Technical Support Document (TSD) for AERMOD-Based Assessments of Long-Range Transport Impacts for Primary Pollutants ------- ------- EPA-454/B-15-003 July 2015 Technical Support Document (TSD) for AERMOD-Based Assessments of Long-Range Transport Impacts for Primary Pollutants U.S. Environmental Protection Agency Office of Air Quality Planning and Standards Air Quality Analysis Division Air Quality Modeling Group Research Triangle Park, North Carolina ------- Preface This document provides the results and interpretation of AERMOD simulations for typical PSD sources. The focus of the analysis is to determine the impact of a range of source types on pollutant concentrations in the near and far field. ------- Contents Preface 4 Contents 5 Tables 6 Figures 7 1. Introduction 8 2. Background 8 3.0 Approach to evaluating near-source and long-range impacts 10 3.1 Source Types and Characteristics 10 3.2 Modeling Assessment of Various Facility Types 11 4. Summary of results 14 4.1 NO2 and SO2 screening 14 4.2 Refined analyses for NO2 and SO2 17 4.3 Refined analyses for PM10 and PM25 20 5. Conclusions 21 6. Additional information 21 References 22 Appendix A 23 Plots from NO2 and SO2 screening analysis 23 Appendix B 34 Plots from NO2 and SO2 refined analysis 34 Appendix C 47 Plots from PM refined analysis 47 ------- Tables Table 1 - Significant Impact Levels (SILs) for Criteria Pollutant by Class 1, 2, and 3 Areas 9 Table 2 - PSD Increment for Criteria Pollutants by Class 1, 2, and 3 Areas 10 Table 3 - Summary of Modeling Scenarios by Source Type from AERMOD Implementation Workgroup (AIWG) 11 Table 4 - Results for Phase 1 screening modeling for NO2 and SO2 by Facility type5 15 Table 5 - Results for Phase 1 screening modeling for NO2 and SO2 by Facility type, continued 16 Table 6 - Summary of NO2 and SO2 emissions for the coal EGU facility 17 Table 7 - Results from the Phase 2 refined analysis for NO2 18 Table 8 - Results from the Phase 2 refined analysis for SO2 19 Table 9 - Summary of 2011 NEI emission data used in refined modeling analysis 20 ------- Figures Figure 1 - Location of receptors for screening and refined runs 13 Figure 2 - Results from the screening analysis for the asphalt plant 24 Figure 3 - Results from the screening analysis for the biomass plant 25 Figure 4 - Results from the screening analysis for the cement kiln 26 Figure 5 - Results from the screening analysis for the coal EGU 27 Figure 6 - Results from the screening analysis for the ethanol plant 28 Figure 7 - Results from the screening analysis for the flare 29 Figure 8 - Results from the screening analysis for the fuel oil turbine 30 Figure 9 - Results from the screening analysis for the landfill gas turbine 31 Figure 10 - Results from the screening analysis for the natural gas compressor station 32 Figure 11 - Results from the screening analysis for the pulp and paper plant 33 Figure 12 - Refined NO2 and SO2 results, ASX, max and design values, 1-hr NAAQS 35 Figure 13 - Refined NO2 and SO2 results, DHT, max and design values, 1-hr NAAQS 36 Figure 14 - Refined NO2 and SO2 results, OAK, max and design values, 1-hr NAAQS 37 Figure 15 - Refined NO2 and SO2 results, SMQ, max and design values, 1-hr NAAQS 38 Figure 16 - Refined SO2 results, ASX, 3 & 24-hr increment 39 Figure 17 - Refined SO2 results, DHT, 3 & 24-hr increment 40 Figure 18 - Refined SO2 results, OAK, 3 & 24-hr increment 41 Figure 19 - Refined SO2 results, SMQ 3 & 24-hr increment 42 Figure 20 - Refined NO2 & SO2 results, ASX, annual NAAQS 43 Figure 21 - Refined NO2 & SO2 results, DHT, annual NAAQS 44 Figure 22 - Refined NO2 & SO2 results, OAK, annual NAAQS 45 Figure 23 - Refined NO2 & SO2 results, SMQ, annual NAAQS 46 Figure 24 - Refined PM25 results, ASX, 24-hr annual increment 48 Figure 25 - Refined PM25 results, DHT, 24-hr annual increment 49 Figure 26 - Refined PM25 results, OAK, 24-hr annual increment 50 Figure 27- Refined PM25 results, SMQ, 24-hr annual increment 51 Figure 28 - Refined PM25 results, ASX, 24-hr NAAQS 52 Figure 29 - Refined PM25 results, DHT, 24-hr NAAQS 53 Figure 30 - Refined PM25 results, OAK, 24-hr NAAQS 54 Figure 31 - Refined PM25 results, SMQ, 24-hr NAAQS 55 ------- 1. Introduction For long-range transport (LRT) applications of distances of more than 50 km from a source, the proposed revisions to EPA's Guideline on Air Quality Models (published as Appendix W to 40 CFR Part 51) include recommendations of a screening approach for addressing PSD increment and removal of CALPUFF as an EPA preferred model for such applications. While there is no proposed replacement refined model =for LRT applications under Appendix W, the information provided in this report indicates that the need for LTR assessments for NAAQS and PSD increment violations for inert pollutants is rare thereby mitigating the necessity for a preferred model for regulatory LRT assessments. This document provides technical details of an EPA modeling assessment that supports this proposed change, including summarizing the model scenarios and approach used to determine the impact of a range of source types on pollutant concentrations in the near and far field. 2. Background The permitting process for the Preventions of Significant Deterioration Program (PSD) requires that a new or modifying source demonstrate that the additional emissions will not cause or contribute to a violation of the NAAQS or PSD increment. The traditional approach for demonstrating compliance is a multi-step process. As described in Section 9 of the proposed version of Appendix W, this two-stage process entails: 1. Conduct a single-source impact analysis where only the new or modifying source is considered in the analysis. The new or modifying source will model its own emissions to determine the appropriate impacts for each applicable pollutant and each averaging time. The predicted impacts from the single source are compared to the Significant Impact Level (SIL) for each averaging time and pollutant. a. If all single source impacts are less than the SIL, then the new or modifying source is generally not expected to cause or contribute to a violation of a NAAQS or PSD increment and the compliance demonstration is considered complete. b. If the single source impacts are greater than the SIL anywhere, then there is a chance that the source will cause or contribute to a violation of a NAAQS or PSD increment and then a cumulative impact analysis should be undertaken. 