SENSITIVITY OF ANNUAL PM2.5 IN ATLANTA TO SO2 EMISSION REDUCTIONS
RESULTING FROM Georgia’s Multipollutant Rule [391-3-1-.02(2)(sss)]

GA EPD performed photochemical grid modeling to determine the
sensitivity of PM2.5 concentrations in the Atlanta nonattainment area to
reductions in SO2 emissions due to the installation and operation of
scrubbers at three large coal-fired power plants in Georgia.  The
analysis was based off of the Visibility Improvement State and Tribal
Association of the Southeast (VISTAS) modeling, using the emissions
inventories for both 2002 and 2009, and CMAQ version 4.5 with the CB-IV
chemical mechanism.  Full annual SO2 sensitivities were conducted by GA
EPD, based on 2002 meteorology and 2009 emissions, using a 12-km
modeling grid, covering the state of Georgia and parts of the
neighboring states (boundary conditions were obtained from simulations
using a larger domain).

Cases Modeled

Sensitivities were modeled relative to 2009 emissions, to evaluate the
efficiency of various control strategies to be implemented by 2009. 
Three large coal-fired power plants in Georgia were modeled to simulate
actual expected reductions from the installation of scrubbers for SO2
removal (assuming a 95% control efficiency for SO2).  Note that Plant
Hammond Units 1-4, Plant Bowen Units 3-4, Plant Wansley Unit 1, and
Plant Yates Unit 1 were not included in the analysis since they were
required to already have scrubbers installed by December 31, 2008. The
following is a detailed list of the three facilities that were modeled:

Scrubbers at Bowen: Installation of scrubbers on Plant Bowen Units 1-2
(average reduction of 218 tons-per-day of SO2 compared to
pre-installation emissions; assuming a 95% control efficiency for SO2).

Scrubbers at Wansley: Installation of scrubbers on Plant Wansley Unit 2
(average reduction of 104 tons-per-day of SO2 compared to
pre-installation emissions; assuming a 95% control efficiency for SO2).

Scrubbers at Yates: Installation of scrubbers on Plant Yates Units 2-7
(average reduction of 137 tons-per-day of SO2 compared to
pre-installation emissions; assuming a 95% control efficiency for SO2).

Modeling Results

The results of the sensitivity analysis are reported in terms of the
average reduction in annual PM2.5 in the 3x3 matrix centered around the
grid-cell containing the monitoring site. Both absolute reductions
((g/m3) and reductions normalized per ton-per-day of precursor
controlled (ng/m3/TPD) are reported. Modeled reductions at each site
were scaled relative to the observed speciated 5-year PM2.5 design value
(weighted for 2000-2004), following the approach recommended in the EPA
modeling guidance (EPA, Guidance on the Use of Models and Other Analyses
for Demonstrating Attainment of Air Quality Goals for Ozone, PM2.5, and
Regional Haze, EPA-454/B-07-002, April 2007).  Sensitivity analysis
results for SO2 emission controls at Plant Wansley (Table 1), Plant
Bowen (Table 2), and Plant Yates (Table 3) at monitors in Atlanta are
summarized below.  The monitor with the highest sensitivity in each
Table is highlighted in bold red.

g/m3)” is equal to the difference between the 2009 future design
value with 2009 uncontrolled emissions (2009_TPD) and the 2009 future
design value with 2009 controlled emissions (2009_TPD * 0.05).  The
“Normalized annual reduction in PM2.5 (ng/m3/TPD)” is calculated by
dividing the “Annual Reduction in PM2.5 (g/m3)” by the “TPD
reduction”.  

The Powder Springs-Macland Aquatic Center PM2.5 monitor (13-067-0004)
did not start operating until February 5, 2003.  As a result, this
monitor was not included in the sensitivity analysis because it did not
have a valid 5-year weighted 2002 design value (based on 2000-2004 data)
since more than 2 years of data was missing.  Without a valid 2002
design value, it is not possible to follow the EPA modeling guidance to
calculate a 2009 future design value or the sensitivity impacts.  Since
2003, this monitor has measures annual PM2.5 concentrations below most
of the other PM2.5 monitors in Atlanta.  Therefore, leaving it out of
the analysis should not have any impact on the conclusions.

Table 1: Summary of sensitivity analysis results for SO2 emission
controls at Plant Wansley.

