1. Good Neighbor SIP Discussion October 24, 2014 Gregory Stella Alpine Geophysics, LLC 1

2. History Discussions began in December of 2013 with the petition for the addition of Illinois, Indiana, Kentucky, Michigan, North Carolina, Ohio, Tennessee, Virginia, and West Virginia to the Ozone Transport Region (176A) Further conversations in May of 2014 where Tad introduced the Good Neighbor SIP process to help with MD’s ozone SIP due in June 2015 Additional, regular conversations held regarding the modeling that has been conducted to support these two items 2

3. What Are We Trying To Resolve? MOG asked to conduct technical analyses to assist in the air quality assessment of an EGU optimization study – Run SCR/SNCR controls at optimized rates – Purpose to review controls to help MD reach attainment 3

4. Current Air Quality Observations We’ve now seen most recent 2014 ozone concentration data showing Harford, MD in attainment of the 75 ppb NAAQS – Many other northeastern monitors also attain 8hr ozone NAAQS 4 4th Highest MDA8 (ppm)8hr Ozone DV (ppm) StateCountySite ID200920102011201220132014*2011-20132012-2014* MarylandHarford2402510010.0830.0960.0980.0860.0720.0670.0850.075 MarylandHarford2402590010.0690.0800.0850.0830.0680.0700.0790.074 *Draft 4 th high as of September 30, 2014

5. 2014 Ozone Season Preview Look at preliminary (draft) 4 th high 8hr ozone design value from AQS network – 122 eastern state monitors show nonattainment of 75ppb NAAQS with 2011-2013 DVs – 17 eastern state monitors (15 counties) remain in nonattainment with draft 2012-2014 DVs – No counties moved from attainment to nonattainment 5

6. Nonattainment Monitors with Draft 2012/2014 8hr Ozone DVs State NameCounty NameAQS Site ID 2011-2013 Design Value (ppm) 2012-2014 Design Value* (ppm) ConnecticutFairfield0900130070.0890.082 MichiganAllegan2600500030.0860.082 ConnecticutFairfield0900100170.0830.079 IndianaLaPorte1809100050.0830.079 ConnecticutFairfield0900190030.0870.079 IllinoisLake1709710070.0800.079 MichiganBerrien2602100140.0820.078 OhioLake3908500030.0800.078 ConnecticutNew Haven0900990020.0890.078 ConnecticutMiddlesex0900700070.0810.078 MichiganMuskegon2612100390.0810.078 ConnecticutTolland0901310010.0770.078 MarylandCecil2401500030.0820.078 ConnecticutNew London0901101240.0840.077 MissouriSaint Charles2918310020.0820.076 MarylandPrince George's2403380030.0810.076 IllinoisCook1703100320.0800.076 6 *Using draft 4 th highest ozone concentrations as of September 30, 2014

7. 176A Petition State Monitors Recent 8hr Ozone Design Values 7 4th Highest MDA8 (ppb) 3yr Design Value (ppb) MonitorCounty2011201220132014* 2011201220132014* 240251001Harford, Maryland 9886726792938575 361030002Suffolk, New York 8983726184858772 90019003Fairfield, Connecticut 8789866179858779 421010024Philadelphia, Pennsylvania 8985686683878073 340150002Gloucester, New Jersey 9287736682878475 250070001Dukes, Massachusetts 7882655876807568 440090007Washington, Rhode Island 748279607378 74 100031007New Castle, Delaware 7882627175807472 330074001Coos, New Hampshire 6871696569708768 500030004Bennington, Vermont 59676250 65646260 * As of 30 Sept 2014

8. Ozone Metrics - Harford, MD 8 Results based on EPA published ozone 8-hr ozone design values and ozone source apportionment modeling from LADCO/IPM 2018 air quality simulations 2018 OSAT Contributions

9. Ozone Metrics – Suffolk, NY 9 Results based on EPA published ozone 8-hr ozone design values and ozone source apportionment modeling from LADCO/IPM 2018 air quality simulations 2018 OSAT Contributions

10. Ozone Metrics – Fairfield, CT 10 Results based on EPA published ozone 8-hr ozone design values and ozone source apportionment modeling from LADCO/IPM 2018 air quality simulations 2018 OSAT Contributions

11. Ozone Metrics – Philadelphia, PA 11 Results based on EPA published ozone 8-hr ozone design values and ozone source apportionment modeling from LADCO/IPM 2018 air quality simulations 2018 OSAT Contributions

12. Ozone Metrics – Gloucester, NJ 12 Results based on EPA published ozone 8-hr ozone design values and ozone source apportionment modeling from LADCO/IPM 2018 air quality simulations 2018 OSAT Contributions

13. MD’s Good Neighbor Optimization MD picked 11 states – IL, IN, KY, MD, MI, NC, OH, PA, TN, VA, and WV – Optimized SCR/SNCR to lowest rate since 2005 – MD predicts 0.9 ppb improvement in 2018 with lowest rate optimization across all states 13

