February
17,
2003
I
ron
and
Steel
Foundries
NESHAP
Docket
EPA
Docket
Center
(
Air
Docket)
U.
S.
EPA
West
(
MD­
6102T)
Room
B­
108
1200
Pennsylvania
Avenue,
NW
Washington,
DC
20460
Electronic
Submittal
(
no
attachments)
Paper
Copy
to
Follow
by
Mail
(
with
attachments)

Attention:
Docket
ID
No.
OAR­
2002­
0034
Dear
Sir
or
Madam:

United
States
Pipe
and
Foundry
Co.,
Inc.
(
U.
S.
Pipe)
is
the
largest
of
four
domestic
producers
of
ductile
iron
pressure
pipe
and
has
six
facilities
in
four
states;
Alabama,
Tennessee,
New
Jersey
and
California.
Last
year
these
producers
manufactured
1.8
million
tons
of
ductile
iron
pipe
for
the
distribution
of
potable
water
in
the
U.
S.
and
for
export
markets.

U.
S.
Pipe
operates
two
significantly
different
types
of
foundries.
One
is
for
the
production
of
ductile
iron
pipe
and
utilizes
water­
cooled
steel
molds
and
very
little
chemical
or
naturally
bonded
sand.
The
other
is
the
more
traditional
kind
of
foundry
utilizing
sand
molds
and
cores
with
both
natural
and
chemically
bonded
sands
for
the
production
of
fittings,
valves
and
hydrants.

U.
S.
Pipe
has
a
long
history
of
involvement
with
EPA's
development
of
the
MACT
and
has
hosted
the
EPA
in
visits
to
our
facilities.
We
are
pleased
to
submit
the
following
comments
on
the
proposed
rule:
National
Emission
Standards
for
Hazardous
Air
Pollutants
for
Iron
and
Steel
Foundries;
Federal
Register,
December
23,
2002
(
Volume
67,
Number
246).

First,
we
will
make
several
comments
related
to
specific
pages
in
the
preamble;
then
comments
will
be
addressed
to
specific
sections
of
the
proposed
rule.

The
Preamble
Preamble
page
78276
While
permanent
mold
ductile
iron
pipe
foundries
do
share
some
processes
and
similarities
with
traditional
iron
casting
sand
foundries,
there
are
substantial
differences
in
emissions
from
pouring,
cooling
and
shakeout
(
PCS).
Pipe
foundries
do
not
use
green
sand,
a
common
molding
material,
thus
eliminating
the
2­
10%
sea
coal,
petroleum
products,
organic
flours
and
clays
that
contribute
to
the
organic
HAP
emissions
from
sand
foundries.
Pipe
foundries
do
use
cores
with
binder
chemicals
in
them
but
at
much
smaller
sand
 
to­
iron
ratios
than
sand
foundries.

During
the
development
of
the
Iron
and
Steel
Foundry
MACT
standard,
USEPA
reviewed
the
melting
operations
(
cupola,
electric
arc
furnace,
and
electric
induction
furnace)
separately,
even
though
each
operation
melts
scrap
metal.
USEPA
even
identified
and
created
limits
for
different
core
systems
such
as
polyurethane
no­
bake
and
furan
core
systems.
Different
mold
systems
were
identified
in
the
preamble
(
green
sand
and
permanent
molds)
but
were
not
individually
reviewed.
We
think
more
effort
should
have
been
expended
to
understand
the
very
different
pipe
making
process.
This
will
be
discussed
in
more
detail
in
our
comments.
2
Preamble
page
78277
In
its
discussion
of
PCS,
EPA
clearly
distinguishes
between
operations
that
bring
molds
to
the
metal
and
those
that
bring
metal
to
the
molds.
This
differentiation
is
important
in
separating
area
sources,
usually
small
foundries
with
pallet
lines
where
no
fixed
pouring
station
exists
and
where
poured
molds
are
moved
manually
to
the
shakeout,
from,
potentially,
major
sources
which
pour
large
castings
in
floor
or
pit
areas
and
molds
and
iron
are
moved
by
overhead
crane
in
buildings
too
large
to
reasonably
capture
and
control.
In
either
case,
pallet
line
or
floor/
pit
pouring,
it
is
often
necessary
to
move
or
reposition
poured
molds
in
a
pouring
area;
thus,
EPA's
requirement
that
molds
remain
stationary
after
pouring
must
be
deleted
to
avoid
compliance
enforcement
confusion.

I
t
is
stated
" 
most
cupolas
employ
afterburners
which
effectively
destroy
organic
HAP."
While
coincidental
destruction
of
HAP
does
occur,
cupola
afterburners
were
installed
primarily
due
to
safety
concerns
related
to
the
uncontrolled
explosive
combustion
of
CO
in
the
confined
space
above
the
cupola
and/
or
the
associated
pollution
control
devices.
Cupola
afterburners
are
designed
to
safely
remove
CO
through
controlled
combustion
with
the
ancillary
recovery
of
heat
for
recuperative
preheat
of
cupola
combustion
air.
EPA
does
not
define
"
afterburner"
either
physically
or
operationally.
Since
VOCs
and
organic
HAPs
are
typically
destroyed
at
a
temperature
lower
than
CO,
organic
HAP
reduction
also
occurs.
EPA,
however,
offers
no
definition
for
"
afterburner".
We
intend
to
propose
a
definition.

Preamble
page
78279
The
preamble
discusses
the
use
of
carbon
monoxide
(
CO)
as
a
surrogate
for
organic
HAP
emissions
from
a
cupola.
CO
is
generated
by
the
operation
of
a
cupola.
Uncontrolled
concentrations
vary
with
operating
conditions,
scrap
materials
being
melted,
coke
sources
and
seasonal
adjustments
to
the
melting
process.
Absolute
outlet
concentration
requirements
will
vary
even
under
operation
of
the
very
best
afterburner
systems.
It
is
more
appropriate
to
require
that
a
minimum
temperature
be
maintained
for
afterburner
systems.

The
requirement
of
a
set
face
velocity
for
the
design
of
capture
and
collection
systems
defies
all
logic
in
engineering
design.
Designs
must
be
based
on
the
source
size
and
configuration,
plume
characteristics
such
as
temperature
and
contaminant
properties,
safety
and
health
considerations
and
physical
constraints
affecting
hood
design
and
location.
Control
velocities
might
be
more
or
less
than
200fpm,
but
it
is
clearly
absurd
to
specify
a
design
number
without
considering
any
of
the
above­
mentioned
design
constraints.

EPA
has
proposed
work
practice
standards
for
inspecting
scrap.
Scrap
inspection
cannot
eliminate
metallic
HAP
except
when
they
are
in
a
form
easily
recognized
on
the
surface
of
a
rail
car
or
truckload
shipment,
e.
g.
shiny
stainless
steel
containing
chrome,
an
unwanted
alloy
contaminant.
HAPs
are
trace
components
in
the
metal
scrap
and
numeric
limits
on
HAP
in
scrap
metal
cannot
be
established.
Some
HAPs
are
valuable
components
in
the
base
iron
for
the
product
requirements.
Some
HAP
metals
are
necessary
to
the
melting
process
and
are
not
only
impurities
in
the
scrap
but
required
elements
for
the
process.
Shipments
can
only
be
inspected
on
the
surface,
as
they
are
now
by
most
foundries,
but
only
rejected
after
an
analysis
upon
melting.
An
inspection
program
fraught
with
potential
compliance
uncertainties
will
achieve
no
environmental
benefit.

While
scrap
inspection
programs
have
been
established
at
many
foundries
to
control
the
quality
and
properties
on
incoming
scrap,
much
of
the
inspection
and
rejection
burden
lies
on
suppliers.
Cursory
overview
of
arriving
loads
of
scrap
does
not
provide
control
of
HAPs.
The
inspection
program
that
the
EPA
has
laid
out
here
is
ambiguous
at
best
 
by
their
own
comments.
This
program
only
serves
to
place
an
extra
burden
on
the
foundry
industry.
Because
of
the
ambiguity,
it
will
be
both
difficult
to
comply
with
and
difficult
to
enforce.
U.
S.
Pipe
supports
the
elimination
of
the
scrap
inspection
program
since
no
floor
has
been
established
for
HAP
reduction.
3
EPA
is
proposing
a
requirement
to
" 
manually
ignite
gases "
from
molds
when
no
data
exists
demonstrating
HAP
reduction.
Also,
some
molds
may
not
light
due
to
insufficient
concentration
of
combustible
gases.
U.
S.
Pipe
requests
that
this
requirement
be
deleted.

EPA
is
proposing
using
production
foundries
for
experimentation
on
reduced
HAP
binder
formulations.
This
is
not
acceptable
when
an
industry
and
government
funded
experimental
foundry
exists
to
do
just
this.
The
Casting
Emission
Reduction
Program
(
CERP)
foundry
in
California,
with
EPA
as
a
participant,
is
continually
working
with
industry
suppliers
and
foundries
to
development
such
materials
and
is
the
proper
place
for
such
experimentation
to
take
place.
U.
S.
Pipe
supports
the
removal
of
all
requirements
for
production
foundries
to
be
used
as
laboratories
for
new
product
testing.
Any
responsibilities
for
new
product
development
and
testing
should
rest
with
the
suppliers.

