EPA-450/2-74-018-C
April 1976
BACKGROUND INFORMATION
FOR STANDARDS OF PERFORMANCE:
ELECTRIC SUBMERGED ARC FURNACES
FOR PRODUCTION OF FERROALLOYS
VOLUMES: SUPPLEMENTAL INFORMATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/2-74-018-C
BACKGROUND INFORMATION
FOR STANDARDS OF PERFORMANCE:
ELECTRIC SUBMERGED ARC FURNACES
FOR PRODUCTION OF FERROALLOYS
VOLUME 3: SUPPLEMENTAL INFORMATION
Emission Standards and Engineering Division
I .S. ENVIRONMENTAL PROTECTION VGEiNCY
Office of Air and Waste Management
Office of Air Quality Planning and Standard!*
Roearch Triangle Park. North Carolina 2771 I
April 1<>76
-------
This report has been reviewed by the Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards, Office of Air and
Waste Management, Environmental Protection Agency, and approved for
publication. Mention of company or product names does not constitute
endorsement by EPA. Copies are available free of charge to Federal
employees, current contractors and grantees, and non-profit organiza-
tions - as supplies permit - from the Air Pollution Technical Information
Center, Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, or may be purchased from the National Technical Informa-
tion Service, 5285 Port Royal Road, Springfield, Virginia 22161.
Publication No. EPA-450/2-74-018-C
ii
-------
Table of Contents
Page
I. Introduction 1
II. Mass Standards 3
A. Rationale for Proposed Standards 3
B. Comments on Achievability of Standards 7
1. Representativeness of Area Sampled 8
2. Anisokinetic Sampling 10
3. Flowrate Determinations 15
4. Manufacturers' Guarantees 19
5. Additional Emission Test Results 22
6. Effect of Quality of Charge Materials 32
C. Summary and Conclusions 35
III. Opacity Standard 38
IV. Alternative Test Procedures for Open Fabric Filter
Collectors 47
V. Economic and Inflationary Impact 51
VI. Summary of Public Comments and EPA Responses 55
References 76
Appendix A - Summary of Emission Data Received in Response
to Section 114 Letters
Appendix B - Six Minute Average Opacity Values
iii
-------
I. Introduction
On October 21, 1974 (39 FR 37470), the Environmental Protection
Agency oroposed standards of performance for new ferroalloy oroduction
facilities. At that time, EPA made available to the public
detailed information on the bases for the standards in "Background
Information for Standards of Performance: Electric Submerged-Arc
Furnaces for Production of Ferroalloys," EPA-450/2-74-018a and b
(October 1974). Copies of the document may be obtained upon
written request from the Emission Standards and Engineering Division,
Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, Attention: Mr. Don R. Goodwin.
This report, which provides supplemental information on the
standards of performance, is being issued in conjunction with
promulgation of the regulation and discusses the issues associated
with the final rulemaking action. As a result of comments
on the proposed standards, the issue of the achievability of the
standard limiting emissions from chrome and manganese alloy
production to 0.23 kg/MW-hr (0.51 Ib/MW-hr) arose and resulted in
EPA's revaluation of the bases for each of the mass standards
of nerformance. Chapters II and IV contain a discussion of the
results of the revaluation including a consideration of
alternative emission test procedures for demonstrating compliance
of open, pressure fabric filter collectors. Chapter II also
contains additional emission test data for well-controlled facilities.
The emission data obtained after proposal of the standards are
presented in Appendix A.
-------
On November 12, 1974 (39 FR 39872), EPA revised the general
provisions aoplicable to standards of performance and Reference
Method 9, the test method for determinino comoliance with onacity
standards. The standards of performance for ferroalloy production
facilities were proposed before Reference Method 9 was revised.
Accordingly, EPA reevaluated the opacity standard to determine
the appropriate level of the standard consistent with the
changes in Method 9. A discussion of the November 12, 1974,
revisions and the revaluation of the ferroalloy opacity standard
is in Chapter III.
A summary of updated control costs is presented in Chapter V,
and a summary of public comments and EPA's responses to them is in
Chapter VI. Eighteen comment letters on the proposed standards
of performance were received and considered. Also, approoriate
Federal agencies and departments were consulted prior to publica-
tion of the final standards of performance, as required by
section 117(f) of the Clean Air Act. All comments received were
helpful in pointing out problems with the proposed standards and
defining areas where a better explanation of the bases for the pro-
vision was necessary.
-------
II. Mass Standards
A. Rationale for proposed standards.
The proposed mass standards of performance for ferroalloy production
facilities were based on ? study of the onerations, emissions, and
control techniques of the industry and the measured emission rates
from well-controlled facilities. As a result of the study, EPA determined
that the best system of emission reduction (considering costs) for open
furnaces is a well-designed fume capture hood that minimizes the
induced air volume and a well-designed fabric filter collector or
high-energy venturi scrubber. In EPA's judgment the emission limitations
achievable by the best systems of emission, .reduction are 0.45 kg/MW-hr -
(0.99 Ib/MW-hr) for high-silicon alloys and 0.23 kg/ MW-hr
(0.51 Ib/MW-hr) for chrome and manganese alloys. The bases for the
selected emission limitations are discussed below.
In the study of the industry prior to proposal of the standards,
EPA conducted emission tests on seven open electric submerged arc
furnaces, two of which were controlled by closed pressure fabric
filter collectors, two by open pressure fabric filter collectors, two
by venturi scrubbers, and one by an electrostatic precipitator. The
emission rates determined for these seven facilities are shown in
Fiqure II-l which is reproduced from the report on this study.* Emissions
in excess of the standards were reported only at facilities which
had either improperly maintained the control system or used systems
*Background Information for Standards of Performance: Electric
Submerged Arc Furnaces for Production of Ferroalloys, EPA-450/2-74-018a
and b.
-------
FIGURE II-I
PARTICULATE EMISSIONS FROM OPEN ELECTRIC SUBMERGED-
ARC FURNACES PRODUCING FERROALLOYS
3.0
2.5
2.0
00
00
1.5
O
I—
or
=L'
o.
1.0
0.5
FURNACE
•ROL EQUIPMENT
JRNACE SIZE, Mw
PRODUCT
T
<
H
ft
c &<
— "I u
1 1 *)
1 1
•oft*
*^
1 1
£ LI L2
r v(2) v(3) \
( 27 27
SiMn
1
-
t
C
|
L
l(
I
~T T~
KEY
EPA
A MAXIMUM
ii DATA POINT
H-H AVERAGE
u MINIMUM
i> DATA POINT
J (1) INCLUDES PARTIAL CONTROL OF TAPPING
FUME.
(2)AP=57 in. W.G.
^ (3) AP = 47 in. W.G.
H (4)AP = 37 in. W.G.
i
C |g
0
R
,1 II
c c
u 'ci|
Pi ' '
rl ' '
v i i
0
5 DATA
POINTS
1 1 W< 1 1 1
3 Nd) T(D U(D Q 0 M
1) V P B B B B V-VENTURISCR
P- ELECTROSTA
1 7 40 18 20 27 ' PRECIPITA
1 H.C. I FeCrSi | 75% 1 Si B- BAGHOUSE
FeCr FeSi
-------
other than the best system of emission reduction. That is, at the
time of the emission tests, facility N had an inadequately designed
scrubber; facility 0 had patched, leaky bags in the fabric filter
collector; and facility T's control system, an electrostatic
precipitator alone, was not representative of the best system of emission
reduction. The emission rates for the seven facilities generally
do not reflect control of tapping fumes. Emission rates which
included tapping fumes could not be measured because no facilities
with both effective tapping fume collection and an independent
control system amenable to emission measurements could be located
at the time of the study. Accordingly, calculation methods were
used to determine the equivalent emissions from a furnace with
effective capture and control of tapping fumes. The calculation
procedure was based on measured uncontrolled emission rates for
tappinq, an assumption of 100 percent capture of tanpina fumes, and
a control device with a 99 percent collection efficiency. Using
a conservatively high value of 68 kq/hr (150 Ib/hr) of uncontrolled tapping
fumes, the effect of including tapping fumes from a 30 MW furnace
tapped for 15 minutes over a furnace cycle oeriod of 75 minutes would
be an increase in the emission rate of 0.0045 kg/MW-hr (0.01 Ib/MM-hr).
From the evaluation of the test data, the control systems, and the effect of
tapping fume control, EPA determined that emissions from a well-
controlled open furnace producing silicon metal or 75 percent
ferrosilicon can be controlled to less than 0.45 kg/MW-hr
(0.99 Ib/MW-hr) and emissions from a well-controlled open
furnace producing high carbon ferrochrome, silicomanganese, or
5
-------
ferrochrome silicon can be controlled to less than 0.23 kg/MW-hr
(0.51 Ib/MW-hr).
The emission limitations require that the gas volume
treated by the control device be limited. Gas volumes from
open furnaces vary with product and hood design because of differences
in the volatility of the furnace charge materials, differences in the
quantity of gaseous by-products of the carbothermal smelting process,
differences in furnace capacity (diameter), and differences in the
opening between the furnace and the hood. As typically d'.-signed in
the U. S., an open f-jrnace has a water-cooled canopy hood located about
1.5 to 2.5 meters (5 to 8 feet) above the furnace crucible rim. The large
vertical opening between the furnace crucible and the hood permits large
quantities of induced air to be drawn into the emission collection
system and may dilute the gases produced in the furnace by as much
as 50 to 1. Typical qas volumes from ooen furnaces hooded in this
manner vary from 50 to 190 standard cubic meters per second (sons)
[100,000 to 400,000 standard cubic feet per minute (scfm)]. The
volume of induced air can be reduced by decreasing the size of the
opening between the furnace crucible and the hood. Demonstrated methods
of reducing the quantity of induced air include: (1) positioning
the hood closer to the furnace, (2) enclosing the area between the
furnace and the hood with refractory-lined or water-cooled doors or with
chain curtains, and (3) air curtains. These methods are in routine use
1 2
at ferroalloy production facilities in the U. S. and in foreign countries. '
By emphasizing reduction of gas volumes during design of the furnace
hood, the air volume requiring treatment by the control device can be
6
-------
substantially reduced, and the standards can be easily met. However,
results of EPA tests at two facilities with furnace hoods not specifically
designed to minimize induced air volumes showed that their emissions could
be controlled to levels below the applicable standards.
On the basis of a thorough evaluation of the emission data and
information from the engineering study, EPA proposed standards of
performance applicable to all the major silicon, chrome, and manganese
alloys based on emission data for five of the alloys. The study
of the ferroalloy industry revealed that although emission rates
from uncontrolled furnaces can vary greatly among alloys, similar
alloys often have similar levels of controlled emissions from a
given type of furnace. Thus, standards of performance applicable
to groups of ferroalloys were developed and proposed. Each group
of ferroalloys consists of alloys having similar controlled
emission rates and control equipment. This similarity in emissions
of proHucts within an alloy group is shown by the common practice
in the industry of combining emissions from several similar alloys
and controlling by a single complex of control devices.
B. Comments on achievability of standards.
Commenters on the achievability of the 0.23 kq/MW-hr (0.51 Ib/MW-hr)
standard expressed concerns in two main areas. First, the commenters
questioned the representativeness of the data used to demonstrate the
achievability of the standard. Specifically, the commenters were
concerned that sampling of only a limited number of compartments or
control devices serving a facility, nonisokinetic sampling, and the
procedures used to determine the total gas volume flow from open fabric
7
-------
filter collectors would bias the data. As additional support for their
position, the commenters also claimed that control equipment vendors
will not guarantee that their equipment will achieve 0.23 ka/W-hr
(0.51 kg/MW-hr). The second area of concern to the commenters was
expressed by their opinion that use of fine ores as charqe materials
will increase the furnace emission rate and preclude achievina the
standard. For these reasons, the commenters argued that the standard
of performance should be promulgated as 0.45 kg/MW-hr (0.99 Ib/MW-hr)
for all alloys. In EPA's opinion, a standard of 0.45 kg/MW-hr (n.99
Ib/MW-hr) would al'.ow installation of systems other than best systems
of emission reduction. The bases for this conclusion are discussed below.
1. Representativeness of area sampled.
In the tests on facilities 0 and U, emissions were sampled in
one of the three stacks discharging the effluent from the baqhouse
while total flow rate from the facility was monitored. The mass emission
rates (kg/hr) of these facilities were calculated assuming the effluent
concentration was the same in all three stacks. Similarly, emissions
from facility M are based on the test results for one of three collectors
and total flow from the complex of three fabric filter collectors.
The commenter argued that these assumptions are not appropriate because
standards of performance should be based on actual, accurate measurements,
EPA tests, and allows testing of, a representative number of
stacks on or compartments in a control device because subsections of a
properly operating control system should perform equivalently.
The procedures used at facilities 0 and U were reasonable oecause the
other spctinns of the control system were of equivalent design and in
8
-------
equivalent or better operating condition to that of the tested section.
The operating conditions of these two facilities were such that there
was no reason to believe that the performance of any section of the
device would be significantly different. The procedure used in
determining the emission rate from facility M also did not result in
emission data prejudicial to industry. The malfunctioning sections
were shut off, and the tested collector was cleaning a greater than
normal proportion of the total gas volume flow. Thus, the concentration
of the emissions possibly was greater than during periods of proper
operation of fabric filter collector.
In addition to possible variations in performance of sections of
the control device, the commenter was also concerned that different
sections of the control device would experience a different inlet
particulate loading. For control systems on electric submerged arc
furnaces, this should not be a problem assuming proper distribution of
the inlet gas stream to the control devices. Particulate matter emissions
from well-controlled electric submerged arc furnaces are known to consist of
particles of less than two micrometers diameter. ' ' Particles of this size do
not have appreciable inertia, hence, they follow the motion of the gas
stream. Accordingly, there will not be any size segregation of the
particulate matter as it travels from the furnace to the control devices.
Although the confidence in the data would be slightly improved by
measuring from all stacks of a control device or all
compartments of a fabric filter collector, there would be a significant
9
-------
increase in the cost of emission testing and determination of
compliance with the standards of performance. EPA does not
believe that this added degree of confidence in the data is sufficient
to justify the increased costs.
2. Anisokinetic sampling.
The commenter noted that EPA was inconsistent in its acceptance
of emission test data obtained under nonisokinetic sampling conditions.
Specifically, the commenter observed that EPA did not use results
from runs conducted nonisokinetically at facilities amenable to
testing according to Method 5, but EPA did use data obtained nonisokineti-
cally at two open fabric filter controlled facilities. The commenter
contended that the standards should not be based on emission data obtained
nonisokinetically since the results do not represent the true emission
level of the facility.
The objective of Reference Method 5 is to establish a uniform
procedure for determining particulate matter emissions with maximum
accuracy and precision under a variety of conditions. There are aspects
of the method which, subject to engineering judgment, can be considered
flexible in certain cases without sacrificing the reliability of the
test results. Thus, understanding the reasons, and justifications, for
the apparent inconsistencies in the data reduction and test procedures
employed requires understanding the reasons for the requirements in the
method.
Reference Method 5 specifies criteria for the velocity of a sample
collection (sampling velocities must equal the gas stream velocities
10
-------
*
within a soecified error ) to ensure collection of a representative
sample of the particles suspended in the moving gas stream. The
errors in measured concentrations determined under conditions of
nonisokinetic sampling are due to the inertia of heavier particles.
