Analysis of Ozone Season NOx Emissions Data for Coal-Fired
EGUs in Four Mid-Atlantic States
EPA Clean Air Markets Division1
December 2020
1 Introduction
The Ozone Transport Commission (OTC) is recommending that EPA require Pennsylvania
to set daily nitrogen oxides (NOx) emissions limits during the ozone season for coal-fired
electric generating units (EGUs) with already-installed selective catalytic reduction (SCR)
or selective non-catalytic reduction (SNCR) controls. Further, the OTC recommends that
these limits be as stringent as limits set by Delaware, Maryland, and New Jersey. The
purpose of this analysis is to explore recent NOx emission rates and trends at coal-fired
EGUs equipped with SCR or SNCR in Pennsylvania, Delaware, Maryland, and New Jersey.
For comparison, coal-fired EGUs not equipped with SCR or SNCR are also included in this
analysis.
This analysis relies on EPA Clean Air Markets Division's (CAMD's) Power Sector Emissions
Data. The data used in this analysis are hourly data from the 2009-2019 ozone seasons for
coal-fired EGUs in these four states, with some additional focus on the more recent ozone
seasons 2015-2019. Partial hours of operation are excluded from this analysis. Hours in
which substitute data were submitted for heat input, NOx emissions rate, and/or NOx mass
emissions have also been removed from this analysis.2 In this analysis, 5,370 substitute
data observations were removed, accounting for just 0.44% of the data.
1 Questions about this analysis should be directed to Justine Huetteman, EPA Clean Air
Markets Division, huetteman.justine@epa.gov. The data sets used in the analysis and the
code used to produce the unit-specific figures in section 3.1 of this analysis are available
upon request.
2 Substitute data are required to be submitted in instances in which the monitor is not
available or is not performing properly, or a quality assurance test of the monitoring
equipment was missed. These substitute data values are usually similar to actual
emissions; however in some rare cases, they may be significantly higher than actual
emissions, sometimes representing the EGU's maximum potential emissions.
1
-------
2 Overview of Recent NOx Emission Rates and NOx Mass Emissions
Average ozone season NOx emission rates at coal-fired EGUs with SCR have decreased
significantly in Pennsylvania since 2015, while average emission rates at SCR-equipped
EGUs in Delaware, Maryland, and New Jersey have been relatively consistent (see Table
2.1). In 2017, a more stringent emissions budget for the units in Pennsylvania (as well as
Maryland and New Jersey) was implemented under the Cross-State Air Pollution Rule
(CSAPR) Update, and Pennsylvania units also became subject to more stringent NOx RACT
requirements. After being higher than the average emission rates for all three of the other
states in 2015-2016, in 2017-2018 the average NOx emission rates for the Pennsylvania
units were below the average rate for the New Jersey units but above the average rates for
the Delaware and Maryland units. The average emission rate for the Pennsylvania units
increased above the average rate for the New Jersey units in 2019 but remained well below
2015-2016 levels.
Table 2.1. Weighted Average Ozone Season NOx Emission Rates at Operating SCR-equipped
Coal-fired EGUs (Ib/mmBtu)
Year
Pennsylvania units
Delaware units
Maryland units
New Jersey units
2015
0.252
0.094
0.059
0.117
2016
0.233
0.078
0.058
0.111
2017
0.099
0.084
0.057
0.112
2018
0.102
0.086
0.064
0.112
2019
0.130
0.082
0.060
0.108
t Each weighted average emission rate is computed as the sum of ozone season NOx
emissions for the group of units divided by the sum of ozone season heat input for the
group of units.
With respect to NOx mass emissions, during the 2015-2019 ozone seasons, Pennsylvania
coal-fired EGUs with SCR produced significantly more NOx emissions than coal-fired EGUs
with SCR in Delaware, Maryland, and New Jersey (see Figure 2.1). The Pennsylvania coal
fleet also operated for more operating hours (see Figure 2.2) and at a higher level of heat
input (see Figure 2.3) than the other states during these ozone seasons. In 2017,
Pennsylvania NOx emissions declined sharply, by more than 50% over the previous year.
The decrease is primarily due to generally lower NOx emission rates, although a reduction
in operating hours and heat input also contributed to the decrease. The reduction in NOx
emission rates could be due to more stringent emission rate limits set under the
Pennsylvania RACT II standard and/or more stringent NOx mass emission budgets set
under the CSAPR Update, both of which took effect in 2017. Nevertheless, despite the
decreases since the 2015-2016 ozone seasons, during the 2019 ozone season Pennsylvania
coal-fired EGUs with SCR emitted more than six times the amount of NOx as the SCR-
equipped units in the other three states combined. While one important reason
Pennsylvania's 2019 emissions were higher than the emissions of the other states' units is
the generally higher heat input consumed by the Pennsylvania units, differences in
emission rates also contributed, with Pennsylvania's SCR-equipped units reporting an
average NOx emission rate during the 2019 ozone season nearly twice the average NOx
emission rate for SCR-equipped units in the other three states (see Table 2.2).
2
-------
Figure 2.1: Total Ozone Season NOx Emissions by State for Coal-fired EGUs with SCR
2i ,
15.00C
£Z
O
¦4->
m
sz
0
1 10,000'
E
HI
o*
5,000
State
DE
MD
NJ
PA
Year
Figure 2.2: Total Ozone Season Operating Hours by State for Coal-fired EGUs with SCR
30,000
U
0
X
CD
c
1
m
CL
O
20,000
10,000
1.
