EPA-450/3-79-026
Primary Aluminum —
Background Information
for Promulgated Amendments
Emission Standards and Engineering Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
March I980
-------
This report has been reviewed by the Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards, Office of Air, Noise,
and Radiation, Environmental Protection Agency, and approved for
publication. Mention of company or product names does not constitute
endorsement by EPA. Copies of this report are available thrpugh the
Library Servies Office (MD-35), U.S. Environmental Protection Agency,
Research Triangle Park, N.C. 27711; or may be obtained, for a fee, from
the National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
Publication No. EPA-450/3-79-026
ii
-------
Background Information
and Final
Environmental Impact Statement
for
Primary Aluminum Plant Amendments
Type of Action: Administrative
Prepared by:
Don R. Goodwin
Director, Emission Standards and Engineering Division
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Approved by:
(Date)
David 6. Hawkins
Assistant Administrator for Air, Noise, and Radiation
Environmental Protection Agency
401 M Street, S. W.
Washington, D. C. 20460
Final Environmental Impact Statement submitted to
EPA's Office of Federal Activities on
Additional copies may be obtained at:
Environmental Protection Agency Library (MD-35)
Research Triangle Park, North Carolina 27711
Telephone (919) 541-2777
This document may be reviewed at:
Central Docket Section (A-130)
Room 2903B, Waterside Mall
Environmental Protection Agency
401 M Street, S. W.
Washington, D. C. 20460
'(Date)
April 1980
(Date)
iii
-------
PRIMARY ALUMINUM
BACKGROUND INFORMATION: PROMULGATED AMENDMENTS
Table of Contents
1. INTRODUCTION
2. COMMENTS AND RESPONSES ....
2.1 Introduction .....
2.2 List of Commenters ....
2.3 Summary of Public Comments
and EPA Responses ....
3. ENVIRONMENTAL, ECONOMIC, AND ENERGY IMPACTS
Page
1
3
3
3
4
23
IV
-------
1. INTRODUCTION
On September 19, 1978, the U.S. Environmental Protection Agency
(EPA) proposed amendments to the standards of performance for new
primary aluminum plants (43 FR 42186) . The standards and amendments
are intended to limit emissions of gaseous and particulate fluorides
into the atmosphere. Also proposed on September 19 were revisions
to Reference Method 14 for measuring fluoride emission rates from
primary aluminum plants. The purpose of this document is to summarize
and discuss comments submitted by the public on the September 19
proposal, and to present the environmental, economic, and energy impacts
of the final amendments.
The standards of performance, including Reference Method 14, were
originally proposed on October 23, 1974 (39 FR 37730), and promulgated
on January 26, 1976 (41 FR 3826). Accompanying these standards were
three documents, entitled Background Information for Standards of
Performance: Primary Aluminum Industry - Volume 1: Proposed Standards
(Report Number EPA 450/2-74-020a); Volume 2: Test Data Summary
(EPA 450/2-74-020b); and Volume 3: Supplemental Information
(EPA 450/2-74-020c). Also, in 1977 EPA published the results of an
emission test program, entitled Air Pollution Emission Test, The
Anaconda Company Sebree Reduction Plant, Henderson, Kentucky (Report
1
This refers to Volume 43, page 42186, of the Federal Register.
-------
Number 77-ALR-6),. This test program was used in developing the
proposed amendments and revisions. Supplemental information for the
proposed amendments is contained in another document, entitled
Primary Aluminum Background Information: Proposed Amendments
(EPA 450/2-78-025a). All of the Federal Register notices and documents
noted above may be obtained from the U.S. EPA Library (MD-35), Research
Triangle Park, N.C. 27711, telephone (919) 541-2777. The final
amendments and revisions have been published in the Federal Register
concurrently with the publication of this document. Both may be
obtained from the EPA library.
-------
2. SUMMARY OF PUBLIC COMMENTS AND EPA RESPONSES
2.1 INTRODUCTION
Upon proposal of the primary aluminum amendments and Reference
Method 14 revisions, the public was invited to participate in the
rulemaking process by submitting comments to EPA's Emission Standards
and Engineering Division. Five comment letters and two telephone
calls were received. Also, a telephone call was placed to a represen-
tative of a primary aluminum company to ask for information on
anemometer calibration. The comment letters and records of telephone
calls may be reviewed in EPA's docket for the primary aluminum
amendments (docket number OAQPS-78-10).