That is, if the velocity in the sampling nozzle is greater than the
gas stream velocity, a convergent gas stream will develop at the
nozzle, and a greater amount of the lighter particles will be collected.
Particles possessing sufficient inertia will not converge into the
nozzle and will not be collected. Conversely, if the velocity in the
nozzle is less than the gas stream velocity, part of the approaching
gas stream is deflected and does not enter the probe. The heavier
particles continue into the probe, resulting in an overestimation
of the concentration. Analysis of the bias resulting from nonisokinetic
sampling of particles has shown that for particles with aerodynamic
diameters of five micrometers diameter or less, and for sampling at
extremely low velocities, the error is insignificant and isokinetic
sampling is not necessary. ' That is, a reasonably representative
sample of particles of less than five micrometers diameter can be
obtained by nonisokinetic sampling.
Sampling of particulate matter emissions in well-defined flows,
such as those found in stacks, can be conducted isokinetically and the
resultant sample concentration will be representative of the effluent
concentration. Emission tests by EPA on we!1-controlled facilities
which discharge emissions through stacks were conducted in accordance
Sampling at nozzle velocities equal to the gas stream velocity is a
condition known as isokinetic sampling.
11
-------
with the procedures of Method 5. As specified by Method 5, data from
runs' conducted at sampling rates outside the range 90 <_ I >_ 110 per-
cent were not used as part of the bases for the standards of performance.
Analysis of the variance of the nonisokinetic sample concentrations
from the mean test results shows no statistically significant difference
in the two groups of data. Thus, consideration of only isokinetic
data was not strictly necessary.
Emissions from open pressure fabric filter collectors cannot be
sampled strictly according to the criteria of reference methods
currently in Appendix A of 40 CFR Part 60. The effluent from
pressurized fabric filter collectors is discharged at low velocities
from large areas. Consequently, isokinetic sampling is difficult
using existing equipment and procedures of Reference Method 5. For this
reason, in the tests of open pressure fabric filter collectors conducted
by EPA, the criteria regarding the relative velocity of sample collec-
tion could not be followed and alternative approaches were used to
sample emissions from the specific facilities.
The only ideal alternative to emission testing using modified
Method 5 procedures was to prohibit use of open fabric filter collectors
because of the inability to evaluate their performance. Factors such
as significantly lower installation costs and ease of identification
and replacement of leaking bags make open fabric filter collectors
superior to closed. Such practical considerations supported avoiding
the loss of an effective control system not amenable to sampling
using Reference Method 5. For these reasons, provisions were
established in §60.266(d) of the regulation to require construction
12
-------
of the control device to allow measurement of emissions and flow
rates using applicable test methods and procedures. This section
permits the use of open fabric filter collectors (or other control
devices) if compliance can be demonstrated by an alternative test
procedure. Under the provisions of 40 CFR 60.8(b), the owner or
operator of an affected facility may upon approval by the EPA
Administrator show compliance with the standards by an "alternative"
procedure. Thus, the approach taken by EPA in emission testing of
open fabric filter collectors is not prejudicial to the source and
allows use of systems that would otherwise be prohibited.
In EPA's judgment, it is possible to representatively measure
emissions from open fabric filter collectors by nonisokinetic sampling
if the effluent does not contain large particles such as would result
from leaky or torn bags. Emissions from well-controlled ferroalloy
furnaces were shown in the joint EPA-Ferroalloy Association "Engineering
and Cost Study of the Ferroalloy Industry" (EPA-450/2-74-008)3 to consist
of particles of less than two micrometers aerodynamic diameter for all
alloys. These data are consistent with size distribution data reported
4 5
in other studies ' and expected size distributions of aerosols formed
by condensation. The mass and, hence, inertia of these particles are
negligible. Thus, sampling of this metal oxide fume at superisokinetic or
subisokinetic rates is not expected to significantly bias the measured
concentration from the actual effluent concentration. Figure II-2 shows
the relationship of sample validity to nozzle and gas stream flow
conditions. Examination of Figure II-2 shows that sampling of
13
-------
C
CC
O!
U
C
C
•c
01
to
CO
C
O
no
s-
O)
O
O
O
-U
us
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Ratio Air Stream Velocity to Satnnle Nozzle Velocity
Superisokinetic
t
Isokinetic
Subisokinetic
Figure 11-2. Effect of Nonisokinetic Samplina on Measured Samnle Concentration
-------
particles of less than five micrometers aerodynamic diameter at non-
isokinetic rates results in deviations from the true concentration of
less than five percent. For the above reasons, EPA does not believe
that superisokinetic samplinn of emissions from properly operating
fabric filter collectors would result in measured concentrations
significantly lower than the actual concentration in the effluent.
The reported concentrations of effluent for facilities M and Q are
considered to be representative of the actual concentrations.
3. Flowrate determinations.
As part of an overall concern with the reoresentativeness of the
test data, the commenter questioned the validity of the procedures
used to determine the total volume flow from open fabric filter
collectors and the validity of the mass emission rate. Specifically,
the commenter argued that the use of induced flows measured on one day
to calculate the total gas volume flow on another day does not give
an accurate assessment of the mass emissions from facility M. In
response to this part of the comment on the achievability of the standards
of performance, EPA reassessed the representativeness of the calculated
total gas volume flow for the open fabric filter collector and concluded
that it was sufficient to adequately determine the mass emission rate
of the facility.
Pressurized fabric filter collectors used by the ferroalloy
industry in the U. S. generally exhaust the gases through a roof
monitor(s) or an open roof. Thtre are open grates at the bottom of the
collector (at the cell plate) to provide cooling of the gases
15
-------
and bags by natural convection. The presence of this dilution air
added after the bags complicates measurements of the mass emission
rate from this type of collector.
The mass emission rate from open fabric filter collectors can be
determined from measurements of the concentration and the average
velocity in the discharge area or from measurements of concentration,
inlet gas volume, and the induced air volume. For facility M, EPA
calculated the total volume flow rate from the collector from measure-
ments of the inlet gas volume flow and the flow rate of air induced
into the baghouse. In this test the inlet gas volumes were measured
during each run, but the volume of air induced into the collector was
measured once during the period of the emission test. The total gas
flow reported for facility M's test runs was the sum of the inlet gas
volume for the run and the induced air volume.
Although the procedure used for determining the total gas
volume flow was not ideal, the reported gas volumes are considered to
be reasonably representative of the actual total gas flow rate. The
basis for this conclusion is that the quantity of cooling air induced
around the bags in an open collector is primarily dependent on the
temperature of the inlet gas stream and the ambient air. Therefore,
equivalent air volumes will be drawn into the collector under
similar meteorological and inlet conditions. During the period of
emission testing at facility M the meteorological conditions and inlet
ges temperatures were fairly uniform and, thus, the volume of induced
air was expected to be constant. Consequently, measurement of induced
air velLnies once durinn the test "'as expected tc orovide ap adeouate
16
-------
estimate for calculation of the total volume flow from the
facility.
Since conducting this test EPA has gained additional
experience and has concluded that, in general, it is preferable
to measure the total gas volume flow during each run of a performance
test. This conclusion, however, does not invalidate the use of
test data obtained at facility M because calculation of gas volumes
from temperature data show that the original assumptions were valid.
Calculation of induced air volumes from enthalpy balances using
the reported inlet and ambient temperatures results in estimates of
gas volumes which are consistent with the measured volumes. The
volumes were calculated assuming adiabatic mixing (no heat loss) and,
hence, perhaps may represent conservatively high values. The calculated
and measured dilution air volumes and the total volume flow from the
collector are shown in Table II-l. For facility M at the time of the
emission testing, the induced air represents a small portion of the
total flow; consequently, small fluctuations do not significantly
affect the reliability of the total volume flow rate value.
The validity of the assumption of relatively stable induced
air flow rates is also shown by calculations of the effects of
ambient temperature fluctuations on induced air volumes and total
gas flow from the collector. During the period of emission testing
o
at facility M, the ambient temperatures fluctuated less than 6 K
(10°F). Using enthalpy balance calculations, it can be shown that
this temperature change has an insignificant effect on the total
gas volume flow relative to the larger fluctuations occurring in the
17
-------
TABLE II-l.
Comparison of Calculated and Measured Dilution Air Volumes for
B-Baghouse at Facility MB
Run No. 123
Inlet Temperature, °k 391 394 3^7
Outlet Temperature, °k 355 378 366
Ambient Temperature, °k 273.1 277 277
Inlet Volume, sons7' 76.9 102.1 96.5
Measured Dilution Volume 30.7 30.7 30.7
sons
Total Volume, sons 107.8 132.8 127.2
Calculated Dilution Volume, 31.5 17.4 30.9
scms
Calculated Total Volume,
scms
108.4 lly.b 1^7.4
Ratio Total/Calculated
Total 1.01 0.90 1.00
scms - standard cubic meters per second
18
-------
inlet gas volumes. Table II-2 compares the calculated dilution air
volumes for two ambient temperatures and the measured variations in
inlet gas volumes for one collector. These data show that the
fluctuations in the inlet gas volume to the collector over a three-day
period are greater by a factor of 30 than the corresponding variations
in induced air volumes. Therefore, relative to the overall variations
in total gas volume discharged by facility M, errors in the estimates
of the induced air volume are minor and the corresponding errors in
the mass emission rate are not significant.
4. Manufacturers' guarantees
The commenters cited as further evidence of the unachievability
of the standard the refusal of manufacturers of fabric filter
collectors to guarantee that their equipment will reduce emissions be-
low 0.23 kg/Mk'-hr (0.51 Ib/MW-hr). It is EPA's opinion that the re-
fusal of control equipment vendors to guarantee reduction of emissions
to less than 0.23 kg/MW-hr (0.51 Ib/MVi-hr) does not demonstrate the
unachievability of the standard. The reluctance of vendors to guarantee
achieving the standard is not surprising considering the variables
which are beyond the control of the vendor. Specifically, the vendor
does not determine the gas volume cleaned by the control device and,
therefore, cannot directly control the facility's mass emission rate
or the operation and maintenance of the collector. In addition, the
lack of experience with performance testing of collectors makes it
difficult to evaluate the performance of the collector over the
guarantee period.
19
-------
TABLE 11-2.
Comparison of Effect of Temperature on Induced Air
Volume vs Inlet Volume VariationsS
Run No. 13
Inlet Temperature, °k 3-'l 397
Ambient Temperature, °k 273.1 277
Inlet Volume, scms 76.9 96.5
Calculated Dilution Air, scms 31.5 30.8
Total Outlet Volume, scms 108.4 127.3
A(l-3) Dilution, Percent total 0.6 0.5
A(l-3) Inlet Volume, Percent total 18 . 18
20
-------
Vendors of control equipment rarely have any control over the
design of the furnace and tanning station collection systems. Since
a fabric filter collector tends to control the concentration of
particulate matter in the effluent, the pass emission rate is
governed by the total volumetric flow rate from the control device.
Compliance with the standards of performance is not accomplished solely
by application of a well-designed control device. Attention must be
directed also toward control of furnace gas volumes. Because of
limited experience with emission testing to evaluate performance of
open fabric filter collectors, the vendors do not have a good estimate
of the performance of these systems over the guarantee period. For
the vendor, establishment of the performance guarantee level is further
complicated by the fact that performance of the collector is contingent
upon its being properly operated and maintained.
Because of these factors, in the past vendors of open fabric
filter collectors have been willing to guarantee performance only at
levels which they are confident of achieving over the guarantee
period. Generally, the guarantee levels were approximately 0.027
to 0.046 g/dscm (0.012 to 0.02 gr/dscf) for control of emissions from
open electric submerged arc ferroalloy furnaces. These guarantee
concentration levels are substantially higher than the concentrations
(<0.003 gr/dscf) measured by EPA on well-designed and properly
operated and maintained open fabric filter collectors serving open
furnaces.
Standards of performance are necessarily based on data from a
limited number of best-controlled facilities and on engineerinq
21
-------
judgments regarding performance of the control systems. For this
reason, there is a possibility of arriving at different conclusions
regarding the performance capabilities of these systems. Consequently,
the question of vendors' reluctance to guarantee their equipment to
achieve 0.23 kg/W-hr (0.51 Ib/MW-hr) was considered in context with the
results of additional recent emission tests on fabric filter collectors.
Recent emission data obtained from the ferroalloy industry show that
effluent concentrations of less than 0.0090 g/dscm (0.004 qr/dscf)
have been measured in properly operating compartments of open fabric
filter collectors. The reported mass emission rates were all less
than 0.23 kg/MW-hr (0.51 Ib/MW-hr). The lowest effluent concentration
guaranteed by vendors for any of the tested collectors was 0.027 g/dsc"i
(0.012 gr/dscf). These data show that the performance of a properly
operated and maintained fabric filter is significantly better than
vendors are willing (or have in the past been willing) to guarantee.
(The data obtained from the industry are discussed in more detail in
section 5.)
On the basis of the above factors concerning vendors' guarantees
of fabric filter collectors and the reported emission test results,
EPA concluded that the reluctance of vendors to guarantee reduction
of emissions to levels below 0.23 kg/MW-hr (0.51 Ib/MW-hr) cannot be
considered as evidence of the unachievability of the standard.
5. Additional Emission Test Results
The Clean Air Act requires that standards of performance reflect
the degree of emission limitation achievable by application of the
22
-------
best system of emission reduction which has been adequately demonstrated.
Accordingly, the data bases for standards are necessarily limited.
Because of the controversy over the achievability of the 0.23 kg/MW-hr
(0.51 Ib/MW-hr) standard for ferroalloy facilities, EPA reviewed the
performance of fabric filter collectors, considering the industry's
experience with recently constructed furnaces.
In EPA's study of the ferroalloy industry, four facilities
controlled by fabric filter collectors were emission tested, although
not all collectors were properly maintained. Recognizing that the
data basis for the oroposed standards was limited and that a number of
well-controlled facilities had started operation since completion of the
original study, EPA obtained additional data to better evaluate the
performance of emission control systems of interest. Under the
authority of section 114 of the Clean Air Act, EPA requested copies
of all emission data for well-controlled furnaces operated by 10
ferroalloy producers. Data were received for five well-controlled
facilities, four of which were controlled by open fabric filter
collectors; the other was scrubber-controlled. In general, these
facilities had close fitting [1.52 meters (five feet) or less separation
between the furnace crucible and the hood] water cooled canopy hoods;
tapping fumes were collected and sent to the control device along with
the furnace emissions.
The emission data submitted by the industry are reported in
Appendix A and summarized in Figures II-3 and II-4. Only the emission
test on the scrubber-controlled facility, facility N2, was conducted
according to the procedures of Method 5. Due to the difficulties
23
-------
1
.30
.25
l/J
c
0
•r—
| .20
-------
.01
c
o
in
LU
I- <4-
a> o
4-J 10
S" i-
O)
-u
u
-u
t-
.001
.0001
Facility
Control Equipment
Furnace Size, MW
Product
EPA
Other Procedure
W
H2
S
7
FeCr
D
19
Si
20-25
Si
cc
L
19
no
1C
75"J
Figure II-4. Emission Concentrations Reported Dy Industry
-------
associated with sample collection in low velocity gas streams and
large open areas, the tests on the four open fabric filter collectors
could not be conducted according to Method 5. The industry used a number
of different approaches to sampling emissions from open fabric filter
collectors as there are no established procedures for testing these
systems. In general, the test procedure consisted of sampling at
superisokinetic rates at a single point in each of a representative
number of compartments in the collector. The emission sampling procedure
used at facility CC differed slightly in that special equipment which
allowed sampling at low velocities and isokinetic rates wa3 used; also,
the dilution air added after the fabric was restricted by blocking off
the grates at the cell plate. The test results for these four facilities,
consequently, are not of comparable accuracy and reproducibility; however.
the data are indicative of performance levels to be expected from well-
designed fabric filter collectors.