Preamble
page
78280
It
is
not
practical
to
require
a
facility
to
test
for
three
(
3)
consecutive
operating
hours.
Equipment
downtime,
productivity,
safety
and
other
factors
typically
require
some
foundry
processes
to
be
temporarily
placed
on
stand­
by.

Preamble
page
78281
The
agency
indicates
that
"
Although
some
variation
exists
in
these
(
foundry
processes)
operations
at
different
foundries,
these
variations
do
not
significantly
alter
the
nature
or
amount
of
the
HAP
emissions
from
the
individual
emissions
sources,
the
types
of
HAP
emitted,
or
the
control
technology
typically
used
to
reduce
HAP
emissions."
U.
S.
Pipe
disagrees.
In
fact,
U.
S.
Pipe
and
The
American
Foundry
Society
(
AFS)
have
argued
that
pipe
foundries
should
be
subcategorized
based
on
the
inherently
different
operations
at
PCS
compared
to
a
traditional
sand
foundry.
A
case
for
subcategorization
of
pipe
foundries
was
presented
to
EPA
and
is
contained
in
the
docket
at
page
II­
D­
10,
Document
No.
II­
D­
108.

During
the
development
of
the
Iron
and
Steel
Foundry
MACT
standard,
USEPA
reviewed
the
melting
operations
(
cupola,
electric
arc
furnace,
and
electric
induction
furnace)
separately,
even
though
each
operation
melts
scrap
metal.
USEPA
even
identified
and
created
different
requirements
for
different
core
systems
such
as
polyurethane
no­
bake
and
furan
core
systems.
Different
mold
systems
were
identified
in
the
preamble
(
green
sand
and
permanent
molds)
but
were
not
individually
reviewed.

Permanent
and
semi­
permanent
molds
should
have
been
reviewed
and
identified
as
insignificant
HAP
sources
since
the
core
sand
to
metal
ratio
is
small.
USEPA
identified
in
the
preamble
that
they
are
concerned
about
organic
HAPs
from
molds
after
molten
metal
is
poured.
Permanent
molds
consist
of
water­
cooled
steel
molds.
Typically,
the
core
sand­
to­
metal
ratio
in
a
permanent
mold
pipe
foundry
averages
about
0.017/
1.
As
the
following
table
shows,
the
ranges
of
sand­
to­
metal
ratios
in
a
typical
green
sand
foundry
are
as
high
as
14/
1
for
mold
sand
and
2.5/
1
for
core
sand.
Permanent
mold
pipe
making
requires
no
mold
sand
thus
making
the
difference
even
greater,
i.
e.
16.4
 
9.1
/
1.
Semipermanent
molds
include
graphite
molds.

TABLE
TYPICAL
GREEN
SAND
FOUNDRY
SAND
TO
METAL
RATIOS
BY
WEIGHT
CORE
MOLD
TOTAL
SAND
14"
Flg
Ell
.90/
1
8.1/
1
9.0/
1
24"
Flg
Ell
1.25/
1
8.2/
1
9.45/
1
30"
Flg
Ell
1.4/
1
6.7/
1
8.1/
1
36"
Flg
Ell
1.5/
1
6.2/
1
7.7/
1
42"
Flg
Ell
1.6/
1
8.1/
1
9.7/
1
60"
Flg
Ell
2.5/
1
13.9/
1
16.4/
1
4
TYPICAL
PERMANENT
MOLD
CASTING
SAND
TO
METAL
RATIOS
BY
WEIGHT
CORE
MOLD
TOTAL
SAND
Avg.
Shell
Core
.017/
1
0.0
.017/
1
Avg.
Cold
Box
.040/
1
0.0
.040/
1
With
the
exception
of
the
cores,
which
are
regulated
in
a
separate
section
to
this
standard,
permanent
steel
molds
are
similar
to
the
continuous
casting
operation
at
integrated
iron
and
steel
manufacturing
facilities.
Both
operations
cast
into
water­
cooled
steel
molds.
Casting
operations
for
integrated
iron
and
steel
manufacturers
are
not
regulated
under
that
industry's
proposed
MACT
standard.

Language
should
be
written
that
identifies
non­
organic
permanent
and
semi­
permanent
molds
as
low
emitters
of
HAP
and,
therefore,
not
required
to
meet
the
standards
for
pouring
areas,
pouring
stations
and
pouring,
cooling
and
shakeout
operations.

The
agency
indicates,
that
in
addition
to
those
sources
proposed
for
regulation
under
this
NESHAP,
" 
numerous
other
ancillary
emission
sources
that
may
contain
trace
quantities
of
HAP." 
would
not
require
regulatory
limits.
As
a
result
of
this
review
the
agency
determined
that
HAP
emissions
from
sand
handling
systems,
mechanical
finishing
operations,
metal
treatment
operations,
holding
furnaces,
and
HAP
metals
from
cooling
lines
and
shakeout
stations
are
de
minimis
and
do
not
require
regulatory
limitations
under
the
NESHAP.
U.
S.
Pipe
agrees
with
the
agency
that
these
sources
are
unlikely
to
contribute
to
ambient
HAP
emissions
from
iron
and
steel
foundries
and,
therefore,
supports
the
agency's
approach.

We
support
and
concur
with
the
EPA
comment
noting
that
binder
systems
are
selected
based
on
performance
characteristics
and
production
requirements.

U.
S.
Pipe
supports
the
determination
to
not
control
holding
furnaces.

Preamble
page
78282
 
3
We
support
not
establishing
opacity
limits
on
exhaust
stacks
because
emission
limitations
for
the
respective
processes
have
already
been
established
in
the
standard.

U.
S.
Pipe
requests
that
EPA
evaluate
long
form
data
for
foundries
that
" 
specify
scrap
selection
as
a
work
practice
to
reduce
emissions ".
This
statement
is
contrary
to
our
experience.
Scrap
inspection
programs
are
in
place
for
quality
interests
only
and
are
not
effective
for
HAP
reduction.
We
agree
that
"
Visual
inspections
are
only
able
to
identify
obvious
off­
specification
materials "
U.
S.
Pipe
also
believes
that
the
most
responsible
process
in
which
to
recycle
oil
and
grease
contaminated
scrap
is
in
a
wellcontrolled
cupola
melting
operation
such
as
found
in
our
plants.
Scrap
inspection
programs,
as
envisioned
here,
could
potentially
affect
the
nation's
scrap
recycling
efforts.

Preamble
page
78284­
7
MACT
for
organic
HAP
emissions
The
preamble
discusses
the
use
of
carbon
monoxide
(
CO)
as
a
surrogate
for
organic
HAP
emissions
from
a
cupola.
CO
is
generated
by
the
operation
of
a
cupola.
Uncontrolled
concentrations
vary
with
operating
conditions,
scrap
materials
being
melted,
coke
sources
and
seasonal
adjustments
to
the
melting
process.
Absolute
outlet
concentration
requirements
will
vary
even
under
operation
of
the
very
best
afterburner
systems.
Since
VOCs
and
organic
HAPs
are
typically
destroyed
at
a
temperature
below
that
of
CO,
it
is
more
than
sufficient
to
require
a
minimum
temperature
and
residence
time
be
maintained
for
afterburner
systems.
5
EPA
reviewed
various
State
requirements
that
regulate
CO
emissions
from
cupolas.
Specifically,
it
cites
the
State
of
Illinois
emission
standard
(
i.
e.,
200
ppmv
CO)
and
identifies
this
limitation
as
"
the
most
stringent
and,
therefore,
the
most
effective
in
organic
HAP
emissions
reduction."
U.
S.
Pipe
believes
that
the
use
of
the
State
of
Illinois
standard
is
inappropriate
since
this
standard
was
contested,
found
to
be
improperly
derived,
and
to
the
best
of
our
knowledge
has
never
been
enforced.

When
this
standard
was
promulgated
it
would
have
been
applicable
to
the
General
Motors
(
GM)
facility
in
Danville,
Illinois.
After
reviewing
the
basis
for
the
standard,
GM
initiated
legal
proceedings
to
contest
this
limitation.
In
1992,
it
was
determined
that
the
basis
for
this
standard
was
severely
flawed
and,
therefore,
GM's
facility
was
issued
an
exception
from
compliance
and
issued
a
much
higher
emission
limitation
(
i.
e.,
2,000
ppmv
CO).
(
See
Illinois
Rules,
Section
216.382,
Exception,
General
Motor's
Ferrous
Foundry
in
Vermilion
County,
added
at
16
Ill.
Reg.
18075,
effective
November
13,
1992).