2. The cumulative impact analysis takes into account all sources affecting the air quality in an area including the project source impact and consideration of background, which includes contributions from natural, nearby, and unknown sources. The cumulative modeling demonstrations for NAAQS and PSD increment are different: a. For a NAAQS compliance demonstration, the emissions from the new or modifying source must be added to the background concentrations and the resulting "design concentration(s)" are compared to the specific NAAQS to determine compliance. When modeled violations are evident, the contribution from the new or modifying source to the modeled violations can be compared to the SIL to determine if the source is causing or contributing to the modeled violation(s). b. For a PSD increment compliance demonstration, the total increase in concentrations from all permitted sources is compared to the 'baseline' concentration. The total increase in concentrations is not allowed to be greater than the applicable PSD increment. 8 ------- As indicated above, the SIL is used as demonstration tool to determine the culpability of a new or modifying source to any NAAQS or PSD increment violations. Table 1 shows the SIL levels for PM2.5, PM10, SO2, NO2, and CO, while Table 2 shows the PSD increment levels of fine particulate matter (PM2.5), coarse particulate matter (PM10), sulfur dioxide (SO2), and nitrogen dioxide (NO2). In 1996, a rulemaking was proposed with Class I specific SILs for NO2 (annual: 0.1 ug/m3), SO2 (annual: 0.1 ug/m3, 24-hr: 0.2 ug/m3, 3-hr 1.0 ug/m3) and PM10 (annual: 0.2 ug/m3, 24-hr: 0.3 ug/m3) (U. S. EPA, 1996). While this rule never went final, such that these Class I SILs have never been promulgated, it is our understanding that they have been used in practice by some states and regions for a screening analysis to eliminate the need for increment analysis are longer transport distances. Pollutant Class 1 Class 2 Class 3 Source Fine Particulate Matter (PM2.5) Annual mean 24-hr maximum 0.061 0.07 0.3 1.2 0.3 1.2 40CFR51.165(b)(2) Particulate Matter (PMio) Annual arithmetic mean 24-hr maximum * * 1 5 * * Carbon Monoxide (CO): 8-hr maximum 1-hr maximum * * 500 2000 * * 40 CFR 52.21 (k)(2) 40CFR51.165(b)(2) Sulfur Dioxide (SO2) Annual mean 24-hr maximum 3-hr maximum 1-hr maximum * * * * 1 5 25 3 ppb (~7.8 ug/m3) * * * * 40CFR51.165(b)(2) (U. S. EPA, 2010b) Nitrogen Dioxide (NO2): Annual mean 1-hr maximum * * 1 4 ppb (~7.5 ug/m3) * * 40CFR51.165(b)(2) (U. S. EPA, 2010a) Table 1 - Significant Impact Levels (SILs) for Criteria Pollutant by Class 1, 2, and 3 Areas. For most PSD compliance demonstrations, the near-source impacts (e.g., those occurring within 50 km of the new or modifying source) are the controlling factor in successfully meeting Clean Air Act requirements. For the inert criteria pollutants, these near-source impacts are assessed with the EPA's preferred dispersion model, AERMOD (Cimorelli, et al, 2005). Due to variations in meteorology that is expected to occur beyond 50 km and the time required for a plume to travel this distance, steady-state plume models like AERMOD are expected to be overly conservative in the far-field. Thus, when long range transport (LRT) is expected to be important (i.e., impacts beyond the nominal distance of 50 km), 1 The annual and 24-hr Class I SILs were remanded and vacated by request from the EPA in January 22, 2013. These values are currently under review and the EPA is developing a proposed rule to update the PM25 SILs. See the PM25 modeling guidance for more discussion on this issue (U. S. EPA, 2014). ------- an alternative model is necessary for assessing impacts for those distances with the current Appendix W recommending the use of the CALPUFF modeling system (U. S. EPA, 2003). Section 6.2.3 of the current version of Appendix W, published in 2005, discusses the regulatory needs for LRT impact assessments. The focus in section 6.2.3 is the need to protect Class I areas and in particular, Class I PSD increments are identified as the most stringent regulatory benchmarks in the PSD program. While refined LRT modeling could also be needed for NAAQS, it is uncommon that a facility can demonstrate compliance for a NAAQS and not also comply with applicable PSD increment(s). Pollutant Class 1 Class 2 Class 3 Fine Paniculate Matter (PM2.5) Annual mean 24-hr maximum l 2 4 9 8 18 Particulate Matter (PMio) Annual arithmetic mean 24-hr maximum 4 8 17 30 34 60 Sulfur Dioxide (SO2) Annual mean 24-hr maximum 3-hr maximum 2 5 25 20 91 512 40 182 700 Nitrogen Dioxide (NO2): Annual mean 1-hr maximum 2.5 NA 25 NA 50 NA Table 2 - PSD Increment for Criteria Pollutants by Class 1, 2, and 3 Areas. 3.0 Approach to evaluating near-source and long-range impacts In order to assess the nature of long range transport aspects of a PSD compliance demonstration, a variety of facility types were modeled across several inert criteria pollutants for a range of meteorology conditions to improve our understanding of the source impacts in the near-field (i.e., within 50 km) and far-field (beyond 50 km). 3.1 Source Types and Characteristics Table 3 provides a summary of the source types that were included in a modeling study conducted by EPA state agencies under the AERMOD Implementation Workgroup (AIWG). In 2011, EPA re-instituted the AIWG with a focus on the new 1-hour NO2 and SO2 NAAQS. The purpose of the workgroup was to provide insights into challenges being brought forward by stakeholders regarding modeling as part of compliance demonstrations for the new standards (Snyder & Thurman, 2012). The workgroup focused on modeling of "real world" examples utilizing existing and newly formed guidance for the NO2 (U. S. EPA, 2010a) and SO2 NAAQS (U. S. EPA, 2010b). The AIWG workgroup was composed of EPA staff from the Regional offices and the Office of Air Quality Planning and Standards (OAQPS) as well as modelers from state, territorial, and local air quality agencies. The workgroup compiled a list of source types or facilities that were of interest to various state and local agencies. 