfm
0 150°Fd
aAt 90 percent of full capacity.
The gas volumes represent typical values obtained
from the industry survey questionnaires.
°Assumed to be the same as for HC FeMn since the furnace
may be designed to produce either product.
The figures shown for the tap fume collection are additive to the open
furnace volume, based on an open furnace configuration with the
collection hood 5 to 7 feet above the furnace deck.
IX-28
-------
Another emission source that must be controlled besides the furnace
itself is the furnace tapnina operation. The method of control assumed
for this cost analysis depends on the furnace type. For open furnaces
the tapping fumes can be collected with a separate hood and vented into
the main control device. For other furnaces a separate fabric filter
control system would be the most probable method of control.
In addition to the costs for the model plants, costs are presented
for a large, totally enclosed furnace that is currently under construction
in North America. These data should represent the most up-to-date costs
experienced by industry for construction of a totally enclosed furnace
in this part of the world.
3. Open Furnace Control Costs
Control costs for the model open furnaces shown in Table IX-11
were developed for two types of control devices - fabric filters and
wet scrubbers.
a. Fabric Filter Control Costs - Estimates of investment and operating
costs required to control open furnaces using fabric filter systems are
shown in Table IX-12. These costs were developed from information sub-
op
mitted to EPA by the Industrial Gas Cleaning Institute (IGCI). The tap-
ping fume control system is vented into the fabric filter, and the costs
for that system are included. The assumptions that form the basis for
these cost estimates will be discussed below. The industry's cost esti-
mates for fabric filter systems are higher than the figures in Table IX-12
because additional eguipment and installation factors are considered. The
industry's cost estimates are shown in Table IX-13 and will be discussed in
the second part of this section.
IX-29
-------
Table IX-12. CONTROL COSTS FOR FABRIC FILTERS ON OPFN FURNACES
Product
HC FeMn
Cost item and SiMn 50% FeSi HC FeCr CaC
2
Capital cost
Fabric filter $ 682,000 $ 832,000 $ 532,000 $ 449,000
Auxiliary equipment 226,000 275,000 176,000 149,000
Installation 1,142,000 1,393,000 892,000 752,000
Total capital cost S 2,050,noO $ 2,500,000 $ l,600,nno $1,350,000
Annual cost
Operating labor
Maintenance (6%)
Electricity
Capital recovery
(15 yr. life, 8% interest)
Administration (2%)
Taxes and insurance (2%)
53,000
123,000
87,000
240,000
41,000
41,000
53,000
150,000
106,000
292,000
50,000
50,000
53,000
96,000
68,000
187,000
32,000
32,000
53,000
81 ,000
57,000
158,000
27,000
27,000
Total annual cost $ 585,000 $ 701,000 $ 468,000 S 403,000
HC FeMn SiMn
Annual cost per ton * 5.91 $ 13.30 $ 14.76 $ P.18 S 4.49
IX-30
-------
4.0
2.0
1.0
0.8
0.4
0.2
0.1
20
40 60 80 100 200
INLET GAS VOLUME TO COLLECTOR, acfm x 103
400
600
Figure IX-1. Capital costs of open furnace control with fabric filters.
IX-31
-------
The capital costs for fabric filter installations as received from
the IfiCI were plotted against the associated collector inlet volumes, and
the graph is shown in Figure IX-1. The capital cost for each model furnace
may be determined from Figure IX-1 by finding the capital cost that
corresponds to the gas volume flow rate for that model. The capital costs
from the IGCI study are based on a new plant situation (I.e., a simple
duct run, no space limitations, etc.). The costs for the furnace hood
and the incremental electrical substation are not included. The capital
costs for the fabric filter installations include the baghouse, fans,
upstream mechanical collector, dust storage bins with 24-hour capacity,
dust hoppers and conveyers, foundation support, ductwork connections, and
stack. The charges for engineering design layout, electrical and piping
tie-ins, insulation, erection, performance testing, and startup are all
included. Fiber glass bags with a temperature resistance of 500 F are
assumed to be used. The baghouse is also assumed to contain one extra
compartment, which permits shutdown for maintenance.
The following assumptions concerning annual costs of operation apply
to operation of the control facility for open furnaces.
1. Replacement parts and maintenance were estimated at 6 percent
of the original plant investment for the purpose of replacing
50 percent of the bags and 10 percent of the air valves per
annum, and for unknown contingencies.
2. Manpower requirements were estimated to be 1/2 man per shift.
3. Electricity costs account for sufficient power to push the gas
into the baghouse with 10 to 12 inches pressure loss for HC
FeCr and 15 to 20 inches pressure loss for FeSi. Electrical
costs were based on 1 cent per kilowatt-hour.
IX-32
-------
4. Depreciation and interest charges are accounted for by the use
of a capital recovery factor based on 15 year life and on 8
percent interest rate.
5. Administrative costs of 2 oercent of original investment and
another 2 percent for property tax and insurance were assumed.
The ferroalloy industry has reported higher costs for fabric filter
installations due to the following factors:
1. The industry's cost figures are based mainly on installations
at existing plant sites. Since these installations must be fit
into the available space, certain cost items such as ducting will
be more expensive.
2. The industry's figures also include items that were not included
in the IGCI cost estimates. These items are the furnace hood cost,
electrical substation expansion costs, equipment startuo costs,
and company engineering and contingency costs.
Including these items and assuming a retrofit installation, the capital costs
can be as much as 50 percent higher than the IGCI costs. Table IX-13 shows
the industry's cost estimates for the model furnaces.
If the average of the IGCI costs and the industry's costs are used,
the annual cost per ton ranges from a low of $5.12 per ton for calcium
carbide to $17.73 per ton for 50 percent ferrosilicon.
b. Wet Scrubber Control Costs - Estimates of the investment and operating
costs required to control open furnaces using wet scrubbers are shown in
Table IX-14. These estimates are derived from information from the
3?
Industrial Gas Cleaning Institute (IGCI) and are based on equipment and
IX-33
-------
Table IX-13. CONTROL COSTS FOR FABRIC FILTERS ON OPEN FURNACES
(REPORTED BY INDUSTRY)
HC FeMn
Cost item and S1Mn 50% FeSi HC FeCr CaC2
Capital cost
Fabric filter $ 1,000,000 $ 1,265,000 $ 700,000 $ 630,000
Auxiliary equipment 360,000 455,000 255,000 220,000
Installation 1,640,000 2,080,000 1,145,000 1,050,000
Total capital cost $3,000,000 $3,800,000 $ 2,100,noo $1,900,000
Annual cost
Operating labor $
Maintenance (6%)
Electricity
Capital recovery
(15 yr. life, 8X interest)
Administration (2%)
Taxes and insurance (2%}
53,000
180,000
87,000
350,000
60,000
60,000
$ 53,000 $
228,000
106,000
444,000
76,000
76,000
53,000
126,000
68,000
245,000
42,000
42,000
$ 53,000
114,000
57,000
222,000
38,000
38,000
Total annual cost $ 790,000 $ 983,000 $ 576,000 $ 522,000
HC FeMn SiMn
Annual cost per ton $ 7.98 $17.95 $ 20.69 $ 11.29 $ 5.74
IX-34
-------
operating requirements to meet the process weight standard published
in the Federal Register of August 14, 1971. The costs have been adjusted
from I6CI data to reflect the gas flows of the model plants presented in
Table IV-11. The costs in Table IX-14 are based on a new plant installa-
tion and do not include the furnace hood or additional electrical
substation costs. The industry's experience confirms the costs as
presented in Table IX-14.
Plots of investment cost data for scrubbers developed by the IGCI
are shown in Figure IX-2 for HC ferrochrome and 50 percent ferrosilicon
furnaces. The cost curve for ferrochrome was used to develop the costs
for all the other alloys except 50 percent ferrosilicon. The investment
costs include a venturi scrubber, a fan with at least 20 percent excess
capacity, and entrainment separator, aftercoolers, a slurry settler, two
filters to dewater the slurry product, and tapping emissions control. The
charges for engineering design layout, electrical wiring, piping, insula-
tion, erection, performance testing, and startup are all included. It
should be noted that the water treatment equipment may not necessarily be
enough to meet EPA's proposed effluent standards.
The annual cost per ton of product ranges from a low of $7.08 per
ton for calcium carbide to a high of $34.38 per ton for 50 percent ferro-
silicon.
4. Totally Enclosed Furnace Control Costs
Capital and annual costs are presented in this section for control
devices on totally enclosed furnaces. Since the furnace has a tight cover
with seals around the electrodes, the gas volume going to the control device
is much smaller than for an open furnace. Thus, the cost of the control
IX-35
-------
Table lk-14. CONTROL COSTS FOR WET SCRUBBERS ON OPEN FURNACES
Product
Cost item
Capital cost
Scrubber $
Auxiliary equipment
Installation
Total capital cost $
Annual cost
Operating labor S
Maintenance (7%)
Electricity
Water
Capital recovery
(15 yr. life, 8% interest)
Administration (2%)
Taxes and insurance (2%}
Total annual cost $
HC
Annual cost per ton
of product $8.
HC FeMn
and SiMn
110,000 $
290,000
1,400,000
1,800,000 $
26,000 $
126,000
290,000
155,000
210,000
36,000
36,000
879,000 $
FeMn SiMn
88 $19.97
50% FeSi
190,000
510,000
2,450,000
3,150,000
26,000
220,000
595,000
298,000
368,000
63,000
63,000
1,633,000
$34.38
HC FeCr
$ 96,000
254,000
1,250,000
$ 1,600,000
$ 26,000
112,000
225,000
118,000
187,000
32,000
32,000
$ 732,000
$14,35
CaC2
$ 87,000
233,000
1,130,000
$ 1,450,000
$ 26,000
102,000
190,000
99,000
169,000
29,000
29,000
$ 644,000
$ 7.08
IX-36
-------
4.0
2.0
1.0
0.8
CNJ
a
- o.e
0.4
0.2
0.1
20
FERROCHROME
40 60 80 100 200
INLET GAS VOLUME, acfm x 103
400
600
Figure IX-2. Capital costs of open furnace control with wet scrubbers.32
IX-37
-------
device is much smaller than for the open furnace. The higher cost of a
covered furnace compared to an open furnace will be discussed in Section 6.
In addition to the furnace control device, a separate fabric filter to
control taphole emissions is included.
a. Wet Scrubber Control Costs for Furnace Gas Cleaning - The capital and
annual cost for two types of wet scrubber control systems are discussed
in this section. Both systems are considered best demonstrated technology,
but one system includes two parallel control systems which increase the
cost. The costs are based on actual costs reported for foreign installations
of HC FeMn furnaces updated to 1973 U.S. dollars. However, European con-
struction costs will be generally lower than similar installations in the
United States because of differences in building standards, labor rates,
safety requirements, accounting procedures, and tax structure.
System A consists of two parallel control devices each capable of
handling the total gas flow. The control device is made up of three stages
of venturi scrubbers. The draft for this system is provided by aspiration
when water is injected at the venturi throats. The capital cost includes
the complete scrubber system and the necessary water treatment facilities
and flare stack. The annual cost includes operating labor and materials,
maintenance, depreciation, interest on capital, administrative costs,
property tax and insurance. Ho credit has been taken for the heat recovery
value of the carbon monoxide (CO), which can be a significant amount pro-
vided conditions permit use of the gas. For example, the heat value of
5000 scfm of gas (assuming 60 percent CO) from the HC FeMn furnace is about
$180,000 per year based on production at 90 percent of full capacity and a
heat value of $0.40 per million Btu.
IX-38
-------
The total reported capital costs for system A was about $630,000, which
was about 9 percent of the total furnace installation cost of $7.2 million.
The reported annual cost for system A was about $225,000.
System B is another foreign installation consisting of a two-stage
venturi scrubber with a fan to supply the necessary pressure. The capital
costs include the control system, flare stack, and water treatment
facilities.
The reported capital cost of control system B was about $260,000, which
was about 6 percent of the total furnace installation cost of 4.5 million.
The reported annual cost was about $77,000. Again, no credit has been taken
for the heat value of the CO. This furnace was installed at an existing
site where some of the existing equipment could be adapted for use with
the new furnace.
b. Fabric Filter Control Cost- A few overseas companies use fabric filters
as the control device on totally enclosed furnaces. This method of control
has not been used in the U.S., and the domestic industry does not expect
to use this method of control for totally enclosed furnaces. The estimated
capital cost for a conventional fabric filter control system consisting of
a radiant cooler, cyclone, fan, fabric filter, dust removal and storage
equipment, water seal tank, and flare stack is about $250,000. However,
this system would have to be specially designed because of the high con-
centration of CO gas. These added design considerations could double or
triple the cost.
C. Tapping Fume Control Cost- The estimated capital and annual costs
presented in Table IX-15 are based on a separate fabric filter control
system for emissions generated during the furnace tapping operation. The
IX-39
-------
Table IX-15. CONTROL COSTS FOR A SEPARATE TAPPING
FUME COLLECTION SYSTEM
Cost item Cost
Capital cost
Fabric filter $ 85,000
Auxiliary equipment 55,000
Installation 260,000
Total capital cost $ 400,000
Annual cost
Operating labor
Maintenance (10%)
Electricity
Capital recovery
(15 yr. life at 8% interest)
Administration (2%)
Taxes and insurance (2%)
$ 10,000
40,000
23,000
47,000
8,000
8,000
Total annual cost S 136,000
IX-40
-------
assumed flowrate was 60,000 acfm at 150°F. The system includes a hood,
fan, fabric filter, and dust removal and storage equipment.
Because the tapping operation can be scheduled with some flexibility,
this control system could serve more than one furnace. Possibly two
tapping fume hoods could be vented to the same fabric filter, and this
would reduce the control cost per furnace. However, for this analysis
a separate tapping fume control system for each furnace has been assumed.
5. Semi-enclosed Furnace Control Costs
The main difference batween a semi-enclosed furnace and a totally
enclosed furnace is the method of sealing the area around the electrodes.
On the semi-enclosed furnace, the seal is made by maintaining the feed
mix around the electrodes. The furnace gases drawn from under the cover
require treatment in the same manner as the totally enclosed furnace gas.
In addition, hoods may be constructed to capture the emissions that
escape around the electrodes. This gas stream can be controlled by com-
bining it with the taphole gases and venting the combined stream to a
fabric filter.
One way to illustrate the cost of control for the semi-enclosed
furnace is to examine the differences in control cost between the semi-
enclosed furnace and the totally enclosed furnace. The capital cost of
semi-enclosed furnace installation is lower than the totally enclosed
furnace by about the amount of the mechanical seals. On the other hand,
electrode and taphole emissions from the semi-enclosed furnace may require
a fabric filter that is about twice as large as the fabric filter for the
taphole emissions from the totally enclosed furnace. For a 30 mw furnace,
IX-41
-------
the cost of these two factors approximately cancel each other. In general,
the control cost for the semi-enclosed furnace is equal to or slightly
greater than that for the totally enclosed furnace. The annual cost for
the semi-enclosed furnace would be higher by the amount of the operating
cost to run the larger fabric filter.
6. Case Study of a Totally Enclosed Furnace
As shown by the previous models, the cost of the air pollution control
equipment for the totally enclosed furnace is considerably less than the
open furnace cost. The main reason is that the gas volume from the totally
enclosed furnace is much smaller.
However, the pollution control equipment is not the only consideration
when comparing the open and totally enclosed furnace. Actually the open
and totally enclosed furnaces require two different sets of process equip-
ment of which the pollution control system is one part.
In order to make a complete comparison of the two furnace types, one
should look at the total system from both the process side and the air
pollution control side. On one hand the totally enclosed furnace requires
a more expensive furnace installation with a larger building, a more complex
feed handling system, and a more complex furnace cover, fin the other hand,
the totally enclosed furnace can be controlled with a much smaller and less
expensive air pollution control system. In this section the costs for a
totally enclosed furnace are compared to the costs for an open furnace to
Illustrate this point.
The costs in this section are for a large totally enclosed furnace
under construction in North America. These costs should he more represen-
tative of the costs that would be experienced at a U. S. location than
IX-42
-------
some of the previously presented costs for totally enclosed furnaces which
were based on overseas installations. The maximum power rating for this
furnace is 33 mw for HC FeMn and 38 mw for SiMn.
The primary control system for the totally enclosed furnace consists
of the sealed furnace cover, a water spray cooler, a mechanical dust
separator, and a variable-throat venturi scrubber followed by a mist
eliminator. The pressure drop across the scrubber is in the range of 75 to
80 inches of water. The gas flow from the furnace is about 6600 scfm, and
the gas flow at the scrubber is about 9700 scfm. The cleaned gas stream,
which is high in CO, can be used as a fuel source in the feed pretreatment
plant or diverted to a stack. A complete water treatment system is
included; the treated water is recycled to the scrubber and the filter cake
of solids is recycled to the sintering plant.
The furnace tapping system is designed with a hood over each of four
tapholes. A total flow rate of 30,000 acfm is combined with another 20,000
acfm vent stream and sent to a fabric filter collector.
Table IX-16 shows the costs for the totally enclosed furnace and its
control equipment compared to the company's estimated costs for an open
furnace with a fabric filter collection system. The prorated share of the
project's utilities, electrical and engineering expense for the control
system is included in the control system cost. In addition to the furnace
collection system and the tapping emission collection system, the company
reported two other cost factors for the totally enclosed furnace that are
different from the open furnace. The first is the incremental furnace
cost which includes such items as a more complex feed system, a taller
building, and larger and more complex electrode columns and electrical equip-
ment. The second item is an incremental feed pretreatment cost which includes
ore and coke dryers and a sinter plant.
IX-43
-------
Table IX-16. COMPARISON OF CAPITAL AND ANNUAL COSTS FOR AN OPEN
AND A TOTALLY ENCLOSED HC FeMn AND SiMn FURNACE
PRODUCING HC FeMn OR SiMn
Totally enclosed
Cost item Open furnace furnace
Comparison of total capital costs9
Basic furnace and associated
process equipment
Incremental furnace cost
Incremental feed pretreatment
Air pollution control systems
$8,500,000
3,500,000
$12,000,000
S 8,500,000
1,400,000
3,000,000
2,100,000
$15,000,000
Comparison of control equipment costs
Capital costs3
Primary system $ 3,500,000 $ 1,700,000
Taphole system (included in above) 400,000
Incremental furnace cost -- 1,400,000
$ 3,500,000 $ 3,500,000
Annual costs
Operating cost $ 143,000 $ 135,000
Maintenance (6%) 210,000. 210,000
Capital recovery 409,000° 390,000°
(at 8% interest)
Administration (2%) 70,000 70,000
Taxes and insurance (2%) 70,000 70,000
$ 902,000 $ 875,000
Annual cost per ton ($/ton)
HC FeMn 9.11 8.84®
SiMn 20.50 19.896
aLetter from Mr. D. J. Maclntyre, Manager, Environmental Affairs, Union
Carbide Canada Limited. March 9, 1973.
Based on 30 row for HC FeMn and 34 mw for SiMn, both at 90% operating rate.
Depreciation lives: 10 years - furnace cover, 15 years - pollution control system
20 years - incremental furnace costs.
Depreciation life: 15 years.
eThis does not include the annualized investment cost or operating cost of the
incremental feed pretreatment equipment. The ferroalloy industry has indicated
that the total manufacturing cost per ton of product is about equal for both
the open furnace with control and the totally enclosed furnace with control and
feed preparation.
IX-44
-------
The decision to use the incremental feed pretreattnent must be made after
evaluation of the overall process. Drying and sintering allow the use of
coke and ore fines and the recovered particulates from the air pollution
control systems. Through the use of these feed pretreatment steps the
furnace can be operated in a smoother and safer manner. Some foreign plants
with totally enclosed furnaces have these additional feed pretreatment steps
and some do not. It is even hard to define exactly what should be included
as incremental feed pretreatment equipment. For example, some open furnaces
have dryers and some do not (depending on the availability of dry materials).
Thus, dryers may or may not be considered as incremental equipment for
totally enclosed furnaces. The incremental feed pretreatment cost could be
considered as part of the air pollution control cost, or could be considered
a process addition for which the economics must be justified in each
individual case.
In Table IX-16 the capital cost for the incremental feed pretreatment
is shown, but these costs are not included in the presentation of the annual
cost of the air pollution control equipment. After an overall evaluation
was made, this particular plant decided that the totally enclosed furnace
with the additional feed pretreatment was the best choice in this case.
It is not possible to generalize from this case to say that in all cases
the totally enclosed furnace with feed pretreatment would be the best choice.
For example, in the case where a furnace is to be added at an existing plant
an open furnace could possibly use the existing feed preparation and
delivery system whereas a totally enclosed furnace might require a new,
separate feed pretreatment system. Also, the open furnace could possibly
be installed in an existing building while the taller, totally enclosed
furnace would probably require a new or expanded building. These or other
differences at any specific site could affect the costs enough to change
the best choice of furnace type to an open furnace.
IX-45
-------
D. ECONOMIC IMPACT
1. Introduction
The impact of abatement costs will be analyzed in this section, and
the model plant approach will be continued through development of income
statements before and after control for the five hypothetical plants. The
ability of the firm to raise the capital necessary for control will then
be discussed. Finally, a general overview of the industry will cover
important trends and prospects.
Sufficient data describing the profitability of the ferroalloy industry,
particularly with regard to individual products, are difficult to obtain.
Three firms that were not diversified have been chosen to represent the
industry, so distortion from other product groups is precluded. Table IX-17
presents aggregate operating data for these three firms in terms of percent
of sales for the period 1963 to 1971. As can be inferred from the table,
operating costs for the 9 years have averaged approximately 93 percent of
sales, with earnings before taxes around 7 percent of sales. Net income
after taxes has averaged approximately 3.89 percent of sales, and cash" flow,
about 8.19 percent of sales. These average results will be used to generate
model income statments for the five hypothetical plants.
2. Model Income Statements
Tables IX-18 and IX-19 present earnings statements for the model plants
based on costs presented in Tables IX-12 and IX-13, respectively. Sales
figures for ferroalloys were calculated using annual capacities of the
30 mw furnaces and the value per ton of shipments from quotations in the
1973 American Metals Market. These values were: HC ferromanganese, $200.00;
silicomanganese, $210.00; HC ferrochrome, $264.00; and 50% ferrosilicon,
$175.00. All prices are based on short tons. These sales figures were
IX-46
-------
CO
ro
CTl
o
co
vo
CTl
>-
o
.
ci
oc
to
LU
0£.
O— ^"^
LU 4->
Oi C
O)
LU U
LU S-
S.
QC
O
U-
00
to
UJ
ai
s
LU
O.
O
I
X
.a
ro
o
O CO i—
•i— 0) P--
4-J CDCT^
CO ro i—
•i- S- 1
4-> O) CO
ro > vo
4J ro CTl
t/i i —
s_
ro
0)
_
CTl
o
r-»
CTI
CTI
vo
CTI
00
VO
CTi
^
VO
CTl
r—
VO
VO
CTl
1 —
Lf)
VO
CTl
r—
^
VO
CTl
CO
VO
CTl
e
OJ
+J
I-H
o
0
O
O
'
O
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
0
vO
°
0
o
o
o
o
o
o
0
o
o
o
o
CO
OJ
ro
to
LO
VO
CO
oo
oo
f^
CO
VO
Lf)
l€
co
^
co
Lf)
If)
1
o
CTl
Lf)
oo
oo
CM
CTl
•
CO
oo
r-.
s
LO
o
CTl
I--
^-
CTl
O) .
« « > o
•O i 1- 5-
O i. IO O.
CO O) S- 0)
C -I-5 "O
co (U co
"O C7)'i— Cf)
0 EC
O « •!— •!—
CJ> CT) E "O
C -0 3
O 1^ ^J
i— T3 X
+-> O) C O)
co CO ro
o
o
o
CO
^
5:
Ln
CM
CO
CO
^
CM
co
•
^f1
s
^:
CO
•
CO
CTl
CM
CO
CO
VO
CO
CO
VO
«^-
-o
c
ro C
o
C •!-
0 +J
•i- ro
-t-> IM
ra T-
•r^ 4-3
u &-
o
^
VO
vo
r-"
C\J
o
5
o
1— —
"••
f~
CO
ro
o
C
+J
ro
ro
O
•o
ro
T3
CD
>
15
>
a>
s-
o
s,
a.
(U
-a
co
CO
X
ro
S_
QJ
E O
O O
O i—
U
i— -r- O
ro
o
2 .
a> co
CO
ro
•O
O)
•o
T3
ro
T3
W
CO
T3
ro
^5
ro U
S-
ai a)
i: Q.
cu
c +->
•r- ro
co si
cu a.
s_ o
3 S-
O) Q.
•r- CL
M- ro
O O
o .a o
ra
o
CO
co
-------
UJ TO
i— cu
o
UJ T3
s: c
8 -
•z. cu
CU
UJ to
o to
o cu
00 OO
•— o
x 'x
*-* -v=t-
CU
S
ra
c
o
•r—
CL
*r"
u
CU
•o
c
ro
X
S-
CU CU
0 4->
in c T-
CU S- 4-
U- 3
4- 0
o c si
in cu 43
Q- ro
O U-
Ol CU
0 4->
ro r—
t- C •!-
O S- 4-
CU 3
U- 4- 0
o c si
3: cu 43
Q- (O
O U.
s-
CU CU
o +->
ro i—
S- C
C 3
S 4- 0
I/) C si
CU 43
CL
*-*
o r-^ to
in uo i--
r^ r*^ oo
* *
oo r-^
o in en
in oo to
in r-~ to
in oo
r— r—
O t~^ LT>
CM «^- CM
i—i—OO
en* CO
CM en in
1 — OO «£J~
to ^t* oo
CM r^
CO
CU
to c
c o
CU •!-
CL 4->
X ro
CU N
•r~
w 4-*
C S-
(O O
E
T3 ro
i—
o -o
10 C
(0
to
•o c
o o
o •<-
CD 4J
IO
4— *r—
cO O U
V) CU
O) -t-> i.
•— to Q.
ra o cu
I/) O Q
o
S- S-
CU 4-J
4-> C
4- O
=C (J
^
CU O
S_ S-
O 4~^
4— c:
CU O
CO U
CU •*->
-M C
4- 0
=c o
cuo
4? "c
CU O
CQ U
"o
cu +->
4- O
•=c o
cu "o
O 4->
4- C
cu o
CO U
'o
s- s_
CU 4->
£ o
«t o
cu o
s- s-
O 4->
4- C
cu o
CQ U
j:
r— *t COtDOOOCMO
o o to oo oo
• — — • c — in
r^ oo en in r^ oo o
i i — t~~ oo i — oo r^ oo i i
i to CM oo r- • • -i i
oo to
O U ••->
C 3 -i- CL O O
tO +-> 4- (j ,—
•M ro W O 4- C 4-
W1 S- CU O T-
0 CU i— c J=
U CU CL. lOOC-MCO
So 10 +-> o cu ia
r— O — ^ 4J c (J
004- S« b^ -^
S- C O -ta C C
£: &3 CU O CU CU ^
O C7) m E i r~ r~~ ^ ^ ^-
oc ^J-ooooooo
•r- (J i— O O t— •«— -r-
i— 4-> C4-CC4-4J+J
3 S- CU 4T 4C 3 3
ECU X-l->tO4->4-»tOXJT3
co- roairacucuracucu
o
o
o
CM
ca
C
o
o
4T
to
o
to
oo
«* •
O
"O
s: in
•r- P-.
O c
0 s
CM •*•*
C& 4-^
S-
tO O
C 4=
O tO
+-*
o
4-> O
S- O
o •*
43 O
to in
o •
IO •!-
00 C^)
- cu
oo u.
&«
«o
s: in
cu
i i -^
c:
O fO
n:
• A
o
.. O
to .
c to
CU CM
fm fl
to to
c
CD O
C 4->
•r—
^ ^ *
0 S-
•— o
f— 4C
O (/)
4-
o
c o
o en
•o'oo
CU LT)
co
(O A
40 S-
C_)
CO CU
cu u_
•
CM
1
X
H-4
cu
_Q
ro
t
r~
E
o
S-
4-
to
4->
C
cu
E
a.
•r—
3
CT
O)
|H~
O
i-
.JJ
c
o
o
o
4->
cu
^
ro
4->
3
•i—
i-
+J
4->
ra
C
o
4->'
ro
•r—
U
CU
a.
cu
•o
S-
ia
cu
in
^_
CO
CU
•a
3
re o
IX-48
-------
00
•z.
UJ
UJ
g
oo
T3
CU
co
o
g
O T-
O
z: cu
l—t 10
< cu
_i s-.
Q. 4J
o
_l
LU to
O CU
CTl OO
i— O
I I —
X X
jQ
fO
c
0
•4—*
Q.
•i—
U
CU
•o
+J
£
CO
i.
0 •!->
•r- CO r—
OO C •!-
U. 3
4- 0
5~5 *r™
o c s~
m cu .a
O. CO
o u_
s_
cu eu
o +->
—
S- C T-
0 S. 4-
eu zs
U- 4- 0
CJ C S-
ZC CU X)
Q. CO
C? Ll-
j_
eu eu
0 +->
(0 i—
s- £
SI 4^ 0
00 C £
CU -Q
CL co
O U.
4-
cu cu
o +•>
CO r—
C C •!-
s: s- 4-
CU =1
U. 4- (J
0 C E
T- ni r->
Q- CO
O U_
S
+•*
HH
o r^ vo
LO LT) 1--.
r*^. r*1-. co
00 1^-
O LO CTt
in co 10
LO h«- *£>
n «*
in oo
r— r—
o t-^ tn
en ^ en
•N •%
en oo
CM en m
t-^ oo ^a-
X ro
CU N
•a +J
c s-
cO O
s
•a ca
r^
0 T3
CO C
CO
(A
•a c
o o
O -I-
cO
(O O O
v> cu
CU -4-> S-
r— (/I Q.
CO O CU
00 CJ Q
o
i. S.
eu 4-)
4- O
«c o
r
eu o
S- S-
o +->
4- C
cu o
CQ 0
"o
S- i-
eu 4->
4— O
< o
cu o
S_ S-
o •*->
4- C
cu o
ca o
£.
t_ s_
CU 4^
4- 0
4— O
«t o
cu 'o
O 4->
4- C
cu o
OO U
JC.
^->* s-^ ^-^ ^-^ ^-^
cou3 u^i — Ln^csio
** — * *• — * LO h^»
r^. oo cri LO i^ co o
r— r^» ro r— oo r*-. oo
i vo CJ ro p^. • • • i i
1 CO ^O ^J" 1 1
r—
VO ^D ^" VO LO ^J~ VO f"~
r^ c\i ro co LO co co CM <*o o^
LO LO C\J CM O^ • • • • •
r— LO OJ
<£> CO CO OJ CO ** O
O^ O~i CD P"»» CO OJ *kO
O ^ VO OJ • • . 1 |
1 « * CO O i— 1 1
1 i— i— r— OJ
, — ^ s — ^ f — „ . — t „ — ^
(^ c*j ^j* co r*«* LO co ^t~
- — ^^ CM in
oo CM 10 i — r^ «*
i •* en in in oo i — i — i i
1 IO CM OO ^. • . , . | |
oo co r^
1
^ co en en "NJ* r^^ «^i- ^j-
3^ o^ to CM ^. to oo en en f^
r-. in CM oo i—
* i — OO i — VO O
r— i— in 00
co uo oo oo oo 10 m
oo CM 10 o co r>- oo
t oo 10 i~>. 10 • • • i i
1 ** ** CO t*** kD 1 1
r- _ _
pr
O +J
U 0 +->
c 3 o *e
•>- "O 3 fc«
O T3 "
cn i- o eu »
C Q. i. E 2
1 -I— Q. O O
10 •«-> 4- 0 i—
+J <0 10 O 4- C 4-
VI S_ 0) O «r-
o cu r— c x:
o cy a. cooc-uio
C O l/l +-> O CU CO
r— O •->. 4J C O
O O 4- fe« b<> *^
S- C 0 4^ C C
c ^s cu o eu cu «
oc -=1-0000000
•r- C2> O r— O O r- -r- -r-
CO CO CO »i— »r- -t— O O
3 s- eu J= ^ zs S
Co. cocucocueucoeucu
^ O I — ZCJSIZOQiCli
o
o
d
CM
(O,
10
c
o
S-
o
o
vo
c •
•H~ p**
00 i—
CD
O c
0 S
CM -^
S-
10 O
c .c
O 10
o
4-> O
S- 0
o •>
JC O
i/i in
0 «
IO T-
00 00
. cu
CO U-
s^
"O
sr m
cu
u- -a
c
O CO
3:
C3
•• O
to .
