U.S. EPA—South Carolina Collaboration to Develop
a Multi-Pollutant, Risk-Based Air Quality
Management Strategy for the Upstate South
Carolina Region
May 26, 2016
I
-------
(This page left intentionally blank)
-------
ACKNOWLEDGEMENTS
Authors
SC Department of Health and Environmental Control:
Tommy Flynn
Maeve Mason
Andy Hollis
US Environmental Protection Agency:
Kimber Scavo
Neal Fann
Julia Gamas
Ali Kamal
Mark Morris
Ted Pal ma
We gratefully acknowledge the work of CAU/TATT and Dean Hybl, Elzbieta Covington, Paul Martin, Greg
Quina, Leslie Coolidge, Brian Barnes, Gregory Green, Martha Keating, Tyler Fox, Alison Eyth, David
Misenheimer, Marc Houyoux, Sharon Phillips, Laura Bunte, Regina Chappell, Brad Akers, Lynorae
Benjamin, Jane Spann, Karen Wesson, Rudy Kapichak, and Kelly Sheckler.
A EPA
United States
Environmental Protection
Agency
D H E C
PROMOTE PROTECT PROSPER
South Carolina Department of Health
and Environmental Control
UPSTATE
Working Together for Our
Physical & Economic Vitality
iii
-------
U.S. EPA—South Carolina collaboration to develop a multi-pollutant, risk-based air
quality management strategy for the Upstate South Carolina Region
Table of Contents
ACKNOWLEDGEMENTS iii
Authors iii
Executive Summary I
Overview I
Section I: Project Summary 3
Choose an Area of Study and Set Goals 4
Assess Current Air Quality Issues in the Upstate Region 4
Decide on the Parameters for the Emissions Inventory 5
Develop a Control Strategy 5
Process Emissions for Modeling 6
Conduct Air Quality Modeling 7
Assess Risk from Air Toxics 7
Assess Ozone and PMzs Benefits 7
Section II: Results & Recommended Next Steps 9
Results 9
BenMAP Risk Assessment 17
Overarching Conclusions 18
Key Take Away/Lessons Learned 19
Recommended Next Steps 21
Section III: Project Template 23
Multi-Pollutant Analysis Template 23
Appendix A: Original Project Description — November 2013 25
Overview 25
Demographic 26
Air Quality Issues in the Upstate Region 26
Tools and Data 28
Emissions Inventory and Baseline and Regional Modeling 28
iv
-------
Control Measure Information and Additional Data 29
Air Quality Modeling 30
Air Toxics Risk Assessment 31
Ozone and PM2.5 Benefits Assessment 31
Insights on Development of the Local Control Strategy 32
Other Planning/Policy Considerations 34
Appendix B: Background on Air Quality Management: Working toward a Multi-Pollutant
Approach 38
Appendix C: 201 I NATA Risk Reduction Analysis - South Carolina Ten at the Top
Counties 40
Background 40
CAU/TATT Air Toxic Emissions 40
CAU/TATT Estimated Cancer Risks 42
Emission Reductions and Estimated Cancer Risk Reductions 44
Appendix D: South Carolina Ten at the Top Counties Cost Analysis 46
Introduction 46
CAU/TATT Criteria Emissions Profile 41
CAU/TATT CoST Analysis Results 49
Geographic Distribution of CAU/TATT Emissions Reductions 50
Appendix E: South Carolina Ten at the Top Counties CMAQ Modeling 59
Introduction 59
Modeled PMzs Reductions 60
Ozone Reductions 64
Other Pollutants 65
Appendix F: Additional Information Regarding Health-Related Benefits 69
Introduction to Benefits Analysis Methods 69
Health Impact Assessment 71
Economic Valuation of Health Impacts 73
Uncertainty Characterization 75
Benefits Analysis Data Inputs 77
Demographic Data 77
Effect Coefficients 78
Baseline Incidence Estimates 82
v
-------
Economic Valuation Estimates 85
Growth in WTP Reflecting National Income Growth Over Time 92
References 95
vi
-------
U.S. EPA—South Carolina collaboration to develop a multi-pollutant, risk-based air
quality management strategy for the Upstate South Carolina Region
Executive Summary
This report describes a collaborative effort between the U.S. EPA, the South Carolina Department of
Health and Environmental Control, and local community and business leaders in ten upstate South
Carolina counties to develop and analyze a multi-pollutant, risk-based air quality management strategy. A
primary goal was to identify and evaluate a local control strategy targeting emissions of ozone and PM2.5
and their precursors while at the same time reducing air toxics of concern for communities to maximize
both health benefits and air quality improvements. This report provides an overview of the data and
analytical steps needed for such an analysis. The results of this project demonstrate that improving air
quality in areas already attaining the NAAQS can yield significant health benefits. This project can also
inform and help attainment areas assess actions to keep ozone and particulate matter levels below the
level of the NAAQS to ensure continued health protection for their citizens, better position such areas
to remain in attainment, and help all areas efficiently direct available resources toward a more cost-
effective strategy. Perspectives from each of the partners in this study are also provided in this report. In
general, local area perspective and expertise play a large role in successfully implementing any voluntary
emissions reduction program. Additionally, this collaborative effort between federal and state technical
staff allowed for knowledge transfer and feedback on new and innovative tools developed during the
course of this project.
Overview
The U.S. Environmental Protection Agency (EPA) and the State of South Carolina's Department of
Health and Environmental Control (DHEC) share an interest in exploring multi-pollutant analysis and
planning as a means to improve air quality effectively, and as a way to make the most efficient use of
available resources. In 2012, the Office of Air Quality Planning and Standards launched the voluntary
Ozone Advance Program. This program provides assistance to areas throughout the country that want
to create a better buffer against future violations of the ozone national ambient air quality standard
(NAAQS) by helping identify and implement pollution reductions strategies designed to help areas from
falling into nonattainment status. The Ozone Advance program was soon followed by the Particulate
Matter (PM) Advance program, which was similarly designed to help areas take steps to reduce local PM
levels. Several areas, including South Carolina joined both programs and EPA soon realized this was an
excellent opportunity to implement and further study multi-pollutant planning.
The EPA's Detroit multi-pollutant pilot project1 provided a framework for analyzing air quality
management programs capable of realizing multiple policy goals. In particular, the Detroit project
demonstrated that it is possible to achieve air quality improvements among an array of pollutants while
also reducing air pollution risk to both the general population and those most vulnerable to air
pollution-related health impacts.
1 Wesson, K., Fann, N., Morris, M, Fox, T., Hubbell, B., 2010. A multipollutant, risk-based approach to
air quality management. Case study for Detroit. Atmospheric Pollution Research, 1, 296- 304.
1
-------
A team at the EPA's Office of Air Quality Planning and Standards, EPA Region 4 in Atlanta, DHEC, and
South Carolina's Upstate Region (Upstate) which includes the nonprofit group Clean Air Upstate
Coalition/Ten at the Top (CAU/TATT) began work together in 201 3 on a multi-pollutant, risk-based air
quality management strategy building upon the information learned in the Detroit pilot project. Section I
of this document is the project summary that documents the work conducted throughout the
collaboration. Section II describes the results and recommended next steps from the analyses and
Section III includes a template which outlines the practical steps we took throughout the process and
highlights additional points to consider. There are several appendices which include details of the
analyses and collaboration.
2
-------
Section I: Project Summary
The following flow chart (Figure I) provides an overview of the process used in the development of this
multi-pollutant, risk-based air quality management strategy and serves as a guide to describe the various
steps.
Figure I:
Adjust NATA
risk results
using local
emission
reductions
Project Scope
Meeting:
determine project
area, year(s) of
study, data
acquisition needs
and requirements.
Develop
conceptual model
*
r
Develop
1
control
strategy for
area
A
Run CoST to
evaluate
control
strategy(ies)
L
Acquire
meteorological
data
Run CMAQ to
develop base
case and test
case runs
Draw conclusions,
publish report,
present findings
-------
Choose an Area of Study and Set Goals
In the summer of 201 3, EPA and DHEC had several conference calls to scope out the parameters for
working together including deciding on the area of study. DHEC, in consultation with their local
partners, recommended studying the CAU/TATT area (http://tenatthetop.org/). It is a group of 10
counties in the upstate of SC that were very proactive in taking steps to implement local measures. The
10 counties are Abbeville, Anderson, Cherokee, Greenville, Greenwood, Laurens, Oconee, Pickens,
Spartanburg, and Union counties. Over the next few months, the team worked together to develop a
project plan and description (See Appendix A) which included the following goals:
(1) identify local emission reduction measures for the Upstate that address multiple pollutants, that are
harmonized with existing or planned federal/state/local measures,2 that are quantifiable, and whose
implementation by DHEC and/or Upstate is achievable;
(2) maintain compliance with the National Ambient Air Quality Standards (NAAQS);
(3) demonstrate that the selected strategy(ies) can reduce population risk from exposure to ozone,
PM2.5, and selected air toxics in the Upstate and can reduce exposure among populations at greatest
level of baseline risk;
(4) transition to a multi-pollutant air quality management strategy; and
(5) foster a spirit of collaboration among EPA, the Upstate, and DHEC that highlights the importance of
a coalition approach.
Assess Current Air Quality Issues in the Ubstate Region
In order to understand the air quality issues in the area, DHEC presented the current air quality data to
EPA (2010-12) for both ozone and PM2.5 and EPA provided its most recent projected air quality for the
year 2020. The EPA also presented the 2005 National Air Toxics Assessment (NATA) data to assess air
toxics in the region and updated that analysis to reflect the 201 I NATA towards the end of the project.
PM2.5 oir quality levels. 2010-2012 air quality data indicated that the Upstate attained the current annual
PM2.5 NAAQS by a narrow margin. Anderson, Greenville, and Spartanburg Counties were designated as
unclassifiable for the 1997 PM2.5 NAAQS (70 FR 944, January 5, 2005). The Upstate was attaining the
2006 and 2012 PM2.5 standards for daily and annual PM2.5. 2010-2012 design values for PM2.5 monitors in
the Upstate indicated that Greenville had a 10.9 jJg/m3 design value (12 jJg/m3) for the annual standard
and a 23 jJg/m3 (35 jJg/m3) design value for the 24-hr standard.
Ozone air quality levels. The Upstate was attaining both the 1997 (0.08 ppm) and 2008 (0.075 ppm)
ozone. 2010-2012 design values indicated that Abbeville (0.064 ppm), Anderson (0.073 ppm), Greenville
0(.069 ppm), Pickens (.071 ppm), and Spartanburg (0.075 ppm) are all in a range of concern for attaining
any future more stringent NAAQS.
2 See the 2004 National Academy of Sciences (NAS) report describing the elements of a multi-pollutant
air quality management plan (AQMP).
4
-------
Projected ozone and PM2.5 air quality levels. EPA provided 2010-2012 design values for the Upstate
counties as well as projected 2020 design values based on EPA's photochemical modeling used in the
Regulatory Impact Analysis (RIA) for the PM2.5 NAAQS Final Rule (Please note that this modeling used a
2007-based modeling platform with projections from the 2007-2009 design values and not all monitors
were included in the model run. Furthermore, some areas had projected design values, but no operating
monitor due to monitor shutdowns). Based on EPA's regulatory modeling, the Upstate counties realized
reductions in their projected design values due to Federal rules that are expected to be in place in 2020
including multiple mobile source rules and the Mercury and Air Toxics (MATS) final rule. Table I in
Appendix A displays this information.
Air toxics. Based on the 201 I National Air Toxic Assessment (NATA) nearly 28,000 tons of air toxics are
emitted each year from the Upstate. According to NATA, the average cancer risk in the Upstate
associated with inhalation of air toxics is about 47 in a million. A majority of this risk is associated with
formaldhyde (62 percent) with acetaldehyde (14 percent) and benzene (8 percent) also key contributing
pollutants. Formaldhyde and acetaldehyde are generally formed with photochemical activity along the I-
85 corridor in the southeast US, while benzene emissions are associated with mobile traffic along the
many interstates in the Upstate. NATA estimates that the about 3,000 people who live in the area are
exposed to cancer risks greater than 60 in a million, with the highest risks in the urban areas of
Greenville. Please see Appendix C for additional details on the 201 I NATA results.
Decide on the Parameters for the Emissions Inventory
The EPA used the 201 I National Emissions Inventory (NEI) Version 2 for ozone and PM2.5 and the 201 I
NATA data for air toxics emissions, concentrations, and risk. Emissions sources that are known to be of
concern, especially if they are likely to be candidates for reductions, will be important to characterize
well. For these sources, such data as emissions factors and stack parameters could be further evaluated
to assure that the source is well characterized, with particular attention being paid to inventorying all
pollutants emitted. Emission summaries for all sources of concern would be valuable, including: (I)
pollutant and sector by the 10-county area, by county, and by any seasonal patterns and particular
geographic areas of interest; and (2) for particular sources/sectors, a more detailed characterization
across pollutants and what controls may be available or planned. For hazardous air pollutants (HAPs),
DHEC worked to improve their inventory for the toxic species that are leading the cancer and non-
cancer risk in the area and engaged in the state review process for the 201 I NATA in 2014-15.
Develob a Control Strategy
DHEC and EPA held several conference calls to assess the emission inventories information in order to
identify those sources affecting potential areas of interest (e.g., monitor locations; populations of
concern) within the 10-county area with a focus on those that are in need of control to reduce
emissions and associated risks. As a first step, EPA used its Control Strategy Tool (CoST -
http://www.epa.gov/ttn/ecas/cost.htm). This tool provided a good place to start, having control
effectiveness and cost information for many criteria pollutant control measures. This tool electronically
connects this control measures information directly to sources listed in the emissions inventory using
the Emissions Modeling Framework (EMF - http://www.ie.unc.edu/cempd/projects/emf/install/). EPA
applied CoST to the 201 I NEI, Version 2. To estimate reductions of HAPs, EPA used a feature of CoST
that allows it to calculate co-benefit reductions of volatile organic and metal HAPs from reductions in
5
-------
the selected VOC and PM criteria pollutants. Controls to HAPs were not applied directly. This
information gave DHEC insight on the sources they thought were important to analyze. The decision
was to focus the analysis on the non-EGU point source sector which included all area sources, and all
non-EGU point sources. EGU point sources were not considered since there are no coal-fired EGUs in
the study area, and the current natural gas fired units are well controlled. The goal of analyzing a
maximum controls strategy on the non-EGU sector was to see the magnitude of potential reductions.
This would then provide context for the local measures that CAU/TATT was considering implementing.
While two of the CAU/TATT measures were part of the CoST maximum controls run, they were
considered separate from the rest of the measures in that CoST run or "robust strategy." See Appendix
D which includes details on the CoST analysis.
In spring 2014, DHEC held meetings with stakeholders representing CAU/TATT to identify available
control options for those sources to develop a local control strategy that targets "multi-pollutant"
reductions, i.e., those that will focus on the toxics of concern for communities within the 10-county
area but maximize those ozone and PM2.5 precursor emissions reductions to gain health benefits and
further reductions in future design values for ozone and PM2.5- This process consisted of a general
introduction to the project, followed by a brainstorming session. DHEC worked with CAU/TATT and
considered various measures to reduce congestion and unnecessary idling (e.g., Right Turn on Red,
Roundabouts, Light Synchronization, Anti-Idling Programs). The strategies identified in this session were
then prioritized based on several criteria, including: availability of data used to quantify the results of
each strategy and the perceived or realized support for each initiative in the CAU/TATT area. Three
particular control strategies of interest to South Carolina emerged.
DHEC and the EPA held several discussions on potential local CAU/TATT measures. While reductions
from these measures may not be at a scale that would make a substantial difference in a regional
inventory, local measures are still important in the overall scheme of multi-pollutant risk-based planning.
See Appendix B for more details. Three local CAU/TATT strategies were analyzed to assess their
potential effectiveness: new gas stoves and gas logs, open-burning curtailment and anti-idling. Two of the
local measures were included in the CoST maximum control (new gas stoves and gas logs and open
burning curtailment) and were part of the resulting control strategy CoST run, but were not considered
part of the robust strategy. DHEC analyzed the anti-idling measure separately.
It is important to consider seasonality when choosing control measures. In general, areas in
nonattainment of an ozone NAAQS may only require control measures to be operated during certain
times of the year when ozone is highest. This will affect the design of a control measure, how it will
normally operate during a typical year, capital and operation and maintenance costs. Another example is
PM control of heating devices such as wood stoves. Most PM problems from wood smoke occur in the
winter as wood stoves are used much more during that season than the rest of the year. In most cases,
the cost of controls will be less if allowed to idle, but there will be costs to reactivate these
controls. For controls that are inherently seasonal such as wood stove controls, most of the costs
related to seasonality are built in to the CoST control measures.
Process Emissions for Modeling
A "base case" 201 I emissions inventory was provided by the EPA to DHEC. A "test case" emissions
inventory which took into account the reductions identified with CoST was also provided. The Sparse
6
-------
Matrix Operator Kernel Emissions (SMOKE) was used to process and merge these emissions
inventories for each source category (e.g., onroad mobile, nonpoint, point, etc.) into the gridded, hourly,
speciation emissions needed for an air quality model. Note that most onroad mobile sector control
strategies would require the Motor Vehicle Emissions Simulator (MOVES) to be run to develop emission
factors, which are combined with activity data by SMOKE to provide air quality model-ready emissions.
Conduct Air Quality Modeling
The modeled predictions of air quality changes are data that can be used to gauge the success of a given
control strategy. These assessments are essential for predicting the effects on local and regional air
quality, attainment of NAAQS standards, and risk and exposure. EPA recommended the CMAQ
photochemical model (Community Multiscale Air Quality Model - www.cmaq-model.org/) at a
horizontal scale of 12x12 km for predicting changes in ozone and PM2.5 concentrations.
EPA and DHEC decided that a "brute-force" model run comparing the base case emissions inventory
and a test case emissions inventory would be most appropriate for this project. The 201 I control
strategy test case included the aforementioned robust strategy and the three local CAU/TATT
strategies (new gas stoves and gas logs and open burning curtailment included in the CoST run; anti-
idling was analyzed by DHEC separately). EPA provided base case model-ready emissions files at a grid
resolution of 12km, test case point and area source emission files, and boundary condition files. This
provided the needed source emissions and characteristics for a refined air quality photochemical
modeling "brute-force" exercise. DHEC conducted the CMAQ modeling for this project, and the results
are presented below in Section 3. Given the lack of financial resources for this project, we were unable
to run CMAQ at a finer grid resolution or model a future year, though this could be useful. In addition,
ideally, we would have included air toxics in the CMAQ model run. We also would have used CMAQ
combined with HAP dispersion modeling results from the AERMOD
(http://www.epa.gov/ttn/scram/dispersion prefrec.htm#aermod) dispersion model to provide modeled
estimates at the census tract for air toxics. See results below in Section II and Appendix E for more
details.
Assess Risk from Air Toxics
To predict the effect of the proposed emissions reductions on air toxic risks, we started with the 201 I
NATA county level risks for each of the CAU/TATT counties. We assumed that a reduction in
emissions would result in a similar reduction in risk for a given pollutant. Because the inventory used for
NATA (201 I NEI) and that developed for the CAU/TATT reduction effort are not the same, we could
not directly apply the tonnage of reductions to the NATA analysis. Instead, we applied the percentage
reductions from the CAU/TATT inventory to NATA point and nonpoint risk results on a pollutant by
pollutant basis. There were no estimated emissions reductions from other source types, so there were
no estimated risk reductions from those. This approach assumes that reductions are equal across all
NATA point and nonpoint source categories. Nevertheless, we feel this approach will provide an
approximate estimate of potential reductions in risk associated with the proposed emissions reductions.
See results below in Section II and Appendix C for more details.
Assess Ozone and PMtj. Benefits
South Carolina DHEC and EPA staff worked collaboratively to estimate the human health benefits of
improving ozone and PM2.5 air quality projected to result from implementing the robust strategy as well
7
-------
as the three local CAU/TATT strategies (new gas stoves, anti-idling and open burning curtailment. We
applied the environmental Benefits Mapping and Analysis Program—Community Edition (BenMAP-CE),
http://www.epa.gov/benmap, to assess the number and economic value of the avoided PM2.5 and ozone-
related health impacts. Calculating the health impacts required four key sources of data described in
Table I below.
Table I: Key Data Inputs for BenMAP-CE Used to Estimate Avoided Health Impacts
Data Input
Source
Air quality changes
Population counts
Risk coefficients
DHEC modeled PM2.5 and ozone changes
U.S. Census data projected to the year 201 I
Concentration-response relationships from U.S. air
pollution epidemiological studies
Baseline rates of death and disease
Centers for Disease Control and Prevention-provided
death rates and Healthcare Cost and Utilization
Program provided hospital visit rates for all other areas
After surveying the epidemiological literature, EPA determined that there were no epidemiological
studies well matched to the 10-county area, and so the project team applied the default studies the
Agency uses for its national-scale benefits assessments to quantify the changes in PM and ozone-related
premature deaths and illnesses (USEPA, 2009; USEPA, 201 3). See results below in Section II and
Appendix F for more details.
8
-------
Section II: Results & Recommended Next Steps
Results
A suite of area and point source control measure were identified using the CoST. The control case
included the following reductions:
• For select non-point (area) sources: The major reduction control technologies were: Low
NOx Burners (1997 AQMD, and RACT to 25 TPY); conversion to low NOx burners in water
and space heaters; open burning curtailment program; conversion of wood fireplaces and stoves
to gas; and reformulation (Ozone Transport Commission or OTC Rule, Phase II, and process
modification). Other controls include application of Control Technology Guidelines (CTGs); low
pressure/vacuum or LPV relief valve use; and solvent utilization changes.
• For select point sources: The major reduction technologies were: low emission combustion;
dry injection/fabric filter system utilization; wet scrubber installation; and permanent total
enclosure installation. Other controls included low NOx burner conversion; Selective Catalytic
Reduction (SCR); add-on controls, work practices and material reformulation/substitution; and
solvent recovery system installation.