10 ------- For each modeled facility, emissions and source characteristics were based on actual facilities from past permitting experiences but were modified to be generic facilities. AIWG participants conducted several modeling scenarios across multiple regions of the country that reflected changes stack height, addition of controls, and modifications of facility boundaries reflecting changes in ambient air. Also for NO2 sources, the modeling scenarios involved comparing the use of available Tier 3 methods under Appendix W: Plume Volume Molar Ratio Method (PVMRM) and Ozone Limiting Method (OLM). For complete details of the AIWG modeling study and results, the full report is available at:http://www.epa.gov/ttn/scram/10thmodconf/review material/AIWG Summary.pdf and http://www.epa.gov/ttn/scram/10thmodconf/review material/AIWG Summary v2.pdf. Table 3 - Summary of Modeling Scenarios by Source Type from AERMOD Implementation Workgroup (AIWG) Pulp & paper plant Base emission (tpy) NO2/SO2 Stack heights (m)2 Fuel oil turbine Landfill gas turbine NG compressor 104/6083 1184/417 80/45 90/0.01 9657/3403 25, 25, 25, 253, 253, 253, 251, 61, 61, 61 13,13,13,10 11, 11, 11, 11, 53 30, 30 ,29, 85, 85, 72, 72, 76, 8, 67, 67 3.2 Modeling Assessment of Various Facility Types The goal of this modeling assessment is to determine what types of facilities may have significant impacts, as defined as modeled concentrations above the SIL, at distances greater than 50 km from the source. There are fundamentally 4 parameters that affect this result: 1) the source configuration (e.g., stack height), 2) the emission rate, 3) the meteorology in the geographic area, and 4) the terrain in the geographic area. The source configurations and meteorology are closely tied when determining maximum impacts from a facility. These two aspects of a modeling compliance demonstration under PSD are more constrained than the possible range of emissions. Any facility, whatever source configuration and meteorology, can have significant impacts at any distance if the emissions are high 2 Primary stacks for NO2, for SO2, for both 3 NO2 only, no SO2 emissions at this stack height. 11 ------- enough. Thus, it is essential to have emissions that reflect reality with respect to the facility type, which makes the AIWG modeling scenarios ideal for this evaluation. Each original AIWG scenario was evaluated with a limited number of meteorological scenarios, which generally originated in the vicinity of the physical location of the facility that the scenario was based upon. However, to expand the usefulness of these scenarios for the purposes of this assessment and to more efficiently evaluate maximum potential impacts, a three-phased analysis of was used here: 1. In the first phase, screening meteorology, generated by the MAKEMET tool included in AERSCREEN (U. S. EPA, 2011), was used to evaluate a source's maximum 1-hour concentration impacts that could occur across these "worst-case" meteorological datasets. The screening meteorology includes conditions ranging from low wind & high stability cases, which would give highest concentrations for near-surface releases, to high wind and highly unstable conditions, which would give the highest concentrations for elevated release, with tall stacks. In addition to using screening meteorology, this initial phase used the SCREEN option in AERMOD, which determines plume centerline concentrations, regardless of the wind direction and source/receptor spatial relationships. When using this option with multiple sources in a single AERMOD run, the estimated concentrations are biased to be higher because each receptor will see the plume centerline from each source, which generally could not occur regardless of the wind direction. For these model simulations, receptors were placed at distances ranging from 100 m to 60 km. Multiple receptor heights were used in order to evaluate the potential presence of terrain downwind, with receptor heights including 0, 25, 65, 100, 150, and 200 m (Figure 1). Due to the plume centerline option being used, only a single receptor at each distance and height was required. Thus, the results from this initial phase represent an extremely conservative estimate of plume impacts. If the results from a facility showed impacts less than the SIL under this initial phase, then it is not expected for it to have impacts greater than the SIL for a refined single-source analysis. 2. The second phase of more refined modeling was pursued for those facilities whose predicted impacts from the first phase were above the SIL at 50 km. For this phase, AERMOD's CO SCREEN option was not used and several sets of actual meteorology were used instead of screening meteorology. One set of the actual meteorology selected was known to cause higher concentrations in surface releases, while one was known to cause higher concentrations with elevated releases. Additional meteorological data sets were used to provide more complete and robust findings. Since the plume centerline concentrations were not calculated in this phase, a polar receptor grid was used with 1-degree radial spacing through 360 degrees. Receptor distances and heights matched those from phase 1. For this second phase, we calculated 1-hour, 3-hour, 24-hour and annual concentrations to compare to the SILs for various averaging periods. 3. The third phase was pursued if results from the second phase still indicated impacts above the SIL at 50 km, wherein the facilities were evaluated against the relevant NAAQS for each averaging period. If the facility's impact was above the NAAQS in the near field, then the emissions were scaled down such that the source's impact satisfied the NAAQS in the near-field and its impacts at 50 km and beyond were evaluated and compared to the SIL for the appropriate averaging periods. This phase was based off the modeling in phase 2 by linearly scaling concentrations rather than scaling emissions and rerunning model simulations. 12 ------- 60000 -60000 Receptor locations -60000 -30000 0 30000 Distance from center, m 60000 200 150' ,100' (L> X 50 4. Receptor locations -•—o—• -•—o—• -B- 0 _ 20000 40000 60000 Distance from center along each radial, m Figure 1 - Location of receptors for screening and refined runs/ 4 Panel A shows the x and y locations, with refined receptors colored in black and screening receptors in red. Panel B shows the x and z locations for all receptors along any x/y radial. 