"
Hunt,
Tim"
<
Tim_
Hunt@
afandpa
To:
Mary
Kissell/
RTP/
USEPA/
US@
EPA,
Scott
.
org>
Jenkins/
RTP/
USEPA/
US@
EPA
cc:
Ken
Hustvedt/
RTP/
USEPA/
US@
EPA,
Dave
06/
03/
2004
12:
40
PM
Guinnup/
RTP/
USEPA/
US@
EPA
Subject:
PCWP
process
unit
and
Q&
As
Mary
Tom
and
Scott,

As
you
know,
we
have
been
working
on
a
way
to
focus
process
unit
testing
to
a
subset
of
units
and
rely
on
other
means
to
quantify
emissions
from
other
units.
We
have
put
together
the
attached
white
paper
to
present
our
two
options
as
well
as
other
features
of
the
rule
relating
to
process
units
such
as
not
testing
units
with
no
detectable
HAPs
and
how
to
focus
testing
of
acrolein
and
benzene
to
the
right
set
of
process
units.
I
am
also
including
an
analysis
(
spreadsheet)
by
NCASI
that
supports
the
groupings
of
process
units
in
Option
2
of
the
white
paper.
In
addition,
I
wanted
to
share
another
concept
for
focusing
efforts
(
both
for
the
regulator
and
regulated)
on
those
process
unit­
HAP
combinations
that
will
drive
any
risk
estimates
(
titled
"
process
unit
focus
paper").
Finally,
in
addition
to
the
questions
we
asked
during
our
April
1st
call,
we
have
put
together
several
other
questions
and
proposed
answers
on
elements
of
the
final
rule
that
remain
unclear.
Some
of
the
solutions
may
be
simple
technical
changes
to
the
rule
before
it
is
published
while
others
could
be
handled
through
guidance
or
EPA
Q&
A.

We
would
like
to
schedule
a
call
or
even
a
meeting
with
you
soon
to
determine
next
steps
given
that
the
whole
testing
issue
is
holding
mills
back
from
starting
to
explore
in
depth
whether
they
qualify
for
the
low
risk
subcategory.
My
Committee
is
meeting
on
June
16th
here
in
DC
which
might
be
a
good
time
to
chat
with
many
of
the
affected
companies
present.
In
addition,
Scott
and
Dave
have
already
agreed
to
call
in
to
the
meeting
to
discuss
a
template
for
risk
assessment
submittals.
Maybe
we
can
just
extend
the
discussion
to
the
HAP
testing/
estimation
and
process
unit
issues,
what
do
you
think?.

Please
let
me
know
if
you
have
any
questions.

Timothy
Hunt
Senior
Director,
Air
Quality
Programs
American
Forest
&
Paper
Association
1111
19th
St.,
NW
Washington,
DC
20036
phone:
202­
463­
2588
tim_
hunt@
afandpa.
org
(
See
attached
file:
Process
unit
white
paper
final
6­
3­
04.
doc)(
See
attached
file:
Process
Unit
NCASI
supporting
analysis
6­
04.
xls)(
See
attached
file:
QA
Wood
MACT
final
6­
04.
doc)(
See
attached
file:
Process
Unit
Focus
Paper
final
6­
3­
04.
doc)
1
Table
1
lists
the
following
HAPs:
acetaldehyde,
acrolein,
arsenic,
benzene,
beryllium,
cadmium,
chromium,
formaldehyde,
lead,
manganese,
methylene
diphenyl
diisocyanate,
nickel
and
phenol.

DC\
672333.1
AF&
PA
White
Paper
On
Process
Unit
Testing
and
Alternatives
To
Demonstrate
Eligibility
For
The
Low
Risk
PCWP
Subcategory
June
3,
2004
A.
Introduction:

In
its
final
rule
establishing
National
Emission
Standards
for
Hazardous
Air
Pollutants
from
Plywood
and
Composite
Wood
Products
(
PCWP),
EPA
created
a
subcategory
of
low­
risk
PCWP
sources.
The
final
rule
provides
that
sources
have
up
to
two
years
to
demonstrate
eligibility
for
the
low­
risk
subcategory.
Criteria
for
demonstrating
low­
risk
eligibility
are
given
in
Appendix
B
to
the
rule.

Section
5
of
Appendix
B
provides
that
a
source
wishing
to
demonstrate
low­
risk
eligibility
"
must
conduct
HAP
emissions
tests
.
.
.
for
every
process
unit
within
the
affected
source
that
emits
one
or
more
of
the
HAP
listed
in
Table
1
to
this
Appendix."
1
Section
5
goes
on
to
mandate,
in
essence,
"
traditional"
stack
testing
under
worst­
case
operating
conditions.

B.
Process
units
with
no
or
very
low
HAP
emissions:

Before
turning
to
the
question
of
what
is
a
"
process
unit,"
it
is
important
to
have
a
clear
understanding
of
which
process
units
do
not
emit
one
or
more
of
the
listed
HAPs
and
per
the
rule,
are
exempt
from
the
testing
requirement.
Many
sources
emit
listed
HAPs
at
levels
well
below
the
method
detection
limits.
AF&
PA
believes
that
sources
that
consistently
have
non­
detectable
emissions
of
the
listed
HAPs
are
not
significant
contributors
to
the
risk
and
need
not
be
tested.
EPA
has
already
made
a
determination
that
since
HAP
metals
can
reasonably
be
anticipated
only
from
direct­
fired
dryers
(
which
are
amenable
to
stack
testing)
and
that
methylene
diphenyl
diisocyanate
(
MDI)
can
reasonably
be
anticipated
only
from
processes
with
MDI
resins,
these
compounds
do
not
need
to
be
tested
for
every
source.
EPA
even
anticipated
the
burdens
of
testing
for
acrolein
and
benzene
from
all
process
units
and
provided
a
mechanism
in
the
rule
to
narrow
the
focus
of
testing
(
see
separate
section
below).

Based
on
data
from
EPA
and
the
NCASI,
the
following
process
units
do
not
emit
any
of
the
Table
I
HAPs
or
generally
emit
them
at
less
than
1
ppm
(
the
limit
of
detection
established
in
the
rule):

°
chippers
°
sanders
°
saws
°
engineered
wood
products
presses
[
See
analysis
from
NCASI]

In
addition,
HAPs
emission
data
does
not
exist
for
several
other
unit
operations
at
PCWP
facilities.
Many
of
these
unit
operations,
such
as
debarkers,
have
not
been
tested
because
there
is
no
reason
to
believe
that
they
would
have
HAPs
emissions.
Also,
many
are
very
difficult
or
impossible
2
It
is
important
to
remember
that
process
units
that
do
not
emit
any
of
the
listed
HAPs
also
do
not
need
to
be
tested.
2
to
test
by
any
means,
let
alone
using
the
methods
for
determining
gas
flow
specified
in
Appendix
B.
Therefore,
these
unit
operations
should
not
need
to
be
tested.