Site Name	TPD reduction	Annual Reduction1 in PM2.5 ((g/m3)	Normalized
annual reduction in PM2.5 (ng/m3/TPD)

Doraville (13-089-2001)	103.88	0.0858	0.826

E. Rivers School (13-121-0032)	103.88	0.0902	0.868

Fire Station No. 8 (13-121-0039)	103.88	0.0906	0.872

Forest Park-Georgia DOT (13-063-0091)	103.88	0.0976	0.940

Gainesville-Fair Street School (13-139-0003)	103.88	0.0612	0.589

Gwinnett Tech (13-135-0002)	103.88	0.0714	0.687

Kennesaw-National Guard (13-067-0003)	103.88	0.0839	0.807

South DeKalb (13-089-0002)	103.88	0.0928	0.893

Yorkville (13-223-0003)	103.88	0.0983	0.947

1Scaled relative to the observed speciated 5-year PM2.5 concentrations
(weighted for 2000-2004).

Table 2: Summary of sensitivity analysis results for SO2 emission
controls at Plant Bowen.

Site Name	TPD reduction	Annual Reduction1 in PM2.5 ((g/m3)	Normalized
annual reduction in PM2.5 (ng/m3/TPD)

Doraville (13-089-2001)	217.72	0.1682	0.772

E. Rivers School (13-121-0032)	217.72	0.1564	0.718

Fire Station No. 8 (13-121-0039)	217.72	0.1553	0.713

Forest Park-Georgia DOT (13-063-0091)	217.72	0.1417	0.651

Gainesville-Fair Street School (13-139-0003)	217.72	0.1692	0.777

Gwinnett Tech (13-135-0002)	217.72	0.1717	0.788

Kennesaw-National Guard (13-067-0003)	217.72	0.2117	0.973

South DeKalb (13-089-0002)	217.72	0.1411	0.648

Yorkville (13-223-0003)	217.72	0.1549	0.711

1Scaled relative to the observed speciated 5-year PM2.5 concentrations
(weighted for 2000-2004).

Table 3: Summary of sensitivity analysis results for SO2 emission
controls at Plant Yates.

Site Name	TPD reduction	Annual Reduction1 in PM2.5 ((g/m3)	Normalized
annual reduction in PM2.5 (ng/m3/TPD)

Doraville (13-089-2001)	137.20	0.1530	1.115

E. Rivers School (13-121-0032)	137.20	0.1673	1.220

Fire Station No. 8 (13-121-0039)	137.20	0.1680	1.224

Forest Park-Georgia DOT (13-063-0091)	137.20	0.1795	1.308

Gainesville-Fair Street School (13-139-0003)	137.20	0.0988	0.720

Gwinnett Tech (13-135-0002)	137.20	0.1207	0.880

Kennesaw-National Guard (13-067-0003)	137.20	0.1589	1.158

South DeKalb (13-089-0002)	137.20	0.1641	1.196

Yorkville (13-223-0003)	137.20	0.1579	1.151

1Scaled relative to the observed speciated 5-year PM2.5 concentrations
(weighted for 2000-2004).

Clean Data Determination

On December 8, 2011, EPA promulgated its determination (76 FR 76620)
that the Atlanta nonattainment area had attained the 1997 annual average
PM2.5 National Ambient Air Quality Standard (NAAQS).  This determination
was based upon quality-assured and certified ambient air monitoring data
for the 2008-2010 period that showed design values below 15.0 µg/m3 as
indicated in Table 4.  

Table 4.  Annual and Three-year Average Ambient PM2.5 Concentrations.

Ambient PM2.5 Concentration (µg/m3)

Site Name	2003	2004	2005	2006	2007	2008	2009	2010	2008-2010 w/out data
substitution	2008-2010 with data substitution

Forest Park-Georgia DOT (13-063-0091)	16.0	16.8	16.6	16.7	15.3	13.7	11.5
13.4	12.9	N/A

Kennesaw-National Guard (13-067-0003)	16.0	15.8	16.3	16.5	15.3	13.5	11.2
12.2	12.3	N/A

Powder Springs-Macland Aquatic Center (13-067-0004)	15.2	15.2	15.5	15.8
14.5	13.1	10.3	12.1	11.9	12.3