All of the comments were reviewed and were useful in developing
the final amendments. A list of commenters and a discussion of
significant comments appear below.
2.2 LIST OF COMMENTERS
Commenter
and Organization
Indra Jashnani
Martin Marietta Company
Baltimore, Maryland
Joseph E. Gibb
Aluminum Company of America
Alcoa Center, Pennsylvania
Docket
Identification
IV-D-1
-2
-------
-3
-4
-5
-6
-7
-8
Bill Stewart
Texas Air Control Board
Austin, Texas
Frank S. Lisella
Public Health Service
Atlanta, Georgia
Lawrence C. Tropea, Jr.
Reynolds Metals Company
Richmond, Virginia
Donald W. Everett
The Anaconda Company
Louisville, Kentucky
Jim Yeager
Anaconda Aluminum Division
Louisville, Kentucky
John Prohaska
S. C. Department of Health and
Environmental Control
Columbia, South Carolina
2.3 SUMMARY OF PUBLIC COMMENTS AND EPA RESPONSES
Monthly Performance Testing
COMMENT: The proposed amendments would require performance
testing at least once each month throughout the life of a new primary
aluminum plant. This requirement would impose a very large economic
burden on the plant. For example, EPA's figures indicate that monthly
testing would require $156,000 per year in addition to the large
capital investment and operating expenses required for the control
systems themselves. Testing at intervals less frequent than monthly
should be sufficient to determine whether a new plant is complying with
the standards. The proposed amendments would impose onerous,
expensive, and unnecessary testing requirements on well-controlled
Mr. Yeager was contacted and asked to provide information on
anemometer calibration.
-------
plants. Several possible alternatives to the proposed monthly testing
requirement are as follows:
(1) An initial performance test, as originally required under
40 CFR 60..8(f) , would be sufficient to demonstrate compliance.
Periodic visual inspections could then be used to determine whether
the control systems were being properly maintained. If the visual
inspections indicated that maintenance was poor, monthly testing
could then be required. This procedure would not impose the burden of
monthly testing on the entire industry.
(2) Criteria should be established for determining when monthly
testing is necessary. For example, testing could be performed on a
six-month basis until a violation occurred, when testing would revert
to a monthly schedule.
(3) The provisions permitting the Administrator, upon application,
to establish an alternative test schedule for primary and bake plant
emissions could be extended to include secondary emissions. For example,
quarterly testing of secondary emissions could be required until a
violation occurred. Monthly testing could then be invoked for some
period of time, possibly six months, until emissions were once again
consistently below the level of the standard. Quarterly testing would
then resume.
RESPONSE: While developing the proposed amendments, it was found
that the operation and maintenance of aluminum plant emission control
systems had seriously deteriorated during the past several years. The
Administrator believes that regular emission testing will help remedy
Vhis refers to Title 40, section 60.8(f) of the Code of Federal
Regulations.
-------
this situation by providing an incentive for good operation and
maintenance throughout the life of the plant. Monthly testing was
selected as a compromise between other testing options. Weekly testing,
for example, would ensure better operation and maintenance than monthly
testing, but at a much greater cost. Alternatively, quarterly or
yearly testing would be less expensive, but would not have ensured
an adequate level of maintenance. Monthly testing, on the other hand,
will ensure satisfactory operation and maintenance of the emission
control systems at a reasonable cost.
In evaluating criticism of the proposed monthly testing requirement,
the Administrator focused his attention on costs. Since the cost of
the monthly testing requirement, at less than 0.5 percent of the
annualized costs of a typical primary aluminum plant, is so small, the
Administrator could not justify changing the proposed monthly testing
requirement. More information on costs appears in Chapter 3.
Concerning the three recommended alternatives to monthly testing,
the Administrator does not feel that any can be implemented. EPA does
not have enough trained personnel to perform periodic visual inspections,
as recommended under alternative number (1). Also, visual inspections
would not always guarantee that the entire emission control system was
operating properly. Alternatives number (2) and (3) would not accomplish
the goal of monthly testing of secondary emissions. The Administrator
believes that testing as often as once each month is necessary to ensure
an adequate level of maintenance between tests.
Performance Check Of Anemometers Following Each Test Series
COMMENT: The proposed Reference Method 14 revisions would require
a performance check of the propeller anemometers to be conducted after
-------
each series of three test runs. These requirements could lead to
substantially increased costs in the sampling program when there are a
number of anemometers to be removed and recalibrated every month. .