In
summary,
AF&
PA
believes
that
the
rule
allows
industry
to
eliminate
sources
from
the
emission
testing
requirements
where
the
units
do
not
emit
significant
quantities
of
the
listed
HAPs.
Furthermore,
AF&
PA
believes
that
sources
emitting
HAPs
at
non­
detectable
levels
should
not
be
included
in
determining
the
average
stack
height
for
risk
estimation
purposes.

C.
Testing
process
units:

The
final
rule
defines
process
unit
as
"
equipment
classified
according
to
its
function
such
as
blender,
dryer,
press,
former
or
board
cooler."
40
C.
F.
R.
§
63.2292.
The
definition
further
specifies
identified
pieces
of
PCWP
equipment
as
"
process
units."
These
identified
"
process
units"
represent
the
largest
emissions
sources
at
an
affected
source
and
tend
to
be
amenable
to
stack
testing
(
e.
g.,
various
types
of
presses
and
dryers).
AF&
PA
understands
from
discussions
with
EPA
staff,
however,
that
the
Agency
also
expects
sources
to
account
for
emissions
from
other
types
of
process
units
when
demonstrating
eligibility
for
the
low­
risk
subcategory.

A
significant
problem
arises
if
all
process
units
are
included.
Emission
testing
is
impractical
for
many
sources
and
unnecessary
for
others,
thereby
severely
discouraging
many
PCWP
sources
from
attempting
to
qualify
for
the
low­
risk
subcategory.
AF&
PA
believes
that
there
are
two
basic
options
for
addressing
the
process
unit
testing
issue.

Option
1:
Selected
Emissions
Testing
The
simplest
approach
 
and
one
that
requires
no
modification
to
the
final
rule
 
is
for
EPA
to
interpret
the
term
"
process
unit"
in
Appendix
B
to
refer
to
those
"
process
units"
that
are
specifically
identified
in
40
C.
F.
R.
§
63.22922.
As
noted
above,
the
most
significant
emission
sources
from
PCWP
units
 
and
the
ones
amenable
to
stack
testing
 
are
specifically
identified
as
"
process
units"
in
the
definitions
section
of
the
final
rule.
For
example,
the
definition
of
"
secondary
tube
dryer"
states,
"
A
secondary
tube
dryer
is
a
process
unit."
40
C.
F.
R.
§
63.2292.
The
named
process
units
are:
conveyor
strand
dryer;
dry
rotary
dryers;
fiberboard
mat
dryer;
green
rotary
dryer;
hardboard
oven;
hardwood
veneer
dryer;
press
predryer;
pressurized
refiner;
primary
tube
dryer;
reconstituted
wood
product
board
cooler;
reconstituted
wood
product
press;
rotary
strand
dryer;
secondary
tube
dryer;
softwood
veneer
dryer;
tube
dryer;
and
veneer
redryer.
EPA
did
not
identify
Group
1
miscellaneous
coating
operations
and
lumber
kilns
as
process
units.

By
interpreting
Section
5
of
Appendix
B
in
this
way,
EPA
will
avoid
placing
sources
in
an
impossible
situation
of
being
forced
to
conduct
"
stack
testing"
on
minor
emission
points
that
are
not
amenable
to
stack
testing,
while
still
ensuring
that
the
major
emission
points
at
the
source
are
tested.
In
fact,
this
is
very
similar
to
what
the
Agency
did
when
it
conducted
its
risk
assessments
and
identified
a
number
of
PCWP
facilities
as
qualifying
for
the
low­
risk
subcategory,
establishing
a
reasonable
precedent
for
calculating
risks
from
PCWP
emissions
sources.
3
This
approach
is
reasonable
because
it
is
unlikely
that
minor
emission
sources
 
such
as
log
vats,
chippers
and
blenders
 
would
be
risk
drivers,
or
even
make
a
meaningful
contribution
to
the
risk
from
the
source.

Option
2:
A
"
Tiered"
Approach
To
Quantifying
Emissions
Another
approach,
that
would
likely
require
rulemaking,
is
for
EPA
to
establish
a
"
tiered"
process
to
identifying
emissions
from
PCWP
process
units
where
they
are
likely
to
occur
in
significant
quantities.
AF&
PA
is
especially
concerned
about
the
burden
and
practicality
of
across
the
board
testing
for
emissions
of
acetaldehyde,
formaldehyde,
and
phenol.
As
noted
in
section
B
above,
EPA
has
already
made
a
determination
that
testing
of
HAP
metals,
MDI,
acrolein
and
benzene
need
only
be
done
at
sources
where
they
are
expected
to
be
found
in
relevant
quantities.
Unlike
these
HAPs
where
testing
is
focused
to
process
units
where
emissions
may
occur,
EPA
has
not
provided
comparable
targeting
for
acetaldehyde,
formaldehyde,
and
phenol.

As
a
preliminary
matter,
AF&
PA
believes
that
the
universe
of
potential
"
process
units"
at
a
PCWP
source
should
be
limited
to
those
identified
in
Table
1
of
the
preamble
of
the
final
rule
as
emitting
one
or
more
listed
HAP.
Notwithstanding
those
units
whose
definitions
in
40
C.
F.
R.
§
63.2292
self­
identify
them
as
a
process
unit,
as
discussed
under
Option
1,
there
is
no
place
in
the
final
rule
itself
that
explicitly
identifies
what
the
Agency
considers
to
be
process
units.
Table
1
of
the
preamble
therefore
provides
the
most
complete
and
comprehensive
listing.
Thus,
the
following
"
process
units"
should
be
evaluated
as
potentially
requiring
emissions
evaluation
(
testing,
emission
estimation,
or
determination
that
no
significant
emissions
emitted)
as
part
of
the
low
risk
demonstration:

P
softwood
veneer
dryers
P
primary
tube
dryers
P
secondary
tube
dryers
P
rotary
strand
dryers
P
conveyor
stand
dryers
P
green
rotary
dryers
P
hardboard
ovens
P
dry
rotary
dryers
P
pressurized
refiners
P
press
predryers
P
fiberboard
mat
dryers
P
board
coolers
P
reconstituted
wood
product
presses
P
engineered
wood
products
presses
P
softwood
plywood
presses
P
hardwood
plywood
presses
P
veneer
redryers
P
sanders
P
humidifiers
P
atmospheric
refiners
P
formers
P
blenders
P
stand­
alone
digesters
P
wastewater
operations
P
storage
(
resin)
tanks
P
lumber
kilns
P
log
vats
P
chippers
P
rotary
agricultural
fiber
dryers
P
agricultural
fiber
board
presses
P
hardwood
veneer
dryers
miscellaneous
coating
operations
P
fiber
washers
P
saws
From
this
list
of
thirty­
four
process
units,
AF&
PA
suggests
four
groupings
to
guide
where
testing
is
3
See
K.
Hanks
(
RTI),
"
Estimates
of
Ancillary
Plywood
and
Composite
Wood
Products
Process
Emissions
for
use
in
Risk
Modeling,"
(
Nov.
7,
2003).
4
needed,
no
testing
is
necessary,
and
emission
estimation
is
sufficient.

Group
I:
Units
that
Should
be
Tested
(
16)

The
following
sixteen
process
units
are
amenable
to
stack
testing,
include
the
sources
that
EPA
has
found
to
be
the
"
risk
drivers"
and
should
be
required
to
test
in
accordance
with
Section
5
of
Appendix
B:

softwood
veneer
dryers
primary
tube
dryers
secondary
tube
dryers
rotary
strand
dryers
conveyor
stand
dryers
green
rotary
dryers
fiberboard
mat
dryers
dry
rotary
dryers
pressurized
refiners
press
predryers
hardwood
veneer
dryers
rotary
agricultural
fiber
dryers
reconstituted
wood
product
presses
agricultural
fiber
board
presses
hardboard
ovens
atmospheric
refiners
Some
facilities
may
have
conducted
HAP
testing
of
these
and
other
process
units
prior
to
the
issuance
of
the
final
PCWP
MACT.
If
the
data
meets
the
testing
criteria
of
Section
5
then
EPA
should
accept
its
use
in
the
risk
assessment.
Repeating
tests
that
have
already
been
done
is
a
waste
of
resources
and
unnecessary
burden.

Group
II:
Use
of
Emission
Factors
in
place
of
Testing
HAP
testing
for
the
remaining
process
units
is
not
straightforward
given
the
diversity
of
process
units
at
PCWP
facilities.
However,
alternative
approaches
exist
that
allow
emissions
to
be
quantified
for
use
in
risk
assessments
without
going
through
site
by
site
testing
of
each
unit.
The
following
three
groupings
would
provide
the
necessary
quantification
of
HAP
emissions
from
specific
process
units
without
the
unnecessary
burden
and
cost
of
testing.

Group
II(
A):
Use
of
Existing
Emission
Factors
(
3)

There
are
three
process
units
for
which
EPA
has
already
developed
conservative
emission
factors.
3
AF&
PA
believes
that
it
would
be
appropriate
for
PCWP
sources
to
use
these
emission
factors
to
develop
conservative
estimates
of
the
emissions
from
these
process
units:

storage
(
resin)
tanks
miscellaneous
coatings
operations
wastewater
operations
5
Group
II(
b):
Estimate
Emissions
(
9)

AF&
PA
believes
that
it
would
be
reasonable
to
allow
PCWP
sources
to
estimate
emissions
using
published
emission
factors,
information
from
NCASI
Technical
Bulletins,
mass
balance,
best
engineering
judgment
or
other
reasonable
estimation
techniques
for
the
process
units
that
contribute
less
than
10%
of
the
Maximum
Allowable
Toxicity
Weighted
Carcinogen
Emission
Rate
and
Maximum
Allowable
Toxicity
Weighted
Noncarcinogen
Emission
Rate.
PCWP
sources
should
be
required
to
provide
the
basis
for
their
emissions
estimates,
with
a
brief
rationale
explaining
why
those
estimates
are
reasonable.
The
process
units
for
which
emission
factors
were
used
would
be
included
in
determining
the
average
stack
height.
Emissions
from
the
following
process
units
will
be
very
difficult
to
test
and
should
be
estimated
by
the
PCWP
source:

humidifiers
stand­
alone
digesters
formers
blenders
veneer
redryers
log
vats
fiber
washers
hardwood
plywood
presses
board
coolers
Group
II(
c):
Hard
to
measure
sources
(
2)

AF&
PA
finds
that
lumber
kilns
and
softwood
plywood
presses
are
particularly
difficult
to
test
given
the
fugitive
nature
of
the
emissions
or
the
lack
of
an
appropriate
sampling
point.
Therefore,
if
a
mill
chooses,
a
conservative
emission
factor
selected
from
those
in
AP­
42
can
be
used
to
estimate
emissions.

lumber
kilns
softwood
plywood
presses
Multiple
and
similar
process
units:

In
addition
to
the
three
groupings
above,
a
facility
should
be
able
to
use
tests
on
one
process
unit
to
develop
emission
factors
to
estimate
emissions
from
similar
process
units
at
the
same
facility
or
company
if
the
units
are
similar
in
design,
function,
raw
material,
and
end
product
process
parameters.
Most
individual
facilities
have
multiple
affected
units
that
are
similar
in
design,
function,
raw
material,
and
quality
specifications
with
only
slight
variations
due
to
configuration,
size
and
brand.

For
example,
virtually
all
indirect­
fired
(
steam
heated)
plywood
veneer
dryers
are
of
nearly
identical
design.
Modern
steam
heated
veneer
dryers
utilize
cross
circulation
with
axial
fans
that
blow
the
heated
air
across
jet
impingers
onto
the
passing
veneer.
After
the
heated
zone,
additional
fans
blow
ambient
air
across
the
veneer
to
cool
it
to
ambient
temperature.
The
only
meaningful
physical
difference
between
two
steam
heated
"
jet
dryers"
will
typically
be
their
capacities;
defined
by
the
number
of
hot
and
cooling
zone
chambers
and
operating
temperature.
Similarly,
all
plywood
presses
use
virtually
the
same
physical
process
to
cure
the
glue
line
forming
a
bond.
Hot
pressing
to
cure
the
glue
line
is
controlled
by
platen
temperature
and
press
time.
Therefore
if
similar
thickness
veneer
is
4
"
If
EPA
approves
that
your
process
unit
will
not
emit
detectable
amounts
of
benzene
or
acrolein,
that
unit
may
be
excluded
from
the
testing
requirement
in
this
table"
(
referring
to
Table
2)
6
used,
the
glue/
resin
compound
is
the
same,
along
with
similar
press
time,
HAP
emissions
should
be
equivalent.
These
independent
variables
can
be
accounted
for
across
multiple
intra­
facility
units
when
facilities
set
up
their
worst
case
testing
conditions
and
have
these
conditions
placed
in
their
Title
V
permits.