13 ------- The AIWG scenarios included NO2 and SO2 emissions only, with NO2 using full conversion. However, PSD increments exist for CO, PM10, and PM2.5. Since CO air quality levels and emissions are currently so low, this pollutant is rarely an issue in PSD permitting and therefore were not evaluated in this assessment. However, PM10 and PM2.5 are necessary to include in this assessment and evaluated along with NO2 and SO2. Since PM emissions were not included in the original AIWG scenarios, an analysis of the EPA's 2011 National Emissions Inventory (NEI, http://www.epa.gov/ttn/chief/net/2011inventory.html) was conducted to determine scaling factors for PM emissions, based on the NO2 and SO2 emissions for a particular facility type. Additional details on this analysis are provided in the next section. 4. Summary of results 4.1 NO2 and SO2 screening The first phase of screening analysis was conducted for NO2 and SO2 only. The results are summarized in Table 4 and Table 5 and in multiple figures are provided in Appendix A. For these screening runs, the maximum concentration for both NO2 and SO2 at each receptor was recorded. In addition, the 8th high 1-hour concentration for NO2 and the 4th high 1-hour concentration for SO2 were also recorded at each receptor. The 8th and 4th high 1-hour concentrations are the concentrations typically used for determining compliance in a refined cumulative modeling analysis. While results are conservative given the plume centerline concentrations calculated in the screening analysis, they give some sense of the distribution of the highest concentrations and provide valuable perspective on the maximum concentrations. Several of the facility impacts were below their respective 1-hour SILs at the 50 km distance. The flare was below the SIL for NO2 and the asphalt plant and biomass burning plant were below the SIL for SO2. The landfill gas turbine and natural gas (NG) compressor were below the SIL at 50 km for both NO2 and SO2. Several other facilities were very close to the SIL at 50 km or had some receptors above and some below the SIL, depending on the receptor height. Only the NG compressor was compliant with any NAAQS, which has negligible SO2 emissions. While several facilities were close to the NAAQS, most were orders of magnitude higher than the NAAQS for both the highest concentration and the associated ranks concentration. The focus of the evaluation here is to determine if source impacts are above the SILs at 50 km. However, as noted in Table 4 and Table 5, most of the facilities are fairly close to the applicable SILs at 50 km but well above the NAAQS in the near field. While the analysis of the SILs at 50 km is largely within the context of the PSD increment, an analysis of PSD increment would be completed in conjunction with an analysis of NAAQS. Thus, it is reasonable to assume that NAAQS compliance in the near field would need to be met before considering the source impacts at 50 km and comparing to the applicable SIL, particularly if the concentrations in the near field are orders of magnitude above the NAAQS (i.e.., the NAAQS is the controlling standard). When the emissions from these facilities are scaled back to meet the NAAQS in the near field, the impacts from most of the facilities at 50 km fall well below the applicable SILs. For example, for the fuel oil turbine, the equivalent of the design value from the screening runs for NO2 (the 8th maximum 1-hr average) was 1115 ppb (not shown), while the standard is 100 ppb and the maximum impact at 50 km was 12.5 ppb. The emissions would need to be reduced by a factor of 11 to pass a NAAQS demonstration. This reduction would also apply to the impact at 50 km, resulting in an 14 ------- Table 4 - Results for Phase 1 screening modeling for NO2 and SO2 by Facility type5 Facility Asphalt plant Biomass Cement kiln Coal EGU Ethanol plant Receptor Elevation 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 NO2 NAAQS (impacts in ppb) Maximum 1- hour impact 291.0 856.8 884.7 408.3 181.1 129.8 17.5 17.7 548.5 899.4 534.6 243.7 148.7 149.1 150.4 152.2 1468.9 6771.0 232.8 970.7 434.3 222.1 517.2 1572.6 391.4 811.9 1735.2 1746.1 1871.6 1109.3 Impact at 50 km (SIL4ppb) 5.7 6.2 4.4 2.3 1.6 1.4 1.0 1.0 1.7 5.5 2.8 1.5 12.6 12.7 12.7 12.7 12.7 21.7 5.0 6.0 3.8 3.8 3.8 6.1 6.4 6.4 7.4 15.6 13.7 6.0 SO2 NAAQS (impacts in ppb) Maximum 1- hour impact 96.1 115.4 50.0 49.7 49.3 49.0 11.2 11.3 351.4 576.2 342.5 156.1 60.8 61.0 61.5 62.2 567.6 2397.5 74.8 74.9 75.8 77.3 1328.8 4388.2 32.2 175.6 1091.4 1223.5 1311.5 777.3 Impact at 50 km (SILSppb) 0.7 0.7 0.2 0.2 0.1 0.1 0.7 0.7 1.1 3.5 1.8 1.0 5.2 5.2 5.2 5.2 5.2 8.8 6.9 6.9 6.9 6.9 6.9 16.7 3.0 3.0 4.7 11.0 9.6 4.2 15 ------- Table 5 - Results for Phase 1 screening modeling for NO2 and SO2 by Facility type, continued5 Facility Flare Fuel oil turbine Landfill gas turbine NG compressor Pulp& paper plant Receptor Elevation 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 NO2 NAAQS (impacts in ppb) Maximum 1- hour impact 4.0 17.2 95.8 132.3 150.8 90.2 78.3 212.4 770.3 808.5 1126.0 907.9 28.1 110.5 230.5 174.0 75.5 42.4 92.6 835.8 351.9 181.9 59.9 43.1 303.0 290.6 815.1 9733.0 10313.8 8546.9 Impact at 50 km (SIL4ppb) 0.3 0.3 0.7 1.5 1.2 0.6 3.2 3.2 5.8 10.8 12.5 7.4 1.0 1.1 1.9 1.1 0.5 0.4 3.2 3.7 1.8 1.2 1.0 1.0 26.7 26.8 26.8 43.7 85.8 65.6 SO2 NAAQS (impacts in ppb) Maximum 1- hour impact 207.2 897.8 5014.2 6924.3 7888.0 4718.9 22.0 64.1 275.6 231.8 403.9 325.1 13.0 65.7 136.9 104.0 45.1 25.3 0.0 0.1 0.1 0.0 0.0 0.0 452.9 603.3 3438.1 2718.0 2542.2 2235.4 Impact at 50 km (SILSppb) 16.4 16.4 37.7 77.3 60.9 29.9 0.9 0.9 1.4 3.5 4.5 2.7 0.4 0.5 1.0 0.6 0.3 0.2 0.0 0.0 0.0 0.0 0.0 0.0 11.3 11.3 18.0 12.0 21.8 18.4 5 The screening runs have several layers of conservatism built into them. Firstly, the screening meteorology is designed to pick the meteorology that will result in the worst case concentrations for any facility configuration. Secondly, plume centerline concentrations have been calculated, giving the highest concentration possible from along a cross section of a plume. Thirdly, the plume centerline concentrations are occurring for each source simultaneously. While there is some physical possibility of the worse-case met occurring and directing a plume such that a receptor is along the plume centerline, the third case is generally not possible. 