The Multi-pollutant Report:
Technical Concepts
ri
r'
. ;*
~zone
Oil1
Toxics
* 'J
m\ >
—
v
"Q
Q.
-------
TABLE OF CONTENTS
Chapter Page
INTRODUCTION 1-1
1: MULTI-POLLUTANT CONCEPTS FOR AIR QUALITY MANAGEMENT 1 -1
Technical Elements for a Multi-pollutant Perspective 1-2
Pollutants May Have Common Emissions Sources 1-2
Control Technologies Can Affect Multiple Pollutants 1-3
Atmospheric Processes Create, Remove, and Transform Pollutants 1-6
Exposure Pathways and Risks Are Affected by Multiple Pollutants 1-7
Pollutants Affect Ecosystems and Other Environmental Concerns 1-9
Using Multi-pollutant Concepts for Air Quality Management 1-10
Summary 1-14
2: MULTI-POLLUTANT AIR QUALITY ISSUES AT THE
NATIONAL LEVEL 2-1
Identifying Areas with Multiple Air Quality Problems 2-1
Ozone and Particle Pollution 2-1
Hazardous Air Pollutants 2-2
Where Multi-pollutant Air Quality Issues Occur 2-3
Correlations Among Multi-pollutant Issues 2-5
Summary 2-6
3: MULTI-POLLUTANT AIR QUALITY AT THE LOCAL LEVEL, AN EXAMPLE 3 -1
Ambient Air Characterization 3-3
Criteria Pollutants 3-4
Fine Particles 3-4
Ozone 3-5
Air Toxics 3-6
Source Contributions to Air Quality Problems 3-9
Summary 3-13
4: CURRENT TRENDS AND PROJECTED IMPROVEMENTS OF AIR QUALITY
AT THE NATIONAL LEVEL 4-1
Decreasing Trends in Multi-pollutant Concentrations 4-1
Clean Air Rules Will Further Improve Air Quality 4-2
Ozone and Particle Pollution 4-4
Air Toxics and Mercury Deposition 4-6
Visibility 4-7
Nitrogen and Sulfur Deposition 4-8
Summary 4-9
5: MULTI-POLLUTANT ANALYTICAL PRODUCTS AND CAPABILITIES 5-1
Integrated Emissions Inventory - 2002, 2005, and 2008 National
Emissions Inventories 5-2
Integrated Monitoring Network - NCore 5-3
One Atmosphere Air Quality Modeling - CMAQ Model 5-4
-------
Multi-pollutant Modeling Platform - 2002 and Projected Future Years
Spatial Predictions of Air Quality Data - CDC/PHASE Project
Summary
-------
Introduction
The U.S. Environmental Protection Agency's (EPA) Office of Air Quality Planning and
Standards (OAQPS) is responsible for a broad set of air quality management activities including
setting standards, developing cost-effective control strategies, and evaluating air quality. To
date, OAQPS has approached air quality management one pollutant at a time based largely on
the legislative directions of the 1990 Clean Air Act Amendments (CAAA). The 1990 CAAA
established a mix of emissions, technology, and ambient air quality goals focused on reducing
criteria air pollutants (CAPs) and hazardous air pollutants (HAPs). The 1990 CAAA also
include programs to reduce acid deposition and protect the stratospheric ozone layer. Air quality
management responsibilities are shared among federal, state, local, and tribal governments.
Major current components of this air quality management system include the following:
• National Ambient Air Quality Standards (NAAQS) for six criteria pollutants
• State Implementation Plans (SIPs) for criteria pollutants including regional haze
• Federal regulations such as the NOx SIP Call, Clean Air Interstate Rule (CAIR), Clean
Air Mercury Rule (CAMR), Clean Air Visibility Rule, and Heavy-Duty Diesel and
Nonroad Diesel Rules
• An air toxics (HAPs) program to develop technology-based Maximum Achievable
Control Technology (MACT) standards and residual risk standards for HAPs
The technical infrastructure to support the implementation of the 1990 CAAA includes research,
monitoring networks, emission inventory development, modeling, exposure and risk assessment,
and cost/benefit analysis.
The current system has resulted in significant reductions in emissions and pollutant
concentrations during a period of strong economic growth. Acknowledging the past successes,
the 2004 National Academy of Sciences (NAS) study "Air Quality Management in the United
States" (National Research Council, 2004) called for modifying current air quality management
practices to integrate assessment, planning, and implementation efforts across all air quality and
environmental issues—that is, a multi-pollutant (and multimedia) focus. The NAS study has
catalyzed reorganization activities and new research themes throughout EPA's air quality
management program. Using NAS's report as a blueprint, OAQPS has begun to transition
toward a comprehensive, multi-pollutant treatment of our nation's air quality problems.
Managing air quality with a multi-pollutant approach is a challenge for our evolving air
quality management system. The adoption of this type of approach to environmental decision-
making requires an improved understanding and appreciation of the scientific complexities of co-
pollutant interactions. In a discussion of the air quality management of fine particles, the North
American Research Strategy for Tropospheric Ozone (NARSTO, 2004) stated
"The current understanding of atmospheric processes shows that PM2.5 problems
are related to ground-level ozone, acid rain, and climate issues and share many of
the same sources. This recognition provides the impetus for integrated and
optimized management strategies that accommodate different atmospheric
responses for each pollutant."
1-1
-------
Continuing to improve the air quality management system requires the addition of a
multi-pollutant component to the existing framework (shown in Figure 1-1). Technical concepts
that need to be addressed include shared emissions sources, multi-pollutant effects from control
technologies, common receptors, and shared atmospheric chemistry, transport, deposition, and
exposure pathways.
Emissions
Sources
"\
Emissions
J
Atmospheric
Environment
Exposure &
Effects
A
Atmospheric
Processing
\.
Pollutant
/ •
Concentrations &
/
Deposition
Human
Health
Meteorology
Visibility &
Climate
J
Ecosystems
Environmental
Decisionmaking
"\
>
J
Technical
Analyses
Societal
Factors
Environmental
Management
Emissions
Reduction
Programs
Figure 1-1. Framework for informing air quality management.
SOURCE: Adapted from NARSTO, 2004a
This report focuses on multi-pollutant concepts as they relate to our air quality
management system. We focus on ozone, fine particles, and air toxics because these pollutants
remain among the most persistent air quality problems affecting human health. However, we
recognize that a multi-pollutant definition is far broader and should include coarse particles,
sulfur dioxide (S02), nitrogen oxide (NOx), and environmental concerns such as regional haze,
deposition to ecosystems, stratospheric ozone protection, and climate change. With this report,
we hope to facilitate a common understanding of multi-pollutant concepts to foster collaboration
within and across the technical and policy disciplines throughout OAQPS; explore multi-
pollutant analytic issues; and illustrate the initial development and implementation of a technical
infrastructure to support a multi-pollutant approach to our programs.
This report is divided into the following chapters:
1. Multi-pollutant Concepts for Air Quality Management. This chapter explores the
science behind the multi-pollutant approach including the links among emissions sources,
control technologies, atmospheric processes, and environmental exposure.
1-2
-------
2. Multi-pollutant Air Quality Issues at the National Level. This chapter explores the
spatial and statistical correlations among ozone, fine particles, and HAPs across the
continental US.
3. Multi-pollutant Air Quality at the Local Level, an Example. This chapter features a
local assessment of multi-pollutant air quality in Detroit, MI.
4. Current Trends and Projected Improvements of Air Quality at the National Level.
This chapter provides current trends in CAPs and HAPs and projected changes to future
air quality.
5. Multi-pollutant Analytical Products and Capabilities. This chapter outlines current
efforts to develop technical infrastructure including a multi-pollutant emissions
inventory, monitoring, and modeling capabilities.
1-3
-------
Chapter 1: Multi-pollutant Concepts for Air Quality Management
A starting point for discussion of a multi-pollutant approach to air quality management is
to recognize the numerous interactions that occur within the atmosphere and across
environmental media. Figure 1-1 illustrates the relationships among pollutant sources (e.g.,
transportation and industry), atmospheric processes (e.g., photochemistry and dispersion), and
effects on human health and ecosystems (e.g., acidification and eutrophication) from exposure to
air pollution. A multi-pollutant perspective requires an understanding of these interactions and
an ability to account for them in analytical assessments to inform the development of programs
and policies within our air quality management system. The relationships among pollutants,
sources, transport and transformation pathways, and environmental effects are complex. This
chapter provides details of these interactions and their relevance for air quality management.
~Transport I Transformation
A Sources
~ Removal
~ Photochemistry
~ Chemical Transformations
~ Prevailing
v Winds
~ Cloud Processes
~Dispersion
^ Vertical .
w Mixing \Vi\\SY».
A Industry
A Transportation
A Agriculture i R
\\>>\\\\\\\\\\ Y Dry
T Deposition A Visibility
Sewage
PI
: I) Forest
Productivity
EffeCtS L • Drink,n9 Water S2&M M • fecTltural
Runoff
Estuaries
Resources
A Soils
1 [ • Aquati
Groundwater L'V Ecosystem
Agricultural
Products
Human
Health
Figure 1-1. Conceptualized depiction of pollution sources, atmospheric
processes, and effects on human health and ecosystems from exposure to air
pollution.
Source: Adapted from the National Science and Technology Council Committee on Environment and
Natural Resources, Air Quality Research Subcommittee, 1999
1-1
-------
Technical Elements for a Multi-pollutant Perspective
It is important that environmental policymakers understand the following technical
elements that form the basis of a multi-pollutant approach:
• Pollutants may have common emissions sources.
• Control technologies can affect multiple pollutants.
• Atmospheric processes create, remove, and transform pollutants.
• Exposure pathways and risks are affected by multiple pollutants (and may be enhanced
by pollutant interactions).
• Pollutants affect ecosystems and other atmospheric concerns in addition to human health.
Each technical element is described in more detail below.
Pollutants May Have Common Emissions Sources
Many human activities emit a variety of air pollutants. Figure 1-2 shows the national-
level contributions of major source categories to emissions of specific criteria and air toxic
pollutants. These air pollutants include primary emissions of stable pollutants (i.e., those that do
not react—or react very slowly—in the atmosphere) and emissions of precursor pollutants (i.e.,
those that react and contribute to secondary formation of pollutants such as ozone or
formaldehyde). For example, emissions from combustion-based sources (e.g., electricity
generation and motor vehicles) include directly emitted criteria pollutants (e.g., carbon monoxide
[CO], primary fine particles [PM2.5], and SO2), air toxics (e.g., benzene, lead, some volatile
organic compounds [VOCs], and trace metals including mercury), precursor emissions (e.g.,
NOx, SO2, some VOCs, and ammonia [NH3]), and greenhouse gases (GHGs) (e.g., carbon
dioxide [CO2]).
The contributions of source categories vary across pollutants. For example, SO2
emissions are dominated by electric generating units (EGUs), other fossil fuel combustion
industrial sources (e.g., boilers), and industrial processes (e.g., petroleum refineries), whereas the
major contributors to primary fine particles are EGUs, fires, road dust (part of "miscellaneous" in
Figure 1-2), and residential wood combustion. It is important to note that the relative
contribution of sources, shown in Figure 1-2 at the national level, differs across regions of the
country and within local areas, both urban and rural. Recognizing the multi-pollutant nature of
emissions sources enables consistent emission estimation procedures and regulatory reporting
requirements across sources and activities, leading to better informed decisions within EPA's air
quality management system.
1-2
-------
¦ EGUs ~ Nonroad
~ Fossil fuel combustion ~ Onroad vehicles ~ Waste disposal
| Industrial processes Residential wood combustion ~ Fertilizer & livestock
~ Miscellaneous ~ Solvent use ~ Fires
VOC
I
PM
S07 I
10
NH, III 11 II
N0< 1
Lead i i ~
X
1—T
I
I
TTTBT
Manganese I
Arsenic I
Cadmium I
Benzene
Formaldehyde
1,3-Butadiene
~I
I ¦
0
20
40 60
Percent of emissions
80
100
Figure 1-2. National-level source contributions to CAPs and HAPs by source
sector for 2002.
Source: U.S. Environmental Protection Agency, 2007a
Control Technologies Can A ffect Multiple Pollutants
Because sources emit more than one pollutant (as shown in Figure 1-2), control
technologies or other approaches to reduce emissions (e.g., reduced demand) have the potential
to affect multiple pollutants. Figure 1-3 shows examples of multi-pollutant controls available at
power plants:
1. Selective catalytic reduction (SCR) - ammonia reacts with NOx on a catalyst to reduce
NOx to nitrogen and can enhance removal of mercury through oxidation downstream in
the process for bituminous-fired units.
