11

PM2.5 “Regulatory” Guidance Overview and Case Studies for Secondary Formation of PM2.5 from Precursors

Kasi Dubbs, P.E., C.M.

Midwest Environmental Compliance ConferenceMay 13, 2015

2

“Regulatory” History on PM2.5 and NSR/PSD (1/4)˃ July 18, 1997 – First PM2.5 NAAQS established (annual

= 15 µg/m3, 24 hr = 65 µg/m3)˃ October 23, 1997 – PM10 surrogate policy memo

(Use PM10 as a surrogate for PM2.5)˃ October 17, 2006 – PM2.5 24 hr NAAQS revised

from 65 µg/m3 to 35 µg/m3

˃ May 16, 2008 – EPA finalizes partial PM2.5 NSR and PSD Implementation Rule PSD and NSR Major Source Thresholds (MSTs) Significant Emission Rates (SERS) for direct PM2.5 as well

as secondary PM2.5 precursors (NOx and SO2) Offset ratios for PM2.5 as well as interpollutant trading

ratios

3

“Regulatory” History on PM2.5 and NSR/PSD (2/4)˃ February 11, 2010 – EPA announces PM10

Surrogate Policy coming to an end (Can no longer use PM10 as a surrogate for PM2.5 in modeling and BACT analyses)

˃ March 23, 2010 – Memo on Modeling Procedures for PM2.5 NAAQS (“Page Memo”) Addresses form of the standard for modeling vs.

form of NAAQS monitoring design value Special attention to background concentrations Speaks to technical complications with

secondary formed PM2.5, but refers to future guidance

4

“Regulatory” History on PM2.5 and NSR/PSD (3/4)˃ October 20, 2010 – EPA finalizes Final PM2.5

PSD Implementation Rule PM2.5 Significant Impact Levels (SILs) PM2.5 Increments PM2.5 Significant Monitoring Concentration

(SMC)˃ January 7, 2011 – National Association of Clear

Air Agencies (NACAA) issues PM2.5 Modeling Implementation Document PM2.5 Emission Inventories Secondary Formation of PM2.5 Background Concentrations for PM2.5

5

“Regulatory” History on PM2.5 and NSR/PSD (3/4)˃ May 16, 2011 - Surrogate policy ended July 21,

2011 – EPA revokes interpollutant trading ratios from May 2008 PM2.5 NSR/PSD Implementation Rule

˃ December 14, 2012 – PM2.5 Annual NAAQS revised from 15 µg/m3 to 12 µg/m3

˃ January 22, 2013 – Court decision: PM2.5 SMC vacated, PM2.5 SILs vacated and remanded

˃ May 20, 2014 – Final EPA guidance for PM2.5 permit modeling Emphasizes qualitative vs. quantitative approaches for

evaluating secondary PM2.5

6

Interpollutant Trading Ratios

˃ Sources applying for permits in PM2.5 nonattainment areas can offset emissions increases of direct PM2.5 emissions with reductions of PM2.5 precursors using interpollutant trading ratios (also called “PM2.5 offset ratios”). For example, an existing source can increase

PM2.5 emissions by X tons in exchange for reducing SO2 emissions by Y tons.

EPA 2008 (but later rescinded)♦ NOx:PM2.5 = 200:1 (Eastern US)♦ NOx:PM2.5 = 100:1 (Western US)♦ SO2:PM2.5 = 40:1 (Nation-wide)

7

Equivalent Primary PM2.5 Emissions˃ PM2.5 offset trading ratios can be used to account for

secondary formation of PM2.5.˃ NACCA introduced an approach in their 2011 report˃ Convert SO2 and NOx emissions into “equivalent” direct

PM2.5 emissions. 100:1 100 TPY SO2 = 1 TPY direct PM2.5 10:1 100 TPY SO2 = 10 TPY direct PM2.5 1:1* 100 TPY SO2 = 100 TPY direct PM2.5 0.5:1 100 TPY SO2 = 200 TPY direct PM2.5

*This is ~100% conversion of SO2 to (NH4)2SO4

˃ Model the total primary PM2.5 including the actual primary PM2.5 plus the total equivalent primary PM2.5 determined from the precursor emissions and offset ratio

˃ Lower PM2.5 offset ratios will produce more secondary PM2.5.

