United
Environmental Protection Agency
REGION/ORD/OAR WORKSHOP ON
AIR TOXICS EXPOSURE ASSESSMENT
SUMMARY REPORT
June 25 - 27, 2002
San Francisco, California
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 15-27,2002
TABLE OF CONTENTS
FOREWORD
.-
vn
EXECUTIVE SUMMARY
WORKSHOP SESSION SUMMARIES
IX
. 1
.
m
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 15-27, 2002
INTRODUCTION 1
Introductory Remarks-Jack Broadbent (Region 9) 1
Air Toxics Exposure Assessment Workshop - Welcome - Larry Cupitt (ORD/NERL) .... 2
Overview of Exposure Assessment and Air Toxics — Ken Mitchell (Region 4) 3
Overview of Air Toxics Exposure Assessment in ORE) - Tim Watkins (ORD/NERL) 5
SESSION I: DESIGNING AN AIR TOXICS EXPOSURE ASSESSMENT - MONITORING vs.
MODELING 7
Monitoring vs. Modeling - An Interactive Group Exercise - Paul Shapiro (ORD/NCER)
and Ted Palma (OAR/OAQPS) 7
Case Studies to Illustrate Uses of Monitoring and Modeling 9
MATES II - A Regional Perspective - Mike Nazemi (South Coast Air Quality
Management District) 9
Minneapolis - St. Paul - A Comparison of Community, Residential, and Personal
Exposure - John Adgate (University of Minnesota) 12
Session I Expert Panel Discussion - Mike Nazemi (SCAQMD), John Adgate (UMN), John
Girman (OAR/ORIA), Larry Cupitt (ORD/NERL), Neil Frank (OAR/OAQPS),
Matt Lorber (ORD/NCEA), Joe Touma (OAR/OAQPS) 15
SESSION II: MONITORING METHODS AND NETWORK DESIGN 19
Air Toxics Monitoring Pilot Project - Barbara Morin (Rhode Island Department of
Environmental Management) 19
New Trends in Monitoring Methods- Don Whitaker (ORD/NERL) 21
Atmospheric Formation and Decay of Air Toxics - Implications for Exposure Assessments
- Deborah Luecken (ORD/NERL) 23
Air Toxics Monitoring Methods and Network Design - Steve Bortnick and Shannon
Stetzer (Battelle Memorial Institute) 25
California Monitoring Program: Statewide Network to the Neighborhood Scale — Jeff
Cook and Linda Murchison (California Air Resources Board) 28
Design Your Own Air Toxics Monitoring Network - Motria Poshyvanyk (Region 5) .... 33
IV
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 15-27,2002
SESSION III: MODELING TOOLS - CURRENT AND FUTURE 35
Air Quality Models - Joe Touma (OAR/OAQPS) 35
Using Emission Inventories for Air Quality Modeling - Joe Touma (OAR/OAQPS) and
Madeleine Strum (OAR/OAQPS) 37
Model Applications: Local and Urban Scale Modeling 38
Local Scale - Barrio Logan Modeling Analysis - Vlad Isakov (California Air Resources
Board) 38
Urban Scale Modeling - Houston Case Study - Joe Touma (OAR/OAQPS) 40
Applying CMAQ Models3 for Air Toxics Assessments - Bill Hutzell (ORD/NERL) 42
Questions and Discussion with Session III Speakers 44
Breakout Groups to Discuss Modeling Topics/Questions 46
SESSION IV: HUMAN EXPOSURE ASSESSMENT 49
Introduction to Human Exposure - Linda Sheldon (ORD/NERL) 49
Air Toxics Exposure in Indoor Environments - John Girman (OAR/ORIA) 51
Multi-pathway Exposure Assessment - Matthew Lorber (ORD/NCEA) 53
What Human Exposure Data and Models are Available? - Haluk Ozkaynak (ORD/NERL)
55
Questions and Discussion with Session IV Speakers 57
Additional Case Studies: Modeling and Monitoring 59
Personal Exposure Monitoring Meets Risk Assessment: The South Baltimore Community
Exposure Study - Devon Payne-Sturges (OA/OPEI) 59
Monitoring: Children's Exposure to Diesel School Bus Emissions - David Brown
(EHHI/NESCAUM) 62
National Air Toxic Assessment (NATA) - The Initial National Scale Assessment - Ted
Palma (OAR/OAQPS) 65
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 15-27,2002
SESSION V: SOURCE APPORTIONMENT 69
Introduction to Source Apportionment Methods - Lynn Hildemann (Stanford University)
69
Source Apportionment Tools - Gary Norris (ORD/NERL) 71
A Modern Example: Northern Front Range Air Quality Study - Chad Bailey
(OAR/OTAQ) 73
Panel Discussion with Session V Speakers 74
SESSION VI: COMMUNICATING THE RESULTS AND WORKSHOP CONCLUSIONS
75
Communicating the Results of Air Toxics Exposure Assessments - Alvin Chun (Region 9)
75
Conclusions and Next Steps - David Klauder (ORD/OSP) 78
APPENDIX A: AGENDA A-l
APPENDIX B: LIST OF PARTICIPANTS B-l
APPENDIX C: SLIDES FROM PRESENTATIONS C-l
APPENDIX D: FLIP CHART NOTES D-l
APPENDIX E: PARTICIPANT EVALUATION SUMMARY E-l
APPENDIX F: AIR TOXICS CASE STUDIES F-l
VI
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
FOREWORD
The ORD/OAR/Regional Training Workshop on Air Toxics Exposure Assessment was the tenth
in a series of Regional Science Topic Workshops sponsored by the Office of Science Policy
(OSP) in the Office of Research and Development (ORD) at the United States Environmental
Protection Agency (EPA). Other workshops in this series included:
• Asthma: The Regional Science Issues
• Communicating Science: Waves of the Future Info Fair
• Fully Integrated Environmental Location Decision Support (FIELDS)
• Non-Indigenous Species
• Pesticides
• Endocrine Disruptors
• Emerging Issues Associated with Aquatic Environmental Pathogens
• Aquatic Life Criteria
• Critical Ecosystems
The objectives of the Regional Science Topic Workshops are to: 1) establish a better cross-
Agency understanding of the science applicable to specific region-selected human health and/or
ecological topics, and 2) develop a network of EPA scientists who will continue to exchange
information on these science topics as the Agency moves forward in planning education,
research, and risk management programs.
Each year, EPA regions identify priority science topics on which to conduct workshops. The
workshops address the science issues of greatest interest to the regions on the selected topic area.
Each workshop is planned and conducted by a team of regional, ORD, and interested program
office scientists, is led by one or more Regional Science Liaisons (RSLs) to ORD, and is
facilitated by a regional chairperson. Participants maintain the cross-Agency science networks
they establish at the workshops through planned post-workshop projects and activities such as
identifying collaborative research opportunities, creating information sharing mechanisms (e.g.,
interactive web sites), and developing science fact sheets for regional use.
For additional information on a specific workshop or on the Regional Science Topic Workshop
series in general, contact David Klauder in ORD's Office of Science Policy (202-564-6496).
vn
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
THIS PAGE INTENTIONALLY LEFT BLANK
Vlll
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
EXECUTIVE SUMMARY
The ORD/OAR/Regional Training Workshop on Air Toxics Exposure Assessment was held on
June 25 - June 27, 2002, in San Francisco, California. The workshop was chaired by Winona
Victery (Region 9) with support from David Klauder (ORD/OSP).
The workshop was organized into six sessions:
/. Designing an Air Toxics Exposure Assessment - Monitoring vs. Modeling
II. Monitoring Methods and Network Design
III. Modeling Tools - Current and Future
IV. Human Exposure Assessment
V. Source Apportionment
VI. Communicating the Results and Workshop Conclusions
The workshop focused on two general exposure assessment questions:
1. What is our inhalation exposure to toxic chemicals of concern in our regions (at a
"screening level of certainty")?
2. What is our inhalation exposure to toxic chemicals of concern in our community (at a
"high level of certainty")?
Scientists from EPA (regions, Office of Research and Development, and Office of Air and
Radiation) and invited speakers from private industry, academia, and state agencies presented
methods, current research, and case studies on monitoring and modeling, human exposure, and
source apportionment. Two breakout sessions focused on designing an air toxics monitoring
network (Session II), and an in-depth discussion of several topics related to modeling (Session
III). Participants also took part in an interactive exercise designed to simulate a real-world
problem to be solved by weighing the advantages and drawbacks of both modeling and
monitoring, as well as communicating the results. The last session included a presentation on
effectively communicating the results of air toxics exposure assessments to the public. The
closing remarks and discussion at the end of the workshop generated a list of action points and
potential workshop outcomes (see Session VI). Planned outcomes include posting of
presentation slides and workshop summary on the OSP internet site. Most participants found the
workshop useful according to the workshop evaluations; many expressed a need for methods and
tools, training, improved communication, and sharing of data, results, and tools application.
IX
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
THIS PAGE INTENTIONALLY LEFT BLANK
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
INTRODUCTION
Welcome: Jack Broadbent (Region 9) and Larry Cuppitt
(ORD/NERL)
Workshop Goals/Logistics: Winona Victery (Region 9)
PLEASE NOTE: Slides from the Workshop presentations are available at:
http://epa.gov/osp/regions/workshops.htm
Introductory Remarks — Jack Broadbent (Region 9)
Jack Broadbent began the workshop by welcoming the participants and speakers, and
emphasized the increasing public concern regarding the subject of air toxics. Implementing a
strategy for urban air toxics is a top EPA priority. To implement such a strategy, it is necessary
to understand how to perform exposure assessments and interpret their results; how to
incorporate modeling and monitoring in conducting such assessments; and how to communicate
assessment results to the public. The workshop also provides a valuable opportunity to meet
persons from across the Agency that are involved in the field of air toxics.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Air Toxics Exposure Assessment Workshop — Welcome - Larry Cupitt
(ORD/NERL)
U.S. EPA Administrator Christine Whitman has emphasized the importance of using sound
science for Agency decisions. Science Topic Workshops are part of the Office of Research and
Development's (ORD) Regional Science Program. Up to three such workshops are held
annually, on topics chosen by the regions. The regions, ORD, and interested program offices all
participate on the workshop planning teams. Recognized as an effective way to understand the
state of the science and identify regional science needs, these workshops also help build liaisons
for information exchange and reveal opportunities for integration and collaboration. Specific
objectives of the Regions/ORD Science Topic Workshops include:
• Creating cross-Agency science networks;
• Providing opportunities to integrate EPA science into regional decision-making; and
Identifying critical science uncertainties and needed science products.
Sound science is critical to understanding how emissions, ambient conditions, exposures, and
ultimate impacts are linked. Air toxics exposure assessments provide information for EPA to
decide whether to take action, to determine which actions will be most effective, and to provide
a measure of success of the actions taken. Exposure science, however, also presents some
challenges. The need to track exposure data is recognized, particularly as more data (e.g., on
bioindicators) continues to become available. EPA also needs to know more about exposure.
These challenges can be met with an aggressive program to make the data useful in protecting
both human health and the environment. In addition, the available indicator data should be
augmented with information on hazards associated with the exposure, sources or pathways of
exposure, and scientifically credible actions to mitigate both exposure and effects.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Overview of Exposure Assessment and Air Toxics - Ken Mitchell (Region 4)
The mission of the EPA is to protect human health and to safeguard the natural environment
upon which life depends. To accomplish this mission, EPA strives to ensure that all Americans
are protected from significant risks to human health and the environment where they live, learn,
and work. Exposure to air toxics can occur at two different scales (national/regional and
community); assessments should take both of these scales into account. Air toxics are a
significant potential risk because they may cause health effects, disperse from one location to
another, persist and/or bioaccumulate, and have multiple potential routes of exposure. In
addition, little is known about risks from exposure to air toxics. The Clean Air Act (CAA) lists
188 hazardous air pollutants (HAPs), but it is likely there are other chemicals that are not yet
recognized as air toxics. The National Air Toxics Assessment (NATA) provided a national-scale
estimate of cancer risk due to thirty-three priority urban air toxics. Although emissions have
decreased in the last ten years, median cancer risks are still relatively high throughout the
country - from less than one per million to more than a hundred per million. Indoor air quality
must also be considered; indoor air can be more polluted than outdoor air, and most people
spend the majority of their time indoors.
Multiple programs and initiatives exist that address air toxics; dose-response assessment and
exposure assessment are particularly important in characterizing overall risk. Human exposure
to air toxics occurs any time there is contact with a chemical; some examples of exposure routes
are inhalation, ingestion, and uptake (absorption) through the skin or eyes. Each exposure
results in a dose, or the amount of chemical that reaches the external barrier (applied dose),
crosses the external barrier (internal dose), reaches an individual organ/tissue (delivered dose),
or reaches cells/membranes to cause adverse effects (biologically effective dose) [slides 17,18].
Chemical contact, uptake/intake rates, pathways and routes of exposure, and the potential and
absorbed dose can all be estimated by exposure assessment. Mathematical formulas were
presented for calculating risk [slides 22,23]. The effective concentration ("C" in the risk
equations) can reflect chronic exposure (low level exposure over long periods of time) or acute
exposure (high level exposure over a short time). Chronic and acute exposure often exhibit
different effects. Cumulative exposure is also important, and provides a more complete
approach than measuring individual sources; it is often measured at the personal exposure level.
Assessments can be conducted at different scales; risk is found to be higher, in general, in
studies done at low levels of geographic scale. Risks are likely much higher in specific places
than estimated by the national assessment, and risk vary from one person to another. As a result,
assessment goals have to be carefully defined to determine which level of assessment is needed.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Finally, uncertainties exist at every step of the exposure and risk assessment process, and must
be considered in the analysis of results.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Overview of Air Toxics Exposure Assessment in ORD — Tim Watkins
(ORD/NERL)
Tim Watkins updated participants on current and planned ORD research in the area of air toxics
exposure assessment. A diagram [slide 2] illustrated how the five ORD laboratories align with
the risk assessment/risk management paradigm. The air toxics research program is based on the
Air Toxics Research Strategy (ATRS) and the Air Toxics Multi-Year Plan (MYP). Both these
documents are undergoing peer review and will be publicly available in the near future. The
ATRS outlines the key questions addressed by exposure assessments, and the MYP defines clear
research goals and serves as the implementation tool for the ATRS.
ORD air toxics exposure assessment activities include research in four main areas:
Source characterization characterizes emissions from sources of air toxics and identifies source
contributions of measured concentrations. ORD is currently measuring emissions to be used in
developing emission factors and source profiles; these will eventually help to improve emissions
inventories. Source apportionment models are also being developed for use in deriving source
apportionment data and tools.
Research is also being conducted on several aspects of atmospheric fate. In the field of
atmospheric chemistry, the current focus is on characterizing the chemical and physical
processes that impact the fate of air toxics. Literature is being reviewed to determine the
chemical mechanisms for the thirty-three urban air toxics, and will ultimately yield chemical
algorithms for incorporation into air quality models. The Community Multiscale Air Quality
(CMAQ) Modeling System was developed in response to a need for a modeling system that
estimates the dispersion and deposition of air toxics at multiple scales (for additional information
and CMAQ website see the summary of Session III in this report). Benzene, formaldehyde,
acetaldehyde, mercury, and dioxin will be incorporated into CMAQ in the near future, and one
or two additional air toxics will be added each year. Measurements of ambient concentrations of
persistent toxics are also used to characterize atmospheric processes that affect the fate of air
toxics. High altitude mercury monitoring, and the National Dioxin Air Monitoring Network
(ND AMN) are two current activities in the area of atmospheric measurements. Data will be
used to understand the fate and long range transport of persistent toxics, and to incorporate into
atmospheric models. In addition, measurement methods are being refined and new methods
developed for measuring air toxics, both in ambient air and for human exposure studies (for
additional information on air toxics measurement methods see the summary of Session II in this
report). The method for measuring acrolein and other carbonyls will be available in the near
future and will be used in the air toxics ambient monitoring network.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
There is also ongoing research in the area of personal exposure to characterize the relationships
among ambient, outdoor, indoor, and personal concentrations, to identify the factors which
influence personal exposure, and as a means of measuring indirect exposures to persistent toxics.
Current work includes both measuring and modeling human exposure to air toxics, as well as
measuring concentrations of persistent toxics in food. Once data have been collected they will
be added to the National Exposure Research Laboratory's (NERL) Human Exposure Database
System (HEDS) (for additional information on HEDS see the summary of Session III in this
report).
Dose-to-target tissue studies are also underway to better characterize exposure-dose-response
relationships, and to incorporate those relationships into dose modeling. Results will be used in
dose response assessments and in the development of an integrated "source-to-dose" human
exposure model.
The importance of matching reported health effects to exposure information was emphasized, as
incompatible exposure and health information can impair risk assessment. Exposure assessment
should consider the health risks of concern - e.g., acute vs. chronic effects, or reproductive
effects. Diagrams were presented to show the impacts and relevance of ORD's exposure
assessment research [slide 16] and to illustrate the design, progress, clients, outputs and
goals/outcomes of ORD's air toxics research program.
Watkins concluded by stating that this workshop would be a valuable opportunity to provide
information on ORD's research to clients and end users, and to connect with regional scientists
to gain a better understanding of their exposure assessment research needs.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
SESSION I: DESIGNING AN AIR TOXICS EXPOSURE ASSESSMENT -
MONITORING vs. MODELING
Co-chairs: Ken Mitchell (Region 4), Paul Shapiro (ORD/NCER), and Ted Palma (OAR/OAQPS)
Paul Shapiro introduced the first session as an opportunity to give participants a way of viewing
air toxics problems in context, and addressing the issues of monitoring and modeling. The Air
Research Coordination Team (RCT), headed by Bob Fegley (ORD/OSP), meets and plans
throughout the year; the RCT welcomes regional involvement in prioritizing science. The
website for the National Center for Environmental Research (NCER) is another useful resource:
http://www.epa.gov/ncer
Shapiro also indicated that the case studies file provided in the binder would become a "living
document" and a resource for participants to be used beyond the end of this workshop. Finally,
he informed participants that Session I would begin with an interactive exercise, then proceed to
presentations of case studies, and end with an expert panel discussion.
Monitoring vs. Modeling — An Interactive Group Exercise - Paul Shapiro
(ORD/NCER) and Ted Palma (OAR/OAQPS)
An overview of modeling and monitoring outlined the typical usage, purposes, and strengths and
weaknesses of both techniques, and several points to consider when deciding which technique to
use.
The problem scenario presented was an increase of nasal and lung irritation and headaches in a
populated community located near a newly built highway and a chrome electroplating facility.
A map [slide 8] illustrated specific locations, relative distances, and the direction of prevailing
winds. Five randomly selected participants represented a planning team consisting of a
monitoring expert, a modeling expert, a meteorologist, a community relations expert, and a data
analyst. The planning team was allotted a budget of $100,000 to address the problem and a time
frame of ninety days, after which a community meeting would be held. The planning team was
also given a price list for services provided by a fictional environmental consulting company
("S&P"), and was tasked with deciding on a combination of modeling and monitoring services
based on their budget, time frame, and goal for the assessment.
At the end of their allotted time, the planning team met with a community reaction panel,
comprised of five other randomly chosen participants and consisting of: the mother of a child
with asthma; a City Council Member; a hospital director; the owner of the chrome plating
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
company; and a representative from a community environmental organization. Following the
simulated "meeting", the community members filled out a community scorecard to rate their
satisfaction with the planning team's assessment and communication of results. The results of
the "scorecard" are presented below:
1 . Confidence Identified Problem
2. Allayed My Fears
3. Know Who Is At Risk
4. Know What to Do to Protect
Them
5. Know How to Control Problem
6. Was modeling data adequate to
satisfy your needs
7. Was Monitoring data adequate
to satisfy your needs
8. Was balance of Monitoring and
Modeling appropriate
9: Know Next Steps
10. S&P Earned Their Fee
1 (low)
X
X
NA
X
2
X
X
X
3
X
X
4
X
5 (High)
Conclusions from the exercise emphasized that each situation is unique, and that competing
needs and interests, limited resources, and varying technical capabilities all play important roles
in real-life scenarios. As evidenced by the simulated community meeting, communication of
results to the public can be as important as the science. Participants were reminded to keep these
issues in mind throughout the remainder of the workshop.
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Case Studies to Illustrate Uses of Monitoring and Modeling
MATES II - A Regional Perspective - Mike Nazemi (South Coast Air Quality
Management District)
The results of the Multiple Air Toxics Exposure Study (MATES II) were presented. South Coast
Air Quality Management's (AQMD's) governing board adopted guiding principles in 1997, and
established a set of environmental justice initiatives [slides 3,4]. The MATES II study was
conducted as part of the "ambient monitoring of air toxics initiative" and was a follow-up of
MATES I, a similar study conducted a decade ago. There are three principal components to
MATES II: 1) toxic air contaminant monitoring in ten fixed locations in the South Coast Air
Basin; 2) a Basin-wide emissions inventory, developed for 1998; 3) dispersion modeling and risk
assessment, used to construct a full-Basin picture of inhalation risk. A panel of experts from
academia, environmental groups, industry, and government agencies served as a review board
and provided technical direction for the study.
Monitoring was conducted for one year at ten sites, based on EPA Guidelines (Neighborhood
Scale Monitoring). Additional sampling was done in fourteen communities using mobile
monitoring platforms. In all, thirty toxic pollutants were measured. The locations of monitoring
sites, equipment used, a complete list of chemicals monitored, and the laboratory procedures
used for analysis were reported [slides 9-12]. Results from MATES II, when compared with
MATES I, indicated that cancer risks in the Basin have decreased dramatically over the last ten
years [slide 10]. Three monitoring sites common to both studies were used for comparison: Los
Angeles (LA), Long Beach (LB), and Rubidoux (RU). Only pollutants common to both studies
were included in calculating accumulated risks. Cancer risks since MATES-I have decreased by
76% in Los Angeles, by 73% in Long Beach, and by 55% in Rubidoux. These and similar risk
results in other Basin locations were illustrated in slides 13 through 21. Cancer risks for each
location were also aggregated into stationary and mobile sources, highlighting the importance of
diesel particulate to the total inhalation cancer risk [slides 22-25]. Seasonal variation was shown
to be significant in risk levels attributed to mobile sources, but not in those from stationary
sources [slides 26, 27].
The MATES II emissions inventory included on-road mobile sources, area and off-road mobile
sources, and more than 2,588 major point sources. The Direct Travel Impact Model (DTIM) was
used to provide hourly gridded emissions and speciation profiles. Stationary sources' emissions
information was collected from Annual Emissions Reporting and from toxic hot spots, and
allocated spatially and temporally. Dry cleaners, gasoline stations, auto body shops, and chrome
platers were the most important contributors to stationary source emissions. Locations of these
facilities in the Basin were correlated with emissions of perchloroethylene, benzene, and
9
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
hexavalent chromium [slides 32-45]. Results were presented by species and by source [slides
46-49], both with and without diesel paniculate matter (the most significant factor in toxic
emissions). Distribution of diesel emissions plotted on a Basin map revealed high
concentrations along roads and shipping lanes [slide 50]. Annual average day emissions were
summarized for diesel paniculate matter, benzene, and 1,3, butadiene for on-road, off-road, and
stationary sources [slide 51].
Meteorological and dispersion modeling were used to model concentrations, and risk assessment
estimated the cancer risk from nearly thirty air toxics in a simulation of a full year. Model
performance was checked against measured concentrations for gases and particulates [slides 58-
59]. Office of Environmental Health Hazard Assessment (OEHHA) unit risk factors (URFs)
were used, and assumptions included a seventy-year lifetime exposure by inhalation risk only,
and the premise that cancer risks are additive. Model results, presented on a map [slide 61],
indicated higher cancer risk in the south-central Los Angeles area, the harbor area, and near
freeways. The model results were similar to monitoring results in their estimates of cancer risk,
and in attributing a high percentage of that risk to mobile sources [slides 62, 63]. The model
generally exhibited an under-prediction bias, possibly due to underestimated on-road emissions.
Nearly ninety percent of the cancer risk was attributed to diesel paniculate matter (PM),
benzene, and 1,3, butadiene (combined). Mobile sources accounted for more than 98%, 90%,
and 95% of the diesel PM, benzene, and 1,3, butadiene, respectively. The modeling study
demonstrated that regional modeling tools can be effectively used for a regional risk assessment.
A micro-scale monitoring study was also conducted in an attempt to capture "hot spots" not
evident by the fixed monitoring, to confirm hot spots revealed by modeling, to respond to public
concerns, and to assess to what extent monitoring was locally representative. Consistent with
the MATES II study results, diesel and other mobile sources are the dominant contributors to
inhalation cancer risk. In one site (Anaheim), local emissions of styrene were considerably
higher than predicted by the model; three styrene-emitting facilities were found to be nearby,
although outside the area considered by the model. A strategy was proposed for directly or
indirectly controlling the emissions of air toxics, and model results were presented predicting the
impact of its implementation [slides 73, 74].
Uncertainties inherent in the MATES II and micro-scale studies included the indirect
measurement of diesel PM, determination of risk values, laboratory and measurement processes,
and model inputs and computational algorithms. However, analysis of the results to reveal
sensitivities to methods and to diesel PM toxicity showed a consistent pattern [slides 81, 82].
10
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Questions and Comments
Question: How did you get the numbers presented on the last slide [Species Apportionment -
Sensitivity to Diesel PM Toxicity]?
Response: By doing a comparative analysis using weighted emissions of the different pollutants.
Question: Why did the model not work in Anaheim?
Response: A group of experts chose the site to be used for modeling, and determined a "wedge
of influence" for known emissions sites. In this case, the sites responsible for the
styrene emissions were outside this wedge of influence, yet it was found that they
were contributing to area emissions.
Question: Have you thought about gasoline paniculate matter (PM)?
Response: Gasoline PM was not measured, so we do not have any information on it.
Question: Did you look at non-cancer endpoints?
Response: No, they were not considered in the study. The study was limited by the available
toxicity data.
11
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Minneapolis - St. Paul - A Comparison of Community, Residential, and
Personal Exposure - John Adgate (University of Minnesota)
A study performed in the Minneapolis-St.Paul area compared outdoor, indoor, and personal
exposures to particulate matter (PM, 5) volatile organic carbons (VOCs), and compared model
results with the monitoring results. The presentation focused on VOCs, as there are no results
yet for the PM2 5 study. Three neighborhoods, each with PM and VOC canister monitoring sites,
were used in the study; personal monitoring was conducted with organic vapor badge monitors
worn by seventy subjects (non-smokers). Several toxics were measured by badges and
monitoring sites, and/or were modeled [slide 7]. EPA's Industrial Source Complex (ISCST3)
model was used for air dispersion modeling, along with meteorological data. Modeled times
were fifty-eight 48-hour periods, to correspond with the measurements, and modeling scale was
at the census tract level.
Point sources, mobile sources, and area sources were modeled. Point sources were major
stationary sources that were inventoried individually; emissions of eighty-two pollutants from
these sources were modeled using the Region Air Pollutant Inventory Development System
(RAPIDS) database. On-road and non-road mobile sources were modeled using traffic and road
data by census tract and emission factors from RAPIDS. Emissions were assigned to census
tracts, and modeled as area sources. Rail and air mobile sources were also considered by
apportioning airport-related and rail emissions from RAPIDS to census tracts near the airport
and along rail lines. Detailed lists of the specific area source categories were presented [slides
13,14]. A summary of the emissions inventory values for point source, mobile source, and area
source emissions showed the mobile sources to be dominant, followed by the area sources [slide
15]. Benzene emissions were also illustrated on maps of the three neighborhoods, with the
locations of major point sources and of the study participants' homes superimposed [slides 15-
18]. Modeled benzene concentrations in the three neighborhoods [slides 19,20] showed the
highest values as well as higher variability in the Phillips area - the one surrounded by the most
highways (see map on slide 17). Sources used for modeling benzene concentrations were
dominated by mobile sources (89%) [slide 21]. Overall, personal benzene levels were higher
than indoor levels, which were higher than outdoor levels [slide 22]. Modeled benzene
concentrations were also compared to measurements by all four measuring techniques/types for
all three neighborhoods [slides 23-26]. The ISCST3 model reasonably predicted outdoor VOC
concentrations in two of three communities, likely because the emissions inventory is more
accurate and/or the area sources less complex. The model was less reliable in the Phillips
neighborhood, where mobile sources dominated. The model appears to over-predict low
concentrations and under-predict high concentrations. And, as expected, it fails to predict the
high VOC concentrations found in indoor and personal air.
12
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Questions and Comments
Question: Did you consider emissions from apartment buildings?
Response: Yes, but we do not know what these show.
Question: Why were personal exposures greater than both indoor and outdoor air
concentrations?
Response: People are exposed to a lot of VOC sources (e.g., when driving, using consumer
products, hobby products) and are in close proximity to those products while using
them. This "personal cloud" effect was also obvious in the PM experiment.
Comment: As an example, a very high exposure in one study was attributed to chemicals used by
that person while making jewelry.
Comment: Personal peaks should be considered to determine health effects levels for VOCs.
Response: Yet we still get one number for the whole year from emissions inventories; this is the
major challenge to the inventory data.
Question: What do you propose adds to the outdoor contribution [to make indoor and personal
levels so much higher]?
Response: Personal activities; for example, covered garages and cigarette smoking are two of
the major sources.
Comment: Cigarette smoke can be a big contributor even as second-hand exposure.
Question: How did you use the badges to get better detection limits?
Response: The badges are standard 3M™ badges; however, the analysis done by our lab is
different, and can give much lower detection limits.
Question: Do you take repeat measurements, and what variation did you observe between repeat
measurements?
Response: We do have repeat measurements, but have not yet finished the analysis to determine
the amount of variation between them.
Question: Was the personal sampling non-working exposure?
Response: There was some working exposure, since the badges were worn for forty-eight hours
each time. Some people showed significant work exposures, particularly for PM.
Question: What is your opinion of where the Clean Air Act is heading, especially since it is
focused on outdoor air? Can we extrapolate this to the nation as a whole using
monitoring or modeling?
13
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Response: We need both processes; right now we are trying to track indoor vs. outdoor exposure
over time.
Question: Do you plan to do any spatial extrapolation, i.e., will you extrapolate from your three
areas, or were you interested only in those areas?
Response: A little of both; although I agree that applicability to other parts of the country is an
important question.
Comment: That area is a "red spot" on the National map.
14
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Session I Expert Panel Discussion — Mike Nazemi (SCAQMD), John Adgate (UMN),
John Girman (OAR/ORIA), Larry Cupitt (ORD/NERL), Neil Frank (OAR/OAQPS), Matt
Lorber (ORD/NCEA), Joe Touma (OAR/OAQPS)
John Girman (OAR/ORIA) - Monitoring and Modeling Indoor Air Toxics
Monitoring is probably used more frequently than modeling for indoor air toxics exposure
assessment, since monitoring data is generally more available than mode! inputs. Mass balance
modeling is used to analyze experimental results and to assess concentration and ventilation
studies. Good mass balance models do exist for indoor air, such as the CONTAMW model
developed by the National Institute of Standards and Technology (NIST), and EPA's RISK
model, developed by the National Risk Management Research Laboratory (NRMRL). Many
monitoring studies are also being conducted on indoor air in office buildings, schools, and
residential buildings. Research needs still exist for both modeling and monitoring indoor air
toxics.
Monitoring needs include:
• Study design
• Sampling equipment
• Analytical support
• Protocols
Quality Assurance/Quality Control (QA/QC)
• Sample protocols (access)
• Time resolution
Modeling needs include:
• Models
• Ventilation rates
• Building volumes (multi-chamber)
• Activity patterns
• Emission rates (a big data "gap")
• Loss rates (deposition, reactions, filtering)
• Outdoor-to-indoor penetration factors
15
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Larry Cupitt (ORD/NERL)
Exposure to air toxics consists of contact with a chemical or stressor integrated over time;
chemical concentrations and sources cannot be separated from time. For some compounds,
indicators exist that can estimate exposure; persistent pollutants, for example, can be measured
in blood samples. Other chemicals, however, degrade or pass through biological systems very
quickly.
Over the coming years there will be more of this type of data, as methods are devised to measure
chemical concentration in the blood or urine. In some cases, sources can be attributed to these
results from known precursors. As such tests become available, it will become necessary to
explain their results to the public in terms of what the results mean and where the chemical(s)
may be coming from.
Neil Frank (OAR/OAQPS) - Monitoring Tools to Support Exposure Assessment at Various
Geographic Scales
The role of ambient monitoring in exposure assessment is to assess community-wide
concentrations, and to quantify ambient conditions in the vicinity of localized hot spots or areas
of concern (e.g., schools). There are limitations to ambient monitoring, however. Monitors do
not directly estimate human inhalation exposure, and inhalation is only one of the possible
exposure routes. Numerous monitoring stations may be needed to capture local structure,
making cost a potential problem. Monitoring cannot give an accurate picture of exposure over
tune, unless combined with modeling. Ambient monitors can be used both to support models by
providing data (e.g., outdoor concentrations) and to evaluate the results of dispersion and
deposition models.
The design of an air toxics monitoring network should match the geographic scale of the
analysis, as well as the intended data use. Regional/national scale monitoring requires
representative community-wide monitoring locations; measurements of typical source impacts;
and appropriate climatological and emissions regimes. A regional or national scale study should
monitor year-round for one year or longer, with samples taken for twenty-four hours at
intermittent intervals. Urban- or local-scale monitoring requires a higher density of sites, both
community-wide and at known or suspected hot spots. Measurements can be taken year-round
or for shorter periods of time, and sampling can be intermittent or semi-continuous (using less
than twenty-four-hour intervals).
Maps were presented showing the location of air toxics monitoring sites across the country
[slides 5,6], although it was noted that not all of the sites run year-round, and even fewer are set
to run for several years. Scatter plots demonstrated the alignment of monitoring with modeling
16
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
results for benzene, formaldehyde, and chromium [slides 7-9]. Data was also presented from
ninety-five benzene sites with sufficient monitoring data to derive a national trend from 1994 to
2000 [slides 10,11]. A significant drop in benzene emissions is evident from 1994 to 1996,
illustrating another use of monitoring data: they can reveal trends to be used as surrogates for
changes in exposure, and for evaluating the effects of changes in emissions regulations.
