UNITED
STATES
ENVIRONMENTAL
PROTECTION
AGENCY
WASHINGTON,
D.
C.
20460
3/
22/
05
OFFICE
OF
AIR
AND
RADIATION
DRAFT
ATTACHMENT
for
Unit
GTP1
Mr.
Phillip
Polyak
Designated
Representative
Dearborn
Industrial
Generation
P.
O.
Box
126
Dearborn,
MI
48121­
0126
Re:
Consolidated
Requirements
for
a
NO
x
Predictive
Emissions
Monitoring
System
for
a
Combustion
Turbine
(
Unit
GTP1)
at
Dearborn
Industrial
Generation
(
Facility
ID
(
ORISPL)
55088)

Dear
Mr.
Polyak:

To
allow
for
easier
implementation,
this
Attachment
consolidates
the
requirements
in
EPA's
Additional
Quality
Assurance
Requirements
for
Conditionally­
Approved
Predictive
Emissions
Monitoring
System
at
Dearborn
Industrial
Generation
(
Facility
ID
(
ORISPL)
55088),
Unit
GTP1,
and
EPA's
September
2,
2003
conditional
approval.

BACKGROUND
On
October
24,
2002,
DIG
petitioned
for
approval
of
a
CMC
Solutions'
Smart­
75
©
PEMS,
which
is
a
hybrid
statistical­
based
computer
software
system
that
utilizes
turbine
sensor
inputs
to
produce
outputs
of
estimated
nitrogen
oxides
(
NO
x)
and
carbon
dioxide
(
CO
2)
emissions.
The
PEMS
is
installed
on
a
170
MW
GE
Frame
7FA,
simple
cycle
combustion
turbine
(
Unit
GTP1)
at
the
DIG
plant
in
Dearborn,
Michigan.
Unit
GTP1
was
installed
in
1999
to
generate
electricity
exclusively
for
commercial
resale.
The
unit
combusts
only
pipeline
natural
gas
and
uses
dry
low­
NO
x
combustion
technology
to
control
NO
x
emissions.
Unit
GTP1
is
subject
to
the
Acid
Rain
Program
regulations,
and
currently
qualifies
as
a
peaking
unit
(
as
defined
in
§
72.2).
According
to
the
Michigan
Department
of
Environmental
Quality,
Unit
GTP1
is
also
subject
to
the
NO
x
Budget
Trading
Program,
under
NO
x
Rules
801­
818
(
also
referred
to
as
R336.1801
­
R336.1818).
NO
x
Rules
801­
818
require
DIG
to
begin
monitoring
and
reporting
NO
x
mass
emissions
and
heat
input
for
Unit
GTP1
in
accordance
with
Subpart
H
of
40
CFR
Part
75,
by
May
1,
2003.

To
meet
the
NO
x
monitoring
requirements
of
the
Acid
Rain
Program,
DIG
elected
to
implement
Part
75,
Appendix
E,
which
applies
exclusively
to
gas­
fired
and
oil­
fired
peaking
units.
The
Smart­
75
©
software
was
installed
on
the
turbine
in
1999
and,
since
that
time,
has
been
functioning
as
a
data
acquisition
and
handling
system
(
DAHS)
which
satisfies
Appendix
E
reporting
requirements.
However,
note
that
DIG
would
be
required
to
install
a
NO
x
continuous
emissions
2
1
The
ozone
season
ends
on
September
30
and,
for
2005
and
thereafter,
starts
on
May
1.

2
A
court
decision
has
mandated
that
the
2004
ozone
season
begin
on
May
31
rather
than
May
1
in
certain
states
(
including
Michigan).
monitoring
system
(
CEMS)
and
a
more
sophisticated
DAHS
on
Unit
GTP1
if
the
unit
should
ever
lose
its
status
as
a
peaking
unit.
Faced
with
this
possibility,
the
October
24,
2002
petition
requested
approval
of
the
PEMS
as
an
alternative
monitoring
system
(
AMS)
under
Subpart
E
of
Part
75.
If
approved
as
an
AMS,
the
PEMS
could
be
used
in
lieu
of
a
NO
x
CEMS
if
Unit
GTP1'
s
peaking
unit
status
should
ever
be
lost.
Approval
of
the
petition
would
also
allow
DIG
to
use
the
AMS
for
Part
75
reporting,
regardless
of
Unit
GTP1'
s
peaking
status.
NO
x
Rules
801­
818
further
require
DIG
to
hold
NO
x
allowances
equal
to
the
ozone
season1
NO
x
mass
emissions
from
Unit
GTP1,
beginning
on
May
31,
20042.

In
its
certification
application,
DIG
submitted
Subpart
E
data
for
two
PEMS
models,
a
simple
model
and
a
complete
model.
Each
model
was
evaluated
against
quality­
assured
data
recorded
by
a
NO
x­
diluent
CEMS,
which
was
temporarily
installed,
certified,
maintained
and
quality­
assured
according
to
Part
75
for
the
purposes
of
the
PEMS
demonstration.
The
two
models
are
identical;
only
the
training
data
set
used
for
each
model
was
different.
For
the
simple
model,
the
first
40
hours
of
quality
assured
data
were
used
to
train
the
model
and
the
remaining
762
hours
were
used
to
test
the
model
using
Part
75,
Subpart
E
statistics.
For
the
complete
model,
all
802
hours
of
quality
assured
data
were
used
to
train
and
to
test
the
model.
Because
it
is
more
rigorous
to
test
a
PEMS
model
on
a
different
data
set
than
the
one
on
which
it
was
trained
,
and
because,
typically,
a
PEMS
may
be
trained
on
as
few
as
40
hours
of
data
in
practice,
EPA
decided
to
evaluate
compliance
with
Subpart
E
based
on
the
simple
model.

EPA'S
DETERMINATION
Under
subpart
E,
the
owner
or
operator
of
a
unit
applying
to
the
Administrator
for
approval
of
an
AMS
must
demonstrate
that
the
AMS
has
the
same
or
better
precision,
reliability,
accessibility,
and
timeliness
(
PRAT)
as
provided
by
a
CEMS.
The
demonstration
must
be
made
by
comparing
the
AMS
to
a
contemporaneously
operating,
fully
certified
CEMS.
Sections
75.41
through
75.46
discuss
the
criteria
for
evaluating
PRAT,
daily
quality
assurance,
and
missing
data
substitution
for
the
AMS.
Section
75.48
details
the
information
that
must
be
included
in
the
application
in
order
to
demonstrate
that
the
criteria
in
§
§
75.41­
46
are
met.

The
following
paragraphs
describe
how
DIG
meets
the
requirements
of
a
subpart
E
AMS
petition.
As
detailed
below,
EPA's
approval
applies
only
to
the
Unit
GTP1
when
firing
pipeline
natural
gas,
and
for
certain
PEMS
outputs,
i.
e.,
lb
NO
x/
mmBtu,
and
NO
x
(
ppm,
dry).
If
a
PEMS
input
parameter
value
goes
below
certain
minimum
or
above
certain
maximum
values,
DIG
shall
report
the
maximum
potential
NO
x
emission
rate
(
MER).
During
startups,
shutdowns,
and
lean/
lean
turbine
operation,
or
if
the
PEMS
alarms,
DIG
must
report
the
NO
x
MER.

