United States
        Er.virsnsser.ta! Protection
        Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-453/R-95-OG8b
Final Report
June 1995
        Air
JV EPA  SECONDARY LEAD SMELTING
        BACKGROUND INFORMATION
        DOCUMENT FOR
        PROMULGATED STANDARDS
               NESHAP

-------
                       TABLE OF CONTENTS


                                                           Page

1.0  SUMMARY	   1-1

     1.1  Summary of Changes Since Proposal   	   1-2

          l.l.l  Applicability	   1-2
          1.1.2  Standards for Process Sources  	   1-2
          1.1.3  Standards for Process Fugitive Sources .   1-2
          1.1.4  Standards for Fugitive Dust Sources  . .   1-3
          1.1.5  Test Methods and Schedule	   1-3
          1.1.6  Monitoring Requirements  ........   1-3

     1.2  Summary of Impacts of Promulgated Action  ...   1-3


2.0  SUMMARY OF PUBLIC COMMENTS 	   2-1

     2.1  Area Source Finding	   2-4

          2.1.1  Exposure Modeling And Risk Assessment  .   2-5
          2.1.2  Impacts	  2-11
          2.1.3  Legal Issues	2-12

     2.2  Applicability	2-14

          2.2.1  General	2-14
          2.2.2  Selection Of The Source Category ....  2-16
          2.2,3  Selection Of Pollutants	 .  2-17

     2.3  Standards for Process Emission Sources  ....  2-19

          2.3.1  MACT Floor Regulatory Interpretation . .  2-19
          2.3.2  Rationale For Metal  HAP Emission
                 Standards	2-21
          2.3.3  Rationale For Organic HAP Emission
                 Standards	2-30
          2.3.4  Rationale For Total  Chloride  Emission
                 Standards	2-40

     2.4  Standards for Process Fugitive sources  ....  2-48

          2.4.1  Identification Of MACT Floor  for Process
                 Fugitive Sources 	   2-49
          2.4.2   Selection of the Lead Emission Limit for
                Process Fugitive Sources  	   2-50
          2.4.3   Selection of Equipment Standards ....   2-50
                              11

-------
            TABLE OF CONTENTS (CONTINUED)

                                                      Page

2.5  Standards for Fugitive Dust Sources	2-52

     2.5.1  Selection Of MACT for Fugitive Dust
            Sources	2-52
     2.5.2  Selection Of Fugitive Dust Control
            Standards	2-54

2.6  Test Methods and Schedule	2-56

     2.6.1  Lead and Metal HAP's	2-57
     2.6.2  THC and Organic HAP's	2-59
     2.6.3  Total Chlorides	2-60
     2.6.4  Face Velocity	2-61

2,7  Monitoring Requirements 	  2-61

     2.7.1  Metal HAP Monitoring Options 	  2-63
     2.7.2  Organic HAP's - THC And Temperature  .   .  2-74
     2.7.3  Total Chlorides - Parameter Monitoring  .  2-75

2.8  Recordkeeping and Reporting Requirements  .  .   .  2-81

2.9  Interaction with Other Rules  	  2-83

     2.9.1  Interaction With NAAQS	2-83
     2.9.2  Interaction with Regulation of Air
            Emissions Under RCRA	2-85
     2.9.3  Interaction With Effluent Limitation
            Guidelines	2-88
     2.9.4  The Title V Operating Permit Program .   .  2-88
     2.9.5  Administrative Procedure Act
            Requirements 	  2-90

2.10  Miscellaneous	2-91

     2.10.1  Definitions 	  2-91
     2.10.2  Compliance Dates  	       2-92
     2.10.3  Presentation of the Standards   .  ..  .   .  2-94
     2.10.4  Impacts of a Loss in Lead-Recycling
             Capacity	2-94
     2.10.5  Environmental Justice	   .  2-95
     2.10.6  Pollution Prevention Considerations .   .  2-96

APPENDIX A - Summary of Public Comments on Notice to
             Withdraw Hydrochloric Acid/Chlorine
             (HC1/C12)  Limits
                         111

-------
                          1. 0   SUMMARY

     On June 9, 1994, the Environmental Protection Agency
 (EPA) proposed national emission standards for hazardous air
pollutants  (NESHAP) for secondary lead smelters in the Federal
Register  (59 FR 29750) under authority of Section 112 of the
Clean Air Act  (the Act).  Public comments were requested on
the proposal in the Federal Register.  There were 31
commenters composed primarily of secondary lead smelters and
their trade associations, State and local air pollution
agencies, equipment vendors, and environmental organizations.
     Additional comments were requested in a supplemental
notice announcing the availability of additional data and the
EPA's plans to withdraw the proposed hydrochloric
acid/chlorine  (HC1/C12) emission standards (60 FR 19556,
April 19, 1995).  The comments that were submitted on the
supplemental notice, along with responses to these comments,
are summarized in Appendix A of this document.  The summary of
all comments and responses serves as the basis for the
revisions made to the standard between proposal and
promulgation.
     Section 1.1 of this chapter summarizes the changes made
in the rule since proposal.  Section 1.2 summarizes the
impacts of the final standards.  Chapter 2.0 of this document
provides a detailed summary of all the comments received on
the June 9, 1994 proposal and the EPA's responses to those
comments.
                              1-1

-------
1.1  SUMMARY OF CHANGES SINCE PROPOSAL
     Substantial changes have been made to the NESHAP in
response to public comments and additional analyses since
proposal.  The major changes are summarized below.
1.1.1  Appli cabi1i ty
     The applicability of the rule has been clarified by
adding a statement that the rule does not apply to primary
lead smelters, lead refiners, and lead remelters.
1.1.2  Standards for Process Sources
     A provision has been added to the organic HAP emission
standards for process sources that will allow owners and
operators of collocated reverberatory and blast furnaces to
comply with the total hydrocarbon (THC) emission limits for
blast furnaces when the reverberatory furnace is not
operating.  These limits are 70 parts per million by volume
(ppmv)  THC, as propane, for new blast furnaces and 360 ppmv
for existing blast furnaces.
     The HC1/C12 emission standards for process sources have
been withdrawn.  There have been no changes to the metal
hazardous air pollutant (HAP) emission standards since
proposal.
1.1.3  Standards for Process Fugitive Sources
     The annual lead performance test requirement for process
fugitive sources has been removed as a result of changes in
the monitoring requirements.  Face-velocity requirements for
process fugitive enclosure hoods,  except for those controlling
dryer transition pieces and refining kettles, have been
changed from 350 feet per minute {fpm) to 300 fpm.  The
volumetric flow rate requirement for refining kettle enclosure
hoods has been deleted.
     Process fugitive sources located in enclosed buildings
have been exempted from the enclosure hood requirements if the
building is ventilated to achieve an air velocity of 250 fpm
at building doorways and the building is ventilated to a
control device.  Emissions from the building ventilation
                              1-2

-------
 control devices must comply with a lead emission limit of
 2.0 milligrams per dry standard cubic meter  (mg/dscm).
     The rule has been clarified to indicate that the exhaust
 from dryers and agglomerating furnaces must also comply with
 the lead emission limit of 2.0 mg/dscm.
 1.1.4  Standards for Fugitive Dust Sources
     The annual lead performance test requirement for fugitive
 dust sources that are enclosed and vented to control devices
 has been removed as a result of changes in the monitoring
 requirements.  The rule has been'clarified to indicate that
 structures meeting Resource Conservation and Recovery Act
 (RCRA) containment building standards satisfy the requirements
 for total enclosures specified in the standards for fugitive
 dust sources.
 1.1.5  Te s t Methods	and Schedu1e
     Velometers have been added as an alternate method for
 determining face velocities of process fugitive enclosure
 hoods and of doorway air velocity for building enclosures.
 1-1-6  Monitoring Requirements
     The requirement for a continuous opacity monitor (COM) on
 all process source baghouses to monitor compliance with the
 metal HAP lead emission standard has been withdrawn.  All
 process,  process fugitive, and building ventilation baghouses
must now undergo regular inspection and maintenance procedures
 specified in a standard operating procedures (SOP)  manual.
 The procedures must include a bag leak detection system and an
 alarm to indicate baghouse failures.   The SOP must also have a
 corrective action plan for responding to the alarms.
     The HC1/C12 monitoring requirements have been withdrawn
because the HC1/C12 emission standards have been removed.
 1.2  SUMMARY OF IMPACTS OF PROMULGATED ACTION
     The EPA has revised the estimated emission reductions and
 cost impacts of the rule since proposal.   One facility was
 incorrectly identified at proposal as employing gas stream
blending.   The estimated organic HAP emission reduction was
 revised from 1,220 megagrams per year (Mg/yr)  to 1,230 Mg/yr

                              1-3

-------
to reflect these updated baseline conditions.  The estimated
metal HAP reductions remain at 53 Mg/yr.  Although the EPA
plan.s to withdraw the HC1/C12 emission standards, HC1/C12
emissions are expected to decrease due to the decline of
polyvinyl chloride  (PVC) plastic in battery feedstock because
PVC is no longer used in lead-acid batteries.
     The total cost estimate for the promulgated rule only
includes costs resulting from controlling organic and metal
HAP emissions.  The estimated cost of controlling organic
HAP's was revised to account for additional costs associated
with gas stream blending because these costs were previously
under-estimated.  The previous estimated capital costs of
reducing organic HAP emissions under the proposed standards
were Si,100,000 and the estimated annual costs were $620,000.
_^<= o-c	,                              _c HAP's is
$3,300,000 and the revised annual cost is $1,700,000.  The
total estimated cost of reducing metal HAP emissions has not
changed since proposal.  There will be not costs associated
with reducing HCl/Clg emissions.
     The total estimated emissions reduction for both metal
and organic HAP's is 1,283 Mg/yr at an annual cost of
$1,860,000 for a cost effectiveness of $1,450 per Mg of HAP.
There is no expected change in the overall economic impact of
the rule since proposal.
     Since proposal, the EPA has also revised the estimated
increase in annual energy consumption.  The increase in energy
consumed by the controls needed to meet the final emission
standards is estimated to be 180 million cubic feet per year
of natural gas.   The increased annual natural gas consumption
estimated at proposal was 130 million cubic feet.  The change
in natural gas consumption is due to a revised analysis of the
costs and equipment needed to perform gas stream blending to
control organic HAP emissions from reverberatory/blast furnace
smelters.
     There have been no other changes in the impacts estimated
at proposal.

                              1-4

-------
                2.0  SUMMARY OF PUBLIC COMMENTS

     A total of 31 letters commenting on the proposed standard
and the Background Information Document (BID)  for the proposed
standard were received.  There were no requests to hold a
public hearing on the proposed standard.  A list of
commenters, their affiliations, and the EPA docket item number
assigned to their correspondence is given in table 2-1.
     For the purpose of orderly presentation,  the comments
have been categorized under the following sections and topics:
     2.1  Area Source Finding
     2.2  Applicability
     2.3  Standards for Process Emission Sources
     2.4  Standards for Process Fugitive Sources
     2.5  Standards for Fugitive Dust Sources
     2.6  Test Methods and Schedule
     2.7  Monitoring Requirements
     2.8  Recordkeeping and Reporting Requirements
     2.9  Interaction With Other Rules
     2.10 Miscellaneous
                             2-1

-------
TABLE 2-1.  LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
             STANDARDS FOR HAZARDOUS AIR POLLUTANTS FOR
             SECONDARY LEAD SMELTERS

Docket Item Number3-  Commenter and Affiliation
IV-D-Ol              Robert N. Steinwurtzel,  Stephen E. Roady
                     Association of Battery Recyclers
IV-D-02              Kathryn Grandfield
                     Sierra Club,  Ozark Chapter
IV-D-03              Neil J. Carmen
                     Sierra Club,  Lone Star Chapter
IV-D-04              Robert H. Colby,  Donald F. Theiler
                     State and Territorial Air Pollution
                     Program Administrators/Association of
                     Local Air Pollution Control Officials
IV-D-05              Michael J. Wax
                     Institute of Clean Air Companies
IV-D-06              Robert H. Collom,  Jr.
                     Georgia Department of Natural Resources
IV-D-07              Richard Thompson
                     GNB Battery Technologies
IV-D-08              Chris F.  Erichson
                     Erichson Company,  Inc.
IV-D-Q9              Langley A. Spurlock
                     Chemical Manufacturers Association
IV-D-10              Nancy D.  Chick
                     Colorado Department of Public Health and
                     Environment
IV-D-11              Jerome F. Smith,  Jeffrey T. Miller,
                     Edwin H.  Seeger
                     Lead Industries Association
IV-D-12              Mark L. Mullin
                     Markair,  Inc.
IV-D-13              Gerald Dubinski
                     Battery Council International
IV-D-14              Gerald A. Dumas,  Lynn L. Bergeson,
                     Ann Classon
                     RSR Corporation
IV-D-15              Wayne H.  Leipold
                     Cyprus Miami Mining Corporation
IV-D-16              Krishna Parameswaram
                     ASARCO,  Incorporated
IV-D-17              John B. Simonieg
                     Process Materials and Equipment
                     Corporation
rv-D-18              Asa Reed, Jr.
                     Private Citizen
                             2-2

-------
 TABLE 2-1.  LIST OF COMMENTERS ON PROPOSED NATIONAL EMISSION
               STANDARDS  FOR HAZARDOUS AIR POLLUTANTS  FOR
               SECONDARY  LEAD SMELTERS  (CONTINUED)

 Docket Item Number5  Commenter and Affiliation
 IV-D-19              Jerry 0.  Schuehler
                      Private Citizen
 IV-D-20              Patricia Leyden
                      South Coast Air Quality Management
                      District
 IV-D-21              Robert J.  Marzoli
                      United Process Control,  Inc.
 IV-D-22              Milton Feldstein
                      Bay Area  Air Quality Management District
 IV-D-23              Thomas Kruzen
                      Ozark Riverkeepers Network
 IV-D-24              Kazie Perkins
                      Private Citizen
 IV-D-25              Ray L. Williams
                      Shamrock  Control and Equipment,  Inc.
 IV-D-26              Rick Tacelli,  Chris Reiner-
                      Auburn International,  Inc.
 IV-D-27              John J. Rigby
                      National  Automotive Radiator Service
                      Association
 IV-D-28              John B. Blatz
                      The Dexter Corporation
 IV-D-29              Troy A. Greiss
                      East Penn Manufacturing  Co.,  Inc.
 IV-D-30              Glen E. Hasse
                      Schuylkill Metals Corporation
 IV-D-31              J.  T.  Healy
	Midwesco  Filter Resources, Inc.	

aThe docket number for this project is A-92-43.  Dockets are
 on file at Air and Radiation Docket and Information Center
 (6102), U. S. Environmental Protection Agency, 401 M Street,
 S.W.,  Washington, DC  20460.

NOTE:     Three of the 31 commenters endorsed letters
          submitted by other commenters.  Commenter IV-D-07
          endorsed the letters submitted by IV-D-01, IV-D-11,
          and IV-D-13.  Commenter IV-D-13 endorsed comments by
          IV-D-11.  Commenter IV-D-15 endorsed  comments by
          IV-D-16.
                              2-3

-------
2.1  AREA SOURCE FINDING
     Thirteen commenters  (IV-D-01, IV-D-02, IV-D-03, IV-D-04,
IV-D-07, IV-D-09, IV-D-10, IV-D-11, IV-D-13, IV-D-14, IV-D-18,
IV-D-22, and IV-D-30) commented on the EPA's decision to
regulate area source secondary lead smelters under the same
standards as major secondary lead smelters.  Six commenters
agreed with the EPA's decision and these comments are
discussed below.
     Seven commenters (IV-D-01, IV-D-07, IV-D-09, IV-D-11,
IV-D-13, IV-D-14, and IV-D-30) disagreed with the EPA's
finding that secondary lead smelter area sources should be
regulated by maximum achievable control technology  (MACT)
standards.  The specific arguments against the EPA's finding
are presented in the following subsections.
     Comment:  Six commenters  (IV-D-02, IV-D-03, IV-D-04,
IV-D-10, IV-D-18, and IV-D-22) agreed with the EPA's finding
that area source secondary lead smelters should be listed and
regulated by MACT standards.  Two commenters (IV-D-02 and
IV-D-04) agreed that area source smelters could be a threat to
public health.  One commenter  (IV-D-10) wrote that experience
has shown that small lead operations may still have an adverse
impact on air quality in the surrounding area.   Another
commenter (IV-D-22)  wrote that substantial data about the
serious impacts of low-level lead exposure indicate that
regulation of area sources is warranted.  One commenter
(IV-D-04)  agreed that there is no technical reason why area
source smelters cannot achieve the same control level as major
source smelters.
     Response:  The EPA agrees that area source smelters
present a threat of adverse health effects warranting both
listing and MACT standards and has received no new data to
indicate that the proposal to regulate area sources should be
revised.  In particular,  the area sources emit some of the
more dangerous hazardous air pollutants (HAP's):  lead,
arsenic, and 1,3-butadiene.  As set out in the preamble,
baseline emission levels from these area smelters are capable

                              2-4

-------
 of posing  substantial adverse health effects.  With respect to
 the EPA's  decision  to impose standards based on MACT rather
 Lhctii SACT  for  these area sources, MACT is available at
 reasonable cost and, moreover, is based on technology that is
 already installed,  with minor exceptions, at all area sources.
 The EPA, therefore, sees no reasonable justification for a
 lesser degree  of  control.
     In addition, the EPA believes that a failure to regulate
 all secondary  lead  smelters would be inconsistent with the
 principles set forth in Executive Order No. 12898, Federal
 Actions to Address  Environmental Justice in Minority
 Populations and Low-Income Populations.  Some secondary lead
 smelters are surrounded by either minority or low-income
 populations.   In  the past,  some secondary lead smelters have
 had significant adverse impacts on these communities (see
 docket item IV-J-2).  Therefore,  the EPA believes that it is
 important  to regulate all secondary lead smelters under MACT
 standards  to help protect these communities.
 2.1.1  Exposure Modeling And Risk Assessment
     Comment'•  One  commenter (IV-D-03)  supported the area
 source finding, but recommended that the EPA use cumulative
 emission reviews and dispersion modeling to evaluate the
 adverse health effects of secondary lead smelters.  The same
 commenter  also asked the EPA to consider the cumulative
 impacts of  secondary lead smelters and other stationary
 emission sources on large population centers when evaluating
 the impact of secondary lead smelters that are area sources.
     Response:  Under section 112 of the Clean Air Act (the
Act),  Congress has directed the EPA to consider the impacts
from specified source categories  of hazardous air pollutants
 (e.g.,  secondary lead smelters).   Although the EPA believes
that the cumulative impact  of other sources of lead compound
emissions  is important,  such assessments are beyond the scope
and intent of this rulemaking.   Consequently,  the subject
analyses were restricted to the impacts of the area sources
within the secondary lead smelter source category.

                             2-5

-------
     Comment:  One commenter  (IV-D-30) wrote that the EPA
based the area source finding on an analysis of only six of
the seven smelters, which is too small a sample to use for
regulating an entire industry.
     Response:  The EPA estimates that six smelters are area
sources and one smelter is near the emission threshold to
qualify as an area source.  The area source finding was based
on an analysis of all seven of these smelters.  The EPA
estimates that the remaining 16 smelters are all major
sources, which the Act requires to be regulated by MACT
standards.  The decision to regulate area sources and to
regulate by MACT standards only affects those smelters that
may qualify as area sources.  Therefore, the area source
finding v~s based on the entire population of affected
OillC j- i^C^ ^
     Comment:  Several commenters (IV-D-07, IV-D-11, IV-D-13,
IV-D-14, and IV-D-30) wrote that the EPA overstated the risk
associated with area source smelters.  The commenters stated
that the EPA should not have based the estimate of the number
of individuals exposed to secondary lead smelter emissions
from area sources by selecting a 50-kilometer (km)  (31-mile)
radius around each smelter in order to define the affected
population.
     One commenter (IV-D-07) suggested that the maximum radius
of influence around each smelter should have only been 5 to
8 km (3 to 5 miles).   Another commenter (IV-D-11) noted that
another EPA study of lead exposure treated individuals living
beyond 2.25 to 5 miles from the source as members of the
nonexposed control group.
     Response:  In the proposal preamble,  the EPA stated that
approximately 17.6 million people reside within 50 km
(31 miles) of the seven secondary lead smelters that are
considered area sources,  and that these people are considered
by the EPA to be exposed to HAP emissions from these smelters.
This 50-km (31-mile)  radius, however, does not necessarily
represent the extent of the area that is affected by smelter

                              2-6

-------
 emissions.   In  fact, it  only designates  the  limits  of  the area
 around  each  smelter  in which the effects  of  the smelter were
 iaodcled.  Tha E?A agrees with the commenters that smelter
 emissions and impacts from smelter emissions are minimal, if
 at all  detectable, at a distance of 50 km from each smeltBr.
     The EPA also agrees that only individuals located at
 shorter distances are likely to be adversely affected  by
 smelter emissions.   However, the radius  of influence around
 each smelter must be determined on a case-by-case basis using
 site-specific information.  There'fore, it cannot be assumed
 that the maximum radius of influence around each smelter is 5
 to 8 km (3 to 5 miles).  The study referred to by the
 commenter (IV-D-11)  was for only one smelter and analyzed only
 exposure to  lead emissions.  In contrast, the analysis
 conducted for this area source finding analyzed the effects of
 14 pollutants found  in smelter emissions  (10 of which  are
 potential carcinogens) and was performed for each area source
 smelter.
     Comment:  One commenter (IV-D-14) wrote that the  record
 did not support or demonstrate how the EPA arrived at  the
 estimate that 250 individuals are exposed to lead levels in
 excess  of the lead national ambient air quality standard
 (NAAQS)  of 1.5 micrograms per cubic meter (jig/m3).
     Another commenter (IV-D-30)  noted that there are no
 residents inside the fenceline of secondary lead smelters and
 all residents outside smelters are exposed to less than
 1.5 /zg/m3 of pollutants.   The same commenter estimated that
 the organic pollutants emitted by secondary lead smelters will
be at concentrations below 1 ^tg/m3 at the fenceline and
probably at undetectable levels beyond the fenceline.
     Response:   The procedures and raw data used in the
 adverse health effects finding are found in docket item
 II-B-33  of Docket No. A-92-43.   In these procedures, each area
 source was modeled using site-specific data and the Human
Exposure Model  (HEM).  Meteorological data were taken from the
nearest  airport to each facility and population data were

                              2-7

-------
taken from the 1990 census.  Exposures were calculated for
population centroids rather than only at the fenceline of each
source because the latter tends to overestimate risks.
Modeling was performed for six smelters that would qualify as
area sources and one smelter that is near the borderline
between an area and a major source.
     The EPA disagrees with the second commenter's (IV-D-30)
assertion that all residents living outside the smelter
fenceline are exposed to pollutants at concentrations below
1.5 /ig/m.3.  The EPA's modeling analysis indicated that about
250 individuals are exposed to ambient lead levels greater
than 1.5 jug/m3.  Moreover, several secondary lead smelters are
associated with lead nonattainment areas (56 FR 56694,
November 6, 1991)  and, therefore, contribute to lead exposures
greater than 1.5 ^g/m3.   Two of these smelters are area
sources.
     The commenter (IV-D-30)  provided no data or analyses to
support the conclusion that organic pollutants emitted by
smelters would be at undetectable levels beyond the fenceline
of a smelter.  The EPA agrees that the pollutant
concentrations are low,  but the EPA model indicates that they
present a risk at these levels.  This modelling analysis is
fully described in docket item II-B-33.
     Comment:  Two commenters disagreed with the risk
estimates used in the EPA's risk assessment for some of the
organic pollutants.  One commenter (IV-D-30) noted that the
unit risk estimate (URE)  used by the EPA to estimate cancer
risks represents the risk to a hypothetical individual
continuously exposed throughout his/her lifetime to a
pollutant concentration of 1 /ig/m3.  However,  the ambient
concentrations of pollutants for which there are URE's are not
presented in the risk assessment found in the preamble.
     Another commenter (IV-D-09) wrote that the upper-bound
cancer risk estimate for 1,3-butadiene used by the EPA in the
risk assessment for area sources is overstated by several
orders of magnitude.   The original 1985 risk assessment for