C
v> to
1=
a> o
C -4->
•r™
S "^
o s_
i— O
r— .e
O l/>
4-
o
c o
o en
n
•a oo
cu m
to
CO "
-Q S-
O
to cu
eu ti-
i-
co o
00 3C
(0
•
CM
r—
1
X
HH
cu
S
CO
i—
E
O
S-
4-
10
^_>
to
o
J3
4->
C
CU
§.
•1*-
^3
cr
cu
r—
O
.^j
C
o
o
o
•4->
cu
XI
CO
-4-)
3
•r—
S-
Ji
CO
c
o
*^3
CO
•r—
O
CU
L.
O-
ai
•a
-------
then used to derive operating costs and profits based on the historical
percentages in Table IX-17. Tables IX-18 and IX-19 then compare the
"before" and "after" affect of imposing control costs on the five
model plants.
3. Economic Impact on Model Plants
As shown in Table IX-18, which is based on the lower costs developed
in Table IX-12, control costs in the absence of product price increases
will reduce profits and cash flows significantly. Reduction in net earnings
after taxes ranges from 42 percent for HC FeMn to 114 percent for 50% FeSi.
Similarly, reductions in cash flows (i.e., the sum of depreciation charges
and after-tax earnings) range from approximately 20 percent for HC FeMn
to 54 percent for 50% FeSi.
However, if the profit figures are based on costs from Table IX-13
which include retrofit expenses and off-site items, the effect is more
drastic. In Table IX-19, silicomanganese and 50 percent ferrosilicon show
a net loss, while HC FeMn and HC FeCr show 57 percent and 53 percent re-
ductions in net income, respectively. Cash flow reductions range from 25 to
76 percent. Although application of aggregate corporate operating ratios
to the calculation of income statements for individual products may distort
the results for any or all of these products, the interpretation of this
analysis would remain the same with better data; i.e., control costs are
significant relative to the thin profit margins and can only be supported
by very substantial price increases.
4. Economic Impact on the Domestic Ferroalloy Industry
The preceding section concluded that pollution control for the model
plants would have to be supported by price increases to maintain even the
IX-50
-------
current low profit margins. In this section, the total economic impact of
emission control on the domestic ferroalloy industry is discussed. The
ability of the industry to pass control costs along to the user in the form
of higher prices is likewise considered.
The better control systems (baghouses and high-energy scrubbers)
installed by industry in recent years have average costs for amortization
and operation that total $9.30 (1971 dollars) per ton for ferroalloys
having relatively favorable dust-removal properties, such as the products
used for the models. Control costs for high-silicon alloys, which have
difficult dust-removal properties, average $61 per ton of product for
recent installations (see Table IX-20). Emission control costs range from
2 to 9 percent of sales for specific ferroalloys.
The ultimate total cost of emission controls to meet state implementa-
tion plans is estimated at approximately $30 million per year for the 2
million tons of ferroalloys produced. This estimate is based on a product
mix characteristic of recent industry experience and on unit control costs
derived from actual plant visits. These data are shown in Table IX-20.
Since annual costs run approximately 25 to 28 percent of required capital
costs, the ultimate capital costs will approach $120 million.
As of 1971 only about 20 percent of the total production capacity
(based on megawatt ratings) can be regarded as well enough controlled to
meet the state implementation plans. Such plants must include good hooding
enclosure and high collection efficiency of captured pollutants. The
amount of particulate emissions from all ferroalloy furnaces has been
estimated to be 150,000 tons annually. Installation and operation of the
most effective control systems on all furnaces would ultimately decrease
IX-51
-------
Table IX-20. PERCENTAGE OF PRODUCTION AND CONTROL COSTS
Control costs
Product
HC FeMn
SiMn
LC FeMn
HC FeCr
LC FeCr
FeCrSi
Up to 30% Si
50% FeSi
65% FeSi
75% FeSi
85-90% FeSi
Si Metal
MgFeSi
Ca alloys0
All other Si
Totals
Portion of productions3
net tons
348,000
194,000
138,000
148,000
126,000
86,000
124,000
494,000
26,000
106,000
10,000
116,000
58,000
14,000
12,000
2,000,000
Historical ,
$/ton
$ 5.00
9.00
7.15
6.15
-
11.25
8.00
10.00
10.00
20.00
61.00
61.00
10.00
61.00
20.00
-
Recent,
$/ton
$ 8.00
18.00
10.00
12.00
-
15.00
16.00
20.00
20.00
30.00
61.00
61.00
20.00
61.00
20.00
-
Average total
cost,
$
$ 2,262,000
2,619,000
1,183,000
1,343,000
-
1,129,000
1,488,000
7,410,000
390,000
2,650,000
610,000
7,076,000
870,000
854,000
240,000
$30,124,000
Based on Price Waterhouse Statistical Reporting Program, 1971,
adjusted from 1,632,835 net tons to 2,000,000 net tons.
The Ferroalloys Association.
°Excluding calcium carbide production.
IX-52
-------
the estimated emissions to 8,000 to 10,000 tons annually, based on 1971
production.
The domestic ferroalloy industry is faced with serious economic
problems in addition to air pollution control. Slackened demand for
many of the products, increased competition from foreign ferroalloy pro-
ducers, rising electrical power and coal costs, and higher wages have
combined to reduce profit margins significantly. A large proportion of
the industry has shut down in recent years as a result.
Foreign imports, mostly manganese and chromium products, accounted
for 19 percent of domestic consumption in 1971. Imports of silicon
products, historically on the order of 3 to 5 percent of domestic consumption
for this product class, jumped to 10 percent in 1971. Several factors account
for increasing foreign competition. Low transportation costs are available
to foreign producers, as their plants are located near the seaports. Some
countries, particularly in northern Europe, have cheap hydroelectric power.
As power costs vary from 15 to 40 percent of the product manufacturing cost,
depending upon the product, a cost difference of 4 to 5 mils per kilowatt-
hour or greater between U.S. power rates and overseas hydroelectric rates
can amount from $10 for HC FeMn to as high as $70 per ton for silicon, which
offers a considerable cost advantage to the foreign production.
High-quality metallurgical coal and coke, which are used in significant
quantities by the industry, have become expensive in recent years. Of
course, foreign producers have been paying higher prices for these coals,
too. Increased prices of ores, machinery, and parts have also been im-
portant factors in raising manufacturing costs for the domestic industry.
IX-53
-------
There are also qualitative factors that may contribute to the
stagnation of the industry. The expectation of increased costs for
emission control, the difference in government policies among various
countries toward taxation, subsidy payments for promotion of industrial
growth, advances in foreign technology leading to production efficien-
cies, and different latitudes in various countries with respect to
pollution control regulations all cannot be measured adequately in
numeric terms, but are postulated as important factors in the economic
setting of the industry today.
Of the groups analyzed, only silicon and ferrosilicon manufacturers
seem to be in a position to pay for emission control by raising product
prices. The quartz used to produce silicon alloys is plentiful domes-
tically,and demand for silicon products, particularly the silicon metals,
appears strong, as measured by an expected annual growth rate of 4 to 6
percent. Lastly, imports are still less than 10 percent of the market
measured in terms of domestic shipments. Some of the control costs for
silicon products could thus probably be passed on to the consumer. How-
ever, silicon and ferrosilicon manufacturers are also facing the greatest
costs.
Faced with low growth rate in demand by the steel industry for their
products and stiff foreign competition, ferromanganese and ferrochrome
producers will be forced to absorb most of their emission control costs.
Only through increased production efficiency in the form of higher
mechanization and larger plants can ferromanganese and ferrochrome
producers hope to retain current profit margins and simultaneously absorb
control costs. It is likely the smaller and more marginal plants (i.e.,
IX-54
-------
those faced with higher power costs, labor problems, and other cost-
associated problems) may either shut down in the face of expected air
pollution abatement costs, shift to other product lines, install more
efficient manufacturing facilities, or build in other countries which
have more lenient standards and lower costs.
IX-55
-------
X. RECOMMENDED RESEARCH AND DEVELOPMENT PROGRAMS
A. INTRODUCTION
Previous sections of this cooperative study have evaluated known control
technology, control technology now being tested in commercial operations,
and the economic effects of using various methods of emission control.
Present knowledge indicates the various known control methods are not
each broadly applicable to controlling emissions from the manufacture of
all products. Existing controls result in the accumulation of substantial
amounts of wastes in the form of either dusts or drained slurries, without
proper disposal, the dry dusts can be entrained by wind, and the drainage
from slurries can become a water pollution problem. With the advent of
new source performance standards and the increasingly stringent levels of
control generally required under the Clean Air Act Amendments of 1970,
the problem of handling and disposing of collected dusts can become more
acute.
The technology of covered furnaces is applicable to the manufacture
of almost all products except some high-silicon ferroalloys. However,
the use of covered furnaces for different product lines has been noted
to vary to some extent from country to country; in Japan, for example,
covered furnaces have been used to produce silicon alloys containing up
>,
to 75 percent silicon. A worldwfde evaluation should be made of furnace
designs, operating techniques, manufacturing limitations and capabilities,
control technology, and economic factors. The goal of such an evaluation
would be to determine the extent to which covered furnaces can be applied
to the manufacture of high-silicon alloys.
X-l
-------
B. RECOMMENDATIONS
The following recommendations are presented for further research and
development.
1. Process Modifications
a. Determine the relationship of the cost of control to
particle size of emissions. This might include an
investigation of agglomeration techniques by sonic,
electrostatic, or other means that would increase
particle size and, presumably, decrease costs.
b. Examine methods that would decrease the rates of furnace
exhaust gas to be handled by control devices, since
costs normally decrease when the volumes decrease.
c. Examine the economics of raw materials (ores, coals,
etc.) used as ferroalloy charge stocks and their
effects on gaseous and particulate emissions. If
possible, develop data on the economic feasibility
of using raw materials that would result in the
lowest quantity of particulate emissions.
2. Applications of Control Techniques
a. Examine other types of available control techniques
and collection systems besides recognized venturi
or other liquid scrubbing systems, baghouse filters,
and electrostatic precipitators. This would include
examination of all available liquid-contacting
nozzle-ejector systems for effectiveness, reduction
X-2
-------
in costs, and applicability to a broader range of
dusts. Examine combination two-stage systems such
as electrostatic precipitator treatment preceding
a low-pressure venturi scrubber. Examine the
possible domestic application of other systems
developed outside the United States, such as multi-
stage scrubbers and the charging chute (shaft kiln).
b. Examine methods for reducing pressure drop across
baghouse collectors.
c. Develop a fiberglass fabric resistant to higher
temperatures or find other fabrics more suitable to
ferroalloy dust-filtering operations, with emphasis
on extended life of the fabric.
d. Develop a preconditioner additive for electrostatic
precipitators that may be more effective than the
currently used ammonia.
3. Waste Utilization
a. Search for new uses of particulates recovered from various
product lines and by various recovery methods. Uses for dry,
fluffy dusts and the settled-slurry fine particulates
should be included in this search. Examine the chemical
content and physical form of the recovery products and search
for additional uses in agriculture, industry, and other
areas not now in evidence. Examine collected materials
for other elements of possible economic value.
X-3
-------
b. Examine in greater depth the possibility of preparing,
by extrusion or other means, recovered participate for
recycling to the submerged-arc furnace. The problem
of possible wind entrainment is particularly severe
with dry dusts. Certain positive values, particularly
chromium and manganese, are contained in the dusts and
might be recovered for use. Although pelletized chromium-
and manganese-containing dusts may not be fully acceptable
as desirable charge stock, the stockpiling of these dusts
in the pelleted form would prevent their becoming entrained
by wind. Such stockpiles could be an important source
of raw material for furnace feedstock when other supplies
are not available.
4. Waste Heat Utilization
a. Investigate methods for using the sensible heat of
reaction gases as well as heat from the combustion of
carbon monoxide therein.
b. Develop methods for achieving greater utilization of the
heat produced within the furnace itself, possibly
through the use of water walls.
5. Emission Measurements
a. A much higher volume sampler is needed at the collector
outlet to more quickly obtain an adequate weight of
particulate. A sampler should be developed that is simple
to apply and that will accurately and quickly reflect the
results of control measures.
X-4
-------
b. Develop continuous-reading monitors of reliable accuracy
for hour-to-hour measurements of outlet particulate
concentrations.
X-5
-------
APPENDIX A
Description of EPA Source Tests
All the data compiled by EPA tests are a part of this Appendix.
Furnaces A through J were tested in the United States. Furnaces K
and L (ferromanganese and siliconnanganese) were tested in a plant
located in Norway. A brief description of the process and a flow
diagram showing test points are included for each plant tested. Re-
sults of each individual sample analysis from tests made on these
furnaces are shown in Tables A-2 through A-5. The source of the
samples shown in these tables is given in Table A-l. The amount of
submicron particulate matter in the gas stream flowing either to or
from an air pollution control system was obtained from the back half
(impinger catch) of the sampling train; this information is presented
in Tables A-6 through A-ll. Wherever inlets and outlets were tested
simultaneously, the percent collection efficiency (impinger catch only)
is shown for each furnace tested. These efficiency rates indicate
that the three types of control devices (baghouses, scrubbers, and
precipitators) used on these furnaces are effective in removing a high
percentage of submicron particulates.
Table A-l2 shows the actual gas flow rates and temperatures of
the furnace gas streams measured during each individual test run.
A-l
-------
FURNACE A
Ferrochrome Silicon
(Uncontrolled)
The initial atmospheric emission tests
made by the EPA contractor were conducted
May 18 to 19, 1971, on an uncontrolled
ferrochrome silicon furnace. The furnace
has a hood for collecting the furnace fumes,
and it was about 100-percent effective in
capturing the fumes. Two exhaust ducts
connected at opposite sides of the hood are
equipped with blowers that discharge the gas
into the open atmosphere through two stacks
terminating above the building roof. The
tap-hole hood was about 95-percent effective
in capturing generated tapping fumes. This
SAMPLE
POINT 1
Figure A-1. Uncontrolled ferrochrome silicon
-fitcnae.?
hood was vented into the uncontrolled furnace
exhaust duct. Figure A-1 shows the uncontrolled exhaust system and
test points. The charge material to the furnace was a mixture of
chrome ores, quartz, coke, and wood chips. The two exhaust stacks
were sampled simultaneously for approximately two hours to cover the
tapping cycle. Both the EPA train and the ASME particulate train
were used for comparison of results. The two tests with the EPA train
showed emissions to be 197 pounds per hour (0.14 gr/scf) and 438 pounds
per hour (0.32 gr/scf). The higher emissions for the second test were
caused by more furance gas blows than normal because one of the two
stoking machines broke down. However, the quantity of emissions for
A-2
-------
both tests was considered lower than normal for this alloy and has
been attributed to screened ores used in the furnace mixture. The
ASME train showed a slightly higher particulate emission rate than
the EPA train (28 percent for the first test and 2 percent for the
second test). The percentage of particulates collected in the
impingers (back half) of the EPA train was approximately 50 percent
during the first day's test at sample point 1 but varied from two
to five percent in all other samples. The EPA train was not com-
pared with the ASME train when the high fraction of sample was
obtained in the impingers.
The mass media particle size of the fumes emitted varied from
0.62 to 0.67 microns (see Table VI-16 and Appendix D)
Visible emissions ranged from 60 to 100 percent. One of the
two stacks serves not only the furnace hood but also the tapping
station. Consequently, the larger volumes of gases in this stack
dilute the concentration of particulates and results in fewer visible
emissions.
Chemical analysis was made of the exhaust gases coming from the
furnace and the particulates collected on the filter of the particu-
late sampling train. Sulfur dioxide ranged from 11 to 17 parts per
million. Carbon dioxide and carbon monoxide were 0.8 percent and
0 percent, respectively. Chemical constituents of the collected
dust are shown in Table VI-17.
A-3
-------
FURNACE B
Chrome Ore/Lime Melt Furnace
(Uncontrolled)
On May 20, 1971, a chrome ore/lime melt fur-
nace was tested. Dried and screened chrome ore
mixed with lime is charged into an open-arc tilt-
ing electric furnace with a pouring spout at the
top. The furnace is periodically tilted and the
melt falls by gravity into a large ladle,
where subsequent ladle reaction with ferrochrome
silicon produces a low-carbon ferrochrome (see
Flow Diagram, Figure V-3, page V-7). Two test
runs were made with results of 50 pounds per hour
(0.14 gr/scf) and 61 pounds per hour (0.175 gr/
scf). At the same time an ASME sampling train
SAMPLE
POINT
\
TAP
LADLE
TILTING
OPEN-ARC
ELECTRIC
FURNACE
Figure A-2. Uncontrolled chrome
ore/lime melt furnace.
was used with test results of 60.4 pounds per hour and 72.3 pounds per
hour. Capture of the fumes by the hooding and exhaust system was
judged very poor. With the use of a high-volume filter to determine
the particulate concentration and an estimate of the volume of the
uncaptured gases escaping the exhaust system, it was shown that emis-
sions escaping the exhaust system varied from 9 to 65 pounds per
hour.
Chemical analysis was made of the gases coming from the furnace
and the particulates collected on the filter of the particulate sampling
train. The analysis showed no sulfur dioxide or carbon monoxide, and
only 0.1 percent carbon dioxide. Chemical analysis of the collected
particulate sample is shown in Table VI-18.
Particle size was determined by the use of a cascade impactor.
A-4
-------
FURNACE C
Silicomanganese Furnace
(Scrubber)
A series of tests were conducted be-
tween July 27 and August 4, 1971, on a
furnace making silico manganese, 2-percent-
carbon grade. The furnace fumes are con-
trolled by two Research Cottrell flooded
disc scrubbers discharging into a common
stack. Figure A-3 shows the flow diagram
of the furnace and scrubber system. Three
test runs were made at each scrubber pres-
sure drop setting of 57,47 and 37 inches H20
pressure. Figure A-4 shows the collection
efficiency curves as a result of these
tests. The mass emission rates in pounds per hour were 9.8 (0.01
gr/scf), 16.7 (0.02 gr/scf)> and 44 (0.05 gr/scf) for pressure drops
of 57 inches H20, 47 inches H20, and 37 inches H20, respectively. The
percent efficiencies were 99.1 (57 inches H20), 99.1 (47 inches Ufl),
and 96.3 (37 inches H20). The efficiencies were the same for 57 inches*
HpO and 47 inches hUO, but the concentration of dust into the scrubber
when testing at 47 inches HJ) was 60 percent higher. Even though pres-
sure drop varied across the scrubbers, velocities and temperatures
remained relatively constant. The average scrubber inlet particulate
loading was 1355 pounds per hour.
Particulate emission rates were also determined for tapping. Tap-
ping time varied for each tapping test from 28 to 32 minutes. There
Figure A-3. Scrubber system serving Silico-
manganese furnace.
A-5
-------
are two tap holes with exhaust systems, but only one tapping station
was in use. Therefore, it was necessary to test both stacks simul-
taneously to obtain the total tapping emissions. Tapping losses for each
tapping period were 59 pounds (32 minutes), 27 pounds (30 minutes), and 17
pounds (28 minutes). The average tapping losses (34 pounds) exceed the con-
trolled losses (average over three different pressure drops) by 22 pounds.
Gas analysis made at the scrubber outlet showed sulfur dioxide
emission levels generally below 1 ppm. Carbon monoxide was negligible,
and carbon dioxide varied between 2 and 3 percent.
The filter catch of the EPA particulate sampling train was analyzed
for chemical constituents by use of a microscope, qualitative electron-
beam X-ray microanalysis and atomic absorption. The quantities of
sample materials were very small, making analysis difficult. The glass
fiber of the filter became intermixed with the particulate matter im-
bedded in the filter, and no accurate silicon analysis could be made.
Details concerning the specific elements can be found by referring
to Table VI-18.
Several samples were obtained by a cascade impactor to determine
particle size. The mass median diameter (MMD) of the particulates at
the scrubber exhaust varied from about 0.2 to 0.7 micron. The MMD
of particulates from the furnace varied from 0.6 to 5.0 microns.
Because of the short sampling time and the varying loading of the gas
stream from the furnace outlet, the MMD particle size reported may not
reflect a true average particle size from the furnace. However, most
of the samples analyzed showed a MMD of less than 1 micron. Table VI-17
A-6
-------
and Appendix D shows the particle size of samples collected.
No emissions were visible from the scrubber when it was operated
at 57 and 47 inches H?0 pressure drop, but a slight trace of emissions
was reported at 37 inches H20.
A-7
-------
0.08
100
0.07
0.06
0.05
a
to
z
1 0.04
te
a
i—
LLl
t 0.03
0.02
EFFICIENCY
0,01
99
98
I
UJ
o
97
27 37 47
VENTURI PRESSURE DROP, in. H20
Figure A-4. Scrubber efficiency as function of pressure drop.
A-8
57
96
-------
FURNACE D
Ferrochrome Silicon
(Baghouse)
CLEAN AIR TO ATMOSPHERE (
Three test runs were made on Aug-
ust 31 to September 1, 1971, on an air
pollution control system with a 12-
compartment push-through baghouse con-
trolling the fume emissions from a FeCrSi
furnace. An air curtain is provided
around the periphery of the hood over
the furnace with the air provided coming [j\rr
from the exhaust system serving the tap-
ping station hood (see Figure A-5). The
larger particle sizes in the exhaust
system serving the furnace settle out
I 1 II 1 I iTTl TiTTll 1 1 1 I 1 1 1 11 1 I I I
12-COMPARTMENT
BAGHOUSF
SPARK
IRJJESTOJ
AIR
COOLER
Figure A-5. Baghouse serving ferrochrome
silicon furnace.
in the spark arrester. A very small amount settled out in the spark
arrester hopper; at the time of the tests, it had not been emptied
after several months of operation. The effluent gas stream then passes
through two indirect forced draft air coolers, although it had not
been necessary to use these as coolers. The dust-laden gas then is
directed into a 12-compartment baghouse with a total of 1728 (11-1/2
inch X 30 foot) fiberglass bags. Mechanical shakers clean the bags in
each compartment once every 78 minutes.
Three EPA particulate sampling trains were spaced equal distances
apart at the baghouse-roof monitor outlet in order to obtain a represen-
tative sample. Simultaneous sampling with the three samplers showed
that particulate emissions to the atmosphere varied considerably in
A-9
-------
three test runs, ranging from 18.1 to 37.9 Ib/hr with an average
of 30 Ib/hr. The average grain loading was only 0.009 grains per
standard cubic foot. The collection efficiency of the baghouse was
96.5 percent. The percentage of participates collected in the im-
pinger (back half) of the EPA train was high for an unknown reason
and quite consistent for nine samples, ranging only from 63 to 74
percent. The impinger water residue from the sampling train was
analyzed by the Sulfaver-Turbidimetric procedure and found to be
33 percent sulfate ion (SO }. Optical emission spectrography has
shown most of the material in the water residue to be Fe, Na, Ca,
Si, Al, Mg, and K. The sample volume through the sampling train
was 88.72 standard cubic feet of dry gas with an SO concentration
of 0.69 mg/cubic foot. The impinger water thus contained enough
S02 to potentially ionize 91 mg sulfate ion. Total material col-
lected in the impinger weighed 20.4 mg. With the possibility that
particulates may have formed from the reactants in the impinger portion,
the amount of particulate matter reported going into the atmosphere
should be based on the fraction collected in the front half (filter and
probe) of the EPA train. In this case, the emissions for the three test
runs would be 11.0 Ib/hr (0.0035 gr/scf), 9.4 Ib/hr (0.0025 gr/scf), and
5.8 Ib/hr (0.0014 gr/scf), resulting in a collection efficiency of 98.7%.
Particle size determinations were made by use of a cascade im-
pactor. The mass median diameter (MMD) for the baghouse exhaust was
approximately 0.7 to 0.8 microns. The MMD for the furnace exhaust was
0.3 and 3.2 microns during taps and between taps, respectively.
The baghouse exhaust flow rate was too low to measure accurately
with a pitot tube. Because air is induced into the baghouse and
A-10
-------
mixed with the cleaned furnace gases after passing through the cloth
filters, it was necessary to measure the induced air by use of a vane
anemometer. The volume of gases from the baghouse was determined by
adding the measured volume of dirty gases from the furnace going into
the baghouse to the measured volume of induced air. Induced air was
found to be over half the volume leaving the baghouse. These volumes
were verified by use of a heat balance. Another method for determin-
ing the amount of dilution air is measuring the concentration of CO
into and out of the baghouse. The C02 readings into and out of the
baghouse were 1.2 and 0.5 percent, respectively, which also compared
very closely. Including induced air, the baghouse outlet volume was
383,000 SCFM.
The quantity of particulates contained in the air induced into
the bottom of the baghouse was determined by use of an ambient air
3
particulate sampler to be 1400 mg/nM , which means that this con-
centration of dust was added to the cleaned gases exiting the baghouse.
Subtracting the amount of dust in the induced ambient air from the
total outlet emissions would only reduce the emissions by slightly
less than one pound per hour.
The baghouse was in good operating condition during the tests,
and no emissions were visible.
A-ll
-------
FURNACE E
HC Ferrochrome Furnace
(Electrostatic Precipitator)
Three test runs were made on
September 21 to 23, 1971, on a
HC ferrochrome furnace equipped
with an electrostatic precipitator
for controlling furnace fumes. The
precipitator is preceded by a gas
conditioning tower because of the
Figure A-6. Electrostatic precipitator
serving HC 'srrochrome furnace.
high resistivity of the ferroalloy fume. The gas conditioner, similar
to a scrubber tower in construction, removes approximately 40 percent
of the furnace fume. Resistivity of the fumes is reduced by spraying
170 gallons per minute of water into the gas conditioner. Water
adsorbing on the dust particles forms a liquid surface film through
which electrolytic conduction of the accumulated charge can occur.
Since this type of dust does not readily adsorb moisture, a small
amount of ammonia is added to enhance the moisture adsorption capacity.
The dry precipitator consists of three sections in series with discharge
electrodes of negative polarity and positively charged collecting
surfaces at ground potential. The furnace hood is equipped with
water-cooled, vertically operated doors to minimize the excess air
required for effective furnace emission collection. Two tapping holes
120 degrees apart are both vented to the air pollution control system.
The tests showed that the air pollution control system removes
15.4 tons per day of dust and fumes. Average emissions from the pre-
cipitator to the atmosphere were 21 pounds per hour with a concen-
A-12
-------
tration of 0.0183 grains per standard cubic foot. The average inlet
loading to the conditioning tower and precipitator was 1312 Ib/hr and
1.87 gr/scf. The collection efficiency was 98.1 percent. The fraction
of particulates collected in the water impinger section of the EPA train
varied from 8 to 20 percent in the outlet sample and from 0.7 to 3.2
percent in the inlet sample.
Particle size distributions of furnace fumes and precipitator ex-
haust fumes are shown in Table VI-16 and Appendix D. Average size (mass
medium diameter, MMD) of individual particles from the furnace was typically
from one to two microns. Frequently, there was little difference between
the sizes measured before and following the air pollution control system.
The MMD of the fumes at the precipitator outlet varied from 0.38 to 2.54
microns.
Chemical analysis was made of the gas and the filter catch of the
sampling train. The average sulfur dioxide analysis from the furnace
was 8 ppm. Carbon monoxide varied from 200 to 500 ppm. The percent
of carbon dioxide varied from 1.6 to 2.4 percent. The major con-
stituents found on the filter using atomic absorption methods of
analysis were found to be Cr, Mg, Al, and Si02. The major conclusion
is that the sample is a mixture of oxides: Si02, Cr^O.,, MgO, and
A120_. The sum of the percent values, after conversion to equivalent
oxide values, is 84 percent, which indicates adequate closure in the
sense that all the major constituents have been taken into account.
The remaining 16 percent could well be accounted for by water of hy-
dration or by the presence of chlorine, carbon, and titanium. Analysis
of a dust sample collected by the precipitator unexpectedly found 4.22
percent sodium and 5.95 percent potassium (see Table VI-17).
A-13
-------
Chemical analysis made by the company of the collected dust samples
showed a difference between the dust collected in the conditioning
tower and that collected in the electrostatic precipitator. This is
shown in Table A-13. It was speculated the conditioning tower collects
the larger dust particles, which are primarily from the mix charge
materials, while the electrostatic section collects the fumes from
the furnace reactions.
Table A-l TYPICAL DUST AND FUME ANALYSIS FOR
FURNACE E28
1
Product
Cr203
FeO
Si02
A12°3
MgO
CaO
C
Conditioning tower
dust, %
28.3
7.6
10.2
25.7
15.8
3.9
6.0
Precipitator
fume, %
5.4
6.7
24.2
7.1
38.8
12.3
2.3
A-14
-------
FURNACE F
Silicon Furnace
(Baghouse)
A series of three test runs were conducted
on January 17 to 20, 1972, on a three-baghouse
system (3744 bags, 11-1/2" diameter X 30' long)
serving a silicon furnace. Figure A-7 shows a
plan view of this system. Baghouse B was se-
lected for sampling simultaneously with four
EPA particulate samplers as the cost would be
high to test all three parallel-operated bag-
houses. The trains were equally spaced in the
roof monitor. A representative
sample was obtained from all 8 com-
partments by traversing over the
monitor 24 sampling points. The
inlet sample was obtained at the
furnace outlet exhaust duct. The
three tests showed the emissions
from Baghouse B only were 11.27,
HOOD OVER Si
FURNACE
ROOF
MONITOR
TRAVERSE
POINTS
BAGHOUSE B
BAGHOUSE A
SAMPLE POINT
TRAVERSE
POINTS
TOTAL
28 POINTS
SECTION OF
FURNACE
EXHAUST DUCT
Figure A-7. Plan view of baghouse system on
silicon furnace.
13.82, and 10.3 pounds per hour. Based on the measured gas flow rates
into all three baghouses, and assuming all three baghouses were equally
as efficient in particulate removal as Baghouse B, the total amount of
emissions to the atmosphere for each test run would calculate to be
28.6, 30.9, and 23.5 pounds per hour. The concentration of particulate
emissions to the atmosphere from Baghouse B was 0.006, 0.006, and 0.004
grains per cubic foot. The three baghouses collect 14.6 tons per day
A-15
-------
of participates from the submerged-arc furnace producing a silicon pro-
duct. The collection efficiency of the baghouse averaged 98.9 percent
for the three test runs.
Two other particulate samplers were used at the baghouse outlet
for the purpose of comparison. A Boubel high-volume source test sampler
was used adjacent to an EPA sampler, and a high-volume ambient air sampler
was suspended at the outlet of the same baghouse compartment. Comparison
of filter catch only for all three samplers found the EPA train collected
59 percent and 67 percent more than the Boubel and high-volume samplers
respectively, during one test. The other two times, the EPA train was
compared only with the high-volume sampler and was found to be 53 per-
cent and 17 percent higher.
An ASME sampler at the baghouse inlet was compared with the EPA
sampler, and the EPA train collected 19 percent more particulates based
on pounds per hour of emissions. The difference may be attributed to
the fact that the ASME testing was started about halfway through the
EPA test. The ASME train continued to be used for a sampling period
equal to the amount of time that the EPA train was used.
The operating condition of the baghouse sampled appeared good
even though there was a small bag leak in one of the end compartments
during the first two tests. The results of testing reflect this higher
emission. Emissions from the baghouse for test runs one and two were
10 to 35 percent higher than those for test run three. Disregarding
the area containing the leaky bag, the baghouse system appeared to be
capable of reducing particulate emissions to approximately 0.004 gr/scf
or approximately 23 Ib/hr.
A-16
-------
Laboratory analysis of the 12 outlet samples of participates
from the impinger train (percent of total) varied from 22 to 77
percent and averaged half the total weight.
