United States
Environmental Protection
Agency
Office of Pollution Prevention
and Toxics (7406)
Washington, DC 20460
EPA 744-R-00-005
April 2000
!
>.
Baltimore Community
Environmental Partnership
Air Committee
Technical Report
."•• _ >-.*.• v • *
-.,-. '*--
Community Risk-Based
Air Screening:
A Case Study in Baltimore, MD
A Product of the
Community Environmental Par*
-------
EPA 744-R-00-005
April 2000
Final Report
Baltimore Community Environmental Partnership
Air Committee Technical Report
Community Risk-Based Air Screening:
A Case Study in Baltimore, MD
U.S. Environmental Protection Agency
Office of Pollution Prevention and Toxics
Washington, DC 20460
April 30, 2000
-------
TABLE OF CONTENTS
Page No.
ACKNOWLEDGMENTS v
INTRODUCTION 1
The Community Environmental Partnership 2
Environmental Setting 4
The Partnership Air Committee and Goals 5
Air Committee Meetings and Work 5
Overview of the Community Air Screening Methodology 6
Understanding What the Baltimore Risk-Based Screening Effort Could and Could Not
Accomplish 8
Summary Flow Chart 10
BUILD PARTNERSHIP (STEP 1) 13
Formed Partnership 13
Clarified Air Committee Goals 14
Developed Plan for Community Outreach 17
EMISSIONS INVENTORY (STEP 2) 19
Overview 19
Sources for Identifying Facilities 20
Sources Used To Collect Emissions and Ambient Air Monitoring Data 22
Database Management 25
INITIAL SCREEN (STEP 3) 27
Overview 27
Initial Screen Procedures 28
Background Information on Risk Screening 29
Collection of Toxicity Information 33
Calculation of the Air Concentration and Potential Dose 33
Calculation of Cancer Risk Estimates and Hazard Quotients 35
Source Inventory Database 36
Selection of Screening Values 36
Comparison of Cancer Risk Estimates and Hazard Quotients to Screening Values .... 37
SECONDARY SCREEN (STEP 4) 43
Overview 43
Completing the Secondary Screen 45
Selection of Health-Based Screening Levels and Endpoints 50
Chemicals with Monitoring Data 50
Results of Secondary Screening 51
Interpretation and Communication of Results 52
-------
TABLE OF CONTENTS (Continued)
Page No.
FINAL SCREEN (STEP 5) 55
Overview 55
Completing the Final Screen 56
Results of the Final Screen 58
DEVELOPED RECOMMENDATIONS AND COMMUNICATED RESULTS TO THE
BROADER COMMUNITY (STEP 6) 63
Overview 63
Recommendations for Acting on Results 63
Communication of the Results 64
GENERAL OBSERVATIONS ON THE SCREENING METHODOLOGY DEVELOPED IN
BALTIMORE 67
The Methodology Provides an Effective Screening Tool for Communities 67
The Methodology Helps Facilitate the Mobilization of Local Resources to Make
Improvements in Local Air Quality 68
The Technical Aspects of Screening Methodology Need Further Refinement 69
Specific Lessons Learned for Each Step of the Screening Methodology 69
REFERENCES 77
LIST OF APPENDICES
APPENDIX A List of Community Environmental Partnership (CEP) Air Committee
Members A-l
APPENDIX B Letters from Partnership B-l
APPENDIX C Sources for Facility Information C-l
APPENDIX D Toxicity Information D-l
APPENDIX E Document for Generic Turner Method for Estimated Exposure from
Near-Ground Releases to Air E-1
APPENDIX F Examples of Emissions, Site, and Monitoring Data Collected by
Committee F-l
APPENDIX G Example of Database Columns Developed for the Community Pilot
Project Air Emissions Database G-l
APPENDIX H Facilities Modeled for Secondary Screen H-l
APPENDIX I Results of Secondary Screening for Target Toxics 1-1
APPENDIX J Partnership Air Committee Report J-l
APPENDIX K Baltimore Air Dispersion Modeling K-l
APPENDIX L Peer Review Comments and Response L-l
11
-------
LIST OF TABLES
Page No.
Table 1 Sources Included and Not Included in the Inventory for the
Baltimore Case Study 21
Table 2 Chemicals Selected from Initial Screen 40
Table 3 Results of Secondary Screening for Target Toxics in Partnership
Neighborhoods 52
Table 4 Emission Rates from Facilities Used in Secondary and Final Screen 57
Table 5 Estimated Air Concentrations of Chemicals from Secondary and Final
Screens 59
LIST OF FIGURES
Figure 1 Case Study Area - Baltimore, Maryland 3
Figure 2 Flow Diagram for Air Screening Methodology 11
Figure 3 Coarse Receptor Grid in Baltimore 48
Figure 4 Fine Receptor Grid in Baltimore 49
Figure 5 Comparison of Unknown to Stationary Sources of Benzene Between the
FMC Monitoring Station and Modeled Concentrations 60
Figure 6 Baltimore Screening Results 62
Figure 7 Generic Air Screening Methodology for the Community 76
in
-------
IV
-------
ACKNOWLEDGMENTS
This report is a summary of work done by the Air Committee of the Baltimore Community
Environmental Partnership from 1996 through 1999. The report was prepared by technical staff of
the U.S. EPA Office of Pollution Prevention and Toxics for the Air Committee of the Baltimore
Community Environmental Partnership. The main EPA authors of the text were David Lynch,
Greg Macek, and Hank Topper. The following EPA staff also contributed significantly to the
writing of the report and to the work in Baltimore: Ethel Brandt, Damon Brown, James Darr, Deb
Forman, Matt Gillen, Todd Holderman, Reggie Harris, Dawn loven, Jim Murphy, Nhan Nguyen,
Terry O'Bryan, and Van Shrieves.
Members of the Air Committee who contributed to the work in Baltimore are listed in
appendix A. Not all members listed in this appendix participated in the review and approval of this
report. The following members of the Air Committee members reviewed, commented on, and
approved this report: Suzanne Bond, Pars Ramnarain, and William Paul, Maryland Department of
Environment; Don Torres and Rubin Dagold, Baltimore City Health Department; Peter Conrad,
Baltimore City Department of Planning; Dr. Michael Trush, Johns Hopkins School of Hygiene and
Public Health; Rev. Richard Andrews and Ed Looker, community residents; Dave Mahler, Condea
Vista; Rebecca Besson, Delta Chemical; Richard Montgomery, Phoenix Services; John Quinn,
BGE; Steve Dyer, Grace Davison; and Charles Nardiello, Arundel Corporation.
Versar, Inc., under contract Nos. 68-W6-0023 and 68-W-99-041, assisted EPA in the work
in Baltimore and in the preparation of the case study. The EPA Work Assignment Managers were
Dave Lynch and Damon Brown. The Project Officers for these contracts were Tom Murray, Cathy
Fehrenbacher, and Cathy Turner, respectively. Pat Wood and David Bottimore were Versar's
Work Assignment Managers for the study. David Bottimore, Pat Wood, Amanjit Paintal, and De-
Mei Meng contributed to the writing of the report. Supporting Versar staff included Teri
Schaeffer, Maggie Wilson, Kelly O'Rourke, Tim Sletten, Bill Jones, and Jay Wind. Word
processing, graphics, technical editing and logistics support were provided by Valerie Schwartz,
Susan Perry, Sandy Paul, Jennifer Baker, Janeice Zeaman, and Kathy Kelly.
The following EPA staff reviewed drafts of the report and provided comments: Carole
Braverman (Region 5), Rich Cook (OAR/OMS), Jeneva Craig (OAR/OAA), Lois Dicker (OPPT),
Andrea Pfahles-Hutchens (OPPT), Elizabeth Margosches (OPPT), Lawrence Martin (ORD), Doris
Maxwell (OAR/OAA), Deirdre Murphy (OAR/OAQPS), Paul Rasmussen (OAR), and Ed Weiler
(OPPT).
A peer review of a draft of this document was conducted by Eastern Research Group
(ERG) under Contract No. 68-W6-0022. Todd Holderman (EPA/OPPT) was the Work
Assignment Manager, Carol Rawie (EPA/OPPT) was the Project Officer, and Linda Cooper was
ERG's Task Manager. Six peer reviewers were selected on the basis of their expertise in air
quality assessment, emissions modeling, and risk screening/assessment. The following experts
served as peer reviewers of this report: Mr. Michael Callahan (EPA/NCEA), Dr. Gail Charnley
(HealthRisk Strategies), Dr. Douglas Crawford-Brown (UNC/Chapel Hill), Dr. Amy D. Kyle
(UC/Berkeley), Dr. Kenneth L. Mitchell (EPA/Region 4), and Dr. Ronald Wyzga (EPRI).
-------
VI
-------
INTRODUCTION
Baltimore Case Study: Risk-Based
Air Screening
• Southern Baltimore, Maryland
• Six-Step Risk-Based Air Screening
Process Applied to 125 Sources
and 175 Chemicals
• Identification of Chemicals of
Concern
• Accomplishments and Limitations
of Screening Effort
This Baltimore Case Study report
describes the work and the results of a risk-
based air screening project in Baltimore,
Maryland. The report was prepared by
technical support staff of the U.S.
Environmental Protection Agency's Office of
Pollution Prevention and Toxics (OPPT) and
Versar, Inc., for the Air Committee of the
Community Environmental Partnership (CEP),
located in southern Baltimore City and northern
Anne Arundel County, Maryland. The
introduction to this case study report describes
the CEP, the Air Committee, the risk-based
screening methodology, and the
accomplishments and limitations of the
screening effort. Following the introduction
are sections that present the application of the
six air screening steps to the assessment of air pollution sources in southern Baltimore. These are
followed by a summary of the results and lessons learned. The public report that was prepared to
communicate the results of the study to the community is included in Appendix J. The results
described in that report and this Baltimore Case Study report provide preliminary answers to
questions raised by community members about the air quality in their neighborhoods.
This is a case study of the work as it was carried out in Baltimore. The screening
methodology described in this report is a work in progress. During the course of the work and in
the effort to summarize the work for this case study, the participants in this effort identified many
areas for improvement. These are noted throughout the text and summarized in the section on
lessons learned. In addition, the case study was reviewed by independent experts in a formal peer
review process. Additional suggestions for improvement were developed from these reviews. A
summary response to the peer review comments and the comments themselves can be found in the
appendices. While recognizing the validity of the suggestion for improvements in the scope and
methods of the Baltimore Study, the Partnership Air Committee is confident that the information
provided to the community in this report is both significant and valid.
EPA technical staff of the Office of Pollution Prevention and Toxics are now using the
suggestions for improvement from the participants and the peer reviewers to develop an improved
screening methodology and a "how-to" manual to help communities interested in understanding
and improving their air quality. This improved methodology and manual will be available in the
spring of 2000. For further information on this work or this case study, please contact the
Community Assistance Technical Team of the Office of Pollution Prevention and Toxics. See
contact information in the box at end of the Introduction (page 10).
-------
The Community Environmental Partnership
On May 3, 1996, the residents, businesses, and organizations of four Baltimore area
neighborhoods—Brooklyn/Brooklyn Park, Cherry Hill, Curtis Bay, and Wagners Point—joined
with local, State, and Federal government agencies in the Community Environmental Partnership
(CEP) to begin a new effort to find ways to improve the local environment and economy. The
five neighborhoods in the Partnership, with a combined population of about 30,000, are located in
southern Baltimore City and northern Anne Arundel County (Figure 1). These neighborhoods
have a broad range of environmental and economic concerns, including concerns that arise from
the concentration of industrial, waste treatment and disposal, and brownfields sites located in and
around the Partnership area. The area has great potential for the development of its
environmental assets and its economy. The neighborhoods border the Chesapeake Bay and are
the site for a new eco-industrial park, a major redevelopment effort that has the potential to
attract new jobs. In this context, the Partnership set out to take a comprehensive look at the local
economy and environment and to build consensus around a plan for action.
The Community Environmental
Partnership started as a pilot for the new
community-based approach to environmental
protection and economic development.1 This
new approach is an effort to address
environmental issues from the perspective of
a neighborhood. It incorporates the local
community's knowledge and allows for the
consideration of a detailed level of
information often missed when policy is
made at the national or State level. The
community-based approach changes the roles
of the community and government: It
empowers the community to take the lead in
the decisions affecting their environment, and
it puts government in the role of an adviser,
providing the information and technical
assistance not available in the community.
Building consensus at the local level also ™
makes it possible to unite the community
around voluntary pollution prevention approaches
requirements.
Community Environmental Partnership
Community Residents
Businesses
Organizations (Local Schools and the
Johns Hopkins School of Public Health)
Local Government (Baltimore City and
Anne Arundel County)
State Government (Maryland Department
of the Environment)
Federal Government (U.S. EPA)
that can go beyond current statutory
See EPA's Framework for Community-Based Environmental Protection, U.S. EPA. February 1999.
EPA237-K-99-001 (U.S. EPA, 1999a).
-------
^Brooklyn
Arundel'Village
» \-m ,
V" .••"««»'« """ '%^^,:"-^~'>'kYV_,_
xs
Figure 1. Case Study Area - Baltimore, Maryland
-------
At the beginning of this effort, the partners agreed to focus on the following four goals:
1. Build the long-term capacity of the community, including residents and businesses,
to take responsibility for their environment and economy,
2. Develop a comprehensive picture of the local environment and economy and
an action agenda based on the needs and wants of the community,
3. Build consensus in the Partnership for the implementation of an action plan that
makes a difference in the local environment and economy, and
4. Encourage and support sustainable economic development in the community.
As its first initiative, the CEP conducted a publicity campaign and, in July 1996, held a
large public meeting to solicit public input and participation for the project. At this meeting,
community residents and businesses discussed and voted on priorities for the Partnership. Five
areas were identified as community priorities: (1) air quality; (2) trash, illegal dumping, and
abandoned housing; (3) economic development; (4) parks and surface water quality; and (5)
community health. Committees were formed to address each of these priorities. This report
details the work of the CEP Air Committee. The John Snow Institute is currently preparing a
separate report on the overall work of the CEP.
Environmental Setting
The case study area of southern
Baltimore is an industrialized area with a large
concentration of industrial, commercial, and
waste treatment and disposal facilities. Among
these are 11 facilities reporting air emissions to
the U.S. Environmental Protection Agency
(EPA) Toxics Release Inventory (TRI). Major
facilities include an agricultural chemicals
manufacturer, other chemical manufacturers,
petroleum storage facilities, a medical waste
incinerator, the city landfill, and a municipal
wastewater treatment plant (publicly owned
treatment works [POTW]). Additional
facilities, including the city waste incinerator, a
large steel mill, and two utility power plants,
are located in neighborhoods located close to the Partnership area. More than 175 chemicals are
emitted from the facilities in and around the Partnership neighborhoods, such as volatile organic
chemicals (VOCs), metals, and others. About 30,000 people reside in the five Partnership
neighborhoods of Cherry Hill, Brooklyn, Brooklyn Park, Curtis Bay, and Wagners Point.
Community Key Issues of Concern
• Air Quality
• Trash, Illegal Dumping, and
Abandoned Housing
• Economic Development
• Parks and Surface Water
Quality
• Community Health
-------
The Partnership Air Committee and Goals
Air quality ranked first in the list of concerns voted on at the community's priority-setting
meeting. The high interest in air quality was an indication of widespread community concern for
the health of residents living in the Partnership neighborhoods and for the possible contribution of
outdoor air pollution to the community's health. Community residents were particularly
concerned about the possible consequences of the combined emissions from all the industrial,
commercial, and waste treatment and disposal facilities located in and around their neighborhoods
(See a full description of the pollution sources covered in the Emissions Inventory section on
page 19.). In response to community concerns, the CEP Air Committee focused its work
primarily on the contribution of air emissions from these types of point and area sources to
outdoor air. Although some of the government partners raised the issue of indoor air quality as a
topic for consideration, the community did not choose to make this a priority for the CEP Air
Committee at that time.
To meet community concerns,
the Air Committee set two overall goals Goals of Air Committee
for its work: (1) to determine if current
levels of toxics in the air in Partnership
neighborhoods resulting from the Levels of Toxics in the Air Resulting from
multiple industrial, commercial, and
waste treatment and disposal facilities in
and around the Partnership area may
To Determine if the Current Aggregate
the Multiple Industrial, Commercial, and
Waste Treatment and Disposal Facilities in
and Around the Partnership Area May
affect community health; and (2) to Affect Community Health
recommend actions to improve
community air quality. The Committee ' To Recommend Actions To Improve
focused on voluntary participation and Community Air Quality
voluntary action on the part of its
members, with the aim of going beyond
regulatory requirements where possible. Compliance, enforcement, and regulatory reform were
not the focus for the Air Committee's work. The Committee's work was also done with the
view of building the long-term capability of the community to understand and address air quality
issues. (See additional discussion of the Air Committee's goals on page 14.)
Air Committee Meetings and Work
The Air Committee held its first meeting in September 1996, and it has continued to meet
monthly since that time. Air Committee meetings have consistently drawn around 20 participants.
Representation of different sectors of the community on the committee was fairly balanced for
most of the Committee's work. Co-chairs, one industry representative and one resident, were
elected to lead the Committee work. Four or five residents, one representative of an
environmental organization, a faculty member of a local university, six or seven government
-------
representatives, and six or seven business representatives attended the meetings regularly. (See list
of Committee participants in Appendix A.) All Committee decisions were made by consensus,
and special efforts were made to provide background information and education to ensure full
participation by all Committee members. Facilitators were not used to help with the meetings.
(During the summer of 1998, after the completion of the screening project but before the release
of the results to the public, most community residents and the representative of the environmental
organization left this committee. After their departure, the Committee relied on the Partnership's
Executive Committee for community representation and direction to complete its work. (See the
John Snow Institute case study for a more detailed discussion of the community participation in
the Partnership. See also the letters exchanged between the environmental organizations and the
Partnership in the summer of 1998 in Appendix B.)
The Baltimore Air Committee began its work by discussing air concerns, conducting an
odor survey, and reviewing a report on local TRI releases. The Committee also invited a dioxin
expert from EPA to give a presentation. After several months of background preparation, the
Committee discussed and agreed on a method to conduct the air screening methodology described
in the following pages.
Overview of the Community Air Screening Methodology
As the Air Committee began its work, it soon
recognized that to answer community questions about air
quality, it would need to find a way to evaluate more than
175 chemicals emitted to the air by more than 125
facilities in or around the Partnership neighborhoods. To
complete this review, the Committee decided to develop a
risk-based screening methodology that could help to set
community priorities. The screening process uses
standard methods to provide information on the potential
health risks associated with chemicals in the air in
Partnership neighborhoods. These methods are consistent
with EPA's general guidance on conducting exposure and
risk assessments (U.S.EPA, 1989; 1992a). The screening
process also builds on established procedures for
modeling human health risks from air pollutants, such as A
Tiered Modeling Approach for Assessing the Risks Due
to Sources of Hazardous Air Pollutants (U.S. EPA,
1992b). Using a risk-based approach helps to identify
those chemicals that may pose the greatest risks by
considering the many factors that influence the potential
human health impacts from air pollution sources. For
example, the methodology considers factors such as the
type of chemical emitted, the quantity emitted, the
distance from source to receptors (residents), local wind
patterns, and the varying toxicity of the different chemicals.
Air Screening Methodology
• Step 1: Formed
Partnership, Clarified
Goals
Step 2: Built Source
Inventory Database
• Step 3: Conducted
Initial Screening
Step 4: Conducted
Secondary Screening
Step 5: Conducted Final
Screening
• Step 6: Developed
Public Report and
Recommendations
6
-------
The Air Committee did not begin its screening work with a completed plan for this
methodology in mind. Rather, the Air Committee developed the methodology in a step-by-step
fashion in response to the need at each stage of the work. This allowed the screening
methodology to be developed in a way that would enable participants to tailor it to the particular
needs of the study area. The Air Committee was able to exchange information with similar air
inventory and risk assessment projects under way in other EPA offices, such as the Chicago
Cumulative Risk Initiative (U.S. EPA, 1999b), the Cumulative Exposure Project (U.S. EPA,
1999c), and the Urban Air Toxics program (U.S. EPA, 1999d). These projects have similar goals
of trying to determine exposures and risks from hazardous air pollutants, but they differ from this
effort in scope. The methodology described in this report (for the Baltimore area) is still a work
in progress. Lessons learned from the experience and areas identified for improvement are
provided in the final section of the report. The development and the implementation of the
screening work drew on the resources of all Committee members.
To develop a practical screening methodology that could be implemented with limited
resources, the Committee used a multistep process. The initial screening used readily available
information and simple and protective risk calculations to review the 175 chemicals. In each
succeeding step, as the number of chemicals remaining in the screening process decreased, the
Committee was able to use more detailed information and more accurate and resource-intensive
methods of analysis.
Overall, the Committee's work in Baltimore can be divided into the six steps briefly
described below. (See Figure 2 at the end of the introductory section for a flow chart of the air
screening methodology.) A more detailed discussion of these steps as they were implemented in
Baltimore is provided in the remainder of this report.
• Step 1: Formed Partnership. Clarified Goals
Formed a broad Partnership committee with representatives from all sectors of the
community, including community residents, local businesses, organizations, schools
and universities, and local, State, and Federal government agencies. Clarified the
goals of the Partnership and developed a plan for work. Also developed an outreach
plan to facilitate communication with the community.
Step 2: Built Source Inventory Database
Created a community-specific inventory of industrial, commercial, and waste
treatment and disposal facility air pollution sources from information available from
sources such as emissions permits, compliance records, and the Toxics Release
Inventory (TRI). Collected ambient air monitoring data for toxics from stations
located in and around the Partnership neighborhoods. Entered these data into a
database to facilitate screening.
7
-------
• Step 3: Conducted Initial Screening
Screened the inventory using toxicity data and a protective calculation of exposure to
identify the chemicals needing further analysis.
Step 4: Conducted Secondary Screening
Used computer air dispersion modeling and local meteorological information to get a
better estimate of the ambient concentrations for the chemicals selected in the initial
screening (Step 3). Compared both modeling and monitoring results to health-based
screening values chosen by the Committee. Chemicals with neighborhood
concentrations higher than the screening values were identified for further analysis.
• Step 5: Conducted Final Screening
Contacted the facilities to obtain the most accurate information available on emissions
of the targeted chemicals and data pertinent to air dispersion modeling. Conducted
the air dispersion modeling again using the refined information. Compared the
resulting estimated airborne concentrations and/or monitored air concentrations to the
screening guidelines. Chemicals exceeding the Committee screening values were
identified as priorities for the community.
Step 6: Developed Public Report and Recommendations
Developed recommendations for improving air quality based on the results of the
screening exercise and developed a report communicating the results and the
recommendations to the community.
Understanding What the Baltimore Risk-Based Screening Effort Could and Could Not
Accomplish
It is important to note up front what the results of the screening analysis could and could
not provide to the community. The screening provided valuable information to the community
and did accomplish the following:
The analysis did identify and inventory all the significant commercial, industrial, and
waste treatment and disposal sources of toxics in outdoor air in and around the
Partnership neighborhoods.
It did provide the best estimates available on the types and amounts of toxics in
outdoor air in Partnership neighborhoods from facility sources, including estimates of
the aggregate concentrations of the same chemical from multiple sources.
-------
• It did compare estimated and measured concentrations to health values and provide
enough risk information to help the community set priorities and chart an effective
course of action for improving air quality.
It did help to establish a community air quality baseline that can be used to evaluate
future progress and to identify potential concerns with new sources.
• It did allow the Partnership to compare the levels in its neighborhoods to other urban
neighborhoods where concentrations of the same chemical have been measured.
• The collaborative work did help to build consensus in the community on air issues
and it did provide education and information to build community capacity to
understand and address air quality issues in the long term.
The information provided from the screening analysis had definite limitations. The
screening could not accomplish the following:
• Most importantly, the analysis could not establish the cause of current instances of
diseases in the community. Chronic illnesses related to environmental causes may be,
in part, due to exposures that occurred many years in the past. This analysis assessed
current air quality and, consequently, it provided information on illnesses that could
possibly occur in the future, not illnesses that are a result of past exposures. Also,
many non-environmental factors that may influence community health were not a part
of this analysis. These include factors such as diet, smoking, access to medical care,
lifestyle, and genetics. All of these contributing factors need to be considered to
effectively address community health concerns.
• Except in some limited cases, this type of screening analysis could not provide
information on the possible effects of the combined mixture of different chemicals in
the air. The science to understand the effects of mixtures of a large number of
chemicals does not currently exist.
• The actual risk from these chemicals in each Partnership neighborhood could not be
determined. This is because (1) much of the screening is based on estimates and not
on actual measurement, (2) actual measurements were taken only in a limited number
of locations in the Partnership neighborhoods, and (3) a study was not conducted for
people living in the Partnership neighborhoods to accurately determine exposures. A
detailed exposure study would consider, for example, time spent in the neighborhood,
age, time spent outdoors, etc. To collect all information necessary for a more
detailed risk analysis would cost more and take longer, and the Committee decided
the additional information might not have contributed significantly to the
community's ability to set priorities for improving air quality.
-------
The analysis could not provide a complete and comprehensive screen of the hazards
associated with the 175 chemicals contained in the Baltimore inventory. Sixty-three
of the 175 chemicals did not have readily available toxicity information and could not
be included in the analysis. In addition, the toxicity information that was available
may be incomplete. New testing, such as the testing for effects on children and for
effects on endocrine systems, may identify additional hazards not considered in this
analysis. Given the limits of toxicity information currently available, the Baltimore
study is a review of known hazards, not all hazards.
The analysis could not provide a complete picture of all aspects of air quality. Three
aspects of air quality that may have significant chronic health effects were not a part
of this study: ground level ozone, which is a by-product of the reaction of certain
chemicals with sunlight; small particulate matter, especially from diesel exhaust; and
short-term peak concentrations of certain chemicals that may contribute to health
problems such as asthma. The Air Committee has recommended further work in
these areas to evaluate their potential effects on the community.
Summary Flow Chart
Figure 2 contains a flow chart with a detailed outline for the six steps of the screening
process. The details for each step are explained in the remaining sections of this report. See also
the flow chart (Figure 7 on page 73) with modifications based on lessons learned from the
Baltimore Case Study.
How to Contact tbeEPA/OPPT
Community Assistance Team
Community Assistance Team :
US Environmental Protection Agency
Office of Pollution Prevention and Toxics
Ariel Rios Building (7406)
1200 Pennsylvania Avenue, N.W,
Washington, DC 20460
Telephone: 20:2-260-6750
10
-------
Step 1 -
Build Partnership
Form Partnership, Clarify Goals,
Develop Community Outreach Plan
Step 2 -
Source/Emission/Monitoring
Inventory
of readily available information.
Toxic Release Inventory, State
Permit & Monitoring Data
Build Facility Source and
Monitoring Data Inventory
Step 3 -
Initial Screen
of readily available information
and conservative scenario.
Derive Concentration Using
Generic Turner Method or Use Monitored
Concentrations
Toxicological Information Input
Estimate Health Risk Values
Screening Criteria Input
Screen
Results
X Chemicals
X Facilities
Step 4 -
Secondary Screen
Modeling air dispersion with
readily available information.
Meteorological Data Input
Location Data Input
ISCST Dispersion Modeling Used to
Derive Neighborhood Air Concentrations
Ambient Monitoring Data
Committee Screening
Values Input
Screen Concentrations with Health
Based Screening Values
Results
X Chemicals
X Facilities
Step 5 -
Final Screen
Modeling air dispersion with
best available information.
Facility Specific More
Accurate Information Input
ISCST Dispersion Modeling Used to
Derive Neighborhood Air Concentrations
Ambient Monitoring Data
Committee Screening
Values Input
Screen Concentrations with Health
Based Screening Values
Results
Priority Chemicals & Facilities
Step 6 -
Recommendations and
Communication
Develop Pollution Prevention and
Risk Management Recommendations
and Communicate Results
Figure 2. Flow Diagram for Air Screening Methodology
II
-------
-------
BUILD PARTNERSHIP (STEP 1)
Stepl -
Build Partnership
Form Partnership, Clarify Goals,
Develop Community Outreach Plan
This section describes the first step of the air screening exercise carried out by the Air
Committee of the Community Environmental Partnership (CEP) in the Baltimore area. Step One
included building the Partnership to carry out the work, establishing Committee goals, and
developing an outreach plan to communicate the Committee's work to the community.
Formed Partnership
The effort to build a working partnership to
address the air quality and other environmental
concerns of the neighborhoods of southern
Baltimore and northern Anne Arundel County
began in the spring of 1995. To form the
Community Environmental Partnership, staff and
managers from the EPA Office of Pollution
Prevention and Toxics (OPPT) met with
neighborhood residents, schools, churches,
businesses, and local political representatives,
including the neighborhood congressional
representatives, as well as leaders and staff from the
city and county, Maryland Department of the
Environment, and EPA Region 3. More than 20
preliminary meetings and briefings were held to
explain and consult on the goals and plans for the
proposed Partnership. Many of these meetings
focused on the question of the potential effect of
the Partnership's work on already established
efforts to address the concerns of the Partnership
neighborhoods. To facilitate government
coordination and cooperation in the Partnership, the
government partners met biweekly for the first 2
years of the project.
Before the Partnership was established, a local group of businesses and the Baltimore
Development Corporation both expressed special concerns. The businesses, represented by the
Participants in Community
Environmental Partnership
Neighborhood Residents
Neighborhood Organizations,
Schools, and Churches
Neighborhood Businesses
Local Political Representatives,
Including Congressional
Representatives
Area Colleges and Universities
All Levels of Government: City,
County, Maryland Department
of the Environment, and EPA
Region 3 and OPPT
-------
local Chemical Industry Council, were concerned that a new partnership might upset or duplicate
the already established industry and neighborhood relationship. The Baltimore Development
Corporation believed that the Partnership might interfere with the city's key brownfields
redevelopment plans for an area within the Partnership boundaries. After these concerns were
addressed, a consensus acceptable to all the participants was built around a sustainable
development approach that considered jobs and a healthy environment. As a part of the
consensus-building process, the Partnership visited 50 area businesses to introduce the project and
to solicit their support. The positive response to these visits set the stage for the consensus
needed for the Partnership. Following a year of discussions and as a culmination of the efforts to
reach agreement, all of the partners met in the office of Baltimore's mayor on May 3, 1996, to
officially launch the Community Environmental Partnership.
As described in the Introduction, the CEP
organized five committees to address community
priorities. Air quality was identified as the top
priority. More than 60 people from all sectors of
the community and all levels of the government
participated actively in the work of these
committees. Small companies and retail businesses
worked primarily on the Economic Development
Committee. Students and teachers from local
schools and the largest number of residents worked
on the Parks and Surface Water Quality and the
Trash committees. These committees focused
primarily on a project that identified an urban
stream flowing into a cove on the Bay shoreline and
initiated efforts to restore the stream and use the
cove as a community wildlife preserve.
Partnership Priorities
Air Quality
Community Health
Trash, Illegal Dumping, and
Abandoned Housing
Parks and Surface Water
Quality
Economic Development
The Air Committee, with approximately 20 participating members, was the largest
committee in the Partnership. The Air Committee met monthly from September 1996 through
completion of this report to the community. (A list of regular Air Committee members can be
found in Appendix A.) In addition to its regular members, meetings usually drew several
additional participants interested in or asked to address a current agenda item.
Clarified Air Committee Goals
The goals for the Air Committee were established in response to the community's
concerns for air quality as expressed at the opening public meeting or by the community residents
and business members of the Air Committee. The concerns expressed included the following:
-------
• The possibility that the cumulative and aggregate effects of the chemical emissions
from the concentration of industrial, commercial, and waste treatment and disposal
facilities in and around the Partnership area may be contributing to poor community
health, including the possibility that not enough attention was paid in the permitting
process to the possible cumulative effects of emissions from multiple facilities. This
concern about the possible contribution of emission sources to community health was
the main concern expressed by the residents working in the Air Committee.
• The possibility that disease incidence in Partnership neighborhoods, especially the
incidence of cancer, are higher than other areas of the city and county. The Health
Committee, a separate Partnership committee, was organized to investigate this
concern.
• The possibility that unreported emissions may exceed permit levels, especially during
weekends or at night. This concern was exacerbated by the frequent occurrence of
strong and unidentified odors in some Partnership neighborhoods.
• The possible disproportionate number of waste treatment and disposal facilities sited
in the Partnership neighborhoods. Both residents and businesses felt that the
reputation and the livabiliry of the community were adversely affected by the large
number of waste treatment and disposal facilities. The location of a regional medical
waste treatment facility in the Partnership area was a special concern to some
residents.
All of these concerns about community health and siting issues were heightened by
Baltimore City's plans to focus its brownfields redevelopment efforts on a part of the Partnership
area. While the city's plans to attract environmentally responsible companies for an eco-industrial
park allayed some concerns, residents and businesses still had concerns about the location of new
facilities in the area because the cumulative effects of existing facilities had not been adequately
characterized.
In response to these community concerns, the Air Committee decided to focus on the
main concern and adopted the following goals:
• To determine if the current aggregate levels of toxics in the air resulting from the
multiple industrial, commercial, and waste treatment and disposal facilities in and
around the Partnership area may affect community health, and
15
-------
• To recommend actions to improve
community air quality.
The Air Committee's choice of goals
narrowed the range of community concerns that
would be addressed by the Committee. The
Committee concluded that siting of waste
treatment and disposal facilities was a local land
use issues and not an appropriate issue for the
Partnership Committee. In addition, the
Committee's focus on potential exposures to
toxic chemicals emitted by industrial, commercial
and waste treatment and disposal facilities meant
that the Committee did not fully consider other
types of potential exposures, including:
Clarified Air Committee Goals
• To Determine if the Current
Aggregate Levels of Toxics in
the Air Resulting from the
Multiple Industrial,
Commercial, and Waste
Treatment and Disposal
Facilities in and Around the
Partnership Area May Affect
Community Health
To Recommend Actions To
Improve Community Air
Quality
Exposure to toxics from mobile
sources, including particulate matter
emissions from diesel truck traffic. (The Committee did analyze data from the State
ambient air monitoring station located in the Partnership area. The monitored levels
represented the aggregate concentrations from all sources including mobile sources.
To help explain these monitored concentrations, the Committee estimated the
contribution of mobile sources to the level of toxics in outdoor air. This estimation is
described in Step 5);
• Exposures to toxics in indoor air; and
• Short-term and peak exposures that might produce acute effects.
The Committee's scope of work also did not include the consideration of additional
factors, other than outdoor air toxics, that might affect community health, such as diet, access to
medical care, exposure to lead paint, etc. The narrowing of the Committee's focus to an
examination of facility emissions and their potential to affect community health was a conscious
Committee choice. The choice was made to respond to the concern of some Committee members
who felt that the inclusion of other sources of toxics would distract attention from the industrial,
commercial, and waste treatment and disposal facility sources that they believed were the main
community concern. In general, the Committee accepted this approach and decided that its work
would have more credibility if it spoke directly to the main community concern.
The limited scope of the Air Committee's investigation eventually produced a dilemma
The Committee wanted to focus on the facility sources and to develop concrete recommendations
to improve community health. However, the limited focus meant that when the Committee
completed its work, it might not be in a position to identify the most effective actions to improve
-------
community health. This would be the case if a source of air pollution not included in the study,
such as mobile sources, turned out to be a significant source for the community. In fact, when the
results of the limited analysis of exposure to facility sources found that these sources were not
likely to be a significant contributing factor to community chronic health concerns, the Committee
did not have enough information about the other sources to develop the most effective
recommendations. (See recommendations in the Air Committee Report, Appendix J.) The
possible contradiction between the limited scope of the Committee's work and its ability to make
recommendations for the improvement of community air quality and health was not adequately
discussed, understood, and agreed to at the beginning of the work.
Developed Plan for Community Outreach
The Coordinating Committee of the CEP asked each of its committees to take 6 months to
collect information, develop recommendations, and report back to the Partnership. While the
Committee's work proceeded, the Partnership continued its overall outreach efforts. In May
1997, the Partnership opened a community storefront office with environmental information,
Internet access, and meeting space. Baltimore Mayor Schmoke joined the community for the
opening celebration. In addition, a monthly newsletter describing the progress of the various
committees and Partnership was established and sent to more than 300 community members.
During its first year, the Partnership organized community cleanups, several educational
presentations, and a major Earth Day event.
On April 30, 1997, the Partnership organized its second large public meeting to give the
five committees the opportunity to present their findings and recommendations to the community.
The Air Committee was not able to complete its screening exercise in time for this meeting;
therefore, the Committee presented preliminary results and committed to make a full report at a
later date. The screening exercise was completed by the end of 1997. The Air Committee then
focused on the production of a report (Appendix J) that could adequately explain and summarize
its work for the community. (See discussion of the preparation of this report in Step 6.) Plans
are now being made to organize meetings in the community to present and discuss this report.
17
-------
-------
EMISSIONS INVENTORY (STEP 2)
Step 2 -
Source/Emission/Monitoring
Inventory of readily available
information.
Toxic Release Inventory, State
Permit & Monitoring Data
Build Facility Source and
Monitoring Data Inventory
Overview
To begin its screening analysis, the Air
Committee first collected all readily available
information relating to local air quality, including
information on facility releases and information on
concentrations of toxic air pollutants measured at
local monitoring stations. The emissions inventory
was conducted over a 2-month period by a technical
subgroup of the Air Committee, which consisted of
representatives from EPA and the Maryland
Department of the Environment (MDE). The
inventory recorded information on the amounts of
chemicals emitted into the air each year by facilities
in and around the Partnership area. EPA and MDE
reviewed the inventory periodically for
completeness and accuracy. The subgroup used a
computerized spreadsheet to compile and manage
the extensive information.
The emissions inventory included a wide variety of industrial, commercial, and waste
treatment and disposal sources of air pollution,2 ranging from small sources such as gas stations,
Emissions and Monitoring Inventory
• Inventory of Emissions Data
from 125 Facilities in Area
Both Emissions Data and
Ambient Air Monitoring Data
Collected from MDE and U.S.
EPA
• Data Organized and Managed
Using Spreadsheet That Was
Used for All Screening Steps
2 Using the terminology of the Clean Air Act, both point (major stationary) and area (small stationary)
sources were included in the emissions inventory for this project. Under the Clean Air Act, "point" or "major
stationary" sources are stationary facilities that emit a regulated air pollutant in an amount exceeding the threshold
level -- 100 or 250 tons per year - depending on the pollutant and type of facility. Typical major stationary sources
include large industrial complexes like power plants, chemical plants, oil refineries, and steel mills. "Area" sources
are smaller stationary sources of pollution that are not inventoried individually but whose emissions are estimated as
a group and reported as a single source category for a geographic area. Examples of "area sources" include gas
stations, dry cleaners, consumer use of solvents, and gas furnaces, fireplaces, and woodstoves which are typically
associated with homes and nonindustrial sources.
In the third step of the Baltimore project, air dispersion modeling was used to estimate airborne
concentrations of chemicals of concern emitted from industrial, commercial, and waste treatment and disposal
facilities. To avoid confusion over terminology, please note that the Clear Air Act definitions for point and area
(continued...)
19
-------
w
/ith annual emissions to air of less than 100 pounds of chemicals, to large facilities with annual
emissions of over 1 million pounds. As discussed in Step 1, mobile sources of air pollution such
as vehicles, small engines (e.g., lawn mowers and other lawn equipment), combustion products
from furnaces, fireplaces, and grills, and ozone and other pollutants emitted or formed in other
regions and transported long distances to the Partnership area were not covered in the inventory.
Table 1 presents a summary of the types of sources included (and not included) in the inventory
for the Baltimore Case Study.
Once the decision was made on the types of facilities to include in the inventory, attention
shifted to the process of building the source inventory and finding databases of emissions and
monitoring data. This effort benefitted from the knowledge and experience of other EPA
programs, such as the Cumulative Exposure Project and the Urban Air Toxics program (U.S.
EPA, 1999c and d). as well as on electronic databases maintained by the EPA and MDE
(described later in Step 2.)
Sources for Identifying Facilities
A number of information sources were
used to identify specific facilities in the
Partnership area and to quantify the annual
emissions of individual chemicals. ZIP Codes
21225 and 21226 defined the Partnership area.
Information was also gathered for facilities
located within 5 miles of the Partnership area
(ZIP Codes 21060, 21061, 21090, 21122, 21219,
21222, 21227, 21230). As a starting point, the
subgroup included all businesses operating in the
Partnership area that were listed by Dun &
Bradstreet (D&B). Each business was listed by
South Baltimore Partnership
Area/Neighborhoods
ZIP Codes 21225, 21226
Neighborhoods
- Cherry Hill
Brooklyn/Brooklyn Park
Curtis Bay
Wagners Point
2(... continued)
sources described above are not the same as those commonly used in air dispersion modeling. In air dispersion
modeling, the terms point and area source have a meaning not related to the amount of the emissions. Point
sources have an exact emission site, such as an exhaust stack and they can be both large and small. Area sources,
in contrast, cannot be associated with an exact emission site. Area source emissions may come, for example, from
evaporation over a large area or from leakage from small multiple locations such as valves. In air dispersion
modeling, sources, both large and small, with emissions dispersed across the site are called area sources.
Emissions from these sites are modeled as though they were uniformly emitted from the entire area covered by the
site. Under the Clean Air Act, all small sources are called area sources regardless of whether their emissions come
from an exact point or are dispersed across a site. Thus a small business with an exhaust stack is an area source
under the Clean Air Act and a point source in the terminology of air dispersion modeling. Similarly a large source
with dispersed emissions, such as a waste treatment facility, would be called a point source under the Clear Air Act
and an area source for purposes of air dispersion modeling. Understanding the different use of these terms will be
helpful when air dispersion modeling is discussed in step three of the screening methodology.
20
-------
Table 1. Sources Included and Not Included in the Inventory
for the Baltimore Case Study
CAA Category
Included in Baltimore
Inventory
Not Included in
Baltimore Inventory
Point (major stationary)
Examples: chemical plants,
power plants, incinerators,
landfills, steel mills, POTWs
Area (small stationary)
(a) Commercial and industrial
chemical use and handling
Examples: dry cleaners,
gasoline stations, print shops
(b) Commercial, industrial,
institutional boilers
Examples: schools, hospitals,
office building heating
(c) Household heating and
chemical use
Examples: furnaces, fireplaces,
lawn chemicals
Mobile Sources
(a) On road
Examples: cars, trucks, buses
(b) Off road
Examples: portable generators,
construction equipment, boats,
lawn mower
21
-------
name, type of business, address, telephone number, number of employees, and standard industrial
classification (SIC) code (U.S. EPA, 1997a). The subgroup compared this list against a list of
facilities permitted by the State of Maryland to emit chemicals to the air, provided by the MDE
Air and Radiation Management Administration (ARMA). The list of permitted facilities and the
EPA Toxics Release Inventory (TRI) were used to make a master list of facilities that might emit
chemicals into the air. The list was then reviewed by:
Partnership members, including residents familiar with businesses operating in their
neighborhoods;
Chemical engineers familiar with the types of businesses and activities that emit
chemicals to the air; and
MDE staff who were aware of the facilities no longer in operation or whose
permits had changed.
Once the final list of facilities operating in and around the Partnership area was obtained,
emissions data in pounds per year (Ib/yr) were collected and entered in the inventory database. A
list of 125 potential facilities was created in this step.
Sources Used To Collect Emissions and Ambient Air Monitoring Data
A variety of database sources were used in compiling the inventory for southern
Baltimore. Various government agencies, at local, State, and Federal levels, maintain these
databases as part of their compliance monitoring systems. The Air Committee accessed pertinent
data sources to obtain data on emissions and concentrations of chemicals in ambient air. The data
sources from MDE and EPA are described below. Appendix F contains examples and
information on accessing these data
sources on the Internet.
Maryland Department of the
Environment. Air and Radiation
Management A dministration
Registered Stationary' Source
Emissions
Registered source data were
provided by MDE. Facilities that are
major sources of volatile organic
compounds (VOCs). sulfur oxides
(SOJ. and nitrogen oxides (NOJ and
facilities with permits to operate were
Emission Inventory Databases
Dun & Bradstreet List of Businesses
Maryland Department of the Environment
Registered Stationary Source Emissions
Toxic Air Pollutant (TAP) Emissions
EPA
- Toxics Release Inventory (TRI)
Facility Index System (FINDS)
Aerometric Information Retrieval
System Facility Subsystem (AIRS/AFS)
-------
included in the registered sources emissions inventory. These facilities are required to provide
annual emissions data for selected chemicals to the MDE. The chemical emissions reported
include certain air toxics and criteria air pollutants (e.g., SOX, NOX, particulate matter).