2. Electrostatic precipitator (ESP) - electrical discharge charges fly ash particles in the flue
gas; charged particles are collected on a surface. The collected particles also include
heavy metal HAPs such as lead, cadmium, arsenic, and nickel.
3. Fabric filter (FF) - flue gas passes through tightly woven fabric, resulting in the
collection of particles including heavy metal HAPs on the fabric. In general, an FF is
interchangeable with an ESP. In some units, an FF is placed downstream of an ESP (in
which case it is referred to as a "COHPAC" unit). For example, a COHPAC unit would
be used if an existing plant cannot meet particle levels with the ESP alone; the added FF
downstream in a COHPAC unit would provide additional particle removal.
1-3
-------
4. Wet scrubber - flue gas desulfurization (FGD) occurs as flue gas comes in contact with
limestone or lime slurry in the scrubber; S02 reacts to form calcium sulfate/calcium
sulfite salts, which are removed along with acid gases such as hydrochloric acid (HC1),
hydrofluoric acid (HF), and sulfur trioxide (SO3); this process also captures soluble, or
ionic, mercury.
FF or
ESP
SCR
FGD
Absorbers
Figure 1-3. Examples of multi-pollutant controls at power plants.
Automobile emissions controls employ all these technologies to reduce air pollution-
causing emissions from automobiles. The major pollutants emitted from a vehicle fall into three
basic categories: tailpipe (or exhaust), evaporative, and life cycle. As shown in Figure 1-4, both
exhaust (VOCs, fine particles, NOx, CO, air toxics, and CO2) and evaporative emissions (VOCs
and air toxics) are multi-pollutant in nature.
When people think of air pollution caused by vehicles, most think of exhaust emissions,
i.e., the products of burning fuel in the engine emitted from the exhaust system. Exhaust
emissions controls fall into three areas:
1. Increasing engine efficiency - has been gradually improved with progress in the
following technologies:
• electronic ignition
• fuel injection systems
• electronic control units
1-4
-------
2. Increasing vehicle efficiency - has come from reducing fuel consumption and related
emissions:
• lightweight vehicle design
• minimized air resistance
• reduced rolling resistance
• improved powertrain efficiency
• increased spark to the spark plug
• regenerative braking
3. Cleaning up emissions - advances in engine and vehicle technology continually reduce
the amount of pollutants generated, but this is generally considered insufficient to meet
emissions goals. Therefore, technologies to react with and clean up the remaining
emissions have long been an essential part of emissions control; some of these include
• air injection
• exhaust gas recirculation
• use of catalytic converters
4. Reducing Vehicle Miles Traveled (VMT) - use of public transportation, flexi-place, and
other innovative voluntary programs can reduce emissions.
Evaporative Emissions
• Volatile Organic Compounds
• Air Toxics
Refueling Losses
• Volatile Organic Compounds
• Air Toxics
Air Conditioning System
• CFCs
Tire and Brake Wear
• PM
Exhaust Emissions
Volatile Organic Compounds
Particles
• Nitrogen Oxides
• Carbon Monoxide
• Air Toxics
• Carbon Dioxide
Figure 1-4. Schematic of multi-pollutant emissions from a vehicle.1
1 Note that chlorofluorocarbons (CFCs) in air conditioning systems only exist in automobiles manufactured prior to
1994. CFCs have been phased out since 1995 and replaced by hydrochlorofluorocarbons that do less damage to
stratospheric ozone.
1-5
-------
More recent efforts to reduce evaporative emissions include capturing vented vapors
from within the vehicle using carbon canisters and reducing refueling losses through the use of
tank filler necks that capture some of these evaporative emissions.
Life-cycle emissions (not shown in Figure 1-4) are produced in activities associated with
the manufacturing, maintenance, and disposal of automobiles and include emissions from
manufacturing plant power usage, volatile solvents used in the manufacturing process, and
outgassing of synthetic materials used to reduce weight and simplify manufacturing.
Emissions controls often result in "co-control" across pollutants emitted by the source;
however, emissions of some pollutants may actually increase as a result of controlling another
pollutant. For example, reformulated gasoline was introduced in the mid-1990s to reduce ozone
concentrations by targeting reductions in emissions of VOCs from vehicles. HAPs such as
benzene and 1,3-butadiene were also reduced in ambient air by reformulated gasoline, but
formaldehyde likely increased. Recognizing multi-pollutant benefits and disbenefits from
specific control technologies or programs allows a more complete characterization of pollutant
releases to the environment and the potential human health and ecosystem impacts. Therefore,
EPA's programs and policies can be developed to result in a potentially more effective and
efficient overall set of controls that address multiple air quality objectives.
Atmospheric Processes Create, Remove, and Transform Pollutants
After being released to the atmosphere, all primary and precursor emissions become part
of a common chemical and physical system. Pollutants in a single air mass experience the same
meteorological conditions. Some react chemically with one another, some may be acted upon by
common oxidants, some may be removed by common physical processes such as rain, and some
subsequently arrive at sensitive receptors as a mixture. Because of these shared processes,
changes in one pollutant can lead to changes in other pollutants.
Figure 1-5 is a simplified diagram that shows the potential interactions among emissions
that lead to ozone, fine particles, air toxics, regional haze, deposition to ecosystems, and climate
change. Source emissions are characterized as either directly emitted, such as benzene or
primary particles, or as precursors, such as NOx and VOC emissions that combine with sunlight
to form ozone. Particles are multi-pollutant in composition because they contain both direct
(e.g., carbon, metals) and secondarily formed components (e.g., sulfates, nitrates, carbon). In
some regions, organic compounds, including secondary organic aerosols, are important
contributors to ambient fine particle concentrations. Some air toxics are primarily gaseous, some
are particles, and others are semi-volatile. Some gaseous air toxics (e.g., benzene, 1,3-butadiene)
react in the atmosphere and are precursors to ozone formation.
1-6
-------
Direct Emissions
Precursor Emissions
Particulate
Matter
Ozone
Air Toxics
Regional Haze
Aquatic/
Terrestrial
Deposition
Climate
Change
Figure 1-5. Simplified diagram illustrating relationships among direct and
precursor emissions that lead to ozone, fine particles, and air toxics formation.
Figure 1-6 illustrates some of the relationships among direct and precursor emissions in
more detail than the simplified diagram in Figure 1-5. Primary emissions (in blue) are
distinguished from those formed secondarily (in red) via atmospheric reactions. The hydroxyl
radical (OH) and ozone play key roles in many of the reaction pathways. Further information on
these atmospheric relationships can be found in textbooks (e.g., Seinfeld and Pandis, 1998).
Atmospheric model development continues to improve the treatment of meteorology, emissions,
atmospheric chemistry processes, and relationships across pollutants to better evaluate emission
reduction programs (Community Modeling & Analysis System Center, 2007a, b).
Exposure Pathways and Risks Are Affected by Multiple Pollutants
Exposure to airborne pollutants through inhalation, skin, and ingestion pathways is
largely a multi-pollutant process because air parcels contain mixtures of pollutants. Not only do
these pollutants interact in the atmosphere (as described earlier), these interactions also affect
human health and ecosystems. Each breath of air contains a mixture of fine particles and gases
that penetrate the lungs. Epidemiological and toxicological studies showing the effects of these
pollutants typically attempt to isolate the effects of specific pollutants such as fine particles or
ozone to determine direct associations of specific health effects and to account for the correlation
among pollutants in the overall mixture.
1-7
-------
Emissions Sources
VOC (HAPs)
CO
HAP
me+als
SVOC
particles > t
Other
primary
PM
Primary <&
secondary
Organic C
Sulfate
Nitrate
Chemical
Deposition
Figure 1-6. Links illustrating chemical relationships among CAPs and HAPs
including mercury, as well as connections across sources, secondarily formed
species, gases, primary particles, and deposition.
Note: hv represents sunlight, HO2 = hydroperoxyl radical, and RO2 = organic peroxyl radicals,
where R symbolizes any number of organic chemical groups
Scientific studies have shown that an increase in fine particle pollution leads to an
increase in respiratory problems such as asthma attacks (in asthma sufferers) and bronchitis, as
well as emergency room admissions and hospitalization for respiratory diseases. Increased
mortality has also been linked to higher particle concentrations; furthermore, people who breathe
high concentrations of fine particles for long periods are more likely to die prematurely. These
health impacts depend on the size and, potentially, the composition of particles; size determines a
particle's ability to penetrate the lungs, and composition determines a particle's toxicity once
deposited in the lungs.
Ozone can affect human health in similar ways. Ozone (and other oxidants) can
penetrate the lungs and cause a number of respiratory problems including increases in asthma
attacks, respiratory symptoms, and emergency room and hospital admissions. When ozone
concentrations are elevated, similar to particle concentrations, mortality rates increase.
Additional research is needed to understand possible synergistic or antagonistic effects
among pollutants in the mixture and the role of pollutants in the mixture as catalysts or carriers
for other pollutants. For example, sulfate particles tend to absorb water and can thus "carry"
other types of particles in the overall mixture (e.g., metals and polycyclic aromatic hydrocarbons
[PAHs]), potentially providing a larger dose of these metals and PAHs into the lungs than
previously thought.
1-8
-------
Pollutants Affect Ecosystems and Other Environmental Concerns
The ambient mix and chemical/physical properties of pollutants can affect ecosystems
and other environmental concerns. For example, ozone attacks biological molecules such as
terpenes from biogenic emissions. As plant leaves absorb ozone, cells within the plants that
control photosynthesis are damaged, potentially leading to reduced plant growth and root
development. Airborne pollutants can also deposit on surfaces, and some pollutants cause
eutrophication in water bodies.
The ability to quantify pollutant interactions depends on whether their effects are
additive, synergistic, or antagonistic. As shown in Table 1-1, a single air pollutant category
(e.g., NOx) can interact with other pollutants in the atmosphere (e.g., ozone, fine particles) and
consequently have multiple effects on health, the environment, and climate.
Table 1-1. Common emissions precursors and their ability to impact human
health, the environment, and climate (independent of relative magnitude or
direction of effect).
so2
NOx
nh3
VOC
CO
Primary
PM+BC
CH4
co2+
GHGs
Health impacts
- direct (VOC HAPs)
•
- direct (criteria)
•
•
•
- indirect (03 and particle
formation)
•
•
•
•
•
•
Ecosystems
- Acidification
•
•
•
- Eutrophication
•
•
- 03 vegetation
•
•
•
•
Radiative forcing
- direct (gas)
•
•
- direct and indirect via 03 and
aerosols
Acidification: the process whereby air pollution, mainly NH3, S02, NOx, is converted into acid substances.
Acid deposition is best known for the damage it causes to forests and lakes, but it also damages
freshwater and coastal ecosystems, soils, and historical monuments.
Eutrophication: the process whereby water bodies receive excess nutrients, such as nitrogen, that stimulate
excessive plant growth.
Radiative forcing: Direct - increase in temperature induced by GHGs; Indirect - change in cloud formation
process associated with particles about which cloud droplets coalesce
S02=sulfur dioxide, NH3=ammonia, PM=particulate matter, BC=black carbon (a constituent of particles),
CH4=methane, C02=carbon dioxide, GHGs=greenhouse gases, 03=ozone
1-9
-------
Using Multi-pollutant Concepts for Air Quality Management
The US air quality management system involves setting goals, collecting and analyzing
air quality and emissions data, developing and implementing emissions control strategies, and
evaluating progress. Figure 1-7 shows how each major amendment to the CAA increasingly
altered each of the major steps in the air quality management process as part of the continual
evolution of EPA's system. As noted by Bachmann (2007), "over time, the system itself has
evolved through legislation and policy to address problems in achieving results, advances in
scientific and technical understanding, and changing socioeconomic and political conditions."
The NAS report and the Clean Air Act Advisory Committee's (CAAAC) Air Quality
Management (AQM) Subcommittee Phase I and II reports identify challenges in the current air
quality management system and call for innovations that will improve air quality with greater
efficiency and effectiveness. Based on these recommendations, the next management challenge
is to incorporate multi-pollutant environmental decision-making into this system through
improved understanding of the interconnected technical and scientific elements, including
environmental data and modeling tools, to appropriately inform each step in Figure 1-7.
Consideration will be given to how multi-pollutant tools can be integrated into the system by
identifying the limitations of the current air quality management system, acknowledging changes
needed to make the process more efficient and effective, and assessing how these tools can lead
to more informed decision-making. For more information, see the textbox on "Multi-pollutant
Air Quality Management Plan Project at OAQPS" on page 1-11 and
http: //epa. gov/air/caaac/aq m. html.
