RESPONSE
TO
SIGNIFICANT
PUBLIC
COMMENTS
ON
THE
PROPOSED
CLEAN
AIR
MERCURY
RULE
Received
in
response
to:

Proposed
National
Emission
Standards
for
Hazardous
Air
Pollutants;
and,
in
the
Alternative,
Proposed
Standards
of
Performance
for
New
and
Existing
Stationary
Sources:
Electric
Utility
Steam
Generating
Units
(
69
FR
4652;
January
30,
2004)

Supplemental
Notice
for
the
Proposed
National
Emission
Standards
for
Hazardous
Air
Pollutants;
and,
in
the
Alternative,
Proposed
Standards
of
Performance
for
New
and
Existing
Stationary
Sources:
Electric
Utility
Steam
Generating
Units
(
69
FR
12398;
March
16,
2004)

Proposed
National
Emission
Standards
for
Hazardous
Air
Pollutants;
and,
in
the
Alternative,
Proposed
Standards
of
Performance
for
New
and
Existing
Stationary
Sources,
Electric
Utility
Steam
Generating
Units:
Notice
of
Data
Availability
(
69
FR
69864;
December
1,
2004)

Docket
Number
OAR­
2002­
0056
6.0
MERCURY
EMISSIONS
MONITORING
US
Environmental
Protection
Agency
Emissions
Standards
Division
Office
of
Air
Quality
Planning
and
Standards
Research
Triangle
Park,
North
Carolina
27711
15
March
2005
i
General
Outline
1.0
INTRODUCTION
AND
BACKGROUND
2.0
APPLICABILITY
AND
SUBCATEGORIZATION
3.0
PERFORMANCE
STANDARDS
FOR
COAL­
FIRED
ELECTRIC
UTILITY
STEAM
GENERATING
UNITS
4.0
PERFORMANCE
STANDARDS
FOR
OIL­
FIRED
ELECTRIC
UTILITY
STEAM
GENERATING
UNITS
5.0
MERCURY
CAP­
AND­
TRADE
PROGRAM
6.0
MERCURY
EMISSIONS
MONITORING
7.0
IMPACT
ESTIMATES
8.0
COMPLIANCE
WITH
EXECUTIVE
ORDERS
AND
STATUTES
9.0
NODA
10.0
OTHER
Appendix
A
LIST
OF
COMMENTERS
6­
1
6.0
MERCURY
EMISSIONS
MONITORING
6.1
SELECTION
OF
MONITORING
METHOD
Comment:

Several
commenters
(
OAR­
2002­
0056­
2429,
­
2634,
­
2718,
­
2922,
­
3565)
noted
that
in
the
proposed
MACT
in
subpart
UUUUU,
EPA's
proposed
40
CFR
part
75
revisions
include
provisions
for
use
of
both
Hg
CEMS
and
a
sorbent
trap
monitoring
system
meeting
Method
324.
However,
EPA
also
proposes
to
prohibit
new
units
that
commence
commercial
operation
more
than
6
months
after
publication
of
the
final
rule
from
using
Method
324.
The
commenters
do
not
believe
that
either
of
EPA's
proposed
alternatives
to
restrict
use
of
Method
324
is
appropriate
or
justified.

Response:

EPA
agrees
with
the
commenters.
In
light
of
comparative
field
data,
EPA
believes
that
monitoring
using
sorbent
media
should
be
as
similar
as
possible
to
monitoring
using
Hg
CEMS.
Therefore,
today's
final
rule
allows
the
sorbent
trap
methodology
to
be
used
for
new
units.

Comment:

One
commenter
(
OAR­
2002­
0056­
2375)
stated
that
if
the
EPA
promulgates
a
requirement
that
all
sources
above
a
certain
emissions
threshold
monitor
with
CEMS,
the
commenter
recommended
that
the
threshold
be
76
lb/
year.

Response:

EPA's
rule
will
require
continuous
monitoring
(
Hg
CEMS
or
sorbent
trap
monitoring
systems)
for
all
units
with
annual
Hg
mass
emissions
greater
than
29
lb/
yr.
For
units
with
Hg
emissions
of
less
than
or
equal
to
29
lb/
yr,
a
low
mass
emitter
option
is
provided
that
is
less
rigorous
but
still
environmentally
conservative.

Comment:

Numerous
commenters
(
OAR­
2002­
0056­
1596,
­
2101,
­
2429,
­
2485,
­
2634,
­
2718,
­
2827,
­
2850,
­
2861,
­
2918,
­
2922,
­
2932,
­
2948,
­
3406,
­
3432,
­
3455,
­
3536,
­
3539,
­
3565)
expressed
concern
that
EPA's
proposal
is
unfairly
and
unjustifiably
biased
against
the
sorbent
trap
method.
The
commenters
did
not
support
Alternative
#
1,
because
it
restricts
the
use
of
sorbent
traps
to
low
emitting
units.
Commenters
were
generally
more
receptive
to
Alternative
#
2,
except
for
the
proposed
quality
assurance/
quality
control
(
QA/
QC)
procedures
for
sorbent
trap
systems
(
most
notably
the
quarterly
relative
accuracy
testing),
which
they
found
to
be
inappropriate,
6­
2
overly
burdensome,
costly,
and
time­
consuming.
Several
commenters
stated
that
EPA
has
no
justification
for
restricting
the
use
of
the
sorbent
trap
method
because
it
has
been
shown
during
EPA­
sponsored
mercury
(
Hg)
monitoring
demonstrations
that
the
method
can
achieve
accuracies
comparable,
and
in
some
cases
better
than
those
achieved
by
Hg
CEMS.
Other
commenters
recommended
that
the
type
of
QA/
QC
procedures
prescribed
for
sorbent
trap
systems
should
be
more
specific
to
the
sorbent
trap
technology
and
should
be
more
clearly
defined.
Finally,
a
number
of
commenters
objected
to
the
proposal
to
report
the
higher
of
the
two
Hg
concentrations
from
the
paired
sorbent
traps,
and
recommended
that
the
results
be
averaged
instead.

Response:

Section
75.81(
a)
of
the
final
rule
adopts
a
modified
version
of
Alternative
#
2,
which
allows
the
use
of
sorbent
trap
systems
for
any
affected
unit,
provided
that
rigorous,
technologyspecific
QA
procedures
are
implemented.
The
operational
and
QA/
QC
procedures
for
sorbent
trap
systems
are
found
in
section
75.15
and
in
appendices
B
and
K
of
the
final
rule.
EPA
also
has
incorporated
the
recommendation
of
the
commenters
to
use
the
average
of
the
Hg
concentrations
measured
by
the
paired
sorbent
traps
meeting
specified
criteria.
And
in
cases
where
one
of
the
traps
is
accidentally
lost,
damaged
or
broken,
the
owner
or
operator
would
be
permitted
to
report
the
results
of
the
analysis
of
the
other
trap,
if
valid.

Recent
field
test
data
from
several
different
test
sites
indicate
that
sorbent
trap
systems
can
be
as
accurate
as
Hg
CEMS.
Recent
field
tests
have
answered
questions
regarding
which
substances
in
the
flue
gas
can
interfere
with
accurate
vapor
phase
Hg
monitoring
by
sorbent
traps.
Sorbent
trap
technology
also
has
evolved,
with
the
addition
of
a
third
section
that
enables
the
traps
to
be
subject
to
enhanced
QA
procedures.
And
the
Agency
has
been
working
with
industry
and
equipment
manufacturer
representatives
to
develop
new
QA
procedures
that
are
more
relevant
to
the
operation
of
a
sorbent
trap
system.
These
improved
QA
procedures
are
included
in
the
final
rule.
In
view
of
this,
EPA
believes
that
it
is
appropriate
to
extend
the
use
of
sorbent
trap
systems
to
all
affected
units.

EPA
notes
that
although
the
restrictions
on
the
use
of
sorbent
traps
have
been
removed,
there
are
some
inherent
risks
associated
with
the
use
of
this
technology.
For
instance,
because
sorbent
traps
may
contain
several
days
of
accumulated
Hg
mass,
the
potential
exists
for
long
missing
data
periods,
if
the
traps
should
be
broken,
compromised,
or
lost
during
transit
or
relative
accuracy
test
audit
(
RATA)
of
a
sorbent
trap
system
is
performed,
the
results
of
the
test
cannot
be
known
until
the
contents
of
the
traps
have
been
analyzed.
If
the
results
of
the
analysis
are
unsatisfactory,
the
RATA
may
have
to
be
repeated.
This
also
may
resulting
in
a
long
missing
data
period.
However,
EPA
believes
that
these
undesirable
outcomes
can
be
minimized
by
following
the
proper
handling,
chain
of
custody,
and
laboratory
certification
procedures
in
the
final
rule.
The
use
of
redundant
backup
monitoring
systems
can
also
help
to
reduce
the
amount
of
missing
data
substitution.

Comment:
6­
3
One
commenter
(
OAR­
2002­
0056­
4891)
stated
that
any
monitoring
method
that
meets
the
EPA
criteria
should
be
allowed.
As
noted
above,
accurate
Hg
monitoring
technologies,
including
accurate
continuous
emissions
monitoring
systems
for
Hg,
are
not
yet
commercially
available.
When
such
equipment
will
be
successfully
tested
and
commercially
available
is
not
known.
To
facilitate
the
use
of
new
Hg
monitoring
technologies
as
they
are
developed,
the
commenter
urges
EPA
to
allow
power
plant
owner/
operators
subject
to
the
Utility
Mercury
Reduction
Rule
(
UMRR;
now
known
as
the
Clean
Air
Mercury
Rule,
CAMR)
to
use
any
monitoring
method
that
meets
EPA's
standard
criteria
for
reliability
and
accuracy.
Unduly
limiting
the
options
available
for
monitoring
emissions
would
only
serve
to
drive
up
the
cost
of
what
is
already
an
extremely
expensive
regulatory
scheme.
Given
the
current
lack
of
accurate
Hg
emissions
monitoring
methods,
it
will
be
important
for
facilities
to
have
the
flexibility
to
use
monitoring
methods
that
are
developed
in
the
future
as
a
result
of
the
adoption
of
UMRR.

Response:

EPA
agrees
with
the
commenter
and
has
specified
a
performance­
based
approach
for
monitoring
criteria.
The
performance­
based
approach
allows
for
use
of
various
suitable
sampling
and
analytical
technologies
while
maintaining
a
specified
and
documented
level
of
data
quality.

6.2
MISSING
DATA
PROCEDURES
Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
were
concerned
that
EPA
had
overlooked
a
potential
problem
with
the
requirement
to
subject
Hg
CEMS
and
Method
324
to
periodic
RATAs
and/
or
other
audits
that
rely
on
comparison
to
EPA
test
methods
for
Hg
(
including
Method
29
and
the
Ontario­
Hydro
method).
Unlike
the
instrumental
reference
methods
routinely
used
to
QA
SO2
and
NOx
CEMS,
the
available
Hg
test
methods
can
take
days
to
complete
and
weeks
for
the
return
of
test
results
from
the
laboratory.
Under
the
current
general
provisions,
a
monitor
that
fails
a
RATA
would
be
deemed
out­
of­
control
beginning
with
the
hour
that
the
required
test
was
conducted
until
the
hour
that
the
test
is
successfully
passed.
This
construct
could
lead
to
significant
implementation
problems
with
respect
to
missing
data
and
requirements
to
calculate
and
report
data.
If
a
source
does
not
know
until
weeks
after
a
RATA
is
completed
whether
the
test
was
passed,
the
source
has
no
means
of
minimizing
missing
data
associated
with
a
failed
test.
Similarly,
once
it
is
clear
that
a
test
has
been
failed,
the
source
must
schedule
and
perform
a
new
test
and
wait
for
results
before
determining
whether
the
monitor
is
back
in
control
and
the
data
are
valid.
Under
these
procedures,
monitoring
systems
that
fail
a
RATA
will
have
significant
amounts
of
missing
data
due
simply
to
the
delay
in
obtaining
testing
results.
Until
a
method
is
developed
that
will
allow
for
onsite
results
of
Hg
RATA
testing,
EPA
must
provide
special
rules
to
avoid
these
unavoidable
implementation
problems.
6­
4
Response:

EPA
Agrees
with
the
commenters.
Based
on
field
testing.,
EPA
intends
to
develop
adequate
criteria
for
a
performance
based
instrumental
reference
method
for
Hg.
Initial
evaluations
of
such
a
method
have
already
begun.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2922,
­
2634,
­
2718)
stated
that
section
63.10020(
c)
states
that
any
period
for
which
a
monitoring
system
is
out
of
control
and
data
are
not
available
constitutes
a
"
deviation."
The
commenters
object
to
labeling
each
period
when
a
monitor
fails
a
QA/
QC
test
and
is
therefore
"
out­
of­
control"
as
a
"
deviation"
of
the
requirement
to
monitor.
Monitoring
systems
no
matter
how
well
maintained
will
occasionally
fail
a
QA/
QC
test.
As
long
as
the
source
takes
appropriate
action,
no
deviation
from
the
requirement
to
monitor
has
occurred.
The
commenters
are
especially
concerned
about
this
provision
given
the
uncertainty
surrounding
the
ability
of
the
Hg
CEMS
to
satisfy
the
proposed
standards
in
PS­
12A
on
an
ongoing
basis.

Response:

EPA
has
decided
to
control
Hg
emissions
using
a
cap­
and­
trade
approach
rather
than
by
using
maximum
achievable
control
technology
(
MACT,
40
CFR
part
63).

Comment:

Several
commenters
(
OAR­
2002­
0056­
1854,
­
2634,
­
2718,
­
2721,
­
2891,
­
2922,
­
3403,
­
3455,
­
3565,
­
2855)
stated
that
the
proposed
missing
data
procedures
seem
to
be
unduly
harsh
and
appear
to
be
unfairly
biased
against
the
use
of
the
sorbent
trap
method.
The
commenters
indicated
that
the
missing
data
routines
should
properly
consider
the
uncertainties
associated
with
Hg
monitoring,
i.
e.,
there
is
a
lack
of
evidence
that
high
percent
monitor
data
availability
(
PMA)
is
achievable
with
these
monitoring
systems.
Other
commenters
suggested
that
EPA
should
remove
the
maximum
potential
concentration
(
MPC)
provision
altogether
for
Hg
monitors
and
fill
in
all
missing
data
periods
using
average
concentrations
until
more
confidence
is
gained
in
the
reliability
of
Hg
monitors.

Response:

The
final
rule
retains
the
basic
missing
data
substitution
approach
for
Hg
that
was
proposed.
This
approach
has
worked
well
in
the
Acid
Rain
and
NOx
Budget
Programs.
The
conservative
nature
of
the
missing
data
routines
has
provided
a
strong
incentive
to
sources
to
keep
their
monitoring
systems
operating
and
well­
maintained.
However,
the
PMA
cut
points
in
the
final
rule
have
been
loosened
slightly
to
account
for
the
present
lack
of
long­
term
Hg
monitoring
experience
in
the
U.
S.
EPA
will
continue
to
collect
and
analyze
CEMS
and
sorbent
6­
5
trap
data
from
various
field
demonstration
projects
and
will
evaluate
the
performance
of
certified
Hg
CEMS
operating
on
similar
source
categories
(
e.
g.,
waste
combustors).
If
the
data
indicates
that
the
PMA
cut
points
should
be
changed
for
Hg
CEMS
or
sorbent
traps,
the
Agency
will
initiate
a
rulemaking
for
that
purpose.

The
suggestion
to
remove
the
MPC
provisions
and
to
fill
in
all
missing
data
periods
using
average
concentrations
until
EPA
develops
better
procedures
was
not
incorporated
in
the
final
rule
for
two
reasons.
First,
when
add­
on
emission
controls
that
reduce
Hg
emissions
either
malfunction
and
are
taken
off­
line,
uncontrolled
Hg
emissions
will
result.
If
the
Hg
CEMS
or
sorbent
trap
system
is
out­
of­
control
during
the
control
device
outage,
an
appropriate
substitute
data
value
must
be
used
to
represent
uncontrolled
Hg
emissions
and
provide
an
incentive
to
fix
the
Hg
monitoring
system.
The
MPC
concept
has
successfully
been
used
in
the
Acid
Rain
and
NOx
Budget
Programs.

Second,
EPA
does
not
agree
with
the
commenters
that
using
the
MPC
for
certain
missing
data
periods
is
always
unduly
harsh
or
punitive.
For
the
initial
Hg
MPC
determination,
the
March
16,
2004
SNPR
provided
three
options:
(
1)
use
a
coal­
specific
default
value;
or
(
2)
perform
site­
specific
emission
testing
upstream
of
any
control
device;
or
(
3)
base
the
MPC
on
720
hours
or
more
of
historical
CEMS
data
on
uncontrolled
Hg
emissions.
The
Agency
believes
that
these
options
provide
adequate
opportunity
for
affected
units
to
develop
appropriate
MPC
values.

Regarding
the
missing
data
routines
for
sorbent
trap
systems,
available
field
test
data
have
indicated
that
these
systems
are
capable
of
performance
that
is
equivalent
to
a
CEMS.
In
view
of
this,
EPA
believes
that
sorbent
traps
should
be
treated
on
a
a
more
equal
footing
with
Hg
CEMS
in
many
areas,
including
the
missing
data
provisions.

Finally,
EPA
notes
that
a
new
missing
data
policy
has
been
posted
on
the
Clean
Air
Markets
Division
web
site.
The
policy
allows
the
four­
tiered
missing
data
algorithms
to
be
applied
hour­
by­
hour,
in
a
stepwise
manner,
based
on
the
PMA.
Previously,
the
Agency's
policy
had
been
to
determine
the
PMA
at
the
end
of
the
missing
data
period
and
to
apply
a
single
substitute
data
value
(
sometimes
the
MPC,
if
the
ending
PMA
was
<
80
percent)
to
each
hour
in
the
missing
data
block.
This
new,
more
lenient
interpretation
of
the
40
CFR
part
75
missing
data
requirements
will
result
in
more
representative
missing
data
substitution
and
minimize
the
use
of
the
MPC.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
Appendix
A
to
the
Preamble­­
Proposed
Changes
to
Parts
72
and
75,
(
Proposed
Rules
March
16,
2004);
page
12418,
the
section
dealing
with
missing
data
seems
to
be
either
incorrect
or
unduly
harsh;
in
either
case,
the
section
requires
clarification.
The
procedure
states
that
missing
data
starts
when
the
traps
that
first
caused
the
problem
were
put
into
service,
but
it
further
states
that
this
missing
data
period
6­
6
ends
only
when
the
next
valid
Hg
concentration
data
are
first
obtained.
Because
the
data
is
not
"
obtained"
until
the
traps
have
been
analyzed
and
the
reports
sent
back
to
the
owner,
this
would
mean
that
the
missing
period
data
would
not
only
include
time
that
the
failed
traps
were
in
service,
it
would
also
include
the
time
period
that
the
next
valid
set
of
traps
were
in
service.

Response:

The
wording
in
the
final
rule
has
been
clarified
to
say
that
the
missing
data
period
would
end
on
the
commencement
of
operation
of
another
pair
of
sorbent
traps
that
contain
valid
Hg
concentration
data.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2922)
stated
that
the
proposed
Method
324
needs
to
include
provisions
for
data
availability
and
missing
data.
Because
it
is
not
reasonable
for
EPA
to
expect
100
percent
data
capture
from
any
method,
EPA
must
specify
at
what
point
sorbent
trap
data
would
need
to
be
filled
in
and
what
method
would
be
used.
EPA
might
also
consider
providing
alternative
minimum
data
collection
and
missing
data
requirements
that
would
to
apply
simply
to
Hg
data,
regardless
of
the
method
of
collection.

Response:

See
the
Missing
Data
discussion
in
the
preamble.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2922)
stated
that
EPA
needs
to
clarify
how
periods
of
startup,
shutdown,
and
malfunction
are
to
be
treated
in
data
collection
and
reporting.
The
proposed
rule
appropriately
states
that
deviations
that
occur
during
periods
of
"
startup,
shutdown,
or
malfunction"
("
SSM")
are
"
not
violations"
if
the
source
was
operating
in
accordance
with
its
SSM
plan.
However,
the
proposal
does
not
explain
whether
or
how
those
data
would
be
excluded
from
the
compliance
calculation,
or
how
they
would
be
treated
with
respect
to
the
data
collection
requirements.
Presumably
EPA
does
not
mean
that
any
12­
month
rolling
average
in
excess
of
the
standard
that
includes
a
period
covered
by
an
SSM
plan
is
not
a
violation.
On
the
other
hand,
data
collected
during
periods
of
SSM
are
not
representative
of
normal
operations
and
should
not
be
included
in
data
averages
used
to
determine
compliance
and
missing
data
substitution
procedures.
EPA
needs
to
give
additional
thought
to
these
issues
and
provide
clear
instructions
in
the
rule
for
how
such
periods
would
be
treated.