A ¦$>
tp cp
# A ^ ,# ^
qjj
j|s cp rp cp
#
State
OE
MD
NJ
PA
Year
3
-------
Figure 2.3: Total Ozone Season Heat Input by State for Coal-fired EGUs with SCR
150,000,000 -
p 100,000,000
=J
CL
£=
I 50,000,000-
o-
DE
MD
NJ
PA
A A iSi iS
,£><£><£>,£>(£> if1 <•£> <£><£> <£>
Year
State
¦
DE
MD
NJ
PA
Table 2.2:2019 Ozone Season Summary ofNOxMass Emissions and NOxEmission Rates at
Coal-fired EGUs with SCR
State NOx Emissions (tons) Average NOx Rate (lb/mmBtu)
DE 48 0.082
MD 881 0.060
NJ 360 0.108
PA 8,156 0.130
3 NOx Emission Rates and NOx Mass Emissions by Operating Level
for Individual EGUs
This section of the analysis focuses on NOx emission rates in the context of operating level,
which can have an important impact on SCR performance and therefore NOx emission rates
and mass emissions. The patterns of operation of some coal-fired EGUs have been changing
in recent years, and considering operating level allows any effects of these changes on NOx
emission rates and the operation of SCR controls to be taken into account. At lower levels of
generation, the SCR may not be effective because temperatures may be too low to promote
the catalytic reaction. Generally, the catalyst works more effectively as temperature in the
unit increases with increasing heat input and correspondingly increasing steam and
electrical output.
4
-------
The plots in this section compare ozone season NOx emission rates, NOx mass emissions,
and operating hours for individual coal-fired EGUs for either two or three years during the
2009-2019 period. Data for the year 2019 are always included for a recent snapshot. The
other years shown are the years during the 2009-2019 period in which the unit achieved
its lowest and third lowest average ozone season NOx emission rates.3 If plots are provided
for only two years, then 2019 represents the year of either the unit's lowest or third-lowest
average emission rate during the 2009-2019 period.
This analysis uses heat input as a proxy for operating level. Specifically, this analysis uses a
heat input factor. Heat input factor was calculated for each hour by dividing the heat input
in a given hour by the individual EGU's observed maximum hourly heat input during the
ozone seasons in the 2009-2019 period. The hourly emission rate data are then divided
into 10 bins according to the heat input factor for the hour (bin 1 contains data for all hours
in which the heat input factor for the hour is between 0-10% of the unit's maximum hourly
heat input, bin 2 contains data for all hours in which the heat input factor is between 10-
20% of the unit's maximum hourly heat input, etc.).
The hourly emission rate data values for each heat input factor bin are displayed by means
of a "box plot" or "box-and-whisker plot." Each "box" represents the middle half of all the
hourly data values in that heat input factor bin - that is, the hourly data values that fall in
the "interquartile range" between the 25th percentile and 75th percentile hourly data
values. The horizontal line in the box represents the median hourly data value. Vertical
lines, or "whiskers," extend to the highest and lowest hourly data values that fall above or
below the top or bottom edges of the box within a distance of up to 1.5 times the
interquartile range. Any outlying hourly data values that fall above or below the top or
bottom edges of the box by a distance of more than 1.5 times the interquartile range are
shown as individual dots. Thus, a lower median data value and lower overall placement of
the box on the chart indicate generally lower hourly emission rates, while shorter vertical
distances between the top and bottom edges of the box and between the top and bottom
ends of the whiskers, as well as fewer outliers, indicate lower variability (or greater
consistency) of a unit's hourly emission rates at a given heat input factor bin. In this way,
each box plot provides visual representations of both the magnitude and variability of a
unit's hourly NOx emission rates at a given heat input factor bin in a single chart.
In all of the emission rate charts in this section, a horizontal dashed line showing the 0.12
lb/mmBtu emission rate limit that Pennsylvania's SCR-equipped units are required to meet
3 Data for the ozone season with each unit's third-lowest emission rate during the 2009-
2019 period are included for comparability with the data sets EPA has used to identify
emission reduction opportunities in the CSAPR Update and the proposed Revised CSAPR
Update. Data for the ozone season with each unit's lowest emission rate during the same
period are also included for greater comparability with the data provided by the OTC to
support its CAA section 184(c) recommendation. EPA notes that for some units the OTC has
provided data for ozone seasons before 2009.
5
-------
- under certain operating conditions, on a 30-day rolling average basis - by the state's
current RACT rules is included for comparison purposes.
3.1 Unit-Level NOx Emission Rates and NOx Mass Emissions
3.1.1 Units with SCR Controls
The plots for SCR-equipped EGUs in this subsection show that most of the units have
generally achieved lower NOx emission rates as operating level increases. Moreover, many
of the units have achieved low NOx emission rates at low- and mid-operating levels, not just
at high-operating levels when temperature is expected to be highest. Many of the EGUs
have shown changes in the numbers of hours spent at the various operating levels over
time. For some of the units, the relationship between NOx emission rates and operating
levels has remained generally consistent across the years evaluated, while for others, the
relationship of NOx emission rates to operating levels appears to have changed in recent
years, as discussed below.
Some of the SCR-equipped EGUs analyzed show different relationships of NOx emission
rates to operating levels in the 2019 ozone season than in the other ozone seasons
evaluated:
• Conemaugh units 1-2 and Keystone units 1-2 in Pennsylvania appear to have operated
somewhat differently in the 2019 ozone season than in the earlier ozone seasons. These
units show generally somewhat higher NOx emission rates and greater variability in
NOx emission rates than the other SCR-equipped EGUs in this analysis, particularly in
the 2019 ozone season. In their years with the lowest average ozone season NOx
emission rates in this analysis, these EGUs had relatively low NOx emission rates at mid-
and high-operating levels; moreover, there was little variability in NOx emission rates at
these operating levels. However, during the 2019 ozone season, these EGUs had higher
NOx emission rates and greater variability in NOx emission rates across operating levels
than in the past, particularly at mid-operating levels. Conemaugh units 1-2 had
relatively low NOx emission rates in bins 5-7 during the 2018 ozone season, but in 2019
had much higher NOx emission rates and higher variability in these bins. The units also
operated more frequently in these bins (particularly bin 5). Keystone units 1-2 have
consistently shown higher NOx emission rates in bins 4 and 5 compared to other
operating levels, but in 2019 the units operated more frequently in these bins
compared to the earlier ozone seasons. Conemaugh units 1-2 and Keystone units 1-2
are the largest individual units of the 39 units in Delaware, Maryland, New Jersey, and
Pennsylvania included in this analysis, and while their total heat input has decreased
somewhat across the years evaluated, the extent of the decrease has been less than for
many of the other units analyzed. Partly because of their increased share of the
collective heat input of the 39 units and partly because of their relatively higher
emission rates, in the 2019 ozone season the collective NOx mass emissions of these
four units (6,215 tons) exceeded the collective NOx mass emissions of the other 35 units
included in the analysis (5,213 tons).