16 ------- impact of 1.4 ppb at 50 km, which is well below the SIL of 4 ppb. The pattern that emerges from this adjustment to emissions indicates that facilities with emissions from the taller stacks (e.g., the cement kiln and the coal EGU) are the only ones that still have some impacts above the applicable SIL at 50 km. Facilities with the lower release points have their maximum impacts in close proximity to the fence line so their impacts at 50 km are much lower after the adjustment to meet the NAAQS. Facilities with emissions concentrated at tall stacks do not see their maximum impacts until much farther downwind because the plumes need more time to impact the surface. Of course, this pattern is significantly influenced by including receptors that are close to stack height. When looking at only the elevated receptors, the emissions scaling necessitated to bring the elevated receptors to NAAQS levels replicates the pattern of maximum near field impacts for facilities with lower release heights. The findings from the screening results is that most types of facilities in most terrain will have their maximum impacts relatively close to the source. When the emissions from these facilities are adjusted to meet the NAAQS, the far-field impacts are unequivocally less than the SIL such that no assessment of LRT would be necessary. While terrain is a complicating factor, the results also show that the only facility type that could pass a NAAQS assessment in the near-field but still have impacts above the applicable SIL at distances of 50 km or greater are those with very tall stacks. 4.2 Refined analyses for NO2 and SO2 Based on the findings from the initial screening analysis, the second phase of refined modeling, which uses 5 years of actual meteorology and computes NAAQS and increment appropriate modeled design values, focuses solely on those facility types with tall stacks. Both the coal EGU and the cement kiln fit this profile, but we focused only on the coal EGU because its SO2 emissions were concentrated at the tallest stack and its NO2 emissions were distributed to the two tallest stacks (150 m and 100 m stack heights), thereby providing the most insightful test case for refined modeling. Table 6 shows the specific emissions for each stack at the coal EGU scenario for NO2 and SO2. Table 6 - Summary of NO2 and SO2 emissions for the coal EGU facility Stack height (m NO2 emissions (tpy) SO2 emissions (tpy) For the refined modeling analysis, we used four meteorological datasets consisting of 5 years of meteorological data from 2006 to 2010, reflecting National Weather Service (NWS) stations located at JFK airport in Ashland, Wl (ASX), Somerset airport in Somerville, NJ (SMQ), Dalhart airport in Dalhart, TX (DHT), and Oakland airport in Oakland, CA (OAK). These datasets were selected based on prior usage of a large set of meteorological datasets regularly used by the EPA and are known to represent a range of meteorological conditions. These data sets also provide spatial variability in the refined modeling analyses. The airports are all ASOS sites, with 1-minute observations processed through AERMINUTE and AERMET using the beta u* adjustment option (U. S. EPA, 2014). The maximum 1-hour concentration (5- year average), NAAQS specific design value (5-year average of 98th percentile and 99th percentile daily 1- 1564 4867 174 87 104 4.3 10.4 .7 10.4 .3 ' The 150 m stack is above GEP height, but was modeled at this height anyway. 17 ------- hour maximum concentrations), annual average (5-year average of each annual average), and 24-hour and 3-hour increment levels (highest first and second values) for SO2. As noted above, a polar grid was used with 1-degree separation and 360 degrees (receptor distances and heights matched those used in the screening analysis). Table 7 - Results from the Phase 2 refined analysis for NO2 Met scenario Receptor Elevation Maximum 1-hr impact (ppb, 1st high) All receptors 50km(SIL4ppb) 98th percentile 1-hr impact (ppb, 8th high) All receptors 50km(SIL4ppb) ASX DHT SMQ 25 65 100 150 200 25 65 100 150 200 25 65 100 150 200 25 65 100 150 200 546.81 505.18 373.36 301.35 311.18 286.22 357.80 340.30 269.71 216.93 199.68 239.83 417.98 397.27 312.91 250.09 245.54 212.50 443.30 425.08 302.98 243.67 253.43 250.14 1.50 1.66 1.34 1.27 1.39 1.60 1.38 1.45 1.32 1.28 1.35 1.50 1.62 1.75 1.53 1.52 1.56 1.91 1.15 1.22 1.20 1.09 1.19 1.32 366.15 344.43 270.46 217.27 217.01 191.64 235.82 210.89 129.94 118.30 128.86 127.65 318.30 302.32 239.16 191.94 146.20 138.62 344.51 370.58 227.72 183.29 189.48 184.82 0.94 1.04 0.97 0.85 0.99 1.15 0.75 0.76 0.77 0.73 0.80 0.96 1.16 1.17 1.15 1.12 1.14 1.19 0.72 0.81 0.81 0.72 0.94 1.08 Plots summarizing the refined modeling analysis for NO2 and SO2 are presented in Appendix B and the results are summarized in Table 7 and Table 8. The most striking difference from the screening and refined analysis is the decrease in the maximum 1-hour values. For NO2, the maximum from the screening (with unsealed emissions) was 1572 ppb, while the maximum from the refined runs was around 546 ppb (again, with unsealed emissions). For SO2, the maximum concentrations decrease even 18 ------- Table 8 - Results from the Phase 2 refined analysis for SO2 Met scenario Receptor Elevation Maximum 1-hr impact (ppb, 1st high) All receptors 50 km (SIL 4 ppb) 99th percentile 1-hr impact (ppb, 4th high) All receptors 50 km (SIL 4 ppb) ASX DHT OAK 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 0 25 65 100 150 200 23.65 23.69 24.08 34.60 68.70 262.24 23.39 23.60 23.97 49.53 87.43 652.11 28.98 29.01 29.23 43.59 82.33 394.67 25.57 25.63 25.75 34.85 49.22 177.58 2.58 2.57 2.57 2.56 2.56 3.20 2.47 2.46 2.46 2.46 2.46 3.15 3.01 2.99 2.97 2.96 2.95 4.04 2.08 2.07 2.06 2.05 2.04 2.40 14.22 14.31 14.68 24.57 57.76 200.78 12.04 12.07 12.40 42.82 53.61 454.48 21.36 21.52 21.75 28.06 67.46 329.24 16.24 16.35 16.47 24.03 38.58 139.80 1.98 