Matt Lorber (ORD/NCEA) - Comments on the Columbus Waste-to-Energy Site
Monitoring and modeling tools can be used alone or in combination to support exposure
assessments. In the context of dioxin reassessment, stack monitoring and modeling have been
used to conduct a risk assessment to support regulation. Monitoring includes stacks, ambient air,
ash, and soil, and can be used to evaluate the relationship between stack emissions and
environmental impacts. These measurements also serve to validate relevant models.
A map illustrated the positions of air and soil sampling sites in the vicinity of a dioxin-emitting
facility near Columbus, Ohio. Stack emissions and on-site soil samples exhibit similar profiles,
characterized by large contributions by the same three dioxin congeners. Soil samples taken off-
site show a very different profile, matching the congener profile for representative background
levels [slide 5].
Data from the same study was also used to validate the ISCST3 model and a soil deposition
model using 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) measurements. Figures were
presented showing concentration isolines near the facility as predicted by the model compared
with observed values [slides 6, 7] to indicate the degree of agreement in results.
Joe Touma (OAR/OAQPS) - Monitoring and Modeling
Monitoring and modeling can be used in a complementary manner in air toxics exposure
assessments, with due regards for the strengths and weaknesses of both techniques. Monitoring
can be used to check the results of models, or to estimate background concentrations. Source
configuration, terrain, and meteorological variations need to be considered when deciding the
number of monitoring sites required. Other considerations when choosing monitoring include:
The existence of a monitoring network for pollutants of interest, and whether it can
be used at the times of concern;
Whether the network, data reduction, and storage meet quality assurance
requirements;
Whether the network will accurately represent the impact of the most important
individual source; and
17
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
• Whether there is at least one full year of data available.
Questions and Comments
Question: There is uncertainty inherent in comparisons of modeling and monitoring due to
different data sets, or types of data. Have you considered combining modeling and
monitoring data, instead of always using monitoring to validate the models?
Response: There was a case study in Japan that provides a good example of using the two in
conjunction. A U.S. naval air facility was located near infectious waste incinerators,
for which no emissions data were provided by Japan; modeling was not possible
without such data. Models were eventually used only to determine where emissions
monitors should be placed.
Comment: Models have also been used to determine where to look (monitor) for radon
contamination.
Comment: There are higher levels of certainty when you combine a model with monitoring data;
uncertainty is expected when monitoring data are not available.
Question: The Agency has QA/QC procedures in place for monitoring, but they are not
available for modeling. What are the QA steps needed for a modeling study?
Response: Input data have a lot of bearing on the accuracy of model outputs, and can be used as
a form of QA. Models are also evaluated using "real-world" data.
Comment: The Agency does have some QA/QC guidelines on models, but the question is
whether they are sufficient.
Comment: Quality assurance is also built into every grant that we award.
Question: Since we need information on exposure, and current models are more advanced than
those of a decade ago, is there any work being done to use these new models and look
backwards at what exposures people might have had in the past?
Response: We do have ways of using models for "back-casting", and have done so with the
emissions inventory. Resource limitations can be a problem, however, so this is not
done extensively. There has not been a real need for back-casting.
Comment: The Superfund program does this regularly to estimate past exposure and determine
whether it can be linked to current health effects. In Superfund sites it is not unusual
for communities to ask for a back-casting, especially if health effects do exist.
Question: Are you referring to estimating, e.g., what the annual risk was fifteen years ago, or
what the total risk has been since then?
Response: I was referring to cumulative risk (the lifetime exposure of a person), since incidence
of disease can be predicted based on past exposure.
Comment: Almost all of us realize that we need to look at both.
Comment: As we build community assessment programs, constraints may limit us to only doing
one or the other, especially at the state or local level.
18
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
SESSION II: MONITORING METHODS AND NETWORK DESIGN
Co-Chairs: Motria Poshyvanyk (Region 5), Neil Frank (OAR/OAQPS), and
Tim Watkins (ORD/NERL)
Motria Poshyvanyk (Region 5) began the monitoring session by thanking the presenters and
asking participants to think about the design of a monitoring study. Although monitoring data
are considered "real" data, they should be examined critically, as one would with model outputs.
Air Toxics Monitoring Pilot Project - Barbara Morin (Rhode Island Department of
Environmental Management)
The Air Toxics Monitoring Pilot Project is intended to help design a national network as well as
to verify modeling results. EPA funded a pilot study in each of the EPA regions: four in urban
areas including Providence, Rhode Island, and six in smaller cities and rural areas. The pilots
were conducted for one year using a one-in-twelve days sampling frequency, standardized
methods, and measuring at least seventeen specified core pollutants: nine VOCs, two carbonyls,
and six metals.
The Rhode Island study used five neighborhood scale sites: an existing photochemical
assessment and monitoring stations (PAMS) and PM2 5 site and a second PM2 5 site located in an
urban area; a site adjacent to 1-95; and two other sites in residential/industrial areas. Nine core
VOCs and seven "max" VOCs were measured, as were 16 other VOC hazardous air pollutants
(HAPs), six metals (not speciated), and three carbonyls (listed on slides 7-10). All sites were
operational for one year, with two sites continuing for a second year, and one for a third year.
All sites measured for all pollutants over a twenty-four-hour period every sixth day; additional
sampling was done at some of the sites to capture any diurnal variations. The methods used for
each group of chemicals were listed [slides 13-16]. Analysis of preliminary data resulted in a
change in method for metals, since most were present in concentrations above the minimum
detection limits (MDL) of a less costly method. A problem observed with metals was the
presence of background concentrations in the monitors' quartz filters. Chromium had the
highest background concentrations, at 55-71% - more than half of the measured concentration.
Chromium also had the highest risk, assuming one hundred percent was hexavalent chromium.
The mobile source site exhibited chromium levels above the cancer health effect benchmark.
Carbonyl results showed little spatial variability but tended to vary seasonally; none approached
the cancer health effect benchmarks. The highest-risk VOCs measured did exceed the risk
benchmarks (obtained from EPA and the state of California) [slides 27, 28]; two were from
mobile sources, two from background, and two from stationary sources. Benzene and 1,3
butadiene were highest at the site near the highway and lowest at the rural site.
19
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Tetrachloroethylene reached very high levels during a six-week period in November 2001,
although the reason could not be identified due to the time lag involved in processing samples.
Wind direction is being considered to try to locate a source, and sector sampling may be used.
It became apparent during the course of the project that improved methods are needed to
measure acrolein, acrylonitrile, ethylene oxide, diesel, and arsenic. In addition, technology that
would allow continuous monitoring of formaldehyde may be useful; formaldehyde peaks came
close to the short-term benchmarks for non-cancer health effects, and continuous analysis would
yield more information.
Questions and Comments
Question: What was the issue with the acrylonitrile method?
Response: Many of the results did not match.
Question: Do you look at the same benchmark for all human populations?
Response: No, we really look at exposures.
Question: Were the monitors located at breathing zones?
Response: No, most were on the tops of one-story buildings, which is a little higher than
breathing zones.
Question: Why do we still use the Total Suspended Paniculate (TSP) method?
Response: This was part of the study's set of criteria. It can be useful if you are interested in
more than the inhalation route of exposure, particularly for metals.
Comment: TSP is also relevant for measuring ecological exposures.
Comment: We should think about whether resources should continue to be invested in TSP
analysis.
Comment: If we do not have a good way to use data, then there is little sense in spending
resources to collect it.
Question: How high was the tetrachloroethylene peak?
Response: Between 1.0 and 1.5 parts per billion (ppb).
20
-------
US Environmental Protection Agency
Rcgion/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
New Trends in Monitoring Methods - Don Whitaker (ORD/NERL)
Presented by Tim Watkins (ORD/NERL) for Don Whitaker
A significant issue in discussing current trends in monitoring methods is determining which
compounds need to be monitored. The Clean Air Act Amendments of 1990 helped define these
by listing 188 hazardous air pollutants (HAPs). This was further refined in the Urban Air Toxics
Strategy to thirty-three compounds believed to present the greatest public health risk - known as
the urban air toxics. Methods for measuring the thirty-three urban air toxics exist, although
some exhibit problems such as poor sensitivity or limit of detection, stability, or recovery
problems. Other methods are too costly, time-consuming, or difficult to use on a large scale
basis. The Office of Research and Development (ORD) is working towards filling some of these
monitoring gaps. Because of limited resources, however, research must be targeted, preferably
toward multiple-use methods (i.e., methods for both ambient and personal monitoring), or those
with known shortcomings. Currently, there are approximately three hundred sites nationwide
collecting ambient data on selected HAPs. EPA will also be conducting air toxics human
exposure studies. Three categories of issues need to be considered when assessing monitoring
methods: analytical, economic, and "ease of use" issues [slides 8-10].
The plans for the National Air Toxics Monitoring network provide insight for current ambient
monitoring methods needs. This is a national network planned to be completed over several
years, with the objective to estimate annual average pollutant values and associated trends. Pilot
sites are currently monitoring eighteen "core" volatile organic carbons (VOCs), carbonyls, and
metals [slide 12]. Nine additional pollutants were also desired for inclusion, but problems with
methods prevented them from being included in the "core" list [slide 14]. Current efforts in
methods development are focused on three of these nine compounds: acrolein, formaldehyde,
and 1,1 -dichloroethylene (DCE) (CAS No. 75-35-4) and other VOCs.
The passive aldehyde and ketone sampler (PAKS) method for acrolein was developed for
personal and residential monitoring, and is a passive diffusion sampler analyzed by high pressure
liquid chromatography (HPLC)-Fluorescence. Advantages of this method include improved
sensitivity, reasonable cost, and better recoveries of acrolein and crotonaldehyde. Acrolein
recovery is lower than desired, however, and cartridge stability needs to be improved. These
problems will be addressed in proposed method development, to include field verification and
comparison with the existing 2,4-dinitrophenyl hydrazine (DNPH) method [slides 16-21].
The semi-continuous formaldehyde monitor was developed with EPA support by Texas Tech
University for application to air toxics monitoring and photochemical modeling. The monitor
can measure formaldehyde concentrations from 0.1 to 50 ppb, runs one complete cycle in ten
minutes, and can operate for seven days without need for consumables, among other features
21
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
[slides 25,26]. It is less labor-intensive and less expensive than the current methods, and allows
collection of more data points for modeling purposes. It can only monitor for one compound,
however, and its size and power requirements limit its applicability to ambient site or stationary
monitoring [slide 29-30].
Improvement of the monitoring capabilities for 1,1,-DCE and other VOCs is also underway to
help with ORD's Regional Monitoring Initiative (RMI): Under the RMI program, EPA Region 8
requested support to develop a method that would attain lower detection limits for 1,1,-DCE and
other VOCs, and to review canister-based sampling and analysis methods for these compounds
(listed on slide 34). Preliminary results showed that the methods can now detect concentrations
in parts per trillion, compared with the previous method's parts per billion. The cleanliness of
canisters and adequate standards were also found to be important factors in the accuracy of
results.
ORD's work is addressing known problems with methods, in order to increase the accuracy of
monitored results. Methods also need to be less expensive and less labor-intensive. In addition,
methods applicable to ambient, personal, and residential monitoring can facilitate comparisons.
Questions and Comments
Question: Are there any plans to increase ORD's focus on method development?
Response: We plan research a few years in advance through our planning process. We would
like to develop a methods program, but it is difficult to change research directions
quickly. Although we (ORD) are aware of the need for methods.
Question: Can these methods be calibrated depending on the environment where the monitoring
would take place (e.g., indoor vs. outdoor)?
Response: I cannot answer this specifically. Don Whitaker and Bill McClenny can answer
specific questions about these methods.
Question: For the DCE method, who decided how low the limit of detection should be? It is
important for a monitoring person designing a study to know that.
Response: The Office of Air Quality Planning and Standards (OAQPS) has developed a
hierarchy of data sources for toxicological data, including the detection limits that
methods should meet (or measure below).
Comment: Regarding the existing carbonyl method (DNPH), the regions have some suggestions
for ORD, if there are plans to update these documents.
Response: Historically we do ask the regions for comments. However, we have lost the funding
and full-time equivalents (FTEs) related to that, so there are no plans right now to
update it.
22
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Atmospheric Formation and Decay of Air Toxics - Implications for Exposure
Assessments — Deborah Luecken (ORD/NERL)
Chemical processes can lead to the formation of compounds in the atmosphere, or their decay
through chemical reactions. Since this can affect the concentrations of air toxics, atmospheric
chemistry may have important implications for both modeling studies and monitoring network
design.
Pollutant decay can be quantified by using a chemical's half life or lifetime [slide 3]. Decay
processes can result from reactions with the hydroxide (OH) radical, ozone (O3), or the nitrate
(NO3) radical. Reactions with OH can affect almost every pollutant, and this is usually the most
important reaction during daytime hours. Reaction rates are dependent on both temperature and
OH concentration, so routine measurements of OH are necessary for this process to be taken into
account. Reactions with ozone are less important than those with OH, and usually only affect
those compounds with double bonds. Ozone concentrations can vary both throughout the day
and seasonally, being higher in summer than in winter. Reactions with the NO3 radical are
important only at night, and only for a few species. Other decay processes that occur in the
atmosphere include photolysis in sunlight — which can be important for some compounds during
the day, but highly variable; and processes involving other reactants in the gaseous or liquid
phase. Slide 9 lists some of the common air pollutants and the atmospheric processes important
in their decay.
Atmospheric processes can also lead to the production of air toxics, some of which can be
produced by other air toxics or VOCs. Atmospheric formation can transform one state of a toxic
to another and is usually only important for certain species. However, it can be a major source
of formaldehyde, acetaldehyde, and acrolein. Both formaldehyde and acetaldehyde can be
formed from any VOC in the atmosphere, and it is estimated that 85-99% of these aldehydes are
due to atmospheric formation, rather than emissions [slides 11,12]. Acrolein can be formed
from the decay of 1,3-dienes through cleavage of the double bond [slide 14]. Other air toxics,
potential air toxics, and some metals can also be formed from atmospheric reactions; thirty of the
188 air toxics listed in the Clean Air Act can be formed in the atmosphere from secondary
sources.
Since atmospheric chemistry can be highly variable, it should be taken into account in
monitoring studies. Monitors could be placed downwind of large sources, and should measure
the major aldehyde precursors, even if they are not themselves air toxics. Atmospheric
chemistry should also be considered in modeling, as it can significantly affect the predicted
concentrations of some chemicals. However, these reactions are not adequately verified for all
pollutants. Atmospheric reactions can also form compounds that may be harmful or important in
23
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
processes such as deposition to water or soil; reactive species which can increase ozone
formation; semi-volatile species, or greenhouse gases. Controls in one type of air toxic should
impact other pollutants as well. Further research is needed to improve understanding of air
toxics chemistry, including the chemistry of transition metals and their products. Better
evaluation of model predictions for formaldehyde and acetaldehyde is also needed, as is an
improved understanding of the toxicity of photochemically-produced compounds. Atmospheric
chemistry would be especially useful to include into air quality models, over long periods of
time and large domains.
Questions and Comments
Question: What are the differences and similarities in indoor and outdoor environments?
Response: No photochemical processes can occur indoors, since they require dissolution
reactions and the formation of radicals. It may be possible for ozone reactions to
occur indoors, however.
Comment: A method has been developed for quantifying the amount of formaldehyde converted
to acetaldehyde.
Comment: Defining the concentration of the hydroxide radical is really difficult, because we
have no good measurements. Models could be used to estimate these concentrations.
Question: There are ways to simplify the mechanisms of these reactions. How detailed does the
chemistry need to be for the purposes of modeling?
Response: You should measure everything, if possible, since every simplification will add a
measure of uncertainty. This also depends on the compounds and reactions, and how
much uncertainty you can accept.
Question: I have never seen precursor data used for anything. How should we use such data?
Response: It can be useful for evaluation of models for ozone, as well as for PM and some other
compounds.
Comment: The expense involved in this can be a problem.
Response: However, if such data has been collected, it should be used.
Question: Some states have mentioned that they collect a lot of data that is never used or looked
at. Why is this data not used?
Response: Sometimes it is collected as part of a regulatory requirement - for the Superfund
program, for example. It would take a lot of work and a lot of people to be able to do
something with that data.
24
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Air Toxics Monitoring Methods and Network Design - Steve Bortnick and Shannon
Stetzer (Battelle Memorial Institute)
Data from the Urban Air Toxics Monitoring Program (UATMP) for VOCs and carbonyls
spanning the years 1996-1999 were analyzed to identify and quantify sources of variability.
Duplicate and replicate samples were used including multiple sites, multiple days within sites,
duplicate samples on each day at the same site, and replicate analyses of each collected sample.
Sources of variability included spatial, temporal, sampling, and analytical variability. Graphs
were presented showing the data in relation to combinations of these sources [slides 5-10].
Temporal variability was the driver for most data variation, and in many cases environmental
variability (i.e., ambient or natural variation) was more significant than monitoring uncertainty
(i.e., measurement of man-made variation). At low ambient levels, however, environmental
components of variability tended to decrease, and monitoring uncertainty, particularly analytical
relative error, would take over.
Data were plotted to show variation as it related to annual concentration means for benzene,
acetaldehyde, and manganese [slides 14-16] and equations presented to demonstrate how the
relationship between variance and concentration can be used to estimate optimal sampling
frequency [slide 13]. Annual concentrations were also plotted against the annual average errors
[slides 18-23] and used to devise a formula for calculating precision as a function of sampling
frequency and mean ambient level.
Bias, or systematic differences, related to the monitoring technologies used was also analyzed,
and twenty-four-hour canister benzene monitors were found to record consistently higher
concentrations than PAMS monitors.
Spatial and inter-compound correlations were also examined. Using data from the MATES II
study, measurements were matched for the same compound and on the same days at different
sites, for a total of forty-five site-to-site combinations per compound. Correlation coefficients
were calculated for each compound, as was the distance between sites. Correlations were
presented against the distance between sites from north to south and from west to east [slides 32-
35]. Compound to compound correlations were calculated for twenty-one compound pairings
per site, including carbonyls vs. other classes, metals vs. other classes, and VOCs vs. other
classes [slides 38-40]. A positive correlation was found between formaldehyde and
acetaldehyde concentrations; such comparisons may be performed with classes of chemicals in
the future, rather than using all 188 pollutants.
Four spatial case studies were presented. In Portland, Oregon, a study was designed to
demonstrate the effect of local point source emissions across a wide range of air toxics:
25
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
carbonyls, metals, and VOCs [slides 42-44]. Five sites were set up to take twenty-four-hour
samples every sixth day for one year. High spatial variability - well over 100 percent relative
error- was observed in compounds that had significant local sources. Variability decreased once
one such site was dropped. Nickel, manganese, and iron exhibited both high overall and high
spatial variability. In Iron County, Missouri, a study demonstrated the effects of controls (source
reductions) on monitoring needs [slide 45-47]. Four monitoring sites were present from 1993
through 1999, all surrounding a lead smelter. Very different values were obtained from the four
monitors from 1993 to 1996. After the implementation of controls in 1996, however, variability
dropped to the extent that it would be possible to only maintain one monitor instead of four. The
Shelby County, Tennessee case study [slides 48-50] demonstrated the impact of a local point
source on spatial variability and the geographic extent of the impact. The data analyzed were
from five sites, two of which were located with a half mile of a battery recycling plant. These
two sites were the major contributing sources of variability in measurements, and variability
dropped dramatically when the two sites were not included in the analyses. In addition, it was
determined that the impact of the plant's emissions became negligible at distances greater than
seven miles. The fourth case study, in Cook County, Illinois [slides 51-53], considered the
impact of local point sources and corresponding monitoring objectives on spatial variability.
Data were analyzed from seven sites, two of which were reported to be point source sites,
located near steel mills. As in the Tennessee study, variability among locations was reduced
when the two point source sites were excluded from the analysis. In addition, similar variability
was exhibited by other closely clustered sites.
The last set of data analyzed was obtained from the Air Toxics Pilot Study. Battelle Memorial
Institute is currently acquiring all data collected at each of the ten pilot cities including ambient
measurements, meteorology, and other ambient data. Following QA/QC the data will be
converted to appropriate data sets for analysis. The database should be complete in 2002 and
may be made available to the public by late 2003. Some issues with the data that have been
observed so far include numerous formats and reporting units, and numerous different
conventions of reporting data below the minimum detectable limits (MDL). Data analyses will
include inter-laboratory variability analysis, monitoring data variability analysis, and MDL and
reporting analysis. A draft report is planned for early 2003; the results will also be presented at
the National Air Toxics Workshop in spring 2003. To obtain more detailed information related
to the data analysis results presented and summarized above, e-mail Mr. Michael Koerber of the
Lake Michigan Air Directors Consortium for a copy of the draft report (ladco@ladco.orgX
Questions and Comments
Question: What are the conventional units?
Response: Most data is either collected in or converted to micrograms per cubic meter (• g/m3).
26
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Comment: In the latest study we were asked to report in those units to reflect local conditions.
Comment: The transformation of data can introduce a new component of uncertainty - we need
to be consistent across all data of interest.
Question: Didn't Battelle receive a contract to do that?
Response: We are trying to put together a database with high QA/QC and make it available
publicly, possibly on a website through the Aerometric Information Retrieval System
(AIRS) database.
Question: Are there atmospheric chemistry reasons why you would expect correlations between
formaldehyde and acetaldehyde?
Response: Yes, they come from the same sources.
Question: Does that eliminate the need for a separate aldehyde or acrolein method?
Response: I would not say that it eliminates that need; measuring both would provide a good
check on the data.
Comment: Similar correlations are seen with butadiene and benzene - they have the same
source. Research is being conducted to look at such correlations in a number of
chemicals.
Comment: We do not know that there is a good correlation between acetaldehyde and acrolein;
their sources are not the same, as with formaldehyde.
Comment: The contribution of various sources is an important issue in air toxics. Scattered
point sources may give localized high concentrations, but that is not necessarily a
correlation. The same issue exists with indoor versus outdoor measurements. It is
important to understand the distribution of exposure that results from the nature of the
source.
Question: What is your opinion of the PAMS program?
Response: It is appropriate for what it was intended to do.
Comment: Toxic endpoints of air toxics are moving towards non-cancer, short-term exposure
endpoints. We may be more interested in distributions than in annual averages.
These exposures may not be linear with respect to time.
Response: It is possible to get that information from the data; analysis will depend on the
objective of the study.
Comment: A method has been developed that can be used for source apportionment.
Comment: A source-receptor model is needed so that we can link sources to health effects.
27
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
California Monitoring Program: Statewide Network to the Neighborhood
Scale — Jeff Cook and Linda Murchison (California Air Resources Board)
Jeff Cook reported the preliminary results of a study intended to further the understanding of
network scale, size, and effectiveness. The monitoring program in California was established
when laws passed in the 1980s required the Air Resources Board (ARB) to identify and regulate
air toxic emissions. Monitoring began in 1985 and was eventually used to develop control
measures for air toxics. Progress since the beginning of the program has included the
development of new and modified methods, improvements in the limits of detection, and a
change in sampling media. N1ST standards are used as the primary standards for calibration; an
aggressive QA program includes a manual and quality control reports, inter-laboratory
comparisons in lieu of round-robin testing, and collocated sampler sites for precision purposes.
The number of air toxics contaminants reported has increased five-fold since 1985 [slide 7; list
of compounds on slide 8], and the current database contains approximately 50,000
determinations. Data are publicly available as annual summaries on the ARB website [slide 12],
including minimum, maximum, and mean concentrations, detection limits, number of
observations, and estimated cancer risk. The summaries show almost a seventy-five percent
drop in risk between 1990 and 2000.
Inter-site variation analyses considered benzene (a motor vehicle pollutant indicator),
perchloroethylene (a point source indicator), carbon tetrachloride (indicator of a globally
pervasive pollutant), and hexavalent chromium (a point source indicator not usually found in
ambient samples). These are four of the eight compounds which make up ninety-nine percent of
the ambient risk. Monitoring networks exist at the statewide, regional, and neighborhood scale
[slides 17-19]. Both variability and trends were examined by simple comparisons at the
statewide and regional levels.
Benzene means from statewide sites generally fell within a factor of two; a factor of four
separated the lowest from the highest concentrations. Peaks were evident in heavily industrial
locations, and near the Mexico border [slide 22]. Means were also comparable among the
regional sites [slide 22]. Plots of moving average concentrations over time were presented for
several areas and were comparable to the national average with the exception of the South Coast
Air Basin, where values were notably higher than the national average [slides 24-28]. The
seasonal profile was consistent in four locations [slide 29], likely due to motor vehicle pollutant
contributions.
Perchloroethylene exhibited less variation among sites than might be expected, considering it is
a point source pollutant. This could be due to a well-mixed background from a multitude of
sources, or indicate that some sources are not detected by the national network. Only one
28
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
location (Burbank) exhibited a significantly higher mean [slide 31]. More detail is revealed
when shifting to a smaller scale and denser network, with levels varying from very low to very
high [slide 32]. Two-year moving averages were overall consistent with the national average
[slides 33-36] with higher values evident in the South Coast area [slide 37]. In plots of seasonal
variability [slides 38, 39], the concentration peaks were not uniform among locations, as was the
case with benzene concentrations. Data also revealed exceptionally high values in the San
Rafael area, due to a dry-cleaning facility [slide 40]. This site illustrates the need for
neighborhood monitoring in some cases.
Carbon tetrachloride concentration means were very similar among both the statewide and the
regional locations - results that were expected since most ambient carbon tetrachloride comes
from background concentrations [slides 43,44]. Two-year moving average concentrations were
very similar to the national average [slides 45-49], and carbon tetrachloride exhibited virtually
no seasonal profile [slide 50].
Hexavalent chromium, another point source pollutant, exhibited less variability in the statewide
locations than would be expected [slide 52], with values in some locations averaging below the
limit of detection. Burbank was an exception, with high values for all three years plotted. Two-
year moving average concentrations were consistent with the national average in all locations
except the South Coast area, which had high values overall, with the highest values in Burbank
[slides 53-57]. Slides 58 and 59 illustrate the same results in two different scales. These show
low concentrations in El Cajon and Chula Vista, two areas near San Diego. Neighborhood site
values from the same area, however, show very high results for some areas, in particular a site
located on Newton Street [slides 60, 61]. The neighborhood results were obtained through
small-scale monitoring during the Barrio Logan case study.
Linda Murchison presented the outcome of a study that began when the Air Resources Board
was invited to monitor in the Barrio Logan community in San Diego in 1999. Monitors were
located at Logan Memorial Academy [slide 2] and measured all criteria pollutants for seventeen
months. The monitoring results were very similar to results from the San Diego region as well
as the statewide average. Most hexavalent chromium values were below the limit of detection
[slide 5]. In December 2001 additional monitoring was performed at six sites near two chrome
plating facilities in an attempt to understand neighborhood exposure. Two weeks of monitoring
revealed very high levels of hexavalent chromium; the source of these emissions was unclear,
however, due to the presence of multiple small sources in the area. Further monitoring was
conducted to determine the source of the problem and to understand neighborhood exposure.
The second round of monitoring included the six original sites; additional monitoring of twelve-
hour chromium and twenty-four-hour metals at 2121 Newton Street; and indoor monitoring at
both chrome plating facilities located nearby. It was noted that both plating facilities had been in
compliance with regulations at the time. The first two weeks of data did not show the very high
values measured in December, but some peaks were still evident, as was a relationship between
29
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
the Master Plating facility and 2121 Newton Street [slides 22-24], Modeling analysis assuming
Master Plating to be the chrome source matched the observed data [slide 25,26] and confirmed
this as a localized source, as concentrations dropped off a short distance from the facility. The
modeling results also helped refine the next stage of monitoring, which was continued only at
2121 Newton Street and an alley adjacent to Master Plating. In addition, weekend monitoring
was conducted, and a restraining order to stop chrome plating at Master Plating became effective
on March 25. Results obtained from March 11 through March 24 confirmed the relationship
between Master Plating and 2121 Newton, and revealed a strong correlation when winds were
from the west. Indoors monitoring revealed that hexavalent chrome made up 90% of total
chrome in the plating facility, and 55-60% of the total in 2121 Newton Street. As expected,
concentrations dropped to close to ambient values when plating was not taking place, after
March 25. Some high values measured after that date were attributed to chrome dust disturbed
by cleaning activities at Master Plating; and a very high value [slide 33 - April 5-6] was the
result of construction of a secondary containment tank, mandated by the restraining order. To
complete the study, a last round of monitoring at sites based on model predictions confirmed that
the second plating facility in the area was not a contributor to the high chromium emissions.
Study results validated the use of both exposure and diagnostic sampling. Frequent sampling
may be necessary in similar situations, and monitoring networks can evolve as new information
is revealed. Statewide implications derived from the study include the issue of fugitive dust as a
contributor to hexavalent chromium - as evidenced by increased values on days when cleaning
and construction were taking place at Master Plating. The state may also need to examine
whether a fume suppressant is adequate to reduce emissions from chrome plating facilities. It
was noted that such studies can get costly - the total cost of the Barrio Logan case study was
approximately $1 million.
Jeff Cook completed the presentation by summarizing the lessons learned and the issues raised
by the study.
Regional Monitoring:
• Can be a reasonable and cost effective means of assessing exposure to motor vehicle
and similarly prevalent pollutants;
• Has supported identification and adoption of control measures of motor vehicles,
fuels, point, and area sources;
• Has adequately addressed changes in air quality levels due to fuel switching and other
motor vehicle controls;
• Can detect well-mixed cumulative emissions from many small sources;
• Can capture one aspect of point source monitoring, yet may overlook sub-regional
exposures to high risk pollutants and the need for additional source controls;
30
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
• A dense regional air monitoring network can provide useful information on exposure
due to neighborhood point sources.
Regional vs. neighborhood monitoring:
• There will be trade-offs between regional and neighborhood monitoring resources.
Sampling frequency, number of sites, and duration, need to be carefully balanced to
achieve the objectives of both.
Neighborhood monitoring:
• Provides a distinct view of disproportionate risk that often exceeds that expressed by
regional monitors;
Is particularly effective when sources are clustered and pollutants are high risk
compounds;
• Can be dynamic; interest is often greater, and objectives are likely to evolve;
• Point source monitoring will be comprehensive and time consuming;
• Should have a point of instigation, such as a community concern. A source category
or cumulative impact of sources must be identified.
Modeling:
• Near-source modeling has a role in assisting neighborhood monitoring design;
• Neighborhood modeling places greater demand on the need for local, facility-by-
facility emissions information.
To help justify the cost incumbent in neighborhood monitoring studies, local findings need to be
leveraged to a larger result. These findings are transferable and can benefit other communities
with similar problems, as evidenced by the Barrio Logan example. The Barrio Logan study
impacted statewide chrome measures. Although not the only driving force, neighborhood
monitoring can be the basis for strengthening emissions controls from point sources.
Questions and Comments
Comment: Even though both plating facilities were in compliance with regulations, the inside
concentrations were very high. This may indicate a need for cross-training for
inspectors. Also, the values may have been high enough to contact the Occupational
Safety and Health Administration (OSHA).
Response: The values were not high enough to contact the California OSHA, but they are
investigating both facilities.
31
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Question: How did you keep the community updated?
Response: We met with the community every couple of weeks. The city has attempted to
relocate Master Plating, and their current goal is to separate the residents from such
facilities.
Question: Did you take blood samples?
Response: We did not. The community has requested blood sampling for lead, however, and it
may be provided through the Health Department.
Question: What particle size dust were you concerned about?
Response: The measurements taken were for total suspended particles. We have not looked
closely at this data yet for particle size.
Question: It seems that such a close look may be required at other places .as well; any thoughts
on whether there may be a way to divide the high cost of these studies nationally?
Response: A big question every time we undertake community monitoring is whether we are
prepared for the results. Rather than doing industry-by-industry case studies, a better
approach may be to develop modeling or state/regional-level monitoring tools that
will better capture results at the community level. The idea of sharing the cost and
information is valid, however.
Comment: EPA has developed a simple computer-based model that anyone can use, and that is
just the type of tool that was just mentioned. This tool is the Metal Finishing Facility
Risk Screening Tool (MFFRST). It can be accessed at:
http://www.epa.gov/ncea/mffrst.htm
The EPA contact is Matt Lorber in NCEA. He can be reached at (202) 564-3243.
32
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Design Your Own Air Toxics Monitoring Network - Motria Poshyvanyk (Region 5)
Motria Poshyvanyk gave a presentation which described the exercise and provided guidelines
and suggestions. One adjustment was made to the directions: to allow for multiple short-term
monitoring sites in the small-scale network, it was assumed that there is no cost to the
establishment of a monitoring site. Other directions applied.
All participants were divided into four groups - two groups for each type of monitoring network,
large-scale and small-scale. Each group was selected to include two or more people with
significant monitoring experience. Other specialists (meteorologists, toxicologists, etc.) were
randomly divided. The groups worked separately, and were all provided with a poster-sized map
to write on. Each group was assigned a facilitator to lead the development of the network. A
volunteer was identified to present the designed monitoring network to the reconvened
participants.
Network #1: Large-Scale Urban Area
Blue Group: Facilitator Barbara Morin
Yellow Group: Facilitators Tim Watkins, Deborah Luecken
Network #2: Small-Scale Local Hotspot
Green Group: Facilitator Neil Frank
Red Group: Facilitator Motria Poshyvanyk
Blue Group
The group identified pollutants of concern as VOCs (benzene, 1,3-butadiene), carbonyls
(formaldehyde, acrolein), and metals (chromium). The group added toxics monitoring to
existing PM25 sites where possible. The two existing PM25 speciation sites were supplemented
with hexavalent chromium, VOC, and carbonyl monitors and one-in-twelve-day duplicate
sampling (S173.5K). VOC and carbonyl monitors were added to ten existing PM25 mass sites in
four demographic areas on a one-in-twelve-day schedule ($300K). Metals analysis was added to
three sites ($24K) and limited acrolein monitoring was performed to test the new technology
($7K).
33
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Yellow Group
The group upgraded some PAMS sites to year-round VOC monitoring: an upwind background
site and a site downwind of airports to the north of the city. VOC monitoring was also proposed
for PM2 5 speciation sites and a PM2 5 mass monitor in ah environmental justice area. Proposed
metals sites included: one metals site near Gary, Indiana; metals and hexavalent chromium near
O'Hare airport; metals and hexavalent chromium at a population oriented urban site. Carbonyls
and acrolein were added to all VOC sites. Finally, two background sites were set up at PM7 5
locations in the distant suburbs.