Taking
the
concept
one
step
further,
a
different
company
could
use
the
test
data
from
a
representative
unit(
s)
at
another
company
to
estimate
emissions
from
a
similar
unit
(
design
and
function)
in
their
company.
A
credible
mechanism
would
need
to
be
established
to
review
and
share
test
data
between
companies.
For
example,
new
test
data
could
be
added
to
AP­
42
or
incorporated
in
a
NCASI
Technical
Bulletin.
However,
this
process
would
need
to
be
expedited
since
mills
only
have
two
years
to
complete
emission
quantification,
conduct
risk
assessment,
and
submit
reports
to
EPA.

D.
Testing
for
Acrolein
and
Benzene:

Appendix
B
of
the
PCWP
MACT
provides
a
mechanism
that
limits
testing
for
acrolein
and
benzene
to
a
subset
of
process
units
(
see
footnote
on
page
497
of
2/
26/
04
signed
rule).
4
It
provides,
without
further
rulemaking,
a
way
for
EPA
to
identify
which
process
units
need
not
be
tested.
An
analysis
conducted
by
NCASI
(
attached)
shows
that
twelve
process
unit
types
were
tested
for
these
HAPs
and
HAPs
were
not
found
at
detectable
levels.
Therefore,
to
help
mills
in
developing
there
testing
plans,
AF&
PA
requests
that
EPA
determine
that
the
following
process
units
be
excluded
from
the
testing
requirements
for
acrolein
and
benzene:

blenders
chippers
engineered
wood
product
presses
fiber
washers
formers
hardwood
plywood
presses
hardwood
veneer
dryers
log
vats
sanders
saws
softwood
plywood
presses
veneer
redryers
In
addition,
the
following
process
units
are
not
expected
to
have
acrolein
or
benzene
emissions
even
though
no
testing
data
is
available
because
no
mechanism
of
formation
of
these
compounds
is
known.
Benzene
and
acrolein
are
typically
a
result
of
applying
heat
to
a
process,
either
direct
fired
or
where
high
temperatures
are
present
that
can
degrade
organics.

storage
(
resin)
tanks
miscellaneous
coating
operations
wastewater
operations
E.
Conclusion:
7
AF&
PA
believes
that
a
modification
to
the
rule
may
be
necessary
to
implement
the
tiered
testing
approach
described
above,
to
the
extent
it
is
not
already
allowed.
For
this
reason
 
as
well
as
because
of
the
generally
minor
nature
of
the
emissions
from
these
miscellaneous
process
units
and
the
fact
that
EPA
itself
disregarded
them
when
conducting
its
risk
assessment
 
AF&
PA
urges
EPA
to
adopt
Option
1
instead,
that
is,
requiring
stack
testing
of
the
process
units
enumerated
in
40
C.
F.
R.
§
63.2292,
.
Should
EPA
decide
to
adopt
an
expanded
tiered
testing
approach,
however,
AF&
PA
urges
the
Agency
to
act
expeditiously
so
that
facilities
can
determine
whether
or
not
they
qualify
for
the
low­
risk
subcategory
sufficiently
in
advance
of
the
compliance
deadline
such
that
the
requisite
controls
can
be
installed
on
sources
that
do
not
so
qualify.
Process
Unit
NCASI
Supporting
Analysis
Data
from
the
NCASI
Wood
Products
MACT
testing
program,
EPA's
AP­
42,
and
some
industry
generated
data
gathered
since
the
MACT
testing
were
evaluated
in
this
spreadsheet.

The
Analysis
worksheet
contains
information
on
detects
and
non­
detects
for
acetaldehyde,
acrolein,
phenol,
and
formaldehyde
for
sources
not
subject
to
control
This
worksheet
contains
minimum
and
maximum
concentration
values.
Some
qualitative
information
is
presented
relative
to
the
general
type
of
source
and
the
ability
to
test
the
source.

The
Summary
worksheet
groups
sources
into
various
categories
based
on
information
in
the
Analysis
worksheet.

The
Acrolein
worksheet
segregates
non­
affected
sources
based
on
acrolein
emissions
(
detects
or
nondetects

The
AffectedSources
worksheet
provides
information
on
acrolein
and
benzene
detects
and
non­
detects
for
affected
sources
(
sources
subject
to
control).
Analysis
AP­
42
information
Sources
listed
in
Table
1
of
the
Summary
of
Final
Rule
which
are
not
a
new
or
existing
affected
source
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
humidifiers
(
humidification
kilns)
acetaldehyde
6
0
0%
3.15
Yes/
No
Unknown
Yes
No
Humidifiers
are
not
'
set
up'
for
testing
by
EPA
methods.
They
are
unlikely
to
meet
EPA
flow
measurement
criteria.
Also
they
are
typically
batch
processes
with
irregular
venting,
creating
further
flow
measurement
problems.
The
vents
on
some
of
these
units
are
little
short
2
to
3
inch
pipes
with
rain
bonnets.
They
seem
to
be
mostly
for
release
of
pressure.

acrolein
6
3
50%
15.3
formaldehyd
e
6
1
17%
2.23
phenol
6
3
50%
0.39
atmospheric
refiners
(
particleboard)
acetaldehyde
6
6
100%
0.37
No
No
No
No
acrolein
6
6
100%
0.26
formaldehyd
e
19
16
84%
0.15
0.59
phenol
6
5
83%
0.32
0.41
atmospheric
refiners
(
fiberboard)
acetaldehyde
3
0
0%
5.6
No
No
No
No
This
refiner
used
recycled
white
water,
thus
substantially
different
than
particleboard
atmospheric
refiners.

acrolein
3
0
0%
0.42
formaldehyd
e
3
0
0%
1.75
phenol
3
3
100%
0.37
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
10
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
11
formers
acetaldehyde
9
9
100%
0.35
Yes/
No
Yes/
No
No
Yes
Formers
are
vented
primarily
for
dust
control
and
reclaim
of
materials.
These
venting
systems
were
not
designed
for
measurement
of
organic
gases
or
for
capturing
exhaust
from
the
formers.
Often
the
system
(
baghouse
or
cyclone)
that
captures
materials
from
the
former
also
collects
gas
streams
from
other
equipment,

such
as
conveyors.
In
most
cases,
only
parts
of
the
former
are
vented
to
a
collection
point
­
the
entire
former
is
not
vented
or
captured.

acrolein
9
9
100%
0.25
formaldehyd
e
26
3
12%
5.1
phenol
9
9
100%
0.29
blenders
acetaldehyde
6
6
100%
0.4
Yes
Yes/
No
No
Yes
Most
blenders
are
not
vented
through
a
testable
vent.
These
sources
are
fugitive
emissions
sources
in
many
or
most
cases.