MDE data were later identified by the type of emission (e.g., stack—controlled emissions
through an elevated exhaust stack; or fugitive—uncontrolled emissions from leaks and
evaporation often near the ground) by EPA and MDE's Air and Radiation Management
Administration. These emissions data were entered into a computerized spreadsheet for easier
organization and use.
Toxic Air Pollutant (TAP) Emissions
MDE provided TAP emissions data for the most recent year available (1995). MDE
collects these data because the State of Maryland developed air toxics regulations for emissions of
TAPs not addressed by national or State ambient air quality standards. Carcinogens are "Class I
TAPs," and other toxics are "Class II TAPs." Regulations are applicable to any source required
to have an air quality permit that discharges a TAP. New construction sources may be required to
report TAP emissions, and the source must provide a statement every year that certifies current
compliance. A list of TAP chemicals and an example of TAP emissions data for the Partnership
area ZIP Codes are included in Appendix F.
Ambient Air Monitoring Data
MDE operates an air monitoring
network throughout the State in
accordance with EPA guidelines to
measure the concentrations of criteria
pollutants and selected air toxics in the
ambient air. This ambient air monitoring
data could be used to represent the
concentrations of chemicals in the air that
the neighborhood residents breathe. One
monitoring station is located in the
Partnership area (Fairfield monitoring
station). This area is a predominantly
industrial zone with significant emissions
from chemical manufacturing and
petrochemical storage facilities. Five
other monitoring sites are located in the
Baltimore area (Glen Burnie, Downtown Baltimore, Fort McHenry, Essex, and Northeast
Baltimore). The Fairfield monitor, as well as other monitors, are positioned so as to provide
readings suitable for estimating exposure over a larger geographic area.
Ambient Air Monitoring Data
Maryland Department of the Environment
Ambient Air Monitoring Data for 41
Chemicals from 1992 through 1996
Five Baltimore Area Monitoring Stations
Fairfield Monitoring Station Located in
Partnership Area
Use of 1996 Average Concentrations for
Risk Screening
23
-------
Data from 1992 to 1996 for the 41 chemicals monitored, along with their Chemical
Abstract Registr}' (CAS) numbers and details of the MDE monitoring program, are presented in
Appendix F. The ambient air monitoring data from the five Baltimore area monitoring stations
were compared to the monitoring station in the Partnership area to determine if Partnership area
concentrations were significantly higher than other areas around Baltimore. A comparison of the
monitored concentrations at the five monitoring sites in the Baltimore area for 1996 is provided
in Appendix J.
The trends in air pollutant concentrations for most of the monitored pollutants were
steady or downward between 1992 and 1996. In order to use information most relevant to the
current levels of chemicals in the air, the Air Committee decided to use monitoring data from
1996 in the screening exercise. Furthermore, annual average concentrations were used for
screening because use of maximum values would have probably been too conservative since they
were not typical of air quality.
U.S. Environmental Protection Agency (EPA)
Toxics Release Inventory (TRI) Data
EPA collects multimedia chemical
release data from selected manufacturing and
waste management facilities in the United
States (U.S. EPA, 1997b). Certain types of
businesses are required to report to EPA on
the use and release of about 650 toxic
chemicals. The data are compiled in the TRI
and are publicly available for use by
communities to identify those facilities that
release toxic chemicals into the air, water,
and other media. Air emissions data,
representing both stack and fugitive air
emissions estimates, were retrieved from TRI
for ZIP Codes 21225 and 21226 (U.S. EPA,
1997b). TRI data for 1994 through 1996
were used when State data were not available.
An example of TRI data is shown in
Appendix F
EPA Websites for Air Emissions Data
EPA's Envirofacts Database provides access to
several EPA databases that provide users with
information about environmental releases to air
in the United States. Data sources used for this
project included:
• Envirofacts Database:
http://www.epa.gov/enviro/index_java.html
• Toxics Release Inventory (TRI):
http://www.epa.gov/enviro/html/tris
• AIRS/AFS:
http://www.epa.gov/enviro/html/air.html
Aerometric Information Retrieval System Facility Subsystem (AIRS/AFS) Data
The Aerometric Information Retrieval System Facility Subsystem (AIRS/AFS) contains
emissions and compliance data on air pollution point sources regulated by the EPA and/or State
24
-------
and local air regulatory agencies under the Clean Air Act. AIRS/AFS contains data on industrial
facilities, power plants, and similar sources. In general, emissions data are provided for criteria
air pollutants (sulfur oxides, nitrogen oxides, particulate matter, carbon monoxide, volatile
organic compounds, and lead) and select hazardous air pollutants. Data available for the
screening typically represented emissions from inventories conducted in 1995 (U.S. EPA,
1997c).
Database Management
In order to effectively store, manage, and use the data collected, a spreadsheet was
created using readily available commercial software (Lotus and Excel). Each database record
consisted of a chemical, a facility, an annual emission rate (from the MDE TAP data and from
TRI), and other information necessary to calculate exposures and risks. The records also
included information such as latitude and longitude of the facilities, emission type (stack or
fugitive), stack description, cancer slope factors, and reference doses. (An example of the
columns of the database is provided in Appendix G.) Data entry was performed by a number of
individuals working on the technical subgroup, and entries were generally checked for errors at
least three times. Source documents (hard copy) from which the data were extracted were also
maintained as backup for the electronic files.
25
-------
26
-------
INITIAL SCREEN (STEP 3)
Step 3-
Initial Screen
of readily available information
and conservative scenario.
lexicological Information Input
Screening Criteria Input
Derive Concentration Using
Generic Turner Method or Use
Monitored Concentrations
Estimate Health Risk Values
Screen
Results
X Chemicals
X Facilities
Overview
This section describes the application of
the initial screening step to emissions sources and
monitored concentrations of toxic air pollutants
in southern Baltimore. With information on 175
chemicals and 125 facilities assembled in the
source inventory database, the Air Committee
needed to develop a method to identify which
chemicals, if any, might be of concern to the
community.
To begin the screening process, the
Committee first needed a method to estimate the
ambient air concentrations in Partnership
neighborhoods that resulted from all the
emissions reported in the source inventor}'. With
the large number of chemicals and facilities
needing review, the Committee decided that
using computer air dispersion modeling to estimate concentrations at this step would require
considerable resources. Instead, for the initial screen, the Committee used a simple and
protective calculation, described in detail below, to estimate air concentrations. The Committee
also included ambient air concentrations measured at the area monitoring station in this initial
screen. These estimated and measured ambient air concentrations (concentrations were estimated
for 175 chemicals, which included monitored concentrations for 41) were then used to develop
Initial Screen
Use of Source Emissions Data
and Generic Turner Method to
Predict Air Concentrations
Estimate of Potential Risk from
Inhalation of Chemicals at
Predicted Ambient
Concentrations
Comparison of Estimated Risk
Against Screening Values
Identification of 29 Chemicals
of Concern
27
-------
very protective estimates of potential risks to human health from inhalation of these chemicals.
The risk estimates for each chemical were then compared to a human health risk-based screening
value chosen by the Committee. Any chemicals with risk estimates above the Committee
screening value were identified as being of potential concern and were kept in the process for
further review. Chemicals with risk estimates below the Committee screening value were
eliminated from the screening process. A total of 29 of the 175 chemicals were identified from
the initial screen for further review. Details of the process and results are described below.
The Committee designed this initial screening step using conservative assumptions about
exposure (i.e., assumptions that tend to overestimate exposure) to make sure that any chemicals
that might be of concern were identified for further review. Using conservative assumptions also
meant that the Committee could not assume that the chemicals flagged in the initial screen
presented a significant risk to the community. More accurate and realistic information was
needed to further evaluate potential risks from exposures to these chemicals; therefore, these
chemicals were "promoted" to secondary screening where more accurate exposure concentration
estimates were developed for risk screening.
Initial Screen Procedures
To complete the initial screen, the Committee entered all information needed to get a
screening-level estimate of exposure and risk into the source inventory database. Toxicity
information on each chemical was collected and added to the source emissions and monitoring
information collected in Step 1. Formulas for calculating conservative ambient air concentrations,
exposures, and risks for each chemical were then added to the database, and the calculations were
made. The Committee then chose a screening value and compared the calculated risk to the
screening value to identify the chemicals that needed further evaluation.3
Input from all the Committee members was used to complete this step. Citizens and local
businesses verified the accuracy of the source inventory as it was entered into the database. The
full Committee participated in the discussion and decision on the choice of screening values and
recommended chemicals for further evaluation that were of significant concern to the community
for reasons other than the exposure or risk calculations. Government partners provided the air
emissions information. Technical experts on the committee assisted with collection of toxicity
information, management of the database, and calculations of exposure and risk for the
Committee's review. The database was designed to be located in the community so that it could
be viewed and updated annually (or more often if warranted).
-' If risk-based concentrations (RBCs) for the chemicals are available, the estimated air concentrations can
be compared directly to the RBCs. This eliminates the need for the risk calculations. For the first step of the
screening, the Air Committee did not have access to RBCs. For subsequent steps of the process, described in Steps
4 and 5. the Committee used EPA Region 3 RBCs (U.S. EPA, 1997d).
28
-------
Technical aspects of the initial screening were designed and carried out in a series of
meetings by a subgroup of the Partnership Air Committee. Technical staff from the Johns
Hopkins School of Public Health, industry, EPA, and MDE formed the technical subgroup. The
subgroup included technical staff with expertise in toxicology, exposure modeling, and risk
analysis. EPA provided information on the toxicity of the chemicals in the source inventory, and
other Committee members reviewed the information. Building on the source inventory database
developed in Step 2, EPA staff added the exposure and risk calculations to the spreadsheet for the
screening exercise.
The technical subgroup then held a screening meeting, where chemicals were either
dropped or selected for further review. For this review, the spreadsheet was used to sort the
inventory by chemical, risk, and quantity emitted. Using this information, Committee members
agreed by consensus which chemicals to eliminate from
further review and which to move forward for more
detailed review in the secondary screening. The actual
decision meeting lasted more than 5 hours. Although the
meeting was open to all Committee members, community
residents did not attend this screening meeting. Residents
reviewed the process and results at the next full Air
Committee meeting.
Background information on risk screening and
additional information on the dose and risk calculations
used in the initial screening step are provided below. In
this risk screening methodology, the initial screening step
developed by the Committee is divided into five separate
substeps: (1) collection of toxicity information, (2)
estimation of ambient concentrations and potential doses,
(3) calculation of cancer risk estimates and hazard
quotients, (4) selection of screening values, and (5)
comparison of calculated risks and hazard quotients to
screening values. These steps are described below
following presentation of information on risk screening.
Background Information on Risk Screening
The screening method used by the Air Committee
follows the basic risk assessment paradigm developed by
the National Research Council (1983):
1. Hazard identification is the process of
determining whether exposure to a
Risks and Hazards
How estimates of hazard and risk
are expressed depends on the nature
of the hazard and the types of data
upon which the assessment is based.
For example, cancer risks are most
often expressed as the increased
probability of developing cancer for
an individual exposed to the
chemical in question (i.e., 1 in
1,000,000 or 10'6). Risk estimates
for adverse effects other than cancer
are usually expressed as the ratio of
an estimated dose or exposure level
to a toxicologic potency value. This
is known as a hazard quotient. A
key distinction between cancer and
other toxicologic effects is that most
carcinogens are assumed to have no
dose threshold (i.e., no dose or
exposure level can be presumed to
be without some risk). Other
toxicologic effects are generally
assumed to have a dose threshold
(i.e., a dose or exposure level below
which adverse effects are not
expected). But there are exceptions.
For example, some carcinogens
have thresholds.
29
-------
3.
chemical can cause an adverse health effect and whether the adverse health effect
is likely to occur in humans.
Dose-response assessment is the process of defining the relationship between the
dose of a chemical received and the incidence of adverse health effects in the
exposed population. From the quantitative dose-response relationship, toxicity
values are derived that are used in the risk characterization step to estimate the
likelihood of adverse effects occurring in humans at different exposure levels.
Exposure assessment identifies populations exposed to a chemical, describes their
composition and size, and presents the types, magnitudes, frequencies, and
durations of exposure to the chemical.
4. Risk characterization integrates hazard and exposure information into quantitative
and qualitative expressions of risk. A risk characterization includes a description
of the assumptions, scientific judgments, and uncertainties embodied in the
assessment.
Cancer Risk Assessment
Assessment of cancer risks wr.s conducted in a manner that was consistent with EPA's
cancer assessment guidelines (U.S. EPA, 1996) and guidance documents such as Risk Assessment
Guidelines for Superfund (U.S. EPA, 1989). The National Toxicology Program publishes the
Annual Report on Carcinogens (DHHS, 1994) mandated by the Public Health Service Act. This
report lists chemicals "known to be carcinogenic" and chemicals "which may reasonably be
anticipated to be carcinogens." Research and regulatory organizations typically employ a
"weight-of-evidence" approach to determine the likelihood that a chemical is a human carcinogen.
Each chemical evaluated is placed into defined weight-of-evidence categories. For example, EPA
(1997e) classifies carcinogens by the five categories listed below4.
• Group A — human carcinogen
• Group B — probable human carcinogen (Bl indicates limited human evidence;
B2 indicates sufficient evidence in animals and inadequate or no
evidence in humans)
• Group C — possible human carcinogen
• Group D — not classifiable as to human carcinogenicity
• Group E— evidence of noncarcinogeniciry for humans
4 EPA's guidelines for cancer risk assessment are currently undergoing revision. The proposed guidelines
recommend significant changes in the way weight-of-evidence and potency determinations are conducted and
expressed. The proposed guidelines emphasize the importance of evaluating the mode of action in the assessment
of potential carcinogens.
30
-------
The International Agency for Research on Cancer (IARC) uses a similar classification
scheme:
Group 1 — carcinogenic to humans;
Group 2A — probably carcinogenic to humans;
Group 2B — possibly carcinogenic to humans;
Group 3 — not classifiable as to carcinogenicity; and
Group 4 — probably not carcinogenic to humans.
When the available data are sufficient for quantification, estimates of a chemical's
carcinogenic potency can be developed. For example, EPA "slope factors" express carcinogenic
potency in terms of the estimated upper-bound incremental lifetime risk per milligram per
kilogram (mg/kg) average daily dose (U.S. EPA, 1997e). Cancer slope factors (CSFs) are
available, where applicable, for both oral (SF0) and inhalation (SF,) exposures. "Unit risk" is a
similar measure of cancer potency for air or drinking water concentrations and is expressed as
risk per microgram per cubic meter (/^g/m3) in air or as risk per microgram per liter (//g/L) in
water for continuous lifetime exposures.5 The term "upper bound" in this context means that the
measures of cancer potency are high-end estimates, so they will be conservative. This may result
in an overestimate of cancer risk when toxicity data are incomplete, which is usually the case.
The upper-bound value is intended to be protective of human health for continuous lifetime
exposures, even though cancer risk may be overestimated. The use of the average or lower limit
values would be more likely to underestimate cancer risk.
Cancer risk is calculated by multiplying the estimated dose by the appropriate measure of
carcinogenic potency, the cancer slope factor. For example, an individual with a lifetime average
daily dose of 0.03 mg/kg-day of a carcinogen with cancer slope factor of 0.02 (mg/kg-day)''
would experience an increased lifetime cancer risk of 0.0006 (also expressed as 6 * 10"4 or 6E-
04) from exposure to that chemical. Similarly, cancer risk could be calculated using an air
concentration multiplied by the unit risk factor. In general, risks from exposures to more than
one carcinogen are assumed to be additive, unless information on interactions points toward a
different interpretation.
Risk Assessment for Other Chronic Health Effects
Because adverse effects other than cancer and gene mutations are generally assumed to
have a dose or exposure threshold, a different approach is needed to evaluate toxicologic potency
and risk for these "systemic effects." The approach for assessing noncancer effects was
consistent with EPA's guidelines (U.S. EPA, 1989). "Systemic toxicity" means an adverse effect
" Slope factors and unit risks are appropriate measures of carcinogenic potency when the dose-response is
thought to be linear. The new proposed guidelines include extensive discussion on the use of a margin-of-exposure
or RfD approach for carcinogens in which there is evidence of a nonlinear dose-response or a dose threshold for the
carcinogenic response.
31
-------
on any organ system following absorption and distribution of a toxicant to a site in the body
distant from the toxicant's entry point. A measure of toxicologic potency for chronic (long-term)
effects is the "reference dose" or "reference concentration." The reference dose (RfD) is defined
as "an estimate (with uncertainty spanning perhaps an order of magnitude) of a daily exposure to
the human population (including sensitive subgroups) that is likely to be without appreciable risk
of deleterious effects during a lifetime" and is expressed as a mg/kg-day dose (U.S. EPA, 1997e).
The reference concentration (RfC) is an estimate (with uncertainty spanning perhaps an order of
magnitude) of a daily inhalation exposure of the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of deleterious noncancer effects during
a lifetime. Conversion of RfCs to the more traditional RfDs is straightforward using a 20 m3/day
inhalation rate and a 70-kg body weight (U.S. EPA, 1997f). RfD values for inhalation were
derived from RfCs and are used in this study. The RfD is usually based on the most sensitive
known effect (i.e., the effect that occurs at the lowest dose) and can exist for both oral exposures
(RfD0) and continuous inhalation exposures (RfD,).6 Although some RfDs are based on actual
human data, they are most often calculated from results obtained in chronic or subchronic animal
studies. The basic approach for deriving an RfD involves determining a "no-observed-adverse-
effect level (NOAEL)" or "lowest-observed-adverse-effect level (LOAEL)" from an appropriate
toxicologic or epidemiologic study and then applying various uncertainty factors and modifying
factors to arrive at the RfD. Uncertainty factors are used to derive RfDs and RfCs to account for
factors that may alter toxicity. In the absence of sufficient toxicity data to assess risk, the
objective is to ensure that estimates are protective of human health, including sensitive
subgroups, rather than underestimating the toxicity of chemicals that may pose health risks.
Evaluating risks from chronic exposures to systemic toxicants can be performed using
either an RfD or an RfC. An expression of risk called a "hazard quotient" (HQ) is the ratio of the
estimated chronic dose to the RfD. Similarly, an HQ can also be calculated as the ratio of the air
concentration divided by the RfC. An HQ of greater than 1 would raise a concern. Hazard
quotient values below one imply that adverse effects are very unlikely to occur. The greater the
extent to which exceeds one, the greater the level of concern. However, it is important to
remember that the hazard quotient is not a probabilistic statement of risk (i.e., an HQ of 0.001
does not mean that there is a one-in-a-thousand chance of the effect occurring). Furthermore, it
is important to remember that the level of concern does not necessarily increase linearly as the
quotient approaches or exceeds unity because the RfD or RfC does not provide any information
about the shape of the dose-response curve.
6 The inhalation reference dose (RfD,) was used in this case study for evaluating the systemic toxicity of
chemicals. A reference concentration (RfC) is another way of expressing the toxicologic potency of a chemical
when the exposure is via inhalation.
-------
Collection of Toxicity Information
To generate the screening-level risk
tes for the chemicals emitted in the P<
area, the Committee collected toxicity information
,. . f ,, , . , ... , . ., n _^ ,. Sources of Toxicity Data
estimates for the chemicals emitted m the Partnership J
• ,, f f , f . r • . - Integrated Risk Information
in the form of cancer slope factors for carcinogenic „
effects and in the form of inhalation reference doses
for other chronic effects. EPA staff collected this TT ,. _-- „ „,,
• r. .• .r i r-. • j j • • • Health Effects Summary Tables
information for the Committee and entered it into J
the source inventory database. EPA's Integrated
Risk Information System (IRIS) was chosen as the
primary source of toxicity information because of
its availability and because of the level of scientific review of the assessments contained in IRIS
(U.S. EPA, 1997e). It should be noted, however, that IRIS does not always reflect the most
recent data and assessment on a chemical. In the absence of toxicity data for a chemical from
IRIS, a secondary source for data used in the assessment was the Health Effects Assessment
Summary Tables (HEAST). HEAST (U.S. EPA, 1997f) summarizes published toxicity data and
provides estimates of toxicologic potency, but the data in HEAST are not subjected to the same
degree of review as those in IRIS. Each source of toxicity data is described in more detail in
Appendix D. It should be noted that for the risk calculations, RfDs and slope factors were used.
These values were derived from the RfCs and unit risk factors contained in IRIS and HEAST.
Conversion of RfCs and unit risks to the more traditional RfDs and slope factors is
straightforward using a 20 nrVday inhalation rate and a 70 kg body weight. Toxicity information
for more than 115 of the 175 chemicals was available from these sources. This included 28
chemicals with cancer slope factors and 93 that had RfDs, of which 57 were based on the
inhalation pathway. This meant that many, but not all, chemicals could be assessed as part of the
screening process.
Calculation of the Air Concentration and Potential Dose
For the initial screen, the Air Committee used the generic Turner method, a standard EPA
procedure, to estimate the annual average air concentration and potential dose rate (PDR) for
each chemical in the source inventory (Turner, 1994). The generic Turner method was chosen
because it is based on a well-known and widely accepted approach in the scientific arena for
estimating concentrations of air pollutants emitted from near-ground point sources, and the results
can be easily used in a computer spreadsheet. (A description of the generic Turner method is in
the text box on the next page.) The initial screen addressed only inhalation exposures to the
general population that may result from air emissions from the facilities included in the source
inventory. Additional sources and pathways were not addressed in this or subsequent steps of the
screening exercise. It should be recognized that persons may also be exposed to certain of the
studied chemicals from other sources (e.g., household products, and other pathways such as
ingestion of contaminated food, soil, or water).
-------
Calculating Ambient Air Concentration
The following is an example of the
use of the generic Turner method in the
Baltimore screening analysis to calculate the
ambient concentration. The TRI-reported
emissions of cadmium in 1994 from
SCM/Millennium Specialty Chemicals was
estimated as 4 Ib/yr. The ambient
concentration was estimated as:
ID'1
The Generic Turner Method
Turner's (1994) sector-averaging form of the Gaussian
algorithm can be used to estimate ambient air
concentrations that could result from point source
emissions. With certain assumptions, a multiple-term
equation describing how a chemical released to the air
is dispersed and diluted downwind from its source can
be simplified as a conversion factor. The assumptions
used are as follows:
• A pollutant release height of 3 meters;
• A person exposed 100 meters from the source;
• Neutral atmospheric stability;
An average wind speed of 5.5 meters per
second;
• A continuous release of the chemical; and
• The wind blowing in one direction 25 percent
of the time.
Using these assumptions, the ambient air
concentrations in units of mg/m3 can be estimated by
multiplying the annual air release (Q) of a chemical in
units of kg/yr by a conversion factor. The procedure of
deriving this conversion factor (4.88 * 10"6) is provided
in Appendix E.
Concentration (mg/m3) = Q (kg/yr) * (4.88 * 10'6)
This conversion factor (4.88 * 10"6) can be
incorporated into a computer spreadsheet program that
estimates ambient concentrations for all near-ground
releases of interest. The ambient air concentration can
be compared to an inhalation risk-based concentration.
The user also has the option of converting ambient air
concentration to annual exposure. The exposure is
used to calculate potential dose and ultimately risk.
Concentration (mg/m3) = 4 Ib/yr *
(lkg/2.21b)* (4.88*10-6) = 8.87 *
mg/m3
Calculating Dose
A very conservative estimate of potential
dose was calculated assuming a distance of
100 meters from the source, an inhalation
rate of 1 m3/hour, an exposure time of 24
hours per day, an exposure frequency of 365
days per year for a lifetime of 70 years.
These assumptions generally do not
represent realistic activity patterns;
therefore, the potential dose rate estimated
is very likely higher than ones actually
expected. Additionally, the potential dose
represents an estimate of the total quantity
of the chemical available for absorption by
the specified route, in this case inhalation.
The actual absorbed dose can differ
significantly from the potential dose,
depending on chemical-specific
pharmacokinetic and metabolic factors. Using the assumptions above, the ambient concentrations
calculated by the generic Turner method were converted to the annual exposure as follows:
Annual exposure (mg/yr) = Q (kg/yr) * 0.043
The procedure for deriving the conversion factor (0.043) for annual exposure shown in the above
equation is provided in Appendix E.
34
-------
A potential dose rate (PDR) for the exposed individual was estimated by dividing the
annual exposure by an average body weight of 70 kg and 365 days per year as follows:
PDR (mg/kg-day) = Annual exposure (mg/yr) / (70 kg * 365 days/yr)
For the previous example, the annual exposure and PDR were estimated as:
Annual exposure (mg/yr) = 4 Ib/yr * (lkg/2.21b) * 0.043 = 0.078 mg/yr
PDR = (0.078 mg/yr ) / (70 kg * 365 days/yr) = 3.1 * 10'6 mg/kg-day
Calculation of Cancer Risk Estimates and Hazard Quotients
Estimates of the cancer risks and hazard
quotients were made for emission sources in the
inventory. When available, cancer slope factors
for inhalation exposures were used in the
calculations. In the absence of cancer slope
factors based on inhalation exposures, oral slope
factors were used in the risk calculations. For the
non-cancer assessments, RfC values were
converted to RfDs based on EPA-approved
procedures (U.S. EPA, 1997 f). Use of an
estimated dose and the associated RfD was
preferred because the risk assessors needed to
evaluate risks for many types of scenarios. RfCs
incorporate exposure assumptions and can only
be used for one exposure route. As a result,
RfCs were converted to RfDs and inhalation
doses were calculated for the scenario being
assessed (see Region 3 RBC table in
Appendix D). In turn, the same estimated doses
could be used in the cancer risk calculation by
combining it with the cancer slope factor. In a
few instances, inhalation cancer slope factors
were not available and slope factors based on the
oral route were used. In those cases, another
uncertainty was introduced to the assessment. It
cannot be assumed that oral and inhalation
exposures, even at equivalent dosage rates, will
result in the same toxicologic response.
Cancer Risk Estimates
These cancer risk calculations were
performed using a cancer slope factor
and a dose estimated from the
inhalation exposure pathway. Another
way of calculating cancer risks is to use
an approach that uses the unit risk
factor and the air concentration. The
resulting estimated risks from either
approach would be the same as long as
the same exposure assumptions are
used (e.g., inhalation rate and body
weight). Future case studies that
implement this methodology will likely
use the unit risk factor approach. The
unit risk factor (available from IRIS) is
expressed in units such as l/(mg/m3),
so multiplication of the unit risk by a
given air concentration (in mg/m3) will
yield a cancer risk. This equation
assumes the exposure is over a lifetime
(70 years). The "lifetime of exposure"
includes assumptions of 20 mVday
inhalation rate, 24 hr/day, 365 days/yr,
70 years exposure duration (equal to a
lifetime), and an adult body weight of
70kg.
35
-------
An example of the risk calculations for the emissions of cadmium from the SCM facility is
shown below:
Cancer risk = Cancer slope factor * Potential dose 6.3 = (mg/kg-day)'' * 3.1 * 10'6 mg/kg-day =1.95*
io-5
Hazard quotient = Potential dose/RfD = 3.1 * 10'6 mg/kg-day / 5 * 10'4 mg/kg-day = 6.2 x If/3
Source Inventory Database
All toxicity values, exposure estimates, and risk calculations used by the Air Committee in
the initial screening were incorporated into the source inventory database. An excerpt from the
database for several chemicals is provided below as an illustration of the major database fields that
were used in the risk calculations. The example shows how risk and hazard estimates are made
for single sources of specific air pollutants using the Turner method.
Pollutant Name
Acetonitrile
Ammonia
Benzene
Carbon tetrachloride
Ethvlbenzene
Hydrochloric acid
Toluene
Inhalation
Cancer Slope
Factor
(mg/kg-day)"'
0.029
0.0525
Inhalation
Reference
Dose (RfD)
mg/kg-day
0.0143
0.0286
0.00171
0.000571
0.286
0.00571
0.114
Maximum
Total Air
Emissions
(Ibs/yr)
4,370
290,000
7,156*
2,820
1,772.8
707,808
262.99
Potential Dose
(mg/kg-day)
(based on
Turner)
3.34e-03
2.21e-01
5.46e-03
2.15e-03
1.35e-03
5.40e-01
2.01e-04
Risk
(dose*SF)
(based on
Turner)
O.OOE+000
O.OOE+000
1.58E-004
1.13E-004
O.OOE+000
O.OOE+000
O.OOE+000
Hazard
(dose/RfD)
(based on
Turner)
2.33e-01
7.74e+00
3.19e+00
3.77e+00
4.73e-03
9.46e+01
1.76e-03
Note: This benzene example is from the Baltimore composting facility. Those releases turned out to be inaccurate,
as described in the subsequent screening steps (see Table 4).
Selection of Screening Values
To set screening values, the Air
Committee chose a risk level of 1 in
1.000.000 (lO'6) for chemicals causing
cancer and a hazard quotient greater than
1 (HQ>1) for chemicals with other
chronic effects. While this was a
consensus decision, there was
considerable discussion on the choice of
screening values. Because the State of
Man-land uses a 10° risk level for the permitting of facilities that have carcinogenic air emissions
Committee members were concerned that the choice of a more stringent screening value might be
Air Committee's Risk Screening Values
• IO-6 for Cancer Risk
• Hazard Quotient >1 for Other Chronic
Effects
36
-------
misinterpreted as a critique of the Maryland standards. Despite this concern, the Committee
decided that the goals of the Committee justified the use of screening values that differ from the
Maryland standards. The Committee decided to stay with the more stringent risk values for
several reasons. The Committee designed the screening exercise to identify priority areas for
voluntary pollution prevention, not to identify permit violations. The Committee recognized
Maryland's concern for misinterpreting of the screening values and decided to make special
efforts to clearly communicate the nonregulatory purposes of the screening exercise. The
Committee also felt that the 10"6 screening value for cancer risk would identify those chemicals
that should be considered in the siting of new facilities. Overall, the Committee chose the 10"6
screening value as part of its effort to design a screening exercise that would err on the side of
extra protection. Using a stringent screening value would help to ensure that any chemicals that
might be of concern to the community would be identified and that chemicals not identified for
further review would be unlikely to present a significant risk to the community.
Comparison of Cancer Risk Estimates and Hazard Quotients to Screening Values
Cancer risks and hazard quotients were calculated for all chemicals emitted in the
Partnership area using the generic Turner method. Ambient air monitoring data were also used in
the initial screen to determine if air concentrations might result in risks that exceeded the
screening levels. Data from the MDE air monitoring station located in Fairfield, north of the
FMC facility, were available for 1992 through 1996. This is the only air monitoring station
located in the Partnership neighborhoods that gathers information on air pollutants. This
monitoring station takes air samples every day and data are available on the annual average,
minimum, and maximum concentrations for 41 toxic chemicals. Of the 41 chemicals monitored, 4
had annual average concentrations in 1996 that resulted in risks that exceeded the Committee
screening values (benzene, 1,3-butadiene, carbon tetrachloride, and methyl chloride). The
Committee next sorted the cancer risk and hazard quotient columns of the database in descending
order to identify the chemicals emitted from the facilities that had cancer risks greater than 10"6
and/or a HQ of >1. To satisfy cancer risk and hazard quotient screening criteria, a chemical had
to exceed the criteria for at least one facility.
The risk screening criteria were exceeded for 25 chemicals in the inventory. These are
listed below:
37
-------
1. Formaldehyde
2. *Aldrin
3. Methyl chloride
4. Benzene
5. Methylene chloride
6. *Acrylamide
7. Cadmium & compounds
8. *Perchloroethylene
9. *Trichloromethane
10. *Trichloroethylene
11. Arsenic & compounds
12. Chromium & compounds
13. Vinyl chloride
14. 1,3-Butadiene
15. Carbon tetrachloride
16. *Ethylene oxide
17. Dioxins & furans
18. Toluene
19. Hydrochloric acid
20. Manganese & compounds
21. Ammonia
22. Hydrogen sulfide
23. *Chlorine dioxide
24. 1,2-Dichloropropane
25. Mercury
* Chemicals with an "*" were not selected for the
chemicals were no longer emitted from the facility
no longer in operation.7
next stage of the screening process for reasons such as the
because of changes in the production process or the facility was
A formal attempt to calculate aggregate exposure from multiple sources was not made in
the initial screening process. The risk screening values of 10"6 for cancer and HQ > 1 for other
effects were used to screen only individual sources. Although the Committee did not develop a
formal procedure for calculating aggregate exposures in the initial screening, it informally
reviewed the risk calculations to see if combining the emissions of individual chemicals from
multiple sources could potentially result in additional chemicals exceeding the screening criteria.
This review was performed by sorting the database by chemical so that all the risk calculations for
each chemical could be viewed at once. If a chemical had no individual facility exceedances of
> 10"6 for cancer risk or HQ >1 for other effects, but would possibly exceed those criteria when
combining the emissions from multiple sources, it would have been selected for further analysis.
However, this informal screening for aggregate exposures did not result in any new concerns.
(See the lessons learned section for a recommendation regarding the development of a more
formal method for screening for aggregate exposures in the initial screening step.)
In addition to the cancer risk and hazard quotient screening criteria, the Committee used
other screening criteria to select chemicals for further review. Several chemicals were chosen for
inclusion because they had very high emission quantities. These chemicals were as follows:
• Sulfur oxides (SO J
• Nitrogen oxides (NO J
• Carbon monoxide (CO)
• Carbonyl sulfide
Xvlenes
A lesson learned for this stage of the screening was the need to keep detailed records of the decisions
made and the reasons for the decisions. This will make it easier to present a more complete summary of the initial
screening step.
-------
Total emission rates in the Partnership area for sulfur oxides, nitrogen oxides, carbon
monoxide, and carbonyl sulfide were greater than 1 million Ib/yr each. Because the total emission
rate for xylenes was relatively high (> 400,000 Ib/yr), it was also selected.
The Committee also used professional judgment to select additional chemicals for review.
This was especially important for those chemicals for which there was no toxicity information
available at the time of the screening exercise.8 Committee members used their diverse
backgrounds and experience in the fields of exposure assessment, toxicology, risk assessment, and
regulation of air emissions to make these judgments. The following chemicals were included
using these more subjective criteria:
• Hydrogen fluoride
• Lead
Nickel
• Stoddard solvent
• Sulfuric acid
• Molybdenum trioxide
The results of the initial screen, including the chemicals of concern and their basis for
selection, are provided in Table 2. With the inventory of chemicals now reduced, the Committee
proceeded to look more carefully at the remaining 29 chemicals. Details of the analysis for the 29
chemicals in the next step of the process (the secondary screen) are provided in the following
chapter.
8 Toxicity data are not available for all chemicals and for all health effects. Such data may not be
available because the chemicals have not been tested and because consensus has not been reached on the toxicity
value. This risk screening exercise was performed in 1997 based on available data at the time. The toxicity data
contained in IRIS and HEAST are regularly updated. However, additional chemicals would not have been
identified from the initial screen even if more current toxicity data were used.
39
-------
Table 2. Chemicals Selected from Initial Screen
Chemical Name
Ammonia
Arsenic
Benzene
1,3-Butadiene
Cadmium
Carbon monoxide
Carbon tetrachloride
Carbonyl sulfide
Chromium compounds (III, VI)
1 ,2-Dichloropropane
Dioxin (2,3,7, 8-TCDD)
Formaldehyde
Hydrochloric acid
Hydrogen fluoride
Hydrogen sulfide
Lead
Manganese
Mercury
Methyl chloride
Methylene chloride
Molvbdenum tnoxide
Nickel
Nitrogen oxides
CAS Number
7664-41-7
7440-38-2
71-43-2
106-99-0
7440-43-9
630-08-0
56-23-5
463-58-1
7440-47-3
78-87-5
1746-01-6
50-00-0
7647-01-0
7664-39-3
7783-06-4
15347-57-6
7439-96-5
7439.97-6
74-87-3
75-09-2
1313-27-5
7440-02-0
1 1 104-93-1
No. of
Facilities
14
5
23
1
3
51
4
1
10
1
3
5
20
1
4
3
7
2
2
7
1
7
79
Basis for Selection3
HQ> 1
Cancer risk estimate > 10"6
Cancer risk estimate >10"6/Monitoring
data
Cancer risk estimate > 10'VMonitoring
data
Cancer risk estimate > 10"6
Over 1,000,000 in total emissions
Cancer risk estimate > 10'6 and HQ > 1
/Monitoring data
Over 1,000,000 in total emissions
Cancer risk estimate > 10'" and HQ > 1
Cancer risk estimate > 10"6
Cancer risk estimate > 10"6
Cancer risk estimate > 10"6
HQ> 1
General criteria: Toxicity concerns +
emission sources
HQ> 1
General criteria: Toxicity concerns +
emission sources
HQ> 1
HQ> 1
Cancer risk estimate > 10'6/Monitoring
data
Cancer risk estimate > 1 0"6
General criteria: Toxicity concerns +
emission sources
General criteria: Toxicity concerns +
emission sources
Emissions > 1 ,000,000 Ib/yr
40
-------
Table 2. Chemicals Selected from Initial Screen (Continued)
Chemical Name
Stoddard solvent
Sulfur oxides
Sulfuric acid
Toluene
Vinyl chloride
Xylene
CAS Number
8052-41-3
SEQ:lllb
7664-93-9
109-88-3
75-01-4
1330-20-7
No. of
Facilities
2
48
13
40
2
49
Basis for Selection3
General criteria: Toxicity concerns +
emission sources
Emissions > 1,000,000
General criteria: Toxicity concerns +
emission sources
HQ> 1
Cancer risk estimates > 1 0"6
Emissions (49 facilities with total
emissions of 433,000 Ib/yr)
a. All chemicals, except four, were selected on the basis of modeling air concentrations from emissions. The four
chemicals selected based on ambient air monitoring data were benzene, 1,3-butadiene, carbon tetrachloride,
and methyl chloride.
b. Chemical not registered by the Chemical Abstract Service. Sequence (SEQ) numbers are assigned arbitrarily.
41
-------
42
-------
SECONDARY SCREEN (STEP 4)
Step 4 -
Secondary Screen
Modeling air dispersion with
readily available information.
Meteorological Data Input
Location Data Input
Ambient Monitoring Data
Committee Screening
Values Input
ISCST Dispersion Modeling Used to
Derive Neighborhood Air
Concentrations
Screen Concentrations with Health
Based Screening Values
Results
X Chemicals
X Facilities
Overview
The 29 chemicals identified in the initial
screen were the starting point for the secondary
screen. Instead of using the Turner method to
estimate concentrations, computer air dispersion
modeling was used in this step to estimate
aggregate concentrations from the facility sources.
This air dispersion modeling provided a more
realistic estimate of exposures than the very
protective calculations used in the initial screen and
was consistent with the kind of tiered modeling
approach recommended in "A Tiered Modeling
Approach for Assessing the Risks Due to Sources
of Hazardous Air Pollutants" (U.S. EPA, 1992b).
New information required for the secondary
screening included facility location information and
local meteorological data. The Air Committee also
chose new screening values, described below, for
this and the final step of the screening exercise.
Chemicals that were identified in the initial screen
based on monitored concentrations were also kept
on the list of chemicals for further review in the
secondary screen.
Modeling efforts for the 29 chemicals were performed by the EPA technical staff using the
Industrial Source Complex Short-term Version 3 (ISCST3) model to estimate ambient
concentrations (U.S. EPA, 1995). This model takes into account emissions from point and area
sources and estimates the dispersion of chemicals in the ambient air by using local meteorology
Secondary Screen
29 Chemicals Identified from
the Initial Screen Used as the
Starting Point
Air Dispersion Model Used To
Estimate Ambient Air
Concentrations
New Committee Screening
Values set at 50 percent of
Risk-Based Concentrations
(RBCs) Developed by EPA
Region 3
Secondary Screen Identified 7
Chemicals of Concern from 23
Facilities
43
-------
data. The output from ISCST3 are estimates of hourly, monthly, and/or annual concentrations at
receptor locations. The time required for modeling depends on the number of sources and the
chemicals selected for modeling. The modeling for this project took several weeks to complete.
General modeling efforts included:
• Building a modeling input file using data from the source emissions inventory
database,
• Developing a grid system for the model,
• Locating facilities and neighborhoods in the modeling grid, and
• Running the model.
Appendix K provides background information on model setup, assumptions, and a
chronology of modeling runs with ISCST3. Modeling scenario 1 in Appendix K is the modeling
for the secondary screen. Scenario 2 represents an intermediate step that included more accurate
information on emissions. Scenario 3 incorporated additional information on the type of
chromium (Cr*3 or Cr*6) emitted by facilities, and added updated data on benzene emissions.
Scenario 4 was used to determine the contribution of individual facilities' benzene emissions to
the total modeled benzene concentration in Wagners Point.
Once the input was completed, estimates were generated of the chemical concentrations in
each neighborhood from all known releases of a chemical, along with estimates of the highest
concentrations modeled anywhere within the grid system. The estimated air concentrations were
compared to the screening values chosen by the Committee. Monitored concentrations were also
compared to the new screening values. For the secondary screening step, the Committee decided
to switch and use the Region 3 risk-based concentrations (RBCs) as the basis for its screening
values. The Region 3 RBCs were calculated to
correspond to a 1 in 1,000,000 (10'6) cancer risk
and/or an HQ of 1. The Committee decided to use
50% of the Region 3 RBCs as its screening value.
These were more protective values than the ones
used in the initial screen.
At this point in the process, the Committee
also decided to group chemicals that have similar
effects (e.g.. neurological effects and respiratory
tract irritants) to look at the possibility of
cumulative effects that might result from exposure
to combinations of different chemicals. Details of
this cumulative screening are discussed below.
Results of the secondary screen showed that
concentrations for 7 of the 29 chemicals were above
the Committee screening values in one or more
Secondary Screen Results
7 Chemicals Above Committee
Screening Values
Modeled Chemicals
Benzene
Chromium
Hydrochloric Acid
Manganese
Monitored Chemicals
Benzene
1,3-Butadiene
- Carbon Tetrachloride
Methyl Chloride
44
-------
Partnership neighborhoods. Four chemicals were identified by modeling, four chemicals were
identified by monitoring, and one chemical was selected by both. The Air Committee decided to
carry these seven chemicals to the final screen. The Committee did not communicate the results
of this step to the community at large. Although the Committee did not reach a consensus on the
communication of these results to the community, the Committee held several discussions on the
interpretation of the screening results. Several draft reports from this screening exercise were
prepared, but they were not approved for release to the community by the Committee.
Communication with the facilities that were not already members of the Partnership, but were
releasing chemicals with estimated concentrations at or above the screening values, was initiated
to encourage participation.
In addition to air modeling, the Air Committee focused on providing Committee members
with background information to ensure that each member understood the steps in the process and
could fully participate in the discussions and decisions. Therefore, the Committee organized a
special meeting devoted to explaining and discussing the basic science of the screening exercise,
as well as toxicology, exposure, risk, and modeling. Residents communicated their concerns
about facility emissions and explored whether air dispersion modeling could provide answers to
their questions. The Air Committee attempted to answer all the questions from the members (and
the Committee) to ensure confidence in the overall screening process.
Completing the Secondary Screen
The 29 chemicals selected from the initial screen were carried through the secondary
screen to determine if they were chemicals of concern.
Air Dispersion Modeling
Air dispersion modeling was conducted
using Version 3 (ISCST3) model. ISCST3 has
been tested, validated, and widely used by EPA and
State government organizations for risk assessment,
regulatory, and permitting purposes. This model
was selected for a variety of reasons, including its
ability to be tailored for local conditions and to
model chemical emissions from multiple sources
(U.S. EPA, 1987). ISCST3 was used to estimate
the ambient concentrations of chemicals emitted
from the wide variety of air pollution sources
associated with industrial activities in and around
the Partnership area. The results of the modeling
were used to determine which air pollution
stationary sources needed further characterization
and which could be screened out as not likely to
Air Dispersion Modeling for Step 4
- ISCST3 Model Used
• Emissions from Point and Area
Sources Considered
• Model Estimated Annual
Average Concentrations at
Receptor Locations
Receptor Sites Included:
Brooklyn/Brooklyn Park
Cherry Hill
Wagners Point
Curtis Bay
45
-------
have a significant impact on human health. For chemicals that were emitted by too many facilities
to feasibly model, enough facilities were chosen so that at least 95 percent of the total mass
emitted was captured. Professional judgment was used to verify that omitted facilities would not
affect the analysis (e.g., low quantities were emitted or facilities were not located near populated
areas).