8

PM2.5 Offset Ratios and PSD (1/2)

˃ The approach for developing PM2.5 offset ratios involves the following steps: Define the geographic area Conduct a series of sensitivity runs with

appropriate air quality models to develop data that correlates modeled PM2.5 concentration changes associated with reductions of direct PM2.5 emissions and PM2.5 precursor emissions

9

PM2.5 Offset Ratios and PSD (2/2)˃ July 2014 Minnesota Modeling Guidance (based

on limited analysis) NOx:PM2.5 = 100:1 SO2:PM2.5 = 10:1

˃ Georgia Work in Progress Sensitivity runs to evaluate how PM2.5 offset ratios

varied by:♦ Distance from the source♦ Season of the year♦ Stack height♦ Grid resolution♦ Location in the state

Goal to provide county-specific offset ratios in modeling guidance

10

EPA May 2014 PM2.5 Guidance (1/2)

Compliance with PM2.5 NAAQS for PSD:

Modeled impact of primary PM2.5 emitted from industrial sources in an area

+ Impact of NOx and SO2 as Secondary

PM2.5+

Background Concentration

11

EPA May 2014 PM2.5 Guidance (2/2)EPA Recommended Approaches for Assessing Primary and Secondary PM2.5 Impacts

Source: Page 21 of EPA’s Draft Guidance for PM2.5 Permit Modeling, March 2013

AERMOD Model

12

Secondary PM2.5 Assessment Methods

˃ Qualitative - Develop “appropriate conceptual description of PM2.5”

˃ Hybrid Qualitative/Quantitative Use of local/region specific “offset ratios” for

precursor emissions Analysis of “offset ratios” for precursor

emissions specific to region (i.e., how readily nitrates and sulfates form in the modeled domain due to environmental influences such as NH3 concentrations)

˃ Quantitative - Photochemical models or other models, as applicable, i.e., CAMx or CMAQ

13Since May 2014 Guidance, What are States Doing?

˃ Some States have already been requiring, and will continue to require, hybrid approach type assessments for secondary PM2.5.

˃ Some States looking for guidance on what can justify qualitative versus hybrid versus quantitative.

˃ A common theme: case by case assessment. State requirements regarding secondary PM2.5 will be a

constantly changing and evolving theme over the coming years.

The analysis your facility may have been allowed to do in one State, may not be acceptable (or desired) in another State.

14

Summary of Secondary PM2.5 Assessments

Project Description

Case Approach Status

Virginia 1 Case 3 Hybrid Qualitative/Quantitative – Georgia EPD Offset Ratios

Permit Issued, No EPA comment

North Carolina 1 Case 3 Qualitative Application Under Review

North Carolina 2 Case 3 Qualitative Application Under Review

Michigan Generating Station

Case 3 Hybrid Qualitative/Quantitative – CSAPR Modeling Ratios Permit Issued, No EPA Comment

Texas Generating Station

Case 3 Hybrid Qualitative/Quantitative – CSAPR Modeling Ratios Application Under Review, Method approved by TCEQ

Texas LNG Plant Case 3 Qualitative Undergoing TX Contested Case Process

Louisiana 1 Case 3 Qualitative Permit Issued, No EPA comment

Alaska Oil/Gas Case 3 NA – relied on March 2010 Page memo using H1H model result

Permit Issued, No EPA comment

Idaho Fertilizer Plant Case 3 Hybrid Qualitative/Quantitative – Applied monitoring-based ratios

Draft Permit Issued, No EPA comment to-date

Oregon 1 Case 3 Simplified Hybrid Qualitative/Quantitative – Used codified offset ratios

Permit Issued, No EPA comment

Washington 1 Case 3 Hybrid Qualitative/Quantitative – Applied monitoring-based ratios

Permit Issued

Georgia 1 Case 3 NA – relied on March 2010 Page memo using H1H model result

Permit Issued, No EPA comment

South Carolina 1 Case 3 Qualitative Permit Issued, No EPA Comment

Georgia 2 Case 3 NA – relied on March 2010 Page memo using H1H model result

Modeling Protocol Reviewed, Application in Progress

Alabama 1 Case 3 NA – relied on March 2010 Page memo using H1H model result

Modeling Protocol Reviewed, Application in Progress

15

Missouri Case Study: CSAPR Response Factors vs Offset Ratios to Quantify Secondary PM2.5

˃ CSAPR utilizes CAMx, a photochemical model, to quantify impacts of SO2 and NOX emissions on PM2.5 concentrations at ambient monitoring locations around U.S.

˃ Can use data for monitors addressed in CSAPR to develop response factors (ratios of expected reduction in PM2.5 concentrations per ton of reduction in precursor emissions)

16

Case Study: Monitor Selection (1/3)˃ Guidance allows use

of existing monitor if representative.

˃ Proposed using PM2.5 data collected at an existing PM2.5 ambient air monitor in El Dorado Springs, Missouri.

17

Case Study: Monitor Options (2/3)˃ Located all PM2.5

monitors near facility.