1.
Precision
Under
§
75.41,
for
the
normal
unit
operating
level,
DIG
must
provide
paired
AMS
and
fully
3
3
Under
§
75.41(
b),
in
preparation
for
conducting
the
required
statistical
tests,
the
data
may
be
screened
for
lognormality
and
time
dependency
autocorrelation.
If
either
is
detected,
certain
calculation
adjustments
are
required.
DIG
detected
neither
lognormality
nor
autocorrelation.
Therefore,
consistent
with
§
75.41(
b),
no
calculation
adjustments
were
made
to
the
data.
certified
CEMS
hourly
data
for
at
least
90
percent
of
the
hours
during
720
unit
operating
hours
for
the
primary
fuel
supply
and
for
at
least
24
successive
unit
operating
hours
for
all
alternative
fuel
supplies
that
have
significantly
different
sulfur
content.
DIG
must
not
use
missing
data
substitution
procedures
to
provide
sample
data.
DIG
may
also
demonstrate,
and
adjust
the
data
to
account
for,
any
lognormality
and
time
dependency
autocorrelation.
DIG
must
pass
three
statistical
tests,
i.
e.,
a
linear
correlation
coefficient
(
r)
>
0.8,
an
F­
test,
and
a
one­
tailed
t­
test
for
bias
described
in
appendix
A
to
part
75.
Further,
DIG
must
provide
two
separate
time
series
plots
for
AMS
and
CEMS
data.
Each
data
plot
must
have
a
horizontal
axis
representing
the
clock
hour
and
calendar
date
of
the
readings
and
must
contain
a
separate
data
point
for
every
hour
for
the
duration
of
the
test.
One
data
plot
must
show
percentage
difference
vs.
time,
and
the
other
data
plot
must
show
AMS
and
CEMS
readings
vs.
time.
Finally,
a
plot
of
the
paired
AMS
(
on
the
vertical
axis)
and
CEMS
(
on
the
horizontal
axis)
concentrations
must
be
provided.

DIG
provided
762
hours
of
historical,
paired
CEMS
vs.
PEMS
data
while
pipeline
natural
gas
was
being
combusted
in
Unit
GTP1.
According
to
DIG,
the
762
hours
represent
more
than
90%
of
the
unit
operating
hours
in
the
three­
year
data
collection
period,
thereby
satisfying
the
requirement
in
§
75.41(
a)(
6).
According
to
DIG,
all
762
hours
of
data
were
quality­
assured,
i.
e.,
no
missing
data
substitution
procedures
were
applied.

The
table
below
shows
the
results
of
the
statistical
tests
for
the
two
approved
PEMS
outputs.
3
PEMS
(
lbs
NOx/
mmBtu)
PEMS
(
NOx
ppm,
dry)

t­
test:
mean
difference
d
=
­
0.001
abs.
value
of
confidence
coefficient
cc
=
0.002
Evaluation:
Since

cc

>
d,
the
model
passed.
t­
test:
mean
difference
d
=
0.024
abs.
value
of
confidence
coefficient
cc
=
0.438
Evaluation:
Since

cc

>
d,
the
model
passed.

r­
coefficient
correlation:
r
=
0.859
Evaluation:
Since
r
>
0.8,
the
model
passed.
r­
coefficient
correlation:
r
=
0.837
Evaluation:
Since
r
>
0.8,
the
model
passed.

F­
test:
variance
of
PEMS
=
0.001328
variance
of
RM
=
0.001900
F
=
0.699
F
critical
=
1.13
Evaluation:
Since
F
critical
>
F,
the
model
passed.
F­
test:
variance
of
PEMS
=
73.216
variance
of
RM
=
124.773
F
=
0.587
F
critical
=
1.13
Evaluation:
Since
F
critical
>
F,
the
model
passed.
4
The
PEMS
output
of
NO
x
emission
rate
in
lb/
mmBtu
passed
all
three
statistical
tests,
although
EPA
calculated
somewhat
different
values
for
the
t­
test
and
F­
test
than
were
provided
in
the
petition.
Because
the
electronic
paired
CEMS
vs
PEMS
data
provided
to
EPA
were
in
units
of
NO
x
ppm
and
%
CO
2,
EPA
recalculated
the
NO
x
lb/
mmBtu
values
using
these
data
before
running
the
statistics.
The
NO
x
emission
rates
were
calculated
using
Equation
F­
6
and
an
F
c
factor
of
1,040
for
natural
gas,
from
Appendix
F
of
Part
75.
Although
the
petition
did
not
address
it,
the
PEMS
output
of
NO
x
ppm
(
dry
basis)
passed
all
three
statistics.
EPA
calculated
these
statistics
because
DIG
desired
this
additional
output.
EPA
also
calculated
the
Subpart
E
statistics
for
the
PEMS
output
of
%
CO
2.
The
%
CO
2
output
passed
the
F­
test,
but
failed
the
"
r"
correlation
and
the
t­
test.
Therefore,
the
%
CO
2
PEMS
output
is
not
approvable;
DIG
will
continue
to
use
Part
75,
Appendix
G
to
report
CO
2.

Further,
DIG
supplied
the
appropriate
data
plots
concerning
the
paired
AMS
and
CEMS
data
under
§
§
75.41(
a)(
9)
and
(
c)(
2)(
i).

2.
Reliability
According
to
§
75.42,
the
owner
or
operator
must
demonstrate
that
the
PEMS
is
capable
of
providing
valid
hourly
averages
for
95.0
percent
or
more
of
unit
operating
hours
over
a
one­
year
period,
and
that
the
system
meets
the
applicable
quality­
assurance
requirements
of
Part
75,
Appendix
B.
The
October
24,
2002
petition
states
that
the
PEMS
provided
98.7%
data
availability
over
the
three­
year
data
collection
period.
EPA
therefore
finds
that
the
PEMS
meets
the
§
75.42
requirements
for
monitoring
system
data
availability.
By
meeting
the
QA/
QC
requirements
described
in
this
letter,
DIG
will
also
meet
the
applicable
Appendix
B
quality
assurance
and
quality
control
(
QA/
QC)
requirements.

3.
Accessibility
and
Timeliness
According
to
§
§
75.43
and
75.44,
the
owner
or
operator
must
demonstrate
that
the
PEMS:
meets
the
recordkeeping
and
reporting
requirements
of
Subparts
F
and
G
of
Part
75;
can
provide
"
a
continuous,
quality
assured
permanent
record
of
certified
emissions
data
on
an
hourly
basis";
and
is
capable
of
"
issuing
a
record
of
data
for
the
previous
day
within
24
hours".
The
PEMS
has
demonstrated
the
ability
to
meet
Subpart
F
and
G
requirements
by
providing
Part
75
quarterly
electronic
data
reports
(
EDRs)
to
EPA
since
1999.
The
software
also
provides
a
continuous
display
of
real­
time
emissions
data
to
the
operator.
In
view
of
these
considerations,
EPA
finds
that
the
PEMS
meets
the
requirements
of
§
§
75.43
and
75.44
.

4.
Quality
Assurance
Under
§
75.45,
DIG
must
demonstrate
either
that
daily
tests
equivalent
to
those
in
Appendix
B
of
Part
75
can
be
performed
on
the
PEMS
or
that
such
tests
are
unnecessary
for
providing
qualityassured
data.
Sections
75.48(
a)(
8)­(
11)
require
DIG
to
submit:
a
detailed
description
of
the
process
used
to
collect
data,
including
location
and
method
of
ensuring
an
accurate
assessment
of
operating
hourly
conditions
on
a
real­
time
basis;
a
detailed
description
of
the
operation,
maintenance,
and
quality
assurance
procedures
for
the
AMS
as
required
in
Part
75,
Appendix
B;
a
description
of
methods
used
to
calculate
diluent
gas
concentration;
and
results
of
tests
and
measurements
necessary
5
to
substantiate
the
equivalency
of
the
AMS
to
a
fully
certified
CEMS.
EPA
finds
that
the
PEMS
will
meet
these
requirements
by
meeting
the
following:

(
a)
The
PEMS
uses
the
following
input
parameters:
load,
gas
flow,
PM1
(
nozzle
1
fuel
flow
ratio),
PM2
(
nozzle
2
fuel
flow
ratio),
PM3
(
nozzle
3
fuel
flow
ratio),
inlet
air
temperature,
and
burner
mode.
The
PEMS
input
parameters
must
stay
within
the
minimum
and
maximum
values
(
inclusive)
in
the
below
table
(
referred
to
as
"
the
PEMS
operating
envelope"),
unless
the
PEMS
is
retrained
according
to
paragraph
(
g)
below,
in
which
case,
the
new
training
values
will
supercede
the
values
in
the
below
table.
Except
for
burner
mode
parameter,
if
any
PEMS
input
parameter
value
goes
below
the
minimum
or
above
the
maximum
table
values
by
5
percent
or
more,
the
PEMS
shall
be
considered
out­
of­
control,
and
the
NO
x
MER
shall
be
used,
calculated
according
to
paragraph
(
h),
starting
with
the
hour
after
the
sensor
value
goes
outside
of
the
PEMS
operating
envelope
and
ending
with
the
hour
after
the
sensor
value
is
back
within
the
PEMS
operating
envelope.
Data
from
each
PEMS
input
parameter
shall
be
maintained
on
site
in
a
form
suitable
for
inspection
for
at
least
three
(
3)
years
from
the
date
of
each
record.
If
the
burner
mode
is
not
steady
state
(
mode
6),
DIG
shall
follow
the
procedures
in
paragraph
(
h).