Response:

These
comments
pertain
to
the
January
30,
2004
NPR,
in
which
both
a
MACT
rule
and
a
NSPS
rule
were
proposed
for
Hg.
The
proposed
MACT
approach
has
not
been
selected
for
6­
7
promulgation.
The
NSPS
has
been
finalized
as
a
series
of
amendments
to
40
CFR
part
60,
subpart
Da.
The
NSPS
clearly
states
that
data
recorded
during
periods
of
unit
startup,
shutdown,
and
malfunction
are
not
included
in
the
calculation
of
the
12­
month
rolling
average
Hg
emission
rate.
However,
the
owner
or
operator
is
required
to
report
the
number
of
hours
excluded
from
the
calculations
for
those
reasons.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2922)
stated
that
EPA
proposes
to
require
sources
that
utilize
a
FGD
system
to
maintain
records
of
scrubber
operating
parameters
for
each
Hg
missing
data
period
in
order
to
show
proper
operation
of
the
scrubber.
Because
recording
of
FGD
parameters
generally
is
not
automated,
this
requirement
could
become
very
burdensome
if
there
is
a
significant
amount
of
missing
data.
As
a
result,
the
commenters
request
that
EPA
consider
allowing
sources
the
option
of
utilizing
parameters
other
than
control
device
operating
parameters,
such
as
documented
compliance
with
an
SO2
permit
limit
using
the
SO2
CEMS,
to
establish
proper
operation
of
the
FGD
during
Hg
missing
data
periods.

Response:

The
final
rule
allows
quality
assured
SO2
data
to
be
used
to
demonstrate
proper
operation
of
an
FGD.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2922)
stated
that
proposed
data
collection
and
missing
data
scheme
under
section
63.10008(
d)(
4)
has
several
significant
flaws.
The
first
flaw
is
the
assumption
that
these
minimum
criteria
are
reasonable
requirements
for
Hg
CEMS.
There
are
no
data
to
support
the
assumption
that
Hg
CEMS
will
be
capable
of
operating
within
the
specified
performance
criteria
for
18
hours
a
day
for
21
days
each
month.
Sources
should
not
be
penalized
for
failing
to
meet
minimum
criteria
that
may
not
in
fact
be
achievable.
As
a
result,
EPA
will
need
to
continue
to
review
this
requirement
in
light
of
additional
data
collected
between
now
and
the
first
compliance
deadline.
EPA
also
should
revise
the
rules
to
allow
use
of
Method
324
as
a
backup
for
any
source
that
chooses
that
option,
and
should
allow
data
from
that
method
to
be
used
to
meet
the
minimum
data
requirements
in
lieu
of
a
Hg
CEMS.
The
second
flaw
in
the
rule,
even
if
you
assume
that
the
Hg
CEMS
can
meet
the
minimum
criteria,
is
the
failure
of
the
rule
to
distinguish
between
unit
operating
hours
and
non­
operating
hours
in
determining
if
a
day
is
complete.
If
the
minimum
requirement
applies
regardless
of
unit
operation,
sources
will
be
required
to
try
to
keep
monitors
running
and
quality
assured,
and
to
calculate
mass
Hg,
even
if
the
unit
is
not
operating
in
order
to
collect
enough
data
for
a
complete
day.
That
is
contrary
to
section
63.10020(
a)
("
you
must
monitor
continuously
at
all
times
that
the
affected
source
is
operating").
That
would
also
result
in
months
with
little
operation,
and
therefore
very
little
actual
Hg
mass
emissions
data,
being
counted
in
the
12­
month
rolling
average
with
the
same
weight
as
months
with
significant
operation.
If,
on
the
other
hand,
the
source
does
6­
8
not
collect
that
data,
any
day
the
unit
does
not
operate
and
most
days
involving
a
startup
or
shutdown
are
likely
to
be
"
incomplete"
and
not
count
towards
the
required
21
days
for
a
complete
month.
Under
the
missing
data
provisions,
that
approach
would
mean
that
valid
data
would
be
thrown
out
simply
because
there
was
not
enough
unit
operation
in
the
month.
Sources
should
not
be
required
to
operate
monitors
when
they
are
not
operating,
and
should
not
be
penalized
for
non­
operation.
As
a
result,
EPA
should
give
more
thought
to
whether
the
data
collection
and
missing
data
provisions
associated
with
the
calculation
of
"
monthly"
averages
is
the
best
approach.

Response:

The
proposed
MACT
rule
has
not
been
selected
for
promulgation.
However,
the
proposed
NSPS
rule
for
Hg,
which
contained
the
same
data
capture
requirements
as
the
MACT
rule,
has
been
finalized.
In
the
final
rule,
the
minimum
data
capture
requirement
for
the
Hg
monitoring
systems
is
75
percent
of
the
unit
operating
hours
in
each
month.
If
this
requirement
is
not
met
for
a
particular
month,
a
substitute
Hg
emission
rate
must
be
reported.
Compliance
with
the
NSPS
emission
limit
for
Hg
is
determined
on
a
12­
month
rolling
average
basis.
The
rolling
average
is
weighted
according
to
the
number
of
valid
hours
of
Hg
data
collected
in
each
month,
except
when
the
75
percent
data
capture
requirement
is
not
met.
When
that
occurs,
the
substitute
Hg
emission
rate
for
that
month
is
weighted
according
to
the
number
of
unit
operating
hours
in
the
month.
Months
with
zero
unit
operating
hours
are
not
included
in
the
rolling
average
calculations.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2922)
stated
that
EPA
proposes
to
revise
40
CFR
75.20(
d)
to
include
Hg
CEMS
in
the
list
of
non­
redundant
backup
monitoring
systems
that
can
be
used
for
up
to
720
operating
hours
without
a
RATA.
The
commenters
request
that
EPA
revise
this
section,
and
other
sections
regarding
use
of
backup
monitoring
systems
and
missing
data,
including
40
CFR
75.80(
f),
to
allow
use
of
sorbent
trap
monitoring
systems
as
a
backup
to
Hg
CEMS
and
vice
versa.
The
EPA
should
also
allow
use
of
additional
paired
traps
as
a
backup
to
a
sorbent
trap
monitoring
system.

Response:

The
final
rule
does
not
prohibit
you
from
using
sorbent
traps
as
redundant
backup
monitors.

6.3
SORBENT
TRAP
OPERATION
AND
QA/
QC
In
view
of
the
many
comments
received
regarding
a
large
number
of
testing
and
QA
provisions
in
proposed
Method
324,
EPA
has
decided
to
revise
and
rename
proposed
Method
324
as
Appendix
K
to
Part
75
in
the
final
rule.
Based
on
comments
received
and
experience
6­
9
gained
from
field
tests
since
proposal,
Appendix
K
retains
certain
provisions
and
revises
others
in
proposed
Method
324
to
include
detailed,
performance­
based
QA
standards
and
procedures
for
sorbent
trap
monitoring
systems.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
section
75.15(
a)­(
e),
Page
12417,
conducting
proportional
sampling
with
a
Method
324
sampling
train
may
be
conducted
at
a
single
point.
If
the
velocity
profile
changes
with
load,
as
may
sometimes
be
the
case,
proportional
sampling
by
determining
the
velocity
at
a
single
point
may
misrepresent
the
total
gas
flow
in
the
stack.

Response:

40
CFR
part
75,
Appendix
K,
requires
that
the
output
from
a
part
75
certified
stack
gas
flow
monitor
be
used
to
maintain
a
proportional
sample
flow
rate
through
a
sorbent
trap
or
cartridge.
All
affected
sources
under
the
final
rule
must
already
comply
with
the
40
CFR
part
75
Acid
Rain
Program
monitoring
requirements
which
include
maintaining
and
quality
assuring
a
stack
gas
flow
monitor.
A
part
75
certified
flow
monitor
must
pass
a
two­
load
RATA
annually
and
a
three­
load
RATA
every
5
years,
and
is,
therefore,
representative
of
the
stack
gas
flow
rate
across
different
loads.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
Method
324
(
40
CFR
part
63),
does
not
meet
EPA's
definition
of
a
CEMS
(
page
12453,
section
72.2
Definitions).

Response:

It
is
true
that
in
section
72.2,
a
sorbent
trap
monitoring
system
is
not
included
in
the
definition
of
a
CEMS,
but
rather
is
defined
separately
as
an
"
excepted"
monitoring
system.
However,
EPA
believes
that
this
distinction
is
more
semantic
than
substantive,
since
a
sorbent
trap
system
samples
the
stack
effluent
continuously,
and
data
from
the
system
are
combined
with
hourly
readings
from
a
certified
40
CFR
part
75
flow
monitor
to
give
a
continuous
record
of
Hg
mass
emissions.
Relative
accuracy
test
data
have
shown
that
a
sorbent
trap
system
can
measure
Hg
concentration
as
accurately
as
a
CEMS.

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
notes
that
EPA
also
stated,
for
the
purposes
of
applying
Method
324,
"
an
intermediate
sampling
rate
of
0.3
to
0.5
L/
min
through
each
sorbent
trap
would
be
used
when
the
unit
is
operating
at
the
normal
load
level,
whether
low,
mid,
or
high.
The
sampling
rate
would
then
be
increased
or
decreased,
as
appropriate,
by
0.1
L/
min
when
the
6­
10
unit
operates
at
the
other
two
load
levels"
(
69
FR
12417).
The
commenter
believed
that
this
sample
rate
adjustment
procedure
is
not
appropriate.
The
proportional
sampling
is
not
necessary,
and
requiring
a
step
change
to
the
sample
rate
adds
further
complication
without
any
appreciable
benefit.
According
to
the
commenter,
a
mathematical
analysis
of
the
difference
between
constant
flow
sampling
and
proportional
flow
sampling
shows
a
negligible
difference
between
the
two,
with
a
percent
difference
of
1.45.
The
commenter
submitted
an
attachment
with
the
simulated
data
(
based
on
a
realistic
scenario).
The
commenter
stated
that
since
the
method
allows
for
a
±
25
percent
change
in
proportional
flow
sampling
as
mentioned
above
in
(
69
FR
4736),
the
percent
difference
between
constant
flow
sampling
and
proportional
flow
sampling
is
negligible
at
1.45
percent,
and
well
within
the
±
25
percent
range.
The
commenter
recommended
that
proportional
sampling
not
be
required
because
of
the
added
complications
in
the
design
of
the
sampling
process,
and
without
the
appreciable
increase
in
the
accuracy
of
the
measurements.
In
the
undesirable
alternative
that
the
EPA
requires
proportional
flow
sampling,
the
commenter
recommended
the
sampling
should
be
done
on
a
continuous
proportional
basis
and
not
on
a
3­
step
proportional
basis.
The
commenter
submited
further
explanation
of
the
reasons
proportional
flow
sampling
is
unnecessary
as
follows.
Since
Method
324
measures
the
concentration
of
the
Hg
in
the
stack,
it
is
independent
of
the
flow
rate.
The
procedure
included
in
the
rule
to
calculate
the
total
mass
of
the
mercury
emissions
includes
the
stack
flow
rate
as
follows:
"(
3)
If
you
use
Method
324
(
40
CFR
part
63,
appendix
A),
determine
the
12­
month
rolling
average
mercury
emission
rate
according
to
the
applicable
procedures
in
paragraphs
(
d)(
3)(
i)
through
(
v)
of
this
section.
(
i)
Sum
the
mercury
concentrations
for
the
emission
rate
period,
(

g/
dscm).
(
ii)
Calculate
the
total
volumetric
flow
for
the
emission
rate
period,
(
dscm).
(
iii)
Multiply
the
total
mercury
concentration
times
the
total
volumetric
flow
to
obtain
the
total
mass
of
mercury
for
the
emission
rate
period
in
micrograms.
(
iv)
Calculate
the
mercury
emission
rate
for
an
input
based
limit
(
lb/
Tbtu)
using
Equation
4
of
this
section.
(
v)
Calculate
the
mercury
emissions
rate
for
an
output­
based
limit
(
lb/
MWh)
using
Equation
5
of
this
section,"
(
69
FR
4724).
The
commenter
pointed
out
that
since
any
changes
in
stack
flow
rate
will
be
taken
into
account
in
this
calculation,
proportional
sampling
is
unnecessary.
The
commenter
asks
EPA
to
clarify
their
intent
on
requiring
proportional
flow
for
sampling
periods
greater
than
12
hours.
The
commenter
contends
that
this
is
not
a
necessary
requirement
for
accurate
measurement
of
the
Hg
emissions.

Response:

When
stack
gas
flow
changes,
Hg
concentration
may
also
change.
Therefore,
the
final
rule
requires
flow
proportional
sampling
to
better
ensure
a
representative
sample.
The
3­
step
adjustment
procedure
in
the
proposed
rule
has
not
been
finalized.
Rather,
the
ratio
of
the
stack
flow
rate
to
sample
flow
rate
must
be
kept
constant
(+
25
percent)
from
hour­
to­
hour.

Comment:

One
commenter
(
OAR­
2002­
0056­
2063)
stated
that,
in
reference
to
the
Preamble­­
Proposed
Changes
to
Parts
72
and
75,
(
Proposed
Rules
March
16,
2004),
Method
324,
since
Method
324
is
a
Reference
Method
and
the
main
vulnerability
on
an
ongoing
basis
is
6­
11
the
sample
volume
measurement,
quarterly
volume
measurement
calibrations
should
be
required,
rather
than
additional
Reference
Method
(
RAA)
testing.

Response:

The
final
rule
does
not
require
quarterly
RAAs,
but
has
instituted
quarterly
calibration
checks
of
dry
gas
meters,
and
sample­
specific
volume
QA/
QC.

Comment:

One
commenter
(
OAR­
2002­
0056­
0544)
suggested
that
should
sorbent
tubes
be
utilized
for
monitoring
purposes,
the
sorbent
media
should
not
be
restricted
to
iodated­
activated
carbon
only.
There
is
a
sorbent
media
used
to
test
for
Hg
vapor
in
ambient
air
method
listed
as
NOSH
6009
used
in
industrial
hygiene.
These
sorbent
tubes
are
readily
available
from
SKC
Inc,
manufacturer,
located
in
Pennsylvania,
part
number
226­
17­
1A.
These
tubes
have
been
used
by
many
Industrial
Hygienists
for
many
years
and
are
time
proven
for
their
reliability.

One
commenter
(
OAR­
2002­
0056­
2867)
asserted
that
EPA
should
publish
the
QA/
QC
for
the
digestion
procedure
in
this
rule
and
disclose
the
entity
that
has
developed
the
analytical
procedure
and
written
Method
324.
The
commenter
believed
this
will
increase
the
understanding
of
the
intent
in
requiring
this
sampling
method
and
analytical
procedure.
The
commenter
stated
the
current
digestion
method
appears
to
be
biased
towards
the
use
of
one
vendor's
procedure.
The
commenter
feels
this
may
give
them
an
unfair
market
advantage
in
supplying
the
traps
to
the
industry.
The
commenter
believed
alternate
analytical
procedures
should
be
allowed,
so
that
one
company
does
not
control
the
market.

Response:

The
sorbent
media
monitoring
requirements
(
formerly
Method
324)
have
been
revised
to
make
them
performance
based,
thus
providing
significant
additional
flexibility
in
choice
of
sampling
and
analytical
approaches,
as
well
as
performance
criteria
used
to
assess
the
quality
and
validity
of
the
monitoring
data
generated.
There
are
numerous
potential
approaches
for
sample
preparation,
depending
upon
the
analytical
technique
selected.
The
performance
based
approach
intentionally
avoids
specifying
that
level
of
detail.

Comment:

One
commenter
(
OAR­
2002­
0056­
3546)
stated
that
they
had
recently
conducted
additional
Hg
sampling
at
the
Navajo
and
Coronado
Generating
Stations
to
enhance
their
understanding
of
Hg
emissions
from
these
two
coal­
fired
facilities
and
the
challenges
of
testing
for
Hg
in
flue
gas.
The
commenter
notes
that
both
stations
were
part
of
the
1999
ICR
where
the
Ontario­
Hydro
method
was
used
for
Hg
sample
collection.
Attachment
1
to
the
commenter's
letter
contains
the
test
results
and
a
brief
description
of
sampling
protocols.
6­
12
Response:

The
Agency
appreciates
the
data
and
is
evaluating
it.

Comment:

One
commenter
(
OAR­
2002­
0056­
2101)
stated
that,
in
reference
to
Method
324,
if
a
separate
particulate
filter
is
used
in
front
of
the
trap,
this
filter
may
easily
absorb
gaseous
Hg
species
during
a
long
sample
run.
Is
the
inlet
filter
material
analyzed
along
with
the
sorbent
material?
If
so,
the
method
measures
particulate
Hg
as
well
as
gaseous.
Unless
iso­
kinetic
sampling
is
done,
the
particulate
fraction
may
be
over
or
under
represented.
If
the
particulate
filter
material
is
not
analyzed,
the
danger
exists
that
gaseous
Hg
has
deposited
on
it
and
will
be
discarded
along
with
the
material.
This
would
cause
serious
under­
representation
of
the
gaseous
phase
Hg.

Response:
40
CFR
part
75,
Appendix
K
in
the
final
rule
specifies
that
sorbent
media
should
be
the
first
thing
to
contact
stack
gas.

Comment:

One
commenter
(
OAR­
2002­
0056­
2101)
stated
that,
in
the
proposals,
CEMs
require
that
stack
flows
be
monitored
exactly
in
order
to
allow
calculation
of
an
accurate
emission
rate.
However,
in
the
proposed
Method
324,
even
when
variable
flow
rates
that
track
emission
volumes
are
required
(
only
for
samples
of
>
12
hours
duration),
the
sampling
flow
rates
need
agree
only
within
±
25
percent
of
stack
velocity
and
only
over
a
3:
1
dynamic
range.
This
wide
error
allowance
will
produce
inaccurate
mass
emission
calculations.

Response:

Appendix
K
of
40
CFR
part
75
requires
that
the
output
from
a
part
75
certified
stack
gas
flow
monitor
be
used
to
maintain
a
proportional
sample
flow
rate
through
a
sorbent
trap
across
all
load
levels,
not
just
the
three
specified
in
the
SNPR.
Most
of
the
affected
sources
under
the
final
rule
are
subject
to
the
Acid
Rain
Program
or
to
the
NOx
Budget
Program
(
or
both)
and
already
have
the
required
stack
gas
flow
monitor.
The
final
rule
retains
the
requirement
to
maintain
a
constant
(+
25
percent)
ratio
of
stack
gas
flow
rate
to
sample
flow
rate
from
hour­
tohour
EPA
will
evaluate
sorbent
trap
system
data
as
time
goes
on,
to
see
whether
the
+
25
percent
criterion
needs
to
be
tightened.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
noted
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
8.2.1
(
Sample
Collection),
page
4738,
an
isolation
valve
is
6­
13
shown
in
Figure
324­
1.
The
method
should
suggest
closing
the
valve
to
prevent
negative
or
positive
flow
due
to
stack
gas
pressure.
This
would
also
allow
the
tube
to
heat
to
the
correct
temperature
before
commencement
of
sampling.

Response:

The
sorbent
media
monitoring
requirements
(
formerly
Method
324)
have
been
revised
to
make
them
performance
based
providing
additional
flexibility
as
well
as
performance
criteria
used
to
assess
the
quality
and
validity
of
the
monitoring
data
generated.
The
drawing
referenced
by
the
commenters
is
now
offered
as
one
example
of
a
suitable
sampling
train
configuration.
There
are
numerous
potential
approaches
for
gas
control
and
the
performance
based
approach
intentionally
avoids
specifying
that
level
of
detail.

Comment:

One
commenter
(
OAR­
2002­
0056­
2101)
noted
that,
in
reference
to
Method
324,
section
8.2.3
(
Flow
Rates),
because
the
method
returns
a
true
dry
concentration,
the
instantaneous
stack
flow
rate
must
also
be
a
true
dry
value.
However,
the
velocity
measurement
is
a
normally
wet
value.
This
requires
one
of
the
following:

°
Verification
that
stack
moisture
levels
do
not
vary
significantly
with
velocity
or
over
time;
°
A
moisture
analyzer
to
allow
calculation
of
dry
flow
rate;
°
Some
other
precise
and
accurate
method
of
determining
current
moisture
levels;
or
°
An
error
analysis
showing
that
the
worst
case
errors
introduced
by
failing
to
perform
this
correction
are
not
significant.