6
-------
• Homer City units 1-3 in Pennsylvania showed generally lower and less variable NOx
emission rates in the 2019 ozone season at mid- and high-operating levels compared to
earlier ozone seasons. Units 1 and 2 operated for fewer hours at high-operating levels
in the 2019 ozone season compared to the earlier years, which contributed to their
decrease in NOx mass emissions. On the other hand, Unit 3 operated for more hours
across most operating levels compared to 2017 and thus emitted more NOx.
• Chalk Point unit 1 in Maryland showed relatively higher NOx emission rates during the
2019 ozone season at a range of operating levels compared to past ozone seasons but
also operated much less frequently.
The remaining SCR-equipped EGUs analyzed show more consistent relationships of NOx
emission rates to operating levels across the different ozone seasons evaluated:
• Brandon Shores units 1-2, Morgantown units 1-2, and Wagner unit 3 in Maryland and
Carneys Point units 1001-1002 and Logan unit 1001 in New Jersey consistently show
low NOx emission rates starting around bins 3 and/or 4 and continuing through the
mid- and high- operating levels. The Maryland EGUs have some of the lowest NOx
emission rates and lowest variability of NOx emission rates across operating levels in
this analysis. In this group, the Brandon Shores, Morgantown, and Wagner units
operated less frequently and thus had much lower total NOx mass emissions in the 2019
ozone season, particularly at high operating levels, than they did in the past, while the
other EGUs operated at similar levels (or, for the Carneys Point units, at higher levels)
in the 2019 ozone season.
• Indian River unit 4 in Delaware and Montour units 1-2 in Pennsylvania generally show
low NOx emission rates starting at about bin 5 across all years analyzed but also
operated much less frequently in the 2019 ozone season than they did in the earlier
years and therefore had lower total NOx mass emissions. NOx emission rates at these
EGUs appear to have ticked up slightly in recent ozone seasons at mid- and high-
operating levels, though these differences are minor as rates remain low relative to
other units.
• Cheswick unit 1 in Pennsylvania shows NOx emission rates at various operating levels
that have been consistent or decreased slightly over the years evaluated, although the
emission rates achieved generally have not been as low as many of the other SCR-
equipped EGUs in this analysis. The most notable change in the data for this unit across
the years evaluated has been a shift in operating hours from the high operating levels
(bins 8 and 9) where the unit achieves its lowest NOx emission rates to mid-operating
levels (particularly bin 4) where the unit's NOx emission rates have been significantly
higher. As a result of this shift in operating patterns, in the 2019 ozone season, the unit
emitted most of its NOx mass emissions in bin 4.
7
-------
Indian River Unit 4 DE
1.000
fi
Z3
O
X
CT»
C"
*3
Q
o
500
01 02 03
Or 06 07
Heat Input Factor Bin
08 09
Indian River Unit 4 DE
Year
M 2012 (fewest)
2 u16 (3rd lowest)
.... 13
1 *TTT+^-
m
c
o
-*-»
m
c
o
'm
m
E
til
<$
01 02 03 04 05 06 t
2012 (lowest;
i 2015 (3rd lowest:
.
05 of; -r oe >-i
Heat Input Factor Bin
2019
03 04 05 06 07 08 03 10
8
-------
Brandon Shores Unit 1 MD
1,000
01 02 03 04 05 06 0? 08 09
Heat Input Factor Bin
Brandon Shores Unit 1 MD
E o
|°
s 0
K o.,
s „„
m
c
o
-*-»
m
c
o
t/i
m
E
til
<$
Year
M 2015 (3rd lowest)
17 (lowest)
.... 13
2015 (3rd lowest;
2017 (lowest;
2019
i . : •
i -Hi;;,
j!j: .•
I l:4-i i •
• •
if ir-++-
01 02 03 04 05 06 07 08 09
¦ ' 0
llilll
•V1 02 0? C4 05 06 07 08 09
01 02 03 0+ 05 06 07 08 09
01 02 03 04 05 06 07 08 09 01 01 >- 0: >6 07 >0 09 01 02 03 04 05 06 07 08 09
Heat Input Factor Bin
9
-------
Brandon Shores Unit 2 MD
Year
III 17 (3rd lowest)
:'19 (lowest)
01 02 03 04 Of Oo 07 08 09
Heat Input Factor Bin
Brandon Shores Unit 2 MD
S 0.2 i j a i
a r. • • -1
01 02 03 04 05 06 0? 0B 09 10 01 02 03 04 05 06 0? 08 09 10
-------
Chalk Point Unit 1 MD
1,500
J
Year
111 2014 (fewest)
2 u 18 (3rd lowest)
...13
01 02 03
04 05 06 07
Heat Input Factor Bin
08
09 10
Chalk Point Unit 1 MD
£11 02 03 M 05 06 07 08 09 10 01
» 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
11
-------
Herbert A Wagner Unit 3 MD
1,250
Year
M 2010 (3rd lowest)
2015 (lowest)
...13
Heat Input Factor Bin
Herbert A Wagner Unit 3 MD
E 0
E
It0
©
cr 0
d> ,,
m
c
o
m
c
.9 4
m
m
£ ?
hi ^
oK
2010 (3rd lowest;
201s (lowest;
2019 I
n^aj
1 • 1
•
I :
1: !