The plan and side view of the baghouse in Figure A-8 shows where
the sampling points were traversed when using four EPA samplers
simultaneously. Each sampler determined the quantity of particulate
emissions from two baghouse compartments.
A-17
-------
SAMPLER i4 REQUIRED)
SAMPLER PLATFORM
SIDE VIEW
/ OPEN GRATING LOCATED
/ AT BOTTOM FOR DILUTION
1 AIR AND WALKWAY
i «rr5r° ° ° °
\ ^T-fro §o
; \oo6
' V3 O O
r S S S S S S S S S S S f S f J
;
J
J
k
2 1
4 3
6 5
8 7
10 9
12 11
14 13
IB 15
17 18
20 19
22 21
24 g 23
1-
"*)
a (
~B > SAMPLING POINTS
-c)
-A
-B
„ ^ ROOF OF
"L BAGHOUSE
-A
-
-A
-B
-C
A'
PLAN VIEW
Figure A-8. Eight-compartment, open-type baghouse
showing sampling points.
A-18
-------
FURNACE G
SiMn Furnace
(Scrubber)
The test conducted on February
1 and 2, 1972, was on an Aeronetics
scrubbing system. As dust-laden gas
(1100-1200 F) flows from the ferro-
alloy furnace, a standard heat-
exchanger (see Figure A-8) transfers
heat from the gas to high-pressure
water that then enters a two-phase
jet nozzle. The pressure and temper-
ature of the scrubbing water entering
CLEAN GAS
TWO-PHASE
fJET NOZZLE
CHEMICAL
ADDITIVES
HEAT
EXCHANGER
SLUDGE
DIRTY GAS
Figure A «). Aeronetics scrubbing system.
the jet nozzle at a rate of 82 gallons per minute averages 320 pounds per
square inch and 375°F. A two-phase mixture (steam and water) occurs
as the high-pressure heated water passes through the jet nozzle which
is located at the inlet of the mixing duct. The mixture thus leaves
the nozzle at high velocity, and as it passes through the long venturi
section, dust-laden gases are intermixed with the moisture. Concurrently,
transfer of momentum of the mixture to the furnace gas stream results
in a pressure rise across the mixing section, which produces the force
to move the fumes from the furnace into the scrubbing system. The tap-
ping hood is served by an exhaust fan that discharges into the top part
of the furnace cover and helps to supply combustion air to the furnace
for the conversion of carbon monoxide to carbon dioxide. The conversion
process in turn provides heat necessary to drive the scrubber exhaust
system. The tapping-hood exhaust system was estimated to be 40-percent
A-I9
-------
effective In capturing tapping fumes.
The emission concentration, including the entire catch of the
EPA sampling train, varied from 0.05 to 0.11 grain per standard
cubic foot and averaged 0.086 gr/scf. Considering only the front
half (probe and filter) of the EPA train, the mass emission rate and
concentration varied respectively from 5.8 to 13.6 pounds per hour
and from 0.04 to 0.10 gr/scf. The gas cleaning efficiency varied
from 92.6 to 97.6 percent. Two of the tests at the control inlet showed
lower-than-normal emission factors for this product; the effeciency
would thus be correspondingly lower. Particle sizing of the particu-
lates in the scrubber outlet and inlet was obtained during the test
by using a Brink cascade impactor. The mass median diameter of the
particulate samples varied from 2.41 to 5.1 in the inlet and from
0.18 to 0.50 micron in the stack outlet.
A-20
-------
FURNACE H
50% FeSi Furnace
(Scrubber)
Tests were made on February 15, 16,
and 17, 1972, on a covered 50 percent
ferrosilicon furnace served by two par-
allel-installed Chemico scrubbers opera-
ting at 80 to 85 inches H20 pressure
drop. The total volume of furnace exhaust
gas is approximately 7000 scfm. Three
test runs were made at the sampling point
located in the common outlet duct of the
BLOWER
5 HOODED
^TAPPING
STATION
BLOWER
Figure A-10. Covered ferrosilicon
furnace with scrubbers.
two scrubbers. Three test runs were also made in the three outlet ducts
of the secondary, uncontrolled exhaust system. The secondary uncontrolled
exhaust system captures the fugitive furnace fumes that escapes from the cover
and discharges them directly to the atmosphere. Also tested was the uncontrolled
tapping station. Figure A-10 shows a schematic diagram of the furnace with the
five exhaust stacks and test points. The blowers on the scrubber exhaust
were injected with kerosene to prevent binding of the rotors. Any resi-
dual kerosene carryover is combusted when flared. All three test runs
at the scrubber outlet were of short duration because the filter in
the sampling train became quickly loaded. The filter location was changed
over to the impinger outlet in the sampling train during test runs 2 and 3
with very little extension of testing time.
The results of the particulate loading in the collected aas before flarina
were 86.1, 11.2, and 8.25 pounds per hour. The corresponding grains per standard
A- 21
-------
cubic foot were 0.856, 0.115, and 0.085. Analysis of the filter catch
showed a high percentage of combustibles that should have been burned
by the flare. When the flare was off, a very distinct emission was
visible. No visible emission occurred with the flare on. An attempt
was made to determine the inlet loading to the scrubbers by collecting
water samples coming out of the scrubbers. The results did not come
close to typical inlet loadings for this size furnace; thus the
efficiency of the scrubber cannot be precisely determined. However,
emission factors indicate that the scrubber is about 98.5 percent
efficient, and the efficiency is even higher if the combustibles are
not included.
Losses from the secondary hooding varied from 136 to 569 pounds per
hour and averaged 342 pounds per hour. Tapping losses averaged 20 pounds
per tap. The average tapping time was 15 minutes.
As an experiment, an IKOR sampler was used for cnmnarisnn of particulate
loading during the samolina of the secondary hnnriinn <=tacks and thp tarmina
35 J
stack. This instrument gives instant readout, and it showed wide
fluctuation of the particul ate loading in the secondary hooding stacks.
Carbon monoxide concentrations appeared stable in the scrubber
exhaust but varied considerably in the fugitive fume exhaust ducts,
ranging from 50 to 75 ppm with peaks up to 130 ppm.
Because the test of the scrubber outlet was somewhat influenced
by kerosene injection into the blower, tests were repeated on July 18 and
19, 1972, without injecting kerosene. Three test runs were made; each lasted
approximately one hour. The results were 3.9, 3.6, and 3.6 pounds per
hour and on a pounds per megawatt basis were 0.09, 0.08, and 0.07.
A-22
-------
FURNACE J
Calcium Carbide
(Scrubber)
Tests were made on February 22, 23,
and 24, 1972, on a scrubber outlet stack,
four fugitive-fume hood stacks, and a
tapping stack all serving a covered cal-
cium carbide furnace. The fume collec-
tion system consists of a pair of
identical Buffalo Forge (centrifugal)
scrubbers, with only one on line and the
other one used as a spare. The tapping
operation is continuous, and the hood
Figure A-l L Covered calcium carbide furnace
with scrubber.
over this area directs all fumes directly to the atmosphere. The molten
product pours directly into molds, then is cooled and dumped from the
molds in an automatic operation. Figure A-11 shows a schematic diagram
of the furnace with the six exhaust stacks and test points.
Three test runs conducted at the scrubber outlet Were very con-
sistent and averaged only one-half pound per hour of particulate emis-
sions for the collected gas. The outlet particulate concentration averaged
0.036 grain per standard cubic foot. Fugitive fumes (fumes uncaptured by the
scrubber exhaust system) amounted to an average of 58 pounds per hour. The
average particulate concentration of these fumes was 0.06 grain/scf.
The average tapping particulate emissions of three test runs was 48
pounds per hour with an average concentration of 0.20 grain/scf.
Samples of the scrubber effluent water showed that an average of 689
A-23
-------
pounds of participate rratter per hour was collected by the scrubber,
indicating a scrubber efficiency of 99.9 percent. No solids measure-
ment was taken on the inlet water.
Flue-gas conditions were stable in the scrubber exhaust, but were
very erratic and unstable at other locations. Carbon monoxide levels
in the fugitive-fume ducts were extremely variable across the area of
the traverse; they ranged from 50 ppm to more than 500 ppm. The levels
of carbon monoxide in the tapping exhaust were more stable, generally
in the range of 35 ppm with occasional peaks up to 150 ppm.
A-24
-------
FURNACE K
Ferromanganese
(Scrubber)
This furnace, rated at 27 MW, was
tested August 12 to 21, 1972, by EPA
personnel in a plant located in Norway.
Figure A-12 shows only one scrubber system,
but an identical system is located on the
opposite side of the furnace. This furnace
operating at its rated loading of 27 mw was
tested under two conditions, first with only
SCRUBBER
one scrubber system in operation, and then WATER
FLARE/
HIP
Figure A-12 .Covered ferromanganese furnace
with sealed electrodes served by three
Venturis in series.
both scrubber systems in operation. The
design is based on the use of water jets only, which eliminates need for the
exhaust fan found in conventional fan-scrubber systems. The last two stages
act as ejectors inducing the movement of gas through the entire exhaust system.
The cleaned gas containing a high percentage of carbon monoxide is either flared
or sold as a fuel to a nearby chemical plant.
Six test runs were conducted with only one scrubber system in operation. The
concentration of particulates for the collected gas in the scrubber outlet
ranged from 0.009 to 0.037 gr/scf and averaged 0.018 gr/scf. The average pounds
per megawatt -hour was 0.031.
Two test runs were conducted with both scrubber systems in operation. The
concentration of particulates in the scrubber outlet was 0.010 and 0.016 and the
average pounds per megawatt -hour was 0.024.
Normally only one scrubber system is used. The company will not
A-25
-------
Install two systems on future furnaces.
Excluding incidents of uncontrolled tapping, no emissions were
visible from the furnace except for a few instances during test runs
1 and 2 when emissions were less than 10 percent opacity from fugitive
fumes.
A-26
-------
FURNACE L
Silicomanganese Furnace
(Scrubber)
This covered furnace equipped with
sealed electrodes is located in Norway.
It is used to make silicomanganese or
ferromanganese. During the tests made by
EPA personnel on August 23 and 24, 1972,
the furnace was making silicomanganese and
operating between 22.5 and 23 megawatts.
The air pollution control system serving
the furnace consists of a two-stage
venturi scrubber, two 200-HP exhaust fans,
2ND
STAGE
VENTURI
Figure A- 1 3, Covered silicomanganese
'furnace witn sealed electrodes served by
two-stage venturi scrubbers.
and two 150-HP booster fans. The pressure drop across the first-
stage venturi is approximately 2 inches f^O and across the second
stage approximately 50 inches I-^O.
Because of operating difficulties in the control system during
the fourth test run, data are limited to the first three test runs.
The particulate concentration for the collected gas in the scrubber-
outlet gas stream ranged from 0.00811 to 0.0113 grain per standard cubic
foot and averaged 0.01 gr/scf (probe and filter catch only). On a
pound-per-megawatt basis the average emission was 0.009. No emissions
were visible from the stack.
A-27
-------
Table A-2. KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
' Furnace
: kilowatts
during
Furnace test
A 9,800
9,300
B 7,200
C 29,000
28,000
20,000
23,000
23,000
28,000
28,000
27,500
27,500
28,000
28,000
27,500
27,500
! ! !
1 ! i
1 Test
Furnace Control point
Product type equipment location
FeCrSi Open None Uncontrolled
stack
Uncontrolled
stack
Uncontrolled
stack
Uncontrolled
stack
Cr ore/ Open None Uncontrolled
lime melt stack
Uncontrolled
stack
SiMn Open Scrubber Outlet
Outlet
Outlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Sample
number
IE
1W
2E
2W
3
5
6
7
8
9
10
11
12
(discarded)
13
14
15
16
17
A-28
-------
Table A-2 (continued). KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
Furnace
kilowatts
during
Furnace test Product
27,500
27,500
28,000
28,000
27,000
27,000
27,500
27,500
28,000
28,000
28,000
28,000
28,000
28,000
D 20,000 FeCrSi
20,000
20,000
20,000
22,000
22,000
22,000
22,000
20,000
i ! 1
Test
Furnace Control point
type equipment location
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
None Tapping
Tapping
Tapping
Tapping
Tapping
Tapping
Open Baghouse Outlet
Outlet
Outlet
Inlet
Outlet
Outlet
Outlet
Inlet
Outlet
Sample
number
18
19
20
21
22
23
24
25
26E
26W
27E
27W
28E
28W
29N
29C
29S
30
31N
31C
31S
32
33N
A-29
-------
Table A-2 (continued). KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
Furnace
kilowatts
during
Firnace test Product
20,000
20,000
20,000
E • 32,000 FeCr (HC)
33,000
33,000
34,000
34,000
34,000
33,000
34,000
34,000
F Silicon
Test
Furnace Control point
type equipment location
Outlet
Outlet
Inlet
Open Precipitator Outlet
Inlet
Inlet
Outlet
Inlet
Inlet
Outlet
Inlet
Inlet
Open Baghouse Inlet
Outlet
Outlet
Outlet
Outlet
Inlet
Outlet
Outlet
Outlet
Outlet
Inlet
Sample
number
33C
33S
34
35
36 E
36W
37
38E
38W
39
40 E
40W
41
42E
42EC
42WC
42W
43
44E
44EC
44WC
44W
45
A-~30
-------
Table A-2 (continued). KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
Furnace
kilowatts
during Furnace Control
lace test Product type equipment
•
i 7,800 SiMn Open Scrubber
7,800
7,800
7,800
7,800
7,800
40,500 FeSi (50%) Covered Scrubber
39,500
39,500
50,200 None
50,200
48,700
48,700
« 01
i
Test
point
location
Inlet
Outlet
Outlet
Outlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Outlet
Outlet
Outlet
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Sample
number
46
47E
47EC
47WC
47W
48
49
50
51
52
53
54
55
56
57N
57S
58N
58C
-------
Table A-2 (continued). KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
Furnace
kilowatts
during
Furnace test Product
48,700
52,300
52,300
52,300
50,000
50,000
50,000
40,500
39,500
42,000
J 24,000 CaC2
23,800
Test
Furnace Control point
type equipment location
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
None Tapping
Tapping
Tapping
Covered Scrubber Outlet
Outlet
Sample
number
58S
59N
59C
59S
60N
60C
60S
61
62
63
64
65
A-32
-------
Table A-2 (continued). KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
Firnace
Furnace
kilowatts
during
test
Product
Furnace
type
Control
equipment
Test
point
location
Sample
number
23,500
22,800
22,800
22,800
23,000
23,000
23,000
23,000
21,500
21,500
None
A-33
Outlet
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
66
67NE
67NW
67SW
68NE
68NW
68SE
68SW
69NE
69NW
-------
Table A-2 (continued). KEY TO SAMPLE NUMBERS FOR TABLES A-3 THROUGH A-12
Furnace
Furnace
kilowatts
during
test
Product
Furnace
type
Control
equipment
Test
point
location
Sample
number
21,500
21,500
24,000
23,800
23,800
None
Fugitive
fume
Exhaust
ducts
Fugitive
fume
Exhaust
ducts
Tapping
Tapping
Tapping
69SE
69SW
70
71
72
A-34
-------
Table A-3.
['ARTICULATE EMISSION CONCENTRATIONS AND RATES FROM UNCONTROLLED TEST POINTS
urnace
A
B
C
D
[
F
G
H
J
Sample
nuniier
IE
1W
2E
2W
3
4
9
11
13
15
17
19
21
23
25
30
32
34
36E
Probe and
cyclone
gr/scf j
0.052
0.023
0.069
0.046
0.117
0.137
0.180
0.386
0.98
0.98
0.99
1.25
0.17
0.56
0.23
0.13
0.06
0.06
0.48
36W j 0.59
38E
38W
40E
40W
41
43
45
46
48
50
52
0.40
0.39
0.43
0.35
0.085
0.052
0.085
0.078
0.47
0.31
0.42
Ib/hr
30.1
18.5
42.7
34.3
41.8
47.6
161.0
319.0
439.0
823.0
875.0
1123.0
134.0
407.0
189.0
199.0
81.5
84.7
322.0
415.0
292.0
253.0
315.0
253.0
297.0
193.0
286.0
266.0
65.0
43.7
59.5
Probe,
and
gr/scf
0.220
0.047
0.376
0.249
0.141
0.175
1.05
0,423
2.22
1.92
2.10
2.06
0.965
1.66
1.91
0.66
0.12
0.40
0.892
1.09
0.819
0.835
1.107
0.782
0.815
0.538
0.742
0.723
2.15
1.38
1.33
cyclone,
filter
Ib/hr
126
37
232
188
50
61
940.4
349.8
1918.0
1611.0
1857.0
1846.0
752.6
1209.0
1553.0
996.4
173.2
594.4
597.0
761.0
600.0
538.0
811 .0
573.0
2840.0
1980.0
2490.0
2470.0
296.1
194.8
188.1
Impi nger
gr/scf
0.005
0.040
0.012
0.014
0.014
0.003
0.01
0.04
0.01
0.03
0.02
0.01
0.02
0.03
0.01
0.79
0.40
0.59
0.03
0.02
0.01
0.01
0.01
0.01
0.009
0.007
0.004
0.025
0.04
0.02
0.02
Ib/hr
3
31
8
10
5
1
8.9
33.9
8.0
25.0
18.0
10.0
13.3
22.0
9.0
1194.6
583.5
882.9
20.0
16.0
6.0
6.0
6.0
5.0
30.0
30.0
20.0
90.0
5.5
2.8
2.8
Total per
sample point
(includes water
: residue)
; gr/scf : Ib/hr
i
' 0.225 129
0.087 , 68
| 0.388 ' 240
, 0.263 198
, 0.154 55
0.178 62
1.06 949.3
i
; 0.464 383.7
' 2.23 1926.0
1.95 1636.0
2.12 1875.0
i 2.07 1856.0
0.982 765.9
1.690 1231.0
1.920 1562.0
1.448 2191.0
0.520 75t>.7
0.990 1477.3
0.922 617.0
i 1.113 777.0
, 0.827 606.0
0.845 . 544.0
1.114 817.0
0.789 578.0
!
0.824 2870.0
0.545 2010.0
0.746 2510.0
0.748 2560.0 i
2.19 301.6
1 .40 197.6
•1.35 190.9
Total from
furnace
gr/scf
0.15
0.29
0.15
0.18
1.06
0.46
2.23
1.95
2.12
2.07
0.98
1.69
1.92
1.45
0.52
0.99
1.07
0.84
0.95
0.82
0.55
0.75
0.75
2.19
1.40
1.35
Ib/hr
197
438
55
62
949
384
1926
1636
1875
1856
766
1231
1562
2191b
757b
1477b
1394
1150
1395
2870
2010
2510
2560
302
198
191
Estimated
emissions
not
captured,
percent
0.5
0.5
53
52
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5b
0.5b
0.5b
0.5
0.5
0.5
5
5
5
5
2
2
2
9C
6C
Does not include tapping fume losses unless otherwise indicated.
Includes tapping fura losses.
cNo inlet sarples obtained with EPA sampler. Percent loss based on emission factor and samples obtained of
fugitive fu ,e hood.
A-35
-------
Table A-4. r.wr:-." ire EMISSION C(«CEKTRATIO:IS KIO RATES TO ATMOSPHERE FROM CONTROLLED TEST POINTS
n
c
D
E
r
c
H
J
Sairple
5
C
7
8
10
1?
14
16
18
?0
??
21
29H
29C
MS
31H
3K
31S
3X1
33C
335
35
37
39
42E
42EC
42WC
42W
44E
4irr
44UC
44U
47E
47EC
47WC
47W
49
SI
S3
S4
55
S6
54
es
66
Scrubber 57" PD
Scrubber 57" PU
Probe
0.010
0.007
Scrubber 57" PD 0.011
Scrubber 57" PD 0.003
Scrubber 57" PD 0.009
i
Scrubber 47" PD
Sarplc
Scrubber 47" PD . 0.015
Scrubber 47" PO O.Ola
Strui,b»r 17" PD
0.011
Scr..bl'er 37" PO 0.013«
Tr.rubber 37" PD 0.037
Scrubber 37" PD
Ba'jhousc*
Gaghouse3
Bjghouse*
Baghouse*
Baghousea
Baghouse*
Bd^ltouse*
Eaghouse
Bughouse*
PreclpHator
Precipitator
PrecipHator
8aghouseb
Baghouse*
Baghouse
Baghouse
Baghouse
P-jllOtKP1"
Baghouse
Baghouse
Ba;houseb
Baghouse
Baghouse
Hajhouse1'
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
Scrubber
O.OCO
0.0035
0.0042
0.0023
0.0038
0.0028
0.0020
0.0023
0.0014
0.0016
0.029
0.009
0.010
0.007
0.0032
0.0015
O.OOIS
0.0067
0.0014
0.0023
0.0014
0.0026
0.0018
0.0017
0.0015
0.040
0.093
0.102
0.668
and filter
| Ib/hr
10.0
6.8
11.2
8.1
8.4
discarded
15.2
17.7
11.8
13.5
35.4
70.6
11 .5
13.8
7.6
12.5
9.2
6.6
7.6
4.6
5.3
36.3
12.5
14.2
4.2
1.9
0.9
0.9
5.2
1.0
1.4
0.9
1.5
1.1
1.0
0.9
S.8
12.4
13.6
67.0
Inplnger
gr/scf
0.007
0.004
0.001
0.002
0.002
0.002
0.002
0 002
0.0006
0 008
0.005
0.008
0.010
0.007
0.008
0.007
0.006
0.004
0.004
0.003
0.002
0.003
0.003
0.007
0.002
0.0005
0.0001
0.004
0.003
0.002
0.003
0.001
0.004
0.002
0.002
0.009
0.008
0.005
0.188
Ib/hr
6.1
3.7
1.6
1.6
l.S
2.0
1.7
1.8
0.6
7.3
4.5
28.3
30.5
21.9
27.2
23.0
19.0
12.8
11.5
12.4
3.1
3.2
3.0
2.4
0.6
0.2
0.1
1.6
1.3
1.1
1.
0.6
2.4
1.3
1.4
1.3
1.0
0.6
19.0
! Total per
sample point
, (Includes water
residue)
gr/scf
0.017
0.011
0.012
0.010
0.011
-•
0.017
0.020
" 0.013 '
0.0144
0.045
0.08S
0.012
0.014
0.009
0.01?
0.010
0.008
0.006
0.005
0.005
0.031
0.012
0.013
0.014
0.005
0.002
0.002
0.011
0.004
0.004
0.004
0.004
0.006
0.004
0.004 ,
0.049
0.101
0.107
0.856
0.115
0.085
0.043
0.031
0.034
Ib/hr
It.l
10. i
12.8
9.7
9.9
"
17.2
19.4
13.6
14.1
42.7
75.1
39.8
44.3
29.$
39.7
32.2
25.6
20.4
16.1
17.7
39.4
15.7
17.2
6.6
2.5
1.1
1.0
6.8
2.3
2.5
2.3
2.1
3.5
2.3
2.3
7.1
13.4
14.2
86.1
11.2
8.3
0.6
0.4
0.5
i
Totll Into
atmosphere
«r/$cf
0.02 '
0.01
0.01
0.01
0.01
0.02
0.02
0.01 '
0.01
0.05
0.09
0.01
0.01
0.005
0.03
0.01
0.01
0.01
0.01
0.01
0.05
0.10
0.11
0.86
0.12
0.09
0.04
0.03
0.03
Ib/hr
16.1
Efficiency
10.5
12.8
9.7
9.9
17.2
19.4
13.6
14.1
42.7
75.1
37.9
32.5
18.1
39.4
15.7
17.2
11.2
13.9
99.0
97.5
99.0
99.0
99.4
98.1
96.4
95.2
98.0
95.0
98.9
97.2
98.6
98.8
99.1
98.5
10.2
7.1
13.4
14.2
86.1
11.2
8.3
0.6
0.4
0.5
99.1
97.6
93.2
92.6
'three samplers used during each run designated N, C, and S. Total enfs:fons are based on results of each sanpl*.
Total for ejch rjn i; the average of tt, C, and S.
brour samplers used d^ing p,ich run designated E. EC. WC, and V. Ea:h s*r.pler covers emissions from two compartmen
Total emissions for each tcit run are the su-*i'3tion of the result* of E. EC, KC, and U.
A-36
-------
Table A-5. PARTICIPATE LOSSES FROM THE FUGITIVE FUME HOOD*
1
Sample
Furnace number
H 57N
57C
57S
58N
58C
58S
59N
59C
59S
SON
60C
60S
0 67NE
67NW
67SE
67SW
68NE
68NU
. 68SE
68SW
69NE
69NW
69SE
69SW
I
; Probe and filter
; gr/scf
0.095
Sample
0.110
0.423
0.337
0.565
0.054
0.358
0.740
0.078
0.077
0.163
0.161
0.037
Sample
0.050
0.079
0.018
0.046
0.029
0.079
0.021
0.037
0.039
|
• Ib/hr
40.6
rejected -
! 45.5
i 186.1
: 157.7
' 225.4
22.3
159.9
275.5
36.0
36.8
63.3
15.8
20.4
rejected -
29.2
8.1
9.5
13.2
15.7
: 9.1
10.8
9.9
20.1
Impinger
gr/scf
0.008
Ib/hr
3.6
Total
sample
gr/scf
0.103
per
point
i
! Ib/hr
1
! 44.2
glass probe broke ;
0.015
0.028
0.016
0.035
0.061
0.064
0.014
0.033
0.031
0.008
0.003
0.004
6.0
12.2
7.2
13.9
25.1
28.6
5.1
15.5
14.4
3.1
0.4
2.1
0.125
0.451
0.353
0.600
0.115
0.422
0.754
0.111
0.108
0.171
0.164
0.041
51.5
198.3
164.9
, 239.3
47.4
188.5
280.6
j 51.5
I 51.2
! 66.4
I
: 16.2
, 22.5
| Total Into
: , atmosphere
!
gr/scf i Ib/hr
0.11 293.06
.
i
0.47 602.5
j
i
!
' 0.43 516.5
j
I
0.13 169.1
0.09 69. 0C
glass probe broke
0.002 '
0.007
0.001
0.005
0.015
0.007
0.003
0.001
0.030
1.1
0.8
0.7
1
1.4
8.5
0.7
1.7
0.5
15.7
0.052
0.086
0.019
0.051
0.044
0.086
0.024
0.038
0.069
30.3
8.9
10.2
' 14.6
24.2
9.8
12.5
10.4
35.8
i
0.05 64.0
0.05 68.5
Fumes not captured by air pollution scrubber system.
Fugitive fume hood discharges uncontrolled fumes to atmosphere through three stacks. Total
Into atmosphere is summation of samples N, C, and S. Total of 293 Ib/hr would be higher If sample
57C were added.
Fugitive fume hood discharges uncontrolled fumes to atmosphere through four stacks. Total Into
atmosphere is summation of samples NE, NW, SE, and SW. Total of 69 Ib/hr would be higher if sample
67SE were added.
A-37
-------
CL.
CL
UJ
a:
UJ
o.
oo
<
o
CO
LU
to
o
o
s
OO
oo
«aC
CJ
h~4
Qi
ai
S
IO
>
o
i— dl
0
O1 IO
a. 3
•2 E
i— • S»
IO
s- c
Q-'o
a.
10 at
o "a.
IO
to
to
cn
c
•r™
a.
E
a;
c
o $-
O 4-*
""E
QJ *^O
O 'O
&.
a.
Q.
lO
•u
•v.
XI
M-
0
-v.
i-
CT)
CL
IO
4-*
"v.
Xt
(J
01
"v.
01
Q.
IO
4-3
XI
1
u
01
t
o.
IO
4-*
"V.
Xt
q-
0
01
CO
O) S-
i—
u
IO
c
S-
u.
IO
en
•
co
LO
00
o
o
rv CM co
r— Pv CO
r> o co
CO VO CM
0 r— 0
O O O
o o o
CM VO CO
o cn o
r— CM ro
O CO 0
o o o
000
LO IO LO
, (v rv
i — CO
IO CO LO
CO CM CM
O r— O
O O O
UJ 3 UJ
vo vo rv
CM CM CM
o
IO
rv
•
CM
cn
o
o
**
co
o
o
o
0
VO
LO
CM
CM
O
0
00
CM
r—
co
o
o
3
CM
(O
CM
•
J±
CM
CO
O
O
IV. LO
co co
r— CO
CO CO
O O
O 0
s °
o •*
CO lO
0 r-
O O
O 0
cn ^t-
rv *)•
co rv
CM i —
0 0
o o
LU 2
CO 60
CM CM
O
CM
fx.
co
CO
•
o
o
cn
CM
rv
co
CO
0
o
LO
o
VO
o
o
LO
co
CM
CM
CO
O
r—
s.
CO
CM
.
o
1"
o
CM
rv
CO
CM
o
o
LO
>JO
,
co
o
o
en
CO
*~~
<£>
O
CM
O
CM
VO
r—
»
CO
r™"
IT)
ro
*
o
r—
CO
r—
,
LO
CO
o
0
CM
0
,3.
o
o
0
cn
rv
*""
rv
CO
o
CO
VO
Q
VO
•
CO
LO
r— •
r—
CM
.
O
XI
CO
LO
,_
r—
CM
O
n
O
CM
,_
,-j.
0
o
0
XI
CM
LO
O
CM
O
0
(-3
XI
CM
•
0
LO
CM
O
CM
•
O
XI
CM
O
LO
CM
0
CM
O
XI
o
0
CO
CM
o
0
XI
CM
1^.
"*"
o
f™-
0
r~
r*^
jQ
CNJ
•
LO
«5f
f—.
cn
t~«
•
O
XI
CM
LO
"*
,__
cn
o
XI
s
o
CO
o
o
o
XI
q_
**
CO
CO
o
CM
rv.
c
o
cn
c
a.
CL
A3
4->
01
O
to
3
O
o
u
cn 01
c c
«r— T—
> Q.
S- o.
01 10
01 4J
^e. 01
U T-
IO 01
4J 10
to xt
cn s_
c x:
•r- v^
CL XI
CL r—
10
3 g
o
M- -r-
O >
(= W
E •<—
3 E
OO UJ
IO XI
A-38
-------
Table A-7. COLLECTION EFFICIENCY ( IMPINGER SECTION ONLY)3
OF SCRUBBER SERVING FURNACE C
Scrubber
pressure
drop,
nches H20
57
47
37
Sample
point
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Sample
No.
9
8
5
7
8
9
10
11
12
Impinger
catch, %
1
16
9
15
1.5
12
1
9
0.5
13
1.7
4
1.8
17
0.5
6
Total
parti culates,
Ib/hr
949
9.7
384
9.9
1636
17.2
1875
19.4
1856
13.6
766
14.1
1231
42.7
1562
75.1
Parti cul ate
from impinger
section only,
Ib/hr
9.5
1.6
34.5
1.5
24.5
2.1
18.8
1.7
9.3
1.8
13.0
0.6
22.0
7.2
7.8
4.5
Collection
efficiency
(back half
only), %
83
96
91
91
81
95
68
42
Efficiencies shown are calculated from samples collected in the impinger section
(back half only) of the EPA sampling train and were based on inlet and outlet
samples of the air pollution control device.
A-39
-------
Table A-8. COLLECTION EFFICIENCY (IMPINGER SECTION ONLY)
OF BAGHOUSE SERVING FURNACE'U
Sample
point
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Sample
no.
a
b
a
c
a
d
Impinger
catch, %
40
71
40
71
40
68
Total
particulates,
Ib/hr
827a
k
38b
827a
32. 5C
827a
H
18°
Particulates
from impinger
section only,
Ib/hr
330
27
330
23
330
12
Collection
efficiency
(back half
only) %
92
93
96
Average of samples 30, 32, and 34 obtained on 8-31 and 9-1-71
^Average of samples 29N, 29C, and 29S obtained on 8-31-71.
cAverage of samples 31N, 31C, and 31S obtained on 9-1-71.
Average of samples 33N, 33C, and 33S obtained on 9-1-71.
A-40
-------
Table A-9. COLLECTION EFFICIENCY (IMPINGER SECTION ONLY) OF PRECIPITATOR SERVING FURNACE E
Sample
point
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Sample
No.