Response:

The
final
rule
requires
that
Hg
concentrations
and
stack
gas
flow
rates
used
to
calculate
Hg
mass
emissions
must
be
on
the
same
moisture
basis.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2101,
­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
8.2.6
(
Moisture
Knockout),
page
4738,
the
section
should
state
that
data
for
the
entire
run
period
must
be
invalidated
in
the
event
that
the
leak
check
fails.
This
is
because
the
dry
gas
flow
meter
may
have
been
sampling
after
the
sorbent
cartridge,
not
through
it,
resulting
in
low
Hg
readings.

Response:

EPA
agrees.
Table
K­
1
in
Section
8.0
of
Appendix
K
specifies
that
failure
to
meet
the
6­
14
post­
test
leak
check
criterion
will
invalidate
the
sorbent
trap
monitoring
data.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2101,
­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
8.3.4
(
Field
Spikes),
page
4738,
this
section
gives
the
impression
that
the
Hg
is
spiked
onto
the
cartridges
before
exposure.
(
This
procedure
should
possibly
be
renamed
to
lab
spike,
rather
than
field
spike.)
This
procedure
yields
very
little
information.
Unless
the
sample
gas
matrix
actually
causes
the
removal
of
existing
Hg
on
the
trap,
or
interference
with
the
analysis,
this
procedure
will
always
generate
100
percent
recoveries.
Even
if
Hg
is
spiked
onto
the
cartridges
after
sampling,
this
does
not
provide
definitive
proof
that
Hg
was
being
properly
captured
under
actual
sampling
conditions.
Commenter
OAR­
2002­
0056­
2101
suggested
that
a
more
valid
type
of
spike
test
would
be
to
spike
the
cartridges
in
situ,
under
actual
sampling
conditions,
towards
the
end
of
a
lengthy
sampling
period.
This
could
be
done
by
adding
a
high
concentration
of
a
gas
containing
elemental
(
or
ionic)
Hg.
Spiking
would
be
done
at
a
low
flow
rate
(
e.
g.,
100
ml/
m).
This
spike
gas
would
be
sent
to
the
probe,
directly
in
front
of
the
cartridges
and
the
spike
duration
would
be
chosen
to
produce
a
significant
additional
loading
on
the
cartridge.
Because
the
spike
gas
is
produced
at
a
rate
significantly
below
the
cartridge
sampling
rate,
all
of
the
spike
would
be
drawn
through
the
cartridge
and
none
would
be
lost
out
of
the
tip
of
the
probe.
Only
in
this
way
can
one
be
assured
that
Hg
is
being
captured
quantitatively
throughout
the
entire
measurement
period.
The
spiking
gas
would
have
to
be
introduced
at
a
known
flow
rate
and
with
a
known
concentration.
A
conventional
saturated
Hg
vapor
generator
would
suffice.

Response:

In
the
final
rule,
Appendix
K
requires
spiking
of
the
third
section
of
each
sorbent
trap
with
elemental
Hg
prior
to
sampling.
The
purpose
of
the
spiking
is
to
serve
as
a
QC
check
of
the
laboratory
performing
the
analyses.
EPA
does
not
agree
with
the
commenter's
proposed
spiking
technique.
Unless
one
can
be
assured
of
100
percent
recovery
of
the
Hg
in
each
trap
(
the
Agency
does
not
believe
this
is
possible),
spiking
directly
in
front
of
the
sorbent
trap
will
cause
the
spiked
Hg
to
become
mixed
with
the
sampled
Hg,
making
it
impossible
to
separate
the
two.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
10.1
(
Calibration
and
Standardization),
page
4739,
the
standards
should
contain
the
same
quantity
of
leaching
agent
as
the
samples
being
analyzed.

Response:

The
sorbent
media
monitoring
requirements
(
formerly
Method
324)
have
been
revised
to
make
them
performance
based,
thus
providing
significant
additional
flexibility
in
choice
of
6­
15
sampling
and
analytical
approaches,
as
well
as
performance
criteria
used
to
assess
the
quality
and
validity
of
the
monitoring
data
generated.
There
are
numerous
potential
approaches
for
sample
preparation,
depending
upon
the
analytical
technique
selected.
The
performance
based
approach
intentionally
avoids
specifying
that
level
of
detail.

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
notes
that
Method
324,
section
11.1
states,
"
The
sorbent
traps
are
received
and
processed
in
a
low­
mercury
environment
(
class­
100
laminar­
flow
hood
and
gaseous
Hg
air
concentrations
below
20
ng/
m3)
following
clean­
handling
procedures"
(
69
FR
4739).
The
commenter
recommended
that
the
EPA
not
require
the
use
of
the
class­
100
laminar­
flow
hood.
The
commenter's
experience
with
EPA
Method
1631
for
low
level
Hg
in
water
indicates
that
processing
samples
in
a
class­
100
laminar
flow
hood
is
unnecessary
to
maintain
Hg
contamination
below
stated
levels.
However,
the
room
in
which
samples
are
processed
may
require
restricted
activity
to
keep
contamination
below
20
ng/
m3.
The
commenter
submitted
the
lab
should
be
able
to
achieve
and
maintain
a
low­
Hg
environment
by
the
means
it
deems
necessary.
In
the
event
that
the
EPA
chooses
to
require
the
class­
100
laminar­
flow
hood,
the
commenter
requested
EPA
to
specifically
define
the
underpinnings
for
the
use
of
a
class­
100
laminar­
flow
hood
and
define
the
requirements
of
a
class­
100
laminar­
flow
hood.

Response:

As
noted
previously,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
With
this
approach,
is
up
to
individual
laboratories
to
utilize
whatever
measures
they
find
necessary
to
achieve
acceptable
background
Hg
levels
in
order
to
achieve
the
performance
criteria
for
the
measurement.

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
notes
that
section
11.14
of
Method
324
states,
"
A
field
blank
is
performed
by
assembling
a
sample
train,
transporting
it
to
the
sampling
location
during
the
sampling
period,
and
recovering
it
as
a
regular
sample.
These
data
are
used
to
ensure
that
there
is
no
contamination
as
a
result
of
the
sampling
activities.
A
minimum
of
one
field
blank
at
each
sampling
location
must
be
completed
for
each
test
site"
(
69
FR
4740).
The
commenter
believed
the
requirement
to
assemble
an
entire
sample
train
to
make
a
field
blank
is
excessive.
The
commenter
asserted
a
field
blank
does
not
need
to
be
attached
to
a
separate
sample
train.
The
commenter
stated
simply
handling
the
field
blank
trap
along
with
the
actual
test
sample
trap
at
the
test
location
and
then
sending
it
to
the
lab
for
analysis
will
fulfill
the
goals
of
the
field
blank.
The
commenter
recognizes
that
the
requirements
for
a
field
blank
should
be
based
on
actual
test
data.
The
commenter
requests
that
the
EPA
publish
the
test
data
on
which
the
field
blank
requirements
are
based.
The
commenter
stated
that
EPA
should
provide
data
that
indicate
that
the
process
of
sample
train
preparation
and
sample
recovery
can
introduce
significant
contamination
into
the
sampling
procedure,
if
this
is
indeed
a
requirement.
6­
16
Response:

As
noted
previously,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
Blanks,
including
reagent
blanks,
method
blanks,
and
field
blanks
can
be
used
at
the
tester's
option
to
assess
sources
and
levels
of
sample
contamination.
Blank
correction,
however,
is
not
allowed.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2101,
­
3455)
noted
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
11.14
(
Field
Blanks),
page
4739,
this
section
does
not
say
what
is
to
happen
with
results
that
exceed
30
percent
of
the
measured
value.
The
data
should
be
considered
invalid.

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
11.14
(
Field
Blanks),
page
4739,
Table
324­
2
the
table
should
be
completed
so
the
corrective
action
is
available
for
all
QA/
QC
failures.

Two
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
11.4
(
Mercury
Reduction
and
Purging),
page
4739,
the
suggested
field
blank
of
30
percent
of
the
measured
value
is
much
too
high.
Surely
it
would
be
more
sensible
to
define
the
field
blank
in
absolute
mass
as
per
table
324­
2.
If
a
trap
has
a
max
capacity
of
1,800

g
then
this
would
equate
to
540

g,
which
would
be
a
serious
contamination
issue.

Response:

As
noted
above,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
Blank
samples
are
not
required,
but
are
rather
used
at
the
tester's
option
to
assess
sources
and
levels
of
sample
contamination.
Blank
correction
is
not
allowed
and
there
are
no
criteria
for
acceptable
blank
levels.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
11.3
(
Dilution
Step),
page
4739,
the
dilution
volume
is
not
specified
but
potentially
very
large
dilutions
are
required
to
fit
into
the
calibration
range
of
Method
1631.

Response:

As
previously
noted,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
With
this
approach,
it
is
up
to
individual
users
to
select
6­
17
suitable
analytical
techniques
and
associated
sample
preparation
steps
in
order
to
achieve
the
performance
criteria
specified
for
the
measurement
and
thus
dilution
steps
are
no
longer
specified.

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
sections
11.0
through
11.4
define
the
analytical
procedure
of
Method
324
and
use
a
water
digestion
method.
The
commenter
submitted
that
the
rule
should
include
an
avenue
for
approval
of
an
alternate
ASTM
analytical
procedure,
if
the
alternate
method
is
shown
to
provide
comparable
results
(
similar
to
use
of
ASTM
methods
for
the
analysis
of
coal
and
ash).

One
commenter
(
OAR­
2002­
0056­
2485)
noted
that,
in
reference
to
Method
324,
section
11.4,
the
selection
of
the
EPA
1631
as
the
preferred
method
to
analyzed
the
sorbent
tubes
has
no
scientific
justification.
This
method
is
extremely
complex
and
only
the
most
specialized
laboratories
have
the
necessary
resources
to
run
this
procedure.
The
commenter
also
noted
that,
in
reference
to
Method
324,
section
11.4,
the
Method
1631
method
has
many
requirements
that
are
not
appropriate
or
needed
to
analyzed
sorbent
traps.
The
1631
method
specifies
the
use
of
amalgamation
on
gold
to
decrease
detection
limits.
This
is
not
necessary
because
the
concentrations
in
the
final
solution
are
sufficiently
high.
The
amalgamation
step
complicates
the
measurement
and
introduces
more
error.

Two
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
noted
that,
in
reference
to
Method
324,
section
11.4,
the
Method
1631
method
has
many
requirements
that
are
not
appropriate
and
needed
to
analyzed
sorbent
traps.
The
1631
method
specifies
an
oxidation
using
BCL.
Is
this
step
necessary
as
the
Hg
species
will
be
already
digested
during
the
leaching
process?
What
is
the
purpose
of
the
BCL?

The
commenters
also
noted
that,
in
reference
to
Method
324,
section
11.4,
the
Method
1631
method
has
many
requirements
that
are
not
appropriate
and
needed
to
analyzed
sorbent
traps.
The
addition
of
NH2OH
after
the
BrCl
oxidation
to
remove
free
halogens
is
clearly
specified
in
the
1631
method
to
overcome
damage
to
the
gold
trap
and
prevent
low
collection
efficiency
(
M163l,
section
11.2).
Method
324
states
that
this
stage
is
omitted
allow
the
free
halogens
will
still
be
present.
What
is
the
justification
in
omitting
such
a
critical
stage
of
the
1631
method?
Furthermore
the
sorbent
leaching
step
uses
a
concentrated
HN03/
H2S04
mixture
at
elevated
temperature
in
a
closed
vessel.
As
mentioned
in
Method
324
in
section
11.2
significant
quantities
of
noxious
and
corrosive
gases
are
released.
These
gases
are
likely
to
be
NOx
fumes
which
may
combine
with
BrCl
to
produce
NOCl
fumes
which
may
also
attack
the
gold
trap
and
lower
its
collection
efficiency.

Response:

As
previously
noted,
the
sorbent
media
monitoring
requirements
have
been
revised
to
6­
18
utilize
a
performance
based
approach.
With
this
approach,
it
is
up
to
individual
users
to
select
any
suitable
analytical
technique
as
long
as
it
can
achieve
the
performance
criteria
specified
for
the
measurement.

Comment:

One
commenter
(
OAR­
2002­
0056­
2485)
noted
that,
in
reference
to
Method
324,
section
11.4,
Method
1631
has
many
requirements
that
are
not
appropriate
and
needed
to
analyzed
sorbent
traps.
The
method
was
designed
for
trace
levels
of
Hg
in
water
not
sorbent
tubes.
The
concentration
range
of
the
method
is
too
low
(
0­
100
ng/
L)
for
the
expected
mass
of
Hg
on
sorbent
traps.
For
example
if
the
flue
gas
Hg
concentration
was
1

g/
m3
sampling
for
4
hours
at
0.4
L/
min
would
yield
96
ng
Hg.
Insufficient
information
is
provided
in
the
method
to
calculate
the
final
concentration
of
Hg
in
solution
after
sample
preparation.
A
simple
calculation
however
would
suggest
that
the
mass
of
Hg
collected
would
have
to
be
diluted
using
1000
ml
to
produce
96
ng/
L
to
be
in
the
concentration
range
of
the
method.
If
one
considers
the
maximum
allowable
mass
collected
on
the
small
trap
(
150

g)
a
dilution
volume
of
l,
500,000
ml.
The
large
sorbent
trap
specifies
1800

g
capacity
so
a
dilution
volume
of
l8,000,000
ml
would
be
required.
Although
these
calculations
represent
the
upper
capacity
limits
they
surely
demonstrate
that
1631
is
not
an
appropriate
method
for
the
analysis
of
sorbent
tubes
and
the
potential
error
of
such
high
dilutions
Response:

As
previously
noted,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
With
this
approach
Method
1631
is
no
longer
specified;
it
is
up
to
individual
users
to
select
an
suitable
analytical
techniques
and
associated
steps
in
order
to
achieve
the
performance
criteria
for
the
measurement.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2101,
­
3455)
noted
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
11.6
(
Instrument
Calibration),
page
4739,
Method
1631
uses
a
calibration
factor
approach
to
calibration.
It
does
not
rely
on
r2
values.
The
method
should
standardize
on
either
a
weighted
or
unweighted
regression
approach
to
calibration
because
the
two
approaches
may
yield
significantly
different
results.
Note
that
the
Method
1631
analytical
approach
is
often
used
with
weighted,
(
1/
r2)
least
squares
curve
fitting.

Response:

Numerous
commenters
stated
that
Method
1631
is
inappropriate
for
the
analysis
of
sorbent
trap
samples.
EPA
concurs,
and
all
references
to
Method
1631
have
been
removed
in
the
final
rule.
The
final
rule
allows
any
suitable
analytical
technique
to
be
used.
For
the
calibration
curve
of
the
analyzer,
the
rule
simply
requires
an
r2
value
greater
than
or
equal
to
6­
19
0.99
and
requires
each
calibration
point
to
be
within
10
percent
of
the
true
value.

Comment:

One
commenter
(
OAR­
2002­
0056­
2889)
stated
that,
in
reference
to
Method
324,
section
11.7,
refers
to
section
15,
but
likely
should
refer
to
section
12.

Response:

Although
the
commenter
is
correct,
the
revisions
to
the
sorbent
monitoring
procedures
have
eliminated
the
need
for
this
reference.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
11.8
(
Continued
Calibration
Performance),
page
4739,
10
percent
drift
on
a
calibration
standard
is
too
high.
If
the
CV­
AAS
or
CV­
AFS
has
drifted
10
percent
then
the
probability
of
passing
a
RA
test
is
lessened
dramatically.

One
commenter
(
OAR­
2002­
0056­
2867)
noted
that,
in
reference
to
section
11.9
of
Method
324,
EPA
states,
"
The
QA/
QC
for
the
analytical
portion
of
this
method
is
that
every
sample,
after
it
has
been
prepared,
is
to
be
analyzed
in
duplicate
with
every
tenth
sample
analyzed
in
triplicate"
(
69
FR
4739).
The
commenter
claims
these
are
highly
excessive
requirements
for
the
QA/
QC
of
the
analytical
portion
of
the
method.
The
commenter
recommended
that
the
QA/
QC
be
reduced
to
one
duplicate
every
tenth
sample
analyzed,
as
allowed
under
other
EPA's
QA/
QC
programs.
In
other
EPA
monitoring
programs
such
as
the
methods
for
water
analysis,
QA/
QC
requirements
are
limited
to
a
duplicate
and
a
spike
for
duplicate
for
every
tenth
sample.
The
commenter
submits
that
if
required,
EPA
could
adopt
the
practices
used
in
water
monitoring
programs.

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that,
in
reference
to
section
11.9
of
Method
324,
the
sorbent
trap
laboratory
blank
requirement
of
3
percent
analysis
set
of
20
sorbent
traps
also
seems
excessive.
The
commenter
recommended
one
blank
per
every
tenth
sorbent
trap.

One
commenter
(
OAR­
2002­
0056­
2889)
stated
that,
in
reference
to
Method
324,
section
11.10,
should
also
require
the
independently
prepared
samples
to
be
within
10
percent
of
the
expected
value.

Response:

As
previously
noted,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach
and
this
level
of
detail
is
no
longer
specified.
Laboratories
conducting
analyses
of
the
sorbent
media
cartridges
for
Hg
must
either
be
ISO
6­
20
certified
or
accredited
or
must
perform
the
spike
recovery
study
described
in
Appendix
K
of
40
CFR
part
75
annually.
Thus,
adequate
QC
procedures
will
be
in
place
to
ensure
the
quality
of
the
data.

Comment:

One
commenter
(
OAR­
2002­
0056­
2889)
stated
that,
in
reference
to
Method
324,
section
11.13,
should
in
the
last
sentence
refer
to
10
percent
of
the
measured
sample
results.

Response:

The
commenter
is
referring
to
a
requirement
for
solution
blanks.
As
noted
above,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
Blank
samples
are
not
required,
but
are
rather
used
at
the
tester's
option
to
assess
sources
and
levels
of
sample
contamination.
Blank
correction
is
not
allowed
and
there
are
no
criteria
for
acceptable
blank
levels.

Comment:

One
commenter
(
OAR­
2002­
0056­
2889)
stated
that,
in
reference
to
Method
324,
section
13.0,
should
specify
the
consequences
of
failing
to
achieve
the
required
sample
rate
per
stack
flow.
Another
commenter
(
OAR­
2002­
0056­
2485)
stated
that,
in
reference
to
Method
324,
Table
324­
2,
the
table
should
be
completed
so
the
corrective
action
is
available
for
all
QA/
QC
failures.

Response:

EPA
agrees.
Table
K­
1
in
Appendix
K
specifies
that
failure
to
achieve
certain
performance
or
acceptance
criteria
will
invalidate
the
sorbent
trap
monitoring
data.

Comment:

One
commenter
(
OAR­
2002­
0056­
2485)
stated
that
section
13.0
of
Method
324
is
confusing.

Response:

EPA
agrees
and
has
clarified
the
flow­
proportional
sampling
requirements
in
Appendix
K
of
40
CFR
part
75.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
13.0
(
Constant
Proportion
Sampling),
page
4740
and
section
6­
21
8.2.4
(
Constant
Proportion
Sampling)
Option
1
&
Option
2,
these
sections
are
confusing.
If
these
criteria
were
to
be
followed,
a
source
could
set
their
flow
rate
based
on
75
percent
of
maximum
possible
stack
flow
and
be
in
compliance
continuously
from
50
percent
of
flow
to
100
percent
of
flow.
However,
the
commenter
believes
that
this
would
result
in
significant
under
reporting
of
full
load
data
and
over
reporting
of
reduced
load
data.
The
results
of
this
could
greatly
reduced
Hg
emission
rates.
Load
following
is
the
only
way
to
make
this
work
and
it
should
have
the
same
standards
as
Method
5
sampling
(+/­
10
percent).
Simply
picking
a
sample
flow
rate
as
suggested
in
the
amended
sections
is
not
appropriate
either.