PEMS
Operating
Envelope
PEMS
Input
Parameter
Minimum
Value
Maximum
Value
Load
(
MW)
85.5
169.1
Gas
flow
(
hscfh)
10,275.3
15,873.7
PM1
(
unitless)
a
0.064
0.531
PM2
(
unitless)
b
0.071
0.427
PM3
(
unitless)
c
0.162
0.609
Inlet
air
temp
(
deg
F)
43
103
Burner
mode
d
6
6
a
PM1
or
Premix
1
=
PM1
nozzle
fuel
flow
/
total
fuel
flow
into
combustion
chamber.
b
PM2
or
Premix
2
=
PM2
nozzle
fuel
flow
/
total
fuel
flow
into
combustion
chamber.
c
PM3
or
Premix
3
=
PM3
nozzle
fuel
flow
/
total
fuel
flow
into
combustion
chamber.
d
Six
burner
modes:
(
1)
Startup
(
0­
26%
load
with
primary
gas
going
in
and
being
fired)
or
shutdown;
(
2­
5)
Lean/
Lean
(
27­
67%
load
with
primary
and
secondary
gas
going
in
and
both
being
fired);
and
(
6)
Steady
state
(
68­
100%
load
with
PM1,
PM2,
and
PM3
all
non­
zero).
Note:
Burner
mode
6,
itself,
is
not
necessarily
a
PEMS
input
because
load,
PM1,
PM2,
and
PM3
inputs
are
sufficient
to
define
burner
mode
6.

(
b)
The
sensors
for
the
PEMS'
input
parameters
must
be
maintained
in
accordance
with
the
manufacturer's
recommendations.
Further,
the
PEMS'
Sensor
Validation
System
identifies
and
reconciles
failed
sensors
by:
comparing
each
sensor
to
several
other
sensors;
determining,
based
on
the
comparison,
if
a
sensor
has
failed;
and
calculating
a
value
for
any
failed
sensor.
DIG
must
check,
and
demonstrate,
that
the
Sensor
Validation
System:
validates
sensor
data
in
this
way
every
minute
of
PEMS
operation;
and
computes
hourly
averages
using
at
least
one
valid
data
point
in
each
fifteen
minute
quadrant
of
an
hour
(
producing
at
least
four
valid
6
data
points
per
hour),
where
the
unit
combusted
fuel
during
that
quadrant
of
an
hour,
to
comply
with
§
75.10(
d)(
1).

(
c)
DIG
shall
implement
a
sensor
validation
alarm
system
to
inform
the
operator
when
sensors
need
repair
and
to
indicate
that
the
PEMS
is
out­
of­
control.
In
setting
up
the
alarm
system,
a
demonstration
shall
be
performed
at
a
minimum
of
four
different
PEMS
training
conditions,
which
must
be
representative
of
the
entire
range
of
expected
boiler
operations.
For
each
of
the
four
or
more
training
conditions,
the
demonstration
shall
consist
of
the
following:

(
1)
For
all
of
the
sensors
used
in
the
PEMS
model,
input
a
set
of
reference
sensor
values
that
were
recorded
either
during
the
training
of
the
PEMS
or
during
a
RATA
of
the
PEMS
(
these
values
will
all
be
within
the
PEMS
operating
envelope).
Verify
that
these
reference
inputs
produce
the
expected
PEMS
output,
i.
e.,
the
expected
NO
x
emission
rate;

(
2)
Perform
one­
sensor
failure
analysis,
as
follows.
Artificially
fail
one
of
the
sensors
and
then,
using
the
calculated
replacement
value
for
that
sensor
(
see
paragraph
(
b),
above),
assess
the
effect
on
the
accuracy
of
the
PEMS.
Calculate
the
percent
difference
between
the
reference
NO
x
emission
rate
from
step
(
1)
and
the
PEMS
output.
Repeat
this
procedure
for
each
sensor,
individually;

(
3)
Identify
the
sensor
failure
in
step
(
2)
that
results
in
the
worst
accuracy.
If
the
highest
percent
deviation
exceeds
+
10.0%,
then
set
up
the
PEMS
to
alarm
when
any
single
sensor
fails.
If
none
of
the
percent
difference
values
exceeds
10.0%,
proceed
to
step
(
4);

(
4)
Perform
two­
sensor
failure
analysis,
as
follows:
Artificially
fail
the
sensor
from
step
(
3)
that
produced
the
worst
accuracy
and
also
fail
one
of
the
other
sensors.
Then,
using
the
calculated
replacement
values
for
both
sensors,
assess
the
accuracy
of
the
PEMS
hourly
average
output,
as
in
step
(
2).
Repeat
this
procedure,
evaluating
each
sensor
in
turn
with
the
sensor
from
step
(
3);

(
5)
Identify
the
combination
of
dual
sensor
failures
that
results
in
the
worst
accuracy.
If
the
highest
percent
deviation
exceeds
+
10.0%,
then
set
up
the
PEMS
to
alarm
when
any
two
sensors
fail.
If
none
of
the
percent
difference
values
exceeds
10.0%,
then
set
up
the
PEMS
to
alarm
with
three
sensor
failures.

The
results
of
this
demonstration
shall
be
reported
in
the
Subpart
H
certification
hardcopy
test
report
and
in
record
type
910
in
the
quarterly
EDR
submittal
for
the
quarter
in
which
the
demonstration
is
performed.
When
the
PEMS
alarms,
the
PEMS
is
out­
of­
control
and
DIG
shall
report
the
NO
x
MER,
calculated
according
to
paragraph
(
h),
starting
with
the
hour
after
the
sensor
validation
alarm
system
alarms
and
ending
with
the
hour
after
the
sensor
value
is
back
within
the
expected
range.

(
d)
A
daily
QA/
QC
test
must
be
performed.
DIG
shall
input
to
the
PEMS
a
set
of
turbine
operating
parameters
collected
during
the
most
recent
RATA
or
training.
The
resulting
7
PEMS
NO
x
lb/
mmBtu
output
shall
be
compared
to
the
reference
method
NO
x
lb/
mmBtu
value,
measured
during
the
most
recent
RATA
or
training.
If
the
PEMS
NO
x
output
is
within
+
10.0%
of
the
reference
method
value,
the
daily
QA/
QC
test
is
passed.
If
the
daily
QA/
QC
test
is
failed,
the
PEMS
is
out­
of­
control.
Subpart
D
missing
data
procedures
shall
be
followed
starting
with
the
hour
after
the
failed
test
or,
if
the
test
is
not
timely
conducted,
the
hour
after
the
test
due
date
and
ending
with
the
hour
in
which
the
test
is
passed.
No
grace
periods
are
allowed.
The
results
of
this
check
(
pass/
fail)
shall
be
reported
in
RT
624
in
EDR
version
2.2.
(
Note:
Use
code
`
04'
in
start
column
53
(
QA
test
code)
for
the
daily
QA/
QC
check.)