36
35
38
37
40
39
Impinger
catch, %
2a
8
i.oa
20
0.8
17
Total
parti culates,
Ib/hr
1394
39.4
1150
15.7
1395
17.2
Parti culates
from impinger
section only,
Ib/hr
27.9
3.1
11.5
3.1
11.1
2.9
Collection
efficiency
(back half
only), %
89
73
74
Average of two test points,
A-41
-------
Table A-10. COLLECTION EFFICIENCY (IMPINGER SECTION ONLY) OF BAGHOUSE SERVING FURNACE F
Sample
point
Inlet
• Outlet
Inlet
Outlet
Inlet
Outlet
Sample
No.
41
42
43
44
46
47
Irnpinger
catch, %
1.1
22
1.3
46
3.4
54
Total
parti culates
Ib/hr
2870
29
2010
28
2560
23
Particulates
from impinger
section only,
Ib/hr
31.6
6.4
26.1
12.9
87.0
12.4
Collection
efficiency
(back half
only), %
80
51
86
A-42
-------
Table A-11. COLLECTION EFFICIENCY (IMPINGER SECTION ONLY)
OF SCRUBBER SERVING FURNACE G
Sample
point
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Sample
No.
48
49
50
51
52
53
Impinger
catch, %
1.8
17.6
1.4
7.7
1.5
4.6
Total
parti culates,
Ib/hr
301.6
7.08
197.6
13.42
190.9
14.21
Parti culates
from impinger
section only,
Ib/hr
5.40
1.25
2.76
1.03
2.75
0.65
Collection
efficiency
(back half
only), %
77
63
77
A-43
-------
Table A-12. PARTICULATE EMISSIONS (IMPINGER SECTION ONLY) FROM FURNACE Ha
?un no.
1
2
3
4
1
2
3
1
2
3
Impinger
catch, %
8.1
11.6
6.1
4.3
5.8
52.9
15.1
1.8
30.0
28.1
4.7
4.1
28.0
1.3
22
41
Parti culates to
atmosphere based
total catch,
Ib/hr
44.2
51.5
198.3
164.9
239.3
47.4
188.5
280.6
51.5
51.25
66.4
105.2
76.6
90.7
86.1
11. 2C
8.25C
Parti culates to
atmosphere based on
impinger section only,
Ib/hr
3.6
6.0
12.2
7.2
13.9
25.1
28.6
5.1
15.5
14.4
3.1
4.3
21.4
1.1
19.0
--
--
Source of
parti culates
Mix seals
Mix seals
Mix seals
Mix seals
Tapping
Tapping
Tapping
Scrubber exhaust
'inlet to scrubber not sampled.
'Three exhaust ducts vent mix seal fumes directly to atmosphere.
"After run No. 1, filter relocated in test train after silica gel impinger.
A-44
-------
Table A-13. FURNACE GAS VOLUMES
Product
FeCrSi
FuCrSi
Cr ore lime melt
Cr ore lime melt
SiMn
SiMn
SiMn
SiMn
SiMn
SiMn
SiMn
SiMn
SiMn
FeCrSi
FeCrSi
FeCrSi
FeCr (HC)
FeCr (HC)
FeCr (HC)
Si
Si
Si
Si
SiMn
SiMn
SiMn
FeSi (50%)
FeSi (BOX)
FeSi (50%)
CaC2
CaC2
CaC2
Average
megawatts
(test period)
9.8
9.3
7.2
7.2
23
28
27.5
28
27.5
27.5
28
27
27.5
20
22
20
33
34
34
17
17
17
17
7.8
7.8
7.8
42
41
45
24
23.8
23.5
Type of
furnace,
0-open
C-closed
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C
c
c
c
c
c
Temperature,
°F
280
261
123
154
599
674
607
631
633
629
655
654
541
331
323
336
460
470
445
300
300
310
307
1133
1100
1106
a
a
a
a
a
a
Gas
volume,
acfm
196,000
202,000
47,000
49,000
224,000
203,000
215,000
216,000
230,000
232,000
205,000
199,000
192,000
273,000
271 ,000
282,000
295,000
297,000
310,000
595,000
630,000
590,000
590,000
51,300
51 ,400
51.700
7,600b
7,600b
7,600b
1,443b
1 ,443b
1 ,443b
a i«o. test conducted at inlet.
Volumes (scfm) based on outlet test.
A-45
-------
APPENDIX B
SAMPLING AND ANALYTICAL TECHNIQUES
The procedures and equipment used by EPA personnel or by contractors
working for EPA to measure particulate and other emissions are described in
the appendix of the Federal Register, 42CFR Part 466, Proposed Standards
of Performance for New Stationary Sources (Vol. 36, No. 159, August 17,
1971). The applicable test methods are reprinted below.
A suitable sampling site and the required number of traverse points
were determined according to Method 1. The volumetric flow rate of the
total effluent was obtained by using Method 2. For each run, the average
concentration of particulate matter was determined by using Method 5.
A limited number of tests were also made for the determination of sulfur
dioxide by Method 6.
Because of the configuration of ducting at certain sample points, it
was sometimes necessary to deviate from the procedures set forth in
Method 1. The sampling points in these situations were located in
the only possible or usable places, but care was taken that representa-
tive samples were nevertheless obtained.
Sample ports and procedures for all tests were approved by the EPA
project officer in charge of testing.
FEDERAL REGISTER, VOL 36, NO. 159—TUESDAY, AUGUST 17, 1971—TEST METHODS
METHOD I-SAMPU: AND VELOCITY TRAVERSES d h exttaotlon of a representative
FOR STATIONARY SOURCES sample
1. Principle and applicability. 1.2 Applicability. This method should be
1.1 Principle. A sampling site and the applied only when specified by the test pro-
number of traverse points are selected to cedures for determining compliance with
B-l
-------
PROPOSED HULE WAKING
o
RJ
cr
TJ .
> 2
frt
||
If! 0>
O3 O
"5 T)
gg
ii
^ +*
o w
c "
o c
S°
0) Q.
W ffl
w 52
*^ QJ
S >
o 2
to •
3 W
£* 0)
u- ro
LI
"iJl!yHl *
|I8
s" ^w«
£ a$
°»
ig^ll
«
-------
PROPOSED RULE MAKING
....
Trns!i"ai?rrii
•> 2. « g S? S « I »
'^sss^sss.!'
-
»!§§
n«» >,
« SH
ff 2 a
S8-?
3«j{
la
.
0. VI
fc •-
•5 I
11
s
CO CO r^ h*
ro to oo*
S
U) u>
in oo
o> o>
». r: to to CM co <*•
r- T-* CO IO CO 00 (7)
t> i/i in M u>
<» c> «s in in
^
o
«H
DO
S
00
bo
c
t
CO
o
•o
o
^ S
wi
^"i
op
. . Oi
ter. In: 1971 Book
23. Philadelphia, 1!
1928-71.
Sigs
•dm . «
M -*S tfl O
(S
1
00
d -a ^
si!
3 3 £P
K S w
£+3 a>
CO Q
CO
O
s
s
ERMINATION OF E
2 — DET
Q
O
S
13
S
H
i
8
s
a
XOCITY
S
iS
•°s
.5 ^
+3 aj
red applicability.
. Stack gas veloc;
the gas density
o
u
IS.
o
» a
£•§
S-
S£S
£ B
n G
S
«
«
MS
1
•o
{3
V
a
>.
1
V
F*
V
J3
+j
0
S
11)
£
O)
P
meas
^
1
helbe or reverse t;
Stausc
to
a> i c;
rj o iS
"^ t. S
13 O.|t
°^8
s?s
a
a
1
„•
lity. This method
en specified by th<
termining complli
[ormance Standard
s§«
a»>.
g-cS
0.0
^•o S
w ;3 ;3
^fl« 0.-0
X ^J p, qj
E-l +J ed W
fi
O
t>
^5
O
to
S
o
W
s
3
+j
S
ft
oS
fi
m
+j
G
C
O
O
z
•o"
(*)
o
>
t£
Ul
K
O
Ul
tt
_i
<
QC
Q
IX.
&S
0) _J3
25
B-3
-------
PROPOSED RULE MAKING
43 OaVrtlivte the pitot tube coefficient then the other pointed down-trcnm. Ufe the
pitot tube only it the two cc efficients differ
by no more than 001.
5. Calculations.
Use Equation 2 -2 to calculate the stack gas
using Equation 2-1
f __<"•
Cl""l~ '•li
equation 2-1
where:
C,,,.t = P!tot tub* coefficient of Type 5
pitot tube.
C»,,j = Pltc units
•<»„„_ ^lb moi^-R^ are used.
Cp-Pitot tube coefficient, (limensionlrss.
T, = Ahtcilulf M:ick r,es temperature, °R.
Ap=Vrlodty head of stack pas, In HiO (sec fig. 2-2).
1',= Abwlutc flock pivs prc^tire. In Up.
M.-
uliir wt
.
t of stack giis, Ib./lfo -mole,
PLAI
DAT
RUN
STA<
BAR
STAT
OPER
MT
E r-
NO.
:K DIAMETER. In.
D.Y:TRIC PRESSURE, in. Hg.
1C PRESSURE IN STACK (P0). In. Hg.
ATORS
SCHEMATIC OF STACK
Traverse point
number
Velocity head,
in. H2O
AVERAGE:
vs;
*
CROSS SECTION
Stack Temperature
iy °F
Figure 2-2 shows a cample recording r.heet
for velocity traverse data. U-c the ftverar<*s In
the last two columns of Figure 2 2 to deter-
mine the average stack evs velocity from
Equation 2-2.
6. References.
Mark, L. S. Mechanical Engineers' Hand-
book. McGraw-Hill Book Co., Inc., New York,
1951.
Perry, J H. Chemical Engineers' Handbook.
McGraw-Hill Book Co., Inc., New York, 1060.
Shlgehara, R. T.. W. P. Todd, and W. S.
Smith. Significance of Errors In Stack Sam-
pling Measurements. Paper presented at the
Annual Meeting of the Air Pollution Control
Association, St. Louis, Mo., June 14-19, 1970.
Standard Method for Sampling Stacks for
Paniculate Matter. In: 1971 Book of ASTM
standards. Part 23. Philadelphia, 1971. ASTM
Designation D-2928-71.
Vcnnard, J. K. Elementary Fluid Mechanics.
John Wiley and Sons, Inc., New York, 1947.
METHOD 3 GAS ANALYSIS FOtt CAIU1ON DIOXIDE,
EXCESS AIR, AND DRT MOLECITLAR WKIGHT
1. Principle and applicability.
1.1 Principle. An integrated or grab gas
sample is extracted from a sampling point
and analyzed for its components using an
Orsat analyzer.
1.2 Applicability. This method should be
applied only when specified by the test pro-
cedures for determining compliance with New
Source Performance Standards.
2. Apparatus.
21 Grab sample (Figure 3-1).
2.1.1 Probe—Stainless steel or Pyrex1
glass, equipped with a filter to remove par-
ttculate matter.
2.1.2 Pump—One-way squeeze bulb, or
equivalent, to transport gas sample to ana-
lyzer.
2.2 Integrated sample (Figure 3-2).
2.2.1 Probe—Stainless steel or Pyrex'
glass equipped with a filter to remove par-
ticulate matter.
2.2.2 Air-cooled condenser—To remove
any excess moisture.
2 2.3 Needle valve—To adjust flow rate.
2 2.4 Pump—Leak-frie, diaphragm type,
or equivalent, to pull gas.
2.2.5 Rate meter—To measure a flow range
from 0 to 0 035 c.f jn.
2.2.6 Flexible bag—Tedlar," or equivalent,
with ft capacity of 2 to 3 cu. ft. Leak test the
bag in the laboratory before using.
2.2.7 Pitot tube—Type S, or equivalent,
attached to the probe so that the sampling
flow rate can be regulated proportional to the
stack gas velocity when velocity Is varying
with time or a sample traverse is conducted.
2.3 Analysis.
2.3.1 Orsat analyzer, or equivalent.
3. Procedure.
3.1 Grab sampling.
3.1.1 Set up the equipment as shown in
Figure 3-1. Place the probe In the stack at a
sampling point and purge the sampling Jme.
Figure 2-2. Velocity traverse data.
1 Trade name.
FEDERAL REGISTER, VOL 36, NO. 157—TUESDAY, AUGUST 17, 1971
Uo.169-Pt.il-
B-4
-------
PROPOSED RULE MAKING
PROBE
PJV
FLEXIBLE TUBING
ERIG
FILTER (GLASS WOOL)
SQUEEZE BULB
Figure 3-1. Grab-sampling train.
RATE METER
5. References
TO ANALYZER Altshuller. A. P., et al. Storage of Oos.t.
and Vapors in Plastic Bags. Int J. Air *
Water Pollution. 6:75-81. 1903.
Conner, William D., and J. S. Nader. Air
Sampling with Pla-stic Bags. JournaJ of the
American Industrial Hygiene As.-delation
25 291 -297. May-iunc 1964.
Devorkln, Howard, et al. Air Pollution
Source Testing Manual. Air Pollution Con-
trol Di-strlct. Los Angeles. November 1903.
METHOD 5.—DETERMINATION OF PNRTICULATE
EMISSION'S FHOM STATIONARY SOURCES
AIR-COOLED CONDENSER
PROBE
QUICK DISCONNECT
FILTER (GLASS WOOL)
1. Principle and opp
1.1 Principle. Participate matter is with-
drawn Isokinetlcally from the soxirce and Its
weight is determined gravimetrically after
removal of uncombined water.
1.2 Applicability. This method Is applica-
ble for the determination of particulate
< missions from stationary sources only when
specified by the test procedures for deter-
mining compliance with New Source Per-
formance Standards.
2. Apparatus,
2.1 Sampling train. The design specifica-
tions of the particulate sampling train used
by EPA (Figure 5-1) are described in APTD-
05S1. Commercial models of this train are
available.
2.1.1 Nozzle—Stainless steel (316) with
sharp, tapered lending edge.
2.1.2 Probe—Pyrex ' glass with a heating
system capable of maintaining a gas tempera-
ture of 250' P. at the exit end during
sampling. When temperature or length
limitations are encountered, 316 stainless
steel, or equivalent, may be used, as approved
by the Administrator.
RIGID CONTAINER
Figure 3-2. Integrated gas - sampling train.
3 1.2 Draw sample Into the analyzer.
3 2 Integrated sampling.
3.2.1 Evacuate the flexible bag. Set up the
equipment as shown In Figure 3-2 with the
bag disconnected. Place the probe in the
stack and purge the sampling line. Connect
the bag, making ST>^ that all connections
are tight and that there are no leaks.
3.2 2 Sample at a rate proportional to the
stack gas velocity.
3.3 Analysis.
3.3.1 Determine the CO?, 02, and CO con-
centrations as soon as possible. Make as many
passes as are necessary to give constant read-
ings. If more than 10 passes are necessary,
replace the absorbing solution.
3.3.2 For Integrated sampling, repeat the
analysis until three consecutive runs vary
no more than 0.2 percent by volume for each
component being analyzed.
4. Calculations.
4 1 Carbon dioxide. Average the three
consecutive runs and report result to the
nearest 0.1 percent CO2.
4.2 Excess air. Use Equation 3-1 to cal-
culate excess air, and average the runs. Re-
port the result to the nearest 0.1 percent
excess air.
_ _ (r0o.)-o.s(%co) . nft
0.264 ( % N,) - (% 0,) +0.5(% CO) X 1UO
equation 3-1
where:
%EA=Percent excess air.
%O,=Percent oxygen by volume, dry
basis.
%N,=Pereent nitrogen by volume, dry
basis.
TtCO=Percent carbon monoxide by vol-
ume, dry basis.
0.264=Ratlo ot oxygen to nitrogen in air
by volume.
4 3 Dry molecular weight. Use Equation
3-2 to calculate dry molecular weight and
average the runs. Report the result to the
nearest tenth.
Ma = 0.44(% C0,)+032(% O.)
+ 0.28(% N,+ %CO)
Equation 3-2
where:
Md = Dry molecular weight, lb./lb.-
mole.
%CO,=Percent carbon dioxide by volume,
dry basis.
%O,=Percent oxygen by volume, dry
basts.
TcNj=Percent nitrogen by volume, dry
basis.
0.44 = Molecular weight of carbon dioxide
divided by 100.
0.32 = Molecular weight ot oxygen
divided by 100.
028=Molecular weight of nitrogen
divided by 100.
FEDERAL REGISTER, VOL. 36, NO. 159—TUESDAY, AUGUST 17, 1971
B-5
-------
PROPOSED RULE MAKING
2.1.3 Pilot tube—Type S, or equivalent,
Bt Inched to probe to monitor stnck gas
velocity.
2.1.4 Filter holder—Pyrex1 glass with
healing system capable of maintaining niijr
temperature to & maximum of 225* f.
21.5 Itnplngurs—Pour Impingers con-
nected In scries with glass ball Joint fittings.
The first, third, and fourth Impingers are of
the Orecnburg-Smlth design, modified by rc-
piaclng the lip with a 'i-lnch. ID gla'a tube
extending to ^-Inch from the bottom of the
fia.sk. The r-ocond Implnger is of the Grccn-
burg-Snnth dc:.Jgu with the sUindard tip.
2 1.6 Metering system—Vacuum gauge,
Icak-ficc pump, thermometers capable of
measuring temperature to within 6" P., dry
gas meter with 2 percent accuracy, and re-
lated equipment, or equivalent, as required
to maintain an Isoklnotic .sampling rate and
to determine sample volume.
HEATED AREA FILTER HOLDER THERMOMETER
PROBE
REVERSE-TYPE
PITOT TUBE
PITOT MANOMETER
ORIFICE
IMPINGERS ICE BATH
BY-PASS.VALVE
VACUUM
GAUGE
MAIN VALVE
CHECK
VALVE
VACUUM
LINE
DRY TEST METER
AIR-TIGHT
PUMP
Figure 5-1. Particulale-sampling train.
2.1.7 Barometer—To measure atmospheric
pressure to ±0.1 In. Hg.
2.2 Sample recovery.
2.2.1 Probe brush—At least as long as
probe.
2.2.2 Glass wash bottles—Two.
2.2.3 Glass sample storage containers.
2.2.4 Graduated cylinder—250 ml.
2.3 Analysis.
2.3.1 Glass weighing dishes.
2.3.2 Desiccator.
2.3.3 Analytical balance—To measure to
±0.1 mg.
2.3.4 Beakers—250 ml.
1 Trade name.
2.3.5 Separatory funnels—500 ml. and
1,000 ml.
2.3.9 Trip balance—300 g. capacity, to
measure to ±0.05 g.
2.3.7 Graduated cylinder—25 ml.
3. Reagents.
3.1 Sampling
3.1.1 Filters—Glass fiber. MSA 1106 BH,
or equivalent, numbered for Identification
and preweighed.
3.1.2 Silica gel—Indicating type. 6 to 16
mesh, dried at 175* C. (350* F.) for 2 hours.
3.1.3 Water—Deionized, distilled.
3.1.4 Crushed Ice.
3.2 Sample recovery
3.2.1 Water—Deionized, distilled.
322 Acetone— Reagent grade.
3 3 Analysis
331 Water— DcionUcd, distilled
332 Chloroform—Ronsent grade.
3 3.3 Ethyl ether—Roaeent grade.
334 Dc.slccant—Drlerite,' Indicating.
4. Procedure.
4.1 Sampling.
4 1.1 After selecting the sampling site and
the minimum number of sampling points,
determine the stack pressure, tempciature,
moisture, and range of velocity head.
4.1.2 Preparation of collection train.
Weigh to the nearest gram approximately
200 g. of silica gel. Label a filter of proper
diameter, desiccate3 for at least 24 hours
and \veigh to the nearest 0.5 mg. In a room
where the relative humidity Is less than
50 percent. Place 100 ml. of water In each of
the first two Impingers, leave the third 1m-
pinger empty, and place approximately 200
g. of prewclghed silica gel in the fourth 1m-
pinger. Save a portion of the water for use
as a blank in the sample analysis. Set up the
train without the probe as In Figure 5-1.
Leak check the sampling train at the sam-
pling site by plugging the inlet to the filter
holder and pulling a 15-ln. Hg vacuum. A
leakage rate not in excess of 0.02 c f m at a
vacuum of 15-in. Hg is acceptable. Attach
the probe and adjust the heater to provide a
gas temperature of about 250° F. at the
probe outlet. Turn on the filter heating sys-
tem. Place crushed ice around the impingers.
Add more ice during the run to keep the tem-
perature of the ga^es leaving the last im-
pinger at 70" F. or less.
4 1.3 Participate train operation For each
run record the data required on the example
sheet shown in Figure 5-2. Take readings
at each sampling point at least every 5 min-
utes and when significant changes in stock
conditions necessitate additional adjust-
ments in flow rate. To begin sampling, po-
sition the nozzle at the first traverse point
with the tip pointing directly into the gas
stream. Immediately start the pump and ad-
Just the flow to isokmetic conditions. Main-
tain Isokinetic sampling throughout the
sampling period. Nomographs are available
which aid In the rapid adjustment of the
sampling rate without other computations.
APTD-0576 details the procedure for using
these nomographs. Turn off the pump at the
conclusion of each run and record the finaj
readings. Remove the probe and nozzle from
the stack and handle in accordance with the
sample recovery process described in section
4.2.
»Dry using Drierite ' at 70° ±10' F.
FEDERAL REGISTER VOL. 36, NO. 159—TUESDAY, AUGUST 17, 1971
B-6
-------
PROPOSED RULE MAKING
PLANT
LOCATION
DATE..
f.UN NO.
SA.V.PLE BOX NOj..
men i>ox NO.__
AVBItNT TEMPI:r,ATURE_
BAROViTRICPRESSURE_
ASSUVED MOISTURE. 8_
IIEA1ER BOX SETTING
MODE LENGTH, in.
NOZZLE DIAMETER, in. ._
PROBE HEATER SETTING_
CFACTOR
SCHEMATIC OF STACK CROSS SECTION
TRAVERSE POINT
KUV.BEB
TOTAL
SAMPLING
TIME
to), min.
AVERAGE
STATIC
PRESSURE
[Psl. in. Ha.
SfACK
TEMPERATURE
(Tsl.'F
VELOCITY
HEAD
I A PS).
PRESSURE
DEFERENTIAL
ACROSS
ORIFICE
METER
UH),
In, H20
GAS SAMPLE
VOLUME
(Vml, It3
GAS SAV.PLE TEMPERATURE
AT DRY GAS METER
INLET
ITm,D).°F
Avg.
OUTLET
(Tmout1.°F
Avg,
Avg.
SAMPLE BOX
TEMPERATURE,
°F
IMPINGER
TEMPERATURE.
"f
4.2 Sample recovery. Exercise care in mov-
ing the collection train from the test site to
the sample recovery area to minimize the loss
of collected sample or the gain of extraneous
particulate matter. Set aside portions of the
water and acetone used in the sample recov-
er7 as blanks for analysis. Place the samples
In containers as follows:
Container No. 1. Remove the niter from its
holder, place in this container, and seal.
Container No. 2. Place loose particulate
matter and acetone washings from all sam-
ple-exposed surfaces prior to the filter in this
container and seal. Use a raaor blade, brush,
or rubber policeman to loosen adhering par-
ticles.
Container No. 3. Measure the volume of
water from the first three Implngers and
place the water in this container. Place^'ater
Figure 5-2. Paniculate Held data.
rinsings of all sample-exposed surfaces be-
tween the filter and fourth impinger in this
container prior to sealing.
Container No. 4. Transfer the silica gel
from the fourth impinger to the original
container and seal. Use a rubber policeman
as an aid in removing silica gel from the
impinger.
Container No. 5. Thoroughly rinse all sam-
ple-exposed surfaces between the filter and
fourth impinger with acetone, place the
washings in this container, and seal.
4.3 Analysis. Record the data required on
the example sheet shown In Figure 5-3.
Handle each sample container as follows:
Container No. 1. Transfer the filter and any
loose particulate matter from the sample
container to a tared glass weighing dish, des-
sicate, and dry to a constant weight. Report
results to the nearest O.S ing.
Container No. 2. Transfer the acetone
washings to a tared beaker and evaporate to
dryness at ambient temperature and pres-
sure. Dessicate and dry to a constant weight.
Report results to the nearest 0.5 mg.
Container No. 3. Extract organic particulate
from the impinger solution with three 25 ml.
portions of chloroform. Complete the en-
traction with three 25 ml. portions of ethyl
ether. Combine the ether and chloroform ex-
tracts, transfer to a tared beaker and evapo-
rate at 70* P. until no solvent remains. Des-
sicate, dry to a constant weight, and report
the results to the nearest 0.5 mg.
Container No. 4. Weigh the spent silic
gel and report to the nearest gram.
FEDERAL REGISTER, VOL. 36, NO. 159—TUESDAY, AUGUST 17, 1971
B-7
-------
PROPOSED RULE MAKING
PLANT.
DATE.
JiU'J N0._
CONTAINER
NUMBER
1
2
3a"
3b"»
5
TOTAL
WEIGHT OF PARflCULATE COLLECTED.
ing
FINAL WEIGHT
^>~
-------
PROPOSED RULE MAKING
o
3
<
>-"
<
No 159-
n
B-9
-------
PROPOSED RULE MAKING
23.1 PlpcMos—Transfer tjpe, 5 ml. and
10 ml. s,\7.cs (0.1 ml. dlvl.-lons) and 25 ml.
fcbo (02 ml. divisions).
2 3.2 Volumetric flasks—50 ml., 100 ml.,
and 1,000 ml.
233 Burettes—5 ml and 50 ml.
234 Erlcnmeycr fln«,h—125ml.
3. Reagents.
3.1 Sampling.
31.1 Water—Deionlzed, distilled.
3.1.2 Isopropanol, 80 percent---Mix 80 ml.
of isopropanol with 20 ml. of distilled water.
3.1.3 Hydrogen peroxide, 3 pciccnt—dilute
100 ml. of 30 percent hydrogen peroxide with
900 ml. of distilled water. Prepare fresh daily.
3.2 Sample recovery.
3 2.1 Water—Deionlzed, distilled.
3 2.2 Isopropanol, 80 percent.
3.3 Analysis.
3 3.1 Water—Deionlzed, distilled.
3.3.2 Isopropanol.
3.33 Thorin indicator—l-(o-arsonophen-
jlazo) -2-naphthol-3, 6-disulfonic acid, diso-
dium fcftlt (or equivalent). Dissolve 0.20 g.
in 100 ml. distilled water.
33.4 Barium perchlorate (0 OliV)—Dis-
Eolve 1.95 g. of barium perchlorato
I3a(ClO,), SHjOJ in 200 ml. distilled water
and dilute to 1 liter with isopropanol. Stand-
ardize with sulfuric acid.
3 3.5 Sulfuric acid standard (O.OINf—
Purchase or standardize against a primary
standard to ± 0.00021V.
4. Procedure.
4.1 Sampling.
4.1.1 Preparation of collection train. Pour
16 ml. of 80 percent Isopropanol Into the
midget bubbler and 15 ml. of 3 percent hydro-
gen peroxide to each of the first two midget
Impingers. Leave the final midget Impinger
dry. Assemble the train as shown In Figure
6-1. Leak check the sampling train at the
sampling site by plugging the probe inlet and
pulling a 10-m. Hg vacuum. A leakage rate
not In excess of 1 percent of the sampling
rat* \s acceptable. Carefully release the probe
Inlet plug and turn off the pump. Place
crushed ice around the impmgers. Add more
ice during the run to keep the temperature
of the gases leaving the last Impinger at
70' F. or less.
4.1.2 Sample collection. Adjust the sam-
ple flow rate proportional to the stack as
velocity. Take readings at least every 5 min-
utes and when significant changes In stack
conditions necessitate additional adjust-
ments In flow rate. To begin sampling, posi-
tion the nozzle with the tip pointing directly
Into the gas stream and start the pump. Sam-
ple proportionally throughout the run. At the
conclusion of each run, turn off the pump
and record the final readings. Remove the
probe from the stack and disconnect It from
the train. Drain the Ice bath and purge the
remaining part of the train by drawing clean
ambient air through the system for 15 min-
utes.
4.2 Sample recovery. Disconnect the Im-
pingers after the purging period. Discard
the contents of the midget bubbler. Pour
the contents of the midget Impingers into
a polyethylene shipment bottle. Rinse the
three midget impingers and the connecting
tubes with dKtlllcd water and add these
un-',hin£s to the same ''oragc container.
43 S.imple analysis. Trnii'-fcr the con-
tents of the storage container to a 50-ml.
volumetric flask. Dilute to the mark with
delonixed, distilled water. Pipette a 10 ml.
aliquot of this solution to a 125-ml. crlen-
mycr flask. Add 40 ml. of Isopropanol and 2
to 4 drops of thorln indicator. Titrate to a
pink endpolnt using 0 01JV barium perchlo-
rate. Run a blank with each series of
fcamples.
5. Calibration.
5.1 Use standard methods and equipment
approved by the Administrator to calibrate
the orifice meter, pitot tube, dry gas meter,
and probe heater.
5 2 Standardize the sulfuric acid with po-
tassium acid phthalate as a primary stand-
ard. Standardise the barium perchlorate
with 26 ml. of standard sulfuric acid con-
taining 100 ml. of ibopropanol.
6. Calculations.
6.1 Dry gas volume. Correct the sample
volume measured by the dry gas rneter to
standard conditions (70* P. and 29 92 in.
Hg) by using Equation 6-1.
v (Tiii
• DO I rn
/ 7 °R \V,Pb.,
V. in. Hg/ TV equation 6-1
where:
Vm.,a=Volmne of gas sample through the
dry gas meter (standard condi-
tions) , cu. ft.
V» = Volume of gas sample throvgh the
dry gas meter (meter conditions),
cu. ft.
Ttld=Absolute temperature at standard
conditions, 530* B.
T»=A\erage dry gas meter temperature,
•B.
Pblt=Barometric pressure at the orifice
meter, in. Hg.
4 = Abso)ute pressure at standard con-
ditions, 29.92 In. Hg.
6.2 Sulfur dioxide concentration.
/ Ib 1 \
( 7.03X10-'^)
\ g.-DlI./
equation 6-2
V.cin = Total solution vohi^nc of sulfur
dioxide, ml.
V«=:Volume of pajnple aliquot
titrated, ml.
Vmtta~Volume of gas sample through
the dry gas meter (standard
conditions), see Equation 6-1,
cu. f».
7. Relaenca.
Atmospheric Emissions from Sulfuric Acid
Manufacturing Processes. U.S. DIIEW, PHS,
Division of Air Pollution. Public Health Serv-
ice Publication No. 999-AP-13. Cincinnati,
Ohio. 1965.
Corbett, P. P. The Determination of SO,
and SO, In Flue Gases. Journ.il of the In-
stitute Of Fuel. 24.237-243. 1961.
Matty, B. E. and E. K. Dlehl. Measuring
Flue-Gas SO, and SO,. Power iW:94-97.
November 1957.
Patton, W. F. and J. A. Brink. Jr. New
Equipment and Techniques for Sampling
Chemical Process Gases. Paper presented at
the 55th Annual Meeting of APCA. Chicago,
HI. May 20-24, 1962.
where:
Cso2 = Concentration of sulfur di-
oxide at standard conditions,
dry basis, Ib./cu. ft.
7. 05 X10-t = Conversion factor Including
the number of grams per gram
equivalent of sulfur dioxide
(32 g./g.-eq.) , 453.6 g./lb., and
1,000 ml./l, Ib.-l./g.-ml.
V,= Volume of barium perchlorate
tltrant used for the sample, ml.
V,,= Volume of barium perchlorate
tltrant used for the blank, ml.
K= Normality of barium perchlo-
rat« titrant, g.-eq./l.