It is important to note that CoST applies controls that are mostly "end of pipe." Opportunities for
emissions reductions from emerging renewable energy, energy efficiency measures, and fuel switching
(for example, to natural gas), and additional local measures have not been considered. Furthermore,
mobile source controls were not applied in CoST. A separate strategy to address anti-idling was
supplied by DHEC. Also, strategies for EGU reductions were not actively sought in CoST since there
are no coal fired EGUs in the region and current natural gas units are well controlled.
Criteria pollutant reductions were quantified using the 201 I Modeling Platform in CMAQ at a horizontal
scale of 12 x 12 km on a domain 100 x 100 cell grid centered around the Upstate. A "brute-force"
emission reduction evaluation method was used, comparing base case and test case model runs.
Boundary conditions for the test domain were derived from the EPA's 201 I 12 km NATA model runs.
CoST results indicated that the following reductions in each pollutant would be expected.
NOx - 1587 tons
PM2.5 - 222 tons
SO2 - 766.32 tons
VOCs - 2727 tons
The total cost of controls was estimated at $20,000,000.
PM Reductions
The following results show the modeled PM2.5 reductions that took place between base case and test
case strategies at the PM2.5 monitors in the Upstate. PM2.5 reductions are at around a 2 percent (%)
reduction at the monitors for the annual standard (Table 2). Temporal reductions are much higher than
average in colder months (quarters I and 4)(Table 3). Speciated reductions show higher reductions in
organic carbon (Table 4). These results, taken together indicate that the wood stove conversion to
natural gas reduction strategy may be effective at reducing annual PM2.5 emissions.
9
-------
Table 2: PM2.5 Annual Standard Reductions at Upstate Monitors
Monitor ID
Base DV
Future DV
% Reduction
450450009
10.6
10.39
1.9
450450015
10.9
10.64
2.4
Table 3: Quarter I and Quarter 4 Reductions in PM2.5 Concentrations
Monitor ID
Date
Base DV
Future DV
% Reduction
450450015
Q4
10.44
9.944
4.8
450450015
Q1
10.15
9.701
4.4
450450009
Q4
9.929
9.5
4.3
450450009
Q1
9.551
9.199
3.7
Table 4: Speciated PM2.5 Relative Reduction Factors
Crustal Elemental NH4
Carbon
Organic
Carbon
S04
N03 Water
Salt
0.999 0.9851 0.9972
0.9615
0.9982
0.9764 0.9985
0.9955
0.9991 0.9843 0.9971
0.9558
0.9982
0.9753 0.9986
0.9944
Ozone Reductions
The following results show the modeled ozone reductions between base case and test case that took
place at ozone monitors in the Upstate area (Table5). While maximum daily ozone reductions can be as
high as approximately 2 ppb. Design value reductions are typically less than I ppb at the monitors (less
than a I percent (%) reduction).
10
-------
Table 5: Modeled Design Value Ozone Reductions at Upstate Monitors
Monitor ID Monitor Name Base DV Future DV % Reduction
450010001 Due West 62 61.7 0.48
450070005 Big Creek 70 69.8 0.29
450210002 Cowpens 67.3 67.2 0.15
450450016 Hillcrest 68 67.3 1.03
450451003 Famoda Farms 65.3 65.2 0.15
450730001 Long Creek 64.5 64.4 0.16
450770002 Clemson 69.7 69.5 0.29
450770003 Wolf Creek 69 68.8 0.29
450830009 North Spartanburg 73.7 73.3 0.54
Estimated Benefits of the Air Quality Management Plan
The control strategy listed above was paired with health incidence and mortality data for the
CAU/TATT area for 2010 - 2012. Changes in air quality were modeled against health impact functions
available in the BenMAP program and were tabulated to estimate the number of avoided deaths and
avoided number of non-mortality related end-points. BenMAP was then used to estimate the valuation
of these avoided health endpoints for the entire modeling domain and then the CAU/TATT area. Tables
6-7 shows the number of avoided deaths and avoided non-mortality endpoints for the CAU/TATT
area. Two studies are typically used to estimate the range of monetary savings from these avoided
health endpoints. Table 8 presents a statistical estimate of the monetary benefits of these avoided health
endpoints.
11
-------
Table 6: Estimated Incidence of Avoided PM2.5 and Ozone-Related Premature Deaths and
Illnesses
Impacts as summed across regions
Author
10 TOP
SC
Interstate
Premature Mortality (30-99)
10
1 1
16
Krewski et al. (2009)
(7.5-13)
(7.8-14)
(12-20)
Premature Mortality (25-99)
23
24
37
Lepeule et al. (2012)
(13-33)
(14-35)
(21-52)
Premature Mortality (
-------
Asthma Exacerbation (6-18)
271
290
440
Ostro et al. (2001), Mar et al. (2004)
(34-540)
(36-570)
(55-880)
Lost Work Days (18-64)
1200
1300
2000
Ostro (1987)
(1 100-
(1 100-
(1700-
1400)
1400)
2200)
Minor Restricted Activity Days
7200
7600
12000
Ostro and Rothschild (1989)
(6100-
(6500-
(9900-
8400)
8800)
14000)
-------
Table 7: Estimated Dollar Values of Avoided PM2.5 and Ozone-Related Premature
Deaths and Illnesses (millions of 201 1$, discounted at 3 percent)
Impacts as summed across region
Health Effect
Pollutant
Ten at the
Top
South
Carolina
Interstate
Domain
Premature Mortality
(Krewski et al. 2009 & Bell et al. 2004 (Ozone))
PM2.5 &
O3
$100
(9.4-280)
$1 10
(9.9-300)
$160
(15-450)
Premature Mortality
(Lepeule et al. 2012 & Levy et al. 2005 (Ozone))
PM2.5 &
O3
$230
(20-650)
$240
(21-680)
$370
(32-1,000)
Non-fatal heart attacks
(Peters et al. vl)
PM2.5
$1.3
(0.23-3.3)
$1.4
(0.24-3.5)
$2.1
(0.37-5.3)
Hospital admissions - respiratory
(Zanobetti et al Kloog et al. & Katsouyanni et al.
(Ozone))
PM2.5 &
O3
$0,057
(-0.024-
0.1 1)
$0,062
(-0.025-
0.12)
$0.1
(-0.040-
0.20)
Hospital admissions - cardiovascular
$0,073
$0,076
$0.1 1
(Zanobetti et al Bell et al. Peng et al.)
PM2.5
(0.0041-
0.017)
(0.035-
0.15)
(0.051 -
0.22)
Emergency room visits for asthma
$0.0027
$0.0029
$0.0047
(Glad et al. Mar et al. Slaughter et al. & Sarnat et
al. Peel et al.
PM2.5
(-0.00043-
0.0061)
(-0.00044-
0.0068)
(-0.00062-
0.01 1)
Glad et al. Wilson et al. Mar and Koenig Ito et
al. (Ozone))
14
-------
Acute bronchitis $0,007 $0.0073 $0,011
(Dockery et al.) PMls (-0.00034- (-0.00036- (-0.00055-
0.019) 0.020) 0.031)
Lower respiratory symptoms $0.0039 $0.0041 $0.0063
(Schwartz and Neas) PMls (0.0013- (0.0013- (0.0021-
0.0078) 0.0083) 0.013)
Upper respiratory symptoms $0.0088 $0.0092 $0,014
(Pope et al.) PMls (0.0019- (0.0020- (0.0031 -
0.021) 0.023) 0.035)
Asthma exacerbation $0,028 $0,034 $0,069
(Ostro et al. Mar et al.) PMls (0.0099- (-0.014- (-0.036-
0.081) 0.097) 0.20)
Lost work days $0.19 $0.18 $0.32
(Ostro) PMls (0.16- (0.16- (0.28-
0.21) 0.21) 0.37)
Minor restricted-activity days PM2.5 & $0.53 $0.57 $0.93
O3
(Ostro and Rothschild) (0.28- (0.29- (0.47-
0.81) 0.88) 01.5)
15
-------
Table 8: Predicted Air Quality Change, Estimated Economic Value of Avoided Deaths and
Illnesses and Net Benefits
PM2.5 Ozone
Total Air Quality Benefits (|jg nr3) (ppb)
Maximum Air Quality Change at Zip Code 0.388 0.202
(PM: 29601; Ozone: 29374)
Maximum Air Quality Change at a Monitor 0.260 0.400
(PM: 450450015; Ozone: 450830009)
Maximum Air Quality Change at a County 0.251 0.106
(PM: Greenville; Ozone: Spartanburg)
TATT Change in Population-Weighted Exposure 0.1 31 0.068
Cost-Benefits Analysis (in Millions of 2010 Dollars)
TATT Cost (in 2010 Dollars) $ 150 $290
(PM: per ug nr3 reduced; Ozone: per ppb reduced)
Total Benefits of Avoided Mortality and Morbidity $99-220 $3.1-4.4
(PM: Krewski-Lepeule; Ozone: Bell-Levy)
Total Control Strategy Cost $20
Net Total Benefits $82-210
Benefit-Cost Ratio 4.1 -10
The small reductions in ozone concentrations are likely due to the minor contributions of ozone
forming NOx from stationary sources in the area. A majority of NOx emissions in the Upstate area are
from mobile sources. Based on the 201 I NEI, these sources were responsible for approximately 77
percent of all NOx emission in the 10 Upstate counties (Figure 2)
16
-------
Figure 2:
Upstate NOx Emissions by Source Category
mobile 64% 29630 tp
\ \ P area 8% 3595 tpy
JVrail5%2218tpy
point 10% 4730 tpy
- nonroad mobile 13% 5950 tpy
This data highlights the importance of working with local governments to encourage transportation
planning which reduces mobile source emissions to meet our state's air quality goals.
BenMAP Risk Assessment
Using upstate_specific epidemiological data, partners used the Environmental Benefits Mapping and
Analysis Program - Community Edition (BenMap-CE) to examine population risk exposure between
base case and test case scenarios. The expected risk reduction for each county in the CAU/TATT is
characterized below (Table 9).
Air Toxics Reduction's Effect on Cancer Rates.
Cancer risk reductions from the air pollution control strategies were quantified using the EPA's 201 I
NATA data3. These reductions are as follows:
3 http://www.epa.gov/national-air-toxics-assessment
17
-------
Table 9:
County
Cancer Risk (in a million)
2011 NATA Risk
Expected Risk
Reduction
Abbeville
44
0.01
Anderson
46
2
Cherokee
45
1
Greenville
48
3
Greenwood
47
0.01
Laurens
45
0.02
Oconee
42
0.01
Pickens
43
0.03
Spartanburg
48
3
Union
47
0.01
Overarching Conclusions
Criteria Pollutants: Implementing local control strategies in an area with a mix of sources is an important
component of successful air quality management programs. Control strategies that are implemented
based on analyzing air quality and health data together result in a reduction of risk to certain populations
in the area. As a result, there are cost savings and air quality improvement to attain and maintain clean
air quality and health benefits.
Toxics: The highest air toxics reductions are about 3 percent in both Greenville and Spartanburg
counties. It's important to note that this analysis does not include potential reductions from "anti-idling"
efforts. The NATA model results for mobile sources do not segregate out the idling emissions, thus we
cannot estimate that portion of the risk that would be reduced. However, we would expect further risk
reductions from this program. Because a majority of the above risks are from secondary formed
pollutants (mainly formaldehye), reduction efforts to reduce precursors such as nitrogen oxides and
other criteria pollutants will have a co-benefit in reducing risks from air toxics.
Establishing partnerships: Bringing together local, state and federal partners to address important multi-
pollutant air pollution challenges results in improvements to air quality and public health. It is important
to include local organizations into this type of project because they are able to provide perspective on
what will and will not work in their communities to reduce emissions in an effective manner.
18
-------
Key Take Away/Lessons Learned
Throughout this process the partners for this project have learned a great deal. Below are some key
points and lessons learned from each of the primary partner's perspectives:
DHEC
• Environmental justice (EJ) concerns are and will continue to be a focus in the future. Not only
has the federal government and the EPA demonstrated this via Executive Orders, policy
initiatives, and major rulemaking; the states too are experiencing this in increased involvement
with the permitting process. DHEC has always prided itself on striving to work with and
encourage EJ community leaders to have a seat at the table and will continue to do this into the
future. In addition, SC has a rich history in very active EJ community groups and leaders.
Together DHEC and these EJ communities and leaders have worked together on several
projects that have resulted in successfully mitigating environmental harms to at-risk communities
across the state. The results of this project need to be shared with local EJ community leaders
so that DHEC can continue its work engaging and educating local EJ communities and leaders to
reduce environmental and health risks.
• This project has been instrumental in developing DHEC's understanding of EPA's BenMAP-CE.
Understanding risk communication is becoming increasingly important in developing
relationships with local community representatives. Enhancing the Department's awareness and
understanding or tools such as BenMAP will enhance the Department's ability to respond to
questions and concerns. In addition, this experience will also aid in the Department's own
understanding of NAAQS development at the national level and how tools like BenMAP are
used to make policy decisions.
• Local, voluntary efforts at developing collaborative approaches to problem solving are an
effective and necessary step in transition to multi-pollutant air quality management. As
evidenced by Early Action Compacts, EPA Advance, and local air quality coalitions, time and
time again DHEC has experienced the benefit of involving all interested parties in problem
solving. While these efforts are often time and resource intensive, the benefits are great.
Experience has taught us the value that local community perspective and expertise can play in
helping to make decisions that serve to promote and protect the health of SC citizens and the
environment.
EPA
Addressing air pollution challenges through a collaborative partnership is crucial for any successful air
quality management approach. This project demonstrated how a multi-disciplinary team comprised of
local, state and federal air quality modelers, health scientists, policy analysts and others could develop a
cost-effective approach to managing air quality. The EPA found it very rewarding to establish and
maintain relationships between parties who shared a common goal. Through this project we were able
to establish a framework that will help focus future similar multi-pollutant risk-based analyses (see
Section III). The framework and this report provide a structure for conducting local risk-based analyses
with the goal of identifying specific actions that can reduce emissions, help areas improve air quality and
19
-------
continue to meet the NAAQS while focusing on cost-effective actions that would provide additional
public health protection for communities.
The results of this project demonstrate that improving air quality in areas already attaining the NAAQS
will yield significant health benefits - and thus should encourage many areas to reduce emissions, either
on their own or in the context of the Ozone and PM Advance4 programs, which promote emission
reductions in attainment areas. The results of this project will inform and help other attainment areas
(I) assess actions to keep ozone and particulate matter levels below the level of the NAAQS to ensure
continued health protection for their citizens, (2) better position areas to remain in attainment, and (3)
help areas efficiently direct available resources toward a more cost-effective strategy. In addition,
reductions in attainment areas may also help reduce the transport of air pollution to downwind
nonattainment areas. Nonattainment areas should also benefit from reviewing these results, which
demonstrate that a risk-based approach to addressing air pollution and population exposure in a given
local area is important to ensuring the public health protection of its citizens.
Key lessons learned include: staff involvement at all levels and throughout the various agencies and
stakeholder groups made continued progress possible; it is important to gather local health and
population data to effectively analyze local control strategies; and localized emissions reductions are still
important even in areas that are in attainment. While the air quality benefits of local strategies may not
show an air quality benefit when a broader-scale attainment demonstration for a SIP is conducted, this
risk-based analysis demonstrates that local regulatory and voluntary control measures are an important
part of an air quality management program. By using a multi-pollutant approach, some strategies with
multi-pollutant benefits might not have been considered if pollutants had been examined one at a time; a
strategy that might seem too expensive to reduce a single pollutant or too difficult to implement might
emerge as cost-effective once all the cobenefits are factored into the analysis.
In addition, EPA benefitted from this working relationship beyond providing analytical support and
gaining insight from the results of the analysis. The EPA staff also benefited from feedback from South
Carolina on the CoST database and how to reflect local conditions in the tool with more accuracy and
from the opportunity to deliver BenMAP-CE training to South Carolina staff, which prepared them for
subsequent domestic and international training classes. Through the collaborative working relationship,
EPA was able to train a new staff member in how to run the tool with location-specific health data.
Working together in this way helped EPA to better understand the strengths and limitations to the
BenMAP-CE program, and improved the ability of staff to provide similar assistance to other analysts
throughout the U.S.
4 The EPA continues to encourage state and local air agencies to join the Ozone and PM Advance
program (http://www3.epa.gov/ozoneadvance/basic.htmh. This project started within the context of that
program. The Advance program promotes local actions to reduce ozone and particulate matter and
encourages states, tribes and local governments to take proactive steps to keep their air clean. EPA is
hopeful that the information gathered during this process will help facilitate additional local emissions
reductions and health benefits in the CAU/TATT Region of South Carolina, and serves as an example of
localities nationwide.
20
-------
While there are opportunities for continued analyses in the Upstate area that would provide more
insight into the most cost-effective multi-pollutant control strategy, the results of this analysis
demonstrate that multi-pollutant control programs can save money and time, and achieve significant
health, environmental and economic benefits while reducing costs and burdens on sources of air
pollution. An integrated air quality control strategy that reduces multiple pollutants can help ensure that
reductions will be efficiently achieved while producing the greatest overall air quality and public health
benefits.
CAU/TATT
CAU/TATT supports a multi-pollutant perspective. CAU/TATT is interested at looking at how actions
impact the total sphere of pollutants, instead of looking at them individually and in ways to easily
communicate this information to the general public.
CAU/TATT also supports the development of tools and resources for local coalitions aimed at allowing
them to easily understand the effect that certain activities will have on air quality and various pollutant
levels. This analysis is helpful in determining whether programs (like Breath Better at Schools) and their
financial costs are supported by the likely outcomes. CAU/TATT also supports the need for better data
tracking tools to determine whether programs have measurable impacts.
Finally, CAU/TATT is interested in learning more about the connection between air quality and health,
especially asthma. As pollution levels continue to decrease, they believe that is becomes increasingly
important to understand how/why certain health outcomes fail to show improvement. CAU/TATT feels
that learning what potential contributing factors can be understood and mitigated is an important next
step.
Recommended Next Stebs
In addition to evaluating and summarizing these results, the DHEC and EPA will seek future
opportunities to share these results with other interested parties. These opportunities might include
seeking publication of these results in academic journals, poster and oral presentations at local and
national conferences, etc. In addition to focusing on the publication of these results the following have
been identified potential next steps to include options related to future modeling efforts and the
development of tools to enhance similar type pilot projects.
Modeling:
• A potential 4 km CMAQ or CAMx model run to better assess community level PM2.5 and ozone
reductions. This resolution may also be appropriate for showing the benefits of an anti-idling
program.
Future year baseline and projected model runs to assess the impact of several planned and on-the-way
rules and regulations (utility MACT, NESHAPs, NSPS, 2015 Ozone NAAQS, etc.). This will provide
directional information on the effect of regional and national controls on the Upstate area. In addition,
as deemed necessary as part of additional air quality modeling to be conducted to assess the impact of
the local control strategy, any local enhancements to the emissions inventory to create the future-year
base inventory will need to be determined by DHEC. This could mean collecting data on any closings or
shut-downs of industrial sources, mobile fleet turnover, and installation of potential control measures
and determining which to include in the future base year projection.
21
-------
• A more targeted control strategy to include options for potential implementation (e.g., local
measures for the mobile source sector) to determine effect.
• Periodic rerun of the modeling and BenMAP to see how the control strategy responds to
future improvements in emissions inventory and modeling.
• To estimate air toxics exposure concentrations for a given risk assessment, we could use the
CMAQ/AERMOD hybrid approach that uses the grid cell modeled concentrations from CMAQ
with the finer-scale gradient provided by AERMOD, which is the methodology used in the
Detroit project. The hybrid ambient pollutant concentrations would be used as surrogates for
the inhalation exposure concentrations of the populations in the study locations. The default
assumption in this approach is that the population of interest is breathing outdoor air
continuously, which overestimates exposure because people are not always at the study
location due to their daily activities, and because indoor air concentrations are expected to be
the same or lower than the outdoor concentrations (when the indoor concentrations are
produced solely by inflow from outside air). The HEM-3 model
(http://www.epa.gov/ttn/fera/fera download.html). which was used in the Detroit project, is
based on this approach.
Planning:
• Both DHEC and CAU/TATT are eager to share the results of this pilot with other local air
quality coalitions. Sharing these results is important on many levels, including, but not limited to
increasing understanding and participation at the local level, while supporting/facilitating
collaboration with state and local community and environmental justice leaders as well as
business and educational institutions to come together to develop tailored approached to
reducing pollution and protecting health. As a result of this study, DHEC will be conducting
CAMx source apportionment modeling to quantify pollution contributions from mobile sources
in different regions of the state and using the results to highlight the importance of mobile
source emissions reduction programs. Providing this information to local governments and air
quality coalitions will hopefully lead to transportation planning and policy decisions which
reduce mobile source emissions and improve air quality.
Tools & Expansion:
• In addition, DHEC is interested in working with EPA and academia to develop tools that might
better analyze the effect of mobile source controls, like anti-idling campaigns and
environmentally focused transportation planning.
• Processes to communicate the findings of a multi-pollutant control strategy to the public.
• Finally, DHEC has also considered whether this pilot project could/should be replicated in
another area of the state - perhaps the Central Midlands area around Columbia or in the
coastal region near Charleston.
• Partner with academia to encourage future epidemiological studies that represent either or
both smaller urban and/or regional rural research areas to produce results that can be used to
replace national scale benefit assessments.
22
-------
Section 111: Project Template
Multi-Pollutant Analysis Template
The following list provides the steps and key questions to explore in completing a project similar to the
USEPA-SC project. The timing specified is an amount of time we suggest if your project plan has a 15-
month timeframe. Each project's timeframe will vary depending on the goal, staff and resources available.