Blenders
are
not
a
significant
dust
source
nor
is
there
any
loss
of
materials
from
blenders,
so
mills
do
not
generally
collect
gas
streams
from
a
blender.
Also
blenders
do
not
really
vent,
but
simply
dump
blended
material
into
a
conveyor.
Some
blenders
are
vented
and
some
are
configured
such
that
a
temporary
test
system
can
be
constructed.

acrolein
6
6
100%
0.28
formaldehyd
e
9
0
0%
4.5
phenol
6
6
100%
0.34
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
sanders
acetaldehyde
14
12
86%
0.89
Yes/
No
Yes/
No
No
No
Like
formers,
sanders
are
vented
primarily
for
dust
control
and
reclaim
of
materials.

Sanderdust
is
a
valuable
fuel
and
systems
are
designed
for
capturing
this
fuel.
(
In
a
few
mills,
sanderdust
is
incorporated
back
into
the
product.)
These
venting
systems
were
not
designed
for
measurement
of
organic
gases.
Sometimes
the
baghouse
used
to
capture
sanderdust
also
collects
gases
from
other
mill
processes.
Sanders
generally
tend
to
be
fairly
well
captured,

since
the
dust
is
so
fine
it
would
otherwise
create
a
nuisance.
Some
mills
have
sanders
that
are
set
up
for
testing
some
do
not.
Newer
mills
are
more
likely
to
have
a
good
test
location
for
sanders.

acrolein
14
14
100%
0.26
formaldehyd
e
20
8
40%
0.62
phenol
14
10
71%
1.16
saws
acetaldehyde
8
7
88%
0.4
Yes
Yes
No
No
Saws
present
a
significant
challenge.
Like
formers
and
sanders,
saws
are
vented
primarily
for
dust
control
and
reclaim
of
material.
So
saws
have
pneumatic
capture
systems
in
many
cases
but
they
are
generally
combined
with
other
sources
before
they
reach
a
baghouse
or
testable
location.
The
exhaust
ducts
that
handle
an
individual
saw,
prior
to
combining
with
other
exhaust
sources,
are
typically
not
suited
for
flow
rate
measurements.
Thus
it
is
very
difficult
to
measure
an
individual
saw.

Saws
that
cut
green
logs
are
generally
not
vented.

acrolein
8
8
100%
0.26
formaldehyd
e
14
9
64%
0.55
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
phenol
8
6
75%
0.41
fiber
washers
acetaldehyde
3
0
0%
2.2
Unknown
Unknown
No
Yes
acrolein
3
3
100%
0.26
formaldehyd
e
3
0
0%
0.77
phenol
3
3
100%
0.31
chippers
(
green
log)
acetaldehyde
2
2
100%
0.36
Unknown
Unknown
No
No
acrolein
2
2
100%
0.25
formaldehyd
e
2
2
100%
0.1
phenol
2
2
100%
0.3
chipper
(
dry
panel
and
veneer)
acetaldehyde
6
5
83%
0.47
Yes
Yes/
No
No
No
These
chippers
may
have
been
influenced
by
the
plywood
press
and
layup
line.
"
Mill
air"
pulled
into
the
chippers
and
exhausted
to
baghouse.
The
green
log
chipper
was
outside
the
mill.
Some
mills
call
these
hoggers.
Like
saws,
they
are
typically
not
vented
separately
but
are
combined
with
other
sources.
They
are
vented
for
dust
control
and
reclaim
of
materials.
They
will
be
difficult
to
test
in
many
mills.

acrolein
3
3
100%
0.28
formaldehyd
e
6
5
83%
0.38
phenol
6
4
67%
0.41
log
vats
acetaldehyde
3
0
0%
3.9
Yes
Yes
Yes
No
Most
of
these
are
open
systems.
Very
few
are
expected
to
be
testable.

acrolein
3
3
100%
0.33
formaldehyd
e
3
3
100%
0.19
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
14
phenol
3
3
100%
0.41
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
lumber
kilns
acetaldehyde
11
0
0%
Yes
Yes/
No
Yes
No
Most
kilns
are
very
difficult
to
test
accurately.
They
operate
on
an
extended
batch
process
­
up
to
3
days
for
western
softwoods,
generally
one
day
for
southern
pine.
The
venting
and
emissions
are
both
irregular
throughout
the
batch,
thus
'
snapshot'
samples,
unless
well
times,
are
of
marginal
use.
Fugitive
emissions
are
difficult
to
quantify.
NCASI
Technical
Bulletin
845
provides
an
overview
of
kiln
testing
­
approaches
and
problems.

acrolein
3
0
0%

formaldehyd
e
20
0
0%

phenol
3
3
100%

storage
tanks
acetaldehyde
no
data
Yes
Yes
No
Yes
acrolein
no
data
formaldehyd
e
no
data
phenol
no
data
wastewater
operations
acetaldehyde
no
data
Yes
Yes
No
No
acrolein
no
data
formaldehyd
e
no
data
phenol
no
data
miscellaneous
coating
operations
acetaldehyde
no
data
Yes
Yes
No
No
acrolein
no
data
formaldehyd
e
no
data
phenol
no
data
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
16
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
stand­
alone
digesters
acetaldehyde
no
data
Yes
Unknown
Yes
No
NCASI
was
unable
to
establish
a
flow
rate
for
the
one
batch
digester
'
tested'
in
the
MACT
program.
NCASI
eventually
decided
to
discard
the
data,
even
though
there
was
some
concentration
information.
This
is
a
pressurized
batch
process
with
a
'
blow'
so
that
the
exhaust
flow
is
highly
irregular.