Green Group
The group gathered additional information: 1) updated source inventory, 2) collected on-site
meteorological data, and 3) conducted a local modeling study ($10K). Potential monitor sites
included schools, playgrounds, parks, as well as areas of predicted high concentrations. Hot
spots were identified via saturation monitoring for several days (twelve-hour sampling) and by
dispersion modeling results. Saturation monitoring consisted of twelve sites operating for three
months on a one-in-six-day schedule; an alternative plan was twelve sites for two weeks daily.
Red Group
The group organized a meeting ($5K) with local groups to confirm the locations of
perchloroethylene operating dry cleaners and to ask residents about perceived hot spots resulting
from S&C emissions and local dry cleaners. Phase I included meteorological measurements to
determine prevailing wind patterns ($20K); a screening dispersion model ($10K) was used to
identify the area of S&C maximum impact. Several short-term sites were set up in a grid pattern
in the northeast quadrant to identify high perchloroethylene areas ($90K). Phase II included the
establishment of three long-term monitoring sites ($31.5K each): one site in a high
perchloroethylene area in the northeast quadrant; one site at the site of maximum S&C impact;
and one site representing a typical mix of perchloroethylene, S&C, and mobile source emissions
in the community. Phase I was estimated at S125K and Phase II at S94.5K.
34
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
SESSION III: MODELING TOOLS - CURRENT AND FUTURE
Co-chairs: Randall Robinson (Region 5), Joe Touma (OAR/OAQPS), and
Tim Watkins (ORD/NERL)
Randall Robinson (Region 5) began session three and outlined the topics to be covered: general
information on air quality models; an overview of how these models have been used in the past;
and future directions in air quality modeling.
Air Quality Models - Joe Touma (OAR/OAQPS)
Models have been used for air quality management since the early 1960s by Agency programs
and regions. Their main role has been in facility permitting. Other uses, including
demonstrating the adequacy of emission limits and policy analysis, utilize the capacity of models
to project conditions into future years. As discussed in earlier presentations, modeling results
can also assist in the selection of monitoring sites. Combined with emissions inventory data and
air quality monitoring, modeling can be used as part of risk assessments leading to development
and implementation of regulations [slide 4].
All air quality modeling systems include two major components: an emissions model and a
meteorology model. The accuracy of any model output is dependent on the quality of the data
put into it. Meteorological and topographic complexity and user expertise also play roles.
Overall, models are most accurate in simulating long-term averages in simple topography.
Numerous types of air quality models are currently in use by EPA, ranging from screening to
refined. Screening models are generally applied prior to refined models; they use one or a few
groups of sources and conservative estimates of concentrations. Refined models include
Gaussian plume models and numerical grid models. Gaussian plume models use meteorology
information obtained at the source and assume it holds true throughout a fifty-kilometer
perimeter. Wind direction is used to predict concentrations following the plume of emissions
from a source. This type of model is most widely used for non-reactive pollutants. For reactive
pollutants, or where there is complex topography, numerical grid models are used. These require
complex wind flow and other meteorological information, and consider the chemical
transformation of compounds in the atmosphere (e.g., ozone).
Models (both screening and refined) are developed by EPA, other government agencies, and
private sources. EPA models in particular undergo extensive evaluation and statistical measures
of performance. The executable and source codes of models developed by government agencies
are publicly available for use by anyone. Some private industry models are also available to use
at little or no charge, as listed in the Guidelines on Air Quality Models. Supporting tools and
35
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
sources of data for input into models are also available, including meteorology and terrain data,
emissions data, and census data.
EPA's National Scale Air Toxics Assessment (NATA) provides an example of model use in
policy analysis. The NATA study used Gaussian models to predict lifetime estimates of
exposure and risk for urban air toxics. The model is run every three years when NATA is
updated and results are used to prioritize data and research needs and to provide a baseline for
measuring future trends. Ambient concentration and emission results from the 1996 NATA were
presented [slides 15, 16].
Resources and information on air quality modeling are provided through the Support Center for
Regulatory Air Models (SCRAM) website, as well as by the Air Pollution Training Institute
(APTI). The website contains the code, user's guides and other guidance documents for many
air quality models, proceedings from the 7th Modeling Conference, a forum with frequently
asked questions, and links to other websites. The APTI provides classroom, on-line, and self-
instructional courses, satellite courses, workshops, and seminars. Website addresses are:
SCRAM website:
http://www.epa.gov/ttn/scram
Air Pollution Training Institute course listings:
http://www.epa.gov/oar/oaqps/eog
The EPA has been a leader in air quality modeling for air toxics assessments, and improvements
continue to be made in providing different models for different situations. A future direction the
Agency might consider is whether we should become involved in the area of homeland defense.
Concern regarding the intentional release of harmful chemicals existed prior to the events of
September 11th, and the Agency does not currently have models to deal with such a situation
where concentrations are paired in time and space. A decision will need to be made on whether
this should be a matter for EPA to undertake, or whether the Agency should stay within its
current focus of general population exposure to air toxics.
36
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Using Emission Inventories for Air Quality Modeling - Joe Touma (OAR/OAQPS)
and Madeleine Strum (OAR/OAQPS)
The emissions inventory is a fundamental building block in developing an air quality control and
maintenance strategy. Inventories are useful in identifying sources and general emissions levels
and in providing input to air quality models. They can also provide information on trends over
time that can be used in EPA reports to demonstrate that goals are being met. In order for
inventory data to be useful as model input, inventory preparers and air quality modelers must
work together to understand the needs of air quality models and limitations of the inventory.
Emissions models and processors commonly used include Sparse Matrix Operator Kernel
Emissions (SMOKE), Emissions Modeling System (EMS95/2000 and 2002), and Emissions
Modeling System for Hazardous Air Pollutants (EMS-HAP) - the latter being the model used in
the NATA study.
EMS-HAP is an emissions processor used in combination with inventory data and the
Assessment System for Population Exposure Nationwide (ASPEN) dispersion model and the
Hazardous Air Pollutant Exposure Model (HAPEM). Emissions inventory locations are
converted using EMS-HAP to air dispersion model coordinate units. Quality assurance is
performed on stack parameters, and default values are available to use if necessary. These are
then used to group and speciate pollutants and spatially allocate non-point source emissions.
Since inventory data is annual, emissions can then be temporally allocated, assigned source
groupings, and formatted in a model-ready format for the ASPEN dispersion model [slides 6-8].
For national-scale modeling, EMS-HAP can assign source locations and stack parameters when
needed, based on information on the percentage of states that did not provide emissions data.
Partitioning information for metals is another feature of EMS-HAP, giving a user the option of
specifying which species of a metal should be used in the analysis. An approach also exists for
speciating chromium compound emissions into hexavalent chromium [slides 9,10]. Speciation
data from various sources is now being collected in order to augment this aspect of the model.
Spatial allocation of non-point source emissions is accomplished by using census tract
population data or a similar surrogate to allocate emissions at the county level; consumer
products are an example of the types of emissions that can be estimated using this approach.
Temporal allocations take into account inventory data, day-of-the-week and seasonal variations,
and temporal profiles. Other functions performed by emissions processors such as EMS-HAP
include assigning source groups to ambient emissions, projecting future emissions scenarios, and
producing model-ready input files.
Emission processors automate the process of preparing the inventories needed to manage an air
program, and ensure that inventories have the appropriate level of detail to support modeling.
37
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Model Applications: Local and Urban Scale Modeling
Local Scale — Barrio Logan Modeling Analysis — Vlad Isakov (California Air
Resources Board)
A conceptual modeling protocol was developed for the Barrio Logan neighborhood near San
Diego, California. Local emissions and local scale modeling results were used to calculate the
inhalation risk at Barrio Logan. The study also served to evaluate the model used. Ultimately
this effort will develop and evaluate methodologies to estimate annual average concentrations of
various pollutants released from multiple sources at the neighborhood scale. The modeling
working group included participants from government agencies, universities, industry, and
environmental groups. Modeling was conducted at the neighborhood scale with receptors
located near emissions sources, and at the regional scale - up to the size of air basins - to adjust
for regional background. Thirty pollutants were modeled at the regional scale and more than one
hundred at the neighborhood scale [slide 5]. Both traditional and advanced models were used to
estimate annual concentrations [slides 6, 7] and to identify those that are best suited for assessing
neighborhood impacts. The study is also expected to develop recommendations and guidelines
for assessing the cumulative impacts posed by air pollutants at the neighborhood scale.
Modeling results and recommendations will be shared with EPA and other interested groups.
Maps of Barrio Logan and the surrounding area were presented that indicated daytime and night-
tune wind directions and the locations of emissions sources - including point sources, road links,
rail lines, and shipping lanes [slides 10-13]. Inhalation risk based on predicted concentrations of
local emissions was estimated for benzene, hexavalent chromium, diesel PM, and all pollutants
combined [slides 14-18], Benzene risk became more significant near highways, and hexavalent
chromium risk was high in two hot spot areas. Combining risk from all sources except diesel
PM showed the hot spot point sources to be the most significant contributors. When all sources
were combined, diesel PM emerged as the largest factor contributing to total risk; shipping lane
emissions contributed more to total diesel PM than highways or rail lines.
Further analysis included simulations with the AERMOD and CALPUFF models, adjustment of
background concentrations using results from regional modeling, and tracer studies to evaluate
model performance. Regional modeling using the Community Multi-Scale Air Quality (CMAQ)
model revealed a hot spot with high concentrations of formaldehyde [slide 20] that correlated
with the results of other models as well as the monitoring annual average. The tracer study used
both samplers and a mobile monitoring station to take measurements from fifty sites in August
and December 2001. The tracer was released at a central location, and samples were taken to
compare model predictions of concentrations at various distances and directions radiating from
38
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
the release site. Measured tracer concentrations showed good agreement with those predicted by
the model [slide 27]. Future model evaluation will include a second tracer study, comparison of
modeling results using a new database, and the publication of guidelines identifying the models
and model options best suited for assessing neighborhood impacts.
Uncertainty analysis was also performed to identify major sources of uncertainty, and to develop
a methodology for assessing uncertainty in annual average concentration estimates. Models
were broken down into their components and available data were used to represent model inputs
as distributions. A dispersion model was applied and statistical sampling used to estimate the
range of possible model results [slides 30-32]. A Monte Carlo simulation calculated the
uncertainty - percent difference - and confidence intervals for each source of uncertainty [slides
33-35]. Emissions were the dominant source of uncertainty in this case, although the model was
sensitive to all factors. The magnitude of uncertainty was greater near the sources and decreased
with distance. The development of modeling guidance was deemed important for the AERMOD
model, which is more complex than ISCST3 and capable of providing a wider range of results.
39
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Urban Scale Modeling - Houston Case Study - Joe Touma (OAR/OAQPS)
The National Scale Air Toxics Assessment (NATA) conducted by EPA modeled ambient
concentrations, exposure, and risk for thirty three air toxic pollutants. The modeling domain
included the continental United States at the census tract level and showed the contributions of
major area and other sources as well as background. NATA results are available at:
http://www.epa.gov/ttn/atw/nata
At the regional scale, refined dispersion model analysis can improve upon the NATA assessment
for urban areas. Improved modeling tools can provide a better description of air quality as well
as a higher degree of resolution than the national scale study. Such comparisons should also
help identify data gaps in the national scale study. Conducting urban-scale assessments for a
number of cities is one of the components of the Integrated Urban Strategy (64FR137, July
1999). Data from the emissions inventory on point/non-point sources and road/non-road mobile
sources were refined for the purposes of the urban study. Specific locations were determined for
point sources, and non-point sources allocated from county-level to one kilometer grids; on-road
mobile sources were allocated using local traffic counts, road locations and emission factors; and
non-road mobile sources were allocated from the county level to one kilometer grids. The
ISCST3 model, used for the urban study, provided better area source representation than the
ASPEN model, included deposition, and added simple chemical transformation processes. This
more detailed analysis provided more realistic patterns and better agreement with monitoring
data than the national scale study. In addition, hot spots were found by the urban scale study that
did not appear in the national scale study. Assessment at this scale, however, still cannot be used
for neighborhood-scale concentrations.
The use of the ISCST3 model in this urban scale study demonstrated that results are improved
when emissions can be placed hi more precise terms. Allocating mobile emissions to actual
roadways is preferable to allocating county mobile emissions using surrogates. This improved
agreement was confirmed by comparing modeled concentrations from the national scale and
urban scale studies with monitoring data from seven sites in Houston. Accurate emissions
allocations, and the consideration of factors such as deposition, become even more important
when models are intended for use in estimating exposure to and risk from air toxics.
40
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Questions and Comments
Comment: Air quality modeling is inseparable from inventory data if we want to assess short-
term exposure. Data used for transportation planning is available from numerous
sources and can also be used for inventory purposes.
Comment: Models are not as easily understood by the community as monitoring data, and this is
one of the reasons why we often opt to use monitoring. We need a better sense of the
meaning of results from modeling studies as well as monitoring programs.
Comment: This is one of the big challenges in air toxics, since we have 188 chemicals' names in
the CAA. We have a good understanding of some, but know little about others.
Question: Is penetration of chemicals from outdoors to indoors taken into account with this
model?
Response: There are other models that can be used for indoor penetration and concentrations.
Comment: Improving the quality of the data used for emissions inventories is important, since
data are frequently used as input for modeling.
Comment: The CAA requires that criteria be set for emissions, but no such requirements exist
for air toxics. For example, inventory data from many states improved considerably
after the states started reviewing the national scale assessment results.
Comment: The NATA study has shown no air toxics "problems" outside of urban areas;
however, it is a national average, and may not reflect local conditions.
Comment: Judging from this study, it would seem that modeling could cost more than a
monitoring study.
Response: It can be more expensive in some cases. Deciding between monitoring and modeling
should depend on the purpose of conducting a study. Monitoring can take more time,
and offers no projections for hypothetical situations or future years. If goals will be
satisfied by monitoring, then it should by all means be done, especially if it is also
more cost-effective.
41
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Applying CMAQ Models3 for Air Toxics Assessments - Bill Hutzell (ORD/NERL)
The concentrations of many air toxics depend on meteorological, physical, or chemical processes
that cannot be treated simultaneously by standard regulatory models. The Community
Multiscale Air Quality (CMAQ) model system was used to determine if it could be adapted to
assess the fate of such compounds.
CMAQ is a Eulerian-based system, or a grid model, that simulates urban and regional scale
transport and chemical processes. A fusion of three models [slide 5], CMAQ uses the "one
atmosphere" approach for air quality and deposition modeling. Current research includes
continental and regional studies including atmospheric fate modeling of emissions, air
concentrations, and deposition for mercury, dioxins, and select hydrocarbons.
CMAQ is being adapted to simulate the deposition of mercury, taking into account its gas and
aerosol species, cloud chemistry, and gas chemistry, and using the Regional Acid Deposition
Mechanism (RADM). CMAQ mercury deposition simulations were performed for the spring
and summer of 1995 and results were compared to observations from the Mercury Deposition
Network [slides 9,10]. Spring results had better agreement with monitoring data than those for
summer, likely because summer rainfall was predicted poorly by the model. Simulated cloud
chemistry agreed with most other models and cloud water model predicted concentrations were
within the range of observed concentrations; however, questions and uncertainties remain about
both these processes. Model behavior for wet deposition showed moderate accuracy in cool
seasons, but poor accuracy when convective precipitation was prevalent. More comprehensive
testing is needed to resolved these issues and improve model predictions.
Several toxic compounds are implicitly simulated in existing chemical mechanisms within
CMAQ [slide 13]. Volatile organic carbons labeled as air toxic will be added to CMAQ and the
chemistry mechanism for ozone will be simplified in order for the model to support the 1999
National Air Toxics Assessment (NATA). Formaldehyde and acetaldehyde were simulated by
running CMAQ using the SAPRC99 mechanism [slides 14, 15]. In the future, the CB-IV
mechanism will be used for these compounds, since it produces similar results but runs faster
than SAPRC99.
CMAQ can also be used to simulate the atmospheric fate of dioxins and furans [slide 16].
Expanding chemical and aerosol species, including loss and degradation processes, simulating
gas to aerosol exchanges, and obtaining reliable emissions data are among the challenges faced
in updating CMAQ to include dioxins. A test scenario run for a two-week period in April 1995
showed predicted deposition and predicted average air concentrations compared to emissions
42
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
[slide 18]. Congeners can be treated individually and an estimates of their contribution can be
obtained [slide 19].
Work is also being conducted in the area of neighborhood modeling of urban air toxics, with the
goal of using modeled air quality to support human exposure models such as the Stochastic
Human Exposure and Dose Simulation (SHEDS) model. Challenges inherent in this effort
include: identifying resolutions that are appropriate for separate pollutants or pollutant groups;
taking into account the multiple sources present in urban environments; determining the impacts
of chemical processes that may affect concentration; and obtaining concentration statistics for
assessing human exposure. A prototype of this model will be applied to a study in Philadelphia
and the results will be verified using techniques such as computational fluid dynamics modeling.
CMAQ would provide more information for input in human population exposure models, as
illustrated [slide 24].
In the future, CMAQ will be adapted to address all thirty-three compounds emphasized by
NATA by adding one to two compounds per year. Other toxics which may be addressed are
inert and involatile metals, halocarbons, and acrolein. More information on CMAQ is available
at:
http://www.epa.gov/asmdnerl/models3/cmaq.html
43
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Questions and Discussion with Session III Speakers
Question: Regarding model uncertainty, sometimes there is variability in model parameters that
is not uncertainty - it is certain, but still variable.
Response: We did not address variability in model options; on most models, users can select the
range of model options. For other parameters, e.g., modeling of emissions, there is
uncertainty, not inherent variability.
Comment: [Joe Touma is] working on a study to look at NATA model uncertainty and would
welcome assistance from ORD.
Question What is the status of the ISC-PRIME model?
Response: ISC-PRIME deals with down-wash, or high local concentrations, due to the presence
of buildings. I believe that you can currently use ISC-PRIME.
Comment: Yes, ISC-PRIME is now available as a technique for down-wash.
Question: When will CMAQ be available for urban scale modeling for all thirty-three toxics?
Response: Possibly in three to four years.
Question: Can you clarify why NATA results show higher concentrations in centroids?
Response: Modeling was conducted at the centroid level and then at smaller scales. At the
smaller scale, hot spots were found that were not evident previously.
Question: Why did the uncertainty analysis show higher accuracy for far-field compared to
near-field predictions?
Response: This may have only been the case for that one application; we may get different
results in subsequent applications.
Question: What would you use for modeling lifetime exposure?
Response: Census tract centroids can be used with the ASPEN model, then another model can
be used to convert them to exposure. A comparison can be made between risk
numbers and exposure numbers.
Question: Is there a way of doing this that does not require such a convoluted approach?
Response: There are numerous tools available to choose from. A decision must be made on
which models to use, followed by comparison of the results.
Comment: CMAQ is just starting to be configured for annual time periods.
44
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Question: How did you account for different species of mercury when comparing observed with
predicted deposition?
Response: All wet deposition from the model was aggregated in order to compare it to observed
values.
Question: In the urban-scale study, did you compare the average concentration (spatially) in a
census tract compared to that in the centroid?
Response: Yes, we will be updating NAT A with this information this summer.
Question: Will that change the way NATA data are reported?
Response: It will not change it for the 1999 assessment, since we are reporting distributions.
Comment: A tract centroid is a population weighted location.
Question: Are there any CMAQ developments planned to address sensitivity?
Response: At present, there is nothing standardized on that issue.
Question: [Which factor] is anticipated to produce the most uncertainty in air toxics models?
Response: ORD is conducting research to determine that.
Response: It is difficult to specify that right now; it can also depend on the pollutant being
modeled.
45
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Breakout Groups to Discuss Modeling Topics/Questions
During the second breakout session participants were divided into four groups to engage in
discussion on any aspect related to modeling. Potential topics and questions to address included:
• How active are the regions in ambient air modeling?
• What types of assessments are being done?
• What are the priorities? What is important?
• Where do the regions want to be with ambient air toxics modeling and skills?
• What obstacles exist? What do the regions need? (Guidance, tools, etc.)
• What is needed to overcome these obstacles?
• Suggested workshop follow-up items.
Following the thirty-minute session, each group presented the major discussion points to the
reconvened participants.
Yellow Group
• Application of models: better ways are needed to communicate models and their
results to the community and general public. It is necessary to keep models simple
and explanations clear and concise.
• Toxicity and regulations rely on model outcomes. When models change, the
outcomes also change, and toxicity and regulations are affected. There are many
political ramifications to merely improving or changing a model.
• Modeling should continue to be supplemented with monitoring. Combined efforts
are needed to improve modeling and to continue to support monitoring.
Blue Group
Regions need air toxics models and tools. ICS is a model currently in use, although it
would benefit from improvement. CMAQ is too experimental and seems to have too
many sources of error for use by the regions. It has not been evaluated for air toxics,
and was primarily developed to support NATA analysis.
Regions need models with the ability to accommodate complex parameters (e.g.,
mountainous terrain). Most regions have technical staff limitations, inhibiting model
development and improvement.
The purpose of models is to determine risk levels, predict future risks, and support
monitoring data; these are important functions in risk assessment and risk
management.
46
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Red Group
Air toxics modeling efforts are not a priority for the regions; we need to know what
the relative air toxics priorities are.
Inventories need to be improved. Better data is necessary in order to obtain better,
more useable model results. In addition, the data and models should be in an
appropriate scale for regional modeling.
Transportation planning system data must be integrated into air toxics modeling,
especially considering the importance of mobile source emissions.
Consistency among models is important. Developers should ensure that certain data
and model standards are met before models can be used and applied.
Green Group
Concern was expressed regarding the limitations of modeling for the regions. Data
limitations can augment the shortcomings of models, such as the unknown effects of
retrofits on diesel and other air toxics emissions levels.
There is a shortage of technical staff involved in model development. Monitoring
and modeling personnel are also needed to communicate within and across regions,
especially information on what spatial scales modeling is conducted.
Questions were raised on when community outreach is necessary. Outreach topics or
issues include: an explanation of model results; awareness and communication of
diesel versus ozone and paniculate matter; and immediate versus long-term effects.
47
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
THIS PAGE INTENTIONALLY LEFT BLANK
48
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
SESSION IV: HUMAN EXPOSURE ASSESSMENT
Co-chairs: Alan VanArsdale (Region 1) and Tim Watkins (ORD/NERL)
Tim Watkins (ORD) introduced Session IV and emphasized its focus on human exposure.
Session presentations were planned to give participants an overview of the subject and updates
on the topics of indoor and multi-pathway exposure, and the modeling data and tools available.
Three human exposure case studies were also presented.
Introduction to Human Exposure - Linda Sheldon (ORD/NERL)
The goal of EPA's air toxics activities is to reduce air pollution emissions in order to improve
ambient air conditions. Lower emissions in ambient air should reduce exposure and result in
improved human health. Exposure is defined as the contact of an individual with a pollutant for
specific time durations, and is the basis for health outcomes. For exposure to occur, an
individual must come in contact with a contaminated media; the pollutant from the contaminated
media is then transferred to the exposure boundary. This pathway can be used to link the
ambient concentration of pollutants to exposure. Exposure can be expressed as aggregate or
cumulative - referring to one or multiple stressors, respectively - and results in an absorbed
dose. The absorbed dose, or amount of stressor that crosses the body barrier, results in the target
dose, or the stressor concentration at the site where effects occur. Biomarkers refer to the
concentration of the stressor or a metabolite present in a biological fluid; there is not always a
clear relationship between biomarkers and exposure.
Exposure research programs develop methods, data, and models to evaluate the relationship
between ambient concentration and exposure; to identify and quantify factors that affect this
relationship; and to estimate distributions of exposure and dose. Flow diagrams illustrated the
scientific elements of exposure - from the formation of the stressor to health effects - and how
exposure relates to the environmental health paradigm [slides 7, 8]. Research on human
exposure must determine the extent and possible health effects of exposure, determine ways in
which to reduce it, and provide a measure of the success of the study. ORD research on
exposure focuses on the development of tools for estimating exposure and dose to address the
greatest risks. Exposure characterization for air toxics must take into account sources, fate, and
transport of chemicals to determine concentrations in micro-environments. Individual exposure
will depend on times spent in these micro-environments, and on the activities and mechanisms
responsible for transfer of the stressor or pollutant [slides 11,12]. Both monitoring and
modeling assessment approaches can be used. Exposure models exist which estimate the sum of
exposure in all micro-environments and can be used in combination with measurement data or
default assumptions [slides 13-15]. Exposure monitoring is used to determine concentrations
49
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
and personal activity data, and to characterize attenuation rates and penetration factors of
pollutants (e.g., outdoors to indoors). These data can then be used to support modeling studies.
Several such studies have been conducted in past years. The Total Exposure Assessment
Methodology (TEAM) studies, conducted in the 1970s, were one of the first attempts to relate
personal and ambient exposures. The results revealed personal exposure to be two to three times
higher, and prompted the start of a new set of research in that area. Large building studies found
VOCs indoors at concentrations three or four times higher than those outdoors. Emissions
testing was also conducted to characterize indoor sources, and the Relationship between Indoor,
Outdoor, and Personal Air (RIOPA) study combined indoor and ambient monitoring with
activity data and other factors that could affect exposure. Commuter studies measured
concentrations inside vehicles and compared them to ambient monitoring site values [slide 19].
Future work planned includes a study in Tampa, Florida, to evaluate source to health effects by
looking at combined resources from air toxics, asthma, and particulate matter.
50
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Air Toxics Exposure in Indoor Environments - John Girman (OAR/ORIA)
Indoor concentrations of many pollutants can be two to five times higher than outdoors, and
numerous activities studies have determined that most individuals spend up to ninety percent of
their time indoors. Outdoor air establishes the baseline for indoor concentrations when it
penetrates through ventilation systems or direct infiltration. Building materials, consumer
products, and occupant activities all add to this baseline to contribute to indoor emissions. As
part of an integrated approach to indoor air toxics, the Office of Radiation and Indoor Air
(ORIA) is studying the sources and compiling a ranking of all indoor air toxics based on recent
exposure data and comparisons of indoor and outdoor concentrations. This is intended to be a
screening, risk-based analysis to assist in setting priorities, and utilizes the same methodology
used for selecting the thirty-three priority HAPs from outdoor sources. Monitoring data were
obtained from studies conducted in the last ten to fifteen years, focusing on typical indoor
environments in non-industrial buildings. The study was limited to individual chemicals, and
did not consider biological contaminants or chemical mixtures (e.g., tobacco smoke). Both acute
and chronic health effects were addressed, assuming exposure by inhalation only. Ten
monitoring studies provided 213 concentration records for 112 air toxics including metals,
aldehydes, VOCs, and semi-volatile organic carbons (SVOCs). Indoor exposure data were
obtained from studies on office buildings, residences, and schools; however, the data used for
residences are becoming dated, and only a few intervention studies currently exist for schools.
Concentration data and penetration ratios were presented for the twelve VOCs with the highest
median indoor concentration values, obtained from a random survey of office buildings [slides
10,11]. Wide ranges of concentrations were exhibited by all twelve compounds, with some
buildings operating with very low concentrations. Indoor to outdoor ratios were much greater
than one (1) in many cases, indicating that some of these compounds have indoor sources (the
penetration factors were near one (1) for all compounds). Frequency distributions were also
illustrated for three chemicals [slides 12-14]. Tetrachloroethene concentrations increased much
faster indoors after the 50th percentile, indicating the presence of important indoor sources.
Benzene, its main source being penetration from the outside, showed parallel outdoor and indoor
concentrations until the 75th percentile. A terpene used in cleaning products, d-limonene, is
almost exclusively an indoor contaminant, and exhibited very low concentrations outdoors.
Relative outdoor and indoor concentrations were also presented for a selected list of eighteen air
toxics [slide 15]; a surprising result was the high indoors component of a number of pesticides
(e.g., aldrin, dieldrin), listed at the far left side of the chart.
The ranking of indoor air toxics has not yet been completed; the study is undergoing peer review
by the Science Advisory Board (SAB). However, some air toxics have emerged as likely to be
ranked high-priority for indoor air, including aldehydes, benzene, halogenated hydrocarbons, and
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
pesticides. Data gaps identified in the course of the study include insufficient data for several
building types (e.g., schools), and for high exposure areas or hot spots. Health effects data are
also lacking as is an inventory of source emissions information.
An integrated approach to air toxics is necessary, particularly when considering exposure.
Indoor and outdoor concentrations are interdependent, and exposure to a certain chemical will
have the same effects, regardless of source. This integrated approach should not greatly impact
ongoing activities and schedules, although the practice of placing monitors on the tops of
buildings, where ventilation outlets are located, may need to be examined. Both sources should
be considered for effective risk assessment; as for risk management, it may make more sense to
apply it to indoor sources where most exposure occurs.
52
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Multi-pathway Exposure Assessment - Matthew Lorber (ORD/NCEA)
Multi-pathway exposure considers not only direct exposure to a contaminant as it is released
from the source, but also includes exposure to the same contaminant after it has crossed to
different media. Assessment of multi pathway exposures is most often applied to persistent
bioaccumulative toxics (PBTs). Although the primary route of exposure to air toxics is by
inhalation, indirect exposure through soil, water, and food pathways can be ten to one thousand
times higher than direct exposure for some compounds.
In 1990 the Methodology for Assessing Health Risks Associated with Indirect Exposure to
Combustor Emissions was published by National Center for Environmental Assessment (NCEA)
and provided the first comprehensive set of fate equations and exposure methods for indirect
pathways. The first public review draft of the Dioxin Reassessment (1992) revealed that
exposure to dioxin from beef consumption exceeded that by inhalation by a factor of four. One
year later, that factor of four was cited in an attempt to stop a trial burn at a hazardous waste
incinerator, which was followed by an eighteen-month moratorium on permitting of similar
incinerators. To obtain a permit, incinerator operators were required to conduct a comprehensive
risk assessment for indirect impact, based on an addendum to the 1990 Methodology document.
SAB review of the addendum was conducted in 1994, and NCEA took on the task of updating
the document and addendum to include SAB comments. Region 4 also required the Columbus
Municipal solid waste incinerator to install Maximum Achievable Control Technology (MACT)
in 1994, because of concerns regarding dioxin exposures of nearby home-consumption farming
families. An update to the draft Dioxin Reassessment was also published in 1994 suggesting that
the disparity between the inhalation and consumption routes of dioxin exposure was a factor of
three rather than four. SAB comments on the document regarding health risk assessment
prompted EPA to begin another revision. The Mercury Report to Congress (1997) included an
exposure assessment and fate modeling data highlighting the complexity in modeling the
impacts of this bioaccumulative compound. Subsequent documents included the 1998 Human
Health Risk Assessment Protocol for Hazardous Waste Combustion; the 1998 update to the 1990
Multiple Pathways of Exposure document; and the latest Dioxin Reassessment (2000).
Fate and transport models consider atmospheric dispersion, deposition, resuspension and other
factors, and predict the movement of contaminants within and between compartments. They
incorporate partitioning models that determine how chemicals are partitioned among media in a
given environment; mixing models that predict concentrations in stationary receiving media
(e.g., soil mixing model); and bioconcentration or biotransfer models that estimate accumulation
of contaminants through food webs in animals and humans [slides 9-14].
53
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Multi-pathway exposure assessment scenarios consider subsistence farming and fishing
populations to determine the impacts of contaminated soils and water bodies, respectively.
Home gardeners are the sub-population chosen as those receiving the baseline or background
levels of contaminants. Major pathways include terrestrial animal consumption; soil pathways,
including the dermal route of exposure; the breast milk pathway, particularly for Hpophilic
organic compounds; and others, such as inhalation (direct exposure) or game animal
consumption.
A risk assessment study using fate modeling was performed for the Columbus, Ohio Waste-to-
Energy Facility (WTEF). In 1992, a decade after beginning operations, the facility was found to
emit 980 g TEQ (gram toxic equivalents) per year, compared to 3000 g TEQ per year emitted by
all other dioxin sources nationwide. Actions were taken to reduce emissions, and a stack test in
1994 indicated a seventy-five percent decrease. EPA headquarters, ORD, and Region 5
conducted a screening assessment of indirect impacts in that same year, leading to the conclusion
that continued emissions "may pose an imminent endangerment to public health and the
environment." The Columbus incinerator shut down in December of 1994. Fate modeling used
to support the EPA's position utilized the air-to-beef model described in the draft Dioxin
Exposure document and assumed a subsistence farming family scenario [slides 19, 20].
Exposure pathways considered beef, milk and vegetable ingestion; soil dermal contact and
childhood soil ingestion; and breast milk ingestion. The exposure duration for adults was
assumed to be seventy years [slide 21]. Air concentrations used were the average from nine
dairy farms located between five and twelve miles from the incinerator [slide 22]. Overall
exposure and cancer risk were estimated for each of the exposure pathways, with cancer risk
being highest for beef consumption (2*10^) and lowest for soil dermal contact (9*10~9) [slide
23]. Exposure from breast milk ingestion was determined to be higher by one order of
magnitude than exposure from beef and milk consumption, and higher by two orders of
magnitude than exposure from inhalation. Breast milk exposure near the incinerator site ranged
between two and more than seven times the background dioxin levels [slide 24],
54
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
What Human Exposure Data and Models are Available? - Haluk Ozkaynak
(ORD/NERL)
Air toxics exposure modeling considers indoor and outdoor sources of pollutants, ambient
concentrations, micro-environment concentrations, personal activity data and personal exposure.
All of the above factors are inter-related and can be input to or products of models. Types of
models include statistical models based on data from personal monitoring studies, and multi-
compartment, deterministic models based on known or assumed physical relationships. The
latter are more complex, and data are lacking for some of the required parameters, such as
outdoor-to-indoor penetration rates. Other models address variability and uncertainty in model
structure and input. These include demographics, ambient concentrations, exposure parameters
and human activity data. These models use simulated individual profiles to calculate micro-
environment concentrations and estimate exposure [slide 5]. .The equation for exposure is
similar across EPA exposure models and is the time-weighted sum of all exposures from the
different micro-environments in which a person spends time [slide 6].