                              2-8

-------
 1,3-butadiene was based  on  studies using the B6C3F1 mouse,
 which  does not provide an appropriate model for human cancer
 risk assessment., according  to more recent data.
     One  commenter  (IV-D-30) noted that other than lead and
 2,3,7,8-tetrachlorodibenzo-p-dioxin  (2,3,7,8-TCDD), the
 pollutants used in the EPA's risk assessment were only
 measured  upstream of the smelters' control equipment during
 the EPA's testing program at secondary lead smelters.
     Response:  Summary concentrations for each area source
 smelter and aggregated concentrations for the area sources are
 in the docket (see docket item II-B-33).  The conclusions for
 this area source finding were based not only on 1,3-butadiene,
 but on the mixture of HAP's emitted.  The EPA is aware of the
 recent data concerning the potency of 1,3-butadiene.  The
 studies are being reviewed.  Until the reviews have been
 completed, the EPA is using the value found on the EPA's
 Integrated Risk Information System (IRIS).
     During the EPA's testing program, uncontrolled metal and
 organic HAP emissions were measured simultaneously with
 uncontrolled and controlled lead and total hydrocarbon (THC)
 emissions.  The EPA used lead as a surrogate to predict metal
 HAP emissions and THC as a surrogate for organic HAP's.   As
 described in docket item II-B-32, the thermal destruction of
 organic HAP's is very closely correlated with the destruction
 of THC.  The metal and organic HAP emissions used in the EPA's
 risk assessment were estimated from the controlled lead and
 THC emissions and the relative amounts of individual metal and
 organic HAP's and their surrogates measured in the
uncontrolled emissions.
     Comment:  Five commenters (IV-D-01,  IV-D-11,  IV-D-13,
 IV-D-14,  and IV-D-30)  presented data to refute the need to
regulate area sources.   One commenter (IV-D-01)  noted that
 data are available indicating that lead exposures, based on
blood lead levels sampled in the period from 1988  to 1991,
have declined substantially compared to similar data from 1976
to 1980.   The frequency of blood lead levels greater than

                              2-9

-------
10 micrograms per deciliter have declined from 77.8 percent to
4.3 percent in the decade between the two studies.
     Four commenters {IV-D-11, IV-D-13, IV-D-14,  and IV-D-30)
noted that studies of children living near secondary lead
smelters found the children's blood lead levels to be within
the range considered normal by the Centers for Disease
Control.  One comrtventer {IV-D-30) noted that emissions testing
by the EPA demonstrate that secondary lead smelter emissions
are less than the NAAQS,  The commenters concluded that these
studies demonstrate that smelters.that are in compliance with
the NAAQS do not pose a threat to public health.
     Response:  The EPA agrees with the commenters that a
great deal of progress has been made toward reducing the
general public's exposure to lead.   However, there is still
concern about current blood lead levels and the risks
presented from atmospheric lead deposition.  The former sites
of two smelters--Interstate Lead Company in Leeds, Alabama and
Dixie Metals in Dallas, Texas--have required Federal clean-up
action under the Comprehensive Environmental Response,
Compensation, and Liability Act, also known as "Superfund."
These actions were required because of lead deposition onto
the soils surrounding these sties.   The lead came from air
emission sources that will be regulated by this rule.
     In addition,  lead exposure was only one component of the
EPA's evaluation of area sources.  The decision to regulate
area sources was also based on an assessment of the effects of
all smelter emissions,  including metal HAP's and organic
HAP's.  The maximum lifetime individual cancer risk resulting
from exposure to arsenic and 1,3-butadiene are of particular
concern.
     The EPA disagrees with the commenter  (IV-D-30) that EPA
emissions testing demonstrates that secondary lead smelter
emissions are less than the NAAQS.   The EPA's testing program
measured stack emissions but performed no ambient monitoring.
Ambient monitoring to determine compliance with the NAAQS lead
concentration is performed at a facility's fenceline.

                             2-10

-------
Moreover,  several  secondary  lead smelters are  located  in  lead
nonattainment areas and at least" two: of-,,these  smelters are
area sources.
     Comment:  Three commenters  (IV-D-11, IV-D-13,, and
IV-D-14) argued  that if individuals are exposed to lead levels
above the  NAAQS  for lead, the EPA should not use a lack of
enforcement  of the lead NAAQS as a reason to regulate area
source  smelters  under a MACT standard.  One commenter
 {IV-D-13)  recommended that the EPA take additional action on
the enforcement  of the NAAQS dire'ctly, rather  than imposing a
new set of requirements that would have the indirect effect of
controlling the  NAAQS.
     Response:   The NAAQS for lead was used as one of several
health-effects benchmark for quantifying exposure to lead
emissions  from lead smelters.  The EPA's analysis indicates
that, even in areas where the NAAQS is achieved, area sources
present a  threat to public health.  The purpose of the rule is
to require technology-based  emission limits (considering  costs
and other  factors) on smelters that present a threat of
adverse effects  to public health.  The EPA has determined that
MACT standards are appropriate.
     The comparison of ambient lead concentrations to the lead
NAAQS was part of the overall analysis of adverse health
effects from area smelters and was not intended to identify
enforcement or compliance problems with the lead NAAQS.  The
goal of the rule is not to achieve compliance with the lead
NAAQS.   The rule, however,  should achieve a significant
reduction in ambient lead concentrations by reducing metal HAP
emissions from (in particular)  process fugitive and fugitive
dust sources.
2.1.2  impacts
     Comment:  One commenter (IV-D-07)  wrote that the EPA
underestimated the economic  impacts of regulating area source
smelters under a MACT standard.   The actual number of smelters
that would qualify as area sources,  and would otherwise be
                             2-11

-------
exempt from the MA.CT standards, is .greater than the number of
smelters estimated by the EPA.
     The same continenter  (IV-D-07) noted that regulating area
sources would require all smelters to apply for Title V
operating permits, even if they are area sources and already
have the equivalent of MACT controls.  This impact was not
considered by the EPA in deciding whether to regulate area
sources.
     Response.:   The EPA's estimate of the number of area
source smelters is based on estimated HAP emissions from
smelting furnace,  process fugitive, and fugitive dust emission
sources.  The procedure used by the EPA for estimating these
emissions is described in the proposal BID.  The EPA has
reviewed these emission estimates and continues to believe
~*^i. ,	                                . , accurate.  The
commenter provided no estimate of the number of smelters that
would qualify as area sources nor any technical basis for the
comment that more than seven smelters would qualify as area
sources.
     Secondary lead smelters are subject to Federal new source
performance standards (NSPS)  for secondary lead smelters
[40 Code of Federal Regulations  (CFR) 60,  subpart L] that
would also require them to obtain Title V operating permits.
Therefore,  there is no additional impact arising from the
Title V operating permits application procedure.
2.1.3  Legal Issues
     Comment;   One commenter (IV-D-14) wrote that the EPA has
asserted only that area source secondary lead smelters may
present a threat of adverse health effects and has not
satisfied the criteria in section 112(c){3} of the Act that
area sources must be found to actually present a threat.  The
same commenter also claimed that the EPA has not demonstrated
that any secondary lead smelters are area sources.  However,
the same commenter agreed that if any secondary lead smelters
are area sources,  they would be underregulated pursuant to
current regulations under the Act.

                             2-12'

-------
     Three  commenters  (IV-D-07,  IV-D-13, and  IV-D-30) wrote
 that area source smelters  should not be regulated because  the
 estimated adverse health and environmental  impacts associated
 with the area sources are  below  the criteria  that the
 Administrator has used to  designate other categories of area
 sources for regulation.  Another commenter  (IV-D-11) noted
 that the EPA's estimate of 0.1 cancer incidence per year
 traceable to emissions from all  of the smelters that were
 modeled represents a risk  of about 1 in 2.5 million and that
 this is below the l - in-1-million'de minimis level established
 in section  112 (c) (9) (B) (i) of the Act.
     Response:  Section 112(c)(3) does not require that actual
 adverse health and environmental effects be demonstrated to
 justify a listing determination.  The statute states that  the
 EPA shall list each category of  area sources that the EPA
 "finds presents a threat of adverse effects to human health or
 the environment....warranting regulation under this section"
 (i.e.,  under technology-based standards).  The reference to
 "threat" is a clear indication that potential harms can be
 considered,  because threats include potential occurrences.
 For this commenter to be correct, the Act would have to be
 worded "finds has caused adverse effects."
     The EPA agrees that the estimated carcinogenic effects of
 these area  sources are somewhat below the criteria the
Administrator has used to designate other categories of area
 sources for regulation at  the time of the proposal.   However,
 the EPA estimates that the upper-bound maximum individual
 lifetime cancer risk associated with any one of the area
 source smelters ranges from 4 in 10,000 to 1 in 1,000
 (59 FR 29755,  June  9,  1994).   This risk estimate does not
 include exposure to lead,  which is a B2 carcinogen,  and is
well above the 1-in-l  million threshold found in
section 112(c)(9)  of the Act.   Therefore,  the Administrator
has concluded that  consideration of all potential health risks
associated with these  area sources warranted regulation of
these sources.   The commenters have provided no information to

                             2-13

-------
change the EPA's conclusions about the level of risk exposure
from the seven sources.
     Tne commenter's  (IV-D-ll) probability estimate of
l-in-2.5 million for the level of risk exposure is incorrect.
As described in the previous paragraph, the EPA's estimates of
the lifetime cancer risks is well above the threshold found in
section 112(c)(9) of the Act.
2.2  APPLICABILITY
     Nine commenters  (IV-D-02, IV-D-03, IV-D-D4, IV-D-07,
IV-D-10, IV-D-16, IV-D-22, IV-D-28, and IV-D-30) commented on
the selection and definition of the source category and the
selection of the regulated pollutants.
2.2.1  General
     Comment:  Two commenters  (IV-D-22 and IV-D-28) requested
clarification oft the applicability of the standard.  One
commenter (IV-D-28) requested that the applicability of the
rule be clarified by adding a statement that the rule does not
apply to primary lead smelters, lead refiners, and lead
remelters.  The commenter also requested that the definitions
of the smelting furnaces be revised to indicate that they are
furnaces operating at temperatures greater than 980 °C
(1800 °F).  Another commenter  (IV-D-22) requested that the
applicability section of the proposed rule be clarified to
indicate whether it applies to recyclers of lead and lead
compounds from lead-acid batteries only or to recyclers of all
scrap lead and lead compounds, including those from lead-acid
batteries.  The commenter believes that the first
interpretation is implied by the proposed rule.
     Response:  The EPA agrees that the recommended changes
would clarify the applicability of the rule and has
incorporated them into the final rule.  The language of the
rule has also been revised to clarify that the rule applies to
recyclers of all scrap lead and lead compounds, including, but
not limited to, material from lead-acid batteries.
     Comment:  One commenter  (IV-D-22} requested that the EPA
establish low-usage thresholds that would exempt facilities

                             2-14

-------
 that only melt or remelt less than 400 tons per year of pure
 (soft! lead or less than 200 tons per year of hard lead,
     gegpoiise:  The rule will be revised to clarify that it
 does not apply to lead remelters and refiners and only applies
 to lead smelters that operate smelting furnaces.  The EPA is
 not aware of any lead smelters that would be subject to this
 rule with a capacity less than 10,000 tons per year.
 Therefore, the low usage thresholds recommended by the
 commenter are not necessary.
     Comment:  One commenter (IV-D-07) requested a
 clarification on how the proposed standards would apply to
 feed dryers.  The commenter indicated that, contrary to
 statements in the preamble, the exhaust from feed dryers is
 not necessarily drawn directly into the reverberatory
 furnaces.  At two facilities operated by the commenter, the
 dryer exhaust is handled through separate baghouse and
ventilation systems.  The commenter requested clarification of
 the implications of this arrangement with regard to the
applicability of the proposed standards.
     Response:  The feed dryer exhausts are subject to the
 lead emission standard of 2.0 mg/dscm and are considered a
process fugitive emission stream.  The EPA has revised the
rule to clarify that all other sources,  including feed dryers
and other smelting furnace types, shall meet a lead emission
standard of 2.0 mg/dscm.
     Comment:   One commenter (IV-D-16)  requested clarification
on how the EPA would apply MACT to new pyrometallurgical and
hydrometallurgical/electrowinning technologies.
     Response:  The final MACT standards  do not  apply to new
smelting or electrowinning technologies alluded  to by the
commenter.   The EPA does not have sufficient information on
these technologies to establish emission  standards under these
national emission standards for hazardous air pollutants
 (NESHAP)  [i.e.,  sections 112(c)  and (d)  of the Act]  at this
time.   New sources using new technology may be evaluated for
regulation at  a later date if they emerge.   In addition,  any

                             2-15

-------
new source using either of these technologies that is a major
source would be subject to section 112(g), which requires
newly constructed major HAP emission sources to install MACT.
The MACT for each new smelter not covered by this rule would
be determined on a case-by-case basis if no applicable
emission standards exist when and if a new technology is
commercialized in the United States.
2.2.2  Selection Of The Source Category
     Comment:   Two commenters (IV-D-04 and IV-D-10)  stated
that additional sources should be' evaluated for inclusion as
sources subject to the proposed standards.  One commenter
(IV-D-04) requested that scrap lead smelters and lead refiners
be evaluated to determine whether MACT standards are
appropriate for these sources.  Another commenter (IV-D-10)
requested that other sources of lead emissions, such as lead
refiners and lead oxide production facilities,  be considered
for inclusion in the source category subject to the proposed
regulation because the production of lead compounds, including
lead oxide, is not included elsewhere on the source category
list.
     Response:  Scrap lead smelters are subject to this rule,
but lead remelters, lead refiners, and lead oxide production
facilities are not.  These are different processes and have
substantially less emissions than secondary lead smelters.
The EPA possesses no information on lead remelters and lead
oxide manufacturers that would support their listing as
categories of major sources for regulation under section 112.
     Comment:   One commenter  (IV-D-30) argued that regulation
of the primary lead smelting industry should be a higher
priority of the EPA than regulating secondary lead smelters.
According to the commenter, the secondary lead industry emits
less than 20 tons per year of lead, but the primary lead
industry emits approximately 300 tons per year of lead and is
not in compliance with the lead NAAQS.  According to the
commenter, regulation of the secondary lead industry prior to
regulation of the primary lead industry puts the secondary

                             2-16

-------
lead industry at an unfair competitive disadvantage.  However,
no data were provided to support these statements.
     Response:  The primary lead smelting industry is
scheduled for regulation by MACT standards by the statutory
deadline of November 15, 1997.
2.2.3  Selection Of Pollutants
     Comment:  Two commenters  (IV-D-01 and IV-D-16) discussed
the surrogate pollutants selected by the EPA for organic HAP's
and metal HAP's.  One commenter (IV-D-01) supported the EPA's
decision to use THC as a surrogate for organic HAP emissions.
Another commenter (IV-D-16) requested that the EPA reconsider
its decision to establish an emission standard for lead rather
than particulate matter (PM)  as a surrogate for metal HAP's.
The commenter stated that because baghouses are the MACT for
metal HAP control and because baghouses control PM without
discriminating among PM species,  it may be more appropriate to
establish a limit for PM rather than lead.  The commenter also
noted that the EPA test data indicate that the ratio of lead
to PM is variable and lead emissions may vary with no
variation in the efficiency of the baghouse in controlling PM.
     Response:   As stated in the proposal preamble
(59 FR 29761),  the EPA evaluated both PM and lead as
surrogates for all metal HAP's for which emission standards
would be established.   The EPA found that compounds of lead
are the most prevalent metal  HAP in secondary lead smelter
emissions.  The EPA also found that lead is concentrated with
other metal HAP's in the smallest and most difficult to
control size fraction of PM from smelters.  In addition, it
was determined that  there is  a stronger correlation of metal
HAP emissions with lead emissions than with PM emissions.   No
data have been received since proposal that would change any
of these conclusions.
     The EPA agrees  that the  ratio of lead to PM is variable
and lead emissions may vary with no variation in control
efficiency for PM.   The EPA is therefore establishing a limit
for lead to ensure that controls are designed and operated to

                             2-17

-------
achieve effective control of lead compounds and other metal
HAP's that are found in the smallest size fractions of PM,
regardless of the overall control efficiency of PM.  The EPA
is not changing its decision to select lead as a surrogate for
other metal HAP's.
     Comment:  Three commenters (IV-D-02, IV-D-Q3, and
TV-D-10) discussed the exclusion of dioxins/furans from the
standards.  Two commenters (IV-D-02 and IV-D-03) disagreed
with the Agency's decision not to regulate these emissions
from secondary lead smelters.  Both commenters requested that
a zero-discharge limit on dioxin/furan emissions be
established.  The second commenter (IV-D-03) based this
request on the assertion that the EPA's Draft Reassessment of
t-hp wealth Effects of Dioxins and Furan? (59 FR 46980)
iiiuj-^auc^ -_,,,, .,      -      	            evel to dioxin/furan
emissions.
     Another commenter (IV-D-10) stated that it was not
possible to comment on the need to regulate dioxin/furan
emissions because it appeared from the preamble that the 1PA
had not adequately evaluated these pollutants.  The commenter
pointed out, however, that the intent of section 112 of the
Act is to analyze all listed HAP's from major sources and that
the EPA could not arbitrarily decide to postpone regulation of
dioxin/furan emissions until after the residual risk from the
source category has been evaluated.  The commenter added that
there would be less concern about dioxin/furan emissions if
the proposed standards for organic HAP emissions from blast
furnaces represented controls above the MACT floor.
     Response:  A zero-discharge limit for any pollutant is
not achievable.  The EPA routinely establishes emission
standards for nonthreshold pollutants.  However, the EPA's
data indicate an emission standard for dioxins/furans would
have no benefit.
     The EPA measured dioxin/furan emissions during the EPA
testing programs.  The dioxin/furan emissions measured from
these sources were much lower than the emissions measured from

                             2-18

-------
 other  source  categories,  such as municipal waste  combustors
 and medical waste  incinerators.  The  concentrations measured
 at secondary  lead  smelters are significantly  lower than the
 dioxin concentrations  in  the emissions  from the best
 controlled municipal waste combustors and medical waste
 incinerators.
     Although the  EPA  believes that the controls necessary to
 meet the  organic HAP emission limitations in  the rule will
 also reduce dioxin/furan  emissions, the Agency does not have
 sufficient data to link emission'levels to specific control
 technologies.  No  technology specifically for dioxin/furan
 emission  control is in use in the secondary lead smelting
 industry.  Therefore,  the EPA has no technical basis to
 regulate  dioxin/furan  emissions from this source category
 beyond  that achieved for  the organic HAP's in general through
 the surrogate THC  emission limit.
 2.3  STANDARDS FOR PROCESS EMISSION SOURCES
     Many comments were received on the proposed emission
 standards for process  emission sources.  These were grouped
 into comments on the MACT floor regulatory interpretation and
 the selection of emission standards for metal HAP's,  organic
 HAP's,   and hydrochloric acid/chlorine (HC1/C12).  Comments on
 the determination  of the MACT floors for each pollutant class
 are discussed within the respective subsections.
 2.3.1  MACT Floor .Regulatory	^Interpretation
     Comment:   One commenter (IV-D-01)  stated that the EPA did
 not have  enough emissions data to select MACT floor emission
 standards for organic HAP's and HC1/C12 emissions.
 Specifically,  the commenter asserts that section 112 £d)  of the
Act mandates the use of data from the 5 best-performing
 sources in the category or subcategory where,  as with
 secondary lead smelters,  there are less than 30 sources.   The
 commenter noted that the EPA does not have data from a
 sufficient number of smelters to meet this criterion for
 organic HAP's  or HC1/C12-   The commenter requested that the
 EPA obtain data from more facilities before proposing MACT

                             2-19

-------
standards for the secondary lead industry for these
pollutants.
     One commenter  (IV-D-22) objected to the EPA's
interpretation of section 112(d) of the Act and the approach
used for the selection of the existing source MACT floor.  The
EPA ranked the control technologies, selected the median
control technology in use at the five best-performing
facilities,  and then determined an achievable emission
limitation for sources using that technology.  According to
the commenter, this interpretation is inconsistent with the
language of the Act, which supposes a ranking based on control
technology only for new sources and a ranking based on
emission limitation for existing sources.  Furthermore, the
commenter noted that, in some cases, the EPA appeared to have
selected the highest (i.e.,  least stringent)  observed emission
level as the proposed emission limit in order to guarantee
that it was continuously achievable by the majority of
sources.
     Response:  Contrary to what is implied by the commenters,
the Act has no minimum data requirements for setting MACT
floors.  Section 112(d)(3)(B)  requires that emission standards
shall be no less stringent than "the average emission
limitation achieved by the best performing 5 sources (for
which the Administrator has or could reasonably obtain
emissions information)  in the category or subcategory for
categories or subcategories with fewer than 30 sources."
     The Act sets statutory deadlines for MACT standards and
vests in the Administrator the discretionary authority to
determine the MACT floors based on the best data and
information reasonably available within these time frames.
The approach for determining the "average emission limitation"
will vary based on the amount of data available within the
time frame of the statutory deadlines set for the MACT
standards in addition to other pragmatic considerations.
Based on the amount of data available for secondary lead
                             2-20

-------
smelters, the EPA used the  following procedure  to meet the
requirements of the Act:
      f"P  Thp ipdnstry was  surveyed to identify the types of
          control technologies being used.
      (2)  The control technology used by  the median of the
          best-performing five sources was selected as the
          floor.
      (3)  The available emissions data from well-operated and
          maintained facilities were then used  to identify an
          achievable emission standard for each such
          technology to represent a MACT floor.
     In cases where there were limited data on  the performance
of a MACT control, the EPA  selected the data representing the
least-stringent emission level as the achievable emission
standard for a properly designed and operated technology in
order to account for normal variability in the  performance of
that technology.  For discussions of how variability was
evaluated, refer to the individual sections on  selection of
emission limits in the proposal preamble and in this BID..
2.3.2  Rationale For ^Metal JSAP Emission Standards
     Eleven commenters {IV-D-02,  IV-D-04,  IV-D-05, IV-D-10,
IV-D-14, IV-D-15,  IV-D-16,  IV-D-20,  IV-D-22, IV-D-24,  and
IV-D-30) commented on the proposed emission standards to
control metal HAP emissions from process sources.
     2.3.2.1  Selection of MACT Floor for Metal HAP's.
     Comment:  Two commenters (IV-D-04 and IV-D-20)  commented
on the EPA's selection of MACT for the control  of metal HAP's.
The first commenter {IV-D-04)  requested that the EPA evaluate
differences in baghouse design,  operating parameters,  and
filter media to determine if any of these are correlated with
differences in metal HAP emission rates.   The commenter also
asked the EPA to evaluate the use of high efficiency
particulate arrestor (HEPA)  filters in establishing the MACT
floor and new source MACT.  The commenter noted that HEPA
filters are in use at battery manufacturing plants and
                             2-21

-------
requested that the EPA consider whether this technology could
be transferred to secondary lead smelters.
     The second commenter  (IV-D-20) also requested that the
EPA evaluate baghouse configuration, including the bag
material and whether the baghouse was fitted with a spark
arrestor, in determining the MACT floor for the control of
metal HAP's.  The commenter recommended that the proposed
standard require spark arrestors to minimize the potential for
burning holes in the filter bags.
     Response:   The EPA collected lead emissions data from
nearly all secondary lead smelters and these data represented
a diversity of baghouse designs, operating parameters, and
filter media.  Within the performance range represented by
well operated and maintained baghouses,  no correlations could
be established between any of these variables and baghouse
performance, as measured by controlled lead concentrations.
     Several secondary lead smelters have HEPA filters
following the sanitary baghouses that control building
ventilation exhausts.  The concentration of lead in exhaust
from these units is not significantly different from the lead
concentration from other sources controlled by a fabric filter
baghouse alone (see docket item II-B-8).   In addition, no
secondary lead smelters are using HEPA filters to control
metal HAP emissions from process emission sources.
     The EPA agrees that spark arrestors can facilitate
baghouse maintenance and improve bag life by minimizing the
potential for holes in bags.  However,  spark arrestors will
not affect the achievable emission levels of a baghouse.  The
monitoring requirements in the rule are designed to detect bag
failures from sparks or any other cause.   The emission
standards and monitoring requirements also have the
flexibility to allow the use of spark arrestors or any other
strategy for maintaining bag integrity.   A requirement for
spark arrestors on all secondary lead smelters would undermine
this flexibility,  would be unreasonable,  and would contradict
                             2-22

-------
 the Ace's antipathy to equipment standards when numerical
 emission standards are available.
     2.3.2.2  Selection of the Lead Emission Limit.  Eleven
 commenters  (IV-D-02, IV-D-04, IV-D-05, IV-D-10, IV-D-14,
 IV-D-15, IV-D-16, IV-D-20, IV-D-22, IV-D-24, and IV-D-30)
 commented on the proposed lead limit of 2.0 mg/dscm
 [0.00087 grains per dry standard cubic foot  (gr/dscf)].
     Comment:  Three commenters  (IV-D-02, IV-D-05, and
 IV-D-15) supported the proposed emission limit of 2.0 mg/dscm
 (0.00087 gr/dscf).  One commenter  (IV-D-05) noted that
 controls are widely available that can reduce emissions to
 levels below the proposed standard.  Another commenter
 (IV-D-24) requested that the lead emission limit should be no
 less stringent than the limits for primary lead smelters and
 that the EPA should be aiming for zero lead discharge.
     Response:  The EPA agrees that technologies are available
 to reduce emissions to levels that are required by the
 proposed standards.  The EPA evaluated the controls available
 in the industry and based the proposed rule on the best
 technologies that are available.  Zero lead discharge is not
 achievable with current smelting or emissions control
 technology.
     Primary lead smelters are subject to another rule-making
 activity and that rule will be based on the technology
 demonstrated for that industry.
     Comment:   Five commenters (IV-D-04,  IV-D-10,  IV-D-20,
 IV-D-22, and IV-D-24)  commented that the lead emission limit
 should be lower than the proposed limit of 2.0 mg/dscm.  Four
 of these commenters (IV-D-04, IV-D-10,  IV-D-20,  and IV-D-22)
 stated that  the proposed lead emission limit was too high
because the  EPA did not base it  on the average emission
 limitation achieved by the five  best-performing sources,  as
required by  section 112(d)  of the Act.   According to one
 commenter (IV-D-10),  the EPA did not follow this guidance in
establishing the proposed lead emission limits because the EPA
evaluated data from all 23  secondary lead smelters.