(
e)
DIG
shall
perform
the
tests
indicated
in
the
`
PEMS
Ongoing
QA/
QC
Tests'
table,
below,
including
monthly
3­
run
relative
accuracy
audits
(
RAAs).
A
monthly
RAA
shall
be
performed
in
every
calendar
month
during
the
ozone
season
when
NO
x
monitoring
is
required
by
Part
75
that
the
unit
operates
for
at
least
168
hours,
except
for
a
month
in
which
a
full
9­
run
RATA
is
performed.
The
monthly
RAAs
shall
be
done
on
a
pounds
of
NO
x
per
million
Btu
(
lb/
mmBtu)
basis,
and
shall
be
performed
using
either
EPA
Reference
Methods
7E
and
3A
in
40
CFR
Part
60,
Appendix
A­
4,
or
a
portable
electrochemical
analyzer.
To
the
extent
practicable,
each
successive
monthly
RAA
shall
be
done
at
different
operating
conditions
from
the
previous
one.
Follow
analyzer
manufacturer's
recommended
maintenance
procedures.

The
minimum
time
per
RAA
run
shall
be
20
minutes.
The
reference
method
traverse
point
selection
shall
be
consistent
with
section
6.5.6
of
Part
75,
Appendix
A.
Alternatively,
a
single
measurement
point
located
at
least
0.5
meters
from
the
stack
or
duct
wall
may
be
used
without
performing
a
stratification
test.

Results
of
the
RAA
shall
be
calculated
using
Equation
1­
1
in
Appendix
F
of
40
CFR
Part
60.
Bias­
adjusted
data
from
the
PEMS
(
using
the
bias
adjustment
factor
from
the
most­
recent
RATA)
shall
be
used
in
the
calculations.
The
results
of
the
RAA
are
acceptable
if
the
performance
specifications
in
the
`
PEMS
On­
going
QA/
QC
Tests'
table,
below,
are
met.
If
the
monthly
RAA
is
failed,
the
PEMS
is
out­
of­
control.
Subpart
D
missing
data
procedures
shall
be
followed
starting
with
the
hour
after
the
failed
test
or,
if
the
test
is
not
timely
conducted,
the
hour
after
the
test
due
date,
and
ending
with
the
hour
in
which
the
test
is
passed.
No
grace
periods
are
allowed.

Report
the
results
of
all
monthly
RAAs
in
the
appropriate
quarterly
electronic
data
report
(
EDR).
Use
EDR
record
type
624,
and
report
the
results
of
each
test
as
either
`
pass'
or
`
fail'.
Report
the
QA
test
code
in
column
53
of
RT
624
as
`
05'.

If
a
chemiluminescent
NO
x
analyzer
(
a
portable
chemiluminescent
analyzer
may
be
used)
is
used
to
perform
the
required
RAAs,
the
procedures
of
Method
7E
in
40
CFR
Part
60,
Appendix
A­
4
shall
be
followed.
The
analyzer
performance
specifications
in
Method
7E
for
calibration
error,
system
bias,
and
calibration
drift
shall
be
met.

If
a
portable
electrochemical
analyzer
is
used
to
perform
the
required
RAAs,
ASTM
Method
8
4ASTM
D6522­
00,
Standard
Test
Method
for
Determination
of
Nitrogen
Oxides,
Carbon
Monoxide,
and
Oxygen
Concentrations
in
Emissions
from
Natural
Gas­
Fired
Reciprocating
Engines,
Combustion
Turbines,
Boilers,
and
Process
Heaters
Using
Portable
Analyzers.

5
GRI
(
Gas
Research
Institute),
Topical
Report,
Development
of
an
Electrochemical
Cell
Emission
Analyzer
Test
Method,
July,
1997.

6
Evaluation
of
Portable
Analyzers
for
Use
in
Quality
Assuring
Predictive
Emission
Monitoring
Systems
for
NOx,
The
Cadmus
Group,
Inc.,
September
8,
2004
D6522­
004,
as
modified
below,
shall
be
followed.
ASTM
D6522­
00
applies
to
the
measurement
of
NO
x
(
NO
and
NO
2),
CO,
and
O
2
concentrations
in
emissions
from
natural
gas­
fired
combustion
systems
using
electrochemical
analyzers.
The
method
was
developed
based
on
studies
sponsored
by
the
Gas
Research
Institute
(
GRI)
5.
It
has
also
been
peerreviewed
approved
by
ASTM
Committees
D22.03
and
D22,
and
accepted
by
EPA
as
a
conditional
test
method
(
CTM­
030).
ASTM
D6522­
00
prescribes
analyzer
design
specifications,
test
procedures,
and
instrument
performance
requirements
that
are
similar
to
the
checks
in
EPA's
instrumental
test
methods
(
e.
g.,
Methods
7E
and
20).
These
checks
include
linearity,
interference,
stability,
pre­
test
calibration
error,
and
post­
test
calibration
error.

Based
on
the
results
of
EPA's
portable
analyzer
study6,
the
following
modifications
to
ASTM
D6522­
00
are
required
to
make
the
method
more
practical
without
sacrificing
accuracy:
(
a)
NO
x
analyzers
must
provide
readings
to
0.1
ppm
to
improve
the
likelihood
of
passing
the
performance
specifications
for
sources
with
low
NO
x
levels;
(
b)
an
alternative
performance
specification
(
e.
g.,
+
1
ppm
difference
from
reference
value)
will
be
applied
to
take
account
of
sources
with
low
concentrations
of
NO
x;
and
(
c)
the
measurement
system
must
be
purged
with
ambient
air
between
gas
injections
during
the
stability
check,
to
reduce
degradation
of
electrochemical
cell
performance
(
see
footnote
6,
below).

The
measurement
system
performance
specifications
as
modified
by
the
EPA
portable
analyzer
study
are
shown
in
the
following
table.

ASTM
Method
D6522­
00
Measurement
System
Performance
Specifications
(
as
Modified
by
EPA
Portable
Analyzer
Study)

Performance
Check
Gas
Acceptance
Criteria
Zero
Calibration
Error
NO,
NO2

3
percent
of
span
gas
value
or
+
1.0
ppm
difference,
whichever
is
less
restrictive
O2

0.3
percent
O2
Span
Calibration
Error
NO,
NO2

5
percent
of
span
gas
value
or
+
1.0
ppm
difference,
whichever
is
less
restrictive
O2

0.5
percent
O2
Interference
NO,
NO2,
O2

5
percent
of
average
stack
NO
concentration
for
each
test
run
(
using
span
gas
checks)

Linearity
NO,
O2

2.5
percent
of
span
gas
concentration
or
+
1.0
ppm
difference,
whichever
is
less
restrictive
9
7When
conducting
this
check
for
three
cells
in
an
analyzer,
the
system
must
be
purged
with
ambient
air
between
gas
injections
to
minimize
the
possibility
of
problems
with
the
electrochemical
cells.
Otherwise,
the
cells
will
be
exposed
to
high
NO
and
NO
2
concentrations
for
prolonged
periods
of
time,
which
can
cause
degradation
in
the
cell's
performance
(
i.
e.,
the
socalled
"
O
2­
starved
exposure").
NO2

3.0
percent
of
span
gas
concentration
or
+
1.0
ppm
difference,
whichever
is
less
restrictive
Stability
7
NO,
NO2
O2

2.0
percent
of
span
gas
concentration
or
+
1.0
ppm
maxmin
difference,
whichever
is
less
restrictive,
for
30­
minute
period

1.0
percent
of
span
gas
concentration
or
+
1.0
ppm
maxmin
difference,
whichever
is
less
restrictive,
for
15­
minute
period
Cell
Temperature
±
5