H-1 m m |
• 1
| rh \
+Hi4+
L-r-' •
- . 1
-f iff
oi ;j,:: o- oe o-; " os j? 10 01 02 03 of -'-s :p :)i, y* i-j 0: oi 0- oe >3 " v.:
2010 (,3rd fewest;
| 2015 (lowest;-
2019 |
ill
Cr
4 05 06 07 08 09 10
37 08
Heat Input Factor Bin
12
-------
Morgantown Unit 1 MD
1,500
Morgantown Unit 1 MD
m o
E
E o
_Q
r 0
ts
en a
O' „
Year
M 2013 (lowest)
2014 (3rd lowest)
...13
01 02 03 04 05 06 07 08 09 10
Heat Input Factor Bin
2013 (lowest;
2014 (3rd lowest;
~ I
M 1
I
i ^ '
• • • „
i li+++ I
2019
1
I
• •
m
c
o
m
g 4
tfi
m
E 2
HI
oK
2013 (lowest;
i 2014 (3rd lowest:
2019
L
,nl
Heat Input Factor Bin
13
-------
Morgantown Unit 2 MD
1,500
£
Z3
X 1,000
o>
C"
*3
Q
o
500
o-
Year
M 2010 (3rd lowest)
2011 (lowest)
...13
01 02 03 04 05 06 07 08
Heat Input Factor Bin
09 10
Morgantown Unit 2 MD
E
E
_Q
0.4
0 3-
® 0
ts
Of 0
° „
I .
—i r r—
1 •
• • •
+UU
• • •
-j
)1 02 03 04 05 06 07 08 09 10
m
c
o
150
m
1100
m
| 50
-i 1 1 r~
...ll
£11 02 03 M 05 06 07 08 09 10 01 02 OS -i- vc :«j 05 !•) 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
Note: Morgantown unit2 has one hourly NOx emission rate above 2.0 Ib/mmBtu in bin 6 in
2011 that is not shown here for visualization purposes.
14
-------
Carneys Point Unit 1001 NJ
3,000
Year
III 17 (3rd lowest)
:'19 (lowest)
01 02 03 04 05 06 0?
Heat Input Factor Bin
Carneys Point Unit 1001 NJ
| 0,3
_Q
r °-Z
ts
en
(f
0,1
I
i+r-
01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09
Heat Input Factor Bin
15
-------
Carneys Point Unit 1002 NJ
3,000
fi
Z3
O
X
CT»
C"
*3
2,000
Q
o
1,000
Year
III 15 (3rd lowest)
:'19 (lowest)
01 02 03 Or 06 07 08 09 10
Heat Input Factor Bin
Carneys Point Unit 1002 NJ
E
0.4
E
0,3
_Q
©
0,2
S
en
0,1
X
O
z
0.0
Ttf
~
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 09 10
m
c
7
m
sz
.9 5
w
m
111
oK
2019 (lowest;
04 05 06 07 08 V> 01 02 03 04 05 06 07 08 09 10
Heat Input Factor Bin
Note: Carneys Point unit 1002 has one hourly NOx emission rate above 2.0 Ib/mmBtu in bin 6
in 2015 that is not shown here for visualization purposes.
16
-------
Logan Generating Plant Unit 1001 NJ
3,000
Year
1111 2015 (fewest)
2017 (3rd lowest)
...13
C;- Or 06 07
Heat Input Factor Bin
08 09
Logan Generating Plant Unit 1001 NJ
E
fc
0.2-
jLI
m
.
S 0-1-
£Z
X
U
z
0.0 ~
*5"
1
o
m
r
1
o
m
m
F
50
tu
rf
z
U
01 02 03 04 05 00 07 08 09 10 01 02 03 04 05 06 07 08 03 10 01 02 03 04 05 06 07 08 09 10
.1
01 02 03 M 05 06 07 08 09 10 01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
17
-------
Cheswlck Unit 1 PA
600
£
=3
• 400
o>
c
+3
o.
O
200
il
Year
1111 2009 (lowest)
2 u 1B (3rd lowest)
.... 13
01 02 03 04 Of :>5 07 08 09 10
Heat Input Factor Bin
Cheswlck Unit 1 PA
E o
|.
® o
OT
o
5?.
m
g 3
-*-»
m
o ^
tfi
m
I1
2009 (lowest;
2013 (3rd lowest;
2019
1 •
=!= •
4- '
01 02 03 04 05 06 07 08 09 10
2009 ilowest;
.
2\J 18 13rd h5rA'est.,
I
2019
L
01 02 03 M 05 06 07 08 09 10 01
» 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
18
-------
Conemaugh Unit 1 PA
2,000
1,500
fi
Z3
O
X
1,000
Q
o
500
01 02 03 Or 06 07 08 09 10
Heat Input Factor Bin
Conemaugh Unit 1 PA
Year
III ;."1S (lowest)
:'19 (3rd lowest)
01 02 03 04 05 06 0? 0B 09 10 01 02 03 04 05 06 0? 08 00 10
2 000
o
m
g 400
m
m
E 2
til
04 05 06 07 08
06 07 08 09 10
Heat Input Factor Bin
19
-------
Conemaugh Unit 2 PA
01 02 03
C;- Or 06 07
Heat Input Factor Bin
Conemaugh Unit 2 PA
£ 0.4
E o ->
jQ
m 0
ts
a 0
° 0
tJl
j i i : :
5-
Year
M 2016 (3rd lowest)
2 u 18 (lowest)
...13
H
03 10 01 02 03 04 05 06 07 08 09 10
2018 (lowest:
01 02 03 M 05 06 07 08 09 10 01 02 0' 0- vc -"3 05 •>;> M 0 :
Heat Input Factor Bin
06 07 08 03 10
20
-------
Homer City Unit 1 PA
750
fi
Z3
O
X
o>5
C"
*3
Q
o
250
o-
01 02 03 04 Of 0-5 07 08 09 10
Heat Input Factor Bin
Homer City Unit 1 PA
Year
III ;."10 (3rd lowest)
:'19 (lowest)
¦ •
! i i i i
01 02 03 04 05 06 07 0B 09 10 01 02 03 04 05 06
38 00 10
3i
II..