FEDERAL REGISTER. VOL. 36, NO. 159—TUESDAY, AUGUST 17, 1971
B-10
-------
APPENDIX C
VISIBLE EMISSIONS REPORTED FROM EPA TESTS AND QUESTIONNAIRE DATA
Visible emissions were read by the EPA project engineer on a random
basis during EPA tests. Figure C-l shows the points of discharge where the
emissions were read. Table C-l reports the emissions in percent opacity
corresponding to the emission points numbered in Figure C-l. Point 2 in the
figure is used to indicate the furnace is without a control device whereas
Point 4 shows that the furnace is equipped with a control device. Covered
furnaces H and J have mix seals. Above the mix seals are located a hood and
uncontrolled exhaust systems to remove the escaping dust-laden gases from
the mix seals. This is shown as Point 6 in Figure C-l. The opacity of these
emissions ranges from 10 to 100 percent. Covered furnaces K and L are
equipped with fixed seals and do not have hooding and uncontrolled exhaust
systems. In these two cases no emissions from the building were notice-
able
In most instances, there were no, or very insignificant, noticeable
visible emissions from the control equipment. Most noticeable emissions
were from uncontrolled furances A and B, the uncontrolled exhaust systems
serving the tapping stations, and the uncontrolled exhaust system serving
to remove the fugitive fumes from the covered furances with mix seals (H
and J).
Table C-2 is a summary of visible emissions reported by the ques-
tionnaires submitted by the industry. The reported opacity of the
C-l
-------
fugitive fumes escaping from the mix seals ranged from 0 to 60 percent.
The reported opacity of uncontrolled emissions from open furnaces
ranged from 10 to 100 percent.
Whenever possible, emissions were read using the visible emission
recording sheet shown in Figure C-2. The second page of Figure C-2 is
in reality printed on the reverse side of the visible emission record,
and the necessary data relative to the opacity are filled in at the time
of the readings.
C-2
-------
Table C-). OPACITY OF EMISSIONS REPORTED FROM EPA TESTS
(I)
urnace
A
B
C
D
E
F
6
H
J
K
L
Test
run
1
2
1
2
1
2
3
4
5
7
8
9
10
11
12
1
2
3
1
2
3
1
2
J
1
2
3
I
2
3
1
2
3
1
2
3
4
5
1
I
1
\ Type
ro uct
FeCrS!
Cr ore/
lime
Belt
SIMn
FeCrS!
rIC FeCr
Silicon
S1Hn
aOi FeSt
CaC2
FeHn
SIMn
of
urnace
0
0
0
0
0
0
0
C
C
C
C
Uncontrolled fumes
Outlet of| Outlet of
furnace tapping
exhaust exhaust
system | system
r1
b
Outlet of
fugitive
fume exhaust
system
po nt
Controlled fumes
i
; Tercent
opaci tv of
' Degree uncap tured
Type of fumes fumes at
control captured roof moni tor-
1
None
50 to 100 b
SO to 100 b ! 100C None
i
50 to lOoi b ' 100C
Venturl(57"PO)' 99 0
\ Venturi (57"PO)
100
100
100
10 to 80
10 to 60
10 to 80
100"
100"
100"
100d
,00"
100"
too"
100"
Venturi (57"PO)>
Venturi(57"PD)j
Venturt(5;"PD)|
0
0
0
0
Venturi (47"PD) 99 0
Ventun(47"PD)'
Ventun(4r'PD)
0
0
»enturt(37"PD)' 99 0
\ Venturi(37"PD)
Venturl(37"PO)
i
0
0
Baghouse 99 0
J 8aghouse
Baghouse
0
0
Preclpltator ' 99
10 to 100
10 to 100
10 to 100
10 to 100
10 to 100
10 to 100
PrecipUator
Preclpltator
Baghouse 95
Baghouse
Baghouse
to 99 10 to 20
10 to 20
10 to 20
Venturi 99 0
99 0
99 0
Ventun 85 ! See point 6
Venturi
Venturi
Scrubber
Scrubber
Scrubber
Venturi 10
Venturi
Venturi
venturf
Ventur!
Venturi 1C
Venturi
Venturi
i
i See point 6
; See point 6
!
!
1
0 0
0
0
0
0
K) 0
Q
0
"PercenT "
opacity of
control
equipment
outlet-
0
0
0
0
0
0
0
0
< 10
< 10
< 10
0
0
0
< 5
5 to 20
5 to 20
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
*Q • open, C a closed.
Tapping fumes vented to main uncontrolled exhaust system.
cUncaptured fumes escaping through roof monitor.
Tlo exhaust system at tapping point.
C-3
-------
table C-2 OPACITY 01 EMISSIONS HEPORfi.ll I R0» yiJf.1 IOMKAIW WUA
"
Questionnaire
82
85
46
44
45
64
65
66
70
71
73
83
6
IS
103
79
84
68
72
91
60
55
54
59
8
91
47
52
5
7
10
11
13
14
44
74
76
100
90
67
69
75
77
53
56
57
61
62
98
3
114
as
Product
«C FeHn
6S-75S FeS<
jjn fsi1
St Fin
FeCrSi
FeCr
c.c2
LC i MC feFIn
CfOrt/HM
melt
SIHnZ
I
Outlet of
furnace
80 to 100
80 to 100
20 to 80
20 to 80
20 to BO
20 to 60
20 to 60
10
80 to 100
20 to 80
20 to 60
20 to 60
20 to 60
20 to 60
20 to 60
20 to 60
80 to 100
5 to 50
60 to 80
60 to 60
ncontrollcd fume
Outlet of
tapptmj
No data report-
ed
"ruiiVtlvc
fuOKS
0 to 20
20 to 40
35
10 to 50
5 to 30
5 to 25
0 to 10
40 to 60
20 to 40
20 to 40
S to 10
15 to 30
25
10 to 20
10 to 20
10 to 20
10 to 20
20 to 40
20 to 40
20 to 60
20 to 30
20 to 40
20 to 30
20 to 40
10 to 35
20 to 40
20 to 40
10 to 35
10
ton troll
Type of
BF^sc rubber
Br scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
Venturl
scrubber
BF scrubber
BF scrubber
BF scrubber
Venturf
scrubber
Ventun
8F scrubber
BF scrubber
BF scrubber
Venturl
scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF scrubber
BF Scrubber
ed funcs
Control
equipment
0
0 to 20
0
'>F scruMxr • Guffllo Forgi icrubb.r.
C-4
-------
ONTROL
EQUIP.
ROOF MONITOR
TAPPING
HOOD
SCRUBBER
,wV\V\//#t\\V'ff\\\vwmvv//U\\Wf\WW/\W«W
OPEN FURNACE COVERED FURNACE
Figure C-1. Emission points where visible emissions were read during EPA tests on open
and covered furnaces.
C-5
-------
ENVIRONMENTAL PROTECTION AGENCY
COMPANY NAME
EQUIPMENT LOCATION ( ADDRESS)_
TIME OF OBSERVATION: FROM
RECORD OF
VISIBLE EMISSIONS
A.M.
.P.M. TO.
A.M.
.P.M. DATE .
s«"Vhour
R. No.
6
4*
4M
4K
4
3*
3V,
3M
3
2*
2V,
2K
2
w
m
IK
1
*
v>
%
0
1
% Min. 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
100
9$
90 I
65
80
75
70
65
60
55
50
45
4o
35
30 ]
25
20
15
10
5
0 - - - --
SlarVhour
R. No.
6
4%
4Vi
4!4
4
3H
3K
3)4
3
2%
2Vi
2V>
2
1H
1V4
1%
1
K
fc
'/.
0
% Min. 21 22 23 24 25 26 27 28 29 3
TOO
95
90
85
80
75
70
65
60
55
50
45
40
35
30
r 25
20
15
16
£
0
start/hoor
R. .No.
5
4%
4V4
4M
4
3X
3tt
3«
3
2N
2H
2%
2
IX
154
1H
1
X
Vt
%
0
% Mm. 41 42 43 44 45 46 47 48 49 £
iop
95
90
65
,8S
75
iJo
68
60
55
~5fl
45
46
sT
."30 r
"25
" K 1
15
TtT"
5
0
0 31 32 33 34 35 36 37 38 39 40
T" 1" "*"
i
rO 51 52 53 54 55 56 57 58 59 60
NOTE: Each small square represents an individual reading of intensity corresponding to that shown in the left-hand column
over a time span of % minute. Insort an "S" in the top row of blank squares to indicate the exact minute of the start of
observation. In the next square after the "S", insert the hour in which the measurement was made. Each page of this form
can thus be used to record 1 hour of measurements.
Figure C-2. Visible emissions recording sheet.
C-6
-------
ource of Air Contaminants.
ype of Air Contaminants
oint of Discharge: Stack | | Other .
oint of Observation:
Distance to Base of Point of Discharge, feet
Height of Point of Discharge Above Ground Level, feet
ackground Description
eather: Clear j | Overcast! | Partly Cloudy I I Other
Wind Direction ; Wind Velocity, mi/hr _
lume Description:
Detached: Yes I I No LJ
Color: Black I I White LJ Other
Plume Dispersion Behavior: Looping | | Coning | | Fanning | |
Lofting | | Fumigating | | See Comments I
Estimated Distance (feet) Plume Visible (Maximum) (Minimum)
lomments
iigned Title
Figure C-2. (continued) Visible emissions recording sheet.
C-7
-------
APPENDIX D
PARTICLE SIZE ANALYSIS
Determination of the size distribution of particles suspended in
gases from a submerged-arc furnace is important and is helpful in
designing an optimum dust-removal system. Unless capture and recovery
is excellent, significant quantities of small particles can be emitted
to the atmosphere since nearly all of the particles are less than 1
micron. Very little published information exists on the size dis-
tribution of particulate matter from ferroalloy furnaces. Person states
20
that most particles are less than 0.5 micron. This was verified In en
EPA-contracted study made by TRW, Inc. This study's data, presented in
this appendix, were obtained during the same time tests were made to
determine mass emission rates.
Samples of particulates in the gas streams from five electric
furnaces and at the outlets of four air pollution control devices were
collected by a cascade Impactor to determine particle sr s d^tr^1 ••
The first nine samples were collected using the Andersen sampler. Because
the particles were smaller than expected, the majority of the particles
passed through the sampler and were deposited on the filter. The
Andersen sampler was therefore discarded in favor of the Brink sampler,
which was used in gathering the remaining samples.
D-l
-------
The basic design of the cascade impactor sampler has been described
in detail in literature. The cascade impactor is constructed of a
succession of jets, and each jet is smaller than the preceeding one. The
sampler is small enough that it can be inserted through a small port in the
stack and positioned in the gas stream so that the sample is collected in
conditions as close as possible to isokinetic.
The gas streams take a turning path through the impactor. When the
velocity of an entrained particle is high enough to overcome aerodynamic
drag and escape from the laminar streamline, the particle impacts on a
collection area of the plate. If the particle is too small and its inertia
too low for removal from the stream, it proceeds with the gas through an
orfice in the next plate, and its velocity and inertial effects are
increased. Fractionation is not based on physical diameter only, but
also on "aerodynamic diameter," which takes into account density and
shape. Aerodynamic diameter, in fact, is probably more related to
atmosphere behavior and the operation of particulate collection devices
than its physical diameter.
The Andersen impactor has a stainless steel sampling head containing
nine circular plates, each with several hundred jet-orifices. The size
of the orifices becomes progressively smaller in each succeeding down-
stream plate, so that the velocity of the gas streams that pass through
the orifices increases correspondingly. The top part of the photograph
of the Andersen sampler in Figure D-l shows the relative size of the
assembled sampler. The lower part of the photograph shows the sampler
when disassembled.
D-2
-------
Ij* J^i.,
rfc
.*4~* <
, ,J_,
Figure D-1. Andersen sampler showing assembled and disassembled sections.
D-3
-------
The second sampler, the Brink cascade impactor, has a single jet
orifice for each stage. Each orifice uses a collection cup as an impac-
tion surface rather than the solid areas of the next plate. Velocities
are higher through the single orifices of the Brink than through the
multiple orifices of the Andersen, so the cups normally have adhesive
coatings. The manufacturer indicates that the Brink impactor collects and
classifies particles down to 0.25 micron, with more cuts in the fine-
particle range than the Andersen sampler. Because it was found that most
of the particles from air pollution control devices were below 0.7 micron,
the Brink impactor was selected to obtain most of the samples. The charac-
teristic diameter of an aerosol particle for each impactor stage (i.e.,
Dpc) has been calculated for pressure drops across the impactor of 5 and 10
inches of mercury (based on a spherical particle of density one). The
impactor has five in-line stages as shown in Figure D-2. The top part
of the figure shows the relative size of the impactor; the lower part
shows an enlarged view of the five jet stages. When the pressure drop
across the impactor is at 5 inches of mercury, the particles collected are
in the size range of 3.4, 2.0, 1.36, 0.69, and 0.42 microns. When the
pressure drop across the impactor is at 10 inches of mercury, the sizes
of the particles collected are 3.06, 1.80, 1.23, 0.63, and 0.38 microns.
During the tests, the cascade impactor was mounted on a probe and
connected to vacuum pumps by rubber tubing. Metering valves were installed
on the inlet side of the pump to adjust the air flow through the samplers.
Magnehelic gauges were inserted in the system to measure pressure drops
across the sampler. Figure D-3 illustrates a typical particle sizing
train. Volumetric flow rates through the sampler were measured for long
D-4
-------
Figure D-2. Brink sampler showing assembled and disassembled sections.
D-5
-------
MAGNEHELIC
GAUGE
STACK
Figure D-3. Particle sizing train.
D-6
-------
durations using the dry gas meter and for short durations using the
pressure drop across the sampler.
The particle size tests made while measuring mass emission rates from
furnaces A, C, D, E and G in the field testing program indicate that the
particle size is largely submicron and that particle size distribution
varies for different products. The mass median diameter (MMD) in microns
is less for ferroalloys that contain silicon as a major component (for
example, ferrochrome-silicon and silicomanganese) than it is for HC
ferrochrome, which has a relatively low silicon content.
Size distributions are defined by two parameters: (1) the intercept
of the curve with the 50 percent probability (mass mean particle diameter),
and (2) the polydispersity factor or geometric standard deviation, defined
as: 0 = diameter of particle at 50% probability
diameter of particle at 15.87% probability
Figures D-4 through D-14 show graphical presentations of particle size
test data, that is, log-probability plots of cumulative percent less than
stated micron size versus the Dpc for each fraction collected by a cascade
impactor. The intercept of these curves at 50 percent probability indicates
that, on a weight basis, one-half of the particles are below the micron
size shown at this intercept and one-half are above.
Table D-l shows the particle size versus the collection efficiency of
EPA-tested control equipment. It is interesting to note that the efficiency
of scrubber G (Aeronetics scrubber system serving a silicomanganese furnace)
becomes markedly less when the particle size is under 0.6 microns. It is
also interesting to note that if an efficiency curve vs. particle size were
to be drawn for the baghouse indicated in the table, there would be a
D-7
-------
slight dip in the curve at the particle size range between 0.3 to 1.0 microns.
This drop in efficiency appears to be typical according to the article
"Design and Performance of Modern Gas Cleaning Equipment" Journal of the
Institute of Fuel, February 1956.
Tables D-2 through D-ll show the particulate emission losses in pounds
per hour for various sources according to particle size ranges. These losses
were calculated from the mass emission rates of particulates based on EPA
tests and from Figures D-4 through D-14 of this appendix.
D-8
-------
Table D-l. PARTICLE SIZE VS. COLLECTION EFFICIENCY
OF ERA-TESTED CONTROL EQUIPMENT
(percent)
Particle
size range,
microns
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
+1.5
Overall
Control equipment
Scrubber3
Cl
98.5
99.3
99.7
99.8
99.9
99.5
C2
93.4
96.5
97.7
98.2
98.2
97.4
C3
90. Ob
92.8
94.1
98.7
98.9
96.1
Baghouse
D
99.4
98.1
97.8
97.7
99.6
98.7
Preci pita tor
E
98.4
98.8
99.1
99.2
99.0
99.0
Scrubber
G
NIL
64
91
98
99
96
Scrubber C tested at theee pressure drops: C, at 57", C9 at 47", and C,
*+ IT" u n to
at 37" H20.
'Assumed. Actual calculations above 93%.
D-9
-------
Table D-2. PARTICULATE EMISSIONS BY PARTICLE SIZE
FROM UNCONTROLLED FeCrSi FURNACE
(Ib/hr)
Particle size
range, microns
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
+1.5
Total, Ib/hr
Sample no.
Al
16
44
49
28
26
163
A2
59
122
119
67
A3
76
126
105
59
53 . 54
420
420
D-10
-------
Table D-3. PARTICULATE EMISSIONS BY PARTICLE SIZE
FROM THE SCRUBBER INLET OF A SiMn FURNACE
(Ib/hr)
Particle
size, microns
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
+1.5
Total, Ib/hr
Sample no.a
C-l
223
427
464
316
427
1857
C-2
484
502
372
202
297
1857
C-3
55
83
in
120
1477
1846
C-4
b
1611
C-5
290
218
181
133
387
1209
C-7
545
61
48
36
519
1209
Average
319
258
235
161
621
1594
aSample numbers C-l and C-2 obtained at port 1, C-3 at port 2, C-4 at port 3,
C-5 at port 4, and C-7 at port 5 (not simultaneously).
1234
~
,., Sample Ports
__ Sampl e Points
Cross section of inlet duct showing sampling ports.
5Data points are at the extremities of the graphs, either less than 15% or
greater than 85% of stated size. Reliable particle sizes could therefore
not be determined from these data.
D-ll
-------
Table D-4. PARTICULATE EMISSIONS BY PARTICLE SIZE FROM
SCRUBBER OUTLET OF A SiMn FURNACE
(Ib/hr)
Particle
size,
microns
0 to 0.1
0.1 to 0.2
0.2 to 0.3
0.3 to 0.4
0.4 to 0.5
0.5 to 0.6
0.6 to 0.7
0.7 to 0.8
0.8 to 0.9
0.9 to 1.0
1.0 to 1.2
1.2 to 1.5
+1.5
Total, Ib/hr
Sample number and scrubber pressure drop,
inches water
C-8
57"
1.65
1.32
0.99
0.66
0.58
0.41
0.33
0.25
0.25
0.16
0.25
0.33
1.08
8.26
C-9
57"
1.59
2.35
1.68
1.01
0.59
0.42
0.25
0.17
0.08
0.04
0.03
0.02
0.16
8.39
C-12
37"
<0.30
0.40
2.12
3.54
4.24
4.24
3.54
3.18
2.48
2.12
3.18
2.80
3.28
35.4
C-13
47"
2.00
2.00
1.42
1.06
0.83
0.59
0.47
0.47
0.35
0.24
0.47
0.47
1.43
11.8
C-14
47"
10.6
15.2
10.6
7.1
5.0
3.5
3.5
2.1
2.1
1.5
2.1
2.8
4.5
70.6
C-15
37"
1.5
5.7
9.2
8.5
7.1
5.7
5.0
3.5
3.5
3.5
3.5
4.9
9.0
70.6
C-16
3/"
5.6
8.5
10.6
7.1
7.1
5.0
4.2
3.5
3.0
2.1
3.5
3.5
6.9
70.6
C-17
37"
1.5
5.0
7.8
7.1
7.1
7.1
5.0
4.2
3.5
3.5
5.0
4.9
8.9
70.6
D-12
-------
Table D-5. PARTICULATE EMISSIONS BY PARTICLE SIZE FROM
UNCONTROLLED TAPPING OF A SiMn FURNACE
(Ib/tap)
Particle size
range, microns
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
1.5 to 2.0
2.0 to 3.0
3.0 to 6.0
6.0 to 10.0
+10
Total Ib/tap
Sample no.
C-18
4.2
6.4
6.9
6.4
4.8
6.4
9.0
3.7
5.2
53. Oa
C-19
11.7
9.5
8.0
5.8
3.7
4.8
5.3
2.1
2.1
53. Oa
C-20
4.2
4.8
5.3
4.8
4.2
5.8
9.0
5.8
9.1
53. Oa
C-21
12.2
9.5
7.4
6.4
3.2
5.3
4.8
2.1
2.1
53. Oa
Average tapping loss per tap based on 3 test runs.
D-13
-------
Table D-6. PARTICIPATE EMISSIONS BY PARTICLE SIZE FROM
A FeCrSi FURNACE AT INLET TO BAGHOUSE
(Ib/hr)
Particle size
range,
microns
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
+ 1.5
Total, Ib/hr
Sample no.
D-l !
161
125
101
71
136
594
D-2a |
1
36 '
48
54
48
408
594
D-3a[
- j
101 j
101
95
65
232
594
D-4b
131
119
89
77
178
594
T n-5^
j u °
i 327
113
71
36
47
594
r i
D-6C
178
125
95
59
137
594
D-7C
296
77
54
30
137
594
D-8d
107
77
71
65
274
594
D-9d
184
83
65
48
214
594
Samples 0-2 and D-3 collected simultaneously between furnace taps.
'Samples D-4 and D-5 collected simultaneously during furnace tap.
"Samples D-6 and D-7 collected simultaneously during furnace tap.
Samples D-8 and D-9 collected simultaneously between furnace taps.
D-14
-------
Table D-7. PARTICULATE EMISSIONS BY PARTCLE SIZE FROM
BAGHOUSE EXHAUST ON FeCrSi FURNACE
(Ib/hr)
Particle size
range,
mi crons
0 to 0.1
0.1 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
1.5 to 2.0
2.0 to 3.0
3.0 to 6.0
6.0 to 10.0
+10
Total, Ib/hr
D-10
~-
< 0.55
1.43
1.87
1.65
1.32
0.66
1.76
0.66
1.10
n.o ;
D-ll
--
<0.66
1.65
2.10
1.76
1.54
1.32
1.43
0.44
0.10
11.0
D-13
0.56
1.78
2.35
1.41
1.03
0.66
0.85
0.75
0.38
0.19
9.96
D-14
0.28
1.31
1.88
1.69
1.22
0.85
0.94
0.85
0.28
0.01
9.31
D-15
1.88
1.88
1.41
0.94
0.66
0.47
0.56
0.75
0.28
0.57
9.40 ,
D-163
- •
<0.70
2.08
2.43
1.56
0.87
0.70
0.26
--
—
8.7b ,
*
D-173
0.26
1.30
1.83
1.56
1.21
0.78
0.70
0.78
0.17
0.11
8.7a
D-16 and D-17 samples obtained simultaneously.
^Average of three test runs.
D-15
-------
Table D-8. PARTICULATE EMISSIONS BY PARTICLE SIZE
FROM HC FeCr FURNACE AT PRECIPITATOR
INLET
(Ib/hr)
Particle
size,
mi crons
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
+ 1.5
E-l
159
125
114
102
638
Sample no.
! E-2 !
83
i 97
180
194
830
E-3
138
207
221
193
625
i
i
i Average
; 126
143
172
163
698
Total, Ib/hr 1138 1384 1384 1302
D-16
-------
Table D-9. PARTICULATE EMISSIONS BY PARTICLE SIZE FROM
HC FeCr FURNACE AT PRECIPITATORS OUTLET
(Ib/hr)
Particle size,
microns
0 to 0.1
0.1 to 0.2
0.2 to 0.3
0.3 to 0.4
0.4 to 0.5
0.5 to 0.6
0.6 to 0.7
0.7 to 0.8
0.8 to 0.9
0.9 to 1.0
1.0 to 1.2
1.2 to 1.5
+ 1.5
Sample no.
E-4
- •
0.38
0.62
0.88
0.75
0.75
0.62
0.62
0.62 '
0.38
0.38
0.88
0.88
4.74
E-5
1.25
0.75
0.50
0.38
0.38
0.38
0.25
0.25
0.13 ;
0.25 i
0.25
0.50
7.23
E-6
0.43
0.57
0.71
0.57
0.57
0.43
0.43
0.43
0.43 ;
0.28
0.57
0.85
7.93
E-7
0.57
0.71
0.71
0.57
0.57
o
0.57
0.57
0.28 ,
0.43
0.28
0.57
0.43
7.94
E-8
0.57
0.85
0.71
0.71
0.71
0.57
0.57
0.43
0.43
0.43
0.57
0.85
6.80
Total, Ib/hr
12.50 12.50 14.20 14.20 14.20
0-17
-------
Table D-10. PARTICLE EMISSIONS BY PARTICLE SIZE FROM SiMn FURNACE
AT AERONETICS SCRUBBING INLET
(Ib/hr)
Particle
size,
microns
0 to 0.3
0.3 to 0.6
0.6 to 1.0
1.0 to 1.5
+1.5
Total, Ib/hr
Sample no.
6-1
2
10
20
27
136
195
6-2
1
3
9
11
164
188
Average
2
7
15
19
150
193
D-18
-------
Table D-ll. PARTICULATE EMISSIONS BY PARTICLE SIZE FROM SiMn FURNACE
AT AERONETICS SCRUBBER OUTLET
(Ib/hr)
Particle
size,
microns
0 to 0.1
0.1 to 0.2
0.2 to 0.3
0.3 to 0.4
0.4 to 0.5
0.5 to 0.6
0.6 to 0.7
0.7 to 0.8
0.8 to 0.9
0.9 to 1.0
1.0 to 1.2
1.2 to 1.5
+1.5
Total, Ib/hr
Sample no.
6-3
0.18
0.53
0.76
0.76
0.70
0.53
0.41
0.35
0.29
0.18
0.35
0.29
0.50
5.83
6-4
0.03
0.09
0.58
1.05
1.23
0.93
0.76
0.41
0.29
0.18
0.18
0.09
0.01
5.83
6-5
3.59
3.09
1.73
1.11
0.74
0.49
0.37
0.25
0.25
0.12
0.25
0.12
0.27
12.38
D-19
-------
!
s
s
3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.3
0.2
10
20
30 40 50 60 70
CUMULATIVE PERCENT < STATED MICRON SIZE
Figure D-4. Particle size distribution of uncontrolled fumes from a FeCrSi furnace.
98
D-20
-------
i
o
50.0
40.0
30.0
20.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
.« 3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0.82
0.56
5.40
»3.40
0.79
2.35
0.60
80
90
95
10 20 30 40 50 60 70
CUMULATIVE PERCENT < STATED PARTICLE SIZE
Figure D-5. Particle size distribution of SiMn fumes entering a scrubber serving an open furnace
D-21
-------
£
i
O-
4.0
3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.33
0.21
0.68
0.34
0.31
0.57
0.44
0.60
10
20 30 40 50 60 70 80
CUMULATIVE PERCENT < STATED MICRON SIZE
90
100
Figure D-6. Particle size distribution of SiMn fumes from a scrubber serving an open furnace.
D-22
-------
M.
Q.
cc
LU
ce
2
0.1
20
30 40 50 60 70 80 90
CUMULATIVE PERCENT< STATED SIZE
100
Figure D-7. Particle size distribution of uncontrolled tapping fumes from SiMn furnace.
D-23
-------
2
0.62
3.20
1.01
0.79
0.26
0.30
0.59
1.30
0.73
10
20 30 40 50 60 70 80
CUMULATIVE PERCENT< STATED MICRON SIZE
90
100
Figure D-8. Particle size distribution of FeCrSi fumes entering a baghouse.
D-24
-------
g
&
'i
DC
UJ
O
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
SAMPLE NO.
D-10
D-11
D-13
D-14
D-15
D-16
D-17
MMD, microns
1.50
1.26
0.74
0.86
0.48
0.83
0.80
10 20 30 40 50 60 70 80
CUMULATIVE PERCENK STATED MICRON SIZE
90
100
Figure D-9. Particle size distribution of FeCrSi fumes from a baghouse serving an open furnace.
D-25
-------
7.0
6.0
5.0
4.0
3.0
2.0
o
LU
a:
-------
I
'I
o
LU
O
I
Q.
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
10 20 30 40 50 60 70 80
CUMULATIVE PERCENK STATED MICRON SIZE
90
95
100
Figure D-11. Particle size distribution of (HC)FeCr fumes from a precipitator serving an open
furnace.
D-27
-------
Q
LU
_l
O
cc
9.0
8.0
'1.0
6.0
5.0
4.0
3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
5
10
SAMPLE NO.
G-l
G-2
MMD, microns
2.4
5.1
20 30 40 50 60 70 80
CUMULATIVE PERCENT< STATED MICRON SIZE
90
95
Figure D-12. Particle size distribution of SiMn fumes entering a scrubber serving an open furnace.
D-28
-------
v>
s
3.0
2.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
u
P 0.3
cc
cu
0.2
0.1
10 20 30 40 50 60 70 80 90
CUMULATIVE PERCENT< STATED MICRON SIZE
100
Figure D-13. Particle size distribution of SiMn fumes from a scrubber serving an open furnace.
D-29
-------
E-I. CHEMICAL ANALYSES OF PARTICIPATE EMISSIONS
FROM
FERROALLOY SMELTING OPERATIONS
1. INTRODUCTION
Chemical analyses were made on particulate samples collected by the EPA
sampling train using the EPA sampling method 5 described in Appendix C of
this report. The samples were obtained from 10 ferroalloy furnaces, and
were sent to private contractors for analysis. Metal analyses were made
by either the optical emission spectrography or the atomic absorption
method. In a few instances, the same samples underwent the electron beam
microanalysis. A total of 87 separate metal analyses covering 25 test runs
were made. These are reported as metals and not as metal oxides which
appear in the samples.
Table E-l shows the number and types of analyses made of the 10 furnaces
tested. In addition to those shown in the table, analyses were made of the
anions, NH,+, and water insolubles of the collected particulates from
Furnaces D,H,K, and L.
E-l
-------
>
Q
LU
_l
O_
*>^
-
_j
c£
§
Ll_
0
CO
LU
Q_
>~
i —
Q
ST
,
-Q 03 <— CU U>
E 4-> 03 T3 03
13 Ol C 03 i—
2: E ti E co
CO
CO
4- T3 3
o co cu "a
CU 4- 4-> •
s- to o o a.
CU i— >> CU -i-
-Q (O r— CU r— CJ
E +-> 03 "O i— CU
=3 CU C 03 O S-
2: E i— >-, -a JQ
jD rn r— CU -i- 3
E •<-> 03 T3 r— S_
=3 CU c: 03 O O
^ E 03 E t/i co
CO
cu
4~ !
O CO Q.
CU 4- E
S- CO O O3
CU r— >, CO
-Q 03 i — CU
E +-> 03 T3 CU
3 CU C OJ S-
2: E 03 E O
CO
i —
4- -Q
O •!- CO
-M CU
S- CO CO
CU 3 >,
JD -Q i— CU
E E 03 T3
=3 O C 03
S O fO E
4—
O ul
U CU
S- -i- CO
CU C ^
-Q 03 i — CU
E CD It! -D
13 S- C 03
Z O 03 E
4- co
O CO +->
CU 4- to
i- co O CU
CD r- >j +J
JI3 O3 i — CU
E -(-> 03 "0 T3
CO CU
S- CU N
CU 4-> >,
-Q . r-
F— ^3" fO
3 D_ C
2T LU 03
CU
U
03
cr
s-
u_
"O
CM
. .- ---_ -^ - -- _-- .----- - _-----_-.-----_- =--.= .- -,»---_ •_ ^
T3
T3
Is-.
•a TS
^ — - , — X
cu cu
S- S-
O 0
C i-
2: <-J
CM CM
OJ CM CO
CO OJ CO
-Q -Q -Q CJ
fO fO O fO _Q "O "O "O ^O "O
c\jcMvococMcn«*cnLr)O
CMCM'si-'^CM'^-CM^l-COCO
=CCQOQLUU.CD3:i>
-C
o.
s_
en
2
4J
O
ai
CL
CO
c
o
•r—
to
CO
>r_
E
Ol
ITS
O
•1—
0.
o
-o
c
03
C
o
•r-
a.
o
to
-Q
*O
U
•P
re
-^
-Q
ai
•a
E
>
01
to
^^
^2
re
c
re
re
4^
O)
re j:
•
to
ai
in
i~«
re
c
re
£
u
•r-
->.
re
s-
X
1
ai
jz
<~
e
o
Ol
ai
Ol
•r-
4^
re
-(->
•r—
re
3
3 (
•
•a
o
jr
ai
c
o
a.
s-
o
-Q
re
o
•r-
O
4^
re
^.^
J3
•o
ai
•r-
Oj
cu
(/I
O)
to
^.
^H
re
c
re
re
4_>
Ol
.