Do factual data exist which support the need to conduct^performance
checks as frequently as every month, or would quarterly checks serve
the same purpose?
RESPONSE: The requirement to check the performance of the propeller
anemometers after each series of three test runs was selected to provide
the best possible information for performance testing. Review of
calibration and cost data, however, indicates that anemometer calibration
checks as often as every month are unnecessary, and that yearly checks
would be adequate. Section 4 of Reference Method 14 has been revised
to require an anemometer performance check 12 months after 'the initial
anemometer installation. The results of this performance check will be
used to determine the schedule of subsequent performance checks.
Recalculation of Average Velocity
COMMENT: If an anemometer fails a performance check, Section 4.1.2.4
of the proposed Reference Method 14 revisions would allow recalculated
values of the average velocity (v) to be used in recomputing the overall
average velocity and total flowrate. This procedure would only be
allowed, however, in situations where recalculation of v would lead to
higher calculated emission rates. No similar provision exists for
recalculating v in situations where doing so would lead to lower emission
levels. This appears to represent a double standard. It would be more
logical to recalculate v for all situations, regardless of whether the
-------
calculated emission rates would increase or decrease. Alternatively, it
would be logical to not allow v to be recalculated in any situation
since there is no reason to believe that the accuracy of measurement
would change with the direction of error.
RESPONSE: This proposed requirement has been deleted from the
final Reference Method 14 revisions because of changes in the test
method which allow longer intervals between post-test performance checks.
Use of Single Sampling Train for Entire 24-Hour Test Run
COMMENT: The proposed Reference Method 14 revisions, by reference
to Methods 13A and 13B, would require a single sampling train to be used
throughout an entire test run. It would be impossible to comply with
this requirement, however, without overloading the silica gel impinger.
Should the impinger be changed during the test run or should the entire
sampling train be replaced?
RESPONSE: Revisions of Methods ISA and 13B which will clarify this
question are under final development. Similar revisions to Method 5 have
already been finalized (42 FR 41754).
Anemometer Installation at Point of Average Velocity
COMMENT: Section 2.1.2.3 of the proposed Reference Method 14
revisions directs the owner or operator of a plant to install the
anemometers at a point of average velocity as determined by a traverse
of the roof opening. Our experience with hand-measured velocity
measurements across a roof opening confirms that the point of average
velocity varies with every change in wind direction or speed; and at low
8
-------
velocity winds, this happens in as little as 15-minute intervals at the
extreme. Other aluminum companies confirm our test results which indicate
that as much as 10 percent of the time air actually enters the building
through the roof. It is recommended that the anemometers be located in
the center of the roof opening, as far as possible from any adjacent
support structure.
RESPONSE: We agree with the recommendation. Section 2.1.2.3 has
been revised to allow, as an option, the owner or operator to install
the anemometers halfway across the width of the roof monitor.
Balancing of Sampling Manifold
COMMENT: Section 4.2 of the proposed Reference Method 14 revisions
appears to require the sampling manifold to be mounted in the center of
the roof opening and to be balanced for flow rate after the potline
has commenced operation. It is recommended that this section be
rewritten to allow the balancing of the manifold to be based on the design
flow of the building. The manifold can be installed on scaffolding at the
down-comer blower in the courtyard. By adjusting the blower to an
average volume, as determined from building design or measured averages
of similar rooms, each of the eight manifold nozzles can be adjusted
with a pi tot tube so that each nozzle draws air at equal velocities.
The adjustable gates can then be permanently sealed. This can all be
done and the manifold installed permanently in the roof opening before
the crane is required to service the operating cells.
RESPONSE: This recommendation has been incorporated into the final
Reference Method 14 revisions.
-------
Anemometer Placement in Segmented Roof Monitors
COMMENT: Section 2.1.2 of the proposed Reference Method 14 revisions
requires that roof monitors less than 130 meters in length use at least
two anemometers. Although some roof monitors are longer than 130 meters,
the individual monitor segments may not be. We have considered that one
anemometer would be required for each 85 meters; and as the segmented
roof monitors are less than 85 meters in length, we have typically been
using only one anemometer per segment. It is recommended that the
proposed amendments be clarified to state that only one anemometer would
be required in a segmented roof monitor.