01 02 03 04 07 08
04 05 06 07 08 09 10
Heat Input Factor Bin
21
-------
Homer City Unit 2 PA
1,500
Z3
o
X
CT»
C"
*3
1,000
Q
o
500
01 02 03
Or 06 07
Heat Input Factor Bin
Homer City Unit 2 PA
Year
III ;."10 (3rd lowest)
:'19 (lowest)
E
E
_Q
©
S
en
d
0.1
o.o
2010 (3rd lowest;
2019 (lowest;
•
•
1
^E±3_cb | i 1 |
_gT ;
01 02 03 04 05 06 0? 0B 09 10
06 0? 08 09 10
m
I*-
m
sz
.2 4
w
m
111
oK
01 02 03 04
07 08
06 07 08 09 10
Heat Input Factor Bin
22
-------
Homer City Unit 3 PA
Year
III 17 (3rd lowest)
:'19 (lowest)
03 05 06
Heat Input Factor Bin
Homer City Unit 3 PA
CQ 0.3
E
E
S 0.2
J °-1i
d
o.oi
. ^
I :
01 02 03 04 05 06 07 08 01 02 03 04 05 06 07 08
10 100
£ 50
til
oK
01 02 03 04 05 06 07 OS 0! :>2 03 04 05 06 07 08
Heat Input Factor Bin
23
-------
Keystone Unit 1 FA
3,000
£ 0.3
E
_Q
=• 0
®
S
CH 0
d
m
c
o
ii 7
m
c
.2 K
m "
m
E „
tu ^
Keystone Unit 1 PA
f=a *
M
Year
1111 2010 (fewest)
2017 (3rd lowest)
...13
01 02 04 0' :-6 07 08 09 10
Heat Input Factor Bin
~
01 02 04 05 06 0? 08
01 02 04 05 06 0? 08 09 10 01 02 04 05 00 07 08 09 10
2010 (lowest;
2017 (3rd lowest:
_
2010
01 02 04 05 06 07 08 09 10 01
01 02 04 05 06 07 08 09 10
Heat Input Factor Bin
24
-------
Keystone Unit 2 FA
2,000
fi
=3
O
X
U)
sz
+3
Q.
O 1,000
01 02 03 0- 1=5 06 07 03 09
Heat Input Factor Bin
Keystone Unit 2 PA
§ 0
O 0
2009 (3rd lowest;
2010 (lowest;
• 0 . i'
I ' ^ [±3 P3 | ! j
: 1 1 \ f
¦t -x-a-g-
• 1 •
:
• i •
+ "-T+A"
Year
M 2009 (3rd lowest)
2 u 10 (lowest)
.... 13
2019
01 02 03 04 05 06 07 08 09 01 02 03 04 05 00 07 08 09 01 02 03 04 05 06 07 08 09
m
g 6i
m
o 4
m
m
E 2
til *
)6 07
01 02 03 04 05 06 07 08 00
Heat Input Factor Bin
25
-------
Montour, LLC Unit 1 PA
01 02 03 Or 06 07 08 09 10
Heat Input Factor Bin
Montour, LLC Unit 1 PA
E o
E o
_Q
— 0
®
ffl 0
en
o
0 1 -
0-0
2009 (lowest;
2013 (3rd lowest;
1 i •
• 1 • 1
y
. $04. i j' ;.:
0 Ji'+i
Year
1111 2009 (fewest)
2 u 18 (3rd lowest)
...13
2019
- .I; !
m
c
o
-*-»
m
c
o
t/i
m
oK
2009 (lowest;
2018 (3rd lowest:
i.lll
2010
01 02 03 M 05 06 07 08 09 10 01
» 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
26
-------
Montour, LLC Unit 2 PA
Year
III ;."18 (3rd lowest)
:'19 (lowest)
01 02 03 04 Of :>5 07 08 09 10
Heat Input Factor Bin
Montour, LLC Unit 2 PA
2018 (3rd lowest;
; ¦
i ; : I
. . •
^44^,1
2019 (lowest;
¦¦¦-I
01 02 03 04 05 06 07 03
03 ... 06 07 08 09 10
Heat Input Factor Bin
27
-------
3.1.2 Units with SNCR Controls
The plots for SNCR-equipped4 EGUs in this subsection show that some of the units have
achieved NOx emission rates comparable to the rates achieved by SCR-equipped units.
Some of the units show NOx emission rates that vary or increase with operating level while
others have relatively flat NOx emission rates across operating levels.
The SNCR-equipped EGUs analyzed generally show fairly consistent relationships of NOx
emission rates to operating levels across the different ozone seasons evaluated:
• Colver unit AAB01 in Pennsylvania shows NOx emission rates that do not appear to
increase significantly with operating level and do not have any clear trend of changes
over time. The unit operated for a similar number of hours in the 2019 ozone season as
it has previously. This unit accounted for the largest amount of NOx emissions in the
2019 ozone season (309 tons) of any of the SNCR-equipped units analyzed.
• Scrubgrass units 1-2 in Pennsylvania show NOx emission rates that do not appear to
increase significantly with operating level and do not have any clear trend of changes
over time. These EGUs operated for fewer hours in the 2019 ozone season than in the
earlier ozone seasons evaluated, contributing to lower NOx emissions.
• Chalk Point unit 2 in Maryland shows NOx emission rates that generally increase with
operating level and are generally consistent over time. The unit's emission rates are
generally the highest of the SNCR-equipped units analyzed, but in the years evaluated
the unit operated in fewer hours than most of those other units. Because of the smaller
amount of operating hours, the unit's 2019 NOx mass emissions were lower than the
emissions from Colver.