1.97 1.96 1.95 1.95 2.47 1.81 1.80 1.78 1.77 1.76 2.28 2.65 2.64 2.63 2.63 2.62 2.97 1.63 1.63 1.62 1.62 1.62 2.12 further from 4388 ppb to 454 ppb. For NO2, the maximum 1-hour concentrations at 50 km from the refined modeling (with unsealed emissions) are now well below the 1-hour SIL at X to Y. However, the SO2 concentrations are still above the SIL at 50 km with the unsealed emissions, ranging from F to G. If the SO2 emissions are scaled such that the near-field results meet the NAAQS, then the impact at 50 km are below the 1-hour SIL at W to V. However, the maximum impacts in the near-field for SO2 are driven by results at the 150 and 200 m receptors, i.e., elevated receptors reflective of terrain features in vicinity of the facility. The elevated receptors were included to evaluate the potential impacts of terrain on the modeling results. However, realistically, a facility with a 150 m stack with large emissions is not likely to be built in the immediate vicinity of terrain near or above stack height, as the facility would not be able to pass a NAAQS compliance demonstration. Thus, it is somewhat unrealistic to consider these elevated receptors at the closest distances. If these elevated receptors are not considered in the near-field, then 19 ------- the EGU would pass the NAAQS in the near-field, but be above the SIL in the far-field. Thus, the near- field concentrations would not indicate a NAAQS violation because there would be no receptors near stack height and that a source of this type could indeed have significant impacts at 50 km or greater and need an LRT assessment. For the other averaging times, the maximum concentrations or estimated increment contributions are well below the SILs and increments at 50 km, even with the unsealed emissions. If emissions were scaled to pass the 1-hour NAAQS, then the concentrations at 50 km for the other averaging periods would be even farther below the benchmarks being considered here. In general, these results show that the longer the averaging period, the less likely that there will be significant impacts at distances of 50 km and greater. 4.3 Refined analyses for PM10 and PM25 Since the AIWG facilities did not include PM10 or PM2.5 emissions, we derived these emissions by scaling from the emission rates for SO2 and NO2 based on emission ratios of these pollutants for EGUs listed in the NEI. The ratios of PM10 and PM2.5 emissions to NOx emissions for all EGUs with NOx emissions greater than 40 tons were calculated. Similarly, emission ratios were computed for EGUs with SO2 emissions greater than 40 tons. PM emissions were significantly lower than NOx and SO2 emissions. The results are summarized in Table 7. On average, the PM10 emissions were about 22% of NOx and SO2 emissions, while PM2.5 emissions were around 19% of NOx and SO2 emissions. The facilities with the greatest PM10 and PM2.5 emission ratios resulted in PM10 and PM2.5 being 50% of NOx and 38% of SO2. Given how close PM10 and PM2.5 emission ratios were, this assessment focused on PM2.5 emissions only, as the PM2.5 standard are more stringent than the PM10 standards. Since both PM10 and PM25 would be modeled as inter pollutants, the model would treat each pollutant equally with respect to dispersion, so modeling both PM10 and PM25 with the approximately the same emission rates would result in roughly the same modeled concentrations. The average emission ratios (20%) were used to scale the PM emissions for the PM analysis. Table 9 - Summary of 2011 NEI emission data used in refined modeling analysis NOx ratios PMlOmax 8.0% PMlOmin 49.4% PMlOmean 22.9% PM25 max 4.8% PM25 min 49.2% PM25 mean 19.5% SO2 ratios PMlOmax 3.1% PMlOmin 37.7% PMlOmean 22.6% PM25 max 2.7% PM25 min 37.6% PM25 mean 18.7% Plots summarizing the refined modeling analyses for PM2.5 are presented in Appendix C. The results from the annual average are well below the annual SILs at 50 km. For the 24-hour concentrations, it is difficult to draw specific conclusions from the results, given the uncertainty in the applicable SIL. The maximum concentrations are certainly well below the increment and below some of the 24-hr Class I SILs that are available. For the NO2-scaled emissions, the NAAQS near-field concentrations are slightly above the NAAQS (16 ug/m3) and right at the lowest PM25 SIL option. The SO2-scaled emissions have slightly higher maximum impacts (20 ug/m3) but are in between the two lowest PM25 SIL options. For the NO2 case, when emissions are scaled down to meet the NAAQS, the 50 km impacts become solidly below the SIL, while the SO2 50 km impacts are still higher than the lowest PM25 SIL option. 20 ------- 5. Conclusions The results from this analysis indicate that for most facility types, if the facility can confirm compliance with the short-term standards in the near-field, then there are not likely to be significant impacts at 50 km. Thus, for most facilities that show compliance in the near-field, no evaluation of LRT would be needed. There are indications, however, that for a select class of facilities, mainly those that have very tall stacks (greater than 100 m), there is a possibility of having an impact that is significant with respect to the short-term NO2, SO2 and PM25 SILs. These types of facilities have their maximum impact much farther from the facility than most, which means a higher emissions rate would be acceptable. The results also indicate that terrain features can be important for these types of facilities, as elevated terrain, near stack height, can see higher plume impacts much closer to the source. When this occurs, the compliance with the short-term standard may be sufficient to decrease long-range impacts and eliminate the potential need for an LRT assessment. Conversely, elevated receptors in the far-field can increase the need for LRT assessments, as these receptors would experience impacts closer to the plume centerline. Given the GEP restrictions on stack heights, there are not likely to be many facilities with stacks tall enough to pass the near-field NAAQS requirement while still having significant impacts at 50km. 6. Additional information Data for the analyses described in this TSD can be obtained by contacting: Chris Owen, PhD Office of Air Quality Planning and Standards, U. S. EPA 109 T.W. Alexander Dr. RTP, NC 27711 919-541-5312 owen.chris@epa.gov 21 ------- References Cimorelli, et al. (2005). AERMOD: A Dispersion Model for Industrial Source Applications. Part I: General Model Formulation and Boundary Layer Characterization. J. App. Meterol, 682-693. Snyder, E., & Thurman, J. (2012). AERMOD Implementation Workgroup NO2 & SO2 modeling. 