RESPONSE: EPA's experience at the Sebree plant shows that the use
of one anemometer would probably yield erroneous results. Consequently,
the Administrator believes that use of only one anemometer in a segmented
roof monitor should not be allowed. In special situations, however, the
Administrator may approve the use of one anemometer as an alternative
procedure.
Performance Testing at Modified Existing Plants
COMMENT: If an existing plant is modified to such an extent that it
becomes subject to performance standards for new sources, and if the
plant has secondary emission controls (i.e., roof scrubbers), then the
use of Reference Method 14 for sampling emissions would be impractical.
This is because Method 14 would require installation of a sampling
manifold in the area of the roof monitor occupied by the scrubbers.
RESPONSE: If it can be demonstrated that the use of Reference
Method 14 in a modified existing plant would be impractical, 40 CFR
60.8(d) of the General Provisions permit the owner or operator to apply
10
-------
to the Administrator for a waiver from the testing requirement or for
permission to use an alternative testing procedure.
Proposed Emission Limit for Soderberg Plants
COMMENT: The original standards required potroom emissions to be
below 0.95 kg/Mg (1.9 Ib/ton) for prebake plants and below 1.0 kg/Mg
(2.0 Ib/ton) for Soderberg plants. The 0.05 kg/Mg (0.1 Ib/ton)
difference between the standards was reasonable in view of the difference
between the two types of plants. EPA has erred in not following this
same reasoning in developing the proposed never-to-be-exceeded limit of
1.25 kg/Mg (2.5 Ib/ton) which applies to both prebake and Soderberg plants.
It is recommended that a never-to-be-exceeded limit of 1.3 kg/Mg (2.6 lb/
ton) be established for Soderberg plants while retaining the proposed
1.25 kg/Mg (2.5 Ib/ton) limit for prebake plants.
RESPONSE: This recommendation has been incorporated into the final
amendments.
Compliance With Proposed 1.25 kg/Mg (2.5 Ib/ton) Limit
COMMENT: The preamble to the proposed amendments states, "...under
no circumstances, however, would performance test results be allowed
above 1.25 kg/Mg (2.5 Ib/ton)." This statement could be construed to mean
that each individual test run, rather than the average of three runs,
must be below the proposed 1.25 kg/Mg limit. This should be clarified
in the final preamble and regulation.
RESPONSE: The preamble to the proposed amendments notes that,
"a performance test is defined in 40 CFR 60.8(f) as the arithmetic mean of
three separate test runs, except in situations where a run must be
discounted or canceled and the Administrator approves using the
11
-------
arithmetic mean of two runs". This will be repeated in the preamble
to the final amendments.
Use of Tapping Crucibles With Aspirator Air Return Systems
COMMENT: The preamble to the proposed amendments states, "...only
tapping crucibles with functional aspirator air return systems (for
returning gases under the collection hooding) should be used..." This
recommendation is totally unnecessary in cases where a company can, with
an alternate tapping procedure, achieve compliance with the emission
standards. Furthermore, the recommendation may encourage a state
environmental protection agency or an EPA Regional Office to unnecessarily
require a company to modify its existing tapping practices. The
recommendation to return aspirator gases under the hooding should be
deleted. Alternatively, the recommendation should be adhered to only
where it is consistent with company tapping practices or necessary to
achieve compliance.
RESPONSE: The proposal preamble states that certain operating and
maintenance procedures, including the use of aspirator air return systems,
are exemplary of good emission control and should be implemented. Although
this statement reflects the Administrator's judgment about which procedures
would enable the standards to be achieved, the regulation itself does not
actually require those procedures to be implemented.
If emissions are below 0.95 kg/Mg (1.9 Ib/ton) for prebake potrooms
and 1.0 kg/Mg (2.0 Ib/ton) for Soderberg potrooms, any combination of
procedures may be used. If emission levels are between 0.95 and 1.25
kg/Mg (1.9 and 2.5 Ib/ton) for prebake potrooms or 1.0 and 1.3 kg/Mg
12
-------
(2.0 and 2.6 lb/ton)x for Soderberq potrooms the regulation requires the
owner or operator of a plant to demonstrate that exemplary operating
and maintenance procedures are used. Otherwise, a violation of the
standard would occur. The Administrator has not defined exemplary
operating and maintenance procedures in the regulation because different
plants, .depending on plant design, may incorporate different procedures.