The interdependencies among pollutants are now being incorporated into the design of
emission reduction programs. For example, a strategy focused on reducing ozone typically
considers some combination of NOx and VOC emission reductions. VOC reductions can then
reduce HAP exposures of benzene, 1,3-butadiene, toluene, and xylenes and may also reduce
concentrations of secondary HAPs such as formaldehyde, acetaldehyde, and acrolein. In
addition, VOC reductions may also have some limited effects on reducing secondarily formed
fine particle mass from the transformation of larger aromatic compounds such as toluene,
xylenes, and ethylbenzene. As discussed previously, reductions in NOx might result in
reductions in secondarily formed ammonium nitrate or nitric acid.
1-10
-------
Monitoring
Inventories
Analysis & Modeling
PAMS Network
Photochemical Grid Modeling
NAAQS
Visibility Goal
PSD increments
Acid Rain SOx Cap
Establish
Goals
Determine
Emissions
Reductions
Track and
Evaluate Results
Attain 3-5 years
Attain 3, 5-10 years
Attain, 5-10 (+2) years or
3, 6, 9,15-17, 20 years (03)
Rule Effectiveness
CEMs
Tools:
• Monitoring (air quality,
visibility, deposition,
emissions)
• Receptor modeling
Scientific Research
Develop Programs
to Achieve
Implement and
SIPs, Local & Tribal Plans
Legislative Mobile Reductions
Federal Fuel Additive Rules
NSPS
NESHAP (Risk)
Subpart D in Nonattainment Areas
• RACT and New Source Review
• Reasonable Progress
• Mobile l/M
Utility NSPS % Reduction
Federal Acid Rain Program
Federal MACT Plus Risk
Tier2, Toxics Mobile Standards
03/PM Area Classfication
Requirements: VOC, NOx RACT,
Federal CTG, l/M, 3%/yr Progress,
New source offsets,
03 Transport Commission
Figure 1-7. Evolution of EPA's air quality management system.
Source: Bachmann, 2007. PSD=prevention of significant deterioration, PAMS=Photochemicai
Assessment Monitoring Stations, CEMs=continuous emissions monitors, NSR=new source review,
NSPS=new source performance standards, NESHAP=Nationai Emissions Standards for
Hazardous Air Pollutants, RACTreasonably available control technology, l/M=inspection and
maintenance, MACT=maximum achievable control technology, CTG=control technique guideline
1970 Clean Air Act Amendments
1977 Clean Air Act Amendments
1990 Clean Air Act Amendments
Sources comply
Permits
Enforcement
NSR Permits
Title V permit programs
Multi-pollutant Air Quality Management Plan Project at OAQPS
EPA is currently working with three areas (Illinois/Missouri [St. Louis], New York, and North Carolina) to integrate non-
traditional planning into air quality management. Many state, local, and tribal governments are moving away from single-
pollutant planning to developing multi-pollutant strategies that address future air quality needs. EPA's "Air Quality
Management Plan Project" encourages state and local governments to create comprehensive air quality plans. Air
Quality Management Plans (AQMPs) address air quality concerns such as attainment and maintenance of criteria
pollutant standards, sector-based emission reductions, improvements in regional haze and visibility, and risk reductions
of HAPs. These plans may include other air quality concerns such as land use, transportation, energy, and climate
change. The goal is to integrate the requirements of the current SIP process into a more comprehensive plan for air
quality in a manner consistent with the 2004 NAS report, "Air Quality Management in the United States," and the 2007
CAAAC recommendations.
To explore some of the technical challenges of implementing an AQMP, EPA is undertaking the Detroit Multi-pollutant
Pilot Project. This project will investigate the methods, tools, and models available to the state, local, and tribal agencies
in developing AQMPs. In particular, the project will explore the local- and regional-scale nature of certain CAPs and
HAPs, and the best tools to use when considering their impacts. The project will also consider the multi-pollutant
impacts of selecting control strategies that will control key HAPS, as well as ozone and fine particles. Using Detroit as
the example urban area, the project seeks to demonstrate an approach for considering multiple pollutants in an
integrated manner for air quality planning. A report will be provided that discusses the tools, data, and methods used;
the approach implemented; and the project results and conclusions. This report will be useful in informing EPA guidance
developed for consideration of multi-pollutant air quality assessments.
1-11
-------
Table 1-2 summarizes how emission reduction programs can lead to multiple changes in
pollutant or atmospheric concerns. Though reductions in precursor pollutants generally lead to
improvements in air quality and other atmospheric concerns, the design of a multi-pollutant
strategy still requires careful consideration of multiple consequences of precursor emission
controls. For example, reducing NOx emissions can result in significant decreases in fine
particles, ozone, nitrate, acid deposition, and watershed eutrophication, and improvements in
visibility. However, reducing NOx emissions can increase mercury deposition/methylation in
sediment. NOx emission reductions can also cause ozone in certain places to increase due to the
NOx titration effect on ozone in VOC-limited areas of the country. Similarly, under certain
conditions, reductions in VOCs, while leading to improvements in ozone, air toxics, and
watershed eutrophication, can lead to particulate nitrate increases caused by the reduction of
peroxyacetyl radicals that would cause more nitric acid to form and be available for atmospheric
conversion to particulate nitrate.
Spatial and temporal patterns among pollutants must also be considered in the design of a
multi-pollutant emissions reduction strategy. Spatial and temporal scales for environmental
decision-making depend on the pollutant of concern and range from global (i.e., long-range
transport) to local scales. For example, Figure 1-8 illustrates the potential need to consider
contributions to fine particles (and their chemical constituents) from all spatial scales when
developing emission reduction strategies. As domestic concentration levels of secondarily
formed pollutants decrease across the US through the reduction of precursors, the relative
contribution of pollutants from international transport becomes more important. Similarly, as
progress is made in "regional" air pollution levels, attention can turn to more localized pollution
sources (e.g., manufacturing or near-roadway emissions). The diurnal and seasonal patterns in
pollutant concentrations, which are a function of the source of pollution, meteorology, and
formation and removal processes, may also need to be taken into account when designing a
multi-pollutant control strategy.
8
c
$
ai
5
-E
8.
--I
L
10=
10"
104
102
Global
w
r CFC. HCFCu GHG
Oontinerflal
Hn. N Ox, PM2 s. 0.H
w
Regional
PM „j. PW2 & 0.3, VOC. NH.j
Local
Duet. PM10< Ultrafine PM
Z>
£-¦
ar
O
1-j
-s
I
&
OJ
Figure
Source:
1-8. Illustrative transport scales for PM and other atmospheric pollutants.
NARSTO, 2004c
1-12
-------
Table 1-2. Possible pollutant/atmospheric relationships associated with emission precursor reductions (NARSTO, 2004b).
Reduction in
pollutant
emissions
Ozone
Change in Associated Pollutant or Atmospheric Issue
PM Composition
Sulfate
Nitrate
Organic
Carbon
PM,
Visibility
HAP
VOCs
HAP
Metals
Acid
Deposition
Watershed
Eutrophication
Hg - dep/
methylation
SO,
I
J
i
J
4s
NOx
I
IT
I
I
I
J
I
voc
HAPs
I
tfid
I
I
CO
NH,
J
I
I
I
Primary
PM-organic C
I
J
Primary
PM-black C
I
I
Primary PM
(crustal/metals)
I
I
Mercury
I
a - Arrow direction denotes relative increase or decrease j of pollutant resulting from a decrease in associated emissions. Large arrow
indicates either well established relationship and/or substantial magnitude of effect. Small arrow implies possible response that is likely
to be of minimal magnitude,
b - Ozone reduction associated with decreased light scattering from decrease in fine particle levels,
c - NOx titration effect on ozone largely limited to VOC-limited urban areas,
d - Associated with effect on decreasing OH and ozone levels.
e - Substitution effect in competition for NH3 in NH3-limited regions (and increase in hydrogen peroxide leading to increased in-cloud S02
production).
f - Associated with reduction of peroxyacetyl radicals leading to increased nitric acid formation.
g - Associated with nitrogen, sulfur, and mercury interactions within sediments.
-------
Summary
The technical and scientific basis for integrating multiple air quality issues to inform
EPA's air quality management system will continue to evolve to better account for the
interdependencies shown in Table 1-2. As we progress, a variety of implications arise for air
program management and the direction of policy. As suggested by the NAS report (and agreed
upon by CAAAC), a truly integrated air quality management framework would maximize health
and environmental benefits using comprehensive emission reduction control strategies. It would
streamline the various programs and their associated requirements across different source
sectors, potentially reducing the cost of achieving health and environmental goals. We will need
to integrate multi-pollutant concepts in the management of air quality problems. A multi-
pollutant technical infrastructure (i.e., data, tools, and models) will need to be built so that more
effective and efficient environmental solutions can be achieved. More details on this
infrastructure are given in Chapter 5. However, moving to a multi-pollutant air quality
management approach will require changing the way we currently solve problems and interact
with one another, and will take considerable time, effort, and support.
This report is a first step in exploring the nature of historical, current, and future multi-
pollutant air quality issues across the US. The next chapter introduces the "current"
characterization of these multi-pollutant issues by identifying areas with multiple problems, such
as areas where ozone and fine particles are above the NAAQS and where air toxics pose high
risks to populations.
1-14
-------
Chapter 2: Multi-pollutant Air Quality Issues at the National Level
This chapter explores the geographic nature of multi-pollutant air quality issues across
the US. The conceptual approach used here to examine multi-pollutant issues is simplified to
focus on only three select pollutant types: ozone, fine particles, and air toxics. This
simplification allows concise articulation and visual representation of the relationships among
the three primary air quality issues affecting human health. Future assessments that extend this
initial investigation to include visibility impairment, nutrient and acid deposition, and secondary
air quality standards will result in a broadening of the identified pollutant relationships to rural
areas. In addition, further extension of this analysis to include such considerations as the
recently revised 24-hour particle standard and more stringent standards for other pollutants (e.g.,
ozone) will bring even more urban areas into this spatial overlap of pollutants (i.e., "nexus").
Identifying Areas with Multiple Air Quality Problems
In this section, national-level spatial relationships between ozone and fine particles are
examined first. Air toxics concentrations at the national level are then examined using National
Air Toxics Assessment (NATA) 1999 results (U.S. Environmental Protection Agency, 2006a).
Finally, the nexus of air toxics, ozone, and particles is shown at the national level, and the
correlations among these multi-pollutant issues are discussed.
Ozone and Particle Pollution
The counties in which ozone and PM2.5 levels exceed current air quality standards are
shown in Figure 2-1. Counties that exceed the ozone standard are outlined in blue; counties that
exceed the PM2.5 annual standard are outlined in red; and counties that exceed standards of both
pollutants are outlined in purple. The areas with the highest ozone and/or particle pollution
concentrations are primarily in, or downwind of, heavily populated urban areas in California,
Texas, the industrial Midwest, the Southeast, and the Northeast. Counties with high
concentrations of ozone or fine particles are more likely to be adjacent to counties with similar
problems.
2-1
-------
o
B »
^ ,o 2 o ^
~ Ozone and PM?s above NAAQS
~ Ozone above NAAQS
~ PM2sabove NAAQS
Monitored; below NAAQS
Not Monitored
/ LJJ a
«
%
Figure 2-1. Counties with ozone and/or fine particle design-value concentrations
above the NAAQS for 2003-2005.
Hazardous Air Pollutants
The ambient air quality monitoring network for air toxics is sparse compared with the
ozone and fine particle networks and does not provide sufficient measurement data to
comprehensively estimate toxics risk nationwide. The best currently available characterization
of the national air toxics picture is NATA, which is the source of cancer risk estimates in this
report.2 Released in February 2006, the 1999 NATA results offer a snapshot of air quality and
the health risks from air toxics resulting from estimated 1999 emissions (U.S. Environmental
Protection Agency, 2006a). This assessment covers 177 of the 187 listed air toxics plus diesel
particulate matter (DPM). The risks estimated in the assessment are associated with inhaling the
pollutants; inhalation is the most significant route of exposure for the majority of air toxics.
J Although the 1999 NATA includes estimates for respiratoiy and neurological non-cancer effects from air toxics,
these results are not used in this report's characterization of air toxics for purposes of comparing problem areas for
toxics with those for ozone and particles. Future multi-pollutant assessments may include these non-cancer impacts.