Therefore,
U.
S.
Pipe
requests
that
the
agency
remove
the
State
of
Illinois
standard
from
the
determination
of
the
MACT
floor.
The
determination
of
the
MACT
floor
for
organic
HAP
emissions
from
cupolas
is
further
discussed
below
in
comments
regarding
§
63.7690(
a)(
5).

MACT
for
HAP
metal
emissions
EPA
states
that
"
We
have
credible
emissions
source
test
data
for
 
12
(
cupolas)
controlled
by
baghouses ".
Our
analysis
of
this
data
in
Appendix
D
of
the
"
Background
Information
for
Proposed
Standards"
raises
concerns
over
the
credibility
of
some
of
this
information.
Three
tests
from
Foundry
WI­
35
were
used
in
the
MACT
floor
determination.
These
tests
all
indicate
such
low
measured
values
as
to
question
whether
sufficient
sample
was
taken
to
achieve
method
detection
limitations.
Also,
only
one
of
these
three
tests
reports
the
cupola
melt
rate
during
testing.
Was
this
a
newly
constructed
unit
with
no
operational
history?
Such
consistently
low
emission
numbers
are
unusual
and
should
be
analyzed
by
EPA
before
using
in
the
MACT
floor
determination.
This
test
could
not
have
been
from
a
system
performing
at
its
"
worst
foreseeable
circumstances"
(
discussed
later).

The
data
from
Foundry
MI­
26
reported
in
Appendix
D
and
also
in
the
docket
OAR­
2002­
0034,
Document
No.
II­
D­
110
is
not
credible
based
on
the
test
report
which
states
on
page
4:
"
Because
of
the
sampling
location,
(
outside,
horizontal
monovent,
Harsell
design)
the
samples
were
collected
anisokinetically."
This
means
the
sample
taken
was
not
representative
since
it
was
taken
at
a
non­
proportional
flow
rate.
Also,
"
The
baghouse
doesn't
follow
the
examples
described
in
Method
5D,
so
a
modified
version
of
the
method
was
employed."
For
these
reasons
these
test
results
should
not
be
considered
in
developing
the
MACT
floor.
See
the
attached
test
report.

EPA
states
that
state
PM
emissions
limitations
are
"
much
more
lenient
than
actual
emissions."
U.
S.
Pipe
believes
that
this
statement
is
misleading
and
inaccurate
and
should
not
be
the
basis
for
excluding
this
information
from
the
determination
of
the
MACT
floor.
Specifically,
in
a
footnote
to
the
preamble,
the
agency
indicates
that
the
three
state
programs
it
reviewed
have
limits
at
0.08
gr/
dscf
or
higher,
and
that
27
of
34
cupolas
tested
were
measured
at
0.07
gr/
dscf
or
lower.
In
review,
these
data
do
not
demonstrate
these
state
standards
to
be
"
lenient"
but,
rather,
standards
set
at
a
level
achievable
"
under
the
most
adverse
circumstances
which
can
reasonably
be
expected
to
recur;"
which
is
the
standard
required
under
the
NESHAP
process.
Even
a
cursory
review
of
these
data
shows
that
seven
(
7)
units
(
i.
e.,
34
minus
27)
(
i.
e.,
20%)
tested
at
or
above
the
0.08
gr/
dscf
emission
limit.
In
addition,
those
units
that
tested
at
0.07
gr/
dscf
are
operating
at
87.5
percent
of
the
emissions
level
allowed
under
these
State
standards.
In
the
light
of
this
brief
analysis,
U.
S.
Pipe
finds
no
basis
for
the
agency
to
describe
these
state
limitations
as
"
much
more
lenient
than
actual
emissions"
and
requests
that
the
agency
consider
these
data
in
the
MACT
floor
analysis.

The
agency,
when
proposing
emissions
limitations
for
several
sources,
discusses
and
applies
a
method
to
account
"
for
the
variability
inherent
in
process
operations
and
control
device
performance."
However,
the
preamble
offers
no
indication
that
this
consideration
was
given
to
all
of
the
proposed
regulated
operations
for
each
emissions
limitation.
Further,
U.
S.
Pipe
believes
that
the
agency,
when
accounting
6
for
such
variability,
has
focused
solely
on
the
control
device
performance
and
not
properly
considered
the
variability
inherent
in
the
process
operations
themselves.

The
court,
in
National
Lime
Association
v.
EPA
1
held,
in
part,
that,
"
to
be
achievable,
we
think
a
uniform
standard
must
be
capable
of
being
met
under
most
adverse
conditions
which
can
reasonably
be
expected
to
recur."
In
Sierra
Club
v.
EPA
2,
the
court
applied
this
principle
to
the
NESHAP
rulemakings
and
found
that:
" 
EPA
would
be
justified
in
setting
the
floors
at
a
level
that
is
a
reasonable
estimate
of
the
performance
of
the
"
best
controlled
similar
unit"
under
the
worst
foreseeable
circumstances.
It
is
reasonable
to
suppose
that
if
an
emission
standard
is
as
stringent
as
"
the
emissions
control
that
is
achieved
in
practice"
by
a
particular
unit,
then
that
particular
unit
will
not
violate
the
standard.
This
only
results
if
"
achieved
in
practice"
is
interpreted
to
mean,
"
achieved
under
the
worst
foreseeable
circumstances."

As
a
result
of
these
decisions,
U.
S.
Pipe
believes
that
the
agency
must
consider,
and
account
for,
the
variability
inherent
in
process
operations
and
control
device
performance
when
determining
the
MACT
floor
for
all
sources,
both
existing
and
new,
and
for
all
regulated
pollutants.
However,
in
the
preamble
the
agency
utilizes
a
method
to
account
for
variability
only
in
a
limited
number
of
circumstances.
U.
S.
Pipe
knows
of
no
basis
for
limiting
this
consideration
and,
therefore,
requests
that
the
agency
reconsider
all
of
the
proposed
emissions
limitations
for
all
sources
to
address
this
variability.

While
reviewing
related
issues,
in
Cement
Kiln
Recycling
Coalition
v.
EPA
3,
the
court
cited
and
summarized
the
previous
decision:

Sierra
Club
permits
EPA
to
account
for
variability
by
setting
floors
at
a
level
that
reasonably
estimates
"
the
performance
of
the
`
best
controlled
similar
unit'
under
the
worst
foreseeable
circumstances."

I
n
addition,
the
court
also
found
that
factors
other
than
the
control
technology
identified
as
MACT
may
influence
the
achievability
of
an
emission
limitation
and,
therefore,
should
be
considered
in
the
MACT
floor
determination.
Specifically,
the
court
found
that:

 
whether
variability
in
the
MACT
control
(
technology*)
accurately
estimates
variability
associated
with
the
best­
performing
sources
depends
on
whether
factors
other
than
the
MACT
control
contribute
to
emissions.
In
other
words,
if
factors
other
than
MACT
technology
do
indeed
influence
a
source's
performance,
it
is
not
sufficient
that
EPA
considered
sources
using
only
well­
designed
and
properly
operated
MACT
controls
(
to
determine
the
MACT
floor*).
[*
­
clarifications
added].

Based
upon
the
court's
instructions
in
Cement
Kiln,
U.
S.
Pipe
believes
that
it
is
improper
for
the
agency
to
determine
an
operation's
inherent
variability
based
upon
the
limited
number
of
sources
that
utilize
the
MACT
technology.
U.
S.
Pipe
believes
that
Cement
Kiln
holds
that
the
agency
must
consider
the
process
variability
of
all
similar
sources
within
the
industry,
regardless
of
their
control
technology,
where
such
process
variability
can
influence
emissions.
However,
in
the
preamble
the
agency
repeatedly
notes
that
sources
without
the
MACT
control
technology
were
"
not
used
in
estimating
the
RSD
(
relative
standard
deviation)
because
available
emissions
source
test
data
clearly
demonstrate
that
the
(
sources)
controlled
with
these
devices
were
not
among
the
best
performing
12
percent
of
sources."
Therefore,
U.
S.
Pipe
requests
that
the
agency
reconsider
all
of
the
proposed
emissions
limitations
to
consider
the
process
variability
of
these
operations
throughout
the
industry.

I
n
summary,
we
believe
that
the
agency
must
consider,
and
account
for,
"
the
variability
inherent
in
process
operations
and
control
device
performance"
when
determining
the
MACT
floor
for
all
sources,
both
existing
and
new,
and
for
all
regulated
pollutants.
In
addition,
when
establishing
these
floor
levels,
U.
S.
Pipe
believes
that
the
agency
must
account
for
those
process
variables
encountered
throughout
the
industry,
other
than
MACT
technology,
that
account
for
"
the
most
adverse
circumstances
which
can
1
National
Lime
Ass'n
v.
EPA,
200
U.
S.
App.
D.
C.
363,
627
F.
2d
416,
431
n.
46
(
D.
C.
Cir.
1980).
2
Sierra
Club
v.
EPA,
334
U.
S.
App.
D.
C.
421,
167
F.
3d
658
(
D.
C.
Cir.
1999).
3
Cement
Kiln
Recycling
Coalition
v.
EPA,
255
F.
3d
855
(
D.
C.
Cir.
2001).
7
reasonably
be
expected
to
recur."
Therefore,
U.
S.
Pipe
believes
that
the
emissions
standards
proposed
in
§
63.7690,
in
order
to
be
"
achievable,"
should
be
reconsidered.