Instead, the Administrator believes that exemplary operating and
maintenance procedures for a particular plant should be determined by
EPA on a case-by-case basis. A plant should not be required to use
aspirator air return systems (or the remaining techniques noted in the
preamble to the proposed amendments) if the standards can be achieved
with other procedures.
Performance Testing of Primary and Bake Plant Emissions
COMMENT: The proposed amendments would allow relief from the
monthly testing requirements for primary and bake plant emissions only
if the contribution of the primary and bake plant emissions to the
total emission rate averages about 2.5 and 5.0 percent, respectively.
These proposed percentage requirements are arbitrary and unnecessary,
and their adoption would prevent many well-controlled plants from
obtaining an exemption. The key issue here is the variability of emissions
and whether primary and bake plant emissions are predictable, not whether
they contribute some arbitrary percentage to overall plant emissions. It
is recommended that EPA delete from the amendments the percentage
guidelines and instead require the owner or operator of a plant to show
that the primary and/or carbon bake plant emissions are predictable. Under
this approach, exemptions would be granted in cases where emissions from
the primary and bake plant control systems are greater than 2.5 and
13
-------
5.0 percent, respectively, if it can be demonstrated that the emission
levels from these systems can be accurately predicted.
RESPONSE: The commenter has misunderstood the preamble to the
proposed amendments. The percentages mentioned in the preamble are
neither guidelines nor requirements, but are intended to help justify
EPA's position that testing of the primary control system and bake plant
as often as once each month may be unreasonable. The preamble states:
The proposed amendments would allow the owner
or operator of a new plant to apply to the
Administrator for an exemption from the monthly
testing requirement for the primary control system
and the anode bake plant. EPA believes that the
testing of these systems as often as once each month
may be unreasonable given that (!) The contribution
of primary and bake plant emissions to the total
emission rate is minor, averaging about 2.5 and
5 percent, respectively; (2) primary and bake plant
emissions are much less variable than secondary
emissions; and (3) the cost of primary and bake
plant emissions sampling is high. An application to
Administrator for an exemption from monthly testing
would be required to include (1) evidence that the
primary and bake plant emissions have low variability;
(2) an alternative testing schedule; and (3) a
representative value for primary emissions to be used
in total fluoride emission calculations.
The regulation states:
An owner or operator may petition the Administrator
to establish an alternative testing requirement that
requires testing less frequently than once each month
for a primary control system or an anode bake plant.
If the owner or operator shows that emissions from the
primary control system or the anode bake plant have low
variability during day-to-day operations, the
Administrator may establish such an alternative
testing requirement. The alternative testing require-
ment shall include a testing schedule and, in the
case of a primary control system, the method to be
used to determine primary control system emissions for
the purpose of performance tests. The Administrator
shall publish the alternative testing requirement in
the Federal Register.
14
-------
Note that neither the preamble nor the requlation requires primary or
bake plant emissions to be a percentage of total emissions.
Permanent Manifold and Anemometer Installations
COMMENT: Section 1.1 of the proposed Reference Method 14 revisions
requires that manifolds and anemometers be permanently installed in each
potroom group. Depending upon the number of installations required at
the plant, it might be more economical not. to permanently mount the
anemometers, but to have specific locations for anemometers to be
placed before test runs are performed; and to move them from potroom
group to potroom group as required.
RESPONSE: Based on this recommendation, the final Reference Method
14 revisions do not require the anemometers to be permanently installed.
Definition of Performance Test
COMMENT: A performance test is currently defined in 40 CFR 60.8(f)
as the arithmetic mean of three separate test runs. Separate runs are
not cost-effective and create manpower and morale problems. Three
consecutive runs are more efficient, accurate, and representative of an
aluminum reduction operation. During the cooperative EPA test program
at the Anaconda aluminum reduction plant in Sebree, Kentucky, the
efficacy of consecutive runs was discussed with EPA and agreed to. It
is recommended that 40 CFR 60.8(f) be modified to provide for the
arithmetic mean of three consecutive test runs.
RESPONSE: The term "separate" means that the results of the test
runs are to be calculated individually. This does not prevent the
three test runs from being conducted consecutively. In the case of
15
-------
emission testing at primary aluminum plants, it would be logical to
conduct the tests consecutively due to the duration and complexity
of the tests. This will be clarified in the preamble to the final
amendments.
Duration of Test Runs
COMMENT: Section 5.3.4 of the proposed Reference Method 14 revisions
specifices an 8-hour period for each test run. This time period appears
to be inconsistent with the 24-hour period stated on page 42187 of the
preamble to the proposed regulation.