˃ Narrowed options based on: Data currentness Data quality Speciation of PM2.5

Meteorological conditions in region

18

Case Study: Monitor Selection (3/3)˃ Which monitor is representative of

the concentration in area of proposed facility? Things to Consider:

♦ Wind direction & air parcel trajectories

♦ Climatology♦ Demographics of region♦ Surrounding sources of precursor

pollutants

19

Case Study: Wind Analysis (1/3)˃ Determine average wind speed and

direction near proposed facility location 5 year analysis for 3 nearby airports

20

Case Study: Wind Analysis (2/3)˃ Narrowed data by putting emphasis on

winter season months Cooler temperatures are more ideal for

nitrate formation and thus, secondary PM2.5 formation.

Wind roses using data from October – March for 5 years.

21

Case Study: Wind Analysis (3/3)

˃ Joplin Airport closest weather station to facility

˃ 20% of time analyzed had wind direction between 170º (S) and 200º (SSW)

˃ Grove & Monett - more variation Stronger SSE and

NW components

22

Case Study: Forward Trajectories (1/2)

˃ Demonstrates path air parcel took from a point of origin.

˃ Used NOAA’s HYSPLIT Model.

˃ Image of December 2013 trajectories.

˃ 13% (4 days) air advected within 17º sector of El Dorado Springs monitor.

23

Case Study: Forward Trajectories (2/2)˃ Findings:

Top 3 monitors with most trajectory hits from October 2013 – March 2014: ♦ Liberty, MO = 20 (11%)♦ El Dorado Springs, MO = 17 (9.3%)♦ Stilwell, OK = 13 (7.1%)

˃ Determined Liberty, MO monitor not representative of concentrations near proposed plant due to close proximity to densely populated Kansas City.

24

Case Study: Back Trajectories˃ Demonstrates path air parcels have taken

prior to arriving at monitor site.˃ Again used NOAA’s HYSPLIT (Hybrid

Single Particle Lagrangian Integrated Trajectory Model).

˃ Only used cold season days when the monitor showed exceedance of 12 µg/m3

(annual NAAQS primary PM2.5 standard) Information from EPA Air Quality website El Dorado Springs = 8 exceedances, 0 hits within

10 km Liberty = 17 exceedances, 1 hit Stilwell = 10 exceedances, 0 hits

25

Case Study: Final Justification for Monitor of Choice˃ Location

84 km northeast of facility♦ Far enough distance to allow secondary PM2.5 formation♦ Close enough to experience similar weather conditions

˃ Wind & Trajectory Analysis South and southwest winds common near Joplin Advect air toward monitor

˃ Data Quality Deployed by the state/local air monitoring

stations (SLAMS), reports to Missouri Laboratory Services Program

26

Case Study: Cross-State Air Pollution Rule (CSAPR)˃ Use CSAPR data to find 2014 base and control case

data in MO for annual and 24-hr PM2.5 design values.

˃ Calculate response factors to estimate the concentration of secondary PM2.5 per ton of NOx and SO2 per year.

27

Case Study: Cross-State Air Pollution Rule (CSAPR)˃ The response factor was multiplied by the

proposed project’s total SO2 and NOx PTE values to find estimated impact on secondary formation of PM2.5 at monitor.

28

Case Study: Cross-State Air Pollution Rule (CSAPR)

˃ Impacts of combined primary and secondary PM2.5 on NAAQS analysis.

29

Case Study: Pollutant Offset Ratio

˃ Using EPA Offset Ratios to Estimate Secondary PM2.5 and treat the same as primary PM2.5: Model using AERMOD or screening model Pollutant offset ratios of 40:1 for SO2 and 200:1

for NOx

Result = 1.49 lb/hr of secondary PM2.5

Use screening model (SCREEN3) to assess downwind impact of emission rate♦ Enter stack information, meteorology and terrain options♦ Does not account for chemical reactions in formation of

PM2.5 just gives estimate of downwind concentration.♦ PM2.5 concentration = 2.93 µg/m3 approximately 2,506 m

from source.

30

Case Study: Pollutant Offset Ratio

˃ Regional emissions of NOx and SO2 are known so used offset ratios to also find secondary PM2.5 emissions from region.

˃ Compared the regional PM2.5 emissions to proposed project’s expected emissions.

˃ Findings: Proposed project would contribute ~2.8% of total regional secondary PM2.5 emissions.

31

Case Study: Secondary PM2.5 Analysis for PSD Permit Conclusions

˃ Using both CSAPR Response Factors and Pollutant Offset Ratios showed low amounts of secondary PM2.5 formed from NOx and SO2

˃ Maximum impacts from primary and secondary PM2.5 would not occur at same time & location, unlikely secondary PM2.5 would result in violation of NAAQS

˃ Compared secondary PM2.5 value found from pollutant offset ratios to regional values obtained from MDNR Plant contributes about 3% to the total regional

value

32

Questions?

Anna Henolsonahenolson@trinityconsultants.comPhone: 253-867-5600