Some
creativity
and
special
consideration
will
have
to
be
used
in
testing
these
sources.
EPA
methods
are
not
likely
to
be
applicable.

acrolein
no
data
formaldehyd
e
no
data
phenol
no
data
rotary
agricultural
fiber
dryers
acetaldehyde
no
data
Unknown
Unknown
Yes
No
acrolein
no
data
formaldehyd
e
no
data
phenol
no
data
agricultural
fiber
board
presses
acetaldehyde
no
data
Unknown
Unknown
Yes
Yes
acrolein
no
data
formaldehyd
e
no
data
phenol
no
data
dry
rotary
dryers
acetaldehyde
13
4
31%
DND
DND
No
No
Yes
No
acrolein
18
8
44%
DND
DND
formaldehyd
e
42
2
5%
DND
DND
phenol
12
10
83%
DND
DND
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
veneer
redryers
(
radio
frequency)
acetaldehyde
6
3
50%
DND
DND
Unknown
Unknown
Yes
No
acrolein
6
6
100%
DND
DND
formaldehyd
e
6
3
50%
DND
DND
phenol
6
6
100%
DND
DND
softwood
plywood
press
acetaldehyde
18
7
39%
DND
DND
Yes/
No
Yes/
No
Yes
Yes
Presses
with
cupolas
will
be
hard
to
test.

Some
sort
of
capture
system
will
need
to
be
manufactured
and
even
then
EPA
Method
flow
requirements
will
be
difficult
to
meet.

NCASI
Technical
Bulletin
768
discusses
the
work
done
to
test
a
cupola
on
one
plywood
press.
Some
plywood
presses
have
powered
fans
that
provide
adequate
test
sites.
Many
of
the
powered
fan
units
will
need
stack
extentions
or
some
modification
for
testing.

acrolein
18
18
100%
DND
DND
formaldehyd
e
32
8
25%
DND
DND
phenol
15
14
93%
DND
DND
hardwood
plywood
press
acetaldehyde
12
12
100%
DND
DND
Yes/
No
Yes/
No
Yes
Yes
Presses
with
cupolas
will
be
hard
to
test.

Some
sort
of
capture
system
will
need
to
be
manufactured
and
even
then
EPA
Method
flow
requirements
will
be
difficult
to
meet.

NCASI
Technical
Bulletin
768
discusses
the
work
done
to
test
a
cupola
on
one
plywood
press.
Some
plywood
presses
have
powered
fans
that
provide
adequate
test
sites.
Many
of
the
powered
fan
units
will
need
stack
extentions
or
some
modification
for
testing.

acrolein
12
12
100%
DND
DND
formaldehyd
12
3
25%
DND
DND
Source
HAP
Number
of
Runs
Number
of
Non­
Detects
%
ND
Minimum
ND
(
ppm)
Max
Detect
(
ppm)
Hard
to
Test?
Fugitive?
Heat
Applied?
Noncured
resins?
Comments
19
e
phenol
12
6
50%
DND
DND
engineered
wood
products
press
acetaldehyde
9
8
89%
0.8
Yes
Yes
Yes
Yes
NCASI
had
a
difficult
time
finding
Engineered
Wood
Products
presses
that
were
testable.
Some
will
be
testable.
Most
will
be
fugitive
or
hard
to
test
because
general
building
vents
convey
the
emissions
from
the
press.

acrolein
9
9
100%
0.26
formaldehyd
e
9
7
78%
0.82
phenol
9
9
100%
0.12
hardwood
veneer
dryers
acetaldehyde
24
3
13%
DND
DND
No
No
Yes
No
acrolein
21
21
100%
DND
DND
formaldehyd
e
24
4
17%
DND
DND
phenol
24
22
92%
DND
DND
DND
­
Did
Not
Determine
file
=
SumMiscSourcesForPhenAcetAcrol
Summary
Sources
listed
in
Table
1
of
the
Summary
of
Final
Rule
which
are
not
a
new
or
existing
affected
source
Sources
that
do
not
emit
detectable
levels
of
acetaldehyde,
acrolein,
formaldehyde,
or
phenol
Green
log
chippers
Sources
that
do
not
apply
heat
or
utilize
uncured
resin,
and
at
which
detectable
levels
of
acetaldehyde,
acrolein,
formaldehyde,
and
phenol
are
less
than
1
ppm
(
1)

Atmospheric
refiners
(
particleboard)
Sanders
(
1)

Saws
Dry
panel
and
veneer
chippers
Sources
that
apply
heat
and
resin
but
do
not
have
levels
of
acetaldehyde,
acrolein,
formaldehyde,
and
phenol
greater
than
1
ppm
Engineered
wood
products
presses
Fugitive
Sources
for
which
NCASI
does
not
have
data
Wastewater
operations
Miscellaneous
coating
operations
Storage
tanks
Sources
not
listed
above
and
for
which
NCASI
does
not
have
data
Stand­
alone
digesters
Rotary
agricultural
dryers
Agricultural
fiber
board
presses
Sources
that
will
be
difficult
to
test,
but
have
greater
than
1
ppm
of
acetaldehyde,
acrolein,
formaldehyde,
or
phenol
(
see
comments
on
'
Analysis'
worksheet
regarding
testing
difficulties)

Humidifiers
Log
vats
Lumber
kilns
Softwood
plywood
presses
Hardwood
plywood
presses
Other
sources
Atmospheric
refiners
(
fiberboard)
Formers
Blenders
Fiber
washers
Veneer
re­
dryers
(
radio
frequency)
Dry
rotary
dryers
Hardwood
veneer
dryers
21
(
1)
At
one
sander
1.16
ppm
of
phenol
was
measured
file
=
SumMiscSourcesForPhenAcetAcrol
Acrolein
AP­
42
and
NCASI
information
Sources
listed
in
Table
1
of
the
preamble
New
or
existing
source
subject
to
controls
in
Table
1A
or
Table
1B?
Acrolein
emission
data
available?
All
available
acrolein
samples
below
detection
limits?
Acrolein
detected
during
sampling?

conveyor
strand
dryers
Yes
These
sources
are
subject
to
fiberboard
mat
dryers
Yes
controls
as
either
new
or
green
rotary
dryers
Yes
existing
sources
hardboard
ovens
Yes
press
predryers
Yes
pressurized
refiners
Yes
primary
tube
dryers
Yes
reconstituted
wood
product
board
coolers
Yes
reconstituted
wood
product
presses
Yes
rotary
strand
dryers
Yes
secondary
tube
dryers
Yes
softwood
veneer
dryers
Yes
atmospheric
refiners
No
Yes
These
sources
are
not
subject
to
dry
rotary
dryers
No
Yes
controls,
but
have
emissions
of
humidifiers
No
Yes
acrolein
lumber
kilns
No
Yes
blenders
No
Yes
These
sources
are
not
subject
chippers
No
Yes
to
controls,
have
been
tested
for
engineered
wood
product
presses
No
Yes
acrolein
and
do
not
emit
acrolein
fiber
washers
No
Yes
at
detectable
levels
formers
No
Yes
hardwood
plywood
presses
No
Yes
hardwood
veneer
dryers
No
Yes
log
vats
No
Yes
sanders
No
Yes
saws
No
Yes
softwood
plywood
presses
No
Yes
veneer
redryers
No
Yes
New
or
existing
source
subject
to
controls
in
Table
1A
or
Table
1B?
Acrolein
emission
data
available?
All
available
acrolein
samples
below
detection
limits?
Acrolein
detected
during
sampling?