Model Description
ISCST3 was designed to calculate ground-level average concentrations and/or total
deposition values emitted from single or multiple stationary sources (U.S. EPA, 1995). ISCST3
uses meteorological data and site-specific parameters (e.g., stack parameters and pollutant
emission rates) to calculate hourly, monthly, and/or annual average concentrations, as well as
deposition values. The calculations can be performed at each receptor (neighborhood) on a
coordinate grid for each source or for combined emissions from select groups or all sources.
For the purpose of ISCST3 modeling, stationary sources in the Partnership area were
divided into point and area sources,9 based on the characteristics of their emissions. Point sources
are generally associated with a specific point defined by the location on the emissions/receptor
coordinate grid. In the modeling exercise, point sources are generally exhaust stacks with a
defined height, diameter, and other associated variables. The emission rates entered into the
model for these types of sources were in units of mass per unit time (e.g., Ib/hr). Area sources in
the context of ISCST3 modeling are emissions that do not originate from a specific point, such as
a stack, but are emitted from an area of known width and length (e.g., evaporation from a
wastewater treatment plant or leaks from a fuel terminal). The emissions rates entered into the
model for these types of sources were in units of mass per area per unit time (e.g., pounds per
square foot per hour [Ib/ft2/hr]). The use of the term "area source" in this context should not be
confused with that of "area source" under the Clean Air Act (i.e., a stationary source of
hazardous air pollutants that is not a "major" source).
Model Setup and Assumptions Used
ISCST3 requires emissions data, meteorological data, and facility information as modeling
input. The emissions of each chemical and stack parameters for each facility studied were
identified from information provided by MDE. (Example shown in Appendix F.) In most cases,
maximum permitted emissions of each chemical for each facility were used as the emissions input
for the secondary screen. The characterization of each emission as stack or fugitive was made
based on professional judgment by an engineer familiar with most of the facilities. Both toxic and
criteria air pollutants were modeled using local meteorological data from the most current years
4 The terms point sources and area sources, when used in the context of dispersion modeling are different
than when used for defining types of sources based on the Clean Air Act (point, area, mobile sources) See
footnote number 2 for further discussion.
46
-------
available (1987-1988, 1990-1992). Generally, meteorological data over a 5-year span are used in
air dispersion modeling to account for temporal variations.
Data to characterize area sources were not available as part of the secondary screen.
Default assumptions based on the best engineering judgment were used as follows: small area
sources (such as gas stations) were assumed to be 50 x 50 meters and 3 meters emissions height.
Large area sources (such as large industrial facilities) were assumed to be 500 x 500 meters and
30 meters emissions height.
Receptor Grid and Model Outputs
ISCST3 was run using a Cartesian coordinate source and two receptor grids. The coarse
grid with 2,000 m spacing was 18,000 16,000 m, or about 110 square miles (Figure 3). This
coarse grid allowed for prediction of air concentrations for 72 receptor locations in a
110-square-mile area around the Partnership neighborhoods. The coarse grid was used in order
to reduce the number of computations when including facilities up to 5 miles away from the
Partnership area. Since no calculations outside of the Partnership neighborhoods were needed for
the distant emissions sources, but the coarse grid could still provide estimates within the
Partnership neighborhoods for these pollutants, use of the coarse grid was preferred over the
much more computationally demanding fine grid covering the same area. The fine grid, with 250
m grid spacing, was 5,000 meters on a side, or about 10 square miles. This fine grid provided
better resolution of the air concentrations (at 700 receptor locations) in the Partnership
neighborhoods (Figure 4).
Selection of Facilities Modeled
For the priority chemicals with multiple emission sources, a subset of 36 sources was
selected to reduce the number of facilities for air modeling. The focus was placed on those
facilities whose emissions accounted for at least 95 percent of the mass of total emissions. For
example, manganese was emitted by seven facilities, but only two facilities were modeled
(Chemetals and Bethlehem Steel) because they accounted for more than 95 percent of the total
mass of manganese emitted in the Partnership area. An additional selection criterion was used in
the case of benzene to cover the range of sources so that some small sources such as gas stations
were included along with the larger sources of emissions. The facilities selected for air modeling
for this stage of the analysis are listed in Appendix H.
Selection of Receptor Sites
ISCST3 was used to estimate ambient air pollutant concentrations for the 4 Partnership
neighborhoods, Cherry Hill, Brooklyn/Brooklyn Park, Curtis Bay, and Wagners Point. The
coordinates used for modeling corresponded with the approximate geographic centers of these
four communities. Recognizing that air pollutants may be transported from outside the
Partnership area, facilities within 5 miles of the Partnership area were included in the emissions
47
-------
Baltimow
Legend
" Discrete Receptors
CD Fine Gnd Points
•fa Emission Sources
« Coarse Grid Point
— County Boundary
X
Figure 3. Coarse Receptor Grid in Baltimore
48
-------
B Discrete Receptors
A Emission Sources
Streets
• Rne Receptors
- II 4.02
Figure 4. Fine Receptor Grid in Baltimore
49
-------
inventor}'. While this approach did not capture pollution transported from other regions of the
United States, it represents an exhaustive attempt to consider local commercial and industrial
stationary' sources (Figure 4).
Selection of Health-Based Screening Levels and Endpoints
For this stage of the screening process, the Committee used 50 percent of the RBCs
calculated by Region 3 as the screening values. The RBCs provided a concentration benchmark
to compare directly to the concentrations estimated by the modeling or measured at the
monitoring stations. The RBCs (U.S. EPA, 1997d) are the concentrations at which either the
cancer risk to an exposed population is 1 in 1,000.000 or the HQ is 1. If the monitored or
modeled concentrations exceeded 50 percent of the RBCs, then the chemical was identified as a
candidate for further analysis. The assumptions built into the RBCs are provided in Appendix D.
A review and potential adjustment of these assumptions was identified for future improvement of
the screening methodology to ensure the protection of sensitive populations.
Grouping of Chemicals with Similar Target Organs or Physiological Systems
The Air Committee reviewed the toxicology information for the 29 chemicals to screen for
possible cumulative effects from exposure to multiple chemicals and to identify chemicals with
similar target organs or physiological systems. On the basis of this review, chemicals with known
neurological effects and chemicals that act as respiratory tract irritants were grouped together.
Cumulative exposures resulting from the chemical groupings did not result in any new concerns.
The chemicals reviewed and results of the cumulative assessment can be seen in Table 1-2 in
Appendix I.
Chemicals with Monitoring Data
Data from a monitoring station located within the Partnership area were available for 4 of
the 29 chemicals: benzene. 1,3-butadiene, carbon tetrachloride, and methyl chloride. (See
Appendix F for an example of air toxics monitoring data.) These data were compared with the
screening concentrations to determine if the monitored levels were greater than the modeled levels
and/or the screening levels. All four chemicals were found at levels above the RBCs, so
additional study of on these chemicals was warranted as part of the final screen. For 1,3-
butadiene. evaluation was based only on monitoring data because there were no significant
stationary emission sources in the Partnership area to model.
50
-------
Results of Secondary Screening
The 29 chemicals selected in the initial screening step, including 18 by risk screening, 5 by
emission quantity, and 6 by professional judgment, were carried through the secondary screen for
further analysis. Monitoring data for any of these chemicals, if available, were examined to
determine whether monitored data or modeled data had higher concentrations. The data with
higher concentrations were compared against the risk screening values before performing the next
step. The estimated concentrations and monitored concentrations, as well as the corresponding
percentage of the screening value for each concentration, were presented in table format for
Committee review. This table is presented in Appendix I. A second table, indicating only
whether or not more information was needed, was also developed for Committee review. (See
Table 3.) Those chemicals having concentrations above the committee screening level were
identified as needing further analysis in the final screen step.
Chemicals Not Requiring Further Evaluation
Secondary Screening Chemicals
Seven of 29 Pollutants Exceeded 50 Percent of
Screening Value in One or More Neighborhoods.
For 22 of the 29 pollutants, estimated
concentrations from modeling or measured
concentrations from monitoring were well
below the Air Committee's screening criteria
in all the neighborhoods. Because of the low
concentrations, the Air Committee concluded
that no further evaluation was needed for the
22 chemicals.
Chemicals Recommended for Further
Evaluation
Concentrations for 7 of the 29
pollutants exceeded 50 percent of their
respective screening values in one or more of
the neighborhoods. Benzene, 1,3-butadiene,
carbon tetrachloride, and methyl chloride
were identified based on the monitored
concentrations. Benzene, chromium,
hydrochloric acid, and manganese were
identified by the modeled concentrations. (Benzene had both monitored and modeled
concentrations greater than 50 percent of its RBC.) The Air Committee recommended further
evaluation for each of these seven chemicals as part of the final screening step.
Benzene
1,3-Butadiene
Carbon Tetrachloride
Chromium
Hydrochloric Acid
Manganese
Methyl Chloride
from monitored and
modeled concentrations
from monitored
concentrations
from monitored
concentrations
from modeled
concentrations
from modeled
concentrations
from modeled
concentrations
from monitored
concentrations
51
-------
Interpretation and Communication of Results
The results of the screening exercise were presented to the Committee in different table
formats, and the advantages of each format were discussed. Draft reports interpreting these
results were also discussed in the Committee. At this point, the Committee did not reach a
consensus on a format for the presentation of the information to the community.
Table 3. Results of Secondary Screening for Target Toxics in Partnership Neighborhoods
Chemical
Ammonia
Arsenic*
Benzene*b
l,3-Butadiene*d
Cadmium*
Carbon monoxide
Carbon
tetrachloride*
Carbonyl sulfide
Chromium
(Hexavalent)*
Chromium
(Trivalent)
1,2-
Dichloropropane*'
Dioxin*
(2.3.7.8TCDD)
Formaldehyde*
Hydrochloric acid
Neighborhood Concentrations (from modeling)
Cherry Hill
Low3
Low
Low
--
Low
Low
Low
Low
Needs more
information
Low
Low
Low
Low
Low
Brooklyn/
Brooklyn Park
Low
Low
Low
--
Low
Low
Low
Low
Needs more
information
Low
Low
Low
Low
Low
Curtis Bay
Low
Low
Low
--
Low
Low
Low
Low
Needs more
information
Low
Low
Low
Low
Needs more
information
Wagners
Point
Low
Low
Needs more
information11
--
Low
Low
Low
Low
Needs more
information
Needs more
information
Low
Low
Low
Needs more
information
State-
Operated
Monitoring
Station Results
Needs more
information
Needs more
information
Needs more
information
52
-------
Table 3. Results of Secondary Screening for Target Toxics in Partnership Neighborhoods
(continued)
Chemical
Hydrogen fluoride
Hydrogen sulfide
Lead
Manganese
Mercury
Methyl chloride*
Methylene chloride*
Molybdenum trioxide
Nickel
Nitrogen oxides
Stoddard Solvent
Sulfur oxides
Sulfuric acid
Toluene
Vinyl chloride*
Xvlene
Neighborhood Concentrations (from modeling)
Cherry Hill
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Brooklyn/
Brooklyn Park
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Curtis Bay
Low
Low
Low
Needs more
information
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Wagners
Point
Low
Low
Low
Needs more
information
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
State-
Operated
Monitoring
Station Results
Needs more
information
a. Low concentrations from modeling; no further work was needed.
b. (*) refers to carcinogens.
c. Areas marked as "Needs more information" had modeled concentrations above 50 percent of the risk-based
concentration (RBC) chosen by the Partnership Air Committee. These chemicals were candidates for further
screening.
d. Modeling was not conducted because facility emissions were not available.
e. 1,2-Dichloropropane is a carcinogen via the oral route.
53
-------
54
-------
FINAL SCREEN (STEP 5)
Step 5-
Final Screen
Modeling air dispersion with
best available information.
Facility Specific More
Accurate Information Input
Ambient Monitoring Data
Committee Screening
Values Input
h,
w
ISCST Dispersion Modeling Used to
Denve Neighborhood Air
Concentrations
+
Screen Concentrations with Health
Based Screening Values
i
Results
Priority Chemicals & Facilities
Overview
The final screening step used the
most accurate information available to
better characterize annual emissions. This
refined information was used to identify
the chemicals and facilities of most
concern to the Partnership neighborhoods.
The final screen began with the seven
chemicals identified in the previous step.
The seven chemicals were emitted from 23
facilities and/or measured at the local
monitoring station. The final screen
identified four of the seven chemicals as
community priorities.
To collect the most accurate
information, members of the Partnership
Air Committee contacted representatives of
the 23 facilities or consulted MDE files to obtain annual emissions measurements or estimates
from the emissions compliance statements submitted to the State each year by permitted
facilities. When this information was not available, TRI emissions to air for the most recent year
(1996) were used. In addition, improved data were solicited on stack heights, facility location,
dimensions, and so forth, which resulted in a more accurate estimate of neighborhood
concentrations by the ISCST3 modeling. Additional information on the type of chromium
emitted to the air was also collected. Based on the new information, neighborhood
concentrations were re-estimated and compared to the Air Committee screening values. Any
chemicals with monitored or modeled concentrations above the screening values were identified
as priority chemicals for the community. Step 6. which follows, describes how the Committee
Final Screen
Started with 7 Chemicals from 23
Facilities
Used Refined Source Emission Data for
More Accurate Modeling
Used Ambient Air Monitoring Data for
Certain Chemicals
Result of Final Screen: 4 Chemicals
(Benzene, 1,3-Butadiene, Carbon
Tetrachloride, and Methyl Chloride)
55
-------
developed recommendations for addressing the priority chemicals and began work to
communicate the recommendations and results of the screening to the community.
Completing the Final Screen
Collection ofToxicity Information
Information on the toxicity of the remaining chemicals was found in the EPA Region 3
RBC table (U.S. EPA, 1997d). The risk screening was conducted as in the secondary screen.
The only new information needed at this step was toxicity information on the type of chromium
(trivalent or hexavalent) emitted from facilities in the Partnership area. Earlier screening steps
used a conservative assumption that all chromium emitted from the facilities to air was the more
toxic hexavalent chromium. The final screening was based on a more accurate estimate of the
form of chromium in the emissions.
Air Modeling
Four of the seven chemicals selected for the final screen (benzene, chromium,
hydrochloric acid, and manganese) had local facility sources. Air dispersion modeling was
conducted for these chemicals using the ISCST3 model, as in the previous step. Modeling
results were used to determine which facility sources should be candidates for voluntary
pollution prevention and emissions reductions.
Modeling Inputs and Assumptions
For the final screen modeling, emission rates, selection of facilities, and stack parameters
were refined with more accurate data. All other modeling inputs and assumptions remained the
same as in Step 4. Instead of using maximum state-permitted emissions, yearly air emissions
were obtained from the annual emissions compliance statements filed with MDE. This emission
information was derived from stack monitoring or engineering estimates and is based on the
expected performance of the facility. A comparison of the emission rates for the secondary and
final screens for the four modeled chemicals can be found in Table 4.
Selection of Facilities Modeled
Twenty-three facilities with emissions of the four targeted chemicals were selected for air
dispersion modeling in the final screen. The facilities, chemicals emitted, and emissions
amounts are listed in Table 4.
56
-------
Table 4. Emission Rates from Facilities Used in Secondary and Final Screens
Facility Name
Amerada Hess
Amoco Oil Co.
Amoco Station
Amoco Station
Baltimore Composting
Baltimore Resco
Bethlehem Steel
BGE Brandon Shores'
BGE Wagner Station'
Bavwav Terminal
Chemetals Corp.
Citgo Station
CONDEA-VistaChem.
Crown Station
Crown Station
Grace Davison
Med NetVMedX Inc.
MOTIVA (Mobil Oil-Maritank)
Norris Farm Landfill
Phoenix Services
MOTIVA (Shell Oil Terminal)
Shell Station
CITCO (Star Enterprises)
Stratus Petroleum
U.S. Gvpsum
Pollutant Name
Benzene
Benzene
Benzene
Benzene
Benzene
Chromium
Hydrochloric acid
Chromium
Manganese
Chromium
Hydrochloric acid
Chromium
Hydrochloric Acid
Benzene
Hydrochloric acid
Manganese
Benzene
Benzene
Hydrochloric acid
Benzene
Benzene
Chromium
Hydrochloric acid
Benzene
Benzene
Hydrochloric acid
Benzene
Benzene
Benzene
Benzene
Chromium
Secondary Screen
Emission Rate (Ib/yr)
NA
4.000
NA
NA
7,156
3.333
6.126,000
848
20,124
909
4.200,000
294
1,300,000
1,120
23,172
61.661
122
3.000
21,000
NA
NA
122
42.300
882
1.051
91.016
1.400
130
NA
NA
26
Final Screen Emission
Rate (Ib/vr)
652
80
66
67
7,156"
67 (+3); 3 (+6)
6,126,000
848 (+3)
20,124
633 (+3): 276 (+6)
4,200.000
204 (+3); 90 (+6)
1,300,000
220
8.758
16,300
61
2,200
12.000
62
44
122 (+3)
6.520
1,440
16
6.952
480
65
348
880
26 (+3)
NA Not available (several benzene sources were discussed as part of the final screen; no data were included in the secondary
screen).
a. Emissions data for I.3-butadiene, carbon tetrachloride, and methyl chloride were not included in this table,
since assessment of risk was based on monitoring data and not emissions from stationary source.
b. This number was determined to be erroneous; however, the emissions did not affect the Partnership
neighborhoods.
c. Estimates were based on design and operating parameters.
57
-------
Results of the Final Screen
Of the four chemicals modeled in the final screen, only benzene emissions were estimated to
result in airborne concentrations in a Partnership neighborhood at levels above the Committee
screening level. Table 5 displays estimated air concentrations of chemicals from the secondary and
final screens.
To help identify the contribution of each of the facility sources of benzene to the modeled
concentrations in the Wagners Point neighborhood, model runs were conducted in a manner such
that each benzene source was considered individually. The ISCST3 model was run repeatedly with
only one benzene source "turned on" at a time. This yielded an estimated maximum airborne
concentration due to the single emissions source under consideration. That value was compared to
the estimated concentration due to all sources to determine the contribution of that source
(percentage of the total). Petrochemical storage facilities in the Wagners Point area were identified
as the primary source of the modeled benzene concentrations.
In addition to determining the contribution of each source to the modeled concentration, the
Air Committee examined monitoring data for benzene in the Partnership area and a comparison of
the two values was performed to determine how closely the modeled concentration matched the
monitored concentration. The monitoring station in Fairfield is about Vz mile from the location of
the highest predicted concentration of benzene in Wagners Point. At this distance the two locations
could be unequally subject to influences, such as nearby benzene sources or differences in wind
direction and frequency, that could confound the comparison of benzene concentrations.
Nonetheless, if it is assumed that the modeling is accurate, then significant differences between
measured benzene concentrations and modeled benzene concentrations could be due to sources of
benzene not captured in the emissions inventory.
The results of this effort were used to develop the pie chart in Figure 5. The pie chart shows
the estimated individual contribution of each facility to the ambient benzene concentrations
measured at the monitoring station located approximately !/2 mile from the Wagners Point
neighborhood (Fairfield). This pie chart allows for a comparison of the modeled facility
contributions (12 percent) to other nonmodeled sources (88 percent). It is well known that mobile
sources make a significant contribution of benzene to urban air (U.S. EPA, 1999e). (Mobile sources
were not modeled by the Air Committee, but their inclusion in future efforts is highly
recommended.) On the basis of this information, the Air Committee concluded that a significant
portion of the unaccounted for benzene concentration monitored at the Fairfield station could be
attributed to mobile sources, likely benzene emitted from mobile sources (cars and trucks) passing
through the area on high-volume routes such as I- 695 and Patapsco Ave and at the 1-895 toll plaza.
A more precise determination of the sources of the measured benzene could not be made because
the Committee was unable to completely determine if all nonmobile sources had been accounted for
There may be additional unregistered local sources or other local sources not included in the
modeling. There may also be some transport of benzene into the Partnership area from beyond the
58
-------
Table 5. Estimated Air Concentrations of Chemicals from Secondary and Final Screens
Rcceplor
Cherry Hill
Wagners Point
Brooklyn
Curtis Bay
Concentration
.Averaging Time
Annual
Annual
Annual
Annual
Benzene
(A/g/m3)
Secondary
0.003
0.19
0.008
0.019
Final
00028
041
00078
0.014
Chromium & Compounds
Total
(A/g/m3)
Secondary
0.0001
0.0006
0.0004
0.0004
Final
NA
NA
NA
NA
Chromium (+3)
(^g/m3)
Secondary
NA
NA
NA
NA
Final
0 00008
0.00026
0.00011
0.00017
Chromium (+6)
(Mg/m3)
Secondary
NA
NA
NA
NA
Final
000001
000001
0.00001
0.00001
Hydrochloric Acid
(/jg/m3)
Secondary
1.5
84
1.5
3.7
Final
1.4
0.89
0.74
066
Manganese
(A/g/m3)
Secondary
0.014
0.054
0.024
0.039
Final
00044
0.016
00072
O.OII
A/g/nr1 = micrograms per cubic meter
NA = Not Applicable.
-------
Unaccounted For
88%
Figure 5. Comparison of Unknown to Stationary Sources of Benzene
Between the FMC Monitoring Station and Modeled Concentrations
60
-------
15-square-mile area considered in the study. It is also possible that the model may have
underestimated the contribution of the modeled facility sources.
Ambient air monitoring data from the monitoring station in Fairfield indicated the presence
of four chemicals (i.e., benzene, 1,3-butadiene, carbon tetrachloride, and methyl chloride) with
annual average concentrations greater than the Committee screening levels. With the exception of
benzene, no significant sources of these chemicals were listed in the emissions inventory. Benzene
is emitted from both stationary and mobile sources; 1,3-butadiene most likely originates from mobile
sources; carbon tetrachloride and methyl chloride are typically present in urban air at levels
monitored in the Partnership area. (See description in the Air Committee Report in Appendix J.)
The results from each screening
step are shown in Figure 6. Initially, the
inventory consisted of 175 chemicals.
As a result of the screening process,
four chemicals of concern were
identified, three from monitoring data
alone and one (benzene) from both
modeling and monitoring data.
Chemicals Identified in Final Screen
Monitoring
Benzene
Methyl Chloride
1,3-Butadiene
Carbon Tetrachloride
Modeling
• Benzene
61
-------
Source/Emission/Monitoring Inventory
175 Chemicals
125 Facilities
Initial Screen
29 Chemicals"
36 Facilities
Secondary Screen
7 Chemicals
23 Facilities
From Modeling
From Monitoring
Benzene
Chromium
Hydrochloric Acid
Manganese
Benzene
1,3-Butadiene
Carbon Tetrachloride
Methyl Chloride
Final Screen
4 Priority Chemicals
17 Facilities
From Modeling
From Monitoring
Benzene
Benzene
1,3-Butadiene
Carbon Tetrachloride
Methyl Chloride
Pollution Prevention/Risk Management
18 chemicals were selected by risk screening,
5 chemicals were selected by emission quantity; and
6 based on professional judgement
Figure 6. Baltimore Air Screening Results
62
-------
DEVELOPED RECOMMENDATIONS AND COMMUNICATED RESULTS TO THE
BROADER COMMUNITY (STEP 6)
Step 6-
Recommendations and
Communication
Develop Pollution Prevention and
Risk Management Recommendations
and Communicate Results
Recommendations
and Communication
Recommendations for Reductions
in Chemical Emissions and Levels
Consideration of Types of
Chemicals and Sources
Communication of Results and
Recommendations to Community
Overview
The final step of the Air Committee's
work focused on the development of
recommendations to improve air quality and
the communication of the results of the
Committee's work to the broader community.
As discussed in the Introduction, work on
these aspects of the screening exercise was
significantly delayed when in the summer of
1998, following the completion of the final
screening step, a key group of members left the
Committee. Following this development, the
Committee continued its work with input and
direction from the Partnership's Executive Committee. At this point, the recruitment of new
members became an additional goal for the Air Committee.
Recommendations for Acting on Results
Recommendations were developed to address the ambient air levels for the chemicals
identified in the final screen.
Benzene in Wagners Point Resulting from Stationary and Area Source Emissions
The Committee recommended work to identify pollution prevention and risk management
efforts to reduce emissions from the contributing facilities. Representatives of the bulk petroleum
facilities were contacted and invited to participate in the work of the Air Committee. Company
representatives and staff from the trade associations representing the companies agreed to
participate and work to identify and implement opportunities to reduce emissions of benzene. In the
spring of 1999, the residents of Wagners Point accepted a buyout offer unrelated to the work of the
Air Committee and relocation of the community began. As a result, Committee work on the
reduction of these benzene emissions was postponed.
63
-------
Benzene and 1,3-Buladiene Levels Attributed to Mobile Sources
Based on its analysis, the Air Committee concluded that mobile sources contributed a
significant portion of the levels of benzene and 1,3-butadiene in the Partnership neighborhoods.
Toxics from mobile sources are both regional and national air problems and cannot be addressed
solely in Partnership neighborhoods. As a result, the Air Committee recommended that the
Partnership consider participating in air quality improvement efforts at the regional level. Both the
MDE and EPA are considering new initiatives to control toxics from mobile sources and community
input will be crucial to those efforts. The Air Committee made plans to invite representatives from
MDE and EPA to speak to the Committee. The Committee will then develop a plan to make the
community's voice heard on these issues.
Carbon Tetrachloride and Methyl Chloride
Recommendations were not developed for these chemicals based on the Committee's
conclusion that their ambient levels were due to natural sources or past uses and not to any current
use or emissions.
Given the limits of the study conducted by the Air Committee, which focused on emissions
from industrial, commercial, and waste treatment and disposal facilities, the Committee also
developed the following recommendations for additional work to address community concerns:
Encourage appropriate actions to reduce odors;
Encourage appropriate action to reduce diesel truck exhaust through means such as
enforcement of current truck traffic restrictions, better diesel motor maintenance for
vehicles regularly using local roads, and rerouting of truck traffic; and
Develop ways to educate the community about the impacts of indoor air pollution.
Communication of the Results
The Air Committee made a major effort to find an effective way to communicate the results
of its work to the broader community. Preparing a report to the community may have taken as
much Committee work and time as conducting the technical screening exercise itself. The effort to
effectively communicate the work of the Committee to the community began at the secondary
screening step of the project. At that stage, several draft reports to explain the results of the
secondary screen were prepared and discussed at length in the Committee. However, a consensus
on the interpretation of the results did not develop, and the effort was halted as the results of the
final screen became available.
64
-------
With the results of the final screen and the recommendations in hand, the Committee began
a new effort to develop a report for the public. At least 10 drafts were prepared. As a part of this
process, the Committee brought together a group of residents not involved in the work to solicit
their input on how to communicate the results of the screening exercise. Questions developed from
this meeting were used to organize the report. The final Air Committee Report, approved in
October 1999, can be seen in Appendix J.
Several factors contributed to the difficulties encountered in the effort to develop the public
report. The work of the screening exercise was a new experience for all of the participants,
including the technical staff working on the Committee. As a result, a considerable amount of time
was spent learning about the process and its consequences. The task of summarizing the work in a
public report brought all issues and questions to the surface, and building a consensus in the
Committee on these issues required time and effort that could not be avoided. It was especially
difficult to develop the understanding and explanation for exactly what the screening exercise
could and could not accomplish. Understanding and explaining this required a review of all the
data and methods used by the Committee. The Committee used an extensive peer review process
to help it understand and clarify the issues raised in the report and to increase its confidence in the
results. This process itself required time and effort. Explaining the relationship of the information
provided in the exercise to the important questions of community health was especially difficult.
In addition, some Committee members did not expect the results found in the screening exercise.
The important discussion of the issues surrounding these expectations also added to the time
required to summarize the work. The Committee also learned that it was not enough to summarize
the results of its work, it also had to understand the community's views on the issues related to air
quality and health. Learning this also took time. In all, the difficulty in drafting and finalizing the
report was a reflection of the amount of educational work that was required to begin building a
consensus on air quality issues in the community.
Despite all the work put into the public report found in Appendix J, the Committee
recognized that it was still not adequate for broad dissemination in the community. While the
Committee was convinced that it was an accurate description of its work and that the results were
important information for the community, they recognized that it was still too long and technical
for wide distribution. As a result, the Committee adopted a plan to present the information in the
report to small groups in the community to get feedback on how to explain the screening exercise
and its results. Plans were made to present the results to the local Ministerial Alliance, groups of
local teachers, the chemical industry's Community Advisory Panel, Parent Teacher Association
groups, a local tenants' association, and other small community groups. The Committee planned to
prepare presentation, summary, and handout materials for these meetings based on the draft report.
Committee presentations to small community groups are now starting. Using feedback from these
meetings and its practice in preparing additional materials to explain the screening exercise and its
results, the Committee plans to hold larger public meetings to disseminate the information
throughout the community, as well as to recruit new members to address the issues recommended
for additional work.
65
-------
66
-------
GENERAL OBSERVATIONS ON THE SCREENING METHODOLOGY DEVELOPED IN
BALTIMORE
As explained in the introduction to this
report, the screening method developed in the
Partnership will undergo further development and
trial. Plans are currently under way for two
additional communities to use and improve this
methodology. A peer review process for the
methodology will also be undertaken both inside
and outside the Agency. Using the experiences of
the additional trials and the peer review, the
screening methodology presented in this case study
will be revised. The revised methodology will then
be disseminated widely as a tool for community
use.
Summary and Lessons Learned
• Methodology Was an Effective
Screening Tool for Southern
Baltimore
Partnership Benefitted from Air
Screening Exercise
• Technical Aspects of Screening
Methodology Need Further
Refinement
Preliminary conclusions based on the
Baltimore experience indicate that the screening methodology developed in the case study and
described in this report may have widespread application in communities concerned about air
quality. This methodology could provide communities with an effective screening tool and with a
process for building a community consensus on actions to improve air quality. Experiences in
Baltimore also point out several key areas where this process can be improved. The observations
and lessons learned, discussed below, will form the starting point for the further testing of the
methodology.
The Methodology Provides an Effective Screening Tool for Communities
Local communities often have difficulty understanding environmental data and reaching
consensus when setting priorities for effective community action. Communities are especially
concerned about aggregate and cumulative exposures from the multiple sources in and around their
communities. The screening methodology developed in Baltimore provides a technical tool to help
communities begin to evaluate the potential impacts of sources of air pollutants in their
neighborhoods and to quickly and effectively identify which chemicals might present higher than
acceptable risk. The screening tool enables a community to go beyond the commonly available
level of knowledge (of amounts and types of emissions) and to use information about the level of
risk those chemicals might present. The methodology helps a community to combine emissions
data, hazard information, exposure modeling, and risk screening in a priority-setting exercise.
Moreover, the screening tool allows communities to begin to evaluate the aggregate exposure to
single chemicals that have multiple sources in a local area and to consider cumulative effects by
identifying multiple chemicals that have similar effects. The tool is designed to provide
information in a relatively short time with limited resources. Use of the risk screening method
allows communities to avoid the costly and time-consuming analysis of a full risk assessment.
67
-------
while providing enough risk information to help a community build consensus on priorities for
improving air quality.
Because the information provided by a risk screening analysis of this kind is limited, a
special effort must be made to explain the uncertainties and limitations to the public. Without the
proper level of educational effort, the risk screening tool could easily be misunderstood for a risk
assessment, and conclusions could be mistakenly drawn that are not supported by the analysis.
This is an inherent limitation of this risk screening analysis that must be taken into account. The
narrow scope of the risk screening focuses on pollution sources to the ambient air, excluding other
important areas of environmental risk in the community such as indoor air. It is important for the
community to understand that the study on which this report is based examined only certain types
of sources and only from the inhalation pathway. Other media (e.g., contaminated soil, drinking
water, lead paint, etc.) and exposure routes should be taken into consideration. A special effort to
place the screening results in a wider context of environmental risks is important to the proper use
of this screening methodology, and may help avoid confusion and misplaced priorities.
The Methodology Helps Facilitate the Mobilization of Local Resources to Make
Improvements in Local Air Quality
The results of the risk screening methodology used in Baltimore include more than
technical facts about chemical risks that were determined using the screening tool. The
methodology also incorporates a collaborative process that can result in better approaches to
building community consensus and can mobilize community resources around concrete actions.
These benefits come from the work that is required to build a partnership and to conduct the
screening exercise. The Partnership attempts to bring all the sectors of the community together,
including governments, and provides a forum for dialogue on air quality issues. It encourages the
communication of information and perspectives among different sectors of the community and sets
the stage for the development of a community consensus. The technical screening process itself
provides a framework for the discussion of all the important air quality issues, as well as the
relevant scientific methods that are involved. A thorough and careful discussion and understanding
of hazard, exposure, modeling, and risk are essential to the success of the partnership approach.
The methodology also emphasizes the need to ensure that all participants can participate fully in
the process, maximizing the potential for consensus and for effective action. Overall, the
methodology is designed to build the long-term ability of the community to understand and address
air quality issues. As much as it is a technical screening tool, the methodology is also an
educational process designed to make the best information and science available to the community.
Because the educational and capacity-building approaches are essential to this
methodology, implementation requires the commitment of appropriate resources. The technical
screening exercise can be done relatively quickly, but the accompanying education of both the
Committee and the broader community will take time and resources. This process cannot be
shortened if consensus and community mobilization are the goal of the process.
68
-------
The Technical Aspects of Screening Methodology Need Further Refinement
• Addition of mobile source modeling. The Baltimore exercise focused on stationary
and area sources. This task will expand capacity of methodology to include mobile
source modeling.
• Review and improvement of source inventory review. Review existing source
inventories to identify additional sources of emissions to ensure that all significant
sources are included.
Identification of best source for toxicity data. Compare available toxicity databases
to identify most accessible and complete source of data for community screening
exercise.
• Expansion of screening methodology to include short-term acute effects.
• Review of screening calculations to determine if they are appropriate for and
protective of sensitive and urban populations.
Development of a method to screen for cumulative exposures in the initial screening
step.
• Expansion of methodology to include indoor air risks, to provide a more
comprehensive picture of air risks.
• Incorporation of GIS mapping to enhance the communication of the modeling and
screening results.
Specific Lessons Learned for Each Step of the Screening Methodology
Step 1. Lessons Learned: Built Partnership. Clarified Goals. Developed Outreach Plan
1. Clarify Expectations About the Results of the Project from the Start. It is important to clearly
explain in detail what the project will and will not be able to accomplish. The limitations of
the work must be completely understood, and the participants must agree that the results are
worth the effort they will be making. Pay special attention to explaining that the information
provided by the screening exercise will need to be combined with other information to
effectively address public health concerns. Also, pay special attention to clarifying the
difference between regulatory enforcement and voluntary pollution prevention actions.
2. Clarify the Roles of All Participating Partners Before Starting. While participants will need
to be flexible to meet unforeseen circumstances, clarifying and agreeing on roles up front will
69
-------
help communication. Participating governments should draft a Memorandum of
Understanding (MOU) clearly outlining the known project tasks and responsibilities. The
process of approving the MOU will give each participating government organization the
opportunity to ensure that enough resources are assigned to the project. An MOU of some
kind for all the partners may be helpful.
3. Choose Government Staff Trained in Outreach and Community Work To Staff the
Partnership Working Committee. Technical staff who lack community outreach training
should work with skilled community outreach staff. It is recommended that further training
be provided to government staff on multimedia and other technical approaches relevant to
community environmental concerns.
4. Establish a Set of Minimum Partnership Representation Requirements That Need To Be in
Place before Beginning a Project. Make sure there are enough willing partners from each
sector of the community who agree with the process and will work in a partnership with a
broad range of stakeholders. All partners also must be committed and willing to work toward
a consensus. Everyone will have personal agendas, but partners must be willing to work with
others to try to find common ground. Representation from the different partners should be
broad, reflecting as many community viewpoints as possible. Do not rely on a single group
or organization to represent the community or businesses. If the minimum requirements
cannot be met, it is better to postpone the project until broader participation can be
developed, because the problems created down the road are likely to make the work
ineffective.
5. Resources Must Match the Capacity of the Community Where the Project Is Located. If a
strong community infrastructure with representation from all sectors of the community
already exists, few resources will have to be devoted to building a partnership. Communities
lacking strong civic infrastructures will require considerable time and resources to develop
the necessary starting point for a successful project.
6. Work on Trust-Building at the Start and Throughout the Project. The partnership will bring
together a broad representation of the community and governments. Trust will be an issue.
This should be brought into the open and dealt with from the beginning. It will also reappear,
especially when difficult issues or decisions must be made, so attention must be paid to
building trust throughout the project.
7. Establish Ground Rules That Reflect the Nature of the Partnership and Show Respect for the
Process. Discussion of these ground rules will provide the key ingredients for trust-building
and the ability to complete work in an open and cooperative manner. Ground rules will
provide an easy reference at difficult parts of the process.
8. Get an Independent Facilitator for the Start-up Process and Working Meetings. It is very
important that someone skilled in facilitating partnerships be assigned to the group to pay
70
-------
attention and to make sure the process is working. The facilitator should understand the
content of the work but should be focused on process, making sure everyone participates
equally, meetings are run and organized well, issues of trust are dealt with, etc. It is not
possible to participate fully in the content of the working meetings and facilitate the process
at the same time. Facilitators can be paid or volunteer and can be found locally, such as a
local school principal or minister, or can come from outside the community from
organizations such as the National Civic League.
9. Set a Minimum Participation Level for Committee Legitimacy for Each Sector of the
Community and Establish It As a Necessary Quorum for Meetings. If the quorum is not met,
then the committee should shift its emphasis to recruitment.
10. From the Beginning of the Project, Identify Some Issues That Everyone Can Agree on and
Organize Small Actions To Make Progress on These Issues. Mixing action with screening
work will help avoid the feeling some will have of never actually doing anything but meeting.
It will also establish the Committee in the community and set the stage for better
communication. The Committee can learn more through action and can recruit new
members, if necessary. Taking action on asthma by setting up workshops through area Parent
Teacher Associations (PTAs) is an example of an action that the Committee could adopt.
Step 2, Lessons Learned: Built Source Inventory Database
1. Include the Means To Estimate or Collect Data on Emissions from Mobile Sources. Mobile
sources were not addressed in the Baltimore exercise primarily because the focus was on
commercial, industrial, and waste treatment and disposal sources. Since mobile sources
contribute significantly to air pollution, future efforts should consider modeling or measuring
emissions from mobile sources.
2. Investigate Existing Urban Source Inventories To Determine the Best Inventory To Use for
the Screening Methodology. The Baltimore methodology included point and area sources.
Other sources may need to be added.
3. Consider the Types of Emission Information Needed for the Screening Exercise As Soon As
Possible After the Project Begins. Information entered into a database from the onset of the
process is much easier to handle and organize than hard copies of information that have to be
physically manipulated.
4. Set up a Personal Computer in a Central Location. Having it set up in a community center
or office will give all participants easy access to the data. Provide training on data entry and
database use and maintenance. Investigate possibilities of accessing the database via Internet
or other forms of live data sharing.
71
-------
5. Create Fields in the Source Inventory Database To Identify the Data Source for Each Entry,
(e.g., from TRI or from the state permitting database). This is especially useful for
determining the most appropriate value when multiple values exist, and for quality control
purposes.
6. Routinely Update the Database. Emission data and other information are likely to change
over time. As new information becomes available, trained personnel should be available to
periodically make the relevant changes.
7. Use Residents, Local Industry, and Government Representatives as Valuable Resources To
Verify the Location and Operational Status of Facilities. A modest investment in equipment
such as a geographic positioning system (GPS) unit, laser range finder, and U.S. Geological
Survey (USGS) topo maps can significantly increase the accuracy of air modeling inputs such
as facility location.
8. Use State Air Toxics Studies Where Available. These documents may contain valuable
information that can be useful in conducting risk screening exercises such as data monitoring,
emission estimates, facility information, and assessment methodologies.
9. Save Significant Time and Effort by Designing Electronic Forms To Collect Various Types oj
Information. These forms can be transferred via e-mail and should be designed to be
compatible with the format of the emission inventory. In the case study, information on the
facilities' stack parameters was collected by hand on hard copy forms and entered into the
emission inventory database. Electronic forms would have allowed this information to be
transferred directly into the database.
Step 3. Lessons Learned: Conducted Initial Screening
1. Identify All the Key Decision Points in the Screening Exercise and Get Clear Committee
Decisions on These Issues Before Starting the Exercise. Focus especially on the decisions for
choosing screening values and their relationship to the purpose of doing the screening
exercise.
2. Make a Special Effort To Provide the Necessary Background Information for Nontechnical
Members of the Committee, Including Training, To Ensure That All Committee Members
Fully Understand the Science of the Screening Process Prior to Step 3. The screening
meetings will be fairly technical and should be conducted with careful preparation and good
facilitation. Such meetings should be held either by a subgroup that reports to the full
Committee, or by the full Committee. These screening meetings should be open and
residents should be encouraged to attend. Translation of the technical language (e.g., using
outreach materials to make sure the community at large understands the process) should be
provided for the nontechnical participants.
72
-------
3. Keep Detailed Records of the Decisions Made and the Reasons for the Decisions. All steps
of the screening process should be well documented for review by any interested community
members.
4. Be Thorough with the Review. Given the level of detail and the amount of information, it
would be better to hold two screening decision meetings. The first meeting should focus on
identifying missing information and familiarizing each person with the process. The second
and final decision meeting can then be more thorough and all points of view can be
considered. Of critical importance is the gathering of toxicity information for the risk
calculations. The database should be as complete as possible so the risk calculations can be
made. This will ensure that all chemicals of concern to the community will be identified in
the screening exercise.
5. Develop and Carry Out a Quality Assurance Method To Ensure That No Inadvertent Errors
Were Made in the Screening Exercise. All data entries and calculations should be checked
for accuracy. This quality control can be designed so that it does not cause too much of a
delay in the work. Perhaps a local college or university can provide quality assurance as a
class project.
6. Prepare a Summary of the Decision Meeting(s) and Provide Outreach Materials to the
Community Explaining the Decisions Immediately. Keep the community informed as the
screening process proceeds. This will start the information transfer to the community and
give the Committee practice in explaining the process, strengths, and weaknesses.
7. Review All the Assumptions of the Screening Process, Including the Generic Turner and ISC
Modeling Methods to Determine if Adjustments Are Needed To Protect Children and Other
Sensitive Populations in the Community.
8. Develop a Formal Method for Evaluating Potential Cumulative Exposures in the Initial
Screening Step. For the initial screening step, the Partnership Air Committee informally
reviewed chemicals with multiple sources to determine if the combination of sources would
reach the 10~6 cancer risk screening value.
9. Try To Make Background Information and Training Available To Ensure That All Committee
Members Fully Understand the Views of Each Member and the Science of the Screening
Process Prior to Step 3. This will take time, careful preparation, and good facilitation of
Committee meetings. The Committee should summarize this exchange of information and
prepare outreach materials to make sure the community at large has the benefit of this
information.
73
-------
10. Develop a Common Interpretation of the Modeling Information and Communicate This
Information to the Community at Each Stage/Step of the Process. The screening process
should not move forward until the Committee can reach agreement on any issues related to
modeling and until community outreach materials are prepared.
Step 4. Lessons Learned: Conducted Secondary Screening
1. At This Stage of the Screening Exercise, Avoid Using Actual Concentration Numbers in the
Presentation of the Screening Results. Using real numbers may create the impression that the
screening analysis is more exact than warranted. The estimation of emissions and the
uncertainties of the modeling used at this stage of the screening exercise are better expressed
simply as "above" or "below" the screening level. The screening is designed to eliminate
chemicals with some confidence, but those found to remain above the screening level need
further information before any conclusions can be drawn about potential effects.
2. Examine the Assumptions That Go into the Calculation of the Region 3 Risk-Based
Concentration Tables (or other sources for risk-based concentrations). Revisit assumptions
for future screening exercises to ensure they are protective of sensitive populations and
appropriate for urban ambient air screening.