Several databases provide exposure and related information [slides 7-13]:
• U.S. EPA Human Exposure Database System (HEDS) - Exposure database
http://www.epa.gov/heds/
• Mickey Leland National Urban Air Toxics Research Center (NUATRC) - Exposure
database
http://w\av.sph.uth.tmc.edu/mleland/
National Health and Nutrition Examination Survey (NHANES) - Biomonitoring
database
http://www.cdc.gov/nchs/nhanes.htm
• U.S. EPA Aerometric Information Retrieval System (AIRS) - Air Quality Database
http://www.epa.gov/airs/
U.S. EPA Consolidated Human Activity Database (CHAD) - Human Activity
Database
http ://www.epa.go v/chadnet 1 /
55
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
• U.S. EPA OAQPS National Air Toxics Assessment (NATA) - Exposure Modeling
and Results
http://www.epa.gov/ttn/atvv/nata/
Planned National Exposure Research Laboratory (NERL) research includes developing new
human exposure and dose estimation models for assessing population health risks and addressing
the exposure of susceptible populations. Variability and uncertainty will be explicitly quantified
for exposure models and a framework will be developed to accommodate both aggregate and
cumulative exposures. NERL's continuing goal in research is to support and enhance science
conducted by the program offices. The Air Toxics Multi-Year Plan outlines the development
and expansion of the SHEDS model through FY05, and a cumulative population exposure model
for a representative set of air toxics by FY08.
The SHEDS model will be designed to quantify both variability (temporal, spatial, or individual
differences in the value of an input) and uncertainty (measure of the incompleteness of
knowledge or information about an unknown quantity), and will give an estimate of the
confidence intervals associated with the model. Activity patterns for one year will be simulated
by using eight people from different age and gender cohorts, and diary entries for one weekday
and one weekend day per season [slide 17], Since age, gender, season, and weekday seem to be
the most important variables for longitudinal activity patterns, this approach should optimize
inter- and intra-person variability. A commuting algorithm will be included using census tracts
and probabilities for work location based on home location, combined with information from a
commuting database [slide 18]. Benzene will be used as a case study for the first operation of
the SHEDS model, which will include a variety of exposure micro-environments and influential
factors [slide 19]. One such factor, in-vehicle exposure, will be estimated using linear regression
of data from California roads [slide 21]. Other micro-environments will include parking
garages, fueling stations, and areas along streets or sidewalks.
Current research needs include determining more micro-environments and population groups of
concern, and measuring or estimating those factors that yield the greatest uncertainty. Personal
exposure measurements are also needed for population cohorts that have limited existing data.
In the field of model development, mechanistic and stochastic models will be needed to predict
source-to-dose relationships, as well as new modeling methods to address high-end exposures to
urban air toxics (i.e., hot spots).
56
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Questions and Discussion with Session IV Speakers
Question: What is your reaction to the [chromium] concentrations at 2121 Newton Street, [in
the Barrio Logan study]?
Response: I am not very familiar with the study; however, we have some understanding of ultra-
fine particles, and there are modeling tools that could be used.
Comment: Working on a similar situation with ambient air around the World Trade Center site,
we had to assume that indoor concentrations were equal to those outdoors.
Response: We could take advantage of some of the existing data from Barrio Logan, since it is a
good example of looking at individual risk; a study could be done to estimate or
model exposure.
Question: Is there a general trend in infiltration rates for particulate matter, as opposed to gases?
Response: Gases, in general, penetrate without much decay. Particulate matter does not
penetrate as well.
Question: A commuter study showed higher [outside-to-inside] ratios in Los Angeles than in
Sacramento? Why were the ratios different?
Response: The background concentrations are higher in Los Angeles; in Sacramento, the
ambient monitoring site was located further away.
Question: There seems to be an infinite number of indoor environments. Is it possible to use a
holistic approach?
Response: We essentially take a holistic approach when conducting personal monitoring, by
capturing everything that the subjects are exposed to.
Response: An inventory of source emissions from various environments would also be useful.
Data are needed in a lot of these categories.
Comment: Characterization of the different types of emissions would be useful, but personal
exposure monitoring - especially when combined with activity diaries - is a good
way to cover all possible micro-environments. Such monitoring can then be used to
go back and check if all micro-environments were captured in the source
characterization.
Comment: There seems to be some amplification of chemical concentrations indoors; chlorine is
a typical example of this.
Response: Outdoors, an emissions plume will be dispersed over time; concentrations often are
higher than expected inside, but should not be higher than outdoor concentrations
unless there are additional indoor sources.
57
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Comment: The exposure, however, could be higher because of the time factor; chemicals will
linger indoors, rather than disperse.
Comment: Air exchange is another factor to consider along with penetration, particularly for
reactive compounds.
Question: In some cases (e.g., in some radon studies) you can put a new family in the same
house and have two very different exposures. Is there any information on the fact
that the way people live in homes affects their exposure?
Response: In the case of radon, temperature is a factor, as are water traps, so such an effect is
not surprising. Other factors such as smoking, hobby activities, or cleaning materials
contribute other contaminants. There is potential for educating the public so they are
aware of these factors.
Comment: Public education could be an immediate intervention step and would not incur large
costs.
Comment: Risk assessors don't always understand the meaning of model results, and it takes a
lot of expertise to utilize activity data appropriately. It would be helpful if ORD were
to produce a guidance document on how to do this.
Response: Doing an uncertainty and variability analysis can help make this as objective as
possible.
Question: Toxicity values don't always match evaluations when we move away from
presumptions. Should we be making a greater effort?
Response: Homologous data and measurements are needed for comparisons - to this end, we try
to make models as flexible as possible when it comes to input data.
Question: If good data exist on indoor emissions and their toxic components, they could be used
to persuade industries to change their products. Are there any plans to pursue this
approach?
Response: Market pressures can work in changing consumer products, and the best approach
may be to educate consumers. For example, the carpet industry introduced a "green
label" program without any input from EPA that has been very successful.
Comment: This is happening already in some industries; a recent study ran into the problem of
not being able to find any consumer products left that contained the compound the
researchers were interested in. Since products change so rapidly, an inventory may
become outdated after only one or two years.
58
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Additional Case Studies: Modeling and Monitoring
Personal Exposure Monitoring Meets Risk Assessment: The South Baltimore
Community Exposure Study - Devon Payne-Sturges (OA/OPEI)
The South Baltimore Community Exposure Study was a comparison of measured and modeled
exposures and risks in an urban community. The investigation included three different exposure
estimation techniques and their impacts on risk estimates and risk characterization. Ambient
concentrations modeled by the Assessment System for Population Exposure Nationwide
(ASPEN) model have been used by EPA to estimate health risks. However, studies on human
exposure have shown discrepancies between ambient concentrations and exposures experienced
by individuals; personal exposures are frequently greater than both outdoor and indoor ambient
concentrations.
The Brooklyn, Brooklyn Park, and Curtis Bay communities in South Baltimore presented a
unique opportunity for this study due to the close proximity of polluting facilities [slide 3].
ASPEN model results (1996) for ambient exposures to twelve VOCs were compared with
outdoor, indoor, and personal exposures to the same twelve compounds, in an effort to
characterize potential public health risks. Personal exposure monitoring was conducted on
thirty-six randomly sampled non-smoking adults using seventy-two-hour weighted exposures
and time-activity questionnaires. The twelve VOC samples were all hazardous air pollutants
(listed on slide 10). In general, personal exposures were somewhat higher than indoor
concentrations and much higher than outdoor concentrations; indoor concentrations were higher
than those outdoors [slide 11]. These general trends did, however, vary by compound. Ratios
(personal and indoor to outdoor) were near one (1) for compounds with outdoor or ambient
sources. Chloroform was one VOC suspected to have significant indoor sources; personal
exposure was eleven times outdoor concentrations; indoor concentrations were nine times higher
than those outdoors [slide 12]. Ratios of ASPEN to actual exposures were also compared [slides
13,14]. ASPEN results were considered acceptable when predicted values were within twenty-
five percent of actual measurements. ASPEN somewhat overestimated personal benzene
exposure, but predicted accurately overall for compounds from primarily outdoor or mobile
sources. Personal and indoor exposures to chloroform and methylene chloride, which have
significant indoor sources, were under-predicted. These results led to the under-prediction of
cancer risk when using the ASPEN model, as opposed to using personal and indoor monitoring
data; a four-fold difference was evident between ASPEN cumulative risk predictions and those
based on personal monitoring [slide 15]. The relative risk profile from each of the twelve VOCs
examined also changed considerably when using ASPEN results versus mean personal
exposures; whereas ASPEN predicted benzene to be the biggest contributor (55%) to cumulative
59
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
risk, monitoring results revealed that chloroform was more significant, contributing 65% to the
total risk [slides 16, 17].
Study results highlighted the need to validate models with actual exposure data, since ambient
VOC concentrations did not adequately address health risks. Exposures to VOCs raise concern
for public health, and risk assessment performed using modeling and personal exposure
monitoring can help prioritize the pollutants that should be targeted for intervention to reduce
exposure and risks. The study examined exposure to only a limited number of air pollutants,
however, and findings are restricted to those compounds; metals, particulate matter, and diesel
PM were not considered, and may change the overall risk picture for the community. It was also
assumed that short-term exposures were representative of the long-term. Study strengths were
that exposure was measured close to the individual, and that the individuals selected were a
population-based random sample. In addition, the cumulative risk analysis provided by the
results was more meaningful and provided the context in which the public was interested.
Overall, the study confirmed the importance of VOCs as a public health issue. It also served as a
way of validating the ASPEN model; the model is a useful tool that predicts ambient
concentrations well. Based on the information obtained on indoor sources and exposure to
certain VOCs, a recommendation was made to implement a national program requiring
manufacturers of consumer products to reveal information about toxic components to
consumers. Such a program would not only provide needed information, but may introduce
market incentives for manufacturers to substitute less toxic components in their products.
Questions and Comments
Question: There is a tank facility located near the neighborhoods; did the residents indicate that
the tank "farms" were the cause of odor problems - particularly during cleaning?
Response: Odors in general have been a problem in the community, and could be from the tank
facility or from the wastewater treatment plant located nearby.
Comment: The chemicals used when cleaning such tank facilities have also been a problem.
Question: At what time of year was the sampling conducted?
Response: Sampling was conducted from January 2000 through June 2001.
Question: Chloroform dominated the indoor risk; what would happen if you conducted an
uncertainty analysis on those results?
Response: We have not done an uncertainty analysis.
Comment: If the residences were using treated (chlorinated) water, that could be an indoor
source of the chloroform.
Response: We don't know if this happens, although it is an issue we plan to examine further.
60
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Question: Were any people from sensitive populations included among the study participants
located within the heavy industrial area?
Response: The study was originally planned for residents within that area; however, they were
re-located after successful lobbying. Following that, we moved the study to adjacent
communities.
Question: What about tobacco smoke?
Response: We limited the study to non-smokers, but we know that some participants were still
exposed to environmental tobacco smoke.
Question: What was the public reaction to the results?
Response: The public did understand the representative sampling results. This was a case where
we had difficulty recruiting enough people to participate in the study.
61
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Monitoring: Children's Exposure to Diesel School Bus Emissions — David
Brown (EHHI/NESCAUM)
This personal exposure study among school children in Connecticut used personal monitoring to
estimate the distribution of diesel exposure among children riding school buses and over the
duration of a school day. Goals of the study included the identification of factors associated with
short term elevations in concentration, and a comparison of the personal exposure findings to
Connecticut ambient air levels.
Personal sampling was conducted using fifteen student volunteers from fifteen schools; sampling
duration began before the morning bus ride and continued until the end of the afternoon bus ride.
Fifteen sample days were collected, and activity logs were kept every fifteen minutes on sample
days. Background samples were also taken outside the schools on sample days, as were
instantaneous "grab" samples outside the buses. A second phase of the study added
measurements on surrogate bus routes for comparison, including rural areas with multiple bus
runs; varying numbers of stops, hills, and idling times; and different bus configurations. The
latter phase was prompted by the findings of a recent asthma study, which revealed asthma
frequencies in rural areas to be as high as those in urban areas, but lower than center cities.
Since students from rural areas spend more time per day in school buses [slide 6], it is suspected
that their exposure to diesel may be related to the high asthma rates.
Fifteen-minute averages of exposure, as well as continuous exposure of a student during a
school day showed a clear trend of high PM]0 concentrations during the morning and afternoon
bus rides [slides 7, 8]. Results from another student showed the same pattern, but included an
additional peak at mid-day [slide 9]; this was attributed to students being in a room by the
playground at the same time as buses were unloading another group of students there. Median
values compiled from nine students monitoring confirmed the exposure to higher concentrations
of these volatile organic hydrocarbons while students were in the buses, as opposed to other
times during the day [slide 10]. Comparison of ambient versus personal concentrations showed
four compounds in common in all monitored student samples: benzene, toluene, 2-butanone,
and methyl tertiary butyl ether (MTBE). These samples showed higher concentrations than in
nearby background monitors in all cases. For example, MTBE concentrations in personal
samples were nearly five times the ambient concentrations [slide 11]. A similar pattern was
observed with three carbonyls: formaldehyde, acetaldehyde, and acetone [slide 12]. In this case,
acetone concentrations showed a two-fold increase in personal samples as compared to nearby
background samples. Ambient data for the state of Connecticut obtained in three-hour intervals
also showed an increase in PM2 5 concentrations during the morning when buses were operating
[slide 13]; this effect was not noticed during the afternoon bus routes, likely because dilutions
increased steadily during the course of the day as the air mass was heated. Individual student
62
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
exposure to PM2 5 was higher than the average daily ambient concentrations for Connecticut in
the majority of cases [slide 15], calculated over the same hours.
The second phase of the study identified a number of factors that affect exposures to diesel.
Each letter [slide 17] represents a bus. Each bus made multiple circuits of the standardized route
over a single day. A certain degree of variability in exposure measures is evident during
replicates of the same route by the same bus [slide 17]. Some of this variability depended on
whether windows were open or closed when the bus was running [slide 18]. Idling time had a
significant effect on increased diesel concentrations, as demonstrated by monitoring of a student
who exited a bus and then re-entered after the bus had been idling for ten minutes [slide 19];
upon re-entry, the PM2 5 concentration was five times higher than before the student left the bus.
The effect of stops en route had a variable effect on exposures, which depended on the time
spent idling during stops [slide 20].
Idling buses were a major contributor to diesel PM during school days. The study demonstrated
that environmental pollution needs to be more fully characterized, as high individual exposures
may be missed by relying on ambient monitoring alone. Stronger collaboration between
environmental and public health agencies and advocates is also necessary. Regulatory and
policy action may need to be considered in some cases even without an absolute cause and effect
association. The state of Connecticut was notified before data from the school bus study were
published, and initiated a program to limit idling of school buses. Plans also include retro-fit of
the buses and moving loading and unloading zones away from areas where children congregate
before and after the school day.
Questions and Comments:
Question: The benzene ambient concentrations seem to be fairly high; are these values typical
for Connecticut?
Response: The background sites were located near each school — a different site for each school.
They may not be representative or typical of the entire state. Connecticut daily
average levels are lower; in an urban/suburban sampling site 24 hour levels of about
0.5ug/m3 with 24-hour maximums of about 1.4ug/m3 were reported. Our
"background" levels near the schools were for the school day only and may be
affected by local sources.
Question: Why were the 75th percentile data used [slide 13, in ambient PM concentrations for
Connecticut] rather than the 95th percentile?
Response: Because of the two-hour exposure, if we see respiratory effects, I would expect them
to be receptor-mediated. One hour is too short when trying to find relationships
between exposure and health responses, such as hospital admissions or asthma
attacks. We are trying to look at the 90th and 95th percentile median values as well.
63
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Note: The speaker misunderstood this question. The questioner wondered why 75th
percentile values were used rather than 95th to show diurnal trends. The 95th would
have shown a stronger peak in the morning due to reduced air mixing. The speaker
assumed that one fourth of the days with that mixing reduction would occur.
Comment: Schools are a convenient place to measure children's exposure risk because most of
them are there for a large portion of the day; thus, we know or can determine where
subjects are for the entire duration of monitoring. They provide a good, "under
control" population for experimental measurements.
64
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
National Air Toxic Assessment (NATA) - The Initial National Scale
Assessment - Ted Palma (OAR/OAQPS)
The National Air Toxic Assessment (NATA) is one of four components of EPA's Air Toxics
Program; the National Scale Assessment, part of NATA, is updated every three years. NATA
activities include emissions inventories, monitoring, air quality modeling, exposure modeling,
and risk characterization, and research on effects and assessment tools. The results of the 1996
assessment were presented in this talk. The initial draft of the 1996 assessment was completed
in the summer of 2000, and a detailed document prepared and sent for SAB review in January
2001. The SAB issued its report in December 2001. The assessment was updated to incorporate
SAB "short-term" comments and was made public via the EPA NATA website in May 2002.
The National Scale Assessment is a tool that can be used by EPA, states, tribes, and local
governments to prioritize resources and data collection and to provide a baseline for tracking
trends and measuring progress against goals. It is important to note that the assessment by itself
cannot be used as the basis for specific regulatory decisions, but must be combined with local
studies. The study considered chronic exposure by inhalation only, based on 1996 emissions
data; average exposures were calculated at the census tract level and presented at the county
level or higher. The assessment focused on thirty-three urban HAPs including diesel PM [slides
6, 7]; sources of indoor origin were excluded. The components of the assessment included the
development of an emissions inventory, modeling (air dispersion and inhalation exposure)
compared with both ambient monitoring, risk characterization, and dose-response assessment
[slide 8].
The emissions inventory is a crucial part of the assessment and used data from the 1996 National
Toxics Inventory (NTI), the 1996 VOC in National Emissions Trends Inventory, and the 1996
Heavy Duty Diesel Rule Inventory. The EMS HAP pre-processor was used to prepare inventory
data for the ASPEN model, which provided values for input to the HAPEM4 exposure model
[slide 10]. Comparison of ASPEN predictions with monitoring showed that the model predicted
benzene accurately, but did poorly for metals [slide 15]. The HAPEM4 model predicts
breathing-level concentrations by building a series of activity patterns for representative age and
gender cohorts using data from the Consolidated Human Activity Database (CHAD). Cohorts'
movements are tracked by the model through one year and thirty-seven micro-environments in
order to determine a composite breathing-level concentration. Micro-environment
concentrations are determined by the model as a function of ambient concentration, and
population exposures are extrapolated from cohort exposures using census population data for
each tract. These are important strengths of HAPEM4, as activity patterns show that the
majority of people spend very large portions of their time indoors. The model also allows for
commuting between tracts, has the capacity for improving predictions by adding information
65
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
such as more cohorts (e.g., children) and better activity patterns. HAPEM4, however, is not
designed to predict extremes in distribution of exposures; in addition, the model still needs to be
fully evaluated against personal monitoring data [slide 20].
HAPEM4 results based on the initial national scale assessment predicted exposure
concentrations that were generally lower than the ambient concentrations predicted by ASPEN
[slide 21]. On-road mobile gaseous HAPs were an exception, with a predicted exposure of
101% of ambient. Exposures were also calculated for each cohort [slide 22]. Both cancer and
non-cancer risk were characterized; median cancer risk (cumulative) was fifty-one per million,
with a range from less than one to more than one hundred per million [slides 23, 24]. Cancer
risks (average) were also attributed to each compound, as well to different sources of pollutants
[slides 25, 26]. National drivers were benzene, chromium, and formaldehyde for cancer, and
acrolein for non-cancer effects; several different compounds were identified as regional drivers
[slide 27] and contributors [slide 28]. Other compounds were neither risk drivers nor
contributors; however, the national assessment alone cannot exonerate HAPs, as it has low
resolution and calculates risk based on inhalation only [slide 29]. Diesel exhaust was identified
as a likely human carcinogen which could dominate risk from all HAPs. ORD will shortly
finalize the Diesel Health Assessment Document. Results from the 1996 National Scale
Assessment are available through the NAT A website:
http://www.epa.gov/ttn/atw/nata/
A final advisory on the National Scale Assessment was issued by the SAB in December 2001.
noting that the effort represented "an important first step towards characterizing the relationships
between sources and risks of HAPs." Most SAB comments were provided as recommendations
- both short- and long-term - on aspects of the assessment including the inventory, modeling.
dose-response, risk characterization, uncertainty, and variability [slide 33, 34]. Future directions
for the assessment include a case study in Houston to include indoor and personal monitoring
and a variability study; publication of the first in a series of NATA technical documents; and
modeling for the 1999 assessment. Ongoing projects include using databases to identify patterns
and data gaps, comparing results with local and urban-scale assessments, and promoting the use
of results to facilitate planning and data-gathering activities.
66
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Questions and Comments:
Comment: The Agricultural Policy/Environmental extender (APEX) or SHEDS model can be
used for local-scale assessments.
Question: Why can the model not be validated using actual exposure data?
Response: We can do that, but we are waiting for new data, since we do not want to validate a
model with the same data used to develop it.
Question: Workplace exposures can be significant, as people spend forty or more hours at their
workplace each week. How are they handled in the national assessment?
Response: Workplace exposure is taken into account in the activity patterns.
67
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
THIS PAGE INTENTIONALLY LEFT BLANK
68
-------
US Environmental Protection Agency
Region/ORD/QAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
SESSION V: SOURCE APPORTIONMENT
Chair: Chad Bailey (OAR/OTAQ)
Chad Bailey (OAR/OTAQ) began Session V by recapping that participants had thus far been
exposed to different types of information on the relationship between exposure and health
effects, and to modeling tools helpful in predicting them. The focus of Session V was source
apportionment, or source-receptor modeling; presentations included an introduction to tools and
methods used in source apportionment, as well as a case study.
Introduction to Source Apportionment Methods - Lynn Hildemann (Stanford
University)
The goal of source apportionment is to determine the contributions of pollution sources to a
location of interest, using a "receptor"-oriented approach. Commonly used methods include
chemical balance modeling (CMB) and two more recently developed methods: principal
component analysis (PCA) and positive matrix factorization (PMF).
The chemical mass balance approach uses previously measured emissions compositions from
known sources to determine which of the sources have contributed to a sample and to what
extent. This technique works only when complete emissions composition information is
available for all sources - it cannot be used when source or data are missing. When enough
information is available, CMB can be used to analyze single receptor samples and identify
specific, well-known sources. Model outputs must be carefully examined, however, to ensure
that problems such as missing sources are identified. This approach is also not able to resolve
the chemical contributions of sources whose emissions have similar chemical compositions.
Tracer species used in CMB must be non-reactive; although trace metals and ions have
traditionally been used, recent work has focused on more stable organic aerosol tracers. An
example of CMB application is a study that used the technique to attribute ambient
concentrations of PM in the Los Angeles area to contributing sources [slides 9, 10].
Principal component analysis (PCA) can be utilized only when numerous receptor samples are
available. The model determines which of several "unknown factors" are contributing to the
composition of each sample. Expert input is then required, as the model user must decide which
factors to retain, and judge what sources are represented by each factor. Occasionally this
technique results in negative values for chemical components, or in factors that cannot clearly be
connected to sources. Another drawback of the model is that it can yield many potential
solutions that cannot be weighed to account for uncertainty. PCA does not require source
emission compositions as input, however, and can help identify sources that were not originally
69
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
considered as potential contributors. A recent application of PC A used indoor samples from
non-smoking residences and assigned emissions to six sources of pollutants [slide 13].
Positive matrix factorization is a recent advance in source apportionment modeling which, like
PCA, does not need source compositions as inputs. Also like PCA, it identifies "factors" and
their contributions to a receptor sample. Unlike PCA, PMF can account for uncertainties in the
input measurements, and can be used even when input data are missing or below the detection
limits. In addition, PMF constrains solutions to be greater than zero, avoiding the problem of
factors with negative components. A 1999 study using PMF modeling examined 178 personal
exposure samples and determined source contributions by regressing PM concentrations against
the factors identified by PMF [slides 16, 17].
A table presented [slide 18] summarized the three methods and showed the trade-offs involved
in using one versus another. A problem for all methods discussed is the difficulty of
distinguishing among sources with similar chemical signatures.
70
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Source Apportionment Tools - Gary Nonis (ORD/NERL)
The National Exposure Research Laboratory is currently involved in developing and testing
several source apportionment modeling tools, including: a new CMB model expected by
November 2002; the EPA Unmix 2.3 model; and the positive matrix factorization (PMF) /
Multilinear Engine (ME) model. The latter two models were discussed, and examples of their
use in air toxics studies were presented.
Unmix 2.3 performs factor analysis constrained to yield non-negative numbers to generate
source profiles (including uncertainties) and source contributions from ambient data. A stand-
alone version is available that runs on the Windows 98™ operating system. Unlike PMF
models, Unmix 2.3 makes no explicit use of ambient data uncertainties. The user-interface
allows input of the number of sources, and provides a "feasible solution" for a set of data. A
recent study used hourly PAMS data from El Paso, Texas, and identified gasoline vapor (11%),
propane (10%), and non-vehicle emissions (79%) as the three benzene emissions sources and
their contributions [slides 5,6].
Positive matrix factorization was used to perform an analysis of the composition of PM
measured in the 1998 EPA Baltimore Exposure Panel study. The exposure study population
consisted of elderly subjects residing in a retirement facility in Baltimore, Maryland.
Community, outdoor, and indoor samples were taken using the versatile air pollutant sampler
(VAPS); additional sampling was conducted using a personal exposure monitor (PEM) for all
the above locations as well as apartment and personal monitoring. When PEM samples yielded
higher soil and trace element oxide (TEO) concentrations than the VAPS samples, data from the
two samplers were analyzed separately [slide 10]. Another surprising finding was a large
difference between organic carbon mass concentration collected with a quartz filter and PM2 5
collected with the VAPS sampler (Teflon filter) [slide 11]; this difference was attributed to an
artifact in measuring. PMF modeling was used to identify the sources of PM25 contributing to
infiltration (outdoors to indoors) and to personal exposure. The average contributions of five
sources [slides 13, 14] showed nitrate, sulfate, and motor vehicle exhaust to be the most
significant. The contribution of organic carbon was very high (68%), influenced once again by
an artifact. The multilinear engine 2 (ME) model was used to identify the sources contributing
to personal exposure, ensuring that as much of the observed concentration as possible is
explained by external factors [slide 15]. External factors of importance were sulfate. crustal, and
an unknown factor. The two major internal factors contributing were a combination of sulfate,
silicon, and calcium (possibly wallboard), and a factor similar to outdoor PM in chemical
profile; personal care products played a smaller role [slide 17]. The unidentified outdoor PM
factor was measured on personal and apartment samples but was absent from the indoor or
outdoor stationary monitors. Activity data may have been helpful in resolving uncertainty, e.g.,
71
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
time spent in vehicles and outdoors may have explained the outdoor PM contributions to
personal exposure [slide 18]. Activity data provided more information on one of the internal
factors (wallboard): although the data were used outside the model, they revealed that
contribution of this factor increased while residents were dusting and using vacuum cleaners
[slide 20].
Planned research stemming from the Baltimore study will examine the organic carbon artifact,
and include time-activity data in the receptor model. Regarding personal exposure to motor
vehicle PM exhaust, efforts are needed to find a tracer, since lead can no longer be used.
72
-------
US Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
A Modern Example: Northern Front Range Air Quality Study - Chad Bailey
(OAR/OTAQ)
The Northern Front Range Air Quality Study (NFRAQS) was conducted in 1996-1997 in the
Denver, Colorado urban region as a multi-sponsor research project managed by Colorado State
University. Study objectives were to attribute air pollution to sources in order to support air
quality management decisions.
Source apportionment focused on carbonaceous aerosol, ammonia (if present), and PM25.
Ambient sites analyzed for carbon (including radioisotopes), ions, and elements (eight sites), as
well as for "extended species" to be used for CMB modeling (two sites). Several potential
sources were used, including four different types of motor vehicle exhaust profiles [slide 5].
Chemical mass balance modeling was selected because a combination of pre-defined source
profiles was available; all sources of analyte species must be included in CMB modeling for
accurate results. Motor vehicle exhaust was found to be the largest emissions contributor,
followed by ammonium nitrate; dust, debris, and ammonium sulfate had smaller contributions
[slide 8]. Apportionment of mobile source emissions revealed gasoline high emitter exhaust and
cold start exhaust to be bigger contributors than diesel exhaust [slide 9]. This result was the
opposite of inventory findings, which estimated diesel percent contribution at about three times
that of gasoline exhaust. NFRAQS results, if accurate, have important implications about
mobile source emissions, especially high emitters; the inventory did not consider high emitting
or cold start gasoline vehicles, and may be apportioning mobile emissions sources incorrectly.
Studies are ongoing to collect more data to check the results of the NFRAQS study.
The use of CMB modeling is the source of some of the limitations of the study. It requires that
all sources of model species be characterized; however, the NFRAQS study did not consider
non-road engines, and may have overlooked an important, chemically different source (a major
rail line). In addition, CMB assumes stable source profiles and the collection and analysis of
source and ambient samples by the same means; this poses a problem for mobile sources, as the
source is measured at the exhaust, but sampling is ambient. Idling versus driving can also
change the exhaust composition.
The NFRAQS study also has implication for human exposure studies. Since CMB requires all
sources to be well-analyzed and modeled, factor analysis approaches may be more useful in
exposure work until a complete indoor source inventory is available.
73
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Panel Discussion with Session V Speakers
Comment: It seems that there are a lot of diesel engines, and since they last a long time there
may also be a time factor to consider.
Response: The longevity of these engines is certainly an issue for consideration.
Question: Are there default systems that allow you to find out how far off you are when using
the CMB model - especially if you are missing a source?
Response: There are statistical measures to determine how good a fit your answer is; the
expertise of the user has a big role in the decision involved.
Question: Have there been any attempts to bring information on natural gas vehicles to the
attention of decision-makers?
Response: The Office of Transportation and Air Quality (OTAQ) has been watching this issue
closely, and a major regulation was passed last year to reduce particle emissions.
Question: Why were the indoor and apartment samples different?
Response: Indoor samples were taken from common areas, as opposed to inside individual units.
Question: What was the source of ammonium nitrate in the Denver study?
Response: There was enough ammonium in the atmosphere to convert nitrate to a particulate
form.
Question: Would using different classes rather than individual chemicals help simplify source
apportionment modeling?
Response: There are methods that can be used to do this, usually as an approach to risk
assessment. However, I would be less comfortable assessing risk based on sources.
Comment: If we had a sense of proportion of the risk we could assign it to different categories
and use that information in source apportionment studies.
Response: The purpose of source apportionment is to identify chemicals and associate them with
sources.
Comment: This type of risk analysis information could be used in terms of personal exposure.
74
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
SESSION VI: COMMUNICATING THE RESULTS AND WORKSHOP
CONCLUSIONS
Chair: Periann Wood (Region 9)
Periann Wood (Region 9) introduced Captain Alvin Chun - a Captain with the United States
Public Health Service on assignment at EPA - who talked about communicating risk assessment
and results to the public.
Communicating the Results of Air Toxics Exposure Assessments - Alvin Chun
(Region 9)
Communicating scientific results to the public is an important function of government scientists,
particularly when it relates to risk from harmful substances. It is usually counter-intuitive to
scientists to share initial results with a wide audience, such as a community — especially when
those results can be considered "bad news." The first reaction to such findings is to verify them
by collecting further data. Although this is a sound scientific approach, it may not be the best
option from the perspective of people who might be affected. Delaying the communication of
data can give the impression that the government Agency involved is hiding information from
the public. Involving the community early (i.e., before the sampling plans have been finalized)
also has the advantage of giving them some confidence or awareness of our efforts, and
developing a relationship with community representatives, who may serve as allies or liaisons in
future public meetings.
Successful meetings require that the experts who are frequently called upon to answer residents'
questions be prepared with direct and relevant information and know what to expect.
Participants in public meetings may be angry because of losses they may be experiencing, the
relevance of the information presented, or they may not understand what the results mean; often,
they may not believe the experts' or government representatives' answers. This is likely to
happen when the public has not been sufficiently included in the process early on, when
scientists are not direct, or do not have enough information to say with certainty what risk
assessment results mean for public health. The matter of defining risk is a frequent point of
contention: the public may understand "safe'' to mean zero risk level, but standards are rarely set
at that level. Letting the public know prior to sampling what the standards are and how they
relate to risk can be very helpful in building trust and preparing scientists to answer related
questions. When limited information is available, safe and unsafe levels should be defined from
the start to avoid any perception that levels were set after data was gathered. Action levels
outlining what the response will be depending on concentrations found should also be
established early, presented, and even negotiated with the public.
75
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
In a case study from a community located next to a Superftmd site, high levels of vinyl chloride
were found near some homes. Vinyl chloride is a Class A carcinogen to which children are
particularly susceptible. The regional office did not immediately communicate their findings
due to concerns regarding potential uncertainties, and to avoid unnecessary alarm. The
community was eventually informed nine months later. A better approach in this situation may
have been for EPA to state their concerns about the data and about prematurely frightening the
residents; then, present their strategy for determining contaminant levels and the actions that
would follow. In addition to earning the trust of the community, early com.-; ;unication offers
control to the citizens, provides an opportunity to explain and discuss the need for further study,
and request permission to sample inside homes. Confident answers by experts are essential in
early intervention meetings, as incomplete or inconsistent answers may give the impression of
incompetence. When answers are not immediately available because further study is needed.
providing the public with a time line of when these results will be available and offering periodic
updates will provide some positive reassurance. Groups who might be especially concerned and
whose support and understanding we may need include parents, community leaders, real estate
agents, and the County Health Department.
Preparation for involving citizens, such as in meetings or one-to-one visits, can also be enhanced
by understanding the losses experienced by people affected - including loss of control, potential
loss of health, property value, quality of life, harm to their pets, confidence, and trust in
government or county officials. Participants in these meetings may also attend with a pre-
conceived perception of government representatives as uninterested, incompetent, or even
dishonest. Acknowledging that poor relationships or experiences with government may be
commonplace can help government officials accept public anger without taking it personally.
Public interactions and meetings can serve as an opportunity to change this opinion over time so
that EPA scientists may later be viewed as committed professionals interested in protecting the
public and resolving the problem. Seriously listening to concerns when they are dealing with
losses is a more effective strategy than interrupting to convey information Although scientists
are valued by their peers primarily for their knowledge, the public responds more favorably to
representatives who are initially empathetic, open, and honest. Such response will help scientists
to understand the problem from the public's perspective, thus earning the trust needed for
citizens to consider the scientist's knowledge. Listening to residents' suggestions may also be
helpful, as people are familiar with their own community; they may provide useful information,
e.g., where to take samples. Once convinced that the interests of the community are a priority
for the scientists, members of the public will be more likely to believe and accept or consider
their data, opinions, and recommendations.
