College of Engineering Update for

Agilent

Impacts of Emission Changes on Air Quality and Acute Health Effects in the

Southeast, 1993 – 2012

Lucas HennemanGeorgia Institute of Technology

A&WMA SS ConferenceCallaway Gardens, Georgia

20 August, 2015

Georgia Tech Student Chapter

• Genesis: Spring 2014• Activities:

– Monthly perspective dinners• Guest speaker at every meeting • We are now looking for speakers for this academic year!

– Student/young professionals socials– AWMA national conference– Student social events (e.g. Atlanta Beltline walking tour)

• Would love to collaborate with other student chapters in the region

• Thanks to the Southern Regional and Georgia Chapter for their help with logistics and funding

A&WMA Annual Conference

Raleigh, NCJune 2015

• Our chapter sent 8 members• Thank you to all our sponsors – Regional and State A&WMA Chapters– Yamaha Motor Manufacturing– Georgia Power– Shaw Industries– R2T Inc.

GT Student Chapter: beyond the first year

• The big test: a change in leadership– Survived one so far

• Needed: speakers, sponsors, events, etc.• Wanted: other student groups in the region to

collaborate with• ACE 2016: New Orleans

HEI Project Genesis

• Growing focus on assessing effectiveness of specific control programs–“Accountability”

• The Health Effects Institute has funded a number of studies in this domain

• Goal: quantify effects of individual regulatory actions on both air quality and public health

• Project is in its 3rd year

This material is based upon work supported by Health Effects Institute and the

National Science Foundation Graduate Research Fellowship under Grant No.

DGE-1148903.

6

Air Pollution Accountability

Health Effects Institute. (2003). Assessing the Health Impact of Air Quality Regulations: Concepts and

Methods for Accountability Research

• Measurements made at each link in the chain lead to improved regulations

• This project investigates relationships between each link

Counterfactual Health Outcomes

Approach

Reduction Estimates

Meteorological Detrending

Detrended Concentrations

Raw Concentrations

Counterfactual Emissions

Emissions Estimates

Counterfactual Concentrations

*S: Sensitivity

*

Atlanta, GA: Heavily impacted by mobile and power plant sources

JSTAnnual

Standard 75ppb

24-hr Standard 35µgm-3

AnnualStandard 12µgm-3

9

Detrended ambient concentrations in Atlanta, GA

• Annual averages relatively insensitive• Primary pollutants decreasing• Ozone distribution shrinking (lows increasing, highs decreasing)• See Henneman et al. (2015), Atmospheric Environment

Sulfate

PM2.5Ozone

NOx

Season Trend

Summer Winter STM Removed Observed

Pollution control policies

Vehicular sources• Inspection and Maintenance (1993 – present)• Georgia Gasoline (2000-2005)• Tier II Gasoline standards (2006 – present) • Heavy Duty Highway Rule (2007 – present)Power plants• Acid Rain Program (1995 – present)• Georgia multipollutant rule (2000 – present)• CAIR (2008 – present)• Fuel switching (all years)All sources within non-attainment area• National Ambient Air Quality Standards

11

Modeled Mobile Emissions

• Model: EPA’s Motor Vehicle Emissions Simulator (MOVES2010b)

• Since 2000, vehicle miles traveled (VMT) increased 12%

• Modeled emissions have decreased 50-70%

• Counterfactuals calculated assuming 1993 emissions rates for all post-1993 vehicles

Actual Counterfactual

12

Electricity Generating Unit (EGU) Emissions

• Measured emissions have decreased for all pollutants since 1995

• Counterfactuals calculated using 1995 emissions factors and heat input

• Reductions traced to fuel changes, controls, and others (e.g. increased efficiency)

Actual Counterfactual

Emissions Reductions by Program

Actual NOx emissions (measured CEM data in the Atlanta NAA)

Acid Rain Program, Georgia rule yy†

NOx SIP Call*, Summertime NOx controls (Georgia rule jjj†)

Clean Air Interstate Rule (CAIR), Georgia Multipollutant Control Rule (Georgia rule sss†), natural gas economics

*Georgia not included in NOx SIP Call†Georgia Rules for Air Quality Control

Counterfactual Health Outcomes

Approach

Reduction Estimates

Meteorological Detrending

Detrended Concentrations

Raw Concentrations

Counterfactual Emissions

Emissions Estimates

Counterfactual Concentrations

*S: Sensitivity

*

15

Pollutant Sensitivities

• Sensitivities: relationship between observed concentration and emission derived using linear regression

• ∆C: difference between observed and counterfactual

• ∆C added to original observed to return counterfactual

16

• Empirical sensitivities capture temporal trend from mobile and EGU emissions

• Mobile emissions dominate contribution to ozone• Care must be taken in interpretation – mobile

emissions of many species (NOx and VOCs) vary together

Mobile and EGU Sensitivities: Ozone

Total Mobile EGU

17

• EGU emissions follow seasonal trend due to contribution of SO2 emissions to SO4

2-

• Mobile emissions have little seasonal pattern (different species contribute depending on season)

• Care must be taken in interpretation – mobile emissions of many species (NOx and VOCs) vary together

Mobile and EGU Sensitivities: PM2.5

Total Mobile EGU

18

Ozone sensitivities vs Ozone concentrations

Positive response to NOx controls (ozone goes down)

Negative response to NOx controls (ozone goes up)

Range of proposed standard

Relatively insensitive

Negative response

19

Counterfactual Ozone Concentrations

• Counterfactual concentrations assume no added emissions controls since 1995 (EGU) and 1993 (mobile)

• Statistical method captures daily variability and lag structure present in observations

• Reductions in peak values• Increases in lowest values

due to decreased NOx titration

• Small effect on median ozone

Counterfactual Concentrations

20

Counterfactual PM2.5 Concentrations

• Counterfactual concentrations assume no added emissions controls since 1995 (EGU) and 1993 (mobile)

• Statistical method captures daily variability and lag structure present in observations

• Reductions both in medians and annual distribution

Counterfactual Concentrations

21

Conclusions• Meteorological Detrending

– Measured and detrended primary pollutants follow trends of utility and mobile emissions

– Detrended ozone does not match decreasing emissions trends, but shows decreasing highs and increasing lows

• Emissions– Reductions attributed to specific regulations, but care must be taken

• Sensitivities Analysis– Ozone more sensitive to mobile emissions– PM2.5 sensitivities to EGU vary by season, and sensitivities to MOB do not

• Counterfactual Concentrations– Ozone: median counterfactual remain constant, annual distribution

increases– PM2.5: Both annual distributions and medians have decreased

Extra Slides

23

EGU load (demand) categorized by controls and fuel types

NOx Control

Fuel Type

Source: EPA Air Markets Database

• Controls documented in Continuous Emissions Monitoring (CEM) data

• Challenges in estimating avoided emissions- Controls installed for

various reasons- Fuel types changes for

various reasons

None

Other

SCR*

NoneCoalOilNatural Gas

* Selective Catalytic Reduction

Periods of NOx controls

• Comparison between CEM emissions and counterfactual yields multiple periods of different control measures

• A few time periods stand out (do not follow trends linked with controls

25

Calculating Ozone Sensitivities

3source3pollutant

3source3pollutant

2source2pollutant

2source2pollutant

1source1pollutant

1source1pollutant0det3

*

*

*

SE

SE

SEßO

PS*: Atmospheric photochemical state