14. 3 rd Lowest Rate Optimization Developed list of SCR/SNCR units based on CEM data reporting Looked at 3 rd lowest rate for optimization – Both SCR/SNCR results – Presumption that lowest rate occurs immediately after installation – 3 rd lowest rate allows for full year of operation after installation and company optimization of controls 14

15. Emission / Air Quality Relationships Ratio development calculation relating 2011 ozone season summer episode (June through August) NOx emissions from EGU sources within each State to its ozone concentration air quality contribution at downwind monitors Designed to establish a NOx emissions to ozone concentration ratio that will be used in later stages of the analysis 15

16. Base Year Source-Receptor Analyses Base year (2011) APCA/OSAT to develop ozone source- receptor relationship data from CAMx simulation 16

17. Concentration to Tons Ratio Calculation Each ton of NOx from all EGUs in upwind State is calculated to have an impact on downwind monitor Assumption that each ton of NOx reduced from upwind state EGUs will reduce downwind monitor ozone concentration by this amount 17 Design Value Adjusted (ppb)ppb/ton Region2011 NOx TonsEGUTotal% of TotalRatio IL 19,4360.331.671.86%1.69E-05 IN 35,5590.472.352.61%1.34E-05 KY 25,9050.321.591.77%1.22E-05 MD 5,4202.4318.7720.85%4.48E-04 MI 22,1300.261.992.21%1.17E-05 NC 15,3130.151.181.31%9.60E-06 OH 29,8470.665.385.98%2.20E-05 PA 43,5072.469.2510.28%5.66E-05 TN 8,3420.080.760.84%9.49E-06 VA/DC 11,0140.696.387.08%6.26E-05 WV 16,4960.842.492.76%5.07E-05 Grand Total 10.2390.00100.00%

18. Emission Inventory Assessment Match 2011 modeling platform of EGU emissions to CAMD reported metadata and identify which units have post-combustion NOx control (SCR / SNCR) installed for operation Using other historical CAMD data (ozone season 2005-2013 emission and operating reports), we made a preliminary determination as to which units in 2011 were operating these post- combustion controls during the summer episode 18

19. SCR/SNCR - Top Reducing Units 2011 Ozone Season NOx RateNOx NOx Rate (lbs/MMBtu)NOx Reduction (Tons) StateFacilityORISUNIT (lbs/ MMBtu)(Tons)Lowest 3 rd LowestLowest3 rd Lowest PAKeystone313620.3645,044.30.0430.045-4,444.3-4,415.2 PAKeystone313610.3744,854.60.0430.052-4,294.7-4,177.8 PAMontour314910.3313,298.40.0580.084-2,720.1-2,465.3 OHW H Zimmer601910.2173,559.50.0560.075-2,638.4-2,338.4 PAMontour314920.3163,132.20.0580.094-2,559.4-2,203.7 KYParadise137830.3322,431.20.1000.108-1,698.0-1,640.2 INGibson611320.2262,043.40.0670.098-1,437.0-1,163.5 WVHarrison Power Station394430.2141,834.20.0660.072-1,269.7-1,213.9 INGibson611310.1821,504.50.0340.071-1,221.1-917.1 WVHarrison Power Station394420.2011,774.70.0660.080-1,190.9-1,072.7 WVHarrison Power Station394410.1911,698.40.0630.079-1,134.2-997.2 ILKincaid Station87620.2041,566.70.0600.062-1,104.9-1,086.4 PAHatfield's Ferry Station317930.4392,848.00.2700.368-1,098.9-466.4 PACheswick822610.2351,690.00.0900.171-1,043.4-462.8 19

20. Emission/Air Quality Change Calculation Calculate the State-level difference in EGU NOx emissions resulting from the application of post- combustion NOx control compared to actual 2011 operations – in cases where post-combustion control was already in operation at lowest rates or units are not identified to have SNCR/SCR installed, this value will be zero Apply this value to each State-monitor where we have generated an emissions to air quality change ratio 20

21. SCR & SNCR Optimization May-Sept Application 2011 Ozone Season2005-2013Lowest Rate3rd Low Rate NOx RateNOx EmissionsLowest3 rd NOx Tons State(lbs/MMBtu)(Tons)NOx RateLowest RateDifference PA Total0.24234,730.20.0950.120-21,095.2-17,574.6 IN Total0.16020,225.20.0860.108-9,300.5-6,522.1 OH Total0.13124,554.40.0870.100-8,178.3-5,742.7 KY Total0.11915,866.30.0680.082-6,867.0-5,001.8 WV Total0.10014,418.90.0530.063-6,789.0-5,300.2 VA Total0.2359,660.90.1300.141-4,304.0-3,867.2 NC Total0.12716,479.50.0970.109-3,880.5-2,342.5 IL Total0.0977,592.70.0580.064-3,021.7-2,569.0 MI Total0.1155,652.30.0880.100-1,292.0-716.4 MD Total0.1547,353.20.1290.144-1,194.0-481.1 TN Total0.0937,102.80.0810.090-886.0-210.4 Grand Total0.141163,636.30.0840.098-66,808.1-50,328.1 21