76
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Questions and Comments
Question: What happens when you know what the community wants or needs, but cannot give
it to them?
Response: If you have paid attention to them and established a good rapport, they may be more
understanding and accept that you cannot help them.
Comment: Communities/people still want to be respected, even if in the end you cannot help
them.
Question: Do you actually tell people that environmental situations are "safe"?
Response: Yes, as long as you explain how you are defining "safe"; or, you can say it without
using the word "safe", and yet convey to them that it is acceptable.
Question: In one case, at another Superfund site, action levels were set and communicated to the
residents prior to delivering the results to residents. A short time after that, the EPA
changed the action levels. How would you deal with that?
Response: You would have to tell the truth and give the reasons for the change, and, in this case,
expect a loss of trust and subsequent anger. You may have to negotiate the levels, or
actions in order for people to feel reassured. Even following negotiations, the level
that you set may be immaterial to the community or more stringent than you had
estimated (and if the goal is reassurance, more stringency may be what it takes).
Comment: The community may be aware that an action such as a level change may have been a
political decision.
Comment: In a case like this it would be very important to talk to the community frequently
during the entire process and even alert them to the things that will influence the final
level.
Comment: There has been talk recently of relaxing the EPA's standards for cleaning up
buildings.
77
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Conclusions and Next Steps - David Klauder (ORD/OSP)
David Klauder (ORD/OSP) brought the workshop to a close and informed participants that all
presentations would be posted on the internet one week after the workshop. The address to the
internet site is:
http://epa.gov/osp/regions/workshops.htm
The workshop planning group was acknowledged for preparing a successful meeting.
One of the goals of this workshop was to build networks that will lead to long-term Region/ORD
partnerships and collaborative projects. The closing discussion centered on the generation of
ideas for activities for possible follow-up activities among workshop participants that would
further these goals. The following is a summary of these ideas.
Suggested Follow-Up Actions
High Priority Research and Research Product Needs
• Greater Regional participation in the August 2002 revision of the Air Toxics Multi-
year Plan (MYP) and possibly other ORD-led research planning activities (Randy
Robinson).
• Identify opportunities to address critical methods development needs (Ken Mitchell).
• Develop tools for community-level assessments and research that would improve
these tools (Paul Shapiro & Ken Mitchell).
Technology Transfer
• Enhance the utility of the National Guidance Document as a technical resource
manual to the Regions (Paul Shapiro & Ken Mitchell).
• Broader Agency involvement in the Regional Urban Air Toxics Initiative pilot
program to help identify tools for community-level assessments (Pam Tsai).
• Conduct workshops and/or training sessions on the application and interpretation of
results from specific methods, models, and other tools presented in this workshop
(Joe Touma & Randy Robinson).
• Conduct follow-up workshops on related topics, e.g., exposure modeling, air risk
assessment, diesel emissions and other mobile sources (from Evaluation Sheets).
78
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
Communications
• Create an Agency expert directory for key Air Toxics science topics (Alan
VanArsdale).
• Improve mechanisms, e.g., EPA web links and Agency Science Inventory, for
informing interested Agency scientists about existing products and relevant research,
including how people are using data and tools (Haluk Ozkaynak).
• Greater cross-Agency involvement on existing air toxics conference calls, e.g. the air
toxics monitoring call organized by Motria Poshyvanyk (Paul Shapiro).
79
-------
US Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
THIS PAGE INTENTIONALLY LEFT BLANK
80
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
APPENDIX A: AGENDA
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
EPA Region 9 Office, 75 Hawthorne St., San Francisco, CA
June 25-27, 2002
Workshop to Focus on Two General Exposure Assessment Questions:
1. What is our inhalation exposure to toxic chemicals of concern in our Region (at a
"screening level of certainty")?
2. What is our inhalation exposure to toxic chemicals of concern in our community (at a
"high level of certainty ") ?
JUNE 25th - MORNING
8:30 AM - 9:45 AM Welcome and Introduction
8:30 Welcome - Jack Broadbent, Director (Air Division - Region 9), Larry Cupitt, Acting
Associate Director of Health (ORD/NERL), and Winona Victery (Region 9)
8:45 Overview of Exposure Assessments - Ken Mitchell (Region 4) and Tim Watkins
(ORD/NERL)
9:45 AM -12:30 PM Session I: Designing an Air Toxics Exposure Assessment -
Monitoring vs. Modeling
Co-chairs: Ken Mitchell (Region 4), Paul Shapiro (ORD/NCER), and Ted Palma
(OAR/OAQPS)
9:45 Monitoring vs. Modeling - An Interactive Group Exercise: Paul Shapiro (ORD/NCER)
and Ted Palma (OAR/OAQPS)
10:15 BREAK
10:30 Case Studies to Illustrate Uses of Monitoring and Modeling
MATES II-A Regional Perspective: Mike Nazemi (South Coast Air Quality
Management District)
Minneapolis - St. Paul - A Comparison of Community, Residential, and Personal
Exposure: John Adgate (University of Minnesota)
12:30 LUNCH
A-l
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
JUNE 25th - AFTERNOON
1:30 PM - 2:30 PM Session I: Designing an Air Toxics Exposure Assessment - Monitoring
vs. Modeling (cont'd)
1:30 Expert Panel
Mike Nazemi (SCAQMD)
John Adgate (UMN)
John Girman (OAR/ORIA)
Larry Cupitt (ORD/NERL)
Neil Frank (OAR/OAQPS)
Matt Lorber (ORD/NCEA)
Joe Touma (OAR/OAQPS)
2:30 PM - 5:30 PM Session II: Monitoring Methods and Network Design
Co-chairs: Motria Poshyvanyk (Region 5), Neil Frank (OAR/OAQPS), and Tim Watkins
(ORD/NERL)
2:30 Air Toxics Monitoring Pilot Project: Barbara Morin (Department of Environmental
Management, Rhode Island)
3:00 New Trends in Monitoring Methods: Tim Watkins (ORD/NERL) (for Don Whitaker)
3:30 BREAK
3:45 Atmospheric Formation and Decay of Air Toxics - Implications for Exposure Assessments:
Deborah Luecken (ORD/NERL)
4:30 Air Toxics Monitoring Methods and Network Design
Steve Bortnick and Shannon Stetzer (Battelle Memorial Institute)
5:15 General Discussion
5:30 Adjourn to Thirsty Bear (brew pub next to the Region 9 meeting room)
JUNE 26th - MORNING
8:30 AM -12:00 PM Session II: Monitoring Methods and Network Design (cont'd)
8:30 California Monitoring Program: Statewide Network to the Neighborhood Scale
Jeff Cook and Linda Murchison (California Air Resources Board)
A-2
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
10:00 BREAK
10:15 Design Your Own Air Toxics Monitoring Network (an interactive group exercise)
Motria Poshyvanyk (Region 5) - lead
12:00 LUNCH
JUNE 26th - AFTERNOON
1:00 PM - 5:30 PM Session III: Modeling Tools - Current and Future
Co-chairs: Randall Robinson (Region 5), Joe Touma (OAR/OAQPS), and Tim Watkins
(ORD/NERL)
1:00 Air Quality Models: Joe Touma (OAR/OAQPS)
1:15 Using Emission Inventories for Air Quality Modeling: Joe Touma (OAR/OAQPS)
1:30 Model Applications
A. Local Scale - Barrio Logan Modeling Analysis: Vlad Isakov (California Air Resources
Board)
2:00 B. Urban Scale Modeling - Houston Case Study: Joe Touma (OAR/OAQPS)
2:30 Applying CMAQ ModelsSfor Air Toxics Assessments: Bill Hutzell (ORD/NERL)
3:00 Q & A with Speakers
3:30 BREAK
3:45 Breakouts to Discuss Modeling Questions — Potential Topics:
* How active are the Regions in ambient air modeling?
* What types of assessments are being done?
* What are the priorities? What is important?
* Where do the Regions want to be with ambient air toxics modeling and skills?
* What obstacles exist? What do the Regions need? Guidance? Tools? Etc.
* What is needed to overcome these obstacles?
* Suggested workshop follow-up items.
5:00 Breakout Reports and Wrap Up Discussion
5:30 Adjourn
A-3
-------
U.S. Environmental Protection Agency
Rcgion/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
JUNE 27th - MORNING
8:00 AM - 12:00 PM Session IV: Human Exposure Assessment
Co-chairs: Alan VanArsdale (Region 1) and Tim Watkins (ORD/NERL)
8:00 Introduction to Human Exposure: Linda Sheldon (ORD/NERL)
8:30 Air Toxics Exposure in Indoor Environments: John Girman (OAR/ORIA)
8:50 Multi-pathway Exposure Assessment: Matthew Lorber (ORD/NCEA)
9:10 What Human Exposure Data and Models are Available?: Haluk Ozkaynak (ORD/NERL)
9:30 Q & A with Speakers
10:00 BREAK
10:15 Additional Case Studies:
A. Personal Exposure Monitoring Meets Risk Assessment: The South Baltimore
Community Exposure Study: Devon Payne-Sturges (OA/OPEI)
10:50 B. Monitoring: Children's Exposure to Diesel Emissions: David Brown (NESCAUM)
11:25 C. Modeling: National Air Toxic Assessment (NATA) - The Initial National Scale
Assessment: Ted Palma (OAR/OAQPS)
12:00 LUNCH
JUNE 27th - AFTERNOON
1:00 PM - 2:30 PM Session V: Source Apportionment
Chair: Chad Bailey (OAR/OTAQ)
1:00 Introduction to Source Apportionment Methods: Lynn Hildemann (Stanford University)
1:20 Source Apportionment Tools: Gary Norris (ORD/NERL)
1:50 Case Study: A Modern Example: Northern Front Range Air Quality Study: Chad Bailey
(OAR/OTAQ)
A-4
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
2:10 Panel Discussion with Speakers
2:30 BREAK
2:45 PM - 4:30 PM Session VI: Communicating the Results and Workshop Conclusions
Chair: Periann Wood (Region 9)
2:45 Communicating the Results of Air Toxics Exposure Assessments: Alvin Chun (Region 9)
4:00 Conclusions and Next Steps - Winona Victery (Region 9) and David Klauder (ORD/OSPj
4:30 Adjourn
A-5
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
THIS PAGE INTENTIONALLY LEFT BLANK
A-6
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
APPENDIX B: LIST OF PARTICIPANTS
EPA Regional Offices
Beth Antley (Attendee)
Region 4
980 College Station Rd.
Athens, GA 30605
Tel: 706-355-8620
Fax: 706-355-8744
E-mail: antley.beth@epa.gov
Mike Bandrowski (Attendee)
Region 9 (AIR-6)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4194
Fax: 415-947-3583
E-mail: bandrowski.mike@epa.gov
Thomas Baugh (Attendee)
U.S. EPA, ORA
Region 4 (14th Floor)
61 Forsyth St. SW
Atlanta, GA 30303
Tel: 404-562-8275
Fax: 404-562-8269
E-mail: baugh.thomasL@epa.gov
Carol Bcllizzi (Attendee)
Region 2
290 Broadway
New York, NY 10007-1866
Tel: 212-637-3712
Fax: 212-637-3901
E-mail: bellizzi.carol@epa.gov
Carol Bohnenkamp (Attendee)
Region 9 (AIR-7)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4130
Fax: 415-947-3579
E-mail: bohnenkamp.carol@epa.gov
Scott Bohning (Attendee)
Region 9 (AIR-7)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4127
Fax: 415/947-3579
E-mail: bohning.scott@epa.gov
Stacy Braye (Attendee)
Region 9 (WST-4)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3345
Fax: 415-947-3533
E-mail: braye.stacy@epa.gov
Jack Broadbent (Speaker)
Region 9 (AIR-1)
75 Hawthorne St.
San Francisco, C A 94105
Tel: 415-947-4141
E-mail: broadbent.jack@epa.gov
John Brock (Attendee)
Region 9 (AIR-5)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3999-
Fax: 415-947-3579
E-mail: brock.john@epa.gov
B-l
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27. 2002
Catherine Brown (Attendee)
Region 9 (AIR-7)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4137
Fax: 415-947-3579
E-mail: brown.catherine@epa.gov
Nicole Buffa (Attendee)
Region 9 (CMD4-1)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4161
E-mail: buffa.nicole@epa.gov
Ray Chalmers (Attendee)
Region 3 (3API 1)
1650 Arch St.
Philadelphia, PA 19103-2029
Tel: 215-814-2061
Fax: 215-814-2134
E-mail: chahners.ray@epa.gov
Tai-Ming Chang (Attendee)
Region 9 (CMD-1)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3831
E-mail: chang.tai-ming@epa.gov
Alvin Chun, Captain (Speaker)
Region 9 (AIR-6)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4190
Fax: 415-947-3583
E-mail: chun.alvin@epa.gov
Kuenja Chung, Ph.D. (Attendee)
Region 6 (6pdq)
Air Quality Analysis Section
1445 Ross Ave.
Dallas, TX 75202
Tel: 214-665-8345
Fax: 214-665-6762
E-mail: chung.kuenja@epa.gov
Larry Cupitt, Ph.D. (Speaker)
Region 4 (D305-01)
109 T.W.Alexander Dr.
Research Triangle Park, NC 27709
Tel: 919-541-0349
Fax: 919-541-3615
E-mail: cupitt.larry@epa.gov
Bruni Davila (Attendee)
Region 9 (SFD7-4)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3162
E-mail: davila.brunilda@epa.gov
Richard Daye (Attendee)
Region 7
901 North 5th St.
Kansas City, KS 66101
Tel: 913551-7619
Fax: 913 551-9619
E-mail: daye.richard@epa.gov
Dana Dean (Attendee)
Region 9
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3872
E-mail: dean.dana@epa.gov
B-2
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Lynda Deschambault (Attendee)
Region 9 (CMD4-2)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4183
Fax: 415-947-3583
E-mail: deschambault.lynda@epa.gov
Kathy Diehl, MSB (Attendee)
Region 9 (AIR-8)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415972-3996
Fax: 415947-3579
E-mail: diehl.kathy@epa.gov
Sylvia Dugre (Attendee)
Region 9 (AIR-2)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4149
Fax: 415-947-3579
E-mail: dugre.sylvia@epa.gov
Todd Ellsworth (Attendee)
Region3(3AP21)
1650 Arch St.
Philadelphia, PA 19103-2029
Tel: 215-814-2195
Fax: 215-814-2124
E-mail: ellsworth.todd@epa.gov
Gwen Eng (Attendee)
ATSDR
Region 9 (HHS-1)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4317
Fax: 415-947-4323
E-mail: eng.gwen@epa.gov
Danny France (Attendee)
Region 4
980 College Station Rd.
Athens, GA 30605
Tel: 706-355-8738
Fax: 706-355-8744
E-mail: france.danny@epa.gov
Gerald Hiatt, Ph.D. (Attendee)
Region 9 (SFD-8B)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3064
Fax: 415-947-3518
E-mail: hiatt.gerald@epa.gov
Ofia Hodoh (Attendee)
Region 4
61 Farsyth Street, SW
Atlanta, GA 30303
Tel: 404-562-9176
Fax: 404-562-9095
E-mail: hodoh.ofia@epa.gov
Peter Kahn (Attendee)
Region 1 (EGA)
11 Technology Dr.
North Chelmsford, MA 01863
Tel: 617-918-8392
Fax: 617-918-8417
E-mail: kahn.peter@epa.gov
Jeff KenKnight (Attendee)
OAQ
Region 10 (OAQ-107)
1200 Sixth Ave.
Seattle, WA 98101
Tel: 206-553-6641
Fax: 206-553-0110
E-mail: kenknight.jeff@epa.gov
B-3
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Robert Kotchenruther, Ph.D. (Attendee)
Region 10 (OEA-095)
1200 Sixth Ave
Seattle, WA 98101
Tel: 206-553-6218
Fax: 206-553-8210
E-mail: kotchenruther.robert@epa.gov
Richard Lessler, Ph.D (Attendee)
Region 9 (AIR-6)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4197
Fax: 415-947-3583
E-mail: lessler.richard@epa.gov
Libby Levy (Attendee)
ATSDR
Region 9 (HHS-1)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4319
Fax: 415-947-4323
E-mail: levy.libby@epa.gov
Brook Madrone (Attendee)
Region 10(OAQ-107)
1200 6th Ave.
Seattle, WA 98101
Tel: 206-553-1814
Fax: 206-553-0110
E-mail: madrone.brook@epa.gov
Kenneth Mitchell, Ph.D. (Speaker)
Region 4
61ForsythSt.,SW
Atlanta, GA 30303
Tel: 404-562-9065
Fax: 404-562-9095
E-mail: mitchell.ken@epa.gov
William Nelson (Attendee)
ATSDR
Region 9 (HHS-1)
75 Hawthorne St., Suite 100
San Francisco, CA 94105
Tel: 415-947-4316
Fax: 415-947-4323
E-mail: wqnl@cdc.gov
Phuong Nguyen (Attendee)
Region5(AR-18J)
77 West Jackson Blvd.
Chicago, IL 6064
Tel: 312-886-6701
Fax: 312-886-5824
E-mail: nguyen.phuong@epa.gov
Francis O'Neill (Attendee)
Region 6 (6MD-HA)
10625FallstoneRd.
Houston, TX 77099-4303
Tel: 2819832181
Fax: 281 983 2248
E-mail: oneill.francis@epa.gov
Cheryl Overstreet (Attendee)
Region 6 (6PD-NB)
1445 Ross Ave.
Dallas, TX 75202
Tel: 214-665-6643
Fax: 214-665-7263
E-mail: overstreet.cheryl@ epa.gov
Coe Owen (Attendee)
Region 9 (AIR-7)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4134
Fax: 415-947-3579
E-mail: owen.coe@epa.gov
B-4
-------
U.S. Environmental Protection Agency
Reeion/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Motria Poshyvanyk (Organizer)
Region5(AR-18J)
77 West Jackson Blvd.
Chicago, IL 60604
Tel: 312-886-0267
Fax: 312-886-5824
E-mail: poshyvanyk.motria@epa.gov
Randy Robinson (Organizer)
Region5(AR-18J)
Metcalfe
77 West Jackson Blvd.
Chicago, IL 60604
Tel: 312-353-6713
Fax: 312-886-0617
E-mail: robinson.randall@epa.gov
Keith Rose, Ph.D. (Attendee)
Region 10 (OAQ-107)
1200 Sixth Ave.
Seattle, WA 98101
Tel: 206-553-1949
Fax: 206-553-0110
E-mail: rose.keith@epa.gov
Margaret Sieffert (Attendee)
Region5(AR-18J)
77 West Jackson Blvd.
Chicago, IL 60604
Tel: 312-353-1151
Fax: 312-886-0617
E-mail: sieffert.margaret@epa.gov
Michael Sivak (Attendee)
Region 2
290 Broadway, 18th Floor
New York, NY 10007
Tel: 212-637-4310
Fax: 212-637-4360
E-mail: sivak.michael@epa.gov
MaryBeth Smuts, Ph.D. (Attendee)
Region 1 (CAP)
Suite 1100
One Congress St.
Boston, MA 02114
Tel: 617-918-1512
Fax: 617-918-1505
E-mail: smuts.marybeth@epa.gov
Wienke Tax (Attendee)
Region 9 (AIR-2)
P.O. Box 86825
Tucson, AZ 85754-6825
Tel: 520-622-1622
Fax: 520-622-1622
E-mail: tax.wienke@epa.gov
Barbara Toole O'Neil (Attendee)
Region 9 (Air-5)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3991
Fax: 415-947-3579
E-mail: tooleoneil.barbara@epa.gov
Pam Tsai, Sc.D. (Attendee)
Region 9 (Air-6)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4196
Fax: 415-947-3583
E-mail: tsai.pani@epa.gov
Alan VanArsdale (Attendee)
Region 1 (EGA)
New England Regional Lab.
11 Technology Dr.
North Chelmsford, MA 01863
Tel: 617-918-8610
Fax: 617-918-9397
E-mail: vanarsdale.alan@epa.gov
B-5
-------
U.S. Environmental Protection Agency
Region/ORD/QAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Winona Victery, Ph.D. (Organizer)
Region 9 (PMD-1)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3736
Fax: 415-947-3558
E-mail: victery.winona@epa.gov
Lawrence Wapensky (Attendee)
Region 8 (8P-AR)
999 18th St.
Denver, CO 80202
Tel: 303-312-6043
Fax: 303-312-6064
E-mail: wapensky.lawrence@epa.gov
Patrick Wilson, Ph.D., M.P.H (Attendee)
Region 9 (WST-5)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3354
Fax: 415-947-3530
E-mail: wilson.patrick@epa.gov
Periann Wood, Ph.D. (Attendee)
Region 9 (A -7)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-947-4187
Fax: 415-947-3583
E-mail: wood.periann@epa.gov
Jeff Woodlee, CIH (Attendee)
Health & Safety
Region 9 (PMD2-1)
75 Hawthorne St.
San Francisco, CA 94105
Tel: 415-972-3740
Fax: 415-947-3558
E-mail: woodlee.jeff@epa.gov
Other EPA Offices
Chad Bailey (Speaker)
OAR (ASD/ATC)
2000 Traverwood Dr.
Ann Arbor, MI 48105
Tel: 734-214-4954
Fax: 734-214-4939
E-mail: bailey.chad@epa.gov
Barbara Driscoll (Attendee)
U.S. EPA /OAQPS, ESD/PPSG
OAR (C439-04)
Research Triangle Park, NC 27711
Tel: 919-541-1051
Fax: 919-541-0942
E-mail: driscoll.barbara@epa.gov
Neil Frank, Ph.D. (Speaker)
OAR ©304-01)
Research Triangle Park, NC 27711
Tel: 919-541-5560
Fax: 919-541-3613
E-mail: frank.neil@epa.gov
John Girman (Speaker)
OAR (6609J)
1200 Pennsylvania Ave, NW
Washington, DC 20460
Tel: 202-564-9317
Fax: 202-565-2071
E-mail: girman.john@epa.gov
David Guinnup, Ph.D. (Attendee)
U.S. EPA/OAQPS
OAR(C404-01)
Research Triangle Park, NC 27711
Tel: 919-541-5368
Fax: 919-541-0840
E-mail: guinnup.dave@epa.gov
B-6
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Barbara Morin (Speaker)
RIDEM
OAR
235 Promenade St.
Providence, RI 02908
Tel: 401-222-2808
Fax: 401-222-2017
E-mail: bmorin@dem.state.ri.us
Peter Murchie (Attendee)
U.S. EPA
OAR
811 SW 6th Ave, 3rd Floor
Portland, OR 97204
Tel: 503-326-6554
Fax: 503-326-3399
E-mail: murchie.peter@epa.gov
Ted Palma (Speaker)
OAR(C404-01)
Bldg. C
Research Triangle Park, NC 27711
Tel: 919-541-5470
Fax: 919-541-0840
E-mail: palma.ted@epa.gov
Devon Payne-Sturges, DrPH (Speaker)
OA(1809T)
Ariel Rios
1200 Pennsylvania Ave., NW
Washington, DC 20460
Tel: 202-566-2316
Fax: 202-566-2363
E-mail: payne-sturges.devon@ epa.gov
Henry Topper (Attendee)
OPPTS
701 Spencer Ave
Santa Rosa, CA 95404
Tel: 707-292-6637
Fax: 202-564-8750
E-mail: topper.hemy@epa.gov
Joe Touma (Speaker)
OAR(D243-01)
Research Triangle Park, NC 27711
Tel: 919-541-5381
Fax: 919-541-0044
E-mail: touma.joe@epa.gov
Office of Research and
Development (ORD)
William Boyes, Ph.D. (Attendee)
NHEERL
ORD (MD-74b)
ERC
86 Alexander Dr.
Research Triangle Park, NC 27711
Tel: 919-541-7538
Fax: 919-541-3335
E-mail: boyes.william@epa.gov
Janet Burke, Ph.D. (Attendee)
ORD (E205-02)
109 Alexander Dr.
Research Triangle Park, NC 27711
Tel: 919-541-0820
Fax: 919-541-7953
E-mail: burke.janet@epa.gov
Robert Fegley (Attendee)
ORD(8104R)
Reagan Bldg.
1200 Pennsylvania Ave, NW
Washington, DC 20460
Tel: 202-564-6786
Fax: 202-565-2915
E-mail: fegley.robert@epa.gov
B-7
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Stephen Graham. Ph.D. (Attendee)
ORD (E-205-2)
109 T.W.Alexander Dr.
Research Triangle Park, NC 27709
Tel: 919-541-4344
Fax: 919-541-7953
E-mail: graham.stephen@epa.gov
Bill Hutzell, Ph.D. (Speaker)
ORD (E243-03)
Research Triangle Park, NC 27711
Tel: 919-541-3425
Fax: 919-541-1379
E-mail: hutzell.bill@epa.gov
David Klauder, Ph.D. (Organizer)
ORD (8104)
Reagan Bldg.
1300 Pennsylvania Ave., NW
Washington, DC 20460
Tel: 202-564-6496
Fax: 202-565-2915
E-mail: klauder.david@epa.gov
Matthew Lorber (Speaker)
NCEA
ORD (8623D)
1300 Pennsylvania Ave., NW
Washington, DC 20460
Tel: 202-564-3243
Fax: 202-565-0078
E-mail: lorber.matthew@epa.gov
Deborah Luecken (Speaker)
ORD (E205-02)
109 T.W.Alexander Dr.
Research Triangle Park, NC 27711
Tel: 919-541-0244
Fax: 919-541-7953
E-mail: luecken.deborah@epa.gov
Gary Norris, Ph.D. (Speaker)
U.S. EPA/NERL
ORD (E205-03)
Research Triangle Park, NC 27711
Tel: 919-541-1519
Fax: 919-541-0960
E-mail: norris.gary@epa.gov
Haluk Ozkaynak, Ph.D. (Speaker)
ORD (MC-8601D)
Ariel Rios
1200 Pennsylvania Ave, NW (Rm. 620W5)
Washington, DC 20460
Tel: 202-564-1531
Fax: 202-565-0075
E-mail: ozkaynak.haluk@epa.gov
Paul Shapiro (Organizer)
ORD (8722R)
Reagan Bldg.
1200 Pennsylvania Ave., NW
Washington, DC 20460
Tel: 202-564-6833
Fax: 202-565-2447
E-mail: shapiro.paul@epa.gov
Linda Sheldon, Ph.D. (Speaker)
HEASD/ NERL
ORD(E205-01)
109 T.W.Alexander Dr.
Research Triangle Park, NC 27709
Tel: 919-541-2454
Fax: 919-541-0239
E-mail: sheldon.linda@epa.gov
Tim Watkins (Organizer)
ORD (D305-01)
Research Triangle Park, NC 27711
Tel: 919-541-5114
Fax: 919-541-0715
E-mail: watkins.tim@epa.gov
B-8
-------
U.S. Environmental Protection Agency
Resion/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Invited Guests
John Adgate, Ph.D. (Speaker)
University of Minnesota
(MMC 807)
Mayo 1260
420 Delaware St. SE
Minneapolis, MN 55455
Tel: 612-624-2601
Fax: 612-626-0650
E-mail: jadgate@umn.edu
Judith Blair (Attendee)
Maresh Brains at Work
4313 Apple Way
Boulder, CO 80301
Tel: 303-545-2259
Fax: 303-444-7213
E-mail: judith@brainsatwork.com
Steven Bortnick, Ph.D. (Speaker)
Battelle Memorial Institute
505 King Ave.
Columbus, OH 43201-2693
Tel: 614-424-7487
Fax: 614-424-4250
E-mail: bortnick@battelle.org
David Brown, Sc.D. (Speaker)
NESCAUM/ EHHI
65 Bulkley Ave. (North)
Westport, CT 06880
Tel: 203-259-5698
Fax: 203-256-8799
E-mail: dbrown@nescaum.org
Michael Coggiola, Ph.D. (Attendee)
SRI International
333 Ravenswood Ave.
Menlo Park, CA 94025
Tel: 650-859-3045
Fax: 650-859-6196
E-mail: michael.coggiola@sri.com
Jeff Cook (Speaker)
California Air Resources Board
P.O. Box 2815
Sacramento, CA 95812
Tel: 916-322-3726
Fax: 916-327-8217
E-mail: jcook@arb.ca.gov
Lynn Hildemann, Ph.D. (Speaker)
Stanford University
Terman Engineering Center, Rm. M-5
Stanford, CA 94305-4020
Tel: 650-723-0819
Fax: 415-947-3583
E-mail: hildemann@ce.stanford.edu
Vlad Isakov, Ph.D. (Speaker)
California Air Resources Board
1001 I St.
Sacramento, CA 95812
Tel: 916-445-2151
Fax: 916-327-8524
E-mail: visakov@arb.ca.gov
Nancy Maresh (Attendee)
Maresh Brains at Work
4313 Apple Way
Boulder, CO 80301
Tel: 303-545-2259
Fax: 303-444-7213
E-mail: nancy@brainsatwork.com
B-9
-------
U.S. Environmental Protection Agency
Region/ORD/QAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Linda Murchison, Ph.D. (Speaker)
California Air Resources Board
P.O. Box 2815
Sacramento, CA 95812
Tel: 916-322-5350
Fax: 916-323-1075
E-mail: lmurchis@arb.ca.gov
Mike Nazemi (Speaker)
S. Coast Air Quality Mgmt. District
(PRDAS)
21865 E.Copley Dr.
Diamond Bar, CA 91765
Tel: 909-396-3187
Fax: 909-396-3252
E-mail: mnazemi@aqmd.gov
Todd Sax, M.S. (Attendee)
California Air Resources Board
1001 I St., PO Box 2815
Sacramento, CA 95821
Tel: 916-322-6159
Fax: 916-323-1075
E-mail: tsax@arb.ca.gov
Shannon Stetzer (Speaker)
Battelle Memorial Institute
505 King Ave.
Columbus, OH 43201
Tel: 614-424-6575
Fax: 614-424-4250
E-mail: stetzers@battelle.org
Jed Waldman (Attendee)
C A Dept. of Health Services
2151 Berkeley Way (EHLB)
Berkeley, CA 94704
Tel: 510-540-3427
Fax: 916-440-5855
E-mail: jwaldman@dhs.ca.gov
B-10
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
APPENDIX C: SLIDES FROM PRESENTATIONS
These slides can be found at
http://epa.gov/osp/regions/workshops.htm
1. Air Toxics Exposure Assessment Workshop - Welcome
2. Exposure Assessment and Air Toxics
3. Overview of Air Toxics Exposure Assessment in ORD
4. MA TESII-A Regional Perspective
5. Minneapolis-St. Paul -A Comparison of Community,
Residential, and Personal Exposure
6. Air Toxics Monitoring Pilot Project
1. New Trends in Monitoring Methods
8. Atmospheric Formation and Decay of Air Toxics - Implications
for Exposure Assessments
9. Air Toxics Monitoring Methods and Network Design
10. California Monitoring Program: Statewide Nehvork to the
Neighborhood Scale
11. Design Your O\\m Air Toxics Monitoring Nehvork
12. Air Quality Models
13. Using Emission Inventories for Air Quality Modeling
14. Barrio Logan Modeling A nalysis
15. Urban Scale Modeling - Houston Case Study
16. Applying CMA Q Mode Is 3 for A ir Toxics A ssessments
17. Introduction to Human Exposure
18. Air Toxics Exposure in Indoor Environments
Larry Cupitt
Ken Mitchell
Tim Watkins
Mike Nazemi
John Adgate
Barbara Morin
Don Whitaker
Deborah Luecken
Steve Bortnick and
Shannon Stetzer
Jeff Cook and Linda
Murchison
Motria Poshyvanyk
Joe Touma
Joe Touma
Vlad Isakov
Joe Touma
Bill Hutzell
Linda Sheldon
John Girman
C-l
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
19. Multi-pathway Exposure Assessment Matthew Lorber
20. What Human Exposure Data and Models are Available? Haluk Ozkaynak
21. Nation Air Toxics Assessment (NA TA) - The Initial National Ted Palma
Scale Assessment
22, Children's Exposure to Diesel Emissions David Brown
23. Personal Exposure Meets Risk Assessment: The South Devon Payne-Sturges
Baltimore Community Exposure Study
24. Introduction to Source Apportionment Methods Lynn Hildemann
25. Source Apportionment Tools Gary Norris
26. A Modern Example: Northern Front-Range Air Quality Study Chad Bailey
27. Communicating the Results of Air Toxics Exposure Assessments Alvin Chun
C-2
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
APPENDIX D: FLIP CHART NOTES
Breakout Session I: Design Your Own Air Toxics Monitoring Network (Day 2)
Goals / Objectives:
Design an Air Toxics Monitoring Network
• Groups #1 (Yellow) and #2 (Blue): Design an Air Toxics Monitoring Network in a Large-
Scale Urban Area
• Groups #3 (Red) and #4 (Green): Design an Air Toxics Monitoring Network in a Small-
Scale Local Hotspot
Group # 1 (Yellow)
Attendees:
Baugh, Thomas
Blair, Judith
Brown, Catherine
Cook, Jeff
Cupitt, Larry
Deschambault, Lynda
Frank, Neil
KenKnight, Jeff
Lorber, Matthew
Mitchell, Kenneth
Nazemi, Mike
Nelson, William
O'Neill, Francis
Region 4
Maresh Brains at Work
Region 9
California Air Resources Board
ORD/NERL
Region 9
OAR/OAQPS
Region 10
ORD/NCEA
Region 4
South Coast Air Quality Management District
Agency for Toxic Substances and Disease Registry
Region 6
D-l
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Palma, Ted
Payne-Sturges, Devon
Poshyvanyk, Motria
Robinson, Randy
Shapiro, Paul
Sivak, Michael
Smuts, Marybeth
Touma, Joe
Victery, Winona
OAR/OAQPS
OA/OPEI
Region 5
Region 5
ORD/NCER
Region 2
Region 1
OAR/OAQPS
Region 9
Group # 2 (Blue)
Attendees:
Bortnick, Steve
Braye, Stacy
Chalmers, Ray
Coggiola, Michael
Daye, Richard
Diehl, Kathy
Fegley, Robert
France, Danny
Girman, John
Hutzell, Bill
Kotchenruther, Robert
Lessler, Richard
Luecken, Deborah
Maresh, Nancy
Battelle Memorial Institute
Region 9
Region 3
SRI International
Region 7
Region 9
ORD/OSP
Region 4
OAR/ORIA
ORD/NERL
Region 10
Region 9
ORD/NERL
Maresh Brains at Work
D-2
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Murchison, Linda
Nguyen, Phuong
Overstreet, Cheryl
Sheldon, Linda
Sieffert, Margaret
Tax, Wienke
Topper, Henry
VanArsdale, Alan
Wilson, Patrick
California Air Resources Board
Region 5
Region 6
ORD/NERL
Region 5
Region 9
OPPTS/OPPT
Region 1
Region 9
Group # 3 (Red)
Attendees:
Bailey, Chad
Bellizzi, Carol
Bohnenkamp, Carol
Boyes, William
Broadbent, Jack
Brown, David
Chung, Kuenja
Dugre, Sylvia
Graham, Stephen
Guinnup, David
Hildemann, Lynn
Isakov, Vlad
Levy, Libby
Madrone, Brook
OAR/OTAQ
Region 2
Region 9
ORD/NHEERL
Region 9
NESCAUM/EHHI
Region 6
Region 9
ORD/NERL
OAR/OAQPS
Stanford University
California Air Resources Board
Region 9
Region 10
D-3
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Norris, Gary
Sax, Todd
Stetzer, Shannon
Tool O'Neil, Barbara
Wapensky, Lawrence
Watkins, Tim
Whitaker, Don
Wood, Periann
ORD/NERL
California Air Resources Board
Battelle Memorial Institute
Region 9
Region 8
ORD/NERL
ORD/NERL
Region 9
Group # 4 (Green)
Attendees:
Adgate, John
Antley, Beth
Bandrowski, Mike
Bohning, Scott
Brock, John
Burke, Janet
Chun, Alvin
Driscoll, Barbara
Ellsworth, Todd
Hiatt, Gerald
Kahn, Peter
Klauder, David
Morin, Barbara
Murchie, Peter
Owen, Coe
University of Minnesota
Region 4
Region 9
Region 9
Region 9
ORD/NERL
Region 9
OAR/OAQPS
Region 3
Region 9
Region 1
ORD/OSP
Rhode Island Dept. of Environmental Mngt.