10th Conference on Air Quality Modeling. U. S. EPA, RTP, NC. U. S. EPA. (1996). 40 CFR Parts 51 and 52: Prevention of Significant Deterioration (PSD) and Nonattainment New Source Review (NSR). FR, Vol. 61, No. 142, July 23,1996, 38249 - 38344. U. S. EPA. (2003). Guideline on Air Quality Models. 40 CFR Part 51 Appendix W (68 FR 18440). U. S. EPA. (2010a, June 29). Guidance Concerning the Implementation of the 1-hour NO2 NAAQSfor the Prevention of Significant Deterioration Program. U. S. EPA. (2010b, August 23). Guidance Concerning the Implementation of the 1-hour SO2 NAAQSfor the Prevention of Significant Deterioation Program. U. S. EPA. (2011, March). AERSCREEN User's Guide, pp. EPA-454/B-11-001. U. S. EPA. (2014, May). Addendm to User's Guide for the AERMOD Meteorological Preprocessor (AERMET), EPA document number EPA-454/B-03-002. Retrieved from RTP, NC. U. S. EPA. (2014). Guidance for PM2.S Permit Modeling, EPA report number EPA-454/B-14-001. RTP, NC, May. 22 ------- Appendix A Plots from NO2 and SO2 screening analysis For all figures in this section, the NO2 SIL is shown in blue and the SO2 SIL is shown in red. 23 ------- - CL- Q. CM O NO2 asphalt, max results 10 20 30 40 downwind dist, km 50 -••o -± 25 » 65 -+- 100 -& 150 60 ^200 halt, eiqth hiqh results 20 30 40 downwind dist, km eo e 200 SO2 asphalt, max results to 20 30 40 downwind dist, km 50 60 e 200 SO2 asphalt, fourth high results to 20 30 40 downwind dist, km 50 Figure 2 - Results from the screening analysis for the asphalt plant. 25 65 +- 100 & 150 60 e 200 24 ------- NO2 biomass, max results 10 20 30 40 downwind dist, km 50 60 - 25 » 65 -+- 100 a 150 200 CM O NO2 biomass, eigth high results to 20 30 40 downwind dist, km 50 60 - 25 » 65 -+- 100 a 150 r 200 SO2 biomass, max results to 20 30 40 downwind dist, km 50 25 65 100 150 60 e 200 SO2 biomass, fourth high results to 20 30 40 downwind dist, km 50 60 - 25 » 65 -+- 100 a 150 200 Figure 3 - Results from the screening analysis for the biomass plant 25 ------- NO2 cement kiln, max results ic 20 30 40 downwind dist, km 50 -••o -± 25 » 65 -+- 100 -& 150 60 ^200 . CL Q. CM O NO2 cement_kiln, eigth high results : 20 30 40 downwind dist, km 50 - 25 » 65 -+- 100 -& 150 60 ^200 SO2 cement kiln, max results ic 20 30 40 downwind dist, km 50 •A 25 *• 65 -+- 100 -& 150 60 ^200 SO2 cement_kiln, fourth high results ic 20 30 40 downwind dist, km 50 25 65 100 150 60 200 Figure 4 - Results from the screening analysis for the cement kiln 26 ------- NO2 coal_egu, max results ic 20 30 40 downwind dist, km ^- 25 » 65 -+- 100 •a 150 60 ^200 NO2 coal_egu, eigth high results : 20 30 40 downwind dist, km 25 65 +- 100 150 60 e 200 jQ Q. CL CM O to ic SO2 coal_egu, max results 20 30 40 downwind dist, km 50 •A 25 *• 65 -+- 100 S 150 60 ^200 SO2 coal_egu, fourth high results ic 20 30 40 downwind dist, km 50 - 25 » 65 -+- 100 -& 150 60 ^200 Figure 5 - Results from the screening analysis for the coal ECU 27 ------- NO2 ethanol, max results ic 20 30 40 downwind dist, km 25 65 +- 100 150 60 e 200 NO2 ethanol, eigth high results : 20 30 40 downwind dist, km 50 25 65 +- 100 150 60 e 200 SO2 ethanol, max results ic 20 30 40 downwind dist, km 50 60 e 200 SO2 ethanol, fourth high results ic 20 30 40 downwind dist, km 50 60 200 Figure 6 - Results from the screening analysis for the ethanol plant 28 ------- NO2 flare, max results 10 20 30 40 downwind dist, km 50 60 e 200 NO2 flare, eigth high results 20 30 40 downwind dist, km - 200 1C SO2 flare, max results 20 30 40 downwind dist, km 50 -••o -± 25 -M- 65 -+- 100 •a 150 60 ^200 SO2 flare, fourth high results : 20 30 40 downwind dist, km 50 -•- o -± 25 » 65 -+- 100 -a 150 60 ^200 Figure 7 - Results from the screening analysis for the flare 29 ------- NO2 fuel oil turbine, max results 1C 20 30 40 downwind dist, km 50 60 e 200 NO2 fuel_oil_turbine, eigth high results : 20 30 40 downwind dist, km 150 60 e 200 SO2 fuel oil turbine, max results to 20 30 40 downwind dist, km 50 25 • 65 +- 100 & 150 60 e 200 SO2 fuel_oil_turbine, fourth high results to 20 30 40 downwind dist, km 50 25 65 +- 100 150 60 200 Figure 8 - Results from the screening analysis for the fuel oil turbine 30 ------- NO2 landfill_gas_turbine, max results 10 20 30 40 downwind dist, km 50 ^- 25 » 65 -+- 100 a 150 60 ^200 NO2 landfill_gas_turbine, eigth high results to 20 30 40 downwind dist, km 50 60 - 25 » 65 -+- 100 -& 150 r 200 SO2 landfill_gas_turbine, max results to 20 30 40 downwind dist, km 50 -••0 -± 25 *• 65 -+- 100 -a 150 60 ^200 SO2 landfill_gas_turbine, fourth high results to 20 30 40 downwind dist, km 50 -* 25 » 65 H- 100 a 150 60 ^200 Figure 9 - Results from the screening analysis for the landfill gas turbine 31 ------- CM O NO2 ng_compressor, max results 10 20 30 40 downwind dist, km 50 60 - 25 » 65 -+- 100 -& 150 200 NO2 ng_compressor, eigth high results to 20 30 40 downwind dist, km 50 -•• 0 -± 25 » 65 -+- 100 -& 150 60 ^200 SO2 ng_compressor, max results 20 30 40 downwind dist, km 50 •A 25 -M- 65 -+- 100 •a 150 60 ^200 SO2 ng_compressor, fourth high results 20 30 40 downwind dist, km 50 -* 25 » 65 -+- 100 -a 150 60 ^200 Figure 10 - Results from the screening analysis for the natural gas compressor station 32 ------- .a Q. Q. 8 NO2 pulp_paper, max results : 20 30 40 downwind dist, km 50 Z_hill -••o -± 25 » 65 -+- 100 a 150 60 ^200 iqth hiqh results : 20 30 40 downwind dist, km 50 -•-o -± 25 » 65 -+- 100 -& 150 60 ^200 CL CL CM O CO SO2 pulp_paper, max results ic 20 30 40 downwind dist, km 50 -••o -± 25 *• 65 -+- 100 •a 150 60 ^200 S02 pul er, fourth high results ic 20 30 40 downwind dist, km 50 - 25 » 65 H- 100 -a 150 60 ^200 Figure 11 - Results from the screening analysis for the pulp and paper plant 33 ------- Appendix B Plots from NO2 and SO2 refined analysis For all figures in this section, the NO2 SIL is shown in blue and the SO2 SIL is shown in red. 34 ------- NO2 