Step I : The initiating agency should determine the project's scope and objectives and convene
appropriate partners (federal, state, local, industry, communities, NGOs) to get agreement on goals and
the project scope. What geographical area do you want to study? Nonattainment area status,
jurisdictional boundaries in your state and/or local government, population density, mix of sources and
current air quality issues are important factors to consider. This includes considering air toxics,
greenhouse gases, transportation, energy and land-use planning and environmental justice
considerations. What are the year(s) of study and the team's data acquisition needs and requirements? It
is also helpful to develop a conceptual model (e.g., workplan) that includes a description of the area of
study and the problems and issues to address.
Timing: 1-2 months
Step 2: Acquire meteorological, emissions and NATA data
Timing: I month
Step 3: Develop control strategy(ies) to analyze and compare (consider all pollutants of interest,
geographic area, and potential non-end-of-pipe measures that could be applied that are not in CoST). An
end-of-pipe measure is typically a control that is applied to a unit or process to reduce its output of
emissions.
Timing: 2-3 months
Step 4: Run CoST and calculate the cost per ton of any measures of interest for which CoST may not
have information, such as local measures and non-end-of-pipe measures (energy efficiency, renewable
energy and fuel switching). Evaluate the cost-effectiveness of the strategy(ies). Steps 3 and 4 can be
iterative.
Timing: I month
Step 5 : Process emissions for modeling and run CMAQ to develop base case and test case (and future
year) runs & review results. For a more robust air toxics analysis, also run AERMOD and perform a
CMAQ/AERMOD hybrid analysis. The CMAQ/AERMOD hybrid air quality model combines the results
of a chemical transport model (CMAQ) and a Gaussian dispersion model (AERMOD) to take advantage
of the chemistry and long-range transport provided by CMAQ and the local-scale gradient provided by
AERMOD. If this approach is taken for air toxics, step 6 is not needed.
23
-------
-determine if the modeling results meet the objectives; if not, additional modeling and/or other analyses
may be warranted
Timing: 1-2 months
Step 6: Adjust NATA risk results using local emission reductions
Timing: Concurrently with CMAQ run
Step 6: Acquire health data
Timing: 2-4 months to occur simultaneously with steps 2-5
Step 7: Run BenMAP-CE. Incorporate results of CMAQ test case and future year and health data in
BenMAP
If multiple strategies were analyzed, compare the air quality and BenMAP-CE results with the cost of
controls to determine the most cost-effective control strategy. Take into consideration cost and
benefits of the strategy options. In particular, consider both the magnitude and distribution of benefits,
assessing the extent to which air quality benefits occur among susceptible and vulnerable subgroups.
Not all benefits are quantifiable (e.g., environmental justice and ecosystems services). Such unquantifiable
benefits should be factored into the strategy selection if applicable.
Timing: 2 months
Step 8: Review results, draw conclusions, and write a report
Timing: 2 months
Step 9: Implement the selected strategy.
Timing: As appropriate for each state and local agency's adoption process
24
-------
Appendix A: Original Project Description -- November 2013
Overview
The U.S. Environmental Protection Agency (EPA) and the State of South Carolina's Department of
Health and Environmental Control (DHEC) share an interest in exploring multi-pollutant analysis and
planning as a means to improve air quality effectively, and as a way to make most efficient use of
available resources.
EPA's Detroit multi-pollutant pilot project provides a framework for analyzing air quality management
programs capable of realizing multiple policy goals. In particular, the project demonstrates that it is
possible to achieve air quality improvements among an array of pollutants while also reducing air
pollution risk to both the general population and those most prone to air pollution-related health
impacts. Two key factors contributed to the success of the project: (I) careful planning that involved
profiling the Detroit metropolitan area, rigorously formulating the overall "air quality problem" to be
addressed, and identifying the data and tools to be utilized; and (2) EPA and the Michigan Department of
Environmental Quality (DEQ) collaborated closely, ensuring that all parties understood the project goals
and were willing to share data. In May 201 3, EPA approached the DHEC with the opportunity to work
together to develop a multi-pollutant analysis modeled after the Detroit pilot.
EPA, DHEC, and South Carolina's Upstate Region (Upstate) which includes the nonprofit group Clean
Air Upstate Coalition (CAU)/Ten at the Top (TATT) are each interested in collaborating to develop and
use a multi-pollutant, risk-based analysis for the region which builds upon the lessons learned in Detroit
while addressing air quality issues unique to the Upstate.
EPA's and DHEC's goals for this project are:
(1) identify local emission reduction measures for the Upstate that address multiple pollutants, that are
harmonized with existing or planned federal/state/local measures,5 that are quantifiable, and whose
implementation by DHEC and/or Upstate is achievable;
(2) maintain compliance with the National Ambient Air Quality Standards (NAAQS);
(3) demonstrate that the selected strategy(ies) can reduce population risk from exposure to ozone,
PM2.5, and selected air toxics in the Upstate and can reduce exposure among populations at greatest
level of baseline risk;
(4) transition to a multi-pollutant air quality management strategy; and
(5) foster a spirit of collaboration among EPA, the Upstate, and DHEC that highlights the importance of
a coalition approach.
EPA, DHEC, and the Upstate will work together to develop a multi-pollutant risk-based analysis. The
conclusions will inform choices that DHEC and the Upstate make as to which specific strategies may be
5 See the 2004 National Academy of Sciences (NAS) report describing the elements of a multi-pollutant
air quality management plan (AQMP).
25
-------
implemented and documented as a supplement to the combined ozone/fine particulate matter (PM2.5)
Advance plan ("path forward") for South Carolina.6
Below we summarize the questions to be answered by this analysis; the tools and data available; the
project schedule; and the roles and responsibilities of EPA, DHEC, and the Upstate. Because we expect
to refine each component of the analytical plan iteratively, this document will evolve over time.
Demographic
South Carolina's Upstate is comprised of Abbeville, Anderson, Cherokee, Greenville, Greenwood,
Laurens, Oconee, Pickens, Spartanburg, and Union counties. Clean Air Upstate Coalition (CAU)/Ten at
the Top (TATT) is a nonprofit group founded in 2005 which is focused on fostering a spirit of
cooperation and collaboration among Upstate public, private, and nonprofit leaders to build a
comprehensive picture of what is important to people in the Upstate as they look at the future.
CAU/TATT has facilitated and coordinated the Upstate Air Quality Advisory Committee and the Clean
Air Upstate initiative in an effort to reduce emissions and stay within federal air quality standards with
representatives from stakeholder groups across the Upstate (including Upstate Forever, Piedmont
Natural Gas, WSPA-TV, BMW Manufacturing, Duke Energy, City of Greenville, and Michelin NA).
Air Quality Issues in the Upstate Region
The Upstate faces a confluence of air quality management challenges.
1. Current PMzs air quality levels. Recent air quality data indicate that the Upstate attains the current
annual PM2.5 NAAQS by a narrow margin. Anderson, Greenville, and Spartanburg Counties were
designated as unclassifiable for the 1997 PM2.5 NAAQS (70 FR 944, January 5, 2005) and that
designation remains in effect. The Upstate is attaining the 2006 and 2012 PM2.5 standards for
daily and annual PM2.5. 2010-2012 design values for PM2.5 monitors in the Upstate indicate that
Greenville has a 10.9 jJg/m3 design value (12 jJg/m3) for the annual standard and a 23 jJg/m3 (35
|jg/m3) design value for the 24-hr standard.
2. Current ozone air quality levels. With the help of Early Action Compacts7 and local stakeholder
involvement, the Upstate is attaining both the 1997 (0.08 ppm) and 2008 (0.075 ppm) ozone
standards but is likely to exceed a more stringent NAAQS (0.070 ppm or less). 2010-2012
design values indicate that Abbeville (.064), Anderson (.073), Greenville (.069), Pickens (.071),
and Spartanburg (.075) are all in a range of concern for attaining any future more stringent
NAAQS.
3. Projected ozone and PM2.5 air quality levels. Table I provides the 2010-2012 design values for the
Upstate counties as well as projected 2020 design values based on EPA's photochemical
modeling used in the Regulatory Impact Analysis (RIA) for the PM2.5 NAAQS Final Rule (Please
note that this modeling used a 2007-based modeling platform with projections from the 2007-
2009 DVs and not all monitors operating today were included in the model run. Furthermore,
some areas have projected Design Values, but no currently operating monitor due to monitor
6 www.epa.gov/ozonepmadvance
7 For more information, see: http://www.scdhec.gov/environment/baq/EarlyActionPlan/index.asp.
26
-------
shutdowns.). Based on EPA's regulatory modeling the Upstate counties realize reductions in
their projected design values due to Federal rules that are expected to be in place from now to
2020 including multiple mobile source rules and the Mercury and Air Toxics (MATS) final rule.
Table I. Current and Projected Design Values for Ozone and PM2.5: SC Upstate
2010-2012 Design Value 2020 Projected Design Value
County Ozone Annual PM2.5 Daily PM2.5 Ozone Annual PM2.5 Daily PM2.5
(PPb) (Mg/m3) (Mg/m3) (ppb) (Mg/m3) (Mg/m3)
Abbeville
64
60.5
Anderson
73
NVa
Cherokee
70
55.4
Greenville
69
10.9
23
9.86
21.7
Greenwood
8.88
18.4
Laurens
Oconee
64
NV
NV
6.40
13.4
Pickens
71
Spartanburg
75
10.7
21
63.1
8.36
17.5
Union
59.5
4. Air toxics. Based on the 2005 National Air Toxic Assessment (NATA) nearly 16,000 tons of air
toxics are emitted each year from the Upstate. According to NATA, the average cancer risk in
the Upstate associated with inhalation of air toxics is about 46 in a million. A majority of this
risk is associated with formaldhyde (56 percent) with benzene (12 percent) and acetaldehyde
(10 percent) also key contributing pollutants. Formaldhyde and acetaldehyde are generally
formed with photochemical activity along the 1-85 corridor in the southeast US, while benzene
emissions are associated with mobile traffic along the many interstates in the Upstate. NATA
estimates that the nearly 21,000 people who live in the area are exposed to cancer risks greater
than 60 in a million, with the highest risks in the urban areas of Greenville and Spartanburg.
Please see Attachment I for additional details on the 2005 NATA results.
5. Environmental Justice. Previous local-scale analyses suggest that there are pronounced gradients
to intra-city air quality—particularly for PM2.5- Moreover, the public health literature indicates
that certain population groups may be more susceptible to air pollution impacts. This portion of
8 No value
27
-------
the analysis will consider how air pollution levels correspond to such population subgroups, and
whether policies can specifically target such populations. For example, children, the elderly, and
people with respiratory or heart diseases are at higher risk of being vulnerable to the effects of
air pollution. Communities that are low-income and/or minority can bear a disproportionate
burden of environmental harm and risks. These overburdened communities can have high
unemployment, low income, and limited access to healthcare and are often located in urban
areas that may have environmental pollution from multiple active or abandoned industrial
facilities. In 2008, EPA conducted an evaluation of communities in SC using social demographics,
environmental, compliance, and health data to identify areas with disproportionately high and
adverse environmental and public health burdens. The evaluation identified several areas in the
Upstate that were overburdened. One specific area was in Greenville, SC. DHEC identified and
contacted stakeholders to engage in collaborative problem-solving to address air-related issues.
The stakeholder group developed and implemented an education and outreach plan for this
area. These efforts not only helped to reduce air emissions, but also established stronger
working relationships with the stakeholders in that area.
6. Key multi-pollutant sources. EPA and DHEC will use the 2005 NATA information and the CoST
tool to provide information regarding which sources may be particularly important contributors
to the emissions in the area and offer potential co-control opportunities from a multi-pollutant
perspective. The NATA and CoST information will assist DHEC in identifying the specific
sources (e.g., mobile, inland port, prescribed burns) that will be added to this section and that
can become the focus of DHEC and the Upstate's efforts to design a local control strategy.
Given the nature of the air quality problem in the Upstate, the main objectives of a "multi-pollutant,
risk-based" control strategy for the local area might be:
• Attainment and maintenance of the recently revised PM2.5 NAAQS and current and potentially
more stringent ozone NAAQS;
• Lowering emissions, ambient levels, and exposures to key air toxics of concern; and
• Maximizing health benefits among those populations at greatest risk of air pollution-related
health impacts.
Tools and Data
This section describes some of the technical tools and data that might be needed to develop a multi-
pollutant, risk-based control strategy. The information offered below is mostly based on what was done
for the Detroit project and the lessons learned from it.
Emissions Inventory and Baseline and Regional Modeling
# 00
Having an emissions inventory that can support a multi-pollutant, risk-based analysis is important. For
this project, we will use the 201 I National Emissions Inventory (NEI) for ozone and PM2.5 and the 201 I
NATA data for air toxics emissions, concentrations, and risk as it becomes available. Below is a list of
additional items to consider when determining where to make adjustments to the current emissions
inventory.
28
-------
• Source characteristics: Emissions sources that are known to be of concern, especially if they are
likely to be candidates for reductions, will be important to characterize well. For these sources,
such data as emissions factors and stack parameters could be further evaluated to assure that
the source is well characterized, with particular attention being paid to inventorying all
pollutants emitted. Emission summaries for all sources of concern would be valuable, including:
(I) pollutant and sector by the 10-county area, by county, and by any seasonal patterns and
particular geographic areas of interest; and (2) for particular sources/sectors, a more detailed
characterization across pollutants and what controls may be available or planned. (Workplan
Item 3)
• Hazardous air pollutants (HAPs): Sometimes there are gaps in emissions inventories with respect
to HAPs since it can be difficult to inventory these data from many sources. For example, in
many cases emissions are reported as total tons of VOC and are not speciated into the
component pollutants. Trying to inventory all 187 HAPs for all sources may not be possible.
Instead, it may be a better use of resources to focus on improving the inventory for the toxic
species that are leading the cancer and non-cancer risk in the area. Monitoring data and
information from any special studies performed in the area may be useful to identify the
pollutants that are of the greatest concern and could be the focus of emissions inventory
improvements. (Workplan Item 3) South Carolina should engage in the current state review
process for the 201 I NATA inventory since EPA plans to use the 201 I NATA platform to run
modeling analyses.
• Baseline modeling: For this study, EPA will provide the 201 I NATA inventories and air quality
and risk analyses for an appropriate multi-pollutant baseline. This will provide the needed source
emissions and characteristics, refined air quality concentrations (census track receptors for
toxics and 12km CMAQ run for ozone and PM2.5), and the air toxic risk assessment that will
allow source attribution of risk drivers for the 10-county area. (Workplan Item 4)
• Future year baseline modeling and emissions projection: EPA will also provide projected emissions
and air quality for 2020 based on the PM2.5 NAAQS RIA for ozone and PM2.5. This will provide
directional information on the effect of regional and national controls on the Upstate area. Table
I, above, also provides directional impact of future controls. In addition, as deemed necessary as
part of additional air quality modeling to be conducted to assess the impact of the local control
strategy, any local enhancements to the emissions inventory to create the future-year base
inventory will need to be determined by DHEC. This could mean collecting data on any closings
or shut-downs of industrial sources, mobile fleet turnover, and installation of potential control
measures and determining which to include in the future base year projection. (Workplan Item
4)
Control Measure Information and Additional Data
In order to maximize potential co-control opportunities for direct and precursor emissions for PM2.5,
ozone, and air toxics, DHEC will identify those sources affecting potential areas of interest (e.g., monitor
locations; populations of concern) within the 10-county area with a focus on those that are in need of
control to reduce emissions and associated risks. As part of this effort, with assistance from EPA, DHEC
will identify available control options for those sources to develop a local control strategy that targets
29
-------
"multi-pollutant" reductions, i.e., those that will focus on the toxics of concern for communities within
the 10-county area but maximize those ozone and PM2.5 precursor emissions reductions to gain health
benefits and further reductions in future design values for ozone and PM2.5.
• To assist DHEC in development of their local control strategy, EPA makes available (not
including any state specific databases) its Control Strategy Tool (COST -
http://www.epa.gov/ttn/ecas/cost.htmV This tool provides a good place to start, having multi-
pollutant information on many sources and allowing the user to electronically connect directly
to the emissions inventory using the Emissions Modeling Framework (EMF -
http://www.ie.unc.edu/cempd/projects/emf/install/). EPA will offer the DHEC CoST-related
support. Most likely it will still be important to add additional information and "multi-
pollutanize" some of the control measures in this database for the sources of most concern,
e.g., it may be necessary to enhance the database with reduction efficiencies for air toxics for
those controls that gain ozone and PM2.5 precursor emissions reductions (VOCs and metal
HAPs). (Workplan Item 2)
• Health and population data: DHEC will gather the appropriate necessary refined health data for
input to BenMAP-CE9 in order for the benefits analysis to be reflective of the demographics and
susceptibility of the underlying population in the 10-country area. This will help determine the
location of vulnerable and susceptible populations and correlations with higher concentrations
of pollutants of concern and quantify the health benefits of emissions reductions. EPA will assist
DHEC with running BenMAP-CE.
• See Health Impact Assessment below. (Workplan Item 3)
Air Quality Modeling
The modeled predictions of air quality changes are data that can be used to gauge the successfulness of
the control strategy. These data are essential for predicting the effects on local and regional air quality,
attainment of NAAQS standards, and risk and exposure. As in the Detroit project and with leveraging
EPA's 201 I NATA effort, EPA expects to apply the CMAQ photochemical model (Community
Multiscale Air Quality Model - www.cmaq-model.orgA at a horizontal scale of 12x12 km for predicting
ozone and PM2.5 and combine the HAP results from the AERMOD
(http://www.epa.gov/ttn/scram/dispersion prefrec.htm#aermod) dispersion model to provide horizontal
resolution at the census tract. Further discussion will determine the best choice of additional air quality
modeling for this work.
Once control options for specific sources are defined and a local control strategy for the area is
designed, then EPA and DHEC will work together to decide on and conduct an appropriate local air
quality assessment to inform the toxics risk assessment (i.e., inform Human Exposure Model (HEM-3))
and the ozone and PM2.5 health assessment (i.e., inform BenMAP). There are several options for the
assessment that EPA will conduct with input from DHEC:
I. Conduct specific air quality modeling of the local control strategy for a small modeling
domain that includes the 10-county area. This would be illustrative so "projections" would be a scaled
9 BenMap-CE is the Environmental Benefits Mapping and Analysis Program-Community Edition.
30
-------
version of the 201 I NATA inventory to serve as a "future base" and then assess the local control
strategy as the "future control" scenario.
2. Adjust the 201 I NATA inventory based on appropriate adjustment factors to translate
emissions changes to air quality concentration changes for the 10-county area (and beyond, if expect
downwind PM2.5 or ozone benefits outside the area).
3. Conduct a qualitative assessment if the emission reductions are not deemed significant or
consider a more localized assessment of specific communities if the control scenarios are largely focused
on single facilities (e.g., separately conduct dispersion modeling of an individual source).
If additional "fine-scale" modeling is deemed necessary, EPA and DHEC will have additional discussions
about source locations, spatial and temporal scales and future year projections at the "fine-scale." We
define "fine-scale" modeling to be modeling of a photochemical model with a horizontal grid resolution
of 4x4 km or smaller and/or application of a dispersion model. (Work Plan Item 5)
The modeled concentrations of air quality can be used to evaluate the impact of the control strategy on
the future year design values (DVs) and on the air quality in the urban area, as well as in the region.
Visualizing the results in programs like ArcGIS is extremely helpful to analyze the geographical impact of
the strategy. Using GIS, one can also overlay the population to better understand the population-
weighted air quality changes, as well as analyze areas of remaining high concentrations with respect to
emission sources.
It will also be important to calculate the change in the predicted, future year design values based on the
local control strategy. To do this more efficiently, EPA has created the Model Attainment Tool Software
(MATS) (http://www.epa.gov/ttn/scram/modelingapps mats.htmV MATS takes the inputs of modeled and
monitored data to predict the 8-hr ozone, the 24-hr PM2.5, and annual PM2.5 DVs. MATS can also be
used to create the spatial fields of ozone and PM2.5 air quality to input into BenMAP for the health
impact assessment.
Air Toxics Risk Assessment
To estimate exposure concentrations for a given risk assessment for this project, we would use ambient
pollutant concentrations as surrogates of the inhalation exposure concentrations for the populations in
the study locations. The default assumption in this approach is that the population of interest is
breathing outdoor air continuously, which is conservative because people are not always at the study
location due to their daily activities, and because indoor air concentrations are expected to be the same
or lower than the outdoor concentrations (when the indoor concentrations are produced solely by
inflow from outside air). The HEM-3 model (http://www.epa.gov/ttn/ferayfera download.html). which
was used in the Detroit project, is based on this approach.
Further discussion between EPA, DHEC, and the Upstate would be beneficial for this assessment.
Ozone and PM^ Benefits Assessment
As in the Detroit project, we would recommend relying upon the environmental Benefits Mapping and
Analysis Program (BenMAP) to assess the avoided PM2.5 and ozone-related health impacts and associated
monetized benefits of alternate policy scenarios. The calculation of health impacts requires four key
31
-------
sources of data: air quality changes, population estimates, risk coefficients, and the baseline incidence
rate for each health endpoint quantified. As described in Hubbell et al. (2009), performing a city-level
health impact analysis calls for local-scale input data to reduce the risk of biasing the analysis with health
data that does not characterize the health status of populations within the Upstate. Spatially resolved
incidence rates and effect coefficients will also be useful to any EJ analysis performed for the 10-county
area, as they will enable us to identify vulnerable and susceptible populations and estimate health impacts
among these at-risk populations. However, we were unable to obtain age-stratified population data
needed to run this analysis correctly so we resolved to use the BenMAP county-level incidence rates.
Air quality changes for ozone and PM2.5 based on the local control strategy emissions reductions are
provided outside of the BenMAP program (standard file format from CMAQ is established and can be
generated for input). However, while EPA can generate population projections for the study area, the
identification of effect coefficients and incidence rates will require more effort. See Table 3: Health and
Socioeconomic Data Inputs.