The effluent concentrations reported for these four fabric
filter collectors are consistent with the levels measured by EPA on
well-designed and properly operated and maintained collectors. In
EPA's emission study, effluent concentrations greater than 0.0090
g/dscm (0.004 gr/dscf) were found to be associated with leaking and
torn bags or improperly positioned bags in a compartment.
The data submitted by the industry show that emissions from
silicon metal or 75 percent ferrosilicon production can be controlled
to less than 0.0068 g/dscm (0.003 gr/dscf) and to 0.0* kn/""-"-r (n.015
Ib/MW-hr). Correction of the results reported for facility CC to
account for the effects of dilution air on the calculated mass
26
-------
emission rate increases the rate to 0.11 kg/MW-hr (0.24 Ib/MW-hr).
Evaluation of other possible sources of errors and uncertainties in
the data results in a maximum estimate of 0.20 kg/MW-hr (0.45 Ib/MW-hr)
for facility CC. All the data submitted by the industry are for
furnaces which are smaller than the maximum size presently being
constructed. The gas volume from an electric submerged arc furnace is
governed by the furnace diameter (size) in addition to the alloy being
produced and the design of the collection hood. Larger gas volumes,
and hence emission rates, are expected as the furnace capacity increases.
Taking into consideration the effect this scale factor would have on
the emission levels, EPA concluded that emissions from large well- "
controlled high-silicon alloy furnaces can be controlled to less than
0.45 kg/MW-hr (0.99 Ib/MW-hr) but that reduction to less than 0.23 kg/MW-hr
(0.51 Ib/MW-hr) is not achievable at this time.
Comparison of the supplemental data and the test data obtained
by EPA shows that the difference in the emission rates for open
furnaces is primarily attributable to a reduction in gas volumes from
the furnaces. In 1971-1972, when EPA conducted the emission tests
for the proposed standard, collection hoods were usually located about
1.8 to 2.5 meters (6 to 8 feet) above the furnace and no chain curtains
or refractory lined doors, or flaps, were used to minimize the volume
of induced air. The supplemental data received from the industry in
January 1976 were for recently constructed furnaces and, in general, the
collection systems were designed to minimize the gas volume carried by
the system. Table 11-3 compares the gas volumes from the older furnaces
tested by EPA with the gas volumes reported by the industry for the
newer furnaces. These data clearly show that substantial reductions in
27
-------
Table II-3. Comparison of Furnace
Gas Volumes from Open Furnaces
Furnace
SCMS
Furnace
Capacity
MW
SCMS/MW
A. Proposal Values
Typical, 75% FeSi 130.4
Typical, Silicon Metal
Facility M
Facility 0*
B> Industry Data
Facility AA
Facility BB
Facility CC
Facility DD*
Facility EE
217.3
192.1
74.6
30.6
67.76
62.29
64.18
129.77.
30
25
17
22
20.
21
15
18
30
4.35
8.69
11.30
3.39
4.54
2.99
4.15
3.57
4.33
Furnace producing 75% FeSi, all otiier facilities producing silicon metal
28
-------
gas volumes can be achieved. Comparison of the data for facilities M
and EE (packet-type silicon furnaces) shows that despite facility EE's
greater production capacity, the measured gas volume is 33 percent
lower than the measured gas volume for facility M. Similarly,
comparison of the measured furnace gas volumes for facilities AA,
BB, and CC (round silicon metal furnaces) with either the typical
value for the industry or facility's M value shows at least a 50
percent reduction in the gas volumes from the furnace.
In EPA's opinion the supplemental data clearly demonstrate that
substantial reductions in gas volumes from open furnaces are possible.
These data also reinforce EPA's previous conclusion that well-designed
and properly operated and maintained fabric filter collectors should
reduce the concentration of emissions from open furnaces to less than
0.0090 g/dscm (0.004 gr/dscf). By use of substantially lower gas
volumes and use of a well-designed and operated fabric filter
collector, emissions from electric submerged arc furnaces producing
high-silicon alloys will be less than 0.45 kg/MW-hr (0.99 Ib/MW-hr).
In EPA's study of the ferroalloy industry, it was determined
that emissions from production of high-silicon alloys would be the
more difficult to control due to the finer size of the particles and
significantly larger gas volumes from the furnace. Comparison of gas
volumes reported for production of silicon metal and 75 percent
ferrosilicon with those from chrome and manganese alloys production
shows that this conclusion is still valid (see Table II-4). Therefore,
the supplemental data do not indicate that reassessment of the use
of the alloy group is necessary. Jue to the lov/er aas volumes
29
-------
from furnaces producing chrome and manqanese alloys, it is still
reasonable to expect a lower mass emission rate for such furnaces.
While data on only one chrome alloy furnace were received in
response to the section 114 letters, data from several tightly hooded
open furnaces were available from EPA's study of the industry. Table II-4
also provides a comparison of the furnace gas volumes with that from
typical industry practice. Jata for all these facilities show that
substantial reductions in gas volumes relative to typical industry
practice values are possible by tiqhtly hooding the furnace. Facilities
L, 0, and U of the original study had gas volumes approximately
1 SCMS/MW (2000 SCFM/-MW) lower than those expected based on typical
hooding practices. Facilities N and N2 are small furnaces (7-9 MW
nominal capacity) wh-'ch are tightly hooded to allow recovery of the
waste heat from the furnaces. Because of stringent control of gas
volumes from the furnace, emissions from facility N2 were less than
0.045 kg/MW-hr (0.09 Ib/MW-hr) despite effluent concentrations of
0.02 g/dscm (0.009 gr/dscf). In EPA's judgment the reductions in
furnace off gas volumes demonstrated at facilities L, 0, and U are
sufficient to enable the source to readily comply with the standard.
EPA would like to emphasize that compliance with the 0.23 kg/MW-hr
(0.51 Ib/MW-hr) standard does not require use of specially designed
collection hoods, as at facilities N and N2. The standard is achievable
by use of efficient control devices and by well demonstrated means of
reducing furnace gas volumes. The water-cooled canopy hood can be
designed to minimize the volume of induced air by enclosing the open
area between the furnace and hood with refractory lined flaps, by use
30
-------
Table I1-4. Comparison of Furnace Gas
Volume By Alloy Being Produced
Product
Furnace
Size
Furnace
SCMS
Reported
Furnace
SCMS
Typical
SCMS/'IW
Silicon lletal (AA-CC)
Silicon Metal (EE)
75% FeSi (DD)
75% FeSia
Si. In (N)
(L)
HC FeCr(N2)
(U)
(T)
Feiln (EPA Study Values)
CaC2 (EPA Study Values)
15-20 MW
30
18
40a
7-9
25-27
7.3
15
33
30
30
71.7
129.77
64.18
123.64
7.55
54.27
4.25
27.84
73.62
101.46
58.04
4.22
4.33
3.57
3.09
0.94
2.17
0.58
1.86
2.23
8.69
ii
4.35
M
3.40
n
2.45
ii
n
3.40
1.98
Values reported by Commenter Z-9.
Gas volumes expected for typically hooded open furnaces as desinned
in 1972. Gas volumes were reported for 30 MW capacity furnaces in
Reference 3.
31
-------
of air curtains, or by lowering the hood to 1.5 meters (five feet)
above the furnace crucible. With only moderate control of the induced
air volume, the gas volume from a 30 MW furnace can be held to around
6,000 dscm (212,000 dscf). If emissions from this hypothetical 30 MW
furnace were controlled to a concentration of 0.0068 g/dscm (0.003 gr/dscf)
by an open fabric filter collector with dilution air cooling, the
mass rate of emissions would be 0.17 kg/MW-hr (0.36 Ib/MW-hr). This
mass rate is significantly below the standard of 0.23 kg/MW-hr (0.51 Ib/MW-hr)
for chrome and manganese alloys.
The control of gas volumes from open furnaces has beei well demon-
strated on chrome ano manganese alloy production, in addition to the
hotter operations of high-silicon alloy production, and effective
control systems are available and well demonstrated. Therefore, the
0.23 kg/MW-hr (0.51 Ib/MW-hr) standard for chrome and manganese alloys
is achievable at a reasonable cost and has been adequately demonstrated.
The standard was not revised upward to 0.45 kg/MW-hr (0.99 Ib/MW-hr)
as suggested by the commenter as that would allow design of control
systems without consideration of controlling the gas volumes.
6. Effect of quality of charge materials.
Industry representatives expressed the opinion that use of fine,
friable ores which will affect furnace operation and emission rates
will prevent new chrome and manganese furnaces from complying with the
0.23 kg/MW-hr (0.51 Ib/MW-hr) standard. In recent years only the
finer sized chrome and manganese ores have been available at reasonable
orices. Accordingly, new furnaces are expected to primarily use finer
ores as feed materials. In 1971-72 during EPA's emission test nronram
32
-------
the use of fine ores was less common and, hence, the commenters
argued that the standards do not consider the effect of fine ores
on emissions.
In order to more fully consider the commenters1 arguments on
fine ores, EPA requested that they provide data or supplemental
information on experiences with operation of furnaces using fine
ores to substantiate their position. At the time of the request the
commenters indicated that no data on emission rates as a function of
ore quality were available but agreed to provide information on
operating experience. The information they submitted did not show that
the use of fine ores should necessarily affect the emission rate of
the furnace and the achievability of the standard. In general, the
material provided by the industry representatives only discussed the
general problems associated with use of fine raw materials and did
not discuss any of the common solutions to these problems. EPA was
aware of these general problems of fine feed materials during the
development of the proposed standards.
Commonly, charge materials are sized and blended before addition
to a typical furnace. The charge materials, thus, generally include
a portion of fine materials. Ideally, the materials are sized and
blended to produce a porous charge which will promote uniform gas
distribution through the mix and furnace operating stability. Very
fine feed materials affect furnace operations due to the
reduced porosity of the charge which results in bridging and non-uniform
descent of charge materials. The fusion and crusting over of the
charge promotes channelization of the gas flow. Collapse of a bridged
33
-------
area results in jets of hot gases emanating from the reaction zone.
These gas blows increase the participate matter emissions from the
furnace and resulting unstable furnace operations. The problem of gas
blows from the furnace can be minimized by adequate stoking of the
charge to allow uniform descent and uniform evolution of the qas
formed in reduction of the ores. In EPA's opinion, the blowina
problem associated with use of fine chrome and manganese ores is
not as significant as the problems experienced in production of
silicon metal. In both cases, stoking of the furnaces at sufficiently
frequent intervals tc maintain a uniform evolution of reaction qases
and unbridged crust will reduce uncontrolled emissions from the
furnace. The effect of use of fine ores as charge material on the
furnace emission rate will vary with the frequency and adequacy of
stoking.
In addition to stoking the furnace, there are other methods of
using finer, or more friable, ores which do not result in unstable
furnace operation and increased emissions. Finer ores can be pelletized
or sintered to reduce the tendency to bridge, to increase charge porosity,
and to minimize entrapment losses. At present, sintering and pelle-
tizing of fine ores is not common practice, but it could become
an important pretreatment process with changing economic and supply
conditions. Another possible approach is to inject tne fine ores into
the furnace through hollow electrodes as is done on calcium carbide
furnaces.
34
-------
EPA did not specifically consider the effect of size of charge
materials on furnace emission rates during the emission study or
development of the standard because this effect should be insignificant
with a properly operated furnace. Such a furnace would be stoked at
sufficiently frequent intervals to minimize gas blows and maintain
stable operation. With such an operation the entrainment losses would
be minimized. Fabric filter collectors are relatively insensitive to
small fluctuations in inlet concentrations. Therefore, small increases
in the furnace uncontrolled emission rate will not affect the
achievability of the standard. The 0.45 kg/MW-hr (0.99 Ib/MW-hr)
standard for high-silicon alloys was not established because of a
larger uncontrolled emission rate due to blowing furnace conditions.
The difference between the standards for the high-silicon and the
chrome and manganese alloy groups primarily results from the larger
gas volumes per megawatt furnace capacity from high-silicon alloy
furnaces. Consequently, EPA does not agree that the use of finer
sized chrome or manganese ores will increase the emission rate of a
controlled furnace or that revision of the standard is required.
C. Summary and Conclusions
In response to comments on various aspects of the data bases for
the standards, EPA reassessed the rationale for the standards. From
this reassessment, EPA concluded that the standards of performance for
ferroalloy production facilities are achievable by application of best
systems of emission reduction. Therefore, the standards of performance
for chrome and manganese alloys and for high-silicon alloys are being
35
-------
promulgated at the proposed levels of 0.23 kg/MW-hr (0.51 lb/MW-hr)
and 0.45 kg/MW-hr (0.99 lb/MW-hr), respectively.
The revaluation of the data bases for the standards showed that
the emission test procedures used would not significantly bias the
results. Review of the assumptions and the modifications made to
the procedures of Method 5 to allow emission testing on fabric filter
collectors showed them to be reasonable considering the nature of the
effluent. Comparison of the induced air volume measured by vane
anemometers with volumes expected from energy balance considerations
showed that fluctuations in induced air volumes would produce only
minor changes in the total gas volume flow rate from the collector.
Therefore, the measured emission levels were considered representative
of those achievable by best systems of emission reduction. Emission
data from recently constructed well-controlled facilities show
comparable levels of emission reduction to those measured by EPA
and confirm the expected performance of the control systems.
The achievability of the standards was further demonstrated by
data received from the industry which showed that substantial reductions
in furnace gas volumes are being achieved by prope"1 design of the
collection hood. These data showed that gas volumes from well-hooded
large silicon metal furnaces can be reduced to 40 percent of the
volumes from typically hooded large silicon furnaces. With such good
control of gas volumes and use of well-designed fabric filter collectors,
emissions from these electric submerged arc furnaces were much less than
0.45 kg/MW-hr (0.99 Ib/MW-hr). In addition, data were available from
36
-------
EPA's earlier emission tests on several tightly hooded open furnaces
producing chrome and manganese alloys. These furnaces had significantly
smaller gas volumes than typically hooded open furnaces and, thus, lower
mass emission rates. Emission tests conducted by EPA show that by
control of gas volumes and by use of a well-designed fabric filter
collector or venturi scrubber, emissions can be reduced to below
0.23 kg/MW-hr (0.51 Ib/MW-hr).
The effect of raw material quality (e.g., size of ore) on furnace
emission rates was considered as part of the overall evaluation of the
achievability of the 0.23 kg/MW-hr (0.51 Ib/MW-hr) standard. EPA
concluded that with proper operating procedures, or other techniques,
emissions from use of fine ores are not greater and the standard is
achievable with best systems of emission reduction.
In general, EPA believes that the reassessment of the bases for
the standards has shown that the standards of performance for both
product groups are achievable. Information provided by the industry
for well-controlled furnaces shows that the type of control system
required by the standards is available at a reasonable cost and has
been adequately demonstrated.
37
-------
III. Opacity Standard
The proposed standards of performance limited the level of
emissions from the control device to less than 20 percent opacity
for both product groups. The proposed opacity standard was
based on data obtained in accordance with the requirements of
Reference Method 9 as promulgated on December 23, 1971 (36 FR 24876).