Response:

EPA
understands
the
commenter's
theoretical
concern
and
will
evaluate
additional
data
from
actual
units
to
see
whether
the
permissible
deviation
criteria
needs
to
be
tightened
in
the
future.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
stated
that
several
provisions,
including
sections
13
and
14,
purport
to
contain
requirements
for
calculations
and
data
analysis.
These
sections
are
inadequate.
These
sections
should
contain
all
of
the
equations
and
calculations
needed
to
conduct
the
method
and
arrive
at
Hg
emissions
values
that
are
either
in
the
units
of
applicable
standards
or
that
can
be
converted
to
those
units.
Procedures
for
incorporating
blank
determinations
should
also
be
included.

Response:

EPA
has
addressed
the
commenters'
concerns
in
finalizing
the
sorbent
monitoring
procedures;
however,
the
final
procedures
do
not
include
provisions
for
blank
correction.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
stated
that
the
proposed
Method
324
appropriately
requires
leak
checks
of
the
sampling
line
with
and
without
the
sorbent
trap
in
place.
The
checks
are
performed
using
a
rotameter.
The
commenters
are
concerned
that
with
a
nominal
flow
rate
of
0.4
L/
min
(
see
Table
324­
1),
2
percent
of
that
flow
rate
(
which
is
the
maximum
leakage
allowed
under
the
proposal)
may
be
too
low
to
be
accurately
read
on
a
standard
rotameter.
Accordingly,
the
commenters
suggest
that
EPA
consider
revising
the
Method
324
to
quantify
the
leak
rate
based
on
readings
from
the
dry
gas
meter
over
a
period
of
at
least
1­
minute.

Response:
6­
22
EPA
understands
the
commenters'
concerns
and
has
revised
the
leak
check
procedures
to
clarify
that
the
dry
gas
meter
should
be
used
to
quantify
the
leak
rate.
The
leak
check
specifications
have
also
been
revised
to
allow
a
leak
rate
up
to
4
percent
of
the
planned
(
pre­
test
leak
check)
or
4
percent
of
the
average
sampling
rate
(
post
test
leak
check)
which
is
consistent
with
Method
5
(
40
CFR
60,
Appendix
A).

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
pointed
out
that
Table
324­
1
(
69
FR
4738)
of
Method
324
states
that
the
maximum
sample
duration
for
the
large
sorbent
trap
is
10
days.
The
commenter
recommended
that
the
maximum
sample
duration
not
be
limited
at
this
time.
The
commenter
stated
that
requiring
short
sampling
periods
adds
to
the
cost
of
the
monitoring
process,
as
the
traps
will
have
to
be
changed
more
frequently.
The
commenter's
experience
indicated
that
changing
the
sample
trap
results
in
approximately
30
minutes
of
unmonitored
time.
The
10­
day
sorbent
trap
stipulation
as
proposed
would
lead
to
increases
in
the
amount
of
unmonitored
time,
as
compared
to
a
longer
sample
period,
which
requires
changing
less
frequently.
The
commenter
submitted
that
the
use
of
shorter
length
traps
also
increases
handling,
analysis,
and
operations
support
compared
to
a
longer
sampling
trap.
This
extra
handling
and
analysis
will
also
introduce
more
opportunity
for
error.
The
commenter
recommended
that
the
sample
period
comport
with
capabilities
of
the
carbon
traps.
The
commenter's
experience
shows
that
the
traps
can
be
designed
and
used
for
longer
sampling
periods.
In
recent
tests,
the
commenter
compared
the
results
of
two
smaller
sorbent
traps
(
one
10­
day
and
one
11­
day
sorbent
trap)
to
one
large
trap
(
a
21­
day
sorbent
trap).
The
commenter
claimed
the
test
demonstrates
that
the
results
are
comparable
and
not
dependent
on
the
length
of
the
sample
period
or
size
of
the
trap
(
the
commenter
submitted
an
attachment
with
the
test
results).
The
commenter
pointed
out
that
the
percent
difference
between
the
weighted
average
concentration
measured
in
the
two
smaller
traps
versus
the
larger
trap
is
13.7
percent.
This
is
well
within
the
RA
range
proposed
by
EPA
of
20
percent
when
comparing
methods
to
Ontario­
Hydro
(
69
FR
12419).
The
commenter
has
plans
for
further
development
of
Method
324
to
lengthen
the
sample
duration.
The
commenter
envisioned
that
the
capabilities
of
the
sorbent
technology
will
improve
with
time,
leading
to
longer
duration
sample
collection.
In
this
context,
the
commenter
recommended
that
EPA
should
instead
focus
on
the
accuracy
of
information
and
not
arbitrarily
limit
the
sampling
period
to
10
days
or
a
month.

Response:

EPA
has
revised
the
sorbent
monitoring
procedure
to
be
performance­
based
and
has
eliminated
the
10­
day
limit
on
sampling
duration
leaving
it
up
to
the
tester
to
identify
a
sampling
duration
that
provides
the
correct
balance
between
convenience
and
performance.

Comment:

One
commenter
(
OAR­
2002­
0056­
2485)
stated
that,
in
reference
to
Method
324,
Table
6­
23
324­
2,
the
table
should
be
completed
so
the
corrective
action
is
available
for
all
QA/
QC
failures.

Response:

EPA
agrees.
Table
K­
1
in
Section
8.0
of
Appendix
K
specifies
that
failure
to
achieve
the
performance
or
acceptance
criteria
will
require
corrective
action
in
some
cases,
and
in
other
cases
will
invalidate
the
sorbent
trap
monitoring
data.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
stated
that
Table
324­
2
sets
out
the
QC
requirements
for
samples.
The
commenters
are
concerned
that
some
of
these
requirements
are
excessive
or
not
sufficiently
explained
or
studied.
For
example,
the
requirement
for
laboratory
blanks
could
be
excessive
for
large
trap
lots.
And,
without
more
data,
it
is
not
possible
to
determine
whether
the
paired
train
criterion
(
which
is
not
set
out
anywhere
else
in
the
rule)
is
subject
to
the
same
problems
as
PS­
12A,
section
8.6.6.
Finally,
the
field
spiking
requirement
is
not
sufficiently
explained.
These
problems
are
explained
more
fully
in
RMB's
comments.

Response:

As
noted
previously,
the
sorbent
media
monitoring
requirements
have
been
revised
to
utilize
a
performance
based
approach.
Blank
samples
are
not
required,
but
instead
are
used
at
the
tester's
option
and,
therefore,
there
are
no
criteria
for
acceptable
blank
levels.
The
paired
train
performance
criterion
has
been
based
on
actual
levels
achieved
during
several
EPA
demonstrations
as
well
as
stakeholder
supplied
data.
The
field
spiking
requirement
has
been
replaced
with
a
cartridge
spiking
requirement
for
QA
and
normalization
of
the
data;
this
new
procedure
has
been
adequately
detailed.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
stated
that
because
Method
324
is
a
test
method
itself,
they
believe
that
EPA
could
justify
a
rule
that
did
not
require
a
RATA
for
validation
of
the
sorbent
trap
system.
The
commenters
are
not
aware
of
any
other
instance
where
EPA
had
required
that
an
EPA
test
method
be
compared
to
another
test
method
using
RATA
procedures
prior
to
use
to
determine
compliance.
Use
of
the
RATA
in
this
case
is
unusual
given
the
mixed
performance
of
the
Ontario­
Hydro
method
that
would
be
used
for
the
RA
test.
Because
Method
324
is
a
new
method
that
has
not
been
employed
in
practice,
the
commenters
are
not
objecting
to
an
annual
RATA
requirement.
Such
testing
in
the
early
years
of
the
program
could
provide
valuable
information
for
improvement
of
both
methods.
However,
the
commenters
would
object
to
any
more
frequent
testing.
Method
324
already
provides
significant
QA/
QC
in
its
sampling
and
analysis
procedures
and
additional
RATA
testing
would
be
unwarranted.
6­
24
Response:

EPA
does
not
believe
that
the
sorbent
media
measurement
procedure
has
the
quality
to
serve
as
a
reference
method.
However,
we
are
confident
in
its
performance
as
a
monitoring
technique.
We
have
reconsidered
the
RATA
requirements
for
sorbent
monitoring
systems
and
concluded
that
a
yearly
RATA
plus
enhanced
performance­
based
QA
measures
in
Appendix
K
should
suffice
to
provide
accurate
measurements.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
stated
that
similar
to
proposed
PS­
12A,
section
8.1.1
of
Method
324
proposes
to
require
testing
for
SO2
and
NOx
stratification
at
the
proposed
installation
location.
This
requirement
is
not
consistent
with
other
EPA
regulations
and
should
be
revised
to
deem
the
location
suitable
as
long
as
the
RATA
is
passed.
As
in
EPA's
other
rules,
stratification
testing
should
only
be
required
if
the
source
uses
a
wet
control
device
(
or
is
otherwise
expected
to
have
stratification)
and
exercises
the
option
to
use
the
short
measurement
line
or
a
single
measurement
point
during
RATA
testing.

Response:

The
provisions
of
Method
324
have
been
revised
and
placed
in
40
CFR
part
75,
appendix
K.
Appendix
K
suggests
(
but
does
not
require)
that
stratification
testing
be
used
to
site
Hg
monitors.

Comment:

Several
commenters
(
OAR­
2002­
0056­
3455,
2485)
noted
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
8.1.6
(
Pre­
Test
Leak
Check),
page
4738,
a
leakage
rate
of
less
than
2
percent
of
the
recommended
flow
rate
of
0.4
L/
min
equates
to
less
than
0.008
L/
min.
This
flow
rate
is
not
measurable
using
a
flow
rate
in
the
range
of
the
method
(
0
­
0.8
L/
min).
A
leak
check
under
vacuum
would
be
more
effective
and
accurate.

Response:

EPA
agrees
that
the
leak
checks
should
be
conducted
under
vacuum.
The
pre­
test
leak
check
specifies
~
15"
Hg
vacuum,
while
the
procedures
have
been
revised
to
specify
that
the
post
test
leak
check
be
done
at
the
maximum
vacuum
reached
during
sampling.
In
addition,
the
leak
check
specifications
have
been
revised
to
permit
a
leak
rate
up
to
4
percent
of
the
planned
(
pretest
leak
check)
or
4
percent
of
the
average
sampling
rate
(
post
test
leak
check)
to
be
consistent
with
Method
5
(
40
CFR
60,
Appendix
A).

Comment:
6­
25
One
commenter
(
OAR­
2002­
0056­
2485)
noted
that,
in
reference
to
Method
324,
section
1.0,
the
title
is
misleading
as
the
sorbent
trap
is
not
dry
during
sampling
and
it
is
quite
likely
that
a
fraction
of
the
particulate
Hg
is
sampled
because
no
filtration
is
used.

Response:

The
provisions
of
Method
324
have
been
revised
and
placed
in
40
CFR
part
75,
appendix
K.

Comment:

One
commenter
(
OAR­
2002­
0056­
2485)
stated
that,
in
reference
to
Method
324,
section
6.1.2,
presumably,
the
large
and
small
sorbent
traps
have
the
same
type
of
sorbent
material.
Suggesting
the
use
of
larger
sorbent
tubes
at
higher
duct
temperatures
would
imply
that
breakthrough
of
Hg
is
potentially
a
problem
at
temperatures
of
375
°
F.
The
length
of
the
probe
holding
the
sorbent
tubes
is
not
specified.
It
is
quite
common
for
the
flue
gas
to
be
saturated
with
water
so
unheated
sorbent
traps
may
be
subjected
to
condensation.
Surely
it
is
advisable
to
heat
the
sorbent
traps
above
the
stack
gas
temperature
for
all
applications.

One
commenter
(
OAR­
2002­
0056­
3455)
stated,
in
reference
to
proposed
Method
324
(
40
CFR
part
63):
It
is
quite
common
for
the
flue
gas
to
be
saturated
with
water
so
unheated
sorbent
traps
may
be
subjected
to
condensation.
Surely
it
is
advisable
to
heat
the
sorbent
traps
above
the
stack
gas
temperature
for
all
applications.
Our
opinion
is
the
specified
operating
temperature
of
the
sampling
probe
for
Method
324
is
too
low.
Experience
and
data
taken
in
the
field
have
pointed
out
that
a
minimum
operating
temperature
of
400
°
F
is
required
to
transport
oxidized
Hg,
unless
the
sample
probe
is
recovered
along
with
the
sorbent
tubes.

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
their
experience
indicates
that
further
development
of
the
Method
324
for
use
on
a
wet
stack
is
necessary
and
can
be
accomplished
in
the
days
ahead.
The
commenter
noted
that
the
original
sampling
probe
does
not
work
in
a
wet
stack.
Condensation
builds
up
in
the
trap,
and
renders
the
sample
suspect.
The
commenter
stated
that
in
order
to
sample
in
a
wet
stack,
the
trap
must
be
inside
a
heated
probe.
The
commenter
pointed
out
that
Method
324
addresses
this
issue
in
section
6.1.2
and
states
that
the
sampling
probe
must
be
heated
in
duct
temperatures
less
than
200
°
F,
which
equate
to
wet
stacks.

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
stated
that
one
important
aspect
of
the
Method
324
measurement
is
avoiding
condensation
in
the
sorbent
trap
by
heating
the
sampling
probe
in
those
conditions
where
the
gas
stream
may
fall
below
the
condensation
point.
Section
6.1.2
requires
use
of
a
heated
sampling
probe
for
effluents
below
200
°
F
as
measured
with
a
thermocouple.
The
commenter
suggest
the
lower
boundary
of
the
range
be
increased
to
250
°
F
to
ensure
that
no
water
droplets
form
in
the
sorbent
trap.

One
commenter
(
OAR­
2002­
0056­
2867)
requests
that,
in
reference
to
Method
324,
6­
26
section
6.1.2,
EPA
confirm
that
using
a
heated
probe
does
not
introduce
differences
into
the
procedure
for
obtaining
a
dry
stack
sample
and
wet
stack
sample.
The
commenter
contends
that
any
introduced
differences
are
not
quantified
at
this
time,
and
should
be
investigated
further.
Questions
the
commenter
requests
be
answered
include:

°
What
are
the
differences
between
the
heated
probe
and
normal
probe?
°
Will
the
differences
introduce
any
variations
into
the
dry/
wet
stack
sampling
process?
°
Does
the
heated
probe
effectively
capture
the
Hg
in
the
stack?
°
If
the
heated
probe
is
used
on
the
wet
stack,
can
it
be
used
on
all
stacks
to
avoid
introducing
differences
between
the
wet
stack/
dry
stack
methods?
°
Is
this
a
sufficient
work
around
for
the
wet
stack
problem?

Response:

EPA
agrees
that
the
commenters'
concerns
are
all
valid.
However,
the
revised
sorbent
monitoring
requirements
are
performance­
based
and
so
achievement
of
the
performance
criteria
will
assure
that
these
issues
have
been
effectively
managed
for
each
sample.
The
final
sorbent
monitoring
provisions
advise
the
user
that
these
are
important
issues
to
consider
in
developing
their
sampling
strategy.

Comment:

One
commenter
(
OAR­
2002­
0056­
2101)
stated
that
the
preamble
should
clearly
say
that
Method
324
is
intended
as
a
measurement
method,
and
not
a
reference
method.
This
commenter
asked
if,
in
reference
to
Method
324,
section
1.0
(
AF
vs
AA),
tests
have
been
done
to
confirm
that
atomic
absorption
(
AA)
gives
answers
equivalent
to
atomic
fluorescence
(
AF)
under
all
circumstances?
Method
1631
is
a
performance
based
AF
method
that
allows
substitution
of
AA
provided
that
QA/
QC
criteria
are
met,
however
this
method
was
originally
intended
for
simple
matrices
like
waste
waters,
not
extracts
from
flue
gas.
(
It
should
be
noted
that
this
reviewer
is
not
aware
of
any
laboratories
that
have
achieved
true
Method
1631
class
performance
with
AA
based
systems,
even
for
the
simpler
matrices.)
EPA
should
exercise
extreme
care
to
ensure
that
AA
techniques
will
always
yield
identical
results
to
AF
techniques.

One
commenter
(
OAR­
2002­
0056­
2485)
stated
that
Method
324
should
also
be
subjected
to
the
same
RA
test
as
the
HgCEM.

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
Method
324
should
be
considered
as
a
viable
reference
method.
The
Ontario­
Hydro
method
is
the
currently
accepted
reference
method
for
measuring
Hg
concentration
in
a
flue
gas.
EPRI's
experience
shows
that
Method
324
has
less
variability
than
the
Ontario­
Hydro
method
in
side­
by­
side
comparisons
(
as
previously
seen
in
Attachment
1).
Method
324
is
also
a
continuous
monitoring
system,
whereas
the
Ontario­
Hydro
method
measures
the
Hg
concentration
for
a
snapshot
of
time.
Thus,
Method
324
has
6­
27
considerable
advantages
over
the
Ontario­
Hydro
method
as
a
reference
method.
Once
EPA
incorporates
Method
324
as
a
reference
method,
a
natural
extension
will
be
to
use
it
to
monitor
Hg
emissions
all
the
time
or
as
a
test
audit
for
Hg
CEMS.
This
argument
is
enhanced
by
the
underpinnings
in
other
CEMS
activities,
where
the
reference
method
is
used
to
periodically
check
the
daily
monitoring
method
or
can
be
used
for
daily
monitoring
in
lieu
of
CEMS.

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
for
applications
where
Hg
CEMS
are
used
and
for
Hg
monitoring
and
audits
to
be
advanced,
a
real
time
instrumental
reference
method
for
RA
audits
is
needed.
EPA
must
develop
such
an
instrumental
reference
method
for
Hg.

Several
commenters
(
OAR­
2002­
0056­
2485,
­
3455)
suggested
that
Method
1631
was
not
an
appropriate
method
to
analyze
sorbent
traps
for
many
reasons.

One
commenter
(
OAR­
2002­
0056­
2485)
noted
that,
in
reference
to
Method
324,
section
2.0,
this
section
states
that
other
recognized
procedures
can
be
used
for
the
analysis
of
sorbent
tubes
such
as
ASTM
D6784­
02
and
method
29.
Neither
of
these
methods
is
suitable
for
this
application.
Both
methods
use
various
impinger
solutions
none
of
which
relate
to
the
chemistry
used
to
leach
the
sorbent
tubes.
Suggesting
the
use
of
these
methods
will
ultimately
create
confusion
and
the
method
open
to
interpretation.
For
example
the
ASTM
method
has
digestion
and
preparations
for
KCl,
H2SO4/
KMnO4
and
H2O2/
HNO3
impingers.
Which
digestion
or
procedure
should
be
used
for
the
sorbent
tubes?
Furthermore
CV­
AAS
and
CV­
AFS
should
ideally
have
matrix
matched
standards
for
calibration
procedures
so
what
type
of
standardization
should
be
used
for
sorbent
tube
leachate?
The
CEN
13506
methods,
EPA
245.7
and
245.1
are
more
appropriate
options
because
they
do
not
use
amalgamation
and
have
wider
dynamic
ranges.

One
commenter
(
OAR­
2002­
0056­
2040)
proposes
an
alternative
analytical
technique
to
analyze
Hg
sorbent
tubes
for
proposed
Method
324.
This
technique
involves
direct
(
no
sample
preparation)
Hg
analyzer
Lumex
RA
915+
with
Attachment
Pyro915/
RP91C
for
testing
of
iodated
carbon
sorbent
from
the
tubes
as
an
alternative
to
chemical
digestion
with
following
atomic
fluorescent
analysis
of
tubes.
The
advantages
of
this
technology
compare
to
proposed
in
Method
324
as
follows:

°
Technology
is
based
on
field
portable
Atomic
Absorption
spectrometry
with
Zeeman
correction
coupled
to
a
furnace
heated
to
800
degrees
C
wherein
Hg
is
converted
from
a
bound
state
to
the
atomic
state
by
thermal
decomposition
in
a
two­
section
furnace.
In
the
first
section
of
the
furnace
the
"
light"
Hg
compounds
are
preheated
and
burned.
In
the
second
section
a
catalytic
afterburner
decomposes
"
heavy"
compounds.
°
Direct,
onsite
testing
of
tubes.
Save
time
for
shipping,
generating
of
chain
of
custody,
and
wait
for
laboratory
results.
°
Testing
results
will
be
available
within
1
hour
after
method
324
testing
tube
removed
from
the
sampler.
Analytical
throughput­
20
tests
per
hour.
°
No
chemical
waste
generated.
°
NIST
traceable
standards
used
for
multipoint
calibration.
6­
28
°
Detection
Level
to
0.5

g/
kg
is
10
times
lower
than
concentration
of
Hg
expected
for
Method
324
tubes.
°
No
compressed
gases
required.