The EPA will survey the epidemiological literature to determine whether existing or new PM or ozone
studies have been conducted for South Carolina or the Upstate. It will also be critical to use ZIP or
tract-level baseline incidence rates where available. Below we have detailed the health endpoints and age
ranges for which we need these data. In general, we need rates stratified by patient sex, ZIP or census
tract FIPS, year, age category, and principal diagnosis.10 If available, race-stratified rates would be useful
to performing a more reliable EJ or distributional analysis.
If appropriate, distributional and EJ impacts could be quantified using a combination of BenMAP and
ArcGIS. EPA has developed approaches for using baseline health data in conjunction with air quality
levels to identify populations at greatest risk of air pollution impacts. The identification of "at-risk"
populations might be a useful first step to developing the air quality management strategies.
Insights on Development of the Local Control Strategy
While the sections above discuss what data is needed to implement and analyze the results of a multi-
pollutant, risk-based control strategy, this section tries to discuss some of the important things to
consider when designing the control strategy. In general, based on the Detroit project results and
"lessons learned," we recommend trying to incorporate these basic goals:
• Aim to achieve population oriented emission reductions, particularly for susceptible and
vulnerable populations;
• Consider selecting control measures that reduce multiple pollutants whenever possible;
• Focus on reducing the toxic pollutants that are driving the cancer and non-cancer incident rates;
• When making decisions based on costs, try to consider the resultant $ per ppb or jJg/m3 of air
quality improvement or $ per health benefit a control measure will potentially supply rather
than simply looking at $ per tons of emissions reduced;
10 If possible, age stratified by the following bins would be most useful: 0-1, 2-6, 7-14, 15-17, 18-24, 25-
34, 35-44, 45-54, 55-64, 65-74, 75-84, 85-99.
32
-------
• Where possible, combine base air quality with health information to better inform decisions,
especially as they relate to EJ issues.
We realize that policy considerations and costs constrictions will also need to be considered when
determining what control measures to include. When possible, we recommend using the basic guidelines
listed above to develop a first draft of the strategy and then analyzing impacts on air quality, DVs, toxic
risk, and health benefits to refine the strategy to better fit the goals of the work. (Work Plan Item 2).
Information regarding the confluence of air toxic and criteria pollutant risk and the location of emitting
sources may prove helpful to constructing an emissions control strategy that satisfies the criteria above
(Figures 1, 2 and 3).
Figure I. Air toxic lifetime cancer risk and criteria pollutant annual mortality risk at each
county
2005 NATA-predicted lifetime cancer risk 2005 PM1S and ozone mortality risk
Lifetl me cancer risk due to air toxics (chances in a million) Percentage of deaths due to PM^5 and Oj
33
-------
Figure 2. Lifetime cancer risk from air toxics formed in the atmosphere and
annual ozone mortality at each county
2005 NATA-predicted lifetime cancer risk from air 2005 ozone mortality risk
toxics formed in the atmosphere
Other Planning/Policy Considerations
• Energy Planning (Energy Efficiency/Renewable Energy)
• Environmental Justice assessnnent?/Transportation & Land-use planning and Climate Change?
(Work Plan Item 2)
34
-------
Roles and responsibilities:
The success of each stage of the project will hinge upon the close collaboration between EPA, DHEC,
and the Upstate. Table 2 includes a workplan and summarizes roles and responsibilities for the project
with a draft schedule.
Work Plan Key Items:
Table 2. Work Plan (EPA and DHEC agree to meet to discuss status on the last Monday of each
month at I pm.)
Schedule
Coordination
Resources
(1) Problem
Formulation & Work
Plan
August/
September
DHEC with EPA
providing technical
guidance
EPA/DHEC
(2)
Brainstorm/Develop
Control information
and control strategies
August/
September/
October/
November
DHEC and
CAU/TATT with
EPA providing
technical assistance
Covered by state's current
SIP funds. EPA could help
with running CoST.
(3) Data Acquisition
for Emissions
Inventories and Health
Data
December
DHEC and
CAU/TATT
Covered by state's current
SIP funds
(4) Emission Inventory
(base and future
projection) using
2011 data and
additional CoST run
with control strategy
December-
May
DHEC & EPA
Expect to leverage existing
EPA inventories and
projections. SC may make
local improvements as
appropriate.
(5) AQ modeling and
post processing
December-
May
EPA leads with
assistance from
DHEC
EPA would conduct in
consultation with DHEC.
Details and options noted
above.
(6) Risk and benefits
assessment
December-
May
DHEC and EPA
jointly perform
assessment
DHEC would conduct in
consultation with EPA and
with CDC involvement
(7) Compilation of
results of risk-based
analysis, and potential
selection of measures
to be implemented
and added to SC
Advance path forward
May-June
DHEC leads with
EPA providing
technical guidance
Covered by state's current
SIP funds
35
-------
(8) Information for
June
DHEC leads with
Covered by state's current
public outreach
EPA providing
SIP funds
guidance
Table 3. Health and Socioeconomic Data Inputs
Unless indicated below, each data set representing total counts (or prevalence) will be segmented
accordingly:
• By gender
• By age group: 0-5, 6-11, 12-19, 20-39, 40-59, 60+
• ZIP level data for three most recent years (to be averaged across years)
DISEASE
ICD 9 CODE
DATA SET
Acute myocardial infarction
410-414
ER and hospital admissions
Conduction disorders
426
ER and hospital admissions
Cardiac dysrhythmias
427
ER and hospital admissions
Congestive heart failure
428
ER and hospital admissions
All cardiovascular
390-429
ER and hospital admissions
All cardiovascular (less
myocardial infarctions)
390-409, 41 1-429
ER and hospital admissions
Heart disease complications
429
ER and hospital admissions
Cerebrovascular disease
430-438
ER and hospital admissions
Hemorrhagic stroke
430-432
ER and hospital admissions
Stroke
430-434
ER and hospital admissions
Ischemic stroke
433-434
ER and hospital admissions
Peripheral vascular disease
440-448
ER and hospital admissions
Respiratory disease
460-519
466, 480-486, 490-
493
ER and hospital admissions
Respiratory illness
464-466, 480-487,
490-492
ER and hospital admissions
36
-------
464-466, 480-487
Chronic lung disease or
COPD
490-496
490-492, 494-496
490-492, 494,
ER and hospital admissions
Chronic lung disease (less
asthma)
490-492
491,492, 494, 496
ER and hospital admissions
Pneumonia
480-486
ER and hospital admissions
Asthma
493
ER and hospital admissions
Lower respiratory infection
466.1, 466.0, 480-
487, 490, 510-51 1
ER and hospital admissions
Other Effects (if available)
Acute bronchitis
466
Prevalence only
Chronic bronchitis
491
Prevalence only
Cough variant asthma
493.82
Prevalence only
Asthma: any exacerbation or
attack
493-493.9
Prevalence only
Days of work lost due to any
cause
Asthma and/or
bronchitis
Incidence data
Socioeconomic data
Poverty data
• Fraction of individuals and households (by race) below the
poverty line (ZIP or tract)
• Fraction of individuals and households (by race) at below 1.5 x
the poverty line (ZIP or tract)
Education
• Fraction of individuals (by race) with less than a grade 12
education (ZIP or tract)
• Fraction of individuals (by race) with a grade 12 education
(ZIP or tract)
• Fraction of individuals (by race) with greater than a grade 12
education (ZIP or tract)
37
-------
Appendix B: Background on Air Quality Management; Working
toward a Multi-Pollutant Approach
In 1955, the first federal air pollution control law was promulgated primarily to fund research into the
scope and sources of air pollution. Since that time, air quality management has evolved in many ways to
include the first Federal Clean Air Act (CAA or Act) in 1963. However, it wasn't until 1970 that the
previous iterations were amended creating what some consider to be the first modern day CAA. The
1970 amendments increased authority of the newly created Environmental Protection Agency (EPA) and
established the basic structure of our nation's present air quality management program. This Act
authorized the establishment of National Ambient Air Quality Standards (NAAQS), the New Source
Performance Standards (NSPS) for new and modified stationary sources, the establishment of National
Emission Standards for Hazardous Air Pollutants (NESHAPs), increased enforcement authority, and
authorized requirements for the control of motor vehicle emissions. The 1970 CAA also established
requirements for State Implementation Plans (SIPs) to achieve the NAAQS and address air quality
concerns.
In June 1989, then President Bush proposed significant revisions to the CAA. The resulting 1990
amendments were enacted in large part to deal with urban air pollution or NAAQS. The NAAQS are
air quality standards set by the EPA for six "criteria pollutants" which are among the most harmful to
public health and the environment. With the 1990 amendments, EPA is required to set NAAQS for each
of the criteria pollutants and review these standards once every five years to determine if they are
appropriate or if new standards are needed to protect public health." Since these last major
amendments, technology and science have continued to evolve such that many now recognize the
importance of a more holistic approach to air quality management. A number of task forces, work
groups, and studies have looked at the current air quality management system and made
recommendations for improvements. In 2004, the National Research Council issued a report that
described some of the challenges that will be faced in future air quality management efforts. They
recommended a multipollutant approach to reducing emissions for both criteria and hazardous air
pollutants.12
It is with these thoughts in mind that the EPA and states have sought opportunities to work together to
seek out opportunities for collaboration; to take an integrated multi-pollutant approach to managing air
quality that is based on risk assessment and simultaneous review of multiple interrelated pollutants. In
response to the need to further explore and understand the technical needs and challenges presented
by implementing a multi-pollutant, risk-based approach to air quality management, the EPA's Office of
Air Quality Standards and Planning (OAQPS) embarked on a case study in 2010 centered on the urban
area of Detroit, Michigan. As part of this case study, two contrasting air quality control strategies were
assessed and compared; known as the 'status quo' and a "multi- pollutant, risk-based" approach aimed
at further reducing population risk from exposure to ozone, PM2.5 and selected air toxics.
In 201 3, the EPA again sought to partner with another state to further explore a multi-pollutant
approach to air quality management in the Southeastern United States. Working with staff of the South
Carolina Department of Health and Environmental Control as well as local community leaders in the
" 42 U.S.C. §7409 (2011).
12 John Bachmann (2007) Will the Circle Be Unbroken: A History of the U.S. National Ambient Air
Quality Standards, Journal of the Air & Waste Management Association, 57:6, 652-697
(http://dx.doi.org/10.3155/1047-3289.57.6.652)
38
-------
South Carolina upstate, a pilot project was developed to assess the effects of examining a control
reduction strategy aimed at addressing multiple pollutants and air toxics to improve air quality and
health. Perspectives from each of the partners in this study are provided in the report. In general local
area perspective and expertise play a big role in successfully implementing any voluntary emissions
reduction program. Additionally, a collaborative effort between federal and State technical staff allowed
for knowledge transfer and feedback on new and innovative tools developed during the course of this
project.
39
-------
Appendix C; 2011 NATA Risk Reduction Analysis - South
Carolina Ten at the Top Counties
Background
The South Carolina Clean Air Upstate Coalition (CAU)/Ten at the Top (TATT) is a group often
counties in the northwest corner of South Carolina and consists of the following counties: Abbeville,
Anderson, Cherokee, Greenville, Greenwood, Laurens, Oconee, Pickens, Spartanburg, and Union. The
following air toxic analysis is based on county level air toxic risk and emission estimates from the 201 I
National Air Toxic Assessment (NATA) as well as emissions reduction data provided by South Carolina as
part of this program.
The 201 I NATA is a risk analysis based on an emissions inventory, the 201 I National Emissions
Inventory (NEI) of major, area, and mobile source emissions for the calendar year 201 I. The analysis
assumes that these emissions occur for 70 years and does not take into account yearly variability in
emissions. The reader is referred to the NATA Technical support document
(http://www.epa.gov/national-air-toxics-assessment/201 I -nata-technical-support-document) and the
NATA website (www, e pa. gov/N AT A) for a more complete compendium of the approach, as well as the
uncertainties and limitations of the NATA analysis.
CAUITATT Air Toxic Emissions
Based on the 201 I NEI, nearly 28,000 tons of air toxics are emitted each year from the South Carolina
CAU/TATT counties. In comparison, statewide emissions of air toxics in South Carolina are estimated
by the 201 I NEI to be nearly 160,000 tons/year. Methanol emissions comprise nearly half the
CAU/TATT emissions, with formaldehyde, toluene, and acetaldehyde emissions comprising over a
quarter of the emissions in the CAU/TATT counties. Figure I depicts the air toxic emissions by
pollutant for the SC CAU/TATT counties. One-third of these emissions are from onroad mobile
sources, and almost another third is from smaller area sources. Major point sources are responsible for
only 4 percent of the emissions in the SC CAU/TATT counties. Figure 2 depicts the air toxic emissions
for each source sector for the SC CAU/TATT counties.
40
-------
Figure I. 201 I NEI - Pollutant contributions to SC CAU/TATT county air toxic emissions
(28,000 TPY).
Figure 2. 2005 NATA - Source Sector contributions to SC CAU/TATT county air toxic
emissions (28,000 TPY).
41
-------
CAU/TATT Estimated Cancer Risks
Based on the 201 I NATA, the average excess cancer risk in the SC CAU/TATT counties associated
with inhalation of air toxics is about 47 in one million. This risk is slightly above the statewide cancer
risk estimate of 44 in one million and the national average cancer risk predicted by NATA of 40 in one
million. A majority of the risks in the CAU/TATT counties is associated with formaldhyde (62 percent),
with acetaldehyde (14 percent) and benzene (8 percent) also key contributing pollutants. While not
directly emitted in large quantities (see above), formaldhyde and acetaldehyde are mostly formed
through photochemical activity. In the Southeast US, such activity is most prevalent along the 1-85
corridor. Figure 3 depicts the primary pollutants contributing to the average cancer risks in the
CAU/TATT counties.
NATA estimates that a majority of the risks in the CAU/TATT counties is from secondarily formed
pollutants and from background sources or transported emissions into the CAU/TATT region. Figure 4
depicts the primary source sectors contributing to the average cancer risks in the CAU/TATT counties.
While NATA estimates that the entire population of the SC CAU/TATT counties is exposed to cancer
risks greater than 30 in one million, about 3,000 residents are exposed to cancer risks greater than 60 in
one million at the census tract level, with the highest risk of 66 in one million in the urban core areas of
Greenville. Figure 5 depicts the cancer risks for all the census tracts in South Carolina.
Figure 3. 201 I NATA - Pollutant contributions to SC CAU/TATT county cancer risks (47
in one million).
42
-------
Figure 4. 201 I NATA -Source Sector contributions to SC CAU/TATT county cancer risks
(47 in one million).
Figure 5. 201 I NATA - Census Tract Cancer Risks- South Carolina.
Total Risk
{In a million)
20-30
30-40
40 -SO
50-60
43
-------
Emission Reductions and Estimated Cancer Risk Reductions
To predict the effect of the proposed emissions reductions on air toxic risks, we started with the 201 I
NATA county level risks for each of the CAU/TATT counties. We assumed that a reduction in
emissions would result in a similar reduction in risk for a given pollutant. Because the inventory used for
NATA (201 I NEI) and that developed for the CAU/TATT reduction effort are not the same, we could
not directly apply the tonnage of reductions to the NATA analysis. Instead, we applied the percentage
reductions from the CAU/TATT inventory to NATA point and nonpoint risk results on a pollutant by
pollutant basis. There were no estimated emissions reductions from other source types, so there were
no estimated risk reductions from those.
This approach assumes that reductions are equal across all NATA point and nonpoint source categories.
Nevertheless, we feel this approach will provide an approximate estimate of potential reductions in risk
associated with the proposed emissions reductions.
The emissions reductions were focused on six air toxic pollutants from point and nonpoint source
sectors. Table I depicts the expected reductions for each pollutant for all the CAU/TATT counties.
Table I. Proposed Air Toxic Emissions Reductions.
Pre-CAU/TATT (TPY)
Post-CAU/TATT (TPY)
% Reductions
Formaldehyde
12
2
-85%
Acetaldehyde
6
1
-84%
Benzene
15
2
-86%
1,3-Butadiene
3
0.3
-87%
Chromium (VI) compounds
0.001
0.00001
-99%
Ethyl benzene
28
1 1
-60%
We estimated risk reductions from the proposed program by applying the percentage emissions
reductions to the county wide risk averages for each of the CAU/TATT counties. Table 2 depicts the
county risk averages from the 201 I NATA and the expected reductions. The highest reductions are
about 3 percent in both Greenville and Spartanburg counties. It's important to note that this analysis
does not include potential reductions from "anti-idling" efforts. The NATA model results for mobile
sources do not segregate out the idling emissions, thus we cannot estimate that portion of the risk that
would be reduced. However, we would expect further risk reductions from this program.
44
-------
Table 2. 201 I NATA- CAU/TATT Average County Total Cancer Risks and Expected Risk
Reductions.
201 1 County Cancer
NATA Risk
Expected % Risk
Reduction
Abbeville
44
0.01
Anderson
46
2
Cherokee
45
1
Greenville
48
3
Greenwood
47
0.01
Laurens
45
0.02
Oconee
42
0.01
Pickens
43
0.03
Spartanburg
48
3
Union
47
0.01
45
-------
Appendix D; South Carolina Ten at the Top Counties Cost
Analysis
Introduction
Emissions reductions opportunities were identified for South Carolina's Clean Air Upstate Coalition
(CAU)/Ten at the Top (TATT) counties using the CoST tool (http://www.epa.gov/ttnecas I /costhtm).
CoST provides information on potential control options to reduce emissions and how much they would
cost to implement. The tool is applied to emissions inventories, and controls are identified and applied
at the unit-Standard Classification Code (SCC) level for point sources and at the county-SCC level for
non-point and mobile sources. CoST can be applied to reduce a single pollutant, or, as in the case of
South Carolina's CAU/TATT to reduce multiple pollutants simultaneously. For South Carolina, CoST
was applied using the 201 I NEI and the pollutants of interest were: NOx, PM2.5, SO2 and VOCs. Two
options exist to reduce emissions using CoST, one is at minimum cost to achieve a target emissions
reduction, and the other is by applying maximum reductions. For the CAU/TATT, the maximum
emissions reductions method was chosen to see what was potentially available in terms of controls and
emissions reductions. While it is possible to use CoST to apply controls to non-EGU point, non-point,
mobile and electricity generating units (EGUs), the group decided to apply CoST to non-EGU point and
nonpoint sources, so the CoST analysis did not include EGUs, mobile sources and open burning SCCs.
An emissions strategy for school bus idling was conducted by South Carolina's Department of Health
and Environmental Control (DHEC) outside of CoST.
States can use CoST to get a sense of the types of controls they could apply and design and to compare
potential control strategies and their emissions reductions. States can also provide updated information
on controls given their knowledge of local conditions. The U.S. EPA welcomes any such updated local
information that can be incorporated into CoST. Currently the tool includes mostly end-of-pipe
controls, although there are ongoing efforts to include opportunities for reductions from renewable
energy, energy efficiency and fuel switching. Emissions control strategies from CoST should be viewed as
illustrative or hypothetical, because each State has a better idea of what controls and emissions
reductions strategies are feasible for them.
As with any tool or model uncertainties exist in results derived from CoST. Data may not always be
available to reflect the specific characteristics of a facility or source group, and thus the variation in
emissions control requirements and resulting emissions. Also, emissions inventories used with CoST are
a reflection of how accurately data was reported by the reporting entities (e.g., industries reporting to
state air agencies, states reporting to EPA). There is also uncertainty regarding which production and
control technologies will become cheaper in the future, so CoST results should not be expected to
provide a prediction of these emerging technologies.
Software to run CoST can be downloaded from the Community Modeling and Analysis System (CMAS)
website. A user manual is available. The Contact person for CoST is David Misenheimer
(misenheimer.david@epa.gov) at the Air Economics Group in the Office or Air Quality Planing and
Standards.
46
-------
CAU/TATT Criteria Emissions Profile
Data from the 201 I NEI shows that there were 8,000 tons of NOx, PM2.5, SO2 and VOCs being emitted
in CAU/TATT counties. Figure I shows the breakdown by pollutant, with 2,300 tons of NOx, 257 tons
of PM2.5, less than 1000 tons of SO2 and about 5000 tons of VOCs.
Figure I. 201 I NEI - Criteria Pollutant Emissions in CAU/TATT Counties
Point source emissions were about 3,000 tons: 1,500 tons of NOx, 60 tons of PM2.5, 1,000 tons of SO2
and about 500 tons of VOCs. These emissions came mostly from IC engines and coal fired boilers. Non-
point source emissions were about 5,000 tons: 780 tons of NOx coming mostly from residential,
commercial and institutional Natural gas use; about 200 tons of PM2.5 mostly from fireplaces, hydronic
heaters and woodstoves; and almost 4,000 tons of VOCs mostly from architectural, motor vehicle and
other coatings. Figures 2 and 3 show emissions for point and non-point sources as a percentage of total
source emissions respectively.
47
-------
Figure 2. 201 I NEI - Criteria Pollutant Point Source Emissions in CAU/TATT Counties
a Percentage of Total Point Source Emissions
PM2.5
2%
Figure 3. 20II NEI - Criteria Pollutant Non-Point Source Emissions in CAU/TATT
Counties as a Percentage of Total Non-Point Source Emissions
-------
Table I is a summary of emissions information from the CAU/TATT. It shows that the largest
contribution to NOx is from internal combustion (IC) engines and boilers, the largest contribution of
PM2.5 is from fireplaces, hydronic heaters and woodstoves, and the largest contribution of VOCs is from
architectural, motor vehicle and other coatings. This last source is also the largest contributor to non-
point sources, and to total emissions in general, whereas IC engines and boilers are a substantial
contributor to non-EGU point sources, and to emissions in general contributing almost 40 percent of
total emissions.