The method required that opacity be read in increments of five
percent and that compliance with the standards be determined on
the basis of single readings. Thus, as proposed any single observa-
tion of opacity greater than 15 percent would be a violation of
the opacity standard for an affected ferroalloy facility.
On November 12, 1974 (39 FR 39872), after proposal of the
standards for ferroalloy facilities, EPA revised Reference Method 9
and the general provisions applicable to standards of performance.
These revisions resulted from a study of the errors associated
with single observations made by qualified observers while
reading plumes according to the prescribed procedures. It
was shown in this study that qualified observers can determine
plume opacities with maximum positive errors of less than 7.5
percent based on the average of 24 consecutive readings. Accordingly,
Reference Method 9 was revised to require that opacity observations
be recorded to the nearest five percent at 15-second intervals
with a minimum of 24 observations. Reference Method 9 was revised
to extend the minimum time over which opacity observations are
made to six minutes to obtain sufficient observations to ensure
acceptable accuracy. The use of sets of opacity observations
38
-------
precludes a single high reading from being cited as a violation.
In addition, §60.11(e) was added to the general orovisions to
orovide a means for an owner or operator to petition EPA to obtain
a higher opacity standard for any facility that demonstrates
comoliance with the mass standard concurrent with failure to
attain the opacity standard. The provisions of §60.11(e) allow
opacity standards to be established at levels which consider the
maximum expected effects of the normal range of operating
variables at well-controlled new facilities.
Because of the revisions to Reference Method 9 and amendments
to the general provisions, the proposed opacity standard for ferro-
alloy facilities was reevaluated to determine the appropriate level
on the basis of six-minute average values and expected variations
in operating conditions at well-controlled new facilities. It
was concluded that the opacity standard should limit emissions
from the control device to less than 15 percent. This opacity
standard will ensure continued proper operation and maintenance
of the control device and continued compliance with the mass
standards. The bases for the recommended level of the standard
are presented below.
The proposed opacity standard was based on 37 hours of
opacity observations made in accordance with the original
Reference Method 9 at seven ferroalloy production facilities.
These data are presented in Volume 2 of "Background Information
for Standards of Performance: Electric Submerged Arc Furnaces
for Production of Ferroalloys," EPA-450/2-74-018bJ On the basis
39
-------
of these data, the concurrent emission test results, and inspection
of the observed control device, it was concluded that facilities
in compliance with the applicable mass standard would have emissions
of 15 percent opacity or less (based on single observation values).
Six-minute averages of the 37 hours of onacity data were
calculated to convert the data to a basis consistent with revised
Method 9. The six-minute average ooacity values of all the observations
are given in Anoendix B to this renort. The maximum six-minute average
ooacitv value observed at each facility is shown in Table III-4. The
maximum ooacity level observed at a well-maintained and cperatad
control device was line oercent, at facility 0. Visible emissions
in excess of this opacity level were observed at facility V, but the
collector was not considered well-maintained due to numerous leaking
bags in several comoartments. The variation between the ooacitv levels
associated with different effluent concentrations shown by the data
in Table III-4 is primarily due to the variations in path length and their
effect on ooacitv. (That is, the larger the nlume diameter for a qiven
effluent concentration, the more light will be attenuated as it passes
through the olume; and hence, the greater will be the onacity of the
olume.) For examnle, if facility O's emissions had been discharged
from several stacks of 0.40 meter diameter instead of from one
stack of 2.87 meters, the plume would have exhibited zero percent
opacity instead of nine percent opacity. These data show that
ferroalloy oroduction facilities which achieve the mass emission
standard should have emissions of less than 10 percent opacity.
40
-------
Table III-4
Facilitv
Code
0
0
R
S
II
V
H
Control Device
Closed nressure
baqhouse
Onen baqhouse
Venturi scrubber
Closed suction
baghouse
Closed oressure
baqhouse
Open baghouse
(ben baqhouse
I
Discharqe Point,
Diameter, meters
2.87
monitor discharge
0.387
0.387
1.12
monitor discharge
monitor discharqe
Emission Concentration
g/acm kg/mw-hr
0.024 0.46
-
0.022 0.005
0.059 0.016
0.00014 0.016
-
-
Maximum Six-Minute
Average Onacity Value
Observed
9a
0
0
0
oa
18b
8b
^Observer not certified.
"Baghouse not nronerlv maintained.
-------
Further, it is expected that facilities which use open fabric filter
collectors will have emissions of less than five percent opacity
because of the substantial dilution of the effluent which occurs
in the collector.
In addition to the Method 9 observations of emissions, an
evaluation was made of the effects of variations in operating
parameters on olume opacity. To ensure that the opacity standard
is established at a level which is not more restrictive than the
mass standard, Bouquer's Law was used to analyze the effects on
plume opacity of variations in effluent concentrations, particle
characteristics, and stack diameters. The values used for the
operating variables in the analysis consisted of measured values
and values that could potentially produce the maximum opacity within
the limitations of the mass emission standards and possible effluent
gas volumes.
The range of possible effluent concentrations and stack
diameters at well-controlled ferroalloy production facilities is
a function of the design of the emission collection system and the
alloy being produced. As typically designed in the U.S., open
electric submerged arc furnaces have a water cooled canopy hood
located about 1.5 to 2.5 meters (five to eight feet) above the furnace
rim. This large vertical opening between the furnace and hood
permits large quantities of air to be drawn into the emission
collection system. The volume of induced air can be significantly
reduced by design of the collection hood to minimize the open
area between the furnace and the hood.
&?
-------
For typically hooded open furnaces producing the various
alloys, effluent concentrations less than 0.023 g/dscm (0.016 g/acm)
[0.01 g/dscf or 0.007 gr/acf] are required for compliance with
either of the mass standards. Effluent concentrations for
tightly hooded open furnaces could possibly be 0.034 to 0.046 g/dscm
(0.015 to 0.02 gr/acf) if substantial dilution of emissions does
not occur in the control device. Although these concentrations
are theoretically allowed by the mass standards, the concentrations
are unlikely if emissions are controlled by a well-designed fabric
filter collector or venturi scrubber. To ensure consideration
of all expected variations in operating parameters for both
typically and tightly-hooded furnaces, concentrations of 0.011 to 0.046
g/acm were considered in the opacity calculations.
The opacity of a plume is affected by the path length through
which the light is attenuated. The dimensions of discharge points
or areas on control devices used at ferroalloy facilities varies
with the type control device and the gas volume being treated.
Ferroalloy production facilities in the U.S. generally use open fabric
filter collectors to control emissions from the electric submerged
arc furnaces. Generally, the qases from these collectors are
discharged to the atmosphere through a monitor or through stub
stacks. The shortest dimension of the monitor or stack is the
path length through which emissions are observed when using
Method 9. For typical roof monitor discharges, these path lengths are
43
-------
expected to vary from 2.0 to 4.3 meters. Some modular construction
collectors may have monitors of smaller dimensions than these. Closed
pressure or closed suction collectors or venturi scrubbers are less
commonly used by the ferroalloy industry, but they might be installed
in a few cases since they are easier to emission test. Stack diameters
for these control devices are related to the size of the facility and
the gas volume being treated. In EPA's study of the industry, the
largest observed stack diameter was 3.0 meters.
Should a facility use a larger than expected stack diameter, thus
causing the opacity of the facility's emissions to exceed the level of
the standard, the owner or operator may request that the EPA establish
a special opacity standard for that facility under the provisions of
section 60.11(e).
In addition to concentration and path length observed, the light
scattering and absorbing characteristics of the particulate
matter are important factors in the mass-opacity relationship.
Particulate matter emissions from ferroalloy electric submerged
arc furnaces consist mainly of condensed metal oxide fumes and
entrained furnace charge materials. Thus, all particles emitted
from the control device are expected to be less than 1.0 micrometer actual
diameter, to vary over a limited size range for each alloy, and
to be spherical. Particle size analysis of emissions from production
of high-silicon alloys showed that the mass mean diameter of
the particulate from the control device varied from 0.1 to 0.4
f
micrometer diameter for the various alloys. These emissions
44
-------
were found to be orimarily transparent, spherical, silicon dioxide
particles.3 Analysis of emissions from chrome and manaanese
alloy production showed that the mass mean diameters of the
particulate varied from 0.2 to 0.6 micrometer diameter for
various alloys, and the emissions consisted of opaque, spherical,
metallic oxide (black, brown) particles.
The light attenuated by a olume was calculated using Bouguer's
q
Law as exnressed by Ensor and Pilat for a Dolvdisoerse aerosol
(the distribution of particle sizes about the mean is considered).
So, variations in particle size of emissions are inherently
considered in the calculation procedure. Plume ooacities were
calculated for both tight!v-hooded and tynically-hooded ooen furnaces
producing either high-silicon alloys (white aerosol m = 1.5) or
chrome and manganese (black aerosol) alloys. For all probable cases,
the calculated nlume opacities are less than 15 percent.8 For
well-controlled, typically-hooded ooen furnaces, the maximum
opacity calculated is less than 10 percent for "high-silicon"
alloys and less than 13 percent for chrome and manganese alloys.
For well-controlled, tightly, hooded open furnaces (concentration s'0.023 g/dscm),
the maximum ooacity calculated-is'less than 10 oercent for both
product groups. Tightly hooded onen furnaces controlled only to the level
of the applicable mass standard are expected to have olume onacities
less than 12 percent for high-silicon alloys and less than 14 percent
for chrome and manganese alloys.
45
-------
On the basis of the projected opacity levels, EPA has concluded
that typical variations in effluent concentration, stack diameter,
and oarticulate matter characteristics do not justify establishing
the opacity standard at a level greater than T5 percent opacity.
The Method 9 opacity observations further indicate that establishing
the opacity standard at a level greater than 15 percent could allow
less effective control of emissions than is reguired by the mass
emission standards. The 15 percent opacity standard is achievable by
well-controlled, typically or tightly hooded open furnaces whose
emissions are in conpliance with the applicable mass standard.
Few, if any, lacilities are expected to require establishment
of special opacity standards under the provisions of §60.11(e).
An example of an atypical facility which would require a special
opacity standard would be any facility which, although in compliance with
the mass standard, has a higher than expected effluent concentration
and discharges emissions to the atmosphere through a larger
than expected stack or monitor. The provisions of §60.11(e) allow
the owner or operator of such an atypical facility to petition
the EPA Administrator for establishment of a special opacity
standard. Thus, no owner or operator of an atypical facility
will be prejudiced by this establishment of the opacity standard
on the basis of typical variations in operating conditions.
46
-------
IV. Alternative Test Procedures for Open Fabric Filter Collectors
The orovisions of §60.266(d) of the promulgated regulation
require the owner or operator to design and construct the control
device to allow measurement of emissions and flow rates using
applicable test methods and procedures. This section does not
permit the use of control devices (such as open pressurized
fabric filter collectors) from which emissions cannot be measured
by reference methods in Appendix A of 40 CFR Part 60, unless
compliance with the promulgated standard can be demonstrated
by an alternative method. EPA has not specified an emission
test procedure for open pressurized fabric filter collectors
because of the large variations in the design of these collectors.
Test procedures can be developed on a case-by-case basis, however.
Provisions in 40 CFR 60.8(b) allow the owner or operator upon
approval by the Administrator to use an "alternative" or
"equivalent" test procedure to show compliance with the standards.
EPA emphasizes that development of the "alternative" or "equivalent"
test procedure is the responsibility of any owner or operator
who elects to use a control device not amenable to testing by
Reference Method 5. From general comments on the standards, it
was apparent that considerable misunderstanding existed concerning
acceptable alternative procedures for demonstration of compliance.
The following discussion is provided to give general guidance in
the major areas of concern.
Pressurized fabric filter collectors are not amenable to representative
samolinq by Reference Method 5 because of, amonq other things, the
47
-------
large areas through which emissions are discharged to the atmosphere
and the low velocities of the effluent. The larqe areas make it
difficult to determine sampling ooints which adequately represent
the entire area. The velocities of gases discharged from these
large areas are very low, typically around 2 meters per second (400 fpm).
Velocities in this range cannot be measured by conventional
sampling equipment. Consequently, isokinetic sampling and
accurate determination of volumetric flow rates using conventional
equipment are not feasible.
Possible approaches to emission sampling in large ereas and low
velocity gas streams require consideration and evaluation of the
condition of the collector and the capabilities of the sampling
equipment. In general, if all compartments in a collector are in
equivalent condition and operating under comparable conditions,
then sampling of emissions from a randomly selected, representative
number of compartments is an acceptable alternative procedure.
Another possible alternative procedure is to subdivide the roof
monitor area into subareas which are then sampled. If all compartments
are not equivalent in condition or operation, then it is necessary
to either repair the collector, sample emissions from all compartments,
or bias the sample by emission testing the compartments in poorer
condition.
One alternative for sample collection when isokinetic sampling
of the effluent is infeasible is sampling at a predetermined average
velocity rate. This alternative, which minimizes many of the
problems of sampling in these systems, consists of determining
48
-------
the average velocity in the area, using devices for low-velocity
measurements, and emission sampling at this rate for the entire
sampling period. If the sampling equipment is not amenable to this
procedure, then nonisokinetic sampling of the effluent may be
considered because of the small particle size of the emissions.
Subisokinetic sampling of emissions is the recommended procedure
as it assures collection of any large particles escaping from
bags which are improperly positioned or leaking. (Subisokinetic
sampling, sampling at lower velocities than the gas stream
velocity, biases the sample toward collection of a greater
amount of particulate matter per unit volume, and hence a greater
concentration than is actually present.) In situations where
Subisokinetic sampling is not possible, superisokinetic sampling
using a large diameter nozzle and a large sample volume may be
permissible subject to approval by EPA. Approval of superisokinetic
sampling is, among other things, subject to verification that all
compartments are in good operating condition and to demonstration
by measurement or reasonable engineering judgment that particles
larger than five micrometers aerodynamic diameter are not present
in the gas stream.
Because of the induction of cooling air into open pressurized
fabric filter collectors and the problems associated with velocity
measurements, alternative procedures are required for determination
of the total gas volume emitted from the control device. When
the average velocities in the subareas are determined during the
emission test, the volumetric flow rate can be calculated from
the known discharge area. The volumetric flow rate can also be
49
-------
established by the measured inlet flow rate and the measured induced
air volumetric flow rate. An example of an alternative test method
acceptable for demonstration of compliance is one in which emissions
are sampled at a predetermined average velocity rate in representative
subareas of the collector, and the volumetric flow rate is calculated
from the average velocity and discharge area. In terms of
general concepts, the test procedure used at facility CC employed
this overall approach. That is, emissions were sampled at the pre-
determined velocity in the center of each compartment above the grate
over the rocker arm and were sampled using a large diarreter nozzle
and large volumes. With the exception of minor details of the proce-
dure, the general approach used at facility CC would be an acceptable
alternative method for demonstration of compliance.
These possible alternative approaches to emission testing of
pressurized fabric filter collectors are intended for general guidance
only. It is emphasized that an "alternative" procedure for demonstration
of compliance is dependent upon specific design features of the collector,
the condition of the collector, and the sampling equipment available.
Due to the costs of testing, the owner or operator should obtain EPA
approval for a specific test procedure or other means for determining
compliance before construction of a new source. Under the provisions
of S60.6, the owner or operator of a new facility may request review of
the acceptability of proposed plans for construction and testing of
control systems which are not amenable to sampling by Reference Method 5.