F
from
initial
temperature
(
f)
DIG
shall
perform
initial
certification
tests
on
the
PEMS
prior
to
May
1,
2003.
These
tests
shall
be
performed
in
the
following
order:
(
1)
Ensure
that
the
Sensor
Validation
System
meets
the
requirements
of
paragraph
(
b).
(
2)
Train
or
retrain,
as
applicable,
the
PEMS
according
to
the
manufacturer's
recommendations.
(
3)
Ensure
that
the
requirements
in
paragraph
(
c)
are
met.
(
4)
Perform
a
RATA,
following
the
procedures
in
section
6.5
of
Part
75,
Appendix
A,
except
use
three
different
operating
levels
(
low,
mid
and
high)
as
defined
in
section
6.5.2.1
of
Part
75,
Appendix
A.
However,
because
the
PEMS
is
only
approved
for
use
at
68
to
100
percent
load,
use
68
percent
load
as
the
lower
boundary
of
the
range
of
operation
and
100
percent
load
as
the
upper
boundary
of
the
range
of
operation.
Use
paired
PEMS
and
reference
method
data
to
calculate
the
results
on
a
lb
NO
x/
mmBtu
basis2.
DIG
shall
apply
to
each
operating
level
the
RATA
performance
specifications
contained
in
the
PEMS
On­
going
QA/
QC
Tests
table
in
paragraph
(
i).
Report
the
RATA
data
and
results
of
only
the
normal
operating
level
in
EDR
record
types
610
and
611
and
keep
the
data
and
results
for
the
other
two
operating
levels
on­
site,
available
for
inspection.
The
RATA
result
for
the
normal
operating
level
determines
when
the
next
RATA
is
due.
(
5)
Conduct
an
Ftest
and
a
correlation
analysis
using
Part
75,
Subpart
E
equations
at
low,
mid
and
high
operating
levels.
Calculations
shall
be
based
on
a
minimum
of
30
runs
at
each
operating
level.
The
F­
test
is
to
be
applied
to
data
at
each
operating
level
separately.
The
correlation
analysis
shall
be
performed
using
all
data
collected
at
the
three
operating
levels
combined.
If
the
standard
deviation
of
the
reference
method
NO
x
data
is
less
than
either
3
percent
of
the
span
or
5
ppm,
the
reference
method
standard
deviation
of
either
3
percent
of
span
or
5
ppm
may
be
used
when
applying
the
F­
test.
When
the
mean
value
of
the
reference
method
NO
x
data
is
less
than
5
ppm,
the
correlation
analysis
(
r­
test)
may
be
waived
at
that
specific
operating
level.
(
6)
Calculate
and
apply
a
bias
adjustment
factor
(
BAF)
at
the
normal
operating
level
according
to
section
7.6
of
Part
75,
Appendix
A.
If
all
tests
and
procedures
in
(
1),
(
3),
(
4)
and
(
5)
are
not
passed,
and
all
the
procedures
in
(
2)
and
(
6),
are
not
completed,
by
May
1,
2003,
the
NO
x
MER,
calculated
with
the
appropriate
NO
x
MPC
value
from
part
75,
appendix
A,
Table
2­
2,
must
be
used
with
the
measured
heat
input
to
calculate
and
report
NO
x
mass
emissions
until
all
tests
and
procedures
in
(
1)­(
5)
are
passed
and
all
the
procedures
in
(
2)
and
(
6),
are
completed.
10
8
The
unit
is
a
low­
emitting
source
if
the
mean
reference
value
during
the
RATA
or
RAA
is
<
0.200
lb/
mmBtu
NO
x
Note:
Because
DIG
has
already
complied
with
the
provisions
of
paragraph
(
f),
it
does
not
need
to
perform
paragraph
(
f),
above,
unless
a
new
NO
x
PEMS
monitoring
system
is
installed.

(
g)
After
initial
certification,
if
a
RATA
is
failed
due
to
a
problem
with
the
PEMS,
or
if
changes
occur
that
result
in
a
significant
change
in
NO
x
emission
rate
relative
to
the
previous
PEMS
training
conditions
(
e.
g.,
turbine
aging,
process
modification,
new
process
operating
modes,
or
changes
to
emission
controls),
the
tests
and
procedures
in
paragraph
(
f)
shall
be
performed
on
the
PEMS
in
the
order
specified
in
that
paragraph.
The
tests
and
procedures
in
paragraph
(
f)
shall
be
completed
by
the
earlier
of
60
unit
operating
days
(
as
defined
in
section
72.2)
or
180
calendar
days
after
the
failed
RATA
or
after
the
change
that
caused
a
significant
change
in
NO
x
emission
rate.
DIG
shall
use
the
appropriate
Part
75
missing
data
procedures
(
see
section
5),
starting
from
the
hour
of
the
failed
RATA
and
ending
the
hour
after
successful
passage
or
completion
of
the
tests
and
procedures,
as
required
above.
DIG
shall
report
the
NO
x
MER,
and
Method
of
Determination
Code
"
55
Other
substitute
data
approved
through
petition
by
EPA"
in
RT
320
for
reporting
lb
NO
x/
mmBtu
emission
rate,
starting
with
the
hour
after
the
change
that
caused
a
significant
change
in
NO
x
emission
rate,
and
ending
the
hour
after
successful
passage
or
completion
of
the
tests
and
procedures,
as
required
above.

(
h)
For
any
hour
or
partial
hour
of
startup,
shutdown,
or
lean/
lean
turbine
operation
(
burner
modes
1
­
5,
in
other
words,
if
dry
low­
NO
x
is
not
operating),
DIG
must
report
the
NO
x
MER,
as
defined
in
§
72.2,
and
calculated
in
accordance
with
§
2.1.2.1(
b)
of
Appendix
A
to
Part
75.
A
NO
x
MPC
of
150
ppm,
in
accordance
with
Part
75,
Appendix
A,
Table
2­
2,
shall
be
used
in
the
MER
calculation.
Method
of
Determination
Code
"
55
Other
substitute
data
approved
through
petition
by
EPA"
shall
be
used
in
RT
320
for
reporting
lb
NO
x/
mmBtu
emission
rate.

(
i)
On­
going
QA/
QC
tests
shall
be
performed
according
to
the
following
table:

PEMS
On­
going
QA/
QC
Tests
Test
Performance
Specification
Frequency
Daily
QA/
QC
<
10.0%
of
reference
NOx
emission
rate
Daily
(
see
paragraph
(
d))

3­
run
RAA

Accuracy
<
10.0%
or

For
a
low
emitting
source8,
results
are
acceptable
if
the
mean
value
for
the
PEMS
is
within
+
0.020
lb/
mmBtu
of
the
reference
mean
value
Monthly
(
see
paragraph
(
e))
11
RATA
For
semiannual
RATA
frequency:
°
RA
>
7.5%
and
<
10.0%
or

For
a
low
emitting
source8,
results
are
acceptable
if
the
mean
value
for
the
PEMS
is
within
+
0.020
lb/
mmBtu
of
the
reference
method
mean
value.