3. Develop and Review Further the Method for Grouping Chemicals wth Similar Effects To
Estimate Cumulative Effects.
4. Keep Detailed Records and Check All Steps for Accuracy.
Step 5. Lessons Learned: Conducted Final Screening
1. Maintain Careful Record of the Information Provided by the Facilities in the Final Screen
and Check for Accuracy. A clear documentation of the differences between the secondary
and final screenings will be important.
2. If There Is a Monitoring Station In or Near the Project Area, Consider the Location of the
Monitoring Station as One of the Model Outputs so Comparison of Monitored and Modeled
Concentrations Can Be Facilitated.
3. If Possible, Verify Modeling Results with Monitoring for Validation.
4. Keep Detailed Records and Check All Steps for Accuracy.
74
-------
Step 6. Lessons Learned: Recommendations and Communication
1. Engage the Committee in the Preparation of Communication Materials That Explain the
Scope and Limits of the Exercise at the Beginning of the Process Before the Results Are In.
This will help everyone on the Committee to understand what will and will not come from
the exercise. The early preparation of communication materials will also help to ensure that a
gap does not exist between the time when the Committee gets the results of its screening
exercise and the communication of those results to the community. This gap allows
individuals to present their own interpretation of results to the community before the
Committee has a chance to communicate the view of the Committee consensus.
2. Establish Outreach Goals As a Core Committee Task. The Committee should combine
community outreach and information collection as equal goals. The Committee should
devote approximately equal time to outreach and screening throughout the project.
3. Develop Outreach Materials and Communicate to the Community at Each Stage of the
Screening Process, Not Just at the End of the Exercise. Communicate regularly to the
community during the course of the screening exercise, perhaps in the form of a newsletter,
press releases, and presentations to small community groups. This will develop the
communication skills of the Committee and help to avoid the problem of having to learn how
to communicate everything when the results come in. Meetings focused on screening and
outreach should alternate, with the Committee providing constant updates and education on
the work to the community. Please see the amended flow chart (Figure 7) for the screening
methodology that incorporates this lesson. This flow chart presents community outreach and
input, providing a more complete picture of the methodology than the flow chart presented in
the Introduction.
4. Communicate Regularly to the Press So They Understand the Process and Are Prepared To
Help with Communication to the Public.
5. Present the Results of the Risk Screening in as Broad a Context as Possible so the
Community Has the Information To Set the Most Effective Priorities. Consider providing
information on areas such as mobile sources and indoor air so that the community has as
complete a picture of air risks as possible.
75
-------
Step 1
Build Partnership
Step 2 -
Source/Emission/Monitoring
Inventory
of readily available information.
FORM PARTNERSHIP, CLARIFY GOALS,
DEVELOP COMMUNITY OUTREACH PLAN
Build Inventory
Step 3 -
Initial Screen
of readily available information
and conservative scenario.
DEVELOP EXPLANATION
AND OUTREACH MATERIALS
Derive Concentration
Estimate Health Risks
[Screen with Health Based Values)
Step 4 -
Secondary Screen
Modeling air dispersion with
readily available information.
RESULTS AND
OUTREACH MATERIALS
Use Dispersion Model to Estimate Concentrations
Screen with Health Based Values
DEVELOP EXPLANATION
AND OUTREACH MATERIALS
Step 5 -
Final Screen
Modeling air dispersion with
best available information.
RESULTS AND
OUTREACH MATERIALS
Use Dispersion Model to Estimate Concentrations
Screen with Health Based Values
DEVELOP EXPLANATION
AND OUTREACH MATERIALS
Step 6 -
Recommendations and
Communication
RESULTS AND
OUTREACH MATERIALS
COMMUNITY
OUTREACH
AND INPUT
COMMUNITY
OUTREACH
AND INPUT
COMMUNITY
OUTREACH
AND INPUT
COMMUNITY
-> OUTREACH
AND INPUT
Figure 7. Generic Air Screening Methodology for the Community
76
-------
REFERENCES
DHHS, 1994. Annual report on carcinogens. Prepared by National Toxicology Program (NTP),
U.S. Public Health Service, Department of Health and Human Services, RTF, NC.
National Research Council (NRC). 1983 Risk assessment in the federal government: Managing the
process. Committee on the Institutional Means for Assessment of Risks to Public Health,
Commission on Life Sciences, NRC. Washington, DC: National Academy Press.
SAL 1999. Modeling Cumulative Outdoor Concentrations of Hazardous Air Pollutants. Revised
Final Report. SYSAPP-99-96/33r2. Systems Applications International, Inc. February 1999.
Turner, D. 1994. Workbook of atmospheric dispersion estimates: an introduction to dispersion
modeling. Second edition. Boca Raton, FL.: CRC Press.
U.S. EPA, 1999a.. Framework for Community-Based Environmental Protection; EPA
237-K-99-001. Office of Policy/Office of Reinvention. February 1999.
U. S. EPA, 1999b. Chicago Cumulative Risk Initiative (CCRI). Website:
http://www.epa.gov/reg5oopa/agenda99/goal_8.htm
U. S. EPA, 1999c. Cumulative Exposure Project. Website:
http://www.epa.gov/cumulativeexposure/air/air.htm
U.S. EPA, 1999d. National Air Toxics Program: Integrated Urban Strategy. Website:
http://www.epa.gov/ttnuatwl/urban/urbanpg.html
U.S. EPA, 1999e. Integrated Urban Air Toxics Strategy. 64 FR 38705. July 19, 1999. U.S. EPA
Office of Air Quality Planning and Standards. Final Urban Air Toxics Strategy.
U.S. EPA. 1998. Science Policy Council Handbook on Peer Review. Office of Science Policy.
EPA 100- B- 98- 001. January 1998. Website: http://www.epa.gov/ordntrnt/ORD/spc/prhandbk.pdf
U.S. EPA, 1997a. Facility Index System (FINDS). Website:
http://www.rtk.net/www/data/fm_gen.html & Data Universal Numbering Sytem (DUNS) Website:
hrtp://www.epa.gov/envirofw/html/finds/duns_num_co.html
U.S. EPA, 1997b. Toxic Release Inventory (TRI). Website:
http://www.rtk.net/www/data/tri_gen.html
77
-------
U.S. EPA, 1997c. Aerometric Information Retrieval System Air Quality Subsystem (AIRS/AFS)
database. Office of Air and Radiation. Washington, DC. Website:
http://www.epa.gov/ttiVairs/afs/index.html
U.S. EPA, 1997d. EPA Region III Risk-Based Concentration Table (RBC Table). Website:
http://www.epa.gov/reg3hwmd/risk/riskmenu.htm
U.S. EPA, 1997e Integrated Risk Information System (IRIS). Website:
http://www.epa.gov/docs/ngispgm3/iris/index.html
U. S. EPA, 1997f. Health Effects Assessments (HEAST) Summary Tables. FY 1997 Update.
EPA-540-R-97-036. July 1997.
U.S. EPA. 1996. Proposed Guidelines for Carcinogenic Risk Assessment. EPA/600/P-92/003C.
Office of Research and Development. National Center for Environmental Assessment. April 1996.
U.S. EPA, 1995. Users guide for the industrial source complex (ISC3) dispersion models. Office
of Air Quality Planning and Standards (OAQPS) Emissions, Monitoring, and Analysis Division,
RTP, NC. Website: http://www.epa.gov/ttn/scram
U.S. EPA. 1992a. Guidelines for Exposure Assessment. EPA/600-Z-92/001. FR57:22888-22938,
May 29, 1992.
U.S. EPA, 1992b. A Tiered Modeling Approach for Assessing the Risks Due to Sources of
Hazardous Air Pollutants. U.S. EPA Office of Air Quality Planning and Standards, Research
Triangle Park, NC. EPA-450/4-92-001. March 1992.
U.S. EPA. 1989. Risk Assessment Guidance for Superfund, Volume I. Human Health Evaluation
Manual (Part A). EPA/540/1-89/002. Interim Final, December 1989. Website:
http://www.epa.gov/oerrpage/superfund/programs/risk/ragsa/index.htm
U.S. EPA. 1987. Guideline on Air Quality Models (Revised) U.S. EPA, Office of Air Quality
Planning and Standards, Research Triangle Park, NC. EPA-450/2-78-027R.
78
-------
APPENDIX A
List of Community Environmental Partnership (CEP) Air Committee Members
-------
REGULAR AIR COMMITTEE MEMBERS
NAME
Richard Anderson
Caroline Bahr
Delores Barnes
Rebecca Besson
John Besson
Ann Bonenberger
Clarice Brown
Peter Conrad
Francis Croft
Ruben Dagold
Stephen Dyer
Steve Farkas
Randy Gaul
Matt Gillen
Terry Harris
Reginald Harris
Albert Hayes
Ed Looker
David Lynch
*Dave Mahler
*Doris McGuigan
Richard Montgomery
Allen Morris
Charles Nardiello
William Paul
John Quinn
Rev. R. Andrews
Pars Ramnarain
Hank Topper
Don Torres
Michael Trush
ORGANIZATION
Concerned Citizens for a Better Brooklyn (CCBB)
Enoch Pratt Library
Concerned Citizens for a Better Brooklyn (CCBB)
Delta Chemical Corp.
Delta Chemical Corp.
Concerned Citizens for a Better Brooklyn (CCBB)
Southern Neighborhood Service Center
Baltimore City Planning Department
Sierra Club
Baltimore City Health Department
Grace Davison
Baltimore Gas & Electric (BGE)
Resident
U.S. EPA
Sierra Club
U.S. EPA
U.S. EPA
Resident
U.S. EPA
Condea Vista
Ministerial Alliance/Maryland Waste Coalition
Phoenix Services
CITGO
Arundel Corporation
MDE/ARMA
Baltimore Gas & Electric (BGE)
Brooklyn United Methodist Church
MDE/ARMA
U.S. EPA
Baltimore City Health Department
Johns Hopkins School of Hygiene & Public Health
* Co-Chairs
A-l
-------
APPENDIX B
Letters from Partnership
-------
CLEANUP
COALITION
July 14, 1998
The Honorable Lynn Goldman
Assistant Administrator
U.S. EPA
401 M St. SW
Washington, B.C. 20642
Michael McCabe
Regional Administrator
U.S. EPA
841 Chestnut Building
Philadelphia, PA 19107
Re: Environmental Partnership Program in South Baltimore
Dear Mr. McCabe and Ms. Goldman:
We are writing with regret to inform you that after two years and many hours of work,
we have decided that we can no longer participate in the Environmental Partnership Program
in South Baltimore. We count ourselves among the founders of this important project and we
have reached this conclusion only after considerable deliberation and a sincere effort to
salvage this troubled effort. We explain our reasons in some detail below, in the hopes that
they will help EPA redesign similar initiatives.
We began this process deeply committed to the Partnership's ultimate goal: the
discovery of more effective ways to reduce pollution through the reinvention of traditional
regulatory programs. But along the way, after countless meetings where we tried repeatedly
to pursue that objective, it became clear to us that other participants in the Partnership
Program did not share this goal, but rather saw the effort as a vehicle for pursuing their own
agendas. EPA, as the convener of this effort, must bear the responsibility for allowing this
dissension to fester, never effectively leading the group to reach consensus on the overall
purpose of the Partnership.
All of us have far too many opportunities to sit in rooms with people who disagree
with us. arguing endlessly about who is right. We long ago learned the pat positions of our
opponents and developed our own automatic responses. What we need — and what we
thought we would get from the Partnership when we first signed on — was a real opportunity
to get beyond rhetoric to results, developing a new and deeper understanding of the
environmental conditions that threaten us and debating the best way to address those
problems.
1 O 7 Scott
St--*Baltimore,
B-l
MD 21201*410 625 O559
-------
The final straw came at the last meeting of the Air Subcommittee. Industry
representatives, who at this point outnumber public interest representatives by a margin of
three to one, informed us at great length that there is no serious pollution problem in South
Baltimore and certainly no evidence that public health is suffering as a result of environmental
contamination, as opposed to the individual lifestyle choices of our families, friends, and
neighbors. In short, we were told that our concerns are fanciful and that we are sick because
we smoke and drive automobiles. Life is just too short to spend being hectored in this
manner.
The only redeeming feature of that meeting was a statement made by Reginald Harris,
the EPA Region III representative to the Partnership. Mr. Harris made an effort to explain to
our opponents why their arguments were unjustified and counterproductive. But this
intervention, as much as we appreciated it, came too little and too late.
As we wrote you a year ago, the Environmental Partnership Program in South
Baltimore failed for three distinct reasons: 1) the absence of tangible and specific goals and
milestones; 2) a process that erects high barriers to effective citizen participation; and 3) a
profound and systematic failure to communicate effectively by EPA line staff. Before you
begin a similar effort elsewhere in the country, we hope that you will carefully consider these
comments and not just move on, finding another group of unsuspecting citizens to participate
in such a pointless exercise.
Sincerely,
Doris McGuigan
Maryland Waste Coalition
Cleanup Coalition
, •
Terry Hams
Sierra Club
Cleanup Coalition
Ann Bonenberger
Maryland Waste Coalition
Concerned Citizens for a
Better Brooklyn
Cleanup Coalition
Rose Hindla
Fairfield/Wagner's Point
Neighborhood Coalition
Cleanup Coalition
Dru Schmidt-Perkins
League of Conservation
Voters
Cleanup Coalition
Dan Pontious"
MaryPIRG
Cleanup Coalition
cc:
Senator Barbara Mikulski, Senator Paul Sarbanes, Governor Parris Glendening,
Congressman Wayne Gilchrest, Senator George Delia, Delegates Timothy Murphy and
Brian McHale, Mayor Kurt Schmoke
Administrator Carol Browner, Deputy Administrator Fred Hansen, MDE Secretary
Jane Nishida, EPA Division Director William Sanders, EPA Region ITI representative
Reginald Harris
B-2
-------
Southern Baltimore & Northern Anne Arundel County
Community Environmental Partnership
Working Together to Improve our Communities
3606 Hanover Street Baltimore, MD 21225
The Honorable Lynn Goldman
Assistant Administrator
U.S. EPA
401 M St.. SW
Washington, D.C. 20542
Michael McCabe
Regional Administrator
U.S. EPA
841 Chestnut Building
Philadelphia, PA 19107
September 11, 1998
RE: Environmental Partnership in South Baltimore
Dear Mr. McCabe and Ms. Goldman:
This letter is a response to the July 14, 1998 letter from the Cleanup Coalition announcing their
withdrawal from the Community Environmental Partnership Air Committee. We are concerned about this
most recent attack on our organization, and we want you to know that those of us who are committed to
the Partnership far out number tfie handful of Partnership members who signed the letter. Three of the
individuals who signed the letter have never attended a meeting or been involved in the Partnership in a
significant way. We are afraid that these individuals represent groups with an agenda to discredit the
efforts of partnerships among residents, businesses and government officials. It appears to us that the
Cleanup Coalition, despite their worthy goals, is more accustomed to maintaining an adversarial approach
than to achieving positive results for the community. When positive and effective efforts like ours do not
come up with results that support their adversarial approach, their only option seems to be to withdraw
and write a letter. The approach of working together to create a win-win situation seems foreign to their
way of thinking.
Both the current letter and last year's letter criticizing our Partnership were timed to appear on the
day before our Air Committee was scheduled to finalize reports for the community. This is clearly not a
coincidence. The members of the Cleanup Coalition appear to be willing to try to block the distribution of
information important for our community's health. Their involvement in the process up to the finalizing
of the committee's most recent report tends to discredit their current position. Perhaps they are opposing
the report because the results do not appear to support their organizational agendas..
Members of the Cleanup Coalition are continuing their opposition to the new approach we have
taken in the Community Environmental Partnership. We have tried to go beyond the adversarial approach
and to build a partnership among all the sectors of our community. We are concerned about the
continuing opposition to this approach. Such opposition makes it difficult for us to focus on positive
community improvements. Valuable time and efforts has been spent responding to these concerns. We
would like to be able to focus more upon building a stronger partnership that will help our community.
B-3
-------
These are the facts about the Partnership:
The Partnership Air Committee and its draft report, contrary to the claims made, does not target
individual life styles or blame community members for their health problems. The Partnership Air
Committee has not, as claimed, spent endless hours in a wasted effort.
The Committee has completed one of the most comprehensive reviews of stationary source
releases ever attempted and it has accomplished this with the voluntary participation of all sectors
of the community. The Air Committee succeeded in pulling together a vast amount of information
and has succeeded in answering questions about local air quality that the community has been
asking for many years. The results of this work will give us a chance to be much more effective in
targeting our ongoing efforts to improve the health of our community.
The three members of the Cleanup Coalition who participated in the Partnership worked with this
committee and agreed with all of its major decisions up until their recent decision to withdraw.
The Partnership has harnessed a tremendous amount of volunteer effort to improve our
communities. We have had hundreds of school children and parents participate in two major park
clean ups and educationals.
We have had volunteer committee members spend countless hours working with state and
federal officials to collect and interpret vital environmental information for the community.
The Partnership has organized pollution prevention, tenant rights, Internet and computer training,
workshops on asthma, ozone, green business, it has continued positive efforts with Congressman
Wayne Gilcrest to pursue a wildlife reserve in the area, and more recently has begun to help local
residents find temporary employment
The Partnership has succeeded in bringing a very broad range of organizations and individuals
together to work in our communities. We have brought MDE, DPW, EPA, Johns Hopkins
School of Public Health, University of Maryland School of Social Work, Chesapeake Bay
Foundation, Save our Streams, Millennium, Chem Metals, FMC, Delta Chem, BFX 4 H, Civic
Works, hundreds of local middle, elementary and special educational and vocational children and
their parents, Brooklyn Homes Tenants Association, the Police Athletic League, and others-all
working together to find constructive solutions to community problems.
The Partnership has begun a major project to create a wildlife preserve and education center for
our communities on the north Brooklyn shore. This project could help change the reputation of
our neighborhoods and give our Region a priceless natural resource.
The Partnership has brought residents and industry together and opened up a broad community
dialogue on important issues.
The Partnership has created an unprecedented partnership of City, County, State, and Federal
governments and brought this partnership into the community to help us answer questions and
solve problems. This has given us a rare chance to work side by side with our government
agencies.
We hope that this partial list will convince you that our Partnership is doing importa- work, or, at least,
convince you to find out more about us. We are determined to continue and to bu on the work we have
begun. We are proud of what we have accomplished and we are excited about our ,uture plans. If you
have any questions about our work, we encourage you to please take-the time to find out as much as you
can about our Community Environmental Partnership. We would like to schedule a meeting with you to
further-discuss our activities and plans. If you can't visit us, please give us a call at 410-354-0352.
B-4
-------
Thank you for your support
Sineefely.. -> .
{Ls. \C J
Executive Committee
Rev. Rick Andrews, Wanda Grimes, Dan Butler
cc: Senator Barbara Mikulski, Senator Paul Sarbanes, Governor Parris Gendening, Congressman
Wayne Gilchrest, Senator George Delia, Delegates Timothy Murphy and Brian McHale, Mayor
Kurt Schmoke
Administrator Carol Browner, Deputy Administrator Fred Hansen, MDE Secretary Jane
Nishida, EPA Division Director William Sanders, EPA Region III representative Reginald Harris
B-5
-------
APPENDIX C
Sources for Facility Information
Envirofacts
TRI
FINDS (includes Dun & Bradstreet Numbers)
AIRS/AFS
-------
Envirofacts
C-l
-------
Envirofacts Database:
Website Address: http://w\vw.epa.gov/enviro/index iava.html
This website provides access to several EPA databases that provide you with information
about environmental activities that may affect air, water, and land anywhere in the United States.
The Environmental Protection Agency (EPA) created the Envirofacts Warehouse to provide the
public with direct access to the wealth of information contained in its databases. The Envirofacts
Warehouse allows you to retrieve environmental information from EPA databases on Air,
Chemicals, Facility Information, Grants/Funding, Hazardous Waste, Risk Management Plans,
Superfund, Toxic Releases, and Water Permits, Drinking Water, Drinking Water Contaminant
Occurrence, and Drinking Water Microbial and Disinfection Byproduct Information (Information
Collection Rule [ICR]). You may retrieve information from several databases at once, or from
one database at a time. Online queries allow you to retrieve data from these sources and create
reports, or you may generate maps of environmental information selecting from several mapping
applications available through EPA's Maps On Demand.
You can also read about the spatial data used by the Maps On Demand mapping applications.
The Locational Reference Tables contain all of the latitude and longitude coordinate information
available through Envirofacts.
C-2
-------
TRI
C-3
-------
TRI Standard Reports - RTK NET
http://www.rtk.net/vvww/data/trLgen.htnil
Toxics Release Inventory
Report
Facility TRI Report
ndustry TRI Report
Parent TRI Report
Offsite Transfer TRI Report
The Toxic Release Inventory (TRI) is a database of information about releases and transfers of toxic
chemicals from manufacturing facilities. Facilities must report their releases of a toxic chemical to TRI if
they fulfill four criteria:
1. They must be a manufacturing facility (primary SIC code in 20 -39);
2. They must have the equivalent of 10 full-time workers;
3. They must either manufacture or process more than 25,000 Ibs of the chemical or use more than
10,000 Ibs during the year;
4. The chemical must be on the TRI list of 350 specific toxic chemicals or chemical categories.
Therefore, not all, or even most, pollution is reported in TRI. However, T Rl does have certain
advantages:
1. It is multi-media. Facilities must report the amounts they release to air, land, water, and
underground separately, and must report how much they send off-site;
2. All quantities are reported in pounds. This is an advantage compared to databases like PCS,
which often report releases as concentrations, or other databases which report releases by
volume of waste. These measures are often impossible to convert into pounds;
3. It is congressionally mandated to be publically available, by electronic and other means, to
everyone. This means that it's relatively easy to obtain TRI data and that the data is weil-known,
becoming a national "yardstick" for measuring progress in pollution and waste generation.
The TRI data is reported by individual facilities, who send their reports to Federal EPA every year These
reports are filled out on a form called "Form R". EPA takes these forms and converts them into an
electronic database To better understand TRI data, it is recommended that you order a copy of one of
these forms from the TRI Hotline (1-800-535-0202). You can also order (for free) a national "data
release", or summary on paper, of TRI data every year from the Hotline.
C-4
2/8/gg 9:27*
-------
FINDS (including Dun & Bradstreet)
C-5
-------
INDS Standard Reports - RTK NET
http://www.rtk.netwww;aata/fin_gen,html
FINDS Facility Index System
Facility FINDS Report
ndustry FINDS Report
FINDS data is a comprehensive listing of facilities regulated under a
variety of EPA programs. The FINDS database provides some basic
information about each facility and a listing of its ID numbers in other
EPA databases. With these ID numbers, you know where to look for more
information (if you can somehow get access to the other EPA databases.)
FINDS has both master records and alias records. A master record describes
the most accurate information for a facility that is known to EPA. An alias
record describes information for a facility as it appears in another EPA
database. A single facility will have one master record and one or many
alias records in FINDS.
The program will search both the master and alias records, unless you search
specifically using a source program type in the Area report. Low detail
searches will display only the master records; High detail adds the alias
records. All that will be retreived in any case is the facility's name,
address, and a few other identifiers — that is all that is in FINDS.
C-6
f 1
2/8/99 9:26 AH
-------
Snvirofacil
Attribute: DUNS NUM CO
'Overview f Query ",' IVkKteJ j* feedback
Definition:
The Data Universal Numbering System (DUNS) value which uniquely identifies a corporate entity.
This attribute is the primary key for the FND_COMPANY and FND_DUNS_SIC CODE entity types and is
the foreign key for the FND_FACILITY entity type.
Definition Source:
FINDS 4.0 Data Element Dictionary, September 22, 1994.
Security:Public
Source System: FINDS
• FINDS Table: FINDS_FACILITY_ALL
• Element: DUNS_NUM_COMPANY
Last Updated: 03/31/95
Remarks: The data in the FND_COMPANY and FND_DUNS_SIC_CODE tables Is only available to
those EPA users who have access to the Internal Envirofacts database. Access to the data in this
table is restricted to EPA users due to the Agency's licensing agreement with Dun and Bradstreet
The information about this attribute is provided for the use of the EPA users who wish to query
the system. Outside users will not be able to access this table and will see an error message when
they try to access this table.
Properties: Mandatory Basic Text
• Length: 9
• Default: None.
Return to:
• FINDS Entity & Attribute Information
• Envirofacts Overview
C-7
2/8/99 1:06 PM
-------
AIRS/AFS
-------
Welcome to AIRS TIN
The Aerometric Information Retrieval
System (AIRS) TIN web site is designed to
provide technical information about the AIRS
data management system primarily to AIRS
users (state and local agency management,
EPA Regional Offices, consultants, and
environmental groups.)
We encourage you to visit the What's New
page to learn about current happenings and
events.
Main Table of Contents
What's New
Year 2000
AIRS Conference '99
AIRS Facility System (AFS)
Air Quality System (AQS)
AQS - Current System
AQS - Re-Engineering Project
AIRS User Registration Form
Instructions for Registration Form
Memos
Events/Training
Contacts
Technical Forum
Search TTNWeb
This site is maintained by the Information Management Group (IMG) of thelnformation Transfer and Program Integration Division
(ITPID), Office of Air Quality andPlanning & Standards, US Environmental Protection Agency (US EPA).
EPA | OAR | OAQPS | TTN | AIRS
http:/A/vww. epa. g ov/ttn/ai rs/
Search | AIRS Webmaster
July 21, 1999
C-9
10/21/99 4 11 PM
-------
AFS Mam
Wysiwyg //112/http://www epa.gov/ttn/airs/afs/index.hti
s Air
Rsmsitsc S. ScsKF;
IOAQPS
AIRS Facility Subsystem (AFS)
AFS contains emissions, compliance and permit
data for stationary sources regulated by the U.S.
EPA and state and local air pollution agencies.
This information is used by states in preparation
of State Implementation Plans (SIPs), to track
the compliance status of point sources with
various regulatory programs, and report
emissions estimates for pollutants regulated
under the Clean Air Act.
This site is designed to keep users of the
system apprised of developments. For general
background information about AFS and AIRS,
see AIRS Basic Facts.
To generate reports of AFS data (major point
sources), see the AIRSDate I*1**TTNJ page.
If you are a user of AFS and need technical
assistance, call 1-800-367-1044 or email
AFSHELPLINE(5)TRCCOS COM
AFS Table of Contents
'99 AIRS Conference
General Policy and FAQ
Compliance Community Info
Emissions Community Info
MACT Community Info
Toxics Community Info
Permits - Title V Community Info
State Emissions Inventory
Software Clearinghouse
AFS Memos
Software and manuals
Events
AIRS Contacts (pdf)
Technical Forum
Back to TTN AIRS Main
Search TTNWeb
EPA | OAR | OAQPS | T'N AIRS
http /Avww epa.gov/ttn/airs/afs/mdex html
Search | AIRS Webmaster
April 18, 1999
C-10
-------
APPENDIX D
Toxicity Information
EPA Region III Risk-Based Concentration Table
MRLs (ATSDR)
IRIS
HEAST
-------
EPA Region III Risk-Based Concentration Table
D-l
-------
US EPA Region 3 Risk Assessment http://www.epa.gov/regJnwma/risKynsKmenu.ntn
Region 3,
Hazardous Site Cleanup Division
RISK ASSESSMENT
> EPA Region III Risk-Based Concentration Table - October 1998 Update (Some files are
in Portable Document Format, PDF, and you will need a PDF reader. You may download a
free copy from the Web, supplied by Adobe Software or use a Reader of your choice. This
link to Adobe is only provided as a convenience to you, and does not represent a product
endorsement. Using this option will cause you to leave the EPA web site. You may return to
this page by navigating through the BACK button on your browser.)
Background Information
Updated Risk Based Concentration Table Cover Memo
RBC Table- PDFfile
RBC Table- Self-extracting Lotus 123 file (54k)
RBC Table- Self-extracting Lotus WK4 file (57k)
RBC Table- Self-extracting Excel file (76k);
fr Use of Monte Carlo Simulation in Risk Assessments
* Selecting Exposure Routes and Contaminants of Concern by Risk-Based Screening
* EPA Region III Guidance on Handling Chemical Concentration Data Near the Detection
Limit in Risk Assessments
^ Assessing Dermal Exposure fromSoil
f EPA Home I Region 3 Home I HSCD I Search Region 3 I Comments ]
URL: http://www. epa.gov/reg3hwmoVnsk/riskmenu.htm
This page last updated on December 24, 1998
D-2
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION III
841 Chestnut Building
Philadelphia, Pennsylvania 19107
SUBJECT: Risk-Based Concentration Table DATE: 10/1/98
FROM: Jennifer Hubbard, Toxicologist
Superfund Technical Support Section (3HS41)
TO: RBC Table Users
Attached is the EPA Region III Risk-Based Concentration (RBC) Table, which we
prepare and post periodically for all interested parties.
IMPORTANT NOTES: To make the RBC Table more accessible and to minimize paper
usage, it is now primarily available through the Internet. The address is
http://www.epa.gov/reg3hwmd7risk/riskmenu.htm. The Table is available in both Lotus
and Excel as "self-extracting" files. These files should be downloaded and then processed
with your computer's "run" function. The files can then be viewed in Lotus or Excel.
If you have technical questions about the lexicological or risk assessment aspects of the
RBCs, please contact Jennifer Hubbard at 215-814-3328 or
hubbard.jennifer@epamail.epa.gov. Other questions can be addressed to Vanessa Sizer or
Terri Fields at 215-814-3041. You can also consult the Frequently Asked Questions,
below.
CONTENTS, USES, AND LIMITATIONS OF THE RBC TABLE
The RBC Table contains Reference Doses (RfDs) and Cancer Slope Factors (CSFs) for
400-500 chemicals. These toxicity factors have been combined with "standard" exposure
scenarios to calculate RBCs—chemical concentrations corresponding to fixed levels of risk (i.e., a
Hazard Quotient (HQ) of 1, or lifetime cancer risk of 1E-6, whichever occurs at a lower
concentration) in water, air, fish tissue, and soil.
The Region III toxicologists use RBCs to screen sites not yet on the NPL, respond rapidly
to citizen inquiries, and spot-check formal baseline risk assessments. The primary use of RBCs is
for chemical screening during baseline risk assessment (see EPA Regional Guidance EPA/903/R-
93-001, "Selecting Exposure Routes and Contaminants of Concern by Risk-Based Screening").
The exposure equations come from EPA's Risk Assessment Guidance for Superfund (RAGS),
while the exposure factors are those recommended in RAGS or supplemental guidance from the
Superfund program. The attached technical background document provides specific equations
Celebrating 25 Years of Environmental Progress
D-3
-------
and assumptions. Simply put, RBCs are like risk assessments run in reverse. For a single
contaminant in a single medium, under standard default exposure assumptions, the RBC
corresponds to the target risk or hazard quotient.
RBCs also have several important limitations. Specifically excluded from consideration
are (1) transfers from soil to air and groundwater, 2) cumulative risk from multiple contaminants
or media, and (3) dermal risk. Additionally, the risks for inhalation of vapors from water are
based on a very simple model, whereas detailed risk assessments may use more detailed
showering models. Also, the toxicity information in the Table has been assembled by hand and
(despite extensive checking and years of use) may contain errors. It's advisable to cross-check
before relying on any RfDs or CSFs in the Table. If you note any errors, please let us know.
It is important to note that this Table uses inhalation RfDs and CSFs rather than RfCs and
inhalation unit cancer risks. This is because the latter factors incorporate exposure assumptions
and therefore can only be used for one exposure scenario. Because risk assessors need to
evaluate risks for many types of scenarios, the factors have been converted to the more traditional
RfDs and CSFs. Unless otherwise indicated in the toxicity-factor source, the assumption is that
RfCs and unit risks should be adjusted by a 70-kilogram body weight and a 20 m3/day inhalation
rate to generate the RfDs and CSFs.
Many users want to know if the RBCs can be used as valid no-action levels or cleanup
levels, especially for soils. The answer is a bit complex. First, it is important to realize that the
RBC Table does not constitute regulation or guidance, and should not be viewed as a substitute
for a site-specific risk assessment. For sites where:
1. A single medium is contaminated;
2. A single contaminant contributes nearly all the health risk;
3. Volatilization, leaching, dermal contact, and other pathways not included in the
RBCs are not expected to be significant;
4. The exposure scenarios and assumptions used in the RBC table are appropriate for
the site;
5. The fixed risk levels used in the RBC table are appropriate for the site; and
6. Risk to ecological receptors is not expected to be significant;
the RBCs would probably be protective as no-action levels or cleanup goals. However, to the
extent that a site deviates from this description, as most do, the RBCs would not necessarily be
appropriate.
To summarize, the Table should generally not be used to set cleanup or no-action levels
at
D-4
-------
CERCLA sites or RCRA Corrective Action sites, to substitute for EPA guidance for preparing
baseline risk assessments, or to determine if a waste is hazardous under RCRA.
SPECIAL NOTES
The RBC Table was originally developed by Roy L. Smith, Ph.D., for use by risk
assessors in the Region III Superfund program. Dr. Smith is no longer with Region III, and the
Table continues to evolve. You may notice some modifications of formatting and conventions
used in the Table.
For instance, besides formatting, the following changes are noteworthy:
• As usual, updated toxicity factors have been used wherever available. However, because
IRIS and provisional values are updated more frequently than the RBC Table, RBC Table
users are ultimately responsible for obtaining the most up-to-date values. The RBC Table
is provided as a convenience, but toxicity factors are compiled from the original sources
and it is those original sources that should serve as the definitive reference.
• Certain outdated and withdrawn numbers have been removed from the Table.
BACK BY POPULAR DEMAND: Changes to the table have been marked with asterisks
(**). This was the most commonly requested feature over the last six months. Changes
may involve a corrected CAS number or a correction in the VOC status, or they may
reflect changes of RfDs and CSFs on IRIS.
• RBCs are no longer rounded to 1E6 ppm. For certain low-toxicity chemicals, the RBCs
exceed possible concentrations at the target risks. In such cases, Dr. Smith rounded these
numbers to the highest possible concentration, or 1E6 ppm. The rounding has been
discontinued so that Table users can adjust the RBCs to a different target risk whenever
necessary. For example, when screening chemicals at a target HQ of 0.1, noncarcinogenic
RBCs may simply be divided by 10. Such scaling is not possible when RBCs are rounded.
• This Table was originally compiled to assist Superfund risk assessors in screening
hazardous waste sites. The large number of chemicals made the Table unwieldy and
difficult to keep current. Many of the chemicals did not typically (or even occasionally)
appear at Superfund sites. Starting with the April 1998 version of the Table, the 600+
chemicals were reduced to some 400-500 chemicals by eliminating many of those atypical
chemicals. Through time, the Table may continue to grow or decrease in size. Comments
on this issue are appreciated. During the last six months, only one request was received for
restoration of a chemical: NuStarhas been restored to the Table. (A list of the deleted
chemicals is attached.)
• At Region III Superfund sites, noncancer RBCs are typically adjusted downward to
correspond to a target HQ of 0.1 rather than 1. (This is done to ensure that chemicals with
D-5
-------
additive effects are not prematurely eliminated during screening.) However, some
chemicals have RBCs at HQs of 0.1 that are lower than their RBCs at 1E-6 cancer risk.
In other words, the screening RBC would change from carcinogenic to noncarcmogenic.
A new feature of this Table is that these chemicals are now flagged with a "!" symbol.
Therefore, assessors screening with adjusted RBCs will be alerted to this situation.
• Earlier versions of this Table included a substitution of inhalation toxicity factors for oral
factors whenever oral factors were unavailable (this applied only to groundwater and air,
but not soil or fish). This practice has been discontinued in order to minimize the
uncertainty associated with such a conversion. The discontinuation of this practice does
not significantly decrease the number of available RBCs.
• CAS numbers and volatility status have begun to be re-checked in accordance with
comments from users. At this time, 85% of the chemicals have been checked for
volatility, and about 99% of the CAS numbers have been venfied.
• Earlier versions of this Table included soil screening levels (SSLs), when those values
were available in draft form. Since the finalization of the SSL Guidance, risk assessors are
urged to consult the final SSL Guidance directly. The Guidance has detailed
recommendations on site-specific sampling and site-specific SSL generation. (Soil
Screening Guidance: User's Guide. April 1996, Publication 9355.4-23; and Soil Screening
Guidance: Technical Background Document. May 1996; EPA7540/R-95/128)
• One user of the Table pointed out that the CAS numbers do not contain the dashes that
are part of their format. CAS numbers have always appeared on the Table without dashes,
but may be converted to their dashed form by placing a dash before the last number
(farthest to the right), then moving two places to the left and placing another dash. For
example, "107131" becomes "107-13-1"; "7440360" becomes "7440-36-0"; "25057890"
becomes "25057-89-0." Region III could add the dashes directly to the Table, but we do
not wish to make this change without feedback from users on whether this would
adversely affect them. Therefore, we are soliciting comments on this issue (see box on
first page for address).
FREQUENTLY ASKED QUESTIONS
To help you better understand the RBC Table, here are answers to our most often-asked
questions:
1. How can the age-adjusted inhalation factor (11.66) be less than the inhalation rate for
either a child (12) or an adult (20)?
Age-adjusted factors are not intake rates, but rather partial calculations which have
different units from intake rates. (Therefore, they are not directly comparable.) The fact
that these partial calculations have values similar to intake rates is really coincidental, an
D-6
-------
artifact of the similar magnitude of years of exposure and lime-averaged body weight.
2. For manganese, IRIS shows an oral RfD of 0.14 mg/kg/dav, but the RBC Table uses 2E-2
mg/kg/day. Why?
The IRIS RfD includes manganese from all sources, including diet. The explanatory text
in IRIS recommends using a modifying factor of 3 when calculating risks associated with
non-food sources, and the Table follows this recommendation. IRIS also recommends
subtracting dietary exposure (default assumption in this case 5 mg). Thus, the IRIS RfD
has been lowered by a factor of 2 x 3, or 6. The Table now reflects manganese RBCs for
both "food" and "non-food" (most environmental) sources.
3. What is the source of the child's inhalation rate of 12 irrVday?
The calculation comes from basic physiology. It's a scaling of the mass-specific 20 m3/day
rate for adults from a body mass of 70 kg to 15 kg, using the 2/3 power of mass, as
follows:
Ircm = mass-specific child inhalation rate (m3/kg/day)
Ire = child inhalation rate (m3/day)
20 m3/day / 70 kg = 0.286 m3/kg/day (mass-specific adult inhalation rate)
0.286 m3/kg/day x (70067) = (Ircm) x (15067)
Ircm = 0.803 rrrVkg/day
Ire = Ircm x 15 kg = 0.803 m3/kg/day x 15 kg - 12.04 nrVday
4. Can the oral RfDs in the RBC Table be applied to dermal exposure?
Not directly. Oral RfDs are usually based on administered dose and therefore tacitly
include a GI absorption factor. Thus, any use of oral RfDs in dermal risk calculations
should involve removing this absorption factor. Consult the Risk Assessment Guidance
for Superfund. Part A, Appendix A, for further details on how to do this.
5. The exposure variables table in the RBC background document lists the averaging time for
non-carcinogens as "ED*365." What does that mean?
ED is exposure duration, in years, and * is the compuler-ese symbol for multiplication.
Multiplying ED by 365 simply converts the duration to days. In fact, the ED term is
included in both the numerator and denominator of the RBC algorithms for non-cancer
risk, canceling it altogether. See RAGS for more information.
D-7
-------
6. Why is inorganic lead not included in the RBC Table?
EPA has no consensus RfD or CSF for inorganic lead, so it is not possible to calculate
RBCs as we have done for other chemicals. EPA considers lead to be a special case
because of the difficult)' in identifying the classic "threshold" needed to develop an RfD.
EPA therefore evaluates lead exposure by using blood-lead modeling, such as the
Integrated Exposure-Uptake Biokinetic Model (IEUBK). The EPA Office of Solid Waste
has also released a detailed directive on risk assessment and cleanup of residential soil
lead. The directive recommends that soil lead levels less than 400 mg/kg are generally
safe for residential use. Above that level, the document suggests collecting data and
modeling blood-lead levels with the IEUBK model. For the purposes of screening,
therefore, 400 mg/kg is recommended for residential soils. For water, we suggest 15 ug/1
(the EPA Action Level in water), and for air, the National Ambient Air Quality Standard.
7. Where did the CSFs for carcinogenic PAHs come from?
The PAH CSFs are all calculated relative to benzo[a]pyrene, which has an IRIS slope
factor. The relative factors for the other PAHs can be found in "Provisional Guidance for
Quantitative Risk Assessment of Polycyclic Aromatic Hydrocarbons," Final Draft, ECAO-
CIN-842 (March, 1993).
8. May I please have a copy of a previous RBC Table?
We do not distribute outdated copies of the RBC Table. Each new version of the Table
supersedes all previous versions.
9. Please elaborate on the meaning of the "W" source code in the Table.
The "W" code means that a RfD or CSF is current!}' not present on either IRIS or
HEAST, but that it was once present on either IRIS or HEAST and was removed. Such
withdrawal usually indicates that consensus on the number no longer exists among EPA
scientists, but not that EPA believes the contaminant to be unimportant.
Withdrawn numbers axe shown in the Table because we still need to deal with these
contaminants during the long delays before replacement numbers are ready. For the
purpose of screening, a "W" value is similar to a provisional value in that neither value has
achieved Agency consensus. The '"W" code should serve as a clear warning that before
making any serious decision involving that contaminant, you will need to develop an
interim value based on current scientific understanding.
If you are assessing risks at a site where a major contaminant is coded C'W," consider
working with your Region EPA risk assessor to develop a current toxichv constant If the
site is being studied under CERCLA, the EPA-NCEA Regional Technical Support group
-------
may be able to assist.
10. Can I get copies of supporting documents for interim toxicily constants which are coded
"E" in the RBC Table?
Unfortunately, Region 3 does not have a complete set of supporting documents. The
EPA-NCEA Superfund Technical Support Center prepares these interim toxicity constants
in response to site-specific requests from Regional risk assessors, and sends the
documentation only to the requestor. The RBC Tables contain only the latest interim
values that we've either requested or have otherwise received. NCEA maintains the
master data base of these chemicals, but will not release documentation of provisional
values unless they are recent. Furthermore, since NCEA's Superfund Technical Support
Center is mainly for the support of Superfund, it usually cannot develop new criteria
unless authorized to do so for a specific Superfund project.
If an "E"-coded contaminant is a chemical of potential concern at your site, we urge you
to work with the EPA Regional risk assessor assigned to the project in order to develop or
obtain documentation for provisional values. EPA Region 3 furnishes documents only
when needed to support Regional risk assessments or recommendations.
Attached is a list of "E"-coded chemicals whose supporting documentation was issued
prior to 1996, indicating that toxicity information may need to be updated.
11. Why is there no oral RfD for mercury? How should I handle mercury?
IRIS gives oral RfDs for mercuric chloride and for methylmercury, but not for elemental
mercury. Therefore, the RBC Table reflects this primary source. Consult your
toxicologist to determine which of the available mercury numbers is suitable for the
conditions at your site (e.g., whether mercury is likely to be organic or inorganic.)