Response:

In
view
of
the
many
comments
received
and
the
Agency's
own
field
testing,
EPA
has
decided
to
rename
proposed
Method
324,
and
to
revise
it
with
detailed,
performance­
based
QA
standards
and
procedures
for
sorbent
trap
monitoring
systems.
These
new
standards
and
procedures
are
now
included
in
40
CFR
part
75,
Appendix
K,
of
the
final
rule.
EPA
believes
that
by
taking
this
action,
there
will
be
less
confusion
and
more
convenience
for
users
affected
by
the
Hg
cap
and
trade
program.
Because
the
sorbent
trap
monitoring
system
requirements
in
Appendix
K
are
performance­
based,
the
results
of
any
analytical
technique
meeting
the
performance
criteria
should
be
comparable
to
any
other
analytical
technique
meeting
the
same
performance
criteria.

EPA
did
not
propose
Method
324
as
a
reference
method,
but
rather
as
a
set
of
analytical
and
QA
procedures
for
sorbent
traps
to
be
used
in
a
Hg
monitoring
program.
The
Ontario­
Hydro
Method
is
the
reference
method
for
Hg
in
part
75
of
the
final
rule.
However,
EPA
is
developing
a
Hg
instrumental
reference
method
as
part
of
the
Agency's
field
tests
of
Hg
monitoring
systems.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
noted
that,
in
reference
to
Appendix
A
to
the
Preamble­­
Proposed
Changes
to
Parts
72
and
75,
(
Proposed
Rules
March
16,
2004);
page
12417,
EPA
stated
that
sorbent
trap
systems
can't
be
calibrated
with
cylinder
gas
(
in
description
of
Alternative
2,
page
12417);
although
this
may
be
true,
it
does
not
mean
that
their
performance
can't
be
checked.
The
lack
of
QA/
QC
and
method
validation
is
a
very
disturbing
approach
to
proposing
a
new
sampling
method.
It
is
common
knowledge
in
the
measurement
industry
that
carbon
traps
can
become
passivated
in
the
presence
of
flue
gas
constituents.
In
addition
to
becoming
passivated,
traps
are
also
prone
to
re­
releasing
Hg
after
long­
term
exposure
to
flue
gas
constituents,
thus
creating
a
situation
that
would
seriously
under
report
Hg
concentrations,
and
unfortunately,
the
use
of
the
second
section
in
the
trap
does
nothing
to
address
either
of
these
problems.

Two
possible
solutions
to
this
problem
exist:
(
1)
one
trap
in
a
pair
could
be
spiked
and
placed
in
a
dual
sample
train,
when
the
train
is
analyzed
the
spike
could
be
subtracted
and
the
result
should
then
match
with
the
second
train
that
was
not
spiked;
or
(
2)
one
of
the
traps
in
the
paired
sample
could
be
exposed
to
a
Hg
calibration
gas
near
the
end
of
its
sample
period,
this
spike
could
then
be
subtracted
from
the
value
of
the
spiked
trap
and
that
value
compared
to
the
trap
that
was
not
spiked.
If
the
value
of
the
reference
trap
and
the
spiked
trap
minus
the
spike
do
not
match
(
i.
e.,
within
+/­
20
percent)
then
the
data
should
be
invalidated.
6­
29
One
commenter
(
OAR­
2002­
0056­
2101)
stated
that,
in
reference
to
Method
324,
the
periodic
RATA
test
against
a
reference
method
will
not
be
effective
in
ensuring
that
the
method
is
producing
accurate
results.
This
type
of
test
is
not
capable
of
detecting
cartridge
passivation
during
lengthy
Method
324
runs.
Passivation
is
expected
to
occur
relatively
late
in
a
sample
run.
Total
sample
volumes
passed
through
a
collection
cartridge
during
a
RATA
test
may
be
only
a
few
percent
of
the
total
volumes
that
will
be
passed
through
the
cartridges
during
normal
sampling.

Response:

Extremely
high
stack
gas
temperatures
(
greater
than
about
350
deg
F)
required
for
passivating
sorbent
material
are
not
expected.
On
wet
stacks,
40
CFR
part
75,
appendix
K,
states
that
sorbent
traps
may
need
to
be
heated
above
the
dew
point
to
prevent
condensation.
Therefore,
matrix
effects
that
cause
passivation
or
loss
of
Hg
trapping
efficiency
should
not
be
a
problem.
However,
EPA
will
be
examining
this
issue
further
during
our
field
testing
of
sorbent
media.

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
in
the
SNPR
Section
B.
3.,
"
Use
of
Mercury
CEMS
and
Sorbent
Trap
Systems,"
the
EPA
proposes
in
Alternative
2
that
the
sorbent
trap
method
can
be
used
by
any
source,
with
an
annual
9­
run
RATA
and
quarterly
3­
run
RAA
required
for
QA.
The
commenter
maintains
that
these
QA
requirements
are
excessive
and
the
results
(
pass/
fail/
bias)
will
remain
unknown
for
days
and
weeks.
The
commenter
also
stated
that
elsewhere,
the
EPA
stated,
"
For
sources
with
annual
Hg
emissions
below
the
specified
threshold
value,
the
QA
requirements
for
sorbent
trap
monitoring
systems
would
be
less,
with
only
an
annual
RATA
being
required"
(
69
FR
12417).
The
commenter
recommends
that
the
QA
requirements
for
sorbent
traps
­
all
units,
irrespective
of
emission
rates
­
be
consistent
and
limited
to
an
annual
RATA,
although
the
RATA
may
be
unnecessary
if
Method
324
is
ultimately
is
adopted
as
a
reference
method.
The
commenter
stated
that
the
accuracy
of
the
sorbent
trap
method
does
not
decrease
with
increasing
unit
size
or
larger
Hg
emission
rate.
The
commenter
also
stated
that
the
EPA
also
stated
"
the
Agency
is
willing
to
consider
replacing
the
RAA
requirement
with
another
type
of
substantive
quarterly
QA
test,
if
commenters
who
favor
the
use
of
sorbent
trap
systems
are
aware
of,
and
can
provide
details
of,
any
such
test
or
procedures"
(
69
FR
12417).
The
commenter
recommends
that
if
quarterly
checks
are
deemed
necessary,
the
EPA
might
include
semiannual
analysis
of
spiked
and
blank
traps,
and
quarterly
sample
flow
checks
and
pump
calibrations.

Response:

See
Sorbent
Trap
Operation
and
QA/
QC
discussion
in
the
preamble.

Comment:
6­
30
Several
commenters
(
OAR­
2002­
0056­
2079,
­
2485,
­
2634,
­
2718,
­
2861,
­
2922,
­
3455,
­
3565)
were
generally
opposed
to
the
proposed
quarterly
relative
accuracy
audits
(
RAAs)
for
sorbent
trap
systems
as
being
too
costly
and
of
little
value.
A
number
of
commenters
suggested
that
EPA
should
revise
proposed
Alternative
#
2
and
specify
QA
procedures
that
are
meaningful
to
the
type
of
measurement
system
that
the
sorbent
trap
actually
is.
For
example,
the
volume
of
stack
gas
sampled
by
the
system
is
an
important
parameter
in
determining
the
Hg
concentration.
Therefore,
procedures
for
quality­
assuring
the
measurement
of
the
sample
volume
could
be
implemented.

Some
commenters
favored
allowing
the
use
of
proposed
Method
324
for
all
affected
units,
and
stated
that
because
Method
324
is
itself
a
test
method,
it
does
not
need
additional
QA
procedures.
Two
commenters
suggested
that
EPA
should
even
take
steps
to
make
Method
324
a
reference
method.
However,
numerous
other
commenters
objected
to
various
provisions
of
proposed
Method
324
and
offered
suggestions
for
improving
it.
Some
of
the
chief
objections
raised
were
as
follows:


The
allowable
analytical
techniques
and
procedures
in
the
method
are
too
exclusive,
and
in
the
case
of
Method
1631,
inappropriate.
Other
analytical
methodologies
should
be
allowed;


The
impinger
and
dessicant
method
of
moisture
removal
is
inadequate;


The
leakage
rate
prescribed
for
the
leak
checks
checks
may
be
too
low
to
measure;


The
method
allows
constant­
rate
sampling
for
collection
periods
less
than
12
hours,
which
may
introduce
bias
if
unit
load
changes
during
the
collection
period;


The
specification
for
flow
proportional
sampling
(
adjust
sample
flow
rate
to
maintain
proportional
sampling
within
+
25
percent
of
stack
gas
flow
rate)
is
not
stringent
enough
and
can
lead
to
inaccurate
concentration
measurement;


The
frequency
for
dry
gas
meter
calibration
is
unspecified;
and

The
method
does
not
include
chain
of
custody
procedures.

A
number
of
commenters
suggested
that
EPA
should
not
require
the
use
of
paired
sorbent
traps
and
should
allow
the
use
of
single
sorbent
traps.

Several
commenters
objected
to
the
proposal
in
section
1.5.4
of
Appendix
B
that
laboratories
performing
Method
324
be
certified
by
the
International
Organization
for
Standardization
(
ISO)
to
have
proficiency
that
meets
the
requirements
of
ISO
9000.
One
commenter
stated
that
having
a
good
blank
and
matrix
spike
program
in
place
is
much
more
indicative
of
a
good
QA/
QC
program
for
Hg
measurement
than
ISO
9000
certification.
Another
commenter
favored
ISO
certification,
but
not
according
to
ISO
9000.
The
commenter
recommended
that
ISO
17025
be
required
instead,
because
it
requires
the
laboratory
to
demonstrate
proficiency,
rather
than
simply
having
an
acceptable
protocol
for
the
analyses.

One
commenter
stated
that
EPA
has
not
explained
the
appropriateness
of
applying
a
bias
6­
31
test
and
adjustment
factor
to
Method
324,
when
it
has
already
satisfied
the
same
standards
for
bias
and
precision
as
the
Ontario­
Hydro
Method
under
EPA
Method
301.
Another
commenter
suggested
that
it
does
not
make
sense
to
subject
Hg
monitors
to
a
bias
adjustment
factor
under
Appendix
A,
section
7.6
when
paired
reference
method
trains
are
allowed
to
differ
by
10
percent
relative
deviation
(
RD),
based
on
a
flawed
definition
of
RD.
The
commenter
asserted
that
it
is
not
reasonable
to
suggest
that
a
Hg
monitor
is
biased
by
comparing
its
readings
to
a
pair
of
reference
method
tests
that
can
differ
by
20
percent.

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
their
current
experience
with
Method
324
(
QSEMS)
indicates
that
component
reliability
has
to
be
improved.
The
commenter's
experience
indicates
issues
with
the
following:

°
Flow
meter
(
Totalizer)
­
Typical
problems
encountered
are
a
delayed
start
to
recording
the
data
and
discontinuous
recording
of
data
after
several
days.
°
Sampler
Pump
­
Readings
vary
by
±
25
percent
when
set
at
constant
flow.
The
flow
rate
also
decreases
overtime
as
particulates
build
up.
°
Currently
O
2
cannot
be
measured
and
recorded
with
this
device.
Correcting
to
3
percent
O
2
is
the
typical
way
to
compare
individual
test
runs
and
to
allow
the
detection
of
leaks
on
line
during
the
sampling
period.
°
The
carbon
trap
plugs
in
high
ash
situations.
The
method
will
not
work
upstream
of
an
ESP
without
an
inertial
separator
or
other
method
to
remove
the
ash.
°
For
wet
stack
application
the
probe
will
need
modification.

The
commenter
asserted
that
these
changes
are
required
before
QSEMS
can
be
relied
on
to
demonstrate
compliance.
The
commenter
notes
the
industry
is
investing
time
and
money
into
the
development
of
these
systems,
to
enhance
reliability
by
2010.
The
commenter
contends
it
is
important
that
the
compliance
deadline
not
be
moved
earlier
than
2010.

Response:

In
view
of
the
many
comments
received
and
the
Agency's
own
field
testing,
EPA
has
decided
to
rename
proposed
Method
324,
and
to
revise
it
with
detailed,
performance­
based
QA
standards
and
procedures
for
sorbent
trap
monitoring
systems.
These
new
standards
and
procedures
are
now
included
in
40
CFR
part
75,
appendix
K,
of
the
final
rule.
EPA
believes
that
by
taking
this
action,
there
will
be
less
confusion
and
more
convenience
for
users
affected
by
the
Hg
cap
and
trade
program.
Today's
rule
also
revises
both
the
definition
of
a
sorbent
trap
monitoring
system
in
section
72.2
and
the
general
guidelines
for
sorbent
trap
monitoring
system
operation
in
section
75.15,
to
be
consistent
with
the
QA
requirements
of
Appendix
K.

The
final
rule
retains
the
annual
RATA
and
bias
test
requirements
for
sorbent
trap
monitoring
systems,
but
the
proposed
quarterly
RAA
requirement
has
been
withdrawn.
The
requirements
to
use
paired
traps
and
flow
proportional
sampling
have
also
been
retained.
Finally,
the
ISO­
9000
certification
requirement
for
the
laboratory
performing
the
Hg
analyses
6­
32
has
been
replaced
with
a
requirement
for
the
laboratory
to
either
comply
with
ISO­
17025
or
to
comply
initially,
and
annually
thereafter,
with
the
spike
recovery
study
provision
in
section
10
of
40
CFR
part
75,
Appendix
K.

Several
commenters
recommended
that
EPA
should
require
QA
procedures
for
sorbent
traps
that
are
more
meaningful
and
reasonable
than
the
procedures
in
the
SNPR.
EPA
agrees
with
these
comments,
and
based
on
the
recommendations
received,
today's
rule
specifies
such
procedures
in
Appendix
K.
Many
provisions
of
Method
324
have
been
included
in
Appendix
K
without
modification,
but
other
provisions
of
the
method
have
been
modified
and
some
new
QA
procedures
have
been
added
to
address
concerns
expressed
by
the
commenters.
Some
of
the
more
significant
differences
between
Method
324
and
Appendix
K
are
as
follows:


Appendix
K
allows
the
use
of
any
sample
recovery
and
analytical
methods
that
are
capable
of
quantifying
the
total
vapor
phase
Hg
collected
on
the
sorbent
media.
Candidate
recovery
techniques
include
leaching,
digestion,
and
thermal
desorption.
Candidate
analytical
techniques
include
ultraviolet
atomic
fluorescence,
ultraviolet
atomic
absorption,
and
in­
situ
X­
ray
fluorescence;


Appendix
K
requires
that
each
sorbent
trap
be
comprised
of
three
equal
sections,
the
first
one
for
sample
collection,
the
second
to
assess
"
breakthrough",
and
the
third
to
allow
spiking
with
elemental
Hg,
for
QA
purposes;


Appendix
K
specifies
the
frequency
of
dry
gas
meter
calibration,
and
the
appropriate
calibration
procedures;


Appendix
K
requires
ASTM
sample
handling
and
chain
of
custody
procedures
to
be
followed.


Spiking
of
the
third
section
of
each
trap
with
elemental
Hg
is
required
before
each
data
collection
period
begins.


The
laboratory
performing
the
analyses
must
demonstrate
the
ability
to
recover
and
quantify
Hg
from
the
sorbent
media

The
measured
Hg
mass
in
the
first
and
second
sections
of
each
trap
is
adjusted
(
normalized),
based
on
the
percent
recovery
of
Hg
from
the
third
("
spiked")
section.

EPA
believes
that
if
these
procedures
are
implemented,
this
will
ensure
the
quality
of
the
data
from
sorbent
trap
systems.

The
final
rule
retains
the
requirement
to
use
paired
sorbent
traps.
The
SNPR
proposed
the
use
of
paired
sorbent
traps
for
the
same
basic
reason
that
paired
Ontario­
Hydro
trains
are
required
for
RATA
testing,
i.
e.,
it
provides
an
important
check
on
the
quality
of
the
data.
The
proposed
rule
would
have
required
the
higher
of
the
two
Hg
concentrations
obtained
from
the
paired
traps
to
be
used
for
reporting.
However,
the
final
rule
requires
the
results
from
the
two
traps
to
be
averaged
if
they
meet
specified
criteria,
and
allows
the
results
from
one
trap
(
if
those
results
are
valid)
to
be
reported
in
cases
where
the
other
trap
is
accidentally
damaged,
broken
or
lost
during
transport
and
analysis.
Thus,
using
paired
sorbent
traps
provides
a
relatively
6­
33
inexpensive
means
of
ensure
against
data
loss
should
one
of
the
traps
become
lost
or
damaged.

The
commenters
generally
objected
to
the
proposed
quarterly
relative
accuracy
(
RA)
testing
of
sorbent
traps,
believing
it
to
be
unnecessary
and
costly.
After
consideration
of
recent
field
data
comparing
the
sorbent
traps
to
Hg
CEMS,
EPA
agrees
that
sorbent
trap
systems
should
be
treated
more
similarly
to
Hg
CEMS.
Therefore,
the
final
rule
removes
the
quarterly
RAA
requirement,
and
requires
only
that
an
annual
RATA
be
performed
on
a
sorbent
trap
monitoring
system.

One
commenter
objected
to
the
proposed
bias
test
requirement
for
sorbent
trap
systems,
citing
the
fact
that
Method
324
had
satisfied
the
same
standards
for
bias
and
precision
as
the
Ontario­
Hydro
Method
under
EPA
Method
301.
EPA
does
not
agree
with
this
comment.
The
fact
that
Method
324
met
the
bias
and
precision
requirements
of
Method
301
does
not
imply
that
Hg
sorbent
traps
will
not
exhibit
low
bias
with
respect
to
a
Hg
reference
method
during
a
RATA.
The
bias
test
in
section
7.6
of
40
CFR
part
75,
appendix
A,
is
a
one­
tailed
t­
test,
which,
if
failed,
requires
a
bias
adjustment
factor
(
BAF)
to
be
applied
to
the
subsequent
emissions
data.

EPA
also
does
not
agree
with
the
commenter
who
stated
that
bias
adjustment
is
not
appropriate
for
sorbent
trap
systems
because
of
the
allowable
10
percent
RD
between
the
paired
reference
method
trains.
The
part
75
bias
test
determines
systematic
error,
not
random
error,
whereas
RD
and
relative
accuracy
are
metrics
used
to
quantify
random
error
in
the
measurement.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
suggested
that,
in
reference
to
Method
324
(
40
CFR
part
63),
section
1.1.2
(
Applicability),
page
4736,
a
Hg
RATA
should
be
performed
on
the
same
long­
term
time
basis
as
the
method's
use
of
a
CEMS,
i.
e.,
an
applicable
RATA
time
period
equal
to
or
greater
than
the
longest
averaging
period.
This
may
mean
multiple
Ontario­
Hydro
paired
train
runs
during
the
Method
324
extended
sampling
time
period.
Moisture
corrections
from
the
Reference
Method
should
be
applied
to
Method
324
data
for
the
entire
averaging
period.
A
short­
term
averaging
period
for
correlation
to
the
Reference
Method
does
not
provide
long­
term
assurance
of
carbon
trap
performance.

Response:

The
final
rule
requires
that
a
minimum
of
9
runs
be
used
to
calculate
a
RA
for
either
Hg
CEMS
or
sorbent
traps.
For
the
RATA
of
a
Hg
CEMS
using
the
Ontario­
Hydro
Method,
or
for
the
RATA
of
a
sorbent
trap
system
(
irrespective
of
the
reference
method
used),
the
minimum
time
per
run
must
be
long
enough
to
collect
a
sufficient
mass
of
Hg
to
analyze.
EPA
has
decided
to
implement
this
approach
as
a
compromise
between
the
desire
to
test
the
sorbent
traps'
long
term
performance
and
the
practicality
of
performing
a
potentially
week­
long
reference
method
test,
with
the
Ontario­
Hydro
Method.
6­
34
Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
proposed
Method
324
(
40
CFR
part
63),
section
1.1.2
(
Applicability),
page
4736,
performance
verification
criteria
must
be
established
to
assure
that
100
percent
cartridge
trapping
efficiency
is
maintained:

°
Over
the
full
permitted
temperature
range
of
the
sampling
train
and
over
the
full
stack
gas
temperature
range;
°
Over
the
full
flow
rate
range
of
the
sampling
train
(
especially
the
combination
of
highest
temperature
range
with
highest
flow
rate
­
which
would
be
expected
to
create
the
highest
probability
of
breakthrough.);
and
°
Over
the
full
range
of
stack
gas
compositions.
This
becomes
particularly
important
in
cases
where
control
technology
is
employed.