Table I. Summary of NEI 201 I Criteria Emissions for CAU/TATT Counties
Pollutant
Sector
NEI 201 1
Emissions
(tons)
Percent
from
Pollutant
Percent
from
Sector
Percent
from
Total
Major Sources
NOx
Non-point
782
22%
16%
10%
Residential, commercial and institutional NG use
PM2.5
Non-point
193
74%
4%
2%
Fireplaces, hydronic heaters and woodstoves
VOCs
Non-point
3,990
86%
80%
50%
Architectural, motor vehicle and other coatings
NOx
Point-non-EGU
1,535
78%
50%
19%
IC engines and boilers*
PM2.5
Point-non-EGU
63
26%
2%
1%
ICI coal powered boilers**
so2
Point-non-EGU
966
100%
32%
12%
ICI coal powered boilers**
VOCs
Point-non-EGU
479
14%
16%
6%
Paper coating
NOx
Total
2,317
100%
29%
PM2.5
Total
256
100%
3%
so2
Total
966
100%
12%
VOCs
Total
4,470
100%
56%
Total
Non-point
4,965
100%
62%
Total
Point-non-EGU
3,044
100%
38%
Total
Total
8,009
100%
Mote: *IC engines are internal combustion engines. ** IC
Industrial/Commercial/lnstitutional coal powered boilers.
CAU/TATT CoST Analysis Results
coal powered boilers are
The result of applying CoST to the 201 I NEI towards maximum emissions reductions of all four criteria
pollutants (NOx, PM2.5, SO2 and VOC) was about $20 million dollars (in 201 I dollars). Table 2
summarizes the results. Costs of reductions by pollutant were as follows: NOx at $2 million (10 percent
of total cost) and 1,600 tons reduced; PM2.5 at $2 million (10 percent of total cost) and 200 tons
reduced: SO2 at $3 million (14 percent of total cost) and 800 tons reduced; and VOC at $ I 3 million (66
percent of total cost) and almost 3,000 tons reduced. Non-EGU point source reductions were almost
$8 million (40 percent of the total cost). Non-point source reductions amounted to $12 million dollars
(60 percent of total cost).
49
-------
Table 2. Summary of CoST Results for CAU/TATT Counties
Pollutant
Sector
Annual Cost
Percent
Emissions
Average
Percent
Percent
Major Reduction Technologies
(2011$)
of tota 1
Reductions
Cost per
emissions
reductions
cost
(tons)
Ton
reductions
from
Pollutant
NOx
Non-point
$ 551,431
3%
347
$ 1,591
44%
22%
Low NOX burner
pm2.5
Non-point
$ 1,335,943
7%
166
$ 8,069
86%
74%
New gas stoves, Bum ban
VOCs
Non-point
$ 10,213,138
51%
2,344
$ 4,357
59%
86%
Reformulation
NOx
Point-non-EGU
$ 1,389,517
7%
1,241
$ 1,119
81%
78%
Low emission combustion
pm2.5
Point-non-EGU
$ 755,964
4%
57
$ 13,321
90%
26%
Dry injection / Fabric fileters
S02
Point-non-EGU
$ 2,857,715
14%
766
$ 3,729
79%
100%
Dry injection, fabric filters, wet scrubbers
VOCs
Point-non-EGU
$ 2,941,515
15%
383
$ 7,676
80%
14%
Permanent total enclosures
NOx
Total
$ 1,940,948
10%
1,588
$ 1,222
69%
100%
PM2.5
Total
$ 2,091,907
10%
222
$ 9,409
87%
100%
S02
Total
$ 2,857,715
14%
766
$ 3,729
79%
100%
VOCs
Total
$ 13,154,653
66%
2,728
$ 4,823
61%
100%
Total
Non-point
$ 12,100,512
60%
2,857
$ 4,236
58%
Total
Point-non-EGU
$ 7,944,711
40%
2,448
$ 3,246
80%
Total
Total
$ 20,045,223
100%
5,304
$ 3,779
66%
Note: This cost is annualized at a 3 percent discount rate.
For NOx point sources almost 70 percent of the emissions reductions came from Low Emission
Combustion at 3 percent of the total cost. About 20 percent of non-point NOx reductions came from
Low NOx burners. Point source PM2.5 reductions from dry injection and fabric filters were 26 percent
of the reductions at 4 percent of the total cost. For non-point PM2.5, new gas stoves were 38 percent of
the reductions at 3 percent of the overall cost, while open burning curtailment was 36 percent of
reductions at 4 percent of total cost. SO2 point source emissions reductions came from dry injection,
fabric filter systems, and wet scrubbers, and were 15 percent of total costs. No SO2 non-point source
controls were applied. Finally, for VOCs, permanent total enclosures were 12 percent of point source
reductions at 14 percent of the costs, and reformulation control measures were 43 percent of non-
point source reductions at 42 percent of costs.
School bus anti-idling emission reductions were also estimated using county level school bus fleet
numbers and documented emission rates for diesel buses13. The total reductions for this pollution
reduction strategy were relatively small (NOx emissions reductions totaled ~3 TPY and PM2.5
reductions totaled ~. I TPY, for instance). These reductions, though small, can be significant when
looking at nearby areas around schools where there are potentially sensitive populations. However, due
to the small scale of these reductions, they were not included in the photochemical modeling.
Geographic Distribution of CAU/TATT Emissions Reductions
Figure 4 shows NOx reductions in CAU/TATT counties. Almost 1,300 tons of NOx reductions came
from Spartanburg County, with Greenville County following at 200 tons. The remaining counties had 50
or less tons reductions.
13 EPA - OTC report, Average In-Use Emissions from Urban Buses and School Buses. October 2008.
https://www3.epa.gov/otaq/consumer/420f08026.pdf
50
-------
Figure 4. NOx Emissions Reductions in CAU/TATT Counties from Point and Non-point
Sources
Figure 5 shows that PM2.5 reductions took place in only three counties. The highest reductions of almost
I 15 tons happened in Greenville County, while Cherokee and Spartanburg Counties saw reductions
close to 60 and 50 tons respectively.
51
-------
Figure 5. PM2.5 Emissions Reductions in CAU/TATT Counties from Point and Non-point
Sources
SO2 Reductions are shown in Figure 6. Cherokee (460 tons) and Union Counties had the highest
reductions at 460 and 120 tons respectively. Anderson County had reductions of 100 tons whereas
Greenville County had 90 tons of PM2.5 reductions.
Figure 6. SO2 Emissions Reductions in CAU/TATT Counties from Point and Non-point
Sources
52
-------
Legend
S02 Reductions
I | Up lo 100 tons
HI Between 100 arid 250 tons
¦ Between 250 and 500 tons
All CAU/TATT counties saw VOC reductions in the analysis. Greenville and Spartanburg Counties had
the highest reductions at 1,200 tons and 700 tons respectively. Anderson County had almost 500 tons
reductions. The remaining counties had between less than 200 and more than 30 tons reductions. VOC
reductions are shown in Figure 7.
Figure 7. VOC Emissions Reductions in CAU/TATT Counties from Point and Non-point
Sources
53
-------
Legend
VOC Reductions
Up to 100 tons
Between 100 Mid 500 tons
Between 500 and VOOOtons
Between 1,000 and t.500 tore
Between i 500 and 2 000 ions
New gas stoves or gas logs and open burning curtailment were emissions controls considered as options
by DHEC. Figures 8 and 9 show emissions reductions from their application in CAU/TATT counties. In
both cases emissions reductions took place in Greenville and Spartanburg Counties, although reductions
were small as compared to other controls, Greenville County saw about 50 tons of PM2.5 reductions
from new stoves or gas logs and about 60 tons reductions from open burning curtailment. For
Spartanburg PM2.5 reductions were 30 and 20 tons respectively.
Figure 8. PM2.5 Emissions Reductions in CAU/TATT Counties from New Gas Stoves or Gas
Logs
54
-------
Figure 9. PM2.5 Emissions Reductions in CAU/TATT Counties from Open Burning
Curtailment
Legend
Sum Ban RM 2.5 Reductions
j Up lo 19 ions
¦ Between 19 and 62 Ions
55
-------
As explained in the previous section, South Carolina conducted a separate emissions reductions analysis
for school bus anti-idling programs. Figures 10 to 12 show NOx, PM2.5, and VOC reductions from this
measure, with all ten counties showing reductions of all three pollutants. Emissions reductions ranged
between 0.6 and 0.16 tons of NOx with Greenville, Spartanburg and Anderson Counties showing the
highest reductions of almost 0.8, .60 and almost 0.5 tons respectively.
Figure IQ. NOx Emissions Reductions in CAU/TATT Counties from School Bus Anti-Idling
PM2.5 reductions from bus anti-idling happened mostly in Greenville, Spartanburg and Andersonville
Counties, of 0.025, 0.020, and 0.015 tons reduced respectively (Figure I I). Other county reductions
ranged between 0.002 and 0.004 tons.
Figure I I. PM2.5 Emissions Reductions in CAU/TATT Counties from School Bus Anti-Idling
56
-------
Legend
Anti-idting PM 2.5 Reductions
~ Up to 0QQ5lon$
1 | Between 0 005 Ions and 0 007 ions
¦ Between 0 007 tons and 0 026 tons
VOC emissions reductions from school bus anti-idling also happened, in their majority, in Greenville
(0.090 tons), Spartanburg (0.07 tons) and Anderson (almost 0.06 tons) Counties. Other county VOC
reductions ranged between 0.008 and almost 0.03 tons. See Figure 12.
Figure 12. VOC Emissions Reductions in CALJ/TATT Counties from School Bus Anti-Idling
57
-------
Legend
Anti-idling VOC Reductions
~ Up to 0 018 tons
| Between o 018 and 0 028 lore
¦ Between 0.028 and 0 092 tons
58
-------
Appendix E: South Carolina Ten at the Top Counties CMAQ
Modeling
Introduction
The photochemical model simulations for this emission reduction strategy used the Community
Multiscale Air Quality Model (CMAQ) version 5.0.2 which is a three-dimensional air quality model
designed to simulate the formation photochemical and secondarily formed pollutants such as ozone and
PM2.5 over regional spatial scales (https://www.cmascenter.org/cmaq/). This simulation used the
201 I National Emissions Inventory (NEI) Modeling Platform Version 2 applied at a horizontal scale of 12
x 12 km on 100 x 100 cell grid centered around the "Upstate" of South Carolina to assess primary and
secondary formed criteria pollutants and was run for the entire 201 I year. A "brute-force" emission
reduction evaluation method was used to assess criteria pollutant reductions. This method compares
the difference between the base case and a test case which includes emission reduction strategies using
two model runs.
EPA provided merged and unmerged 201 I NEI CMAQ ready emissions files for the base case. For the
test case, emissions inventory files produced by the CoST tool were provided by EPA for non-EGU
point source and area source sectors. For a summary of total reductions between the base case and test
case, see Appendix D: South Carolina Ten at the Top Counties Cost Analysis. Source sectors involved
in the emissions reductions test case strategy were processed by DHEC using the Sparse Matrix
Operating Kernal Emissions (SMOKE) (https://www.cmascenter.org/smoke/) program. Meteorological
data used for the simulation was processed using Weather and Research Forecasting (WRF) model
version 3.4 and the Meteorology-Chemistry Interface Processor (MCIP) version 4.2. In order to reduce
run time and file sizes, a sub-CONUS domain was chosen (shown in Figure I below). Modeling smaller
domains can sometimes be less accurate if boundary conditions are not represented effectively, so
boundary conditions for the project domain were extracted from the EPA's 201 I 12 km NATA model
runs. The base case modeling run was evaluated for model performance by comparing outputs to
observational data and was found to be acceptable based on recent air quality policy applications.14,15
Summary information for mean bias (MB), mean gross error (MGE), normalized mean bias (NMB),
normalized mean gross error (NMGE), root mean squared error (RMSE) and the Pearson correlation
coefficient (r) for hourly ozone and quarterly PM2.5 data is included in Tables I and 2. A bubble plot for
mean bias for ozone, evaluated at monitor sites within the domain is included as Figure I.
Table I. Performance Statistics for Hourly Ozone
MB MGE NMB NMGE RMSE r
3.298263 10.05897 0.097767 0.298166 13.08218 0.731095
14 Simon, H., K.R. Baker, and S.B. Phillips, 2012. Compilation and Interpretation of Photochemical
Model Performance Statistics Published between 2006 and 2012, Atmospheric Environment,
61, 124-139.
15 EPA, Air Quality Modeling Technical Support Document for the 2008 Ozone NAAQS Cross-State Air
Pollution Rule Proposal, November 2015
59
-------
Figure I. Ozone Mean Bias at Monitors in the Test Domain
Mean Bias for Ozone Season 2011 (Observed - Modeled) (ppb)
Nashville
u Knox-iJIe .
• 8 *
Greens^rot) *
TENNESSEE . • .AshevWe
7 • ' f
Chattanooga •
NORjfa »
Charlptte CAROLINA
•
Huntsvllle ,
. | V , -J • * J. r ' _ • •
*
* Atlanta •
Birmingham • •
?0UTH
CAROLINA
Tuscaloosa
0 • ">
ALABAMA
Chane9lon
O
Montgomery 0 GEORGIA
Columbus
Savannah
* 1 Pelagic
Sargassum
Habitat
Restricted Area
Map data ©2016 Google, INEGI
u-
Table 2. Performance Statistics for Quarterly PM
2.5
MB
MGE
NMB
NMGE
RMSE r
-0.07738
0.745125
-0.00714
0.068779
0.825169 0.940994
Base case arid test case reductions of ozorie and PM2.5 were quantified using the Modeled Attainment
Test Software (MATS) (https://www3.epa.gov/scrarn00l/modeiingapps_mats.htm) which calculates the
relative reduction factors of design values at target monitoring locations.
Modeled PM2 5 Reductions
The following results show the modeled PM2.5 reductions that took place between base case and test
case strategies at the PM2.5 monitors in the Upstate. A spatial representation of reductions of mean
annual PM2.5 is included as Figure 2.
60
-------
Figure 2. Mean Annual Reductions in PM2.5 Concentrations
Mean Difference PM2.5 Annual
0.388
0.339 "
0.290 "
0.241 "
CO
£
"Bi 0.193 "
3
0.144 "
0.095 "
0.046 "
*
-0.002
PM2.5 reductions for the annual standard were calculated using MATS, These reductions are at around 2
percent (%) at the two monitors operating in the study area (Table 3).
Table 3. PM2.5 Annual Standard Reductions at CAU/TATT Monitors
Monitor ID Base DV Future DV % Reduction
450450009 10.6 10.39 1.9
450450015 10.9 10.64 2.4
The MATS software also evaluates quarterly reductions of PM2.5. Of note, temporal reductions are much
higher than average in colder months (quarters I and 4)(Table 4 and Figure 3).
61
-------
Table 4. Quarterly Reductions in PM2.5 Concentrations
Monitor ID Date Base DV Future DV % Reduction
450450015
Q1
10.15
9.701
4.4
450450015
Q2
11.04
10.96
0.7
450450015
Q3
11.98
11.96
0.2
450450015
Q4
10.44
9.944
4.8
450450009
Q1
9.551
9.199
3.7
450450009
Q2
11.12
11.04
0.7
450450009
Q3
11.84
11.83
0.1
450450009
Q4
9.929
9.5
4.3
-------
7.7506-1
¦
67816-1
1
5 $136-1
4 8446-1
| 3*756-1
39066-1
1 938C-I
|
9 6886-2
1
0 00060
MATS also provides PM2.5 speciated reductions. The relative reduction factors between test case and
base case are higher reductions in organic carbon (Table 5).
Figure 3. Mean Quarterly Reductions in PM2.5 Concentrations
Mean PM2.S Difference jan-Mar
0775
¦
0678
|
0581
0464
I
0 386
0291
0194
|
0097
1
0000
Mean PM2.5 Difference Apr-Jun
Mean PM2.5 Difference Jul-Sept
Mean PM2.5 Difference Oct-Dec
Table 5. Speciated PM2.5 Relative Reductions Factors
Crustal
Elemental Carbon
NH4
Organic Carbon
S04
N03
Water
Salt
0.999
0.9851
0.9972
0.9615
0.9982
0.9764
0.9985
0.9955
0.9991
0,9843
0.9971
0.9558
0.9982
0.9753
0.9986
0.9944
These results, taken together, may indicate that the wood stove conversion to natural gas reduction
strategy may be driving the annual PM2.5 emission reductions since wood combustion is linked to
63
-------
atmospheric organic carbon16, and wood stoves and fireplaces are used primarily in the colder months
of quarters I and 4. Programs which encourage wood stove and fireplace conversions to natural gas
could prove to be a useful strategy to pursue to reduce PM2.5 concentrations in the area.
Ozone Reductions
The following results show the modeled ozone reductions that took place between base case and test
case strategies at the ozone monitors in the Upstate. A spatial representation of reductions of maximum
8-hour daily max (MDA8) ozone concentrations, during ozone season (April I through October 31), is
included as Figure 4.
Figure 4. Maximum MDA8 Ozone Reductions
Difference in Max MDA Ozone Values
1.955 '
1
1.441 '
1.184 '
>
1 0.927 "
0.670 '
0.413 '
I
0.156 '
1
-0.101
The largest domain wide reduction in ozone was approximately 2 ppb, but most of the area saw less
than a I ppb reduction in ozone. This is the maximum difference; the form of the standard is 4th high
maximum value.
Ozone reductions using the MATS software use the 4th high maximum value consistent with the
standard. The following results show the modeled ozone reductions between base case and test case
that took place at ozone monitors in the Upstate area (Table 6).
16 EPA, Particulate Matter (PM2.5) Speciation Guidance Document, July 22, 1998.
http://www3.epa.gov/ttnamti i/files/ambient/pm25/spec/specpin2.pdf
64
-------
Table 6. 4th High MDA8 Reductions at CAU/TATT Monitors
MonitorJD
Monitor_Name
Base_DV
Future_DV
% Reduction
450010001
Due West
62
61.7
0.48
450070005
Big Creek
70
69.8
0.29
450210002
Cowpens
67.3
67.2
0.15
450450016
Hillcrest
68
67.3
1.03
450451003
Famoda Farms
65.3
65.2
0.15
450730001
Long Creek
64.5
64.4
0.16
450770002
Clemson
69.7
69.5
0.29
450770003
Wolf Creek
69
68.8
0.29
450830009
North Spartanburg
73.7
73.3
0.54
While maximum daily ozone reductions can be as high as approximately 2 ppb, design value reductions
are typically less than I ppb at the monitors (less than a I percent (%) reduction). As such, the reduction
strategies tested in this exercise do not seem to be very effective at reducing ozone concentrations, and
other reductions strategies should be identified in the case of demonstration attainment.
Other Pollutants
Volatile Organic Compounds
The following two figures (Figures 5 and 6) show the maximum and mean percent decreases in VOC
between the base case and the test case, respectively.
65
-------
Figure 5. Maximum Percent Decrease in VOCs
Maximum Percent Decrease in VOCs
24.539 "
¦
21.484 "
1
18.430
15.375
c
aj
S 12.321 -
a_
9.266 •
1
6.212 "
|
3.158 "
1
0.103
Figure 6. Mean Percent Decrease in VOCs
Mean Percent Decrease in VOCs
6.932E0 -
¦
6.066E0 "
1
5.199E0 "
4.333E0 "
£=
QJ
ai 3.466E0 "
Q_
2.600E0 •
I
1.733E0 "
|
8.669E-1 "
1
3.952E-4
66
-------
Anthropogenic ozone production in SC is primarily driven by NOx emissions17, so these reductions in
VOCs have little effect on ozone reductions. However, the reduction of VOCs translates to a reduction
in air toxics, which is important from an public health standpoint
NO2 and SO2
The following two figures (Figures 7 and 8) show the maximum decreases in NO2 and SO2 between the
base case and the test case, respectively.
Figure 7. Maximum Hourly NO2 Reduction
96
91
86
81
76
71
66
61
56
51
46
41
36
31
26
21
16
11
Maximum Hourly N02 Reduction
17
33
49
65
81
97
11.036
n
9.657 '¦
6.898
a 5.519
2.760
1.381
.
17 Southeastern States Air Resource Managers, Inc., Emissions and Air Quality Modeling for SEMAP, Final
Report. October 15, 2014.
https://epd.georgia.gov/air/sites/epd.georgia.gov.air/files/related files/document/appendix d.pdf
67
-------
Figure 8. Maximum Hourly SO2 Reduction
Maximum Hourly S02 Reduction
96
91
86
81
76
71
66
61
56
51
46
41
36
31
26
21
16
11
6
1
15
29
43
57
71
85
99
2.720E0
2.380E0
2.040E0
1.700E0
1.360E0 "
1.020E0
6.806E-1 '
3.407E-1
n
There were modest reductions in these criteria pollutants, with a maximum hourly reduction of around
I I ppb of NO2 and just under 3 ppb of SO2.
68
-------
Appendix F: Additional Information Regarding Health-Related Benefits
Introduction to Benefits Analysis Methods
In the "Ten at the Top" (TATT) analysis we follow a "damage-function" approach in calculating health
co-benefits of the modeled changes in environmental quality. This approach estimates changes in
individual health and welfare endpoints (specific effects that can be associated with changes in air quality)
and estimates values of those changes assuming independence between the values of individual
endpoints. Total benefits are calculated simply as the sum of the values for all non-overlapping health
and welfare endpoints. The "damage-function" approach is the standard method for assessing costs and
benefits of environmental quality programs and has been used in several recent published analyses (Levy
et al., 2009; Hubbell et al., 2009; Tagaris et al., 2009).
Tables I and 2 summarize the human health and environmental benefits categories contained within the
total monetized benefits estimate and those categories that were not quantified due to limited data. The
list of unquantified benefit categories is not exhaustive, and neither is the quantification of each effect
complete. In order to identify the most meaningful human health and environmental co-benefits, we
excluded effects not identified as having at least a causal, likely causal, or suggestive relationship with the
affected pollutants in the most recent comprehensive scientific assessment, such as an Integrated Science
Assessment (ISA). This does not imply that additional relationships between these and other human
health and environmental co-benefits and the affected pollutants do not exist. Due to this decision
criterion, some effects that were identified in previous lists of unquantified benefits in other EPA
Regulatory Impact Assessment (R.IA) have been dropped (e.g., UVb exposure). In addition, some
quantified effects represent only a partial accounting of likely impacts due to limitations in the currently
available data (e.g., climate effects from CO2, etc).