                             2-23

-------
     One commenter  (IV-D-20) noted that the average lead
emission level of the top five emission tests presented in the
supporting documents is less than 1.0 mg/dscm
(0.00044 gr/dscf).  This commenter also noted that data from
two local secondary lead smelters indicate that emissions
below 0.4 mg/dscm (0.0002 gr/dscf) are consistently achievable
and that these smelters should serve as the basis for the MACT
lead emission limit for new sources.  The commenter requested
that the EPA establish a lead emission limit for new sources
no higher than 0.5 mg/dscm  (0.0002 gr/dscf) and no higher than
1.0 mg/dscm (0.00044 gr/dscf) for existing sources.
     One commenter  (IV-D-10) believed that compliance data
from only the five best-performing sources should be
considered in determining an achievable emission level.
r^_^^_-_^                                 ,jld be sufficient to
establish emission limits because they were obtained from
tests required to demonstrate compliance.  The commenter
asserted that the EPA had made an arbitrary decision on
averaging times and had violated EPA guidance on emissions
testing because the EPA had not used the performance of only
the five best-performing sources to develop emission standards
that were continuously achievable.
     Response:   All smelting furnaces use baghouses to control
metal HAP emissions.  The EPA evaluated different parameters
of baghouse design and operation and could not find
differences in performance that could be correlated with these
differences at the emission concentrations observed.  All of
the baghouses evaluated with lead emissions below 2.0 mg/dscm
were determined by the EPA to be well-designed and well-
operated (see docket item II-B-32).   The EPA determined,
however, that a well-designed and well-operated baghouse will
show variable performance over time, and that this variability
cannot be predicted.  The EPA also determined that a limit
more stringent than 2.0 mg/dscm was not achievable on a
continuous basis with this technology in this source category.
For example, some sources showed higher emissions when tested

                             2-24

-------
by the EPA than indicated by compliance data collected before
EPA testing  (see docket item II-B-32).  Because of this
variability, the best five emissions  tests do not represent
the long-term performance of baghouses.  Therefore, the lead
emission limit was set at a level that was determined to be
representative of the performance of  well-designed and well-
operated baghouse control technology, considering normal
variability in performance.
     Comment:  One commenter (IV-D-20) stated that the
proposed lead emission limit is not protective of public
health.  The commenter noted that process sources are the
largest sources of metal HAP emissions and that the EPA should
propose standards that achieve a reduction in metal HAP
emissions from process sources.  The  commenter noted that the
lead emission limit for lead-acid battery plants is
1.0 mg/dscm (0.00044 gr/dscm)  -- one-half the proposed
emission limit of 2.0 mg/dscm  (0.00087 gr/dscf).  The
commenter stated that an emission limit that is twice as high
as the NSPS emission limit for another lead source is not
sound public policy.
     The same commenter (IV-D-20) also asserted that the
proposed lead emission limit would result in violations of the
1.5 /ig/m3 State and Federal ambient air quality standard for
lead.
     One commenter (IV-D-22)  noted that if the proposed lead
limit of 2.0 mg/dscm (0.00087  gr/dscf) represents the average
emission limitation achieved in practice by the five best-
performing sources in the category,  then it cannot also
represent the best control level of all sources (i.e.,  MACT
for existing sources cannot be the same as MACT for new
sources under these circumstances).   The same commenter also
noted that the limit should be more stringent than 2.0 mg/dscm
if it is achievable and cost-effective; it should not be a
level that all sources can "comfortably" achieve.
     Response:   The commenter (IV-D-20)  provided no technical
basis or analysis for the comment that the proposed lead

                             2-25

-------
emission limits would lead to violations of ambient air
quality standards of 1.5 jig/in^.  All sources currently operate
with MACT for control of metal HAP emissions and, in any case,
process sources are not the largest source of secondary lead
metal HAP emissions.  According to the estimated baseline
emissions presented in table 4-3 of the proposal BID, actual
metal HAP emissions from process sources are less than
20 percent of both process fugitive and fugitive dust
emissions, on an industry-wide basis.  The EPA also believes
that process fugitive and fugitive dust emissions play a
larger role in violations of ambient air standards than stack
emissions because they are ground-level emissions.  The final
rule will also regulate process fugitive and fugitive dust
emissions.
     Lead-acid battery manufacturing is not comparable to
secondary lead smelting, so that the NSPS for lead-acid
battery manufacturing cannot be compared to the secondary lead
smelter NESHAP.  The sources at lead-acid battery
manufacturing plants that are subject to the 1.0 mg/dscm
emission limit are relatively low-temperature sources (grid
casting, paste mixing,  and assembly operations)  with
inherently lower lead emissions.  The lead-emitting sources at
secondary lead smelters are considerably different from those
at lead-acid battery plants.  Lead reclamation at lead-acid
battery plants is performed in melting pots similar to the
refining kettles found at secondary lead smelters and is the
only comparable emission source.  However, the NSPS lead limit
for these lead reclamation facilities is 4.5 mg/dscm, which is
more than twice as high as the final NESHAP lead limit for all
sources at secondary lead smelters.
     The EPA disagrees that MACT for existing sources cannot
be the same as MACT for new sources.  The technology on which
both new and existing source MACT is based is identical.  In
this case, all secondary lead smelters use the same metal HAP
control technology and differences in performance below
2.0 mg/dscm lead emissions cannot be correlated with

                             2-26

-------
 differences  in baghouse design  or  operation.   The  EPA also
 determined that  a  limit more  stringent  than 2.0 mg/dscm was
 mjL ctt-uievaljlt; on  a  continuous  basis with this technology in
 this source  category.  In such  a situation, where  there are no
 discernable  differences in performance,  the performance of the
 best source  is no  different from the average performance of
 the "top" five sources; the performance is the same  once
 normal variability is taken into account.  Based on  the
 monitoring data  collected during the EPA test  program and a
 comparison of the  EPA data to compliance data, the EPA
 believes that the  lead emission limit is characteristic of
 only well-operated and well-maintained  baghouses (see docket
 item II-B-32).   Furthermore,  owners and operators of  secondary
 lead smelters will have to perform regular baghouse
 inspections  and maintenance to  maintain compliance with the
 standard.
     Comment:  One commenter  (IV-D-22)  recommended that the
 EPA include  an alternative percent control limit {99-percent
 reduction)  for the lead emission limit  to minimize conflicts
 with the California Air Toxic Control Measure  {ATCM).
     Response:  There is no conflict between the secondary
 lead NESHAP and the California  ATCM.  The MACT controls
 necessary to meet  the proposed  emission  limit  are capable of
 achieving a  99-percent reduction.  The ATCM,  however,  also
 requires measurement of lead emissions at the  control  device
 inlet as well as the outlet in  order to  calculate the  percent
 emission reduction.
     Comment:  Three commenters (IV-D-15, IV-D-16,  and
 IV-D-30)  recommended that the lead emission limit should be
 less stringent than the proposed limit of 2.0 mg/dscm
 (0.00087 gr/dscf).
     One commenter (IV-D-15)  stated that although the  text of
 the proposal specifically states that this is a lead limit,
many people may infer that this is an attainable limit  for
 total  emissions from baghouses,  even though it is 1/25  of the
 50-mg/dscm (0.022-gr/dscf}  PM emission limit  in the NSPS for

                             2-27

-------
secondary lead smelters.  The same commenter added that -the
limit is so low that EPA reference Method 12 may not be able
to accurately measure emissions at these levels and very long
sampling times may be required to measure emissions.
     Response:  The EPA does not believe that there should be
any confusion between the lead compound emission limits in the
NESHAP and the PM emission limits found in any other
applicable rules.  The rule is clear that the emission
standards are for lead compounds, as measured by EPA
Method 12.  The data available to the EPA, including
compliance data and EPA test data, indicate that lead compound
emissions below 2.0 mg/dscm can be accurately measured using
the sampling times prescribed in EPA reference Method 12.
     Comment:  One commenter (IV-D-16) stated that the data do
not support the proposed lead emission limit of 2.0 mg/dscm
(0.00087 gr/dscf) and do not support the conclusion that the
proposed limit is continuously achievable (emphasis added by
commenter).  The commenter claimed that a single run during a
test of a baghouse at East Penn Manufacturing Company in which
the lead concentration was 3.3 mg/dscm was treated as an
outlier and ignored, although the baghouse appeared to be
operating normally.  The commenter noted that the grain
loading at this baghouse's inlet was lower than average but
the outlet lead concentration was the highest measured,  even
though one would expect the outlet concentration to be also
lower than average.
     The same commenter (IV-D-16) also noted that the preamble
states that a baghouse with a good inspection and maintenance
program "may still emit, on average, an emission stream with
an opacity of 5 or 10 percent," and that such an opacity would
correspond to a lead emission rate of 20 mg/dscm, 10 times
higher than the proposed emission limit.  The commenter stated
that the EPA is indicating that 10-fold exceedances of the
standard can occur despite the application of MACT controls.
     Response:   Compliance with the lead emission standards
will be based on the average lead emissions  measured during

                             2-28

-------
 three  runs  rather  than individual  runs.  The average of the
 three  runs  at  the  East Penn baghouse in question was
 1,8 ir.g/d3crr. and this source would  have been in compliance with
 the proposed 2.0 mg/dscm lead limit.  However, the high
 variability in lead emissions from this baghouse suggested by
 the single  run at  3,3 mg/dscm was  not typical of the
 variability observed at other baghouses tested by the EPA.
 The lead emissions measured in the remaining two runs of that
 test were 1.5  and  0.7 mg/dscm.  In addition, this baghouse had
 a substantially lower inlet PM grain loading  (8.4 mg/dscm
 compared to 30.5 to 104 mg/dscm) and lead removal efficiency
 (70 percent compared to 85 to 99 percent) than the other
 baghouses tested.  The EPA believes that the lower performance
 was associated with the lower inlet grain loading observed at
 this baghouse  compared to others tested by the EPA (see docket
 item II-B-32).
     Contrary  to the statements by the commenter (IV-D-16),
 baghouse efficiency increases as the inlet concentration of PM
 increases.  This is because baghouses rely on the layer of PM
 that collects  on the dirty side of the filter bags (known as a
 filter cake) to enhance filtration.  A lower inlet grain
 loading means  that it will take longer for the filter cake to
 accumulate after a cleaning cycle  and baghouse performance
 will be more variable.  Therefore, the commenter is incorrect
 in stating  that the outlet concentration should also be lower
 when the inlet concentration is lower.
     The statement in the preamble that a baghouse with a good
 inspection/maintenance program may still emit 5 or 10 percent
 opacity was based on the assumption that the baghouse was not
 fitted with a  continuous opacity monitor (COM) or bag leak
 detector and that opacity was only monitored by visual
 observation, which has a lower detection threshold of
 5 percent opacity.   The EPA agrees that without good operation
 and maintenance,  exceedances of the standard can occur despite
 the application of MACT controls.   Therefore,  aggressive
monitoring coupled with a comprehensive inspection and

                             2-29

-------
maintenance program is required by the standard to
continuously achieve maximum performance from those controls
consistent with the proposed lead emission limit.
     Comment:  Another commenter  (IV-D-30) commented that the
proposed lead standard represents a significant tightening of
the current NSPS PM standard for secondary lead smelters
(40 CFR 60 subpart L).  The current PM standard of
0.022 gr/dscf represents a lead emission concentration of
0.0055 gr/dscf, based on the assumption that 25 percent of PM
is lead.  The proposed lead standard, therefore, represents an
85-percent reduction in lead emissions compared to the
existing NSPS, according to the commenter.  To meet the
proposed lead standard, smelters will also need to reduce PM
errn PR-i ons to levels significantly below the level required by
L.JIC iVOi — .
     Response:  The NSPS PM limit, established in 1974, is
based on the use of a venturi-type scrubber to control PM
emissions from process sources (see docket item II-A-32).  All
secondary lead smelters are now using baghouses to control PM
and metal HAP emissions, and these are substantially more
effective than venturi scrubbers.   Consequently, the NESHAP
lead limit reflects the use of the improved control
technology.
2.3.3  Rationale,For Organic HAP Emission Standards
     Eleven commenters  (IV-D-01,  IV-D-02, IV-D-03, IV-D-05,
IV-D-07, IV-D-10, IV-D-1B,  IV-D-20, IV-D-22, IV-D-23, and
IV-D-24) commented on the proposed emission standards to
control organic HAP emissions from process sources.
     2.3.3.1  Selection ofMACT Floor for Organic HAP's.
Three commenters (IV-D-Oi,  IV-D-03 and IV-D-22)  objected to
the method of establishing the MACT floor.
     Comment:  One commenter (IV-D-01) stated that the EPA did
not have enough emissions data to select MACT floor emission
limits for organic HAP's.  Specifically, the commenter noted
that the EPA only has emissions data from a single test at
each of the smelter configurations subject to the proposed

                             2-30

-------
 organic  HAP  emission  standards.  The  cornmenter requested  that
 the  EPA  obtain  data from more  facilities before proposing MACT
 SL.ajuuo.jLuS  for the  secondary lead industry.
     Response:  The EPA evaluated the control technology  in
 use  at each  smelter and determined the floor technology for
 each furnace combination.  The EPA then sponsored a testing
 program  to collect emissions data from six different smelters
 to characterize the performance of the organic HAP controls
 identified as MACT for each furnace configuration.  The
 sources  tested by  the EPA were representative of the smelters
 in each  furnace combination and the data were collected under
 normal operating conditions.   Therefore, the EPA does not
 believe  that it is necessary to collect any additional
 emissions  data.  No additional data were submitted during the
 public comment period to refute the proposition that the
 proposed standards are achievable,
     Comment:  One commenter (IV-D-03) recommended that the
 EPA determine a single MACT floor for all smelter types and
 not differentiate among smelting furnace configurations.  The
 commenter  recommended that blast furnaces be allowed
 additional time (6 months to 1 year)   to comply with emission
 standards  under a single MACT  floor because of the greater
 cost associated with controlling organic HAP emissions from
 these furnaces.
     Another commenter (IV-D-22)  also disagreed with the EPA's
 decision to establish separate new source MACT for each
 furnace  configuration.  The commenter requested that the EPA
 consider requiring new secondary lead smelters to install a
particular furnace configuration as new source MACT because
 there is an environmental advantage to using some furnace
 configurations over others,  according to the commenter.
     Response:   Section 112(d)(1)  of  the Act gives the
Administrator the authority to distinguish among classes,
types,  and sizes of sources within a  category when
establishing standards.   The EPA established separate MACT
emission standards for each furnace configuration because

                             2-31

-------
smelting furnaces differ substantially, based on
configuration, in both emission profile and quantity and
achievable control levels for organic HAP's.  Separate MACT
floors were also established because different furnace types
serve different purposes in the secondary lead smelting
industry.
     The EPA agrees that some furnace types have lower organic
HAP emissions potential than other types.  However,
establishing a single floor that did not differentiate among
furnace types would effectively prohibit blast furnace
operation because blast furnaces have a higher organic HAP
emissions potential than other furnace types.  Blast furnaces
are needed because they account for a significant fraction of
total secondary lead production and account for nearly all the
hard lead produced by secondary lead smelters.  Hard lead
accounts for 40 percent of the lead used in lead-acid
batteries.  A smelter is planned for the future that will use
collocated reverberatory and rotary furnaces; the rotary
furnace will be used in place of a blast furnace to process
reverberatory furnace slag,  but this configuration has yet to
be demonstrated.  Standards cannot be established that
prohibit a technology unless an equally viable alternative
exists at a reasonable cost; this is not the case with blast
furnaces.  Therefore,  the EPA will not establish a single new
source MACT or emission standards that would effectively
prohibit the use of blast furnaces.
     Comment:   One commenter (IV-D-01) asserted that the blast
furnace tested by the EPA represented the best-controlled
facility rather than the median of the top five because the
afterburner controlling the blast furnace had the longest
residence time of those known to the EPA.  Therefore,  this
facility should not have been used to establish the MACT floor
for existing blast furnaces.
     The same commenter also contended that the EPA
underestimated the cost impact of blast furnace controls above
the MACT floor because the EPA only accounted for added fuel

                             2-32

-------
 costs.  According to the  conunenter, the EPA did not account
 for additional gas-stream cooling nor the replacement of
 ejs.j.toLiiiy equipment with equipment made of materials designed
 to withstand higher operating and gas-stream temperatures.
 The commenter requested that the EPA reevaluate the cost
 estimates for the blast furnace controls.
     Response:  Residence time is not the only variable that
 determines afterburner performance; other important variables
 are temperature and mixing  (turbulence).  For example, data
 available to the EPA from other industries demonstrate that an
 afterburner with a fixed  temperature and residence time can
 achieve a range of 70- to 99-percent efficiency by varying
 mixing alone (see docket  item II-B-31).  The EPA has
 temperature and residence time data for several blast furnace
 facilities controlled by  afterburners,  but no data on mixing.
 In the absence of information on mixing, it is difficult to
 precisely identify the best performer and the median of the
 top five performers.  Therefore,  the EPA relied on
 temperature,  residence time, and performance data to select
 the best controlled source and the median of the top five.
     One blast furnace was controlled by an afterburner at
 870 °C (1600 °F)  with a residence time of 1.5 seconds (see
 docket item ll-D-4).  No  other sources exceeded this
 temperature.   Therefore,   this source was selected as new
 source MACT because it appears to be the best-controlled
 facility.  Data on the performance of afterburners from a
 spectrum of industries controlled by afterburners indicate
 that this source should be able to achieve 98 percent control
 of THC (see docket item II-B-31).
     The afterburner at Schuylkill Metals Corporation was
 operated at 700 °C (1300  °F) and 2,5 seconds residence time
 (see docket item II-D-7).   The blast furnace at Schuylkill
Metals was selected for testing because it was considered to
be the median of the five best-controlled blast furnace
 smelters.  During testing, the afterburner achieved 85 percent
 control of THC.  Because  this is  less than the predicted

                             2-33

-------
performance of the best-controlled facility, it cannot also
represent new source MACT.  The Schuylkill facility, however,
is believed by the EPA to represent the existing source MACT
floor based on temperature, residence time, and demonstrated
performance.  Because of the high incremental costs associated
with replacing many existing afterburners, the EPA decided not
to establish standards for blast furnaces above the MACT
floor.
     In estimating cost impacts for blast furnaces controlled
by afterburners, it was assumed that the temperature of a
pre-existing afterburner could not be increased above the
reported operating temperature by more than 100 °C  (180 °F).
If the temperature of an existing afterburner had to be
increased by more than 100 °C to meet the MACT floor or a
control level above the floor, it was assumed that the
afterburner would need to be replaced.  The cost to install a
new afterburner where none previously existed or to replace an
existing afterburner included the costs for additional gas-
stream cooling and more heat-resistant materials.  These
costs, along with added fuel costs, were considered at
proposal and are included in the cost estimates presented in
chapter 6 of the proposal BID.  The EPA continues to believe
that all relevant control costs were included in the control
cost estimates for blast furnaces.
     Comment;  One commenter  (IV-D-07) contended that the EPA
has overestimated the number of reverberatory/blast furnace
facilities that perform gas-stream blending to control
emissions from collocated reverberatory and blast furnaces.
The commenter was aware of only three facilities that perform
gas-stream blending and that one facility was incorrectly
identified by the EPA in the proposal BID as performing
gas-stream blending.
     The same commenter (IV-D-07)  noted that gas-stream
blending would restrict blast furnace operation to periods
when the reverberatory furnace is also operating.  The
commenter also noted that each facility would need a back-up

                             2-34

-------
afterburner to control  the blast furnace when the
reverberatory furnace was not operating.  Furthermore, the
proposed standards would require the blast furnace to be
controlled to levels more stringent than those required for
new blast furnaces at blast furnace-only smelters.
     Two cornmenters  (IV-TD-01 and IV-D-07) commented that the
EPA underestimated the  costs of installing gas-stream blending
at existing reverberatory/blast furnace facilities because
installation would require extensive reworking of the
ventilation and control systems for the furnaces.  One
commenter (IV-D-07) also noted that gas-stream blending makes
it more difficult to control the draft on each furnace and
that it may result in greater fugitive emissions from each
furnace.  The ventilation system at the commenter"s facility
is designed so that each furnace can be independently operated
and controlled and gas-stream blending would not be adopted
regardless of the outcome of the rulemaking,  according to the
commenter.
     Response:  The EPA agrees that it incorrectly identified
one smelter (facility No. 14 in the proposal BID) as
performing gas-stream blending.   Three facilities perform gas-
stream blending.   One of these smelters operates two
reverberatory and two blast furnaces.  The updated
information,  however, does not affect the MACT selection for
new and existing reverberatory/blast furnace smelters.  The
best performing smelter and the median of the top 5 smelters
both employ gas-stream blending.
     The final rule will allow that during periods when the
reverberatory furnace is not operating, facilities with
collocated furnaces may achieve the same THC standards as new
and existing blast furnaces at blast furnace-only smelters.
This will include separate emission tests and monitoring
requirements to demonstrate compliance with the THC standards
while operating only the blast furnace.
     Gas-stream blending has already been demonstrated in the
industry with no apparent problems in controlling process