agricultural
fiber
board
presses
No
No
These
sources
have
not
miscellaneous
coating
operations
No
No
been
tested
for
acrolein
rotary
agricultural
fiber
dryers
No
No
stand­
alone
digesters
No
No
storage
tanks
No
No
wastewater
operations
No
No
AffectedSources
AP­
42
and
NCASI
data
were
used
for
this
analysis
Acrolein
Detectable?
Benzene
Detectable?
conveyor
strand
dryers
No
Data
fiberboard
mat
dryers
Steam
Heated
Yes
No
Direct
Fired
Yes
Yes
green
rotary
dryers
Yes
Yes
hardboard
ovens
Yes
No
press
predryers
Yes
No
pressurized
refiners
Yes
No
primary
tube
dryers
Steam
Heated
No
Yes
(
1)
Direct
Fired
Yes
Yes
reconstituted
wood
product
board
coolers
Yes
No
reconstituted
wood
product
presses
All
except
hardboard
Yes
(
2)
No
Hardboard
Yes
No
rotary
strand
dryers
Yes
Yes
secondary
tube
dryers
No
Yes
(
3)
softwood
veneer
dryers
Yes
(
4)
Yes
(
5)

(
1)
one
detectable
sample
run
at
one
MDF
dryer,
see
acrolein/
benzene
white
paper
(
2)
outside
of
the
hardboard
industry,
only
2
of
26
sample
runs
were
detect
for
acrolein
(
3)
one
detectable
sample
run
out
of
6
(
4)
2
or
3
detectable
sample
runs
(
5)
4
or
5
detectable
sample
runs
1
PCWP
MACT
Implementation
Questions
The
following
are
priority
questions
that
AF&
PA
and
its
member
companies
have
identified
that
need
to
be
addressed
before
PCWP
MACT
can
be
implemented
in
a
reasonable
manner.

1.
§
63.2262(
h)
specifies
how
to
determine
the
control
efficiency
across
an
emission
control
system
during
the
initial
performance
test.
The
equation
for
determining
percent
reduction
contains
a
term
(
CE)
for
capture
efficiency.
The
text
says
that
the
CE
is
to
be
"
determined
for
reconstituted
wood
product
presses
and
board
coolers
as
required
in
Table
4
to
this
subpart."
However,
the
rule
is
silent
about
how
to
determine
the
CE
for
control
systems
on
processes
other
than
wood
product
presses
and
board
coolers.
Are
we
to
assume
100
%
capture
efficiency
for
dryers
and
use
a
CE
of
1.0
in
the
equation,
or
attempt
to
measure
the
capture
efficiency
of
dryers
or
other
processes?

The
capture
efficiency
for
processes
other
than
for
reconstituted
wood
product
presses
and
board
coolers
should
be
assumed
to
be
100%
in
equation
1
of
§
63.2262(
h).
The
control
efficiency
calculations
used
to
determine
the
floor
for
the
MACT
rule
assumed
a
capture
efficiency
of
100%
for
all
dryers.

Here
are
the
locations
in
the
final
rule
where
changes
would
be
needed
to
make
this
clear.

From
page
388,
63.2262
(
h)

CE
=
capture
efficiency,
percent
(
applicable
only
to
reconstituted
wood
product
presses
and
board
coolers
as
required
in
Table
4
to
this
subpart)

From
page
389,
63.2262
(
i)

CE
=
capture
efficiency,
percent
(
applicable
only
to
reconstituted
wood
product
presses
and
board
coolers
as
required
in
Table
4
to
this
subpart).

From
page
423,
63.2292
Capture
device
means
a
hood,
enclosure,
or
other
means
of
collecting
emissions
into
a
duct
so
that
the
emissions
can
be
measured.
This
definition
is
applicable
only
to
reconstituted
wood
product
presses
and
board
coolers
as
required
in
Table
4
to
this
subpart.

From
page
423,
63.2292
Capture
efficiency
means
the
fraction
(
expressed
as
a
percentage)
of
the
pollutants
from
an
emission
source
that
are
collected
by
a
capture
device.
This
definition
is
applicable
only
to
reconstituted
wood
product
presses
and
board
coolers
as
required
in
Table
4
to
this
subpart.

2.
§
63.2250(
a)
states
that
"
You
must
be
in
compliance
with
the
compliance
options,
....
in
this
subpart
at
all
times,
except..."
All
processes
exhibit
variability,
and
2
may
not
meet
the
compliance
options
for
every
moment.
For
example,
when
valves
change
on
an
RTO,
there
is
a
fraction
of
a
second
that
the
emissions
bypass
the
RTO
and
are
not
treated.
The
units
can
meet
the
90%
control
efficiency
over
an
hour
average,
but
not
every
second.
Similarly,
emissions
from
the
press
are
cyclic,
peaking
every
time
the
press
opens,
and
vary
with
product
thickness.
The
press
may
be
able
to
meet
the
production
based
compliance
option
(
PBCO)
over
an
hour
long
average,
but
not
every
second
of
operation.
It
may
meet
the
PBCO
for
an
average
mix
board
thicknesses,
but
not
for
some
individual
products.
The
rule
does
not
specify
what
the
averaging
period
should
be
for
either
PBCO
or
Add­
on
Control
Systems
Compliance
Options.
The
rule
does
specify
a
6
month
averaging
period
for
the
Emissions
Averaging
Compliance
Option.
What
averaging
period
should
be
used
for
the
Add­
on
Control
Systems
Compliance
Options
and
Emissions
Averaging
Compliance
Option?

To
be
consistent
for
all
of
the
compliance
options,
a
6
month
averaging
period
should
be
used
for
the
Add­
on
Control
Systems
Compliance
Options
and
Emissions
Averaging
Compliance
Option.
Many
mills
may
implement
emission
averaging
to
be
able
to
take
advantage
of
the
longer
averaging
period.