Attachments
D-9
-------
FT'A Rpginn III RBC Table 10/1/98 1
Sg»ilc *fl»ctl 1 • RBC tf HI tt 0
Risk-based concentrations
Ambient
air
ug/m3
8 1E-001 C
7.3E+O01 N
3 7E+O02 N
5.1E+001 N
2 1E-002 N
2 1E-002 N
1 4E-003 C
2.6E-002 C
7.8E-002 C
55E+O02 N
3 7E+000 N
3JE+000 N
3 7E-004 C
3 7E+000 N
22E-001 N
7 3E-002 N
1 OE+002 N
1 1E+000 N
1 5E+000 N
1 8E+000 N
1 5E+000 N
2 1E-001 N
4 1E-004 C
5 1E-002 N
33E+001 N
28E-002 C
5 7E-002 C
5 1E-001 N
1 5E+001 N
9 1E+001 N
1 1E+002 N
3.7E+002 N
22E-001 C
3 7E-002 N
27E-005 C
1 5E+004 N
1 1E+003 N
3 7E-002 C
7.5E-004 C
1 BE+002 N
57E-003 C
1 8E-001 C
28E-005 C
45E-001 C
2 1E+001 N
Fish
mg/kg
2 7E+001 N
1 4E+002 N
8 1E+OOO N
1 4E+002 N
27E+001 N
70E-004 C
5 8E-003 C
3 9E-002 C
2 OE+002 N
1 4E+000 N
1 4E+000 N
1 9E-004 C
1 4E+003 N
8 1E-002 N
2 7E-002 N
55E-001 C
54E-001 N
68E-001 N
54E-001 N
54E-001 N
2 1E-003 C
1 2E+001 N
1 4E-002 C
29E-002 C
95E+001 N
54E+000 N
34E+001 N
4 1E+001 N
1 4E+002 N
1 1E-001 C
1 4E-002 N
1 4E-005 C
54E+003 N
4 1E+002 N
1 9E-002 C
2 7E+000 N
6 8E+001 N
29E-003 C
4 5E-002 C
1 4E-005 C
2 3E-001 C
1 2E+002 N
1 < RBC-c
Soil
Industrial
mg'kg
4 1E+O04 N
20E+005 N
1 2E+O04 N
20E+O05 N
4 1E+O04 N
1 3E+000 C
1 1E+001 C
7 2E+001 C
3 1E+OO5 N
2 OE+003 N
2 OE+003 N
34E-001 C
20E+006 N
1 2E+002 N
4 1E+001 N
1 OE+003 C
8 2E+002 N
1 OE+003 N
8 2E+002 N
8 2E+002 N
3 8E+000 C
1 8E+004 N
26E+001 C
5 2E+001 C
1 4E+005 N
8 2E+003 N
5 1E+004 N
6 1E+004 N
20E+005 N
2 OE+002 C
20E+001 N
2 5E-002 C
8 2E+006 N
6 1E+005 N
34E+001 C
4 1E+003 N
1 OE+005 N
5 2E+000 C
8 2E+001 C
26E-002 C
4 1E+002 C
1 8E+005 N
Residential
mg/kg
1 6E+003 N
7 BE+003 N
4 7E+002 N
7 8E+003 N
1 6E+003 N
1 4E-001 C
1 2E+000 C
8 OE+000 C
1 2E+004 N
7BE+001 M
- 78E+001 N
3 8E-002 C
7 8E+004 N
4 7E+OOO N
1 6E+000 N
1 1E+O02 C
3 1E+001 N
39E+001 N
3 1E+001 N
3 1E+001 N
4 3E-001 C
7 OE+002 N
2 9E+000 C
5 8E+000 C
5 5E+003 N
3 1E+002 N
2 OE+003 N
2 3E+003 N
7 8E+003 N
2 2E+001 C
78E-001 N
28E-003 C
3 1E+005 N
2 3E+004 N
38E+000 C
1 6E+002 N
3 9E+003 N
58E-001 C
9 1E+OOO C
29E-003 C
4 6E+001 C
7 OE+003 N
-------
MRLs (ATSDR)
D-ll
-------
- Minimal Risk Levels for Hazardous Substances (MRLs) http://www.atsdr.cdc.gov/mrls.html
MRLs
Agency for Toxic Substances and Disease Registry
Division of Toxicology
ATSDR Contact Person for MRLs
Minimal Risk Levels (MRLs)
for
Hazardous Substances
The Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) [42 U.S.C. 9604 et seq.], as amended by the
Superfund Amendments and Reauthorization Act (SARA) [Pub. L. 99-499], requires that the Agency for Toxic Substances and Disease
9 Registry (ATSDR) develop jointly with the U.S. Environmental Protection Agency (EPA), in order of priority, a list of hazardous substances
r^ most commonly found at facilities on the CERCLA National Priorities List (NPL) (42 U.S.C. 9604(i)(2)); prepare toxicological profiles for each
substance included on the priority list of hazardous substances, and to ascertain significant human exposure levels (SHELs) for hazardous
substances in the environment, and the associated acute, subacute, and chronic health effects (42 U.S.C. 9604(i)(3)); and assure the
initiation of a research program to fill identified data needs associated with the substances (42 U.S.C. 9604(i)(5)).
The ATSDR Minimal Risk Levels (MRLs) were developed as an initial response to the mandate. Following discussions with scientists within
the Department of Health and Human Services (HHS) and the EPA, ATSDR chose to adopt a practice similar to that of the EPA's Reference
Dose (RfD) and Reference Concentration (RfC) for deriving substance-specific health guidance levels for non-neoplastic endpoints. An MRL is
an estimate of the daily human exposure to a hazardous substance that is likely to be without appreciable risk of adverse noncancer health
effects over a specified duration of exposure. These substance-specific estimates, which are intended to serve as screening levels, are used
by ATSDR health assessors and other responders to identify contaminants and potential health effects that may be of concern at hazardous
waste sites. It is important to note that MRLs are not intended to define clean-up or action levels for ATSDR or other Agencies.
The toxicological profiles include an examination, summary, and interpretation of available toxicological information and epidemiologic
evaluations of a hazardous substance. During the development of toxicological profiles, MRLs are derived when ATSDR determines that
reliable and sufficient data exist to identify the target organ(s) of effect or the most sensitive health effect(s) for a specific duration for a given
route of exposure to the substance. MRLs are based on noncancer health effects only and are not based on a consideration of cancer effects.
Inhalation MRLs are exposure concentrations expressed in units of parts per million (ppm) for gases and volatiles, or milligrams per cubic
meter (mg/m3) for particles. Oral MRLs are expressed as daily human doses in units of milligrams per kilogram per day (mg/kg/day).
-------
- Minimal KISK Levels for Hazardous Substances (MRLs) http://www.atsdr.cdc.gov/mrls.html
ATSDR uses the no-observed-adverse-effect-level/uncertainty factor approach to derive MRLs for hazardous substances. They are set below
levels that, based on current information, might cause adverse health effects in the people most sensitive to such substance-induced effects.
MRLs are derived for acute (1-14 days), intermediate (15-364 days), and chronic (365 days and longer) exposure durations, and for the oral
and inhalation routes of exposure. Currently, MRLs for the dermal route of exposure are not derived because ATSDR has not yet identified a
method suitable for this route of exposure. MRLs are generally based on the most sensitive substance-induced end point considered to be of
relevance to humans. ATSDR does not use serious health effects (such as irreparable damage to the liver or kidneys, or birth defects) as a
basis for establishing MRLs. Exposure to a level above the MRL does not mean that adverse health effects will occur.
MRLs are intended to serve as a screening tool to help public health professionals decide where to look more closely. They may also be
viewed as a mechanism to identify those hazardous waste sites that are not expected to cause adverse health effects. Most MRLs contain
some degree of uncertainty because of the lack of precise toxicological information on the people who might be most sensitive (e.g., infants,
elderly, and nutritionally or immunologically compromised) to the effects of hazardous substances. ATSDR uses a conservative (i.e.,
protective) approach to address these uncertainties consistent with the public health principle of prevention. Although human data are
preferred, MRLs often must be based on animal studies because relevant human studies are lacking. In the absence of evidence to the
contrary, ATSDR assumes that humans are more sensitive than animals to the effects of hazardous substances and that certain persons may
be particularly sensitive. Thus, the resulting MRL may be as much as a hundredfold below levels shown to be nontoxic in laboratory animals.
Proposed MRLs undergo a rigorous review process. They are reviewed by the Health Effects/MRL Workgroup within the Division of
Toxicology; an expert panel of external peer reviewers; the agency wide MRL Workgroup, with participation from other federal agencies,
including EPA; and are submitted for public comment through the toxicological profile public comment period. Each MRL is subject to change
as new information becomes available concomitant with updating the toxicological profile of the substance. MRLs in the most recent
->-> toxicological profiles supersede previously published levels. A listing of the current published MRLs is provided as follows.
ATSDR Contact Person
For additional information regarding MRLs, please contact:
Dr. Selene Chou
Division of Toxicology
Agency for Toxic Substances and Disease Registry
1600 Clifton Road, Mailstop E29
Atlanta, Georgia 30333
Telephone (404)639-6308 or 1-888-42-ATSDR (1-888-422-8737)
FAX (404)639-6315
E-Mail: cjc3@cdc.gov
Note: Information is in landscape format. Please use scroll bar on the bottom of the screen to access all the information. You can
also search the index of substances quickly by using the "Find" button.
10/27/99931 AM
-------
I )R - Minimal Risk Levels for Hazardous Substances (MRLs)
http://www. atsdr.cdc.gov/mrls. html
ATSDR MINIMAL RISK LEVELS (MRLs)
Name
o
ANTHRACENE
AROCLOR 1254
ARSENIC
BENZENE
BIS(CHLOROMETHYL) ETHER
April 1999
Dura-
Route tion
MRL
Fac-
tors Endpoint
Draft/ Cover
Final Date CAS Nu
ACENAPHTHENE
ACETONE
ACROLEIN
ACRYLONITRILE
ALDRIN
AMMONIA
Oral
Inh.
Oral
Inh.
Oral
Inh.
Oral
Oral
Inh.
Oral
Int.
Acute
Int.
Chr.
Int.
Acute
Int.
Chr.
Acute
Acute
Int.
Chr.
Acute
Chr.
Acute
Chr.
Int.
0 . 6 mg/kg/day
26 ppm
13 ppm
13 ppm
2 mg/kg/day
0.00005 ppm
0.000009 ppm
0.0005 mg/kg/day
0 . 1 ppm
0 . 1 mg/kg/day
0.01 mg/kg/day
0.04 mg/kg/day
0.002 mg/kg/day
0.00003 mg/kg/day
0 . 5 ppm
0 . 3 ppm
0 . 3 mg/kg/day
300
9
100
100
100
100
1000
100
10
100
1000
100
1000
1000
100
10
100
Hepatic
Neurol .
Neurol .
Neurol .
Hemato .
Ocular
Resp .
Hemato .
Neurol .
Develop .
Repro .
Hemato .
Develop .
Hepatic
Resp.
Resp .
Other
Final 08/95 000083
Final 05/94 000067
Final 12/90 000107
Final 12/90 000107
Final 04/93 000309
Final 12/90 007664
Oral Int.
Oral Chr.
Oral Chr.
Inh. Acute
Int.
Inh. Int.
10 mg/kg/day
0.02 ug/kg/d
0.0003 mg/kg/day
0.05 ppm
0.004 ppm
0.0003 ppm
100 Hepatic
300 Immuno.
3 Dermal
300 Immuno.
90 Neurol.
100 Resp.
Final 08/95 000120
Draft 12/98 011097
Draft 10/98 007440
Final 09/97 000071
Final 12/89 000542
-------
IRIS
D-15
-------
EPA's Integrated Risk Information System (IRIS) - Home Page
http://vAvw.epa.gov/docs/ngispgrn3/insyindex.html
&EPA
lfc**1 State
Integrated Risk Information System
Welcome to the IRIS home page, brought to you by the U.S. Environmental
Protection Agency (EPA) and its Office of Research and Development, National Center
for Environmental Assessment. IRIS is a database of human health effects that may
result from exposure to various substances found in the environment. Click on the
Substance File List button to go to a list of the available substance files; then click
on any file name on the list to open that file. For more information about IRIS, read
this Introduction.
Click here for What's New on IRIS, which highlights the most recent changes to IRIS
files.
See the Glossary of Risk Assessment-Related Terms and the list of Acronyms and
Abbreviations for more information explaining terms used in IRIS files.
A list of Toxicological Review support documents are available online. They are
provided in the Adobe Acrobat Portable Document Format* (PDF).
Background Information on methods used by EPA for deriving values in IRIS is available
here. Information on Limitations to the Use of IRIS is here. For information on
downloading IRIS, see the Stand Alone (Downloadable) IRIS Database page.
Here are some links to other sources of environmental health information.
EPA is continuously seeking to improve the IRIS home page and the scientific content
of IRIS. We welcome your comments and suggestions for improvements. Send
comments to the IRIS webmaster by email to Iris.Webmastergfrepa.qov
For technical questions about the scientific information content in IRIS, please call
the U.S. EPA Risk Information Hotline at telephone 1-513-569-7254, or fax to
1-513-569-7159, or email to RIH.IRIS@epamail.epa.Qov.
Navigation hints:
From the opening list of substances, you can click on individual substance names or
the list can be searched with your web browser, such as Netscape or Internet Explorer
by typing the name or Chemical Abstracts Service (CAS) Registry Number at the "Find"
D-16
9:27 Al
-------
HEAST
D-17
-------
Environmental Protection
Agency
Emergency Raspo
nse
PW7-921199
July 1997
Supertund
EPA Health Effects Assessment
Summary Tables
PB97-921195
FY1997 Update
D-18
-------
APPENDIX E
Document for Generic Turner Method for Estimated
Exposure from Near-Ground Releases to Air
-------
ESTIMATING AMBIENT INHALATION EXPOSURES
DUE TO NEAR-GROUND RELEASES OF PMN CHEMICALS
Turner's (1970) sector averaging form of the Gaussian algorithm can be used to estimate
concentrations resulting from a point source release:
(2.03X0
C -
Where:
C = Concentration in ambient air (mg/m3)
Q = Release rate (mg/sec)
H = Release height (m)
X = Receptor distance from source (m)
6z = Vertical dispersion coefficient (m)
u = Mean wind speed (m/sec)
Using the following assumptions. Equation No. 1 can be reduced to Equation No. 2:
Cone. = (Q) (6.165 x 10"4)
H = 3m
X = 100m
6z = 5m (assumes neutral atmospheric stability)
U2 = 5.5 m/sec
Cone. = (Q) (6.165 x 10
-4\
Since Equation No. 1 and Equation No. 2 use units of mg/sec for Q and air releases may
be reported in units of kg/yr, a conversion factor must be included in Equation No. 2. Assuming
a continuous release, kg/yr can be converted to mg/sec by multiplying by 0.0317
(mg''sec)/(kg/yr). Thus, the revised Equation No. 2 is listed below as Equation No. 3.
It is unlikely that any long-term releases would be blown continuously in the same
direction. It would be more reasonable to assume that, as a reasonable worst case, the wind
blows in one direction 25 percent of the time. Thus, the corrected Equation No. 3 is listed below
as Equation No. 4.
E-l
-------
Cone. = (<21T)(4.88 A- 10~6)
Annual exposure can be estimated using Equation No. 5.
EXPOSURE = (O (//?) (D) (F)
Where:
C = Result from Equation No. 4
IR = Assumed to be 1 rrr/hr
D = 24 hrs/day
F = 365 days/yr
Using the above parameters in Equation No. 5, annual exposure can readily be estimated
using Equation No. 6.
Annual Exposure = mg/yr = (O (0.043)
Note that because the exposure estimate is an annual average, it does not matter whether
the release occurs on a long-term or short-term basis. The average annual exposure is the same
for both situations assuming the annual amount released is the same.
E-2
-------
APPENDIX F
Examples of Release, Site and Monitoring Data Collected by Committee
Registered Source Emissions from MDE
TAP Emission Data from MDE
MDE Ambient Air Monitoring Station Description and Data
Data Retrieved from TRI
Data Retrieved from FINDS
Dun and Bradstreet Facility Data
MDE Facility Data
-------
Registered Source Emissions Data from MDE
F-l
-------
Registered Source Emissions - Zip Codes 21225 and 21226
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
7-11 Station
7-11 Station
5617 Ritchie Highway & Church Street
A -A Recycle & Sand
6931 Baltimore Annapolis Boulevard
Amerada Hess Terminal
6299 Pennington Avenue
Amoco Station
Amoco Station
Amoco Asphalt Terminal
3901 Asiatic Avenue
Amoco Terminal
6101 Pennington Avenue
Amoco Terminal
801 East Ordnance Road
Ansam Metals Corp
1026 East Patapsco Avenue
Arundel Corp
4th & Frankfurst Avenue
Arundel Elementary School
2400 Round Way
Atotech USA
1900 Chesapeake Avenue
3704 South Hanover Street
1927 Benhill Avenue
Baltimore City Composting Facility
5800 Quarantine Road
600 Carbon Avenue
Volatile Organic Chemicals
Toxic Air Pollutants
Volatile Organic Chemicals
Toxic Air Pollutants
Paniculate Matter
Sulfur Oxides
Paniculate Matter
Sulfur Oxides
Nitrogen Oxides
Carbon Monoxide
Volatile Organic Chemicals
Tnvir- Air PnllntanK
Volatile Organic Chemicals
Toxic Air Pollutants
Volatile Organic Chemicals
Toxic Air Pollutants
Paniculate Matter
Sulfur Oxides
Mitrogen Oxides
Carbon Monoxide
Volatile Organic Chemicals
Paniculate Matter
Sulfur Oxides
Nitrogen Oxides
Carbon Monoxide
Volatile Organic Chemicals
Tnirir Air Pnllirtantc
Volatile Organic Chemicals
Toxic Air Pollutants
Paniculate Matter
Sulfur Oxides
Nitrogen Oxides
Carbon Monoxide
Volatile Organic Chemicals
Tnirir Air Pollutant*
Paniculate Matter
Sulfur Oxides
Nitrogen Oxides
Carbon Monoxide
Toxic Air Pollutants
Paniculate Matter
Sulfur Oxides
Nitrogen Oxides
Carbon Monoxide
Toxic Air Pollutants
Paniculate Matter
Nitrogen Oxides
Carbon Monoxide
Antimony Compounds
Chromium Compounds
Nitnc Acid
7mr r.nmpnnnrK _
Toxic Air Pollutants
Toxic Air Pollutants
Volatile Organic Chemicals
Toxic Air Pollutants
Toxic Air Pollutants
6,200 Ibs
NR
2.920 Ibs
NR
1.560 Ibs
320 Ibs
NR
4.460 Ibs
5 1,980 Ibs
18,320 Ibs
1 ,660 Ibs
172,280 Ibs
NR
23,000 Ibs
NR
22,560 Ibs
NR
360 Ibs
10.940 Ibs
5, 100 Ibs
1,460 Ibs
2.420 Ibs
NR
2.400 Ibs
27,600 Ibs
9.600 Ibs
1 ,200 Ibs
1,000 Ibs
NR
85,040 Ibs
NR
560 Ibs
580 tos
400 Ibs
2.880 Ibs
780 Ibs
NR
40. 120 Ibs
1 .500 Ibs
740 Ibs
240 Ibs
NR.
360 Ibs
1,080 Ibs
1,080 Ibs
360 Ibs
MB
3,980 Ibs
17,780 Ibs
240 Ibs
2,673 Ibs
1 Ib
1 Ib
1 IK
2.520 Ibs
NR
360 Ibs
NR
360 Ibs
NR
1,820 Ibs
NR
~" Data Source
1995 Emissions Statement
No Reporl
1995 Emissions statement
No Report
1995 Emissions Statement
1995 Emissions Statement
No Report
1995 Emissions St; iment
1995 Emiss-ons Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
Nn Ropnrt
No Report
1995 Emissions Statement
No Report
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
No Rppnrt
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
Nn Rpnnrt
1995 Emissions Statement
No Report
1995 Emissions Statement
1995 Emissions Statement
1 995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
Nn Rppnrt
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1995 Emissions Statement
1 995 Emtssjons-Slalfirnent
iggS^Bmlssions Statement
/ 1994 EPA Form R
/ 1994 EPA Form R
\ 1994 EPA Form R
x 1QQ4 PPfl Fnrm B
1995 Emissions Stalenjejit
No Report
1995 Emissions Statement
No Repnrt
1995 Emissions Statement
1995 Emissions Statement
F-2
-------
TAP Emissions Data from MDE
F-3
-------
Premise
Number
02-0073
02-0044
02-0044
02-0044
02-0044
02-0044
02-0044
02-0044
02-0044
02 0044
02-0044
02-0044
02 0044
02-0044
02-0044
02-0044
02-0055
02-0055
02-0055
02-0056
02-0056
02-0056
02-0056
02-0056
02-0056
02-0309
02-0309
02-0309
02-0309
02-0309
02-0309
Plant name
VALLEY PROTEINS
TAP Emissions in Zip Codes 21225 & 21226
Zip
Code Pollutant
21226 Chlorine dioxide (AAL-15.5)
Street Address
1515 OPEN STREET CURTIS BAY
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
REICHHOLD CHEMICAL
SOUTHERN STATES CORP
SOUTHERN STATES CORP.
SOUTHERN STATES CORP
CHEMETALS CORPORATION
CHEMETALS CORPORATION
CHEMETALS CORPORATION
CHEMETALS CORPORATION
CHEMETALS CORPORATION
CHEMETALS CORPORATION
AMOCO OIL COMPANY
AMOCO OIL COMPANY
AMOCO OIL COMPANY
AMOCO OIL COMPANY
AMOCO OIL COMPANY
AMOCO OIL COMPANY
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
6401 CHEMICAL ROAD BALTIMORE 21226
ORDINANCE ROAD « PENNINGTON 21226
ORDINANCE ROAD 4 PENNINGTON 21226
ORDINANCE ROAD « PENNINGTON 21226
711 PITTMAN ROAD,
711 PITTMAN ROAD,
711 PITTMAN ROAD,
711 PITTMAN ROAD,
711 PITTMAN ROAD,
711 PITTMAN ROAD.
16 ORDINANCE ROAD
16 ORDINANCE ROAD
16 ORDINANCE ROAD
16 ORDINANCE ROAD
16 ORDINANCE ROAD
16 ORDINANCE ROAD
02-0316
02-0318
02-0316
02-0316
02-0316
02-0316
02-0318
02-0316
02-0316
02-0316
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
US COAST GUARD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
CONCRETE ROAD
Ethyl benzene
Ethylene glycol
Melhyl Isobutyl kelone
Malelc anhydride
Toluene
2-Buloxyethanol
Elhanol,2-(2-butoxyelhoxy) (X)
Xylene
Isophorone dllsocyanate
Toluene 2,4-dllsocyanate
Isopropyl alcohol
n-Butyl alcohol
sec-Butyl alcohol
Methyl ethyl ketone
Toluene, 2.6-dllsocyanate
Manganese & compounds (Fumes)
Copper & compounds
Phosphoric acid
CURTIS BAY 21226 Manganese & compounds (Fumes)
CURTIS BAY 21226 Hydrogen chloride (AAL-1171-
CURTISBAY 21226 Ammonia (AAL-450,300V»137
CURTIS BAY 21226 Sulfurte scld 5TVv6.K '—
CURTIS BAY 21226 Hydrogen peroxide
CURTIS BAY 21226 Hydrogen suinde
CURTIS BAY 21226 Elhyl benzene
CURTIS BAY 21226 Toluene
CURTIS BAY 21226 Phenol
CURTIS BAY 21226 Xylene
CURTIS BAY 21226 Methyl tertlary-butyl-elh*f(X)
CURTIS BAY 21226 Benzene
21226 Methylene Wstphenytlsocyanate)
21226 Ethylene glycol
21226 Methyl Isobutyl kelone
21226 Toluene
21226 2-Butoxyelhanol
21226 Perchloroethylen*
21226 Sodium hydroxide
21226 Manganese Oxide
21226 Xylene
21226 Carbonic Ackj Olsodlum Sail
CAS n
10049044
100414
107211
108101
108316
108883
111762
112345
1330207
4098719
584849
67630
71363
78922
78933
91087
7439965
7440508
7664382
7439965
r-JJ647010
664417
Emissions
Ibs/hr tons/year
00833 02570
02500 00174
0 0070 0 0056
1 0200 01055
00020 00006
0 6700 0 0804
00590 00176
03900 00448
1 8000 0 ?665
00000 00000
0 0020 0 0004
07500 00500
02700 00625
02200 00156
00160 00355
00011 00002
0 0007 0 0002
0 0007 0 0002
00000 00000
50057 219250
1 5600 68330
6 8000 29 7840
04130 18105
03000 1 3140
00016 00070
00790 03470
1 1100 48730
00000 00000
03000 1 3000
23400 102000
0 4620 2 0200
00000 00000
01600 01700
0 4000 0 2835
0 4000 0 3870
0 0560 0 0585
00200 00165
0 0020 0 00?5
00020 00120
0 0600 0 0575
00012 00050
-------
MDE Ambient Air Monitoring Station Description and Data
F-5
-------
10/22/99 FRI 16:03 FAX
MARYLAND DEPARTMENT OF THE ENVIRONMENT
AIR AND RADIATION MANAGEMENT ADMINISTRATION
TOXICS MONITORING IN MARYLAND
EPA Method TO-14 -Toxics Monitoring Network
We are collecting 24 hour canister samples every 'sixth day on the
EPA schedule at these sites:
1. Essex
2. North East Police Station (NEPS)
3. Old Town Fire Department
4. Ft. McHenry
5. FMC
6. Glenn Burnie
also (samples collected by the local or State agency, analyzed by
ARMA)
7. Flag Plaza, Pittsburgh
8. AMSL, Philadelphia
9. Lums Pond, DE.
10. Washington, DC.
11. Chester, PA.
12. Marcus Hook, PA.
13. Chester, WVA.
14. Three sites in Ohio (on a variable schedule).
We have been designated as the Quality Assurance laboratory for EPA
Region III and split QA samples with Virginia and EPA Regions I and
II.
Samples are collected using either an EPA designed and fabricated
sampler or the XonTech Model B10A Ambient Air Collection Sampler.
Samples are collected into evacuated (less than 1 mm Hg absolute
pressure) 6 liter SUMMA treated stainless steel sampling canisters
and filled to a pressure of about 2 atmospheres over the day
(midnight to midnight) . At midnight of the sampling day the
sampler starts a pump which pulls ambient air through a stainless
steel sampling cane and pushes the air through a mass flow
controller, a shut off valve and into the canister. The flow rate
of 8.3 millimeters per minute is maintained throughout the 24
hours.
At the end of the 24 hour period the shut off valve is closed to
trap the sample in the canister and the sampler turns off. Between
the sampling dates AFJ1A personnel visit the site, close the manual
valve on the canister, remove the canister from the sampler and
place a new canister on the sampler for the next sampling date.
The canisters are returned to the laboratory for ana.lysis using an
EnTech Model 2000 Preconcentrator and a Hewlett-Packard Model 5890
gas chromatograph with a Model 5971 mass selective detector
(GC/MSD) . We are following EPA COMPENDIUM . METHOD TO-14, "The
Determination of Volatile Organic Compounds (VOCs) in Ambient Air
Using SUMMA Passivated Canister Sampling and Gas Chromatographic Analysis".
F-6
-------
FORTY ONE COMPOUNDS DETERMINED BY GC/KSD
USING EPA METHOD TO-14
Benzene
Bromomethane
1,3-Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroethene
Chloroform
Chloromethane
chloromethylbenzene
1,2-Dibromoethane
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
Dichlorodifluoromethane
1,1-Dichloroethane
1,2-Dichloroethane
1,l-Dichloroethene
cis-l,2-Dichloroethene
1/2-Dichloropropane
cis-l,3-Dichloropropene
trans-l,3-Dichloropropene
l/2-Dichloro-l/l,2/2-tetrafluoroethane
Ethylbenzene
l-Ethyl-4-methyl benzene
Hexachloro-l,3-butadiene
Methylene chloride
styrene
1,1,2/2-Tetrachloroethane
Tetrachloroethene
Toluene
1/2,4-Trichlorobenzene
1,I/1-Trichloroethane
1,1,2-Trichloroethane
Trichloroethene
Trichlorofluoromethane
1,1,2-Trichloro-l/2/2-trifluoroethane
1,2,4-Trimethylbenzene
1,3,5-Trimethylbenzene
o-Xylene
m & p-Xylene
F-7
-------
1^003
10'22/99 FRI 16:04 FAA
GENERAL AIR QUALITY
The U.S. Environmental Protection Agency .(JSPA). has .established
National Ambient Air Quality Standards (NAAQS) for 'six., criteria
pollutants: (1) sulfur dioxide, (2) particulate matter, (3)
carbon monoxide, (4) nitrogen dioxide, <5) ozone, and (6) lead.
The primary standards were established to protect public health,
and the secondary standards were' developed to protect against
non-ht.alth effects such as damage to property and vegetation.
The Department operates an air monitoring network throughout the
State in accordance with EPA guidelines to measure'the
concentrations of the criteria pollutants in the ambient air.
These measurements have been used to project statewide ambient
air quality and have indicated that south Baltimore meets the
ambier.t air quality standards for sulfur dioxide, particulate
matter, carbon monoxide, nitrogen dioxide, and lead.
Ground level ozone continues to present a problem for the
Baltimore/Washington area, which is classified as a non-
attainment area for ozone. The primary contributors to the
formation of ozone are emissions of oxides of nitrogen, primarily
from combustion equipment, and.emissions of Volatile Organic
Compounds (VOC) such as paint solvents and gasoline vapors.
-------
10/22/99 FRI 16:06 FAX
A brief description of this complicated method is as follows:
The sample canisters, 'along with a canister of zero air, a QA
mixture and a canister of'standard gas, are placed on a sixteen
position sampling, manifold. Pollutants in the air (zero air,
standard, QA mix or sample) are concentrated using an EnTech Model
2000.Preconcentrator. A glass bead trap is cooled to -150 C and
then 500 ml of sample is pulled through the trap:. The organic
compounds in the air freeze out on the glass beads While the normal
air constituents (including methane) pass through the trap. A
second smaller trap is then cooled to '-1.60 C. The first trap is
heated to 180 C and the collected compounds are driven off of the
trap onto the second focusing, trap using a smaller vpluae of gas (7
ml) . After the compounds are transferred to the Ifocusing trap,
this trap is heated to 75 c and the .compounds are driven off of the
trap into the GC.
The compounds pass through the GC and are separated so that the
most volatile pass through first. The compounds then pass from the
GC to the MSD where they are detected. Two analysis of zero air
are made first to show that the analytical system is clean. Then
the system is calibrated'using three levels of a 41 compound known
standard mixture. These compounds are specified in the TO-14
Method and are listed on the attached table. Then a QA mixture
containing four compounds of known amounts prepared by a different
method are analyzed to :double check the system.. After the system
is shown to be clean, is calibrated and passes the QA check, the
samples are analyzed in the sample manner as the standards.
Results are obtained by comparing the signal amounts from the
samples to .the known signals from the standard mixture. This is
done using the HP software that is used with the GC/MSD and reports
are obtained for each sample with the 41 compounds listed in parts
per billion.
F-9
-------
TWENTY FOUR HOUR TOXICS AIR SAMPLING
PARTS PER BILLION V/V, 1992
T)
O
COMPOUND5
Dichlorodifluoromethane
Chloromethane (Methyl chloride)
1,2-dichloro-1 1,2,2-tetrafluoroethane
Chloroethene (Vinyl chloride)
1,3-Butadiene
Bromomethane (Methyl bromide)
Chloroethane (Ethyl chloride)
Trichlorofluorometnane
11 -Dichloroethene
Methylene chloride (Dichloromethane)
1,1,2-trichloro-1,2,2-trifluoroethane
1,1 -Dichlproethane
cis 1,2-Dichloroethene
Chloroform (Trichloromethane)
1,2-Dichloroethane (EDC)
1,1,1 -trichloroethane
Benzene
Carbon tetrachloride
1,2-Dichloropropane
Trichloroethene (Trichloroethylene)
cis-1,3-dichloropropene
trans-1,3-dichloropropene
1,1,2-Trichloroethane
Toluene
1,2-Dibromoethane (Ethylene dibromide)
Tetrachloroethene (Perchloroethylene)
Chlorobenzene
Ethylbenzene
meta & para-Xylene
Styrene
1, i ,2,2-Tetrachloroethane
ortho-Xylene
1 -Ethyl-4-methyl benzene
1,3,5-Trimethylbenzene
1,2,4-Trimethylbenzene (Pseudocumene)
1,3-Dichlorobenzene
Chloromethylbenzene
1,4-Dichlorobenzene(p-Dichlorobenzene)
1,2-Dichlorobenzene(o-Dichlorobenzene)
1 '2 4-Trichlorobenzene
Hexachloro- 1,3-butadiene
OLD TOWN
F.M.C. CORP.
FORT McHENRY
MIN.
0.00
0.14
0.00
0.00
0.00
0.00
0.00
3.33
0.00
0.12
0.43
0.00
0.00
0.00
0.00
0.15
0.54
0.04
0.00
0.00
0.00
0.00
0.00
0.83
0.00
0.05
0.00
0.14
0.49
0.02
0.00
0.18
0.00
0.08
0.22
0.00
0.00
0.00
0.00
0.00
o.oo
MAX.
1.47
1.35
0.03
0.17
0.51
0.14
0.00
39.34
0.13
6.75
3.89
0.00
0.05
0.16
0.07
2.66
2.71
0.24
0.02
0.29
0.01
0.00
0.02
6.91
0.07
0.56
0.03
1.09
3.24
0.32
0.02
1.18
0.64
0.78
2.68
0.15
0.00
0.17
0.04
0.02
O.O1
AVG.
0.71
0.59
0.01
0.01
0.26
0.02
0.00
9.95
0.02
0.50
1.16
0.00
0.00
0.06
0.01
0.51
1.09
0.12
0.00
0.03
0.00
0.00
0.00
2.35
0.00
0.16
0.00
0.38
1.26
0.10
0.00
0.52
0.24
0.30
0.93
0.01
0.00
0.05
0.00
0.00
o.oo
MIN.
0.29
0.26
0.00
0.00
0.00
0.00
0.00
0.22
0.00
0.08
0.16
0.00
0.00
0.00
0.00
0.17
0.20
0.05
0.00
0.00
0.00
0.00
0.00
0.30
0.00
0.00
0.00
0.05
0.16
0.00
0.00
0.08
0.02
0.00
0.05
0.00
0.00
0.00
0.00
0.00
o.oo
MAX.
1.30
2.11
0.03
0.15
0.37
0.63
0.06
4.28
0.09
5.59
2.59
0.00
0.01
0.16
0.09
0.74
6.03
1.71
0.11
0.17
0.00
0.00
0.01
9.74
0.07
0.20
0.11
1.00
4.05
0.22
0.03
1.14
0.44
0.23
0.76
0.03
1.33
0.09
0.05
0.10
O.05
AVG.
0.64
0.78
0.01
0.01
0.10
0.05
0.00
0.55
0.02
0.40
0.39
0.00
0.00
0.04
0.01
0.37
1.32
0.20
0.00
0.02
0.00
0.00
0.00
2.00
0.01
0.10
0.04
0.27
0.97
0.07
0.00
0.35
0.09
0.09
0.26
0.01
0.03
0.03
0.01
O.O1
o.oo
MIN. I
0.19
0.16
0.00
0.00
0.00
0.00
0.00
0.29
0.00
0.09
0.07
0.00
0.00
0.00
0.00
0.22
0.20
0.03
0.00
0.00
0.00
0.00
0.00
0.28
0.00
0.00
0.00
0.06
0.17
0.00
0.00
0.09
0.02
0.02
0.06
0.00
000
0.00
0.00
0.00
o.oo
MAX.
1.28
1.23
0.03
0.17
0.21
0.14
0.00
1.06
0.14
3.16
0.26
0.00
0.04
0.11
0.07
1.05
1.29
0.20
0.01
0.13
0.00
0.03
0.00
2.89
0.07
0.21
0.08
0.50
1.46
0.45
0.02
1.81
0.15
0.14
0.44
0.05
0.00
0.12
0.03
0.01
o.oo
AVG.
0.68
0.62
0.02
0.01
0.09
0.03
0.00
0.45
0.03
0.30
0.13
0.00
0.00
0.05
0.01
0.39
0.53
0.12
0.00
0.03
0.00
0.00
0.00
1.10
0.01
0.09
0.01
0.18
0.60
0.05
0.00
0.36
0.05
0.06
0.17
0.01
0.00
0.03
0.00
0.00
o.oo
-------
Data Retrieved from TRI
F-ll
-------
Ait Rclcaicj Reported m 1RI for 2I22V6
DBS
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
64
65
66
67
68
69
70
Facility Name
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
CONSOLIDATED PH
ATOTECH USA INC
ATOTECH USA INC
ATOTECH USA INC
ATOTECH USA INC
ATOTECH USA INC.
ATOTECH USA INC.
ATOTECH USA INC.
Address
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
61 10 ROBINWOOD
6110 ROBINWOOD
61 10 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
61 10 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
61 10 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
6110 ROBINWOOD
61 10 ROBINWOOD
61 10 ROBINWOOD
61 10 ROBINWOOD
6110 ROBINWOOD
61 10 ROBINWOOD
61 10 ROBINWOOD
51 10 ROBINWOOD
5110 ROBINWOOD
;1 10 ROBINWOOD
5110 ROBINWOOD
5110 ROBINWOOD
5110 ROBINWOOD
I900CHESAPEAK
900 CHESAPEAK
900 CHESAPEAK
900CHESAPEAK
900 CHESAPEAK
900CHESAPEAK
900 CHESAPEAK
TRI Facility ID
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL61I8R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
2t225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21225KNSCL6118R
21226MTCHM1900C
21226MTCHM1900C
21226MTCHM1900C
21226MTCHM1900C
21226MTCHM1900C
21226MTCHM1900C
21226MTCHM1900C
Slate
County
FIPS
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24003
24510
24510
24510
24510
24510
24510
24510
ZIP
###*
«##*
*##*
tana
**#*
«#*#
«w*
**#*
**#*
****
t###
«#*»
IHUIM
*#**
****
tttM9
*##*
WWjl
ttJMHt
«**
fWMW
*##*
ttffit
n#9M
###*
moat
*#*»
*##*
UUtM
mat*
X££4
**#•
*#**
*#**
###£
**#*
*##*
####
##*#
*##*
*##*
loot*
UtoWW
ttttti
f*W
Preterre
Latitude
39.208
39.208
39.208
39.208
39 208
39208
39208
39208
39208
39208
39 208
39208
39 208
39208
39208
39 208
39208
39208
39.208
39208
39208
39208
39208
39.208
39.208
39.208
39.208
39208
39208
39208
39208
39208
39.208
39208
39.208
39.208
39208
39.208
39239
39.239
39239
39239
39.239
39239
39239
Preferred
Longitud
766269
766269
766269
76 6269
766269
766269
76 6269
76 6269
76 6269
76.6269
76 6269
76 6269
76 6269
766269
76 6269
76 6269
766269
76 6269
766269
76 6269
76.6269
76 6269
76 6269
76 6269
766269
766269
76 6269
76 6269
76 6269
766269
766269
76 6269
76 6269
76 6269
76.6269
76 6269
76.6269
76 6269
76 5724
76 5724
76 5724
76.5724
765724
765724
76 5724
Standarc
Industry
Code
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2834
2819
2819
2819
2819
2819
28191
2819
\
Chemical Name
METHANOL
DICHLOROMETHANE
METHYL ISOBUTYL KETONE
HYDROCHLORIC ACID
METHANOL
ACETONITRILE
DICHLOROMETHANE
METHYL ISOBUTYL KETONE
HYDROCHLORIC ACID
METHANOL
ACETONITRILE
DICHLOROMETHANE
METHYL ISOBUTYL KETONE
HYDROCHLORIC ACID
AMMONIA
SULFURIC ACID
METHANOL
ACETONITRILE
DICHLOROMETHANE
METHYL ISOBUTYL KETONE
HYDROCHLORIC ACID
AMMONIA
SULFURIC ACID
METHANOL
ACETONITRILE
DICHLOROMETHANE
METHYL ISOBUTYL KETONE
TOLUENE
HYDROCHLORIC ACID
AMMONIA
SULFURIC ACID
METHANOL
ACETONITRILE
DICHLOROMETHANE
HYDROCHLORIC ACID
DICHLOROMETHANE
TOLUENE
HYDROCHLORIC ACID
ANTIMONY
BARIUM
CHROMIUM
ZINC (FUME OR DUST)
NITRIC ACID
NITRIC ACID
ANTIMONY COMPOUNDS
Yrf
Carcinoge
\
Y
J
v|
\
Y
Y
N
N
N
Y
Y
N
J
N
N
N
Y
Y
N
N
N
N
N
Y
Y
N
N
N
N
N
N
Y
Y
N
Y
N
N
N
N
Y
N
N
N
N
Total Air
Emission
1500
7000
500
500
808
9
5200
21
3
500
500
2900
500
500
500
0
500
500
2900
500
500
500
10
35
210
2900
450
60
600
125
0
300
42
1769
360
1815
255
250
7079
500
500
500
500
500
4434
Fugitive
Air
Emission
750
4000
250
250
740
6
2700
4
1
250
250
1500
250
250
250
0
250
250
1500
250
250
250
5
20
10
1500
200
10
300
100
0
200
2
915
180
250
5
0
250
250
250
250
250
25
25
tack Air
imission
750
3000
250
250
68
3
2500
17
2
250
250
1400
250
250
250
0
250
250
1400
250
250
250
5
15
200
1400
250
50
300
25
0
100
40
854
180
1565
250
250
6829
250
250
250
25
25
418
eporlin
ear
1987
1987
1987
1987
1988
1988
1988
1988
1986
1989
1989
1989
1989
1989
1989
1989
1990
1990
1990
1990
1990
1990
1990
1991
1991
1991
1991
1991
1991
1991
1991
1992
1992
1992
1992
199'
199-
1994
1987
1987
1987
1987
1987
1988
1988
Page I
-------
Data Retrieved from FINDS
F-13
-------
TABI F 1 FACF1 ITY DAT A IN 7.IP (>1">H 21225
Cfff
1
J
4
e;
6
7
K
9
0
2
5
i
N
9
0
21
22
2J
24
2?
26
27
2H
29
JO
Jl
32
i)
14
J?
.16
37
JK
39
411
41
42
4.1
44
4?
46
47
48
41
*fi
51
52
«J
n>
MI)| i')85378876
1 l|l|>UM:i)i>94xx4095
MD|)i>M37:74fi
t II >l>9K518fi94S
Mill >')85 187621)
MDD985407H7
MDI W3R2fl43
MMII9H5198577
Mill 1981 106111
Ml 11)981917178
i n>iio:24iii>45
Mill MRS 169162
Mill 1985198940
Ml >l 198517:218
Ml II >02241271d
Ml 11 1985 18081,4
Ml iD'1908 io<>6i
MHD98II7I260S
Mill 198071 1818
MI)D9809|S:05
Ml 11 1000604090
Ml 11198 54 11061
Ml 11 1985180484
Ml 11)985402932
Ml 11 19851701,14
Ml H 1985370477
Mill 1001062197
Ml i|)i)02487585
MDU98M88875
Ml )|)9XI lot.073
MDI 1985181116
Ml 11 104567475]
Ml )|)|I07996044
Mill 19(15392014
MDU9S2567240
Ml l|1001085974
MI)|)98IIO:79I
M|)|«)K26I6020
MDP018942884
MDD9854I2857
1.11)1)00106:811
MDDOI0089365
(1)0981044647
11)1)985408640
IPI 1985396944
(|)|)074i;:0968
||)|)()51940|77
II 1|)9805544D6
f!>(>98 5 179205
B
A( Mr. ANIISKJDCORP
Al HAN ENGINE POWER INC
AMERICAN DISH SVCE OF- RAJ. TIMOR
AMERICAN IANK IRAMSPoRI INC
AMOCO S7I5-TANKS
AMOCO «848M-TANKS
AN SAM METALS CORP
II &G noi)V WORKS INC
BAJ TRUCK 4 EQUIP REF'AIR SRR
I1AI I1MORE IIARBOR TUNNEL
IIROOKJ.VN MEDIC ALCTR
HRooKI YN MOTORS (NC
HROOKI VN PARK JR HIGH
BROOKLYN SVCCTR
BROWNS I10DY & FENDhR
I1R()\WS HONDA CITY HONDA
CAJ'ITOI. AWARDS & PRO
CIII-.M1CAI SPECIALTIES MFC.; CORP
CIIESAJ'EAKE i POTOMAC TEI.E CO
CIIESAPF.AFO; * POTOMAC JhLE CO
CLEAN AMERICA INC
COASTAL TANK LINES FNC
C( )NCRE 1 E TRANSPORT INC
COPANOS COJOIFN D
CRANE KIRBY INC
CROWN SI A
CROWN STA
l)ENTOcn)ECITEMCO
DREVER CO lit AT TREATING Dl V
EXECUTIVE RADIATOR SERVICE INC
FORTMC1D-.NRY TlfNNEL
UISCHEI. MACHINE CO
L\RF3ISON WALKER REFRACTORIES B
OSEITIJ HOCK FNC
IOUSING AinilORJTY OF BAI.TIK)OR£ CHY
IA CONST CORP . BRCXlFaW
ARMS STEEL * LUMBER CO INC
K&T AlfTOBODY
ki T BODY SHOP
KANASCOLTI)
KVWASCO LTD
KNIPP & CO
S LEES BODY SHOP INC
ORD BALTIMORE CLEANERS
1ARSIIALL BODY SHOP
LA.SSIC AUTO BODY SPECIALIST
1AI LACK INC
IODLRN IRASHMOVA1 INC
IORI.OCK PETROLEUM EQUIPMENT SVC INC
EENANBIISINSS FORMS
r.