Preamble
page
78292,
3
EPA
states
that
the
most
common
control
method
for
organic
HAP
in
a
pouring,
cooling
and
shakeout
area
or
line
is
the
ignition
of
mold
off
gas.
Light
off
is
not
used
for
organic
emission
control
and
no
data
exists,
to
our
knowledge,
which
indicates
this.
Light
off
is
usually
spontaneous
but
frequently
does
not
occur
due
to
insufficient
concentrations
of
mold
gas.
Ignition
cannot
be
ensured.
This
requirement
must
be
eliminated.

Preamble
page
78298
EPA
is
requiring
the
use
of
bag
leak
detectors
on
baghouses
to
determine
when
they
are
performing
properly
on
a
continuous
basis.
Two
of
our
U.
S.
Pipe
members
utilize
Harsell
baghouses
or
an
open
construction,
single
compartment
baghouse
with
no
stack.
Exhaust
is
discharged
through
a
continuous
roof
top
monovent.
Bag
leak
detectors
are
not
feasible
for
this
design
since
the
single
compartment
and
monovent
can
be
120
to
300
feet
long.
State
and
local
Title
V
permit
conditions
require
visible
emission
readings
to
be
taken
and
recorded
and
these
should
be
sufficient
for
compliance
certification.

The
proposed
rule
§
63.7681
 
Sources
Subject
to
this
Subpart
§
63.7682(
b)
­
Identification
and
Definition
of
`
Affected
Sources'
The
proposed
rule,
at
§
63.7682(
b),
divides
foundries
into
two
affected
sources:
a
`
metal
casting
department'
and
a
`
mold
and
core
making
department.'
This
apportionment
does
not
accurately
reflect
the
operation
of
most
foundries.
Specifically,
most
foundries
are
organized
into
three
general
types
of
operations:

`
Mold
and
core
making
departments'
as
defined
in
the
proposed
rule;

`
Melting
departments'
where
raw
metals
are
processed,
melted,
and
the
molten
metal
is
tapped;
and
`
Casting
processing
departments'
where
the
cast
metals
are
poured,
cooled
and
cleaned.

Therefore,
U.
S.
Pipe
recommends
that
the
rule
be
modified
to
reflect
these
departments
as
the
"
affected
sources"
regulated
in
this
NESHAP.

Further,
U.
S.
Pipe
believes
that
the
regulation
should
clarify
that
a
foundry
may
contain
multiple
affected
sources
of
the
same
type.
Specifically,
a
foundry
may
have
multiple
melting
departments
that
are
unrelated
(
e.
g.,
a
cupola
and
a
furnace),
multiple
mold
and
core
making
departments
(
e.
g.,
a
green
sand
system
and
a
furan
warm
box
system,
or
two
separate
green
sand
systems),
and
multiple
casting
processing
departments
(
e.
g.,
a
floor
casting
department
and
a
continuous
casting
department,
or
several
unrelated
continuous
casting
departments).
These
similar
departments
can
be
both
operationally
and
physically
removed
from
each
other
and
should
not
be
considered
as
one
operation
or
system.

Therefore,
U.
S.
Pipe
recommends
that
the
definitions
at
§
63.7762
be
amended,
as
follows,
to
divide
the
current
definition
of
"
metal
casting
department"
into
separate
definitions
for
"
melting
department"
and
"
casting
processing
department."
In
addition,
the
definitions
of
the
`
affected
sources'
should
reflect
the
potential
for
the
existence
of
multiple
departments
of
the
same
type.

Proposed
definitions:
"
Mold
and
core
making
department
means
the
area
of
a
foundry
and
associated
equipment
in
which
all
operations
needed
to
produce
molds,
mold
sections,
and
cores
are
done,
including
those
operations
performed
in
mold
or
core
making
and
mold
or
core
coating
lines.
A
foundry
may
have
multiple
mold
and
core
making
departments."
8
"
Melting
department
means
a
metal
melting
furnace
at
a
foundry
and
the
associated
equipment
in
which
all
operations
needed
to
melt
metal
are
done,
including
preparation
of
furnace
feed,
melting
metal,
and
transferring
molten
metal
to
pouring
stations.
A
foundry
may
have
multiple
melting
departments."

"
Casting
processing
department
means
the
area
of
a
foundry
and
associated
equipment
in
which
all
operations
needed
to
pour
and
produce
mechanically
finished
castings
are
done,
including
pouring
metal
into
molds,
cooling
molds,
and
separating
castings
from
molds.
A
foundry
may
have
multiple
casting
processing
departments."

Additional
sections
of
the
regulation
will
have
to
be
adjusted
to
reflect
this
change,
including
the
citing
of
"
affected
sources"
at
§
63.7682(
b)
and
(
c).

§
63.7683(
a)
 
Compliance
Dates
The
proposed
rule,
at
§
63.7683(
a),
would
require
that
existing
sources
comply
with
these
regulations
no
later
than
three
(
3)
years
after
they
become
final
rules.
U.
S.
Pipe
endorses
this
compliance
schedule
since
several
of
the
proposed
emission
limitations,
work
practice
standards,
and
operation
and
maintenance
requirements
will
require
significant
investment
and
engineering
as
well
as
considerable
onsite
construction
activities.

§
63.7683(
b)
and
(
c)
 
Compliance
Dates
The
proposed
rules,
at
§
63.7683
(
b)
and
(
c),
do
not
properly
account
for
new
sources
that
commence
construction
or
reconstruction
in
the
period
between
the
proposal
of
this
standard
and
the
finalizing
of
these
rules.
The
rules
should
indicate
that
sources
constructed
or
reconstructed
during
this
period
might
be
exempted
from
achieving
compliance
for
three
(
3)
years.

As
currently
proposed
the
regulations
would
require
that
any
"
new"
source
achieve
compliance
with
the
standard
by
the
date
the
rule
becomes
final
[
see
§
63.7683(
b)]
or
at
initial
startup
[
see
§
63.7683(
c)].
However,
the
Clean
Air
Act,
at
§
112(
i)(
2),
states
that:

 
a
new
source
which
commences
construction
or
reconstruction
after
a
standard,
limitation
or
regulation
applicable
to
such
source
is
proposed
and
before
such
standard,
limitation
or
regulation
is
promulgated
shall
not
be
required
to
comply
with
such
promulgated
standard
until
the
date
three
(
3)
years
after
the
date
of
promulgation
if
­

(
A)
the
promulgated
standard,
limitation
or
regulation
is
more
stringent
than
the
standard,
limitation
or
regulation
proposed;
and
(
B)
the
source
complies
with
the
standard,
limitation,
or
regulation
as
proposed
during
the
3­
year
period
immediately
after
promulgation.

Therefore,
U.
S.
Pipe
requests
that
§
63.7683
be
amended
to
reflect
this
provision
of
the
Clean
Air
Act.

§
63.7690(
a)(
1)
 
PM
Emissions
Limitations
 
Existing
Metal
Melting
Furnaces
and
Scrap
Preheaters
The
proposed
rules,
at
§
63.7690(
a)(
1),
would
require
that
particulate
matter
(
PM)
emissions
from
existing
metal
melting
furnaces
and
scrap
preheaters
not
exceed
0.005
gr/
dscf.
U.
S.
Pipe
believes
that
this
proposed
limitation
is
untenable
in
that
it
results
in
minimal
reductions
of
HAP
emissions
at
a
prohibitive
cost.
U.
S.
Pipe
believes
that
the
agency
has
the
authority,
valid
reasoning,
and
the
duty
of
responsible
regulation
to
reconsider
this
proposed
requirement.

Specifically,
many
melting
furnaces
are
currently
controlled
by
the
use
of
wet
scrubbers
that
are
approximately
97.8%
effective
at
removing
PM
from
the
exhaust
stream.
However,
some
furnaces
utilize
fabric
filters
that
are
approximately
99.8%
effective.
Therefore,
the
agency
selected
the
use
of
fabric
filters
as
the
MACT
Floor
technology.
Although
this
is
consistent
with
a
strict
reading
of
§
112(
d)
of
the
CAA,
it
results
in
an
excessive
burden
to
attain
a
trivial
gain.

§
63.7690(
a)(
2)
 
PM
Emissions
Limitations
 
New
Furnaces
9
The
proposed
rules,
at
§
63.7690(
a)(
2),
would
require
that
PM
emissions
from
metal
melting
furnaces
and
scrap
preheaters
at
a
new
metal
casting
(
i.
e.,
metal
melting)
department
not
exceed
0.001
gr/
dscf.