If an acceptable "alternative" test procedure is not developed by the
owner or operator, then total enclosure of the pressurized fabric filter
collector and testing by Method 5 is required.
50
-------
V. Economic and Inflationary Impact
As explained in Chapter II of this report, the mass emission
standards for both product groups were reevaluated in view of comments
received on the achievability of the standards and the representativeness
of the data bases. Data received in January 1976 from the industry
indicated that gas volumes from open furnaces producing silicon metal
(and other hiqh-silicon alloys) were substantially lower than those
typically observed during EPA's 1972 study of the industry. Consequently,
the control costs estimates for a large silicon metal furnace were
revised to reflect this information. Because no data indicating
substantial reductions in gas volumes from chrome and manganese alloy
furnaces were received, the control costs for these furnaces were not
reassessed. However, all the control cost estimates were updated to
1975 dollars from the previous basis of 1972 dollars.
The control cost estimates for a large silicon metal furnace were
adjusted to reflect a gas flow of 214 actual cubic meters per second,
aons, at 478°K (454,000 acfm at 400°F) from a 30 IW caoacitv furnacp.
The previous control cost estimate was based on a gas flow of 354 acms
at 478°K (750,000 acfm at 400°F) from a 25 MW caoacitv furnace. The
costs reported in Volume 1 of "Background Information for Standards 6f
Performance: Electric Submerged Arc Furnaces for Production of
Ferroalloys" were adjusted using a scale factor to account for gas
volume differences and inflation indices. The investment costs for a
fabric filter collector for the 30 MW furnace are estimated to be
$2,644,000 in 1^75 dollars. Annual operating costs were similarly
51
-------
adjusted and are $909,000 in 1975 dollars. The annual operating costs
per ton of alloy produced reflect the slightly greater productivity
of the larger furnace. The costs are summarized in Table V-l.
The estimates of control costs for furnaces producing 50% FeSi,
HC FeMn, SiMn, HC FeCr, and CaC2 were updated to 1975 dollars from the
original basis of 1972 dollars using inflation indices. These costs
are 45-50 percent dinner than the 1972 estimates due to a 31 percent
increase in general construction costs and substantial increases in the
cost of electrical power.
At the time of the original economic analysis, domestic ferroalloy
prices were frozen at levels one-half to one-third the selling prices
of imported ferroalloys. Since 1972, price controls on ferroalloys have
been removed, and quoted prices have increased 190 percent for 50
percent ferrosilicon to 250 percent for silicon metal. The updated
and revised annual operating cost estimates now represent a smaller
proportion of the product selling price. Consequently, the conclusion
of the original economic analysis that the control costs are reasonable
and can be passed on to consumers is not affected by the updating of the
control costs.
The economic impact of the standards of performance is minimal
when the costs of achieving a typical State process weight regulation
are considered. The same control system must be installed to meet
typical State regulations or the standards of performance, although
the standards of performance may require slightly more efficient capture
of tapping fumes than some State regulations. Based on the possible
52
-------
increase in costs due to tapping fume control, the five year costs
and the fifth year annualized costs of implementation of the standards
were estimated as $2.5 million and SO.84 million, respectively.
Therefore, the inflationary impact of the standards of performance on
the price of ferroalloys is insignificant. Further, the conclusion
of the economic analysis remains that the costs are reasonable and
should not bar entry to the market or expansion of facilities for
small businesses.
53
-------
Table V-I - SUMMARY OF CONTROL COST FOR STANDARDS OF PERFORMANCE
FOR FERROALLOY PRODUCTION FACILITIES*
AFFECTED FACILITY
Electric Submerged
Arc Furnaces
STANDARD OF
PERFORMANCE
0.45 kg/MW-hr (SI.
FeS1, CaS1, SMZ)
0.23 kg/MW-hr (HC
FeCr, charge chrome.
FeMn, S1Mn, CaC2,
FeCrSI, FeMnSI,
Silvery Iron)
For both groups:
< 15Z opacity from
control device
No visible emissions
escaping the tapping
station during 602
of the time when
tapping 1s taking
place.
< 1055 opacity from
dust handling equip-
ment.
ESTIMATED COST
BASIS FOR COST ANALYSIS
TYPICAL
SIZE
30 MM 502 FeSI
30 MW S1
30 MW HC FeMn
30 MW SIMn
30 MM HC FeCr
30 MW CaC2
33 MW HC
FeMn (sealed)
38 m S1Mn
(sealed)
30 MM CaC2
(sealed)
CONTROL
EQUIPMENT
Fabric f1lter.b
Fabric f1lter.b
Fabric F1lter.b
Fabric f11ter.b
Fabric filter.5
Fabric filter.6
Scrubber.
Scrubber.
Scrubber.
INVESTMENT
COST, $
3,013.000
2,644,000d
2,489,000
2,489,000
1.965.000
1,703.000
4,585 ,000C
4,585,000C
2,201,000C
ANNUAL
COST.$/YR
961 .000
909,000d
803.000
803.000
646,000
564.000
1,139,000
1.139,000
626,000
ANNUALIZED
COST,
S/TON
20.23
53.73
8.12
18.26
12.76
6.26
11.51
25.89
6.88
PRODUCT
PRICE.
$/TON
330
850
440
480
700
171
440
480
171
A costs except for the sealed furnace. Actual economic Impact will be minimal when State standards are considered since State and
ieral standards require the same types of control equipment.
Scrubbers could also be used. Capital costs are about the same except for 50* FeSI which 1s 50% higher. Annual costs are greater by
about 50%.
Includes adjustment for Increased cost of sealed furnace. ,.,,.. «,,.
"Costs based on gas flow of 214 acm at 480°K (454,000 acfm at 400°F) through fabric filter.
-------
VI. Summary of Public Comments and EPA Responses
Eighteen comment letters were received on the proposed standards,
nine of which came from the ferroalloy industry, five from State and
local air pollution control agencies, and four from Federal agencies.
The comments were carefully considered, and where determined by the
Administrator to be appropriate, changes were made to the proposed
regulation.
Table VI-1 lists the name and affiliation of each person who
commented and Table VI-2 presents a key to the code for affiliation
category. The summary of the comments and EPA's responses follows
Table VI-2.
55
-------
Ferroalloy Production Facilities Conrnent Summary
Table VI-1. Li si of Commentators
Comment No.
Z-l
Z-2
Z-3
Z-4
Z-5
Z-6
Z-7
Z-8
Z-9
Z-10
Z-ll
Z-12
Z-l 3
Z-15
Z-16
Z-l 7
Z-18
Commentator
J. W. Gallion
J. W. Cooper
J. L. Biggane
W. Simmons
R. T. Bailey
C. R. Allenbach
G. A. Watson
W. R. Meyer
L. C. Wintersteen
S. D. Dor-emus
A. L. Bayles
E. R. Zausner
W. C. Roundtree
R. L. Duprey
0. D. Jordan
R. D. Turner
E. L. Lantz
J. ,'laloney
Affiliation Code
Oklahoma State Deoartment of Health A
Alabama Air Pollution Control A
Commission
State of New York, Department of A
Environmental Conservation
State of California, Air Resources A
Board
Foote Mineral Company D
Union Carbide Corporation D
The Ferroalloys Association E
Commonwealth of Virginia, State A
Air Pollution Control Board
Airco Alloys D
U. S. Department of the Interior C
Allison L. Bayles & Associates F
Federal Energy Admir.i strati on C
General Counsel of the Department C
of Commerce
EPA - Division of Stationary B
Source Enforcement
Ohio Ferro-Alloys Corp. D
Chromium Mining and Smelting Corp. D
International Minerals and D
Chemical Corporation
Satralloy Inc. J
5fi
-------
Table VI-2. Code to Affiliation Category
Code Affiliation Category
A State or Local Air Pollution Control Agency
B U. S. Environmental Protection Agency
C Other Federal Government Agencies
D Ferroalloy Industry
E Professional Ferroalloy Industry Association
F Consulting Firms
57
-------
Summary of Comments and Responses to the Proposed
Standards of Performance for Ferroalloy Production
Facilities (39 FR 37470)
Preamble
1. The proposed standards appear to be reasonable and enforceable.
Z-3(A), Z-12(C)
Response: No response necessary.
2. Review of the nronosed regulations was hampered by the unavaila-
bility of the Background Document. Z-10(C), Z-13(C)
Response: EPA rerognizes that the unavailability of the documents
made meaningful review difficult. Consequently the end of the
comment period was extended from December 5, 1974 to December 30,
1974.
3. Agree with the decision not to require use of sealed furnaces
at this time. Z-10(C)
Response: No response necessarv.
4. The preamble alleges that sealed furnaces are inherentlv superior
from an air pollution control aspect. This statement, however, has
not been adequatelv demonstrated and ignores other sources of emis-
sions due to sealed furnaces. For example, most sealed furnaces
require special raw material preparation which may also contribute
to air pollution. This misimpression should be corrected. Z-ll(F)
Response: EPA does not agree with the commentator that the preamble
inaccuratelv discussed the air pollution aspects of sealed furnaces.
Sealed furnaces used in conjunction with venturi scrubbers or fabric
filters are inherently superior to an open furnace from both an air
pollution control aspect and an energy conservation asoect even when
thev use similar control equipment. Comparison of eouivalent open
and sealed furnaces reveals that due to the lower air volumes which
require treatment, the controlled emissions from a sealed furnace
are between 0.5 to 2 percent of the mass of particulate emitted bv
comparable open furnaces. Furthermore, the air pollution control
device for a sealed furnace requires less than 10 percent of t>e
power required for the control devices on open furnaces.
58
-------
The alleged additional pollution due to rav material preparation
would not necessarilv occur. Observations of sealed furnaces in
Norwav by EPA engineers showed that sealed furnaces do not inherently
require raw material preparation. For production of certain arade
ferroalloys, raw material preparation was practiced. Emissions
from sintering and pelletizing can be controlled bv a venturi scrubber
or fabric filter collector. If the controlled emissions from sinter-
ina and pelletizing are considered in the mass emission rate of the
sealed furnace, it is still superior to an onen furnace from air
pollution control aspect.
The decision to allow open furnaces was based on economic considera-
tions of reduced product flexibility and possible decreased inter-
corporate competition. It was the Administrator's judgment that
tnese factors would make the costs of the additional emission
reduction unreasonable. Accordingly the standards were established
at levels such that open furnaces could be used.
5. Agree with the Department of Commerce's position on opacitv stan-
dards. Z-10(C)
Response: EPA's position is unchanged from its consideration of
this issue in the "EPA Response to Remand Ordered bv U. S. Court
of Appeals for the District of Columbia in Portland Cement Assoc.
vs. Ruckelshaus (486 F. 2d 375, June 29, 1973)." See also the
response to comment 3 of section 60.252.
6. The cost of compliance for modified facilities should be considered
in determining the reasonableness of the control costs. Z-13(C)
Response: A source is considered "modified" under the Clean Air
Act only if the change results in increased emissions. If the
electric submerged arc furnace's or the entire shop's emissions
are controlled to the pre-modification level, the esa furnace would
net be considered modified and hence would not be subject to the
standards of performance. Therefore, few existing furnaces would
be subject to the standards of performance. The costs for upgrading
the emission control system of any furnace that is "modified" were
not developed because a realistic general assessment of the changes
required is not feasible. The cost of controlling a "modified"
esa furnace emissions will vary with each facility's control sys-
tem and the physical space available. The ease of design and
costs of hoods and other equipment for efficient capture of pollu-
tants will vary significant!v depending on the building housina
the furnace. The amount of upgrading of the control device reauired
will depend on the aualitv of the original system.
59
-------
7. The nroposed regulations reouire design of the control system
to meet the most stringent expected level of control. A variance
should be issued to permit short term oroduction of alloys which
exceed the aoplicable standard because the control svstem v/as
not designed for control of that alloy's emissions. Z-13(C)
Response: It is entirely consistent with the requirements of the
Act to establish standards at levels which require design of the
control system for the most stringent level of control expected.
According to the industry it is only feasible to produce members of
the same product family in the same furnace. Since members of the
same product family have comparable uncontrolled emission rates and
the same mass standard, the degree of control required will he
comparable for all members of the product family.
8. A significant amount of disposal of collected material by landfill
probably will not occur because of the possibility of use of re-
cycle techniques developed by the steel industry. Z-10(C)
Response: No response necessary.
9. The preamble incorrectly stated that the ferroalloy industry dis-
poses of collected dust by landfill. The preamble discussion should
be corrected (amended) because the industry has actively sought pro-
ductive uses for these wastes for a number of years. Z-7(E)
Response: Waste disposal oractices of the ferroalloy industry were
reassessed as a result of discussions with this commentator. Availa-
ble information still indicates that in the U. S., the ferroalloy
industry primarily disposes of solid wastes by landfill rather than
by recycle of these wastes. However, productive uses for these
wastes are being sought by the industry as evidenced by research
in this area. The preamble discussion was intended primarily to
describe prevailing practices, not future '"aste disoosal oractices,
and to orovide guidance for nroner disposal.
10. EPA has failed to discuss the reasons for the standards for ferro-
alloy being markedly different than the standards for electric arc
furnaces in the steel industry since the two orocesses are so
similar. Z-13(C)
60
-------
Response: F.PA disagrees that the two processes are so similar
that either similar standards of performance are required or
that a justification for the difference between the standards
is required. The two standards differ because of major differences
in the location of collection hoods over the furnaces, the control
systems employed, and s.'iop conditions.
The maximum separation between the collection hood and any elec-
tric submerged arc furnace used in the production of ferroalloys
would be 6-8 feet on an open furnace. Tightly hooded ferroalloy
furnaces do not have a significant separation between the furnace
and the hood. Steel electric arc furnaces can use a building
evacuation, combination direct shell evacuation-canopy hood, or
combination side draft hood-canopy hood control system to
capture emissions and control them to the level of the standards.
'.''ith all these control systems-capture of at least part of the
emissions must be done by a canopy hood located 30 to 40 feet
above the furnace. This separation between the furnace and the
hood is necessary to allow free movement of the crane which
charoes raw materials to the furnace. The difference in the
separation between the two tvpe furnaces and their respective col-
lection systems affects the abilitv of the hoods to effectively
capture emissions. The standards requiring effective capture of
furnace emissions, consequently, differ because of differences
in the location of the canopy hood for the two processes. In
electric arc furnace shops, the fumes from charging, tapping, and
other shop activities rise and accumulate in the upper areas of
the building, thus obscuring visibility. Because of the poor
visibility within the shop, the performance of the emission capture
system can only be evaluated at the point where emissions are
discharged to the atmosphere. Ferroalloy electric submerged arc
furnaces do not reouire a large free snace between the canopv hood
and the furnace. Visibility around the electic submerged arc fur-
nace is good. Conseauentlv, the performance of the collection
device on a ferroalloy furnace mav be evaluated at the collection
arsa rather than at the noint of discharge to the atmosphere.
For ferroalloy furnaces a mass standard is preferable to a con-
centration standard of performance because the latter would have
allowed use of larger than necessary exhaust volumes and would have
precluded use of sealed furnaces. Sealed ferroalloy furnaces can
have exhaust gas volumes as lov as 2 to 5 percent of that from an
open furnace of equivalent size. Mass emissions from a controlled
sealed furnace are much less than from a controlled open furnace,
even though the emissions from a controlled sealed furnace are
more concentrated. A concentration standard based on use of sealed
furnaces would allow poor control of onen furnaces. <\lso a concen-
tration standard for ferroallov furnaces would permit the designer
of the control system to neglect consideration of the volume of
gases exhausted.