FACn.lTY ADDRESS
1417SIIANOVERST
1401 CHERRY HJLLRD
4701 BELLE GROVE RD BI.DG 0
6150 ORDNANCE POINT RD
101 W PATAPSCO AVE
5502 RJTCIITE IFWY
1 026 E PATAPSCO AVE
1 SEWARD AVE
601 W PATAPSCO AVE
FRANKFURST AVE NEAR CHJLL)S ST
3721 POTEESTSTE 1
2900 S HANOVER ST
200 HAMMONS LANE
900 E PATAPSCO AVE
5I6PONTIAC AVE
58IORTTCIOEFTWY
1835SFIANOVERST
PATAPSCO AVE
1401 N RITCHIE FFWY
206-21 1 FRANKLEST
1300CFTJLDST
527CFIESAPEAJCE AVE
200 FRANKFURST AVE
61 10 ROB1NWOOD PARK
600 W PATAPSCO AVE
5701 RJTCFDE HWY
3550POTEEST
3417 S HANOVER ST
6201 ROBFNWOODRD
6n38BELLEGROVERJD
KEITTI A LELAND AVEN1 ES
5511 MAGIEST
1200 PATAPSCO AVE
5500 BELLE GROVE RD
4|40IOTHSTRf.ET
200 FRANKFURST AVE
1437 NINTH STREET
1 HAMMONDS LN
5826 RJTCFOE IFWY
6II8ROBINWOODRFJ
61 10 ROBFNWOODRD
3401 S HANOVER ST
6033 BELLE GROVE RD
5634 RJTCFTTE HWY
3570 2ND ST
3570 2ND ST
4801 BELLE GROVE RD
901 BALTIC AVE
4700 BELLE GROVE RD BI.DG 16
3917SHANOVERST
0
CITY
BALTIMORE
FIALIIMORE
BALTIMORE
CURTIS BAY
HA1 TTMORE
B,\I,TTMORJE
BALTIMORE
BROOKLYN PARK
BAL TTMORE
BALTIMORE
BALTIMORE
BALTIMORE
DAI TTMORE
BAI.HMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
GLENBURNTE
BALTIMORE
BALTIMORE
BALTIMORE
BAI.TTMORE
BAF.TFMORE
BAI.TTMORJ5
BALTIMORE
BALTIMORE
BALTIMORE
BROOKLYN PARK
BAI.TFMOR£
BALTIMORE
BROOKLYN
BALTIMORE
BALTU1ORE
BALTIMORE
BROOKLYN
BAI.TTMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BAJ TTMORE
RALTTMORJE
BRCXJF1YN
BALTIMORE
BAI.TIMORE
BAL1TMORE
BALTIMORE
£
STATE
MD
MD
MD
NtD
MD
NOD
MD
N«D
MD
Kffi
MD
MD
MD
MD
MD
MD
MD
MD
NO
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
NfD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
E?
CODE
21225
21225
21225
21225
21225
21225
21225
212252229
21225
212251636
21225
21225
21225
21225
21225
21225
21225
21225
21225
71225
2122.5
21225
21225
212251600
21225
212251605
21225
21225
21225
21225
21225
21225
21225
21225
21225
21225
21225
21225
21225
21225
21225
212253835
21225
21225
21225
21225
21225
21225
21225
21225
21225
IEOION
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
1
3
3
3
3
3
3
COUNTY CODE
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
COUNTY NAMH
\NNE ARUNDEL
ANNE ARUNDEL
ANNE ARirNDEI
ANNE ARUNDEI
ANNE ARUNDFJ
ANNE ARUN11F1.
ANNE ARUNHFL
ANNE ARUNDEL
ANNE ARUNDI-T
ANNE ARIWDEI
ANNE ARUNDEI
ANNE ARUNDEL
ANNE ARUNDFL
ANNE ARilNbTl.
ANNE ARUNDEI
ANNE ARUNDEL
ANNE ARUNDEI.
ANNE ARUNDEL
ANNE ARirNDEI
ANNE ARirNDEI.
ANNE ARUNDEI.
ANNE ARUNI >EL
ANNE ARUNDFL
ANNE ARIJNDEI.
ANNE ARUNDEF
ANNE ARUNDEL
ANNE ARUNDEL
ANNE ARirNDEL
ANNE ARUNDEL
ANNE ARUNDEL
ANNE ARirNDFI.
ANNE ARUNDEL
ANNE ARUNDEI.
ANNE ARUNDI-L
ANNE ARUNDhl.
ANNE ARUNI1FL
ANNE ARUNDEL
ANNE ARUNDEL
ANNE ARUNDEI.
ANNE ARUNDEI.
ANNE ARIJNDEI
ANNE ARUNDEL
ANNE ARUNDEL
ANNE ARUNDEL
ANNE ARUNDEI.
ANNE ARUNDhl.
ANNE ARUNDEL
ANNE ARUNI)I-.I.
ANNE ARUNDEI
ANNE AKUNDEI
ANNE ARlrNDEL
OOF Po
YSTF.MS
I
I
|
j
2
j"
I
"|
-i
]
2
2
-,
2
~" i
2
1
2
1
1
2
2
1
2
~
~"
"
2
3
2
2
1
1
2
1
|
1
001- I'll
ECORD.".
i
I
I
3
|
]
3
|
1
1
|
1
2
i
i
3
~2
2
1
1
3
3
4
3
\
I
i
" 4
4
1
1
1
1
1
2
1
5
3
3
1
i
1
1
1
1
-------
Dun and Bradstreet Facility Data
F-15
-------
fl
G
i \m i : i Am MIPS \vmiPorFNTi\i FNVIORNMT NIAJ RF.I F<\SFS INZIP(CODF 21226
1
i
4
*
6
7
8
1
10
11
12
13
14
If
16
17
IS
19
20
21
22
23
24
25
26
27
2H
29
30
.11
32
33
.14
j<
36
37
IS
39
40
41
42
41
44
45
46
47
4S
49
EI.A\V ARF. CORP
IIP AMI- RICA INC
CAMPHFI 1. BODY
CFNIRAI OIL ASPHALT CORP
CHEMEIALS INCORPORATED
CHEMETAI.S INCORPORATED
CHESAPEAKE COATING INC
CHESAPEAKE CORPORA I ION
CIIFSAJ'LAKF. CORPORATION
CHESAPEAKE PRINTING* MAILING
CHESAPEAKE PR INI ING & MAILING
CHESAPEAKE VENTURE. INC
CHEVRON USA INC
en Y OF HAL IIMORF:
Cl FAN AMERICA INC
col ONIAJ PIPELINE COMPANY
COI OR PRFLUDF. INC
col IIMHK FLEET SERVICE INC
•OMII >G US INC
•OMMTRCIAI. EOUIPNU-NI COMPANY
•OMMFRCIAL TESTING* ENGRCiCO
•ON TRACT MATERIALS PROCESSING
'ORROS1ON REPAIR * SERVICES
OT 1MAN COMPANY
D
' -NNINGION AVE
MFNTIII L AVF.
FOR! SMAI LW(X)DRD
Gl IDDFNRD
CURTIS AVF
PFNNINGION AVF
ARUHDFLCOVE AVE
ARUNDF1 COVE AVE
PENNINGTON AVE
PENNINOTON AVE
PENNINGFON AVE
ONST
CHESAPEAKE AVE
BENTIILL AVE
BRANDON SHORES RD
CARBON AVE
PI IT MAN RD
MENH1LL AVE
FORT SMALL WOOD RD
QUANTINERD
CHEMICAL RD
• 3815 LEO ST
EPATAPSCOAVE
HAWKINS POINT RD
FORT ARMISTEADRD
EPATAPSCOAVE
NORTIIT1RIDGE AVE
.EO ST
ASIATIC AVE
MTTMAN RD
PITTMANRD
BENTIILL AVE
Pint.nNRD
PITIMANRD
E ORDNANCE RD
E ORDNANCE RDSTE 40
CHESAPEAKE AVE
CHESAPEAKE AVE
ASIATIC AVE
"IIILDSST
FAIRFIEI DRD
ENERGY PKY
COVE AVF.
PITTMAN RD
NORT10JR1DGEAVE
.NERGYPKYSTE 1002
BENTIILL AVE
ENNINGTON AVE
CARBIDE RD
C
BAI. IIMORF
BALTIMORE.
BAI TIMORE
BAI TIMORE
BAI TIMORE
BAI TIMORE
BAI IIMORF
BAI TIMORE
BALTIMORE
BALTIMORE
BALIIMORE
BALIIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BAI TIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BAI. TIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BAI TIMORE
B«J TIMORE
BALIIMORE
BAI TIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALTIMORE
BALIIMORE
B«J TIMORE
BALTIMORE
BALTIMORE
BALTIMORE
D
STATE
MD
MD
MD
Kfl)
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
KfD
MD
MD
NflJ
MD
MD
MD
MD
MD
MD
MD
MD
MD
MD
NO
NtD
MD
MD
Kffi
MD
MD
MD
MD
MD
KID
MD
MD
MD
E
ZIP
212261620
212261434
212261802
212261803
212261402
212261423
212261703
212261703
212261619
212261618
212261617
21226
212261012
212261434
212261746
212261007
212261721
212261434
212261717
21226
212261622
21226
212261158
212261610
212261803
212261523
212261536
212261521
212261506
212261792
212261792
212261434
212261721
212261721
212261741
212261741
212261051
212261013
212261507
212261016
212261516
212261733
212261607
212261792
212261536
212261798
212261433
212261430
212261704
E
TEL
4103557626
4103556821
4103543900
4103541613
4103543971
4103548001
4107895506
4107895506
4103550700
4103555112
4103551869
4103553700
4103554700
4107875300
4103554455
4107890436
4103542424
4103600359
4103550196
4103557788
4103542532
4103563641
4103541884
4103550624
4103557200
4103557621
4103556363
4107898800
4107898800
4103554200
4107899400
4107899440
4107688757
4107660008
4106251370
4105763795
4103962800
4103540751
4103558155
4104377725
4107968795
4107898800
4104850399
4102558688
4103541600
4103552838
4I063I7R9I
G
SIC
CODE
3731
3471
5012
5169
3442
1761
7538
7534
5171
4225
7692
7692
2899
3471
4911
5093
3826
3842
5541
4953
4953
7699
7538
5561
2813
2813
5171
7538
2951
2819
2819
3479
2653
5093
2752
2759
4491
5032
4953
4953
4613
2844
7538
2819
7538
8731
5169
1799
4491
II
BUSINESS DESCRIPTION
SHIPBLDINGREPAIRN
PLATING POLISHING
AUTOOTHRMTR VHCLS
CHEM ALLD PRDTS N
MTt DOORS SASH TR
RFNGSDNGSHTMTL
GNRL ATMTVE RPR SHP
TIRERFRDNGRPRSH
PETRO BLK STNS TMN
GENRL WRJISO STORAO
WELDING REPAIR
WELDING REPAIR
CHEM PRPRTNS NEC
PLATING POLISHING
ELECTRIC SERVICES
SCRAP WASTE MTRLS
ANALYTCL INSTRMNTS
SRGCL APPL SUPPLS
GASI.NE SVC STATIONS
RE FUSE SYSTEMS
REFUSE SYSTEMS
REPAIR SVCS NEC
GNRL ATMTVE RTR SHP
RCRTNL VHCLE DLRS
INDUSTRIAL GASES
INDUSTRIAL GASES
PETRO BLK STNS TMN
GNRL ATMTVE RPR SHP
ASPHPVNGMX1 BLCK
IND INORG CHEM NEC
TND INORG CHEM NEC
MTL CTNG ALLD SVCS
CRRGTD SLD FBR BXS
SCRAP WASTE MTRLS
COMMRCL PRTNG LITII
COMMRCL PRINTNG ME
MARINE CARGO HNDLNG
BRCK STN RLTD MTR
RE FUSE SYSTEMS
REFUSE SYSTEMS
REND PETRO PPELINES
TOILET PREPARATIONS
GNRL ATMTVE RPR SHP
IND INORG CHEM NEC
GNRL ATMTVE RPR SHP
COMMRCL PHYS RSRCH
CHEM ALLD PRDJS N
SPCL TRD CNTRS,NE(^>
MARINE CARGO ItfJbLNG
i f
SALES
$252.411
$500,000
$150,000
$300000
$15,000
$4,561.641
$200,000
$190,000
$0
$0
$2,422.735
$0
$0
$750.000
$0
$4,100,000
$0
$183.224
$1,100,000
$0
$0
$812,510
$140,000
$137.132
$0
$0
$0
$140.000
$0
$60,000.000
$0
$1.700,000
$0
$0
$1.000,000
$0
$1,500,000
$0
$0
$4.000.000
$0
$0
$140,000
$85.000,000
$0
$0
$1.800.000
$1,400.1)00
$390.000
J
TOTAL
EMPLOYEES
4
8
2
3
5
45
4
3
0
0
30
0
0
15
0
27
0
6
10
0
0
7
3
1
0
0
0
3
0
285
0
9
0
0
17
20
2
0
0
39
0
0
3
430
0
0
II
22
4
K
NAME OF OWNER
1SEPIIG SMIIH
FORGE SCHUMANN IR
IICHAELAGRO
OUIS ROMM
A V JD COMB
1IINL Al VEY
JINKY ALVF.Y
FRED FLINT
:HRIS moos
HARLES KEN HARRIS
RONALD PELLE HER
HARLES MISERFMIJINO
W LOWMAN
DAVID SIMON
OHNSENATORE
OIINSGANGA
EGGYOT-JEIL
IMOTIIY S MURDOCK
""
11KE ROGERS
ATRICK SMITH
AL KELLER
RANDY ROBERTS
ROBERT A SHANK
RICHARD L MULIIOI LAND
RICHARD MULHOLLANU
1EORGE SCHUMANN )R
ROBERTS ARGABRIGHT
;REG ISAAC
GARY RANK IN
GARY RANKIN
W1LLIAMKROH
I AMES D FREDERICK
BURIONUSKLAR
CURTIS COLICIIER
TOM TILLER
STAN GARLAND
RICHARD Ml Jl.llol LAM)
Bll LY LGREFR
DR EDWIN AL HERS
CHETZANESKI
DAVID II MURDOCK
n:i::r..\ xi s
-------
MDE Facility Data
F-17
-------
MDE Facility Data
Facilin Name:
Street Address:
Citv:
Countrv:
Zip:
Contact Name:
Contact Ph.#s:
IMDE Permit #
SlCcode(s):
Type of Business:
Latitude:*
Longitude*:
Census Tract #
Emission Rate:
Process/Equipment Generating Emission (MDE Code)
Emission Control Equipment Present? (Yes/no)
Total Amounts Emitted in Prior Years
Enforcement/Compliance History (RTK.net)
Data Element for any onsite monitoring capacity and information:
Source Category:
StackHeight:
Stack: Elevation of Stack Base (meters)
Stack Exit Velocity:
Stack Inner Diameter:
Stack Exit Temp.:
Stack: Height of Adj. Bldg.
Stack: Width of Adj. Bldg.
Stack:Length of Adj. Bldg.
Fugitive:Elevation of Area Source
Fugiti\e:Effective Emission Height of Area Source
Fugitive: Width of Square Area Source
FMC Corp.
1701 East Papatsco Axe.
Baltimore
21226
Michael Altman
(410)354-5706
24-00073
2879
Agricultural Chemical Mfg.
250500
24-0073-2-0209
Yes
Incinerator (Hg3 Waste)
52ft.
25 fps
40 in.
120°F
* LTM: Zone 1 8: Easting 3636; Northing 4343.3
F-18
-------
APPENDIX G
Example of Database Columns Developed for the
Community Pilot Project Air Emissions Database
-------
Column
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
0
R
s
T
U
V
\v
AB
AD
AE
AF
AG
AN
AC)
AP
AQ
AR
AS
Column Header
Name
Street Address
City
County
ZIP Code
Contact Name
Contact Phone Number
SlCCode(s)
Type of Business
Latitude
Longitude
MDE Coordinate East
MDE Coordinate North
Census Tract Number
TRI Facility ID Number
MDE Permit Number
Number of Employees
Pollutant Name (*=on-site monitoring)
CAS Number
Carcinogen (Y/N)
TRI Chemical (X=Yes)
OSHA Chemical (X=Yes)
Cancer Slope Factor (QSTAR) (mg/kg/day)
Reference Concentration (RfC) (mg/m"')
Inhalation Cancer Slope Factor mg/kg-day
Inhalation Reference Dose (RfD) mg/kg-da>
Reference Dose (RfD) mg/kg/day
Total Air Emissions (tons/yr) (1994) TRI
Total Air Emissions (Ibs/yr) (1994) TRI
Total Air Emissions (tons/yr) (1995) MDE
Total Air Emissions (Ibs'yr) (1995) MDE
Total Air Emissions (tons/yr) TAP
Total Air Emissions (Ibs/vr) TAP
G-l
-------
Column
AT
AU
AV
AW
AY
AZ
BC
BD
BE
BF
BG
Bl
BO
Column Header
Total Air Emissions (tons.-\r) TAP*
Tola! Air Emissions (lbs/\ r) TAP*
Maximum Total Air Emissions (Ibs/vr)
Potential Dose (mg kg-da\ ) Turner - Vent
Risk (dose'SF) (based on Turner)
Hazard (dose/RfD) (based on Turner)
Stack Emissions (Ibs/yr) (1995)
Stack Emissions ( 1995) tons/yr
Fugitive Emissions (Ibs/yr) ( 1995)
Fugitive Emissions ( 1995) tons/yr
Monitored Concentrations (ppb) Avg. (Max.) 1996
Primary Data Source
Enforcement Compliance History (RTKNET)
G-2
-------
Emission Inventory Database (Few Pages From the Database as example)
Name
FMC Agricultural Chemic al
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural C hemn al
FMC Agricultural Chemical
F MC Agric ultural C hemiral
FMC Agricultural Chemical
I FMC Agric ultutal C hemic al
I FMC Agricultural Chemic al
1FMC Agricultural Chemical
FMC Agricultural Chemic al
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemic al
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agric ultural C hemical
FMC Agricultural Chemical
FMc, Agricultural Chemical
jFMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agncultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMC Agncultural Chemical
FMC Agricultural Chemical
FMC Agricultural Chemical
FMc Agricultural Chemical
Rhone-Poulenc Specialty Chemicals
I Rhone-Poulenc Specialty Chemicals
JRhone-Poulenc Specialty Chemicals
Rhone Poulenc Specialty Chemicals
Rhone-Poulenc Specialty Chemicals
(Rhone-Poulenc Specialty Chemicals
I Rhone Poulenc Specialty Chemicals
i Rhone-Poulenc Specialty Chemicals
Rhone-Poulenc Specialty Chemicals
Rhone-Poulenc Specialty Chemicals
Rhone-Poulenc Specialty Chemicals
Rhone Poulenc Specialty Chemicals
Rhone-Poulenc Specialty Chemicals
Rhone-Poulenc Specialty Chemicals
Ato
ech U S A
ech U S A
ech USA
ech USA
ech U S A
Strpet Address
1 701 Fasl Patapsco Ave
1701 Easl Patapsco Ave
1701 Easl Patapsco Ave
1701 Easl Patapsco Ave
1701 Easl Patapsco Ave
1701 East Patapsco Ave
1701 East Palapsco Ave
1701 Easl Palapsco Ave
1701 Easl Palapsco Ave
1701 Easl Patapsco Ave
1701 East Patapsco Ave
1701 East Patapsco Ave
1701 East Patapsro Ave
1701 East Palapsco Ave
1701 East Palapsco Ave
1701 Easl Patapsco Ave
1701 Easl Patapsco Ave
1701 East Patapsco Ave
1 701 East Patapsco Ave
1 701 East Palapsco Ave
1701 Easl Palapsco Ave
1701 Easl Patapsco Ave
1 701 Easl Patapsco Ave
1701 Easl Patapsco Ave
1701 East Patapsco Ave
1701 East Patapsco Ave
1701 East Patapsco Ave
1 701 East Patapsco Ave
1 701 East Patapsco Ave
1 701 East Patapsco Ave
1701 East Patapsco Ave
1 701 East Patapsco Ave
1701 East Patapsco Ave
1701 East Patapsco Ave
1 701 East Patapsco Ave
1701 Easl Patapsco Ave
1701 East Patapsco Ave
1701 Easl Patapsco Ave
1701 East Patapsco Ave
3440 Fairtield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairtield Rd
3440 Fairtield Rd
3440 Fairtield Rd
3440 Fairfield Rd
3440 Fairfield Rd
3440 Fairfield Rd
1900 Chesapeake Ave
1900 Chesapeake Ave
1900 Chesapeake Ave
1900 Chesapeake Ave
1900 Chesapeake Ave
City
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Ballimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
Baltimore
County
Zip
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226
21226-1012
21226-1012
21226-1012
21226-1012
21226-1012
Contact Name
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
John Sanderson
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Itshak Rosner
Ronald Pelletier
Ronald Peiletier
Ronald Pelletier
Ronald Pelletier
Ronald Pelletier
Contact Ph its
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355 6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410-355-6400
410 355-6400
410-355-6400
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-2600
410-355-3700
410-355-3700
410-355-3700
410-355-3700
410-355-3700
SIC Code
2869.5191
2869 5191
2869 5191
2869. 5191
2869 5191
2869.5191
2869.5191
2869 5191
2869 5191
2869. 5191
2869, 5191
2869.5191
2869.5191
2869.5191
2869.5191
2869 5191
2869.5191
2869 5191
2869, 5191
2869 5191
2869.5191
2869.5191
2869.5191
2869 5191
2869 5191
2869.5191
2869.5191
2869 5191
2869. 5191
2869. 5191
2869 5191
2869. 5191
2869 5191
2869.5191
2869. 5191
2869. 5191
2869 5191
2869 5191
2869. 5191
2869
2869
2869
2869
2869
2869
2869
2869
2869
2869
2869
2869
2869
2869
2819. 2899
2819 2899
2819. 2899
2819. 2899
2819 2899
Type of Business
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Farm Supplies
Latitude
392321
392321
392321
392321
392321
39 2321
392321
392321
392321
39 2321
392321
39 2321
39 2321
392321
392321
39 2321
39 2321
39 2321
392321
392321
39 2321
392321
392321
392321
392321
392321
392321
392321
392321
392321
39 2321
39 2321
39 2321
39 2321
39 2321
39 2321
39 2321
39 2321
392321
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2358
39 2386
39 2386
39 2386
39 2386
39 2386
Lonqitude
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
765843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
" 76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5843
76 5785
76 5785
76 5785
765785
76 5785
76 5785
76 5785
76 5785
765785
76 5785
76 5785
76 5785
765785
76 5785
76 5724
76 5724
76 5724
76 5724
765724
MDE
Coord
East
919
'919
919
919
919
919
919
919
919
" 919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
919
913
913
913
913
913
MDE
Coord
North
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
509
512
512
512
512
512
Census
Tract It
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250500
250600
250600
250600
250600
250600
EXAMPLE) 123
-------
Emission Invpntnry Database (Few Pages From the Database as example)
AC UK S T U V W AB AD AE AF
1
INI F .irilily IU H
1
21. ','i,l MI r Rl 70 If
?12..'h( MI i Rl 7IJIE
4 2I22U Mi I IU 70 IE
', 2l22r,l Mi CHI /OIF
6 21221,1 Mi , R1701E
1
M
2I22M Mi (. RI701E
212261 Ml TR170H-
' 21 2261 Ml Rl "f)IC
10 2122131 MI . HI 70IF
1 1 2 1 22bT Mi ( R 170 IF
r; J2122W MI i RKOIL
n
1-1
IS
16
17
18
19
?0
21
22
23
2 12261" M< LR1701E
2122i;l Mi. LR1701E
212,'bl Mi LRI701E
2122151 Ml i.HI 70 IE
2122BF MI.CRI701E
2122bl MLLRI701E
212261 ML I R170IE
212261 MLLR1701E
212261 Mi. CR170IE
2I226F MCI RI70IE
212261 MI LR1701E
24 j 21.? .WML t R1701E
25
26
27
28
29
30
31
32
33
34
35
36
2l226f Ml CR1701E
21226FMLLR170IE
2I226F ML I. R 170 IE
21226FMI.CR1701E
21226FMCCR1701E
2t226FMLfR1701E
21226FM<~L R1701E
21226FML CR1701E
21226F MCLRI701E
21226FMCCR1701E
21226FMCCR1701E
21226FMCCR1701E
37 J21226FMCLR1701E
38
39
-to
41
42
43
44
45
46
21226FMLCR1701E
21226FMCCR1701E
21226FMCCR1701E
21226LCLL 3440F
21226LCLC 3440F
2I226LLLC 3440F
21226LCLC 3440F
21226LCLC 3440F
21226LCLC 3440F
47 21226LCLC 3440F
48
49
21226LCLC 3440F
21226LCLC 3440F
50 21226LCLC 3440F
51 J21226LCLC 3440F
52
53
21226LLLC 3440F
21226LCLC 3440F
54 2I226LCLC3440F
55
56 21226MTCHM1900C
S7 ,'21226MTLHM1900C
58 1
59 !21226MrCHM1900C
MF1F
Pern, it ti
Numbpf nf
F rnpluypes
28r
Pollulanl Namp ('"on Slip monitoring)
Acetir HCIL) niplhyl ester melhyl ac elate
2H5 Af Plonp
2115
285
Acc'lunitnlp
Arnrnoriicl
285 Rpri^prip
285 Bpii70yl pproxide
285 Larbon munoxidp (CO)
285 f arhon Iptiachloiidp
285
2Hr,
285
2H5
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
285
105
105
105
105
105
105
105
105
105
105
105
105
105
105
L fllPC hoi
( hlnnnp
Lhlorolorm (Inchlorornethanp)
Lhlororriplhanp (Mplhyl chlondp)
Cyanide S rornpounds
fcthannl (Elhyl alcohol)
Elhion
Ftliylben?ene
Ethylene gfycol
Meplane
Hydrochloric acid (Hydrogen chloride)
Hydrogen cyanide (Hydrocyanic acid)
Hydrogen peroxide
sopropyl alcohol
Methanol
Melhyl isobutyl ketone
Melhylene bromide
Nitrogen oxides (NOx)
Nilropheriol-2 (Nitrophenol o-)
Participates
Phenol
Phosphoric acid
Pyndine
Sodium cyanide (Na(CN))
Sodium hydroxide
Sodium sulfate (solution)
Sulfur oxides (SOx)
Sullunc acid
Toluene
Volatile Organic Compounds (VOCs)
Xylene
Acrylic acid
Ammonia
Carbon monoxide (CO)
Dioxane (1 4-)
Ethylene oxide
Formaldehyde
Glycol ethers
Hydrochloric acid (Hydrogen chloride)
Isopropyl alcohol
Melhanol
Mitrogen oxides (NOx)
SulFur oxides (SOx)
Sulfunc acid
Volatile Organic Compounds (VOCs)
Carbon monoxide (CO)
Chromium & compounds
Nitric and
Nitrogen oxides (NOx)
Z nc & compounds (fume or dust)
f ASS
79209
67641
75058
7664417
71432
94360
630080
56235
120809
7782505
67663
74873
57125
64175
563122
100414
107211
142825
7647010
74908
7722841
67630
67561
108101
74953
0-01-1
88755
0 01-2
108952
7664382
110861
143339
1310732
7757826
0-01-3
7664939
108883
0-01-4
1330207
79107
7664417
630080
123911
75218
50000
0 00-5
7647010
67630
67561
0-01-1
0-01-3
7664939
0-01-4
630080
7440473
7697372
0-01-1
7440666
Carcinogen (Y/N) [
For source see
column AJ|
N(R)
Y-lnhal (R)
N(S)
Y-lnhal (R)
Y-lnhal (R)
N(R)
Y-lnhal (R)
Y-lnhal (R)
N(R)(S)
N(R)
N(S)
N(R)(S)
N(R)(S)
N(R)(S)
Y(R)
N(S)
N(R)
N(R)(S)
N (R) (S)
N(S)
Y(R)
Y(R)(S)
N(S)
N(R){S)
N(S)
Y-lnhal (R)
N(S)
N(R)(S)
TRI
Chemical
dehsted
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
OSHA-
X
X
X
X
X
X
Cancer Slope Factor
QSTAR (qT) (mg/kg-
day)-1
0
0
0
0
0029
0
0
0 13
0
0
00061
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0011
1 02
0045
0
0
0
0
d
0
0
Reference
Concentration
(RfC) mq/m3
0
0
005
0 1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
002
0003
0
0
0
008
0
b
6
001
0
0
0
0
b
04
0
0001
0 1
0
0
0
0
002
0
0
0
0
0 0005
0
Inhalation
Cancer Slope
Factor (mg/kg-
day)-1
0029
00525
0 0805
00063
035
00455
420E*001
Inhalalion Reference
Dose (RFD) mg/kg
day
00143
00286
000171
0000571
0286
000571
0 000857
00229
na
na ]
0 00286
na
0114
na
0 000286
00286
na
000571
na
na
na
0000000571
na
EX.AMPLE1 123
-------
Emission Inventory Dalabase (Few Pages From Ihe Database as example)
AO AN AO AP AQ AR AS AT AU AV . .... AW AY
Reference Dose
(RID) mq/kq/day
1
0 1
0006
0
0003
0
0
00007
0
0 1
001
0
002
0
00005
0 1
2
0
0
002
0
0
05
008
001
na
0
na
06
0
0001
004
0
0
na
0
02
na
2
05
0
0
0
0
02
na
0
0
05
na
na
6
na
0
007
0
na
03
Total Air Emissions
(tons/yr) (1994) TRI
0 1785
0 1945
00915
00055
108075
0091
00005
00375
00215
22835
34335
03575
0543
00005
1 0545
02955
0013
0043
0 1475
0015
0005
0 1205
00025
00005
00005
00005
Total Air Emissions
(It.s/yi) (1994) TRI
357
389
183
11
21615
182
1
75
43
45670
6867
715
1086
i
2109
591
26
86
295
30
10
241
5
1
1
1
Total Air
Emissions (tons/
yr) (1995) MDE
1467
74 77
778
8382
2354
036
'i 64
018
525
012
889
Total Air
Emissions (Ibs/
yr) (1995)
MDE
29340
149540
15560
167640
47080
720
3~280
360
10500
240
17780
Total Air
Emissions (tons/
VOP) TAP
02001
1 086
2 185
00131
01752
" 00013
08935
03257
01156
00902
23392
00357
12598
"0 8864
" 1 6114
1 855
353904
0 027
1 1 6876
1 2106
0203
00558
0 0486
" 0 643
00159
0004
78141
4924
0014
0014
0029
00665
003
Total Air Emissions
(Ibs/yr) (?) TAP
4002
2172
4370
262
3504
26
1787
651 4
231 2
1804
46784
71 4
25196
17728
3222 8
3710
707808
54
23375 2
24212
406
1116
972
1286
31 8
8
156282
9848
28
28
58
133
60
.
Total Air Emissions
(tons/yr) (?) TAP'
0 00004
000004
0 00004
" " dbo004
0 00004
Total Air Emissions
(Ibs/yr) (?) TAP'
008
008
0 08
008
008
Maximum
Total Air
Emissions (Ibs/
yr)
4002
~~ ~ 2172
" 4370
389
3504
26
29340
1787
651 4
21615
1804
4678 4
71 4
25196
008
17728
3222 8
3710
707808
008
008
54
233752
2421 2
1086
149540
406
" 15560
0 08
111 6
97 2
008
1286
31 8
"167640
8
156282
47080
9848
28
86
~720
58
295
30
10
241
133
60
3280
360
5
10500
240
1
1
17780
1
Potential Dose (mg/
kg-day) Turner -
Vent
0 0003055095794
00016580879723
00033360241431
0 0002969595862
0 0002674926452
00000019848199
0 022397928686
00013641819551
00004972737132
00165007235361
00001377159623
00035714543137
0 0000545062068
0019234431192
00000000610714
00013533417851
00024602605511
00028321852565
05403351434015
00000000610714
00000000610714
00000412231816
00178444465788
00018483253216
00008290439861
0 114157677427
0 0003099372545
00118783834477
00000000610714
'00000851945754
00000742017269
00000000610714
00009817224366
0 0000242758736
0 1279750771958
0000006107138
00119304467993
0 035940507244
00075178869018
00000213749831
00000656517337
00005496424217
0 0000442767506
00002252007145
00000229017676
0 0000076339225
00001839775328
00001015311696
00000458035351
0 0025039265879
00002748212109
00000038169613
00080156186504
00001832141406
0 0000007633923
0 0000007633923
0013573114248
0 0000007633923
Risk (dose'SF)
(based on
Turner)
OOOE+000
0 OOE+000
OOOE+000
0 OOE+000
7 76E-006
0 OOE+000
OOOE»000
7 16E-005
OOOE+000
OOOE + 000
1 11E-005
2 25E-005
OOOE + 000
OOOE + 000
0 OOE+000
0 OOE+000
OOOE+000
OOOE+000
OOOE+000
0 OOE+000
OOOE + 000
OOOE+000
OOOE + 000
0 OOE+000
OOOE+000
0 OOE+000
OOOE + 000
0 OOE+000
0 OOE+000
0 ODE + 000
0 OOE+000
OOOE+000
bOOE+OOO
0 OOE+000
OOOE + 000
0 OOE+000
OOOE + 000
OOOE+000
OOOE+000
0 OOE+000
0 OOE+000
0 OOE+000
OOOE+000
7 88E-005
1 04E-006
OOOE+000
OOOE+000
OOOE + 000
OOOE + 000
OOOE + 000
OOOE+000
OOOE+000
OOOE + 000
OOOE + 000
321E 005
OOOE'OOO
OOOE + 000
OOOE + 000
EXAMPLE1 123
-------
,ion Inventory Database (Few Pages From the Database as example)
A.' HI Fit) OF RF BG Bl BQ
II, 17,11(1 (dosPl
RID) (h.r.cil in
IIHHPI)
r RR
f RR
(1 23 CHH401G
o oio38'J20?3
n rjt;42H4.i75
ERR
ERR
? 7891 KI2/i4l
ERR
ERR
ERR
ERR
ERR
ERR
ERR
nnn.17 ii%4i
fcRR
ERR
94 629622312
00000712618
ERR
ERR
ERR
0 OHO 7 128961
ERR
ERR
ERR
ERR
ERR
0 0297883131
ERR
ERR
ERR
ERR
FRR
ERR
0 1046530421
ERR
ERR
0074737703
00022955152
ERR
ERR
ERR
ERR
ERR
00322202334
ERR
ERR
ERR
ERR
ERR
ERR
ERR
1 3369391461
ERR
ERR
ERR
'.I. iik j Sl.itk
F missions
(Ihs/yr)
(19V,)
Emissions
(1995)
Inns/yf
Fugitive
Emissions
(Ibs/yi) (1'iqS)
riigili,p
Emissions
( 199r>) Inns/
Yr
MuniloiPd
Emissions
(ppb) Avg
(Max ) 1996
Primary Data Source
TAP
TAP
TAP
TAP
TAP
TAP
1995 MDE Emissions Statempnt
TAP
TAP
TAP
TAP
TAP
TAP
TAP
TAP (sported value=0 0000)
TAP
TAP
TAP
TAP
TAP (reported value = 0 0000)
TAP (reported value=0 0000)
TAP
TAP
TAP
1994 TRI Form R
1995 MDE Emissions Statement
TAP
1995 MDE Emissions Statement
TAP (reported value=0 0000)
TAP
TAP
TAP (reported value=0 0000)
TAP
TAP
1995 MDE Emissions Statemenl
TAP
TAP
1995 MDE Emissions Statement
TAP
TAP
TAP
1995 MDE Emissions Statement
TAP
1994 TRI FormR
1994 TRI FormR
1994 TRI Form R
1994 TRI Form R "
TAP
TAP
1995 MDE Emissions Statement
1995 MDE Emissions Statement
1994 TRI Form R
1995 MDE Emissions Statement
1995 MDE Emissions Statement
1994 TRI FormR
1994 TRI Form R
1995 MDE Emissions Statement
1994 TRI Form R
Enforcement
Compliance
History (RTKnel)
EXJVMPLE1 123
-------
APPENDIX H
Facilities Modeled for Secondary Screen
-------
Appendix H - Baltimore Facilities and Pollutants Modeled for Secondary Screen
Facility Name
Amoco Oil Co.
Baltimore City Composting
Baltimore Resco
Bethlehem Steel
BGE- Brandon Shores
BGE- Wagner Station
Ba\\va\ Terminal
Pollutant Name
Toluene
Benzene
Ammonia
Benzene
Carbon tetrachloride
Toluene
Vinyl chloride
Arsenic
Cadmium
Chromium
Formaldehyde
Hydrogen chloride
Hydrogen fluoride
Mercury
Cadmium
Chromium
Lead
Manganese
Carbon monoxide
Nitrogen oxides
Sulfur oxides
Arsenic
Cadmium
Chromium
Lead
Mercury
Nickel
Hydrogen Chloride
Hydrogen Fluoride
Dioxins and Furans
Carbon monoxide
Nitrogen oxides
Sulfur oxides
Arsenic
Cadmium
Chromium
Lead
Mercun
Nickel
Hydrogen Chloride
Hvdrogen Fluoride
Dioxins and Furans
Benzene
Emission
Rate
(Ib/yr)
9.746
4.000
206.660
7.156*
2,820
8.436
5.720
630
703
3.333
4.355
6.126.000
77.651
15.837
551
848
958
20.124
2.114.980
45.987.400
93.865.380
1.443
178
909
1.468
290
978
4.200.000
5.200.000
0.0062
816.140
27.567.540
35.993.240
462
64
294
477
91
2.167
1.300.000
160.000
0.0019
1.120
H-l
-------
Facility Name
Brooklyn Sen ice Station
Chemetals Corp.
Citgo Station
CONDEA-VistaChem.
FMC Agricultural Chemical
Grace Da\ ison
Hohelmann Port Ser
J.S. Lee's Bod> Shop. Inc.
Med Net/MedX Inc.
Mobil Oil(Maritank)
Norns Farm Landfill
Phoenix Services
Pon international
Ouebecor Printing
SCM Chem - Millennium
Millennium (cont.)
Shell Oil Terminal
MOTIVA
! ' S Coast Guard
1 ' S G% psun
Pollutant Name
Toluene
Ammonia
Hydrochloric acid
Manganese
Sulfuric acid
Benzene
Toluene
Benzene
Hydrochloric acid
Carbon tetrachlonde
Chloromethane
Hydrochloric acid
Toluene
Ammonia
Chromium
Molybdenum trioxide
Nitrogen oxides (NO\)
Sulfuric acid
Stoddard solvent
Toluene
Dioxins & Furans
Hydrochloric acid
Benzene
Toluene
1 .2-Dichloropropane
Benzene
Methvl chloride
Methylene chloride
Vinyl chloride
Dioxins & Furans
Hydrochloric acid
Hydrogen sulfide
Toluene
Carbon monoxide (CO)
Carbonyl sulfide
Sulfur oxides (SOx)
Sulfuric acid
Benzene
X\ lenes (m-.o-.p-)
Beni-.ene
To! -ne
Toluene
Chromium
Emission
Rate
(lb/yr)
141
59.568
23.172
61.661
3.621
122
186
3.000
21.000
1.787
4.678
707.808
15.628
290.000
122
1.180
237.780
3.000
30.380
263
0.00000199
42.300
882
5.291
2.365
1.051
2.365
11.388
2.628
0.00282
91.016
2.640
3.250.000
19.028.940
1.562.400
2.306.640
39.900
1.400
1.500
130
199
8.054
26 2
* This number was determined to be erroneous. However, the emissions
Did not impact the Partnership neighborhoods.
H-2
-------
APPENDIX I
Results of Secondary Screening for Target Toxics
-------
Table I-1. Results of Screening for Target Toxics
Kstimated air levels (in micrograms per cubic meter of air (/^g/m1), based on modeling of facility
emissions in four South Baltimore neighborhoods plus the location with the highest estimated air levels.
Exposure guidelines and monitoring results data are provided for comparison. The concentration as a percentage of the applicable comparison
guideline is shown below the concentration (in parentheses).
Chemical
Ammonia
Arsenic
Hen/cue
1,3-Buladicne
Cadmium
Carbon monoxide
Carbon tetrachloride
Carhonvl sulficle
Screening Comparison
Concentrations (standards and
guidelines)
100/ig/m'
(!•:(' A guideline- IRIS RIC)
Carcinogenic
0 00041 ^g/rn1
Non-carcinogenic
I.I //g/m1
0.22 /(g/m'1
(EPA guideline - derived from
IRIS)
0.0064 //g/m'
(EPA guideline, derived from
IRIS)
0.00099 Aig/m1
10,000 /jg/m' as 8-hour average
([•:PA NAAQS standard)
O.I2A
-------
Table l-l. Results of Screening for Target Toxics (continued)
( 'hcmical
( 'hritmium
(as 1 lc\a\ nlenl Ibrni)
C'liroinium
(as 1 rivalent form)
l.2-l)icliloropropane
Dioxin
(2.3.7.8-K 1)0)
I'ormaldcliN de
1 1\ droclilorie acid
Ihdrogcn fluoride
IKdrogcn sullldc
Lead
Manganese
Succmng C'omparison
Concentrations (standards and
guidelines)
I) 00015 /ig'm'
(1 I'A guideline - derived from
IRIS)
0 0021 //g/m'
0 092 //g/m'
0 000000054 //g/m'
(L(]iiivalent to 5.4\l()*)
(LPA guideline - derived from
IILAS1)
0.14 //g/m'
21 / Hill
0.0001
(67%)
(35%)
0 0002
(••1%)
0 0000000004I9
(4 19x10")
('--!%)
0.00089
(<"!%)
1.51
(7%)
(22%)
0 09554
<
-------
Table 1-1. Results of Screening for Target Toxics (continued)
Chemical
Mercury
Methyl chloride
Melhylcnc chloride
MoKbdenuin trioxide
Nickel
Nitrogen oxides
Sloddard solvent
Sulfur oxides
Sulfuric acid
1 olucne
Screening Comparison
( oncentrations (standards and
guidelines)
0.31 /jg/m1
0.99 ^g/m'
(KPA guideline - derived from
III-ASI)
>•""*"'
IR^g/m1
(l-ll'A guideline - derived from
IRIS reference dose for
molybdenum)
73 /jg/m1
3.700 ^g/m' as annual mean not I
be exceeded
(I-PA NAAQS standard)
5,250Mg/m'
(Maryland standard - ambient air
level derived from ACGIII
TLV/IOO)
80 A'g/rn1 as annual mean
(l-:i'A NAAQS standard)
lO^g/m1
(Maryland standard - ambient air
level derived from ACGIII
TLV/IOO)
420 //g/m1
IHI'A guideline -IRIS RfC)
Neighborhood Concentrations (from modeling)
Cherry Hill
0.00325
(1%)
0.001
(-1%)
0.00081
0.0002
(
-------
Table 1-1. Results of Screening for Target Toxics (continued)
( 'hcmical
Vim 1 chloride
\\ lene
.Screening Comparison
( 'oiicenlralions (standards and
guidelines)
0.021 //g/m'
(1-,1'A guideline - derived I'rom
III-'ASI)
7.300 /%)
27.17/ig/m'
(UA%)
-------
Table 1-2. Evaluation of Combined Exposures for Substances Known to Cause Respiratory Effects
Concentrations and percentages of guidelines for each substance, along with a sum of the individual percentages to provide an estimate of the
possible impact from simultaneous exposures
Chemical
Screening Comparison
Concentrations (standards and
guidelines)
Neighborhood Concentrations (from modeling)
Cherry Hill
Rrookl\ n/
Hrooklyn Park
Curtis Ray
Wagners Point
Point with Highest
Concentration
Stalc-operatcc
monitoring
station results
RESPIRATORY EFFECTS
Ammonia
Formaldehyde
1 lydrochloric acid
llxdrogen fluoride
Nitrogen dioxide
Sulfur dioxide
Sulfuric acid
lOO^g/m1
(I.I'A guideline- IRIS KIT)
1 20^/111'
20 /ig/m1
(LPA guideline- IRIS RfC)
7 /vg/m1
(Man land Standard - Acceptable
Amhient Level)
25 ,/g/m'
TLV/IOO
3,700 /jg/m1 as annual mean not t
he exceeded
(LPA NAAQS standard)
80 //g/m' as annual mean
([•PA NAAQS standard)
10/yg/m1
(Maryland standard - ambient air
level derived from ACGIII
TLV/IOO)
Total Respirator.' Fffects Using higher limit
Using lower limit
0.073
(<•!%)
0.00089
(^1%)
1.51
(7%)
(22%)
0.09554
(<1%)
) 1 43
(<1%)
2.48
(3%)
0.004
(<\%)
;
0.00034
(
-------
APPENDIX J
Partnership Air Committee Report
-------
November 9,1999 Draft
Report from the Partnership Air Committee
1. What is this report?
For the past several years the Air Committee of the Community Environmental Partnership (CEP)
has been working to get a better understanding of the air quality in south Baltimore and northern
Anne Arundel County. The first step of this effort has now been completed. This report
summarizes the work that has been done and the steps planned for the future. Supporting data is
available in a full technical report.