• AES Warrior Run unit 001 in Maryland shows NOx rates that increase slightly with
operating level but generally remain low at all levels, and the rates are consistent over
time. In the 2019 ozone season, the unit operated at about the same operating level as it
has in previous ozone seasons.
• Northampton unit NGC01, Panther Creek units 1 and 2, and Seward units 1 and 2 in
Pennsylvania show NOx emission rates that generally increase with operating level and
are generally consistent over time. In the 2019 ozone season, these EGUs all operated
4 The Colver, Northampton, and Panther Creek units report "ammonia injection" controls;
EPA has grouped these units with the units reporting SNCR controls. Chalk Point unit 2
reports selective autocatalytic reduction (SACR) controls, which rely on simultaneous
injection of ammonia and hydrocarbons (such as natural gas) to reduce NOx through a
catalytic reaction without a separate catalyst; EPA has also grouped this unit with the
SNCR-equipped units for purposes of this analysis. Indian Point unit 4 reports both SCR and
SNCR controls; EPA has grouped this unit with the SCR-equipped units.
28
-------
for fewer hours across mid- and high-operating levels than they have in the past,
contributing to lower NOx emissions.
AES Warrior Run Unit 001 MD
2,000
=i
0
X
01
c
"4=i
1,500
ffi 1,000
a.
o
500
Year
BB 3 (3rct lowest)
' 15 (lowest)
::i9
05 06 07
Heat Input Factor Bin
AES Warrior Run Unit 001 WD
01 £k 00 07 OS 09 10
o: 03 05 06 07 08 09 10 01 02 03 05 06 07 08 09 10
Beat Input Factor Bin
29
-------
Chalk Point Unit 2 MD
300
fi
Z3
O
X .
CT»-
C"
*3
200
Q
o
100
04 Of U-5 07
Heat Input Factor Bin
Chalk Point Unit 2 MD
| 0.3
_Q
— 0.2
I 0,:
d
• jl
o.o-J
^; i a|
jUai
W-TZ—
"r
i
sJTT^ji
Year
M 2017 (3rd lowest)
2 u 1B (lowest)
.... 13
• i I
y-
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 03 10 01 02 03 04 05 06 07 08 09 10
v> A
£= 4
o
t/i
m
E 2
111
oK
J.
Cr 4 05 06 07 08 09 10
37 0?
Heat Input Factor Bin
30
-------
Colver Green Energy Unit AAB01 PA
3,000
=3
O
X
U)
sz
+3
2,000
° 1,000
Year
M 2010 (3rd lowest)
2011 (lowest)
...13
01 02 03 04 05 06 07
Heat Input Factor Bin
Colver Green Energy Unit AAB01 PA
08 09 10
E
0.
E
0,
_Q
©
0.
S
en
0.
X
O
z
0.
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 00 07 08 03 10 01 02 03 04 05 06 07 08 09 10
m 3i
o
2 2I
o
m
E 1i
til
.1
01 02 03 M 05 06 07 08 09 10 01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
Note: Colver unitAABOl has one hourly NOx emission rate above 1.9 Ib/mmBtu in bin 2 in
2011 that is not shown here for visualization purposes.
31
-------
Northampton Generating Plant Unit NGC01 PA
1,000
=3
O
X
U)
sz
+3
g- 500
01 02 03 04 05 06 07 08 00 10
Heat Input Factor Bin
Northampton Generating Plant Unit NGC01 PA
E
E
_Q
0.2
0.1
ts
en
d
0.0
Year
M 2013 (3rd lowest)
2 u 1B (lowest)
.... 13
|0
• • •
01 02 03 04 05 00 07 08 09 10 01 02 03 04 05 06 07 08 03 10 01 02 03 04 05 06 07 08 09 10
m
J3 60
m
.9 40
w
m
111
oK
20
Cr
¦)7 08 09 10
Heat Input Factor Bin
32
-------
Panther Creek Energy Facility Unit 1 PA
200
=3
O
X
U)
sz
+3
150
100
Q.
o
50
_.i
h L
01 02 03 04 Of 0-5 07 08
Heat Input Factor Bin
Panther Creek Energy Facility Unit 1 PA
09 10
Year
M 201? (lowest)
2013 (3rd lowest)
...13
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 03 10 01 02 03 04 05 06 07 08 09 10
m
c
o
-*-»
m
sz
o
tfi
m
E
til
01 02 03 04 05 06 07 08 09 10 01 0;
01 02 03 04 05 06 07 08 OS 10
Heat Input r actor 3iri
33
-------
Panther Creek Energy Facility Unit 2 PA
03 04 05 0)3 07 08 09 10
Heat Input Factor Bin
Panther Creek Energy Facility Unit 2 PA
Year
15 (lowest)
:'19 (3rd lowest)
03 04 05 06 07 08 09 10 03 04 05 06 07 08 09 10
m
c 30
o
0 ^
(f)
m
1 10
oK
03 04 vr 06 07 08 08 10 03 -j-
Heat Input Factor Bin
:•? 08 09
34
-------
Serubgrass Generating Plant Unit 1 PA
3,000
2,000
=3
O
X
U)
sz
+3
Q.