2-2
-------
Figure 2-2 shows county-level nationwide health risk estimates for air toxics for 1999. In most
of the country, the lifetime cancer risk from air toxics is less than 25 in a million. This means
that out of one million people, fewer than 25 are likely to develop cancer as a result of inhaling
air toxics from outdoor sources if they are exposed to 1999 levels over the course of their
lifetime. Most urban locations have an air toxics lifetime cancer risk greater than 25 in a million,
while a few counties have a cancer risk greater than 50 in a million.
Modeled Cancer Risk
(per million people)
7-10
11-25
26 - 50
IB 51'75
76 - 136
Figure 2-2. Modeled cancer risk per million people by county from the 1999
National Air Toxics Assessment.
Where Multi-pollutant Air Quality Issues Occur
Figure 2-3 demonstrates the nexus of health-related air quality issues by mapping the
spatial relationships of ozone, fine particles, and air toxics. The figure is based on the model
results for cancer risk from the 1999 NATA study in combination with ambient ozone and fine
particle air quality concentration data. As in Figure 2-1, high ozone and fine particle pollution
areas are defined as those experiencing concentrations greater than their respective NAAQS for
the period 2003-2005. Areas with air toxics problems are defined as those with estimated cancer
risks higher than the 90th percentile of the 1999 NATA model results for all counties (i.e., greater
than 35 in a million), representing counties that include 55 percent of the total population
(approximately 163 million people). The map aggregates county-level data to the metropolitan
level, i.e., the combined statistical area (CSA) level or core-based statistical area (CBSA) level
where possible, based on the maximum county-level values in each area. County-level data are
2-3
-------
shown where aggregation to CSA or CBSA levels is not possible (i.e., where a county is not in a
CSAorCBSA).
Ozone
Ozorte &
PMIS/
Toxics
Figure 2-3. Nexus of fine particles, ozone, and air toxics.
Note: Nexus is defined as areas with ozone and/or fine particle concentrations above the NAAQS
for 2003-2005 and/or with modeled county-level cancer risk estimates from NAT A 1999 in the top
10 percent for all counties. Data are aggregated to the CSA or CBSA levels where possible based
on the maximum county-level values; otherwise, county-level data are shown.
Figure 2-3 shows that some metropolitan areas experience high ozone, particle pollution,
and air toxics cancer risk. These areas include southern and central California, urban areas in the
industrial Midwest and the Northeast corridor, as well as some parts of the Southeast (e.g.,
Atlanta, Charlotte). The nexus of high ozone, particle pollution, and air toxics cancer risk
typically occurs in urban areas. Notably, a number of areas where residents are predicted to have
relatively high cancer risk associated with air toxics do not have ozone or particle pollution
problems. While urban areas are most likely to have all three issues, some niral areas have high
concentrations of at least one of the three air quality issues.
2-4
-------
Correlations Among Multi-pollutant Issues
Figure 2-4 shows the distribution of air toxics cancer risk by the presence or absence of
ozone and particle pollution by county. Overall, people living in counties with high ozone and
particle pollution levels have higher average cancer risk from air toxics than people living in
counties with lower ozone or particle pollution concentrations. However, the highest estimated
cancer risks from 1999 NATA (i.e., average county excess cancer risk levels at or above 35 in a
million) can occur across any categories of ozone and particle combinations.
150
c
o
E 100-
0
a.
CO
'on
S_
CD
o
c
CO
O
Figure 2-4. Distribution of air toxics cancer risk as a function of ozone and
particle problems.
Note: Counties with ozone and fine particle problems (as defined in Figure 2-3) are more likely to
have air toxics problems as well. The plot shows modeled cancer risk distributions for residents of
counties with different types of air quality problems. The box shows the extent of the 25th and 75th
percentiles, the notch is the median, the whiskers extend to 1,5*interquartile range (IQR), and
individual outliers beyond this are shown as asterisks or circles.
As shown in Figure 2-5, many counties experience concentrations very close to (above
or below) both the ozone and fine particle standards. Pollutant concentrations exist along a
continuum, ranging from relatively clean air in the bottom left quadrant to relatively polluted air
in the upper right quadrant. Differences in county concentrations are due to variations in
population, emissions, meteorology, topography, and transport. For example, many counties
downwind of large urban areas are seen in the bottom right quadrant, while those near the Pacific
Ocean and those isolated from emissions are located in the bottom left quadrant. Southern and
central California have some of the highest concentrations because of high emissions, mountains
that trap air, and meteorological conditions conducive to ozone and particle formation. In
2-5
-------
addition, this figure also reflects the degree of the air toxics probl em in those counties that
measure ozone and particles. Higher air toxics risk values are associated with areas in which
ozone and particle concentrations are also high, consistent with the results shown in Figure 2-4,
although risk values vary significantly among counties in each of the four quadrants.
30
1
Cancer risk (NATA 1999)
25
s
Ol
0
<10
10-40
0
40-70
70-100
0
>100
cu
"O
c
.2
CO
CD
c
o
N
o
20
E
-g15
8
a
oS
8
O
Russell, AL;
Fayette, KY; and
Hancock, WV
OQ3 b
rP O
3 Cl-
io
Hawaii. Oregon.
Coastal N. California.
N. Dakota, and
1
Nebraska \
fWfljrrtt
eg Yjma.c
PM2.5 standard
(15M9/m3)
L
1
and Oce
NJ _
25
50
75
O, 2003-2005 design value (ppb)
100
125
150
Figure 2-5. Scatter plot of county-level maximum 8-hour ozone and PM2.5
concentrations from 2003 to 2005 color-coded by cancer risk level estimates
from NATA 1999.
Summary
In this chapter, the development of a spatial nexus of particle pollution, ozone, and air
toxics was discussed. Particle and ozone problems were defined based on their NAAQS, and air
toxics were represented at the national level by the upper end of cancer risk estimates from
NATA 1999. The development of this nexus revealed some interesting correlations between
measured particle and ozone concentrations and modeled air toxics risk. While this chapter
focused on a national summary, the next chapter looks more closely at Detroit, MI, an area that
has a nexus of PM, ozone, and air toxics problems.
2-6
-------
Chapter 3: Multi-pollutant Air Quality at the Local Level, an Example
As discussed in the previous chapter and shown in Figure 3-1, one metropolitan area that
has a nexus of fine particle, ozone, and air toxics air quality issues is Detroit, MI. An essential
first step in developing a comprehensive plan to address such multi-pollutant air quality issues is
to establish a conceptual model for the area of interest. A conceptual model is based on available
technical data and analyses for an area, and assists air quality planners in determining which
control programs would be most beneficial for reducing the pollutants of most concern (see the
textbox "Value of Conceptual Models for Multi-pollutant Air Quality Management"). Detroit is
rich in technical data; several special studies have been conducted for this area.3 This chapter
discusses the technical data and analyses that outline a conceptual model that could be used to
inform the development of effective multi-pollutant control strategies for the Detroit area.
Toxics &
Ozone
Ozone
Ozone
PM„
oxics &
PM,<
Toxics
Figure 3-1. Multi-pollutant nexus of air quality issues for southeastern Michigan.
3 These studies include, but are not limited to, the Detroit Air Toxics Initiative (Michigan Department of Environmental Quality,
2007a), the Detroit Exposure and Risk Assessment Study (U.S. Environmental Protection Agency, 2007b), LADCO technical
analyses (LADCO, 2007a, b), and recent literature studies (Rizzo, 2005; Trepat et al., 2007).
3-1
-------
Value of Conceptual Models for Multi-pollutant Air Quality Management
The conceptual model for an area is based on the application of technical data, tools, and models that
reflect our best understanding of that area's atmospheric environment (as shown in the figure below). It
compiles and summarizes the most advanced science to inform the development of control strategies to
meet an area's air quality goals. In fact, areas are expected to develop a conceptual model as part of their
State Implementation Plan (SIP) for criteria pollutants such as ozone and fine particles. Chapter 11 of
EPA's Guidance on the Use of Models and other Analyses for Demonstrating Attainment of Air Quality
Goals for Ozone, PM2.5, and Regional Haze is devoted to conceptual models and walks through questions
that should be answered about a specific area to have a full understanding of that area's air quality
problems and thereby determine the best approach for developing and evaluating control strategies to
achieve attainment. For details on conceptual models in practice, Chapter 10 of the 2004 NARSTO report
on fine particles provides conceptual model descriptions for nine North American regions including the upper
Midwest/Great Lakes region (NARSTO, 2004a).
Air quality management efforts that extend the current single-pollutant conceptual models to reflect multiple
pollutants can be challenging. However, as illustrated in this chapter, air quality management solutions to
multiple pollutant problems can be attained by integrating multiple technical data sets, tools, and models that
adequately reflect the region's atmospheric environment and sources. These technical components are
essential to the development of conceptual models, which in turn guide the selection of air quality models
and technical assessments needed to design and implement effective control strategies for multi-pollutant
air quality planning.
The
Atmospheric
Environment
Atmospheric
Processing
Chemical
Physical
Meteorology
Atmospheric
conditions
Transport
Gas/Particle
Man made/
Natural
Emissions
Atmospheric
Concentrations
Spatial variability
Temporal variability
Chemical
composition
Size
¦=>
Atmospheric
Science
Analyses
Air quality
modeling
Receptor
modeling
Emission
characterization
Ambient air
characterization
(measurements)
Conceptual
Model
Development of a conceptual model is based on technical data and analysis reflecting the best
understanding of the nature of an area's air quality issues.
Source: NARSTO, 2004a
3-2
-------
Ambient Air Characterization
Development of a conceptual model for Detroit begins with a characterization of the
existing ambient monitoring data. Figure 3-2 shows the monitoring network for southeastern
Michigan for multiple pollutants. Table 3-1 summarizes the highest monitored values of criteria
air pollutants in the Detroit area for 2005 as outlined in the Michigan Department of
Environmental Quality (MDEQ) annual report (MDEQ, 2007b) and indicates the average cancer
risk for Wayne County as predicted by NATA 1999. Note that 8-hour ozone and both the annual
and daily fine particle concentrations exceed the NAAQS, and the overall cancer risk of 63.2 is
in the top 90®" percentile of risk across US counties.
Warren
Detroit - E. Seven Mile
~
Livonia
Detroit
Detroit - Linwood
Dearborn
Detroit Metropolitan. Wayne County
Locater Map
Detroit - Newberry
^ Detroit - W, Lafayette
Detroit - W. Fort
' Detroit - W. Jefferson
~
River Rouge
Figure 3-2. Michigan Air Sampling Network for southeastern Michigan for 2005.
3-3
-------
Table 3-1. Summary of maximum pollutant levels in the Detroit urban area in
2005 (MDEQ, 2007b).
Pollutant
Duration and Unit
Highest
Value
NAAQS
CO
8-hour (ppm)
2.6
9.0
Lead
24-hour (|ig/m3)
0.117
NA
N02
Annual (ppm)
0.017
0.053
Ozone
1-hour (ppm)
0.118
0.12
8-hour (ppm)
0.103
0.08
pm2,
Annual (|ig/m3)
18.6
15
24-hour (|ig/m3)
79
35
PM10
Annual (|ig/m3)
40
revoked
24-hour (|ig/m3)
95
150
S02
Annual (ppm)
0.007
0.03
24-hour (ppm)
0.045
0.14
Air Toxics
NATA 1999 Model
(cancer risk per
million)
63.2
NA
NA = not applicable. Shading indicates concentrations above the NAAQS for ozone and
PM2 5 and the relatively high risk estimated by NATA 1999.
Criteria Pollutants
As shown in Figure 3-2 and Table 3-1, several monitoring sites in the Detroit area
measure pollutants for which NAAQS exist, such as CO, ozone, lead, NOx, SO2, and particles.
While the Detroit area did not exceed the NAAQS standard for several pollutants (e.g., CO, NO2,
SO2), there were measured exceedances of the 8-hour ozone NAAQS and the 24-hour and annual
fine particle NAAQS.
Fine Particles
To understand more about the sources of the high fine particle concentrations, speciation-
monitored data at the Allen Park and Dearborn sites are explored. Fine particles consist of
multiple chemical constituents and are usually speciated into sulfate, nitrate, ammonium, organic
carbon, elemental carbon, and crustal components. While sulfate, nitrate, and ammonium are
3-4
-------
mostly secondarily formed, elemental carbon and the crustal components are primary species
usually deposited directly from the source. In Figure 3-3, annual average concentrations for
2005 of the major chemical components of fine particles are shown for the Allen Park and
Dearborn sites as percentage of fine particle mass. The annual average fine particle mass in
2005 measured at Allen Park was about 15 |ag/m3, while at Dearborn it was about 18 |ag/m3.