2. What is the Air Quality Committee of the South Baltimore Community Environmental
Partnership?
The Air Committee is one of five committees organized by the CEP to get a better understanding
of the environment and economy in south Baltimore and northern Anne Arundel County. A list of
Air Committee members and their affiliation is attached to this report. The job of the Air
Committee is to collect information on the quality of the air in the Partnership neighborhoods and
make suggestions for how the air quality can be improved. Air quality ranked first in the list of
concerns voted on at the July, 1996, community meeting that began the Partnership. This high
interest in air quality is an indication of the widespread community concern about the health of the
community in the Partnership neighborhoods and the possible contribution of the environment to
those health concerns. The CEP Air Committee has about twenty members including local
residents, industry managers and officials from the U.S. EPA, the Maryland Department of
Environment (MDE), Baltimore City, Anne Arundel County, and The Johns Hopkins University.
All committee members are committed to work together to improve the air quality in the
Partnership neighborhoods. The committee has met regularly since its inception. Meetings have
been held at the Partnership Office since its opening in March, 1997. Meetings are open to the
public.
3. What aspects of air quality were studied?
Many community members believe that chronic health problems in Partnership neighborhoods,
especially certain types of cancers, may be attributed to outdoor air pollutants released by the
factories, utilities, waste facilities and vehicles in and around the Partnership area. Since certain
chronic health problems may be caused by long-term exposure to these pollutants, the committee
decided to start its work by studying annual ambient concentrations of outdoor air pollutants in
Partnership neighborhoods from these sources.
It is important to note, that there are three other aspects of air quality that may have significant
chronic health effects that were not a part of this study: ground level ozone, which is a byproduct
of the reaction of certain chemicals with sunlight; small particulate matter, especially from diesel
exhaust; and short term peak concentrations of certain chemicals that may contribute to health
J-l
-------
problems such as asthma. The Air Committee has recommended further work in these areas to
evaluate their potential effects on the community. See recommendations in Question 12 below
4. What chemicals are present in outdoor air and where do they come from?
The committee reviewed emission reports from over 125 facilities in and around the Partnership
area and air monitoring reports from MDE. 175 chemicals released to, or measured in, the
outdoor air in the Partnership neighborhoods were identified during this review. The chemicals
originate from a wide variety of sources, including factories, utilities, waste facilities and vehicles.
5. How were the chemicals in outdoor air evaluated?
Given the resources available, the Air Committee decided to use a screening method that could
provide the community with information to help identify chemicals that might be a concern. The
Committee screening method used two kinds of available information. First, the Committee used
available information on air pollutant concentrations from the state air monitoring station located
in Fairfield, north of the FMC facility. This is the only air monitoring station located in the
Partnership neighborhoods that gathers information on air pollutants. This monitoring station
takes air samples every day. Records of the ambient concentrations of 41 different chemicals are
available from these samples. The second kind of information used for the screening analysis
was the information on air emissions reported by facilities to the EPAns Toxics Release Inventory
and to the Maryland Department of Environment under the state permitting program. The
Committee used air dispersion computer modeling to estimate the concentrations of air toxics in
Partnership neighborhoods that result from these permitted emissions. At the request of the
community for information on the possible effects from multiple sources, the Committee used
current EPA modeling methodology to combine all the sources for each chemical to get an
estimate of the aggregate exposure levels in each Partnership neighborhood. For example, there
are twenty stationary sources of benzene in and around the Partnership neighborhoods. These
sources were combined in the modeling program to provide an estimate of the total benzene
concentration in each neighborhood.
Both the concentrations measured at the monitoring station and the estimated concentrations from
area facilities were compared to "screening values" chosen by the committee. A screening value
is an air concentration that the committee is confident does not pose a significant human health
risk. The committee used U.S. EPA and MDE health effects information to select a screening
value for each chemical. Screening values can be based on either cancer risks or risks from other
toxic effects. All of the screening values used in this study are based on cancer risk because these
offered the most protective values (i.e. the lowest corresponding concentrations) for the subject
chemicals. For each chemical, the Committee chose a screening value that corresponds to an
increased cancer risk of one in one million under the assumed conditions of exposure. This is
consistent with risk management goals used by various EPA programs, including the ambient air
program. For pollutants that may cause cancer. EPA programs use a risk management range of
one in one million to one in ten thousand under their reasonable maximum exposure scenarios to
guide their decision-making. The screening values used in this analysis are not enforceable
standards and were used for committee screening purposes only. Enforceable State standards are
J-2
-------
applied to individual facilities and are based on an increased cancer risk of one in one hundred
thousand outside the facility. The Committee screening values are more conservative than the
State standards and cannot be directly compared. Once screening values were chosen, the
Committee compared them to the measured and modeled concentrations in the Partnership
neighborhoods. All neighborhood chemical air concentrations found above the screening values
are identified in this report. They are discussed in question seven below.
6. What community questions can and cannot be answered with this information?
It is important to recognize that there are limitations to the information that this kind of screening
analysis can provide. Most significantly, a study of this kind cannot tell the community what the
actual risks from these chemicals are in each of the Partnership neighborhoods. This is true
because much of the screening is based on estimates and not on actual measurements, because the
actual measurements were taken only in Fairfield and not in all of the Partnership neighborhoods,
and because no study was made of the people living in our neighborhoods to get a better idea of
their actual exposure. This would take into consideration things like the time spent in the
neighborhood, ages, time spent outdoors, etc. The Air Committee decided that collecting all the
information necessary for a more detailed risk analysis would be both expensive and time
consuming and may not add that much to the community's ability to set priorities. (See section
Question 11 for more background on the limitations of the method used.)
Finally, the Air Committee air screening exercise does not provide sufficient information to
explain current or future incidences of cancer and other diseases in the Partnership neighborhoods.
There are many contributing factors affecting community health that were not considered in this
study. These include things like lifestyle, diet, smoking, access to medical care, and heredity. In
addition, the Air Committee looked only at current levels of chemicals, not at exposures that
occurred ten or twenty years ago when ambient air pollutant emissions and ambient concentrations
were higher than today's levels. Current incidences of cancer may be caused, in part, by these past
exposures. It is also important to recognize that the analysis in this report is based on the
assumption that reduced emissions are associated with reduced risk.
Despite the limitations, the screening analysis provides valuable information to the community.
The analysis identifies and inventories all the significant commercial, industrial, and waste
treatment and disposal facility sources of chemicals in outdoor air in the Partnership
neighborhoods. It provides the best estimates available on the types and amounts of chemicals in
outdoor air in Partnership neighborhoods, including estimates of the aggregate concentrations of
the same chemical from multiple sources. The analysis compares the estimated and measured
concentrations to health values and provides enough risk information to help the community set
priorities and chart an effective course of action for improving air quality. It also helps to
establish a community air quality baseline that can be used to evaluate future progress and identify
potential concerns with new sources. The analysis also allows the Partnership to compare the
levels in its neighborhoods to other urban, suburban and rural neighborhoods where the same
chemicals have been measured. In sum. the Air Committee study was designed to identify aspects
of air quality where prevention efforts would be most effective in contributing to improving the
future health of the community. This information must be combined with a much broader effort to
J-3
-------
address all the factors contributing to community illness to effectively address community health
concerns.
7. What were the results of the evaluation?
Of the 175 chemicals analyzed in the effort, only four exceeded the Committee Screening Values.
Concentrations of benzene, 1,3-butadiene, carbon tetrachloride, and methyl chloride measured at
the monitoring station were found to exceed the Committee screening values. The benzene level
in Wagner's Point modeled from the emissions from area industries and other facilities was also
above the Committee screening value. All other measured concentrations and concentrations
modeled in Partnership neighborhoods were found to be below the Committee screening values.
Except for the benzene level in Wagner's Point, the emissions from all the industries and other
facilities in and around the Partnership neighborhoods resulted in modeled concentrations that
were below the committee screening values. Vehicles and other mobile sources are a significant
source for benzene and 1,3-butadiene. Carbon tetrachloride is primarily due to past uses
(ToxFAQs,Sept. 1995); methyl chloride concentrations are primarily due to past uses and natural
sources (OAQPS,,Dec.l994). Additional details on the sources and other information on each of
the four chemicals found to be above committee screening levels are given below:
As explained in Question 5 above, the Air Committee chose a screening value that corresponds to
an increased cancer risk of one in one million under the assumed conditions of exposure. The
total risk level for the four chemicals found to be above the Air Committee screening value
correspond to an increased cancer risk of 6 in one hundred thousand. While this risk estimation is
not a characterization of actual health risks, it can provide a relative indication of the potential
health concerns associated with these chemicals. EPA programs use a risk management range of
one in one million to one in ten thousand under their reasonable maximum exposure scenarios to
guide their decision-making for carcinogens.
Benzene: The Committee determined that most of the benzene in outdoor air originates from cars
and other mobile sources. Other sources of benzene in Partnership neighborhoods include a
chemical plant in Fairfield. petroleum product terminals, and gas stations. Except for the
Wagner's Point neighborhood, the modeled benzene concentrations from the industrial and
commercial facilities were below committee screening levels. In Wagner's Point, mobile sources
and bulk petroleum facilities account for most of the benzene. Benzene exposure can cause a
distinct form of leukemia, known as acute myelogenous leukemia, and is classified by the EPA as
a known human carcinogen (Group A). For more details on the health effects of benzene, see the
attached fact sheet (ToxFAQs, Apr. 1993).
1.3-Butadiene: In the Baltimore area, this chemical is emitted almost entirely by cars and other
mobile sources. At the time of the analysis, 1,3-butadiene is classified as a probable human
carcinogen by the U.S. EPA (Group B2; data in humans exist but are considered inadequate alone;
data from rat and mouse studies are sufficient to indicate a carcinogenic potential in humans). For
more details on the health effects of 1.3-butadiene. see the attached fact sheet (ToxFAQs,
Sept.1995).
J-4
-------
Carbon Tetrachloride: Monitored levels at Fairfield are due almost entirely to past emissions of
this chemical which is now being phased out of use due to its effects on the earth's stratospheric
ozone layer. Levels found at Fairfield are typical of urban areas where it has been measured.
Long term exposure to carbon tetrachloride can produce liver and kidney damage. Carbon
tetrachloride has been classified by the EPA as a probable human carcinogen (Group B2; data in
humans exist but are considered inadequate alone; data from rat, mouse and hamster studies are
sufficient to indicate a carcinogenic potential in humans) For more details on the health effects of
carbon tetrachloride, see the attached fact sheet (ToxFAQs, Sept. 1995).
Methyl Chloride: Also known as chloromethane, monitored levels in Fairfield are primarily the
result of natural processes in the environment. Methyl chloride is present in air all over the world.
Levels at Fairfield are similar to levels in other U.S. cities where air monitoring for methyl
chloride has occurred. Long-term exposure to methyl chloride may produce liver, kidney, spleen
and brain damage. Methyl chloride has been classified by the EPA as a possible human
carcinogen (Group C), but has not been associated with any particular form of cancer in humans.
For more details on the health effects of methyl chloride, see the attached fact sheet
(OAQPS,Dec. 1994).
8. How does outdoor air quality in Partnership neighborhoods compare with other
Baltimore locations and with other urban communities?
Benzene, 1,3-butadiene, carbon tetrachloride and methyl chloride are regional air pollutants and
have been measured by MDE at six monitoring locations within the Baltimore region. The bar
charts in Figures 1 through 4 on page 8 compare the concentration levels for each of the chemicals
at the six monitoring locations. The Air Committee screening levels for the chemicals are also
shown on the bar charts. Data from these locations are intended for use in an overall
characterization of these chemicals in the broader Baltimore area rather than to support detailed
assessment of specific neighborhoods. Interpretation of data from individual sites is complicated
by differences in meteorological conditions that can affect readings as well as by siting that may
have been chosen to complement other monitoring sites.
As illustrated in these bar charts, levels of 1,3-butadiene measured at Fairfield are the second
lowest of the six locations. The levels of benzene, carbon tetrachloride, and methyl chloride
measured at Fairfield are higher than those measured at the other locations. Given the small
difference in the concentration levels measured at the different monitoring stations and given the
uncertainties in the risk calculations, the risk levels associated with the measured concentrations at
the six monitoring stations are too close to differentiate. In other words, the risks of the four
chemicals may be essentially the same throughout the Baltimore area.
The committee also compared the level of these chemicals to levels in other cities where similar
measurements were made. Measured levels in these cities are similar to Baltimore levels. Levels
of carbon tetrachloride and methyl chloride in Baltimore were below the levels estimated for cities
in the Agency for Toxic Substances and Disease Registry fact sheets for these chemicals. See
Table 1 on page 9 for details of these comparisons.
J-5
-------
9. Is air pollution in Partnership neighborhoods getting better or worse?
Emission and monitoring reports reviewed by the Committee demonstrate that air quality in
Partnership neighborhoods and the surrounding region has been improving for several years
(Maryland Department of Environment 1992-1996; USEPA, Dec. 10, 1996). This improvement is true
for dozens of toxic chemicals as well as for common air pollutants for which there are national
ambient standards. Information reviewed included emission reports submitted to the state and to
EPA and air monitoring reports prepared by the state.
10. What can we do if we want to improve our air quality?
There are several ways that community members can work to improve air quality. First, the
Partnership Air Committee would like to continue its work to learn more about the other parts of our
air that are not included in this report. One of the Committee recommendations listed below calls
for more work on odors, truck exhaust and truck routing. Volunteers are needed to work on these
areas. Also, it is possible to address the emissions of the four chemicals identified in this report.
Since most of these emissions are associated with mobile sources, that means getting involved in the
national debate on controlling vehicle emissions. EPA is now working on these issues and
community input will be crucial to the decisions made. The Partnership Air Committee plans to
invite representatives from EPA and MDE to speak to the committee and then the committee will
develop a plan to make the community s voice heard on these issues. The Committee also
recommends further work with the local companies that are contributing to the levels of benzene in
our air and to help them find ways to further reduce their emissions. Committee volunteers are
needed to work on this as well.
11. What are the limits of the analysis used?
The committee utilized a conservative (i.e. one that is designed to overestimate concentrations and
risks) screening method to reach these results. The resulting risk calculations do not correspond to
actual exposure scenarios nor do they represent estimates of risk to actual persons. The analysis
simply provides a systematic approach and a common standard to compare the relative importance
of the measured or modeled chemical concentrations.
It is important to point out key limitations of the study that are due in part to the current state of the
science used. 1) The study addresses only cancer risks to a hypothetical adult population resulting
from inhalation exposure to specific individual chemicals. The study does not address other routes
of exposure or possible toxicologic interactions among the multiple chemicals to which people are
exposed. 2) The study does not specifically address sensitive segments of the population such as
children. 3) The screening values used by the Committee were based on cancer effects. Because of
incomplete information on the potential toxic effects of some chemicals, there may be other health
effects, such as birth defects and endocrine disruption, that could lead to lower screening levels
Significant scientific uncertainty and controversy exists around the issue of very low dose effects for
endpoints like endocrine disruption. Please see section 6 for additional explanation of the limits of
the studv.
J-6
-------
12. What does the air quality committee recommend?
The Committee believes that the community should encourage the continual reduction of emissions,
especially through pollution prevention measures. In addition, the Committee has proposed four
recommendations. Community volunteers are needed to work on these recommendations.
1) Work with local facilities to reduce benzene emissions especially through more pollution
prevention
2) Encourage appropriate actions to reduce odors. See attached page with results of the
Committee Odor Survey listing known sources of odors in community
3) Encourage appropriate action to reduce diesel truck exhaust through means such as the
enforcement of current truck traffic restrictions, better diesel motor maintenance for vehicles
regularly using local roads, and the rerouting of truck traffic. See attached page with listing
of diesel vehicles regularly using Partnership streets.
4) Develop ways to educate the community about the impacts of indoor air pollution
13. What else is being done to improve air quality?
On the local level, additional monitoring and air sampling work to get more accurate information on
exposures is now underway. These measurements should add more information to the community's
understanding of local air quality. The Partnership should continue to review data from MDE and
any other local agencies with pertinent air quality information.
On the national level, EPA has proposed an ambitious new schedule for addressing risks from air
toxics in urban areas that would, among other things, set new standards for dozens of categories of
small, stationary sources not targeted under the agency's existing air toxics program. Under the
strategy,"area" sources, such as institutional and commercial boilers, municipal landfills, paint
stripping operations, and sewage treatment works, would face new requirements for cutting air
toxics by 2009. with some rules taking effect as early as 2005 (USEPA,1999). The plan also calls
on the agency to assess emission reductions from mobile sources and determine whether additional
regulations are needed to cut air emissions from these sources. The agency is working to finalize
these rules.
REFERENCES:
Maryland Department of Environment. Ambient Air Monitoring Data for 41 Chemicals from 1992
through 1996.
OAQPS. December 1994. Office of Air Quality Planning and Standards. USEPA. Methyl Chloride
(Chloromethane). 74-87-3. Part II.
J-7
-------
ToxFAQs. April 1993. Benzene. Agency for Toxic Substances and Disease Registry , U.S. Dept. of
Health and Human Services. Public Health Service. Division of Toxicology. Atlanta. GA.. 30/33.
ToxFAQs. September 1995. 1,3-Butadiene. Agency for Toxic Substances and Disease Registry .
U.S. Dept. of Health and Human Services, Public Health Service, Division of Toxicology, Atlanta,
GA., 30333.
ToxFAQs. September 1995. Carbon Tetrachloride. Agency for Toxic Substances and Disease
Registry , U.S. Dept. of Health and Human Services, Public Health Service, Division of
Toxicology, Atlanta, GA., 30333.
USEPA. July 19, 1999. National Air Toxics Program: The Integrated Urban Strategy; Federal
Register Volume 64, No. 137,
USEPA, OPPT. Dec. 10, 1996. TR1 Indicators Report prepared by Steve Hassur: "Preliminary
Analysis for the Baltimore Community Environmental Partnership Air Working Group
-------
1996 Annual Average vs Screening Levels for Priority Chemicals
050
040
P 030
£
o020
o
010
000
Figure 1. 1,3-Butadiene Monitored Concentration VS
Screening Level Concentration
l23-?lH I°-3.!
J022|
[00064^ [0.0064^ [aooe4J^ (apo64_|pP [0.0054^ [0.0064
Fairfield Glen Bumle Essex Downtown Fort McHenry NE Baltimore
I | 1,3-Buladlene H Screening Level
350
300
250
I200
c
§ 1 50
1 00
050
000
Figure 2. Benzene Monitored Concentration VS
Screening Level Concentration
iL-l
Fairtield Glen Bumie Essex Downtown Fort McHenry NE Baltimore
[ | Benzene ^ Screening Level
Figure 3. Carbon Tetrachloride Monitored Concentration VS
Screening Level Concentration
1 00
080
2 060
i 040
000
Fairtield Glen Bumie Essex Downtown Fort McHenry NE Baltimore
[ Carbon Tetrachloride f| Screening Level
Figure 4. Methyl Chloride Monitored Concentration VS
Screening Level Concentration
250
100
000 :
Fairtield Glen Bumie Essex Downtown Fort McHenry NE Baltimore
Methyl Chloride ^ Screening Level
-------
TABLE J-I
AVAILABLE MONITORING DATA
Annual Average Concentrations in ug/m3
Hen/enc
Carbon
1 etrachloridc
1 ,3 Butadiene
Methyl Chloride
Baltimore
3.4
0.94
022
2.5
Fremont, CA
4.1
0.48
0.34
NA
Fresno, CA
4.5
0.49
0.43
NA
Los Angeles
7.3
0.50
1.04
NA
ATSDR
NA
0.63 background
1.3 - 3.8 in cities
NA
2. 1 background
6.2 in cities
Louisiana
3.8
1.0
NA
1.31
NYAir Toxics
2.67
I.I
NA
NA
Texas Air Toxics
1.9
0.57
0.91
NA
-------
APPENDIX K
Baltimore Air Dispersion Modeling
-------
Baltimore CEP Short Term Air Modeling- Summary of Model Set-up and
Assumptions Used.
The Industrial Source Complex Short Term (ISCST3) model was run for five different scenarios.
The manner in which the model was set up and the assumptions used were similar for each scenario. In
general the changes between scenarios were limited to differences in the number of sources modeled, the
type of pollutants modeled, pollutant emission rates, and modeling averaging times. Discussed first below,
is the basic model setup and assumptions used common to all five scenarios. This is followed by individual
discussions of the unique aspects for each of the five scenarios.
Model Setup and Assumptions Used (All Scenarios)
Both toxic and criteria air pollutants were modeled over five separate years
(1987,1988,1990,1991,1992) or a subset of those years. All facilities were located in the southeast portion
of the Baltimore Metropolitan area. Receptors locations were the same for all scenarios.
The following lists the model configuration/set-up used.
• Urban dispersion mode
• Flat, simple terrain
• No wet or dry plume depletion and no wet scavenging
• Regulatory default options used
• Assumed sea-level for all source base elevation heights
• Assumed sea-level for all receptors elevations (model assumes all receptors on flat terrain).
• No source grouping
• Calculate average concentrations only, no deposition
• Four discrete receptors*
• Fine cartesian grid with 250m grid spacing (700 receptors)**
• Coarse cartesian grid with 2000m grid spacing (72 receptors)**
• Hourly emission rate assumed to be annual rate divided by 8760 hours per year
• Assume no flagpole receptor heights
• No building downwash
• Surface weather data from Baltimore-Washington International
• Upper air weather data from Sterling VA.
Location of discrete receptors:
(decimal deg.)
* Corners lor coarse grid
(deg min sec)
Cherry Hill
Brooklyn
Wagners Point.
Curtis Bay
Latitude
39 1727.3
39 1736.2
3909 53.3
39 1002.2
Latitude
39.2484
39.2332
39.2303
. ... 39.2250
Longitude
763920.5
7628 129
763909.9
76 28 03 4
Longitude
76.6237
76.6040
76.5689
76.5903
** Corn
(
Latitude Longitude
Corners for fine grid: 391509.2 763750.1
(deg min sec) 391512.7 763339.8
391130.3 763745.0
391133.8 763335.0
Note: The layout of the grid is depicted (with sources locations) in Figure 1 in Source Data Summary and Assumptions.
Source Data Summary and Assumptions
Twenty nine pollutants were modeled, from a total of 36 sources. Table K-l lists all sources and
their location. Note, as will be discussed later, pollutants modeled differed by source and by each of the
five scenarios.
K-l
-------
Table K-l. List of Sources Modeled and their Location
Source Name
Amerada Hess
Amoco Oil Co.
Amoco Station (Patapsco Ave)
Amoco Station (Ritchie Hyw)
Baltimore Resco
BGE- Brandon Shores
BGE- Wagner Station
TOSCO (Bay wax Terminal) (BP)
Bethlehem Steel
Chemetals Corp
CONDEA-VistaChem
FMC Agricultural Chemical
Grace Davison
Hobelmann Port Services
J.S. Lee's Body Shop. Inc.
Valle\ Proteins
Delta Chemical
Crown Station (Ritchie Hyu )
Crown Station (Potee St)
Baltimore City Composting
Brooklyn Service Center
Cilgo Station
Shell Station
I' S. Coast Guard
MOTIVA (Mobil Oil) (Mantank)
Med Net/MedX Inc
Noms Farm Landfill
Phoenix Services
Pon International
Quebecor Printing
SCM Chem - Millennium
MOTIYA (Shell Oil Terminal)
C1TCO (Star Enterprises)
Stratus Petroleum
L.S G\psum
Latitude
39.209800
39.211600
39.238700
39.219800
39.270803
39.189101
39.178500
39.229996
39.219000
39.194901
39.235796
39.231695
39.209298
39.238796
39.217502
39.214600
39.230600
39.217400
39.239400
39.205700
39.234800
39.216800
39.218400
39.204000
39.235503
39.208804
39.288102
39.202197
39.289599
39.171003
39.206098
39.233803
39230001
39.241303
39 204002
Longitude
76.584898
76.584394
76.611800
76.614700
76.630401
76.534601
76.527401
76.572702
76.476594
76.564601
76.578195
76.581602
76.569402
76.571600
76.642904
76.588500
76.566700
76.614400
76.611200
76.560100
76.597600
76.615200
76.614700
76 569700
76.577899
76.569994
76481500
76.557398
76.507399
76.632399
76.545903
76.567701
76.568995
76.576094
76.561201
K-2
-------
The location of all sources listed and receptors are shown in Figure 1. For many of the sources
there was limited information available about the characteristics/nature of the air emission release. As a
result a number of assumptions were used. The following briefly outlines any key characteristics of each
source and briefly describes how each souse was modeled, included any assumptions used.
• Amoco Oil: Twenty identical stacks and one fugitive source. All pollutants modeled out of one stack.
• Baltimore City Composting: Stack parameters given for a composting reactor, and area source
parameters given for a composting area. There was no breakdown of emissions between these two
sources. Characteristics of the composting reactor stack ( low exit velocity, low stack height, ambient
air exit temperature and large stack diameter) seem to indicate that the stack may actually be a ceiling
exhaust fan(s). Assumed all emissions from the composting area, with a release height of 3 meters.
• Baltimore Resco: Straightforward to model. One stack- all pollutants modeled out of this stack..
• Bethlehem Steel: Complicated source- several stacks and fugitive sources listed. For several
pollutants no information was provided on the breakdown of emissions between the fugitive sources
and the stack sources. For modeling purposes assumed all emissions to be from the BOF scrubber
stack.
• BGE Brandon Shores: Source consists of two identical boilers and two similar stacks. Modeled all
pollutants out of one stack.
• BGE Wagner: Source consists of four utility boilers and four separate stacks. Three stacks are
similar, one stack has a significantly higher exhaust temperature. Pollutants were modeled out of one
stack which best represented the three similar stacks.
• TOSCO (BP Terminal) (Bayway Terminal): One stack and five fugitive sources. Modeled as a point
source.
• Brooklyn Service Center (Patapsco Citgo): No stacks, modeled as an area source.
• Chemetals Corp: 15 identical stacks- modeled out of one stack.
• Citgo Station: No stacks- modeled as an area source.
• CONDEA Vista Chemical Company: Boiler and process line emissions indicated. Assumed all
pollutants except NO2 and SO2 are emitted from the process line only. Thus both the process line and
boiler stack data provided were used to model the source.
• FMC Agricultural: Hazardous waste incinerator- modeled as a point source.
• Grace Davison: One stack-modeled as a point source.
• Hobleman Port Services: One stack- modeled as a point source.
• J.S. Lee's Body Shop: One stack- modeled as a point source.
• Med Net/Medx Inc: Medical waste incinerator (one stack)- modeled as a point source
• Mobile Oil Co. Terminal (Maritank): Five stacks, and area source- modeled as a grouped point
source.
• Norris Farm Landfill: Modeled as a point source since there is a stack based venting system.
• Phoenix Services Inc.: Incinerator (one stack)- modeled as a point source.
• Pori International: One stack- modeled as a point source.
• Quebecor Printing: Two sets of stack parameter are listed, one for solvent recovery stacks and one
for ceiling fans. Temperatures are similar for each set, velocity is higher for the fans than for the
stacks. This source was modeled using the solvent recovery stack only, since increased buoyancy due
to higher temperature in the recovery stacks will make up for the lower velocity of the ceiling fans.
This source only operates during a 4-5 mo. block each year thus, hourly emission rates used were
adjusted to reflect the shorter operating period.
Ki
o
-------
» SCM Chemicals (Millennium): Information available for only one stack- all pollutants modeled as a
point through a single stack, including boiler emissions.
. MOTIVA (Shell Oil Terminal): One stack (no area source information)- modeled as a point source.
• Shell Station: Modeled as an area source.
• U.S. Coast Guard: No stack information, assumed toluene emissions from an area source. Estimated
the size of the area source as the typical dimensions of a Coast Guard vessel (assumed painting of a
ship in dry dock). For NO2 emissions used stack parameters used to represent the boiler at Valley
Proteins.
• U.S. Gypsum: One stack-modeled as a point source.
• Amerada Hess: Stack listed appears to represent emissions from fuel/oil storage and loading only
(stack exit temp, was 77 deg. F). This stack was used to model benzene emissions. For NO2 emissions
stack parameters from Amoco Oil were used as they better represent a flare (combustion process).
• CITCO (Star Enterprises( & Stratus Petroleum: No stack/release parameters given. Modeled both
as a point sources using average of stack parameters from other terminals (Maritank), MOTIVA,
Bayway and Shell Oil).
• Amoco & Crown Stations: No stack/release parameters given. Modeled as an area sources using the
average of area source parameters given for the Shell Station, Citgo Station and the Brooklyn Service
Center.
• Valley Proteins: Six stacks listed, 4 boiler stacks and two for a cooker. Assumed all emissions (NO2)
from boiler stacks. Modeled NO2 out of one stack, which represented the average of the four stacks
listed. Characteristics of the four stacks were similar, thus an average was used.
• Delta Chemical: No stack information provided for SO2 emissions. Modeled as a point source usinti
stack parameters from US Gypsum.
Tables showing the stack and area source parameters used for the modeling effort are given in the
Appendix.
K-4
-------
Figure 1. Model Setup- Location of Sources and Receptors
Baltimore County
• l
Legend
; Discrete Receptors
I I Fine Grid Points
•ir Emission Sources
• Coarse Grid Point
— County Boundary
N
S
K-5
-------
Scenario One (Sept 97)
Twenty eight pollutants were modeled from a total of 22 facilities. The averaging periods used
were annual and 24 hours. Table K-2a, shown below lists by facility the pollutants modeled and the
corresponding annual emission rates used.
Table K-2a. Baltimore Facilities and Pollutants Modeled- Scenario One
Facility Name
Amoco Oil Co.
Baltimore Cu\ Composting
Baltimore Resco
Bethlehem Steel
BGE- Brandon Shores
BOH- Vv agner Station
Pollutant Name
Toluene
Benzene
Ammonia
Benzene
Carbon tetrachloride
Toluene
Vinvl chloride
Arsenic
Cadmium
Chromium
Formaldehyde
Hydrogen chloride
H\drogen fluoride
Mercury
Cadmium
Chromium
Lead
Manganese
Carbon monoxide
Nitrogen oxides
Sulfur oxides
Arsenic
Cadmium
Chromium
Lead
Mercun
Nickel
Hydrogen Chloride
Hvdroeen Fluoride
Dioxins and Furans
Carbon monoxide
Nitroaen oxides
Suliur oxides
Arsenic
Cadmium
Emission Rate
(Ib/yr)
9.746
4.000
206.660
7.156*
2.820
8.436
5.720
630
703
3.333
4.355
6.126.000
77.651
15.837
551
848
958
20.124
2.114.980
45.987.400
93.865.380
1.443
178
909
1.468
290
978
4.200.000
5.200.000
0.0062
816.140
27.567.540
35.993.240
462
64
K-6
-------
Facility Name
BGE- Wagner Station (cont.)
TOSCO (Baywax TermmaK (BP))
Brooklyn Service Station
Chemetals Corp.
Citgo Station
CONDEA-Vista Chem.
FMC Agricultural Chemical
Grace Davison
Hobelmann Port Ser.
J.S Lee's Body Shop. Inc.
Mud Net/MedX Inc.
MOTIVA (Mobil Oil) (Mariiank)
Norris Farm Landfill
Phoenix Services
Pori International
Quebecor Printing
SCM Chem.- Millennium
Pollutant Name
Chromium
Lead
Mercury
Nickel
Hydrogen Chloride
Hydrogen Fluoride
Dioxins and Furans
Benzene
Toluene
Ammonia
Hydrochloric acid
Manganese
Sulfuric acid
Benzene
Toluene
Benzene
Hydrochloric acid
Carbon tetrachloride
Chloromethane
Hydrochloric acid
Toluene
Ammonia
Chromium
Molybdenum tnoxide
Nitrogen oxides (NOx)
Sulfuric acid
Stoddard solvent
Toluene
Dioxins &. Furans
Hydrochloric acid
Benzene
Toluene
1 .2-Dichloropropane
Benzene
Methyl chloride
Methylene chloride
Vin> 1 chloride
Dioxins & Furans
Hydrochloric acid
Hydrogen sulfide
Toluene
Carbon monoxide (CO)
Carbonv 1 sulfide
Emission Rate
(Ib/yr)
294
477
91
2.167
1.300.000
160.000
0.0019
1.120
141
59.568
23.172
61.661
3.621
122
186
3.000
21.000
1.787
4.678
707.808
15.628
290.000
122
1.180
237.780
3.000
30.380
263
0.00000199
42.300
882
5.291
2.365
1.051
2.365
11.388
2.628
0.00282
91.016
2.640
3.250.000
19.028.940
1.562.400
K-7
-------
Facility
Name
Millennium (com.)
MOTIVA (Shell Oil Terminal)
Shell Station
L'.S. Coast Guard
U.S Gvpsun
Pollutant Name
Sulfur oxides (SO\)
Sulfurie acid
Benzene
Xvlenes (m-.o-.p-)
Benzene
Toluene
Toluene
1 Chromium
Emission Rate
(Ib/yr)
2.306.640
39.900
1.400
1.500
130
199
8.054
26.2
This number was determined to be erroneous. However, the emissions did not
impact the Partnership neighborhoods.
Scenario Two (Oct 97)
Four pollutants were modeled from a total of twelve facilities. The averaging periods used were
annual, 24 hours and 8 hours. Table K-2b, shown below lists by facility the pollutants modeled and the
corresponding annual emission rates used.
Table K-2b. Baltimore Facilities and Pollutants Modeled- Scenario Two
Facility Name
Baltimore Resco
Bethlehem Steel
BGE- Brandon Shores
BGE- V. aener Station
Chemetals Corp.
CONDLA-VistaChem
FMC Agricultural Chemical
Grace Da\ ison
Med Net'MedX Inc
Norns Farm Landfill
Phoenix Ser\ ices
I S G\ psum
Pollutant Name
Chromium
Hydrogen chloride
Chromium
Manganese
Chromium
Hydrogen Chloride
Chromium
Hydrogen Chloride
Hydrochloric acid
Manganese
Hydrochloric acid
Hvdrochlonc acid
Meth\ 1 chloride
Chromium
Hydrochloric acid
Meth\ 1 chloride
H\drochlonc acid
Chromium
Emission Rate
(Ib/yr)
70
6.126.000
848
20.124
909
4.200.000
294
1.300.000
8.901
16.707
12.000
2.600
150
122
6.250
130
6.952
26.2
-------
Scenario Three (Jan 98)
Three pollutants (benzene and speciated chromium - Cr+3 and Cr+6) were modeled from a total
of twenty two facilities. The averaging periods used were annual, 24 hours and 8 hours. Table K-2c,
shown below lists by facility the pollutants modeled and the corresponding annual emission rates used.
Table K-2c. Baltimore Facilities and Pollutants Modeled- Scenario Three
Facility Name
Amoco Oil Co.
Baltimore City Composting
Baltimore Resco
Bethlehem Steel
BGE- Brandon Shores
BGE- Wagner Station
US Gypsum
Grace Davison
TOSCO (Bay way Terminal)
(BP)
Citgo Station
CONDEA-Vista Chem.
Amoco Station (Ritchie Hwy)
Amoco Station (Patapsco Ave)
Crown Station (Ritchie Hwy)
Crown Station (Potee St)
MOTIVA (Mobil Oil)
(Maritank)
Morris Farm Landfill
Star Enterprises
Stratus Petroleum
MOTIVA (Shell Oil Terminal
Shell Station
Amerada Hess
Pollutant Name
Benzene
Benzene
Cr+3
Cr+6
Cr+3
Emission Rate
(Ib/yr)
80
7,156*
67
3
847.152
Cr+6 0.848
Cr+3
Cr+6
Cr+3
Cr+6
Cr+3
Cr+6
Cr+3
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
633
276
204
90
25.999974
2.6e-5
122
220
61
2.200
67
66
62
44
1.440
16
348
880
480
65
652
This number was determined to be erroneous. However, the emissions did
not impact the Partnership neighborhoods.
K-9
-------
Scenario Four (Jan 98)
Benzene was modeled using the facilities and emission rates listed in Table K-2c. Each facility
was modeled separately for 1990 and 1991 only. This was done to determine each facilities contribution
to the average annual benzene concentration at Wagners Point in 1990 and at the receptor with the highest
overall concentration in 1991. Note, the 1990 average annual benzene concentration at Wagners Point and
the 1991 highest receptor concentration were the highest overall values calculated for all years modeled in
Scenario 3.
K-10
-------
APPENDIX L
Peer Review Comments and Response
-------
Peer Review Comments and Response
This appendix presents the results of the peer review of the November 5, 1999, draft of Air
Committee Technical Report - Risk-Based Air Screening: A Case Study in Baltimore, MD (the Baltimore
Case Study report). The materials presented summarize the main issues raised by the six peer reviewers
and the subsequent activities initiated by EPA and the Baltimore Air Committee to respond to and revise
the document. This appendix includes a brief overview of the scope and purpose of the peer review, the
charge given to the peer reviewers, a list of the peer reviewers, copies of their complete comments, and the
responses to comments. The comments and responses are organized into three major categories: (1) main
issues raised in the peer review, (2) suggestions for improvements to the risk screening methodology that
will be prioritized for future implementation, and (3) suggestions for clarifying the Case Study report.
Scope of Peer Review
The peer review of the draft document Air Committee Technical Report - Risk-Based Air
Screening: A Case Study in Baltimore, MD was conducted to evaluate the technical procedures used in the
risk screening process in Baltimore. Technical experts from the Federal government, academia, and
industry were identified and asked to review the Case Study document and the methodology used.
Although the review focused primarily on the risk screening steps that were developed in the course of the
Baltimore study, the peer reviewers were also asked to evaluate the methodology (emissions inventory,
initial screen, secondary screen, final screen) and the stakeholder participation process and provide
comments on potential improvements. The charge to the peer reviewers is presented on page L-19.
Background
EPA requires that all major scientific and technical products developed for use in decision making
undergo peer review. The policy applies to both internal and external products that support research,
regulatory, or other Agency decisions. The Peer Review Handbook (U.S. EPA, 1998), published under the
auspices of the Science Policy Council (SPC), provides Agency-wide guidance on the process for
conducting peer reviews.
The goal of the Peer Review Policy and this Handbook is to enhance the quality and credibility of
Agency decisions by ensuring that the scientific and technical work products underlying these
decisions receive appropriate levels of peer review by independent scientific and technical experts.
Peer review is intended to uncover any technical problems or unresolved issues in a preliminary
(or draft) work product through the use of independent experts. This information is then used to
revise that draft product so that the final work product will reflect sound technical information and
analyses. Peer review is a process for enhancing a scientific or technical work product so that the
decision or position taken by the Agency, based on that product, has a sound, credible basis (U.S.
EPA, 1998).
L-l
-------
Selection of Peer Reviewers
Candidate experts were identified and selected to conduct the peer review on the basis of their
expertise in the topic areas covered in the document (air quality assessment, emissions modeling, and risk
screening/assessment, etc.). These experts were selected in a manner that ensured objectivity; the peer
reviewers were independent and had no actual or perceived conflict of interests. Six experts, selected from
a wide range of organizations including academia, consulting firms, industry, and government
organizations, are listed below:
Michael A. Callahan
U.S. EPA National Center for Environmental Assessment
Gail Charnley, Ph.D.
HealthRisk Strategies
Douglas Crawford-Brown, Ph.D.
Department of Environmental Science and Engineering
University of North Carolina at Chapel Hill
Amy D. Kyle, Ph.D.
School of Public Health
University of California, Berkeley
Kenneth L. Mitchell, Ph.D.
U.S. EPA Region 4
Ronald E. Wyzga. Sc.D.
Electric Power Research Institute
Peer Review Comments
The full written comments from the six peer reviewers are attached at the end of this appendix.
Response and Reconciliation
The responses to the comments received on the Baltimore Case Study report are organized into
three major categories: (1) main issues raised in the peer review, (2) suggestions for improvements to the
risk screening methodology that will be prioritized for future implementation, and (3) suggestions for
clarifying the Case Study report.
L-2
-------
1. Main Issues Raised by the Peer Reviewers
In summarizing the peer review comments, EPA and the Baltimore Air Committee identified seven
issues to address. These issues were selected because they raise important questions about the Baltimore air
screening exercise and its conclusions. A statement of the seven issues and EPA's responses follow:
Issue 1.1 Two reviewers pointed out that confidence in the ability of the screening process to identify all
chemicals of concern needs to be better demonstrated in the report. As pointed out by the
reviewers, if the screening process is valid, the screening concentrations should decrease or
remain the same as chemicals proceed through the screening. Each subsequent step in the
screening process uses better information to more accurately characterize the concentrations.
Reviewers suggested that the report should explicitly illustrate the decrease in the concentrations
of the chemicals at each step to build confidence in the screening process. Conversely, if the use
of better information results in higher concentrations, then the earlier steps of the screening
process may not be designed to be sufficiently protective and the confidence in the screening
may be misplaced. If the concentrations go up with the chemicals selected for review, then
concentrations for chemicals eliminated might also be higher with better information. The
concentrations may, in fact, go above the screening levels and consequently, the process may
eliminate chemicals that may be of concern. Reviewers point out that, in fact, concentrations
for one of the selected chemicals, benzene, increased in the final step of the process. This higher
concentration needs to be explained or the validity of the process will be in question.
Response It is agreed that there is a need to better demonstrate the validity of the screening by
demonstrating the decrease in concentrations as one advances to later stages of the methodology.
While this is generally true, and could be shown with the examples of chromium, hydrochloric
acid, and manganese, the increase in estimated benzene concentrations from the secondary to the
final screen raises questions. The means by which the validity of the methodology will be
demonstrated will be clarified in the "How To" manual.
The final screen for benzene resulted in the discovery of additional sources of emissions or
increased annual benzene emissions over those used in the secondary screen in the Wagner's
Point area near the modeled receptor location. These increases in emissions account for the
increase in modeled airborne concentrations between the secondary and final screens. This
reflects an error of omission in the construction of the emissions inventory that was discovered
when a closer examination of the facilities emitting the chemical was conducted. Had the
facilities and their correct annual emissions been identified in the initial inventory steps, no
increases in benzene concentrations between secondary and final steps would have been
observed.
The results for benzene were unusual, but not unexplainable . It is unlikely that an increase in
exposure from secondary to final screen would be observed unless additional sources identified
were in close upwind proximity to the receptor. In the case of benzene in Wagner's Point, this
condition was met because the neighborhood is in close proximity to petrochemical storage
facilities all emitting benzene. The updated information resulted in increased emissions and
estimated concentrations.