o 1,000
01 02 03 06 07 08 09 10
Heat Input Factor Bin
Serubgrass Generating Plant Unit 1 PA
® 0'
E
o
1 0
9%
ujjXjj, gig c
Year
1111 2009 (fewest)
2015 (3rd lowest)
...13
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 00 07 08 03 10 01 02 03 04 05 06 07 08
m
c
o
-*-»
m
sz
o
tfi
m
75
50
E «
HI ^5
oK
! : 1 r~
01 02 03 04 05 06 07 08 09 10
01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
35
-------
Serubgrass Generating Plant Unit 2 PA
2,000
fi
Z3
O
X
CT»
C"
*3
1,500
1,000
Q
o
500
01 02 03 15 08 0? 08
Heat Input Factor Bin
Serubgrass Generating Plant Unit 2 PA
E o
E
!. o
©
I°
c 8
o
-*-»
m 6
c
o
'm J
w *
E
Ui 2
2009 (lowest;
2011 (3rd lowest;
»
* (.11
:: ¦""! : : :: "¦¦¦: : ; : ¦: ¦¦¦ : ;¦ ¦¦ ¦: ¦ ::¦ ¦ i ¦::: : ¦: ¦¦¦ :: ¦ ¦: ¦¦ ::
2009 (lowest;
i 2011 (3rd lowest:
I
Year
2008 (lowest)
;:11 (3rd lowest)
.... 13
2019
—i 1 r—
2019
01 02 03 04 05 06 07
01 02 03 04 05 06 07 08 09
Heat Input Factor Bin
36
-------
Seward Unit 1 PA
1,000
750
£
=3
O
X
U)
•e 500
Q
o
250
Year
M 2013 (3rd lowest)
2014 (lowest)
.... 13
01 02 0- 0- 05 06 07 08
Heat Input Factor Bin
Seward Unit 1 PA
E
E
_Q
0.2
® 0,1
en
d
2 100
O
-*-»
w 75
c
o
'm 50
w -J
E
tu 25
o"
z o
' :¦. :: *
• • :
¦ 1 | ¦
I • • !
[ I 1 ? fj, [ [ i_ 1 J-
-- ~r~ ~ ["lj tp
01 02 03 04 05 06 07 08
01 02 03 04 05 06 07 08
01 02 03 04 05 06 07 08
¦
-I
¦¦ill
.
||H
¦¦¦¦¦
¦-¦ll
01 02 03 04 05 06 07 08 01
01 02 03 M 05 06 07 08
Heat Input Factor Bin
37
-------
Seward Unit 2 PA
Year
M 2012 (fewest)
2013 (3rd lowest)
.... 13
01 02 03 0^ Of 06 0? 08 09
Heat Input Factor Bin
Seward Unit 2 PA
ffi 0.2
E
E
V 0.1
ts
en
d
0.0
T
\:-i
5
T
a?
3^-
01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 00 01 02 03 04 05 06 07 08 09
m
S 90
m
1 *
m
m 30
„.ll
15 06 07 08 09
01 02 03 04 05 06 07 08 09
Heat Input Factor Bin
38
-------
3.1.3 Units Without SCR or SNCR Controls
The plots in this subsection are for coal-fired units without SCR or SNCR controls. Because
there are no longer any coal-fired units operating without SCR or SNCR controls in
Delaware, Maryland, or New Jersey, all the units covered in this section are located in
Pennsylvania. EPA notes that Brunner Island units 1-3, which are capable of combusting
coal and are included in the analysis for that reason, combusted primarily natural gas in
each of the ozone seasons evaluated.
Some of the units without SCR or SNCR controls show NOx emission rates that increase
across operating levels while others show rates that are relatively flat across operating
levels. Most of the units show little change in the NOx emission rates achieved at various
operating levels across the years evaluated. Gilberton units 031-032 are exceptions,
showing different patterns of emission rates and operating levels across the years
evaluated. However, notwithstanding the relative inconsistency across years, the Gilberton
units generally show low emission rates at all operating levels in all the years evaluated.
39
-------
Brunner island, LLC Unit 1 PA
01 02 03 04 Of :>5 07 08 09 10
Heat Input Factor Bin
Brunner Island, LLC Unit 1 PA
Year
M 2017 (fewest)
2 u 18 (3rd lowest)
..13
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 08 07 08 09 10 01 02 03 04 05 06 07 08 09 10
» 60
E 40
E 20
til
4 05 08 07 08 03 10
Heat Input Factor Bin
40
-------
Brunner island, LLC Unit 2 PA
600
fi
=3
O
X
or
c
+3
400
Q.
o
200
04 Of Oo 07
Heat Input Factor Bin
Brunner Island, LLC Unit 2 PA
Year
M 201? (lowest)
2 u 1B (3rd lowest)
...13
2019
. I ¦ • I I
n i ! 1 ' 1
m
C
o
-*-»
m
c
o
t/i
m
E _
til -3
oK
2017 (lowest;
i 2018 (3rd lowest:
'
¦
llllll
2019
..ll
Cr 4 05 06 07 08 09 10 01 02 03 04 05 £16 07 08 09 10
Heat Input Factor Bin
4 05 06 07 08 03 10
41
-------
Brunner island, LLC Unit 3 PA
400
fi
=3
O
X
U)
sz
+3
q"200
Year
III ;."18 (lowest)
:'19 (3rd lowest)
01 02 03 0^ Or 06 0?
Heat Input Factor Bin
Brunner Island, LLC Unit 3 PA
08 09
01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09
m
c
o
t/i
m
tl 25
01 02 03 04 05 06 07
3 04 05 06 07 08 09
Heat Input Factor Bin
42
-------
Ebensburg Power Company Unit 031 PA
3.000
01 02 03 04 05 06 0? 08 09
Heat Input Factor Bin
Ebensburg Power Company Unit 031 PA
m
0
E
0
E
0
®
ts
0
en
0.1
X
O
z
0.0
Year
IB 2011 (fewest)
2013 (3rd lowest)
...13
2011 (lowest;
2013 (3rd lowest;
2019
P
:
1
J • .
1i--l-
01 02 03 04 05 00 07 08 09 01 02 03 04 05 00 07 08 00 01 02 03 04 05 00 07 08 09
m
H 90
is
1 *
m
m 30
¦I.