Response:

The
extremely
high
stack
gas
temperatures
required
for
passivating
sorbent
material
are
not
expected
to
be
encountered
in
utility
stacks.
On
wet
stacks,
Appendix
K
of
40
CFR
part
75
states
that
sorbent
traps
may
need
to
be
heated
above
the
dew
point
to
prevent
condensation.
Therefore,
matrix
effects
that
cause
passivation
or
loss
of
Hg
trapping
efficiency
should
not
be
a
problem.
However,
the
Agency
will
be
investigating
this
issue
further
during
our
field
testing
of
sorbent
media.

Comment:

Commenter
OAR­
2002­
0056­
2922
stated
that
section
75.15
contains
the
Special
Provisions
for
using
the
sorbent
trap
monitoring
method
(
Method
324).
Section
75.15(
e)
specifies
proportional
sampling
and
then
further
explains
the
proportional
sampling
procedure
as
a
change
in
the
sampling
rate
in
relation
to
load.
This
procedure
is
flawed
and
is
in
conflict
with
Method
324.
Any
proportional
sampling
method
will
require
stack
flow
rate
or
load
input
into
the
sampling
device.
The
sorbent
traps
will
be
located
at
a
stack
or
duct
location
and
manually
making
the
required
flow
changes,
at
odd
times
of
the
day,
will
just
not
be
feasible.
EPA
should
make
this
section
consistent
with
Method
324
and
allow
for
automated
input
of
stack
flow
or
load
data
into
the
sorbent
sampling
system
and
allow
for
automated
flow
rate
adjustment.

Response:

EPA
agrees
with
the
commenter
and
the
final
rule
requires
the
flow
control
valve
and
air­
tight
sample
pump
to
be
controlled
by
the
40
CFR
part
75
certified
flow
monitoring
system.
The
final
rule
also
requires
that
the
data
acquisition
and
handling
system
ensure
that
the
sampling
rate
is
proportional
to
the
stack
gas
volumetric
flow
rate.

Comment:
6­
35
Commenter
OAR­
2002­
0056­
2922
stated
that
section
75.15(
i)
should
have
the
words
"
or
partial
hours"
added
to
the
first
sentence.

Response:

EPA
believes
that
the
suggested
change
is
unnecessary
in
light
of
the
definition
of
"
unit
operating
hour"
in
section
72.2
which
includes
"
any
hour
(
or
fraction
of
an
hour)."

6.4
QA/
QC
PROCEDURES
FOR
HG
CEMS
Comment:

One
commenter
(
OAR­
2002­
0056­
3406)
stated
that
given
the
early
stages
of
the
development
of
Hg
CEMS
technology,
it
is
inappropriate
to
specify
a
mandatory
performance
specification.
The
commenter
believed
facilities
subject
to
this
requirements
should
have
the
option
of
proposing
and
complying
with
an
alternative
performance
standard
that
provides
substantially
similar
assurances.

Response:

The
final
rule
contains
an
alternative
relative
accuracy
performance
specification
for
low
emitting
sources.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
suggested
that,
in
reference
to
PS­
12A,
the
final
rule
should
change
the
cycle
time
(
from
the
proposed
fixed
15
minutes)
to
a
field
determined
cycle
time.
Elemental
Hg
can
achieve
a
15­
minute
cycle,
however,
currently
that
cycle
time
would
be
difficult
for
oxidized
Hg
to
achieve.
EPA
should
continue
to
work
with
vendors
with
a
goal
of
achieving
15
minute
cycle
times
for
all
forms
of
Hg.
A
TGM
blend
will
also
take
longer.

Response:

EPA's
field
testing
indicates
that
15
minute
response
times
and
cycle
times
are
achievable
for
both
elemental
and
oxidized
Hg.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
suggested
that,
in
reference
to
PS­
12A,
the
CEMS
should
be
challenged
with
a
calibration
gas
blend
of
Hg0
and
Hg+
2
at
a
frequency
consistent
with
40
CFR
part
60,
Appendix
B,
PS­
2,
introduced
at
the
sample
acquisition
point,
prior
to
any
filtration;
to
access
sample
transport
for
data
validation.
6­
36
Response:

EPA
agrees
with
the
need
to
periodically
check
the
performance
of
the
entire
monitoring
system.
EPA's
field
testing
has
shown
that
a
single
injection
performed
daily
using
the
appropriate
concentration
of
oxidized
Hg
is
sufficient
to
check
the
upscale
calibration
of
the
Hg
CEMS
from
the
probe
tip
through
the
analyzer
and
the
efficiency
of
the
HgCl2
to
Hg
converter.
The
final
rule
allows
either
elemental
or
oxidized
Hg
to
be
used
to
perform
daily
calibration
checks.

Comment:

Commenter
OAR­
2002­
0056­
2922
stated
that
in
section
60.4171(
c)(
2)
it
is
not
clear
whether
a
source
installing
a
new
Hg
monitoring
system
must
submit
a
Certification
Application.
In
many
instances,
the
Hg
monitoring
system
will
be
a
completely
separate
Hg/
diluent
system
that
uses
full
concentration,
dry
basis
analysis.
The
data
acquisition
system
and
other
inputs,
like
stack
flow
rate
will
be
associated
with
a
previously
certified
part
75
system.
It
is
not
clear
whether
such
a
system
would
require
a
certification
application
to
be
submitted.
It
is
also
not
clear
how
much
of
the
system
would
need
to
be
included
in
the
certification
application.

Response:

Section
60.4171(
c)(
2)
states
that
a
certification
application
is
not
required
if
the
system
has
been
previously
certified
under
the
Acid
Rain
Program
or
under
an
applicable
State
or
Federal
NOx
mass
emission
reduction
program
that
adopts
the
requirements
of
40
CFR
part
75,
subpart
H.
Therefore,
e.
g.,
if
a
flow
monitoring
system
was
previously
certified
under,
e.
g.,
the
Acid
Rain
Program
(
part
75),
a
certification
application
for
that
flow
monitoring
system
is
not
required.

The
final
rule
removes
all
requirements
for
a
Hg
emission
rate
(
or
Hg­
diluent)
monitoring
system.
However,
either
a
sorbent
trap
monitoring
system
or
a
Hg
concentration
monitoring
system
is
required
by
the
final
rule.
A
Hg
concentration
monitoring
system
consists
of
a
Hg
pollutant
concentration
monitor
and
an
automated
data
acquisition
and
handling
system.
Because
neither
a
Hg
concentration
monitoring
system
nor
a
sorbent
trap
monitoring
system
was
required
by
40
CFR
part
75
until
this
rulemaking,
a
certification
application
must
be
submitted
for
either
of
these
systems.

Comment:

Several
commenters
(
OAR­
2002­
0056­
1969,
­
1975,
­
2040,
­
2046,
­
2063,
­
2068,
­
2073,
­
2079,
­
2160,
­
2181,
­
2206,
­
2224,
­
2244,
­
2252,
­
2259,
­
2260,
­
2267,
­
2296,
­
2365,
­
2379,
­
2380,
­
2429,
­
2485,
­
2578,
­
2718,
­
2721,
­
2827,
­
2830,
­
2833,
­
2835,
­
2844,
­
2862,
­
2867,
­
2877,
­
2889,
­
2891,
­
2898,
­
2900,
­
2918,
­
2922,
­
2929,
­
2947,
­
3200,
­
3413,
­
3436,
­
3443,
­
3444,
­
3454,
­
3458,
­
3487,
­
3559,
­
3568)
were
in
general
agreement
on
the
following
points.
6­
37
Although
many
vendors
of
Hg
CEMS
have
recently
upgraded
their
instrument
systems
and
these
changes
should
eventually
improve
the
accuracy
and
reliability
of
Hg
CEMS
and
reduce
the
labor
needed
for
instrument
maintenance,
these
new
instrument
systems
have
not
been
tested
extensively
in
demonstration
programs.
Therefore,
the
ability
of
these
instrument
systems
to
achieve
the
proposed
relative
accuracy,
calibration
error,
and
calibration
precision
requirements
has
not
been
adequately
demonstrated.
Therefore,
EPA
does
not
yet
have
a
basis
or
data
to
guide
the
setting
of
specifications
for
calibration
error,
linearity,
or
relative
accuracy.
It
appears
that
the
proposed
performance
specifications
mirror
those
for
SO2
and
NOx
monitoring.
EPA
should
commit
to
collecting
data
and
evaluating
these
specifications
as
soon
as
calibration
gases
are
available,
so
that
the
specifications
can
be
adjusted
if
necessary,
prior
to
program
implementation.
EPA
should
require
operators
of
Hg
CEMS
to
conduct
procedures
that
include
but
are
not
necessarily
limited
to
daily
zero
and
span
audits,
quarterly
relative
accuracy
tests
and
three­
point
elemental
Hg
linearity
tests,
and
absolute
calibration
audits.
Analytically,
there
is
clearly
a
need
to
challenge
the
entire
system
often
with
a
form
of
oxidized
Hg.
This
Hg
chloride
(
HgCl2)
reference
gas
would
be
highly
desirable
to
check
integrity
of
the
sample
interface.
However,
further
research
needs
to
be
required
to
enable
the
development
of
an
accurate
oxidized
Hg
standard.
One
device,
the
Hovacal,
may
have
the
potential
of
delivering
known
concentrations
of
HgCl2.
EPA
should
recognize
and
accept
this
type
of
calibration
system
in
the
proposed
regulation.
There
are
concerns
with
the
proposed
RATA
process,
particularly
the
length
of
time
and
amount
of
money
that
may
be
required
to
comply
with
the
Hg
monitoring
requirements
on
an
annual
basis.
The
final
monitoring
requirements
must
be
technically
achievable
and
capable
of
measuring
Hg
emissions
with
precision,
reliability,
and
accuracy
in
a
cost­
effective
manner.
The
decision
to
report
Hg
concentration
on
dry
or
wet
basis
needs
more
consideration,
as
well
as,
the
evaluation
of
gaseous
interferences.
Lastly,
many
of
the
equations
and
calculations
are
incomplete
or
contain
errors
and
many
sections
need
further
clarification.

Response:

In
the
final
rule,
the
same
tests
are
required
for
initial
certification
and
on­
going
QA
of
Hg
CEMS
as
were
proposed
in
the
SNPR.
However,
note
the
following
changes
to
some
of
the
procedures
and
performance
specifications:


For
the
7­
day
calibration
error
test,
either
elemental
Hg
standards
or
a
NISTtraceable
source
of
oxidized
Hg
(
referred
to
as
"
HgCl2
standards"
in
the
SNPR)
may
be
used;


Quarterly
3­
level
"
system
integrity
checks"
(
which
were
called
"
converter
checks"
in
the
SNPR)
using
a
NIST­
traceable
source
of
oxidized
Hg
may
be
performed
in
lieu
of
the
quarterly
linearity
checks
with
elemental
Hg;


Daily
calibration
error
checks
may
be
performed
using
either
elemental
Hg
standards
or
a
NIST­
traceable
source
of
oxidized
Hg.
The
daily
performance
specification
has
been
made
the
same
as
for
the
7­
day
calibration
error
test;


The
monthly
converter
check
at
three
points
has
been
replaced
with
a
weekly
6­
38
system
integrity
check
at
a
single
point,
and
the
weekly
test
is
not
required
if
daily
calibrations
are
performed
with
a
NIST­
traceable
source
of
oxidized
Hg.


When
the
Ontario­
Hydro
Method
is
used,
paired
trains
are
required,
the
results
must
agree
within
10
percent
RD,
and
the
results
should
be
averaged.

Note
that
EPA
plans
to
analyze
RATA
data
from
Hg
monitors
and
may
initiate
a
future
rulemaking
to
adjust
the
relative
accuracy
performance
specifications
and
to
propose
a
performance­
based
RATA
incentive
system
similar
to
the
reduced
frequency
incentive
system
in
40
CFR
part
75
for
SO2,
NOx,
CO2,
and
flow
monitors.

EPA
disagrees
with
the
commenters
who
stated
that
there
is
no
data
available
to
justify
the
proposed
performance
specifications
for
Hg
monitors.
Such
data
have
been
collected
from
several
field
test
sites
and
for
several
different
types
of
Hg
concentration
monitors,
which
show
that
Hg
CEMS
can
meet
the
proposed
calibration
error
and
linearity
standards,
and
can
meet
a
20
percent
RA
standard.
Therefore,
except
for
the
daily
calibration
error
specification,
which
has
been
tightened
based
on
the
available
data,
the
final
rule
promulgates
the
proposed
calibration
error,
linearity
check,
and
RATA
performance
specifications,
as
proposed.

EPA
has
retained
the
requirement
to
check
the
converter
periodically
with
HgCl2
standards,
because
it
is
essential
to
ensure
that
all
of
the
vapor
phase
Hg
is
being
measured.
The
frequency
of
the
check
(
which
is
referred
to
as
a
"
system
integrity
check"
in
the
final
rule)
has
been
increased
from
monthly
to
weekly,
based
on
supportive
comments
to
check
the
entire
system
more
often,
but
the
requirement
to
perform
a
3­
point
check
has
been
reduced
to
a
singlepoint
test.
And
the
weekly
test
is
not
required
if
a
NIST­
traceable
oxidized
Hg
source
is
used
for
daily
calibrations.

There
are
several
different
devices
available
that
can
provide
oxidized
Hg,
including
the
HOVACAL
and
the
MerCAL.
The
HOVACAL
has
been
successfully
applied
in
the
laboratory
and
field
to
generate
and
deliver
known
concentrations
of
HgCl2
to
Hg
CEMS
to
achieve
the
requirements
of
the
40
CFR
part
75
system
integrity
check.
Moreover,
oxidized
Hg
gas
standards
such
as
are
produced
by
the
HOVACAL
and
MerCAL
are
currently
scheduled
to
be
independently
tested
by
NIST,
to
verify
their
suitability
as
reference
gas
standards.

Comment:

Commenter
OAR­
2002­
0056­
2922
stated
that
the
requirement
of
proposed
Appendix
A,
section
2.2.3
cannot
be
met.
The
commenter
is
not
aware
of
any
device,
other
than
the
Hovacal
that
might
be
able
to
deliver
"
known
concentrations
of
HgCl2"
as
required.
The
performance
of
the
Hovacal
does
not
appear
to
be
well
documented.

Response:

Several
different
sources
of
gaseous,
oxidized
Hg
are
available
and
have
been
6­
39
demonstrated,
incuding
the
HOVACAL.
The
HOVACAL
has
been
successfully
applied
in
the
laboratory
and
field
to
generate
and
deliver
known
concentrations
of
HgCl2
to
Hg
CEMS
to
achieve
the
requirements
of
the
converter
check
as
called
for
in
the
40
CFR
part
75
Hg
CEMS
precertification
requirement.
Moreover,
oxidized
Hg
gas
standards
such
as
the
HOVACAL
and
MerCAL
are
currently
scheduled
to
be
independently
tested
by
NIST
to
verify
their
suitability
as
reference
gas
standards.

Comment:

Commenter
OAR­
2002­
0056­
2922
stated
that
in
Appendix
A,
section
3.1,
the
specification
should
read
"
5%
of
the
span
value
if
the
span
value
is
20
micrograms/
dscm
or
greater,
or
1
microgram/
dscm
if
the
span
value
is
less
than
20
micrograms/
dscm."
The
specification
as
now
written
penalizes
any
monitor
with
a
span
value
between
10
and
20
micrograms/
dscm.

Response:

Appendix
A,
section
2.1.7.3
defines
span
for
Hg
CEMS
in
multiples
of
10
micrograms/
dscm.
Therefore,
there
will
be
no
Hg
span
values
between
10
and
20
micrograms/
dscm.
Appendix
A,
Section
3.1
is
retained,
as
proposed,
in
the
final
rule.

Comment:

Commenter
OAR­
2002­
0056­
2922
stated
that
Equation
F­
29
appears
to
make
the
calculation
correctly.
However,
the
description
in
section
9.1.2
is
difficult
to
understand.
A
complete
set
of
equations
and
nomenclature
should
be
provided
instead
of
the
cross
references
and
replacement
values.

Response:

Proposed
Equation
F­
29
applies
only
to
Hg­
diluent
monitoring
systems.
In
the
final
rule,
EPA
has
deleted
Equation
F­
29
and
all
provisions
related
to
this
type
of
monitoring
system.
The
final
rule
requires
Hg
mass
emissions
to
be
determined
as
the
product
of
the
Hg
concentration
and
the
stack
gas
flow
rate.

Comment:

Commenter
OAR­
2002­
0056­
2922
stated
that,
both
PS
12A
and
Method
324
are
designed
to
measure
vapor­
phase
Hg.
Consistent
with
that,
EPA
should
make
clear
in
the
40
CFR
part
75
RATA
requirements
(
as
EPA
did
in
PS
12A,
section
8.6.2)
that
the
filterable
portion
of
the
reference
method
sample
is
not
included
when
making
the
comparison
to
the
CEMS.
Consistent
with
that
change,
EPA
should
also
remove
the
requirements
in
sectioin
75.59(
a)(
7)
to
record
and
report
RATA
results
related
to
particle
bound
Hg,
or
justify
the
collection
and
6­
40
submission
of
that
additional
data.

Response:

The
final
rule
clearly
states
in
40
CFR
75.22
that
only
the
vapor
phase
Hg,
and
not
the
filterable
portion
of
the
reference
method
sample
is
included
when
making
the
comparison
to
the
CEMS.
However,
EPA
has
retained
in
section
75.59(
a)(
7)
the
recording
and
reporting
of
both
gaseous
and
particle­
bound
Hg,
when
particle­
bound
Hg
is
provided
by
the
reference
method.
The
particle­
bound
Hg
emissions
data
is
required
in
today's
final
rule
to
provide
information
to
assess
the
need
for
possible
future
regulation
of
particle­
bound
Hg.

6.5
MERCURY
MONITOR
AVAILABILITY
Comment:

One
commenter
(
OAR­
2002­
0056­
3560)
stated
that
no
correlation
has
been
established
between
the
basis
for
the
standards
and
the
required
test
methods.
Plants
that
submitted
Hg
data
to
the
EPA
used
the
Ontario­
Hydro
Method
to
calculate
their
Hg
emissions.
EPA
then
used
these
data,
based
on
the
Ontario­
Hydro
Method,
to
create
the
Hg
standards.
Whether
the
proposed
monitoring
systems
are
comparable
to
the
Ontario­
Hydro
Method,
which
was
used
to
set
the
Hg
emissions
limits
is
unknown.
If
they
are
not,
the
Hg
emissions
limits
could
result
in
a
standard
that
is
not
obtainable
from
the
outset.
This
would
leave
facilities
open
to
notices
of
violations/
penalties
when
their
failure
to
comply
is
a
result
of
an
inconsistency
between
analytical
methods
 
the
Ontario­
Hydro
Method
used
to
set
the
Hg
emissions
limits
versus
Hg
CEMS,
sorbent
trap
monitoring
systems,
and
long­
term
sampling
monitoring
that
are
to
be
employed
to
determine
if
units
are
meeting
the
emissions
limits
for
compliance
purposes.
Discerning
whether
the
test
methodology
utilized
to
create
the
Hg
emission
standards
(
Ontario­
Hydro
Method)
correlates
to
any
of
the
required
test
methods,
and
especially
when
Hg
CEMS
are
not
yet
commercially
available,
is
impossible.

Response:

EPA
disagrees
and
believes
Hg
CEMS
and
sorbent
trap
monitoring
systems
have
been
demonstrated
to
be
reasonably
comparable
to
the
Ontario­
Hydro
Method
through
various
field
demonstrations.

Comment:

One
commenter
(
OAR­
2002­
0056­
1854)
noted
that
the
analyzer
can
be
no
more
than
100
feet
from
the
sample
probe.
The
commenter
state
that
this
means,
in
their
case,
that
QA
and
maintenance
procedures
must
take
place
on
the
stack
platforms
where
personnel
will
be
exposed
to
extreme
weather
conditions
from
time
to
time.
The
commenter
also
stated
that
their
experience
would
also
indicate
that
additional
staffing
by
specially
trained
personnel
will
be
6­
41
required
because
these
monitors
cannot
be
left
unattended
for
long
periods
of
time.

Response:

Mercury
CEMS
demonstration
testing
conducted
by
EPA
so
far
has
provided
evidence
from
a
number
of
CEMS
at
a
number
of
different
utilities
that
the
analyzer
does
not
have
to
be
located
on
the
stack
platform.
Regarding
on­
going
system
maintenance,
the
level
required
will
ultimately
depend
on
the
monitoring
system
and
the
emission
characteristics.
We
have
noticed
a
marked
improvement
in
amount
of
time
needed
for
hands­
on
monitor
attendance
over
the
time
frame
of
demonstrations,
including
automation
of
daily
check
procedures
and
capability
for
remote
system
adjustment,
and
expect
significantly
more
improvement
in
the
coming
months.