When comparing the chemical species contributions with the mass between these two sites, the
amount of sulfates, nitrates, and carbon is reasonably consistent on a percentage basis. However,
crustal material at the Dearborn site is a much higher contributor to mass—about 11 percent
compared with less than 0.6 percent at Allen Park. The relatively high amount of crustal
material at the Dearborn site indicates very local contributions of crustal material (or the metal
oxides associated with crustal material). Causes of site-to-site differences in species components
of fine particles for any area should be taken into account when considering sources to include as
part of potential control strategies.
Allen Park Dearborn
~ Sulfates ¦ Nitrates ¦ Elemental Carbon ¦ Crustal ~ Organic Carbon Mass
Figure 3-3. 2005 annual average concentrations (|jg/m3) of fine particle chemical
constituents at Allen Park and Dearborn sites.
Ozone
Of the seven monitors in the Detroit area, only Warren and New Haven have ozone levels
averaged over 2003-2005 that continue to exceed the ozone NAAQS of 0.085 ppm (Figure 3-4).
Because ozone is not emitted but is formed from the chemical interactions of other pollutants, it
is called a "regional" pollutant. Because of this, it becomes important to understand the
atmosphere's responsiveness to the reduction of ozone precursors, such as NOx and VOCs. This
is usually done through sensitivity analyses with atmospheric models. An area is called "NOx-
limited" when reducing NOx will lead to decreases in ozone. In such an environment, the most
effective control strategy will focus on controlling NOx emissions. When the opposite is true,
the area is called "VOC-limited." Detroit is a "VOC-limited" area, with the limiting factor in
ozone production being VOC concentrations. In the Detroit urban area, the most effective
control strategy would focus on VOC reductions.
3-5
-------
£
o.
CL
II
z
Q
5
ce
Lii
a
z
a
u
0.1 QO
C DBS
a 070 -¦
a o&s -¦
..""naaqs"^
a ow
New Haven Warren
Port Huron -*• Allen Park
E. Seven P-li le
Oak Park
Linwood
H 1 1 h
H h
H h
13K 1507 1333 1939 2000 20Q1 2(102 2003 20&i 2MS
Figure 3-4. Trends in ozone levels for monitoring sites in southeast Michigan for
1996-2005.
Source: Michigan Department of Environmental Quality, 2007b
Air Toxics
While there is a large amount of air quality monitoring data in Detroit for NAAQS
pollutants, the amount of air toxics measurements are limited. To address this issue, the MDEQ
undertook an intensive air quality sampling program to measure levels of over 200 air toxics in
the Detroit area. This study, called the Detroit Air Toxics Initiative (DATI) Study, was
conducted from April 2001 to April 2002 at six locations within Wayne County and one location
in Southfield. The location of these monitors is shown in Figure 3-5. A monitor was also
placed in Ypsilanti for comparison purposes.
Based on the measurements taken during the DATI study, MDEQ released the DATI
Report which detailed the study and summarized findings. In this report, the air toxics found to
be risk drivers for both cancer and non-cancer in Detroit are shown in Figures 3-6 and 3-7.
Thirteen air toxics were identified as contributing the most to risks in the Detroit area. These
were 1,4-dichlorobenzene, acrylonitrile, arsenic, benzene, formaldehyde, methylene chloride,
naphthalene, manganese, nickel, cadmium, carbon tetrachloride, acetaldehyde, and 1,3-butadiene
(not in order of priority). Two other pollutants, acrolein4 and DPM, were added to the list as
important air toxics based on additional data.
4 Acrolein information is based on actual monitored data in Detroit as part of EPA's Detroit Exposure Aerosol Research Study
(DEARS).
3-6
-------
Southfield
NE Detroit
N. Delray
S. Delray
River Rouge
Ypsilanti
Figure 3-5. DATI monitoring sites for air toxics in the Detroit area.
Source: MDEQ DATI Report, 2005
South Delray Allen Park North Delray River Rouge Southfield
~ Naphthalene
E9 Nickel (TSP)
0 Methylene Chloride
~ Formaldehyde
H Cartoon Tetrachloride
SCadmium (TSP)
DD Benzene
¦ Arsenic (TSP)
0 Aery Ion itrile
Co Acetaldehyde
B M-Dichlorobenzene
Deaftom ~ 1,3-Butadiene
Figure 3-6. Additive cancer risk by monitoring site in the Detroit area for 2002.
Source: MDEQ DATI Report, 2005
3-7
-------
Figure 3-7. Non-cancer hazard quotient by site and chemical in the Detroit area
for 2002.
Source: MDEQ DATI Report, 2005
Figure 3-6 shows that additive risks for the 12 carcinogens designated as having an
individual risk greater than one-in-a-million varied among the sites.5 Within the Detroit area,
there is about a five-fold difference between sites with the highest and lowest risk from these
carcinogens. If consideration is given to the fact that high concentrations of methylene chloride
at Allen Park and naphthalene and benzene at South Delray may have been isolated occurrences,
then formaldehyde seems to be one of the more important carcinogens across the Detroit area.
In Figure 3-7, the non-cancer benchmark is shown via hazard quotient (HQ) across
monitoring sites in the Detroit area. Monitored levels of six compounds were found to be
present at levels greater than one-tenth of the chronic health protective benchmark value,
indicating an HQ greater than 0.1. These six compounds were acetaldehyde, acetonitrile,
acrylonitrile, benzene, manganese, and naphthalene. Only two of these compounds, manganese
and naphthalene, exceeded their hazard quotient (HQ>1) at any of the sites.
While most of these air toxics were found at levels similar to those in other large,
industrialized urban areas of the US, concentrations of a few air toxics were particularly high at
some sites in the Detroit area. These include methylene chloride at Allen Park, benzene and
naphthalene at South Delray, and manganese at South Delray, North Delray, Dearborn, and
River Rouge. It is interesting to note the unusually high concentrations of manganese, a metal
air toxic and a directly emitted particle. Relating this information to the high crustal component
in the speciated fine particle data shown in Figure 3-3 suggests that a source of directly emitted
5 The Integrated Risk Information System (IRIS) cancer benchmark for formaldehyde used by MDEQ is not the one currently
recommended by OAQPS.
3-8
-------
fine particles could also be emitting manganese. Targeting such a source could provide multi-
pollutant benefits by reducing fine particle concentrations as well as reducing the risk from this
air toxic.
Source Contributions to Air Quality Problems
The relationship between ambient measured concentrations of pollutants and the sources
of these pollutants is important to understand when the release or control of one pollutant will
affect the release or chemical formation of another pollutant. Therefore, in developing an
effective multi-pollutant control strategy, the nature of these source-receptor relationships must
be understood. Figure 3-8 shows the location of monitoring sites in the Detroit metropolitan
area relative to local point and mobile emission sources. Point source (i.e., large facility)
emissions of NH3, NOx, SO2, and VOCs all contribute to fine particle concentrations, while
emissions of NOx and VOCs contribute to ozone, and many of the VOCs are also HAPs. The
annual average daily traffic volumes represent motor vehicle emissions (e.g., higher traffic
volume indicates higher emissions) and help to show the spatial distribution of motor vehicle
emissions of VOCs, NOx, and fine particles.
Macomb County
Oakland County
Detroit S96ILoa^
South Lyon-Howelt-Bnghton
Wayne[IGounty
Ann Arbor
Monitoring Sites
V Dearborn
Allen Part
Other Sites
Monroe
1 Port Huron
Point Source Emissions
PM2.5 (tonnes/year)
• 10
© 100
(3)1000
NOx (tonnes/year)
10
• 100
10.000
VOC (tonnes/year)
• 10
• 100
1.000
S02 (tonnes/year)
• 100
• 1.000
10.000
Annual Avg Daily Traffic
0 -10000
10001-20000
20001 - 50000
50001 - 100000
100001 - 175000
— 175001 - 220000
Figure 3-8. Map of monitoring sites and pollutant emissions in the Detroit vicinity
for 2001.
Source: Adapted from Brown et al, 2001
3-9
-------
From MDEQ's DATI report, Table 3-2 describes all emission sources, including point,
area, and mobile sources within four miles of each monitoring site. As shown, a variety of
sources are in the Detroit area and more than one monitor could be impacted by the same source
category. Figure 3-9 shows the breakdown of several major pollutants among primary source
categories from the 2002 National Emission Inventory (NEI, U.S. Environmental Protection
Agency, 2007b). These data indicate that large industrial sources in Detroit emit S02, fine
particles (including manganese), NOx, and VOCs. This figure also indicates that mobile sources
are a significant source of VOCs, especially benzene, formaldehyde, and 1,3-butadiene. If VOCs
emitted by mobile sources in Detroit were controlled to lower ozone, such a strategy could
provide multi-pollutant benefits by also reducing the high cancer risk from benzene,
formaldehyde, and 1,3-butadiene. Selecting multi-pollutant control strategies that result in
benefits across both CAPs and HAPs are important in an area such as Detroit.
Table 3-2. Significant emission sources within four miles of each monitoring site
in the Detroit area (MDEQ DATI Report, 2005).
Site (AIRS ID)
Point and Area Sources
Mobile Sources
Houghton Lake
(261130001)
Fireplaces/wood stoves, Christmas tree farming, oil and gas
production
US-127, boating,
snowmobiling
Southfield
(261250010)
Paint manufacturing, metal heat treating, machine shop, auto paint
shop, asphalt, ready-mixed concrete
1-696, Telegraph, and
Lodge
Ypsilanti
(261610008)
Equipment manufacturing, waste water treatment plant (WWTP),
commercial printing, plastic products, power generation plants
1-94
Allen Park
(261630001)
Bulk petroleum stations, refuse services, quarry, metal fabrication,
chemical manufacturing/processing, power generation plants,
plastic resin manufacturing
1-75
River Rouge
(261630005)
Steel plant, drywall manufacturing, WWTP, sewage incinerator,
asphalt plant, oil refinery, coke batteries, coke by-product
production facility, power generation plants, coal-and oil-fired
combustion, paint shops, assembly plants (heavy industrial)
1-75
N. Delray
(261630015)
Two steel mills, used oil reclamation plant, asphalt plant, oil
refinery, coke batteries, coke by-product production facility,
WWTP, sunroof manufacturer, power generation plants (heavy
industrial)
1-75
N.E. Detroit
(261630019)
Automotive manufacturing and stamping, chemical preparations,
power generation plants, foundry, metal coating, refuse systems
1-94
S. Delray
(261630027)
Coke battery, asphalt plant, oil refinery, coke by-product
production facility, steel mill, power generation plants (heavy
industrial)
1-75
Dearborn
(261630033)
Auto and steel manufacturing, power generation plants, asphalt
plant, oil refinery, coke batteries, coke by-product production
facility (heavy industrial)
Between 1-75 & 1-94
3-10
-------
| EG Us ~ Miscellaneous Residential wood combustion
~ Fossil fuel combustion ~ Nonroad ~ Solvent use
| Industrial processes ~ Onroad vehicles ~ Waste disposal
VOC
PM
PM
2.5
10
so2
NOx
Benzene
Formaldehyde
1,3-Butadiene
Chromium
Manganese
20 40 60
Percent of emissions
1
II
II
HI | _1 Ml
ii n
1 1
III
1
1 -1 1
II
ii i
Hi 1
¦ III
II 1
II
1 1 1
Ml 1 II HI
80
100
Figure 3-9. Emissions of criteria pollutants, their precursors, and key HAPs in
the Detroit area for 2002,
Source: U.S. Environmental Protection Agency, 2007b
Though the ambient monitoring data alone can give some indication of source emissions
causing these multi-pollutant air quality problems, these data can also be used with receptor
models to estimate individual source contributions to pollutants at a monitoring site. Receptor
models are statistical tools that identify the covariance in concentrations across species to isolate
"factors" that correspond to emissions sources or transported pollution. Currently, two
established receptor models are widely used for source apportionment: chemical mass balance
(CMB) and positive matrix factorization (PMF). Both have been used to characterize source
contributions to ambient fine particle levels (e.g., NARSTO, 2004a; Lee and Flopke, 2006;
Brown et al., 2007).