OAR/OAQPS
Region 9
D-4
-------
U.S. Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
Ozkaynak, Haluk ORD/NERL
Rose, Keith Region 10
Waldman, Jed California Department of Health Services
Woodlee, Jeff Region 9
D-5
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
Breakout Session II: Discussion of Modeling Questions (Day 2)
Goals / Objectives:
• Discuss Modeling Questions - Potential Topics:
* How active are the Regions in ambient air modeling?
* What types of assessments are being done?
* What are the priorities? What is important?
* Where do the Regions want to be with ambient air toxics modeling and skills?
* What obstacles exist? What do the Regions need? Guidance? Tools? Etc.
* What is needed to overcome these obstacles?
* Suggested workshop follow-up items.
Report Out from Breakout Groups
Region 2 - 'squeaky wheels'
NATA - how to use: highlights air toxics
• use/focus on non-road issues
• NATA at large scale - useful
• weak agency support - "an embarrassment"
Air Modeling
Region 10
1. Idaho, grid (Boise): assessment: building emissions inventory
2. Region wide
• exposure (Portland & Seattle)
• goal - develop inventory
• also using grid system
• ARB to supply photochemical modeling
• Seasonal: estimate annual average
• to use as background concentration
• especially for reactive VOCs
• 4x4 km: Statewide
• air basins
D-6
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
• local scale - SAC, LA, BA (maybe)
• OTAQ - mobile sources important to NATA
• improving inventories
• where is exposure? -> near/on roadways
• Oregon/Washington DEP $
• linked-based inventories
• travel counts/demand
• travel planning systems
Needs
RIO - FTE $
• Guidance for use of models: which to use? what does it entail?
• use it uniformly
• Two or Three NAAQS standards to work with
• concentration - related
• numbers to work toward - target
• How will we support work in Air Toxics and enhance Air Toxics reporting requirements
• Need local inventories to support local models and local process, to get it and be able to
use it
• Reporting Rule - uniform reporting; "CERR'~ for Air Toxics
• Report locations and release parameters
Problem
• Not doing more for Air Toxics
• Use central technology to get BEST reduction
• treat HAPs similar to criteria pollutants
• look at reactivity and toxicity
• Accelerate mobile source air toxics rule application and new rules
• Are Air Toxics a Priority?
D-7
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Group # I (Yellow)
Attendees:
Bandrowski, Mike
Bellizzi, Carol
Bortnick, Steven
Brown, David
Dugre, Sylvia
Hiatt, Gerald
Isakov, Vlad
KenKnight, Jeff
Maresh, Nancy
Morin, Bu-bara
Murchie, Peter
Morris, Gary
O'Neill, Francis
Ozkaynak, Haluk
Palma, Ted
Shapiro, Paul
Touma, Joe
Tsai, Pam
Whitaker, Don
Woodlee, Jeff
Region 9
Region 2
Battelle Memorial Institute
NESCAUM/EHHI
Region 9
Region 9
California Air Resources Board
Region 10
Maresh Brains at Work
Rhode Island Dept. of Environmental Mngt.
OAR/OAQPS
ORD/NERL
Region 6
ORD/NERL
OAR/OAQPS
ORD/NCER
OAR/OAQPS
Region 9
ORD/NERL
Region 9
D-8
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Group # 2 (Blue)
Attendees:
Blair, Judith
Bohning, Scott
Boyes, William
Braye, Stacy
Burke, Janet
Coggiola, Michael
Diehl, Kathy
France, Danny
Hildemann, Lynn
Kahn, Peter
Levy, Libby
Luecken, Deborah
Madrone, Brook
Mitchell, Kenneth
Owen, Coe
Poshyvanyk, Motria
Robinson, Randy
Rose, Keith
Sheldon, Linda
Smuts, MaryBeth
Topper, Henry
Wapensky, Lawrence
Watkins, Tim
Wood, Periann
Maresh Brains at Work
Region 9
ORD/NHEERL
Region 9
ORD/NERL
SRI International
Region 9
Region 4
Stanford University
Region 1
Region 9
ORD/NERL
Region 10
Region 4
Region 9
Region 5
Region 5
Region 10
ORD/NERL
Region 1
OPPTS/OPPT
Region 8
ORD/NERL
Region 9
D-9
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Group # 3 (Red)
Attendees:
Bailey, Chad
Bohnenkamp, Carol
Brock, John
Brown, Catherine
Chalmers, Ray
Chung, Kuenja
Cook, Jeff
Cupitt, Larry
Ellsworth, Todd
Fegley, Robert
Graham, Stephen
Kotchenruther, Robert
Lessler, Richard
Murchison, Linda
Nazemi, Mike
Nguyen, Phuong
Payne-Sturges, Devon
Stetzer, Shannon
Tax, Wienke
Toole O'Neil, Barbara
OAR/OTAQ
Region 9
Region 9
Region 9
Region 3
Region 6
California Air Resource Board
ORD/NERL
Region 3
ORD/OSP
ORD/NERL
Region 10
Region 9
California Air Resources Board
South Coast Air Quality Mgt. District
Region 5
OA/OPEI
Battelle Memorial Institute
Region 9
Region 9
D-10
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Group # 4 (Green)
Attendees:
Adgate, John
Antley, Beth
Baugh, Thomas
Broadbent, Jack
Chun, Alvin
Daye, Richard
Deschambault, Lynda
Driscoll, Barbara
Frank., Neil
Girman, John
Guinnup, David
Hutzell, Bill
Klauder, David
Lorber, Matthew
Nelson, William
Overstreet, Cheryl
Sax, Todd
Sieffert, Margaret
Sivak, Michael
VanArsdale, Alan
Victery, Winona
Waldman, Jed
Wilson, Patrick
University of Minnesota
Region 4
Region 4
Region 9
Region 9
Region 7
Region 9
OAR/OAQPS
OAR/OAQPS
OAR/ORIA
OAR/OAQPS
ORD/NERL
ORD/OSP
ORD/NCEA
Agency for Toxic Substances and Disease Registry
Region 6
California Air Resources Board
Region 5
Region 2
Region 1
Region 9
California Dept. of Health Services
Region 9
D-ll
-------
U.S. Environmental Protection Agency
Region/ORD/QAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
THIS PAGE INTENTIONALLY LEFT BLANK
D-12
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
APPENDIX E: PARTICIPANT EVALUATION SUMMARY
Meeting participants agreed that the information they gained from the workshop would help them
perform their job better, especially topics on exposure assessment, monitoring, and modeling.
Some attendees felt that topics concerning risk modeling and source apportionment were not very
useful to them. Though it was agreed that the presentations were effective in communicating
regional issues and ORD science to address those issues, the majority of participants felt that the
workshop should have contained more regional air toxics exposure assessment issues, and some
would have liked to see more emphasis on risk communication.
Participants agreed that the presentations were sufficiently tailored to suit their information needs.
Meeting attendees were split regarding whether breakout sessions were effective in providing an
opportunity to further explore the topics using the information learned. Some felt that breakout
groups should have been smaller to allow for more interaction. In addition, about half of the
meeting attendees felt that there was not enough time allotted to breakout sessions, though the
other half felt that ample time was allotted. Many participants agreed that an interactive meeting
format was favorable and should be more frequently utilized in future workshops. Overall,
participants agreed that the meeting was useful and successful, and consistently offered positive
feedback and praise to the organizers for a great workshop.
E-l
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
THIS PAGE INTENTIONALLY LEFT BLANK
E-2
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
APPENDIX F: AIR TOXICS CASE STUDIES
Mini-Case Studies of
Air Toxics Monitoring and Modeling
This is a compendium of mini-case studies of air toxics monitoring and modeling projects at
different scales and in different locations. It is meant to provide some sense of the range of
monitoring and modeling approaches that has been used and contacts where you can get more
information on specific projects of interest. We hope that people will add suggestions for projects
that could be written up and added to this group and even provide write-ups in the format that is
being used. The more case studies there are, the richer the resource will be.
Questions, suggestions, and contributions can be addressed to the following:
Ken Mitchell
USEP A/Region 4/APTMD
Tel: 404-562-9046
Email: mitchell.ken@epa.gov
Ted Palma
USEPA/OAQPS/ESD/REAG (C404-01)
Tel: 919-541-5470
Fax:919-541-0840
Email: palma.ted@epa.gov
Paul Shapiro
USEPA/ORD/NCER (8722R)
Tel: 202-564-6833
Fax: 202-564-2447
Email: shapiro.paul@epa.gov
F-l
-------
U.S. Environmental Protection Agency
Rcgion/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.20'. i 2
CONTENTS
-1. Atsugi, Japan
2. Columbus, OH
3. Kenova, KY, WV, OH
4. Los Angeles International Airport
5. Milwaukee, WI, and Sacramento, CA
6. Minneapolis-St. Paul, MN
7. National Air Toxics Assessment (NATA)
8. Portland, OR
9. Port Neches, TX, Calcasieu, LA, and Little Rock, AK
10. San Francisco Bay Area
11. South Coast Air Quality Management District, CA
12. San Diego, CA
F-2
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
1. Atsugi, Japan
The Shinkampo Incinerator in Atsugi Japan Case Study
I. Project Name and Location, Project Sponsor, Brief Overview
The United States Naval Air Facility at Atsugi, Japan (NAF Atsugi) is located in the Kanto Plain
area on the island of Honshu, Japan. Directly to the south of the facility, in the Tade River
Valley, was the Shinkampo Incinerator Complex (SIC). The Incinerator is no longer in operation,
having closed in May of 2001. While operating, three incinerators were licensed by Japan to burn
general industrial waste and infectious industrial waste (medical), and were incinerating up to 90
tons a day. The pollution control devices included precipitators and scrubbers. The 4-5 acre
facility was located in a small river valley, and the NAF Atsugi is positioned on a plateau at the
end of the valley. While the incinerator stacks were about 25 m high from the valley base, they
were only about 15 m higher than the ground level of the NAF Atsugi. Further, these stacks were
only 250 m away from the nearest high-rise housing unit and about 1000 m from a school and day
care center. The predominant wind direction is from south to north during the spring and summer;
conversely from north to south during the fall and winter. When blowing from south to north, the
plume moves directly onto the base where exposures could occur. The NAF Atsugi was not
permitted to test the stacks, and was not provided with stack test information by the owners of the
facility. Subsequently, their evaluation of environmental impacts focused on environmental
monitoring on NAF Atsugi, including extensive air and soil testing programs. Numerous organic
and inorganic contaminants were measured in these programs. Also, the Navy conducted
additional analyses to understand the origin of what they were measuring. Specifically, wind
speed and direction measurements were taken along with all air monitoring data, as well as
observational data regarding the incinerator operation (operating/shut down, color of smoke in
plume, etc.). Interestingly, they found that only a handful of the measured contaminants could be
unambiguously attributed to the incinerator, including dioxins, PM10, hydrochloric acid, lead,
cadmium, and arsenic. Some important risk driving contaminants, such as benzene, showed
much less correlation with wind direction suggesting other on-base sources, such as the regular
deployment of the airplanes as part of the airbase operations. To further enhance their
understanding of exposure and incinerator emissions, they also conducted air dispersion modeling
using ISC3. Using the network of air measurements and the wind speed data, they calibrated the
ISC3 on several contaminants including dioxin to be able to estimate on stack emission rates, and
to develop a refined concentration term for human exposure assessment. Interestingly, dioxin
emissions were estimated to total 18 g TEQ/yr, which is not particularly high for incinerators,
although air concentrations are some of the highest regularly measured in the world. This was
because of the low height of the stack and the proximity to exposure locations.
F-3
-------
U.S. Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
The National Center for Environmental Assessment (NCEA)/ORD was enlisted to support the
Navy in their efforts at this site in 1997, mostly in their interpretation of dioxin air data, design of
a dioxin soil testing program, and assessing risk due to dioxin exposure. The Navy is expected to
finalize its comprehensive risk assessment on this site during 2002, and to then make publicly
available its extensive wealth of monitoring data and modeling studies.
II. Contact name and Information
Matthew Lorber
National Center for Environmental Assessment (8623D)
US EPA
1200 Pennsylvania Ave, NW
Washington, DC 20460
ph: 202564-3243
email: lorber.matthew@epa.gov
III. Uses of Modeling and Monitoring
This is a unique example where modeling was only used to supplement an extensive monitoring
program. In many circumstances such as this one, stack test information is combined with
modeling to understand exposure. Because of the cost and other factors such as uncertainty
(enough samples? downwind conditions?, and so on), monitoring is usually not relied upon for
exposure assessing and most often just used to verify, or "truth test" results of modeling. In this
case, however, literally millions of dollars were spent determining exposure point concentrations
for exposure and risk assessing.
F-4
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
IV. Key Results
Two abstracts containing dioxin air and soil data have been submitted to the Dioxin 2002
conference, to be held in August of 2002. NAF Atsugi contractor reports have additional
information.
Figures showing percent of time wind was blowing from the incinerator to the air monitor, clearly
showing the impact of the incinerator —>
Table 1. Summary of TEQ results for the categories of soil samples.
Description
Exposure Study Areas
Reference Study Areas
Trend - downwind and impacted
Trend - all other samples
n
28
12
11
22
TEQ. pg/g
15
27
266
14
TEQ range
<1-90
13-62
66 - 642
8-83
Descriptions: "Exposure Study Areas" - locations
such as schools, apartment buildings where exposure
could occur; "Reference Study Areas" - remote sites
on NAF Atsugi where only deposition would likely
have caused soil elevations, if any; "Trend -
Downwind & Impacted" - all trend samples were
randomly spaced in a study design to evaluate trends.
The downwind & impacted samples were a cluster of
11 samples all downwind and near, within 300
meters, to the incinerator. The other 22 trend
samples were spaced up to a kilometer away from the
incinerator.
-&=.
B. O.
O o-
20 40 60 80
Percent Downwind
100
F-5
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27. 2002
1500 1700
Concentration in
Figure showing the results of the
air monitoring exercise, for the 6
contaminants unambiguously
originating from the incinerator.
The concentrations are modeled
concentrations per g/sec unit
emissions. These measurements
get multiplied by the "multiplier"
identified to get predicted air
measurements.
VI. Lessons Learned
The fact that stack monitoring was not allowed made this a unique circumstance where ambient
air monitoring, wind speed/direction monitoring, and modeling were used to understand exposure
to this source. The air monitoring was extremely expensive, but may have been uninformative
had not wind speed and direction measurements accompanied all air concentration measurements.
Certainly in every similar circumstance, all efforts must be made to conduct stack sampling to
understand what is being emitted from the source.
F-6
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
2. Columbus, OH
Columbus Municipal Solid Waste Incinerator Case Study
I. Project Name and Location, Project Sponsor, Brief Overview
In 1992, a stack test revealed that the Columbus Municipal Solid Waste-to-Energy Incinerator
(WTE) was emitting at a rate equal to about 1000 grams of dioxin toxic equivalents (TEQs) per
year. To put this in perspective, EPA's Dioxin Sources Inventory estimated that all known and
quantified sources were emitting about 12,000 g TEQ/yr in 1987 and 3,000 g TEQ/yr in 1995. As
a result of this alarming stack test, Region 5 began developing the scientific basis for an
Emergency Order to require the owners of the Columbus WTE to install MACT controls well
ahead of the federally-mandated schedule. The Region enlisted the National Center for
Environmental Assessment (NCEA/ORD) to assist in the conduct of a risk assessment on the
impacts of these dioxin emissions. This assessment entailed use of air modeling (using stack
emission data) to simulate the arrival of dioxins at nearby farms, the routing of these dioxins
through the food chain, and finally the estimation of exposure and risk to a farm family
consuming home-grown foods. The Emergency Order was instated in 1994, and it required the
owners of the Columbus WTE to install MACT by 1997. During 1994, owners of the Columbus
WTE made some process changes (temperature of combustion, installation of quenches) in an
attempt to reduce dioxin emissions. To evaluate the effectiveness of these procedures, additional
stack tests and ambient air monitoring were conducted by the State of Ohio EPA in the spring and
summer of 1994. In December of 1994, the Columbus WTE shut down, citing lack of funding to
maintain operations, much less comply with the Emergency Order to install MACT. In 1995, the
Region enlisted the support of NCEA in the design of a soil monitoring study to understand the
potential long-term impacts of dioxin emissions. NCEA has continued to use this wealth of data
to study the relationships between dioxin emissions and environmental impacts. One evaluation
focused on the relationships between all the media which had been sampled, including stack
emissions, incinerator ash. air, and soil. A second evaluation used the stack emission, air, and
soil data in model validation/calibration exercise. This exercise involved use of the ISC3 model,
to see how well it would predict measured dioxin air concentrations, and then it used predicted
deposition rates in a soil concentration model to see how well predicted soil dioxin concentrations
would match measured concentrations.
F-7
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
II. Contact name and Information
Matthew Lorber
National Center for Environmental Assessment (8623D)
US EPA
1200 Pennsylvania Ave, NW
Washington, DC 20460
ph: 202564-3243
email: lorber.matthew@,epa.gov
III. Uses of Modeling and Monitoring
This case study represents a comprehensive example of the interplay between modeling and
monitoring, and even more, how these exercises and evaluations play into the regulatory decision-
making process at EPA. Such efforts are possible, and maybe uniquely possible, with the suite of
compounds generally called "dioxin-like compounds" or just "dioxins", for these reasons: 1)
dioxins are one of the most toxic chemicals ever produced by man, and as such, one of the best
studied class of compounds for human health effects, environmental fate, human exposure, source
characterization, and other disciplines, 2) dioxins are persistent in the environment, including in
air, soils, and sediments, and they bioaccumulate in terrestrial and aquatic animals. This
increases the likelihood of success for just about any monitoring program. 3) these same
tendencies for persistence and bioaccumulation also make dioxin the ideal class of compounds to
model. The models are being used to predict long-term trends, which tends to be an easier chore
than asking them to predict short term events.
F-8
-------
U.S. Environmental Protection Agency
Region/ORD/QAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
IV. Key Results
The results from the Columbus WTE studies by NCEA are covered in three publications. These
publications and a sample of some of the results are:
1. Lorber, M.; Cleverly, D.; Schaum, J. 1996. A screening level risk assessment of the indirect
impacts from the Columbus waste to energy facility in Columbus, Ohio. Proceedings of an
International Specialty Conference, sponsored by the Air and Waste Management Association
and the United States Environmental Protection Agency, held April 18-21, 1996 in Washington,
D.C. published in, Solid Waste Management: Thermal Treatment & Waste-to-Energy
Technologies, VIP - 53. pp. 262-278. Air & Waste Management Association, One Gateway
Center, Third Floor, Pittsburgh, PA 15222.
Exposure and risk results for the Columbus WTE risk assessment.
Exposure Pathway
Soil Dermal Contact
Vegetable Ingestion
Inhalation
Beef Ingestion
Milk Ingestion
Lifetime Average Daily Dose
ng TEQ/kg-day
6*10~8
,i*io-5
6*10-6
i*io-3
5*10-4
Excess Cancer Risk
9*1 0-9
2*4 0-6
9*1 0-7
2*10"*
8*10-5
2. Lorber, M, P. Pinsky, P. Gehring, C. Braverman, D, Winters, W. Sovocool. 1998.
Relationships between dioxins in soil, air, ash. and emissions from a municipal solid waste
incinerator emitting large amounts of dioxins. Chemosphere, Volume 37:2173-2197.
Result: Isolines of equal soil TEQ concentration around the Columbus WTE location.
F-9
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
3. Lorber, M., A. Eschenroeder, R. Robinson. 2000. Testing EPA's ISCST-Version 3 Model on
Dioxins: A comparison of predicted and observed air and soil concentrations. Atmospheric
Environment 34:3995-4010.
Results: Isolines of predicted air concentrations compared against measured air concentrations of
TEQ for one sampling date in March, 1994.
VI. Lessons Learned
If encountering a similar situation - an incinerator emitting large amounts of dioxin-like
compounds - here are some of the things we can do differently and better to understand and
remedy the situation:
1) More air sampling within 1 km while incinerator is operating to better understand the
plume.
2) Search for more exposure matrices to sample, such as impacted farm animals (cows, free
range chickens).
3) If the incinerator has been operating long enough, consider identifying and sampling
blood of a potentially exposed population along with a control population.
4) Further study the environmental transformations of specific dioxin-like compounds from
source to environmental matrix (soil particularly).
F-10
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
3. Kenova, KY, WV, OH
Tri-State Geographic Initiative Case Study
L Project Name and Location, Project Sponsor, Brief Overview
Tri-State Geographic Initiative
• Located along the Big Sandy and Ohio Rivers at the convergence of Kentucky,
West Virginia, and Ohio
Sponsored by:
Ohio Environmental Protection Agency
West Virginia Division of Environmental Protection
Kentucky Department for Environmental Protection
U.S. Environmental Protection Agency Regions 3, 4, and 5
II. Contact Name and Information
Ken Mitchell
Chief, Air Toxics Assessment and Implementation Section
U.S. Environmental Protection Agency, Region 4
404-562-9065
mitchell.kenffiepa.gov
III. Key Questions Addressed and Purpose of the Project
In 1991, Kentucky, West Virginia. Ohio, and the United States Environmental Protection Agency
partnered to study environmental problems in the highly industrialized area where the three states
meet. This project came to be known as the Tri-State Geographic Initiative or TGI. Public
concern, and the availability of time, resources, and data led the partners to focus on the impact of
air, drinking water, and fish consumption on human health in the area.
This case study will focus on air quality which emerged as a priority in the study. The TGI
technical steering committee established the Air Toxics Project through which air monitoring,
F-ll
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
dispersion modeling, and assessment of risks associated with these air pollutants was carried out.
The results of these analyses will help set priorities for improving environmental quality and
public health in the area.
Resource constraints made it impossible to study the entire 2300 square mile area concurrently,
so the major sources of air pollution in the area were grouped into six clusters which could be
studied individually. One of these clusters consisted of several industrial sources in the Kenova,
West Virginia area. The Kenova cluster study area straddles the banks of the Big Sandy River,
and includes not only industrial complexes (including a refinery), but also residences, schools,
day care, and other commercial concerns.
IV. Uses of Modeling and Monitoring
Air monitoring and modeling were conducted for the July 1996 to July 1997 period.
• Monitoring:
7 fixed sites sampled every 12-14 days for volatile organic chemicals,
semi-volatile organic analytes, metals, and acid/base gases to assess
chronic human health concerns. A triggered sampler was located at one of
the fixed locations to collect VOC samples when concentrations were
elevated.
• • A mobile laboratory collected continuous VOC samples over 15 weeks at 4
locations to assess acute risks
Modeling
Area specific meteorological data were collected
• • Air quality effects were modeled based on reported releases from four
major chemical facilities in the Kenova area using four scenarios:
• • Daily meteorological data and highest reported emissions to
represent a worst case scenario
• • Daily meteorological data and daily emissions to represent a typical
scenario
• • Daily meteorological data and daily emissions during periods when
episodes were known to have occurred
• • Emissions from mobile sources and small businesses were not modeled
F-12
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
V. Key Results
• Long-term air monitoring
• • Hazard indices exceeded 1 at all stationary sampling sites indicating an
area-wide concern for non-cancer effects in humans with chromium being
the primary contributor to the risk
• • All sampling sites exceeded 1E-6 cancer risk, and all but one site were
lower than 1E-4 cancer risk, with benzene and chromium driving the risk
• Modeled air risks
Of the 4 facilities analyzed, only one had modeled emissions that resulted
in risks above the levels set in the risk management plan( hazard index > 1,
or cancer risk > E-6)
VI. Lessons Learned
At air monitoring locations, risks predicted based on air monitoring results tended
to be higher than risks estimated based on modeling results.
Modeling indicates that highest modeled industrial source risks are near facility
boundaries, and that risks are typically lower in residential areas further removed
from the facilities.
Modeling results are only as good as the data provided by the industries, some of
which are supportive, and some of which may provide only typical or average
data.
Modeling can provide a better indication of daily values, since monitoring does not
occur every day and may miss episodic emissions.
Modeling allows the pinpointing of sources of particular chemicals; monitoring
reports values that are a composite of emissions from many sources.
Because relatively few such studies have been conducted, interpreting the results is
difficult - Were these results typical or atypical nationally? Would a plant-
specific, regional, or national policy be most appropriate to address the risks?
F-13
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
4. Los Angeles International Airport
I. Project Name and Location, Project Sponsor, Brief Overview
Los Angeles World Airports: Air Quality and Source Apportionment Study of the Area
Surrounding Los Angeles International Airport:
Vol I. Technical Work plan, Vol. II. Quality Assurance Project Plan were prepared for LAWA
using contractors from COM and academic experts (John Watson, Desert Research Institute, and
Ron Henry, University of Southern California), dated November 10, 2000. Both are draft
pending external peer review funded by EPA to be completed in a peer review workshop
expected to occur in July 2002. Winona Victery is the contact in Region 9; EPA is funding peer
review; further implementation will require Federal funding (expected to be $470,000 for one-
month pilot) and year-long study $2-3,000,000. LAWA unable to fund after 9/11.
II. Contact Names and Information
Sabrina Johnson, EPA, OAR, HQ, 202-564-1173, iohnson.sabrina@epa.gov
Roger Johnson, LAWA, 310-646-9640, riohnsonfSiawa.com
Joellen Lewtas, EPA, ORD, NERL, 206-553-1605, lewtas.ioellen@epa.gov
Winona Victery, EPA, Region 9,415-972-3736, victerv.winona'g'epa.gov
III. Key Questions Addressed and Purpose of the Project
Los Angeles World Airports (LAWA) proposes to conduct an air quality study to develop
detailed information about the role of the Los Angeles International Airport (LAX) in emitting air
pollutants and the impact the emissions have in the neighborhoods around LAX. This is the first
attempt at performing such a study near an airport.
F-14
-------
U.S. Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
IV. Uses of modeling and monitoring
Emissions data collected by both on-site and off-site monitoring and comprehensive
characterization of emission sources and mass emissions with those sources for facilities. Source
apportionment techniques will include chemical composition receptor modeling, spatial gradient
analysis, time series analysis, emissions inventory development, and air dispersion modeling.
Monitoring will be performed to identify source-dominated baseline site and neighborhood sites.
Analytes include gases, particle mass and chemistry, substrate sampling, and meteorology.
Dispersion modeling will attempt to determine how much of each pollutant of concern in nearby
neighborhoods is contributed by LAX relative to non-LAX emitters, what is the uncertainty of
these source contributions? What are the dominant source categories that contribute to each
pollutant of concern from LAX and non-LAX emitters? How do source contributions from
different source types differ by time of day, day of week, and time of year?
V. Key Results
Study is on hold pending external peer review and availability of funds.
VI. Lessons Learned
LAWA contractor hired monitoring and modeling experts at the request of the LAX Technical
Work Group to draft the technical work plan. EPA and State of California Air Resources Board,
and South Coast Air Quality Management District, and FAA served on this workgroup. Work by
the first group of LAWA staff was poor and the lead Camp Dresser McGee consultant was hired
by the airport to replace the LAWA staff. LAWA agreed to accept EPA and California Air
Resources Board recommendations for external authors. The Technical Work Group who had
recommended the experts then had to wait for the re-draft of the original document. EPA, CARB
and South Coast AQMD did not feel comfortable asking for peer-review of the LAWA-funded
document. When LAWA funding was no longer available, the airport had already committed
publicly about performing this study with EPA oversight, we agreed that external peer review was
necessary and that EPA could fund that. LAWA has been hopeful that we can still perform the
Air Quality and Source Apportionment Study.
F-15
-------
U.S nvironmental Protection Agency
Rei •..:• ORD/QAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
5. Milwaukee, WI, and Sacramento, CA
I. Project Name and Location, Project Sponsor, Brief Overview
Testing of a Model to Predict Human Exposure to Aldehydes Arising from Mobile and Point
Sources. EPA STAR Grant No. R 826787-01-0
J. H. Raymer*, G. Akland, T. Johnson, D. J. Smith, D. A. Whitaker, and T. Long
II. Contact Name and Information
James H. Raymer
RTI International
3040 Comwallis Road, RTF, NC 27709
Telephone: (919) 541-5924
e-mail: jraymer(g)jti.org
III. Key Questions Addressed and Purpose of the Project.
The main hypothesis to be tested is that a mathematical model can be used to predict personal
exposure distribution to aldehydes. Additional hypotheses to be tested are that (a) personal
exposure levels of aldehydes exceed outdoor concentrations; (b) indoor aldehyde concentrations
exceed outdoor concentrations; and (c) the composition of oxygenated fuel results in significant
differences in population exposures to aldehydes.
IV. Uses of Modeling and Monitoring
Milwaukee, Wisconsin (because of the use of ethanol in the gas ne) and Sacramento, California
(because of the use of methyl /-butyl ether or MTBE) were chostu for the field studies. These
two field studies, conducted for approximately 40 days each during the summers of 1999 and
2000, were both directed towards acquiring representative personal monitoring data that will be
used to estimate the exposures of urban and suburban residents to selected aldehydes, volatile
organic compounds (VOCs), and carbon monoxide (CO). Aldehydes (formaldehyde,
acetaldehyde, acrolein^ propionaldehyde, butyraldehyde, crotonaldehyde, glyoxal, methylglyoxal)
were sampled using DNPH silica cartridges, VOCs (ethanol, MTBE, benzene, toluene, xylenes)
were sampled using sorbent tubes, and CO was measured using real-time monitors. Air exchange
in each home was evaluated.
Through the collection and analysis of VOCs and CO in addition to the aldehydes, the design
permits a source apportionment of the aldehyde contribution to exposure that originates from
direct emissions from mobile sources, those which are photochemically produced from mobile
source emissions, combustion processes, and contributions of other sources and environments
F-16
-------
U.S. Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
which also contribute to these exposures. Data from the Global Positioning System (GPS) were
collected and evaluated as a means of participant tracking. Supplemental data were obtained
from indoor and outdoor pollutant monitors at the residences of the volunteers, from ambient
pollutant and meteorological monitors at fixed-site locations in each city, from real-time diaries
completed by the technicians and volunteers, and from questionnaires completed by the
volunteers. The field design was model-based, that is, the monitoring field data were gathered for
input into the model or to test and validate modeled predictions. Measured and predicted
exposures will be compared for determining uncertainties of the modeled exposures.
Each field study (city) had two components. In the first component (Phase A), integrated
personal exposures for 38 and 33 volunteers, for Sacramento and Milwaukee, respectively, were
measured once during a summer season. This study is similar in design to other personal
exposure studies except that the volunteers will be randomly chosen according to selection
criteria related to location of residence. The study population was selected according to spatial
gradients away from the downtown area. Twenty four-hour measurements of the corresponding
indoor and outdoor concentrations at the residence of each subject were also made. In the second
component (Phase B, "scripted") the exposures of a technician to the same set of pollutants was
measured as the technician followed a set of prepared instructions, called scripts, to follow
throughout the day.
These scripts outlined the activities and microenvironments to which the technician will be
exposed on a given day, e.g., commuting downtown, driving in street canyons, walking inside,
walking outdoors around a commercial area away from a roadway. The script provided
instructions to the technician specifying (1) the duration of an air sample to be taken during each
sampling period, (2) the general and microenvironmental location during the sampling period,
and (3) the general activities undertaken during the period. Personal exposure monitors
(aldehyde-DNPH, VOCs, and CO) were used to measure 1-hr and 12-hr exposures to each
compound as the technician simulated the activity patterns of typical city residents. Using the
data generated from the scripted study and information about the activities of the study
participants in the personal monitoring study, the 24-hour exposures of the participants will be
predicted using an existing EPA model, pHAP, and compared to those measured.
V. Key Results
To date, all of the samples have been collected, analyzed, and assembled into a database that also
contains the meta data. A series of stepwise linear regression (SLR) analyses were performed on
these data to identify the factors that best predicted exposure to each pollutant. The aldehydes
measured in Sacramento by microenvironment are shown in Figure 1. The highest concentrations
of aldehydes were found indoors at a restaurant, indoors at a residence, transportation in a car,
indoors at a grocery store, and outdoors within 10 yards of the roadway. The percent of the
variability for formaldehyde and acetaldehyde explained by the microenvironment is shown in
Table 1. Much higher total correlations were observed for formaldehyde.