                             2-35

-------
fugitive emissions.  Nonetheless, the proposed rule does not
require the use of any specific technology to control
emissions,- smelters may use whatever technology they choose in
order to meet the emission standards.
     The cost of installing gas-stream blending has been
re-evaluated since proposal by including the cost of
additional ductwork, a refractory-lined mixing
chamber/afterburner, a larger cooling system, and increased
baghouse capacity.  The revised annual cost to perform gas-
stream blending for an individual- smelter ranges from $120,000
to $350,000, depending on whether a completely new baghouse is
required.  The revised total annual cost for the industry to
perform gas-stream blending is $1.16 million (see docket
item IV-B-3),  The revised cost-effectiveness estimate for the
I_	 .„                     _            organic HAP
reduction for the entire industry.  The estimate presented at
proposal was $510 per Mg of organic HAP's.  However, these
increased costs will not affect the economic impacts of the
standards.
     2.3.3.2  Se.lejgtion, of _the _THC ^mission Limit.   Nine
commenters  (IV-D-01, IV-D-02, IV-D-03,  IV-D-05, IV-D-07,
IV-D-10, IV-D-18, IV-D-20, and IV-D-24)  commented on the
selection of the proposed THC standards for process sources.
     Comment:  Two commenters  (IV-D-05 and IV-D-20)  supported
the EPA's proposed THC emission limits.   The first commenter
{IV-D-05) stated that controls are widely available that can
reduce smelter THC emissions to below the limits in the
proposed standards, but did not provide any data to support
this statement.
     Two commenters (IV-D-02, IV-D-24)  requested that all
smelter types be required to meet the same THC emission limit
of 20 parts per million by volume (pprrtv)  and that higher
limits for blast furnaces or exceptions for "obsolete" furnace
types should not be allowed.
     Response:  Differences in furnace types preclude a
uniform THC standard.   Blast furnaces are not considered an

                             2-36

-------
 obsolete furnace type; they are the only type that can be
 operated independently of other furnace types on an economical
 nasis at a one-furnace smelter.  Blast furnaces also account
 for a significant fraction of total secondary lead production
 and for nearly all of the hard lead produced by secondary lead
 smelters.  Cost-effective controls are not available for blast
 furnaces and rotary furnaces that will allow them to meet the
 20-ppmv THC limit that is achievable for reverberatory/blast
 furnace smelters.
     Comment:  One commenter (IV-D-01) commented that the
 proposed standards for existing and new blast furnace-only
 facilities and reverberatory/blast furnace smelters are not
 based on a sufficient quantity of data and requested that the
 EPA collect more data prior to setting emission limits.  The
 proposed emission limits are based on a single test consisting
 of two or three runs and, according to the commenter,  this
 single test was not adequate to represent normal variation in
 emissions.   The commenter also noted that the EPA has no
 emissions data representing MACT for new blast furnaces and
 has based the proposed emission limit on data from other
 industries.
     Response:   The proposed THC emission limit for existing
 blast furnaces was based on THC emission measurements at a
blast furnace smelter with MACT floor controls.  The test
 consisted of two 3-hour THC emission measurements.   The
 average THC concentrations during the two 3-hour runs were
 300 ppmv and 360 ppmv.  The THC emissions were measured by a
 continuous emission monitor and were recorded as a series of
 one-minute averages.   In both runs,  emissions fluctuated up
 and down several times per hour.   Each run represented several
 furnace charging and tapping cycles and there were no
differences in furnace or control device operation between the
 two runs.   In order to account  for the variation in average
THC concentration,  the higher 3-hour average emission
 concentration was selected as the proposed emission limit.
Furthermore,  the 3-hour average format of the THC emission

                             2-37

-------
standards minimizes the effects of variations in emissions at
a single facility.
     The proposed THC emission limit for reverberatory/blast
furnace smelters was also based on THC emission measurements
at a smelter with MACT floor controls.  The test consisted of
three 3-hour runs.  The average THC concentration during each
run was 3, 5, and 20 ppmv; there were no differences in
furnace or control device operation among the three runs and
the variation among the three runs is not significant at these
low THC concentrations.  The highest 3-hour average THC
concentration was selected as the proposed emission limit to
account for variation in emissions.  No data have been
received since proposal to indicate that the blast furnace or
reverberatory/blast furnace THC emission limits are not
achievable.
     The Administrator used all available information to
develop the proposed THC emission limits for existing and new
blast furnace-only smelters.  The Act does not necessarily
restrict the establishment of MACT emission levels to the
documented performance of plants within the regulated
industry.  The Act allows the Administrator to exercise
judgement and consider the performance of technologies in use
at similar emission sources.  Because no emissions data were
available from the facility that represented new source MACT,
the Administrator considered the documented performance of
afterburners on similar gas streams.
     At the facility representing new source MACT, data were
available on the temperature and residence time of the
afterburner.  Other EPA studies from a broad spectrum of
industries with gas streams of similar characteristics
indicate that this afterburner is capable of 98-percent
efficiency.  This is equal to a controlled THC concentration
of 70 ppmv, based on an uncontrolled concentration of
3,500 ppmv.  The 3,500-ppmv uncontrolled concentration is
based on uncontrolled THC concentrations measured by the EPA
at the outlets to two blast furnaces.  No new data have been

                             2-38

-------
 submitted  to  indicate  that  the proposed  limits are not
 achievable.
         isiit .  Three commenters  (IV-D-03,  IV-D-10 and IV-D-18)
 recommended  that the emission  standards  for existing blast
 furnaces be  the same as for new blast furnaces.  One of these
 commenters  (IV-D-03) claimed that the current proposal would
 allow existing blast furnace smelters to emit excess organic
 HAP emissions of 1,34 million  pounds.  The second commenter
 (IV-D-10) noted that only two  additional facilities would be
 impacted if  the same standards we're adopted for new and
 existing sources and that this does not  appear to be a
 significant  economic impact on a national basis.
     Response :  The EPA considered proposing the same emission
 limits for new and existing source blast furnaces, but
 determined that the incremental control  costs for existing
 sources would be prohibitive.  If existing sources were
 required to meet the same THC emission limits as new sources,
 all but one blast furnace smelter would  need to install new
 afterburners.  The annual organic HAP control costs for blast
 furnace smelters would triple  from $580,000 to $1.7 million on
 a national basis,  but the organic HAP emission reduction
 achieved would only increase by 20 percent, from 580 megagrams
 per year (Mg/yr)  to 690 Mg/yr.  This represents a high
 incremental cost effectiveness of over $10,000 per Mg of
 organic HAP.
     Based on the estimated impacts on operating smelters and
 the high incremental costs,  the EPA decided not to propose
more stringent standards for existing blast furnaces.  No
 information has been received since proposal indicating that
 these impacts have been estimated incorrectly.  In addition,
no data or analysis were provided by the commenter to support
 the claim of excess organic HAP emissions of 1.34 million
pounds.   Therefore, the EPA is not revising the proposed THC
standards for existing blast furnaces.
                             2-39

-------
     2.3.3.3  PlasticRemoval to Control Organic HAP
Emissions.
     Comment:  Four commenters  (TV-D-02, IV-D-03, IV-D-23, and
IV-D-24) recommended that the EPA require smelters to remove
polyvinyl chloride (PVC) plastic from the feed material prior
to incineration to control organic HAP and dioxin/furan
emissions.  One coinmenter (IV-D-02) acknowledged that this
would be an added cost to the smelters but that it was
necessary to protect public health.  Another coinmenter
(IV-D-23) suggested that the. EPA'offer an incentive to the
first battery manufacturer or secondary lead smelter to devise
a way to remove plastics from the feed material.
     Response:  The PVC plastic in battery cell separators is
the primary source of chlorinated hydrocarbons in the furnace
feed.  PVC separators have already been replaced with non-PVC
technology by battery manufacturers.  Consequently, emissions
of HC1/C12 are expected to decrease in the future.  No other
significant sources of chlorine have been identified in
typical secondary lead smelter feed material.
2.3,4  Rationale For Total Chloride Emission Standards
     Six commenters {IV-D-01,  IV-D-Q7, IV-D-14, IV-D-18,
IV-D-22, and IV-D-30)  commented on the proposed emission
standard to control HC1/C3.2 emissions from process sources.
     Comment:  One commenter (IV-D-1'4) requested that the EPA
withdraw the proposed standards for HC1/C12 emissions.  The
commenter stated that PVC plastic separators in lead-acid
batteries, the cause of these emissions, are being phased out
and will disappear from the feed stock to secondary lead
smelters by the time the standard would become effective.
Therefore, requiring controls at this time would not be
cost-effective.  The commenter also requested that a public
notice be published and a public comment period be allowed
before the EPA promulgates any requirements for alternative
methods for achieving the MACT standards.
     Response;  In the preamble for the proposed standards,
the EPA stated, "All smelting furnaces that process broken

                             2-40

-------
batteries are potential sources of HC1 and Cl2 emissions.
Many used lead-acid batteries contain polyvinyl chloride  (PVC)
plastic separators between the battery grids, although the use
of PVC plastic as a separator material has been discontinued
by most battery manufacturers"  (59 FR 29754).
     Information gathered since proposal indicate that the
number of lead-acid batteries in the scrap inventory that
contain PVC plastic separators has declined  from approximately
1 percent of the total available scrap in 1990 to less than
0.1 percent in 1994 (see docket item IV-D-34).  This trend is
expected to continue because PVC separators  are no longer
manufactured in the United States.  No other sources of
chlorides have been identified in the feed stocks to the
smelting furnaces.  Consequently, the EPA also expects
emissions of HC1 and Cl2 to follow a similar decline.
     The EPA estimates that by the 1997 compliance date, the
amount of PVC in the battery scrap inventory will be reduced
to a point such that HC1/C12 emissions will be no more than
10 percent of the emissions measured in the EPA's 1992 testing
program.   Furthermore, HC1/C12 emissions are expected to
continue to decline after this date as the amount of PVC in
the scrap inventory continues to decline.  In light of this
new information,  the EPA has determined that establishing
HC1/C12 emission standards and monitoring requirements would
not be cost effective and would achieve little,  if any,
environmental benefits.  In essence,  the elimination of PVC
plastic separators has resulted in maximum available reduction
of HC1/C12 emissions.   Therefore, the EPA has withdrawn the
HC1/C12 emission standards and associated monitoring
requirements and they are not included in the final rule.
     The EPA published a notice in the Federal Register
announcing the availability of this new information and
requesting public comment.  This notice also indicated that
the EPA was considering withdrawing the HC1/C12  emission
standards and monitoring requirements.   A summary of the
                             2-41

-------
public comments received and the EPA's response to those
comments are presented in appendix A of this BID.
     2.3.4.1  Selection of MACT.  Six commenters  (IV-D-01,
1V-D-07, IV-D-14, IV-D-18, IV-D-22, and IV-D-30) commented on
the selection of MACT to control HC1/C12 emissions.
     Comment:  One commenter (IV-D-18)  stated that the use of
fluxing agents for the control of HC1/C12 is satisfactory if
periodic sampling of the exhaust stream verifies the
effectiveness of the control method and the total life-cycle
environmental risk of the process was not increased.
     Another commenter (IV-D-22) commented that the EPA should
re-examine the proposed MACT for HC1/C12 control for new
sources.  The commenter noted that 99.9-percent control of
^cn'rl erases from hazardous waste incinerators is routinely
B^H-LCVC^	  ..      .       ,.-....          jiiould be considered
in the development of MACT emission standards for new sources
on the basis that this constitutes control of a similar
source.
     Response:  For the reasons discussed in section 2.3.4,
the EPA has withdrawn the HC1/C12 emission standards and
monitoring requirements.   Nevertheless, at proposal the EPA
evaluated the non-air environmental impacts associated with
fluxing and found them to be negligible; therefore, the total
life-cycle environmental risk of the fluxing process probably
would not be increased.
     During the EPA testing program, both fluxing and wet
scrubbers achieved emission levels representing greater than
99 percent control of HC1/C12-   However, the EPA test method
used to measure HC1 and Cl2 emissions has some uncertainty at
low emission levels, i.e., below 30 mg/dscm for total
chlorides.  Given the uncertainty in the test methods
available for measuring HC1/C12 emissions,  there is probably
no significant difference in performance indicated by the
99.9-percent control observed at the hazardous waste
incinerator and the greater than 99-percent control observed
at secondary lead smelters.

                             2-42

-------
      Comment:  Three  commenters  (IV-D-01,  IV-D-14, and
 IV-D-30) qiiestioned whether the EPA had sufficient data to
 conclude that  control of HC1/C12 emissions from reverberatory
 furnaces through fluxing is feasible.  One commenter  (iv-D-01)
 commented that the EPA did not have enough data to conclude
 that  fluxing in reverberatory furnaces could be based on the
 use of  fluxing in rotary furnaces.  The same commenter also
 noted that the question of whether fluxing can consistently
 remove  HC1 from the exhaust stream at the proposed levels is
 unresolved.
      Four commenters  (IV-D-01, IV-D-07, IV-D-14, and
 IV-D-30) stated that  improper fluxing can interfere with
 efficient furnace operation and lead production.  One
 commenter (IV-D-01) added that the addition of too much flux,
 especially limestone, can reduce the efficiency of the furnace
 over  time and that the addition of soda ash is also
 problematic because of the risk of an explosive reaction.
 Another commenter  (IV-D-07) added that the metallurgical
 consequences of fluxing makes this infeasible for the control
 of HC1/C12.
      One commenter (IV-D-01)  stated that it is the experience
 of secondary lead smelters that the amount of chlorides
 available to be emitted is variable among smelters and over
 time.   The same commenter also contended that the EPA did not
 account for differences in furnace configuration,  feed
materials,  fluxing materials,  and material handling.
According to the commenter, the EPA also did not account for
 the fact that fluxing agents are inherent in some feed stocks.
The commenter requested that the EPA consider these variables
before proposing HC1/C12 emission standards.
     Response:   The EPA measured HC1/C12 emissions from all
 furnace types in use today and these data indicated that the
addition of fluxing agents to the feed material could be an
effective control of HC1/C12  emissions.   These tests included
a test at a reverberatory furnace.   During this test,  fluxing
agents were added to the furnace charge materials  in the form

                             2-43

-------
of powdered soda ash in varying amounts.  As the amount of
fluxing increased, HC1/C12 emissions were found to decrease.
TI"ic=e data were used to establish that fluxing is an effective
HC1/C12 control for reverberatory furnaces.  The data
available indicated that the proposed emission levels could be
achieved through fluxing.  If fluxing agents are inherent in
some feedstocks, then not as much additional fluxing materials
will need to be added to achieve compliance with the proposed
standards.
     All of the data obtained by 'the EPA indicating that
fluxing can control HC1/C12 emissions were obtained during
periods of normal furnace operation and, therefore, represent
demonstrated control technology.  No other data were provided
to indicate how normal levels of fluxing can interfere with
furnace operation or efficiency.  In addition,  no information
or data were provided to explain how an explosive reaction may
occur from the addition of soda ash to a furnace; soda ash is
a commonly used fluxing agent.
     However, other data obtained by the EPA since proposal
indicate that the method by which the fluxing agents are added
to the feed material may be an important factor in determining
the success of fluxing.  The operators of a blast furnace
smelter measured HCl emissions during normal furnace operation
and at elevated levels of fluxing and observed no significant
HCl emissions control  (see docket item IV-D-33).   At this
smelter', flux was added to the blast furnace in the form of
limestone gravel.  At the blast furnace tested by the EPA,
powdered fluxing agents were mixed with flue dust before
charging the flue dust back to the furnace.  The flue dust
contains chlorides that are recycled to the furnace before
being emitted as HCl.  The EPA believes that the close
physical association between the chlorides in the flue dust
and the fluxing agents at the smelter tested by the EPA
resulted in greater HCl control compared to the other blast
furnace smelter.  The EPA is not aware of any other blast
                             2-44

-------
 furnace smelters that combine flue dust and fluxing agents in
 the  same manner as at the smelter tested by the EPA.
     Tile I-iCl/Cl^ data obtained since proposal indicate that
 blast-furnace smelters would need to either install a new flue
 dust and flux handling system or install a wet acid gas
 scrubber in order to meet the proposed HC1/C12 emission
 standards.  The cost and economic impacts of either option
 would be significantly greater than those estimated at
 proposal.  The estimated annual cost of a wet scrubber is
 about $850,000 for a single smelter and seven blast furnace
 smelters would be affected.  The cost impact of a new flue
 dust and flux handling system would be less but would still be
 significant,
     In any case,  as discussed in section 2.3.4, the EPA is
 not promulgating the HC1/C12 emission standards and monitoring
 requirements.
     Comment:  One commenter (IV-D-01) pointed out that
 desulfurization and dechlorination of the feed material is not
 a feasible HC1/C12 control option for blast furnaces because
 finely divided materials (necessary for desulfurization and
 dechlorination)  cannot be fed into blast furnaces.
     Response:   Data collected by the EPA indicated that
 certain fluxing practices could potentially control HC1/C12
 emissions from blast furnaces and dechlorination would not be
 necessary.  However,  as discussed in section 2,3.4, the EPA is
 not promulgating the proposed HC1/C12 emission standards and
monitoring requirements.
     2.3.4.2  SelectionofEmissionLimits.
     Comment;  Three commenters (IV-D-02,  IV-D-05, and
 IV-D-20)  supported the proposed HC1/C12 emission limit of
 15 mg/dscm.   One commenter (IV-D-05)  noted that controls are
widely available that can reduce HC1/C12 emissions to levels
below the limits specified in the proposed standards.   Another
 commenter (IV-D-20)  noted that a test at the scrubber outlet
 of a smelter located in California's South Coast Air Quality
                             2-45

-------
Management District  (SCAQMD) showed that HC1 emissions were
below detectable levels.
     Response:  No data were provided to support the claim
that controls are widely available to reduce HC1/C12 emissions
from secondary lead smelters to levels below the limits
specified in the standards.  As discussed in section 2.3.4,
the EPA has withdrawn the proposed HC1/C12 emission standards
because of the decrease in HC1/C12 emissions arising from the
decrease of PVC in the battery scrap inventory.  Consequently,
requiring HC1/C12 emissions controls would achieve little, if
any, environmental benefit.  Furthermore, the test methods
available to measure HC1 and Cl2 emissions have some
uncertainty, specifically a negative bias, at emission levels
below 30 mg/dscm for total chlorides; therefore, the emissions
from the smelter in the SCAQMD may not have been as low as
indicated by the test results.
     Comment:  One commenter  (IV-D-Ol) contended that the
proposed emission limit of 15 mg/dscm is too stringent and is
based on limited testing at only two facilities.  The
commenter added that sufficient data are not available to
demonstrate that the limit can be met on a continuous basis by
all furnace types using the MACT controls on which the EPA
based its cost and economic impact estimates.  For these
reasons, the commenter proposed that the HC1/C12 emission
limit be increased to 50 mg/dscm.  No data were submitted to
support the commenter's proposed HC1/C12 emission limit of
50 mg/dscm.
     Response:  The EPA measured HC1/C12 emissions at four
secondary lead smelters and these tests represented all
furnace types.  Three test runs lasting a minimum of one hour
per run were conducted at each source.  The data collected by
the EPA indicated that fluxing practices could potentially
control HC1/C12 emissions and that control could be achieved
continuously.  However, as discussed above in this section,
data obtained by the EPA since proposal indicate that fluxing
may not adequately control HC1/C12 in all cases, depending on

                             2-46

-------
how  the  flux  is added  to  the furnace.  Regardless, as
discussed  in  section 2.3.4, the EPA plans to withdraw the
proposed HC1/C12 emission standards.
     2.3.4.3  Cost Impacts.
     Comment::  Four commenters  (IV-D-01, IV-D-07, IV-D-14, and
IV-D-30) stated that the  EPA underestimated the  cost impacts
of the proposed HC1/C12 emission standards because more
facilities than predicted by the EPA would need  to install
scrubbers.  They commented that the EPA's cost impact
estimates are based on fluxing, but the ability  of facilities
to use fluxing rather than scrubbers to control  HC1/C12
emissions has been overestimated by the EPA.
     The same commenters  noted that the EPA's data indicate
that the cost of scrubbers would be prohibitive, with capital
costs of $1.7 million and annual costs of $850,000 at a
.typical  smelter operating a reverberatory furnace.  The
commenters also pointed out that the scrubbers would only be
useful for a few years because HC1/C12 emissions would decline
anyway from the phase-out of PVC separators in lead-acid
batteries.
     Response:  As discussed in 2.3.4.2, data initially
collected by the EPA indicated that certain fluxing practices
could potentially control HC1/C12 emissions on all furnace
types.  Subsequent data has indicated that such  control may
not be effective in all cases depending on how the flux is
added to the furnace.   The EPA agrees that scrubbers would be
a more costly control option than fluxing if they were
necessary.  However,  for  the reasons described in
section 2.3.4, the EPA is not promulgating the HC1/C12
emission standards.
     2.3.4.4  Other Impacts.
     Comment:   Two commenters (IV-D-07 and IV-D-14)  claimed
that the EPA underestimated the water quality impacts of the
proposed standards.  They both argued that more smelters than
estimated by the EPA would install scrubbers that would
generate wastewater streams.