For
annual
RATA
frequency:

RA
<
7.5%
or

For
a
low
emitting
source8,
results
are
acceptable
if
the
mean
value
for
the
PEMS
is
within
+
0.015
lb/
mmBtu
of
the
reference
method
mean
value
Annual
or
semiannual
and
after
each
PEMS
training
(
see
paragraphs
(
f)
and
(
g)

Sensor
validation
system
(
minimum
data
capture)
Check
for
production
of
at
least
1
valid
data
point
per
15
minutes
(
see
paragraph
(
b))
Daily
Bias
adjustment
factor
If
davg
<

cc

,
bias
test
is
passed
After
each
RATA
(
see
paragraphs
(
f)
and
(
g)

PEMS
training
(
Linear
correlation
and
F­
test)
r
>
0.8,
and
Fcritical
>
F
According
to
paragraphs
(
f)
and
(
g)

Sensor
validation
compliance
alarm
system
set­
up
(
see
paragraph
(
c))
After
each
PEMS
training
(
see
paragraphs
(
f)
and
(
g)

The
daily
QA/
QC
test
is
described
in
paragraph
(
d).
The
3­
run
RAA
is
discussed
in
paragraph
(
e).
On­
going
RATAs
shall
be
performed
at
the
normal
operating
level
according
to
the
procedures
in
Part
75,
Appendix
B,
section
2.3.1
and,
as
discussed
in
paragraph
(
f),
shall
be
calculated
on
a
lb/
mmBtu
basis.
Immediately
prior
to
a
RATA,
the
BAF
shall
be
set
to
1.000.
On
a
daily
basis,
DIG
shall
check
that
the
sensor
validation
system
is
set
to
provide
one
valid
data
point
per
15
minute
period,
as
discussed
in
paragraph
(
b).
After
each
RATA,
DIG
shall
calculate
and
apply
a
bias
adjustment
factor
at
the
normal
operating
level
according
to
Part
75,
Appendix
A,
section
7.6.
DIG
shall
train
or
retrain
the
PEMS
according
to
paragraphs
(
f)
and
(
g).
After
each
training,
DIG
shall
perform
a
normal
operating
level
RATA
and
bias
test,
described
in
paragraph
(
f),
and
the
compliance
alarm
demonstration
in
paragraph
(
c).

5.
Missing
Data
Substitution
Under
§
75.46,
DIG
must
demonstrate
that
all
missing
data
can
be
accounted
for
in
a
manner
consistent
with
the
applicable
missing
data
procedures
in
Subpart
D.
The
DIG
petition
states
that
Unit
GTP1
currently
meets
Appendix
E
requirements,
including
Appendix
E
missing
data
procedures.
When
DIG
discontinues
the
use
of
Appendix
E
and
begins
to
use
the
PEMS
as
an
approved
Part
75
AMS,
all
of
the
Subpart
D
missing
data
procedures
for
NO
x
emission
rate
shall
be
immediately
implemented
(
except
where
alternate
procedures
are
required
in
this
final
approval).
The
Subpart
D
missing
data
procedures
include,
but
are
not
limited
to:
the
initial
missing
data
procedures
in
§
75.31,
determination
of
monitor
data
availability
(
§
75.32),
and
the
standard
missing
data
procedures
in
§
75.33.
12
6.
Additional
Requirements
A
monitoring
plan
is
due
45
days
prior
to
the
initial
certification
tests
(
§
75.62)
described
in
paragraph
(
f)
above.
DIG
shall
submit
the
operating
envelope
for
Unit
GTP1
to
the
Michigan
Department
of
Environmental
Quality
and
to
EPA
Region
5
for
inclusion
in
the
hardcopy
monitoring
plan.
Any
time
changes
are
made
to
the
PEMS
operating
envelope,
the
complete,
revised
PEMS
operating
envelope
shall
be
submitted
in
a
hardcopy
monitoring
plan
by
the
applicable
deadline
in
§
75.62(
a)(
2).
More
information
on
monitoring
plan
submittals
and
other
submittals
can
be
found
at:
http://
www.
epa.
gov/
airmarkets/
monitoring/
submissions/
monplan.
html.

DIG
shall
follow
the
EDR
version
2.2
reporting
instructions,
found
at:
http://
www.
epa.
gov/
airmarkets/
reporting/
edr21/,
in
conjunction
with
the
required
PEMS
record
types,
and
the
supplementary
PEMS
EDR
reporting
instructions
attached
to
this
petition
response,
to
report
data
from
the
PEMS.
Monitoring
Data
Checking
(
MDC)
software
that
can
be
used
to
quality
assure
the
electronic
reports
prior
to
submission
is
found
at:
http://
www.
epa.
gov/
airmarkets/
reporting/
index.
html.

If
there
are
any
further
questions
or
concerns
about
this
matter,
please
contact
John
Schakenbach
of
my
staff
at
202­
343­
9158
or
at
(
schakenbach.
john@
epa.
gov)
.

Sincerely,

Sam
Napolitano,
Director
Clean
Air
Markets
Division
cc:
John
Schakenbach,
EPA,
CAMD
Louis
Nichols,
EPA,
CAMD
Constantine
Blathras,
EPA
Region
5
Karen
Kajiya­
Mills,
MI
DEQ
7/
30/
03
EDR
REPORTING
[
PREDICTIVE
EMISSIONS
MONITORING
SYSTEMS
(
PEMS)]

I.
Introduction
Table
A­
15,
below
includes
the
essential
EDR
record
types
for
units
that
have
received
approval
under
Subpart
E
of
Part
75
to
use
predictive
emissions
monitoring
systems
(
PEMS)
to
report
NO
x
emissions.
The
scope
of
Table
A­
15
is
limited
to
affected
oil
and
gas­
fired
units
(
i.
e.,
boilers
and
combustion
turbines)
that:


Have
a
single
unit­
single
stack
exhaust
configuration;
and

Use
Appendix
D
methodology
to
quantify
unit
heat
input;
and

Use
Appendices
D
and
G
to
account
for
SO
2
and
CO
2
mass
emissions
(
if
the
units
are
in
the
Acid
Rain
Program);
and

Do
not
co­
fire
oil
and
gas.

For
PEMS
reporting,
EDR
version
2.2
must
be
used,
since
fuel­
specific
missing
data
substitution
for
NO
x
emission
rate
is
required.
For
hourly
NO
x
emission
rate
reporting,
RT
320
is
used.
Hourly
200­
level
records
are
not
reported
for
either
NO
x
concentration
or
diluent
gas
(
O
2
or
CO
2)
concentration.

For
units
that
burn
more
than
one
fuel
type,
separate
PEMS
are
required
for
each
fuel.
Each
PEMS
should
be
reported
as
a
separate
monitoring
system
with
a
unique
monitoring
system
ID
in
RT
510.
Each
PEMS
will
require
its
own
set
of
certification,
recertification,
and
quality
assurance
tests.

II.
Interpreting
Table
A­
15
In
Table
A­
15,
the
first
column
identifies
the
record
type.
The
second
column
gives
a
brief
description
of
the
record
type.
The
third,
fourth,
and
fifth
columns
indicate
whether
the
record
type
must
be
reported
for
a
particular
type
of
submittal.
The
third
column
header,
"
MP,"
refers
to
monitoring
plan
submittals.
The
fourth
column
header,
"
CT,"
stands
for
certification
or
recertification
applications.
The
fifth
column
header,
"
QT,"
refers
to
electronic
data
report
submittals.
The
letter
codes
in
columns
3
through
5
are
defined
as
follows:

Y
This
record
type
is
required
for
this
type
of
submittal
(
monitoring
plan,
certification/
recertification
application
or
electronic
data
report)

N
This
record
type
is
not
appropriate
for
this
type
of
submittal.

O
This
record
type
is
appropriate,
but
optional
for
this
type
of
submittal.

A
This
record
type
may
be
required
for
this
submittal.
If
any
doubt
exists
as
to
the
need
to
submit
this
record
type,
consult
the
appropriate
EDR
instructions.

T
This
record
type
is
required
each
time
a
quality
assurance
test
(
e.
g.,
a
RATA)
is
performed.
2
Table
A­
15
EDR
RECORD
TYPES
FOR
UNITS
WITH
PEMS
Record
Type
Description
MP
CT
QT
Program
Applicability
and
Comments
100
Facility
Identification
Y
Y
Y
ARP,
Subpart
H
101
Record
Types
Submitted
O
O
O
ARP,
Subpart
H
102
Facility
Location
and
Identification
Information
Y
Y
Y
ARP,
Subpart
H
300
Operating
Data
N
N
Y
ARP,
Subpart
H
°
Report
one
RT
300
for
each
hour
in
the
quarter,
except
when
a
unit
does
not
operate
during
the
entire
quarter.
°
For
each
operating
hour,
report
the
fuel
combusted
in
column
64.