J=
a.
s-
CO
g
4->
O
Ol
Q.
to
c
o
to
to
•r—
E
Ol
o
'£
a.
o
^^^
_Q^
•o
-------
FURNACES A & B
£-11. CHEMICAL ANALYSIS OF EMISSIONS
FROM
A FERROCHROMESILICON FURNACE (A)
AND A CHROME ORE/LIME MELT FURNACE (B)
1. INTRODUCTION
Participate fumes and gaseous emissions are generated during the
smelting and pouring of a commercially important class of ferroalloy
materials called reactive metals. The particulate portion of these
emissions has been collected on glass fiber filters, strategically
placed in the air stream of an exhaust system. Fourteen such filters
from Furnaces A and B were analyzed by microscope, X-ray diffraction,
atomic absorption, electron beam X-ray microanalysis, and optical
emission spectroscopy.
The analytical results are presented in the following sections where
it is shown that the particulate specimens from the two furnaces are
distinctly different from the standpoints of chemical composition and
crystallographic structure. The samples from Furnace A consist princi-
pally of non-crystalline fused silica (Si02) with impurities. Impurities
present in concentrations;-! weight percent are Mg, Cr, and Zn in decreas-
ing order. In contrast, the samples from Furnace B contain crystalline
material with the inverse spinel structure such as typified by Fe304.
Instead of consisting mostly of Si02 as seen for the Furnace A specimens,
the specimens from Furnace B consist of chromium, silicon, magnesium,
iron, and zinc all in the 4 to 18 weight percent range along with chemical-
ly combined oxygen. This is, the specimens from Furnace B consist of
metal oxides.
E-3
-------
FURNACES A & B
2. TEST RESULTS
2.1 Optical Examination
The specimens were examined at magnifications up to 100X. Figure E-l
shows Specimen 25W-1 and shows the manner in which all specimens were
divided for individual analysis. The specimens from Furnace B were yellow-
brown and distinctly different from the gray-colored Furnace A specimens.
2.2 X-Ray Diffraction Analysis
X-ray diffraction occurs when a crystalline substance is exposed to a
beam of X-rays. The angle between the diffracted beam and the incident
beam is always 20, or twice the angle of incidence. By using monochromatic
X-rays of wavelength X, the interplanar spacing d of various planes in
a crystal can be found by using Bragg's Law, X = 2d sin 0. An electronic
detector or photographic film is used to record 0 angles and the intensities
of the diffracted beams. Every crystalline substance has an unique X-ray
pattern comprised of many 9 angles (usually converted to d-spacings) and
associated intensity values. Over 22,000 X-ray diffraction patterns have
been published to date.
The diffraction samples were prepared by removing the powders from
the individual filters, thoroughly mixing each powder manually in a plastic
container with a wooden tongue depressor, and pressing into 1/2-inch
diameter pellets under 80,000 psi. This method of specimen removal from
the quartz (Si02) filter in no way disturbed the filter. No filter particles
mixed with the specimens removed. In fact, a small quantity of powder
remained on the filter after removal of the specimens. These pellets were
analyzed on a G.E. XRD-5 X-ray unit. The instrumental settings used are
listed in Table E-2.
The diffraction patterns from all samples were weak; therefore, a
chromium tube was used as a source of X-rays in order to reduce background
radiation due to X-ray fluorescence. The use of the chromium X-ray tube
and pulse height analysis maximized the signal/noise ratio.
E-4
-------
FURNACES A & B
FILTER
SIZE OF SAMPLE
TAKEN FROM EACH
FILTER FOR ELECTRON
BEAM MICROANALYSIS
25W-1
SAMPLE FOR ATOMIC
ABSORPTION ANALYSIS
PETRI DISH
SAMPLE FOR X-RAY
DIFFRACTION ANALYSIS
SAMPLE FOR OPTICAL-
EMISSION SPECTROSCOPY
ANALYSIS
Figure E-1. Drawing of condensate from ferrochrome operation showing
area analyzed.
E-5
-------
FURNACES A & B
TABLE E-2.
X-RAY DIFFRACTOMETER SETTINGS
X-ray Source: Chromium Tube; 50KVP, 20 ma, no filter
Beam Slit: 1°
Soller Slit: Medium Resolution
Exit Slit: 0.1°
Table Speed: 2°/Min; Chart Speed 2"/rnin.
Detector: Flow proportional
Scale: Linear 100
Pulse Height Selector: El = 2V with Gain x 16.
AE = 6V
E-6
-------
FURNACES A & B
The diffraction results are summarized in Figure E-2 which shows the
sample identification, and the d-spacings and relative intensities of the
diffracted beams. These patterns were then compared with tables of known
*
diffraction patterns.
The X-ray results can be summarized as follows:
1. All specimens were largely non-crystalline as evidenced by an
absence of a diffraction pattern or very weak diffuse patterns
with very few lines. No X-ray diffraction patterns were obtained
from Specimens 25W-1, 25W-2, 25E-1, and 25E-2; they were completely
**
non-crystalline.
2. Another eight samples had weak patterns, but the patterns could
not be correlated in a meaningful way with any known pattern
from the diffraction file. In a few instances, a force-fit
might have been possible but the choices were hydrated crystals
such as Ca3Al8(PO,J8(OH)6-15H20. It seemed unlikely that a
highly hydrated and complex crystal would have formed during the
few microseconds available for emissions to condense from the
gaseous high temperature effluent. These eight patterns were,
therefore, classified as unknown.
3. Recognizable patterns were obtained from both specimens from
Furnace B. The patterns belong to the naturally occurring class
***
of compounds called spinels. It was not possible to positively
tell which particular spinel oxide was present but the best fit
to the X-ray data include:
Joint Committee on Powder Diffraction Standards, Powder Diffraction File.
Swarthmore, Pennsylvania, 1969.
The specimen numbers designate the Furnace, A or B, and duct, west
or east, from which specimens, 1 or 2, were taken simultaneously. For
example, specimen 25W-1 is Sample #1 taken from the west (W) duct of
Furnace A . Specimen 25E-1 was taken from the east duct of Furnace A
at the same time specimen 25W-1 was taken.
The spinel group includes a large number of oxides of the general
formula AB2C\. The more familiar members of the spinel group are
MgAl^, ZnFe2Ou, CdFe^, FeAl204, CoAl^, NiAlzO^, MnAl204, and
ZnAl204. Inverse spinels have the same X-ray pattern, are more common
1n nature, and include FeO-Fe203 also written as Fe3(V
E-7
-------
FURNACES A & B
jH-r> -i-H-r-
to
-1J
(O
-
o
(O
2
x
2
^
en
-------
FURNACES A & B
Chromite: FeO-[Crx Al1_x]2 03 0.64
-------
FURNACES A & B
The individual elemental concentrations obtained from the aqua regia
leach and the filtrate from the Na2C03 fusion were then summed for each
sample; these results are compiled in Table E-3.Samples 6E-1 and 6E-2 were
found to be Insoluble in the Na2C03 flux and as a consequence of this, it
was not possible initially to obtain Si02 or further elemental analysis on
these two samples using the above method. These two samples were found to
be soluble in a potassium pyrosulfate flux. The samples were, therefore,
fused with approximately 0.5 gram of potassium pyrosulfate, and the resultant
fused samples were then dissolved in dilute HC1. The solution was then
filtered; the filtrate made up to 100 cc volume, and subsequently analyzed
by Atomic Absorption Spectroscopy (A.A.)- The residue on the filter paper
was then ignited in a muffle furnace at 900°C and the residue back weighed
as Si02. The results of these analyses were added to the results found
for the acid extracted portion of the sample, and are tabulated in Table E-3.
Since the optical-emission spectroscopy analyses discussed later showed
that Specimens 6E-1 and 6E-2 contained considerable quantities of calcium,
A.A. analyses for calcium were run on these two specimens also. The instru-
mental parameters used for each element are listed in Table E-4. Nitrous
oxide-acetylene flames were used for Cr and Mg to eliminate inter-element
interferences. Because of the small amount of sample collected (30-120 mg)
and the desirability of determining the toxic elements (Mn, Cd, Pb, As, Hg,
Be) in low concentration levels, it was found necessary to use the entire
sample for each of the above analyses.
The choice of these elements was based on the combined considerations
of (i) expected presence in condensate, (ii) toxicity, and (iii) availability
of atomic absorption lamps. The results are summarized as follows:
1. The samples from Furnace A invariably contain at least 66 wt%
Si02 with an average value of 73.6 wt%. In contrast, the
specimens from B contained only -6 wt% Si02 as found in a
supplemental optical-emission analysis. The Si02 did not come
from the filter paper because the sample was removed from the
quartz (Si02) paper prior to analysis.
Walter Slavin, Atomic Absorption Spectroscopy, Interscience Publishers,
New York, New York, pp 79-189, 1968.
E-10
-------
FURNACES A & B
«c
:D
• Q. °3
ro
I U_ <;
LU O
l/>
LU CO UJ
_1 i—i O
CQ co <:
< >- Z
h- —I CC
x>
o
(O
o
=*=
O)
a.
•6
CO
Q
111
C\, •— «3-*3-rocnmini£>r~-'— CM j CMCO
Q .......••••
CMI— «d-vocMO'^>J3inr^Lr>(_< cxi oo
oo r^ooi • r">- r-^ F^-. ^o ^ fr~- ^ r\
en CD m *J"
cxj 00 r-^ in cvj .— r- vo «3- •— cnCM cno
o-— oooi— «— «— '—| CM o CD m m
CDOOOOOOOCDOOO O O
v v
oorou5«3-cn'«j-orvjcorocsio
c: v£»*3~m^^rocoinLnr--Ln^3'ro cnio
O'— OCDOOOOCDOO COi—
cu u-,cor~^o^'roocr\cMCM ~
u_ ^-'— roro^3-^j-«d~roCM^j-c\jcMc-ri
OOOOOOOOOOOO «*' iri
r^cocMini— ^orocncMr^i— •— v£>ro
en ............
^j-r~~r^vor^^^'— •— ooiocom
Oi = = = = = = zr = :
O
O
cu o
co
o
fO
o, g
ro CO CM
o o o
000
O
CM
O
O
O
o
V
ro
O
CD ^
CD
00
o
o -
o
o
V
CO CO
o o
o o
CD O
O o
V
r-. ro m
O O O
000
O O O
o o o
V
*f ro
o o
O o
O O
O CD
V
r-^ ro
VO «^~
o o
o o
<~> CD
in oo
O ro r—
O o O
o o o
O CD CD
cnin1— icoo«i-cn«4-»3-
invoTOincMor^-covo
OOCDCDO-— OOO
i-» i— i~~- r>-
CO r^ ro o
O CD O r—
CM
o
O O O O CD O O
co «*• CM co ic CM m
vo en CM o vo oo •—
CD O O O
VO
en
CO
CM r— r—
CM CM CM CM
I I
in in in in
I
3:
I
U'
CM CM
I
O O CM CO
CM
1
LU LU LU
CM
I
ininininLnLnin LULU
CMCMCMCMCMCMCMCMCMCMCMCM
0)
o
(6
t-
3
O)
O> O
4-
(U
I- -I— " Cl)
rsco a) E
-o s- s_ o
00) i.
+-> cu jr x:
cou. o
o- — • i/i ~— •
O) 3
.O CU f
u ;*»
3 c-Q
-o cu.
W S- CO ro S.
i/i a. 4-> u>
o c cu E -1
•O C Irtr-
•i- D> C O)
O.r- CU
ai-a cr-o
x: <*- cu
** cu ° "> >CL
"
i —
* *- QL
i- o-=-
c « raw «
°oo °- *- v>
(O -Q. O T3
E-n
-------
FURNACES A & B
TABLE E-4.
INSTRUMENTAL PARAMETERS
ELEMENT
Cr
Mn
Cd
Pb
Hg
Be
V
Mg
Fe
Zn
Al
Ca
WAVELENGTH
o
(A)
3579
2801
2288
2833
2537
2348
3184
2852
2483
2138
3093
4227
FUEL OXIDIZER
SYSTEM
N20
air
air
air
air
N20
N20
N2°
air
air
N20
N2°
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
acetylene
SLIT
WIDTH (A)
2
2
7
7
7
7
7
7
2
7
7
7
HOLLOW CATHODE
CURRENT (MA)
10
10
4
4
4
12
15
4
12
8
13
8
E-12
-------
FURNACES A & B
2. The chromium content ranged from 0.68 to 2.08 wt% from Furnace
A and the average value was 1.3 wt%. The value from Furnace
B is much higher at 7.2 wt%.
3. The average manganese content was 0.067 wt% in Furnace A and
0.062 wt% in B. The values are virtually the same.
4. The magnesium content varied considerably among different
samples. It was highest, 11.6 wt%,in the two 25E-1 specimens.
These specimens also had the lowest Si02 levels (67.5 and
66.6 wt%) from among the samples from Furnace A.
5. The iron content in Furnace B was a full factor of 10
higher than in Furnace A- The average values in Furnace B
and A are 3.75 and 0.34 wt%, respectively.
6. The zinc content in Furnace A varied from 0.30 to 1.68 wt%
with an average of 1.04 wt%. The corresponding value for
Furnace B is 2.20 wt%.
7. The average aluminum content from Furnace A was 0.10 wt%
and,hence, lower than the 0.545 wt% found in the two samples
from Furnace B.
8. Mercury, beryllium, and vanadium, all toxic elements, were
below the detectability limits of 0.05 wt%, 0.001 wt%, and
0.03 wt%, respectively.
9. The cadmium levels from Furnace A varied from below 0.0003
wt% to 0.002 wt%. The condensate from Furnace B contained
an average of 0.0024 wt% cadmium which was somewhat higher
than that for Furnace A j,ut still relatively low. Cadmium
is a toxic element.
10. Calcium analyses of Specimens 6E-1 and 6E-2 from Furnace B
were decided on only after it was seen from the optical-emission
spectroscopy results that the calcium levels were very high
compared to the Furnace A specimens. The A.A. analyses for
E-13
-------
FURNACES A & B
calcium yielded 10.6 wt% and 10.7 wt% for the two specimens.
Since the A.A. technique is more exacting than the optical-
emission technique, it is the A.A. calcium values which should
be considered as being the true calcium concentrations in the
Furnace B specimens.
11. The total concentration of elements from Furnace A samples
1s virtually 100% when all the metal values are converted to
*
equivalent oxide percentages. This means that all the major
elements in the emissions from this furnace were detected and,
in addition, a few minor but toxic elements (V, Hg, Be, Cd) were
also detected.
12. The total concentration of elements from Furnace B after
conversion to equivalent oxide percentages is 70%, a somewhat
less satisfactory mass balance situation than for Furnace A .
This lack-of-closure should not be taken to signify the presence
of an additional but undetected element. No additional element
of any consequence was detected in either the electron microprobe
or optical-emission techniques. It is concluded that all the
major elements are accounted for in Table E-3 and that the lack-of-
closure in samples from Furnace B is due to errors associated
with the extreme difficulty encountered in dissolving the samples.
2.4 Electron Beam X-Ray Microanalysis
The electron microprobe is an advanced piece of equipment which uses
a small beam of electrons to produce characteristic X-ray emissions from a
sample volume with a radius of -1 micron. Curved crystal X-ray spectrometers
are used to analyze the resultant characteristic X-ray spectra. In these
analyses, the electron beam was defocused to a diameter of 200 microns
(0.008 inch) to cover a larger segment of the sample.
Equivalent oxide percentages are obtained by multiplying the weight
percent metal in Table E-3 by the ratio Mo/Mm where Mo is the molecular
weight of the metal oxide and Mm is that of the metal. The oxide
formulae were taken to be A1203, ZnO, Fe^O,,, MgO, A203, and CaO. Thus
for Ca, the equivalent oxide percentage is 10.65 x (40+16)/40. Justi-
fication for this conversion is based on electron microprobe results.
E-14
-------
FURNACES A & B
The electron beam Impinged in vacuum upon the untouched sample surface
as shown in Figure E-l. An examination was made of the complex spectrum of
X-rays given off by the specimen under electron beam excitation, and it was
found that the entire spectrum could be identified uniquely on the basis of
the elements shown in Table E-4. All portions of the X-ray spectrum in the
wavelength range 1-100A covering all elements except H, He, Li, and Be were
taken into account.
The silicon and oxygen signals did not originate from the silica
filters although the latter were present in the electron microprobe chamber.
The electron beam penetrated about 2 microns (and absolutely no more than
20 microns) into the sample from the top surface. The total sample thickness
was about 0.02 inch (-500 microns). Thus, the silica filter material was
-500 microns away from the effective sensing depth of the electron beam.
The major outcome of the electron microprobe analyses was that the
main elements were identified for the atomic absorption analysis already
discussed. Thus, Fe, Cr, Si, Al, Ca, Mg, and Zn were found on the untouched
samples and were, therefore, selected along with other elements for A.A.
analyses.
A second outcome of the electron microprobe analyses was the detection
of oxygen at roughly the 50% level in the samples from both furnaces. This
means that the metals are present as oxides and is the basis for the
conversion of the metal percent values in Table E-3 to equivalent oxide
percents. The 50% oxygen value was strictly applicable only to the top
2-20 microns of the untouched samples where the analyses were made. However,
1t was assumed that the sample was essentially a mixture of oxides
throughout its depth. Such an assumption seemed reasonable when the
source of the samples was taken into account.
The 50% value was obtained in a 10"6 torr vacuum. Thus oxygen was not an
occluded atmospheric gas but was present as an oxide.
E-15
-------
FURNACES A & B
Table E-5. ELECTRON BEAM X-RAY MICROANALYSIS RESULTS FROM
QUALITATIVE ANALYSES
Camnlp
Elements Positively Identified in X-ray Spectra
25E-r
25E-23
Fe, Cr, 0, Si, Al. Ca, Mg, Zn, Na, Ba, K, -, - -
Fe, Cr, 0, Si, Al, Ca, Mg, -, -, -, -, Ni, -, -
Fe, Cr, 0, Si, Al, Ca, Mg, Zn, Na, -, -, Ni, Cl, S
Ferrochromesilicon furnace (Furnace A).
Chrome ore/lime melt furnace (Furnace b).
E-16
-------
FURNACES A & B
The concentrations are not given in the electron microprobe table (Table E-5)
because, although the elements shown were present throughout the depth
of the samples, their concentrations (particularly the metals) varied with
depth (i.e., the samples were non-uniform). Thus, the atomic absorption
analyses were used to determine the quantitative analyses on properly
composited samples while the electron microprobe qualitatively identified
the elements.
2.5 Optical-Emission Spectroscopy
Optical-emission spectroscopy or arc-spark spectroscopy consists of
electrical excitation of the electrons of the elements in the sample. When
the electrons return to their ground state, light is emitted. The emitted
light is passed through a prism or diffraction grating to separate it into
its component wavelengths. The spectrum is then analyzed electronically
or optically on a photographic plate. Each line occurring at a definite
wavelength position on the spectrum designates a specific element, and
the intensity of light at that wavelength is proportional to the quantity
of that element present.
Portions of the samples were subjected to optical-emission analyses
to provide (1) a check on the analytical procedures (particularly the
lack-of-closure in the atomic absorption analyses from Furnace g ), and (ii)
« more sensitive approach to trace element analysis than that provided by
electron beam X-ray microanalysis. The spark emission results for the
isajor elements agreed well with the atomic absorption and electron beam
results and, in addition, identified numerous trace impurities not found
1n the other approaches. The results are compiled in Table F-6.
E-17
-------
FURNACES A & B
>-
3
§
to
s;
a.
o
Ol
ra
a.
CO
£
0
<_)
3
CJ
c
z
TO
10
c
!_
(U
U-
5
Di
s:
to
o
QJ
n
r
«j
v)
0
o
LO
0
O
0
o
0
0
o
o
0
o
o
o
ID
0
0
0
ID
0
o
in
0
o
0
00
o
CO
o
CM
«3-
O
*!-
ra
I
!X
LO
CM-
CD
0
LO
o
o
o
§
o
o
o
o
o
o
o
in
0
o
o
LO
0
0
LD
o
o
o
«*
o
0
CM
O
CM
vo
§
CM
1
3C
LO
CM
o
o
LO
0
0
0
CSI
0
o
0
0
o
0
0
0
o
LO
o
o
o
LD
0
o
LO
o
o
o
vo
o
CM
o
•=»•
0
CM
VO
0
*3-
CM
1
:z
LO
CM
O
o
LO
o
o
8
o
0
o
0
LO
o
ID
o
o
ID
o
o
0
**
o
n-
o
**-
o
CM
VO
o
*»•
CM
n
ID
CM
o
o
o
0
o
o
o
0
o
o
o
o
o
LD
0
o
o
LD
0
o
LD
o
o
o
10
o
CM
o
«r
0
CM
VO
o
•el-
CM
1
3
LT>
CM
o
0
0
0
CM
o
o
o
o
o
o
g
o
ID
o
o
o
LO
o
o
ID
o
o
o
00
o
CM
o
"
CM
vo
o
*!•
1
LU
LD
CM
o
o
CM
o
o
CM
o
o
0
o
o
o
0
o
o
ID
o
o
0
ID
o
o
LD
o
0
o
00
o
CM
o
«d-
o
CM
o
§
1
LU
LO
CM
o
o
LD
o
o
o
CM
o
o
o
o
o
o
8
o
LD
o
o
o
LD
o
o
LD
o
o
o
CO
o
CM
o
o
LD
CM
vo
§
1
UJ
LO
CM
o
o
LO
o
o
o
CM
§
0
§
o
o
o
o
LO
o
o
o
ID
o
o
LO
o
o
o
«a-
o
CM
o
o
LD
C*J
CO
o
*»•
1
UJ
ID
CM
o
0
iD
o
o
o
CM
§
o
o
o
o
o
o
o
LO
o
o
o
LO
o
o
LO
o
o
o
«*
o
CM
o
o
CM
vo
o
«<•
CM
1
UJ
LD
CM
o
o
LD
o
o
o
o
o
o
0
o
o
o
o
o
LO
8
0
LO
o
o
to
o
o
o
vo
o
o
o
CM
vo
o
«3'
CM
1
LU
LD
CM
o
o
CM
o
o
o
o
o
o
o
o
8
o
LO
o
o
0
LO
o
o
LO
o
0
o
«3-
o
CM
o
0
CM
«r
o
«3-
CM
UJ
LO
CM
o
o
LD
o
o
o
f>
o
o
m
o
o
o
o
o
fl
o
o
LO
o
o
ro
o
o
0
CM
CM
o
o
n
o
n
JO
1
LU
vo
o
0
LO
o
0
o
ro
o
o
n
o
o
o
o
o
fl
o
o
LO
o
o
ro
o
o
o
CM
ro
o
o
CM
o
ro
CM
I
LU
VO
ra
u
01
a>
-a
XI
OJ
u
CD
ai
x> z: z: s: s: z: z:
O- a. o. CL. a. a.
4-> o. a. o. o. a. a.
o
c o o o o CM o
o o o LO o
3 CM CM CM I—
•o
O)
01
E
OJ
I J3 rjj "X3
. tooa <->
at
(J
ro
c
s_
ai
VJ
ro
c
s_
c
o
o
O
o
0
O)
1-
ra
T3
vll
ai
u
vll
Q.
O
ai
u
ra
e
i.
-------
FURNACE C
III. CHEMICAL ANALYSIS OF EMISSIONS
FROM
A SILICOMANGANESE FURNACE
1. INTRODUCTION
Particulate fumes and gaseous emissions are generated during the
smelting and pouring of a commercially important class of ferroalloy
materials called reactive metals. The particulate portion of these
emissions has been collected on glass fiber filters, strategically
placed in the air stream of an exhaust system. Six such filter samples
taken from a scrubber serving a SiMn furnace and an uncontrolled tapping
station were analyzed by microscope, qualitative electron beam X-ray
microanalysis, and atomic absorption. The analytical results are presented
in the following sections. Sample specimen designations preceded by
MIE are filter samples collected at the scrubber outlet; those preceded
by WTE and ETE are samples collected at the tapping station.
2. TEST RESULTS
2.1 Optical Examination
The specimens were examined at magnifications up to 30X. The
particulate matter could be seen intimately mixed with the quartz filter
fibers, and it was obvious that the particulate matter could not be
physically separated from the filter pad upon which it had been collected.
Under tungsten filament the specimens appeared as follows:MlE-6 (light
gray), M1E-T (dark gray), M1E-12 (yellow-brown) and ETE-2 (partly dark
brown, partly light brown).
E-19
-------
FURNACE C
2.2 Electron Beam X-Ray Microanalysts
The electron microprobe is an advanced piece of equipment which uses
a small beam of electrons to produce characteristic X-ray emissions from
a sample volume with a radius of —1 micron. Curved crystal X-ray
spectrometers are used to analyze the resultant characteristic X-ray
spectra. In these analyses, the electron beam was defocused to a diameter
of 150 microns (0.006 inch) to cover a relatively large area of each
specimen and thereby obtain data which would be representative of the
entire sample. The electron beam impinged in vacuum upon the untouched
sample surface. An examination was made of the complex spectrum of X-rays
given off by the specimen under electron beam excitation, and it was found
that the entire spectrum could be identified uniquely on the basis of the
elements shown in Table E-7.A11 portions of the X-ray spectrum in the
wavelength range 1-100A covering all elements except H, He, Li, and Be
were taken into account.
The analyses were conducted on small portions of the filter pads which
were not later digested for the atomic absorption analyses. The small
samples for electron probe analyses are still intact. The qualitative
analysis results are summarized in Table E-T.Several points seem germane:
I. The major elements are manganese, magnesium, calcium, and
potassium. The silicon signal could have come from either
the filter pad or from the particulate matter. The fibers
of the pad were visible in the optical microscope which is
attached to the electron probe.
E-20
-------
FURNACE C
2. Distinct signals, equivalent to several weight percent,
were found for sulfur, chlorine, carbon, sodium, and
potassium.
3. The presence of sodium and chlorine frequently suggests
salt and could have come from handling with bare hands.
However, it must be stated that the filters were not
handled with bare hands during the chemical analysis
effort.
2.3 Atomic Absorption Analyses
Atomic Absorption (A.A.) means that a cloud of atoms in the
un-ionized and unexcited state is capable of absorbing radiation
at wavelengths that are specific in nature and characteristic of
the element in consideration. The atomic absorption spectro-
photometer used in these analyses consists of a series of lamps
which emit the spectra of the elements determined, a gas burner
to produce an atomic vapor of the sample, a monochromator to
isolate the wavelengths of interest, a detector to monitor the
change of absorption due to the specimen, and a readout meter to
visualize this change in absorption.
E-21
-------
FURNACE C
The filters with samples were weighed, and the sample weights
calculated by subtracting the tare weights written on the outside
of the Petri dish sample containers from the total weights.
Specimen identifications were as follows: MlE-6, M1E-7, M1E-12,
and a composited specimen consisting of filters WTE-3, ETE-1,
and ETE-2.
E-22
-------
FURNACE C
CO
UJ
CO
§
CK
O
I
X
o
UJ
_J
UJ
o-
OJ
XI
ro
Source
o
c^
CO
o
c
:*:
ro
<0
CO
re
o
re o •»->
°V ^ '5
UJ i-
K— OJ 10
5 & "re
J= Q-
4-> CT>
••- C l-
3 -r- OJ
•O -r- r-
rjj v> -i-
+J O <*-
•r- O.
VI E QJ
O O QJ
Q.OI-
QJ
0)
to
OJ
I/)
OJ
•o
O)
O (/)
J- OJ
O. CD
O -i-
J- -O
u
'e o
c
OJ
-C M
c
o
'•fj
3
'o
1/1
rjj 4->
I- CD
O
CVJ
QJ 3
4J O
re CM o
QJ I r—
J- O I
0> r- r-
S-
QJ
re i/i
f— OJ
to :>,
re •—
S ">
c
CM re
^ i
UJ T-
re
JC
in
01
O)
»
3:
a
II
II
O-
Q- O
re xi
co «
ui
6
(/I
QJ
CJ>
-O
OJ
o
c
u
re
o
o
o
OJ
o
i~
O)
ro
O
O
o
Q.
ro
O
E-23
-------
FURNACE C
The filters (both pad and participate) were extracted for 2 hours with
50 mis of 1:1 H^SO^; this solution was decanted and saved. The filters were
then extracted for 2 hours with a boiling dilute Aqua Regia solution.* This
solution was then combined with the H2$04 solution, filtered, taken to con-
stant volume in a volumetric flask, and analyzed by A.A. The instrumental
parameters used for the individual elements are given in Table E-8 and the
results of the A.A.** analyses are shown in Table E-9.
The A.A. results are normalized to compensate for the portion removed
for electron beam X-ray microanalyses. Silica (Si02) analyses were not per-
formed since the entire Si02 filter pad with the intimately mixed specimen
was digested in acid in each case. In order to determine if the acid diges-
tion process chemically attacks the Si02 filter, an unused filter will be
exposed to the digestion process and A.A. analyses run on a blank to deter-
mine background concentrations. A remote possibility exists that some K or
Na could have come from the filter material if it is not pure Si02.
Note also that in Table E-9 only one specimen (M1E-6) adds up to 100%.
The other specimens most likely consist of metal oxides or a mixture of
metals and metal oxides; oxygen and silicon are therefore the likely
missing chemical components needed to bring the totals in all four cases
to 100%.
A comparison of these results with the results reported for Furnace A
immediately brings to light
certain differences between the two emissions. The emissions from
Furnace C contain, in general, less chromium but more manganese,
sodium, potassium, chlorine, and,to be particularly noted, sulfur, than
Furnace A emissions.
R. J. Thompson, G. B. Morgan, and L. J. Purdue, "Analysis of Selected Ele-
ments in Atmospheric Particulate Matter by Atomic Absorption", Atornic
Absorption NEWS Letter. Volume 9, No. 3, 1970
**
Walter Slavin, Atomic Absorption Spectrescopy, Interscience Publishers,
New York, New York, pp 79-189, 1968.
E-24
-------
Table E-8. Instrumental Parameters
FURNACE C
Element
Cr
Mn
Mg
Fe
Al
Ca
Ba
Na
K
Zn
Wavelength
(I)
3579
2801
2852
2483
3093
4227
5536
5890
7665
2139
Fuel Oxidizer
System
Air-Acetylene
Air-Acetylene
N^O-Acetylene
Air-Acetylene
N20-Acetylene
NgO-Acetylene
NgO-Acetylene
Air-Acetylene
Air-Acetylene
Air-Acetylene
Slit 0
Width (A)
2
2
7
2
7
7
7
20
20
7
Hollow Cathode
Current (Ma)
10
10
4
12
13
8
7
10
10
8
E-25
-------
FURNACE C
c
o
C
cu
o
o
o
C
O)
OJ
Ol
OS
00
•r-
in
o
CL
o
to
§
a>
JO
IO
s
o
h-
M
*
TX
CO
ro
CJ
,_
OJ
a
c
2:
*-
c
Q>
f^ •
•r- O
U Z
QJ
CL
co
CM OO CM
•* CO OO
0> °° 5
' 0 CD
V
CO CM OO
r-- co **•
r^ •& CQ
• — ^ r-~
CM ^f OO
1 1 1
• • •
O O O
t£) ijfji &)
0 0 O
O CM CO
r- C3 O
oo CM CM
CM r— ,
vo 1
.^- ,»LU
It CO 1—
O 1 UJ
CLU.I
o -xl:
CJ> — ro
,
-a
QJ
1_
0
QJ
Q.
4-3
O
oo
QJ
00
t"^
ro
c
c:
o
_u
^_
•r-
OO
Ul
0
z*'
1
<*-
00
c
o
•r-
•*-*
jq
• r-
S-
c
0
o
i.
o
*4—
c
3
0
(J
(J
rO
O
+J
0)
JJ->
13
C
•(•"•
E
-M
oo
QJ
^>
-M
to
-o
Ol
I *
L>
OJ
5^
O
(_>
C
O
0) OJ
J= -C
K— -M
Ul
h-
O
z
1
c^
i [
(j
Ul
O.
QJ
QJ -C
(J 4->
ro
C 0)
S- C
'S >
i.