RESPONSE: The new aluminum plants EPA has examined operate on
24-hour cycles; that is, 24 hours are required for the major potroom
operations and events to occur in the area beneath the sampling manifold.
Consequently, the cost estimates located on page 42187 of the preamble
were based on 24-hour emission sampling periods. It is essential that
the sampling period represent all potroom operations and events, including
tapping, carbon setting, and tracking. This is not inconsistent with
Section 5.3.4 of Reference Method 14 which states:
Each test run shall last 8 hours or more; if more than
one run is to be performed, all runs shall be of approximately
the same (+.10 percent) length. If question exists as to the
representativeness of an 8-hour test, a longer period may be
selected. Conduct each run during a period when all^normal
operations are performed underneath the sampling manifold.
During the test period, all pots in the potroom group shall
be operated such that emissions are representative of normal
operating conditions in the potroom group. (Emphasis added.)
Definition of Potroom Group
COMMENT: The proposed definition of potroom group would require
many sampling manifolds to be installed in pjlants ..where only a few
16
-------
pots are ducted to a primary control system. Some existing plants
utilize as many as 36 individual primary control systems. The proposed
definition would require 36 sampling manifolds to be installed in
similarly designed new plants. This would be unnecessary and wasteful.
It is recommended that the proposed amendments be revised such that only
one sampling manifold is required for each pot type, provided that
similar operating practices are employed. Similar types of pots operated
in the same plant by the same supervisory management would have similar
roof monitor emissions. Thus, it would not be necessary to sample
emissions throughout the entire plant. Simple visual inspections of pot
hooding in potroom areas where testing is not conducted would reveal
whether the hooding is maintained as well as in potroom areas where testing
is conducted.
RESPONSE: The definition of potroom group is based on the design
of the Anaconda plant in Sebree, Kentucky, and on similar plants where
only two primary control systems are associated with each potline.
Plants planned for the future are expected to be similarly designed. If
a new or modified plant is designed differently, and if it can be demon-
strated that testing of the plant would be impractical and costly due to
a large number of primary control systems, then the Administrator would
approve, upon application, an alternative test procedure.
Although some existing plants use as many as 36 individual primary
control systems, it is emphasized that these regulations do not apply to
existing plants. If testing is to be conducted at an existing plant
which is significantly different in design from the Sebree plant, alternate
test procedures should be developed.
1.7
-------
Prior Notice of Performance Tests
COMMENT: Under 40 CFR 60.8(d) it is assumed that the 30-day prior
notice of any performance test is limited to the initial test required
to obtain an operating permit. Thereafter, the seven-day notice of
performance tests, as required under section 60.195 of the proposed
amendments, would apply to monthly testing. Is this assumption correct?
RESPONSE: Yes. 40 CFR 60.8(d) has been revised to read as
fol1ows:
The owner or operator of an affected facility shall provide
the Administrator at least 30 days prior notice of any performance
test, except as specified under other subparts, to afford the
Administrator the opportunity to have an observer present.
(44 FR 33580)
The amended regulation for new primary aluminum plants
(40 CFR 60, Subpart S) states:
Following the initial performance test as required under
§60.8(a), an owner or operator shall conduct a performance
test at least once each month during the life of the affected
facility, except when malfunctions prevent representative
sampling, as provided under §60.8(c). The owner or operator
shall give the Administrator at least seven days advance
notice of each test. The Administrator may require
additional testing under section 114 of the Clean Air Act.
Calculation of Aluminum Production Rate
COMMENT: Since production records are compiled on a monthly basis,
section 60.195(f) of the proposed amendments for determining production
rates would cause unnecessary recordkeeping. It is recommended that the
section be revised as follows:
The rate of aluminum production is determined by dividing
the number of hours in the month into the weight of
aluminum tapped from the affected facility during the
month in which the performance test was performed.
18
-------
RESPONSE: If the rate of aluminum production were determined on a
calendar-month basis, as the commenter suggests, then in situations where
testing is conducted at the beginning of a month, the final test results
would not be known until the end of the month. This delay could allow
emissions to be above the standard for nearly an entire month before a
violation could be determined and corrective actions taken. It is
preferable that the test results be known as soon as possible after the
testing is completed, as provided for in the proposed and final regulations.