Table 3-3 shows the results of applying PMF to 2004 fine particle chemical constituent
data from the Allen Park and Dearborn sites. These data indicate that mobile emissions are
significant contributors to fine particle concentrations at both sites. This conclusion is also
supported by a recent study that the Lake Michigan Air Directors Consortium (LADCO)
undertook to investigate the organic carbon portion of fine particles in the Detroit area using
chemical markers that are identifiers for specific source categories. While the LADCO work is
still under development, early results also indicate that mobile sources (gas and diesel) contribute
heavily to ambient organic carbon in the Detroit area. It is interesting to note that recent EPA
modeling has shown that mobile sources (both nonroad and onroad) also play a large role in
ozone formation. These findings suggest the use of a multi-pollutant control strategy for the
Detroit local area that would allow reduction of fine parti cles and ozone through mobile source
controls.
3-11
-------
Table 3-3. PMF-based source contributions to air quality at the Dearborn and
Allen Park sites in Detroit for 2004 (Rizzo, 2005).
Source
Contribution in |ag/m3
(Percent of total fine particles is in parentheses)
Dearborn
Allen Park
Vehicles
5.3 (25%)
5.9 (35%)
Secondary Ammonium Nitrate6
3.7(18%)
3.5 (21%)
Secondary Ammonium Sulfate6
4.6 (22%)
5.0 (30%)
Vegetative Burning
0.9 (4%)
0.9 (5%)
Road Salt
0.8 (4%)
0.4 (2%)
Steel (Metals Processing)
1.1 (5%)
0.3 (2%)
Soil
1.4(7%)
0
Diesel source
1.3 (6%)
0.2(1.1%)
Industrial (Utility and Petroleum
Refineries)
1.7(8%)
0.7 (4%)
Tolal l-'ine Panicle Mass
2() S
10
In Table 3-3, source contributions from vehicles, secondary ammonium nitrate, and
secondary ammonium sulfate, are similar at the two sites. However, larger contributions of
sources such as steel and industrial to the Dearborn monitor suggest a significant local source
contribution. This is in agreement with the monitoring data discussed earlier in this chapter, for
which the amounts of ambient crustal material and manganese (a pollutant known to be emitted
from steel mills) were much higher at the Dearborn site. In this way, a receptor model can help
identify sources of fine particles in a local area and enable a better understanding of the sources
that contribute to the gradients seen in fine particle concentrations in an urban area. This
information allows informed decisions on potential control strategy selections at a given site and
aids in the implementation of an air quality management plan.
Receptor modeling can also be applied to identify and quantify sources that contribute to
multiple pollutant issues. Figure 3-10 illustrates a multi-pollutant PMF analysis at the Allen
Park site using collocated ambient air toxics and fine particle data. This figure shows that
mobile sources contribute to both fine particles and air toxics at the Allen Park site. Factors
identified in this work included direct emissions from diesel, mobile sources, a steel facility, an
industrial facility emitting carbonaceous aerosols, and crustal material. Other factors represent
transported and chemically aged pollutants, including the aged mobile sources factor, and two
regional transport factors representing winter and summer seasons, respectively. This type of
application of multi-pollutant receptor modeling in a local area may prove valuable in assessing
source contributions to multiple pollutant problems, and aiding policy makers in determining the
most effective control strategies that can be applied to improve air quality.
6 Pollutants such as ammonium nitrate and ammonium sulfate are formed in the atmosphere from chemical reactions of precursor
species NH3, NOx, and S02. Sources of these precursor pollutants include mobile sources and EGUs.
3-12
-------
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
,6®
oe
4?
oe
>^N
/y
XX
?>
*b *o ~*o
^ ^ J? h<
& -S~ .S~ .¦£? _
-------
Chapter 4: Current Trends and Projected
Improvements of Air Quality at the National Level
The 2004 NAS study acknowledged the air quality improvements brought about by
EPA's implementation of the CAA, but additional changes are needed to continue to improve air
quality for multiple pollutants. The NAS report also indicated that solutions to future air quality
issues will benefit from a multi-pollutant, airshed-based approach. This chapter illustrates the
reductions in criteria and hazardous air pollutant concentrations based on ambient measurements.
It also shows the expected improvements in particles, ozone, air toxics (including mercury),
visibility, and nitrogen/sulfur deposition resulting from "on-the-books" federal programs based
on projections from air quality modeling.
Decreasing Trends in Multi-pollutant Concentrations
Under the CAA, EPA established the NAAQS to protect public health and welfare
(visibility impairment and damage to crops, vegetation, and buildings). Over the past three
decades, EPA has partnered with state, local, and tribal agencies to implement programs aimed at
reducing emissions of those pollutants that contribute to poor air quality. The national-level
trends in CAPs and selected air toxics shown in Figure 4-1 indicate the progress resulting from
these programs.
Figure 4-1 shows the average national picture of criteria pollutant concentrations relative
to the standard for each CAP and for air toxics concentrations relative to species-specific cancer
benchmarks recommended by OAQPS. Since at least the mid-1990s, average PMi0, N02, CO,
SO2, and Pb concentrations have been less than 60 percent of their standards. We have data for
some pollutants for 26 years but for some others we have only 7 years worth of ambient
measurements. The changes observed between first and last years are dramatically different
among the pollutants, ranging from 1 to 4 percent reductions per year among the CAPs, while air
toxics range from a 2 percent increase per year for acetaldehyde to a 6 percent reduction per year
for benzene. Of the six pollutants for which EPA establishes national ambient standards, only
two—ozone and fine particles—remain persistent, widespread problems with average
concentrations above, or close to, the NAAQS. In contrast, multiple air toxics, including
benzene, carbon tetrachloride, 1,3-butadiene, acetaldehyde, acrolein, and arsenic are above levels
of concern nationally.
4-1
-------
1990 1995
Year
2000
2005 1994
1996
1998
2000
Year
2002
2004
Sulfur Dioxide Annual Average Normalized to NAAQS 13-Butadiene Normalized to the Cancer Benchmark
Ozone Annual 4|h Max 8-hour Average Normalized to NAAQS -------- Acetaldehyde Normalized to the Cancer Benchmark
CO Annual 2nd Max 8-hour Average Normalized to NAAQS " Benzene Normalized to the Cancer Benchmark
Nitrogen Dioxide Annual Average Normalized to NAAQS Tetrachloroethylene Normalized to the Cancer Benchmark
PM2.5 Weighted Annual Average Normalized to NAAQS
PM10 Weighted Annual Average Normalized to NAAQS
Lead Max Quarterly Average Normalized to NAAQS
22
20 o
18 §
a>
16 Hr
0)
14 5
«|
Q)
10 =
Figure 4-1. National-level trends in CAPs (left) and selected air toxics (right) relative to
the NAAQS and cancer benchmarks.
Note: The trend periods for CAPs and air toxics vary in this figure. These few species were
selected as illustrations for air toxics because they have some of the longest trend records, highest
data quality, and most monitoring sites among air toxics species. Currently, ozone and particle
pollution trends are better understood than trends in air toxics.
Clean Air Rules Will Further Improve Air Quality
EPA recently promulgated a number of federal regulations to reduce multiple air
pollutants. In 2006, EPA implemented the "Clean Air Rules", which included the Clean Air
Interstate Rule (CAIR), the Clean Air Mercury Rule (CAMR), and the Clean Air Visibility Rule
(CAVR). These rules target emissions of NOx, S02, and mercury from power plants (see the
following textbox for more information). Reductions in these pollutants will help improve
multiple air quality problems such as ozone, particle pollution, air toxics, atmospheric deposition
of mercury to waterways, acid rain, and visibility in Class I areas. In addition, EPA promulgated
the Clean Air Nonroad Diesel Rule in 2004 aimed at reducing air pollution from construction,
agricultural, and industrial diesel-powered equipment. These Clean Air Rules will reduce
emissions across a range of pollutants of concern and thus provide a good example of how the
multi-pollutant approach can improve air quality across the US. The results in the rest of this
chapter illustrate how these and other existing CAA programs will simultaneously reduce
individual pollutants based on modeling results.
4-2
-------
EPA's Clean Air Rules: www.epa.gov/cleanair2004/
Clean Air Interstate Rule (CAIR). This rule provides states with a solution to the problem of power
plant pollution that drifts from one state to another. This rule mandates that power plants in the
eastern half of the country reduce SO2 and NOx emissions, which will reduce ozone and particle
pollution concentrations and decrease haze. The rule uses a cap-and-trade system to reduce the
target pollutants by 70 percent.
States Covered by CAIR
I -i States not covered by CAIR
^ States controlled for fine particles (annual SOa and NO,)
Stares controlled for both fine particles (annual S02 and NO.,) and ozone (ozone season NOs)
fjj States controlled for ozone (ozone season NO,)
Clean Air Mercury Rule (CAMR). This rule builds on CAIR to significantly reduce mercury emissions
from coal-fired power plants, the largest remaining domestic source of human-caused mercury
emissions. When fully implemented, these rules, using a cap-and-trade system, will reduce utility
emissions of mercury from 48 tons to 15 tons a year, a reduction of nearly 70 percent.
Clean Air Visibility Rule (CAVR). This rule targets SO2 and NOx and requires emission controls
known as best available retrofit technology, or BART, for selected industrial facilities built between
1962 and 1977 emitting air pollutants that reduce visibility by causing or contributing to regional haze.
Clean Air Nonroad Diesel Rule. This rule will reduce emission levels from construction, agricultural,
and industrial diesel-powered equipment by more than 90 percent and will also remove 99 percent of
the sulfur in diesel fuel by 2010. This rule specifically targets DPM emissions and may also reduce
S02.
Source: U.S. Environmental Protection Agency, 2007c
Figure 4-2 shows the projected changes in pollutant emissions between 2001 and 2020
including the reductions resulting from implementation of the Clean Air Act Programs. As
shown, with the exception of ammonia, emissions of all pollutants are expected to decline over
this timeframe with significant reductions between 30 and 50 percent for NOx, S02, and VOCs.
These declines demonstrate the effectiveness of CAA Programs; however, Figure 4-2 also shows
the large remaining emissions across the eastern and western US in 2020. Our projections
indicate increases in ammonia emissions of 20 percent in the 12 western-most states and 7.5
percent in the 38 eastern-most states from agriculture and related sources over this time period.
4-3
-------
20
o
eg
o
cnj
o
~ 0
o
o
CNJ
E
o
L_
<4—
0)
O)
S -20
sz
o
0)
o>
CD
c
-------
fine particle issues will continue in Midwestern cities such as Chicago, IL; Detroit, MI; and
Cleveland, OH; as well as Birmingham, AL.
Ambient data from 2003-2005
Model Predicted Air Quality in 2020
County Ambient Air Quality
~ PM25 above NAAQS
~ Ozone above NAAQS
~ Ozone and PM25 above NAAQS
Figure 4-3. Past, current, arid projected improvements in ozone and fine particle
air quality for 1999-2001, 2003-2005, and 2020.
Note: Air quality has improved substantially in the past five years and is predicted to improve even
more in the future as a result of the recently promulgated Clean Air Rules and other existing CAA
Programs.
Source: U.S. Environmental Protection Agency, 2006b
Ambient data from 1999-2001
4-5
-------
Air Toxics and Mercury Deposition
Figure 4-4 shows the results of projecting toxicity-weighted HAP emissions from 1990
through 2020 (Strum et al., 2005). Overall, total air toxics emissions are projected to decline by
about 48 percent in 2020 from 1990 levels. As shown, all source sectors except for "Area &
Other" (i.e., stationary sources that do not meet the major source threshold, non-industrial
sources such as residential heating and use of consumer solvents, and fires) decrease between
1990 and 2020.7 These results suggest that the "Area & Other" source sector will contribute as
much as the "Major" source sector in the future. The emerging prominence of the "Area &
Other" source category for HAPs may indicate that it is the most promising sector for future
multi-pollutant reduction efforts. It is also important to note that the onroad and nonroad mobile
source categories emissions will see reductions in the future at the national level but may still be
significant contributors to cumulative risk in urban and local areas.
8
7
6
5
4
3
1990
2002
2020
~ Major
~ Area & Other
~ Fires-Prescribed & Wild
~ Non-Road Mobile
~ On-Road Mobile
Figure 4-4. Projections of risk-weighted air toxics emissions (scaled toxicity
weighted emissions) for 1990, 2002, and 2020.
Source: Strum et al., 2005
1 Some important caveats of these projections are important to understand. Because the MACT program compliance dates are
prior to 2010 and the impact of the residual risk program is not included in these projections, the emissions from major sources
decline from 1990 to 2010 but begin to increase again in 2010. In addition, this study did not include the impact of area source
standards that had not yet been proposed by 2004 and, therefore, future emissions from the "Area & Other" sector may be lower
than reflected in these projections. Finally, future-year mobile source emissions are also expected to decrease even more than
reported as a result of future programs that had not been accounted for at the time of this analysis, including the Mobile Source
Air Toxics Rule, additional standards for small nonroad gasoline engines, and standards for commercial marine vessels and
locomotives.