U.
S.
Pipe
believes
that
this
proposed
limitation
does
not
properly
account
"
for
the
variability
inherent
in
process
operations
and
control
device
performance."
When
determining
the
proposed
PM
emissions
limitation
for
new
furnaces
and
cupolas,
there
is
no
indication
that
any
statistical
method
or
other
means
was
utilized
to
account
"
for
the
variability
inherent
in
process
operations
and
control
device
performance"
for
these
new
sources.
Rather,
it
appears
that
the
agency
adopted
this
MACT
floor
based
on
testing
conducted
at
a
limited
number
of
sites
that
utilize
the
MACT
technology.
U.
S.
Pipe
believes
that
any
emission
standard
derived
from
a
database
limited
in
this
manner
cannot
meet
the
threshold
of
considering
and
accounting
for
the
variability
inherent
in
furnace
operations
throughout
the
industry,
nor
the
variability
expected
in
the
control
device's
performance
over
these
operating
conditions.

A
properly
developed
emission
standard
must
consider
the
"
worst
foreseeable
circumstances
faced
by
any
unit
in
a
given
source
category."
Therefore,
U.
S.
Pipe
believes
that
the
agency
cannot
determine
the
new
source
MACT
floor
by
solely
relying
upon
the
emissions
tests
of
those
new
sources
that
utilize
the
MACT
technology,
unless
those
tests
were
performed
under
conditions
that
reflect
the
worst
foreseeable
in
the
industry
(
i.
e.,
worst
scrap,
worst
coke,
etc.).

§
63.7690(
a)(
3)
 
PM
Emissions
Limitations
 
Existing
Pouring
Stations
The
proposed
rules,
at
§
63.7690(
a)(
3),
would
require
that
PM
emissions
from
existing
pouring
stations
not
exceed
0.010
gr/
dscf.
The
very
low
HAP
emissions
from
permanent
mold
pipe
making,
as
discussed
elsewhere
in
this
document,
make
the
requirement
to
install
capture
and
control
equipment
technologically
and
economically
infeasible.

§
63.7690(
a)(
5)
 
CO
Emissions
Limitations
 
Cupolas
The
proposed
rules,
at
§
63.7690(
a)(
5),
would
require
that
carbon
monoxide
(
CO)
emissions
from
new
and
existing
cupolas
not
exceed
200
parts
per
million
by
volume
(
ppmv).
In
the
preamble,
EPA
states
"
Cupola
afterburners
are
typically
operated
at
an
ignition
temperature
of
1300
0
F
or
higher
to
combust
the
CO
in
the
cupola
exhaust
stream 
Given
that
thermal
destruction
of
most
organic
compounds
occurs
at
1200
0
F
or
lower,
we
believe
that
organic
HAP
are
effectively
controlled
by
an
afterburner "
EPA,
however,
goes
on
to
use
source
compliance
test
data
to
propose
a
CO
emission
limit
of
200
ppmv
for
existing
and
new
cupola's,
but
proposes
to
require
compliance
to
be
demonstrated
on
a
3­
hour
average
using
CEM's.

First,
U.
S.
Pipe
contends
that,
based
upon
our
preamble
discussion,
this
limitation
was
derived
from
an
illegitimate
state
standard.
Therefore,
U.
S.
Pipe
recommends
that
this
limitation
be
replaced
with
a
requirement
that
an
afterburner
meeting
specified
combustion
parameters
be
maintained.
Second,
as
discussed
in
the
preamble,
the
results
of
testing
for
organic
HAPs
indicate
only
"
very
low
and
highly
variable"
concentrations
of
organic
HAPs
from
these
operations.
Third,
there
is
one
U.
S.
Pipe
foundry
that
presently
is
equipped
with
a
CEM
for
CO,
and
historical
data
has
firmly
established
that
a
3­
hour
average
of
200
ppmv
cannot
be
achieved.
In
fact,
our
permit
limit
is
550
ppmv
averaged
over
a
10
hour
period
(
normally
an
operating
day).

Although,
in
normal
operation,
iron
is
tapped
continually,
the
iron
melting
process
is
far
from
steady
state.
Every
charge
introduced
creates
turbulence
and
can
momentarily
extinguish
any
combustion.
It
is
not
unusual
for
a
CO
CEM
to
show
spikes
of
5000
ppmv
and
spikes
up
to
10,000
ppmv
do
occur.
One
of
our
trade
association
member
foundries
has
a
CO
CEM
and
has
recently
spent
approximately
two
(
2)
million
dollars
to
upgrade
its
afterburner
performance.
Even
after
this
upgrade,
it
cannot
achieve
200
ppmv.

Based
on
this
information
and
the
experience
of
U.
S.
Pipe,
it
is
apparent
that
stack
test
data
cannot
and
should
not
be
used
to
establish
a
limit
using
CEMs.
Further,
it
should
be
evident
that
an
emission
limit
of
200
ppmv
cannot
be
achieved
in
practice,
nor
based
on
EPA's
own
statements,
is
such
a
limit
necessary
to
reduce
organic
HAP
emissions
from
a
cupola.
U.
S.
Pipe
recommends
that
the
HAP
control
for
cupola/
afterburner
emissions
be
based
on
the
afterburner
being
designed
for
a
minimum
of
0.3
seconds
retention
at
1300
0
F
or
higher.
10
§
63.7690(
a)(
8),
(
b)(
5)
 
TEA
Emissions
Limitations
 
Cold
Box
Mold
and
Core
Making
Lines
The
proposed
rules,
at
§
63.7690(
a)(
8),
would
require
that
triethylamine
emissions
from
cold
box
mold
and
core
making
lines
be
reduced
to
a
level
that
does
not
exceed
1
ppmv
using
wet
acid
scrubbers
maintaining
scrubbing
liquid
3­
hour
average
flow
rate
above
a
minimum
flow
rate
established
during
performance
testing
and
by
maintaining
a
3­
hour
average
pH
below
a
maximum
pH
established
during
performance
testing.
Nowhere
in
the
preamble,
does
it
state
the
flow
rate
or
pH
of
the
wet
acid
scrubbers
that
were
evaluated
for
MACT.
Normally
these
scrubbers
are
operated
to
provide
a
solution
that
is
recyclable
and
has
a
pH
of
2.5
to
4.5.
Flow
rates
are
specified
by
the
manufacturer
of
the
scrubber,
along
with
pollutant
removal
efficiencies.
Discussions
with
industry
professionals,
leads
U.
S.
Pipe
to
believe
that
the
wet
acid
scrubbers
that
were
evaluated
for
MACT
were
running
with
a
fresh
acid
solution
at
a
pH
lower
than
2.0.
Triethylamine
recyclers
will
not
accept
material
with
a
pH
less
than
2.0,
since
it
is
considered
"
hazardous
waste"
by
definition.
If
the
foundry
industry
must
operate
wet
acid
scrubbers
at
a
pH
of
2.0
or
less
to
achieve
compliance,
the
recycle
industry
will
be
eliminated,
more
natural
resources
will
be
used
and
the
acid
solution
will
have
to
be
disposed
of
as
a
hazardous
waste
creating
another
burden
on
the
environment.
Recycling
costs
range
from
$
1.85
to
$
2.10
per
gallon,
depending
on
transportation
costs.
Disposal
costs
for
waste
scrubber
solution
run
from
$
1.00
to
$
3.00
per
gallon.
Contrary
to
disposal
of
waste
scrubber
solution,
recycling
is
considered
pollution
prevention.

The
proposed
standard
states
that
EPA
Method
18
will
be
used
for
compliance
testing
of
the
wet
acid
scrubbers.
The
detection
limit
of
EPA
Method
18,
1
ppmv,
has
been
selected
as
the
compliance
limit
for
triethylamine
emitted
from
wet
acid
scrubbers.
In
the
preamble
the
agency
states,
"
We
have
reliable
performance
test
data
for
seven
acid
wet
scrubbers
at
six
foundries".
None
of
the
performance
test
data
listed
was
done
using
EPA
Method
18,
other
test
methods
were
listed
but
not
the
required
method.
The
data
is
assumed
to
be
reliable,
but
in
fact
is
not
comparable
to
EPA
Method
18.
This
comparison
is
like
comparing
"
apples
to
oranges".
Since
reliable
performance
data
using
EPA
Method
18
does
not
exist,
meeting
a
1
ppmv
compliance
limit
using
EPA
Method
18
has
not
been
proven
to
be
technically
feasible.
Also,
setting
a
compliance
limit
at
the
detection
limit
of
EPA
Method
18
does
not
take
into
account
the
variability
of
the
test
at
these
low
levels.
Additionally,
the
cost
of
testing
triethylamine
emissions
using
Method
18
ranges
from
$
7,000
to
$
8,000
for
a
single
test.
These
costs
would
be
reduced
if
the
foundry
industry
could
use
alternate
analytical
test
methods
as
was
done
by
the
EPA
to
develop
the
wet
acid
scrubber
MACT.