C 0)
S^ oo
00 00
(O QJ
•4-J
cn oo
c >,
•i— OO
S- +J
QJ 00
00 3
ro
•»-> -C
OJ X
-M
3 "O
O 0)
i- ^
QJ O
c*j 5^.
J3 4->
3 C
S- 0
U (J
oo c:
+j 3
-o
QJ "O
•(-> QJ
0 +->
O) (J
i — QJ
o 'o
(J
00
QJ oo
i — QJ
CL i —
e Q.
00 (O
00
S-
O> S-
4-» QJ
r~- +•>
4- -r-
QJ
S- QJ •
ns i. c
rO o
CM -n-
i — CM 4-»
1 1 03
UJ UJ i-
r— J— QJ
i> i 1 1 CX
o
•o -o
c: c co
(W ro C
^ l "^
1 1 Q.
UJ LU ro
•— )- -M
S^ LU
QJ
•• f r_J
UD OO ro
1 1 C
LU Ul S-
i— 1— Z3
s: s t-
E-26
-------
FURNACE C
EN RONMENTAL PROTECTION AGFj ^Y
Reserr~.. Triangle Park, North Carolina 27711
Reply to
Atln of: Date: 11-17-71
Subject: Analysis of Samples for Mercury
To: K. W. Grimley, Division of Applied Technology
THRU: R. E. Lee, Jr., Chief, 3SFAB y\L^
1. Origin: Furnace C
2. Date collected: 7/31 - 8/3/71
Date analyzed: 11/10/71
3. These seven samples of scrubber exit water were analyzed
for mercury using flameless atonic absorption.
4. Tests were run on both the clear liquid and the liquid
and sediment. There was no mercury detected in any of the
samples run.
5 The lower detectable limit is~.008 ug Hg/g.
Kathryn il. i'lac^eod
Source Sample £>na Fuels
Analysis Broach, DAS
cc: R. Neligan
A. Altshuller
J . i-icGinnity
D. Shearer
D. von Le.'hmden
D. Slaughter
R. Atherton
V7. Kelly
F. Uilshire
E-27
-------
< £3
FURNACE C
» s
a: is*
s s
CM «M
v* in o £>
»— t* ft
O (O in S M» § ^ £»
is. 5, ^r m MI *-••
m 10 r*. esj m V
O O f O
•fe
a.
s
i
ft g S S P
«O is. r^ r^. ut in
o o O O O O
51
9 V V
<9 9 9
ra
en
(U
•O
ai
a)
JC
-
5
o r> f^
I
a.
0)
o
x
OJ
0)
rtJ
o o o o o o o
V V V
S S S 8 S? S
fs. 10 r* *n m o»
•o
O)
•^
(U
(J
I
I/I
/?
S °
0000
S»-» M *— ^- 4" A « £*< Jo*> 3"
? 3s i «j»i t£ CtS Cj~
i* 1/lt.lXkvll.lXl-JlU
E-28
-------
ENVIRONMENTAL PROTECTION AGENCY FURNACE C
.'•<> .'o
• Dale: September 10, T
Ore and Slag Samples - Furnace C
To: Or. K. t. Lee
Source Sample and Fuels Analytical Branch
Division of Atmospheric Surveillance
1. Origin: Samples of two representative manganese ores and two
ferromanganase slags from the prociuction of silicomanganese alloy
in Furnace C during the MCA- EPA emission tests
(July 27-A-.v;st 5, U'71).
2. Samp.les : Sample 1 - Manganest Ore
Sample 2 - Manganese Ore
Sample 1 - Slag
Sample 2 - Slag
3. Analysis: Please analyze each sample separately for trace metals
by both, neutrcr, activation analysis and emission spectroscopy (on contract,
if necessary). Also analyze a representative portion for beryllium by
atomic absorption.
Trace ratals are dsfinod as ^h° following: Sb, As, Ba4 Be,.B, Cd,
Ca, Crs Cu, Fe, Pb, Hg, fin. Hg, Ni , K, Se, Si, Na, Sr, S, Sn, V and Zn.
4. Results: Forward results to the author.
D. R. Patrick
Chemica1
cc: VI. Basbagill
J. Dealy
S. Blacker
...pi. ?rl.rit» 1 MnxiM trinity I Jll~.it. Ock.r.
T™. fet. SI.. 1. >i CJ »t V ",. XI 'b tt :n LU ?b S« » F tl AS Sn re St Ml « C« 51 HI Al )
tn ~V» | O 1-3M <7M UX) M]»c 200 <100 100 O09 200 OO « »J 1 200 <1 to » 100 <100 100 <30 • M « 200 <1 <30 JX JOO MOO n 2000 (1 2000 }1
ttMfk«
OB
oa
ots
oa
1 "luw«**«-. Table E-ll. METAL ANALYSIS OF ORE AND SLAG SAMPLES.
E-29
-------
FURNACE D
IV. CHEMICAL ANALYSES OF EMISSIONS
FROM
A FERROCHROMESILICON FURNACE
1. INTRODUCTION
Particulate fumes and gaseous emissions are generated during the process-
ing of a commercially important class of ferroalloy materials called reactive
metals. The particulate portion of these emissions is collected on glass
fiber filters strategically placed in the air stream of a ventilation system.
Six such filters from samples collected on a FeCrSi furnace were analyzed by atomic
absorption and qualitative electron beam X-ray microanalysis. Each of the six
filters prior to compositing was examined microscopically.
2. TEST RESULTS
2.1 Optical Examination
The loaded filters were examined at magnifications up to SOX. Under
tungsten filament illumination the separate filters appeared as follows:
ABD-1M Dark gray powder with black particles-no quartz (baghouse inlet)
fibers from the collector pad visible.
ABD-2M Light gray powder with very few black particles- (baghouse inlet)
no quartz fibers from the collector pad visible.
ABD-3M Dark gray powder with black particles-quartz fibers (baghouse inlet)
from the collector pad visible.
ANE-1M Light gray powder with black particles-quartz fibers(baghouse outlet)
from the collector pad visible.
ACE-1M A few black particles among the quartz fibers. (baghouse outlet)
ASE-1M A few black particles among the quartz fibers. (baghouse outlet)
E-30
-------
FURNACE D
The optical examination revealed that:
1. Four filters had trapped a heterogeneous participate materia.
consisting predominantly of a gray powder and a minor amount
of black particles.
2. The amount of sample collected in four cases was so small that
the fibers from the filters could still be seen. In fact, in
two such samples, only a small amount of the black particles
could be seen against a background that was predominantly the
filter material.
Two different techniques were necessary to form composite samples:
1. Simple Blending of Loose Powders
Samples ABD-1M, ABD-2H, and ABD-3M were shaken, lightly scraped
and copious amounts of loose gray material were gathered, blended,
and designated as Baghouse Inlet Duct Sample ABD-M. A negligible
amount of the collector filter material was included in the blended
sample.
2. Dissolution in a Common Reagent
Samples ANE-1M, ACE-1M, and ASE-1M were submerged (particulate
matter and filter pads) in a common solution of sulfuric acid.
A control experiment was also run on a unused filter pad to
determine the contributions of the filter. The composited sam-
ple in this case was labeled Baghouse Outlet Stack Sample ABE-M.
Small samples for electron beam X-ray microanalysis were cut
from every specimen prior to formation of any composite samples.
E-31
-------
FURNACE D
2.2 Electron Beam X-Ray Microanalysis
The electron microprobe is an advanced piece of equipment which uses a
small beam of electrons to produce characteristic X-ray emissions from a sam-
ple volume with a radius of ~1 micron. Curved crystal X-ray spectrometers
are used to analyze the resultant characteristic X-ray spectra. An examina-
tion was made of the complex spectrum of X-rays given off by the specimen
under electron beam excitation, end it was found that the entire spectrum
could be identified uniquely. All portions of the X-rcy spectrum in the
wavelength range 1-100A covering all elements except H, He, Li, and Be were
taken into account.
In these analyses, the electron beam was defocused to a diameter of "150
microns (0.006 inch) to cover a relatively large area of the specimen and to
insure that both the gray condensate and the black particles were analyzed.
The electron beam impinged in vacuum on the untouched surfaces of three specimens:
1. Sample ABD-1M
In this sample, the layer of particulate material was far too
thick to allow penetration of the electron beam into the
collector (filter) pad. In other words, only the condensed
particulete material was analyzed in this case.
2. Sample ABD-3M
The layer of particulate was sufficiently thin that a contri-
bution from the collector pad may be present.
3. Sample ANE-1M
A contribution from the collector was definitely present in
this case because the fibers from the collector could be seen
in the optical microscope viewing system attached to the
electron microprobe.
The qualitative results are compiled in Table E-12 and provide the basis for
selection of elements for quantitative analyses. Note that a total of 15 ele-
ments were found* and that the stack sample (ANE-1M contained a small but
distinct amount of both sulfur and chlorine. Special mention is made of these
The spectral scans were conducted in a manner such that all elements except
H, He, Li, Be, B, N could be detected.
E-32
-------
FURNACE D
>-
_J
•<
o:
o
I
X
O
a:
o
LU
<
o-
o
o
•r—
CO
co
1 —
c
-
to
to
CO
1 U Q.
Q Q $- 0 LU S- E
CO CD -i- 13 Z -i- (O
o
PJ
ro
4->
O
(O CM O
01 I •—
S- O
CD r—
us
•3-
D Ifl
— l/l
I 0)
E-33
-------
FURNACE D
elements because they were not included in the quantitative analyses which will
be described in the next paragraph. Note also that oxygen was detected at
about the 50%, thereby suggesting that the particulate material was a mixture
of oxides.
2.3 Atomic Absorption Analyses
Atomic Absorption (A.A.) means that a cloud of atoms in the un-ionized and
unexcited state is capable of absorbing radiation at wavelengths that are speci-
fic in nature and characteristic of the element in consideration. The atomic
absorption spectrophotometer used in these analyses consists of a series of
lamps which emit the' spectra of the elements determined, a gas burner to pro-
duce an atomic vapor of the sample, a monochromator to isolate the wavelengths
of interest, a detector to monitor the change of absorption due to the speci-
men, and a readout meter to visualize this change in absorption.
As stated previously, the two sets of samples were composited two differ-
ent ways for the atomic absorption analyses. The detailed procedures for the
physically blended powders are as follows:
1. The particulate material from three specimens was either shaken
loose or scraped from the filter pads with a wood tongue de-
presser and blended in a polyethylene container.
2. Duplicate portions of the blended powder were digested in hot
HC1-HN03.* After cooling, the suspension was filtered.
3. The filtrate (soluble portion) was analyzed for the elements of
interest by atomic absorption. The precipitate (non-soluble por-
tion) was analyzed by "large beam" electron microprobe analysis and
flame photometry and found to be free of sodium or potassium. This
action was done because potassium acid sulfate (KHSO.) v/as used in
the next step.
4. The precipitate was blended with a known quantity of KHS04 and ignited
in a 850°C muffle furnace to form a fused mass which subsequently was
dissolved in HC1. Solution was not complete, and a filtration step was
needed to separate the solution from a precipitate.
5. The solution was analyzed for the elements of interest by atomic
absorption, and the results from this step were added to those from
Step 3 to yield the total percentage of each element in the parti-
culate sample.
*
The hot solution used was 8 ml concentrated HC1, 32 ml concentrated HNCL
and 40 ml distilled water.
E-34
-------
FURNACE D
6. The precipitate from Step 4 was checked for SiO? by a gas
evolution technique.* This technique selectively decomposes
and volatilizes Si02 through reaction with hot H2S04, HH03
and HF in a platinum crucible. The portion of the sample
that still remained after all these steps was labeled an
insoluble residue in Table E-13.
A different procedure was needed for those samples in which the quantity
of condensable particulate was insufficient for a physical separation. In
this case the following procedure was used:
1. Three entire collector pads, with material in and on them,
were digested in a common hot ^$04 solution. An unused
collector pad was submerged in a second identical solution.
2. The steps described previously were followed for both the
unknown and the unused sample. The results for the latter
were corrected to account for the fact that three used pads
were used with the unknown samples but only one unused pad
was employed as a blank.
3. The concentrations of elements in the condensable particulate
material was obtained by subtracting the results for the
"blank" from the total.
The results of'the atomic absorption analyses are compiled in TableE-13.
The following are observations.
1. Both samples are predominantly silicon dioxide, Si02. This
conclusion is directly seen in the results for the Inlet Duct
Sample where 76.4% of the material is SiOp. The concentrations
of the remaining elements are all low in comparison, and magnesium
is the highest at an average 5.44« level. The sum of all the
percentage values is 100%, and this indicates excellent closure
(mass balance). The 100% value is achieved when all the metal
percent values are converted to their equivalent oxide percent
values.**
**
N. H. purman, Editor, Standard Methods of Chemical Analysis, 6th Edition,
Volume 1, D. Van Nostrand Company, Princeton, N. J., p. 950.
Equivalent, oxide percentages are obtained by multiplying the weight percent
metal in Table 2 by the ratio Mo/Mm where Mo is the molecular weight of the
metal oxide and Mm is that of the metal.
E-;
-------
FURNACE D
LU O
UJ h-
ro
""l"
LlJ
Ol
-Q
re
c
fvj
O)
U.
O
0
*"
O O
O CO
r- 0
0 O
|| 1 | «*• o
c
•ss
ro
Z
QJ
a.
ro
CO
i— 0 O
o o o
a\
o
CM
f~
a>
• 0 CL
LiJ ro E
CO 4-> rO
CC 00 00
if> in
CM CNJ
0 0
CM CM
O O
Ol
0.
__ , ^
t OJ *->
C3rZ
rO
=> S-
to ro
Ol Q.
Ol
0) r—
> JQ
•r- ro
•»-> to
ro c
•4J 01
•r- TJ
-(-> C
C O
ro U
OJ O)
o .c
•r- 4J
CO
I
o
U
ro
i-
I
X
Q)
CO
CO
OJ
QJ
-Q
C
O
CJ
ru
OJ
Q)
C
o
T3
QJ
N
ra
C
ro
CO
ro
2
JZ
+J
•r-
2
Ol
to
to
CJ
U
3
t_
CJ
F
3
c:
'r-
4->
ra
a.
QJ
01
o
01
xi
ai
01
-O
ra
-Q
O
i-
D.
ro
ia
O.
QJ
CO
I
CO
.
•*J o
o. in
E i
O) —-
ro
Oi
•a
01
+j
u
3
-a
c
o
to
Q.
a>
o
ro
>.
«£•
01
-a
o
1 *
"0
0)
E
QJ
QJ
to
+->
ro
.c:
4^
(LI
3
X?
•r-
to
O)
S~
Ol
-C
1—
*l~
^_
ro
E
•i —
i.
a.
Ol
J3
O
4->
XI
C
rj
0
f-
x>
c
ro
S-
01
N
>>
ro
C
ro
XJ
C.
iTJ
•s
E
Z3
XI
o
CO
•
E
r>
c
i
3
ro
M-.
O
CO
•M
C
3
O
E
ro
C
o
to
QJ
0>
QJ
-t->
QJ
Q.
O .
CJ Ol
C E
•r- 01
M- -5
O to
E-36
-------
FURNACE D
2. The Stack Sample, in comparison with the Inlet Duct Sample, con-
tains relatively more of every metal cation except magnesium.
The absolute amount of 'the Stack Sample was far less and this
had an impact on the sensitivity values. Thus the lower limits
for barium and titanium are 4% and 8% in the Stack Sample (rather
than 0.4 and 0.8%) because the total sample mass was limited to
"11 milligrams.
3. It must be emphasized that the values have been corrected to
account for the contributions from the filter pads. In other
words, the 12.7ft Ma value is for the particulate matter collected
on a filter and riot for the filter pad.
Metals analysis was also made on two chrome ore samples using the optical
emission spectrography method. These are shown in Table E-14.
E-37
-------
co
CM
I
OJ
O O
o o
ui in
FURNACE D
•v
CO
g
CJ
UJ
Q.
CO
*
•o
(U
4->
(0
o>
•r—
(O
OJ
-o
O)
O)
OO
I/)
S §
O)
Q-
O
I
UJ
Ol
§ I
v v
Q.
(LI
O
X
Ol
I-
"O
O)
CD
O
O)
s_
I
in
ns
o o
o o
in in
§ §
Q.
O.
s s
o o
to
(U
Q:
V^? •?.?
CO 03
E-38
-------
FURNACE D
ENVIRONMENTAL PROTECTION AGENCY
Office of A1r Programs
Research Triangle Park, North Carolina 27711
Reply to
Attn of: AID
Date, fay 5, 1972
subject: sojj Analysis of Partlculate Samples from Furnace D
To: Wlnton Kelly, Chemical Engineer, Petroleum 4 Chemical Section
Two sample fractions (organic extract and Impinger water residue)
were selected for sulfate analysis from each of the two sample tests,
ANE-1 and ABD-2. These samples were first analyzed by the Barium Perch-
lorate titratlort method, which proved to be too In-sensitive. Re-analysis
by the more sensitive Sulfaver-Turb1dimetric procedure gave the following
results: (MO M» en"
Original ng iU4 Weight . Sample
Test No. Beaker No. Sample Weight Sample Percent SO, Fraction
ANE-1
ANE-1
ABD-2
ABD-2
31
43
35
47
8.1
18.8
35.8
42.4
<0.25
6.75
<0.25
5.55
< 3
33
< 1
13
•*
Organic
Extract
Impinger
Water
Residue
Organic
Extract
Impinger
Water
Residue
BAGHOUSE
OUTLET
BAGHOUSE
INLET
An acid-base tltratlon of beakers 35 and 47, using 0.1009 N NaOH,
used less than one drop per sample, Indicating both samples were very
near neutrality. Beakers 31 and 43 were not subjected to an acid-base
tltratlon due to a lack of sample volume after the SO? analysis.
Remnants of test ABD-2 (beakers 35 and 47) have been sealed to pre-
vent contamination and are located at the IRL Building sample storage
area. Any questions regarding these samples or the data can be directed
to me (X277) or found 1n the laboratory notebook located in Room 26, at
the IRL Building.
ank M11shire
Chemist
Petroleum & Chemical Section
Emission Testing Branch, ATD
cc: Mr. W. Grimley
Mr. H. Crist
E-39
-------
FURNACE D
ENVIRONMENTAL PROTECTION AGENCY
Research Triangle Park, North Carolina 27711
Rtply to
Attn of:
Dau: 5/3/72
Subject: Terro-Alloy Samples
To: Vfinton Kelly, ETB
Petroleum & Chemical Section
The results of analyses on the above samples collected
from Furnace D are attached.
Your samples are identified as follows:
Battelle No, EPA Test No,
Sample Fraction
B-272
ABD-2
B-273
B-274
B-275
B-276
B-277
B-278
B-279
/•BD-2
ACE- 1
ABD- 1
ACE-1
ABD-1
ANE-2
ANE-2
Solvent extraction of im-
pinger water >
Impinger water residue J
Solvent extraction of im- /'-
pinger water J
Impinger water residue ^
Impinger water residue
Solvent extraction of im-
pinger water
Impinger water residue
Solvent extraction of im-
pinger water
BAGHOUSE
INLET
BAGHOUSE
OUTLET
BAGHOUSE
INLET
BAGHOUSE
OUTLET
BAGHOUSE
INLET
BAGHOUSE
OUTLET
Howard L. Crist
Chief, Source Sample Analysis Section
SSFAB, DAS
Attachment
cc: W. Grimley
E-40
-------
Table E-15. METAL AND OTHER ANALYSES OF COLLECTED PARTICULATES
FROM FURNACE D
(ug/sample)
Determinations requested:
Type analysis: VAS, OES
anions, water insolubles, organic material, trace
metals, pH
Be
Cd
As
V
Mn
Ni
Sb
Cr
Zn
Cu
Pb
B
Li
Ag
Sn
Fe
Sr
Na
K
Ca
Si
Sample designation
B272
<0.1
<5
10
3
100
30
<5
1000
150
10
200
3
<5
3
30
700
<5
1000
2000
150
100,000
B273
<0.1
<5
<10
<1
200
30
<5
1000
300
10
300
3
<5
1
10
300
<5
2000
2000
700
50,000
B274
<0.1
<5
<10
<1
<1
<1
<5
<1
<10
<1
<5
1
<5
<0.1
<5
5
<5
<10
<10
20
200
B275
<0.1
<5
<10
<1
200
30
<5
600
50
7
300
3
5
0.1
10
300
<5
800
1000
400
40 ,000
B276
<0.1
<5
<10
<1
10
2
<5
3
10
5
5
1
<5
<0.1
5
20
<5
100
30
100
200
B277
<0.1
<5
<10
<1
80
5
<5
600
30
5
30
5
<5
1
20
300
<5
300
800
200
120,000
B278
<0.1
<5
<10
<1
10
2
<5
10
20
7
5
1
<5
0.1
5
50
<5
100
30
200
200
B279
<0.1
<5
<10
<1
<1
<1
<5
<1
<10
<1
<5
<1
<5
<0.1
<5
1
<5
<10
<10
<1
5
E-41
-------
Table E-14 (continued),
METAL AND OTHER ANALYSES OF COLLECTED PARTICIPATES
FROM FURNACE D
(ug/sample)
Determinations requested:
Type analyses: VAS, OES
anions, water insolubles, organic material, trace
metals, pH
Mg
Bi
Co
Ge
Mo
Ti
Zr
Ba
Al
S04
Cl
NH4
N03
Water
insol-
uble
Organic
PH
Sample designation
B272
5000
5
1
<3
3
20
3
50
2000
<500
30
<30
<100
223,000
16,500
6.2
B273
15,000
2
1
<3
<1
3
<1
20
1000
24,000
1000
120
200
191,700
2000
6.1
B274
10
<1
<1
<3
<1
<1
<1
<1
10
< 500
<20
70
100
3500
1400
5.4
B275
12,000
<1
3
<3
<1
1
<1
10
1200
22,000
60
700
100
41,300
2000
4.5
B276
5
<1
<1
<3
<1
3
<1
2
40
6800
<20
1000
<100
2200
1000
3.2
B277
4000
<1
5
<3
<1
5
<1
50
1200
<500
<20
<30
<100
121,500
4000
6.3
B278
30
<1
10
<3
<1
5
<1
2
60
10,000
<20
750
100
None
1000
2.7
B279
<1
<1
<1
<3
<1
1
<1
<1
<3
<500
20
<30
<100
None
400
5.3
E-42
-------
FURNACE E
V. CHEMICAL ANALYSES OF EMISSIONS
FROM
AN HC FERROCHROME FURNACE
Particulate fumes and gaseous emissions are generated during processn.b
of an important class of ferroalloy materials called reactive metals. The
participate portion of the emissions is collected on glass fiber filters stra-
t^nically placed in the air stream of a ventilation system. Two such filters of a
sample collected at the outlet of a furnace producing HC ferrochrome were analyzed
by the combined techniques of a electron beam X-ray analysis and atomic absorption
analysis, and the results are detailed in the following paragraphs.
2. TEST RESULTS
2.1 Optical Examination and Compositing
The two samples were gray and were labeled WCD and ECD. Small portions
were cut for electron microprobe analyses. The remainder was shaken and a
copious amount of loose powder was gathered, blended, and designated preci-
pitator Inlet Duct CD-M.
2.2 Electron Beam X-Ray Microanalysis and Atomic Absorption
The electron microprobe is an advanced piece of equipment which uses a
small beam of electrons to produce characteristic X-ray emissions from a sam-
ple volume with a radius of ~1 micron. Curved crystal X-ray spectrometers
are used to analyze the resultant characteristic X-ray spectra. In these
analyses, the electron beam was defocused to a diameter of 150 microns
(0.006 inch) to cover a larger segment of the sample. The electron beam
Impinged in vacuum upon the untouched surfaces of small pie-shaped pieces of
sample-covered filter pads. An examination was made of the complex spectrum
of X-rays given off by the specimen under electron beam excitation, and it
was found that the entire spectrum could be identified uniquely. All portions
of the X-ray spectrum in the wavelength range 1-100A covering all elements
except H, He, Li, and Be were taken into account.
E-43
-------
FURNACE E
Atomic Absorption (A.A.) means that a cloud of atoms in the un-ionized
and unexcited state is capable of absorbing radiation at wavelengths that are
specific in nature and characteristic of the element in consideration. The
atomic absorption spectrophotometer used in these analyses consists of a
series of lamps which emit the spectra of the elements determined, a gas bur-
ner to produce an atomic vapor of the sample, a monochromator to isolate the
wavelengths of interest, a detector to monitor the change of absorption due
to the specimen, and a readout meter to visualize this change in absorption.
The qualitative electron microprobe results are in Table E-16, along with
the quantitative atomic absorption results. The latter were generated on the
composited samples mechanically separated from the filter (collector) pads
by shaking and lightly scraping the filters. A negligible amount of collector
filter material was included in the blended sample, and therefore no unused
filter pad was needed.
The sum of the percent values, after conversion to equivalent oxide
values, is 84% and indicates adequate closure in the sense that all the major
constituents have been taken into account. The remaining 16% could well be
accounted for by the presence of chlorine, carbon, and titanium.
The major conclusion is that the sample is a mixture of oxides: SiCL,
Cr203, MgO, and A1203.
X-ray diffraction analyses were not executed, and hence it is not known
if the oxide mixture is amorphous (non-crystalline), crystalline (spinel
structure), or partially amorphous-parti ally crystalline.
A sample was obtained from the precipitator and analyzed for metals
by the Optical Emission Spectrography (OES) method. This sample does not
necessarily represent true emissions from the furnace since some particulates
are removed by the conditioning tower preceding the precipitator.
E-44
-------
FURNACE E
Table E=16. QUALITATIVE ELECTRON BEAM X-RAY MICROANALYSIS AND ATOMIC -ABSORPTION
RESULTS FROM SAMPLE COLLECTED AT PRECIPITATOR INLET DUCT (FURNACE E)
P
o
o
CJ
o
to
c
c
*
*
•z.
CO
n]
o
5
£
Cn
C
£
J-
U
Ol
"Q.Z:
ooo
H-
_l
0
0^
s: ^
o
^ !
x;
UD
ID
O
cn
3C
CO
_J "^
o
i
00 o
C. •!-
o •<->
i. o a.
*J 4- L.
o go
LU <:<
H « high >10% Quantitative atomic absorption
analyses were not performed.
M « medium 1-10%
i— t— O
s
g
-------
s =
O cc
CL .
UJ co
CC UJ
O"°
UJ "•
o '
o: i
Ss
u
i
S? So
Q ^
^ o
a.
I
FIIRMfl
UJ
o u7
£«£
o t~ z
o ss 3
s^i
g ^^
o Si
1 > o
to ° "
UJ P CX
1 1 J j — UJ
in
Z X ^
a" ^
0
Ij
UJ
CL
S
OO
sis
C/3
UJ
U_ — '
°1
55g
£SS
g"jg
Ka^^
0 S
t^
0
fe
o
c_
1
1
\l
(
L
1
\J
^t
I
^
\
e^
^N
-A —
H
UJ
S
0
CE E,
CO
0
£3 ^ ^
cv> 2
UJ UJ
cc a
cc
UJ
or
u.
o
UJ
H-
a
Q
P
O UJ
U_ LU
0 CC
1— CO
LU *^
I 1
Or
UJ
0
u7
_J
CO
«c
o
_J
O-
a_
-------
FURNACE E
^
^
%
H
^
bsj
^
_v
JS
•v
1
^
^
^
1
v^
>-
I'
V
-a
Y
V
•<
V
t
(^
H
i an
nan
ann
nnnn
an
3
ZD
O
E
3
-------
FURNACE F
VI. CHEMICAL ANALYSIS OF EMISSIONS
FROM
A SILICON METAL FURNACE
I. INTRODUCTION
Particulate samples were collected using an EPA sampling train at the
inlet and outlet of a baghouse serving a silicon metal furnace. Chemical
analyses were made of the particulates collected in each section of the
sampling train. For the use of Table E-18, sample numbers are identified
as follows:
Description of
sample or sample
fraction
Sample No.
Baghouse inlet
Baghouse outlet
Probe wash residue
Glass fiber filter
Impinger aceton wash residue
Impinger water residue
Chloroform-ether extraction residue
165,188
(166,167,168,169),189
170,190
171,191
172,192
137,152
138,153
139,154
140,155
141
Sample numbers 137 through 141 were for a test run at the baghouse outlet.
Samples 152 through 155 were also for a test run at the baghouse outlet.
Samples 165 through 172 were for a test run at the baghouse inlet. Samples
166, 167, 168 and 169 were combined into one sample. Sample numbers 188
through 192 were analyzed representing a test at the baghouse inlet.
The samples were water leached for the anion and NH4 analysis, and they
were acetone leached for the total organic analysis; portions of these
leaches were combined with a portion from an additional acid leach-scrubbing
step to obtain a sample for optical emission spectrograph (OES). A portion
of each filter was water and acetone leached for the organics, anions, and
NH.+; a separate portion of each filter was acid extracted for the OES sample.
The analysis of the filter and residue samples are shown in Table E-18.
Results are reported in micrograms per entire filter or residue. The weights
E-48
-------
FURNACE F
of the samples analyzed are:
Wt. in Mg
137
138
139
140
141
152
153
154
155
165
166
167
168
169
170
171
172
188
189
190
191
192
153.6
10.4
23.1
21.2
3.4
13.1
16.8
27.5
25.0
486.3
1017.4
701.8
923.0
1404.4
30.7
59.6
68.6
9.5
18.8
20.6
23.2
1.0
E-49
-------
FURNACE F
w
I
H
&
a
•s
EH
" •£«
H s-g.
i? ifS
^4 U -U
e §
U d
VI 3
a I
rtrjv f i o j; o in
V V OOO OVV Vy C*fJOOf>OO
V v H V i *rt i i i 4if> i t^iiitiot/'jf-t^vcjc*!!
i-* r-f r4 IA n
o\
r4
VJ
•jOHOOtncjOnm t-i OQ'^onc><3tr|Oc>rtHnrt'^ OOH^AOCJOOOOO
/-jV'-i'-tVr-nA rHi »nV r«.Vf-«.-
-------
C*J
o^
, ^
r-t
OS
r-l
I
O
a\
r4
1
OS
oo
rH
1
CO
CO
I-I
1
CM
i-t
1
-1
3
-1
t
OS
vO
r-l
CO
r-l
1
A
V0
1
*
VO
1-4
VO
(H
Element
r-l in
? v1 $ V 'v V V V1 ^ ° v1 .
1-1
OinOrlO.-linrHOr-lin
VVrMVr-I^VV<'> 1
iH
OinOr-l3v'1'"'tsv'r"d t^ m
V v"v ^.j VV o>V>-i'n
1^ r-l
I-I
o i o o m o to ° H "> ' ' ""> r •' r* n r1 '~< ' o o «-i o in o o o o t > o
IN enVVcnV>*Vi-l'1i-OI
V f "• V V in r-l C^ '-O r-l
«^ V '• •*
r i o^ O
r-l iH
Ol 1 IvpOOOOOIOvovqiNMI 1 1 ' IOOOOOOOI
O VOtnOOO OVV VV OOOOcnt'O
in i-icor^voco oinco^V-^cM
»v .. ,
r-l r-< ^.' t^
o\
1-4
• • en
1 cnOnC'O00""1'-10^^^^'-1 ' OOr-IOOOOOOOO 1
y •• »»
r4 i-l
V v 1-1 o VV Vlr-tv
•> V v
3
l-l . CM
Vln r-l^l-VCM-J-OOst-VcMVvVlt-ir-lV
-------
Furnace G
CHEMICAL ANALYSIS OF EMISSIONS
FROM
SILICOMANGANESE FURNACE
I. INTRODUCTION
Inlet and outlet scrubber samples collected by the use of the EPA
particulate sampler were chemically analyzed using optical emission (OES)
and visual absorption spectrophotometry (VAS). For the purpose of
identifying the results shown in Table E-19, the following information is
given:
Description of samples or sample
Sample No. fraction analyzed
B-338
B-339
B-340
B-341
B-342
B-343
B-344
B-345
B-346
B-347
B-348
B-349
B-350
B-351
Probe wash residue
Glass fiber filter
Glass fiber filter
Glass fiber filter
Impinger water residue
Chloroform-ether extract
Impinger acetone wash residue
Probe wash residue
Glass fiber filter
Impinger water residue
Chloroform-ether extract
Impinger acetone wash residue
Blank glass fiber filter
Blank glass fiber filter
Sample numbers from B-338 through B-344 represented portions of one test
collected at the scrubber inlet. Samples from B-345 through B-351 represents
portions of one test collected at the scrubber inlet.