61
-------
A mass standard of performance was considered for electric arc
furnaces in the steel industry, but was not selected due to
other considerations. These other factors include: (1) effective
capture of tapping and charqinq fumes requires exhausting large
gas volumes through the canopy hoods to maintain adequate
capture velocities and (2) restricting the gas volumes could
in some hot climate areas and with building evacuation control
systems result in a conflict with OSHA requirements. These
factors made a concentration standard the only practical
method of regulating the electric arc furnace emissions on a
consistent basis nationwide.
11. Particle size, shaoe, density, color etc. will var« from dav to
day as well as with the product. Conseouentlv a functional mass-
opacity relatiorship cannot be developed. Z-17(D)
Resoonse: The joint EPA - The Ferrballov Association study ana-
lyzed particulate matter samnles of various ferroalloys for aero-
dynamic particle size, particle density, elemental composition,
crystalline structure, and color. The conclusion drawn from these
data was that the range of variation was not so great as to pre-
clude establishment of an opacity standard which is indicative
of prooer operation and maintenance of the control device.
Section 60.261
The definition of "control svstem" in paragraph (1) should be
restricted to the gas cleaning device. Z-5(D)
Response: EPA disagrees with the comment. The ancillary eouip-
ment such as the hoods, ducts, fans, dampers, etc., which capture
or transport the emissions is part of the overall control system
(and its costs). However, confusion over the intent of the pro-
visions showed that separation of the definition into the terms
"control device" and "capture system" was advisable. The final
regulation limits emissions from the control device and limits
emissions escaping capture by the capture svstem.
Section 60.262
i. The wording of paragraph (a) implies that the mass emissions
standards are applicable to the "dust handling eguioment" as
well as the control device. The word "facilitv" should be changed
to "fjrnace" and this section should be rewritten. 7-14(B)
Response: This section nas been revised as suggested.
62
-------
2. The standard for ferrochrome silicon should be 0.45 kq/MH-hr
instead of 0.23 kg/'W-hr because of the considerable energy
savings (280 vs. 1800 KW) that can be realized. Z-P(D), Z-7(E),
Z-9(D), Z-18(n)
Response: The standard for FeCrSi is based on emission data
from sealed as well as ooen furnaces. Both furnaces tested met
the nronosed standard of 0.21 ka/MK'-hr (0.51 Ib/flW-hr). The
standard of performance for furnaces producinq ferrochrome sili-
con reflects the degree of emission reduction achievable by apoli-
cation of the tested systems(fabric filter collectors on onen
furnaces or venturi scrubbers on sealed furnaces). The standard
of performance was not intended to allow use of less efficient
control systems. The argument that increasing the standard to
0.45 kq/MW-hr (0.99 Ib/MW-hr) would result in a substantial
energy savings assumes that electrostatic precipitators rather
than baghouses would actually be installed. From previous dis-
cussions with members of the" ferroallo» indstury, EPA concluded
that due to operational problems use of orecipitators is unlikelv
regardless of the level of the standard. The actual result of
relaxation of the standard would be installation of less efficient
control systems and/or neglect of the air volumes beinq treated.
If reduction in energy consumption is of prime concern, then use
of closer fitting hoods or sealed furnaces should be considered.
5. The control device's opacity standard should be based on a func-
tional mass-opacity relationship. -Without such a relationship
the standard should only be used as a "triggering mechanism" to
indicate a possible violation of the concentration standard.
Z-13(C)
Response: EPA.believes that a complete functional mass-opacity
relationship need not be established in order to develop opacity
standards which require proper operation and maintenance of the
control system. In cases where both mass or concentration and
opacity standards are established, it is EPA's policy to not
require application of a more efficient control system than
necessary to achieve compliance with the mass or concentration
standard. These opacity standards are established at levels
based on consideration of conditions that would produce maximum
obscuration of light at typical well-controlled facilities.
Opacity levels associated with improper and maintenance of the
control device are also considered in the determination of the
level of the standard. Thus, as structured by EPA, the opacity
standards will be exceeded or.ly by facilities which fail to
properly operate and maintain the control equipment.
63
-------
Use of opacity limits only to create a "rebuttable oresumption
of a violation" would allow an owner or operator to routinelv
inadequately ooerate or maintain the pollution control eouipment
except during periods of emission testing. It takes two weeks
or longer to schedule a typical Method 5 performance test. If
only small repairs were required, e.g. fabric filter bag replace-
ment, remedial action could be delayed until shortlv before the
emission test is conducted. For control systems using scrubbers,
the energy input could be reduced (the pressure drop through the
system) when stack tests were not being conducted, which would
result in the release of significantly more particulate matter
than if the system were being prooerly operated. For these rea-
sons, owner or operators are required to oroperly maintain air
pollution control equipment at all times [40 CFR 60.7(d)] and
to meet opacity standards at all times except during periods of
startup, shutdown, and malfunction [40 CFR ">0.11(c)l and during
other periods of exemption specified in the applicable standards.
In development of the opacity standard, EPA determined that opacity
and particulate matter emission rates were quantitatively related
at ferroalloy production facilities. Data collected during the
standards development process showed that a typical well-controlled
open furnace witn an emission rate of 0.45 kg/MW-hr (0.99 Ib/MW-hr;
will have a plume opacity below 15 percent (based on six-minute
average value). Specifically, the opacity standard is based on
observation of emissions at seven ferroalloy facilities as well as
consideration of the maximum effect of the normal range of particle
characteristics and stack diameters on plume opacity. At four of
the seven facilities, the opacity observations were conducted
concurrently with the particulate matter emission testing. The
maximum six-minute average opacity value observed was nine percent
(read across a 10 foot diameter stack) and was associated with a
mass emission rate of 0.45 kg/MW-hr (0.99 Ib/MW-hr). Facilities
with lower emission rates or smaller stack diameters had no visible
emissions. The observed apparent plume opacities are consistent
with plume opacities predicted by Bouguer's Law for the same stack
diameters and effluent concentrations. Thus, the opacity of emissions
from ferroalloy facilities is sufficiently related to the mass
emissions to be a reliable mpasure of emissions. These data, and
projections of plume opacities, clearly indicate that a well-operated
and maintained control system on an open furnace that meets the
applicable mass emission standard will have visible emissions
substantially below the opacity standard of 15 percent. Opacity
levels in excess of the standards are indicative of improper
operation and maintenance of the control system and emissions in
excess of the applicable mass standard.
4. The same opacity standard is proposed for both categories of
mass emission rate standards. Separate opacity standards for
each mass emission rate standard would be more appropriate.
Z-13(C)
-------
Response: A single opacity standard for both mass emission rate
standards was recommended to simplify the regulation and its enforce-
ment. Separate opacity standards were not established because the
mass standards differ more due to differences in effluent gas volumes
for the product categories than due to differences in effluent
concentration. Since opacity is concentration dependent (among other
variables), these variations in gas volumes resulted in an overlap
of expected opacity values for the two groups of products. A single
opacity standard, therefore, was established on the basis of maximum
expected effluent concentration and maximum expected stack diameter.
This standard requires proper operation and maintenance of the
control device and is achievable by facilities subject to either of
the mass standards.
5. Fugitive emissions from the furnace coyer are a minor fraction
of total emissions and are an insianificant contribution to
degradation of ambient air nualitv. Paragraph (3)(i) is thus
unnecessary. Z-5(D), Z-ll(F)
Response: Emissions from the electric submerged-arc furnace
hood or cover and the tapping station hood can, if not properly
cantured, be a significant fraction of the total emissions due
to the furnace. (In some cases the amount of particulate matter
escaoing the collection hoods is greater than the emission rate
from the control device.) The intent of standards of performance
is to require application of best demonstrated control tech-
nology. Adequate capture of emissions at the point of generation
is an integral feature of the control level sought by the stan-
dards .
6. The fugitive emission standard for the furnace collection hood is not
technically feasible and the standard is based on insufficient
data. Therefore paragraph (3)(i) should be deleted. Z-9(D)
Response: The fugitive emission capture requirement for the
furnace is technically feasible based on observations at well-
controlled facilities. The standard is based on observations
et facilities observed in the survey for well-controlled facili-
ties and long term observation at four facilities. The canopy
hoods over four open furnaces were observed one hour each for a
total of four hours. No visible emissions were observed to
escape any of these furnace hoods. The covers of several
sealed furnaces were observed also for visible emissions
during the emission test program. Mb visible emissions were
observed to escape the cover of a sealed furnace with oroperly
maintained seals. Deletion of this requirement without addina
a standard limitina emissions exiting from the shop would
allow inadequate capture of the furnace fumes. Ooacity standards
on shop emissions are impractical and unnecessary for this source.
See comment 13 for further discussion on this noint.
-------
Paragraph (3)(i) requirement is not practical on a 100? of the
time basis. Violent reactions which are a function of the alloy
chemistry can cause puffing from the hood openings. ThG 1003
control reauirernent will result in increased consumption of
electricity. Z-5(D), Z-ll(F), Z-16(D)
Response: Observed facilities, which are discussed in the
above comment, showed that the furnace cover/hood standard
is technically and economically feasible. The energy consump-
tion from this requirement has been considered in the economic
analysis and determined to have an insignificant impact. In
addition, power consumed for capture and control of the furnace
emissions can be reduced by use of hoods that can be positioned
closer to the furnace than 6-8 feet or by use of sealed furnace-
See also the response to comment 5.
8 Paragranh (3){ii) should be modified to consider blowing taps
which are not a malfunction, failure of the process equipment,
or failure of the process to operate normally and thus not exempted
from this standard by the provisions of §60.8(c) (38 FR 28564).
EPA should observe a "blowing tap" before a standard on the
tapoing station is promulgated. Z-5(D), Z-7(E), Z-ll(F), Z-15(D)
Resoonse: Periods during which a "blowing tap" occurs have been
exempted from compliance with the tapping station emission cap-
ture requirements.
9. Paragraph (3)(ii) requirements do not recognize that hood capture
efficiency is affected by the following conditions: (1) a blow-
ing tap" which will usually persist throughout the entire tan
period, (2) poling or lancing of the taphole which causes exces-
sive emissions and may be necessary throughout the entire tapping
period, and (3) removal of metal and slag from the spout which may
require as much time as the tapping period. Emissions exiting
from the building as a direct result of the above conditions
will average an percent opacity. Paragraph (3)(ii) therefore
should be rewritten to allow emissions from the building during
tapping periods to average 40 percent opacity. Z-5(D), Z-7(E),
Z-9(D)
Response: According to available data on roof monitor emissions,
the suggested standard of *0 percent opacity on emissions from
the building would permit very poor capture of furnace and tap-
ping fumes. The existence of other emission sources and/or a
mixture of old and new furnaces within the building would make
enforcement of this standard difficult. Blowing taps are
exempted from the requirements of proposed §50.262(a)(3)(n)
[50.262(a)(5)] because this is a process malfunction condition
which is not wholly preventable. Periods in which the tapning
hood is swung aside for polinq/lancing or removal of metal or
66
-------
slag from the snout are failures of the process to operate in
a norm?! or usual manner. As malfunctions, these periods are
not subject to the standards. All poling or lancinf operations
which can be conducted with the taoping hood in place are sub-
ject to the capture requirements. Emissions from normal polings
or lancinq of the tap hole were considered in the develooment of
the standard.
in. Enforcement of the provisions of naraqraph (3)(ii) will be
difficult because the tapping stream is highly radiant thus
makinq observation difficult. In addition, shop conditions
can further complicate observation of the tap. Z-ll(F)
Response: The tanning station requirement is supported bv
observations of one tap hood during two tanning periods.
Enforcement of this ^revision will not be complicated by
shon conditions. EPA engineers have observed and evaluated
the collection efficiencies of tapping hoods at eight ferro-
alloy facilities in the U. S. and Japan. In none of these
facilities did shop conditions preclude evaluation of the
capture efficiency of the hood. Neither is the radiance
of the molten metal a factor in the assessment because the
observer is looking at the tap hood above the molten metal,
not the metal. Further evidence of the lack of a nrohlem
from the radiance of the metal is workers in the tanning
area do not use dark qlasses to improve vision or to nrotect
eyesight.
11. Paragraph (3)(iii) should be amended to allow greater than
10% opacity for neriods not in excess of one minute to ac-
count for unavoidable nuffs. Z-6(D), Z-7(E), Z-P(D)
Response: The standard recommended by the commentator is
not supported by the available onacitv data. Observation
of dust handling enuinment in the steel and asphalt indus-
tires found no visible emissions are associated with the
system. The 10 percent onacity standard on dust handlinq
equipment is sufficiently greater than observed values so
that a time exemption is not necessary.
12. The provisions of naragraphs (3)(i) and (ii) exceed EPA's
statutory authority, the fact that some emissions may es-
cane the building does not confer jurisdiction which other-
wise does not exist to requlate emissions within that build-
ing. EPA can exercise its regulatory authority over these
enissions when thev escape into the air. Z-S(D) , Z-7(E),
Z-ll(F), Z-13(C), Z-17(D)
67
-------
Response: It is the opinion of the Office of General Counsel
of EPA that the within the building regulations are within
the statutory authority of the Clean Air Act. Section 111
of the Act does not make direct discharge of the pollutant
into the atmosohere an express or implicit requirement for
regulating the emissions. The intent of the within the
building regulation is to ensure that the industry installs
efficient hoods to capture the voluminous emissions during
ore reduction and tapping operations of a ferroalloy furnace.
This only practical way of ensuring this is to restrict the
amount of visible emissions escaping capture by the hood.
13. The within the shop opacity standards should be replaced
with a zero equivalent opacity standard on the shoo roof
to facilitate delegation of enforcement authority to States.
Most State's air pollution control legislation limits en-
forcement of standards to the air outside of the working
area. Z-l(A)
Response: EPA disagrees that shop opacity standards would
be a better method of regulating these fugitive emissions.
Other emission sources and activities within the building
can cause visible emissions from the shop. An opacity
standard on the shoo roof is Impractical because of emissions
from non-regulated activities and sources within the shop
and intermingling of emissions from existing and new furnaces
in the same building. Since the vast majority of new furnaces
will be installed in buildings with other emission sources
and/or existing sources, a shop opacity standard could seldom
be enforced.
The Office of General Counsel of EPA believes that enforce-
ment authority for the within the shop standards can be dele-
gated to states. States delegated enforcement authority for
standards of performance would be operating under the
authority of the Clean Air Act, not under the authority of
the respective state act. Therefore, the state air pollution
control agency can be delegated enforcement authority without
modification of existing legislation.
14. The proposed shop standards are impractical, unnecessarily
complicated, and not subject to any standard of objective
measurement. It cannot be argued that the emissions would
be invisible when exiting the building since in that case
the emissions could not be considered to be a significant
source of pollution. Z-13(C)
68
-------
Resoonse: The intent of the within the buildinq regulation
is co ensure'adequate capture of all enissions Generated by the
operation of the ferroalloy furnace. The direct method of
accomplishing this intent is to control emissions at the
point of generation where it is easier to control rather
than after substantial dilution has occurred. It is more
difficult and expensive to collect these emissions after they
escape from the furnace area. Emissions from shops contain-
ing poorly or uncontrolled ferroalloy furnaces and tapping
stations are highly visible. These emissions may vary from
0 to 100 percent opacity depending on the production processes
occurring at the time and the capture efficiency of the hoods.