F-17
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
For comparison, the aldehyde concentrations measured in various microenvironments in
Milwaukee are shown in Figure 2. The highest aldehyde median concentrations were found
indoors at a store, indoors at a restaurant, and indoors at the technician's residence. Analogous
SLR analyses have not yet been performed for Milwaukee. This information will be of use to the
U.S. EPA hi determining the significance of microenvironment on human exposure.
VI. Lessons Learned
Indoor exposures to formaldehyde and acetaldehyde are higher than outdoor exposures,
especially in areas with food and food preparation.
Further lessons await completion of the data analysis.
F-18
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27. 2002
Figure 1. Sacramento Study Aldehyde Results, (n = number of total observations; IND = indoors, OUT
= outdoors; c.f. Table 1)
Sacramento Study
Aldehyde Results
g
-&
£
I »
5
CJ
I
2 100 -
15
seraax
*!*** Median
Maximum
37
J
I
' ri-f
5 20 1
Microenvi raiment
Table 2. Results of Stepwise Linear Regression Analyses Performed on Formaldehyde
and Acetaldehyde Concentrations - Sacramento Scripted Activity Study
Compound
Formaldehyde
(n = 191)
Acetaldehyde
(n = 192)
Binary Predictor Variable Regression
Coefficient
Constant 7.5
Indoors - store 12.4
Indoors - residence 13.0
Indoors - auto parts store 29.8
Indoors - restaurant 10.8
Using solvents 34.5
Smoke from forest fires 29.5
Indoors - church 17.5
Indoors - service station 11.5
Indoors - furniture store 16.1
Indoors - building supply store 10.0
Constant 13.9
Air conditioning on 26.0
Drinking alcoholic beverage 25.1
Indoors - grocery store 92.1
Indoors - residence 22.6
Eating 20.3
Cumulative
R2 Value
0.000
0.172
0.354
0.433
0.500
0.558
0.601
0.616
0.629
0.638
0.650
0.000
0.238
0.292
0.331
0.360
0.378
F-19
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
6. Minneapolis-St. Paul, MN
I. Project Name and Location, Project Sponsor, Brief Overview
A Comparison of Community, Residential, and Personal Exposure in Minneapolis-St. Paul,
Minnesota
EPA STAR Grants R825241 and R827928
II. Contact name and Information
Gregory Pratt John L. Adgate
Minnesota Pollution Control Agency Univ. of MN School of Public Health
520 Lafayette Rd Div. of Environ, and Occup. Health
St. Paul, Minnesota 55155 MMC 807
651.296.7664 420 Delaware St SE
gregory.pratt@pca.state.mn.us Minneapolis, MN 55455
612.624.2601
jadgate@umn.edu
III. Key Questions Addressed and Purpose of the Project
The purpose of this research was to estimate outdoor air concentrations of VOCs in the
Minneapolis-St. Paul metropolitan area and compare them with outdoor, indoor, and personal
monitoring. Based on preliminary modeling three neighborhoods were chosen for outdoor
ambient, indoor, and personal monitoring. The modeled outdoor concentrations were then
compared with the monitored outdoor concentrations to try to understand the reasons for
agreement or lack of agreement. The modeling was done using the most commonly used EPA
regulatory air dispersion model, ISCST3 (Industrial Source Complex Short Term version 3) and
VOCs emission estimates from point, area, and mobile sources (mobile and area sources using
1997 and point using 1999 emission inventories). The point sources were modeled as point
sources, and the area and mobile sources were resolved to the census tract level and modeled as
area sources.
IV. Uses of Modeling and Monitoring
Only outdoor air VOC concentrations were modeled. Modeled concentrations were used to assist
in selecting neighborhoods for outdoor, indoor, and personal monitoring. Three communities
were selected: two (East St. Paul and Phillips) with relatively high and one (Battle Creek) with
relatively low estimated ambient concentrations. Model estimates can be compared with outdoor
measurements (matched in space and time) to provide information on model accuracy.
The monitoring data consisted of personal, indoor, and outdoor measurements of VOCs. The
personal, indoor, and some outdoor measurements were made with personal organic vapor
monitors (OVMs). Other outdoor measurements were made with sampling canisters analyzed by
F-20
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
GC-MS (the Federal Reference Method). There were 75 non-smoking study participants from
three neighborhoods in the metropolitan area, each of whom had indoor (stationary residential)
and personal measurements up to six times over three seasons. The monitoring data can be used
to understand the relationships between indoor, outdoor, and personal VOC measurements. The
data can also be used to look at the variability in indoor and personal measurements for a given
home or person over time (i.e., longitudinal variability).
V. Key results
1. The OVM monitoring badges were an effective and relatively inexpensive method for
obtaining 48 hour average VOC measurements for -12 compounds;
2. OVMs compared well with canisters for most VOCs in this study;
3. Generally for measured VOCs: Personal > Indoor > Outdoor;
4. Air dispersion modeling with ISCST reasonably predicted outdoor VOC concentrations if
the emission inventory was accurate;
5. The model tended to predict average concentrations and was less accurate at high or low
concentrations. This is expected since the emissions inventory cannot capture the details
of all of the temporal and spatial variability in emissions;
6. The ISCST model performed best in Battle Creek (relatively simple emissions scenario)
and worst in Phillips (more nearby and relatively complex emissions);
7. The model failed to predict the higher VOC concentrations found in indoor and personal
air.
VI. Lessons Learned
Modeling can be effective in predicting outdoor concentrations of VOCs. The most critical
model input is the emissions inventory—without an accurate inventory the model predictions are
not reliable.
An indoor air model and a personal exposure model would be necessary to estimate indoor air
concentrations and personal exposures more accurately. Indoor air modeling requires an accurate
inventory of indoor air pollution sources, and personal exposure modeling requires an accurate
assessment of each microenvironment encountered by the study subject as well as a record of the
amount of time spent in that microenvironment. We collected time-activity information on these
subjects, but have limited data on sources in these homes and no air exchange data for these
households. We also do not have data on air concentrations in microenvironments outside the
home.
The outdoor air modeling presented in this case study is regulatory state of the art. It would be
possible to use a more refined model, but the accuracy of the emissions data does not warrant that
level of sophistication at this time.
F-21
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
7. National Air Toxics Assessment (NATA)
I. Project Name and Location, Project Sponsor, Brief Overview
National Air Toxic Assessment - Initial National Scale Assessment
II. Contact name and Information
Dave Guinnup, USEPA OAQPS, ESD/REAG (C404-01)
Research Triangle Park, NC 27711; 919-541-5368; guinnup.davefg).epa.gov
III. Key Questions Addressed and Purpose of the Project
The National-Scale Air Toxics Assessment, which is based on 1996 emissions data, produced
results that are useful in understanding the quality of air and its possible effect on human health
nationwide. The assessment looked at 33 air pollutants (a subset of 32 air toxics from the Clean
Air Act's list of 188 air toxics plus diesel particulate matter). The primary goal of the
national-scale assessment was to identify those air toxics which are of greatest potential concern,
in terms of contribution to population risk. The results will be used to set priorities for the
collection of additional air toxics data (e.g., emissions data and ambient monitoring data).
IV. Uses of Modeling and Monitoring
The initial national-scale assessment is comprised of four major technical components: 1)
compiling a national emissions inventory of air toxic and diesel PM for the year 1996 from
outdoor sources; 2) estimating 1996 air toxics and diesel PM ambient concentrations; 3)
estimating 1996 population exposures; 4) characterizing potential public health risks.
The 1996 National Toxics Inventory (NTI) is the underlying basis for the 1996 emissions used in
the national-scale assessment. The NTI contains air toxics emission estimates for four
overarching source types: major, area and other, onroad mobile, and nonroad mobile.
To develop nationwide estimates of annual average ambient concentrations of air toxics, EPA
used the Assessment System for Population Exposure Nationwide (ASPEN) model (developed
and used in FPA's Cumulative Exposure Project). The modeling domain for the national
modeling eftort is the contiguous United States, Puerto Rico and the Virgin Islands. The ASPEN
model, which is based on the ISCLT2 dispersion model, estimates annual average ambient
concentration of each air toxic pollutant at the centroid of each census tract within the geographic
domain. In an effort to apply a "reality check" on the ASPEN estimates, results were compared to
available ambient air monitoring data. EPA selected a representative subset of seven air toxics
(benzene, perchloroethylene, formaldehyde, acetaldehyde, cadmium, chromium, and lead) mainly
because these air toxics have the largest number of monitoring sites. In general the ASPEN model
was found to underpredict impacts for most components.
F-22
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
The HAPEM4 model was used to predict the nationwide inhalation exposures. Through a series
of calculation routines, the model makes use of census data, human activity patterns, ambient air
quality levels, climate data, and indoor/outdoor concentration relationships to estimate an
expected range of "apparent" inhalation exposure concentrations for populations of concern. The
final step in the assessment include characterizing both cancer and noncancer effects on public
health due to inhalation of study air toxics.
V. Key Results
The assessment developed a list of national priority air toxics. The pollutants that are predicted to
have the greatest impact upon the largest number of people include: Benzene, Chromium,
Formaldehyde, Acrolein. In general results of the assessment will be used to: identify air toxics
of greatest potential concern; set priorities for collection of additional air toxics data at the EPA,
state, local, and tribal level; roughly characterize the relative contributions to air toxics
concentrations and population exposures of different types of air toxics emissions sources (e.g.,
mobile, large industrial, smaller industrial); establish a baseline for tracking trends over time in
modeled ambient concentrations of air toxics; and establish a baseline to measure progress toward
meeting goals for reductions of inhalation risk from ambient air toxics.
VI. Lessons Learned
A detailed peer review of the assessment by the SAB identified both the scientific strengths and
weakness of the assessment. One possible lesson learned involved communicating the results of
the assessment and assuring that these results are not misused. It is important that limitations and
uncertainty on the data be defined and presented along with the results to prevent misuse or
overuse of the study results.
F-23
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
8. Portland, OR
I. Project Name and Location, Project Sponsor, Brief Overview
Portland Air Toxics Assessment (PATA); Portland, OR (Multnomah, Clackamas and Washington
counties); Oregon DEQ, METRO, USEPA (OTAQ and OAQPS)
II. Contact Name and Information
Peter Murchie
Policy, Planning and Standards Group
OAQPS/ESD
tel: 503-326-6554
murchie.peter@epa.gov
Arlene Rosenbaum and Ed Carr, ICF Consulting
III. Key Questions Addressed and Purpose of the Project
Project will help Oregon develop and implement risk/geographic-based air toxics program/rules.
Goal is to have modeled and monitored data with risk characterization to help identify and
prioritize risk reduction strategies.
IV. Uses of Modeling and Monitoring
Five urban monitoring sites in place from 1999 - 2001; refined stationary and mobile (emissions
allocated to major roads) inventories for 1999 will be used to run CALPUFF; HAPEM5 will be
run with ambient results; CALPUFF outputs will be compared with monitoring data to validate
model. Modeled and monitored ambient data and risk characterization will be used to compare to
the National Scale Assessment.
V. Key results
Monitoring information and 1996 National Scale Assessment results were used to identify
pollutant drivers for project scoping. Plan to have modeled and monitored data and comparison
with monitored values to be completed by September 30, 2002.
F-24
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
9. Port Neches, TX, Calcasieu, LA, and Little Rock, AK
Air Toxics Local Scale Assessment Case Studies
I. Project Name and Location, Project Sponsor, Brief Overview
Project name: Regional Air Impact Modeling Initiative: Pilot Studies - Initial Phase
Locations: Port Neches, Texas
Calcasieu, Louisiana
Little Rock, Arkansas
Project Sponsor: U.S. EPA Region 6
II. Contact Name and Information
Contact name: JeffYurk
Institution: U.S. EPA Region 6
Address: U.S. EPA Region 6
1445 Ross Ave
Mail code 6PD-O
Dallas, TX 75202
Phone number: 214-665-8309
E-mail address: vurk.ieffrev@epa.gov
III. Key Questions Addressed and Purpose of the Project
The U.S. Environmental Protection Agency (EPA), Region 6, has established a regional air
impact modeling initiative (RAIMI) pilot program for estimating the combined health risks to
individual receptors within a neighborhood as a result of aggregate exposure to multiple
contaminants from multiple sources and multiple exposure pathways. The initial phase includes
the estimation of potential aggregate inhalation risks associated with modeled air concentrations
from significant local emission sources within the Port Neches Assessment Area, Jefferson
County, Texas.
F-25
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
The overall strategy for completing the RAIMI pilot study was conceived to efficiently maximize
the usefulness of existing guidance, risk assessment tools, and databases to suffice the following
project design goals:
• Useful as a permitting tool to support EPA, state, and local permitting authorities -
independently or combined — evaluate and demonstrate protect! veness of cross
program permitting decisions and support holistic, tailored permit strategies with the
flexibility to be either area (i.e., industrial complex), facility, or source-specific;
• Provide a standardized and consistent means by which all permitting authorities could
account for and assess aggregate health effects to multiple contaminants from multiple
sources, which are often subject to multiple permitting schemes, but cumulatively
impacting the same receptor neighborhoods;
• Provide the necessary level of detailed information to prioritize, and simultaneously
begin identifying potential solutions, for sources resulting in unacceptable risks by
estimating combined health effects resulting from multiple contaminants and sources,
but at a community level of resolution that is specific to definable individual locations,
and generated in a fully transparent fashion such that aggregate risk levels are
completely traceable to each contaminant, each pathway, and each source;
• Calculate and track risks from literally hundreds of sources and contaminants based on
actual emissions data submitted by facilities to the state agency, and as new or refined
data becomes available, it can be directly incorporated into the assessment to obtain
revised risk estimates on practically a real time basis;
• Serve as a versatile and dynamic project platform, allowing for the rapid
identification, characterization, assessment, and management of aggregate
environmental exposures that pose the greatest health risks to the public.
IV. Uses of Modeling and Monitoring
1. North Little Rock, Arkansas: Modeling was used to evaluate chronic exposure at the site.
Monitoring was used in conjunction with modeling to select locations to evaluate acute
exposure at the site.
2. Calcasieu, Louisiana: Modeling used to cite monitors and track emissions which exceed state
standards to their sources of origin.
3. Port Neches, Texas: Canister monitors had identified the area as a chronic exposure problem
for over ten years. Mobile monitoring verified the canister results and also indicated acute
exposure problems. Modeling was used to track measured emissions back to their source(s).
A monitoring to modeling study conducted with this study revealed data gaps associated with
the emissions inventory for both source location and estimated emission rates. Also,
F-26
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
uncertainties were found to be associated with both monitoring and modeling. The magnitude
of the effects of data gaps and uncertainties was found to be chemical- and site-specific.
When data gaps were filled for two of the major chemicals emitted in the area, modeling to
monitoring comparisons were within 15% of each other.
V. Key results
Major program changes for which specific, immediate benefits of the application of the RAIMI
include:
1. Prioritization of resources based on the contribution to human health risk (management of
"worst first")
2. Improved community outreach and involvement by allowing citizens access to more
comprehensive environmental analysis
3. Objective, scientific basis for evaluating new facility sitings, operational modifications
and plant expansions
The RAIMI studies completed to date have focused on evaluation of volatile organic compounds
released to the air over three specific communities. Data generated from application of the
RAIMI facilitated regulatory decisions that focused on environmental benefits.
1. North Little Rock, Arkansas, had a history of citizen complaints surrounding a creosote
plant.
• One facility, multiple emission sources
No significant potential health impacts identified
• Odor problem existed, naphthalene identified
Based on modeling, State elected not to pursue additional permit requirements.
2. Calcasieu, Louisiana had been identified by the National Environmental Justice Advisory
Committee as an area of high concern. The Agency of Toxic Substances Disease Registry
had measured dioxin levels above background in residents' blood.
• 18 major facilities, 2500 point sources
• RAIMI modeling used to validate placement of air monitors and to track excessive
monitored air pollution back to individual sources.
F-27
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
3. Port Neches, Texas had excessively high monitored concentrations of air pollutants dating
back over 10 years.
• 16 major facilities, 1500 point sources
• RAIMI modeling used to locate specific emission source(s) that resulted in
excessive monitored air toxics. Prioritization of permitting and enforcement
actions was narrowed to focus on two facilities and three individual sources.
VI. Lessons Learned
A review of the design goals at the end of the Pilot Study indicates that design goals were
substantially achieved as RAIMI provided:
A flexible mechanism to consider a range of risk management alternatives to
ensure protectiveness;
• A standardized and consistent means to prioritize sources based on risk impacts;
• A sound methodology ensuring that each elemental component of risk or modeled
air concentration is fully traceable to the culpable source; and
• The flexibility to be revised as new or more complete emissions data sets become
available.
Certain of these design goals could be more fully achieved, however, given the following:
• Emissions inventories that are more complete, particularly with regard to
emissions speciation, and inclusion of emissions from minor sources, would
significantly increase confidence in results;
• Data management tools could be developed and incorporated into the RAIMI
approach to improve the ability to access, compare, manipulate and revise data
among the emissions characterization data sets, air dispersion modeling results and
risk modeling results;
• Aspects of the technical approach can be modified to reduce the amount of
extraneous data generated.
F-28
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27. 2002
10. San Francisco Bay Area
I. Project Name and Location, Project Sponsor, Brief Overview
Characterization of Urban Air Toxics Sources in Support of HAPs Emission Control Strategies.
Location: SRI International, Menlo Park, CA. Sponsor: EPA Office of Research and
Development, National Center for Environmental Research, under STAR grant R827927, Deran
Pashyan COTR.
II. Contact Name and Information
Dr. Michael J. Coggiola, SRI International, Menlo Park, CA, 650-859-3045,
niichael.coggiola@sri.com
III. Key Questions Addressed and Purpose of the Project
This research program leverages SRI's development of a continuous emissions monitor (CEM)
for dioxins and furans supported by the U.S. Department of Energy (DOE). The detector uses a
pulsed nozzle gas inlet, resonance enhanced multiphoton ionization (REMPI), and time-of-fiight
mass spectrometry (TOF-MS). Using this Jet-REMPI approach, detection limits in the low 20
parts-per-trillion have been obtained. The extreme sensitivity and chemical specificity of this
instrument, and the nearly universal nature of REMPI and mass spectrometry, provide a new
analytical capability. With a single instrument, the spatial and temporal distribution of a majority
of the most toxic organic HAPs can now be concurrently measured at levels that are of
toxicological interest.
IV. Uses of Modeling and Monitoring
This instrument will provide direct detection and identification of the most HAPs and HAP
mixtures in urban air in real time, i.e. one to several minutes of averaging. The objectives in this
combined laboratory and pilot field study are to establish a viable means of measuring the
emission rates, and temporal and spatial distributions of urban air toxics and HAPs using our
ultra-sensitive CEM. Because our CEM is capable of directly measuring in real-time the
concentration of specific HAPs and urban air toxics at levels far below present analytical
instruments, we can potentially identify and characterize critical emission sources over a wide
geographical area under a variety of ambient monitoring conditions.
V. Key Results
SRI's spectral library now includes more than 120 compounds, including many of the most
common urban air toxics. Using this library, we have begun a preliminary local field study by
collecting urban air samples onto sorbent tubes for off-line analysis. We anticipate that these
analyses will ultimately allow us to not only improve the sampling and analysis protocols, but
also to establish some approximate air toxics levels in the San Francisco Bay Area. This latter
F-29
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
data will permit us to determine if our current instrumental sensitivity is sufficient for real-time
field measurements, or if further improvements will be required to acquire this type of field data.
To test our sampling and analysis approach, samples were taken from a variety of sources,
including automobile exhaust, and ambient air. In ambient air, we collected samples on a carbon
filter element for 3.5 hours, to obtain enough sample to use GC/MS as a survey tool. Only
toluene shows in the GC/MS scan as a small signal. To use Jet-REMPI, we diluted the same
sample by a factor of 20,000 to avoid overloading the instrument. This dilution is equivalent to a
direct measurement with a sampling time of about 1 sec. All of the BETX (benzene, toluene,
ethyl benzene, and the xylenes) compounds were readily seen in the Jet-REMPI instrument, and
the three isomers of xylene could be individually quantified. Our preliminary study showed that
not only can many urban air pollutants, such as BETX, be detected quantitatively using Jet-
REMPI combined with long-term sorbent sampling, but more importantly, because of the high
sensitivity of REMPI, these same compounds could also be detected in the field in near real-time
without preconcentration.
In parallel with the urban air sampling effort, we have also performed a pseudo-field study with
support from the Department of Energy. SRI had a unique opportunity to perform a series of
pseudo-field measurements without the time and expense associated with transporting the
instrument to a field site. This was possible by using the REMPI apparatus that SRI built for Dr.
Brian Gullett at the EPA's NRMRL. Since that apparatus is essentially a duplicate of the one
currently in use at SRI, measurements taken with the system in conjunction with EPA's
combustion facility allowed us to make these "field-like" measurements. Among the most
intriguing observations was the detection of several interesting species that were identified in a
nominally "clean" methane flame. Benzene and phenol (at 150 ppt) were both positively detected
in the off-gas stream of the reactor. In addition, clear evidence was also found for aniline
produced in the EPA reactor under the test conditions. For all species, the measured spectra are
essentially identical to those recorded using a "clean" test gas mixture and all were easily
detected in the exhaust stream. Furthermore, their presence could be entirely attributed to the
methane combustion chemistry as the background levels were not detectable in the absence of the
flame.
This study further demonstrated the usefulness of Jet-REMPI for rapidly detecting and identifying
organic species at trace levels under "field-like" operating conditions.
F-30
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27,2002
VI. Lessons Learned
There are several important lessons learned from both of these simple studies. It seems clear that
our new analytical method is very well suited to measuring air toxics with high temporal and
spatial resolution. Our current instrument, however, is much too large and bulky to take to the
field. In addition, it is clear that more effort is required to develop a direct air sampling inlet
suitable for field studies. SRI is currently pursing the development of a more compact Jet-
REMPI system based on a variety of commercial components, including fixed frequency and
broadband tunable laser systems, and compact time-of-flight mass spectrometers. Once
developed, such a system could be very useful in field studies of urban air toxics.
F-31
-------
U.S. Environmental Protection Agency
Region/QRD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
11. South Coast Air Quality Management District, CA
I. Project Name and Location, Project Sponsor, Brief Overview
Multiple Air Toxics Exposure Study (MATES-II) in the South Coast Air Basin of California.
II. COD ract Name and Information
Mike Nazemi - South Coast Air Quality Management District (SCAQMD)
mnazemi@aqmd.gov
http://mvw.aqmd.gov/matesiidf/matestoc.htm
http://www.aqmd.gov/matesiidf/chapter2.doc
http://www.epa.gov/ttn/chief/conference/eilO/toxics/nazemitoxics.pdf
III. Key Questions Addressed and Purpose of the Project
This was a Environmental Justice (EJ) study involving many stakeholders.
IV. Uses of Modeling and Monitoring
Four Counties (several 1000 sq km); modeling was done at a local scale as well. Two of the three
neighborhoods were about 4 km squares; the third was about a 2-3 km square.
V. Key results
There were three main components: Monitoring; Emission Inventory Development; and
Modeling. The monitoring effort was an enhancement of existing South Coast monitoring in the
area (increase in number of sites and sampling frequency). Monitoring spread across a four
county region from April 1998 - March 1999. Looked at 30 plus pollutants.
Monitoring networks were not dense enough to answer EJ issues or provide culpability. Thus
they supplemented with modeling. First developed a detailed emissions inventory. Modeled
study area with Urban Airshed Model (UAM) at a 2km resolution. Also did microscale modeling
using ISCST3. Study showed diesel PM (DPM) was biggest health concern for air toxics. Also
looked at results without DPM.
VI. Lessons Learned
The biggest lesson appeared to be one of communication issues with the many stakeholders
involved; especially as the project wound down.
F-32
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
12. Barrio Logan, San Diego, California
I. Project Name and Location, Project Sponsor, Brief Overview
Project Name: Barrio Logan
Project Location: San Diego, California
Project Sponsor: California Air Resources Board
II. Contact Name and Information
Linda Murchison, Ph.D.,
Assistant Division Chief
Planning and Technical Support Division
California Air Resources Board
1001 I St., PO Box 2815
Sacramento, CA 95812
(916)322-5350
lmurchis@arb.ca.gov
The Barrio Logan Pilot Study was developed to address concerns about air quality in a
predominantly minority and low-income community in San Diego, and to begin to develop
monitoring and modeling protocols for evaluating environmental justice concerns.
The Barrio Logan community of San Diego was selected for study because it is located in a large
urban area, near major freeways and industrial sources, as well as neighborhood sources such as
gas stations, dry cleaners and automotive repair facilities. In the initial phase of the study, ARB
conducted ambient air quality monitoring at the Memorial Academy Charter School. Monitoring
began in October 1999 and concluded February 2001. Results from the first six months of data
suggested toxic and criteria pollutant concentrations were similar to those measured in other
urban areas of San Diego. Further analysis on all monitoring results are currently being
conducted.
The Barrio Logan Pilot Study is also the first to be conducted for ARB's Neighborhood
Assessment Program (NAP). The goals of the NAP are to assess the cumulative impact of air
pollution sources on communities and to develop guidelines for evaluating strategies for reducing
air pollution impacts at a neighborhood scale. In order to estimate air pollutant concentrations,
ARB staff developed an approach using regional (UAM, Models-3) and micro-scale (ISCST3,
AERMOD) air quality models to assess ambient and near-field pollutant concentrations. Model
results are currently being evaluated using tracer measurements, additional air toxics monitoring,
and uncertainty analysis.
F-33
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27.2002
EMPACT
AIR MONITORING PROJECTS
From 1998 to 2001, the Environmental Monitoring for Public Access and Community Tracking
(EMPACT) Program funded 33 grants to local government agencies (Metro Grants). Metro
Grants support locally proposed and managed environmental monitoring and communication
projects that emphasize active partnerships between local and state government, research
institutions, non-governmental organizations (NGOs), the private sector, and the federal
government. Metro Grants receive up to $400,000 in federal funds and are required to contribute
a local match to encourage sustainability beyond the federal funding period.
To learn more about other EMPACT projects and to order EMPACT technology transfer
handbooks, please visit www.epa.gov/empact. All project abstracts, progress reports and final
reports are posted to www.epa.gov/ncer. For more information, you may contact Madalene
Stevens, EPA Project Officer, at 202-564-2278, stevens.madalene@epa.gov.
The following is a summary of the nine EMPACT Metro Grants conducting air-monitoring
programs.
1. AirBeat: Time-Relevant Communication of Ozone and Particulate Air Pollution, A Pilot
Project to Raise Awareness and Promote Exposure Reduction - Boston, MA
(Local Contact: Jennifer Charles, JenEnviro@aol.com)
http://airbeat.org/
This project developed and implemented real-time ambient air pollution monitoring and data
management techniques for ozone, fine particulate matter (PM2 5), black carbon soot (BC), and
visibility to allow time-relevant communication of these data to the public in a way that can be
readily available, easily understood, applied by members of the community to reduce human
exposure, and used to increase public awareness and understanding of pollution sources, health
effects, and precautionary measures. These data are targeted to the urbanized Roxbury
neighborhood of Boston. This project is the basis for a technology transfer handbook currently
being developed.
2. Burlington Ecolnfo: Community Based Environmental Monitoring in the Burlington
Ecosystem, The Next Step in Building a Sustainable City - Burlington, VT
(Local Contact: Betsy Rosenbluth, brosenbluth@vahoo.com)
http://www.uvm.edu/~-empact/
The goal of this project is to engage citizens in developing environmental information accessible
to a broad cross-section of residents and to use this information to inform collaborative actions
that address priority problems in their urban ecosystem. This is a multi-media project that includes
both ah- and water monitoring data. Hourly ozone measurements are included on the web site.
F-34
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27 2002
3. ECOPLEX: Environmental Conditions On-Line for the Dallas-Fort Worth MetroPLEX -
Denton, TX
(Local Contact: Kevin Theusen, kevin.thuesenrg>cityofdenton.com)
http://www.ecoplex.unt.edu/
The goal of this project is to inform citizens of the current, historical, and near-term forecasts of
environmental conditions to which they are exposed, including water, land, sun and air. The web
site features ozone data and ozone alerts.
4. AirlnfoNow - Tucson, AZ
(Local Contact: Beth Gorman, bgorman@deq.co.pima.az.us')
http://www.airinfonow.corn/
This project aims to develop a unified approach among the collaborating agencies for
environmental data collection, management, reporting, and education using air quality as a pilot
medium. The web site reports live air quality index values for PM10 and PM,5, ozone, and carbon
monoxide, features a web cam visibility photo, and has an extensive education component. The
web site, phone hotline, and newsletter are available in both English and Spanish.
5. Paso del Norte Environmental Monitor - El Paso, Texas
(Local Contact: Ricardo Dominguez, rdominguez@elpasompo.org')
Project web site not yet developed. See http://ozonemap.org/ and http://air.utep.edu/ for related
sites.
This project is using currently monitored ozone, carbon monoxide and particulates combined with
local weather, current traffic conditions and international bridge crossing delays to compile
information into a predictive traffic model to be broadcast to the public in order to encourage
alternative modes of transportation. This project encompasses the tri-state / bi-national area of El
Paso, TX, Sunland Park, NM and Ciudad Juarez, Chihuahua, Mexico and the web site will be
presented in both English and Spanish. This project is the basis for a technology transfer
handbook currently being developed.
6. The Tulsa Air and Water Quality Information System (TAWQIS) - Tulsa, Oklahoma
(Local Contact: Monica Hamilton, mhamilton@ci.tulsa.ok.us)
http://e-tulsa.org/
This project assists people in the metropolitan community and the surrounding regions of
Northeastern Oklahoma connect environmental data to their daily lives and promote public
involvement in environmental policy. This objective will be reached by providing environmental
information, and public awareness and educational programs. This project includes both air and
water monitoring data. The web site features real-time ozone data.
F-35
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
7. Providing Timely Public Access to Daily Air Quality Information About Birmingham,
AL and its Regional Environment - Birmingham, Alabama
(Local Contact: Sam Bell, SBell@.icdh.org)
http://vortex.nsstc.uah.edu/empact_bhm/
This project will develop and implement a greatly improved and sustainable program of local air
quality monitoring and timely and effective public access to useful information about metro
Birmingham as well as regional (southeastern and eastern US) air quality and related
meteorology. The public outreach will also include a program for the promotion of public
awareness and education about air quality and related health effects. The web site features the
current local air quality index; timely ozone, PM10, PM2 5, and SO2 levels; and extensive mapping
capabilities.
8. Rapid Mapping for Clean Air in Commerce City - Commerce City, CO
(Local Contact: Lynn Robbio Wagner, wagnertojtchd.org)
Project web site not yet developed
This project will demonstrate innovative methods for providing timely reporting of the spatial and
temporal distribution of air pollutants in a heavily industrialized urban community. This
methodology will utilize data from real-time measurements of meteorological parameters and
concentrations of air pollutants, atmospheric dispersion models (ADMs), and a Geographic
Information System (GIS) to map the spatial distributions of selected air pollutants.
9. Real-Time Monitoring and Communication of Levels of Fine Particles, Ozone and Black
Carbon in Northern Manhattan - New York, New York
(Local Contact: Swati Prakash, Swati@weact.org')
Project web site not yet developed
The primary objective of this project is to develop and implement real-time monitoring, data
management, and public communication of ambient levels of fine particulate matter (PM2 5),
ozone and black carbon soot (BC) in the urbanized Northern Manhattan neighborhoods of Harlem
and Washington Heights. The project will use instrumentation that is sensitive to diesel fuel
emissions, including an Aetholometer for real-time monitoring of black carbon. Data will be
communicati i to residents in a way that is accessible, understandable, and can be utilized to
reduce their exposure and health risks.
F-36
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment June 25-27, 2002
Regional Community Assessment
Monitoring and Modeling Studies
This table represents information provided by the regions for the past and on-going air toxics activities which evaluate and
address concerns about air toxics at the community level. Some of these activities are only monitoring sites supporting the air
toxics monitoring network. Other activities include complete participation by local communities in assessing and addressing
the local air toxics concerns. Many of the projects include development of local emission inventories, modeling, and monitoring
information. EPA is developing a complete database to be located on the TIN website which will include more detailed
information on these air toxics studies.
For additional information, contact:
Barbara Driscoll
Policy, Planning and Standards Group (C439-04)
USEPA/OAQPS/ESD
Tel: 919-541-1051
Fax: 919-541-0942
Email: driscoll.barbara@epa.gov
F-37
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
5/31/02
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Region 1
Ongoing/Planned '
Merrimack Valley, MA
5 cities/towns
i New Haven, CT
i
I
i
Lawrence, MA
Manchester, NH: area around
airport
Manchester, NH-school
ongoing
Assessment to identify reduction , Air toxics/Stationary,
priorities ; Mobile, Indoor
Toxics Inventory Development j Greenhouse gases and
• toxics
Inventory and Modeling
Inventory of sources and
emissions (1996)
Monitoring-inside and out
All Sources
Air Toxics-All Sources
i VOCs. Particulates and
I Aldehydes
Susan Lancey
Region I
617-918-1656
Barbara Driscoll
OAQPS
i 919-541-1051
Mary Beth Smuts/
Barbara Driscoll
Allen Jarrell/
Barbara Driscoll
Bob White-
NHDES/Doug
Koopman-Air
Rick Rumba-
NHDES/Dr. Rosemary
Caron
i
Review emissions data for use by existing
community involvement networks in 5 diverse
Merrimack Valley towns to clarify, prioritize,
and select the most crucial toxics issues; these
will then be addressed for reduction by
appropriate interventions and/or regulatory
strategies.
Beginning development of emission inventory
and local action plan.
Project will evaluate whether the cumulative
risks due to air pollution (incinerators)
contribute to high asthma rates.
Evaluation of 7 sq mile urban area will use
information to target and conduct inspections.
Indoor/outdoor monitoring in school to correlate
with asthma records.
F-38
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Providence, Rl
Past
Olneyville, Providence, Rl
(1990-91)
Chelsea, MA
Bridgeport, CT( 1987-1988)
Springfield/Chicopee MA
(1987-1988)
Monitoring (2000 pilot site)
Monitoring
EPCRA Inventory
Region 2
Ongoing/Planned
Air toxics-area sources
Air Toxics- All Sources
Air Toxics
Air Toxics
Protecting Communities from Assessment to identify reduction
Toxics, South Camden, NJ priorities
Air toxics/Stationary,
Mobile
Barbara Morin-RIDEM
Barbara Morin-
RIDEM/MSmuts-
Regl
Dwight Peavey-
EPCRA
Marlon Gonzales
Region 2
212-637-3769
OAQPS contact to be
| identified; in interim,
Barbara Driscoll
One of 10 cities picked for new air toxics
monitoring.