69
-------
Table I. Human Health Effects of Pollutants Affected by CAU/TATT Emissions
Reductions
Benefits
Category
Effect Has
Effect Has
Source of
Specific Effect
Been
Quantified
Been
Monetized
More
Information
Improved Human Health
Reduced incidence
Premature mortality based on short-
~
~
of premature
mortality from
term exposure (all ages)
Premature respiratory mortality based
exposure to ozone
on long-term exposure (age 30-99")
Reduced incidence
of morbidity from
exposure to ozone
Hospital admissions—respiratory
(age > 65)
Emergency department visits for
asthma (all ages)
Asthma exacerbation (age 6-18)
~
~
~
~
~
~
ozone ISAd
Minor restricted-activity days (age 18-
65)
School absence days (age 5-17)
~
~
~
~
Decreased outdoor worker productivity
—
—
(age 18-65)
Other respiratory effects (e.g.,
medication use, pulmonary
inflammation, decrements in lung
functioning)
Cardiovascular (e.g., hospital
ozone ISAd
admissions, emergency department
visits)
Reproductive and developmental
effects (e.g., reduced birthweight,
restricted fetal growth)
—
—
Reduced incidence
of premature
mortality from
exposure to PM2.5
Adult premature mortality based on
cohort study estimates (age >25 or age
>30)
Infant mortality (age <1)
~
~
~
~
Reduced incidence
Non-fatal heart attacks (age > 18)
~
~
of morbidity from
exposure to PM2.5
Hospital admissions—respiratory (all
ages)
Hospital admissions—cardiovascular
~
~
~
~
(age >20)
Emergency department visits for
asthma (all ages)
Acute bronchitis (age 8-12)
~
~
~
~
ozone ISAd
Lower respiratory symptoms (age 7-
14)
Upper respiratory symptoms
(asthmatics age 9-11)
Asthma exacerbation (asthmatics age
6-18)
Lost work days (age 18-65)
~
~
~
~
~
~
~
~
Minor restricted-activity days (age 18-
65)
Chronic Bronchitis (age >26)
~
~
70
-------
R Effect Has Effect Has Source of
Cat"6^5 Specific Effect Been Been More
8 y Quantified Monetized Information
Emergency department visits for — —
cardiovascular effects (all ages)
Strokes and cerebrovascular disease — —
(age 50-79)
Other cardiovascular effects (e.g., other — —
ages)
Other respiratory effects (e.g., — —
pulmonary function, non-asthma ER
visits, non-bronchitis chronic diseases,
other ages and populations')
PM ISAC
Reproductive and developmental — —
effects (e.g., low birth weight, pre-term
births, etc.)
Cancer, mutagenicity, and genotoxicity — —
effects
PM ISA c.d
The benefits analysis in this chapter relies on an array of data inputs—including air quality modeling,
health impact functions and valuation functions among others—which are themselves subject to
uncertainty and may also contribute to the overall uncertainty in this analysis. As a means of
characterizing this uncertainty we use Monte Carlo methods for characterizing random sampling error
associated with the concentration response functions from epidemiological studies and economic
valuation functions. Second. While the contributions from additional data inputs to uncertainty in the
results are not quantified here, this analysis employs best practices in every aspect of its development.
To assess economic value in a damage-function framework, the changes in environmental quality must
be translated into effects on people or on the things that people value. In some cases, the changes in
environmental quality can be directly valued, as is the case for changes in visibility. In other cases, such as
for changes in ozone and PM, a health and welfare impact analysis must first be conducted to convert air
quality changes into effects that can be assigned dollar values.
We note at the outset that EPA rarely has the time or resources to perform extensive new research to
measure directly either the health outcomes or their values for regulatory analyses. Thus, similar to
Kunzli et al. (2000) and other recent health impact analyses, our estimates are based on the best
available methods of benefits transfer. Benefits transfer is a means of adapting primary research from
similar contexts to obtain the most accurate measure of benefits for the environmental quality change
under analysis. Adjustments are made for the level of environmental quality change, the socio-
demographic and economic characteristics of the affected population, and other factors to improve the
accuracy and robustness of benefits estimates.
Health Impact Assessment
Health Impact Assessment (HIA) quantifies changes in the incidence of adverse health impacts resulting
from changes in human exposure to specific pollutants, such as PM2.5. HIAs are a well-established
approach for estimating the retrospective or prospective change in adverse health impacts expected to
result from population-level changes in exposure to pollutants (Levy et al. 2009). PC-based tools such as
the environmental Benefits Mapping and Analysis Program (BenMAP) can systematize health impact
analyses by applying a database of key input parameters, including health impact functions and population
projections. Analysts have applied the HIA approach to estimate human health impacts resulting from
71
-------
hypothetical changes in pollutant levels (Hubbell et al. 2005; Davidson et al. 2007, Tagaris et al. 2009).
EPA and others have relied upon this method to predict future changes in health impacts expected to
result from the implementation of regulations affecting air quality (e.g. U.S. EPA, 2015a). For this
assessment, the HIAs are limited to those health effects that are directly linked to ambient PM2.5
concentrations. There may be other indirect health impacts associated with implementing emissions
controls, such as occupational health impacts for coal miners.
The HIA approach used in this analysis involves three basic steps: (I) utilizing CMAQ-generated
projections of PM2.5 and ozone air quality and estimating the change in the spatial distribution of the
ambient air quality; (2) determining the subsequent change in population-level exposure; (3) calculating
health impacts by applying concentration-response relationships drawn from the epidemiological
literature (Hubbell et al. 2009) to this change in population exposure.
A typical health impact function might look as follows:
Ay = y0 ¦ (e^'Ax — l) ¦ Pop
where yo is the baseline incidence rate for the health endpoint being quantified (for example, a health
impact function quantifying changes in mortality would use the baseline, or background, mortality rate
for the given population of interest); Pop is the population affected by the change in air quality; Ax is the
change in air quality; and (B is the effect coefficient drawn from the epidemiological study. Tools such as
BenMAP can systematize the HIA calculation process, allowing users to draw upon a library of existing
air quality monitoring data, population data and health impact functions. Figure A-1 provides a simplified
overview of this approach.
72
-------
Figure I. Illustration of BenMAP Approach
Baseline Air Quality Post-Polio' Scenario Ai r Quality
Economic Valuation of Health Impacts
After quantifying the change in adverse health impacts, the final step is to estimate the economic value of
these avoided impacts. The appropriate economic value for a change in a health effect depends on
whether the health effect is viewed ex ante (before the effect has occurred) or ex post (after the effect
has occurred). Reductions in ambient concentrations of air pollution generally lower the risk of future
adverse health effects by a small amount for a large population. The appropriate economic measure is
therefore ex ante Willingness to Pay (WTP) for changes in risk. However, epidemiological studies
generally provide estimates of the relative risks of a particular health effect avoided due to a reduction in
air pollution. A convenient way to use this data in a consistent framework is to convert probabilities to
units of avoided statistical incidences. This measure is calculated by dividing individual WTP for a risk
reduction by the related observed change in risk. For example, suppose a measure is able to reduce the
risk of premature mortality from 2 in 10,000 to I in 10,000 (a reduction of I in 10,000). If individual
WTP for this risk reduction is $ 100, then the WTP for an avoided statistical premature mortality
amounts to $1 million ($100/0.0001 change in risk). Using this approach, the size of the affected
population is automatically taken into account by the number of incidences predicted by epidemiological
studies applied to the relevant population. The same type of calculation can produce values for statistical
incidences of other health endpoints.
73
-------
For some health effects, such as hospital admissions, WTP estimates are generally not available. In these
cases, we use the cost of treating or mitigating the effect as a primary estimate. For example, for the
valuation of hospital admissions we use the avoided medical costs as an estimate of the value of avoiding
the health effects causing the admission. These cost of illness (COI) estimates generally (although not in
every case) understate the true value of reductions in risk of a health effect. They tend to reflect the
direct expenditures related to treatment but not the value of avoided pain and suffering from the health
effect.
We use the BenMAP-CE tool version I.I (U.S. EPA, 2015b) to estimate the health impacts and
monetized health co-benefits for the Mercury and Air Toxics Standards. Figure A-2 shows the data
inputs and outputs for the BenMAP-CE tool.
Figure 2. Data Inputs and Outputs for the BenMAP-CE Tool
Census
Population Dam
Population
Projections
2016
Woods & Poole
Population
Projections
Modeled Baseline
and Post-Control
2016 Ambient
PM25 and Os
Concentrations
Incremental Air
Quality Change
PMj 5 & O- Health
Functions
Prevalence Rates
PM2; & Oj-Related
Health Impacts
Economic
Valuation
Functions
Monetized PMIS
and 03-retated
Benefits
Blue identifies a user-selected input within the BenMAP model
Green identifies a data input generated outside of the BenMAP model
74
-------
Uncertainty Characterization
As for any complex analysis using estimated parameters and inputs from numerous models, there are
likely to be many sources of uncertainty affecting estimated results, including emission inventories, air
quality models (with their associated parameters and inputs), epidemiological health effect estimates,
estimates of values (both from WTP and COI studies), population estimates, income estimates, and
estimates of the future state of the world (i.e., regulations, technology, and human behavior). Each of
these inputs may be uncertain and, depending on its role in the co-benefits analysis, may have a
disproportionately large impact on estimates of total monetized co-benefits. For example, emissions
estimates are used in the first stage of the analysis. As such, any uncertainty in emissions estimates will
be propagated through the entire analysis. When compounded with uncertainty in later stages, small
uncertainties in emission levels can lead to large impacts on total monetized co-benefits.
The National Research Council (NRC) (2002, 2008) highlighted the need for EPA to conduct rigorous
quantitative analysis of uncertainty in its benefits estimates and to present these estimates to decision
makers in ways that foster an appropriate appreciation of their inherent uncertainty. In general, the
NRC concluded that EPA's methodology for calculating the benefits of reducing air pollution is
reasonable and informative in spite of inherent uncertainties. Since the publication of these reports, EPA
continues to improve the characterization of uncertainties for both health incidence and benefits
estimates. We use a Monte Carlo analysis to assess uncertainty quantitatively, as well as to provide a
qualitative assessment for those aspects that we are unable to address quantitatively.
We used Monte Carlo methods to characterize both sampling error and variability across the economic
valuation functions, including random sampling error associated with the concentration response
functions from epidemiological studies and random effects modeling. Monte Carlo simulation uses
random sampling from distributions of parameters to characterize the effects of uncertainty on output
variables, such as incidence of premature mortality. Specifically, we used Monte Carlo methods to
generate confidence intervals around the estimated health impact and dollar benefits. The reported
standard errors in the epidemiological studies determined the distributions for individual effect
estimates.
In benefit analyses of air pollution regulations conducted to date, the estimated impact of reductions in
premature mortality has accounted for 85 percent to 95 percent of total monetized benefits. Therefore,
it is particularly important to attempt to characterize the uncertainties associated with reductions in
premature mortality. The health impact functions used to estimate avoided premature deaths associated
with reductions in ozone have associated standard errors that represent the statistical errors around
the effect estimates in the underlying epidemiological studies. In our results, we report credible intervals
based on these standard errors, reflecting the uncertainty in the estimated change in incidence of
avoided premature deaths. We also provide multiple estimates, to reflect model uncertainty between
alternative study designs. EPA estimates PM-related mortality without assuming a health effect threshold
at low concentrations, based on the current body of scientific literature (U.S. EPA-SAB, 2009a, U.S. EPA-
SAB, 2009b).
Key sources of uncertainty in the PM2.5 health impact assessment include:
gaps in scientific data and inquiry;
variability in estimated relationships, such as epidemiological effect estimates, introduced
through differences in study design and statistical modeling;
75
-------
errors in measurement and projection for variables such as population growth rates;
errors due to misspecification of model structures, including the use of surrogate variables,
such as using PMio when PM2.5 is not available, excluded variables, and simplification of
complex functions;
biases due to omissions or other research limitations; and
additional uncertainties from benefits transfer method using BPT estimates.
In Table 4, we summarize some of the key uncertainties in the benefits analysis.
Table 4. Primary Sources of Uncertainty in the Benefits Analysis
1. Uncertainties Associated with Impact Functions
The value of the ozone or PM effect estimate in each impact function.
Application of a single impact function to pollutant changes and populations in all locations.
Similarity of future-year impact functions to current impact functions.
Correct functional form of each impact function.
Extrapolation of effect estimates beyond the range of ozone or PM concentrations observed in the
source epidemiological study.
Application of impact functions only to those subpopulations matching the original study population.
2. Uncertainties Associated with CMAQ-Modeled Ozone and PM Concentrations
Responsiveness of the models to changes in precursor emissions from the control policy.
Projections of future levels of precursor emissions, especially ammonia and crustal materials.
Lack of ozone and PM2.5 monitors in all rural areas requires extrapolation of observed ozone data from
urban to rural areas.
3. Uncertainties Associated with PM Mortality Risk
Limited scientific literature supporting a direct biological mechanism for observed epidemiological
evidence.
Direct causal agents within the complex mixture of PM have not been identified.
The extent to which adverse health effects are associated with low-level exposures that occur many
times in the year versus peak exposures.
The extent to which effects reported in the long-term exposure studies are associated with historically
higher levels of PM rather than the levels occurring during the period of study.
Reliability of the PM2.5 monitoring data in reflecting actual PM2.5 exposures.
4. Uncertainties Associated with Possible Lagged Effects
The portion of the PM-related long-term exposure mortality effects associated with changes in annual
PM levels that would occur in a single year is uncertain as well as the portion that might occur in
subsequent years.
5. Uncertainties Associated with Baseline Incidence Rates
Some baseline incidence rates are not location specific (e.g., those taken from studies) and therefore
may not accurately represent the actual location-specific rates.
Current baseline incidence rates may not approximate well baseline incidence rates in 2016.
Projected population and demographics may not represent well future-year population and
demographics.
6. Uncertainties Associated with Economic Valuation
Unit dollar values associated with health and welfare endpoints are only estimates of mean WTP and
therefore have uncertainty surrounding them.
Mean WTP (in constant dollars) for each type of risk reduction may differ from current estimates
because of differences in income or other factors.
7. Uncertainties Associated with Aggregation of Monetized Benefits
Health and welfare benefits estimates are limited to the available impact functions. Thus, unquantified
or unmonetized benefits are not included.
76
-------
PM2.5 mortality benefits represent a substantial proportion of total monetized co-benefits (over 90
percent), and these estimates have following key assumptions and uncertainties.
1. We assume that all fine particles, regardless of their chemical composition, are equally
potent in causing premature mortality. This is an important assumption, because PM2.5
produced via transported precursors emitted from EGUs may differ significantly from direct
PM2.5 released from diesel engines and other industrial sources, but the scientific evidence is
not yet sufficient to allow differential effects estimates by particle type.
2. We assume that the health impact function for fine particles is linear within the range of
ambient concentrations under consideration. Thus, the estimates include health co-benefits
from reducing fine particles in areas with varied concentrations of PM2.5, including both
regions that are in attainment with fine particle standard and those that do not meet the
standard down to the lowest modeled concentrations.
Benefits Analysis Data inputs
In Figure 5-2, we summarized the key data inputs to the health impact and economic valuation estimate.
Below we summarize the data sources for each of these inputs, including demographic projections,
effect coefficients, incidence rates and economic valuation. Our approach here is generally consistent
with the Regulatory Impact Analysis for the Ozone NAAQS RIA (U.S. EPA, 2015a).
Demographic Data
Quantified and monetized human health impacts depend on the demographic characteristics of the
population, including age, location, and income. We use projections based on economic forecasting
models developed by Woods and Poole, Inc. (Woods and Poole, 2008). The Woods and Poole (WP)
database contains county-level projections of population by age, sex, and race out to 2030. Projections
in each county are determined simultaneously with every other county in the United States to take into
account patterns of economic growth and migration. The sum of growth in county-level populations is
constrained to equal a previously determined national population growth, based on Bureau of Census
estimates (Hollman et al., 2000). According to WP, linking county-level growth projections together and
constraining to a national-level total growth avoids potential errors introduced by forecasting each
county independently. County projections are developed in a four-stage process:
1. First, national-level variables such as income, employment, and populations are forecasted.
2. Second, employment projections are made for 172 economic areas defined by the Bureau of
Economic Analysis, using an "export-base" approach, which relies on linking industrial-sector
production of non-locally consumed production items, such as outputs from mining,
agriculture, and manufacturing with the national economy. The export-based approach
requires estimation of demand equations or calculation of historical growth rates for output
and employment by sector.
3. Third, population is projected for each economic area based on net migration rates derived
from employment opportunities and following a cohort-component method based on
fertility and mortality in each area.
4. Fourth, employment and population projections are repeated for counties, using the
economic region totals as bounds. The age, sex, and race distributions for each region or
77
-------
county are determined by aging the population by single year of age by sex and race for each
year through 2016 based on historical rates of mortality, fertility, and migration.
Effect Coefficients
The first step in selecting effect coefficients is to identify the health endpoints to be quantified. We base
our selection of health endpoints on consistency with EPA's Integrated Science Assessments (which
replace the Criteria Document), with input and advice from the EPA Science Advisory Board - Health
Effects Subcommittee (SAB-HES), a scientific review panel specifically established to provide advice on
the use of the scientific literature in developing benefits analyses for air pollution regulations
(http://www.epa.gov/sab/). In general, we follow a weight of evidence approach, based on the biological
plausibility of effects, availability of concentration-response functions from well conducted peer-
reviewed epidemiological studies, cohesiveness of results across studies, and a focus on endpoints
reflecting public health impacts (like hospital admissions) rather than physiological responses (such as
changes in clinical measures like Forced Expiratory Volume (FEVI)).
There are several types of data that can support the determination of types and magnitude of health
effects associated with air pollution exposures. These sources of data include toxicological studies
(including animal and cellular studies), human clinical trials, and observational epidemiology studies. All of
these data sources provide important contributions to the weight of evidence surrounding a particular
health impact. However, only epidemiology studies provide direct concentration-response relationships
which can be used to evaluate population-level impacts of reductions in ambient pollution levels in a
health impact assessment.
For the data-derived estimates, we relied on the published scientific literature to ascertain the
relationship between PM and adverse human health effects. We evaluated epidemiological studies using
the selection criteria summarized in Table 5. These criteria include consideration of whether the study
was peer-reviewed, the match between the pollutant studied and the pollutant of interest, the study
design and location, and characteristics of the study population, among other considerations. The
selection of C-R functions for the benefits analysis is guided by the goal of achieving a balance between
comprehensiveness and scientific defensibility. In general, the use of results from more than a single
study can provide a more robust estimate of the relationship between a pollutant and a given health
effect. However, there are often differences between studies examining the same endpoint, making it
difficult to pool the results in a consistent manner. For example, studies may examine different
pollutants or different age groups. For this reason, we consider very carefully the set of studies available
examining each endpoint and select a consistent subset that provides a good balance of population
coverage and match with the pollutant of interest. In many cases, either because of a lack of multiple
studies, consistency problems, or clear superiority in the quality or comprehensiveness of one study
over others, a single published study is selected as the basis of the effect estimate.
78
-------
Table 5. Criteria Used When Selecting C-R Functions
Consideration
Comments
Peer-Reviewed
Research
Study Type
Peer-reviewed research is preferred to research that has not undergone the peer-
review process.
Among studies that consider chronic exposure (e.g., over a year or longer),
prospective cohort studies are preferred over ecological studies because they
control for important individual-level confounding variables that cannot be
controlled for in ecological studies.
Study Period Studies examining a relatively longer period of time (and therefore having more
data) are preferred, because they have greater statistical power to detect effects.
More recent studies are also preferred because of possible changes in pollution
mixes, medical care, and lifestyle over time. However, when there are only a few
studies available, studies from all years will be included.
Population Attributes The most technically appropriate measures of benefits would be based on impact
functions that cover the entire sensitive population but allow for heterogeneity
across age or other relevant demographic factors. In the absence of effect estimates
specific to age, sex, preexisting condition status, or other relevant factors, it may be
appropriate to select effect estimates that cover the broadest population to match
with the desired outcome of the analysis, which is total national-level health
impacts. When available, multi-city studies are preferred to single city studies
because they provide a more generalizable representation of the C-R function.
Study Size
Study Location
Pollutants Included in
Model
Measure of PM
Economically
Valuable Health
Effects
Non-overlapping
Endpoints
Studies examining a relatively large sample are preferred because they generally
have more power to detect small magnitude effects. A large sample can be obtained
in several ways, either through a large population or through repeated observations
on a smaller population (e.g., through a symptom diary recorded for a panel of
asthmatic children).
U.S. studies are more desirable than non-U.S. studies because of potential
differences in pollution characteristics, exposure patterns, medical care system,
population behavior, and lifestyle.
When modeling the effects of ozone and PM (or other pollutant combinations)
jointly, it is important to use properly specified impact functions that include both
pollutants. Using single-pollutant models in cases where both pollutants are
expected to affect a health outcome can lead to double-counting when pollutants are
correlated.
For this analysis, impact functions based on PM2.5 are preferred to PM10 because of
the focus on reducing emissions of PM2.5 precursors, and because air quality
modeling was conducted for this size fraction of PM. Where PM2.5 functions are not
available, PM10 functions are used as surrogates, recognizing that there will be
potential downward (upward) biases if the fine fraction of PM10 is more (less) toxic
than the coarse fraction.
Some health effects, such as forced expiratory volume and other technical
measurements of lung function, are difficult to value in monetary terms. These
health effects are not quantified in this analysis.