EPA believes that these standards are no more complicated or
impractical than any other standards of performance. Assess-
ment of compliance with the no visible emissions escaping the
control device requirements is easily accomplished by an obser-
ver with a stopwatch. Since the threshold of visibility varies
within a narrow range, different observers can objectively
assess the performance of the capture hood. The promulgated
regulation also includes requirements for monitoring of flow
rates through the hood exhaust ducts. Proper operation and
maintenance of the hoods will be determined from the flow
rate data. The monitoring requirements will make the hood
capture requirements more enforceable on a day-to day basis.
15. The inside the shop opacity standards make use of a buildinq
evacuation type control system impractical. This approach
for controlling fugitive emissions should not be eliminated
accidentally by p standard designed to require control of
fugitive emissions by an alternative aoproach. Z-15(D)
Response: Neither the proposed nor the promulgated regula-
tions precluded use of a building evacuation control system.
Tne regulations define "control system" such that for a build-
ing evacuation system the entire building is considered to be
the control system. The no visible emissions restrictions,
therefore, would be applicable to the entire buildinq. EPA
believes that use of building evacuation systems will be
very infrequent because of the large air volumes that must
be exhausted to achieve effective control, and acceptable
'"orkina conditions.
69
-------
16. The within the shop visiblity standards are based on an
extremely subjective judgnent which depends on shop con-
ditions and the observer'. Z-5(D), Z-ll(F), Z-13(C)
Response: Visibility conditions in well-controlled ferro-
alloy production facilities are very similar and are very good.
Thus, particulate matter escaping capture by the hood is
readily observable in all cases. Variations among shop condi-
tions will not significantly affect the observers' ability to
assess the performance of the capture hoods. The variation
between the ability of unbiased observers to detect visible
emissions should be minimal.
17. The within the shoo opacity standards will require modifi-
cation of Method 9 to allow taking of such observations.
Z-4(A)
Resposse: The within the shop no visible emission require-
ments do not require modification of Method 9 to allow such
a determination. Opacity levels of emissions are not being
assessed within the building, therefore, the criteria of
Method 9 are not applicable to this determination.
18. What type of instruments are being excluded by the phrase
"...shall not be visible without the aid of instruments"?
Z-8(A)
Response: The intent of this phrase is to limit assessment
of compliance to the unaided aye of the enforcement official,
except use of eyeglasses is permissible. Use of instruments
of any kind for detecting emissions escaping capture is pre-
cluded. An example of excluded instruments is portable
transmissometers.
19. The provisions of 60.262(a)(3)(i) will greatly increase capital
and operating costs of the collection system. To ensure cap-
ture of all emissions 100? of the time would require increasing
the air volume by 200% relative to existing volumes used. The
pollution associated with the increased electrical power con-
sumption should be considered in determining the overall
environmental imoact of the furnace cover standard. Z-16(D).
70
-------
Response: EPA's analysis of the power consumed in achieving
compliance with the capture requirement standards does not
agree with the commentators conclusions. In their joint
study (Engineering and Cost Study of the Ferroalloy Industry,
EPA-450/2-74-008), EPA and the Ferroalloy Association agreed
on the volumes necessary for effective capture of emissions
from the model furnace. These exhaust volumes are greater
than those measured at well-controlled facilities. The
greatest power consumption required by these conservatively
high air volumes is 10« of furnace power for a scrubber con-
trolled open furnace. All other cases were for significantly
lower power consumption rates. Actual power consumed due to
the standard is expected to be less than 5 percent of the
furnace power and overall the standard results in a beneficial
environmental impact.
20. The 0.99 Ib/MW-hr standard does not allow for the normal
deterioration in the fabric filter's performance. The
standard should be 1.5 Ib/MW-hr. Z-17(D)
Response: The suggested revision of the standard is unneces-
sary because the emission tests were conducted on well-sea-
soned fabric filter collectors. Some of these collectors
had been inroperation for several years. Since these facili-
ties can comply with the standard, any other well-designed
and maintained fabric filter collector can comply with the
standard.
21. Brief periods of emissions >10 percent opacity should be
allowed during transfer of dust from the fabric filter to
the dust disposal system. Z-17(D)
Response: The standard limiting emissions from dust handling
equipment is based on opacity observations of these operations
at steel plants and asphalt concrete plants. No visible emis-
sions (zero percent opacity) were observed at any time. The
need for the requested upward revisions is removed further
by the revision of Method 9 to base compliance on 6 minute
averages of the observations. Thus, the promulgated standard
provides sufficient accommodation for an occasional puff of
escaping the transfer system.
71
-------
22. Visible emissions cannot be limited to 40 percent of the
tapping period. A more realistic .approach to control of
these emissions would be to regulate emissions from the
roof or to require capture after the emissions have left
the tapping station. Z-17(D)
Response: EPA observations of emissions escaping capture
by the tapping hood at facilities in the U.S and abroad
indicate that'the standard is achievable. EPA does not
agree that an opacity standard limiting emissions from the
shop is more realistic. Standards on shop emissions, suffer
from the major disadvantage of intermingling of emissions
with emissions from nonregulated sources. Thus, compliance
with the standard cannot be determined. Regulating emis-
sions before fiey intermingle makes evaluation of only the
source of interest possible.
23. Furnace and tapping fume hoods are essentially constant rate
devides. Fume generation from furnaces and tapping fluctuate
considerably. Consequently, capture efficiencies will vary
considerably for a well designed hood. Z-18(D)
Response: The standards of performance are based on obser-
vations at well-controlled facilities and normal yariations
in capture efficiencies are already incorporated in the
standards.
Section 6n.263
1. The allowed 20 percent CO by volume represents an explosion
hazard and should be reduced. Z-2(A)
Response: The occurrence of gas streams with CO concentrations
of 10-20% (vol.) is unlikely because CO emissions from open
furnaces are very dilute (<1%) and closed furnaces are very
concentrated (>80%). The intent of this reauirement is to
force flaring of any gas stream which will support combustion
(nominally 12% CO but is a function of the composition of
the dilution gases).
72
-------
Section 50.264
1. Continuous monitors are an unnecessary and costly exoense.
Observation by the human eye should he sufficient and the
monitoring requirement should be deleted. Z-5(D), Z-6(D),
Z-7(E), Z-9(D), Z-11CF), Z-17CD), Z-18(D)
Response: EPA does not aqree that continuous opacity monitors
are an unnecessary expense. The monitoring requirements are
an aid in ensuring that sources which have demonstrated-com-
pliance with the applicable standards are properly operated
and maintained so as to remain in compliance. The costs of
ooacity monitors for ferroalloy facilities have been con-
sidered and the economic impact has been determined to be
insignificant. Observation by the human eye alone is not
considered sufficient because surveillance can only be con-
ducted on an infrequent basis.
2. The purpose of the continuous monitoring requirements is not
clear. The use of monitors will not decrease the time the
malfunction persists because of the existing intensive
maintenance procedures of the industry. Z-ll(F)
Response: The purpose of the opacity monitoring requirement
is to show whether or not the owner or operator properly
operates and maintains the control device. The owner or
operator may utilize the provisions of §60.13{i) to petition
the Administrator to allow alternative methods of demonstrating
that proper operation and maintenance of the control device
is conducted on a routine basis. See also the discussion in
comment 1 of this section.
Meaningful monitoring of baghouse emissions would require
use of one monitor ner each stub stack or compartment and
such extensive monitoring would be very costly. Observation
by the human eye should be the monitoring technique employed.
Z-15(D)
Response: The baghouse can be designed to reduce the number
of opacity monitors requir°d for meaningful monitoring of
emissions. In addition, §G0.13(g) allows the owner or
operator to petition the Administrator to permit use of
fewer monitors than the number of separate effluent gas
streams, if it can be shown that emissions can be meaning-
fully monitored by the suggested nrocedure. Observation
by the human eye alone is not considered sufficient because
73
-------
surveillance can only be conducted on an infrequent basis.
4. Available opacity monitors are not applicable to many types
of control devices such as open baghouses or control devices
which have gas streams with condensed water vapor. Z-ll(F)
Resoonse: EPA disagrees that an exemotion from the opacity
monitoring requirements is necessary for open baghouses.
The opacity of emissions from pressurized (open) baghouses
can be monitored using present instrumentation at a reasonable
cost using a number of approaches. These approaches include
monitoring of several compartments by a conventional path-
length transmissometers, or other approaches. In adJition to
monitoring schemes based on conventional pathlength trans-
missometers, a long path transmissometer could be used on a
pressurized baghouse. Transmissometers capable of monitoring
distances up to 15H meters are commerciailly available and
have been demonstrated to accurately monitor opacity. The
use of long path transmissometers on pressurized baqhouses
has yet to be demonstrated, but if properly installed there
is no reason to- believe that the transmissometer will not
accurately and representatively monitor emissions. The
best location for a long path transmissometer on a pressurized
baghouse will depend on the specific desian features of both;
therefore, the location and monitoring procedure must be
established on a case-by-case basis and is subject to the
Administrator's approval.
Control systems which have condensed water vapor in the
effluent gas stream may be exempted from the opacity
monitoring requirements by §60.13(i)(l).
5. Maintenance and operating costs of the opacity monitors are
excessive, especially for small firms. Z-ll(F)
Response: The costs for continuous monitoring of or-acity
were considered in the economic analysis and the economic
impact was determined to be insignificant. The caoital
and operating costs for the ppacity monitor are less than
one percent of the total capital and operating control costs
for the model furnace installation. This requirement is
not discriminatory against small firms because the monitor-
ing costs are insignificant for both large and small ferro-
alloy firms.
74
-------
6. The time period for reporting of emissions should be
increased. One minute is too short a time period in
which to take corrective measures. Z-6(D), Z-7(E),
1-17(0}
Response: The reporting requirements have been changed
to require reporting of all 6-minute periods wbere the
average onacity exceeds 15 percent.
Section 60.265(b)
1. This paragraph is unclear as to whether the power is measured
on the input or output side of the transformer. If the
power loss is considered in the nower input there will be a
slight decrease in the allowable emissions. Z-2(A)
Response: These provisions have been clarified to specify
that the nower is measured as kilowatts and mav be measured
either at the input or output side of the transformer. Since
the emission rate is partially determined by the power applied
to the electrodes, it is this power rate that should determine
the allowable emissions from the facility.
Section 60.266fe)
1. Cs should be expressed in terms of Kg/dscm, not Kg/dcm.
Z=2(A). Z-S(D)
Response: Agreed. The typographical error has been corrected.
Section 60.266(f)
1. The dimensions for E do not work out to Kg/MW-hr. Shouldn't
tn be expressed in terms of Kg/hr? Z-2(A), Z-6(D), Z-14(B)
Response: Agreed. This typographical error has been corrected,
75
-------
REFERENCES
1. Background Information for Standards of Performance: Electric
Submerged Arc Furnaces for Production of Ferroalloys.
Environmental Protection Agency, Research Triangle Park, N. C.
EPA-450/2-74-018a and b. October 1974.
2. Payton, R. N. Innovations in Ferroalloy Baghouse System Design.
JAPCA. 26;18-22. January 1976.
3. Dealy, J. 0., and A. M. Killin. Engineering and Cost Study of
the Ferroalloy Industry. Environmental Protection Agency,
Research Triangle Park, N. C. EPA-450/2-74-008. May 1974.
4. Silverman, L., and R. A. Davidson. Electric Furnace F^rrosilicon
Fume Collection (Pilot Plant Study). Journal of Metals, p. 1327-
1335. December 1955.
5. Muhlrad, W. Probleme des Fumes Emises para les Fours
Flectronetallurnioues. (The "roblem o^ the Smoke Flitted by
Electrometallurgical Furnaces). Chaleur et Industrie (Heat
and Industry). 422_:237-255. 1960.
6. Mercer, T. T. Aerosol Technology in Hazard Evaluation. New
York, Academic Press, 1973. p. 363.
7. Bloomfield, B. D. Source Testing, In: Air Pollution, Volume
II Analysis, Monitoring, and Surveying, Stern, A. C. (ed.).
New York, Academic Press, 1968. p. 504-508.
8. Kelly, V. E., Memorandum to Jack R. Farmer, Subject: Analysis
o* Test Vo. 7, Union Carbide Corporation, Alloy, 'test Virginia,
January 22, 1976 .
9. Meyer, J. S. Memorandum to Gene W. Smith, Subject: Reevaluation
of Opacity Standards of Performance for Ferroalloy Production
Facilities, February 18, 1976.
10. Ensor, D. S., and M. J. Pilat. Calculation of Smoke Plume Opacity
from Particulate Air Pollutant Properties. JAPCA. 21:496-501
August 1971.
76
-------
Appendix A
Summary of Emission Data Received
In Response to Section 114 Letters
-------
This appendix summarizes the emission test results received
fron the ferroalloy industry in response to letters, sent under
the authority of section 114 of the Clean Air Act, requesting data
on we11-controlled facilities. Data from five facilities were
received in this survey. A brief description of the facilities
tested and the emission test procedure is also provided.
Description of Facilities
N2. Open, 7-9 MII capacity electric submerged arc high carbon
ferrochrome furnace. Furnace load durinq tests was 7.3 C1H. The
air pollution control system consists of a furnace hood (close fitting)
and an Aronetics* two phase jet scrubber. Scrubbing water was
supplied at 172 liters/rnin. Tested hy EPA using Method 5.
AA. Ooen, 19-21 M1-' capacity electric submerged arc silicon
metal furnace. The air pollution control system consists of a
furnace hood, a separate tapping hood, a set of hairpin radiative
gas coolers, and an open pressure baghouse with a monitor discharge.
Emissions from the furnace and tapping hood are controlled by the
baghouse. Tested by company using a Method 5 tyoe train and an
"out-of-stack" Gelman* filter, and sampled at superisokinetic rate. At time
of test hanhouse operatinq at lower than desinn air-to-cloth ratio
because only one of two furnaces operatinq.
PR. Hpen, 21 *W capacity electric suhmeraed arc silicon metal
furnace. The air pollution control system consists of a furnace
* Mention of trade names or commercial products does not
constitute endorsement or recommendation for use by the Agency.
-------
hood, a tapping hood, and an open pressure baahouse with a monitor
discharge. Emissions from both the furnace and tapping hood are
controlled by the baghouse. Tested by the company using a Method 5
type train and an "out-of-stack" Gelman* filter, and sampled at
a superisokinetic rate.
CC. Open, 19 MM capacity electric submerged arc silicon metal
furnace. Furnace load durinn the tests was 15 HW. The air pollution
control system consists of a furnace hood, a separate tapping hood,
and an open pressure modular baghouse with individual monitors
discharge. Furnace and tapping emissions are controlled by the
baghouse. Tested by company using a high-volume modified Method 5
type train with four-inch sampling nozzles. Samples were collected
at the center of each compartment at an isokinetic samplina rate
for the sampling site. During the emission test the open grates at
the cell plate were blocked off (at least partially) so no diultion
air would enter the baghouse during the test.
PD. Open, 18 MW electric submerged arc 75 percent ferrosilicon
furnace. The air pollution control system consists of a furnace
hood, a tapping hood, and an open pressure baghouse with a monitor
discharge. Emissions from both the furnace and tapping hoods are
controlled by the baghouse. Tested hy baghouse manufacturer usinq
high volume samelers in three of nine compartments. For the period
of the emission testing the or>en orates at the cell plate were blocked
off so no dilution air would enter the compartments during the test.