                             2-47

-------
     The second commenter (IV-D-14) also noted that both
scrubbers and fluxing would have other significant adverse
environmental impacts that would offset the benefits of
controlling HC1/C12 emissions.  According to the commenter,
both options would result in large amounts of solid waste,
either scrubber sludge or slag, that could only be landfilled.
Large amounts of energy would also be needed to operate the
scrubber, treat the blowdown, and haul the solid waste to a
landfill.  Fluxing would increase the energy needed to operate
the furnace because furnaces would be less efficient.  The
commenter noted that the EPA is required to consider non-air
quality health and environmental impacts under section 112 of
the Act and cross-media impacts under the Pollution Prevention
Act of 1990 and the EPA's own Pollution Prevention Strategy.
     ^		                              jielters than
estimated at proposal would need to install scrubbers in order
to meet the proposed HC1/C12 emission limits.  Data obtained
since proposal indicate that fluxing as an HC1/C12 control may
not be as effective as estimated at proposal.  Therefore, the
water consumption, wastewater treatment, and scrubber sludge
disposal impacts would be greater than estimated at proposal.
However, as described in section 2.3.4, the EPA is not
promulgating the HC1/C12 emission standards and these impacts
would not occur under the final rule.
2.4  STANDARDS FOR PROCESS FUGITIVE SOURCES
     Six commenters (IV-D-01, IV-D-02, IV-D-03, IV-D-07,
IV-D-20, and IV-D-30)  commented on the proposed emission
standards for process fugitive sources.  Two commenters
(IV-D-02 and IV-D-20)  supported the EPA's proposed standards
for process fugitive sources without further elaboration.  One
commenter (IV-D-03) supported the EPA's determination of MACT
for process fugitive sources to control metal HAP and organic
HAP emissions.  The remaining comments on the proposed process
fugitive emission standards are summarized in the following
sections.
                             2-48

-------
2.4.1   Identification Of MACT Floor for.__Pro_cess__Fugitive
Sources
     Comment:  One commenter  (IV-D-14) disagreed with the
approach the EPA used to develop the MACT floor for control of
process fugitive emissions, in particular the EPA's reliance
on the  Cooperative .Assessment Program,Manual for the Secondary
Lead Industry  (CAP Manual) .  According to the commenter, the
EPA has not demonstrated that the CAP Manual specifications
establish the MACT floor.  The commenter noted that the CAP
Manual was developed in 1983 to ensure workplace safety and
was never officially issued by the Occupational Safety and
Health Administration (OSHA).  The commenter also asserted
that the EPA has not demonstrated that the CAP Manual
specifications are still state of the art and whether there is
a correlation between workplace emissions and ambient
emissions.  The commenter also asserted that it is
inappropriate to develop standards for ambient air emissions
based on control methods that were developed for the
workplace.
     The same commenter {IV-D-14)  also argued that the EPA has
not demonstrated that the proposed specifications reflect the
performance at the five best-controlled secondary lead
smelters.   The commenter noted that the EPA measured face
velocities at only one smelter and assumed that all other
smelters were comparable.
     Response:  The basis for the process fugitive sources
MACT floor is not the CAP Manual.   The CAP Manual was intended
only as a reference document for descriptions of commonly
applied practices;  other references,  such as the Industrial
Ventilation Manual,  recommend similar practices.   The EPA
measured face velocity at several sources at one smelter as an
indicator of process fugitive hood control performance (see
docket item IV-A-2}.   The face velocities were 300 feet per
minute  (fpm)  at the slag tap hood,  330 fpm at the refining
kettle hood,  and 1,000 fpm at the drying kiln transition piece
hood.   Because the face velocity at the slag tap of 300 fpm

                             2-49

-------
was less than the proposed velocity of 350 fpm, the final rule
will require a face velocity of 300 fpm for slag and lead tap
hcc-in, furna.cc charging hoods, and agglomerating furnace
hoods.  However, the required face velocity for the drying
kiln transition piece hood will remain at 350 fpm because this
face velocity has been demonstrated to be achievable.  The
face velocity recommended for these types of sources by the
IndustrialJVentilation^Manual is 200 to 500 fpm.
     It was not necessary to conduct measurements at more than
one smelter because all smelters 'have similar hood
configurations and baghouse capacity dedicated to controlling
process fugitive emissions, after accounting for differences
in smelter size.  Therefore, all smelters have the capacity to
achieve similar face velocities to those at the smelter tested
by the EPA.
2.4.2  Selection of the Lead Emission Limit forProcess
Fugitive Sources
     Comment:   One commenter (IV-D-30)  argued that emission
standards for process fugitive sources would be redundant
because OSHA already requires lead-in-air and personnel
monitoring.
     Response:  The lead-in-air and personnel monitoring
required by OSHA are intended to minimize worker exposure to
lead emissions.   The proposed process fugitive emission
standards, on the other hand, are intended to minimize metal
HAP emissions  to the ambient air.  These are separate
objectives and separate exposure pathways.  To illustrate,
OSHA regulations contain no emission limit on captured process
fugitive emissions that are vented to the atmosphere.  By
contrast, the  NESHAP imposes a lead emission limit of
2.0 mg/dscm on process fugitive emissions.
2.4.3  Selection of Equipment Standards
     Comment:   Two commenters (IV-D-07 and IV-D-14)  requested
that the proposed hood requirements for process fugitive
emission sources be waived for smelters that have all process
equipment located in buildings that are ventilated to a

                             2-50

-------
 control  device.  Additional  regulation  of hood performance  is
 unnecessary  in  these  situations, according  to the  commenters.
     Response:  The EPA agrees and those sources with all
 process  equipment  in  controlled buildings are not  subject to
 the hood requirements in the final rule.  However, as an
 alternative, the EPA  has established a  doorway velocity
 requirement  (250 fpm)  for buildings ventilated to  a control
 device and a lead  emission limit  (2.0 mg/dscm) for the control
 device.  These  requirements  are necessary to ensure that the
 building represents control  equivalent  to a hood ventilated to
 a baghouse.  The doorway velocity required  in the  final rule
 has been confirmed by measurements at a smelter with a
 representative  enclosure and ventilation system  (see docket
 item IV-A-2).
     Comment:   One commenter (IV-D-07)  stated that hood face
 velocity requirements of 150 fpm for refining kettle hoods and
 250 fpm  for all other process fugitive  source hoods are
 adequate and effective in minimizing fugitive emissions.
 Consequently, the  commenter  requested that the EPA revise the
 proposed hood velocity requirements (250 fpm and 350 fpm,
 respectively) for  these sources.
     Three commenters  (IV-D-01,  IV-D-07, and IV-D-14)
 requested that  the EPA withdraw the proposed volumetric flow
 rate requirement for  refining kettles.  Two of the commenters
 (IV-D-01 and IV-D-07)  pointed out that current hood designs
 are able to effectively control emissions at lower volumetric
 flow rates than specified in the proposed rule.
     One commenter {IV-D-14}  argued that both the volumetric
 flow rate and the  face velocity requirements are not feasible
 and would compromise the refining process.   According to the
 commenter,  these rates would pull heat away from the kettle
 and cause solids to form on the surface of the molten lead;
 these solids would compromise the refining process and
 contribute to fugitive dust emissions.   The commenter also
noted that the EPA has not demonstrated that the proposed
requirements could be continuously maintained because

                             2-51

-------
ventilation rate varies greatly as a 'result of opening and
closing the hood doors during the refining process.
     Response:  The hood velocity requirements proposed by the
EPA provide greater control than those proposed by the
commenters  (IV-D-01 and IV-D-07).  The Industrial Ventilation
Manual also states that these velocities are adequate and
achievable.
     The EPA agrees that face velocity is the only relevant
factor in a total-enclosure-type hood.  Therefore, the
volumetric flow rate requirement for refining kettles have
been deleted; however, the face velocity requirement have not
changed.  The specified face velocity has been demonstrated
with no adverse effects on the refining process.
     The initial face velocity determination must be made with
all access doors open and will,  therefore,  measure the minimum
face velocity that a smelter will achieve;  subsequent face
velocity should be greater since not all doors will be open at
one time.  Thereafter, the smelter should be in compliance
under all operating conditions.
2.5  STANDARDS FOR FUGITIVE DUST SOURCES
     Six commenters (IV-D-02, IV-D-04, IV-D-10, IV-D-14,
IV-D-15, and IV-D-30)  commented on the proposed emission
standards for fugitive dust sources.
2.5.1  Selection Of MACT for FugitiveDustSources
     Comment:  One commenter (IV-D-15) was concerned with the
selection of total enclosure and ventilation to a baghouse as
an equivalent MACT for controlling fugitive dust sources.  The
commenter noted that enclosures for high-temperature sources,
such as those found at smelters, require many more air
exchanges than recommended by the American Society of Heating,
Refrigeration, and Air Conditioning Engineers to maintain
reasonable temperatures in the workplace.   This requires
larger fans and control devices with higher capital and
operating costs.  The commenter noted that the EPA needs to
account for this factor if the cost of this alternative is
estimated.

                             2-52

-------
      Response:  Total  enclosures with reasonable workplace
 temperatures have been demonstrated  in all  climates.  The
 standard does not specify the use of total  enclosures as the
 only  control option.   Rather, the NESHAP provides the total
 enclosure  option as an equivalent MACT alternative to a
 partial-enclosure option.  The cost estimates  for this
 standard are based on  the least-cost alternative  (i.e.,
 partial enclosure and  good housekeeping).
      Comment:  The same commenter  (IV-D-15) agreed with the
 EPA's finding that the use of dust suppressants or sealers on
 storage piles is often a viable fugitive dust  control, but
 pointed out that the chemicals added must be compatible with
 the process chemistry.  This factor may require the use of
 more  expensive dust suppressants.
      Response:  Facilities may comply with  the standard using
 water; alternatively,  they may use chemical suppressants
 compatible with their  production processes.  The suppressant
 costs are based on water usage and include  runoff collection
 and treatment.  Increased costs from using more expensive
 chemicals would be offset by savings from decreased water
 usage and treatment.   However, the standard does not specify a
 particular dust suppressant and facilities may use water if
 they prefer.
     Comment:  Another commenter (IV-D-10)  disagreed with the
 EPA's determination that partial enclosures coupled with
 pavement cleaning are  as effective as total enclosures
ventilated to baghouses for the control of fugitive dust
 emissions.   The commenter recommended that, at a minimum,
 total enclosures ventilated to a baghouse represent new source
MACT.
     Response:   No data have been provided to indicate that
 total enclosures vented to baghouses are more effective than
partial enclosures coupled with good housekeeping and pavement
 cleaning.   The EPA conducted a technical analysis of fugitive
dust control measures at smelters (see docket item II-B-28).
 It concluded that partial enclosures with appropriate wetting

                             2-53

-------
and pavement cleaning cost less and are as effective as total
enclosures in controlling fugitive dust emissions when coupled
with monitoring and recordkeeping to ensure that these
activities are performed.
     Comment;   One commenter  (IV-D-02) supported the proposed
standards to control fugitive dust emissions but requested
that the standard operating procedures (SOP) manual be made
available for public review and comment,  in addition to review
by the EPA.
     Another commenter  (IV-D-10) also supported the
requirement for sources to develop an SOP manual to control
fugitive dust.  However, the commenter requested that the EPA
establish minimum standards of acceptability to guide States
when -reviewing these plans, such as frequency of pavement
.--Lcai-L^ 	,                              compliance deadlines
for paving fugitive dust areas.
     Response.-  The SOP manual will be available for public
review and comment during the Title V permitting process that
each facility must undergo.  Each facility must demonstrate
compliance with all provisions of the rule, including the
requirement to pave all traffic areas, within two years of
publication of the final rule.
     The proposed rule specifies the minimum frequency of
pavement cleaning (twice per day).  The final rule specifies
that the storage piles  (including the active face)  be kept
sufficiently moist to prevent the formation of dust.
2.5.2  SelectionOfFugitive Dust Control Standards
     Comment:   One commenter  (IV-D-04) recommended that the
required frequency of pavement cleaning should be equal to the
frequency of pavement cleaning at the best-performing sources.
In addition,  the commenter requested that all vacuuming
equipment be fitted with HEPA filters to avoid re-emitting
metal HAP's.
     Response:  The minimum frequency of pavement cleaning
specified in the rule (twice per day for all areas)  is equal
to that achieved at the best-performing sources, according to

                             2-54

-------
 the MACT  definition  in the Act.  The rule allows pavement
 cleaning  equipment to be chosen by the facility; emissions
 from  that equipment  would be addressed during the SOP approval
 and Title V operating permit application process, which
 provides  for public  comment.  The SOP would be approved only
 after demonstrating  that MACT fugitive emission control
 measures  have been included.
      Comment:  One commenter (IV-D-14) requested that the EPA
 exempt from the proposed fugitive dust control requirements
 those areas that are subject to tertiary air controls under
 the Resource Conservation and Recovery Act  (RCRA) subpart DD
 containment building standards in order to prevent
 inconsistent regulation of these areas.
      The  same commenter also requested that the EPA allow
 smelters  with all fugitive dust sources enclosed in buildings
 ventilated to baghouses followed by HEPA filters to
 demonstrate compliance with the proposed lead emission limits
 through alternative means, such as certifying that the HEPA
 filters are properly functioning.  The commenter argued that
 the proposed requirement for annual stack testing would have
 no benefit at these facilities because emissions are below
 detectable levels.
      Response:  In the revised rule,  the EPA has included an
 alternative means of compliance for those areas that are
 subject to tertiary air controls under RCRA subpart DD
 containment building standards because these controls are
 equivalent to the MACT level of control.
      The  EPA data indicate that sources controlled by HEPA
 filters have measurable lead emissions; however,  lead
 emissions below detectable levels would indicate compliance
with  the  lead emission limit.   The final  standards do not
 require an annual lead test.   However,  each baghouse must be
 fitted with a continuous particulate monitor.   (See
 section 2.7 of this BID for more information on the final
metal HAP monitoring requirements.)
                             2-55

-------
     Comment;   One commenter (IV-D-30) questioned the need to
establish emission standards for fugitive dust sources because
GSKA auaudaiua already require -extensive personnel lead-in-air
workplace monitoring and ambient lead concentrations are
regulated and monitored under the lead NAAQS.
     Response:  The NAAQS monitoring requirements are only
intended to ensure compliance with ambient air .quality
standards.  The Act specifies that NESHAP require MACT,
regardless of specific ambient air levels.  Compliance with
OSHA standards would not demonstrate that MACT emission levels
are being achieved because OSHA does not specify emission
standards for ventilation systems, nor guard against emissions
to ambient environment.  Furthermore, the NESHAP also controls
all metal HAP's but the NAAQS only regulates lead.
2.6  TEST METHODS AND SCHEDULE
     Nine commenters (IV-D-01,  IV-D-03, IV-D-06, IV-D-07,
IV-D-14, IV-D-15, IV-D-22, IV-D-29, and IV-D-30) submitted
comments on the proposed test methods and schedule.  Comments
requesting clarification or additional information for the
tests for all classes of pollutants are discussed below.
Comments applicable to the tests for each class of pollutants
are presented in the following sections.
     Comment:   One commenter (IV-D-06) requested that whenever
Method 3 is designated as an applicable test method, the EPA
change the term "gas analysis"  to "gas analysis for the
determination of dry molecular weight" to more accurately
convey the intent of the requirement.
     One commenter {IV-D-06) recommended that the EPA present
the conversions required in § 64.547(d) of the proposed rule
for correction to 4 percent carbon dioxide in a mathematical
form rather than a text form in order to avoid confusion.
     One commenter {IV-D-22) requested that the EPA provide a
simple procedure for adjusting test methods currently used to
demonstrate compliance with existing State or local standards
in lieu of requiring EPA test methods.  The commenter stated
that this measure would save facilities the expenditures

                             2-56

-------
required for additionally conducting the EPA test methods.
The commenter also cited an existing State PM test method
 (California Air Resources Board Method 5) that could be
modified to measure lead at a cost 'savings of $2,000 to
$3,500 per test.  The same commenter also asked that the
proposed rule "include standard methods of calculation for
determining emission stream content  (for example, total
hydrogen chloride, chlorine, or hydrocarbon)."
     Response:  The EPA agrees that the term "gas analysis"
should be changed to "gas analysis for the determination of
dry molecular weight" and has made this change in the final
rule.  The EPA has also added the conversions for the .
correction to a constant carbon dioxide concentration in a
mathematical form, as well as in a text form.  However,
methods for determining emission stream contents are already
included in the Code of Federal Regulations under the
applicable test methods (i.e.,  Method 25A for THC and
Method 26 for HC1 and Cl2) .  In addition, as discussed in
section 2.3.4,  the HC1/C12 emission standards have been
withdrawn.
     The general testing requirements contained in the General
Provisions  allow for the use of an alternative test method
provided that the method is validated according to the
procedures  in EPA Method 301.  Section 63.7 of the General
Provisions  (40 CFR 63)  specifies the procedures for obtaining
approval for the use of an alternative test method.
2.6.1  Lead and Metal HAP's
     Comment:  One commenter (IV-D-06)  asked that the EPA
delete the  references to Method 12 in § 63.543(a), 63.544(d),
and 63.545(e)  because it is already prescribed in the
applicable  test method in § 63.547(a).   This commenter also
requested that the EPA correct  reference to § 63.545 (f)
to § 63.545(e)  in the first sentence of § 63.547(a) because
§ 63.545(e)  is the paragraph that contains the statement of
the emissions standard.
                             2-57

-------
     Response:  The EPA agrees and has made the requested
changes where they are still applicable in the revised rule.
     Coiuraent;  One commenter  (IV-D-15) stated that Method 12
may not be accurate or may require very long sampling times at
the proposed lead emission limit of 2.0 mg/dscm.  The
commenter, however, did not make any specific recommendations
for changing the standards.
     Response:  The EPA test data and compliance data indicate
that reference Method 12 accurately measures lead emissions
below 2.0 mg/dscm using the sampling times prescribed in the
method.
     Comment;  Five commenters (IV-D-01, IV-D-03, IV-D-14,
IV-D-29, and IV-D-30)  submitted comments on the need for an
annual lead compliance test.
     One commenter (IV-D-03) supported the proposed
requirement for annual compliance tests of lead emissions for
process fugitive and fugitive dust sources.  Another commenter
(IV-D-30) requested that the EPA waive the proposed annual
compliance test requirement for lead; the testing costs could
not be justified because the lead NAAQS requires the facility
to continuously monitor ambient lead concentrations.
     Two commenters (IV-D-01 and IV-D-29) requested that the
EPA waive the annual compliance test requirement under certain
circumstances.  One commenter (IV-D-29) recommended that the
EPA waive the annual compliance test requirement for sources
that demonstrate both (1)  control device effectiveness by
complying with the proposed COM requirements; and (2) no
significant changes in lead air concentrations from the data
collected and reported under NAAQS.  The other commenter
(IV-D-01) proposed that sources could conduct the compliance
test every 2 years if previous tests demonstrated compliance
at a level less than half of the lead standard.
     Response:  The NAAQS monitoring requirements are intended
to ensure compliance with ambient air quality standards; NAAQS
monitoring cannot demonstrate compliance with numerical MACT
emission limits.

                             2-58

-------
     The proposed  rule would have required an annual lead
 compliance  test .for process fugitive sources, but not for
 process sources.   For process sources, the proposed rule would
 have required the  use of a COM to monitor compliance with a
 site-specific opacity limit.  However, based on comments
 received on the proposed monitoring requirements and described
 in section  2.7, the final rule requires bag leak detection
 systems for all baghouses controlling process, process
 fugitive, and fugitive dust sources.  Continuous opacity
 monitors have been removed as a monitoring option in the final
 rule because they  only provide a limited measure of process
 source baghouse performance and do not have the sensitivity to
 indicate the performance of process fugitive or fugitive dust
 baghouses.  In contrast, the more sensitive bag leak detection
 systems provide a  continuous measure of baghouse performance
 for both process and process fugitive sources.  The final rule
 also requires owners and operators to take specific corrective
 actions if  the bag leak detection system indicates any
 degradation in baghouse performance.  The EPA has withdrawn
 the annual  lead test requirement because the final monitoring
 requirements are sufficient to detect -and correct any
 degradation in baghouse performance.  Therefore,  the benefits
 of annual lead compliance tests would be very limited.
     Lead emissions below 2.0 mg/dscm,  the lead standard,  tend
 to be variable.   If the EPA tests and compliance data from
 secondary lead smelters had indicated that emissions
 significantly below this level could be achieved on a
 consistent basis,  then a lower emission standard would have
been established,  rather than allowing smelters to skip
 compliance tests.
 2-6-2  THC and Organic HAP's
     Comment:  One  commenter (IV-D-06)  requested that the EPA
designate Method 1 in § 63.547 (c) (1)  of the rule as an
applicable method  for selection of the sample location when no
traverse sample points are needed (i.e.,  when only a single
                             2-59

-------
sample point is needed) when determining compliance with the
THC emission limits in § 63. 543 (c),  (d) , and  (e) .
     P.esponse ;  "The EPA agrees with the eommenter and has made
the requested change in the revised rule.
     Comment :  One eommenter (IV-D-06) asked the EPA to  -
include a procedure or equation in § 63.547•><=• rule, as it provides no additional benefit.
     Comment :  One eommenter  (IV-D-06) recommended that the
equation for calculating total HC1/C12 emissions in
§ 63. 547 (b) (5)  of the proposed rule be numbered and include a
conversion factor for English units as well as metric units.
     Response:   The EPA agrees that numbering the equations
would clarify the rule and this will be done for the equations
in the final rule.  The EPA also agrees that providing
equivalent english units would facilitate calculating total
HC1/C12 emissions.  However, as discussed in section 2.3.4,
the EPA is not  promulgating the HC1/C12 emission standards.
     Comment :  Two commenters (IV-D-Ol, IV-D-07) asked that
the EPA allow for testing by Method 26 instead of Method 26A
for configurations that do not have a wet scrubber.  They
stated that Method 26A, which costs more than Method 26, is
not needed for these configurations because of the absence of
wet scrubber mist emissions.
     Response :   The EPA agrees that Method 26 is an
appropriate test method in the absence of wet scrubber mist
emissions.  However,  as discussed in section 2.3.4, the EPA is
not promulgating the HC1/C12 emission standards and an HC1
test method (i.e., Method 26) is, therefore, not required.

                             2-60

-------
 2.6.4   FaceVelocity
     Comment:  Two commenters  (IV-D-01 and  IV-D-07) requested
 nhac tine ZPA allow the use of velometers at hood face openings
 as an alternate method for determining face velocities.  They
 stated  that accessing the ductwork for hoods for velocity
 testing using the proposed method  (Method 2) may be difficult
 at some facilities.  One commenter  (IV-D-07) stated that
 kettle  ventilation systems are often installed in close
 quarters with many bends and transitions in the duct that
 preclude locating a test port that meets the spacing
 requirements of Method 2.  One commenter (IV-D-01) added that
 the proposed method is more expensive than velometers.  No
 information was given by either commenter on the relative
 accuracy of velometers compared to the proposed method.
     Response:  The EPA agrees to allow the use of velometers
 as an alternate method for determining face velocities and has
 incorporated these changes in the final rule.  The
 specifications for anemometers (velometers)  were adapted from
 EPA Reference Method 14  (40 CFR 60, appendix A).  The general
 testing requirements contained in § 63.7 of the General
 Provisions also specify procedures for applying for the use of
 an alternative test method.
 2.7  MONITORING REQUIREMENTS
     Many comments were received on the proposed monitoring
 requirements.  Several commenters supported a policy that
 continuous emission monitors (CEM's)  be required whenever
possible;  these comments are discussed in more detail in the
 following paragraphs.   Other comments related specifically to
 the proposed monitoring requirements for metal HAP's,  organic
HAP's,  and HC1/C12;  these comments are discussed in
subsections for each pollutant class.
     Comment:   Four commenters (IV-D-02,  IV-D-03,  IV-D-05,  and
IV-D-23) supported the use of CEM's in general and requested
that they be required for monitoring all emission sources at
secondary lead smelters.   One commenter (IV-D-03)  stated that
CEM's are needed on stationary sources because they are more

                             2-61

-------
likely co detect air quality violations than do on-site
inspections.  This commenter and two others  (IV-D-02 and
Iv-D-23) claimed chat smelters' self-reported emissions are
not credible and, consequently, CEM's should be required.
Another commenter (IV-D-05) stated that CEM's are widely
available for demonstrating compliance with the proposed
standards; however,  no further information was provided.
     Response:  The EPA's selection of the enhanced monitoring
requirements was determined by examining a hierarchy of
monitoring options available for specific processes,
pollutants, and control equipment.  The options ranged from
continuously monitoring the emissions of a specific pollutant
or pollutant class (the most accurate monitoring alternative)
to the continuous monitoring of a related process or control
device parameter.  Each option was evaluated relative to its
technical feasibility, cost, ease of implementation, and
relevance to the associated process emission limit or control
device for this industry.  The EPA then selected the most
appropriate monitoring option that was available, technically
feasible, and cost-effective for a specific application.
Although the EPA agrees that CEM's provide effective
monitoring alternatives when they meet these three criteria,
CEM's are not available at a reasonable cost for the pollutant
classes regulated by this NESHAP for this industry.
Therefore, the proposed rule did not require CEM's as the only
monitoring option and included alternative monitoring
requirements.  The EPA proposed several monitoring options for
organic HAP's and HC1/C12, including CEM's.
     In order to ensure accurate self-reported emissions, the
Act provides for civil and criminal penalties for falsifying
records  (42 USC 7413).  Enforcement of the Act is also
strengthened by provisions for noncompliance penalties
(40 CFR part 66) and citizen suits (40 CFR part 54}.  These
penalties, combined with stringent recordkeeping and reporting
requirements, will deter operators from falsifying monitoring
reports.