3.
Can
plants
with
Green
Rotary
Dryers
operate
them
as
Dry
Rotary
Dryers?

Yes.
As
noted
in
the
preamble,
green
rotary
dryers
and
dry
rotary
dryers
are
essentially
the
same
in
terms
of
equipment
design.
Plants
have
the
option
to
permit
a
unit
that
is
nominally
a
green
rotary
dryer
(
with
controls)
as
a
dry
rotary
dryer
without
control
requirements
under
an
alternative
operating
scenario.
Wood
supply
changes
may
provide
an
opportunity
for
facilities
to
reduce
HAP
emissions
through
the
use
of
dryer
material
than
was
either
envisioned
during
the
permitting
process.
Additionally,
seasonal
wood
supply
changes
may
periodically
allow
a
unit
to
be
operated
within
the
work
practice
requirements
of
a
dry
rotary
dryer.

4.
Can
the
piping
involved
in
pneumatic
transfer
of
fiber
following
a
single
stage
(
primary)
tube
dryer
be
considered
a
"
secondary
tube
dryer"?

No.
Supplemental
heat,
either
indirect
or
direct,
needs
to
be
added
to
the
air
stream
or
applied
to
the
pneumatic
transfer
tubing
before
a
system
can
be
classified
as
a
secondary
dryer.

5.
Appendix
B
specifies
FTIR
methods
to
measure
benzene
emissions
(
specifically
EPA
Method
320
and
ASTM
Method
D6348­
03).
These
methods
are
complicated
and
may
require
multiple
test
methods
when
method
320
is
not
appropriate
for
measuring
other
HAPs
(
specifically
acetaldehyde,
acrolein,
formaldehyde
and
phenol).
NCASI
has
updated
its
Method
IM/
CAN/
WP
99.02
for
selected
HAPs
at
wood
products
facilities
to
measure
benzene.
The
updated
method
is
designed
as
a
self­
validating
method.
Updated
QA/
QC
procedures
(
Section
6.10)
require
duplicate
sample
runs,
spiked
sample
runs
and
field
spike
samples.
Benzene
can
be
included
in
the
spike
solutions
as
are
the
other
compounds
of
interest.
In
Section
2
of
the
method
(
Applicability),
3
benzene
is
listed
in
Table
2.1
as
a
compound
that
"
the
method
is
applicable
to
determining
the
concentration
of...."
Considering
that
EPA
approved
the
method
for
the
other
HAPS,
there
is
no
reason
EPA
should
exclude
benzene.
Can
other
methods
such
as
the
updated
NCASI
Method
IM/
CAN/
WP
99.02
be
used
to
sample
for
benzene?

Yes.
The
updated
NCASI
method
is
appropriate
and
was
not
intentionally
excluded
from
the
PCWP
rule.
This
is
a
technical
correction
that
will
be
made
as
the
rule
is
published.
Other
appropriate
methods
can
be
used,
subject
to
obtaining
approval
from
EPA
as
provided
in
the
MACT
General
Provisions
at
63.7(
e)(
2)
ii.

Note:
other
implementation
questions
have
already
been
discussed
with
EPA
for
their
consideration.
1
AF&
PA
white
paper
offers
alternatives
to
testing
each
process
unit
for
each
listed
HAP.
Focusing
HAP
Quantification
for
PCWP
Sources
to
Process
Units
that
Drive
Risk
Estimates
Based
on
extensive
risk
analysis
conducted
by
EPA
and
industry
to
date,
it
is
expected
that
most
of
the
risk
will
come
from
a
small
set
of
process
units
and
a
small
set
of
HAPs.
Current
evidence
shows
that
acrolein
will
be
the
driver
for
non­
cancer
risk
and
acetaldehyde
for
cancer
risk.
The
vast
majority
of
these
two
HAPs
come
from
direct­
fired
dryers
and
a
limited
number
of
other
sources.
However,
at
least
initially,
EPA
is
requiring
testing1
from
a
large
number
of
process
units
at
plywood
and
composite
panel
facilities.
Based
on
existing
information
and
engineering
judgment
AF&
PA
believes
these
analyses
will
show
that
many
of
these
sources
of
the
named
HAPs
will
never
be
the
"
risk­
driver"
and
thus
not
affect
the
ultimate
determination
of
whether
the
source
qualifies
for
membership
in
the
low
risk
subcategory.
Even
EPA's
preliminary
work
under
the
residual
risk
program
for
various
source
categories
confirms
that
a
handful
of
HAPs
from
a
handful
of
sources
will
make
up
most
of
the
risk
pool.

As
a
result,
EPA
should
establish
a
simple
listing
mechanism,
probably
through
rulemaking,
that
identifies
each
process
unit­
HAP
combination
that
would
not
appreciably
contribute
to
the
overall
risk
from
the
source
category.
AF&
PA
recommends
that
EPA
use
criteria
similar
to
that
used
in
the
final
rule
to
exclude
the
eight
named
facilities.
If
the
process
unit's
emissions
of
named
HAPs
are
less
than
10%
of
the
overall
source
category's
risk,
then
it
should
be
automatically
added
to
the
low
risk
unit
list.
Industry
could
support
the
listing
process
by
providing
sufficient
data
to
EPA
to
meet
the
established
criteria.
For
example,
if
3
to
7
representative
tests
or
estimates
showed
that
HAP
emissions
from
that
process
unit
were
less
than
10%
of
the
risk
threshold
(
contributes
less
than
0.1
to
the
HI
of
1)
at
any
of
the
mills
tested,
then
it
would
be
added
to
the
list.
From
then
on,
other
mills
would
not
have
to
include
that
unit­
HAP
combination
in
their
risk
assessment.
Over
time,
the
list
would
expand
as
more
information
comes
into
EPA
demonstrating
which
additional
process
unit­
HAP
combinations
should
be
excluded
from
future
risk
assessments.

This
option
would
be
helpful
to
both
the
regulators
as
well
as
the
regulated
industry
to
focus
limited
resources
on
the
key
emissions.
EPA
(
or
the
state)
will
be
reviewing
risk
assessments
provided
by
the
mills
with
documentation
on
the
inputs
and
outputs
to
the
models.
These
reports,
and
EPA's
reviews,
could
be
simplified
if
the
presentation
is
focused
on
those
emission
units
contributing
the
vast
majority
(
95%)
of
the
risk.
From
industry's
perspective,
substantial
effort
(
including
the
original
testing)
would
go
into
documenting
insignificant
risks
that
ultimately
will
have
no
affect
on
the
outcome.
In
addition,
this
option
creates
an
incentive
for
the
industry
to
do
early
testing
to
narrow
the
list
of
HAPs
and
process
units
before
having
to
test
or
estimate
emission
at
all
their
facilities.