I
nlet
loading
is
another
factor
that
is
brushed
over
in
the
preamble.
No
inlet
loadings
are
listed
for
the
outlet
emissions,
which
were
below
detection
limits.
Outlet
emissions
will
be
directly
affected
by
inlet
loading.
Manufacturers
of
wet
acid
scrubbers
specify
percent
removal
efficiency
rather
than
a
specific
pollutant
concentration
due
to
this
relationship.

I
n
the
preamble,
the
agency
states
"
establishment
of
a
standard
of
1
ppmv
triethylamine
emission
will
result
in
triethylamine
emission
reduction
of
146
tons
per
year
from
two
foundries
that
do
not
presently
control
emissions".
Why
should
the
whole
foundry
industry
carry
the
burden
of
meeting
the
1ppmv
triethylamine
emission
limit
for
two
foundries,
which
have
no
triethylamine
emission
control?

U.
S.
Pipe
proposes
that
wet
acid
scrubber
flow
rates
be
maintained
within
the
manufacturers
specifications,
wet
acid
scrubber
solution
pH
be
maintained
within
the
limits
set
for
the
purpose
of
recycling,
and
the
performance
of
all
triethylamine
removal
systems
be
measured
as
percent
removal
specified
by
the
manufacturer
rather
than
a
specific
triethylamine
concentration.

§
63.7690(
b)(
1)
 
Capture
and
Ventilation
Requirements
The
proposed
rules,
at
§
63.7690(
b)(
1),
would
require
that
a
capture
and
ventilation
system
that
maintains
a
face
velocity
of
200
feet
per
minute
be
maintained
at
each
emissions
source
subject
to
an
emissions
limit.
U.
S.
Pipe
believes
that
it
is
arbitrary
and
capricious
to
impose
a
single
uniform
standard
on
emission
sources
that
are
significantly
diverse.
In
addition,
we
believe
that
such
a
requirement
has
no
underlying
MACT
floor
determination,
may
be
an
entirely
unnecessary
burden
for
some
sources,
fails
to
consider
the
ancillary
energy
and
environmental
impacts,
and
fails
to
account
for
variable
operating
conditions.
11
Further,
the
preamble
fails
to
justify
the
establishment
of
a
capture
and
ventilation
system
"
that
maintains
a
face
velocity
of
at
least
200
feet
per
minute"
through
a
MACT
floor
determination.
The
preamble
fails
to
cite
any
study
or
information
collection
request
that
evaluates
and
justifies
the
establishment
of
such
a
requirement.

I
n
addition,
proposing
such
a
requirement
for
"
each
emission
source
subject
to
an
emissions
limit"
represents
an
unnecessary
burden
for
some
operations.
Apart
from
this
proposed
requirement,
foundries
utilizing
clean,
low­
emitting
processes,
may
have
no
valid
reason
to
vent
emissions
from
these
operations.
U.
S.
Pipe
believes
that
when
processes
are
operated
in
a
manner
that
limits
worker
exposure,
at
or
below
the
standards
established
by
the
Occupational
Safety
and
Health
Administration
(
OSHA),
then
there
exists
no
basis
for
requiring
stricter
capture
and
ventilation
standards.

U.
S.
Pipe
also
believes
that
the
agency
has
also
failed
to
consider
the
additional
environmental
and
energy
cost
that
would
be
incurred
as
a
result
of
this
requirement.
Specifically,
when
in­
plant
air
is
exhausted
from
the
facility
it
must
be
replaced.
This
make­
up
air
will
require
additional
ducting,
mechanical
air
supply
systems,
and
possible
conditioning
(
i.
e.,
heating
or
cooling)
prior
to
introduction
into
the
facility.
The
operation
of
these
systems
can
significantly
increase
a
facility's
energy
requirements
and
result
in
increased
emissions
attributable
to
heating
or
cooling
the
make­
up
air.

Therefore,
U.
S.
Pipe
recommends
that
this
provision
be
modified
as
follows:

"(
1)
For
each
emissions
source
subject
to
an
emissions
limit
under
paragraph
(
a),
you
must
capture
and
vent
emissions
through
a
capture
system
that
maintains
a
face
velocity
of
at
least
200
feet
per
minute
consistent
with
good
engineering
practices.
You
must
operate
each
capture
system
at
or
above
the
lowest
value
or
settings
established
as
operating
limits
in
consistent
with
your
operation
and
maintenance
plan."

§
63.7700(
a)
­
Work
Practice
Standards
 
Selection
and
Inspection
of
Scrap
The
proposed
rule,
at
§
63.7700(
a),
would
require
that
foundries
prepare
and
operate
according
to
a
written
plan
for
the
selection
and
inspection
of
the
scrap
materials
that
are
melted.
U.
S.
Pipe
believes
that
the
specific
requirements
of
this
provision
exceed
a
properly
established
MACT
floor
and
do
not
properly
acknowledge
the
realities
of
operating
a
foundry.
The
proposed
regulation
should
be
modified
to
recognize:
1)
that
no
foundry
currently
operates
such
an
extensive
inspection
and
selection
program;
2)
that
many
foundries
process
large
amounts
of
scrap
metal;
and
3)
that
foundries
are
often
the
"
last
resort"
for
the
recycling
of
oily
or
impure
scrap
metals
and
some
foundry
operations
(
e.
g.,
cupolas)
are
both
designed
for,
and
capable
of,
handling
these
materials.

The
proposed
rule
would
require
a
documented
inspection
of
"
all
incoming
scrap
shipments."
U.
S.
Pipe
believes
that
the
preamble
fails
to
establish
a
basis
for
setting
the
MACT
floor
at
this
level
and
does
not
present
a
beyond­
the­
floor
analysis
to
justify
this
requirement.
As
noted
in
the
preamble,
the
majority
of
foundries
do
use
some
type
of
scrap
selection
or
inspection
program
to
ensure
the
quality
of
the
scrap
metals
used,
however,
U.
S.
Pipe
is
aware
of
no
foundry
that
performs
documented
inspections
of
"
all"
scrap
shipments
for
reducing
emissions.
In
fact,
the
preamble
states
that
in
response
to
EPA's
inquiry
on
this
matter
"
many
responses
were
general
in
nature,
such
as
"
use
clean
scrap,"
"
follow
scrap
specification,"
or
"
inspect
scrap.""
Therefore,
U.
S.
Pipe
does
not
believe
that
a
MACT
floor
has
been
properly
established
to
require
the
inspection
of
"
all
incoming
scrap."
In
addition,
many
foundries
receive
numerous
shipments
of
large
quantities
of
scrap
materials.
A
requirement
to
inspect
and
document
each
of
these
shipments
would
be
excessively
burdensome
and
cost
prohibitive.

The
proposed
rule
includes
restrictions
on
the
scrap
materials
that
would
limit
the
introduction
of
painted
parts
and
galvanized
materials,
and
a
prohibition
on
the
use
of
oily
turnings.
U.
S.
Pipe
believes
that
these
restrictions
do
not
recognize
that
restrictions
on
the
use
of
painted
parts,
galvanized
materials,
and
oily
turnings
may
result
in
these
materials
no
longer
being
recycled.
Although
these
materials
may
12
introduce
some
contaminants
into
the
process,
U.
S.
Pipe
believes
that
their
use
in
the
foundry
is
preferential
to
these
materials
being
abandoned
or
utilized
in
an
unregulated
or
lesser­
regulated
manner.
U.
S.
Pipe
believes
that
the
use
of
these
materials
in
a
regulated,
controlled
process
is
their
best
environmental
fate,
and
that
the
emissions
limitations
imposed
upon
the
foundry
operations
will
adequately
protect
the
environment.

Finally,
the
proposed
regulation
would
also
require
that
the
scrap
inspections
be
performed
at
the
"
best"
vantage
point.
U.
S.
Pipe
believes
that
this
requirement
is
nebulous
and,
hence,
compliance
may
be
difficult
to
ensure.
U.
S.
Pipe
recommends
this
requirement
be
deleted.

§
63.7700(
b)
­
Work
Practice
Standards
 
Mold
Vent
Ignition
The
proposed
rule,
at
§
63.7700(
b),
requires
that
foundries
"
manually
ignite
the
gases
from
each
mold
vent
that
do
not
ignite
automatically."

I
t
must
be
clear
that
this
requirement
has
no
meaning
where
permanent
mold
casting
is
the
process
involved.
There
is
no
mold
vent
associated
with
the
casting
of
iron
pipe
since
there
is
no
mold
sand.
The
permanent
steel
mold
is
open
on
both
ends
allowing
rapid
mixing
of
oxygen
with
any
organic
HAP
that
may
be
released
during
pouring
and
cooling.