                             2-62

-------
 2.7.1  Metal  HAP Monitoring Options
     At proposal,  the EPA had  concluded  there are no  CEM's  for
 metal  WAP-s or  for lead, the proposed metal HAP  surrogate.
 Therefore, the  EPA proposed that each smelter operate and
 maintain a COM  in  each stack or duct that receives  controlled
 smelting furnace process emissions.  During the  initial lead
 compliance test, the COM would be used to monitor and record
 the opacity of  the smelting furnace exhaust reduced to
 6-minute averages.  Thereafter, the owner or operator would be
 required to maintain a 6-minute average  opacity  not more than
 2 percent  opacity  above the maximum 6-minute average  opacity
 recorded during the initial lead compliance test.   A  2 percent
 opacity difference was added to allow for normal instrument
 drift.  Exceeding  this site-specific opacity limit  would
 constitute a violation of the  2.0 mg/dscm emission  standard
 for lead compounds.
     Many  comments  were received on the proposed metal HAP
monitoring requirements.  These are discussed in detail in the
 following paragraphs.
     Comme n t:  Twenty commenters (IV-D-01, IV-D-02, IV-D-03,
 IV-D-04, IV-D-07,  IV-D-08,  IV-D-10, IV-D-12, IV-D-14,  IV-D-15,
 IV-D-16, IV-D-17,  IV-D-1B,  IV-D-19, IV-D-20, IV-D-21,  IV-D-22,
IV-D-26, IV-D-29,  and IV-D-31)  provided comments on the use of
an opacity standard to monitor metal HAP emissions  from
process sources.
     Four commenters (IV-D-02,  IV-D-03,  IV-D-10,  and  IV-D-18)
supported the proposed opacity requirements to monitor process
emissions.   One commenter (IV-D-02) stated that COM's are
essential for ensuring smelters'  compliance with the proposed
lead standard.  A second commenter (IV-D-10) stated that small
sources may have inadequate procedures for baghouse
maintenance and that the use of COM's will enable immediate
detection of baghouse problems.  One commenter (IV-D-03)
supported the proposed COM requirement if the opacity limits
are kept below 5 to 10 percent.  This commenter stated that
COM's provide greater assurance of  compliance with the metal

                             2-63

-------
HAP standard than do baghouse inspection and maintenance
programs.  One commenter  (iv-D-18) supported the use of COM's
as Iuii9 as the EPA demonstrates that the selected opacity
level will ensure that HAP emissions are below the level of
toxicological risk concern to the local populace and
environment.
     One commenter (IV-D-04) supported the proposed
requirement for COM's for process sources; however, instead of
the site-specific opacity compliance determination specified
in the proposed rule, the commenter requested an upper opacity
compliance limit of 5 percent based on a 6-minute average.
The commenter argued that this standard would be more
equitable because most smelters can currently meet a
3-percent, 6-minute average limit coupled with an additional
2 percent allowance for drift.
     Twelve commenters (IV-D-01, IV-D-07, IV-D-08, IV-D-12,
IV-D-14, IV-D-15, IV-D-16, IV-D-19, IV-D-21, IV-D-22, IV-D-26,
and IV-D-31)  stated that COM's cannot be used to demonstrate
or monitor compliance with a numerical metal HAP emission
limit.  Nine commenters (IV-D-07, IV-D-08, IV-D-12, IV-D-14,
IV-D-16, IV-D-21, IV-D-22, IV-D-26, and IV-D-31) objected to
the proposed opacity monitoring requirements for process
baghouses because the EPA failed to demonstrate that lead, PM,
and opacity are correlated.  Another commenter  (IV-D-22)
stated that if the EPA cannot show this relationship, a COM
will not demonstrate compliance with the lead standard;
therefore, the commenter requested that the standard be
written in terms of particulate emissions rather than lead.
One commenter {IV-D-14}  stated that the relationship of PM to
opacity is imprecise because particulate size and shape, which
affect the opacity reading, vary with furnace feed material
and operating conditions.   All nine commenters stated that
COM's would be particularly insensitive to low PM
concentration levels, such as those associated with the
proposed lead standard.
                             2-64

-------
     Thirteen  commenters  (IV-D-01, IV-D-07, IV-D-08,  IV-D-12,
 IV-D-14,  IV-D-16,  IV-D-17, IV-D-19, IV-D-20,  IV-D-21,  IV-D-26,
 IV-D-29,  and IV-D-31) recommended that the metal HAP
 monitoring requirements allow  for alternatives to COM's to
 monitor baghouse performance,  including bag leak detection
 systems  (such  as tribo-electric devices and light
 backscattering devices), pressure drop monitors, and
 systematic baghouse  inspection and maintenance programs.  The
 commenters stated  that other monitoring systems, such as bag
 leak detection devices, are more accurate and reliable than
 COM's, can be  used to indicate which baghouse cell is  leaking,
 and are easier than  COM's to keep properly calibrated.  These
 commenters also indicated that other systems  are less  costly
 than COM's, both in  capital costs and operating costs.
     One  commenter (IV-D-26)  reported that a  baghouse  leak
 detection system would have an initial cost of $7,000  to
 $14,000 and negligible maintenance costs.  At proposal, the
 estimated initial  cost of a COM was $37,000 and annual
 operating costs, including operation,  maintenance,  and
 recordkeeping  and  reporting,  were $16,500.
     One  commenter (IV-D-16)  stated that section 114 of the
Act does not require the EPA to establish a numerical  emission
 limit that must be monitored on a continuous basis.  The
 commenter added that section 112 of the Act allows for the use
of work practice (non-numerical)  standards.  Therefore,
according to the commenter,  there is no need  for the EPA to
establish a numerical opacity limit that can be used as a
surrogate to demonstrate continuous compliance with the lead
emission standard.   Furthermore,  the requirements of the Act
can be met by establishing work practice standards for process
baghouses similar to those proposed for process fugitive
baghouses, according to the commenter.
                             2-65

-------
     Response:  At proposal, the EPA had concluded that COM's
alone would be sufficient to monitor compliance with the lead
emission standard.  Based on further analysis and information
received from public comments, the EPA now agrees that COM's
cannot be used to monitor compliance with a lead emission
standard for this industry and that opacity can only be used
as a gross indicator of baghouse performance.  The EPA also
agrees that better bag leak detection systems are available
and effective for monitoring baghouse performance.  The EPA
recognizes that these bag leak detection systems have lower
capital and operating costs than COM's.  The EPA also
recognizes that work practice standards based on baghouse
inspection and maintenance programs are also applicable to
    oss baghouses and are consistent with the requirements of
     Bag leak detection systems, such as tribo-electric and
light back scattering devices, can be used to monitor baghouse
performance by indicating bag leaks or tears.  Because bag
leak detection systems are more sensitive than COM's, they can
detect the onset of bag degradation prior to gross baghouse
failures (e.g., torn bags) and can identify the specific
baghouse cells in which a problem exists.  Baghouse inspection
and maintenance programs can further improve baghouse
performance by ensuring proper baghouse operation.  These
programs include monitoring of pressure drop across cells as
well as inspecting bags and other baghouse components for
early identification of any required maintenance.  For these
reasons, the EPA has revised the metal HAP monitoring
requirements for process baghouses.
     The final metal HAP monitoring requirements are use of a
bag leak detection system coupled with a comprehensive
baghouse inspection and maintenance work-practice standard to
ensure that the baghouse is operating properly to control
metal HAP emissions.  Because COM's provide a less accurate
measure of baghouse performance than bag leak detection
systems, they have been removed from the monitoring

                             2-66

-------
 provisions.   There  should be no  significant  added  burden  from
 the  inspection  and  maintenance requirements  because  nearly  all
 secomiduiy  lead  smelters already  perform regular baghouse
 inspections;  there  will be minor added  burden  for
 recordkeeping and reporting to -ensure that inspection and
 maintenance  is  being performed.
     The EPA agrees that the ratios  of  metal HAP's to lead  and
 lead to PM are  variable; however, EPA data indicate  that  metal
 HAP, lead, and  PM emissions are  positively correlated
 (i.e., increases in emissions of'one are associated  with
 increases  in emissions of the others).   In addition, lead is  a
 surrogate  for metal HAP's because metal HAP  emissions are more
 closely correlated  with lead emissions  than with PM  emissions.
 Therefore, the  metal HAP emission standard for process sources
 remains in terms of lead emissions.
     The EPA agrees that opacity readings and lead
 concentrations  may  not be closely correlated and that COM's
 cannot be  used  to demonstrate or monitor compliance  with  a
 numerical  metal HAP emission limit for  this  industry.
 Although the  EPA still believes  that opacity may be  used  as an
 indicator  of  gross baghouse performance, the EPA has
 determined from information collected since proposal that bag
 leak detection  systems may be used as a more reliable
 indicator  of  baghouse performance.  When used in conjunction
 with a baghouse inspection and maintenance program,  a leak
 detector can  ensure that a baghouse  is achieving a consistent
 level of performance on a continuous basis.
     The Agency is currently reviewing the state-of-the-art
 for PM CEMs.   It appears that other devices using the same
 analytical techniques as the bag leak detectors,  i.e.,
 triboelectric and light scattering,   can be calibrated to
measure PM emissions continuously.  The Agency is currently
 studying the  technical feasibility of these PM CEMs.   This
 study is not yet complete,  so the use of a PM CEM cannot be
mandated at this time.   However,  using these types of devices
as bag leak detectors appear to be the next logical step for

                             2-67

-------
the Agency to take in its possible movement toward the use of
PM CEMs.
     Consent:   Six commenters (IV-D-01, IV-D-07, IV-D-08,
IV-D-12, IV-D-16, and IV-D-19) stated that the objective of
the proposed opacity monitoring requirement for process
sources was unclear.  They requested that the EPA clarify
whether it was to (1) demonstrate compliance with the metal
HAP standard or  (2)  ensure proper baghouse functioning.
     Response:  At proposal, the EPA had concluded that COM's
could be used to demonstrate compliance with a metal HAP
emission standard by establishing a site-specific opacity
limit during the initial lead compliance test.  However, based
on information collected from public comments and further
analysis since proposal, it is apparent that COM's cannot be
used in that capacity for this industry.  The correlation
among lead concentration, PM concentration, and opacity is not
strong enough to support using a COM to demonstrate compliance
with the metal HAP emission standard.  Therefore, as discussed
in the preceding response, the metal HAP monitoring
requirements have been revised to remove the COM requirement,
The final rule requires a bag leak detection system coupled
with a baghouse inspection and maintenance program as a means
to ensure proper baghouse functioning following an initial
lead compliance test.
     Comment:   One commenter  (IV-D-15)  objected to the
2-percent allowance in the proposed opacity limit for COM
drift because it is less than the allowable drifts specified
in 40 CFR Part 60, Appendix B, Performance Specification 1,
and Part 60, Appendix F.  Another commenter (IV-D-07) added
that a COM may not measure any opacity during the compliance
test and that subsequent violations could be triggered by
instrument drift.
                             2-68

-------
     Response:  Performance  specification  1  in 40 CFR part  60,
appendix B allows a maximum  zero drift of  2  percent  opacity
and nictxiiiLuiu calibration drift of 2 percent opacity.  These
drifts are not additive because they are for the lower and
upper ends, respectively, of the COM output  ranges.
Performance specification 1  also allows a maximum calibration
error of 3 percent opacity.  However, this error is  only
relevant for determining compliance with an  absolute opacity
limit  (e.g., 5 percent opacity).  It is not  relevant for
monitoring a site-specific opacity limit, as in the proposed
rule, that is determined and monitored with  the same
instrument.  Therefore, the  calibration error is not additive
with the zero drift or calibration drift and the 2 percent
allowance for drift specified in the proposed rule is
sufficient.
     Appendix F does not address drift of COM's.  The proposed
rule allowed for 2 percent opacity to account for drift so  a
violation could not be triggered solely by drift.  Regardless,
as discussed above,  the COM requirement has been withdrawn.
The revised rule specifies that bag leak detection systems  be
calibrated according to EPA guidance,  if available,  or
according to the manufacturers'  written guidelines.
     Comment:  One commenter (IV-D-06)  stated that using the
single highest 6-minute average opacity to establish an
opacity limit would be too lenient because this value may be a
statistical outlier that is not indicative of compliance with
the lead standard.  The commenter requested that the rule
contain a statistically calculated 6-minute average opacity
limit that would better reflect lead emission levels measured
during the lead compliance test.  The same commenter also
requested that the EPA delete the phrase "6-minute average"
from the proposed monitoring requirements for process sources
because there is an implication that only one 6-minute average
needs to be obtained.   The commenter stated that removing the
phrase will require all collected opacity data to be
evaluated.

                             2-69

-------
     Response:   As discussed above, the COM requirement has
been withdrawn.  The revised requirement for a bag leak
detection system is not tied to compliance with the lead
emission limit.  However, the bag leak detection system must
include an alarm to indicate baghouse leaks or tears.  An
alarm by itself will not indicate or constitute a violation of
the standard.  Failure to take corrective action to respond to
an alarm or failure to follow the inspection and maintenance
procedures is a violation of the standard.
     Comment:  Six commenters (IV-D-01, IV-D-07, IV-D-14,
IV-D-15, IV-D-16, and IV-D-30) declared that the proposed
monitoring requirements would be more stringent than the
actual emission standards and did not account for normal
variability in baghouse opacity.
     Four commenters (IV-D-07, IV-D-14, IV-D-15, and IV-D-16}
stated that the proposed opacity monitoring requirement may
force a source to comply with a more stringent lead standard
than the proposed 2.0 mg/dscm because the opacity measured
during the initial lead compliance test would become an
enforceable limit,  even if the measured lead emissions were
well below the proposed limit of 2.0 mg/dscm.
     One commenter (IV-D-16)  also stated that the proposed
monitoring requirements would establish emission standards
more stringent than MACT because compliance with the lead
standard is based on a 3-hour averaging period, but compliance
with the opacity standard is based on only a 6-minute
averaging period.  Shorter averaging periods effectively
create more stringent standards.
     Five commenters, (IV-D-01,  IV-D-07,  IV-D-15,  IV-D-16 and
IV-D-30) also objected to the proposed monitoring requirement
because it does not account for normal variability in baghouse
opacity, which could lead to violations of the standard.  One
of the commenters (IV-D-30)  added that the EPA test data do
not represent a level that can be continuously achieved on a
long-term basis.  Two commenters (IV-D-01 and IV-D-07)
suggested that fluctuations in furnace feed and other factors

                             2-70

-------
 may  cause  sooting  episodes, which may not  necessarily  occur
 during  the initial  compliance  test.  Because  sooting episodes
 do not  entail  significant  increases  in metal  HAP  emissions,
 they could cause a  violation of  the  opacity standard even
 though  the lead limit was  not  being  exceeded.
      Response:  As  discussed above,  the COM requirement has
 been withdrawn.  The revised requirement for  a bag leak
 detection  system allows each source  to set the system
 parameters (e.g., alarm level) to account  for normal
 variations in  emissions.   As discussed above, these systems
 will  be used to monitor baghouse performance  and  are not tied
 to compliance  with  the lead emission limit.   Sooting episodes
 should not affect particulate  matter readings because  the soot
 will  be controlled  by the  metal HAP  control device (i.e., the
 baghouse).
      Comment:  One  commenter (IV-D-03) objected to the use of
 COM's on process sources because they do not  represent
 state-of-the-art monitoring.   The commenter requested  that
 CEM's be required for monitoring metal HAP emissions to ensure
 the accuracy of reported emissions.  However, two commenters
 (IV-D-Ol and IV-D-07) specifically stated that there are no
 CEM's for  metal HAP's.
     Response:  The EPA has no data  to indicate that CEMs for
metal HAP's, other  than for mercury, are commercially
available  at this time.  For this rule,  the EPA would be
 interested in a metal CEM  to demonstrate continuous compliance
with the lead standard.  The EPA does not believe a mercury
monitor is  relevant to the lead standard,  so the use of a
mercury CEM was not pursued.  The development of a metal CEM
which can measure lead is  currently in the research to early
prototype  stages,  so mandating the use of one now is not
possible.   Consequently,  CEM's are not required in the
proposed or final rule.  However, as discussed above,  the COM
requirement has been replaced with a requirement for a bag
leak detection system,  which provides a better measure of
baghouse performance than a COM.

                             2-71

-------
     Comment:  Two commenters  (IV-D-14 and IV-D-16) stated
that the performance of COM's may be adversely affected by
conditions in the furnace exhaust  (e.g.,  moisture).  One
commenter  (IV-D-30) stated that COM's have historically had
operating problems that could cause operating delays and
erroneous violations of the proposed opacity standard.
Another commenter  (IV-D-31) stated that COM's cannot respond
to the velocity changes that occur in baghouses.   Finally, one
commenter  (IV-D-14) added that because the proposed location
for the COM  is upstream of any scrubber,  the COM would not
measure actual emissions to the ambient air.
     Response;  The EPA agrees that COM's may be adversely
affected by  furnace exhaust; however, properly maintained
rOM'!= have been demonstrated as an effective monitoring
uc^iiiiwj-w-j                 _              .>e at several
smelters.  Baghouse velocity changes would not affect the
monitoring of baghouse functioning using a COM because opacity
is not tied  to gas stream velocity.  However, as discussed
above, the revised rule requires bag leak detection systems
instead of COM's.  Bag leak detection systems are used by at
least four facilities in this industry.
     False indications of a baghouse failure (i.e., false
alarms) do not constitute a violation as long as the
appropriate  corrective actions are taken, recorded, and
reported.
     EPA test data indicate that scrubbers designed for acid
gas control  do not provide any significant additional metal
HAP's control when they follow a baghouse.   However, scrubbers
may emit PM, which could affect the reading of COM's or bag
leak detectors.  Consequently, the required location for the
monitoring device is still between the baghouse and any
scrubber.
     Comment:  One commenter  ([IV-D-30]}  objected to the
proposed opacity monitoring requirement for process sources on
the grounds  that it is unnecessary because the NAAQS require
continuous lead monitoring.  The commenter also stated that

                             2-72

-------
 the NAAQS  lead monitoring  requirement makes  the opacity
 proposal excessive and a waste of money.
      Response:  The NAAQS  monitoring requirements are  intended
 to ensure  compliance with  ambient air quality standards  for
 lead.  The NAAQS monitoring  cannot demonstrate that the  MACT
 level  of control for metal HAP'S is being achieved on  a
 continuous basis.
      Comment:  One commenter (IV-D-16) requested that  the
 opacity monitoring proposal  be withdrawn because the EPA had
 improperly evaluated the cost impact of the  requirement.  The
 commenter  stated that the  EPA had overestimated the
 cost-effectiveness of COM's  by overestimating the level  of
 lead reductions attributable to COM's.  One  commenter
 (IV-D-22)  stated that COM's  would not be cost-effective  for
 small  operations that do not recycle batteries.
     Response:  As stated  above, the COM requirement has been
 replaced with a requirement  for a bag leak detection system
 and an inspection and maintenance program.   Both COM's and bag
 leak detection systems have  been demonstrated and are  in use
 in this industry and the costs cannot, therefore,  be
 considered prohibitive.  Small facilities that do not perform
 smelting,  such as lead remelters and refiners, are not subject
 to the standards.
     Comme_nt:  One commenter  (IV-D-03) supported the proposed
baghouse inspection and maintenance requirements for process
 fugitive sources.  Another commenter (IV-D-06) recommended
 changing the proposed baghouse inspection requirements to
limit inspections to the clean side of the baghouse.   The
commenter stated that visual inspections of  the clean sides of
baghouses would easily detect any baghouse leaks and that most
baghouses have a glass window for such monitoring.   The
commenter also stated that this change would reduce exposure
of the inspector to lead dust.
     One commenter (IV-D-15)  stated that the proposed baghouse
inspection requirement could not be easily performed unless it
was only a visual observation of exhaust opacity.   The

                             2-73

-------
cormnenter objected to an internal inspection because it would
require the baghouse to be taken off-line, during which time
cne process would need to be shut down.
     Response:  The EPA's survey of secondary lead smelters
indicates that most smelters perform full baghouse inspections
for both process and process fugitive sources at least weekly,
and generally more often.  Most secondary lead baghouses are
divided into two or more cells.  Therefore, individual cells
can be taken off-line for inspection, without shutting down
the whole baghouse or the process.  As discussed above, the
rule requires a comprehensive baghouse inspection and
maintenance program that is consistent with current practices
at most smelters, together with a bag leak detection system
for both process and process fugitive source baghouses.  These
requirements may require inspecting the dirty sides of some
baghouses but they are consistent with current practices and
are necessary to ensure that the baghouses perform optimally
on a continuous basis.  The EPA agrees that worker exposure to
dust should be minimized and proper precautionary measures
should be taken when working on the dirty sides of the
baghouses.
     Comment:  One commenter (IV-D-20)  requested that the
proposed rule require the installation of ambient air
monitoring equipment for lead to demonstrate compliance with
the federal ambient air standard.
     Response:  Ambient air quality is regulated for lead
under the NAAQS rather than under this standard.
2.7.2  Organic HAP's - THC And Temperature
     The proposed organic HAP monitoring requirements allowed
smelter owners or operators to monitor either THC
concentration directly with a CEM or monitor incinerator or
afterburner temperature.  If the owner or operator selected
the second option,  they would have to maintain, at a minimum,
the same temperature as measured during the initial THC
compliance test.
                             2-74

-------
      Comment:  Two  commenters  (IV-D-02 and  IV-D-03)  supported
 the proposed process  source  THC monitoring  requirements.  They
 oui->pG.L leJ  \.l\e requirement  for  continuous monitoring  of
 incinerator or afterburner temperature because these
 parameters are correlated  to THC emissions.  One of  these
 commenters  (IV-D-03), however, requested that the EPA consider
 state-of-the-art CEM1s--particularly fourier transform
 infrared  (FTIR) instruments--as a means of  ensuring  compliance
 with  the process source THC  standard.  According to  the
 commenter, FTIR is  economically feasible and reliable for some
 industries and may  be so for the secondary  lead smelter
 industry.
      Response:  The proposed requirements for THC or
 temperature monitoring are equivalent and both are adequate
 for meeting the statutory  obligation for continuous  emission
 or parameter monitoring; however, the use of a THC CEM is a
 more  expensive monitoring  option than temperature monitoring.
 No FTIR instruments are commercially available at this time
 that  can be used as an in-stack CEM.  The EPA cannot mandate
 the use of instruments that  are not commercially available.
 Therefore, the EPA  has not included FTIR as an option in the
 final monitoring requirements.
 2.7.3  To ta1 Chiori de s	-	Parameter	Moni t oring
     The EPA proposed four monitoring options to demonstrate
 compliance with the proposed HC1/C12 emission standards.  In
 the first option,  the owner or operator would be required to
monitor and maintain the same ratio of fluxing agents to feed
material as during  the initial HC1/C12 compliance test.   In
 the second and third options, owners or operators using a
 sulfur dioxide (SC^) scrubber to control HC1/C12 would be
 required to monitor and maintain either the same scrubber
 operating parameters  (media pH and injection rate)  or 862
 concentration,  respectively,  as during the initial HC1/C12
 compliance test.   In the final option,  owners or operators
would be required to monitor HC1 emissions directly using an
HC1 CEM.  Alternatives to an HC1 CEM were proposed because an

                             2-75

-------
HC1 CEM has not been demonstrated for this industry and the
costs are expected to be prohibitive.  Data available to the
EPA indicated that the other monitoring options were equally
effective and less costly.
     Several comments were received on the proposed monitoring
requirements to demonstrate compliance with the proposed
HC1/C12 emission standards for process sources.  These
comments are discussed in the sections below.
     2.7,3.1  Fl uxjng Mgn.lt or ing Op t i on.
     Comment:  Two commenters (IV-D-18 and IV-D-20)
conditionally supported the EPA's proposal to allow operators
to monitor the amount of flux added to the furnaces to control
HC1/C12 as a means of demonstrating compliance with the
HC1/C12 emission standard.  One commenter  (IV-D-18) supported
the monitoring of HC1/C12 emissions through fluxing if the
effectiveness of this control technique were periodically
verified by testing the exhaust stream.  Another commenter
(IV-D-20) supported the fluxing monitoring requirement and
additionally requested that sources be required to install a
volumetric feeder for lime or soda ash to ensure a steady
supply of fluxing materials.
     Two commenters (IV-D-03 and IV-D-04) stated that the
proposed use of fluxing to control HC1/C12 was a good
approach, but it provides no mechanism for demonstrating
continuous compliance with the HC1/C12 standard.
Consequently, both commenters requested that the EPA require
HC1 CEM monitoring to verify compliance.  One commenter
{IV-D-04) supported an HC1 CEM requirement because it would
allow a smelter to minimize the use of fluxes to control
HC1/C12 emissions by adjusting the quantity of fluxing
materials added according to monitored HC1 emissions.
     One commenter (IV-D-22) objected to the proposed fluxing
monitoring option because the proposal preamble did not
provide sufficient data to support the efficiency of fluxing
for controlling HC1/C12-  The commenter stated that the EPA
needs to establish a relationship between fluxing agents and