Comment:

One
commenter
(
OAR­
2002­
0056­
3548)
stated
concerns
about
the
technical
feasibility
of
operating
and
maintaining
CEMS
for
Hg.
The
proposed
monitoring
requirements
are
beyond
the
capabilities
of
current
monitoring
equipment.
Although
the
technology
is
developing
rapidly,
the
proposed
technology
is
analogous
to
EPA
proposing
to
require
40
CFR
part
75
monitoring
requirements
for
flow,
SO2,
and
NOx
in
1970
with
a
compliance
deadline
of
1974.
This
being
the
case,
it
is
imperative
that
EPA
provide
as
much
flexibility
as
possible
in
allowable
monitoring
methods.

Response:

EPA
believes
that
Hg
CEMS
and
sorbent
trap
monitoring
systems
meet
the
proposed
monitoring
requirements.
The
final
rule
contains
flexibility
by
allowing
sources
to
account
for
their
Hg
emissions
by
using
Hg
CEMS,
sorbent
trap
monitoring
systems
(
or
a
combination
thereof),
and,
in
some
cases,
using
low
mass
provisions.

Comment:

One
commenter
(
OAR­
2002­
0056­
2915)
stated
concerns
about
the
technical
feasibility
of
operating
and
maintaining
CEMS
for
Hg.
The
commenter's
concerns
regard
the
precision,
reliability,
and
accuracy
of
the
monitoring
alternatives
identified
by
the
EPA
in
the
proposed
Hg
rule.
One
area
of
major
concern
involves
the
proposed
detection
limits.
The
commenter
asserts
that
as
the
allowable
Hg
emissions
levels
grow
smaller
(
particularly
in
the
case
of
new
or
well­
controlled
existing
units),
it
becomes
technically
more
difficult
to
measure
Hg
levels
in
the
emissions
from
an
electric
generating
unit
(
EGU)
and
to
determine
how
the
inherent
measurement
uncertainties
will
impact
an
EGU's
compliance
demonstration
with
the
Hg
limits.
Further,
should
CEMS
for
Hg
be
required
by
the
Hg
rule,
the
commenter
would
have
concerns
with
the
proposed
RATA
process,
particularly
the
length
of
time
and
amount
of
money
that
may
be
required
to
comply
with
the
Hg
monitoring
requirements
on
an
annual
basis.
The
commenter
asserts
that
the
final
monitoring
requirements
must
be
technically
achievable
and
capable
of
measuring
Hg
6­
42
emissions
with
precision,
reliability,
and
accuracy
in
a
cost­
effective
manner.

Response:

EPA
believes
that
Hg
CEMS
will
provide
adequate
precision,
reliability,
and
accuracy
for
emissions
trading,
however,
may
not
be
necessary
for
all
units.
Consistent
with
the
low
mass
emissions
(
LME)
provisions
for
SO2
and
NOx,
the
final
rule
provides
a
less
rigorous
monitoring
option
for
low
Hg
emitters.

Comment:

One
commenter
(
OAR­
2002­
0056­
2830)
stated
it
is
not
appropriate
to
require
CEMS
monitors
on
new
units
if
operations
begins
more
than
six
months
after
publication
of
the
final
rule.
CEMS
have
not
been
commercially
demonstrated,
and
as
commented
on
earlier,
the
commenter
believes
that
they
cannot
be
commercially
demonstrated
within
4
years
of
the
rule
being
finalized.

Response:

EPA
disagrees
with
commenter
and
believes
that
Hg
CEMS
and
sorbent
trap
monitoring
systems
can
be
commercially
demonstrated
before
rule
implementation.
However,
the
requirement
for
new
units
to
use
Hg
CEMS
has
been
withdrawn.
New
units
may
use
sorbent
trap
monitoring
systems
instead,
under
both
40
CFR
part
75,
subpart
I,
and
under
40
CFR
part
60,
subpart
Da.

Comment:

One
commenter
(
OAR­
2002­
0056­
2429)
requested
the
EPA
provide
additional
technical
information
on
both
CEMS
that
meet
PS12
requirements
and
sorbent
traps
that
meet
Method
324
requirements
for
various
plant
configurations
and
conditions
such
as
wet
stacks.
Their
units
are
equipped
with
wet
scrubbers,
fabric
filters,
and
low
NOx
burners.

Response:

EPA
will
provide
additional
information
when
available.

Comment:

One
commenter
(
OAR­
2002­
0056­
2899)
stated
that
currently
no
Hg
CEMS
have
been
demonstrated
to
be
accurate
and
reliable.
The
commenter
stated
that
although
continuous
systems
are
available
from
different
manufactures
none
have
been
used
in
continuous
operation
for
an
extended
period
of
time.
The
commenter
asserts
that
most
have
been
used
in
pilot
or
short
term
full
scale
tests
and
generally
compared
to
the
Ontario­
Hydro
impinger
method,
which
is
a
short
term
test
ranging
from
minutes
to
hours,
and
this
gives
only
snapshots
of
the
continuous
6­
43
monitors
performance.
The
commenter
noted
the
current
generation
of
continuous
monitors
are
based
on
wet
chemistry
requiring
almost
constant
maintenance
and
calibration.
The
commenter
stated
that
although
dry
chemistry
systems
are
under
development
and
testing
none
are
yet
commercially
viable.
According
to
the
commenter,
at
present
continuous
Hg
monitors
not
are
ready
to
be
used
for
continuous
compliance.
The
commenter
stated
that
the
EPA
should
allow
the
option
of
using
a
periodic
measurement
system
such
as
EPRI's
QuickSEM
or
other
system
instead
of
a
continuous
monitor
for
compliance
with
this
rule.

Response:

EPA
disagrees
with
the
commenter
and
believes
that
field
tests
have
demonstrated
Hg
CEMS
to
be
accurate
and
reliable.
The
Hg
CEMS
have
performed
adequately
for
several
months
and
meet
the
Ontario­
Hydro
Reference
Method
specifications.
Furthermore,
several
dry
chemistry
Hg
CEMS
are
currently
being
tested
at
sites
that
represent
the
most
challenging
conditions
and
the
Agency
plans
to
share
with
industry
the
results
of
such
experiences
to
facilitate
the
selection
of
appropriate
monitoring
methodology.
EPA
is
also
confident
that
substantial
advancement
of
Hg
CEMS
will
occur
before
the
implementation
of
the
rule
and
as
other
monitoring
techniques
may
become
available,
is
allowing
the
use
of
systems
that
can
meet
performance­
based
specifications.

Comment:

One
commenter
(
OAR­
2002­
0056­
5495)
stated
that
CEMS
for
measuring
low
Hg
emissions
from
waste
coal
is
not
a
proven
technology.

Response:

EPA
disagrees
with
the
commenter
and
believes
that
field
tests
have
demonstrated
Hg
CEMS
to
be
accurate
and
reliable
at
low
Hg
concentrations.
The
Hg
CEMS
have
performed
adequately
for
several
months
and
meet
the
Ontario­
Hydro
Reference
Method
RATA
specifications
at
two
low
Hg
concentration
(
0.5
­
2
µ
g/
dscm)
coal­
fired
sources.
EPA
is
also
confident
that
substantial
advancement
of
Hg
CEMS
will
occur
before
the
implementation
of
the
rule
and
as
other
monitoring
techniques
may
become
available,
is
allowing
the
use
of
systems
that
can
meet
performance­
based
specifications.

Comment:

One
commenter
(
OAR­
2002­
0056­
2830)
agreed
with
the
EPA
that
compliance
be
monitored
through
the
use
of
CEMS
or
other
continuous
measurement
methods
(
e.
g.,
sorbent
trap)
for
all
affected
sources.
However,
the
commenter
is
concerned
that
monitoring
and
recording
technology
has
not
evolved
to
the
level
of
reliability
necessary
to
collect
emissions
data
for
compliance
purposes.
As
entered
in
the
EPA
Docket,
the
EPA­
funded
project
termed
"
Long­
Term
Evaluation
of
Mercury
Continuous
Emission
Monitoring
Systems"
­
December
11,
6­
44
2003,
it
was
determined
that
the
current
capabilities
of
Hg
CEMS
are
not
at
a
level
that
can
be
relied
on
in
a
regulatory
compliance
environment.
In
almost
all
cases,
the
analyzers
failed
to
meet
the
20
percent
relative
accuracy
criteria
for
the
first
two
phases.
In
Phase
III,
there
were
some
instruments
that
passed
the
relative
accuracy
criteria.
The
Carbon
Bed
technology
was
only
tested
in
Phase
ill
of
the
program
and
did
exceptionally
well
as
compared
to
the
other
CEM
systems
and
against
the
Ontario­
Hydro
Method.
However,
there
are
concerns
that
the
facility
was
an
"
optimum
test
site."
A
substantial
test
for
Hg
analyzer
systems
should
come
under
test
conditions
that
are
strenuous
and
challenging
in
order
to
identify
shortcomings
of
the
systems.
The
test
should
be
completed
at
a
facility
where
the
sampling
location
has
a
representative
particulate
loading
and
sulfur
concentrations.
The
systems
should
also
be
challenged
under
wet
stack
conditions.

Response:

EPA
is
currently
field
testing
Hg
CEMS
and
sorbent
trap
monitoring
systems
under
strenuous
and
challenging
conditions
where
the
sampling
location
is
under
wet
stack
conditions
and
has
a
representative
particulate
loading
and
sulfur
concentration.
Preliminary
data
from
these
field
tests
seems
to
indicate
that
monitoring
and
recording
technology
are
getting
closer
to
the
level
of
reliability
necessary
to
collect
emissions
data
for
compliance
purposes,
supporting
the
Agency's
confidence
that
by
the
end
of
the
demonstration
technology
will
provide
the
answers
to
the
challenges
presented.

Comment:

One
commenter
(
OAR­
2002­
0056­
3469)
stated
that
lack
of
control
and
monitoring
technology
impedes
speedy
compliance.
The
EPA's
proposed
Hg
monitoring
technologies
(
e.
g.,
CEMS
12A
and
Method
324)
are
not
yet
commercially
available
and
do
not
yet
provide
accurate
data.
It
is
not
known
when
they
will
be
successfully
tested
and
commercially
available.

Response:

EPA
disagrees
with
commenter;
Hg
CEMS
and
sorbent
trap
monitoring
systems
are
commercially
available
and
provide
accurate
data,
as
demonstrated
by
various
completed
and
ongoing
field
tests.
In
addition,
EPA
believes
these
technologies
will
significantly
advance
before
compliance
is
necessary.

Comment:

Is
it
currently
feasible,
or
will
it
be
feasible
within
the
compliance
timeframes
of
the
proposed
rule,
to
accurately
monitor
a
source's
Hg
emissions
by
species?

Response:
6­
45
The
final
rule
requires
the
measurement
of
total
vapor
phase
Hg,
but
does
not
require
separate
monitoring
of
speciated
Hg
emissions
(
i.
e.,
elemental
and
ionized
Hg).
Because
of
the
potential
impact
of
Hg
speciation
on
local
versus
broader
geographical
deposition,
the
Agency
considers
separate
monitoring
of
these
emissions
as
a
need
to
be
addressed.
However,
at
least
two
current
monitoring
technologies
can
accurately
monitor
speciated
Hg
emissions.
The
Agency
will
continue
to
test
speciated
Hg
monitoring
technologies.
If
these
technologies
are
adequately
demonstrated,
the
Agency
may
consider
a
proposed
rulemaking
within
four
to
five
years
after
program
implementation,
which
should
provide
enough
lead
time
for
development
and
installation
of
these
monitoring
systems.

6.6
MERCURY
DILUENT
SYSTEMS
Comment:

One
commenter
(
OAR­
2002­
0056­
3455),
in
reference
to
section
72.2
­
Definitions,
page
12453,
asked
why
the
proposed
Hg
emissions
units
of
measurement
are
the
same
as
NOx
­
diluent?
The
Hg
concentration
measurements
are
orders
of
magnitude
below
NOx
emissions,
thus
applying
a
diluent
correction
with
the
additional
uncertainties
of
measurement
further
complicates
the
direct
emissions
reporting
uncertainties.
Mercury
is
a
resident
pollutant
in
the
fuel,
it
can
be
measured,
and
measurement
should
parallel
the
same
regulation
requirements
as
SO2.

Response:

The
final
rule
removes
all
mention
of
Hg­
diluent
monitoring
systems
and
requires
the
hourly
Hg
mass
emissions
to
be
calculated
in
the
same
manner
as
is
done
for
SO2
under
the
Acid
Rain
Program,
i.
e.,
as
the
product
of
the
Hg
concentration
and
the
stack
gas
flow
rate.
The
final
rule
also
better
accommodates
Hg
analyzers
that
measure
on
a
wet
basis.

EPA
believes
that
the
rule
can
be
considerably
simplified
and
shortened
without
losing
any
flexibility
by
deleting
the
provisions
related
to
Hg­
diluent
monitoring
systems
and
allowing
only
Hg
concentration
monitoring
systems
and
sorbent
trap
systems
to
be
used.
Therefore,
the
final
rule
removes
all
mention
of
Hg­
diluent
monitoring
systems
and
requires
the
hourly
Hg
mass
emissions
to
be
calculated
in
the
same
manner
as
is
done
for
SO2,
i.
e.,
as
the
product
of
the
Hg
concentration
and
the
stack
gas
flow
rate.

6.7
LOW
EMITTING
UNITS
Comment:

One
commenter
(
OAR­
2002­
0056­
2900)
questioned
the
need
for
continuous
monitoring
if
the
EPA
requires
coal­
fired
EGUs
to
meet
MACT
standards.
The
commenter
urges
the
Agency
to
allow
periodic
monitoring
and
parametric
monitoring
approaches
rather
than
restricting
sources
to
the
use
of
Hg
CEMS
or
sorbent
trap
monitoring.
6­
46
Response:

The
Agency
has
decided
to
use
a
cap­
and­
trade
program
to
control
Hg
emissions.
Complete
and
accurate
accounting
of
Hg
emissions
is
required
for
a
credible
cap
and
trade
program.
Therefore,
Hg
CEMS
or
sorbent
traps
are
required
in
the
final
rule.
However,
qualifying,
low
emitting
sources
may
comply
with
today's
monitoring
requirements
by
using
conservative
default
Hg
emission
factors
and
annual
or
semi­
annual
stack
testing.

Comment:

Numerous
commenters
(
OAR­
2002­
0056­
2101,
­
2162,
­
2267,
­
2634,
­
2718,
­
2861,
­
2867,
­
2900­
2918,
­
2922,
­
3432,
­
3509,
­
3513,
­
3565,
­
2855)
requested
that
EPA
provide
a
less
rigorous,
cost­
effective
monitoring
option
for
low
emitting
units.
Affected
units
could
meet
a
low­
emitter
criterion
based
on
a
combination
of
unit
size,
operating
time,
and/
or
control
device
operation.
Any
marginal
decrease
in
accuracy
from
less
rigorous
monitoring
would
have
a
minimal
impact
overall,
because
these
units
represent
only
a
small
percentage
of
the
nationwide
Hg
mass
emissions.

Response:

Consistent
with
the
low
mass
emissions
(
LME)
provisions
in
section
75.19
for
SO2
and
NOx,
sections
75.81(
b)
through
(
g)
of
the
final
rule
provide
a
less
rigorous
monitoring
option
for
low
Hg
emitters.
These
provisions
allow
sources
with
estimated
annual
emissions
of
29
lb/
yr
(
464
oz/
yr)
or
less,
representing
about
5
percent
of
the
nationwide
Hg
mass
emissions,
to
use
periodic
emission
testing
to
quantify
their
Hg
emissions,
rather
than
continuously
monitoring
the
Hg
concentration.
For
units
with
Hg
emissions
of
9
lb/
yr
(
144
oz/
yr)
or
less,
annual
emission
testing
is
required.
For
units
with
Hg
emissions
greater
than
144
oz/
yr
but
less
than
or
equal
to
464
oz/
yr,
semiannual
testing
is
required.
For
reporting
purposes,
the
owner
or
operator
is
required
to
use
either
the
highest
Hg
concentration
from
the
most
recent
emission
testing
or
0.50
µ
g/
scm,
whichever
is
greater.
If,
at
the
end
of
a
particular
calendar
year,
the
reported
annual
Hg
mass
emissions
for
a
unit
exceed
464
ounces,
the
unit
is
disqualified
as
a
low
mass
emitter
and
the
owner
or
operator
must
install
and
certify
a
Hg
CEMS
or
sorbent
trap
monitoring
system
within180
days
of
the
end
of
that
year.
The
final
rule
also
contains
special
low
mass
emitter
provisions
for
common
stack
and
multiple
stack
exhaust
configurations.

The
Agency
believes
that
a
low
mass
emitter
provision
can
be
beneficial
to
both
EPA
and
industry.
It
is
cost­
effective
for
industry,
in
that
it
allows
periodic
stack
testing
to
be
used
to
estimate
Hg
emissions
instead
of
requiring
continuous
emission
monitoring.
In
the
context
of
a
cap
and
trade
program,
a
low
emitter
provision
can
provide
environmental
benefit,
because
it
requires
conservatively
high
default
emission
factors
to
be
used
for
reporting,
thereby
removing
Hg
allowances
from
circulation.
Also,
allowing
a
subset
of
the
affected
units
to
use
less
rigorous
monitoring
reduces
the
administrative
burden
of
program
implementation,
allowing
EPA
to
focus
its
attention
on
the
higher­
emitting
sources.
6­
47
Selecting
an
appropriate
low
emitter
cutoff
point
is
of
critical
importance.
On
the
one
hand,
if
the
cutoff
point
is
too
low
(
i.
e.,
too
exclusive)
this
would
not
be
cost­
effective
for
the
regulated
sources
and
would
greatly
increase
the
burden
on
the
regulatory
agencies
to
implement
and
maintain
the
program.
On
the
other
hand,
if
the
cutoff
point
is
too
high
(
i.
e.,
too
inclusive),
this
would
create
inequities
in
the
trading
market.

Over
the
years,
EPA
has
used
a
de
minimis
concept
to
either
exempt
low­
emitting
sources
from
monitoring
or
to
allow
these
sources
to
use
less
rigorous,
lower
cost
techniques
to
monitor
emissions
instead
of
installing
CEMS:


In
the
preamble
of
the
1993
Acid
Rain
Program
(
ARP)
final
rule
(
see
58
FR
3593,
January
11,
1993),
EPA's
Acid
Rain
Division
(
now
the
Clean
Air
Markets
Division)
first
used
the
de
minimis
concept
to
exempt
certain
new
utility
units
from
the
Acid
Rain
Program
(
i.
e.,
units
less
than
or
equal
to
25
MW
that
burn
only
fuels
with
a
sulfur
content
less
than
or
equal
to
0.05
percent
by
weight);


EPA
also
allows
gas­
fired
and
oil­
fired
peaking
units
to
use
the
less
costly
methodology
in
40
CFR
part
74,
appendix
E,
to
estimate
NOx
emissions
instead
of
using
CEMS,
because
the
Agency's
analyses
indicated
that
projected
NOx
emissions
from
these
units
represent
less
than
1
percent
of
the
total
NOx
emissions
from
Acid
Rain
Program
units.


In
1998,
EPA
promulgated
LME
provisions
in
section
75.19
for
SO2
and
NOx
(
see
63
FR
57484,
October
27,
1998).
These
provisions
require
the
use
of
conservatively
high
default
emission
rates
to
quantify
SO2
and
NOx
emissions.
EPA
determined
the
appropriate
SO2
and
NOx
mass
emissions
thresholds
or
"
cutoff
points"
for
unit
to
qualify
as
a
low
mass
emissions
methodology,
considering
inventory
and
regulatory
changes
that
had
taken
place
since
the
original1993
Acid
Rain
rulemaking.
The
selected
threshold
values
were
based
on
a
de
minimis
concept,
i.
e.,
the
SO2
and
NOx
emissions
from
the
units
that
could
potentially
qualify
to
use
the
LME
methodology
represented
less
than
or
equal
to
1
percent
of
the
emissions
from
all
affected
units.