Filter samples B-339, B-340, B-341 and B-346 had in the packages mailed
to the contractor doing the analysis some loose material that had been
E-52
-------
shaken from the filters during shipping and handling. The loose material
and the filters themselves were analyzed separately for convenience, but
the data were combined into a single reported valie for each determination.
The appropriate blank corrections were made in the data for the filter
samples, as indicated in the footnotes to Table E-18. The very large
blank values for certain elements led to high "less than" values for the
filters with sample even though these elements are almost certainly
present in the sample, as indicated by their presence in the loose material.
However, some particles of the glass fiber filters may also be in the
loose material, and so it is very difficult to determine the true net
amount of these elements in the actual sample.
Material balance for these samples is not good, but this may be due to
the silicon content which cannot be determined for reasons given above.
E-53
-------
CO
8
c
t.
3
u.
I
«4
«
u
&
M
. *-* M
£ -o
8- S
S " H
|S &
g£ «
S W A*
M O
{3 * '
«e « *o
- g r* *
0 « •* u
§ M g S
e<3 (4 g
H •-•; o o
sx s :
:s o i
°T3 * S
M « U C
M O O E
sg. a 1
sJ g s
.< ^ M C
til
«
J3
«
I
M
1
7
B
U
S.
%
Absorptl
•3
4B
^v
8
I
cctrography
w
g
«
2
r-t
«
V
vl
u
,§•
1
Analysis
i
~* *o M o (^ I
V v V V V \f j V V V V V vD cni? o w jj ^ y v V
OOr*OOOOO%OOO O O O N W OOC4 m oeoo-*o^o
rH
Is* f-
o -* *&• tr* t*. o r*
ocM^Of-ir*i OOVOOCMCM w«-toi-»m cn
vvvveo v V v "O^V/^T/ » vr^^m^oo*Di
§m«~«eMr» CMO eo »oo ^>~u> *n
OtHO OOeNOOtslO OOt-l O U OOOM OincMOr^l
CO (>.
^ eoc3 ***3gA ** *** AP*° -* ** S •-* ^ '•"•* C>" P*?1 * *
1 > "* w 6
* u
i
•o
s
IM
•^
«
4 M
n) p O O O
u ^ *« IH o.
ffli: E-54
U !
^*a.'ufl'
-------
Furnace H
VII. CHEMICAL ANALYSIS OF EMISSIONS
FROM
A 50% FERROSILICON FURNACE
Chemical analyses were made on particulate samples collected from the
outlet gas stream of a venturi scrubber serving a semi-covered furnace producing
50% ferrosilicon. The first tests were made on this scrubber on February 2,
1972, while kerosene was injected into the exhaust system blower, some kerosene
was collected in the sample. The samples for these tests are identified as
253 through 257.
The second group of tests were made on July 19, 1972, without kerosene
injection into the blower. Combustible material in the sample collected with
kerosene injection was about 67 percent and without kerosene injection, about
50 percent.
Sodium, potassium, and calcium analysis appeared to be too high, so a
check analysis was made on filter samples 134 and 140 by the atomic
absorption method. The results in milligrams were:
Sample
134
140
Sample
weight,mg
98.3
103.1
Na
83
81
K
11
11
Ca
33
35
This represents the amount of these elements extracted from the collected
particulate and the filter material.
E-55
-------
ce§
S"
CO
o: i
3s
j [2: cj
O I— z
ozg
CO 00
eso
§*
E-5
"50
Sis
,!\o
18
xa
IN
00
o
cr>
UJ UJ
a oc
o
h-
O
to
-------
FURNACE H
1 '
-i-H-r -
r
i i
•7-*
>\
I
-------
CC :
O i
&!
Vo
_1 o I— Q
»_> o uj 5 ^-
z
t?
^
't
^
k<
o
z
1 I *
i § I
o « 3
^RNAQE H
^
CO »
Si
O*>
UJ UJ
cc a:
o
s
^
g
*3
—I CC
E-58
-------
•e S
C/)
p
CD
O)
O
X
C/D
U.
3
U_
co
0>
JO
Q-
O
Nl
O
JO
c:
>
5
T3
O
CQ
cn
O
J
1
J
n
J
"i
J
n
J
n
J
n
J
n
J
n
U
h
J
n
J
i
J
n
J
~i
J
n
J
~i
J
1
J
~l
J
~1
J
~1
J
-\
J
~\
J
~i
J
n
J
n
J
n
J
~i
J
i
pi
n
J
1
J
n
J
1
J
n
J
~i
J
j
-]
J
n
J
~i
J
n
J
~i
J
~1
J
n
J
n
J
n
J
~1
J
~]
J
n
J
~l
J
~i
J
~i
J
~i
J
n
J
n
J
n
J
n
J
~1
J
~i
J
n
J
n
j
n
J
~i
J
n
J
~1
J
n
J
n
J
~l
J
n
J
i
j
n
J
n
J
n
J
n
J
-]
J
n
J
n
J
1
J
"i
J
n
J
n
J
h
pi
n
pi
1
J
n
J
~i
J
n
J
~1
J
~\
J
n
u
h
J
~i
u
n
J
n
J
n
J
n
j
n
J
~i
J
n
J
~i
J
n
J
n
J
-j
J
n
J
n
J
n
J
~i
j
n
j
n
J
~1
J
"I
u
h
J
h
J
h
J
~i
J
M
j
h
U
h
j
~i
j
n
j
n
J
n
J
I
J
~i
J
n
J
n
J
i
J
i
J
1
J
i
J
n
J
1
J
1
J
n
J
n
J
i
J
i
J
n
J
i
J
i
J
n
J
i
J
~l
J
1
J
i
J
1
J
n
j
n
J
1
J
n
J
n
j
i
j
n
J
i
J
h
LJ
n
j
i
j
n
j
^
n
j
~i
j
n
j
^
~i
j
~i
j
~i
j
-\
j
~i
j
n
j
~1
J
~i
J
n
J
n
J
n
J
~i
J
n
j
n
J
n
J
n
J
~1
J
~i
J
1
J
n
J
n
J
h
U
3-
J
n
J
n
J
~l
J
n
J
n
J
~i
J
j
1
J
~i
J
~l
J
n
J
1
J
n
j
3-
n
J
"i
J
n
J
~i
J
n
J
~i
J
n
j
n
J
H
LJ
9-
h
J
n
j
h
pi
n
J
~i
J
n
J
n
J
n
J
h
J
~i
J
h
her Comments:
o
X3
U^
CC
X
L£>
CO
iA
-
-
^
1
1
n
n
n
n
h
n
xs
IV,
n
x|
sTI
•V)
%
n
n
n
n
n
^
n
-------
5
K
o £
LU fe
/V UJ
KS
o s
c/5 S
X
£
Ei
I
a
^t i
O f— z
o zg
;j?s
cyce
£££
«S3
N
o
fe
LU
O
.a
w
6
-, EURNACE H
CD
Z
O
g
cc
UJ
o
u.
u.
o
I—
u
UJ
I
-------
FURNACE H
-------
FURNACE H
E-62
-------
FURNACE H
E.63
-------
VIII. CHEMICAL ANALYSIS OF EMISSIONS
FROM
A FERROMANGANESE FURNACE
Chemical analyses were made on participate samples collected from
the outlet gas stream of a venturi scrubber serving a sealed furnace
producing ferromanganese.
Samples of three of the six test runs were analyzed for trace
metals, SO.", Cl", NO,", NH. , pH, organics and water insolubles. Chemi-
cal analyses were made of the particulates collected in each section of
the sampling train. A total of five separate analyses were made to
determine the metals in each portion of the sampling train for one
test run.
The samples were water leached for the anion and NH. analysis, and
they were acetone leached for the total organic analysis; portions of
these leaches were combined with a portion from an additional acid
leach-scrubbing step to obtain a sample for optical emission spec-
trography (OES). A portion of each filter was water and acetone leached
for the organics, anions, and NH.+; a separate portion of each filter
was acid extracted for the OES sample.
The analysis of the filter and residue samples are shown in Tables
E-24 and E-25.
E-64
-------
os 5
o cc
UJ to
ry uj
=0 ^
3
1
I
K
^
is
UJ O
-Jo
Q- uJ
o •"
iS
K
o
s
Hj
u
"*;
^
;
I
t 5
.' Z
^
UJ _
o uj i/
~5e i
h- 5: oo ;
i"H 4
Iss '
O I— z ^
0 zg ^
1 i , I CO
Sd^ I
£s°5 §
z 8« ^
W 0^
"S«t
' • o S
Sgt ,"
te°a ^
it&
S^UJ ',
* s< *•
S200
"§g
Sx"
*gg
£& »
s x
3
^*"
^
1 V
I
**
1
^
s\^
^
\ <
^
1
^ vs
ui 5
x ^
^ ^
, $
^ R
V ,1
Pi >
^ ^
Vj
^t
y
^ -
I*
W
N ti,
^^
N,
,
Q
Oc\
1
i-
g
} >
53
''J
^
g
N
"<
1
)
CS
' <;
ft
- C
*- V.
; ^
1 v-
^ *
- ^
^
N
fV, S
^ v
\
q
^
s
^
V
S
*
- ^
^
^
: v
~ ^
i ^
5 ^
;|
M^
jH
^ ^
;^
\
v
^
r\
S
r>j
^l
CN
^
^
X
"vj
^
5
^
V
^
\
1
1
I
1
"|CE K,
m
z
o
CO
o
£ te
O "J
cc
62
-------
FURNACE K
CO
to
>•
to -
? LJJ S °
S 5 -2 1
u. -i B sr
C3 LU trt —
1- "J o S
o: CJ —
uj ^ E S
^ S ^5i
NT
' ; 5J
U f-
Y. V
V
I
Q
V
IT"
-(i
V
o
00
E-66
-------
O o:
Q. ._
UJ co
CC £
°» -c
Si
\
°. 00 iZ
UJ
CC
UJ O
-"a
LJ_ UJ
g
to
£ o
£ £
t cc
5 <
K
\0
i 0 o
t? So
UJ _JO
(X Q-3:
E-i
rvj
Q
UJ
£ £
S ^
u. o.
U- UJ
o tr
I— so
" 35
UJ >.
-------
FURNACE K
CO t/> ^>
pp^t
=!£ll
E E
aii
_| _
8
E-68
-------
IX. CHEMICAL ANALYSIS OF EMISSIONS
FROM
A SiMn FURNACE
Particulate samples were taken at the outlet of a venturi scrubber
serving a silicomanganese furnace (sealed). Samples were collected with
an EPA sampling train by EPA method 5 (described in Appendix B of this
report). Chemical analyses were made of the collected particulates in each
fraction of the sampling train. Metal analyses were made of the two test
runs. Analyses of organics, anions, NH. , and water insolubles were made
of three test runs.
The samples were water leached for the anion and NH. analysis, and
they were acetone leached for the total organic analysis; portions of
these leaches were combined with a portion from an additional acid leach-
scrubbing step to obtain a sample for optical emission spectrography (OES).
A portion of each filter was water and acetone leached for the organics,
anions and NH.+; a separate portion of each filter was acid extracted for
the OES sample.
The analyses of the filter and residue samples are shown in Tables
E-26, E-27, and E-28.
E-69
-------
-------
FURNACE L
-------
-------
FURNACE L
<£ S 3!
j£ | I
< 5-° 1
feuJIf
&?e 1
K c-L~a*
~&
s
L_J_
V
II
§••3
&i>
^ ^x.
v! v
o: sj
Vv I ^K
S!
^ V
—-—(—
"i V
ii t- -
*^,i o
-'VT-J^-^
V
-a
i i
•
-------
-------
FURNACE L
S2
•I
saf
™
3
s 5
It
CL O,
-------
APPENDIX F
Table F-l. UNITED STATES IMPORTS OF REACTIVE METALS AND ALLOYS*
December 1971
Issued;
TSUSA
KO .
607.
3500
607.
3600
|
607.
3700
607.
''.700
!
CO.'^C'DITY
FF.RR OKA is G ANESE ,
under } 'is Carbon,
Commercial
FERROMANGANESE,
1-4% Carbon,
Commercial
F.ERR OMANG ANESE
over 4% Carbon,
Commercial
FKRROS1LJCON
MANGANESE,
Cornrnoi" ciul
COUKTKY
or
ORIGIN
IMPORTS FOR CONSUMPTION
CURREXT MOKTH
Short
T or>r> '••*
Franr« I 648
Japan
Rep. S, Africa
Sweden
West Germany
TOTAL
Canada
France
Italy
Japan
Norway
Rep. S. Africa
Sweden
West Germany
Yugoslavia
TOTAL
Belgium
Canada
France
India
Japan
Norway
Rep. S. Africa
TOTAL
Canada
France
Jr.pan
Mexico
Norv.-ay
Sweden
Yugoslavia
TOTAL
„
1)2
-
760
-
235
-
-
.-
-
851
-
-
1086
-
~
-
-
-
1345
100
1445
-
-
-
--
';8 ?-9'iV
659
2346
299?.T>
} 1C
:•; 1 4
39:>1
i
M
-------
Table F-l. Continued
December 1971
r - -
TSUSA
K'O .
632.
3200
1
607.
3300
607.
3000
632.
J800
i
•
COMMODITY
MANGANESE
METAL
Commercial
FERROCHROME,
over 3% Carbon,
Commercial
FERROCHROME,
under 3% Carbon,
Commercial
CHROME METAL,
Commercial
COUNTRY
0?
ORIGIN
Australia
Canada
Cyprus
Japan
Netherlands
Rep. S. Africa
TOTAL
Belgium
Brax.il
Canada
Finland
France
Japan
Norway
Rep. S. Africa
Sweden
West Germany
TOTAL
Canada
France
India
Japan
Norway
Rep, S. Africa
Sweden
Turkey
Wcsl Germany
TOTAL
Canada
France
Japan
Netherlands
Un, Kingdom
V/ c fi t G e r r/*a ny
TOTAL
IJ-? PORTS FOR COWSUMPTIO:'
CURRENT HONTJI
S7hort j value
Tons''-'* £1000
10
-
-
114
-
82
206
_
-
-
-
-
-
-
-
220
55
275
_
-
~
-
-
-
-
-
142
142
-
-
41
-
35
-
76
27
-
-
45
-
34
106
_
-
-
-
-
-
-
-
51
33
64
_
-
~
-
-
-
-
-
215
215
-
-
87
-
63
-
150
JAN.__;
Short
10
24
15
800
50
1979
2878
310
1382
535
10903
4256
12992
301
7174
220
3 3697
49550
184
3086
2260
7389
3457
14632
5433
1320
5033
40594
17
90
605
2
818
102
1 6 3 4
:o D".v:
Vr.i'AK
27
2
6
343
23
828
1229
25
291
359
1138
1125
2923
85
956
51
1620
8373
43
-','/"
!',
b
393
] 1 4 7
6
1420
J ',. 2
2963
L _
F-2
-------
Table F-l. Continued
December 1971
'SUSA
NO.
07.
000
,07.
ilOO
607.
5200
607.
B.'iOO
632.
J200
,07.
5500
COMMODITY
FERROSILICON,
8 --60% Silicon,
Commercial
FERROSILICON,
60-80% Silicon
Commercial
FERROSILICON
80-90% Silicon,
Commercial
FERROSILICON
90% and over,
Silicon
SILICON METAL,,
Commercial
C M R O M I LI M S J I , I CO N
COUNTRY
OP
ORIGIN
Canada
i;'r.c> jn,u
Japa.n
Norway
W. Germany
TOTAL
Canada
China
Denmark
France
Japan
Norway
Rep. S. Africa
Sweden
W. Germany
Yugoslavia
TOTAL
Canada
Rep. S. Africa
TOTAL
TOTAL
Canada
Norway
TJn. K-'ngdom
TOTAL
Sxv u u e n
TOT AJ,
IMPORTS FOR CONSUMPTION
CURRKK
Short
Tons >'•••'<
74
16
50
55
195
44
69
165
-
278
„
-
-
-
-
-
-
| CUMULATIVE,
P MONTH j JAN. TO D/vTE
Value
S l.OOO
6
7
18
17
48
17
28
32
«t
-
77
_
-
-
-
-
-
-
-
-
Short
T.ons**
6038
1387
3587
685
277
11974
791
28
44
2837
50
2568
319
3115
444
2224
12420
60
14
74
-
-
174
22
2
198
772
772
Value
SJ.OOO
422
492
1111
213
75
2313
216
7
17
1 1 29
10
737
62
541
161
776
3656
18
3
21
|
i
!
1
I
-___]
74 i
8 !
^ !
i
84
207
207
F-3
-------
Table F-l. Continued
December 1971
I
TSUSA
NO.
607.
7000
I
1 . ,
607.
6000
607.
6500
629.
2800
628.
1500
6.79.
0500
,S20.
6000
COMMODITY
FERROVANADIUM
FERROTITANIUM
Commercial
FERRO TUNGSTEN
TUNGSTEN,
Unwr ought
COLUM3IUM METAI
TANTALUM METAL
ZIRCONIUM METAL
COUNTRY
OF
ORIGIN
W. Germany
TOTAL
France
Italy
Un. Kingdom
TOTAL
Bxaxil
Sweden
TOTAL
Sweden
W. Germany
TOTAL
i
W. Germany
TOTAL
Canada
Japan
Mexico
Un. Kingdom
W. Germany
TOTAL
Canada
Japan
Netherlands
Sweden
Un. Kingdom
W. Germany
TOTAL
IMPORTS FOR CONSUMPTION
CURRENT MONTH
Short
Tons**
-
_
^
-
-
—
5
-
5
1
-
1.
(P
-
M
(P
_
-
_
-
-
_
(F
2047
44532
-
-
-
-
46579
Value
S1000
-
_
_
-
-
..
24
-
24
7
-
7
OUNDS &
-
..
OUNDS V.
-
~
-
-
~
„
OUNDS &
1126
144284
-
-
-
-
145410
CUMULATIVE,
JAN. TO DATK
Short
Tons**
69
69
50
20
17
87
6
12
18
7
44
:>!
DOLLAR:
450
450
:DOLLAP
763
792
22341
16020
175
40091
Value
$1000
360
360
23
15
116
154
29
69
98
66
539
605
5)
7227
7727
S)
ZZfjH
10456
47739
212355
4981
278799
DOLLARS)
4798
164124
12867
7209
240
36692
225930
-
4546
549800
11659
10805
1457
54572
632839
.
F-4
-------
Table F-l. Continued
DLH:.::OK>cr J971
TSUSA
| NO.
629.
iiSOO
I
!
~
COMMODITY
TITANIUM. METAL,
Commercial
Government
Purcho.ee
COUNTRY
OF
ORIGIN
Austria
C/ct'UdUa
Japan
Netherlands
Un. Kingdom
USSR
W. Germany
TOTAL
TOTAL
i
"
IMPORTS FOR CONSUMPTION
CURRENT MONTH
Short
Tons**
_
•11
37
-
-
437
-
485
-
485
Value
.,..$1.0.00.
5
69
-
-
67
-
141
-
141
CUMUIv'.TIVK,
JAN. TO u.Vra
Short
Ton si*
5
117
2498
4
1ZO
1335
66
4145
4145
Value
__£1£LQ.O.._.
3
129
433]
3
130
316
68
4980
-
4980 i
i
i
1
i
i
r-, C > '.-,
F-5
• >!• f, -,- r. >:'
-------
Blocking Chrome
Charging
Charge chrome
Chrome ore/Time melt
Condensed fumes
Covered furnace
APPENDIX G
Glossary
A high-silicon (10 to 12 percent) grade
of high-carbon ferrochromlum used as an additive
1n making chromium steel to block (I.e., stop)
the reaction 1n the ladle.
The process by which raw materials (charge) such
as ores, slag, scrap metal reducing agents, and
limestone are added to the furnace.
A grade of high-carbon ferrochrome, so called
because 1t forms part of the charge 1n the
making of stainless steel.
A melt of chromium ore and Hme produced 1n
an open-arc furnace (tilting) and used as an
Intermediate charge material in the production
of low-carbon ferrochrome.
Minute solid particles generated by the condensa-
tion of vapors from solid matter after volatiliza-
tion from the molten state.
An electric furnace with a water-cooled cover
over the top to limit the admission of air to burn
the gases from the reduction process. The furnace
may have sleeves at the electrodes (fixed seals)
with the charge introduced through ports in the
furnace cover, or the charge may be introduced
through annular spaces surrounding the electrodes
(mix seals).
G-l
-------
Electrolytic process
Exothermic process
Ferroalloy
A low-voltage direct current causes the simple
ions of the metal contained in an electrolyte of
modest concentration to plate on the cathodes as
free metal atoms. The process is used to produce
chromium and manganese metal, which are included
with the ferroalloys.
Molten silicon or aluminum or a combination of
the two combines with oxygen of the charge, generating
considerable heat and creating temperatures of several
thousand degrees in the reaction vessel. The process
is generally used to produce high-grade alloys with
low carbon content.
An intermediate material used as an additive or
charge material in the production of steel and other
metals. Historically, these materials were ferrous
alloys, hence the name. In modern usage, however,
the term has been broadened to cover such materials
as silicon, calcium silicon, calcium carbide, etc.,
which are produced in a manner similar to that used
for the true ferroalloys.
G-2
-------
Induction furnace
Open-arc furnace
Open furnace
Pre-baked electrode
Induction heating is obtained by changing the
frequency to electric conductors composing the
charge, and may be considered as operating on the
transformer principle. Induction furnaces, with
low frequency or high frequency, are used to produce
small tonnages of specialty alloys through remelting
of the required constituents.
Heat is generated in an open-arc furnace by the
passage of an electric arc either between two elec-
trodes or between one or more electrodes and the
charge. The open-arc furnace consists of a furnace
chamber and two or more electrodes. The furnace
chamber has a lining which can withstand the
operating temperatures and which is suitable for the
material to be heated. The lining is contained
within a steel shell which, in most cases, can be
tilted or moved.
An electric furnace with the surface of the
charge exposed to the atmosphere, whereby the
reaction gases are burned by the inrushing air.
An electrode purchased in finished form, avail-
able in diameters up to about 152 cm (60 inches).
These electrodes come in sections with threaded ends
and are added to the electrode columns.
G-3
-------
Reducing agent
Self-baking electrodes
Sintering
Slag
Stoking
Carbon-bearing materials such as metallurgical
coke, low-volatile coal, and petroleum coke used
in the electric furnace to provide the carbon which
combine with oxygen from the metallic oxides in the
charge to form carbon monoxide.
The electrode consists of a steel casing filled
with a paste of carbonaceous materials quite similar
to those used to make pre-baked amorphous carbon
electrodes. The heat from the passage of current
within the electrode and heat from the furnace
itself bakes the electrode and volatilizes the
asphaltic binders in the paste.
The formation of larger particles, conglomerates,
or masses from small particles by heating alone, or
by heating and pressing, so that certain constit-
uents of the particles coalesce, fuse, or other-
wise band together.
The more or less completely fused and vitrified
matter separated during the reduction of a metal from
its ore.
The means by which the upper portion of the charged
materials in the furnace are stirred up. This loosens
the charge and allows free upward flow of furnace gases.
6-4
-------
Submerged-arc furnace
Tapping
Tapping period
Tapping station
In ferroalloy 'reduction furnaces, the elec-
trodes usually extend to a considerable depth into
the charge; hence, such furnaces are called
"submerged-arc furnaces." This name is used for
the furnaces whose loads are almost entirely of
the resistance type.
A process whereby slag or product is removed
from the electric submerged-arc furnace.
That period of time during which product or
slag flows from the electric submerged-arc furnace.
The general area where molten product or slag
is removed from the electric submerged-arc furnace.
G-5
-------
°F
acfm
scf
scfm
psig
gr
ppm
mg/m
ton
gross ton
kv-a
kw
Kw-hr
mw
mw^hr
Silvery iron
50% FeSi
MgFeSi
Si
CaSi
SMZ
HC FeMn
ABBREVIATIONS
temperature, degrees Fahrenheit
actual cubic feet per minute
standard cubic feet measured at 70°F and 29.92 in. Hg.
standard cubic feet per minute at 70°F and 29.92 in. Hg.
pounds per square inch guage
grain (1 grain equals 64.8 milligrams) -
7000 grains equal 1 pound
parts per million
micrograms per cubic meter (0°C and 760 mm Hg)
milligrams per cubic meter (0°C and 760 mm Hg)
weight of 2,000 pounds avoirdupois, also, short ton
or net ton
weight of 2,240 pounds avoirdupois
kilovolt-ampere
kilowatt, 1,000 watts
kilowatt hour
megawatt, million watts
megawatt hour
15% to 20% ferrosilicon
50% ferrosilicon
magnesium ferrosilicon
silicon metal
calcium silicon
silicon manganese zirconium
high-carbon ferromanganese
G-6
-------
MC FeMn
LC FeMn
SiMn
FeMnSi
Chg Cr
HC FeCr
LC FeCr
FeCrSi
Cr ore/lime melt
Chemical Symbols
A1203
CO
co2
CaC2
CaO
Cr203
Fe
H20
MgO
Mn
P
SO,,
medi urn-carbon ferromanganese
low-carbon ferromanganese
silicomanganese
ferromanganese silicon
charge grade ferrochrome
high-carbon ferrochrome
low-carbon ferrochrome
ferrochrome-si1i con
melted chrome ore and Hme (CaO) 1n oxide form
alumium oxide
carbon monoxide
carbon dioxide
calcium carbide
calcium oxide (quick lime)
chromium oxide
iron
water
magnesium oxide
manganese
phosphorus
sulfur dioxide
G-7
-------
H. REFERENCES
1. DeHuff, Gilbert L. Ferroalloys. In: Mineral Facts and Problems.
U.S. Bureau of Mines Bulletin 630, 1965 edition, p. 330.
2. Ferroalloys: Steel's All-Purpose Additives. 33/The Maqazine of
Metals Producing p.39-56. February 1967.
3. Minerals Yearbooks. U.S. Bureau of Mines, 1963-1971.
4. U.S. Department of Commerce, Bureau of the Census.
5. Trends in the Use of Ferroalloys by the Steel Industry of the United
States. Washington, D.C., National Materials Advisory Board (NAS-NAE),
July 1971.
6. Watson, George A. (The Ferroalloys Association). Ferroalloys — A Billion
Dollar Industry. Mining Congress Journal, p. 140-143. February 1971..
7. World Stainless Steel Statistics. London, Metal Bulletin Books, Ltd.,
1972 edition.
8. A Systems Analysis Study of the Integrated Iron and Steel Industry.
Battelle Memorial Institute. Columbus, Ohio. Contract No. PH 22-68-65.
May 15, 1969.
9. Mantell, C. L. Electrochemical Engineering. 4th edition. New York,
McGraw-Hill, 1960.
10. Durrer, R. and G. Volkert. The Metallurgy of Ferro-Alloys. Revised
edition. 1972.
11. Paschkis, V. and John Person. Industrial Electric Furnaces and Appliances.
2nd edition. New York, Interscience Publishers, 1960.
12. Elyutin, V. P. et al. Production of Ferroalloys Electrometallurgy. 2nd
edition. Washington, D.C., National Science Foundation and Department of
the Interior (translated from Russian by the Israel Program for Scientific
Translations), 1957.
H-l
-------
13. Hopper, Rex T. The Production of Ferromanganese. Journal of Metals.
p. 88-92. May 1968.
14. Wowk, Z. B. Silicon Alloy Production in Canada. (Presented at the 1971
Electric Furnace Conference. Toronto, Canada.)
15. Minerals Yearbooks. U.S. Bureau of Mines, 1960-1970.
16. Questionnaires used in the ferroalloy industry study conducted by EPA
and The Ferroalloys Association.
17. Person, R. A. Control of Emissions from Ferroalloy Furnace Processing.
Union Carbide Corp. (Presented at the 27th Electric Furnace Conference.
Detroit. December 10-12, 1969.)
18. Silverman, L. and R. A. Davidson. Electric Furnace Ferrosilicon Fume
Collection (pilot plant study). Journal of Metals, p. 1327-1335.
December 1955.
19. Retelsdorf, H. J., et al. Experiences with an Electric Filter Dust
Collecting System in Connection with a 20 MW Silicochromium Furnace.
Source unknown, p. 66-79.
20. Participate Pollution System Study. Midwest Research Institute.
Kansas City, Mo. EPA Contract No. CPA 22-69-104. 1971.
21. The News and Courier. Charleston, S.C. April 12, 1971.
22. Electric Submerged-Arc Furnaces for the Production of Ferroalloys and
Calcium Carbide, Test Data Summary for New Source Performance Standards.
Environmental Protection Agency. 1973.
23. Test summary results developed by study team from contractor's field
testing reports.
24. Dobryakov, M. Z., et al. Operation of a Gas-Cleaning System on a Closed-
Top Electric Furnace. Steel' in the USSR. p. 401-402. May 1971.
H-2
-------
25. Dry Purification of Reaction Furnace Gases According to the SKW Filter
Candle Process. Dortmund, Germany, Friedrich Uhde G.m.b.H., May 1971. 5 p.
26. Dealy, J. 0. and A. M. Kill in. Observations of Covered Ferroalloy
Furnaces Operating in Belgium and Norway. Unpublished trip report.
27. Kalika, P. W. How Water Recirculation and Steam Plumes Influence
Scrubber Design. Chemical Engineering, p. 133-138. July 28, 1969.
28. Scott, J. W. Design of a 35,000 K.W. High Carbon Ferrochrome Furnace
Equipped with an Electrostatic Precipitator. The Metallurgical Society
of AIME. New York, N.Y. Paper No. EFC-2.
29. Lund, Herbert F. (ed.). Industrial Pollution Control Handbook. New
York, McGraw-Hill, 1971.
30. Federal Register. Vol. 36, No. 158. August 14, 1971.
31. Calcium Carbide -- Salient Statistics. Chemical Economics Handbook.
724.5020C. January 1972.
32. Air Pollution Control Technology and Costs in Nine Selected Areas.
Industrial Gas Cleaning Institute. EPA Contract No. 68-02-0301.
September 30, 1972.
33. Moody's Industrial Manual. Standard and Poor's Corporate Reports.
1963-1971.
34. Durkee, K. R. International Trip Report, Survey of Japanese Ferroalloy
Kurnaces. August 9, 1973.
35. Hyland, G. R. Test for particulate emissions using an IKOR Continuous
Particle Monitor (informal report) June 6, 1972.
36. Ferrari, Renzo. Experiences in Developing an Effective Pollution Control
System for a Submerged Arc Ferroalloy Furnace Operation. Journal of
Metals, p. 99. April 1968.
H-3
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/2-74-008
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Air Pollution Control Engineering and Cost
Study of the Ferroalloy Industry
5. REPORT DATE
May 1974
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James O. Dealy and Arthur M. Killin
a. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG-VNIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Office of Air Quality Planning and Standards
Control Programs Development Division
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
Control Programs Development Division
National Environmental Research Center
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
-:h Triangle Park. N.C. 27711
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Report includes a number of studies of several ferroalloy plants and
provides information of the following aspects of the industry:
1. Atmospheric emissions from production of ferroalloys and
calcium carbide.
2. Methods and equipment used to limit these emissions.
3. Cost and economic impact of air pollution controls.
4. Industry characteristics such as growth rate, raw materials,
processes, consumer products, and number and location of
producers.
Most of the information herein was gathered from industry questionnaires
and EPA source tests.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
I).IDENTIFIERS/OPEN ENDED TERMS C. COSATI 1 leld/Group
13. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (This Report/
None
21 NO OF PAGES
410
20 SECURITY CLASS (Thispage)
None
22. PRICE
CPA Farm 2220-1 (t-73)
H-4
*U.S. Government Printing Office: 1974—747-795/364 Kegion NO. 4
-------