* Mention of trade names or commercial products does not
constitute endorsement or recommendation for use bv the Agency.
A-2
-------
EE. Open, 25-30 MW electric submerged arc silicon metal
furnace. Furnace load during the tests was 27.4 MW. The air pollution
control system consists of a furnace hood with refractory lined doors
(or flaps), a separate tapping hood, and an open pressure baghouse with
a roof monitor discharge. Emissions from both the furnace and tapping
hoods are controlled by the baghouse. Inlet gas volume measurements
conducted by company to verify the collection system were within range
of design parameters.
A-3
-------
Table 1
Facility N2
Chromium Mining and Smelting Corporation
Furnace No. 23
Run No.
Date
Test time, min.
Average Power Input, MW
Stack Effluent
Flowrate « DSCFM
Temperature, °F
Water Vapor » Vol. %
Participate Emissions
Probe and Filter Catch
Cone, gr/dscf
Emission rate, Ib/hr
kg/hr
Ib/MW-hr
kg/MW-hr
1
6/26/74
360
7.30
10,911
141
22.87
0.01
0.94
0.43
0.13
0.06
2
6/27/74
360
7.38
7260
147
20.42
0.007
0.44
0.31
0.06
0.03
Averaq
360
7.34
9086
144
21.65
0.0085
0,66
0.37
0.09
0.045
A-4
-------
Table 2
Facility AA
Inter!ake, Inc.
Beverly, Ohio
Date
Average Power Input, MW
Average
7/17/72
Unknown
Stack Effluent
Flowrate, DSCFM
Temperature, °F
Water Vapor, Vol. %
Particulate Emissions
Cone., gr/dscf
Emission rate, Ib/hr
kg/hr
192,020
217
0.0009
3.7
1.68
A-5
-------
Run No.
Date
Test time, min.
Average Power Input, MW
Stack Effluent
Flowrate , DSCFM
Temperature, °F
Water Vapor, Vol. %
C02, Vol. %
02, Vol. %
CO, Vol. %
Particulate Emissions
Probe and Filter Catch
Cone. , gr/dscf
Emission rate, lb/hra
kg/hr
Ib/MW-hr
kg/MW-hr
Table 3
Facility BB
Interlake, Inc.
Selma, Alabana
1 2
10/29/75 10/30/75
180 240
21 21
124,866
240 215
3.0 3.0
0.1 0.2
20.1 20.2
0.2 0.2
0.00049 0.00052
0.68 0.68
0.31 0.31
0.032 0.032
0.015 0.015
Average
210
21
228
3.0
0.2
20.2
0.2
0.00050
0.68
0.31
0.032
0.015
a Calculated assuming no dilution air added by dampers or around bags.
A-6
-------
Table 4
Facility CC
Kawecki Berylco Industries, Inc.
National Metallurgical Division
Springfield, Oregon
Run No.
Date
Test time, min.
Average Power Input, MW
Stack Effluent
Flowrate, DSCFM
Temperature, °F
Water Vapor, Vol. %
Parti cul ate Emissions
Probe and Filter Catch
Cone. , gr/dscf
Emission rate, Ib/hr
kg/hr
Ib/MW-hr
kg/MW-hr
1
10/1/75
83
15
134,300
220
-
0.0024
2.76
1.25
0.184
0.083
2
10/1/75
83
15
120,800
229
-
0.00214
2.22
1.01
0.148
0.067
3
10/1/75
83
15
141,000
233
-
0.00189
2.28
1.03
0.152
0.069
Average
83
15
132,033
227
-
0.00214
2.42
1.10
0.149
0.073
A-7
-------
Table 5
Facility DD
International Minerals & Chemical Corporation
Tennessee Alloy Company
Kimball, Tennessee
Run No.
Date
Test time, min.
Average Power. I'inut, MW
1
4/9/74
60
18
2
4/9/74
60
18
3
4/10/74
60
18
Averai
60
18
Stack Effluent
Flowrate, DSCFM
Temperature, °F
Water Vapor, Vol. %
C0?t Vol. %
02, Vol. %
CO, Vol. %
Particular Emissions
Probe and Filter Catch
Cone., gr/dscfa
Emission rate, Ib/hr
kg/hr
Ib/MW-hr
kg/MW-hr
138,200 138,400
>200 >200
1.7
1.3
131,400 136,000
>200 >200
1.8
1.6
0.00029
0.34
0.154
0.019
0.009
0.00025
0.29
0.132
O.C16
0.007
0.00021
0.23
0.104
0.013
0.006
0.00025
0.29
0.13L
0.016
0.007
a Values are not correct but report contained insufficient data to
correct values.
A-8
-------
Table 6
Facility EE
Union Carbide Corporation
Metals Division
Alloy, West Virginia
Average
Date
Test time, min.
Average Power Input, MW 27.4
Stack Effluent
Furnace, DSCFM 275,000
Taphole, DSCFM 40,000
a o
Temperature , F 350
a Inlet to cooler
A-9
-------
Appendix B
Six Minute Average Opacity
Values
-------
Facility 0
Visible Emissions: Six minute opacity averages*
Plant: Ferrosilicon open electric submerged arc furnace
Company Name: Kureka Seitetsu Company Ltd.
Plant Location: Toyama, Japan
Date: 10/11/73 Time: 2:15 P-m.-4:02 p.m.
Type of Discharge: Stack
Type of Control: Two parallel settling chambers in series with baghouse
Stack Diameter: Stack #1 2.87 meters C9.4 feet)
Stack #2 2.90 meters (9.5 feet)
Stack #3 2.99 meters (9.8 feet)
Calculated from reading at one minute intervals
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
11
Stack #1
(% opacity)
5.8
9.2
5.8
5.0
6.7
6.7
3.3
2.5
5.0
5.0
5.8
Stack #2
(% opacity)
5.0
6.7
5.0
4.2
5.8
5.0
4.2
3.3
5.0
5.0
5.0
Stack #3
(% opacity)
5.0
6.7
5.0
5.0
5.8
6.7
3.3
3.3
5.0
5.0
5.0
-------
10/11/73
2:15p.m.-4:02 p.m.
Period
(6 minute
average)
12
13
14
15
16
17
18
Stack #1
(% opacity)
6.7
5.0
5.0
5.0
3.3
4.2
5.0
Stack #2
(% opacity)
6.7
5.0
5.0
5.0
3.3
5.0
5.0
Stack #3
(% opacity)
7.5
5.0
5.0
5.0
5.0
5.0
5.0
B-2
-------
Facility Q
Visible Emissions: Six minute opacity averages(*)
Plant No. 9: Ferrochrome silicon open electric submerged arc furnace
Company Name: Airco Alloys and Carbide
Plant Address: Niagara Falls, New York
Date: 2/20/74
Types of Discharges: Baghouse monitor, building roof monitor
Type of Control: Baghouse
Calculated from readings at thirty second intervals
B-3
-------
Baghouse Monitor
2/20/74
Time: 9:35 a.m. - 1.35 o.m.
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Observer A
(percent opacity)
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Observer B
(percent opacity) Comments
0.0
0.0
0.0 light snow
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
B-4
-------
Baghouse Monitor
2/20/74
Time: 9:35 a.m, - 1:35 o.m.
Period
(6 minute
average)
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Observer A
(percent opacity)
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
0.0
0.0
0.0
Observer B
(percent opacity) Comments
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
0.0
0.0
0.0
R-5
-------
Facility R
Visible Emissions: Six minute averages
Plant: Ferrochrome-silicon submerged electric arc furnace
Company Name: Nippon Kokan K. K. Toyama Works
Plant Location: Toyama, Japan
Date: 10/3/73 - 10/6/73
Type of Discharge: Flare stack
Type of Control: Two series venturi scrubbers followed by a demister
Stack Diameter: 0.39 meters (1.27 feet)
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
10/3/73
1:30-2:30
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/4/73
10:30-11:30
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/5/73
11:30-12:30
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/6/73
10:15-11:15
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
B-6
-------
Facility S
Visible Emissions: Six minute opacity averages
Plant: Ferrochrome sealed electric submerged-arc furnace
Company Name: Showa Denko K. K. (s)
Plant Address: Toyama, Japan
Date: 10/8/73 and 10/9/73
Type of Discharge: Flare Stack
Type of Controls: Baghouse on furnace
Stack Diameter: 0.39 meters (1.27 feet)
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
10/8/73
12:40-12:52
(% opacity)
0.0
0.0
rain 81 min.
0.0
0.0
0.0
10/9/73
9:00-10:00
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/9/73
10:00-11:00
-(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/9/73
11:00-12:00
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/9/73
12:00-12:30
-(% opacity)
0.0
0.0
0.0
0.0
0.0
B-7
-------
Facility U
Visible Emissions: Six minute opacity averages
Plant: High carbon ferrochrome tightly hooded electric submerged
arc furnace
Company: Awamura Metal Industry Company (u)
Plant Address: Uji City, Japan
Date: 10/10/73 and 10/11/73
Type of Discharge: Baghouse stack
Type of Control: Two parallel banks at cyclones followed by a bag filter
Stack Dimensions: 1.78 x 1.12 meters (5.85 x 3.66 feet)
B-8
-------
Baghouse Stack
10/10/73
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
10/10/73
7:00-9:00
(% opacity)
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/10/73
10:30-12:30
(% opacity)
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/10/73
12:30-2:30
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/10/73
2:30-3:30
(% opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
B-9
-------
Baqhouse Stack
10/11/73
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
10/11/73
8:00-9:45
(% opacity)
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
0.0
0.0
0.0
0.0
0.0
-
—
10/11/73
10:45-12:45
(% opacity)
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10/11/73
12:45-2:45
(% opacity)
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
B-10
-------
Facility V
Visible Emissions: Six minute opacity average calculated from readings
at thirty second intervals
Plant: No. 9 silicomanganese open electric submerged arc furnace
Company Name: Union Carbide (v)
Plant Address: Alloy, West Virginia
Date: 2/26/74
Types of Discharge: Baghouse monitor, building monitor, capture hood
Types of Control: Baghouse
B-ll
-------
Baghouse Monitor
2/26/74
10:20 a.m.-l:20 p.m.
Period Observer A Observer B
(6 minute average) (percent opacity) (percent opacity) Comments
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
13.7
11.7
8.8
4.1
7.0
4.6
2.5
3 7
4.6
5.4
7.0
13.3
11.7
12.7
15.0
17.5
16.7
15.8
12.7
10.0
6.7
5.4
6.7
5.4
11.7
11.3
8.8
4.2
6.7
3.8
2.5
2.5
5.0
5.4"
7.0
14.2
10.8
12.0
15.0
16.3
15.1
12.9
12.9
10.0
8.3
6.7
6.7
5.0
B-12
-------
Baghouse Monitor
2/26/74
10:20 a.m.-l:20 p.m
Period Observer A Observer B
(6 minute average) (percent opacity) (percent opacity) Comments^
25 7.0 8.3
26 5.8 6.3
27 5.0 5.0
28 5.0 5.0
29 8.8 7.4
30 10.0 10.8
B-13
-------
Baghouse Monitor
2/26/74
1:20 p.m.-4:20 p.m.
period Observer A Observer B
(6 minute average) (percent opacity) (percent opacity) Comments
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
12.5
13.3
14.2
15.0
15.8
14.2
11.3
8.3
9.2
14.2
12.1
11.3
10.8
11.3
12.1
13.5
14.2
10.8
10.8
10.0
10.0
9.2
12.5
9.2
15.0
13.8
13.3
12.0
15.8
13.3
10.8
8.8
8.8
15.4
11.3
11.3
10.8
11.3
9.6
9.6
11.3
10.0
10.0
9.6
10.0
9.6
10.8
7.9
B-14
-------
Baghouse Monitor
2/26/74
1:20 p.m.-4:20 p.m.
Period Observer A Observer B
(6 minute average) (percent opacity) (percent opacity) Comments
55 7.5 7.9
56 10.0 10.4
57 10.0 7.5
58 7.9 7.9
59 5.4 7.9
60 5.8 10.0
B-15
-------
Facility M
Visible Emissions: Six minute opacity averages calculated from readings
at thirty second intervals
Plant: No. 7 silicon open electric submerged arc furnace
Company Name: Interlake, Inc.
Plant Address: Beverly, Ohio
Date: 2/22/74
Types of Discharge: Baghouse monitor, building roof monitor
Types of Control: Baghouse
Note: No data were taken for roof monitor discharge due to smoke fron
another source obscuring roof monitor.
B-16
-------
Baghouse Monitor
2/22/74
8:50 a.m.- 2:50 p.m.
Period
(6 minute
average)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Observer A
(percent opacity)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.3
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
0.0
Observer B
(percent opacity) Comments
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
B-17
-------
Baghouse Monitor
2/22/74
8:50 a.m.-2:50.p.m.
Period
(6 minute
average)
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Observer A
(percent opacity)
0.0
0.0
0.0
-
-
-
-
0.0
1.2
0.0
0.0
0.0
0.0
-
-
-
-
-
0.0
0.0
0.0
0.0
0.0
Observer B
(percent opacity)
0.0
0.0
-
-
-
-
-
0.0
0.8
0.0
0.0
0.0
0.0
-
-
-
-
-
0.0
0.0
0.0
0.0
0.0
Comments
wind shift toward
baghouse cause smoke
from other building
to obscure monitor
baghouse obscured
by smoke
B-18
-------
Baghouse Monitor
2/22/74
8:50 a.m.-2:50 p.m.
Period
(5 minute
average)
48
49
50
51
52
53
54
55
56
57
58
59
60
Observer A
(percent opacity)
0.0
0.0
-
0.0
-
-
-
0.0
0.4
0.0
-
-
_
Observer B
(percent opacity)
0.0
0.0
-
0.0
-
-
-
0.0
0.0
0.0
-
-
-
Comments
baghouse obscured
by smoke
baghouse obscured
by smoke
B-19
-------
TECHNICAL REPORT DATA
(Please read laa/ucttons on the reverse before completing}
i REPORT NO.
EPA-450/2-74-018c
3 RECIPIENT'S ACCESSIOf*NO.
4 TITLE ANDSUBTITLE
Backqround Information for Standards of Performance:
Electric Submerged Arc Furnaces for Production of
Ferroalloys Volume 3: Supplemental Information
s REPORT
Aoril
6 PERFORMING ORGANIZATION CODE
7 AUTHORIS1
8. PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
U. S. Environmental Protection Aqency
Office of Air Quality Planning and Standards
Research Triangle Park, N. C. 27711
10 PROGRAM ELEMENT NO.
11 CONTRACT/GRANT NO
12 SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15 SUPPLEMENTARY NOTES
16. ABSTRACT
This volume is the third in a series of background information on standards of
performance for electric submerged arc furnaces for production of ferroalloys. The
volume contains: (1) supplemental information on the basis for the mass standards
of performance, (2) a summary of the issues on the proposed standards and EPA's
responses, and (3) a reevaluation of the opacity standard with regard to recent
revisions to the opacity method CReference Method 9 of Appendix A to 40 CFR Part 60).
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Electric submerged-arc furnaces
Standards of performance
Air pollution control
IS DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21 NO OF PAGES
107
2O SECURITY CLASS (This page >
Unclassified
22 PRICE
EPA Form 2220-1 (9-73)
------- |