ASX_NAAQS, max results T 20 30 40 downwind dist, km -•- 65 -+- 100 •& 150 60 200 NO2 ASX_NAAQS, eigth high results 20 30 40 downwind dist, km 60 200 so2 ASX NAAQS, max results 20 30 40 downwind dist, km 60 200 SO2 ASX_NAAQS, fourth high results i: T 20 30 40 downwind dist, km 50 25 65 100 150 60 ^200 Figure 12 - Refined NO2 and SO2 results, ASX, max and design values, 1-hr NAAQS 35 ------- NO2 DHT_NAAQS, max results 20 30 40 downwind dist, km 5C 60 200 NO2 DHT_NAAQS, eigth high results 20 30 40 downwind dist, km 50 Z hill 60 200 so2 DHT NAAQS, max results 20 30 40 downwind dist, km 60 200 SO2 DHT_NAAQS, fourth high results i: 20 30 40 downwind dist, km 50 z_mn -•- o 25 60 200 Figure 13 - Refined NO2 and SO2 results, DHT, max and design values, 1-hr NAAQS 36 ------- NO2 OAK_NAAQS, max results i 20 30 40 downwind dist, km 5C 60 150 200 NO2 OAK_NAAQS, eigth high results 20 30 40 downwind dist, km 50 Z hill 60 200 so2 OAK NAAQS, max results 20 30 40 downwind dist, km 25 60 200 SO2 OAK_NAAQS, fourth high results i: 20 30 40 downwind dist, km 50 z_mn -•- o 25 60 200 Figure 14 - Refined NO2 and SO2 results, OAK, max and design values, 1-hr NAAQS 37 ------- NO2 SMQ_NAAQS, max results 20 30 40 downwind dist, km 5C 60 200 NO2 SMQ_NAAQS, eigth high results 20 30 40 downwind dist, km 50 60 200 so2 SMQ NAAQS, max results 20 30 40 downwind dist, km 60 200 SO2 SMQ_NAAQS, fourth high results i: 20 30 40 downwind dist, km 50 z_mii -•- o 25 60 200 Figure 15 - Refined NO2 and SO2 results, SMQ, max and design values, 1-hr NAAQS 38 ------- so2ASXH1H,3hr 20 30 40 downwind dist, km to so2 ASX H2H, 3hr 20 30 40 downwind dist, km 50 •-0 200 to so2ASXH1H,24hr 20 30 40 downwind dist, km 50 Z hill 60 200 to so2ASXH2H,24hr 20 30 40 downwind dist, km 50 Z hill 60 200 Figure 16 - Refined SO2 results, ASX, 3 & 24-hr increment 39 ------- so2DHTH1H,3hr 20 30 40 downwind dist, km 5C Q- 50 CM so2 DHT H2H, 3hr 20 30 40 downwind dist, km 50 100 150 200 so2DHTH1H, 24hr 20 30 40 downwind dist, km 6G i: so2 DHT H2H, 24hr 20 30 40 downwind dist, km 50 Z hill 100 150 200 60 Figure 17 - Refined SO2 results, DHT, 3 & 24-hr increment 40 ------- so20AKH1H,3hr 20 30 40 downwind dist, km 5C Z hill 50 so2 OAK H2H, 3hr 20 30 40 downwind dist, km Z hill 50 10 so2OAKH1H,24hr 20 30 40 downwind dist, km Z hill 6G 200 i: so2 OAK H2H, 24hr 20 30 40 downwind dist, km 50 Z hill 6C 200 Figure 18 - Refined SO2 results, OAK, 3 & 24-hr increment 41 ------- 8 10 so2 SMQ H1H, 3hr 20 30 40 downwind dist, km 5C : 50 - 100 |- 150 200 to so2 SMQ H2H, 3hr 20 30 40 downwind dist, km 50 Z hill 50 10 so2SMQH1H,24hr 20 30 40 downwind dist, km Z hill 6G 200 1 3 so2 SMQ H2H, 24hr 20 30 40 downwind dist, km 50 60 200 Figure 19 - Refined SO2 results, SMQ, 3 & 24-hr increment 42 ------- NO2 ASX ANN, max results o 20 30 40 downwind dist, km 0 50 100 150 200 so2 ASX ANN, max results 1C 20 30 40 downwind dist, km 50 6C Figure 20 - Refined NO2 & SO2 results, ASX, annual NAAQS 43 ------- 2.0 1.5 1.0 Q. CL 8 -z. NO2 DHT_ANN, max results 20 30 40 downwind dist, km 50 100 150 200 so2 DHT ANN, max results 1C 20 30 40 downwind dist, km 50 6C Figure 21 - Refined NO2 & SO2 results, DHT, annual NAAQS 44 ------- NO2 OAK_ANN, max results 10 20 30 40 downwind dist, km 100 150 200 so2 OAK ANN, max results 1 3 20 30 40 downwind dist, km 50 60 Figure 22 - Refined NO2 & SO2 results, OAK, annual NAAQS 45 ------- o NO2 SMQ_ANN, max results 20 30 40 downwind dist, km 50 so2 SMQ ANN, max results 20 30 40 downwind dist, km Figure 23 - Refined NO2 & SO2 results, SMQ, annual NAAQS 46 ------- Appendix C Plots from PM refined analysis For all figures in this section, the PM25 increment is indicated on each plot. For the NO-scaled emissions, this is shown in blue and the SO2-scaled emissions are shown in red. 47 ------- E10 so2-scaled ASX_PM25 H1H, 24hr 10 20 30 40 downwind dist, km 50 Z hill 50 CO no2-scaled ASX PM25H1H, 24hr 20 30 40 downwind dist, km 50 Z hill 50 60 CO no2-scaled ASX_PM25, ann avg 1C 20 30 40 downwind dist, km 50 Z hill so2-scaled ASX_PM25, ann avg 20 30 40 downwind dist, km •-0 50 Figure 24 - Refined PM25 results, ASX, 24-hr annual increment 48 ------- co CD so2-scaled DHT_PM25 H1H, 24hr 20 30 40 downwind dist, km 5C Z hill 100 150 200 no2-scaled DHT PM25H1H,24hr 20 30 40 downwind dist, km 50 CO no2-scaled DHT_PM25, ann avg 20 30 40 downwind dist, km 50 Z hill 100 150 200 6G so2-scaled DHT_PM25, ann avg 1C 20 30 40 downwind dist, km 50 6C Figure 25 - Refined PM25 results, DHT, 24-hr annual increment 49 ------- so2-scaled OAK_PM25 H1H, 24hr co 20 30 40 downwind dist, km 5C no2-scaled OAK PM25H1H,24hr 20 30 40 downwind dist, km 50 no2-scaled OAK_PM25, ann avg 20 30 40 downwind dist, km 50 Z hill 100 150 200 6G so2-scaled OAK_PM25, ann avg 1C 20 30 40 downwind dist, km 50 Z hill 100 150 200 6C Figure 26 - Refined PM25 results, OAK, 24-hr annual increment 50 ------- so2-scaled SMQ_PM25 H1H, 24hr 20 30 40 downwind dist, km 5C no2-scaled SMQ PM25H1H,24hr 20 30 40 downwind dist, km 50 60 no2-scaled SMQ_PM25, ann avg 20 30 40 downwind dist, km 50 6G so2-scaled SMQ_PM25, ann avg 20 30 40 downwind dist, km 50 Figure 27 - Refined PM25 results, SMQ, 24-hr annual increment 51 ------- so2-scaled ASX_PM25_24HR_NAAQS 8th high, 24hr 20 30 40 downwind dist, km 50 no2-scaled ASX_PM25_24HR_NAAQS 8th high, 24hr 1 3 20 30 40 downwind dist, km 50 6C Figure 28 - Refined PM25 results, ASK, 24-hr NAAQS 52 ------- so2-scaled DHT_PM25_24HR_NAAQS 8th high, 24hr 20 30 40 downwind dist, km 50 no2-scaled DHT_PM25_24HR_NAAQS 8th high, 24hr 1C 20 30 40 downwind dist, km 50 Figure 29 - Refined PM25 results, DHT, 24-hr NAAQS 53 ------- so2-scaled OAK_PM25_24HR_NAAQS 8th high, 24hr 20 30 40 downwind dist, km 50 no2-scaled OAK_PM25_24HR_NAAQS 8th high, 24hr 1 3 20 30 40 downwind dist, km 50 6C Figure 30 - Refined PM25 results, OAK, 24-hr NAAQS 54 ------- so2-scaled SMQ_PM25_24HR_NAAQS 8th high, 24hr 100 150 200 20 30 40 downwind dist, km no2-scaled SMQ_PM25_24HR_NAAQS 8th high, 24hr 1 3 20 30 40 downwind dist, km 50 6C Figure 31 - Refined PM25 results, SMQ, 24-hr NAAQS 55 ------- 56 ------- United States Office of Air Quality Planning and Standards Publication No. EPA- 454/B-15-003 Environmental Protection Air Quality Analysis Division [July, 2015] Agency Research Triangle Park, NC 57 ------- |