U.
S.
Pipe
also
believes
that
this
provision
inaccurately
assumes
that
it
is
reasonable
to
ensure
the
ignition
of
the
gases
at
"
each"
sand
mold
vent,
and
that
the
"
each"
sand
mold
vent
will
emit
an
ignitable
gas.
Sand
mold
vents,
which
allow
for
the
escape
of
gases
during
the
pouring
and
cooling
of
a
casting,
may
emit
an
ignitable
gas.
However,
the
regulation
should
acknowledge
that
casting
molds
may
have
literally
hundreds
of
mold
vents,
and
that
some
of
these
mold
vents
will
be
located
in
inaccessible
areas
(
e.
g.,
over
the
center
of
a
large
casting).
In
addition,
some
mold
vents
may
emit
such
minimal
quantities
of
gaseous
byproducts
that
they
may
not
ignite.

Moreover,
the
proposed
regulation
would
require
that
a
"
pouring,
cooling
and
shakeout
line"
in
an
existing
foundry
comply
with
this
requirement.
U.
S.
Pipe
believes
that
this
provision
may
be
misconstrued
to
require
such
ignition
in
the
cooling
and
shakeout
areas.
In
fact,
mold
vent
ignition
operations
are
conducted
at
or
near
the
"
pouring
station."

Therefore,
U.
S.
Pipe
requests
§
63.7700(
b)
be
modified,
where
appropriate
to
the
casting
process,
as
follows,
to
require
that
foundries
provide
an
ignition
source
at
or
near
the
pouring
area
or
pouring
station
such
that
mold
gases
may
ignite.

(
b)
For
each
pouring,
cooling,
and
shakeout
line
At
or
near
each
pouring
station
in
an
existing
metal
casting
department
and
each
pouring
area
in
a
new
or
existing
metal
casting
department,
you
must
manually
ignite
the
gases
from
each
mold
vent
that
do
not
ignite
automatically
provide
an
ignition
source
such
that
the
mold
vent
gases
may
ignite.

§
63.7700(
f)
 
Work
Practice
Standards
 
Mold
and
Core
Making
Lines
The
proposed
rule,
at
§
63.7700(
f),
requires
that
mold
and
core
making
lines
other
than
furan
warm
box,
phenolic
urethane
cold
box,
and
phenolic
urethane
no
bake:
1)
conduct
a
study
of
reduced­
HAP
binder
formulations;
2)
adopt
a
reduced­
HAP
binder
system
or
demonstrate
that
"
all"
available
alternatives
are
technically
or
economically
infeasible;
and
3)
conduct
a
study
every
five
(
5)
years
until
a
reduced­
HAP
system
is
adopted.
U.
S.
Pipe
believes
that
this
proposed
standard
has
no
legal
basis
within
the
Clean
Air
Act;
presents
an
unachievable
compliance
requirement;
and
presents
an
excessive
burden
without
providing
a
beyond­
the­
floor
economic
demonstration.

§
112(
h)
of
the
Clean
Air
Act
outlines
the
authority
of
the
agency
to
establish
"
work
practice
standards"
within
a
MACT
NESHAP.
However,
in
§
112(
h)(
2)
the
use
of
"
work
practice
standards"
is
limited
to
those
situations
when:
(
A)
a
hazardous
air
pollutant
or
pollutants
cannot
be
emitted
through
a
conveyance;
or
(
B)
the
application
of
a
measurement
methodology
to
a
particular
class
of
sources
is
not
practicable.
U.
S.
Pipe
believes
that
neither
of
these
thresholds
has
been
met
to
allow
for
this
proposed
"
work
practice
standard."
13
Further,
the
basic
structure
of
§
112
of
the
CAA
establishes
that
the
MACT
standards
are
to
represent
a
one­
time
identification
and
employment
of
emission
control
systems
that
reflect
the
pollution
abatement
technologies
currently
available.
The
CAA
anticipates
and
provides
for
any
necessary
adjustments
to
a
MACT
standard
via
the
mechanisms
of
§
112(
f).
Nowhere
does
§
112
provide
authority
for
the
agency
to
mandate
a
continuing
program
to
reevaluate
products
or
technologies.
Therefore,
U.
S.
Pipe
believes
that
the
requirement
to
"
conduct
a
study"
every
five
(
5)
years
is
improper
and
must
be
removed.

I
n
addition,
U.
S.
Pipe
believes
that
the
requirement
to
demonstrate
that
"
all
available
alternatives"
to
a
current
binder
system
are
infeasible
presents
an
unachievable
standard.
Since
there
are
an
infinite
number
of
potential
binder
materials
and
combinations
of
materials,
no
regulated
source
can
possibly
demonstrate
that
"
all
alternatives"
have
been
considered.

The
regulation
also
fails
to
define
the
term
"
reduced­
HAP
binder."
Thus,
the
provision
as
written
would
require
those
facilities
that
already
utilize
binders
systems
that
contain
extremely
low­
HAP
concentrations
to
continue
to
evaluate
alternatives.
Likewise,
casting
facilities
may
be
subjected
to
varying
interpretations
of
the
term
"
reduced­
HAP
binder"
and
subject
to
potential
compliance
actions.

Finally,
the
Casting
Emissions
Reduction
Program
(
CERP)
has
conducted
numerous
evaluations
of
binder
systems.
CERP,
of
which
EPA
is
a
member,
reports
that
HAP
emissions
testing
"
costs
approximately
$
200,000
to
$
250,000
per
binder
system
()".
These
costs
will
further
increase
with
the
need
to
conduct
trial
production
operations
to
ensure
that
product
quality
is
maintained
when
changing
to
a
replacement
binder.
Thus,
the
review
of
even
a
limited
number
of
binder
systems
will
result
in
an
excessive
cost
burden
to
a
foundry
operation.
U.
S.
Pipe
believes
that
it
is
incumbent
upon
the
agency
to
provide
a
beyond­
the­
floor
analysis
to
demonstrate
that
this
proposed
requirement
considers
the
cost
of
achieving
any
potential
reductions.

§
63.7720(
b)
 
General
Compliance
Requirements
These
requirements
are
adequately
addressed
through
the
existing
Title
V
Major
Source
Operating
Permit
and
the
Synthetic
Minor
Operating
Permit
conditions.
Any
additional
requirements
would
be
overly
burdensome
and
unnecessary.

§
63.7730
 
Initial
Compliance
Requirements
§
63.7732(
b)
­
Determining
Compliance
with
the
Emissions
Limitation
for
Particulate
Matter
The
proposed
rule,
at
§
63.7731(
b),
requires
the
use
of
Method
5,
5B,
5D,
5F,
or
5I
to
determine
compliance
with
the
emission
limitations
for
particulate
matter
(
PM).
U.
S.
Pipe
requests
that
this
provision
be
clarified
to
indicate
that
compliance
with
the
PM
standard
is
to
be
determined
based
on
the
"
front­
half"
catch
and
that
any
"
back­
half"
condensable
matter
is
excluded
from
the
determination.

The
preamble
indicates
(
page
78286)
that
the
PM
emission
standards
were
developed
based
on
data
that
included
only
the
"
front­
half"
catch.
Therefore,
it
is
appropriate
that
compliance
with
these
limitations
also
be
determined
based
on
the
same
methodology.
U.
S.
Pipe
is
concerned
that,
in
the
future,
persons
may
misinterpret
the
compliance
determination
requirement
to
include
the
"
back­
half"
PM
catch,
thus
increasing
the
stringency
of
the
emission
limitations.
Since
the
court
4
has
held
that
changing
a
sampling
method
can
change
the
stringency
of
a
limitation,
U.
S.
Pipe
believes
it
is
appropriate
that
the
regulation
be
amended
to
ensure
proper
understanding.

§
63.7732(
c)
 
Determining
Compliance
with
the
Emissions
Limit
for
Carbon
Monoxide
Based
on
our
previous
comments
on
the
CO
limit
for
cupolas,
it
is
proposed
that
this
section
be
revised
to
require
source
testing
(
every
five
years)
and
not
CEMs.

4
Appalachian
Power
Company
v.
EPA,
341
US
App
DC
46;
208
f.
3d
1015,
decided
April
14,
2000.
"
We
have
recognized
before
that
changing
the
method
of
measuring
compliance
with
an
emission
limitation
can
affect
the
stringency
of
the
limitation
itself.
Portland
Cement
Ass'n
v.
Ruckelshaus,
158
US
App
DC
308,
486
F.
2d
375,
396­
97
(
DC
1973),
discussed
in
Clean
Air
Implementation
Project
v.
EPA,
105
F.
3d
at
1203."
14
§
63.7734(
5)
 
Demonstrating
Initial
Compliance
with
Emission
Limitations
Based
on
our
previous
comments
on
the
CO
limit
for
cupolas,
it
is
proposed
that
this
section
be
revised
to
address
source
testing
(
every
five
years)
and
not
CEMs.