                             2-76

-------
 furnace  feed, stream composition, and emissions in order to
 determine  the appropriate fluxing levels  for compliance
 purposes.  The commenter also stated that it may not be
 possible to maintain a constant ratio of  feed to flux and that
 a constant ratio may not be adequate for monitoring because
 the chlorine content of the feed may be variable.
     Response:  At proposal, the EPA's data indicated that
 normal fluxing was a feasible option for  controlling HC1/C12
 emissions.  However, as described in section 2.3.4, the EPA is
 not promulgating the HC1/C3.2 emissions standards and
 associated monitoring requirements because subsequent
 information obtained since proposal indicates fluxing may not
 be effective in many cases and HC1/C12 emissions will be
 declining, in any case, because of the phase out of PVC
 separators.
     The EPA agrees that a volumetric feeder would have made
 it easier to maintain a constant ratio of feed to flux, as
 required by the proposed monitoring requirements.  However,
 only a few smelters use volumetric feeders because standard
 industry operating procedures are sufficient to maintain a
 constant ratio of feed to flux.  Therefore, the choice of
 which mechanism to use to maintain that ratio would have been
 left to  the individual smelter operator,  according to the
proposed rule.
     The EPA included HC1 CEM's as one of several monitoring
 options but did not require them as the only option in the
proposed rule because the costs would be prohibitive for this
 industry and other less costly options were available.   The
 economic savings represented by using less flux in response to
 lower HC1 emissions would not have been sufficient to justify
 the cost of an HC1 CEM.
     2.7.3.2  Scrubber Operating Parameters Monitoring .Option.
     Comment:  One commenter (IV-D-30)  objected to the EPA's
proposed option to allow operators to monitor compliance by
maintaining the same scrubber parameters  (media pH and
injection rate)  as measured during the initial compliance

                             2-77

-------
test, if they are using wet scrubbers to control HC1/C12
emissions.  The commenter stated that insufficient data have
been provided on the feasibility of regulating scrubber pH to
within one pH unit and that such control is not feasible.
     A second commenter  (IV-D-22) objected to the proposed
3D-percent variance allowed for the scrubber liquid injection
rate.  The commenter requested a requirement for a constant
ratio of the liquid injection rate to the chlorinated
hydrocarbon content of the furnace feed.  This alternative
would (1) allow the source to reduce the injection rate when
the chlorinated hydrocarbon content of the feed was low; and
(2) increase the injection rate to adequately control HC1/C12
when the chlorinated hydrocarbon of the feed was high.  If the
^ourr<= could not test the feedstocks for the chlorinated
11^0.4-o^c.^^^-.                  „           ..ate of chlorinated
hydrocarbon variability would provide an upper-limit estimate
that correlated to a minimum required injection rate.
     Response:  The proposed rule required only that the
3-hour average scrubber pH level be maintained to within one
pH unit of the average pH recorded during the initial HC1/C12
test; deviations greater than one pH unit were allowed as long
as the 3-hour average is maintained within one pH unit.
     EPA data indicate that pH can be maintained within this
range and no data have been provided to indicate otherwise.
Wet scrubbers are greater than 99-percent effective in
removing HC1/C12 from exhaust gases.  The EPA does not believe
that a 30-percent drop in injection rate would affect scrubber
efficiency for HC1/C12 removal.  However, as discussed in
section 2.3.4, the HC1/C12 emissions standards and associated
monitoring requirements have been withdrawn.
     2.7.3.3  Sulfur Dioxide CEM Monitoring Option.
     Comment:  Three commenters (IV-D-04, IV-D-06, and
IV-D-30)  objected to the second monitoring option for smelters
using an SC>2 scrubber to control HC1/C12 emissions.  This
option allowed operators to monitor compliance by monitoring
and maintaining the same SC>2 concentration as during the

                             2-78

-------
 HC1/C12  compliance test.   One commenter (IV-D-30)  stated that
 requiring  an  SC>2  CEM would significantly increase operating
 costs.
     Two commenters (IV-D-04  and IV-D-06)  objected to the
 proposed SC>2  monitoring  option because SC>2 inlet
 concentrations  and emissions  reductions are not correlated
 with HC1/C12  emissions.   They stated that  HC1/C12  emissions
 are dependent upon the vinyl  chloride concentration of the
 feed and SC>2  emissions are dependent upon  fuel  type.
 Consequently, HC1/C12 emissions will vary  more  than S02
 emissions  and their concentrations  will not be  correlated.
 These commenters  stated  that  the proposal  would allow for the
 establishment of  an S02 monitoring  compliance level even when
 SC>2 inlet  concentrations  were at or near zero;  consequently,
 no HC1/C1.2 emissions  control  would  be required.
     Both  commenters  (IV-D-04 and IV-D-06)  also stated that an
 SC>2 monitoring  system should  only be allowed in situations  in
 which it can  be demonstrated  that (1)  reductions of SC>2
 emissions  are sufficient  to verify  that total acid gas
 reduction  is  related to SC>2 reduction,  and (2)  the proportion
 of S02 entering the scrubber  is  proportional to the
 concentration of  other acid gases.
     Response:  As  discussed  in  section 2.3.4,  the HC1/C12
 emission standards  and associated monitoring requirements have
 been withdrawn.   However,  sources selecting the S02 monitoring
 option under  the  proposed  rule would have  incurred no
 additional monitoring expenses because  smelters  with  SC>2
 scrubbers  typically have SC>2  CEM's.
     The EPA  agrees that SC>2  emissions  are  not  necessarily
 correlated with HC1/C12 emissions.   However, the SC>2
monitoring option was not  intended  to establish  a  correlation
between S02 and HC1/C12 emissions;  rather  it was to establish
that the S02  scrubber is being operated in  the  same manner  as
during the initial HC1/C12  compliance test.  Acid  gas
scrubbers will preferentially remove  HC1 and Cl2 before they
remove SC>2  because HC1 and Cl2 are more  soluble  in  scrubber

                              2-79

-------
media than S02-  Therefore, a scrubber that is effectively
controlling SC>2 will also be controlling HC1/C12-
     The statement that 803 emissions from secondary lead
smelters are dependent on fuel type is incorrect; the majority
of SC>2 emissions are from the lead-sulfur compounds in the
battery paste.  All smelters use coke, natural gas, or propane
as fuel.  Natural gas and propane have no sulfur.  Coke
contains some sulfur, but SC>2 emissions due to the sulfur in
coke are still less than emissions of sulfur from battery
paste.
     2.7.3.4  Hydrochloric Acid CEM Monitoring Option.
     Comment:   Two commenters (IV-D-02 and IV-D-03) supported
the monitoring of HC1 using a CEM and requested that the EPA
require this monitoring option for facilities that fail to
comply with the other HC1 monitoring options.
     Response:  As discussed in section 2.3.4, the proposed
HC1/C12 emissions standards and associated monitoring
requirements have been withdrawn.  Smelters that did not
select one of the other HC1 monitoring options under the
proposed rule would have been required to comply with the HC1
CEM requirements.  However, an HC1 CEM would not have been
required as a penalty for noncompliance with the other
monitoring options.
     2.7.3.5  Ambient Monitoring for HC1 and Chlorine
Emissions.
     Comment:   One commenter (IV-D-03) requested that the EPA
consider the use of a corrosion monitoring system of metal
plates as a possible alternative to the first three HC1
monitoring options (i.e., monitoring fluxing parameters,
scrubber parameters,  or S02 concentration).  These metal
plates of copper, bronze, and aluminum react with HC1/C12 and
would be placed along the property fenceline.  After 12 to
18 months,  the plates would be tested for corrosion, based on
a pre- and post-weighing, to verify the reliability of the
first three monitoring options.   This suggestion was based on
a monitoring program at a hydrochloric acid/potassium sulfate

                             2-80

-------
 fertilizer plant that was conducted by the  commenter's State
 Air Control Board.
     Response:  As discussed in section 2.3.4, the proposed
 HC1/C12 emissions standards and associated  monitoring
 requirements have been withdrawn.  It is unlikely that HC1/C12
 concentrations at the fence-line are high enough to cause
 measurable corrosion.  Regardless, this technique could not
 confirm that the source of any corrosion is acid gases from
 the smelter.  In addition, there is no way  to establish a
 correlation between corrosion and violations of a numerical
 HC1/C12 emission standard.
 2.8  RECORDKEEPING AND REPORTING REQUIREMENTS
     Comment:  Four commenters  (IV-D-02, IV-D-03, JV-D-06, and
 IV-D-30} submitted comments on the proposed recordkeeping and
 reporting requirements.  One commenter (IV-D-03) supported all
 of the proposed notification requirements and stated that they
 are consistent with other NESHAP rules.
     Two commenters (IV-D-02 and IV-D-03)  supported the
 proposed requirement for retaining records  for five years.
 One of these commenters (IV-D-03)  stated that the requirement
 is consistent with other NESHAP rules,  encourages compliance,
 and provides a means of verifying facility  compliance with the
 proposed standards.
     Two commenters (IV-D-02 and IV-D-03)  supported the
 proposed standard of quarterly reporting in order to
 facilitate compliance with the proposed standards.
     One commenter (IV-D-06)  recommended that the EPA revise
 the proposed quarterly reporting requirements for all 3-hour
block averages from the continuous monitoring system or
 control device so that only the 3-hour block averages that are
 in violation of any of the proposed standards or monitoring
 requirements must be reported.
     One commenter (IV-D-30)  requested that the EPA delete the
proposed standard of quarterly reporting.   The commenter
 stated that this requirement is excessive  and unnecessary
                             2-81

-------
because records are currently kept at-each -smelter and they
can be inspected at any time by regulatory personnel.
     Response:  Section 63.10(e)(3) of the General Provisions
requires that the excess emission and monitoring system
performance reports and summary excess emission reports
include the total duration of excess emissions  (or parameter
exceedences, if a parameter is being monitored) and does not
require reporting of all 3-hour block averages for the
reporting period.  The reporting requirements in the proposed
rule have been revised to clarify that only excess emissions
or parameter exceedences must be included in the reports so
that this requirement will be consistent with the General
Provisions.
     Based on a consideration of the revised monitoring
requirements, the EPA agrees that semi-annual reports are
appropriate for this source category.  The reporting
requirements have been revised to require only semi-annual
reports.  However, owners and operators must begin to submit
quarterly reports if the source experiences excess emissions.
Owners and operators may return to a semi-annual reporting
frequency by meeting the requirements of § 63.10(e)(3)(ii) of
the General Provisions.  These require the source to have been
continually in compliance for 1 year and to continue to comply
with all recordkeeping and monitoring requirements.
     Comment:  Commenter (IV-D-Ol) expressed a concern that
the increased continuous emissions or parameter monitoring
would not allow for any excursions in emissions that might
occur during startup or shutdown periods or in connection with
emergencies or malfunctions.  The commenter requested that the
final rule address the application of MACT standards and allow
some flexibility for smelter operators during these periods.
     A second commenter (IV-D-30) noted that the proposed
standards do not include any provisions for "Acts of God" that
may cause upsets to the pollution control devices and
monitors.
                             2-82

-------
     Response:  Section 63.6(e) ,of the General Provisions
requires facilities to develop and implement written startup,
shutdown, and malfunction plans.  The purpose of these plans
is to ensure that owners operate and maintain affected
sources, including air pollution controls, in a manner
consistent with good air pollution control practices during
periods of startup, shutdown, and malfunction.
     Sections 63.6 (f) and 63.6 (h) of the General Provisions
state that all non-opacity and opacity emission standards
apply at all times except during 'periods of startup, shutdown,
and malfunction and as otherwise specified in an applicable
subpart.  If a pollution control device or monitor is caused
to malfunction by natural events, such as inclement weather,
then the source would not be subject to the appropriate
emission standards, but would be required to follow the
startup, shutdown,  and malfunction plan.  Section 63.8(b)
contains provisions for monitoring system malfunctions that
are not in the startup, shutdown, and malfunction plan and for
monitoring devices that are determined to be "out of control."
2.9  INTERACTION WITH OTHER RULES
     Several comments were received on the interaction between
the proposed rule and other environmental regulations and
standards.
2-9.1  Interaction With NAAOS
     Comment:  One commenter (IV-D-10)  discussed the
interaction of the proposed emission standards and the lead
NAAQS.   The commenter noted that a number of areas in the
United States are not in attainment with the lead NAAQS and
that the proposed standards did not discuss whether secondary
lead smelters cause or contribute to nonattainment in these
areas.   The commenter stated that if any secondary lead
smelters cause or contribute to nonattainment situations, then
the proposed standards may make it difficult for States to
reach attainment.   This is because the proposed standard will
not require further controls on process sources,  which are the
largest source of lead emissions, according to the commenter.

                             2-83

-------
     The same commenter also indicated that the EPA informed
the commenter about plans to abandon the NAAQS in favor of a
NESHAP approach.  According to the commenter,  the EPA has
found that with the abolition of leaded gasoline, the only
lead nonattainment areas are in the vicinity of lead smelters
or other large stationary sources of lead.  Therefore, the EPA
believes that a control technology-based approach will be more
effective in reducing public exposure to lead.  The commenter
concurs with the EPA's concept of focusing on stationary
sources.  However, the commenter is concerned that under the
proposed MACT emission standards for lead and with the
abolition of the lead NAAQS, the States will have no mechanism
under which to gain further lead emission reductions from
secondary lead smelters.  Under the proposed emission
OL.U.^^LA	,                    _             - stack emissions of
lead.
     Furthermore, according to the commenter,  the EPA did not
include lead refineries and sources that produce lead
compounds in the regulated category.  Therefore, it is
possible that lead emissions in nonattainment areas may not
decrease sufficiently in some areas to reach attainment with
the lead NAAQS.  No further action would be taken on these
categories until the residual risk assessment, but this will
not occur for at least 8 years from the present.  It is
theoretically possible, according to the commenter, that the
proposed MACT standards may delay progress in reducing ambient
concentrations of lead.
     The commenter requested that the EPA examine and discuss
the relationship between the lead NAAQS and the proposed MACT
standard and determine whether the proposed MACT standards may
actually delay progress in reducing ambient concentrations of
lead.
     Response:  The EPA did not discuss at proposal the issue
of whether secondary lead smelters cause or contribute to
violations of the NAAQS, except in the context of the area
source finding, because that is not the focus of the NESHAP

                             2-84

-------
program.  The NESHAP program establishes technology-based
emission  standards and the NAAQS regulate ambient air quality.
Ilcwcvcr,  according to the November 6, 1991 Federal Register.
there are 17 lead non-attainment areas;  secondary lead
smelters  were located in 12 of these areas.
     The  commenter is incorrect in stating that process
sources are the largest sources of lead emissions; as
described in Chapter 4 of the proposal BID, fugitive dust
sources are the largest sources of metal HAP emissions,
followed  by process fugitive sources.  Process sources are
actually  the smallest sources of metal HAP emissions because
all process sources are already controlled by MACT technology
(i.e., baghouses).  The standards require further controls on
fugitive  dust sources and process fugitive sources, and
establish emission limits for all sources ventilated through a
control device or a stack.  If further lead emissions
reductions are necessary because the area around a smelter is
still in  nonattainment for lead after a smelter has complied
with the  NESHAP, then a State may still require more stringent
controls,  emission limits, or operating restrictions as part
of a State Implementation Plan.
     The  EPA currently has no plans to abandon the lead NAAQS
in favor  of a NESHAP approach.  The NESHAP and NAAQS programs
are both  still needed because they regulate different aspects
of air quality.
     Lead refineries and sources that produce lead compounds
(such as  lead oxide production facilities)  are not secondary
lead smelters.  Regulation of these other sources under
section 112 would require an independent finding that they are
major sources of HAP air emissions.  The EPA has no data to
support such a finding.
2.9.2  InteractionwithRegulation of Air Emissions Under RCRA
     At proposal,  the EPA stated that air emissions from
secondary lead smelting  furnaces are potentially subject to
regulation under RCRA because the feed material is often
classified as a  hazardous waste (50 FR 29776).   The furnace

                             2-85

-------
emissions are exempt from regulation;'however, the EPA agreed
to reexamine the exemption as part of a settlement agreement.
The KFA proposed to continue the exemption because the air
emissions would be adequately regulated under the proposed
NESHAP (at least until the section 112 (f) residual risk
determination) and requested public comment.
     The EPA noted that there was also a potential overlap
between the RCRA rules, specifically 40 CFR 264.1101(1) and
265.1101(i), which regulate storage facilities, and the
proposed standards to control fugitive dust emissions.  The
EPA requested comment on whether the RCRA requirements for
storage units were incompatible with the proposed NESHAP.
     Comment:  Five commenters  (IV-D-03, IV-D-07, IV-D-14,
IV-D-18,  and IV-D-30) responded to the request for comment on
whether air emissions from secondary lead smelters should also
be regulated under RCRA.
     One commenter (IV-D-30) noted that only the storage of
hazardous waste prior to reclamation, treatment, and disposal
is subject to RCRA.  The reclamation equipment  (i.e., the
process equipment) is not subject to RCRA, according to the
commenter.
     Three commenters  (IV-D-03, IV-D-07, and IV-D-18) agreed
with the EPA's determination that further RCRA regulation of
air emissions from secondary lead smelters would be
unnecessary.  All three commenters noted that regulation under
both programs would not achieve an increased environmental
benefit compared to regulation under only one program.  One
commenter (IV-D-03),  however,  added that if the proposed
NESHAP fails to obtain adequate protection for the public,
then RCRA regulation may be appropriate in the future.
Another commenter (IV-D-18)  added that including all the
requirements for air emissions in the Title V operating permit
would facilitate permit preparation and processing for
facilities and regulatory personnel,  compared to requiring
additional RCRA air emission permitting.
                             2-86

-------
      One  commenter  (IV-D-14)  recommended  that  the EPA
 harmonize the  fugitive emissions  standards with  the RCRA
 ctcxr.d-rdc tc ensure that both programs are administered
 consistently,  particularly with respect to the requirements
 for  containment buildings.  The same commenter also
 recommended that the EPA exempt from the  fugitive dust
 requirements those areas subject  to RCRA  subpart DD
 containment building standards.
      Response:  RCRA rules presently control storage of all
 lead-bearing materials because they are hazardous wastes
 before they are processed by  smelting.  Hazardous wastes
 generated by the smelting process also are subject to
 regulation.  Air emissions from smelting  hazardous wastes
 (including spent batteries, which were held to be solid and
 hazardous wastes by the Eleventh  Circuit  in Ilcov.EPA) could
 be regulated under RCRA.  However, in 1991 the Agency chose to
 temporarily exempt such emissions  [see existing
 40 CFR 266.100(c) (1) and (3)], and to reexamine the
 appropriateness of that exemption as part of this proceeding.
      The  EPA believes it is appropriate to continue the RCRA
 exemption.  The MACT standards, coupled with the area source
 listing determination, provide substantial protection of human
 health and the environment from air emissions  from secondary
 lead  smelters  engaged in normal lead recovery  activities.
 Integration of the two statutes, required by RCRA
 section 1006,   is best served by continuing the RCRA exemption
 because RCRA regulation of such air emissions  would be largely
 or completely  duplicative of the MACT standards.  A
 determination  as to whether the RCRA exemption should be
permanent can be made at the time of the residual risk
 determination  required by section 112 (f)  of the Act.  The MACT
 standards, at  the least,  will function as reasonable interim
 controls.
     The  EPA has also sought to integrate certain of the RCRA
 storage requirements with the MACT standards.   It has done so
by allowing secondary lead smelters with RCRA  containment

                              2-87

-------
buildings (the standards for which appear in 40 CFR part 264,
subpart DD (for permitted facilities) and part 265, subpart DD
(for interim status facilities) to use these buildings as a
means of complying with certain MACT standards.  The Agency's
technical judgement is that the containment buildings provide
as stringent control as the other MACT option and, therefore,
are a valid alternative means of meeting the standard.
2.9.3  Interaction With	Effluent Limi t at i on	Guide1ines
     Comment:  One commenter (IV-D-07) disagreed with the
EPA's conclusion that most smelters with reverberatory
furnaces would use fluxing rather than scrubbers to control
HC1/C12 emissions.  The commenter also disagreed with the
EPA's conclusion that it was unnecessary to reexamine the
effluent limitation guidelines for secondary lead smelters.
According to the commenter, most smelters would be forced to
use scrubbers rather than alter the fluxing of the furnaces.
As a result,  the water quality impacts of the proposed rule
would be more significant than estimated by the EPA.
Therefore, the commenter believes that the effluent limitation
guidelines should be revised to account for the increased use
of scrubbers at these facilities.
     Response:  Scrubbers are currently in use in this
industry for acid gas control with no apparent conflict with
the current effluent limitation guidelines for this industry.
2.9.4  The	Title V Operating Permit Program
     Three commenters (IV-D-03, TV-D-16, and IV-D-22)
discussed the need for secondary lead smelters to obtain
Title V operating permits required by 40 CFR 71.  One
commenter (IV-D-03) agreed that all smelters should be
required to obtain operating permits.
     Two commenters (IV-D-16 and IV-D-22) disagreed with the
EPA's statement in the preamble that all HAP sources must
obtain a Title V operating permit from the State or from the
EPA if the State does not have an approved permit program.
One commenter (IV-D-16)  noted that, in fact, a smelter is
obligated to obtain an operating permit from the State only

                             2-88

-------
when the State has an approved Title V operating permit
program in place.  Otherwise, a smelter is obligated to submit
a Title V permit application to the EPA only if the EPA has
promulgated a substitute program for that State under
section 502(d)(3) of the Act.
     The second commenter  {IV-D-22) pointed out that all major
sources of HAP's are required to obtain Title V operating
permits, but area sources of HAP's have a 5-year deferral for
obtaining a Title V operating permit.  After 5 years, there
will be further rulemaking by the EPA to determine the
ultimate permit requirements for area sources under the
Title V operating permit program.
     The same commenter  (IV-D-22) also noted that the General
Provisions of part 63 (subpart A) require any source subject
to a section 112(d) emission standard to obtain a Title V
operating permit unless the individual standard to which it is
subject exempts the source from permitting.  The 5-year
deferral for area sources under Title V is also allowed unless
the individual standard overrides the deferral.  The commenter
recommends that the EPA explicitly exempt area source smelters
from Title V operating permit requirements in the final rule,
or at least exempt those area sources already subject to
permitting requirements with a State or local agency.  The
commenter cited the resource-intensive nature of the Title V
operating permit program and the comparatively smaller
emissions from area sources as a justification for this
recommendation.
     Response;  The EPA agrees that smelters in States with no
approved Title V program would be obligated to submit a
Title V permit application to the EPA only if the EPA has
promulgated a substitute program for that State.  Individual
States establish the schedules for Title V permit
applications;  however,  the deadline must be within 12 months
of the EPA approval of a State's Title V program or the EPA's
promulgation of a substitute program for that State.
                             2-89