Metric Units of Measure
COMMENT: The duct work described in the proposal is defined in
terms of metric units of measure which do not correspond to
American standard sizes. This may cause an additional inflationary impact.
RESPONSE: Since the duct work will be custom-made,.EPA does not
expect the use of metric units to be a problem. If you wish, convert
the metric units to English units with the formula: meters x 3.281 = feet.
Basis of Standard
COMMENT: The proposed standards are not based on health-related
criteria. Furthermore, since ;the proposal does not indicate the actual
fluoride concentrations allowed, it is impossible to relate the standards
to worker health criteria established by the National Institute for
Occupational Safety and Health.
RESPONSE: The Clean Air Act requires standards of performance to
reflect:
...the best technological system of continuous
emission reduction which (taking into consideration
the cost of achieving such emission reduction, any
nonair quality health and environmental impact and
energy requirements) the Administrator determines
has been adequately demonstrated.
• 19
-------
Note that the Act does not require health-related criteria to be
considered. Nevertheless, public health and welfare were considered
several years ago when EPA initially decided to control emissions of
fluorides from primary aluminum plants. In particular, two studies on
fluorides were considered. One, the March 1970 Report of the Secretary
of Health, Education, and Welfare to the United States (91st) Congress,
concluded that inorganic fluorides are highly irritant and toxic gases
which, in low ambient concentrations, damage plants and animals. In
this report fluorides were the only pollutants, other than the criteria
pollutants (i.e., the pollutants for which National Ambient Air Quality
Standards have been established), specifically named as requiring Federal
action. The other study was the U.S. Senate Committee on Public Works'
report on the Clean Air Amendments' of 1970 (Senate Report No. 91-1196,
September 17, 1970). This report included fluorides on a list of
contaminants which have broad national impact and require Federal action.
Additional information considered by EPA may be found in the document :
entitled, Background Information for Standards of Performance: Primary
Aluminum Industry - Volume 1: Proposed Standards (EPA 450/2-74-020a),
pp. xvii to xxii.
The standards are intended to limit emissions of fluorides from the
plant itself and should not be related to worker health criteria. The
regulation and Reference Methods ISA and 13B contain instructions for
obtaining actual fluoride concentrations and converting to the units of
the standards.
20
-------
Characterization of Potroom Emissions
COMMENT: The definition of potroom emissions is incomplete in the
proposed amendments. What are the chemical characteristics of these
emissions?
RESPONSE: The proposed amendments do not contain the entire
regulation, as promulgated on January 26, 1976 (41 FR 3826). The
promulgated regulation defines "total fluorides" as:
...elemental fluorine and all fluoride compounds as measured
by reference methods specified in [40 CFR 60.195] or by
equivalent or alternative methods [see 40 CFR 60.8(b)].
The final amendments do not alter this definition. Fluorides are further
characterized in Volume 1 of the background information document for the
original standards (EPA 450/2-74-020a).
Number of Sampling Manifolds Required
COMMENT: Section 2.2.1 of Method 14 states, "[a] minimum of one
manifold system shall be installed for each potroom group ..." The
required number of manifolds is not clear for situations where a potroom
group consists of two or more potroom segments ducted to a common control
system. Is one sampling manifold required for each potroom segment;
is one manifold (i.e., eight sampling zozzles) to be divided equally
among the potroom segments; or is only one manifold to be installed in only
one potroom segment?
RESPONSE: Only one sampling manifold is required for each potroom
group. In situations where a potroom group consists of a group of
potroom segments ducted to a common control system, the manifold would
be installed in only one potroom segment. The manifold may not be
21
-------
divided among potroom segments; however, additional sampling manifolds may
be installed in the other segments, if desired.
When only one manifold is located in a potroom group, care must be
taken to assure that operations are normal in the potroom segments where:
manifolds are not located, but which are ducted to the same control system.
During normal operation, most pots will be in operation, no major upsets
will occur, and the operating and maintenance procedures followed in each
potroom segment, including the segment tested, will be the same. Otherwise,
the emissions levels measured in the tested potroom segment may not be
representative of emission levels in the other potroom segments.
22
-------
3. ENVIRONMENTAL, ECONOMIC, AND ENERGY IMPACTS
The potential increase in fluoride emissions into the atmosphere
resulting from the amendments is shown in Table 1. Although the table
shows a slight increase in emission levels, EPA believes that this
increase may be more than offset by the improved operating and mainte-
nance procedures required by the monthly testing requirement. No other
environmental impacts are associated with the amendments.