South DeKalb (13-089-0002)	15.0	16.1	15.5	15.4	14.8	12.7	11.4	12.3	12.1
N/A

Doraville (13-089-2001)	15.4	15.5	15.8	14.5	15.1	13.1	11.7	12.2	12.3	N/A

E. Rivers School (13-121-0032)	16.1	16.1	15.9	15.4	15.7	13.0	11.6	12.2
12.3	13.0

Fire Station No. 8 (13-121-0039)	17.7	17.6	17.0	18.4	--	7.6	12.1	14.5
11.4	13.6

Georgia Tech (13-121-0048)*	--	--	--	15.1	15.5	14.3	--	--	--	--

Gwinnett Tech (13-135-0002)	16.2	16.1	16.1	16.9	14.2	12.4	11.6	12.3	12.1
12.5

Gainesville-Fair Street School (13-139-0003)	14.7	14.0	14.5	13.8	13.4
11.8	10.2	11.4	11.2	11.9

Yorkville (13-223-0003)	13.8	13.4	14.6	13.9	14.3	11.9	9.9	11.2	11.0	11.6

*The Georgia Tech monitor stopped operation on December 29, 2008.

Discussion

The Atlanta PM2.5 nonattainment area redesignation request and
maintenance plan accounted for controls from Georgia’s Multipollutant
Rule [391-3-1-.02(2)(sss)].  This rule requires flue gas desulfurization
(FGD) and selective catalytic reduction (SCR) controls on the majority
of coal-fired EGUs in Georgia.  The required FGD control implementation
date varies by EGU, starting on December 31, 2008.  SO2 controls were
operating for the entire year of 2010 as follows:

Plant Hammond – controls on 4 of 4 units;

Plant Bowen – controls on 3 of 4 units;

Plant Wansley – controls on 2 of 2 units;

Plant Yates – controls on 1 of 7 units.

Table 5 shows the annual SO2 emissions for Plant Hammond, Plant Bowen,
and Plant Wansley for the period 2003 through 2010.  Also, the table
includes the electrical generation, SO2 emissions per 106 MWh
generation, the average SO2 emissions per 106 MWh generation for
2003-2007, the uncontrolled SO2 in 2008-2010 (assuming average SO2
emissions per 106 MWh generation for 2003-2007 x generation in 2008,
2009, 2010), and the controlled SO2 in 2008-2010 (difference between
uncontrolled and actual SO2 emissions).

In addition, Table 5 shows the annual SO2 emissions for Plant Yates
(Unit 1) for the period 2008 through 2010.  Also, the table includes the
uncontrolled SO2 in 2008-2010 (SO2 emissions in 2008, 2009, 2010 divided
by 0.05), and the controlled SO2 in 2008-2010 (difference between
uncontrolled and actual SO2 emissions).

The “AVG 08-10 Controlled” value in Table 5 is the amount of SO2
emission reductions that can be attributed to Georgia’s Multipollutant
Rule [391-3-1-.02(2)(sss)] between 2008 and 2010.  By multiplying the
controlled SO2 emissions for each facility by the normalized annual
sensitivity in Tables 1 - 3 will quantify the impact of SO2 emission
controls from each facility on annual PM2.5 concentrations.  

Table 5.  Controlled SO2 emissions from Plant Hammond, Plant Bowen,
Plant Wansley, and Plant Yates (Unit 1) in 2008 - 2010.

Facility

2003	2004	2005	2006	2007	2008*	2009*	2010*

Hammond	SO2 (tpy)	35,900	37,700	39,500	40,600	47,800	12,500	900	2,427

	generation 	4.79	4.11	4.64	4.27	5.09	4.32	3.73	3.23

	SO2/gen	7,494.8	9,172.7	8,512.9	9,508.2	9,391.0	2,893.5	241.3	751.4

	AVG 03-07



	8,815.9





Uncontrolled





38,084.8	32,883.4	28,475.4

	Controlled





25,584.8	31,983.4	26,048.4

	AVG 08-10







27,872.2











	Bowen	SO2 (tpy)	164,900	165,900	186,500	206,400	196,800	148,100	54,800
7,618

	generation 	21.9	21.9	23.3	23.7	24	23.4	22.9	24.7

	SO2/gen	7,529.7	7,575.3	8,004.3	8,708.9	8,200.0	6,329.1	2,393.0	308.4

	AVG 03-07



	8,003.6





Uncontrolled





187,285.1	183,283.2	197,689.8

	Controlled





39,185.1	128,483.2	190,071.8

	AVG 08-10







119,246.7











	Wansley**	SO2 (tpy)	94,000	99,000	101,500	96200	93,900	74,300	7400
2,343

	generation 	13.7	14	15.1	14.8	16.1	15.5	11.8	9.8

	SO2/gen	6,861.3	7,071.4	6,721.9	6,500.0	5,832.3	4,793.5	627.1	239.1

	AVG 03-07



	6,597.4





Uncontrolled





102,259.4	77,849.1	64,654.3

	Controlled





27,959.4	70,449.1	62,311.3

	AVG 08-10







53,573.3











	Yates, Unit 1	SO2 (tpy)