Although the benefits associated with each individual health endpoint may be
analyzed separately, care must be exercised in selecting health endpoints to include
in the overall benefits analysis because of the possibility of double-counting of
benefits.
79
-------
When several effect estimates for a pollutant and a given health endpoint have been selected, they are
quantitatively combined or pooled to derive a more robust estimate of the relationship. The BenMAP-
CE Manual Appendices provides details of the procedures used to combine multiple impact functions
(U.S. EPA, 2015b). In general, we used fixed or random effects models to pool estimates from different
studies of the same endpoint. Fixed effects pooling simply weights each study's estimate by the inverse
variance, giving more weight to studies with greater statistical power (lower variance). Random effects
pooling accounts for both within-study variance and between-study variability, due, for example, to
differences in population susceptibility. We used the fixed effects model as our null hypothesis and then
determined whether the data suggest that we should reject this null hypothesis, in which case we would
use the random effects model. Pooled impact functions are used to estimate hospital admissions and
asthma exacerbations. For more details on methods used to pool incidence estimates, see the BenMAP-
CE Manual Appendices (U.S EPA, 2015b), which are available with the BenMAP-CE software at
http://www2.epa.gov/benmap.html.
Effect estimates selected for a given health endpoint were applied consistently across all locations
nationwide. This applies to both impact functions defined by a single effect estimate and those defined by
a pooling of multiple effect estimates. Although the effect estimate may, in fact, vary from one location
to another (e.g., because of differences in population susceptibilities or differences in the composition of
PM), location-specific effect estimates are generally not available.
The specific studies from which effect estimates for the primary analysis are drawn are included in
Tables A-6 and A-7. In all cases where effect estimates are drawn directly from epidemiological studies,
standard errors are used as a partial representation of the uncertainty in the size of the effect estimate.
We refer readers interested in further details regarding each study to the Regulatory Impact analysis for
the Ozone NAAQS RIA and the Regulatory Impact Analysis for the PM RIA (EPA, 2012; EPA, 2015).
PMzj Premature Mortality Effect Coefficients
Both long- and short-term exposures to ambient levels of PM2.5 air pollution have been associated with
increased risk of premature mortality. The size of the mortality risk estimates from epidemiological
studies, the serious nature of the effect itself, and the high monetary value ascribed to prolonging life
make mortality risk reduction the most significant health endpoint quantified in this analysis.
Table 6. Health Endpoints and Epidemiological Studies Used to Quantify PM2.5-i"elated
Health Impacts
Relative Risk or Effect Estimate
Endpoint
Study
Study
Population
Premature Mortality
Premature mortality— Krewski et al. (2009)
cohort study, all-cause Lepeule et al. (2012)
Premature mortality— Woodruff et al. (1997)
all-cause
> 29 years RR = 1.06 (1.04-1.06) per 10 |ig/m3
> 24 years RR = 1.14 (1.07-1.22) per 10 |.ig/m3
Infant (< 1 OR = 1.04 (1.02-1.07) per 10 |ig/m3
year)
Chronic Illness
80
-------
Nonfatal heart attacks
Peters etal. (2001)
Pooled estimate:
Pope etal. (2006)
Sullivan etal. (2005)
Zanobetti et al. (2009)
Zanobetti and Schwartz
(2006)
Adults 0 18 OR = 1.62 (1.13-2.34) per 20 \ig/m3
years)
(3 = 0.00481 (0.00199)
(3 = 0.00198 (0.00224)
(3 = 0.00225 (0.000591)
(3 = 0.0053 (0.00221)
Hospital Admissions
Respiratory
Cardiovascular
Asthma-related
emergency department
visits
Zanobetti et al. (2009)—ICD
460-519 (All respiratory)
Kloog etal. (2012)—ICD 460-
519 (All Respiratory
Moolgavkar (2000)—ICD
490-496 (Chronic lung
disease)
Babin etal. (2007)—ICD 493
(asthma)
Sheppard (2003)—ICD 493
(asthma)
Pooled estimate:
Zanobetti etal. (2009)—ICD
390-459 (all cardiovascular)
Peng etal. (2009)—ICD 426-
427; 428; 430-438; 410-414;
429; 440-449 (Cardio-,
cerebro- and peripheral
vascular disease)
Peng etal. (2008)—ICD 426-
427; 428; 430-438; 410-414;
429; 440-449 (Cardio-,
cerebro- and peripheral
vascular disease)
Bell etal. (2008)—ICD 426-
427; 428; 430-438; 410-414;
429; 440-449 (Cardio-,
cerebro- and peripheral
vascular disease)
Moolgavkar (2000)—ICD
390-429 (all cardiovascular)
Pooled estimate:
Mar etal. (2010)
Slaughter etal. (2005)
Glad etal. (2012)
> 64 years (3=0.00207 (0.00446)
(3=0.0007 (0.000961)
18-64 years 1.02 (1.01-1.03) per 36 |ig/m3
< 19 years (3=0.002 (0.004337)
<18
> 64 years
RR = 1.04 (1.01-1.06) per 11.8 \ig/m3
(3=0.00189 (0.000283)
(3=0.00068
(0.000214)
(3=0.00071
(0.00013)
(3=0.0008
(0.000107)
20-64 years RR=1.04 (t statistic: 4.1) per 10 |ig/m3
All ages RR = i.04 (1.01-1.07) per 7 ng/m3
RR = 1.03 (0.98-1.09) per 10 ng/m3
(3=0.00392 (0.002843)
Other Health Endpoints
Acute bronchitis
Dockeryetal. (1996)
8-12 years OR = 1.50 (0.91-2.47) per 14.9 ng/m3
Asthma exacerbations
Work loss days
Pooled estimate:
Ostro etal. (2001) (cough,
wheeze, shortness of breath)
b
Mar etal. (2004) (cough,
shortness of breath)
Ostro (1987)
6-18 years1
OR = 1.03 (0.98-1.07)
OR = 1.06 (1.01-1.11)
OR = 1.08 (1.00-1.17) per 30 ng/m3
RR = 1.21 (1-1.47) per
RR = 1.13 (0.86-1.48) per 10 ng/m3
18-65 years (3=0.0046 (0.00036)
81
-------
Acute respiratory Ostro and Rothschild (1989)
symptoms (MRAD) (Minor restricted activity 18-65 years (3=0.00220 (0.000658)
days)
Upper respiratory ^ Asthmatics, , ,
^ Pope et al. (1991) » 1.003 (1-1.006) per 10 ug/m3
symptoms v v 1 9-11 years ^ 1 v ^6/
symptoms^'F310Schwartz and Neas (2000) 7-14 years OR = 1.33 (1.11-1.58) per 15 |ig/m3
Table 7. Health Endpoints and Epidemiological Studies Used to Quantify Ozone-Related
Health Impacts a
Study
Population
Relative Risk or Effect Estimate ((3)
Endpoint
Study
(with 95th Percentile Confidence
Interval or SE)
Premature Mortality
Premature mortality—
short-term
Smith et al. (2009)
Zanobetti and Schwartz
(2008)
All ages
(3 = 0.00032 (0.00008)
(3 = 0.00051 (0.00012)
Premature respiratory
mortality-long-term
Jerrett et al. (2009)
>29 years
(3 = 0.003971 (0.00133)
Hospital Admissions
Respiratory
Pooled estimate:
Katsouyanni et al. (2009)
Pooled estimate:
Glad et al. (2012)
> 65 years
(3 = 0.00064 (0.00040) penalized splines
(3 = 0.00306 (0.00117)
Asthma-related
emergency department
visits
Ito et al. (2007)
Mar and Koenig (2010)
0-99 years
(3 = 0.00521 (0.00091)
(3 = 0.01044 (0.00436) (0-17 yr olds)
(3 = 0.00770 (0.00284) (18-99 yr olds)
Peel et al. (2005)
Sarnat et al. (2013)
Wilson etal. (2005)
(3 = 0.00087 (0.00053)
(3 = 0.00111 (0.00028)
RR = 1.022 (0.996 - 1.049) per 25
Other Health Endpoints
Pooled estimate:a
Asthma exacerbation
Mortimer et al. (2002)
Schildcrout et al. (2006)
6-18 years
(3 = 0.00929 (0.00387)
(3 = 0.00222 (0.00282)
School loss days
Pooled estimate:
Chen et al. (2000)
5-17 years
(3 = 0.015763 (0.004985)
Acute respiratory
symptoms (MRAD)
Gilliland et al. (2001)
(3 = 0.007824 (0.004445)
Ostro and Rothschild (1989)
18-65 years
(3 = 0.002596 (0.000776)
a The original study populations were 5 to 12 years for Schildcrout et al. (2006) and 5-9 years for the
Mortimer et al. (2002) study. Based on advice from the SAB-HES, we extended the applied population to
6-18 years for all three studies, reflecting the common biological basis for the effect in children in the
broader age group. See: U.S. EPA-SAB (2004a) and NRC (2002).
Baseline Incidence Estimates
Epidemiological studies of the association between pollution levels and adverse health effects generally
provide a direct estimate of the relationship of air quality changes to the relative risk of a health effect,
rather than estimating the absolute number of avoided cases. For example, a typical result might be that
82
-------
a 10 ppb decrease in daily ozone levels might, in turn, decrease hospital admissions by 3 percent. The
baseline incidence of the health effect is necessary to convert this relative change into a number of
cases. A baseline incidence rate is the estimate of the number of cases of the health effect per year in
the assessment location, as it corresponds to baseline pollutant levels in that location. To derive the
total baseline incidence per year, this rate must be multiplied by the corresponding population number.
For example, if the baseline incidence rate is the number of cases per year per million people, that
number must be multiplied by the millions of people in the total population.
Table 8 summarizes the sources of baseline incidence rates and provides average incidence rates for the
endpoints included in the analysis. For both baseline incidence and prevalence data, we used age-specific
rates, where available. We applied concentration-response functions to individual age groups and then
summed over the relevant age range to provide an estimate of total population benefits.
Table 8. Baseline Incidence Rates and Population Prevalence Rates for Use in Impact
Functions, General Population
Rates
Endpoint
Parameter
Value
Source
Mortality
Daily or annual mortality rate
projected to 2015
Age-, cause-, and
county-specific rate
CDC Wonder (2004-2006)
U.S. Census bureau
Hospitalizations
Daily hospitalization rate
Age-, region-, state-,
county- and cause-
specific rate
2007 HCUP data files3
Asthma ER Visits
Daily asthma ER visit rate
Age-, region-, state-,
county- and cause-
specific rate
2007 HCUP data files3
Chronic Bronchitis
Annual prevalence rate per
person
1999 NHIS (American Lung
Association, 2002b, Table
• Aged 18-44
• Aged 45-64
• Aged 65 and older
0.0367
0.0505
0.0587
4)
Annual incidence rate per
person
0.00378
Abbey et al. (1995, Table 3)
Non-fatal Myocardial
Infarction (heart
attacks)
Daily non-fatal myocardial
infarction incidence rate per
person, 18+
Age-, region-, state-, and
county- specific rate
2007 HCUP data files3;
adjusted by 0.93 for
probability of surviving
after 28 days (Rosamond et
al., 1999)
Asthma
Exacerbations
Incidence among asthmatic
African-American children
• daily wheeze
• daily cough
• daily dyspnea
0.076
0.067
0.037
Ostro etal. (2001)
83
-------
Acute Bronchitis
Annual bronchitis incidence
rate, children
0.043
American Lung Association
(2002c, Table 11)
Lower Respiratory
Symptoms
Daily lower respiratory
symptom incidence among
children15
0.0012
Schwartz et al. (1994,
Table 2)
Upper Respiratory
Symptoms
Daily upper respiratory
symptom incidence among
asthmatic children
0.3419
Pope et al. (1991, Table 2)
Work Loss Days
Daily WLD incidence rate per
person (18-65)
• Aged 18-24
• Aged 25-44
• Aged 45-64
0.00540
0.00678
0.00492
1996 HIS (Adams,
Hendershot, and Marano,
1999, Table 41); U.S.
Bureau of the Census
(2000)
Rates
Endpoint
Parameter
Value
Source
School Loss Days
Rate per person per year,
assuming 180 school days
per year
9.9
National Center for
Education Statistics (1996)
and 1996 HIS (Adams et al.,
1999, Table 47);
Minor Restricted-
Activity Days
Daily MRAD incidence rate
per person
0.02137
Ostro and Rothschild
(1989, p. 243)
a Healthcare Cost and Utilization Program (HCUP) database contains individual level, state and regional-
level hospital and emergency department discharges for a variety of ICD codes.
b Lower respiratory symptoms are defined as two or more of the following: cough, chest pain, phlegm,
and wheeze.
The baseline incidence rates for hospital and emergency department visits that we applied in this analysis
are an improvement over the rates we used in the proposal analysis in two ways. First, these data are
newer, and so are a more recent representation of the rates at which populations of different ages, and
in different locations, visit the hospital and emergency department for illnesses that may be air pollution
related. Second, these newer data are also more spatially refined. For many locations within the U.S.,
these data are resolved at the county- or state-level, providing a better characterization of the
geographic distribution of hospital and emergency department visits. Newer and more spatially resolved
incidence rates are likely to yield a more reliable estimate of air pollution-related hospitalizations and
emergency department visits. Consistent with the proposal RIA, we continue to use county-level
mortality rates. We have projected mortality rates such that future mortality rates are consistent with
our projections of population growth (U.S. EPA, 2015b).
For the set of endpoints affecting the asthmatic population, in addition to baseline incidence rates,
prevalence rates of asthma in the population are needed to define the applicable population. Table 9 lists
the prevalence rates used to determine the applicable population for asthma symptom endpoints. Note
that these reflect current asthma prevalence and assume no change in prevalence rates in future years.
84
-------
Table 9. Asthma Prevalence Rates Used for this Analysis3
Asthma Prevalence Rates
Population Group
Value
Source
All Ages
<18
0.0780 American Lung Association (2010, Table 7)
0.0941
5-17
0.1070
18-44
0.0719
45-64
0.0745
65+
0.0716
African American, 5 to 17
0.1776 American Lung Association (2010, Table 9)
0.1553 American Lung Association15
African American, <18
a See ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHIS/2000/.
t Calculated by ALA for U.S. EPA, based on NHIS data (CDC, 2009)
Economic Valuation Estimates
Reductions in ambient concentrations of air pollution generally lower the risk of future adverse health
effects for a large population. Therefore, the appropriate economic measure is WTP for changes in risk
of a health effect rather than WTP for a health effect that would occur with certainty (Freeman, 1993).
Epidemiological studies generally provide estimates of the relative risks of a particular health effect that
is avoided because of a reduction in air pollution. We converted those to units of avoided statistical
incidence for ease of presentation. We calculated the value of avoided statistical incidences by dividing
individual WTP for a risk reduction by the related observed change in risk.18
WTP estimates generally are not available for some health effects, such as hospital admissions. In these
cases, we used the cost of treating or mitigating the effect as a primary estimate. These cost-of-illness
(COI) estimates generally understate the true value of reducing the risk of a health effect, because they
reflect the direct expenditures related to treatment, but not the value of avoided pain and suffering
(Harrington and Portney, 1987; Berger, 1987). We provide unit values for health endpoints (along with
18 To comply with Circular A-4, EPA provides monetized benefits using discount rates of 3 percent and
7 percent (OMB, 2003). These benefits are estimated for a specific analysis year (i.e., 2016), and most of
the PM benefits occur within that year with two exceptions: acute myocardial infarctions (AMIs) and
premature mortality. For AMIs, we assume 5 years of follow-up medical costs and lost wages. For
premature mortality, we assume that there is a "cessation" lag between PM exposures and the total
realization of changes in health effects. Although the structure of the lag is uncertain, EPA follows the
advice of the SAB-HES to assume a segmented lag structure characterized by 30 percent of mortality
reductions in the first year, 50 percent over years 2 to 5, and 20 percent over the years 6 to 20 after
the reduction in PM2.5 (U.S. EPA-SAB, 2004c). Changes in the lag assumptions do not change the total
number of estimated deaths but rather the timing of those deaths. Therefore, discounting only affects
the AMI costs after the analysis year and the valuation of premature mortalities that occur after the
analysis year. As such, the monetized benefits using a 7 percent discount rate are only approximately 10
percent less than the monetized benefits using a 3 percent discount rate.
85
-------
information on the distribution of the unit value) in Table 10. All values are in constant year 2006
dollars, adjusted for growth in real income out to 2016 using projections provided by Standard and
Poor's. Economic theory argues that WTP for most goods (such as environmental protection) will
increase if real income increases. Many of the valuation studies used in this analysis were conducted in
the late 1980s and early 1990s. Because real income has grown since the studies were conducted,
people's willingness to pay for reductions in the risk of premature death and disease likely has grown as
well. We did not adjust cost of illness-based values because they are based on current costs. Similarly,
we did not adjust the value of school absences, because that value is based on current wage rates. For
these two reasons, these cost of illness estimates may underestimate the economic value of avoided
health impacts in 2016. Readers interested in learning more about the basis for the economic value
estimates below may refer to the Ozone and PM NAAQS Regulatory Impact Analyses (EPA, 2012; EPA,
2015).
86
-------
Table 10. Unit Values for Economic Valuation of Health Endpoints (2010$)a
Central Estimate of Value Per Statistical
Incidence,
Income Level
Health Endpoint 2000 2016 Derivation of Distributions of Estimates
Premature Mortality (Value of a $6,800,000 $9,300,000 EPA currently recommends a central VSL of $6.3m (2000$) based on
Statistical Life) a Weibull distribution fitted to 26 published VSL estimates (5
contingent valuation and 21 labor market studies). The underlying
studies, the distribution parameters, and other useful information
are available in Appendix 5B of EPA's current Guidelines for
Preparing Economic Analyses (U.S. EPA, 2000).
Chronic Bronchitis (CB) $370,000 $510,000 The WTP to avoid a case of pollution-related CB is calculated as
where x is the severity of an average CB case, WTP13 is the WTP for a
severe case of CB, and $ is the parameter relating WTP to severity,
based on the regression results reported in Krupnick and Cropper
(1992). The distribution of WTP for an average severity-level case of
CB was generated by Monte Carlo methods, drawing from each of
three distributions: (1) WTP to avoid a severe case of CB is assigned
a 1/9 probability of being each of the first nine deciles of the
distribution of WTP responses in Viscusi et al. (1991); (2) the
severity of a pollution-related case of CB (relative to the case
described in the Viscusi study) is assumed to have a triangular
distribution, with the most likely value at severity level 6.5 and
endpoints at 1.0 and 12.0; and (3) the constant in the elasticity of
WTP with respect to severity is normally distributed with mean =
0.18 and standard deviation = 0.0669 (from Krupnick and Cropper
[1992]). This process and the rationale for choosing it is described in
detail in the Costs and Benefits of the Clean Air Act, 1990 to 2010
(U.S. EPA, 1999).
87
-------
Table 10.
Unit Values for Economic Valuation of Health Endpoints (2010$) (continued)
Central Estimate of Value Per
Statistical Incidence,
Income Level
Health Endpoint
2000
2016
Derivation of Distributions of Estimates
Non-fatal Myocardial
Infarction (heart
attack)
No distributional information available. Age-specific cost-of-illness values reflect lost
earnings and direct medical costs over a 5-year period following a non-fatal MI. Lost
earnings estimates are based on Cropper and Krupnick (1990). Direct medical costs are
based on simple average of estimates from Russell et al. (1998) and Wittels et al. (1990).
3% discount rate
Age 0-24
Age 25-44
Age 45-54
Age 55-65
Age 66 and over
$86,190
$96,238
$101,562
$181,208
$86,190
$86,190
$96,238
$101,562
$181,208
$86,190
Lost earnings:
Cropper and Krupnick (1990). Present discounted value of 5 years of lost earnings:
age of onset: at 3% at 7%
25-44 $8,774 $7,855
45-54 $12,932 11,578
55-65 $74,746 66,920
Direct medical expenses: An average of:
1. Wittels etal. (1990) ($102,658—no discounting)
_ 2. Russell et al. (1998), 5-year period ($22,331 at 3% discount rate; $21,113 at 7%
7% discount rate
Age 0-24
Age 25-44
Age 45-54
Age 55-65
Age 66 and over
$84,117
$94,238
$99,033
$170,332
$84,117
$84,117
$94,238
$99,033
$170,332
$84,117
discount rate)
Hospital Admissions
Chronic Obstructive
Pulmonary Disease
(COPD)
$17,961
$17,961
No distributional information available. The CO I estimates (lost earnings plus direct
medical costs) are based on ICD-9 code-level information (e.g., average hospital care costs,
average length of hospital stay, and weighted share of total COPD category illnesses)
reported in Agency for Healthcare Research and Quality (2000) (www.ahrq.gov).
Asthma Admissions
$9,627
$9,627
No distributional information available. The CO I estimates (lost earnings plus direct
medical costs) are based on ICD-9 code-level information (e.g., average hospital care costs,
average length of hospital stay, and weighted share of total asthma category illnesses)
reported in Agency for Healthcare Research and Quality (2000) (www.ahrq.gov).
88
-------
Table 10. Unit Values for Economic Valuation of Health Endpoints (2010$) (continued)
Central Estimate of Value Per
Statistical Incidence,
Income Level
Health Endpoint 2000 2016 Derivation of Distributions of Estimates
All Cardiovascular
$26,682
$26,682
No distributional information available. The CO I estimates (lost
earnings plus direct medical costs) are based on ICD-9 code-level
information (e.g., average hospital care costs, average length of
hospital stay, and weighted share of total cardiovascular category
illnesses) reported in Agency for Healthcare Research and Quality
(2000) (www.ahrq.gov).
All respiratory (ages 65+)
$26,632
$26,632
No distributions available. The CO I point estimates (lost earnings
plus direct medical costs) are based on ICD-9 code level
information (e.g., average hospital care costs, average length of
hospital stay, and weighted share of total COPD category
illnesses) reported in Agency for Healthcare Research and
Quality, 2000 (www.ahrq.gov).