301
Quarterly
Cumulative
Emissions
N
N
Y
ARP
°
Quarterly
NOx
emission
rate
is
the
arithmetic
average
of
the
RT
320,
col
42
values
302
Oil
Fuel
Flow
N
N
Y
ARP,
Subpart
H
°
For
ARP
units,
must
be
paired
with
RT
313
when
reporting
SO2
mass
emissions.

303
Gas
Fuel
Flow
N
N
Y
ARP,
Subpart
H
°
For
ARP
units,
must
be
paired
with
RT
314
when
reporting
SO2
mass
emissions.

307
Cumulative
NOx
Mass
Emissions
N
N
Y
Subpart
H
313
SO2
Mass
Emissions
(
Oil)
N
N
Y
ARP
314
SO2
Mass
Emissions
(
Gas)
N
N
Y
ARP
320
NOx
Emission
Rate
Estimation
N
N
Y
ARP,
Subpart
H
°
(
See
supplementary
reporting
instructions)

328
NOx
Mass
Emissions
N
N
Y
Subpart
H
°
(
See
supplementary
reporting
instructions)

330
CO2
Mass
Emissions
Data
N
N
A
ARP
°
Report
RT
330
for
hours
in
which
Equation
G­
4
is
used
to
determine
hourly
CO2
mass
emissions
for
gas
or
oil­
fired
units.

331
CO2
Mass
Emissions
Estimation
Parameters
N
N
A
ARP
°
Report
RT
331
if
you
estimate
CO2
mass
emissions
using
fuel
sampling
and
Equation
G­
1
504
Unit
Information
Y
Y
Y
ARP,
Subpart
H
505
Program
Indicator
for
Report
Y
Y
Y
ARP,
Subpart
H
506
EIA
Cross
Reference
Information
Y
Y
Y
ARP,
Subpart
H
507
Peaking
Unit
or
ARP
Gas­
Fired
Unit
Qualification
Data
A
A
A
ARP
508
Subpart
H
Reporting
Frequency
Change
N
N
A
Subpart
H
510
Monitoring
Systems/
Analytical
Components
Table
Y
Y
Y
ARP,
Subpart
H
°
(
See
supplementary
reporting
instructions)
3
Record
Type
Description
MP
CT
QT
Program
Applicability
and
Comments
520
Formula
Table
Y
Y
Y
ARP,
Subpart
H
°
Report
formulas
for
SO2
and
CO2
mass
emissions
(
ARP
units,
only),
NOx
mass
emissions
(
Subpart
H
units),
and
unit
heat
input
rate.

531
Defaults
and
Constants
Y
Y
Y
ARP,
Subpart
H
°
(
See
supplementary
reporting
instructions)

535
Unit
and
Stack
Operating
Load
Data
Y
Y
Y
ARP,
Subpart
H
Required
for
any
unit
using
load­
based
missing
data
procedures
for
NOx
or
fuel
flow
rate.

536
Range
of
Operation,
Normal
Load,
and
Load
Usage
Y
Y
Y
ARP,
Subpart
H
°
Report
RT
536
to
define
operating
range
and
normal
load
for
RATA
testing
540
Fuel
Flowmeter
Data
Y
Y
Y
ARP,
Subpart
H
550
Reasons
for
Monitoring
System
Downtime
or
Missing
Parameter
N
N
A
ARP,
Subpart
H
°
(
See
supplementary
reporting
instructions)

556
Monitoring
System
Recertification,
Maintenance,
or
Other
Events
N
N
A
ARP,
Subpart
H
°
Report
RT
556
for
recertification
of
the
PEMS
or
fuel
flowmeters
°
(
See
supplementary
reporting
instructions)

585
Monitoring
Methodology
Information
Y
Y
Y
ARP,
Subpart
H
°
(
See
supplementary
reporting
instructions)

586
Control
Equipment
Information
A
A
A
ARP,
Subpart
H
587
Unit
Fuel
Type
Y
Y
Y
ARP,
Subpart
H
610
RATA
and
Bias
Test
Data
N
Y
T
ARP,
Subpart
H
°
Report
RTs
610
each
time
a
RATA
is
performed
for
certification,
recertification
or
for
on­
going
QA/
QC.
°
(
See
supplementary
reporting
instructions)

611
RATA
and
Bias
Test
Results
N
Y
T
ARP,
Subpart
H
°
Report
RT
611
each
time
a
RATA
is
performed
for
certification,
recertification
or
for
on­
going
QA/
QC.
°
(
See
supplementary
reporting
instructions)

624
Other
QA
Activities
N
N
Y
ARP,
Subpart
H
°
Report
RT
624
for
PEMS
daily
QA/
QC
and
for
PEMS
periodic
accuracy
checks
using
a
reference
method,
or
a
portable
analyzer.
°
(
See
supplementary
reporting
instructions)

627
Fuel
Flowmeter
Accuracy
Test
N
A
T
ARP,
Subpart
H
°
Report
only
for
fuel
flowmeters
that
are
certified
and
quality
assured
by
periodic
accuracy
tests
according
to
Section
2.1.5.1
or
2.1.5.2
of
Appendix
D.

628
Fuel
Flowmeter
Accuracy
Test
for
Orifice,
Nozzle
and
Venturi
Flowmeter
N
A
T
ARP,
Subpart
H
°
Report
only
for
orifice,
nozzle
and
venturi­
type
flowmeters
that
are
quality
assured
by
periodic
transmitter/
transducer
calibrations.
4
Record
Type
Description
MP
CT
QT
Program
Applicability
and
Comments
629
Fuel
Flow­
to­
load
Ratio
Test
Baseline
Data
N
N
A
ARP,
Subpart
H
°
Report
if
quarterly
fuel
flow­
to­
load
ratio
test
in
Section
2.1.7
of
Appendix
D
is
used
to
extend
fuel
flowmeter
accuracy
test
deadlines.

630
Quarterly
Fuel
Flow­
to­
load
Ratio
Test
Results
N
N
A
ARP,
Subpart
H
°
Report
if
quarterly
fuel
flow­
to­
load
ratio
test
in
Section
2.1.7
of
Appendix
D
is
used
to
extend
fuel
flowmeter
accuracy
test
deadlines.

696
Fuel
Flowmeter
Accuracy
Test
Extension
N
N
A
ARP,
Subpart
H
°
Use
RT
696
to
claim
allowable
extensions
of
fuel
flowmeter
accuracy
test
deadlines.

697
RATA
Deadline
Extension
or
Exemption
N
N
A
ARP,
Subpart
H
°
Report
when
claiming
a
RATA
deadline
extension
Appendix
B,
Section
2.3.3.

699
QA
Test
Extension
Based
on
Grace
Period
N
N
A
ARP,
Subpart
H
°
Report
when
claiming
a
QA
test
deadline
extension
under
Appendix
B,
Section
2.2.4.

900
Certifications
Y
Y
Y
ARP
901
Certifications
Y
Y
Y
ARP
910
Comments
Y
Y
Y
ARP,
Subpart
H
°
(
See
supplementary
reporting
instructions)

920
Comments
O
O
O
ARP,
Subpart
H
940
Certifications
Y
Y
Y
Subpart
H
941
Certifications
Y
Y
Y
Subpart
H
999
Contact
Information
O
O
O
ARP,
Subpart
H
5
SUPPLEMENTARY
EDR
REPORTING
INSTRUCTIONS
FOR
PEMS
For
a
unit
with
an
approved
petition
to
use
a
predictive
emissions
monitoring
system
(
PEMS),
use
the
following
supplementary
instructions,
in
conjunction
with
the
EDR
version
2.2
Reporting
Instructions
document,
to
prepare
the
required
EDR
submittals.

RT
320
Monitoring
System
ID
(
10).
Report
the
monitoring
system
ID
(
from
RT
510,
column
13)
of
the
PEMS
used
to
determine
the
NO
x
emission
rate
during
the
hour.