-------
     Area sources do not automatically have a 5-year deferral
for obtaining a Title V operating permit; this deferral is
automatic only for sources subject to NESHAP that were
promulgated under part 61 (i.e., prior to the 1990 amendments
of the Act).  Otherwise, this deferral must be granted in a
specific emission standard.   Section 63.1 (c) (2) (iii) of the
General Provisions states, "If a standard fails to specify
what the permitting requirements will be for area sources
affected by that standard, then area sources that are subject
to the standard will be subject to the requirement to obtain a
Title V permit without deferral."
     The EPA's area source finding concluded that the health
risks posed by area source smelters warranted the same
emigpion standards (i.e.,  MACT) under this NESHAP as the major
ow^^^.   .   .                           -c sources should be
subject to the Title V operating permit program and no
deferral is being granted in this emission standard.
Regardless, secondary lead smelters that are area sources
would still need to obtain Title V permits even if they were
exempt under this NESHAP because nearly all of them are
subject to the 40 CFR 60 subpart L NSPS and most are also
subject to permitting requirements with a State or local
regulatory agency.
2.9.5  Administrative Procedure Act Requirements
     Comment:  One commenter (IV-D-14) noted that only the
preamble was published in the Federal Register.   The commenter
believes that the Administrative Procedures Act requires
publication of the proposed rule because the proposed rule
contains technical information, such as monitoring
requirements, that was not included in the preamble.  The
commenter argued that interested parties were, therefore,
deprived of adequate notice of the proposal's potential impact
and of the opportunity to comment on it.   The commenter
requested that the EPA publish the proposed rule in the
Federal Register and reopen the comment period.
                             2-90

-------
      Response:   The  Office  of Air  Quality  Planning  and
 Standards  has  reviewed  its  responsibility  to  adequately  inform
 ths  affected public.  The decision and actions  taken  to  reduce
 the  amount of  printed material  in  the federal	Register and
 still ensure that  the proposed  regulatory  text  and  support
 documents  are  accessible for public  comment satisfy the
 statutory  requirements  of the Administrative  Procedures  Act
 (APA),  the Federal Register Act (FRA), and the  requirements  of
 the  Clean  Air  Act Amendments of 1990.  Access to  the  rule  and
 support material was identified in the preamble to  the
 proposed rule.
      The Agency uses many connections to increase access to
 information not printed in  the  Federal Regisiteri  the
 Technology Transfer Network's  (TTN)  recently-signed-rule
 bulletin board; direct  access to the Air and  Radiation Docket
 and  Information Center; distribution of the proposed  rule  and
 preamble to trade associations;  distribution  to plaintiffs in
 court-ordered  regulatory actions;  distribution  to the small
 business ombudsperson system in each State; and,  if necessary,
 through contact with the EPA.   The general response to this
 process has so far been positive.  In addition, the commenter
 was provided a copy of  the proposed  rule at the time  of
 proposal.
 2.10  MISCELLANEOUS
     Several comments were received  on a number of  issues
 relating to the proposed standards that do not  fit  into  the
 categories presented above.   These comments are presented  in
 the following sections.
 2.10.1  Definitions
     Coininent:   One commenter (IV-D-06) recommended that  a
 definition of THC be included in the definitions in the
proposed standards.  The commenter pointed out  that the  test
method proposed for determining compliance with the proposed
THC emission standards  (Method  25A) does not preclude methane
 or ethane.   Both of these compounds are excluded from the
definition of volatile  organic  compounds (VOC).  If the  EPA

                             2-91

-------
does not intend to require the inclusion of these compounds,
then the definition of THC should correspond to the definition
of VOC by including the same list of exempt compounds.
     Response:  The statutory definition of VOC (40 CFR 51)
includes a list of compounds that are negligibly
photochemically reactive, including methane and ethane.  This
definition is intended for developing inventories of VOC
emissions that contribute to the formation of ozone and bears
no regulatory relationship to THC as measured by reference
Method 25A.  Method 25A measures total gaseous organic
concentration (i.e., THC) expressed in terms of propane, and
this measurement includes methane and ethane.  The EPA is
establishing emission limits for THC as a surrogate for total
organic HAP's and is not regulating individual organic
species.  Because THC is determined by an EPA reference test
method, there is no need for a definition of THC analogous to
the definition of VOC.
2-10.2  Compliance Dates
     Three commenters (IV-D-01,  IV-D-02, and IV-D-14)
commented on the proposed compliance dates.
     Comment:  One commenter (IV-D-02) requested that
compliance with the proposed standards be phased in as quickly
as practical, rather than requiring compliance only within
2 years of the promulgation date.  The commenter noted that
compliance with some of the requirements, especially those
that result in little cost to the smelters, could be achieved
before 2 years and that phasing in compliance would achieve
emission reductions sooner.  The same commenter noted,
however, that blast furnace smelters may need more than
2 years if they are required to meet a more stringent THC
limit of 20 ppmv proposed by the commenter.
     Response;  The EPA selected the proposed 2-year
compliance schedule based on the estimated period of time a
smelter would need to design, fabricate, install,  and test a
new add-on control device such as an afterburner,  baghouse,  or
scrubber.  Compliance with the standards for nearly all of the

                             2-92

-------
 pollutant  types  and  emission  sources regulated by  this NESHAP
 can be achieved  through  the use of add-on controls that may
 require 2  years  to implement.  Compliance with some standards
 can also be achieved through  process modifications or work
 practice controls.   It would  be impractical and would achieve
 little additional emission reductions to specify different
 compliance dates that would be dependent on which  type of
 controls the smelter operator selected.  Therefore, the
 proposed 2-year  compliance schedule has not been revised.  The
 THC limit  for blast  furnaces  has not been revised and the EPA
 does not believe that merely  extending the compliance date
 would allow smelter  operators to achieve a higher level of
 control,
     Comment:  One commenter  (IV-D-14)  noted that a smelter's
 schedule to design and install appropriate control equipment
 will depend on the speed with which the applicable permitting
 authority  approves necessary permits.  Furthermore, these
 permit approvals are often delayed for reasons beyond the
 smelter's  control.   Therefore, the commenter recommended that
 the EPA revise the compliance date to be no earlier than
 18 months  following  the date that all necessary permits are
 approved.
     Response:  Section €3.6(i)(4)  of the General Provisions
 allows operators of  existing sources to request an extension
 of up to one additional year to comply with emission standards
 if the additional period is needed to install air pollution
 controls.  Adopting  the approach recommended by the commenter
 would require the EPA to adopt provisions to ensure that the
 operators  have submitted appropriate permit applications in a
 timely manner,  but such provisions would be impractical to
 implement.   Therefore,  the EPA did not  revise the 2-year
 compliance schedule.
     Comment:   One commenter  (IV-D-Ol)  argued that data on the
public health risks associated with secondary lead smelters
 indicate that these risks are low enough to support a decision
 to allow the industry 3 years to install MACT controls and

                             2-93

-------
come into compliance with the proposed standards.  The
proposed regulation allows only 2 years.  The commenter
referred to section 112(i)(3) of the Act.
     Response:   There is nothing in section 112 (i) (3) of the
Act to indicate that public health risks should be considered
in establishing the compliance schedule for this NESHAP.  The
compliance schedule for this NESHAP was based on the estimated
time needed to implement MACT controls.  No information has
been provided to indicate that this schedule is not feasible,
so it has not been revised.
2.10.3  Presentationof the Standards
     Comment:  One commenter  (IV-D-22) requested that the
regulation include a summary table of the numerical emission
limits.  The commenter noted that this would allow a facility
to quickly and easily locate all of the emission limitations
that apply to a given configuration of equipment.
     Response:   The EPA agrees that a summary table of the
numerical emission limits would be useful and one has been
included in the final regulation.
2.10.4  Impacts of a Loss in Lead-Recycling Capacity
     Comment:  One commenter  (IV-D-27) expressed concern that
overly stringent emission standards for secondary lead
smelters could cause facility closures, which would lead to a
reduction in smelter capacity and a reduction in the recycling
of scrap lead.   According to the commenter, this would lead to
greater disposal of lead scrap in landfills and more mining of
virgin lead.  The commenter requested that the EPA consider
the likelihood of a decrease in recycling of lead scrap if
overly stringent emission standards are adopted for secondary
lead smelters.   The commenter also requested that the EPA
establish emission standards that do not result in the
reduction of secondary lead smelter capacity,  but did not
propose how the EPA should accomplish this.
     Response '•   The proposal preamble discusses the effects of
the standards on the total industry lead smelting capacity
(59 FR 29760).   Total smelting capacity for the industry is

                             2-94

-------
expected to decrease by less than 1 percent.  It is also not
expected that the proposed NESHAP would reduce the
opportunities for recycling scrap lead or lead-acid batteries.
2.10.5  Environmental Justice
     Two commenters  {IV-D-03 and IV-D-10) submitted comments
on the proposed secondary lead NESHAP and the issue of
environmental justice.  One commenter (IV-D-03) asserted that
there have been disproportionate impacts on the communities of
color and poor people in two communities in Texas, one of
which is the former site of a secondary lead smelter and the
second of which is the site of an operating secondary lead
smelter.  The commenter claimed that these are both examples
of "lead-dumping" on poor people of color due to lead smelter
emissions.  The commenter also expressed concern that the
proposed NESHAP has too many loopholes that significantly
weaken the ability of the proposal to help bring an end to
environmental racism.
     According to the commenter (IV-D-03),  the draft rule can
be improved to help end environmental injustice in minority
communities and poor neighborhoods by requiring cumulative
emission reviews and modeling.  The commenter also noted that
the current proposal allows blast furnace smelters to have
higher organic HAP emissions of 1.34 million pounds and that
this will allow the poor and minorities to continue to suffer
environmental injustice.   The commenter also noted that the
EPA has failed to calculate in the proposed NESHAP the
economic and social costs of increased incidence of cancer,
lung disease,  and other health problems in areas around
secondary lead smelters.
     Another commenter (IV-D-10)  noted that many lead
nonattainment areas are located in urban areas with low income
residents and,  therefore,  the regulation of these sources has
important implications for environmental justice.
     Response:   The former site of the secondary lead smelter
in West Dallas,  Texas has been cleaned up (remediated)  and the
smelter has been dismantled.   No further exposure from this

                             2-95

-------
site is possible.  With respect to the second smelter located
in Texas, the EPA cannot alter the location of an existing
nm.Trce,  However, the proposed and final NESHAP achieves
significant reductions of HAP's and related criteria
pollutants.  The EPA is not aware of any loopholes in the
NESHAP that would prevent these reductions from being
achieved.
     Under the final NESHAP, organic HAP emissions from blast
furnace smelters will total no more than 302 Mg (332 tons) per
year.  The commenter did not provide the basis for the claim
that the proposed and final NESHAP would allow blast furnace
smelters to have excess organic HAP emissions of 1.34 million
pounds (610 Mg) compared to a more stringent standard.  The
final NESHAP should achieve reductions in lead compound
ciu-Looa.w^_ . .    __                          ~ve dust sources and
these reductions should help reduce ambient lead
concentrations in nonattainment areas.  Furthermore, the
NESHAP will reduce the risks to all individuals near smelters
by regulating emissions from both major and area source
smelters.
     The Act requires that the NESHAP be a technology-based
standard, rather than a health risk-based standard.  There are
no data on the incidence of health problems around secondary
lead smelters on which economic and social costs could be
based.  However,  under the NESHAP, HAP emissions will be
significantly reduced, along with carbon monoxide emissions,
and this will reduce the threat to public health and the
environment presented by smelters.  The commenter did not
provide any details on how cumulative emission reviews and
modeling can be used in conjunction with the NESHAP to help
end environmental injustice.
2.10.6  Pollut ion Prevent ion Cons iderat ions
     Comment:  One commenter (IV-D-03) responded to the EPA's
request for comments on several pollution prevention options
presented in the proposal preamble.  The commenter supported
allowing the smelters to pursue each of the pollution

                             2-96

-------
prevention options, as long as compliance with the proposed
emission standards is demonstrated.  With respect to plastic
separation to reduce organic HAP and HC1/C12 emissions, the
commenter argued that recycling options for these plastics
must continue to be pursued as long as these materials are
being received by smelters.
     Response:  Nothing in the proposed or final rule will
prevent or discourage smelters from pursuing pollution
prevention options.  In fact, the reduction of HC1/C12
emissions will result from the replacement of PVC separators
with those of other materials.  Even though this was
undertaken by battery manufactures and not the smelters, this
is an example of pollution prevention.  Regardless of the
control technology adopted, all smelters must demonstrate
compliance with the emission standards and monitoring
requirements of the final rule.
2.10.7  Section	112	Prohibition ofAdHoc Standards
     Comment:  One commenter  (IV-D-16) objected to the
proposed monitoring requirements because they violate the
prohibition against ad hoc standards in section 112 of the
Act.  The commenter asserted that the proposed opacity
monitoring requirements for compliance with the lead standard
are ad hoc because (1)  there is no correlation between PM and
lead; (2)  there is no correlation between PM and opacity;  and
(3) they would establish lead standards more stringent than
MACT because the opacity limit would be based on the opacity
measured during the initial lead compliance test and would
become an enforceable limit,  even if lead emissions were well
below the proposed lead standard.   The commenter stated that
the proposed monitoring requirements for compliance with the
proposed THC and HC1/C12 standards also constitute ad hoc
standards,  based on the same rationale presented for the
proposed opacity monitoring requirements;  however,  the
commenter did not elaborate further.
                             2-97

-------
     Response:   The EPA does not believe that there is any
legal standing to the term "ad hoc standards" and notes that
there is no mention of this term in section 112 of the Act.
The EPA followed the guidance provided in section 112 of the
Act and in 40 CFR part 70 in developing the emission standards
and monitoring requirements.
     As discussed in detail in section 2.7.1 of this document,
the EPA agrees that COM's cannot be used to demonstrate
compliance with the metal HAP emission standards for process
sources.  The revised metal HAP monitoring requirements are
baghouse leak detectors coupled with baghouse inspection and
maintenance programs to ensure that baghouses are properly
functioning on a continuous basis.  As discussed in
section 2.3.4 of this document, the EPA plans to withdraw the
HC1/C12 emission standards.  Owners and operators may use a
THC CEM during initial compliance tests to adjust control
parameters such that THC emissions are at or slightly below
the levels of the emission standards.  Therefore, owners and
operators are not required to meet emission levels any more
stringent than the final emission standards.
                             2-98

-------
                          APPENDIX A
       SUMMARY OF PUBLIC COMMENTS ON NOTICE TO WITHDRAW
          HYDROCHLORIC ACID/CHLORINE (HC1/C12)  LIMITS

     Information gathered since proposal indicates that
polyvinyl chloride (PVC) plastic contained in lead-acid
battery scrap feedstock has decreased to negligible quantities
in recent years.   The PVC in feedstock is a precursor to HC1
and Cl2 emissions.  A supplemental Federal Register notice was
published on April 19, 1995 {60 FR 19556) to amend the
proposed standards (59 FR 297750)  by withdrawing the HC1
limits.  A total  of eight comment letters were received in
response to the supplemental notice.   Six commenters supported
the amendment (IV-D-39, IV-D-40, IV-D-41, IV-D-43, IV-D-45,
and IV-D-46),  while two commenters opposed it (IV-D-42 and
IV-D-44).   A list of  all of the commenters,  their
affiliations,  and the EPA docket item number assigned to their
correspondence is given in table A-l.

    TABLE A-l.  LIST OF COMMENTERS ON HC1 LIMIT WITHDRAWAL

 Document Item Number  Commenter and Affiliation
IV-D-39                M.L.  Sappington
                       Electrowinning Technologies, Inc.
IV-D-40                J.W.  Walton
                       State of Tennessee,  Department of
                       Environment and Conservation,  Division ot
                       Air Pollution Control
IV-D-41                L.L.  Bergeson
                       Weinberg, Bergeson & Neuman
IV-D-42                J.P.  Valenti
                       Disposal Safety Incorporated
IV-D-43                E.  Roy Baggett
                       Sanders Lead  Co.,  Inc.
IV-D-44                N.J.  Carmen
                       Sierra Club,  Loan Star Chapter
IV-D-45                R.N.  Steinwurtzel
                       Association of  Battery Recyclers
IV-D-46                K.  Parameswaren
                       ASARCO Inc.
                             A-l

-------
     Comment:  Six coinmenters  (IV-D-39, IV-D-40, IV-D-41,
IV-D-43, IV-D-45, and IV-D-46) agreed that because batteries
manufactured in the United States are no longer made with PVC
plastic separators, the PVC content in the used lead-acid
battery feedstock is decreasing dramatically and will dwindle
to negligible levels by the time HC1 emission controls would
be installed under the rule.  These commenters argued that
because PVC is the source of chlorides, emissions of HC1 and
Cl2 will decrease even in the absence of a Federal rule.  The
controls are costly and would provide negligible, if any,
environmental benefit.  One commenter  {IV-D-45) stated that
their organization provided the EPA with data showing a
tenfold reduction in the percent of PVC-containing batteries
available for recycling in the United States between 1990 and
1995.
     Response: The EPA agrees with the commenters that HC1/C12
emissions will decline because PVC is no longer used as a
separator material.
     Comment;  One commenter  (IV-D-39) supported the
withdrawal of the HC1/C12 standard because an HC1/C12 standard
would discourage the use of pollution prevention technologies
at secondary lead smelters.  According to the commenter, the
economic benefits of battery paste desulfurization, which
increases furnace efficiency and controls S02 emissions, would
be diminished because a scrubber to control HC1/C12 would also
control SC>2 emissions.  Consequently,  installing scrubbers
would halt the move toward battery paste desulfurization
technology as an S02 emission control.  Battery paste
desulfurization provides a significant reduction in energy use
in the furnace and offers the ability to recover a marketable
product instead of generating wastes  (sludge and wastewater)
from a scrubber.
     The commenter {IV-D-39) argued that emission control
strategies that prevent pollution generation are preferable to
those that remove the pollutants at the point of release.
Compared to paste desulfurization, scrubbers adversely affect

                              A-2

-------
the environment by generating increased wastewater and solid
waste and by increasing energy consumption and combustion gas
emissions at off-site power plants.
     According to the same commenter  (IV-D-39), feed
desulfurization is also an essential initial processing step
if the electrolytic recovery of lead from battery paste is to
be implemented.  Recovery of lead through electrowinning is an
important emerging technology which has many environmental
advantages and has strong EPA and State level support.  Lead
electrowinning would greatly reduce the air emissions
associated with pyrometallurgical furnaces and minimize the
production of slag which must be land disposed.  Since feed
desulfurization is the first step in this process, any move
away from desulfurization in the industry would delay or
potentially stop progress toward the commercialization of this
environmentally superior process, according to the commenter.
     Response: The EPA agrees that the HC1/C12 standard should
be withdrawn.  The EPA also agrees that battery paste
desulfurization may be used to control some fraction of SC>2
emissions.  However,  EPA data indicate that some secondary
lead smelters that practice battery paste desulfurization also
have SC>2 scrubbers (see docket items II-B-5 and II-B-7) .
Therefore, the impact of an HC1/C12 emission standard on the
practice of paste desulfurization is difficult to predict.
However, the EPA agrees that,  compared to paste
desulfurization,  scrubbers may increase energy consumption and
generate increased wastewater and solid waste.
     The EPA agrees that lead electrowinning would reduce air
emissions associated with pyrometallurgical furnaces.
However, because HC1/C12 emissions are produced from
combustion processes,  electrowinning technologies would
produce no HC1/C12 emissions.   Consequently,  an HC1/C12
standard would not be expected to discourage the use of this
technology.   In any case,  the EPA is not promulgating HCl/Clj
emission standards,  nor are any of the standards applicable to
                              A-3

-------
emerging non-pyrometallurgical processes such as
electrowinning.
     Comment:  Two commenters  (IV-D-43 and IV-D-45) supported
the withdrawal of the HC1/C12 standard due to the cost of
scrubbers.  One of these commenters  (IV-D-43) stated that no
economic analysis had been conducted on the impact of
installing scrubbers to comply with the standard.  The other
commenter  (IV-D-45) stated that the high cost of scrubbers is
unjustified because the effect"of these scrubbers on HC1/C12
emissions would be negligible due to the decrease of PVC
content in the battery feedstock.
     Response:  As stated in section 2.3.4,3, the EPA
previously believed that fluxing could be used to control
HC1 /n_
-------
regulate HC1/C12 emissions.  The commenter requested that the
EPA establish standards that result in the maximum control of
acid gases from secondary lead smelters by mandating
technologies such as flue gas scrubbers and practices such as
battery paste desulfurization and furnace fluxing to remove
chlorides.
     Response: The EPA has received data  (see docket
item IV-D-47) indicating that the personal protective
equipment in a smelter's feedstock amounts to less than 1 ton
per year.  If all of this equipment consisted of PVC and all
available chlorides were emitted, less than 0.5 tons per year
of HC1/C12 emissions would result.  However, the actual PVC
and chloride content of personal protective equipment is
significantly below the level estimated by the commenter.  For
this reason and because of the decrease in the PVC content of
the battery feedstock,  the EPA has concluded that total
HC1/C12 emissions from secondary lead smelters will be
negligible and no HC1/C12 standard is warranted.
     The EPA is considering under a separate action whether
the disposal of PPE in the smelting furnaces is allowed under
the BIF rule.
     Scrubbers presently installed in the secondary lead
smelting industry are in place due to  Prevention of
Significant Deterioration or New Source Review permitting
activities.   Sulfur dioxide is not a HAP and the EPA is
prohibited from using an HC1/C12 standard as a means of
enforcing S02 control requirements.
     Comment:  One commenter (IV-D-44)  agreed that HC1/C12
emissions are declining as a result  of the decrease of PVC
battery separators in the feedstock.   The commenter stated,
however,  that the EPA should require the HC1/C12 standard for
a period of 3 to 5 years,  or until sources can demonstrate no
HC1/C12 emissions using emission performance tests.   The
commenter stated several reasons for retaining the HC1/C12
standard during this period.   According to the commenter,
HC1/C12 is  easy to control and the required controls are not

                             A-5

-------
too expensive "as long as a facility is producing HC1 gases."
Because HC1/C12 is a strong acid, its emissions cause adverse
heaitn ettects and property damage.  These emissions are of
particular concern with respect to environmental justice
because many smelters are located in low-income, minority
neighborhoods, according to the commenter.
     Response:  Data obtained by the EPA indicate that the
percent of PVC-containing batteries available for recycling in
the United States decreased tenfold between 1990 and 1994,
with PVC content accounting for less than one tenth of one
percent of the total battery feedstock (see docket
item IV-D-34).  Because PVC separators are no longer used in
batteries, this trend will continue and the PVC content of the
battery feedstock will decline to negligible levels by the
1997 effective date of this rule.  Because the PVC content of
the battery feedstock is the source of HC1/C12 emissions, an
HC1/C12 standard would have negligible, if any, environmental
benefits.  Consequently, HC1/C12 controls, performance tests,
or temporary emission standards would not be economically
justifiable.
     For example,  a large secondary lead smelter (100,000 tons
per year lead production capacity)  would process about
40,000 lead-acid automobile batteries per day.  In 1990, about
1 percent of the batteries available for recycling
(400 batteries per day at a large smelter) contained PVC
plastic separators.  Each battery with PVC separators would
contain about l pound of PVC separator material.  The PVC used
in battery separators is nearly all PVC homopolymer with very
little modifiers added, so this PVC would be about 57 percent
chlorine by weight.  Therefore, each battery with PVC
separators would contain about 0.6 pounds of chlorine.  Based
on these numbers,  a large smelter in 1990 would emit about
240 pounds of HC1/C12 per day, or about 45 tons per year.
This estimate assumes that all of the available chlorides are
released as HC1/C12 air emissions and none are retained in the
slag.

                              A-6

-------
     In 1994, the frequency ct batteries available for
recycling that contained PVC separators had dropped to
0.1 percent of all batteries available for recycling.  This
frequency translates to a maximum HC1/C12 emission rate from a
large smelter of about 4 tons per year.  Because no new
automotive batteries are being manufactured with PVC
separators, this trend will continue and HC1/C12 emissions
from a large smelter will be no more than 1 or 2 tons per year
by the rule's effective date in 1997.  Emissions of HC1/C12
will continue to decline after that date as very old batteries
with PVC separators become less and less frequent among those
available for recycling.
                             A-7

-------

-------