—,—,—,—
01 02 03 04 05 06 07 08 09
Heat Input Factor Bin
43
-------
Gilberton Power Company Unit 031 PA
3.000
01 02 03 04 05 08 07 08
Heat Input Factor Bin
Gllbertcn Power Company Unit 031 PA
E
E
©
S
en
d
0.1
0.0
r-4
l%4
Year
M 2012 (3rd lowest)
2015 (lowest)
.... 13
09 10
E-3 ' 9 . . [b
04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 03 10 01 02 03 04 05 06 07 08 09 10
m
B 40
m
130
V)
® 20
E
tu 10
I
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 08 03 10
Heat Input Factor Bin
44
-------
Gilberton Power Company Unit 032 PA
3,000
01 02 03 04 05 08 07 08
Heat Input Factor Bin
Gllbertcn Power Company Unit 032 PA
09 10
Year
M 2012 (3rd lowest)
2015 (lowest)
.... 13
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 OS 09 10 01 02 03 04 05 06 07 08 09 10
g 31
o
'm 0
E
til 10
oK
I
I
Heat Input Factor Bin
45
-------
Mi Carmel Gogeneratiori Unit SG-101 PA
3,000
Year
M 2016 (3rd lowest)
2017 (lowest)
.... 13
05 08 07
Heat Input Factor Bin
Mt Carmel Cogeneratlon Unit SG-101 PA
E 0.5
1».4
V 0.3
| 0.2
oK°-1
..
03 04 05 00 07 08 09 03 04 05 06 0? 08 09
03 04 05 00 07 08 09
m
c
o
-*-»
m 4
o
tfi
m
E 2
til
<$
I
)4 05 06 07 08 09
03 04 05 06 07 08 09
Heat Input Factor Bin
46
-------
St. Nicholas Degeneration Project Unit 1 PA
2,000
=3
O
X
U)
sz
+3
o 1,000
01 02 03 0* 05 06
Heat Input Factor Bin
St. Nicholas Cogeneratlon Project Unit 1 PA
07 08
Year
1111 2008 (fewest)
2 u 10 (3rd lowest)
.... 13
£ 0.3
E
I 0.1
£o.
6
1
T '
'&±X:L
01 02 03 04 05 06 07 08 01 02 03 04 05 06 07 08 01 02 03 04 05 06 07 08
m
c
o
-*-»
m
c
o
t/i
m
I 2..
<$
fa
I
: 03 04 05 06
08 01 02 03 04 05 06 07 08
Heat Input Factor Bin
47
-------
WPS Wesiwood Generation. LLC Unit 031 PA
1,250
01 02 03 04 05 06 07 08 09
Heat Input Factor Bin
WPS Westwood Generation, LLC Unit 031 PA
E
J= 0,2
_Q
©
® 0.1
en
i „„
.si"
Year
¦¦l 2016 Oowest)
2018 (3rd lowest)
...13
01 02 03 04 05 06 07 08 09 10 01 02 03 04 05 06 07 OS 09 10 01 02 03 04 05 06 07 08 09 10
c 40
o
m
c
o
t/i
m
E
til
01 02 03 04 05 06 07 08 09 10
01 02 03 04 05 08 07 08 03 10
Heat Input Factor Bin
48
-------
3.2 Evaluation of Hourly NOx Emission Rate Outlier Data
The plots in this section provide more information on the high outlier data values from the
earlier analysis. For purposes of the analysis, high outliers for a given unit in a given year
are defined as hourly emission rate data values that exceed the 75th percentile data value
for the respective heat input factor bin for that year by more than 1.5 times the
interquartile range (75th - 25th percentile). The purpose of the outlier analysis was to
evaluate whether there was any clustering of outliers that could indicate a pattern of
operation or whether, instead, the outliers simply indicate unplanned variability.
Figure 3.1 shows the total number of high NOx emission rate outliers that occur on each
day of the 2019 ozone season for all SCR-equipped coal-fired EGUs in each of the four states
analyzed. Outliers do not appear to be clustered around particular days. They occur with
some regularity across the 2019 ozone season for both Maryland and Pennsylvania.
Figure 3.2 performs the same analysis for Conemaugh units 1-2 and Keystone units 1-2 in
Pennsylvania and Brandon Shores units 1-2 in Maryland. The four Conemaugh and
Keystone units had the largest amounts of 2019 ozone season NOx emissions of all the units
in this analysis, as noted in section 3.1.1, while the two Brandon Shores units had the
largest amounts of 2019 ozone season NOx emissions of all the units in Delaware,
Maryland, and New Jersey. The outliers for each unit at each of the three plants are spread
across the ozone season. There are different distributions for the two units at each plant
and no particular indication of clustering on individual days for any of the six units.
EPA notes that the levels of individual hourly NOx emission rate values that are considered
outliers can vary across facilities. The Brandon Shores units have outliers fairly frequently
throughout the ozone season; however, these EGUs generally have much lower NOx
emission rates and less variability in NOx emission rates than the Conemaugh and Keystone
units. In fact, many points that are considered outliers for the Brandon Shores units in the
2019 ozone season occurred at emission rate levels that would not be considered outliers
at the Conemaugh and Keystone units, because the Conemaugh and Keystone units have
higher NOx emission rates and higher variability in NOx rates.
Figure 3.2 also highlights in blue the days on which select monitors in the OTC exceeded
the 2015 ozone NAAQS (>70 ppb) during the 2019 ozone season.5 There does not appear to
be a correlation between the occurrence of outliers and high ozone days in the 2019 ozone
season.
5 EPA used the list of monitors on page 6 of Attachment 2 in the OTC's recommendation in
the table titled "Part 2A - Measured ozone data through 2019 ozone season" for Figure 3.2.
49
-------
Figure 3.1: Ozone Season 2019 NOx Rate Outliers at Coal-fired EGUs with SCR
Slate
OE
MD
NJ
PA
Jul Aug
Date
Figure 3.2: Ozone Season 2019 NOx Rate Outliers by Unit: Brandon Shores units 1-2,
Conemaugh units 1-2, and Keystone units 1-2
if)
s_
m
o
-Q
E
Brandon Shores
Conemaugh
Keystone
State
MD
PA
May Jim Jul Aug Sep C-rtUa. Jun Jut Sep i.ia. Jur- Jul Aug Sep Oct
Date
Note: Blue sections represent high ozone days at select monitors in OTC states.
50
------- |