F­
Factor
(
26).
Leave
this
field
blank.

Average
NOx
Emission
Rate
for
the
Hour
(
36).
Report
the
average
unadjusted
NO
x
emission
rate
for
the
hour
(
lb/
mmBtu),
rounded
to
three
decimal
places,
as
determined
by
the
PEMS.
For
hours
in
which
you
use
missing
data
procedures,
leave
this
field
blank.

Adjusted
Average
NOx
Emission
Rate
for
the
Hour
(
42).
For
each
hour
in
which
you
report
NO
x
emission
rate
in
column
36,
apply
the
appropriate
adjustment
factor
(
1.000
or
the
BAF)
to
the
unadjusted
average
emission
rate,
and
report
the
result
rounded
to
three
decimal
places.
For
each
hour
in
which
you
use
missing
data
procedures,
report
the
appropriate
substitute
value.

Formula
ID
(
50).
Leave
this
field
blank.

Method
of
Determination
Code
(
53).
Report
"
03"
when
you
use
the
PEMS
to
determine
the
NO
x
emissions
rate.
Report
"
12"
when
you
report
the
fuel­
specific
maximum
NO
x
emission
rate
(
e.
g.,
during
hours
of
startup
or
shutdown
or
when
NO
x
controls
(
if
any)
are
not
functioning
properly).
During
hours
when
you
use
other
missing
data
procedures,
report
the
appropriate
MODC
listed
in
the
EDR
instructions.

RT
328
NOx
Methodology
for
the
Hour
(
45).
Report
"
NOXR­
PEMS".

RT
510
The
PEMS
monitoring
system
consists
of
either
one
or
two
data
acquisition
and
handling
system
(
DAHS)
components.
For
single­
component
PEMS
systems
or
for
systems
where
the
PEMS
software
and
standard
DAHS
software
have
the
same
manufacturer/
provider,
model
or
version
number,
etc.,
report
one
RT
510
for
the
PEMS
system.
If
the
PEMS
software
and
the
standard
DAHS
software
have
different
manufacturer/
providers,
model
or
version
numbers,
etc.,
report
each
as
a
separate
RT
510
with
the
same
PEMS
monitoring
system
ID.

Component
ID
(
10).
Report
the
three­
character
alphanumeric
ID
for
each
DAHS
component.

Monitoring
System
ID
(
13).
Create
a
unique
three­
character
alphanumeric
ID
for
each
PEMS
monitoring
system.
Define
a
separate
NOX
PEMS
system
for
each
fuel
type.
For
sources
6
switching
from
NO
x
CEMS
or
Appendix
E
to
PEMS,
do
not
re­
use
the
CEMS
or
Appendix
E
system
ID
numbers.

System
Parameter
Monitored
(
17).
Report
"
NOX"
for
the
system
parameter
monitored.

Primary/
Backup
Designation
(
21).
Report
"
PE"
to
indicate
that
this
is
a
predictive
emissions
monitoring
system.

Component
Type
Code
(
23).
Report
"
DAHS"
as
the
component
type
code.

Sample
Acquisition
Method
(
27).
Leave
this
field
blank.

Manufacturer
(
30).
Report
the
name
of
the
manufacturer
or
developer
of
the
software
component.

Model/
Version
(
55).
Report
the
model/
version
of
the
software
component.

Serial
Number
(
70).
Report
the
serial
number,
if
applicable
 
otherwise
leave
blank.

RT
531
Parameter
(
10).
Report
"
NORX"
as
the
parameter
monitored.
(
You
should
report
one
531
record
for
each
fuel
type.)

Default
Value
(
14).
Report
the
fuel­
specific
maximum
potential
NO
x
emission
rate
(
MER),
in
units
of
lb/
mmBtu.

Units
of
Measure
(
27).
Report
"
LBMMBTU".

Purpose
or
Intended
Use
(
34).
Report
"
MD"
for
missing
data.

Type
of
Fuel
(
37).
Report
the
fuel
type
code
for
the
fuel.
(
See
the
EDR
Instructions
for
RT
531
for
the
list
of
available
codes.)

Indicator
of
Use
(
40).
Report
"
A"
for
any
hour.

Source
of
Value
(
41).
Report
"
DEF"
for
default
value.

RT
550
Parameter
(
10).
Report
"
NOX".

Monitoring
System
ID
(
14).
Report
the
monitoring
system
ID,
from
RT
510,
of
the
NOX
PEMS
system.

RT
556
Component
ID
(
10).
Report
the
PEMS
component
ID
subject
to
recertification/
diagnostic
7
testing,
if
a
specific
component
is
involved.
If
the
event
is
system,
not
component,
specific,
leave
this
field
blank.

Monitoring
System
ID
(
13).
Report
the
monitoring
system
ID,
from
RT
510,
of
the
NO
x
PEMS
system.

Event
Code
(
16).
Report
code
"
99"
(
i.
e.,
"
Other").

Code
for
Required
Test
(
19).
Codes
for
PEMS
systems
are:

80
PEMS
daily
QA/
QC,
sensor
validation
system
check,
train
or
retrain
(
if
manufacturer
recommends),
sensor
validation
compliance
alarm
check,
statistical
tests,
and
normal
operating
level
RATA
and
bias
test;

81
PEMS
daily
QA/
QC,
and
PEMS
check
with
reference
method
or
portable
analyzer;

Beginning
of
Conditionally
Valid
Period
(
31,
39).
If
conditional
data
validation
is
used,
report
the
date
and
hour
that
the
probationary
PEMS
daily
QA/
QC
test
was
successfully
completed
according
to
the
provisions
of
§
75.20(
b)(
3)(
ii).

Note:
For
PEMS,
you
may
only
use
conditional
data
validation
if
the
"
event"
in
column
16
requires
RATA
testing.
If
you
elect
to
use
conditional
data
validation,
you
must
complete
the
RATA
within
the
allotted
time
in
§
75.20(
b)(
3)(
iv).

RT
585
Parameter
(
10).
Report
"
NOXR"
as
the
parameter
code
associated
with
the
PEMS.
Report
one
RT
585
for
each
generic
fuel
type
combusted.

Monitoring
Methodology
(
14).
Report
"
PEMS"
as
the
monitoring
methodology
for
the
PEMS.

Missing
Data
Approach
for
Methodology
(
28).
Report
"
FSP75"
for
the
fuel­
specific
missing
data
approach
for
the
PEMS
methodology.

RT
610
Units
of
Measure
(
33).
Report
"
2"
(
lb/
mmBtu)
as
the
units
of
measure.

Value
from
CEM
System
Being
Tested
(
34).
Report
the
average
value
recorded
by
the
PEMS,
for
each
RATA
run.

RT
611
Units
of
Measure
(
34).
Report
"
2"
(
lb/
mmBtu)
as
the
units
of
measure.

Arithmetic
Mean
of
CEM
Values
(
35).
Report
the
arithmetic
mean
of
all
the
RTs
610
PEMS
values
associated
with
the
RATA.
8
Number
of
Load
Levels
Comprising
Test
(
133).
Report
"
1"
or
"
3"
(
if
certification
or
recert).

BAF
for
a
Multiple­
Load
RATA
(
134).
Leave
this
field
blank.

RT
624
Component
ID
(
10).
Report
the
PEMS
software
component
ID
from
RT
510.

Monitoring
System
ID
(
13).
Report
the
NO
x
monitoring
system
ID
from
RT
510.

Parameter
(
16).
Report
"
NOX".

QA
Test
Activity
Description
(
30).
Fill
in
appropriately.

Reason
for
Test
(
51).
Report
"
Q".

QA
Test
Code
(
53).
Report
one
of
the
following
codes,
as
appropriate:

04
PEMS
daily
QA/
QC
05
Periodic
check
of
PEMS
accuracy
with
a
portable
analyzer,
or
reference
method
RT
910
Text
(
4).
Briefly
describe
the
PEMS.