4-6
-------
Figure 4-5 shows the significant reductions in projected mercury deposition across the
US from 2001 to 2020 as a result of implementation of the Clean Air Rules: CAIR and CAMR.
Almost 115 tons of mercury were emitted by all sources in 2001 with EGUs emitting 42 percent
of the total. Total mercury emissions in 2020 are projected at roughly 77 tons, reflecting a net
reduction of almost 38 tons (32 percent) from 2001 levels. EGU reductions account for 24 tons
(62 percent) of the total reduction.
2001 Base Case
Deposition in
Micrograms ISq. Meter
Less Than 1
¦ l-5
¦ 5-10
¦ 10-15
¦ 15-20
M Over 20
states
2020 with CAIR, CAMR and Other
Control Programs
Figure 4-5. Projected improvements in mercury deposition from 2001 to 2020.
Note: Deposition of mercury will decrease significantly as a result of implementing CAIR and
CAMR.
Visibility
In Figure 4-6, the difference in estimated visual range between measurements taken in
2002 and projected visual ranges based on Community Multiscale Air Quality (CMAQ) model
predictions in 2020 is shown for a subset of 114 IMPROVE sites. On the map, circle size
corresponds to the amount of improvement in visual range (i.e., larger circles correspond to more
improvement). All 114 sites show improvement in visual range. Sites in the eastern US and the
Pacific Northwest show the most improvement. Improvement in the east can be attributed to a
combination of CAIR and mobile source regulations that are projected to reduce particle
pollution in this region. Control measures implemented to achieve compliance with the
NAAQS, the Acid Rain Program, and the Regional Haze Rule also contribute to reductions in
particle pollution and improvements in visual range at Class I sites.
4-7
-------
nO
Improvement in Visual Range from 2002 to 2020 (km)
0.00 14.45 28.90
Figure 4-6. Projected improvement in visual range (km) from 2002 to 2020.
Nitrogen and Sulfur Deposition
Figure 4-7 illustrates the spatial distribution of CMAQ-predicted changes in nitrogen and
sulfur deposition nationwide from 2001 to 2020 (U.S. Environmental Protection Agency,
2006b). These maps show both wet and dry deposition. In addition, both oxidized and reduced-
form nitrogen deposition are included in the total. From 2001 to 2020, control measures that are
part of Title IV (Acid Rain Program) and several mobile source rules (including Tier I and
Tier II vehicle controls, heavy-duty diesel and nonroad engine standards, and lower volatility and
low-sulfur fuels) are expected to provide dramatic reductions nationwide in SO2 and NOx
emissions that are principal contributors to sulfur and nitrogen deposition. In addition, the NOx
SIP Call and CAIR influence significant additional emission reductions of NOx and SO2 in the
eastern US. However, ammonia emissions (another source of nitrogen) from agriculture are
expected to increase by 2020. The NOx, SO2, and ammonia emission changes shown led to the
predicted regional reductions in nitrogen and sulfur deposition seen in these figures. Local
increases in nitrogen deposition are associated with locations where increases in ammonia
emissions are projected to occur.
4-8
-------
Percent Change in Nitrogen Dep: 2001 to 2020
60.0 112
45.0
30.0
15.0
0.0
-15.0
-30.0
-60.0 1
1
Percent Change in Sulfur Dep: 2001 to 2020
Figure 4-7. CMAQ-predicted changes in nitrogen and sulfur deposition from
2001 to 2020 (percent change).
Summary
Reductions in the concentrations of CAPs and air toxics have been shown to improve
historically using ambient data. This chapter shows the expected air quality improvements in
particles, ozone, and mercury resulting from "on-the-books" federal programs based on
projections from air quality modeling. While air quality projections for air toxics are not yet
available, projections of weighted emissions indicate that air toxics from all sources other than
"Area & Other" will be reduced in the future as a result of federal programs.
4-9
-------
Chapter 5: Multi-Pollutant Analytical Products and Capabilities
A key component of the air quality management cycle is technical information collection
and analysis. Having the right tools to do the job is vital, and these tools must be able to
successfully integrate across pollutants, across media, and across spatial scales. Figure 5-1
demonstrates how technical data and tools allow those multi-pollutant concepts discussed in
Chapter 1 to be considered as part of the regulatory and policy development process. Following
the recommendations of the 2004 NAS study, OAQPS is emphasizing a more integrated and
multi-pollutant approach for air quality management. For example, the multi-pollutant AQMP
pilot projects highlighted in Chapter 1 will illustrate how these technical elements can be
employed in the air quality management process to address multi-pollutant air quality issues.
emissions
Vohctes, tnchisfry, ult&es, cws
EmteikHt ittfe* vary met (fens and
dunng ^ norma* totaling ccrv
maffuncCon) m unusual nigral
xnet prcK&idai, unpaved reaKfeu
space; mm an bo especially Jrigfi
ii!S®ns ftido^procBaa
wife
i
Physical Processing
5H.lt.' r>! if-s nl&Sh® Igcal
Transport of prtwtarts yp to 8000*
of kilometers
Chemical Processing
;•#. co.
I*#., ft. mm «* wiitwmiH}
* Cycfcsltsmsrnemiolaiieirafea
• Sulfur diMhfefSGh} « Putia"OCs •
] SO, and NO,
1»» mt wxlueon
*
* Indoor^ and an
* Routfrx othof than air
Maullh eHocIs are caused by acute ^abort-
Wm} and titvtmtc ijicpcwwrss
Visibility
Impairment
I
Health
Effects
mmmmm,. oimt potential ha*»s)
Ewysfem Exposure
Ecosystem
Ff'ects
Mstp rials
OtgredatiOrt
is.f „ rrranurmms. but I i
Figure 5-1. Technical elements of the air quality management system.
Source: National Research Council, 2004
5-1
-------
OAQPS is playing a lead role in developing technical products and capabilities to inform
regulatory and policy efforts that will help to identify effective multi-pollutant solutions to
environmental problems. These products and capabilities include, but are not limited to, the
following:
1. Integrated emissions inventory
2. Integrated monitoring network
3. "One atmosphere" air quality modeling
4. Multi-pollutant modeling platform
5. Spatial predictions of air quality data
This chapter provides details about each of these products and capabilities, some of
which are available now and some of which are on the horizon. It should be noted that several of
these products are already used for various EPA activities such as NAAQS designation and
implementation work, regulatory impact analyses, risk assessments, monitoring network design,
and source sector analyses.
Integrated Emissions Inventory - 2002, 2005, and 2008 National Emissions Inventories
In 2001, at EPA's International Emissions Inventory Conference workshop "One
Atmosphere, One Inventory, Many Challenges", EPA proposed combining CAP and HAP
inventories. Since then, EPA has taken steps toward this "one inventory" idea with its 2002 NEI.
The 2002 NEI is EPA's latest comprehensive national emission inventory for the entire US. It
contains emission measurements and estimates for seven CAPs, their precursors, and 188 HAPs.
The NEI contains emissions data for all major contributors to air pollution, including point
sources, mobile sources, and non-point sources.
For point sources, the NEI includes emissions for individual processes at an industrial
facility. For mobile and non-point sources, the NEI contains county-level emission estimates.
The NEI is developed using the latest data and best estimation methods including data from
Continuous Emissions Monitors (CEMs), data collected from all 50 states and many local and
tribal air agencies, and output from EPA's latest models such as the MOBILE and NONROAD
models.
In order to ensure adequate resources to complete the NEI reengineering effort (see next
paragraph), EPA developed the 2005 National Emissions Inventory (Version 1) as a reduced
effort version based on the 2002 NEI (Version 3). Only the highest value-added adjustments
from readily available datasets were used to make additions for the 2005 NEI:
• Inclusion of 2005 electric utility data from EPA's Clean Air Markets Division
(CAMD)
• Inclusion of 2005 National Mobile Inventory Model (NMIM) results for onroad and
nonroad emissions
• Information about facility closures from state/local and tribal agencies
• Inclusion of 2005 wildfire/managed burn data
5-2
-------
EPA is in the process of creating the 2005 NEI Version 2, which will be completed in spring
2008. The 2005 NEI Version 2 includes updates for non-EGU stationary sources with the
following data:
• HAP emissions data received from States and industry to support the MACT program,
including the recent Risk and Technology Review rulemaking
• 2005 State, local, and tribal data submitted to EPA under the Consolidated Emissions
Reporting Rule (CERR)
• HAP data from the Toxic Release Inventory (TRI) for missing facilities and pollutants
• Off-shore platform emissions data from Mineral Management Services (MMS)
OAQPS is also undertaking a project to re-engineer the processes and systems currently
used to build the NEI. The Emissions Inventory System (EIS) will result from this project and
will be used to generate the NEI beginning with the 2008 inventory cycle. The goals of this
project are to (1) provide better quality assurance tools so that data can be checked before being
submitted by the states; (2) implement more efficient processes to significantly reduce the time it
takes to develop the inventory; and (3) create a central location for users to access the resources
they need to develop and maintain their inventories. The 2008 NEI will be available in June
2010.
Integrated Monitoring Network - NCore
EPA has proposed a new national multi-pollutant air monitoring network strategy called
the National Core Monitoring Network (U.S. Environmental Protection Agency, 2007d). NCore
will measure multiple pollutants in a geographically diverse network of monitoring sites. The
information provided by this network should help improve our understanding of the relationships
among air quality problems. These new multi-pollutant monitoring sites will anchor the current
national monitoring network. The sites will measure important precursor gases, basic
meteorology, as well as ozone, particles, CO, NOx, and SO2. Although CO, NOx, and SO2 are
criteria pollutants, NCore is designed foremost to measure trace-level concentrations in
representative, well-mixed locations to support accountability studies and exposure- and health-
based assessments, and to evaluate air quality models. In addition, the sites will also measure
VOCs, air toxics, and other key measurements, such as meteorological variables. By measuring
multiple pollutants and other measurements at a single location, EPA and its partners can
maximize the multi-pollutant information available. This approach greatly enhances the
foundation for future health studies and NAAQS revisions. The sites will be placed in broadly
representative urban locations (about 55 sites) and rural locations (about 20 sites) throughout the
country to help characterize urban and regional patterns of air pollution (Figure 5-2).
5-3
-------
Figure 5-2. Map of candidate NCore sites.
Source: U.S. Environmental Protection Agency, 2007d
The NCore network will incorporate new instrument methods that meet the goals of
(1) using automated methods that continuously monitor air quality to improve temporal
resolution and (2) continuing to lower detection limits to provide information at lower
concentrations. These improvements are particularly important for health-based studies because
they provide scientists with more data (e.g., every hour rather than a daily average) of higher
certainty (e.g., smaller bias and better precision).
"One Atmosphere" Air Quality Modeling - CMAQ Model
Atmospheric modeling of multiple pollutant issues simultaneously is important from an
air quality management perspective because of the relationships among sources, transport, and
transformation processes of a number of CAPs and air toxics. "One atmosphere" modeling will
allow the assessment of pollutant control measures that affect more than one pollutant or issue.
For example, installing scrubbers at electric utilities to remove S02 also can result in reduced
NOx or mercury emissions. Reducing emissions of ozone precursors (i.e., NOx and VOCs) will
not only affect ozone, but may also affect the formation of certain secondary toxic VOCs. Thus,
"one atmosphere" modeling provides the ability to examine the effects of multi-pollutant control
strategies and the co-benefits of programs across CAPs and air toxics.
The CMAQ model is designed to simulate the formation and fate of ozone, oxidant
precursors, and particle concentrations over national, regional, and urban spatial scales. Key
chemical and physical processes treated by CMAQ include:
-------
• gas-phase photochemistry,
• secondary aerosol (e.g., sulfate) formation through gas-phase and in-cloud aqueous-phase
processes,
• partitioning of nitrate between nitric acid gas and particle nitrate,
• horizontal and vertical transport of pollutants, and
• removal of pollutants through wet and dry deposition.
The current version of CMAQ (v4.6) has a number of scientific updates and
advancements compared with earlier versions. Most notably, this version is capable of "one
atmosphere" modeling in which ozone, primary and secondary particles, mercury, and other
selected toxic pollutants are treated in an integrated manner in a single model simulation.