Monthly testing costs are shown in Table 2, and the economic and
energy impacts of the amendments are shown in Table 3. The costs in
Tables 2 and 3 are estimates and will vary somewhat according to labor
costs and other variable economic factors. Nevertheless, it is apparent
that the amendments will have an insignificant impact on the operating
expenses of a new primary aluminum plant.
23
-------
Table 1. FLUORIDE EMISSIONS FROM A
NEW PRIMARY ALUMIUM PLANT
Fluoride Emissions:
Without Amendments^
With Amendments ^
Difference
Mg/yr
105
105-138
0-33
(ton/yr)
(114)
(114-150)
(0-36)
1
It is assumed that the average capacity of a new plant is 110,000 Mg/yr
(120,000 tons/yr).
"If the amendments were not developed, fluoride emissions would be
limited to 0.95 kg/Mg (1.9 Ib/ton).
The amendments will limit fluoride emissions from prebake plants to
between 0.95 and 1.25 kg/Mg (1.9 and 2.5 Ib/ton). New plants are
expected to be of the prebake design.
24
-------
Table 2. PERFORMANCE TEST COSTS FOR A
NEW PRIMARY ALUMINUM PLANT
(1979 DOLLARS)
Cost for Single Test:
Primary Emissions
Secondary Emissions
Bake Plant Emissions
2
Annual Costs:
Primary Emissions
Secondary Emissions
Bake Plant Emission?
Total
First Year
$4200
5100
4200
$134,400
244,800
33,600
$412,800
Succeeding
Years
$4200
5100
4200
$ 67,200
244,800
16,800
$328,800
1
The following assumptions are made: each performance test consists
of the average of three 24-hour test runs; testing is performed by
two crews of plant personnel working 13-hour shifts; emissions from
the primary control system are ducted to a single stack; each affected
facility (i.e., potroom group) has only one secondary emissions
sampling manifold; the plant consists of four affected facilities
and one anode bake plant.
"It is assumed that after six months of operation of the plant an
alternative sampling program is approved for primary control system
and bake plant testing. The alternative program requires quarterly
testing. (These assumptions are necessary for making the cost
estimates, but are not intended to represent EPA policy. The
alternative sampling program is to be determined on a case-by-case
basis.)
25
-------
Table 3. ECONOMIC AND ENERGY IMPACTS
OF THE AMENDMENTS ON A NEW
PRIMARY ALUMINUM PLANT '
1
Increase In:
Capital Costs2
3
Annual i zed Costs
Energy Consumption
First Year
$2500
$415,000
0 ,
Suceeding
Years
$2500
$330,000
,0
It is assumed that the plant consists of four affected facilities
(i.e., potroom groups) and one anode bake plant.
It is assumed that two test anemometers will be located in each
affected facility and that the anemometers will be replaced yearly.
(Actually, it is unlikely that each anemometer will need to be
replaced every year; however, some plant operators may choose to
do so rather than to check the anemometers' performance, as would
otherwise be required.) These costs are only intended to represent
investments incurred by the amendments and do not cover items which
would otherwise be required by the original regulation (e.g., the
sampling manifold).
o
It is estimated that the amendments would cause annualized costs
to increase by about 0.45 percent (0.25<£/lb) during the first
year and about 0.35 percent (0.20<£/lb) during suceeding years of
operation of a new plant, based on the February 1979 selling price
of aluminum of 53<£/lb. It is unlikely that all of these costs
will be passed directly to the consumer. Thus, commercial aluminum
prices are likely to rise less than the percentages given.
26
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before compteting)
1. REPORT NO.
EPA 450/3-79-026
2.
4. TITLE AND SUBTITLE
Primary Aluminum - Background Information for'Promulgated
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
March 1980
Amendments
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Standards Development Branch
Emission Standards and Engineering Division
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This document summarizes and responds to comments submitted by the public on
the proposed amendments to the standards of performance for new primary aluminum
plants. The standards and amendments are intended to limit emissions of gaseous
and particulate fluorides into the atmosphere. The document also presents the
final environmental, economic, and energy impacts of the final amendments.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Air pollution
Pollution control
Standards of performance
Primary aluminum plants
Fluorides
Aluminum
Air pollution control
13B
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
26
Unlimited
20. SECURITY CLASS (This page).
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
------- |