22. Optimization Application Calculation at each monitor was conducted for each upwind State with ozone contribution (significant or not) 22

23. June-August Optimization NOx All EGU Emissions 23

24. June – August Optimization Ozone Harford, MD Monitor 24 Lowest Rate Strategy (ppb)3rd Lowest Rate Strategy (ppb) Region Tons RedEGUDeltaTotal % of Total Tons RedEGUDeltaTotal % of Total IL-1,8080.300.031.641.86%-1,5220.300.031.651.86% IN-5,7040.400.082.282.58%-4,0070.420.052.302.59% KY-4,6870.260.061.541.74%-3,5470.270.041.551.75% MD-7232.110.3218.4420.90%-2492.320.1118.6521.03% MI-7970.250.011.982.24%-4670.250.011.982.24% NC-2,3490.120.021.151.31%-1,5540.130.011.161.31% OH-5,2540.560.105.285.98%-3,6450.580.085.305.98% PA-13,3601.710.768.499.62%-11,1071.840.638.629.71% TN-5380.070.010.750.85%-1330.080.000.760.85% VA/DC-2,4040.540.156.237.06%-2,1430.560.136.247.04% WV-4,5430.610.232.262.56%-3,5870.650.182.312.60% Grand Total-42,1668.471.7688.24100.00%-31,9628.951.2888.72100.00%

25. Air Quality Effectiveness Calculation Apply emission change to each State-monitor and generate an emissions to air quality change ratio From optimization results, developed EGU reduction effectiveness factor from 2011 modeling and OSAT results – Looks at effectiveness of ton reduced on downwind ozone concentration change – Calculated for each monitor 25

26. OSAT/ Reduction Efficiency (SCR/SNCR) Jun-Aug 2011 Optimization Harford, MD Monitor Compare total 2011 reduction of 1.28 ppb with MD estimate of 0.9 ppb in 2018. MD/PA comprise 58% (0.74 ppb) of ozone reduction under 3 rd lowest rate strategy. 26 Efficiency NOx TonsChange inFactor StateReducedO3 (ppb)ppb/ton red PA 11,1070.635.66E-05 WV 3,5870.185.07E-05 VA/DC 2,1430.136.26E-05 MD 2490.114.48E-04 OH 3,6450.082.20E-05 IN 4,0070.051.34E-05 KY 3,5470.041.22E-05 IL 1,5220.031.69E-05 NC 1,5540.019.60E-06 MI 4670.011.17E-05 TN 1330.009.49E-06 Total 31,962 1.28

27. Efficiency Comparison Optimization of EGU controls in MD has almost an order of magnitude greater impact on Harford monitor than any other state 27

28. 2018 Attainment Results Ran EPA attainment model (MATS) on LADCO 2018 modeling platform – Based on EPA EGU modeling with IPM – Includes results of onroad Tier3 NPRM – Compare to final Tier3 EPA dv modeling, MD scenario application, current draft 3yr dvs 28

29. 2018 Ozone DVs Eastern US Modeling Domain LADCO Modeling DV (ppb)Tier3MD 3C2012-14 Location2011 DV2018 DV 3yr DV Harford, Maryland90.080.475.8774.775 Suffolk, New York83.378.778.52 Fairfield, Connecticut83.777.875.16 New Haven, Connecticut85.777.872.28 Sheboygan, Wisconsin84.377.071.57 Philadelphia, Pennsylvania83.376.677.32 Gloucester, New Jersey84.376.571.13 Hamilton, Ohio82.076.271.57 Jefferson, Kentucky82.076.069.25 Wayne, Michigan78.775.874.13 Saint Charles, Missouri82.375.672.56 Allegan, Michigan82.775.574.44 Allegheny, Pennsylvania80.774.972.91 Franklin, Ohio80.374.770.71 Oldham, Kentucky82.074.666.48 Arlington, Virginia81.774.069.43 Fairfax, Virginia82.374.069.25 Cecil, Maryland83.073.872.82 MD 2018 results lower than LADCO or Tier3 Why? 29

30. MD’s Path to Attainment 30

31. “The Bottom Line” Case / StrategyReductionOzone dv MD 2018 DV79 ppb Tier 3~ 0.8 ppb78.2 Add'l OTR Measures~ 1.277.0 Add’l MD Only Controls~ 1.475.6 <- Already Attainment EGU Optimized (MD/PA)~ 0.575.1 58% of Total Optimization Attainment achieved without Upwind State Controls EGU Optimized (Upwind )~ 0.474.7 MD 2018 Scenarios DV74.7 WOE~ 0.5 MOVES2014/MEGANLower (?) PA NOx RACTLower Unit RetirementsLower 31

32. Conclusions Current Harford monitoring data points to attainment of 75 ppb NAAQS 2018 modeling data projects Harford attainment without additional upwind controls Is there other justification for upwind control? 32

33. Gregory Stella Managing Partner Alpine Geophysics, LLC gms@alpinegeophysics.com 828-675-9045 33