§
63.7740
 
Monitoring
Requirements
U.
S.
PIPE
requests
that
EPA
take
a
pragmatic
approach
in
establishing
monitoring
requirements
for
this
proposed
rule
taking
into
consideration
cost,
the
relative
benefit
of
the
monitoring
proposed
and
consistency
with
monitoring
requirements
already
established
by
existing
rules
and
operating
permit
conditions.

As
currently
proposed,
the
monitoring
provisions
require
the
installation,
operation
and
maintenance
of
a
substantial
number
of
continuous
monitoring
systems
to
either
directly
measure
emissions
from
a
particular
process
(
CEMS)
or
to
monitor
relevant
operating
parameters
to
ensure
emission
control
systems
are
operating
at
optimal
performance
levels
(
CPMS).
While
U.
S.
Pipe
recognizes
that
these
systems
are
technically
feasible,
we
believe
the
use
of
CEMS
and
CPMS
is
cost
prohibitive.
Furthermore,
U.
S.
Pipe
believes
these
monitoring
systems
will
provide
minimal
compliance
assurance
beyond
other
already
available
and
less
costly
monitoring
methods.

CEMS
for
CO
and
VOC
Emissions
§
63.7740(
f)
requires
the
installation
and
operation
of
carbon
monoxide
CEMS
for
all
cupolas
at
affected
sources.
Similarly,
§
63.7740(
g)
requires
the
installation
and
operation
of
VOC
CEMS
for
scrap
preheaters
and
new
pouring,
cooling
and
shakeout
lines.
The
proposed
rule
further
requires
that
periodic
performance
testing
be
performed
to
demonstrate
continuous
compliance
with
proposed
emission
limitations.

U.
S.
Pipe
recognizes
EPA's
desire
that
MACT
standards
contain
continuous
compliance
determination
methods.
However,
in
other
MACT
standards
EPA
has
not
required
the
use
of
CEMS
for
similar
emission
units
and
emission
control
systems
due
to
their
substantial
cost
and
minimal
benefit
over
other
effective
monitoring
methods
(
see
discussion
on
Monitoring
in
the
preamble
to
40
CFR
63
Subpart
RRR,
NESHAP
for
Secondary
Aluminum
Production,
FR
Vol.
64,
No.
28,
p.
6982).
U.
S.
Pipe
requests
that
EPA,
consistent
with
its
decision
to
not
require
CEMS
for
affected
sources
under
the
Secondary
Aluminum
Production
MACT
Standard,
eliminate
the
CEMS
requirement
in
this
proposed
rule.

EPA
has
stated
that
MACT
standards
contain
two
levels
of
compliance
assurance.
First,
affected
sources
must
perform
"
stack"
tests
to
demonstrate
that
the
emission
unit
subject
to
an
emission
limitation
is
capable
of
achieving
compliance.
Second,
affected
sources
must
perform
some
level
of
ongoing
monitoring
to
assure
that
compliance
continues
to
be
achieved.
This
on­
going
monitoring
may
include
direct
measurement
of
HAPs
or
a
HAPs
surrogate
or
it
may
involve
monitoring
of
the
process
or
control
device
to
ensure
it
is
operating
properly.
In
this
instance,
an
assurance
of
compliance
exists
because
the
source
has
previously
shown
by
performance
testing
that,
when
operated
properly,
the
emission
unit
will
achieve
required
emission
levels
and
the
ongoing
process
or
control
device
monitoring
ensures
that
the
equipment
is
operating
properly.
As
EPA
has
cited
within
the
preamble
of
the
proposed
rule,
that
the
destruction
of
organic
HAPs
emitted
by
the
operation
of
cupolas,
scrap
preheaters
and
new
PCS
lines,
is
effectively
achieved
by
thermal
oxidation
at
or
below
a
temperature
of
1200
F.
EPA
also
correctly
states
that
cupola
afterburners
are
typically
operated
at
or
above
a
temperature
of
1300
F
in
order
to
combust
CO
in
the
cupola
exhaust
stream.
EPA
goes
on
to
state
" 
we
believe
that
organic
HAP
are
effectively
controlled
by
an
afterburner
that
effectively
oxidizes
CO,"
(
see
FR,
Vol.
67,
No.
246,
p.
78284).
Similarly,
the
use
of
thermal
oxidizers
to
control
VOC
emission
levels
from
scrap
preheaters
and
new
PCS
lines,
if
operated
at
temperature
levels
in
excess
of
1200
F,
would
also
be
expected
to
effectively
control
organic
HAP.
U.
S.
Pipe
believes
that
continuous
monitoring
of
afterburner
temperature,
in
combination
with
periodic
performance
testing
and
the
development
and
implementation
of
O,
M
&
M
plans
for
these
control
devices
(
as
required
elsewhere
in
the
proposed
rule)
represents
an
effective
alternative
to
CEMS.
This
monitoring
strategy
is
also
consistent
with
existing
regulatory
and
permit
requirements
for
many
affected
sources
and
will
minimize
the
potential
for
redundant
or
conflicting
monitoring
requirements.
15
Furthermore,
as
discussed
earlier
in
these
comments,
EPA
has
failed
to
provide
sufficient
data
that
maintaining
a
CO
concentration
of
200
ppmv
is
an
effective
strategy
for
assuring
the
destruction
of
organic
HAP,
nor
has
it
demonstrated
that
the
use
of
CEMS
is
technically
feasible
for
the
extreme
conditions
associated
with
cupola
and
PCS
operation.

§
63.7762
 
What
definitions
apply
to
this
subpart?

We
propose
to
add
the
following
definition
for
"
Afterburner".
"
Afterburner
means
a
vessel,
either
stand
alone
or
in
the
outlet
stack
of
an
iron
melting
cupola,
designed
to
thermally
oxidize
carbon
monoxide,
generated
during
the
melting
process,
to
carbon
dioxide
and
water
by
mixing
air
and
cupola
off
gas
containing
10­
20%
carbon
monoxide
during
normal
melting
operation.
Burners,
fired
by
a
secondary
fuel
(
e.
g.
fuel
oil
or
natural
gas),
are
used
during
initial
startup
and,
during
normal
cupola
operation,
to
maintain
operating
temperatures,
usually
about
1300
º
F.

The
proposed
rule
defines
`
capture
system'
to
mean
a
system
that
captures
gases
and
fumes
and
"
then
convey(
s)
the
captured
gases
to
a
control
device."
U.
S.
Pipe
believes
that
this
definition
should
be
revised
to
remove
the
supposition
that
all
captured
emissions
must
be
conveyed
to
a
control
device.
Some
emission
sources
may
be
vented
to
outside
(
or
inside)
the
facility
without
the
need
for
a
control
device.
Specifically,
some
foundry
emission
sources
may
be
capable
of
attaining
emission
standards
due
to
process
changes
or
material
substitutions
(
e.
g.,
the
use
of
low­
HAP
binders).
U.
S.
Pipe
believes
that
this
revision
properly
defines
the
purpose
of
a
`
capture
system'
without
assuming
the
ultimate
fate
of
the
gases
and
fumes.
Therefore,
we
recommend
that
this
definition
be
revised
as
follows:

"
Capture
system
means
the
collection
of
components
used
to
capture
gases
and
fumes
released
from
one
or
more
emissions
points
and
then
convey
the
captured
gas
stream
to
a
control
device
away
from
the
emission
source.
A
capture
system
may
include,
but
is
not
limited
to,
the
following
components
as
applicable
to
a
given
capture
system
design:
duct
intake
devices,
hoods,
enclosures,
ductwork,
dampers,
manifolds,
plenums,
and
fans."

U.
S.
Pipe
is
aware
that
trade
groups,
with
which
we
have
affiliation,
specifically
the
American
Foundry
Society
(
AFS)
and
The
Ductile
Iron
Pipe
Research
Association
(
DIPRA),
are
also
submitting
comments
on
these
proposed
regulations.
Rather
than
repeat
these
comments
herein,
U.
S.
Pipe
is
incorporating
by
reference
all
AFS
and
DIPRA
comments.

I
n
summary,
U.
S.
Pipe
requests
that
the
agency
consider
these
comments
in
developing
the
final
regulations
for
the
Iron
and
Steel
Foundry
NESHAP.
U.
S.
Pipe
offers
our
continued
support
in
this
rule
development
and
will
provide
any
clarification
that
is
needed
in
assisting
agency
personnel
in
understanding
the
metal
casting
industry.
In
addition,
U.
S.
Pipe
again
invites
the
relevant
members
of
EPA
staff
involved
with
the
development
of
this
standard
to
tour
several
foundry
operations
in
order
to
better
understand
the
concerns
of
U.
S.
Pipe
and
its
members.

Sincerely,

John
F.
Pleasant
Director
of
Environmental
Services
Attachments:
16
1.
Sturgis
Foundry
Monovent
Baghouse
Test
2.
Letter
to
A.
Vervaert,
January
29,
2001,
esp.
Attachment
G
"
A
Case
for
Subcategorization
of
Permanent
Mold
Pipe
Foundries"