In
1999,
EPA
obtained
Hg
mass
emissions
estimates
for
the
1,120
utility
units
affected
by
the
SNPR,
as
the
result
of
an
information
collection
request
(
ICR)
that
appeared
in
the
Federal
Register
on
April
9,
1998.
These
data
show
that
if
a
low
Hg
mass
emission
threshold
of
9
lb/
yr
were
selected,
228
units,
representing
1
percent
of
the
total
annual
Hg
emissions
from
coal­
fired
electric
utility
units
in
the
U.
S.,
could
potentially
qualify
to
use
the
low
emitter
option.
However,
EPA's
analysis
also
indicated
that
by
raising
the
cutoff
point
to
29
lb/
yr,
almost
twice
the
number
of
units
(
435),
representing
just
5
percent
of
the
total
annual
Hg
emissions,
could
potentially
qualify
as
low
emitters.
Therefore,
EPA
has
decided
to
adopt
the
29
lb/
yr
as
the
qualifying
low
mass
emission
threshold
for
Hg.

Although
the
5
percent
threshold
represents
a
departure
from
the
traditional
de
minimis
value
of
1
percent,
the
Agency
believes
that
allowing
units
with
Hg
emissions
of
29
lbs/
yr
or
less
6­
48
to
use
the
low
mass
emitter
option
is
a
better
choice,
for
both
economic
and
environmental
reasons.
For
continuous
monitoring
methodologies,
the
annualized
cost
per
unit
will
be
about
$
89,500
for
testing,
maintenance,
and
operation.
For
sorbent
trap
methodologies,
the
annualized
cost
per
unit
will
be
about
$
113,000
for
testing,
maintenance,
and
operation.
For
a
unit
that
emits
between
9
lb/
yr
and
29
lb/
yr
of
Hg,
if
the
owner
or
operator
elects
to
use
the
low
emitter
option,
today's
rule
would
require
two
stack
tests
per
year
(
at
$
5,500
each),
and
an
estimated
$
1,500
annual
cost
for
technical
calculation,
labor,
and
other
associated
costs,
for
a
total
annual
expenditure
per
unit
of
around
$
12,500.
Therefore,
for
the
approximately
207
units
with
Hg
mass
emissions
between
9
and
29
lb/
yr,
the
potential
savings
associated
with
the
implementation
of
the
low
emitter
option
could
be
as
high
as:
$
89,500
­
$
12,500
=
$
77,000
*
207
units
=
$
15,939,000
per
year
if
LME
is
used
instead
of
Hg
CEMS.
Alternatively,
if
LME
is
used
instead
of
sorbent
traps,
the
potential
savings
could
be
even
higher:
$
113,000
­
$
12,500
=
$
100,500
*
207
units
=
$
20,803,500
per
year.
This
is
achieved
without
losing
the
environmental
integrity
of
the
program
or
compromising
the
cap,
because
the
default
Hg
concentration
values
used
for
reporting
are
conservatively
high,
and
for
units
with
flue
gas
desulfurization
(
FGD)
systems
or
add­
on
Hg
emission
controls,
the
rule
requires
the
maximum
potential
concentration
(
MPC)
to
be
reported
when
the
controls
are
not
operating
properly.

As
a
further
justification
of
the
5
percent
low
emitter
threshold
for
Hg,
EPA
notes
that
there
are
two
important
differences
between
the
Hg
low
mass
emission
provisions
in
section
75.81
and
the
LME
provisions
in
section
75.19
for
SO2
and
NOx
(
which
are
based
on
a
1
percent
threshold).
First,
under
section
75.19,
default
emission
rates
are
used
exclusively,
and
there
is
no
real­
time
continuous
monitoring
of
the
SO2
or
NOx
emissions.
However,
under
section
75.81,
the
stack
gas
volumetric
flow
rate,
which
is
used
in
the
hourly
Hg
mass
emission
calculations,
is
continuously
monitored.
Second,
the
LME
provisions
in
section
75.19
allow
you
to
either
use
generic
default
NOx
emission
rates
without
performing
any
emission
testing,
or,
if
you
test
for
NOx,
you
are
only
required
to
determine
a
new
default
emission
rate
once
every
5
years.
Under
section
75.81,
emission
testing
is
required
initially
to
qualify
as
a
low
emitter,
and
retesting
is
required
either
semiannually
or
annually
thereafter,
depending
on
the
annual
emission
level.

6.8
RECORDKEEPING/
REPORTING
REQUIREMENTS
Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
Appendix
A
to
the
Preamble­­
Proposed
Changes
to
Parts
72
and
75,
(
Proposed
Rules
March
16,
2004);
page
12420
(
Calculation
of
Mercury
Mass
Emissions)
all
of
the
calculations
for
both
CEMS
and
sorbent
methods
contain
a
significant
error.
Rounding
procedures
for
equation
F­
28
(
hourly
emissions)
require
rounding
of
hourly
emissions,
in
ounces,
to
one
decimal
place.
This
would
result
in
all
but
the
largest
sources
registering
as
0.0
oz/
hr.
Final
rounding
for
reporting
purposes
may
be
to
two
or
three
significant
figures.
Intermediate
results
that
are
used
in
subsequent
calculations
should
not
be
rounded
to
minimize
propagation
of
numerical
errors.
For
example,
if
hourly
emissions
for
a
particular
unit
were
constant
and
calculated
as
0.35
oz/
hr,
rounding
to
0.40
6­
49
oz/
hr
would
result
in
excess
emissions
of
438
oz/
year.
Introducing
artificial
arithmetic
artefacts
into
the
emission
calculations
of
a
high
value
pollutant
such
as
Hg
is
unwise
and
unnecessary.

Response:

EPA
agrees
with
the
commenter
and
has
required
in
the
final
rule
that
equation
F­
28
results
and
the
recordkeeping/
reporting
of
hourly
Hg
mass
emissions
be
rounded
to
the
nearest
thousandth
of
an
ounce.
EPA's
40
CFR
part
75
policy
is
that
intermediate
results
that
are
not
reported,
but
are
used
in
subsequent
hourly
calculations
should
not
be
rounded
to
minimize
propagation
of
numerical
errors.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
stated
that,
in
reference
to
Appendix
F
to
the
Preamble­­
Proposed
Changes
to
Parts
75,
(
Proposed
Rules
March
16,
2004);
page
12472,
the
calculation
procedure
in
Appendix
F
to
Part
75
 
Conversion
Procedures
contains
serious
errors.
In
particular,
equation
F­
28
is
erroneous
and
does
not
yield
correct
results.
The
commenter
noted
that
the
Hg
concentrations
used
for
mass
determination
are
not
referenced
to
any
specific
CO2
or
O2
levels.
This
provides
further
support
for
the
position
that
reporting
on
this
basis
is
superfluous.

Response:

EPA
has
corrected
equation
F­
28
in
the
final
rule.
The
final
rule
does
not
reference
Hg
concentrations
to
any
specific
CO2
or
O2
levels.

Comment:

One
commenter
(
OAR­
2002­
0056­
3455)
noted
that,
in
reference
to
Appendix
F
to
the
Preamble­­
Proposed
Changes
to
Parts
75,
(
Proposed
Rules
March
16,
2004);
page
12472,
Equation
F­
28
contradicts
Equation
3
(
FR
page
4724)
by
requiring
hourly
averaging
of
the
Hg
concentrations
and
hourly
totals
of
the
stack
flow
before
performing
a
calculation.
Significant
flow
and
concentrations
can
take
place
over
the
course
of
an
hour,
resulting
in
an
erroneous
calculation.
This
form
of
calculation
may
be
appropriate
for
the
sorbent
method.
However,
for
CEMS,
a
more
accurate
and
unbiased
calculation
is
available
by
multiplying
concentration
and
volumetric
readings
together
at
the
data
rate
of
the
CEMS
and
then
summing
to
produce
hourly
emissions.

Response:

EPA
understands
the
concern
of
the
commenter.
However,
because
one
hour
periods
are
the
basic
Hg
emissions
reporting
increment,
EPA
requires
sources
to
calculate
their
quarterly
and
annual
Hg
mass
emissions
using
hourly
quantities.
This
allows
EPA
to
recalculate
6­
50
quarterly
and
annual
Hg
mass
emissions
using
the
same
reported
hourly
numbers
that
sources
use.
If
sources
used
sub­
hourly
increments
to
calculate
mass
emissions,
EPA's
recalculated
mass
emissions
based
on
reported
hourly
values
would
not
agree
with
the
reported
numbers.

Comment:

One
commenter
(
OAR­
2002­
0056­
2889),
although
not
supporting
Hg
trading,
did
support
using
the
electronic
reporting,
under
the
acid
rain
program
emission
reporting
system,
set
up
for
SO2
and
NOx
for
Hg
emissions
reporting.
This
would
consolidate
the
emission
reporting
requirements
under
the
NSPS
and
Acid
Rain
Program,
which
regulate
the
same
facilities.
The
addition
of
Hg
to
those
pollutants
already
reported
to
the
Clean
Air
Markets
Division
would
benefit
EPA,
States,
public,
the
industry.

Response:

The
final
rule
requires
the
same
type
of
electronic
reporting
that
is
used
under
the
Acid
Rain
and
NOx
Budget
Programs.

6.9
OTHER
Comment:

One
commenter
(
OAR­
2002­
0056­
3200)
supports
EPA's
rolling
12­
month
averaging
calculation
for
compliance
determinations.
Mercury
is
not
an
acute
health
hazard
and
concerns
arise
from
long­
term
chronic
exposure.
Thus,
Hg
control
lends
itself
well
to
a
compliance
program
with
long­
term
averaging.

Response:

The
Agency
has
decided
to
use
a
cap­
and­
trade
approach
to
control
Hg
emissions.
However,
note
that
the
proposed
NSPS
for
Hg
has
been
finalized
under
40
CFR
60,
subpart
Da,
and
compliance
with
the
Hg
emission
limit
in
40
CFR
60.45a
is
determined
on
a
12­
month
rolling
average
basis.

Comment:

Two
commenters
(
OAR­
2002­
0056­
2068,
­
2422)
stated
that
EPA's
proposed
methods
for
measuring
Hg
emissions
from
coal
fueled
power
plants
must
address
the
detection
limits
of
those
methods
and
how
those
detection
limits
will
impact
compliance
demonstrations
with
the
new
source
MACT
limits.
The
EPA
must
take
into
account
the
ability
of
existing
technology
to
detect
Hg
emissions
at
such
minute
levels,
and
must
also
discuss
the
range
of
test
results
which
may
be
allowable
over
time.
Without
such
clarification,
units
will
be
unable
to
reliably
determine
if
their
test
results
are
within
an
acceptable
range
of
compliance,
or
if
they
violate
the
limits.
One
6­
51
commenter
(
OAR­
2002­
0056­
2422)
adds
that
regardless
of
the
monitoring
alternatives
specified
by
EPA,
the
final
rule
must
address
the
detection
limits
of
that
testing
methods).
As
the
allowable
Hg
emission
levels
grow
smaller,
it
becomes
scientifically
more
difficult
to
assess
whether
the
emission
limits
are
being
met.

Response:

In
field
tests
of
Hg
monitoring
methodologies,
EPA
has
observed
that
low
concentrations
of
Hg
can
be
detected
and
accurately
measured
with
the
currently­
available
monitoring
systems.
Mercury
compliance
will
be
determined
under
a
cap­
and­
trade
program,
rather
than
MACT.

Comment:

One
commenter
(
OAR­
2002­
0056­
3439)
stated
that
the
proposed
supplemental
rule
(
dated
March
16,
2004)
defines
the
minimum
acceptable
test
run
duration
for
the
reference
test
method
(
Ontario­
Hydro)
as
2
hours.
Given
the
very
low
concentration
of
Hg
in
the
gas
stream,
the
commenter
recommended
a
test
run
of
longer
duration
or
sufficient
analysis
of
available
test
data
be
conducted
to
provide
adequate
confidence
in
the
2
hour
minimum
test
run
duration.

Response:

EPA
has
data
to
indicate
that
2
hours
is
sufficient
for
application
to
controlled
sources.
However,
in
the
final
rule,
the
2­
hour
minimum
RATA
run
length
provision
has
been
withdrawn.
For
the
RATA
of
a
Hg
CEMS
using
the
Ontario­
Hydro
method,
or
for
the
RATA
of
a
sorbent
trap
system
(
irrespective
of
the
reference
method
used),
the
final
rule
simply
specifies
that
the
minimum
time
per
run
must
be
long
enough
to
collect
a
sufficient
mass
of
Hg
to
analyze.

Comment:

One
commenter
(
OAR­
2002­
0056­
2063)
recommended
with
regard
to
QA/
QC
requirements
that
annual
RATA
tests
using
a
Reference
Method,
either
wet
chemistry­
based
or
Method
324
(
paired
trains)
are
appropriate
for
either
a
CEM
or
Method
324.

Response:

EPA
agrees
with
the
commenter
with
the
exception
that
we
do
not
consider
that
sorbent
trap
monitoring
procedure
has
the
quality
to
serve
as
a
reference
method.
We
are,
however,
confident
of
its
performance
as
a
monitoring
method
as
long
as
a
yearly
RATA
is
performed
and
passed
and
the
performance
criteria
in
Appendix
K
are
met.

Comment:

One
commenter
(
OAR­
2002­
0056­
2883)
believed
that
the
EPA
should
hold
workshops
to
6­
52
assist
both
purchasers,
States,
and
vendors
on
various
monitoring
systems.

Response:

EPA
agrees
with
the
commenter
and
will
provide
workshops
to
assist
both
purchasers,
States,
and
vendors
on
any
issues
related
to
the
final
rule.

Comment:

One
commenter
(
OAR­
2002­
0056­
2889)
stated
that
the
proposed
PS­
12A
only
accounts
for
vapor
phase
Hg
emissions.
This
allows
facilities
to
ignore
particulate­
bound
Hg
emissions
when
using
a
CEMS.
The
State
of
Massachusetts
provided
test
data
to
showed
that
particulatebound
Hg
can
constitute
as
much
as
40
percent
of
total
Hg.
The
commenter
urged
EPA
to
require
total
Hg
emissions
as
the
basis
of
compliance
demonstrations
per
their
state
rules.

Response:

Nationwide,
97
percent
of
Hg
emissions
from
the
outlets
of
coal­
fired
utilities
are
vapor
phase.
Currently
there
is
no
established
technology,
other
than
the
Ontario­
Hydro
Method
to
measure
particulate­
bound
Hg.
Therefore,
today's
final
rule
only
regulates
vapor
phase
Hg
emissions.

Comment:

One
commenter
(
OAR­
2002­
0056­
2867)
stated
that
for
applications
where
Hg
CEMS
are
used
and
for
Hg
monitoring
and
audits
to
be
advanced,
the
commenter
submits
a
real
time
instrumental
reference
method
for
RA
audits
is
needed.
The
commenter
stated
that
EPA
must
develop
such
an
instrumental
reference
method
for
Hg.

Response:

EPA
agrees
with
the
commenter.
Based
on
field
testing,
EPA
intends
to
develop
an
instrumental
reference
method
for
Hg.
Initial
evaluations
of
such
a
method
have
already
begun.

Comment:

One
commenter
(
OAR­
2002­
0056­
1842)
stated
that
monitoring
Hg
to
the
degree
of
accuracy
required
for
trading
($
35,000/
lb)
is
a
daunting
task.
Add
to
this
the
fact
that
EPA
has
not
addressed
particulate
Hg
and
suggested
just
ignoring
this
quantity.
The
rationale
is
that
particulate
Hg
is
only
three
percent
of
the
total.
At
$
35,000/
lb
the
ignored
quantity
translates
into
$
105,000
for
a
300
MW
boiler,
$
1,050,000/
yr
for
a
3,000
MW
plant
and
$
105,000,000
for
the
whole
industry.
Mercury
measurement
is
money
measurement.
This
number
may
even
be
larger.
Fine
particulate
emissions
from
power
plants
are
not
measured.
The
commenter
believed
6­
53
emissions
are
much
larger
than
EPA
estimates.
Therefore
particulate
Hg
emissions
from
plants
with
old
inefficient
precipitators
are
likely
to
be
much
higher
than
emissions
from
the
average
plant.
Also
some
Hg
control
technologies
create
particulate
Hg.
The
RJM
concept
is
to
condense
acid
mist
on
fine
particulate
in
order
to
create
acid
deposition
sites.
Under
the
EPA
scheme
all
the
Hg
could
then
be
discharged
and
not
counted.
Also
there
is
a
10
times
differential
between
fine
particulate
emissions
from
old
precipitators
and
new
ones.
This
means
it
will
be
necessary
to
measure
particulate
Hg.
Periodic
Method
5
sampling
will
result
in
filter
catches
which
can
then
be
analyzed
for
particulate
Hg.
The
accuracy
will
be
a
function
of
the
sampling
interval
and
the
variations
from
sample
to
sample.
A
few
stack
tests
per
year
may
establish
particulate
Hg
plus
or
minus
50
percent.
A
stack
test
each
week
would
reduce
the
inaccuracy
to
some
smaller
range
e.
g.,
10
percent.
However
continuous
method
5
sampling
would
improve
accuracy
even
more.
Annual
Hg
quantities
then
will
be
the
integration
of
the
results
from
sorbent
traps,
Hg
CEMS,
method
5
tests
for
particulate
Hg
and
material
balances
involving
the
amounts
in
the
coal,
ash,
fly
ash,
and
wastewater.

Response:

Nationwide,
97
percent
of
Hg
emissions
from
the
outlets
of
coal­
fired
utilities
are
vapor
phase.
Currently
there
is
no
established
technology,
other
than
the
Ontario­
Hydro
Method
to
measure
particulate­
bound
Hg.
Therefore,
the
final
rule
only
regulates
vapor­
phase
Hg
emissions.
EPA
will
collect
the
particulate
Hg
concentrations
from
any
source
using
a
reference
method
capable
of
providing
particulate
Hg
concentrations
and
may
initiate
a
future
rulemaking
to
require
that
these
emissions
be
monitored
and
reported.

Comment:

Several
commenters
(
OAR­
2002­
0056­
2634,
­
2718,
­
2861,
­
2922,
­
3565)
noted
that
proposed
40
CFR
75
subpart
Da
section
60.50a(
j)
would
require
Hg
CEMS
and
sorbent
trap
systems
to
perform
some
QA/
QC
requirements
"
in
accordance
with"
Procedure
1
of
40
CFR
part
60,
Appendix
F.
This
presents
some
issues
for
sorbent
trap
systems
because
Procedure
1
does
not
include
all
of
the
information
necessary
to
perform
those
tests.
In
addition,
with
the
use
of
40
CFR
part
60
QA/
QC
requirements
in
subpart
Da,
new
units
that
are
subject
to
both
the
NSPS
and
the
cap­
and­
trade
program
would
be
subject
to
both
the
specified
40
CFR
part
60,
Appendix
F,
and
the
40
CFR
part
75,
Appendix
B,
QA/
QC
requirements.
EPA
should
avoid
imposing
these
duplicative
and
inconsistent
requirements
by
explicitly
stating
in
the
subpart
Da
revision
that
Hg
CEMS
and
sorbent
trap
systems
meeting
the
requirements
of
40
CFR
part
75
do
not
have
to
comply
with
40
CFR
part
60,
Appendix
F,
procedures
set
out
in
40
CFR
60.50a(
j).
EPA
should
use
the
subpart
Da
NOx
revision
in
40
CFR
60.47(
c)(
2)
as
a
model.

One
commenter
(
OAR­
2002­
0056­
3469)
stated
that
to
minimize
duplication
and
additional
costs,
the
EPA
should
adopt
monitoring
and
reporting
standards
that
are
consistent
with
those
for
other
air
quality
programs.
6­
54
Response:

In
the
final
rule
for
Hg
(
40
CFR
60,
subpart
Da),
EPA
has
made
it
clear
that
the
provisions
of
40
CFR
part
60,
Appendix
F
do
not
apply
to
sorbent
trap
monitoring
systems.
Rather,
sorbent
trap
monitoring
systems
must
meet
the
applicable
QA
requirements
in
Appendices
B
and
K
of
40
CFR
part
75.
An
annual
RATA
of
each
sorbent
trap
system
is
required,
in
addition
to
a
number
of
system­
specific
QA
tests
and
procedures.
For
the
qualityassurance
of
data
from
Hg
CEMS,
the
NSPS
clearly
states
that
sources
subject
to
both
40
CFR
60,
subpart
Da
and
40
CFR
part
75,
subpart
I
may
implement
the
QA
procedures
in
40
CFR
part
75,
Appendix
B
in
lieu
of
the
procedures
in
40
CFR
part
60,
Appendix
F.