Hotspot analysis of jewelry industries in
neighborhood.
Ranked HAPs based on quantity of release and
health effects.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Responding to high-exposure area citizens'
concerns, an array of inventory, modeling and
monitoring tools will be used to characterize
risks and to trigger appropriate risk reduction
strategies.
F-39
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Syracuse, NY
Puerto Rico
Indoor Air
Monitoring (2000 pilot site)
Vicinity of World Trade Center, ; Monitoring
NY
Variety of Air Toxics
Past
| Staten Island, NJ
I (1987-1991)
Ambient and meteorological data VOCs, PM and Metals
collection, indoor air sampling, '
| El developed !
Region 3
Ongoing/Planned
\ Delaware Air Toxics ; Assessment to characterize risk i Air Toxics/Stationary,
I Assessment Study (DATAS) j Mobile
Baltimore Trattic Study
Regional monitoring site, indoor > Mobile Sources w/
and personal monitoring { Indoor Air included, PM,
! Toxics
John Fileppelli
Conrad Simon, Robert
Kelly, Rudolph K.
Kapichak, Carol
Belizzi
Melik A. Spain
Region 3
215-814-2299
GregNizich
OAQPS
919-541-3078
OTAQ
Current study assessing impact of multiple
pollutants on indoor environment.
One of 10 cities picked for new air toxics
monitoring.
Monitoring ambient and worker exposure to air
toxics at WTC site and near vicinity. Data
potentially to be used for future risk assessment.
Qualitative risk assessment performed for
indoor and ambient air. Initiated on citizen
complaints.
Development of monitoring network and
speciated HAP inventory to integrate with
ambient air quality modeling and meteorological
components to define regional/local risks.
Assess mobile source impacts on indoor and
outdoor air pollution concentrations of PM and
gaseous toxics in a home and school.
F-40
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors Region Lead/ Description
OAQPS Contact
West Virginia site
Urban sites and two special
studies
Philadelphia, PA
Past
Baltimore Community
(1996-2000)
Chester, PA (1995)
Monitoring (2000 pilot site)
Monitoring (2001 pilot sites)
Monitoring/Modeling/Risk
Estimation
Inventory, ISCST3 model
Multi-media risk study (criteria
and TRI inventory, ISCST2 and
CAL3QHC models used)
Southern Delaware County Air Monitoring and risk assessment
Monitoring Project
(1995-2000)
28 Air Toxics
28 Air Toxics
Air Toxics/Stationary,
Mobile, Indoor
Air Toxics
Ted Erdman
Ted Erdman
! Ray Chambers, Reg 3
I 215-814-2061
Greg Nizich, OAQPS
919-541-3078
Hank Topper with
OPPT
Air toxics, surface and Dianne McNally,
groundwater, fish tissue,
lead, RCRA/
Superfund facilities,
noise, odor
Volatile Toxics, Metals,
PM
Patrick Anderson
Ted Erdman (R3) and
PADEP
One of 10 cities picked for new air toxics
monitoring.
Two year data collection for eleven urban sites
throughout region and one year data collection
in two areas (three monitors/ area).
Uses modeling to better define city's health
risks and estimate potential reductions, to
inform public and seek government/business
support for reduction measures.
Risk based screening w/ results prioritizing
chemicals and facilities. Report issued in April
2000.
Risk assessment based on ambient pollutant
concentrations were modeled with
ISCST2/CAL3QHC. Recommendations made
for lead paint, targeting sources of air emission,
and voluntary emission reductions.
Long-term monitoring project of three sites
(Chester, Marcus Hook and Swarthmore).
F-41
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Kanawha Valley Toxics
Screening Study (July 1987)
| Pilot Multi-media
I Environmental Health
i Characterization Study of South
I and Southwest Philadelphia
(1997)
An Environmental
Characterization of the District
of Columbia (1997)
Tri-States Initiatve (Kenova,
WV)
Kanawha Valley, WV
Philadelphia PA
Multi-media risk study | 20 Air Toxics, drinking
(point/area inventory and ISCLT j and surface water, haz.
modeling) and monitoring waste
Inventory evaluation, health and
demographic assessment
Multi-media inventory evaluation
See description under Region 4
Air Toxics, drinking and
surface water, haz. waste,
lead, radon
Air Toxics, criteria
pollutants, surface water,
haz. waste
Air Toxics
i Air Toxics
Dianne McNally and
WVDEP
Len Mangiaracina
Jeff Burke
Recommendations include improving inventory,
modeling and monitoring techniques, evaluate
non-cancer risks. WV developed toxics
regulation to address 14 pollutants.
Conclusions include automobiles are biggest
source of pollution, cancer mortality 40% higher
than national rate. Reccos include improving
communication between local agencies and
community to empower community to make
environmental decisions.
Conclusion is that indoor air is primary problem.
Data used in EPA, 1995, Summary of Urban Air
Toxics Risk Assessment Screening Studies to
Support the Urban Area Source Program.
Data used in EPA, 1995, Summary of Urban Air
Toxics Risk Assessment Screening Studies to
Support the Urban Area Source Program.
F-42
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors Region Lead/ Description
OAQPS Contact
Region 4
\ Ongoing/Planned
\ Chattanooga, TN Monitoring
Louisville, KY Monitoring
Air Toxics
Air Toxics
Mobile, AL
Charlotte and Mecklenburg
County
Tampa, FL Monitoring
Monitoring, modeling, inventory Air Toxics
Air Toxics
Vivian Doyle,
Region 4
; Vivian Doyle,
i Region 4
i Van Shrieves
j Region 4
Chris Stoneman,
OAQPS;
Lee Page Reg 4
Alan Powell Reg4
i Van Shrieves,
Region 4
Community Based Environmental Protection
Project to assess the effect of air toxics on the
Chattanooga Community.
Community Based Environmental Protection
Project to assess the effect of air toxics on the
Rubbertown Community in Louisville.
Community Based Environmental Protection
Project to assess the effect of air toxics on the
community in Mobile County, Alabama.
Pilot project to test success of concentrated
regional environmental dialogue and
cooperation in a smaller confined area.
This is one of the 10 national air toxics
monitoring sites under the National Air Toxics
Monitoring Program. The information
developed under this project, in conjunction
with the national data analysis project, will be
used to develop a national air toxics monitoring
strategy.
F-43
-------
U.S. Environmental Protection Agency
Reaion/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Southeast Florida Air Toxics
Study (Broward, Dade, and
Palm Beach Counties)
Piedmont of NC,SC,GA
; Monitoring and inventory
Monitoring
I Tristate Geographic Initiative, j Monitoring, modeling, and
i Greenup Industrial Cluster inventory
i Huntsville, AL
Jacksonville, FL
Knox County, TN
; Modeling and inventory
Monitoring and inventory
Hinds, Jackson, Harrison, and Monitoring
Lee Counties, MS
Monitoring
' Air Toxics
Air Toxics
Air Toxics
Air Toxics
Air Toxics
i Air Toxics
! Air Toxics
Van Shrieves,
Region 4
Van Shrieves,
Region 4
Jackie Lewis,
Region 3,4, and 5
Leonardo Ceron,
Region 4
Van Shrieves,
Region 4
Van Shrieves,
Region 4
Van Shrieves,
Region 4
Three-county monitoring effort to characterize
ambient concentrations of selected air toxics and
to develop a comprehensive emissions inventory
in southeast Florida.
Three-state monitoring effort to characterize
ambient concentrations of selected air toxics in
rural and smaller cities in the southeast.
Community Based Environmental Protection
Project to assess the effect of air toxics on the
community in the vicinity of Greenup, KY.
Update and expand existing hazardous air
pollutant inventory. Conduct dispersion
modeling to evaluate impact of air toxics on
residential areas.
Operation of six mobile and three stationary
monitoring sites. Conduct a comprehensive
emissions inventory.
State monitoring initiative to quantify overall
impact of toxics emission sources in areas of
high concentration of air toxics.
Analysis of trace metals from TSP samples in an
area of the county's highest paniculate
concentration.
F-44
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
1 Location Monitoring/Inventory/ Pollutants/Sectors
Modeling
Jefferson County, KY Inventory and modeling Air Toxics
\-
Mobile, AL Monitoring (Part of National Air Toxics
Estuary Program)
Kingsport, TN Monitoring Air Toxics
Nashville, TN Monitoring Air Toxics
South Florida Monitoring Air Toxics
(Everglades, FL)
! Charlotte, NC Monitoring Air Toxics
Region Lead/
OAQPS Contact
Van Shrieves
Region 4
Thomas Dzomba,
Region 4
Van Shrieves,
Region 4
Van Shrieves,
Region 4
John Ackermann,
Region 4
Van Shrieves,
Region 4
Description
Emissions inventory, dispersion modeling, and
risk characterization will be conducted with
focus on area and mobile sources.
Collection of data to determine the chemistry of
precipitation for monitoring of geographical and
temporal long-term trends. Analytes include
acidity pH, base cations and mercury. Study is
part of the National Atmospheric
Deposition/National Trends/ Mercury
Deposition Network.
Study focusing on air toxics monitoring to
characterize air quality.
Operation of two population oriented monitoring
sites to obtain a better understanding of air
quality conditions.
Atmospheric transport and deposition studies on
mercury including comparisons of methods.
Operation of a mobile monitoring lab in order to
secure a better understanding of the influence of
a large urban center on regional atmospheric
mercury, volatile organic, and carbonyl
compound levels.
F-45
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Past
Tristate Geographic Initiative, : Monitoring, modeling, and
Kenova Industrial Cluster inventory
Louisville, KY
Augusta, GA
Tampa Bay, FL
Atlanta, GA
Jacksonville FL (1986-1987)
Monitoring, modeling, inventory
Monitoring, modeling, inventory
Monitoring and deposition
Monitoring using open path,
continuous gas chromatography,
and conventional sampling
technology. Modeling and
emissions inventory.
Air Toxics
Air Toxics
Air Toxics- Mercury
Air Toxics
Air Toxics/Ozone
Precursors
Air Toxics
Jackie Lewis,
Region 3,4, and 5
Van Shrieves,
Region 4
Van Shrieves,
Region 4;
Danny France,
SESD
John Ackermann,
Region 4
Van Shrieves,
Region 4;
Bob Stevens,
OAQPS
Community Based Environmental Protection
Project to assess the effect of air toxics on the
community in the vicinity of Kenova, WV.
Evaluation of potential usage of open path
monitoring and meteorological technology to
support toxics release inventory.
Conducted study to evaluate mercury releases
from a chlor-alkali facility. Results from the
study were used to set new MACT for this
facility category.
Completed one year atmospheric deposition
study of persistent toxics in rainfall.
Data used by EPA in A Screening Analysis of
Ambient Monitoring Data for the Urban Area
Source Program. Monitoring network and data
was a pilot study to help establish the national
Photochemical Assessment Monitoring Stations
(PAMS) network.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
F-46
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Region 5
Ongoing/Planned
Indianapolis Public School #22 Assessment and Reduction
Local Air Risk Assessment and
Risk Reduction Project,
Indianapolis, IN
Air toxics/Stationary,
Mobile
NATA Point Source Inventory
Refinement in RAPCA
Jurisdiction, Dayton, OH
Cleveland OH
Assessment
Air toxics,
VOCs/Stationary
: Devils Lake, Wl
Inventory/Modeling/Risk
Reduction
Inventory, modeling
Indoor, Outdoor and
Mobile
Mercury
Randy Robinson
Region 5
312-353-6713
Lara Autry
OAQPS
919-541-5544
Michael Compiler
Region 5
312-886-5112
Lara Autry
OAQPS
919-541-5544
Bill Long ORIA ,
Jack Barnette Reg 5,
Janet Cohen OTAQ,
Steve Fruh OAQPS
Erin White
Supplements a more detailed local air toxics risk
assessment on a neighborhood and sources
surrounding a local public school. Proposal
includes a pollution prevention audit of a coking
facility, an environmental audit of the public
school, and investigates other risk mitigation
opportunities.
Refine point source inventory to include all
sources and enable modeling to locate hotspots,
and to select monitoring and risk reduction
locations.
Intended to demonstrate an approach in which
local stakeholders, with advice and support from
EPA work collaboratively to reduce risks from
air toxics in short term while also putting
longer-term strategies in place.
Characterize atmosphere emissions, transport,
deposition, and bioaccumulative
interrelationships for TMDL development
applications.
F-47
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Indiana Harbor, IN
Detroit, Ml
i Detroit, Ml
| Chicago IL
Twin Cities, MN
Region-wide
i Flint, MI
I SE Chicago, IL (1989)
Past
Toxics contained in
contaminated sediments
Allergens and certain
Volatiles
Inventory/Modeling/
Monitoring/Risk Assessment
Personal/Indoor air monitoring
Monitoring (2000 pilot site)
i Monitoring
Personal exposure monitoring , Indoor, Outdoor
; Emissions Inventory
development of 189 toxics
pollutants
i Point, Area, Mobile
Inventoried 21 TRI facilities and j Chemicals released from
54 facilities from MI inventory, j plant except for NAAQS
Used ISCL2 dispersion model. j
i ISCLT and CDM-2 models used. : 30 Air Toxins
George Bollweg
Christopher Saint,
ORD
Suzanne King, Region
5/ Ellen Wildermann,
OAQPS
Michele Palmer
Suzanne King
Characterize community impacts from
volatilized dredged toxic sediments.
Asthma/Children's Study
One of 10 cities picked for new air toxics
monitoring.
Cumulative Risk Initiative around airport in
Chicago.
Characterize Risk.
Database available for multiple uses, including
atmospheric deposition.
Title VI request to evaluate power plant permit.
Developed estimate of human health risks
resulting from emissions under 1992 permit for
• -cnesee Power Plant, and risks from multiple
sources in vicinity. Only considered inhalation.
All point and area sources in an 817 sq mile area
were modeled using ISCLT and CDM-2.
Greatest risk attributed to coke oven emissions.
F-48
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
9 Facilities, MI
Cook County, 1L and Lake
• County, IN
Detroit, MI and Windsor,
Ontario, Canada (1985-1988)
Minneapolis/St. Paul, MN
I Columbus, OH (1989)
Lake Michigan Basin (1991)
Applied IRAP model (based on
RCRA guidance developed in
Region 6)
Used TR1 and RAPIDS
inventories; hazard ratios from
CEP model
Inventory and ISCLT3 model
i
Air Toxics
Inventory and ISCLT3 for point
sources and CDM-2 for area
sources
Air Toxics
Air Toxics
Air Toxics
Selected 9 facilities which were MWCs and
WWBs throughout Michigan. Conducted risk
assessment and identified cancer and non-cancer
risks for inhalation pathway, resident exposure
scenario, and subsistence fisherman scenario.
Response to TSCA petition to evaluate air
permitting process. Hazard screening report
covers two counties in two states. Evaluates
hazard not risk with toxicity-weighted emission
estimates and "hazard ratios" from modeled and
outdoor air monitoring data.
Developed inventory of emissions and modeled
using ISCLT3. Formaldehyde was the largest
contributor to cancer risk, next was coke oven
emissions, 1,3-butadiene, and carbon
tetrachloride.
Developed inventory, modeled and determined
cancer risk. Over 61 % of risk due to road
vehicles.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Data used in EPA. 1994. A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
F-49
-------
U.S. Environmental Protection Agency
Reuion/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Columbus and Akron, OH
(1987)
Cincinnati, OH (1989-1991)
Region 6
Ongoing/Planned
Vary by area
j OAQPS/Region 6/Texas
j 4-Area Air Toxics Project
I Port Neches
; Port Arthur
i Texas City
I El Paso
i
j Adopt a School Bus Program,
! TX
Reduction
Air Toxics
Air Toxics
Vary by area
Air Toxics (diesel PM),
PM,NOx/Mobile
Ruben Casso,
Region 6
Mark Morris, OAQPS
Steven Pratt Reg 6
214-665-2140
Jim Blubaugh, OTAQ
202-564-9244
Yvonne Chandler
OAQPS
919-541-5627
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
The assessments will focus on State and Federal
agency experience to identify roadblocks, needs
and potential solutions in assessing and
addressing air toxics risk.
Replace pre-1977 buses, develop alternative fuel
infrastructures and provide PM retrofit for diesel
school buses to reduce student exposure to
newer diesel PM and NOx.
F-50
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Louisiana
Calcasieu Parish
Oklahoma
PoncaCity
Oklahoma
Tulsa
i Oklahoma City
The enhanced sampling which
began in January 2001. For
VOCs, a 24-hour average sample
will be collected every sixth day
for three years. Fordioxins,
furans, and coplanar PCBs, a 30-
day average sample will be
collected every other month for
one year.
Inventory is done by state.
Modeling & risk assessment to
be conducted by EPA in
cooperation with ODEQ.
Inventory to be done by state.
Modeling & risk assessment to
be conducted by EPA in
cooperation with ODEQ.
Ambient air monitoring study in
Oklahoma City in December
2000 and Spring 2001.
107 different volatile
organic compounds
(VOCs), dioxins, furans,
and coplanar PCBs at
five sampling locations
throughout the airshed
Refineries
Urban area - mix of
stationary/mobile
sources
Dioxin
Region 6 contacts
Steve Thompson
Sunita Singhvi
i Ruben Casso,
j Region 6
Ruben Casso,
Region 6
Steve Thompson,
Region 6
EPA Region 6, the Louisiana Department of
Environmental Quality and local industry are
contributing to an extensive air toxics sampling
effort in Calcasieu Parish. A year of sampling
has been conducted and an evaluation and report
of the latest monitoring data in light of the state
ambient air toxics standards will be completed.
Local air toxics emissions/impacts/risk
assessment in the Ponca City area.
Project would also serve to build state capacity
(tools, knowledge, ability).
Potential project based on the results/lessons
learned from Ponca City, OK pilot project.
The Tulsa area is larger and has more of a
mobile source component.
The purpose of the study is to obtain data which
may be useful in determining background urban
levels of dioxin in the air where no major dioxin
emission sources are located.
F-51
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
I New Mexico
i Corrales, NM
Rio Rancho, NM
Region 6 has 5 sites in the
National Dioxin Air Monitoring
Network in Arkansas,
Oklahoma, and Texas
Channelview, TX - Community , Citizen monitoring by tedlar bags i VOCs
Involvement Joint Effort -EPA and canisters i
and TNRCC and Harris County i
To be determined
Under discussion
Project is likely going to be a 30-
day monitoring effort.
12 months of ambient monitoring j At least 18 hazardous air
' pollutants
Dioxin
LA/TX
Voluntary Episodic Release
Reduction Project
Episodic Releases
i Chemical Industry
Ruben Casso,
Steve Thompson,
Region 6
Mark Sather,
Region 6
Steve Thompson
Region 6
William Rhea-
Region 6
Barry Feldman,
Region 6
Citizen complaints alleging health effects from
emissions in the area.
New Mexico requested Region 6 assistance to
help investigate potential levels and impacts of
local air toxics near Corrales, NM
Small cities air toxics monitoring project with
coordinated with stationary and mobile source
emission inventory work.
EPA Region 6 has 5 sites in the national dioxin
air monitoring network in Arkansas, Oklahoma,
and Texas in cooperation with the state
environmental agencies. National Dioxin Air
Monitoring Network The data are intended to
help determine background dioxin levels where
no major dioxin emission sources are located.
An effort to educate citizens on air sampling, air
data availability and air program
implementation.
Region 6 worked with selected facilities to
voluntarily seek and obtain the reductions in the
incidence of, and emissions from, episodic
emissions from facility upsets and malfunctions.
F-52
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Houston Ozone Nonattainment
Area
i Recent
; Port Neches, TX
North Little Rock/Little Rock,
AR
Past
Houston, TX( 1990)
Episodic Releases
VOC
Monitoring inventory modeling
and risk assessment completed
30-day toxics monitoring and
risk& health screening
Local assessment of
chemical plant area
Wood treatment facility
Inventory based on Houston-
Galveston ozone SIP, and RAM
dispersion model used
Air Toxics
Jim Yarbrough,
Region 6
Ruben Casso, Jeff
Yurk, Region6
Mark Morris, OAQPS
Steve Thompson and
Ruben Casso,
Region 6
I
Investigating the feasibility of documenting
significant episodic releases of VOC emissions
from facility malfunctions/upsets and start-
up/shutdowns in the Houston area. This pilot
project will seek the cooperation of area
stakeholders to help identify, track and
document these emissions and their impact on
local air quality.
Region 6 initiated study of HON facilities
evaluating ways to reduce risk.
EPA Region 6 and the Agency for Toxics
Substances and Disease Registry (ATSDR)
provided risk and health screening evaluations,
respectively of 30-days of ambient air toxics
monitoring data collected by the State of
Arkansas around the Koppers' Industries wood
treatment facility.
Past monitoring and emission inventory
development. Modeled results using traditional
approaches (population distributions or land use
distribution), compared with CIS.
F-53
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Texas - Dallas/Fort Worth and
j Houston/Harris County
Combustion Facilities
) Chambers, Harris, Jefferson,
j and Orange County, TX
j (1987-1991)
I
; Baton Rouge, LA (1988-1992)
Monitoring data and modeling ' 12 Air Toxics
. ISCST-3 dispersion model used
on a 100 meter grid.
Air Toxics
Air Toxics
! Region 7
Ongoing/Planned
j St. Louis, MO
ASPEN model and ambient
; monitoring data.
i Air toxics, VOCs, NOx,
! PM/Stationary, Mobile,
Indoor
JefFYurk,
Cynthia Kaleri,
Steve Thompson
Jim Hirtz,
Region 7
913-551-7472
Yvonne Johnson,
OAQPS
919-541-2798
Asbestos, chromium, benzene, formaldehyde,
ally) chloride, aery Ion itrile, and ethylene
dichloride of concern.
Evaluated health risks in communities closest to
incinerators. Developed a risk-based tool to
evaluate these sites. Trace back to large
facilities were reductions should be made.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Improve outdoor monitoring capability and
prepare outreach materials providing real-time
data on key pollutants (e.g. formaldehyde) and
their significance, for the public to use in
reducing exposure and in making decisions on
reduction strategies. Also monitor
formaldehyde indoors, if possible.
F-54
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Iowa rural monitoring site
Region 8
Ongoing/Planned
Denver County
Green Fleets Outreach
and Awareness Program
Denver, CO
Colorado
Parchute, Colorado
Monitoring (2000 pilot site)
Emission inventory, modeling
and monitoring
Air toxics/Stationary,
Mobile and Indoor
Reduction
Air Toxics (diesel PM),
PM/Mobile
Monitoring (2000 pilot site)
Risk Assessment
Air Toxics
Victoria L. Parker-
Christensen, Reg 8
303-312-6441
Peter Murchie, OAQPS
503-326-6554
Jim Blubaugh, OTAQ
202-564-9244
Anne-Marie Patrie,
Region 8
303-312-6524
Peter Murchie,
OAQPS
503-326-6554
Jim Blubaugh, OTAQ
202-564-9244
Lawrence Wapensky
One of 10 locations picked for new air toxics
monitoring.
Initiate implementation phase, based on
completed assessment. Analyze existing
programs, engage voluntary support to pilot new
reduction strategies.
Reduce diesel emissions by involving
public/private fleet owners/operators of both on-
and off-road equipment in voluntary education
and recognition program.
One of 10 cities picked for new air toxics
monitoring.
Pilot air toxics risk assessment of natural gas
field, dehydration units and other local sources.
F-55
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
i Pueblo, Colorado
Monitoring/Modeling/Risk
Characterization
Point, Mobile, Air Toxics Lawrence Wapensky
Region 9
Ongoing/Planned
West Oakland California Air Assessment and reduction
Toxics/Environmental Justice
Funding Proposal, CA
Air Toxics (diesel PM),
PM/Mobile
Henderson, Nevada (Clark
County) Air Toxics/
Environmental Justice Funding
Proposal, NV
Children's Environmental
Health Protection Monitoring;
Community Health Program:
Barrio Logan (SD), CA
Assessment
Air toxics, PM/
Stationary, Mobile
Inventory, Modeling,
Monitoring, Risk
Characterization
Toxics, Diesel PM
Richard Grow, R9
415-947-4104
JoLynn Collins,
OAQPS
919-541-5671
Chad Bailey, OTAQ
734-214-4954
Roy Ford,
Region 9
415-972-3997
Peter Murchie,
OAQPS
503-326-6554
CARB
Dale Shimp (CARB)
Characterize and quantify emissions form
Rocky Mountain Steel Mill, Comanche power
plant, local cement plant, potential incineration
of mustard nerve agent at Pueblo Army Depot,
mobile, fugitive and growth.
Document traffic/idling diesel truck Port-related
traffic patterns and impacts. Inform citizens of
available resources, research mitigation options,
involve owner/operators.
Initiate voluntary gaming industry energy
efficiency program, with collateral PM
reduction, plus air toxic assessments to form
basis for program in very rapidly growing
county.
Community level assessment conducted by
CARB as part of its overall risk-based program
(monitoring completed).
F-56
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
I Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors Region Lead/ Description
OAQPS Contact
Children's Environmental
Health Protection Monitoring;
Community Health Program:
Boyle Heights (LA), CA
Children's Environmental
Health Protection Monitoring;
Community Health Program:
Wilmington (LA), CA
Children's Environmental
Health Protection Monitoring;
Community Health Program:
Crockett (Contra Costa),
Fruitvale (Oakland), and
Fresno, CA
Maricopa County, AZ
Pearl City, HI
Inventory, Modeling,
Monitoring, Risk
Characterization
Inventory, Modeling,
Monitoring, Risk
Characterization
Inventory, Modeling,
Monitoring, Risk
Characterization
Design a comprehensive
inventory and monitoring
network
Toxics, Diesel PM
Toxics, Diesel PM
Toxics, Diesel PM
Toxics, PM, VOCs
Future work not decided yet.
One monitor in place.
CARB
1 Dale Shimp
j (CARB)
CARB
Dale Shimp
i (CARB)
CARB
Dale Shimp
! (CARB)
j Doug McDaniel
| (USEPA R9);
Barbara Driscoll
(OAQPS)
PM(currently); toxics to Roy Ford
be installed I USEPA R9
Community level assessment conducted by
CARB as part of its overall risk based program
(expected to complete monitoring in March
2002)
Community level assessment conducted by
CARB as part of its overall risk based program
(expected to complete monitoring in May 2002).
Community level assessment to be conducted
by CARB as part of its overall risk based
program. Expected to start monitoring at
Crockett in Oct. 2001, and at Fruitvale in Nov.
2001. Monitoring schedule for Fresno is to be
determined.
Joint Air Toxics Assessment Project:
collaborative efforts among Tribes, State,
County, OAQPS, and R9; special emphasis on
three Indian reservations located within the
study area; is also one of the six potential
project areas for the R9 Air Toxics/EJ Initiative;
initial workplan to be developed.
R9 Air Toxics/EJ Initiative: one of the six
potential project areas; initial scoping; initial
workplan to be developed.
F-57
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
! Clark County, NV
SF Bay Area, CA
Los Angeles Airport, C A
LA Alameda Corridor, CA
LAWA Source Apportionment
Study at LAX
San Jacinto, CA
Area of focus would be Toxics
Henderson City NV ;
Monitoring and inventory
development
West Oakland; Richmond (one ! Toxics
dioxin monitor)
Initial focus would be on active Toxics
R9 involvement in current and !
future studies and activities at
LAX
Initial focus on active R9
involvement with Gateway
Cities' Diesel Emission
Reduction Program; possibly
work with CARB in the
Wilmington area including
Federal Measures at the Ports of
LA and Long Beach.
Monitoring at LAX and
neighboring communities
Part of the ten cities monitoring
pilot project (four urban sites and
six rural sites)
Diesel emissions
Toxics and criteria
pollutants
Toxics
Ken Israels
USEPA R9
Richard Grow
USEPA R9
Pam Tsai
USEPA R9
Valerie Cooper
USEPA R9
Winona Victery/ Pam
Tsai
USEPA R9
Sharon Nizich
OAQPS
R9 Air Toxics/EJ Initiative: one of the six
potential project areas; initial scoping
completed ; initial workplan to be developed;
Working closely with CCAQMD.
R9 Air Toxics/EJ Initiative: one of the six
potential project areas; continue scoping efforts;
draft initial workplan to be developed.
R9 Air Toxics/EJ Initiative: one of the six
potential project areas; initial scoping and data
gathering completed; draft initial workplan to be
developed.
R9 Air Toxics/EJ Initiative: continue scoping
evaluation (one of the six potential project
areas); draft initial workplan to be developed.
Project is put on hold after Sept. 11; LAWA is
seeking funding from EPA.
One of the six small city/rural sites selected for
the FY2000 national air toxics monitoring pilot
project.
F-58
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Descriptipn
Los Angeles, CA
Fresno FACES
Past
Flight Path Paniculate Fallout
Study in the Area of LAX
(2000)
Air Monitoring Study in the
Area of LAX (1999)
South Coast Air Basin, CA
(1998-1999)
Air sampling in and outside bus
Indoor and personal sampling,
neighborhood monitoring and
EPA supersite
Monitoring during weeks of
April-May 2000 in Inglewood
One-month monitoring in the
vicinity of LAX
Monitoring, emission inventory
update, and modeling
Diesel, PM, CO, N2O,
formaldehyde/mobi le
sources
PM, toxics, ozone,
environmental tobacco
smoke, biological
agents/indoor
Combusted oil soot
particles
Air Toxics, Diesel PM
Air Toxics
OTAQ
Tracy Hysong
CARB
Henry Hogo
| SCAQMD
Henry Hogo
SC AQMD
Henry Hogo
SC AQMD
School bus exposure assessment to quantify in-
vehicle, outside vehicle, near vehicle and
ambient exposures to diesel exhaust.
Began in Oct/Nov 2000. Study of the effect of
air pollution on 450 asthmatic children.
Examines short term effect of daily air pollution
on the symptoms, medication use, and lung
function of these children and longer-term effect
on the progression of asthma. OTAQ adding
additional funding to look at mobile hot spots.
Expect to have final reports in 2005.
The study was conducted as a follow-up to an
earlier study that found abundant combusted oil
soot particulates around LAX.
The study was conducted to address public
concerns about air pollution that may be
attributable to LAX operations.
Multiple Air Toxics Exposure Study-H; quantify
the magnitude of population exposure risk from
"existing" sources; monitoring at ten fixed sites
and 14 microscale community locations.
F-59
-------
U.S. Environmental Protection Agency
Reeion/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27,2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Phoenix, Tucson, Casa Grande, Monitored for a year in
i and Payson, AZ (1994-1995) residential back yards,
! Atmospheric Simulation Model
used
Los Angeles CA (1995-1998) 3 monitors in Los Angeles
County
Contra Costa CA (1985)
San Francisco CA (1987-1993)
South Coast CA( 1986-1993)
I Southern CA (1987-1989)
Santa Clara CA
Inventory and modeled using
LONGZ
Air Toxics
10 Air Toxics
16 Air Toxics
Air Toxics
Air Toxics
Air Toxics
Air Toxics
Monitored and modeled air toxics in residential
areas. 1,3-butadiene, benzene, formaldehyde,
arsenic, carbon tetrachloride, acetaldehyde, and
chloroform greatest risk.
Average concentrations compared with I in a
million cancer risks. Most of risk due to 1,3-
butadiene, formaldehyde, and benzene.
Modeled air concentrations. No health risk
assessment done. Highest concentrations were
methylene chloride, formaldehyde, benzene, and
perch lorpethylene.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Data used in EPA, 1994, A Screening Analysis
of Ambient Monitoring Data for the Urban Area
Source Program.
Data used in EPA, 1995, Summary of Urban Air
Toxics Risk Assessment Screening Studies to
Support the Urban Area Source Program.
F-60
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Region 10
Ongoing/Planned
Regional Scale Air Dispersion
Modeling for Hazardous and
Toxic Air Pollutants in the
Treasure Valley, Boise, ID
Assessment of Emission
Inventories and Exposure of Air
Toxics Using an Automated
Grid Modeling System
Portland, OR
Seattle, WA
Anchorage, AK (statewide)
http://www.state.ak.us/dec/daw
q/aqi/toxicstrategy.htm
Assessment
Assessment
Inventory/Modeling/
Monitoring
Air Toxics/Stationary,
Mobile
Air Toxics, PM, VOCs/
Stationary, Mobile
Urban Air Toxics and
diesel PM
Monitoring (2000 pilot site)
Inventory development,
monitoring
Benzene/ Indoors and
Ambient
Brook Madrone, RIO
206-553-1814
Peter Murchie,
OAQPS
503-326-6554
Brook Madrone, RIO
206-553-1814
Peter Murchie,
OAQPS
503-326-6554
Paul Koproski,
Brook Madrone,
Peter Murchie
ODEQ/USEPA
Brook Madrone
Alaska Department of
Environmental
Conservation
Develop regional scale modeling capacity of
Idaho DEQ's new Monitoring, Modeling and
Emission Inventory group, using more
technologically refined tools transferred from
WA, to measure Treasure Valley HAP.
Expand automated air modeling system to
include metropolitan Seattle and Portland, and
also provide technology transfer assistance in
use of the system to Idaho.
Study will use 96/99 NTI and additional
monitoring data to characterize localized risk
information and identify/prioritize reduction
management options. Will also provide
information needed for State risk based air
toxics program development (regulation
development).
One of 10 cities picked for new air toxics
monitoring.
Looking at benzene and other pollutants,
indoors and out.
F-61
-------
U.S. Environmental Protection Agency
Region/ORD/OAR Workshop on Air Toxics Exposure Assessment
June 25-27, 2002
Location
Monitoring/Inventory/
Modeling
Pollutants/Sectors
Region Lead/
OAQPS Contact
Description
Tacoma, WA
Bellingham, WA
Port Angeles
Seattle, WA
Past
Ambient and stack monitoring
Air Toxics
Ambient and stack monitoring of an incinerator
in Bellingham. Incinerator shut down in the
90s.
EPA and state conducted series of studies to try
and determine why asthma rates were 6 times
higher than expected.
Data used in Radian Corp. 1995, Model City
HAP Analysis Memorandum
F-62
------- |