352	175.9	286

	Uncontrolled





7,040.0	3,518.0	5,726.0

	Controlled





6,688.0	3,342.1	5,439.7

	AVG 08-10







5,156.6

Source: EPA Clean Air Markets Division

* SO2 control, required by Georgia Rule (sss), was operational on one or
more EGUs for partial or full year

** does not include emissions from facility’s gas-fired units

Figure 1.  Map of Plant Hammond, Plant Bowen, and Plant Wansley relative
to the ten Atlanta PM2.5 monitors.

The location of Plant Hammond, Plant Bowen, Plant Wansley and Plant
Yates relative to the ten Atlanta PM2.5 monitors is shown in Figure 1. 
The Yorkville monitor (13-223-0003) has the highest normalized
sensitivity to SO2 emissions from Plant Wansley when compared to all the
other Atlanta monitors (see Table 1).  Therefore, a conservative
approach would be to use the SO2 normalized sensitivity at the Yorkville
monitor for all the monitors in Atlanta.

The Kennesaw-National Guard monitor (13-067-0003) has the highest
normalized sensitivity to SO2 emissions from Plant Bowen when compared
to the other Atlanta monitors (see Table 2).   Therefore, a conservative
approach would be to use the SO2 normalized sensitivity at the
Kennesaw-National Guard monitor for all the monitors in Atlanta.

Since we did not perform SO2 emission sensitivities from Plant Hammond
and Plant Hammond is located farther away from all the Atlanta monitors
when compared to Plant Bowen, a conservative approach would be to use
the Plant Bowen SO2 normalized sensitivities from the Kennesaw-National
Guard monitor to represent the impact from Plant Hammond at all the
Atlanta monitors.

The Forest Park monitor (13-063-0091) has the highest normalized
sensitivity to SO2 emissions from Plant Yates when compared to all the
other Atlanta monitors (see Table 3).  Therefore, a conservative
approach would be to use the SO2 normalized sensitivity at the Forest
Park monitor for all the monitors in Atlanta.

The maximum impact on annual PM2.5 concentrations at each Atlanta PM2.5
monitors due to SO2 emission reductions that can be attributed to
Georgia’s Multipollutant Rule [391-3-1-.02(2)(sss)] controls at Plant
Bowen, Plant Hammond, Plant Wansley, and Plant Yates is calculated in
Table 6. 

g/m3)

Plant Bowen	119,246.70	326.70	0.973	0.318

Plant Hammond	27,872.21	76.36	0.973	0.074

Plant Wansley	53,573.25	146.78	0.947	0.139

Plant Yates	5,156.60	14.13	1.308	0.018

TOTAL



0.550 



Table 7.  Three-year Average Ambient PM2.5 Concentrations with/without
data substitution and with/without Rule 391-3-1-.02(2)(sss). 

Ambient PM2.5 Concentration (µg/m3)

Site Name	2008-2010 without data substitution and with Rule 

391-3-1-.02(2)(sss)	2008-2010 with data substitution and with Rule 

k

Š

Œ

š

Œ

š

!

"

K

~

Ž

³

Ä

î

ó

 h¼f

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

 h¼f

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

摧⡳ 欀㽤

혈\鐄䃿㨋帑通$갇

摧⡳ 欀

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

摧⡳ 欀塤

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

혈\鐄䃿㨋帑通$갇

h

	N/A

Powder Springs-Macland Aquatic Center (13-067-0004)	11.9	12.3	12.5	12.9

South DeKalb (13-089-0002)	12.1	N/A	12.7	N/A

Doraville (13-089-2001)	12.3	N/A	12.9	N/A

E. Rivers School (13-121-0032)	12.3	13.0	12.9	13.6

Fire Station No. 8 (13-121-0039)	11.4	13.6	12.0	14.2

Gwinnett Tech (13-135-0002)	12.1	12.5	12.7	13.1

Gainesville-Fair Street School (13-139-0003)	11.2	11.9	11.8	12.5

Yorkville (13-223-0003)	11.0	11.6	11.6	12.2



Conclusion

 PM2.5 design values that are all well below the 1997 annual PM2.5 NAAQS
of 15.0 g/m3 (see Table 7).  Therefore, removal of these controls
would not have resulted in 2008-2010 annual PM2.5 design value over the
1997 annual PM2.5 NAAQS.