Benzene, formaldehyde, hydrazine, chloroform, chromium (III and VI), and cadmium are among
the more than 30 toxic gases and metals now simulated by CMAQ. Therefore, transitioning to
this version of CMAQ will allow us to simultaneously evaluate impacts across multiple
pollutants as part of our regulatory and policy assessments that were previously completed
individually and separately for CAPs and air toxics.
CMAQ is used in regulatory and policy assessments to project future nonattainment and
air quality impacts for use in cost/benefit analysis as part of Regulatory Impact Assessments
(RIAs). Inputs to CMAQ include emissions from anthropogenic and biogenic sources,
meteorology, and estimates of pollutant concentrations transported into the area being modeled
(i.e., boundary conditions). CMAQ provides outputs of gridded concentrations and deposition
on an hourly basis for the user-defined modeling domain (i.e., the area covered by the model
simulation). Pollutant concentrations predicted by CMAQ are output for each of the multiple
vertical layers included in the model simulation. The standard hourly CMAQ predictions can be
post-processed to create gridded fields of daily, monthly, and annual average concentrations and
total deposition. In addition, ozone and fine particle species predictions from CMAQ can be
combined with ambient data to project ozone and fine particle design values for future years and
control case emissions scenarios and associated air quality changes that can be input to BenMAP
for estimating health and environmental effects and their monetary benefits.
Multi-pollutant Modeling Platform - 2002 and Projected Future Years
Based on the 2002 NEI and CMAQ model, we have developed a 2002-based multi-
pollutant "modeling platform". A modeling platform is defined as a structured system of
connected modeling-related tools and data that provide a consistent and transparent basis for
assessing the air quality response to changes in emissions and/or meteorology. As part of this
development effort, we conducted emissions inventory processing and national, regional, and
local air quality modeling to produce concentration and deposition estimates for CAPs and HAPs
with a detailed performance evaluation based on ambient monitoring data. The reports from this
effort are planned to be completed in 2008. To meet near-term regulatory needs, the
"CAP-only" version of this platform is serving as the baseline for air quality modeling of ozone
and fine particles as part of the RIA in support of EPA's Locomotive and Commercial Marine
Final Rule and the Ozone NAAQS Revisions Final Rule.
5-5
-------
As shown in Figure 5-3, application of this modeling platform yields simultaneous
concentration predictions for criteria pollutants (ozone and fine particle species) and air toxics
(e.g., formaldehyde, benzene, and DPM) as well as deposition of mercury, nitrogen, and sulfur.
The maps illustrate the CMAQ model predictions for a single summer day based on a consistent
set of emissions and meteorological inputs. This multi-pollutant modeling platform enables us to
consider CAP/HAP interactions through both chemical interactions (e.g., VOCs, particle-related
metal HAPs) and control interactions (joint emission reductions) as discussed in Chapter 1.
Therefore, this modeling platform allows "co-benefit" assessments to be conducted for current
and future policies or rules and informs the development of multi-pollutant control strategies.
We are conducting this type of multi-pollutant modeling as part of the technical effort for
the Detroit Multi-pollutant Pilot Study that will be completed in 2008. We expect that
demonstrating this modeling capability will greatly enhance our ability to simulate and evaluate
multi-pollutant control strategies and/or multi-pollutant impacts of our programs, thereby
supporting our new vision of air quality management.
Spatial Predictions of Air Quality Data - CDC/PHASE Project
Currently, ambient monitoring data and air quality modeling results are used separately to
inform regulatory and policy assessments. However, these data and results are limited in various
ways. Monitoring data are point measurements that may not be representative of air quality
across a broad geographic area resulting in "unmonitored" locations for which we need
information. Alternatively, air quality models provide more complete geographic coverage, but
the resolution of their predictions may not be adequate for some assessments such as
neighborhood scale risk assessments and may also be limited by the quality of the emissions and
meteorological inputs. Therefore, to derive a more spatially complete and accurate measure of
air quality across multiple pollutants, EPA is exploring how best to combine or "fuse" these
disparate data sets to better inform air quality management activities.
McMillan et al. (2007) provided an overview and policy motivation for developing
spatial models based on ambient and predicted air quality data. A number of air quality
management activities such as NAAQS risk/exposure assessments and comparisons with health
outcomes data are underway. An example in this area is the Public Health Air Surveillance
Evaluation (PHASE) project, which is a multi-disciplinary collaboration between EPA, the
Centers for Disease Control and Prevention (CDC), and three Environmental Public Health
Tracking Network (EPHTN) state agencies. The objective of this project is to develop and
evaluate the use of fused air quality predictions to associate public health tracking data with
ozone and fine particles as part of the EPHTN program.
5-6
-------
Ozone and PM2.5 Species
July 8 Daily Max 8-Hr Ozone
July 8,2002 0:00:00
Min= 17.4 at (86,112). Max= 118.1 at (25,45)
July 8 Sulfate
4.00 112
3.50
3.00
2.50
2.00
1.50
1.00
0.50
¦ 16.0112
* 14.0
July 8 Primary Organic Carbon
daily Avg
x=2002ac_tox_v4.61_L3th_us36b.dailyavg.aconc.07
July 8,2002 1:00:00
Min= 0.00 at (1,45), Max= 15.82 at (47,40)
Daily Avg
x=2002ac_tox_v4.61_L3th_us36b.dailyavg.aconc.07
July 8,20021:00:00
Min= 0.0 at (139,39), Max= 17.8 at (100,53)
Figure 5-3. Multi-pollutant modeling outputs for CAPs, HAPs, and deposition for
an example summer day in 2002.
July 8,20021:00:00
Min= 0.00 at (129,9), Max= 0.37 at (77,34)
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
ug/m3
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
ug/m2
July 8 Formaldehyde
Daily Avg Primary + Secondary
y=2002acJox_v4.61_L3th_us36b.dailyavg.aconc.tox.07
July 8,20021:00:00
Min= 0.08 at (25,33), Max= 15.94 at (43,96)
July 8 Benzene
Daily Avg
y=2002ac_tox_v4.61_L3th_us36b.dailyavg.aconc.tox.07
July 8,2002 1:00:00
Min= 0.00 at (33,11),Max= 2.32 at (130,67)
July 8 Mercury Deposition
Daily Total All Species
z=2002ac_tox_v4.61_L3th_us36b.dailysum.dep.07
July 8,2002 1:00:00
Min= 0.00 at (67,112), Max= 55.88 at (127,67)
Air Toxics
July 8 Diesel PM
0.20112
0.18
0.16
0.14
0.12
Deposition
July 8 Sulfur Deposition
Daily Total
z=2002ac_tox_v4.61_L3th_us36b.dailysum.dep.07
0.20112
0.18
July 8 Nitrogen Deposition
Daily Total
z=2002ac_tox_v4.61_L3th_us36b.dailysum.dep.07
Min=
July 8,2002 1:00:00
0.00 at (62,106), Max= 0.28 at (81,78)
Daily Avg
y=2002ac_tox_v4.61_L3th_us36b.dailyavg.aconc.tox.07
1.60112
Min=
July 8,2002 1:00:00
0.00 at(1,112), Max= 3.34at(23,46)
w=2002ac_tox_v4.6l_L3th_us36b.aconc.o3_8hrmax_LST.ioapi
120.0112 i—n 7=r~l— 7
5-7
-------
For the CDC/PHASE work, a statistical technique is being used to combine ambient
monitoring data with output from photochemical models (i.e., CMAQ) to predict ambient air
concentrations. These fused predictions of air quality are particularly useful in areas without
monitors or for days on which monitors do not operate. This technique allows more accurate
prediction where and when monitoring data are not available. Although this work currently
focuses on ozone and fine particles, we are working to extend this technique to air toxics and
individual fine particle species as part of the Detroit Multi-pollutant Pilot Study. Results from
this modeling study will be combined with ambient monitoring data to provide improved spatial
and temporal air quality characterizations as part of the risk and exposure assessments conducted
for the Detroit Exposure and Aerosol Research Study (DEARS).
Summary
The tools needed to analyze multi-pollutant data are either available currently or being
developed in OAQPS; they are key components of the air quality management cycle's technical
information, collection, and analysis steps. As this suite of tools becomes fully available, the
analyses needed to support a multi-pollutant approach for air quality management will be better
realized.
5-8
-------
REFERENCES
Bachmann J. (2007) Will the circle be unbroken: a history of the National Ambient Air
Quality Standards. Air & Waste Manag. Assoc. 's EM.
Brown S.G., Wade K.S., and Hafner H.R. (2007) Summary of recent ambient air quality and
accountability analyses in the Detroit area. Summary report prepared for the U.S.
Environmental Protection Agency, Research Triangle Park, NC, by Sonoma
Technology, Inc., Petaluma, CA, STI-905317.02-3240, September.
Community Modeling & Analysis System Center (2007a) CMAQ review process. Web site
hosted by the CMAS Center, University of North Carolina at Chapel Hill.
.
Community Modeling & Analysis System Center (2007b) CMAQ version 4.6. Web site
hosted by the CMAS Center, University of North Carolina at Chapel Hill.
.
LADCO (2007a) Midwest Regional Planning Organization projects. Web site prepared by
the Lake Michigan Air Directors Consortium, .
LADCO (2007b) Data analysis at LADCO. Web site prepared by the Lake Michigan Air
Directors Consortium, .
Lee J.H. and Hopke P.K. (2006) Apportioning sources of PM2 5 in St. Louis, MO using
speciation trends network data. Atmos. Environ. 40, S360-S377.
McMillan N., Holland D., Morara M., and Young G.S. (2007) Spatial prediction using
combined sources of data, (in press).
Michigan Department of Environmental Quality (2005) Detroit air toxics initiative, risk
assessment report. November. http://www.michigan.gov/documents/DATI_-
_complete_final_report_l l-9-05_142053_7.pdf.
Michigan Department of Environmental Quality (2007a) Detroit air toxics initiative.
.
Michigan Department of Environmental Quality (2007b) 2006 annual air quality report.
Prepared by the Michigan Department of Environmental Quality, Air Quality
Division, Lansing, MI, November, .
NARSTO (2004a) Particulate matter assessment for policy makers: aNARSTO assessment,
P. McMurry, M. Shepherd, and J. Vickery, eds., Cambridge University Press,
Cambridge, England, .
NARSTO (2004b) Particulate matter assessment for policy makers: Chapter 3: Atmospheric
aerosol processes, P. McMurry, M. Shepherd, and J. Vickery, eds., Cambridge
University Press, Cambridge, England.
.
-------
NARSTO (2004c) Particulate matter assessment for policy makers: Chapter 1: Perspective
for Managing PM, P. McMurry, M. Shepherd, and J. Vickery, eds., Cambridge
University Press, Cambridge, England.
.
National Research Council (2004) Air quality management in the United States, The National
Academies Press, Washington, D.C.
.
Rizzo M.J. (2005) A source apportionment analysis of the Dearborn Speciation Trends
Network site. Prepared by the U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards.
Seinfeld J.H. and Pandis S.N. (1998) Atmospheric chemistry and physics: from air pollution
to climate change, J. Wiley and Sons, Inc., New York, New York.
Strum M., Pope A., Palma T., Mason R., Shedd S., Cook R., Thurman J., and Ensley D.
(2005) The projection of hazardous air pollutant emissions to future years: methods &
results. Presented at the Emission Inventory Conference, Las Vegas, NV, April, by the
Office of Air Quality Planning & Standards, the Office of Transportation and Air
Quality, and the Computer Sciences Corp.
.
Trepat E., Kim E., Paatero P., and Hopke P.K. (2007) Source apportionment of time and size
resolved particulate matter measured with a rotating drum sampler. Atmos. Environ.
41 (28), 5921-5933.
U.S. Environmental Protection Agency (2006a) Technology Transfer Network, 1999
National-Scale Air Toxics Assessment, 1999 assessment results.
.
U.S. Environmental Protection Agency (2006b) EPA's Regulatory Impact Analyses (RIA) for
the revised ozone and PM NAAQS and proposed Regional Haze Rule. Report
prepared by the U.S. Environmental Protection Agency, Research Triangle Park, NC.
.
U.S. Environmental Protection Agency (2007a) 2002 National Emissions Inventory data &
documentation, .
U.S. Environmental Protection Agency (2007b) Detroit exposure and aerosol research study.
.
U.S. Environmental Protection Agency (2007c) Clean Air Rules of 2004. Web site prepared
by the U.S. Environmental Protection Agency, .
U.S. Environmental Protection Agency (2007d) NCore multipollutant monitoring network.
Web site prepared as part of the Technology Transfer Network, Ambient Monitoring
Technology Information Center, .
------- |