All respiratory (ages 0-2)
$11,233
$11,233
No distributions available. The CO I point estimates (lost earnings
plus direct medical costs) are based on ICD-9 code level
information (e.g., average hospital care costs, average length of
hospital stay, and weighted share of total COPD category
illnesses) reported in Agency for Healthcare Research and
Quality, 2000 (www.ahrq.gov).
Emergency Room Visits for Asthma
$415
$415
No distributional information available. Simple average of two
unit CO I values:
(1) $311.55, from Smith et al. (1997); and
(2) $260.67, from Stanford et al. (1999).
(continued)
89
-------
Table 10. Unit Values for Economic Valuation of Health Endpoints (2010$) (continued)
Central Estimate of Value Per
Statistical Incidence,
Income Level
Health Endpoint 2000 2016 Derivation of Distributions of Estimates
Respiratory Ailments Not Requiring Hospitalization
Upper Respiratory Symptoms
(URS)
$32 $32
Lower Respiratory Symptoms
(LRS)
$17 $21
Asthma Exacerbations
$47 $57
Combinations of the three symptoms for which WTP estimates are
available that closely match those listed by Pope et al. result in seven
different "symptom clusters," each describing a "type" of URS. A dollar
value was derived for each type of URS, using mid-range estimates of WTP
(IEc, 1994) to avoid each symptom in the cluster and assuming additivity
of WTPs. In the absence of information surrounding the frequency with
which each of the seven types of URS occurs within the URS symptom
complex, we assumed a uniform distribution between $9.2 and $43.1.
Combinations of the four symptoms for which WTP estimates are available
that closely match those listed by Schwartz et al. result in 11 different
"symptom clusters," each describing a "type" of LRS. A dollar value was
derived for each type of LRS, using mid-range estimates of WTP (IEc, 1994)
to avoid each symptom in the cluster and assuming additivity of WTPs. The
dollar value for LRS is the average of the dollar values for the 11 different
types of LRS. In the absence of information surrounding the frequency with
which each of the 11 types of LRS occurs within the LRS symptom complex,
we assumed a uniform distribution between $6.9 and $24.46.
Asthma exacerbations are valued at $45 per incidence, based on the mean
of average WTP estimates for the four severity definitions of a "bad asthma
day," described in Rowe and Chestnut (1986). This study surveyed
asthmatics to estimate WTP for avoidance of a "bad asthma day," as
defined by the subjects. For purposes of valuation, an asthma exacerbation
is assumed to be equivalent to a day in which asthma is moderate or worse
as reported in the Rowe and Chestnut (1986) study. The value is assumed
have a uniform distribution between $15.6 and $70.8.
(continued)
90
-------
Table 10. Unit Values for Economic Valuation of Health Endpoints (2010$) (continued)
Central Estimate of Value Per
Statistical Incidence,
Income Level
Health Endpoint 2000 2016 Derivation of Distributions of Estimates
Acute Bronchitis
$389
$476
Assumes a 6-day episode, with the distribution of the daily value
specified as uniform with the low and high values based on those
recommended for related respiratory symptoms in Neumann et
al. (1994). The low daily estimate of $10 is the sum of the mid-
range values recommended by IEc (1994) for two symptoms
believed to be associated with acute bronchitis: coughing and
chest tightness. The high daily estimate was taken to be twice the
value of a minor respiratory restricted-activity day, or $110.
Work Loss Days (WLDs)
Variable (U.S.
median = $141)
Variable (U.S.
median = $141)
No distribution available. Point estimate is based on county-
specific median annual wages divided by 52 and then by 5—to
get median daily wage. U.S. Year 2000 Census, compiled by
Geolytics, Inc.
Minor Restricted Activity Days
(MRADs)
$55
$67
Median WTP estimate to avoid one MRAD from Tolley et al.
(1986). Distribution is assumed to be triangular with a minimum
of $22 and a maximum of $83, with a most likely value of $52.
Range is based on assumption that value should exceed WTP for a
single mild symptom (the highest estimate for a single
symptom—for eye irritation—is $16.00) and be less than that for
a WLD. The triangular distribution acknowledges that the actual
value is likely to be closer to the point estimate than either
extreme.
91
-------
Growth in \A/TP Reflecting Notional Income Growth Over Time
Our analysis accounts for expected growth in real income over time. Economic theory argues
that WTP for most goods (such as environmental protection) will increase if real incomes
increase. There is substantial empirical evidence that the income elasticity19 of WTP for health
risk reductions is positive, although there is uncertainty about its exact value. Thus, as real
income increases, the WTP for environmental improvements also increases. Although many
analyses assume that the income elasticity of WTP is unit elastic (i.e., a 10 percent higher real
income level implies a 10 percent higher WTP to reduce risk changes), empirical evidence
suggests that income elasticity is substantially less than one and thus relatively inelastic. As real
income rises, the WTP value also rises but at a slower rate than real income.
The effects of real income changes on WTP estimates can influence benefits estimates in two
different ways: through real income growth between the year a WTP study was conducted and
the year for which benefits are estimated, and through differences in income between study
populations and the affected populations at a particular time. Empirical evidence of the effect of
real income on WTP gathered to date is based on studies examining the former. The
Environmental Economics Advisory Committee (EEAC) of the Science Advisory Board (SAB)
advised EPA to adjust WTP for increases in real income over time but not to adjust WTP to
account for cross-sectional income differences "because of the sensitivity of making such
distinctions, and because of insufficient evidence available at present" (U.S. EPA-SAB, 2000). A
recent advisory by another committee associated with the SAB, the Advisory Council on Clean
Air Compliance Analysis, has provided conflicting advice. While agreeing with "the general
principle that the willingness to pay to reduce mortality risks is likely to increase with growth in
real income (U.S. EPA-SAB, 2004b, p. 52)" and that "The same increase should be assumed for
the WTP for serious non-fatal health effects (U.S. EPA-SAB, 2004b, p. 52)," they note that
"given the limitations and uncertainties in the available empirical evidence, the Council does not
support the use of the proposed adjustments for aggregate income growth as part of the
primary analysis" (U.S. EPA-SAB, 2004b, p. 53). Until these conflicting advisories have been
reconciled, EPA will continue to adjust valuation estimates to reflect income growth using the
methods described below, while providing sensitivity analyses for alternative income growth
adjustment factors.
Based on a review of the available income elasticity literature, we adjusted the valuation of
human health benefits upward to account for projected growth in real U.S. income. Faced with a
dearth of estimates of income elasticities derived from time-series studies, we applied estimates
derived from cross-sectional studies in our analysis. Details of the procedure can be found in
Kleckner and Neumann (1999). An abbreviated description of the procedure we used to
account for WTP for real income growth between 1990 and 2016 is presented below.
Reported income elasticities suggest that the severity of a health effect is a primary determinant
of the strength of the relationship between changes in real income and WTP. As such, we use
different elasticity estimates to adjust the WTP for minor health effects, severe and chronic
health effects, and premature mortality. Note that because of the variety of empirical sources
used in deriving the income elasticities, there may appear to be inconsistencies in the
19 Income elasticity is a common economic measure equal to the percentage change in WTP for
a I percent change in income.
92
-------
magnitudes of the income elasticities relative to the severity of the effects (a priori one might
expect that more severe outcomes would show less income elasticity of WTP). We have not
imposed any additional restrictions on the empirical estimates of income elasticity. One
explanation for the seeming inconsistency is the difference in timing of conditions. WTP for
minor illnesses is often expressed as a short term payment to avoid a single episode. WTP for
major illnesses and mortality risk reductions are based on longer term measures of payment
(such as wages or annual income). Economic theory suggests that relationships become more
elastic as the length of time grows, reflecting the ability to adjust spending over a longer time
period. Based on this theory, it would be expected that WTP for reducing long term risks
would be more elastic than WTP for reducing short term risks. The elasticity values used to
adjust estimates of benefits in 2016 are presented in Table I I.
Table I I. Elasticity Values Used to Account for Projected Real Income Growth3
Benefit Category Central Elasticity Estimate
Minor Health Effect 0.14
Severe and Chronic Health Effects 0.45
Premature Mortality 0.40
a Derivation of estimates can be found in Kleckner and Neumann (1999) and Chestnut (1997).
COI estimates are assigned an adjustment factor of 1.0.
In addition to elasticity estimates, projections of real gross domestic product (GDP) and
populations from 1990 to 2020 are needed to adjust benefits to reflect real per capita income
growth. For consistency with the emissions and benefits modeling, we used national population
estimates for the years 1990 to 1999 based on U.S. Census Bureau estimates (Hollman, Mulder,
and Kalian, 2000). These population estimates are based on application of a cohort-component
model applied to 1990 U.S. Census data projections (U.S. Bureau of Census, 2000). For the
years between 2000 and 2016, we applied growth rates based on the U.S. Census Bureau
projections to the U.S. Census estimate of national population in 2000. We used projections of
real GDP provided in Kleckner and Neumann (1999) for the years 1990 to 2010.20 We used
projections of real GDP (in chained 1996 dollars) provided by Standard and Poor's (2000) for
the years 2010 to 2016.21
Using the method outlined in Kleckner and Neumann (1999) and the population and income
data described above, we calculated WTP adjustment factors for each of the elasticity estimates
20 U.S. Bureau of Economic Analysis, Table 2A (1992$) (available at
http://www.bea.doc.gov/beaydn/0897nip2/ tab2a.htm.) and U.S. Bureau of Economic Analysis,
Economics and Budget Outlook. Note that projections for 2007 to 2010 are based on average
GDP growth rates between 1999 and 2007.
21 In previous analyses, we used the Standard and Poor's projections of GDP directly. This led to
an apparent discontinuity in the adjustment factors between 2010 and 201 I. We refined the
method by applying the relative growth rates for GDP derived from the Standard and Poor's
projections to the 2010 projected GDP based on the Bureau of Economic Analysis projections.
93
-------
listed in Table 12. Benefits for each of the categories (minor health effects, severe and chronic
health effects, and premature mortality) are adjusted by multiplying the unadjusted benefits by
the appropriate adjustment factor. Note that, for premature mortality, we applied the income
adjustment factor to the present discounted value of the stream of avoided mortalities occurring
over the lag period. Also note that because of a lack of data on the dependence of COI and
income, and a lack of data on projected growth in average wages, no adjustments are made to
benefits based on the COI approach or to work loss days and worker productivity. This
assumption leads us to underpredict benefits in future years because it is likely that increases in
real U.S. income would also result in increased COI (due, for example, to increases in wages
paid to medical workers) and increased cost of work loss days and lost worker productivity
(reflecting that if worker incomes are higher, the losses resulting from reduced worker
production would also be higher).
Table 12. Adjustment Factors Used to Account for Projected Real Income
Growth3
Benefit Category 2016
Minor Health Effect 1.15
Severe and Chronic Health Effects 1.29
Premature Mortality 1.25
a Based on elasticity values reported in Table I I, U.S. Census population projections, and
projections of real GDP per capita.
94
-------
References
Agency for Healthcare Research and Quality (AHRQ). 2000. HCUPnet, Healthcare Cost and
Utilization Project.
American Lung Association (ALA). 2010. Trends in Asthma Morbidiy and Mortality. American
Lung Association Epidemiology and Statistics Unit, Research and Program Services
Division.
Burnett RT, Smith-Doiron M, Stieb D, Raizenne ME, Brook JR, Dales RE, et al. 2001. Association
between ozone and hospitalization for acute respiratory diseases in children less than 2
years of age. Am J Epidemiol 153(5):444-452.
Chen L, Jennison BL, Yang W, Omaye ST. 2000. Elementary school absenteeism and air
pollution. Inhal Toxicol 12(1 I ):997-1016.
Cropper, M. L. and A. J. Krupnick. 1990. The Social Costs of Chronic Heart and Lung Disease.
Resources for the Future. Washington, DC. Discussion Paper QE 89-16-REV.
Davidson K, Hallberg A, McCubbin D, Hubbell BJ. 2007. Analysis of PM2.5 Using the
Environmental Benefits Mapping and Analysis Program (BenMAP). J Toxicol Environ
Health 70: 332—346.
Dockery, D.W., J. Cunningham, A.I. Damokosh, L.M. Neas, J.D. Spengler, P. Koutrakis, J.H.
Ware, M. Raizenne, and F.E. Speizer. 1996. Health Effects of Acid Aerosols on North
American Children-Respiratory Symptoms. Environmental Health Perspectives
l04(5):500-505.
Eisenstein, E.L., L.K. Shaw, K.J. Anstrom, C.L. Nelson, Z. Hakim, V. Hasselblad and D.B. Mark.
2001. Assessing the Clinical and Economic Burden of Coronary Artery Disease: 1986-
1998. Medical Care 39(8):824-35.
Gilliland FD, Berhane K, Rappaport EB, Thomas DC, Avol E, Gauderman WJ, et al. 2001. The
effects of ambient air pollution on school absenteeism due to respiratory illnesses.
Epidemiology 12( I ):43-54.
Hollman, F.W., T.J. Mulder, and J.E. Kalian. January 2000. Methodology and Assumptions for the
Population Projections of the United States: 1999 to 2100. Population Division Working
Paper No. 38, Population Projections Branch, Population Division, U.S. Census Bureau,
Department of Commerce.
Hubbell BJ, Hallberg A, McCubbin D, Post, E. 2005. Health-Related Benefits of Attaining the 8-
Hr Ozone Standard. Environ Health Perspect I I 3: 73—82.
Ito, K. 2003. Associations of Particulate Matter Components with Daily Mortality and Morbidity
in Detroit, Michigan. In Revised Analyses of Time-Series Studies of Air Pollution and
Health. Special Report. Health Effects Institute, Boston, MA.
Jerrett M, Burnett RT, Pope CA, III, et al. 2009. Long-Term Ozone Exposure and Mortality. N
Engl J Med 360:1085-95.
95
-------
Kleckner, N., and J. Neumann. June 3, 1999. Recommended Approach to Adjusting WTP
Estimates to Reflect Changes in Real Income. Memorandum to Jim Democker, U.S.
EPA/OPAR.
Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi, Y, et al. 2009. Extended follow-up and
spatial analysis of the American Cancer Society study linking particulate air pollution and
mortality. HEI Research Report, 140, Health Effects Institute, Boston, MA.
Kunzli, N., R. Kaiser, S. Medina, M. Studnicka, O. Chanel, P. Filliger, et al. 2000. Public-health
imjpact of outdoor and traffic-related air pollution: A European Assessment. The Lancet
356(9232):795-80l.
Laden, F., J. Schwartz, F.E. Speizer, and D.W. Dockery. 2006. Reduction in Fine Particulate Air
Pollution and Mortality. American Journal of Respiratory and Critical Care Medicine
173:667-672.
Levy J I, Baxter LK, Schwartz J. 2009. Uncertainty and variability in health-related damages from
coal-fired power plants in the United States. Risk Anal, doi: 10.1 I I I /j. 1539-
6924.2009.01227.x [Online 9 Apr 2009]
Moolgavkar SH, Luebeck EG, Anderson EL. 1997. Air pollution and hospital admissions for
respiratory causes in Minneapolis St. Paul and Birmingham. Epidemiology. 8(4):364-370.
Moolgavkar, S.H. 2000. Air Pollution and Hospital Admissions for Diseases of the Circulatory
System in Three U.S. Metropolitan Areas. Journal of the Air and Waste Management
Association 50:1 199-1206.
Moolgavkar, S.H. 2003. Air Pollution and Daily Deaths and Hospital Admissions in Los Angeles
and Cook Counties. In Revised Analyses of Time-Series Studies of Air Pollution and
Health. Special Report. Boston, MA: Health Effects Institute.
National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed Air
Pollution Regulations. Washington, DC: The National Academies Press.
National Research Council (NRC). 2008. Estimating Mortality Risk Reduction and Economic
Benefits from Controlling Ozone Air Pollution. National Academies Press. Washington,
DC.
B., M. Lipsett, J. Mann, H. Braxton-Owens, and M. White. 2001. Air Pollution and
Exacerbation of Asthma in African-American Children in Los Angeles. Epidemiology
12(2):200-208.
B.D. 1987. Air Pollution and Morbidity Revisited: A Specification Test. Journal of
Environmental Economics Management 14:87-98.
B.D. and S. Rothschild. 1989. Air Pollution and Acute Respiratory Morbidity: An
Observational Study of Multiple Pollutants. Environmental Research 50:238-247.
L., P. E. Tolbert, M. Klein, et al. 2005. Ambient air pollution and respiratory emergency
department visits. Epidemiology. Vol. 16 (2): 164-74.
Ostro,
Ostro,
Ostro,
Peel, J.
96
-------
Peters, A., D.W. Dockery, J.E. Muller, and M.A. Mittleman. 2001. Increased Particulate Air
Pollution and the Triggering of Myocardial Infarction. Circulation 103:2810-2815.
Pope, C.A, III, D.W. Dockery, J.D. Spengler, and M.E. Raizenne. 1991. Respiratory Health and
PMio Pollution: A Daily Time Series Analysis. American Review of Respiratory Diseases
144:668-674.
Ransom, Michael, and C. Arden Pope. 1992. M.R. Ransom and C.A. Pope, III, Elementary school
absences and PMio pollution in Utah Valley. Environ. Res. 58, pp. 204—219.
Russell, M.W., D.M. Huse, S. Drowns, E.C. Hamel, and S.C. Hartz. 1998. Direct Medical Costs of
Coronary Artery Disease in the United States. American Journal of Cardiology
81 (9): I I 10-1 I 15.
Samet, J.M., S.L. Zeger, F. Dominici, F. Curriero, I. Coursac, D.W. Dockery, J. Schwartz, and A.
Zanobetti. 2000. The National Morbidity, Mortality and Air Pollution Study: Part II:
Morbidity, Mortality and Air Pollution in the United States. Research Report No. 94,
Part II. Health Effects Institute, Cambridge MA. June.
Schwartz J. 1994a. PM(I0) Ozone, and Hospital Admissions For the Elderly in Minneapolis St
Paul, Minnesota. Arch Environ Health. 49(5):366-374.
Schwartz J. 1994b. Air Pollution and Hospital Admissions For the Elderly in Detroit, Michigan.
Am J Respir Crit Care Med. 150(3):648-655.
Schwartz J. 1995. Short term fluctuations in air pollution and hospital admissions of the elderly
for respiratory disease. Thorax. 50(5):531 -538.
Schwartz, J. 1993. Particulate Air Pollution and Chronic Respiratory Disease. Environment
Research 62:7-1 3.
Schwartz, J. 2005. How sensitive is the association between ozone and daily deaths to control
for temperature? Am J Respir Crit Care Med. Vol. 171 (6): 627-31.
Schwartz, J., and L.M. Neas. 2000. Fine Particles are More Strongly Associated than Coarse
Particles with Acute Respiratory Health Effects in Schoolchildren. Epidemiology I 1:6-10.
Sheppard, L. 2003. Ambient Air Pollution and Nonelderly Asthma Hospital Admissions in
Seattle, Washington, 1987-1994. In Revised Analyses of Time-Series Studies of Air
Pollution and Health. Special Report. Boston, MA: Health Effects Institute.
Smith, D.H., D.C. Malone, K.A Lawson, L.J. Okamoto, C. Battista, and W.B. Saunders. 1997. A
National Estimate of the Economic Costs of Asthma. American Journal of Respiratory
and Critical Care Medicine 156(3 Pt l):787-793.
Tagaris E, Liao KJ, Delucia AJ, et al. 2009. Potential impact of climate change on air-pollution
related human health effects. Environ. Sci. Technol. 43: 4979—4988.
Tolley, G.S. et al. 1986. Valuation of Reductions in Human Health Symptoms and Risks.
University of Chicago. Final Report for the U.S. Environmental Protection Agency.
January
97
-------
U.S. Environmental Protection Agency - Science Advisory Board (U.S. EPA-SAB). 2009b. Review
of EPA's Integrated Science Assessment for Particulate Matter (First External Review
Draft, December 2008). EPA-COUNCIL-09-008. May. Available on the Internet at
.
U.S. Environmental Protection Agency (U.S. EPA), 2015b. Environmental Benefits Mapping and
Analysis Program—Community Edition (Version I.I). Research Triangle Park, NC.
Available on the Internet at .
U.S. Environmental Protection Agency (U.S. EPA), 2015a. Regulatory Impact Analysis, 2015
National Ambient Air Quality Standards for Ground-level Ozone, Chapter 6. Office of
Air Quality Planning and Standards, Research Triangle Park, NC. October. Available at
.
U.S. Environmental Protection Agency Science Advisory Board (U.S. EPA-SAB). 2009a.
Consultation on EPA's Particulate Matter National Ambient Air Quality Standards:
Scope and Methods Plan for Health Risk and Exposure Assessment. EPA-COUNCIL-09-
009. May. Available on the Internet at
.
Vedal, S., J. Petkau, R. White, and J. Blair. 1998. Acute Effects of Ambient Inhalable Particles in
Asthmatic and Nonasthmatic Children. American Journal of Respiratory and Critical
Care Medicine 157(4): 1034-1043.
Wilson, A. M., C. P. Wake, T. Kelly, et al. 2005. Air pollution, weather, and respiratory
emergency room visits in two northern New England cities: an ecological time-series
study. Environ Res. Vol. 97 (3): 312-21.
Wittels, E.H., J.W. Hay, and AM. Gotto, Jr. 1990. Medical Costs of Coronary Artery Disease in
the United States. American Journal of Cardiology 65(7):432-440.
Woodruff TJ, Parker JD, Schoendorf KC. 2006. Fine particulate matter (PM2.5) air pollution and
selected causes of postneonatal infant mortality in California. Environmental Health
Perspectives I 14(5):786-90.
Woods & Poole Economics Inc. 2008. Population by Single Year of Age CD. CD-ROM. Woods
& Poole Economics, Inc. Washington, D.C.
98
------- |