L-3
-------
Issue 1.2
Response
Issue 1.3
Response
Issue 1.4
The sources of other chemicals eliminated by the screening process were better known, not as
numerous as benzene, and could be confidently eliminated from review. For example, the release
of a "specialty" chemical, i.e., a chemical with a unique use, is only going to be found in
association with the specific industrial process it is used in. If the facility employing that process
is captured in the emissions inventory, it is unlikely that other emissions of the "specialty"
chemical will go unaccounted for. On the other hand, a commodity chemical such as benzene
could be used in and released from a multitude of processes, and because a larger number of
facilities could potentially be releasing the chemical, there is a greater possibility that emissions
could be missed due to an incomplete inventory.
We will suggest in the How To manual that in future screening exercises it will be important for
chemicals like benzene with multiple sources to pay special attention to the source inventory and
to keep chemicals with widespread sources in the process until all the sources are properly
characterized. The comment and the experience with benzene point to the importance of
building an accurate source inventory. The confidence in the screening process depends on this
accuracy.
Two reviewers raised concerns that routes other than inhalation may produce important different
results for the chemicals considered. They state that ingestion can be a significant contributor
to risk for many products of combustion processes, e.g., for mercury and dioxin. Concerns were
raised that the cumulative assessment could easily show that some chemicals excluded at the
lower screens should have been carried forward into the higher screens if this route of exposure
had been considered.
The focus of this investigation, as designed by the community, was on local industrial,
commercial, and waste treatment and disposal sources. It was the judgement of the Air
Committee that the most significant exposure pathway for these local sources was inhalation.
Several peer reviewers commented that non-inhalation exposure pathways such as fish and beef
ingestion can be significant for some of the chemicals studied. We agree, but it was the
judgement of the Air Committee that the contribution of the local sources to these non-inhalation
pathways was not significant and, as a result, they were not included in the study. A more
comprehensive picture of community risk would certainly include these pathways. The OPPT
Community Assistance Technical Team recognizes that developing a more comprehensive
understanding of risk is an important issue for communities. In its goal to develop the most
comprehensive screening tools possible, the team plans to make the ingestion pathway a priority
item for improving and expanding the Baltimore screening methodology.
The report should have a formal variability and uncertainty analysis.
EPA agrees that variability and uncertainty analyses would strengthen the overall risk screening
results, but such an analysis was beyond the scope of this screening-level assessment. For future
community assessments. Internet citations of available uncertainty analysis methodologies will
be included and can be incorporated if resources permit.
The analysis does not address the project's goal because it does not look at aggregate risk from
multiple chemicals.
L-4
-------
Response EPA agrees with the comments that the Baltimore report does not look at the aggregate risk of
all the chemicals and that analysis of aggregate risk information is of value to the community.
Because of the importance of this issue, the effort to find an effective way to estimate aggregate
risk will be made a priority for future work. The Baltimore project did provide information on
aggregate risk for the four chemicals that were identified in the final screen. The Air Committee
report, found in Appendix J, states that the total risk level for the four chemicals found to be
above the Air Committee screening value corresponds to an increased risk of 6 in 100,000.
To see if any additional review of the data might provide important information for the
community, EPA reviewed the information developed by the Baltimore Air Committee and
estimated the aggregate cancer risk for the 12 chemical carcinogens analyzed in the secondary
screening step. Aggregate concentrations for these 12 chemicals were measured and/or
estimated with air dispersion modeling for the second step of the screening process. Since the
Turner calculation used in the initial screen did not calculate aggregate concentrations, it was
not possible to estimate, using this approach, the aggregate risk for the chemicals in the initial
step of the screening process.
The individual and aggregate risks for 12 carcinogens analyzed in the secondary screen, as they
were calculated in the Baltimore screening exercise, are included in the table below. The last
column provides the best estimate for the total aggregate risk from the 12 chemicals. The second
column displays the estimated risk for the four chemicals that were identified as community
priorities in the final step of the screening process. As displayed in the table, the addition of all
the chemicals adds a risk of about 3 in 1,000,000 to the aggregate risk of 6 in 100,000 for the
chemicals identified in the final screen. Please note that the risk estimates for the chemicals at
the secondary screening level are based on maximum permitted emission rates and not on the
best available information used in the final step of the screening process. Please see a fuller
description of the limits of these risk screening estimations in the Air Committee Report,
Appendix J.
Aggregate Cancer Risk Estimates
Pollutant Name
Arsenic
Benzene
Butadiene, 1,3-
Cadmium
Carbon tetrachloride
Chloromethane (Methyl chloride)
Chromium +6
Chromium +3
Dichloropropane, 1,2-
Methylene chloride
Dioxins & furans (2,3,7, 8-TCDD)
Formaldehyde
Vinyl chloride (Chloroethene)
Aggregate Risks
Risk Based on
Modeled Cone
3.55E-007
1.75E-006
O.OOE+000
1.48E-007
1.70E-007
1.85E-008
6.02E-008
O.OOE+000
3.00E-009
3.57E-010
2.68E-008
5.95E-009
2.64E-007
2.80E-006
Risk Based on
Monitored Cone
O.OOE+000
1.44E-005
3.60E-005
O.OOE+000
7.40E-006
1.17E-006
O.OOE+000
O.OOE+000
O.OOE+000
O.OOE+000
O.OOE+000
O.OOE+000
O.OOE-rOOO
5.90E-005
Best Estimate
3.55E-007
1 .44E-005
3.60E-005
1.48E-007
7.40E-006
1.17E-006
6.02E-008
O.OOE-t-000
3.00E-009
3.57E-010
2.68E-008
5.95E-009
2.64E-007
5.98E-005
L-5
-------
Issue 1.5 Reviewers make the point that the toxicity information for many chemicals is inadequate and that
the toxicity information for the protection of children and infants from the effects of toxic
substances is particularly inadequate. Given these inadequacies, reviewers state that it is not
responsible to represent the toxicity database as sufficiently complete to allow for full
assessment of the likely health significance of hazardous air pollutants, and assessments based
on the current toxicity database should be represented as a likely underestimate.
Comment Toxicity data for more than 115 of the 1 75 chemicals were available from the two main sources
used for this assessment, EPA's Integrated Risk Information System (IRIS) and Health Effects
Assessment Tables (HEAST). IRIS was chosen as the primary source of toxicity information
because of its availability and because of the level of scientific review of the assessments
contained in IRIS. IRIS is widely recognized by the scientific community as a preferred source
of chronic toxicity data for environmental risk assessments. In the absence of toxicity data for
a chemical from IRIS, the secondary source for data used in the assessment was HEAST.
Toxicity data available included 28 chemicals with cancer slope factors and 93 with RfDs with
57 of the 93 based on the inhalation pathway. Because of the available data, many, but not all,
of the chemicals in the Baltimore inventory could be assessed as part of the screening process.
A more complete literature search for toxicity data was beyond the scope of the Baltimore
screening-level assessment.
EPA agrees with the reviewers' comment that the toxicity information available for the
Baltimore screening was not comprehensive. The information did allow the community to
address known chronic hazard concerns. EPA also agrees that the limits of the analysis resulting
from the incomplete toxicity data should be made clear. Language further stressing this point
has been added to the Case Study.
For future screening-level community assessments, efforts will be made to identify additional
sources of toxicity information readily available to communities via the Internet or other means.
An effort will also be made to make new toxicity information from expanded testing initiatives,
such as the High Production Volume Challenge Program, available to communities.
Issue 1.6 Reviewers raised concerns that because measured and modeled airborne concentrations of the
same chemical were different, the modeling was not accurate, and that results using estimated
airborne concentrations are of questionable value. It was also suggested that monitoring must
be done to verify the modeling.
Response Several reviewers raised questions about the validity of the air dispersion modeling used in the
Baltimore project. While we agree with the reviewers on the need for adequate monitoring to
support air dispersion modeling, we believe that modeling can provide important and valid
information. In the Baltimore project, limited resources did not allow for additional monitoring.
Air dispersion modeling was used to estimate concentrations in the absence of measured values
obtained from monitoring. Air dispersion models are the primary tools used to simulate the
chemical and physical processes in the atmosphere that affect the movement of pollutants from
the source to the receptor (Turner, 1994). Such models are the most widely used techniques for
estimating the impact of pollutants from point sources (U.S. EPA, 1987). Air dispersion models
have been tested and validated and are widely used by EPA and State government organizations
for risk assessment, regulatory, and permitting purposes. The modeling methods used are
L-6
-------
Issue 1.7
Response
generally considered to be applicable for assessing impacts of a source from the facility fence
line out to a 50 km radius of the source being modeled (U.S. EPA, 1992).
Such models can provide information to help target air monitoring. Models can also predict the
average concentrations of any released pollutants at any given location. Air monitors, on the
other hand, can only measure pollutants that occur at that particular monitor. Air dispersion
models can provide information concerning the concentration a pollutant is likely to reach. Air
monitors can only measure the concentration on the day the monitor collects a sample Most
importantly, air dispersion models provide information needed for risk management (for
example, indicate what facility released a particular pollutant in unacceptable amounts).
In addition to general questions on the value of air dispersion modeling, several reviewers noted
the discrepancy between the concentrations measured at the monitoring station located in the
target area and the modeled concentrations. In several cases the measured concentrations are
much higher than modeled concentrations. This led reviewers to question the accuracy of the
modeling overall. The issue of the difference between the measured and modeled concentrations
is discussed on page 53 of the Case Study report and illustrated for benzene in the pie chart in
Figure 5 on page 55. We do not believe that the differences question the validity of the air
dispersion modeling. The modeling did not include mobile sources and the Air Committee
concluded that the difference between monitored and modeled concentrations could largely be
explained by the contribution of mobile sources to the monitored measurements. As noted in the
text, the modeling of mobile sources is strongly recommended for future air screening exercises.
Although geographical areas cannot be directly compared, the recently released report
summarizing a study of air quality in Southern California, the MATES-II Report, generally
confirms the Air Committee conclusion on the contribution of mobile sources to the measured
concentrations. In this report mobile sources are estimated to account for at least 90 percent of
benzene emissions. (Draft MATES II Report of the South Coast Air Quality Management
D i s t r i c t, R e f e r e n c e study, November 1999. Available at
http://www.aqmd.gov/newsl/MATES_II_results.htm).
Turner, D. 1994. Workbook of atmospheric dispersion estimates: an introduction to dispersion modeling.
Second edition. CRC Press, Inc. Boca Raton, FL.
U.S. EPA, 1987. Guideline on Air Quality Models (Revised). U.S. EPA Office of Air Quality Planning
and Standards, Research Triangle Park, NC. EPA-450/2-78-027R.
U.S. EPA, 1992. A Tiered Modeling Approach for Assessing the Risks Due to Sources of Hazardous Air
Pollutants. U.S. EPA Office of Air Quality Planning and Standards, Research Triangle Park, NC.
EPA-450/4-92-001. March 1992.
With respect to cumulative target organ analysis, the section on grouping chemicals according
to "similar organs or physiological systems" needs to be reconsidered because only respiratory
and neurological effects were evaluated.
EPA agrees with the comment that the attempt to identify chemicals with cumulative effects did
not follow the procedures for a hazard index calculation. The Baltimore risk screening exercise
was only a limited attempt to identify chemicals acting on the same target organs, which might
potentially have cumulative effects. Neither a hazard index or cumulative risk assessment was
attempted. Hazard index and cumulative risk assessment require information on the mechanism
of toxicity so that chemicals with the same or a similar mechanism can be grouped and the
L-7
-------
impact of their toxicities summed. The information necessary for such an assessment was not
available for this screening level assessment. The need to provide guidance on identifying
chemicals with cumulative effects is included on the list of improvements for future work.
-------
2. Suggestions for Improvements to the Risk Screening Methodology That Will be Prioritized for
Future Implementation
The following comments raised issues that call for future improvements in the risk screening
methodology. These comments are organized by the steps in the risk screening process and are presented in
tabular form.
Suggestions for Future Improvements
SCOPING ISSUES:
Use facilitator in Partnership interaction activities (meetings, decisions, etc.)
Limit inclusion of indoor air
Look at multiple pathways of exposure
PARTNERSHIP (STEP 1):
Get agreement up front for a risk management plan
EMISSIONS INVENTORY (STEP 2)
Broaden beyond industrial, commercial, waste, especially to mobile (maybe use ASPEN)
Include wastesites and landfills in source inventory
INITIAL SCREEN (STEP 3)
Improve access to toxicity data
Use recent California effort to derive acute toxicity values
Consider sensitive population analysis
Add cumulative/aggregate inhalation exposures to screening
INITIAL SCREEN (STEP 3) (continued)
Identify in advance a process for addressing issues in toxicity
Suggested method for accounting for aggregate at initial screen
Use consistent, conservative screening values throughout all screening steps (e.g., the Region 3 RBC
values).
SECONDARY SCREEN (STEP 4)
Place one of the grid receptors on a school, hospital, nursing home or other sensitive population
FINAL SCREEN (STEP 5)
Expand monitoring as most important conclusion
Conduct air monitoring to validate air dispersion modeling predictions
Discuss detection limits for monitoring information used in screening
Consider persistence of chemicals in environment
Consider using 24 hour, 70 year exposure for urban populations to ambient air
L-9
-------
3.
Suggestions for Clarifying the Case Study Report
The following comments generally called for clarification in the Case Study report. In most cases the
comments were determined by EPA to warrant attention and the document was revised to add the clarification
needed.
Suggestion 3.1.
Response
Suggestion 3.2
Response
Additional discussion is needed to explain the way the terms RfD and RfC are used in the
document.
EPA added text on page 32 to help clarify how toxicity data were used in the initial screen.
For the noncancer assessments, RfC values were converted to RfD values based on EPA-
approved procedures. EPA scientists preferred to use an estimated dose and the associated
RfD because risk assessors needed to evaluate risks for many types of scenarios. RfCs
incorporate exposure assumptions and can only be used for one exposure scenario. By
using the RfD, the same estimated doses (based on inhalation exposures) could also be
used in the cancer risk calculation by combining it with the cancer slope factor. As a
result, RfCs were converted to RfDs and inhalation doses were calculated for the scenario
being assessed (see Region 3 RBC table in Appendix D). Conversion of RfCs to the more
traditional RfDs is straightforward using a 20 mVday inhalation rate and a 70 kg body
weight.
Clarification is needed on the types of air pollution sources that were included in the
emissions inventory used as the basis for the risk screening. Clarification should be added
to address confusion over point sources and area sources.
The emissions inventory for the Baltimore Case Study focused on industrial, commercial,
and waste treatment and disposal sources of air pollution, ranging from small sources such
as gas stations with annual emissions to air of less than 100 pounds of chemicals, to large
facilities with annual emissions of over 1 million pounds. Many of these are known as
point sources, such as power plants, steel mills, chemicals plants, and other large facilities.
Mobile sources of air pollution, such as vehicles and small engines were not covered in
the inventory. The table below (also presented on page 19 of the revised report) provides
a summary of the types of sources included (and not included) in the inventory for the
Baltimore Case Study.
It is also worth clarifying the use of the term "area source," which is used in two different
contexts in the report. Area sources are smaller stationary sources of pollution that are not
inventoried individually but whose emissions are estimated as a group and reported as a
single source category for a geographic area. Examples of area sources include gas
stations and dry cleaners. Another somewhat different use of the term area source applies
to air dispersion modeling when the emission from a source could not be associated with
an exact emission point, such as an exhaust stack. The emissions from these sources were
modeled as though they were uniformly emitted from the entire area covered by the site.
Within the description of the air modeling procedure, these are referred to as area sources.
Care should be taken not to confuse the use of area source in the context of air dispersion
modeling with the definition of area source used in defining the size of sources.
L-10
-------
Source Inventory Table
CAA Category
Included in Baltimore
Inventory
Not Included in
Baltimore Inventory
Point (major stationary)
Examples: chemical plants, power
plants, incinerators, landfills, steel
mills, POTWs
Area (small stationary)
a) Commercial and industrial
chemical use and handling
Examples: dry cleaners, gasoline
stations, print shops
b) Commercial, industrial, and
institutional boilers
Examples: school, hospital, office
building heating
c) Household heating and chemical
use
Examples: furnaces, fireplaces,
lawn chemicals
Mobile Sources
a) On road
Examples: cars, trucks, buses
b) Off road
Examples: portable generators,
construction equipment, boats, law
mowers
Suggestion 3.3
Response
Suggestion 3.4
Response
Explain that some carcinogens have thresholds.
The text has been revised to more accurately represent the threshold/nonthreshold
characteristic of chemical toxicity. A change to the document was made in the text box
that appears on page 26 that adds: "But there are exceptions. For example, some
carcinogens have thresholds."
Provide clarification on Figure 5 and the discrepancy between the modeled and the
monitored concentrations of benzene in the Partnership area.
EPA has revised the report to provide additional discussion of the contribution of
inventoried emission sources to the benzene concentrations monitored at the Fairfield
station. The annual emissions from individual benzene sources are contained in the
ISCST3 input file. Initially all benzene emissions were included in the modeling run and
the maximum annual average concentration in the approximate geographic center of each
neighborhood was calculated. To determine the contribution of each individual benzene
L-ll
-------
Suggestion 3.5
Response
Suggestion 3.6
Response
Suggestion 3.7
source to the total ambient air concentration in the neighborhoods, the model was run
repeatedly with only one benzene source "turned on" at a time. This yielded an estimated
maximum airborne concentration due to the single emissions source under consideration.
That value was compared to the estimated concentration due to all sources to determine
the contribution of that source (percentage of the total). Using this same approach,
emission sources could be grouped together, if desired, as when many small sources are
being considered.
The Partnership had both monitored and estimated annual average concentrations for
benzene in one of the Partnership neighborhoods (Fairfield). A comparison of the two
values was performed to determine how closely the predicted concentration matched the
monitored concentration. The monitoring station in Fairfield is about i/i mile from the
location of the highest predicted concentration of benzene in Wagners Point. At this
distance the two locations could be unequally subject to influences, such as nearby
benzene sources or differences in wind direction and frequency, that could confound the
comparison of benzene concentrations. Nonetheless, if it is assumed that the modeling is
accurate, then significant differences between measured benzene concentrations and
modeled benzene concentrations could be due to sources of benzene not captured in the
emissions inventory. The unaccounted-for emissions could be due to unregistered
stationary sources or, more likely, benzene emitted from mobile sources (cars and trucks)
passing through the area on high-volume routes such as 1-695 and Patapsco Ave and at the
1-895 toll plaza. It is well known that mobile sources make a significant contribution to
benzene concentrations in urban air.
Clarification is needed on the methodology used for selection of the receptor locations for
the ISCST3 modeling, including the geographical area considered for modeling and the
receptor grids.
The Partnership area was defined by neighborhoods (Cherry Hill, Brooklyn/Brooklyn
Park, Curtis Bay, Wagners Point) and by ZIP Codes 21225 and 21226. The coordinates
of the neighborhoods corresponded with their approximate geographic centers of these
towns. Page 43 of the report provides additional details on the receptor grids and the four
Partnership neighborhoods used as the primary receptor locations. Recognizing that air
pollutants may be transported from outside the Partnership area, facilities within 5 miles
of the Partnership area were included in the emissions inventory. While this approach did
not capture pollution transported from other regions of the United States, it represents an
exhaustive attempt to consider local commercial and industrial stationary sources.
One commenter suggested that EPA should create a summary table for the 29 chemicals
showing the concentrations and screening values used in each step.
It was determined that such a table would be very complicated and would not help the
reader to interpret the outcome of the initial screen. EPA did not make the suggested
change to the report because similar tables were included in Appendix I for the secondary
screen, which involved fewer chemicals.
The document needs clarification on the sources available for toxicity data because the
gaps could hinder the assessment of a chemical's human health effects.
L-12
-------
Response
Suggestion 3.8
Response
Suggestion 3.9
Response
Suggestion 3.10
Response
The document was revised to inform the reader of the availability of toxicity data for the
chemicals emitted in the Partnership area. Toxicity data for more than 115 of the 175
chemicals were available from the two main sources used for this assessment, EPA's
Integrated Risk Information System (IRIS) and Health Effects Assessment Summary
Tables (HEAST). These were the best readily available sources of toxicity information
for this assessment. Specifically, changes on pages 28 through 30 were made to better
describe the sources of toxicity data considered for the screening process. Toxicity data
available included 28 chemicals with cancer slope factors and 93 that had RfDs, of which
57 were based on the inhalation pathway. This meant that many, but not all, chemicals
could be assessed as part of the screening process.
IRIS was chosen as the primary source of toxicity information because of its availability
and because of the level of scientific review of the assessments contained in IRIS. In the
absence of toxicity data for a chemical from IRIS, the secondary source for data used in
the assessment was HEAST. These are widely recognized by the scientific community as
the preferred sources of toxicity data for environmental risk assessments. It is
acknowledged that these sources are not comprehensive, but they do allow the community
to address known hazard concerns. A more complete literature search for toxicity data
was beyond the scope of this screening level assessment. The best readily available
sources will also be recommended for future screening level community assessments, but
efforts will be made to identify additional sources of toxicity information readily available
to communities via the Internet or other means.
Clarification is needed on the initial screening approach and how it addresses only one
source at a time.
The initial screen addressed emissions from individual sources because it used the Turner
equation to estimate resulting air concentrations and exposures. Only in subsequent steps,
where ISCST3 modeling was used, could estimates be provided for air concentrations of
chemicals emitted from multiple sources.
The "professional judgment" that was applied for screening is not well documented and
needs clarification.
EPA revised the document to clarify the discussion of the chemicals identified from the
initial screen and the subsequent elimination of select chemicals based on professional
judgment. We added text after the table on page 34 that says: Chemicals with an "*" were
not selected for the next stage of the screening process because they were no longer
emitted from the facility because of changes in the production process or the facility that
had emitted them was no longer in operation.
Additional clarification is needed on the conservative nature of toxicity data, which often
have many safety factors built in.
The document was revised on page 28 to better explain the toxicity data used in the
screening and the potential for overestimating risks. For example, EPA slope factors
express carcinogenic potency in terms of the estimated upper-bound incremental lifetime
risk per milligram per kilogram (mg/kg) average daily dose. Cancer slope factors (CSFs)
are available, where applicable, for either oral (SF0) or inhalation (SF,) exposures. Unit
L-13
-------
Suggestion 3.11
Response
Suggestion 3.12
Response
Suggestion 3.13
Response
risk is a similar measure of cancer potency for air or drinking water concentrations
and is expressed as risk per microgram per cubic meter (Mg/nv) in air or as risk per
microgram per liter (^g/L) in water for continuous lifetime exposures. The term
upper bound in this context means that the measures of cancer potency are high-end
estimates, so they will be conservative. This may result in an overestimate of cancer
risk when toxicity data are incomplete, which is usually the case. Uncertainty and
modifying factors are a few included in deriving the toxicity values, which makes the
resulting toxicity values (e.g., RfDs, RfC, etc. more conservative. Upper-bound
values are intended to be protective of human health for continuous lifetime
exposures, even though cancer risks may be overestimated. The use of the average or
lower limit values would be more likely to underestimate cancer risk.
Clarify the use of term "actual risk" in the report.
No changes were made to the document in response to this comment. In this context, use
of "actual" was intended to inform the reader of the uncertain nature of risk assessments
such as this, so it was important to note that these estimates could not be considered to be
the "actual" risks.
The definition of a reference dose should be expanded to make it clearer.
EPA agrees with the comment and the clarification was added to the report. Specifically,
the following text is now included on page 28:
A measure of toxicologic potency for chronic (long-term) effects is the "reference dose"
or "reference concentration." The reference dose (RfD) is defined as "an estimate (with
uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without appreciable risk of
deleterious effects during a lifetime" and is expressed as a mg/kg-day dose (U.S. EPA,
1997e). The reference concentration (RfC) is an estimate (with uncertainty spanning
perhaps an order of magnitude) of a daily inhalation exposure of the human population
(including sensitive subgroups) that is likely to be without an appreciable risk of
deleterious noncancer effects during a lifetime. Conversion of RfCs to the more
traditional RfDs is straightforward using a 20 m3/day inhalation rate and a 70 kg body
weight. RfD values for inhalation were derived from RfCs and are used in this study. The
RfD is usually based on the most sensitive known effect (i.e., the effect that occurs at the
lowest dose) and can exist for both oral exposures (RfD0) or for continuous inhalation
exposures (RfD,).
The example source inventory database table should be modified. It carries too many
significant figures for a risk assessment and the last two columns on risk and HQ should
have two significant digits.
The purpose of this table was to provide an illustration of the database that was used for
managing the data used in the screening process for the Baltimore Case Study. It is not
desirable to change it in the report because the same change would have to be made in the
database as well. Also, EPA recognizes that the number of significant figures is limited
and that their presentation could imply a level of precision in the estimates that does not
exist. For example, the aggregate risk estimates presented earlier were 5.98 per 100,000
L-14
-------
Suggestion 3.14
Response
Suggestion 3.15
Response
Suggestion 3.16
Response
Suggestion 3.17
Response
but EPA rounded that to 6 per 100,000 because we recognized the uncertainty in such
estimates. Therefore, EPA will not change the report or database at this time, but will
make the issue of the appropriate number of significant digits a future improvement issue
for the database.
The report should be clarified to indicate that non carcinogenic screening values were also
used in the risk screening.
EPA's risk screening methodology included both cancer and non-cancer effects (as
reflected on page 29) by selecting screening levels that correspond to both types of
endpoints. For the initial screen, the risk screening values of 10"6 for cancer and HQ > 1
for other chronic effects were used to screen individual sources. The secondary and final
screens used the Region 3 RBCs as the basis for the screening levels. The RBCs are
developed by EPA Region 3 scientists to reflect the concentrations at which either the
cancer risk to an exposed population is 1 in a 1,000,000 or the HQ is 1. Therefore, all
phases of the screening considered cancer and noncancer endpoints.
Incorporate the Air Committee Report into the Case Study.
The Air Committee Report has been revised and is presented as an appendix to the
Baltimore Case Study report. EPA chose not to combine the two reports. The Air
Committee Report was prepared by the Partnership and has very specific wording that was
developed through a consensus-building process. EPA chose to present that report in its
entirety as an appendix to the Baltimore Case Study report.
Add more detail, including citations, to make the document clearer and more transparent
including information from the literature on similar risk screening methodologies.
EPA agrees and has added to the report many more citations for data and approaches used
by other studies that we considered in developing the methodology. The intent is to
provide the reader with information on the sources of information, particularly Internet
Web sites, that were accessed to obtain information. Also, the "How to" methodology
document that is being developed can be considered to be a companion piece to this report.
That document will add more specifics on the types of data sources available for use in
studies such as these.
Clarify which monitoring data were used in the screening.
EPA added information on page 22 about the monitoring data available for the Partnersh ip
area. 1996 annual average concentration data (the most current year available) from the
Fairfield monitoring station were generally used in the screening. The use of maximum
values would have probably been too conservative since they were not typical of air
quality and would not have been representative of the concentrations of chemicals in the
air that the neighborhood residents breathe. Data were available from 1992 to 1996 for
the 41 chemicals monitored from the five Baltimore area monitoring stations (Glen Burnie,
Downtown Baltimore, Fort McHenry, Essex, and Northeast Baltimore) and the one station
located in the Partnership area. These data from the Fairfield monitoring station were
used in the screening to represent concentrations in the Partnership area. Table 1-1 in
L-15
-------
Suggestion 3.18
Response
Suggestion 3.19
Response
Suggestion 3.20
Response
Suggestion 3.21
Response
Suggestion 3.22
Response
Appendix I summarizes the monitored concentrations that were used in the screening
process.
Clarification is needed to indicate that the predicted concentration at a grid receptor is the
sum total for a chemical from all modeled sources.
EPA agrees and has made this clarification in the text (page 41) to reflect that the ISCST3
modeling performed in the secondary and final screens considered multiple sources that
release a chemical.
Clarification is needed on the inhalation rate used in the calculations in the initial screen.
The 1 mVhr inhalation rate is a part of the overall Turner methodology as described in
Appendix E. This is contrasted against the 20m3/day inhalation rate used in the
conversion of unit risks to cancer slope factors. The ImVhr rate used in the Turner
calculation is the standard method that EPA/OPPT used for previous assessments.
Revising this methodology is beyond the scope of the Baltimore Case Study. This is an
issue for EPA/OPPT to consider in general, and for the technical team to consider in
improving the air screening methodology. For instance, the inhalation rate might be
slightly revised because EPA's Exposure Factors Handbook reports 13.8 mVday as the
median breathing rate, which could be used for both the Turner methodology and the unit
risk factors.
Clarification is needed on Table 4 which indicates "NA" for.. number of secondary screen
emission rates while these same chemicals have values for final screen emission rates.
The final screen for benzene resulted in the discovery of additional emissions sources that
were not part of the secondary screen. Therefore, Table 4 reflects the increased annual
benzene emissions over those used in the secondary screen in the Wagners Point area near
the modeled receptor location. These increases in emissions account for the increased
modeled airborne concentrations of benzene in the final screen.
For the Fairfield monitor, the document should state whether it is a source-oriented
monitor or a community-based monitor. This same comment holds for the other Baltimore
area monitors mentioned in Appendix J.
The Fairfield monitor, as well as other toxic air pollutant monitors in the Baltimore area,
are positioned so as to provide readings suitable for estimating exposure over a larger
geographic area. This text change was included on page 22 of the Case Study report.
Appendix G should be revised for accuracy. Extraneous information that is not used in
the screening process should be removed.
EPA reviewed the list of columns detailed in Appendix G of the Baltimore Case Study
report and made them consistent with the example spreadsheet. We agree with the
comment that extraneous information (i.e., not used in the screening process) should be
removed. For the version of the spreadsheet that is included in the Baltimore Case Study
report, some of the columns have been deleted. Similarly, the spreadsheet used to manage
data for future assessments is being revised as part of the "How to" manual. We hope that
L-16
-------
Suggestion 3.23
Response
Suggestion 3.24
Response
Suggestion 3.25
Response
Suggestion 3.26
Response
these changes will make the spreadsheet more manageable and applicable to all stages of
the screening.
Appendix J should be enhanced to more fully describe the Clean Air Act requirements to
address air toxics in urban air, including a more thorough discussion of MACT standards,
the residual risk program, cleaner fuels, etc.
EPA did not make any changes in response to this comment because the issues are beyond
the scope of this screening effort.
Appendix K discusses the use of 5 years of modeled data for the screening. Clarification
is needed on the multiple modeling scenarios.
Appendix K provided background information on model set-up, assumptions and a
chronology of modeling runs with ISCST3. Modeling scenario 1 in Appendix K is the
modeling for the secondary screen. Scenario 2 represents an intermediate step that
included more accurate information on emissions. Scenario 3 incorporated additional
information on the type of chromium emitted by facilities and added updated benzene
emissions. Scenario 4 was used to determine the contribution of individual facilities'
benzene emissions to the total modeled benzene concentration in Wagners Point. Both
toxic and criteria air pollutants were modeled using local meteorological data from the
most current years available (1987-1988, 1990-1992). Generally, it is recommended that
meteorological data over a five year span be used in air dispersion modeling to account
for temporal variations. The highest predicted values either for the receptor locations (1
of 4) or for any given year (1 of 5) were typically used to make the screening as
conservative as possible.
Provide clarification on the rationale for the selection of the discrete neighborhood
receptors (e.g., Cherry Hill at a given lat/long)?
The document was revised on page 43 to describe the receptor locations. The Partnership
area was defined by neighborhoods (Cherry Hill, Brooklyn/Brooklyn Park, Curtis Bay,
Wagners Point) and by ZIP Codes (21225,21226). The coordinates of the neighborhoods
corresponded with their approximate geographic centers, which were used in the ISCST3
modeling to estimate ambient air pollutant concentrations for those four communities.
The document should include a fuller description of the airsheds and meteorology of the
area.
No changes were made to the document in response to this comment. EPA feels that this
issue is addressed sufficiently in the modeling methodology. Both toxic and criteria air
pollutants were modeled using local meteorological data from the most current years
available (1987-1988,1990-1992). Generally, it is recommended that meteorological data
over a 5-year span be used in air dispersion modeling to account for temporal variations.
L-17
-------
Instructions/Charge
to ERG for
Peer Review of
Baltimore Screening Methodology
L-18
-------
Instructions/Charge to ERG for Peer Review of Baltimore Screening Methodology
I. General Instructions
A. Conflict of Interest:
The Reviewer(s) shall not be a resident of the geographic area which is the subject of the report or the
reviewer shall not be currently involved or have previously participated in technical support work affiliated
with this document. In addition, the reviewers should not be affiliated with private organizations or
stakeholders involved in this effort to the point that there may be a perceived conflict of interest.
B. Scope of Review:
The Case Study under review describes a risk-based air screening exercise carried out by the Air
Committee of the Baltimore Community Environmental Partnership. The work of the Baltimore Air Committee
consisted of the development of both a risk-based screening methodology for analysis of neighborhood air
quality and also a partnership building process designed to increase participation and build the community's
long-term ability to address air quality concerns. Peer reviewers are asked to provide feedback, as appropriate,
on both of these aspects of the project. Questions on the risk-based screening methodology are given in
General Charges 1 and 2 and in the Specific Charges. A question on the partnership building component is
provided in General Charge 3.
As the work in Baltimore progressed, lessons learned and suggestions for improvements were identified
and included in the case study. In Charge 2, peer reviewers are asked to comment on the improvements
identified in the case study.
EPA would like the reviewers to focus on content issues related to the above. An editorial or quality
control review is not requested.
II. Project Goals
The goals listed below were adopted by the Baltimore Air Committee as a guide to its work. Peer
reviewers are asked to comment on the work of the Air Committee in light of these goals.
L-19
-------
A. To determine if the current aggregate levels of toxics in the air in the Partnership neighborhoods
resulting from the multiple industrial, commercial and waste facilities in and around the Partnership may
adversely affect community health.
B. To recommend actions to improve air quality in the Partnership neighborhoods. (Recommendations
to be based on the information on risk-based priorities provided by the screening exercise.)
C. To build the long-term capacity of the community, including residents and businesses, to take
responsibility for their environment and economy.
III. Charges
A. General Charges
1. Did the screening methodology, as applied in Baltimore, achieve goals A and B?
The report identifies various technical improvements to the screening methodology. These are listed
below. Could the methodology (emissions inventory, initial screen, secondary screen, final screen),
as modified with the improvements identified below, help other communities seeking to understand
and improve air quality? Please comment on both the appropriateness of the improvements listed
below and their priority. Are there other improvements that should be considered?
(a) Addition of mobile source modeling: The Baltimore exercise focused on stationary and area
sources. This task will expand capacity of methodology to include mobile source modeling
(b) Review and improvement of source inventory Review: Review existing source inventories
to identify additional sources of emissions to insure that all significant sources are included
(c) Identification of best source for toxicity data: Compare available toxicity data bases to identify
most accessible and complete source of data for community screening exercise
(d) Expand Baltimore methodology to include short term acute effects
L-20
-------
(e) Review screening calculations to determine if they are appropriate for and protective of
sensitive and urban populations
(f) Development of a method to screen for cumulative exposures in the Initial Screening Step
(g) Expand methodology to include indoor air risks to provide a more comprehensive picture of
air risks
(h) Incorporation of GIS mapping to enhance the communication of the modeling and screening
results
3. Are the partnership and community participation aspects of the screening exercise described in the case
study and in the lessons learned section appropriate to achieve goal C? Could this screening exercise be used
in other geographic areas to reach this goal. Can you identify any improvements or changes in the screening
exercise that would help accomplish this goal?
B. Specific Charges: Please provide us feedback on the following aspects of the methodology, given
project goals A and B:
1. The Emissions Inventory: Were the inventory of sources and the release and monitoring data used in
the Baltimore screening exercise sufficient and appropriate to reach the goals of the committee? What
additional sources do you think should be included in a source inventory to expand the scope of the
methodology for use in other communities?
The initial screen: a) Were the methods for calculating airborne concentrations, potential dose, and risk
appropriate and scientifically justified?; b) Was the screening criteria that was applied to identify
chemicals for further analysis appropriate?
3. The secondary screen and the final screen: a) Was the modeling approach for developing estimates of
neighborhood concentrations from multiple sources technically sound?: b) Was the screening criteria
that was applied appropriate?; c) Were the assumptions built into the Region III risk-based
concentrations appropriate.
L-21
-------
Does the draft Committee Report (found in the appendix) adequately and accurately describe the
screening exercise and its results0
5. Is the screening methodology as used in Baltimore sufficiently protective of sensitive populations?
What would you suggest, if anything, for improving this aspect of the screening methodology?
L-22
-------
Peer Review Comments
Michael A. Callahan
U.S. EPA National Center for Environmental Assessment
L-23
-------
Peer Review Comments for
Baltimore Community Environmental Partnership Air Committee Technical Report
Michael Callahan
Senior Science Advisor - U.S. EPA, National Center for Environmental Assessment
A. General charges:
1. Did the screening methodology, as applied in Baltimore, achieve Goals A & B?
I think that at best, this project can only be termed a partial success in Goal A and a failure in Goal
B. The methods for data collection worked well, analysis less well, getting consensus terribly, and the rest,
particularly in dealing with the various agendas on the Committee, not well at all. Without all of the parts
working well, this or any future project based on this methodology cannot be thought of as an overall
success.
Hindsight can be valuable, especially if this methodology is to be applied to other cities and
situations. One of the things I thought planted the seeds for the discontent realized later in the project was
the stark contrasts between the questions the community had (pages 12-13), and the much narrower scope
agreed upon for this project (bottom of page 13). I realize that many of the concerns of the community were
not immediately answerable due to, among other things, lack of a workable methodology. On the other
hand, even if the community agrees to the narrowed scope, and even if the project went off perfectly, there
would still remain a feeling in the end that the community's questions were not answered. The paragraph
on the top of page 56 talks about the need for the community to understand the limitations of this tool, but
what about EPA's need to understand the questions the community is asking, and helping them get
answers? If we have the "hammer" in this methodology, do we also have to see every question either as a
nail or irrelevant?
If this methodology is to be applied to other communities, it is important that EPA find a way to at
least address the other questions (which are very common ones communities ask), or every project will
have a certain community dissatisfaction as a result. This is somewhat like "bait and switch," with the
questions answered not being the questions asked. It may take the community a while to figure this out, but
when they do, trust is lost, probably permanently.
L-24
-------
It is not clear from the writeup (page 14) who exactly the "some Committee members" were that
had the concerns about distracting the focus of the group from speaking "directly to the main community
oncern," but in retrospect it seems a flawed decision. Apparently not everyone on the Committee
understood the implications of only looking at air toxics emissions from facilities. It's not even clear that
this was indeed "the main community concern," since concerns about air pollution also included odors and
concerns about "midnight releases." Future applications of this methodology will have to take great pains to
make sure everyone actually understands and agrees to what steps are to be taken, and the implications.
There also should have been, again looking in retrospect, a contingency discussion. "What happens if we
find no levels of chemicals above our health benchmarks? What happens if we can't document any permit
violations9 What if we do find something of concern? What are the next steps?"
In terms of general peer review question 1. Goal A was only partly successful on the surface. If
viewed from the larger view of the community's concerns, it failed. A lot of data was collected and models
run. but they only covered part of the picture (a significant part, nonetheless). The limitations of the data
and methods did not allow the project to make a statement such as "the air levels of toxics are in a range
EPA sees as safe, based upon conservative assumptions (<10"6 cancer risk and <1.0 HI). Community
concerns are directly focused on the safety of residents, and scientific temporizing is not satisfying to the
community. Moreover, although data collection was successful, analysis and interpretation of results failed
spectacularly. The last sentence on page 53, "A consensus on the interpretation of the results did not
develop, and the effort was halted..." is a marvelous understatement. In looking forward to future
applications of this methodology, we can also look forward to this type of disagreement unless specific
ground rules and contingencies are built into the planned interpretation of the data. Questions like, "What if
we find this? How will that be interpreted?" should (again, with hindsight) be discussed before any data are
collected.
I think asking if Goal B was successful is a question that answers itself. If the Committee could not
even agree on interpretation of the data, how could they recommend logical steps for the community to take
other than generic ones0 Only generic remedies would be quite unsatisfactory to the community after their
expectations were raised by all the neighborhood data being collected, since they probably knew the
generic steps beforehand (or at least the Committee could have listed them early on). Specifically, I can
find no real recommendations in the "recommendations" section in pages 52-54. On benzene,
recommendations were "postponed." For mobile source chemicals, the partnership was told to participate in
nebulous "air quality improvements at the regional level," with discussion of what that means to be
supplied later. For carbon tetrachloride and methylene chloride, ''Recommendations were not developed...."
As a batting average, this record is close to - if not exactly - .000.
L-25
-------
2. Various technical improvements in the methodology.... Are there other improvements that should
be considered?
There appears to be a serious imbalance between the technical methods used for data collection and
analysis on one hand, and the development of the rest of the methodology (called Partnership and
Community Participation in Goal C), including listening, interaction, teaching, negotiations, etc., on the
other. This methodology simply will not work if the technical side is built up to the exclusion of the other
side, as suggested by the list of bullets under this charge. Is the overall mission of this methodology project
to build a new computer or GIS-based tool and release it to make the world a better place? Or is it to collect
the tools and methods, and provide them, along with advice, to the communities to help them better analyze
their situation and hopefully to better solve their own problems? I get the impression it is the latter, but the
story reads like the former. Where are the questions about how to make the non-technical side better?
That being said, in my opinion, community assessment is a cumulative-risk-type operation.
Anything that improves the ability to see, understand, interpret, and explain the "big picture" about what
people are exposed to and where possible threats to health are coming from, is helpful. Mobile source
monitoring would be helpful in the context that it can be linked to actual exposures and legitimate
recommendations (which need to be thought about beforehand). As for toxicity data, there are no magic
data banks that have the answers we have been seeking lo these many years. The usual ones, IRIS, HEAST,
RBES tables, etc., are sufficient for now; they have to be, since there isn't much else out there. When new
tox data become available, I'm sure it will be widely publicized within the toxicology, risk, and public
health communities. Meanwhile, the methodology should note that before the methodology is applied at a
new location, the currency of the tox data should be checked by someone who is knowledgeable about such
things.
If acute effects are to be included in the methodology, a lot more work needs to be done on how the
concentration values are to be obtained. Long term modeling for an area for chronic effects is one thing, but
trying to evaluate acute effects possibly from a small pocket of air is quite another, and a modeling-only
approach will probably not satisfy the community (there will be too many anecdotal incidents, for one
thing). The issue of odors will have to be added to the acute effects analysis, also. The issue of acute effects
will almost certainly require some on-site monitoring. All in all, it is a big, costly, addition to the
methodology, but EPA may have to start moving in that direction if it wants to be relevant in answering the
communities' environmental questions.
In terms of the screening methodology calculations, I do not believe EPA will be able to get away
with saying "this is not a risk assessment" very much longer. The questions being asked by the
communities (e.g., pages 12-13) have significant risk components, and to do calculations and say "this is
not risk assessment" (and rightly so!) will eventually be viewed as avoiding answering the communities'
questions and concerns. The technology exists now to estimate concentrations, develop exposure scenarios,
etc. Within a short time, the ability to do multiple chemical modeling, at least on a screening level far better
L-26
-------
than the Generic Turner Model, will be commonplace. EPA should aim its methodology at that. After all,
ue are no longer doing calculations on a piece of paper with an adding machine. This project took many
years and there was ample opportunity to do fairly sophisticated technical analysis. We should start from
that point, and analyze the chemicals that need to be analyzed, not reach back for tools like the GTM to get
rid of things that might add to the cumulative risk.
Including indoor air methods may be the single best improvement to the methodology in terms of
developing realistic and useful recommendations about how to improve the community's health. It is a
mixed blessing, however, as many persons do not want anyone telling them anything about their own
lifestyle or the way they keep their homes, which has a large influence on indoor air concentrations. It is
invasive of one's lifestyle, expensive (NHEXAS=$1 7M), and often finds things that individuals would
rather not see pointed out. But, it gets results. Adding indoor air methodology should not be taken on as an
issue without eyes wide open as to cost and potential for highly charged discussions (case in point: the
community representatives' leaving the Baltimore project was - according to their letter- due at least in
part to their feeling that the analysis was moving in this direction, if only by suggestion of others on the
Committee that lifestyle issues were important).
GIS mapping is a worthwhile addition to the methodology, and will probably be critical within a
year. Communities will not have the capabilities to do their own GIS work in the short term, but perhaps
within a few years the software will be available for tomorrow's PCs. Meanwhile, EPA should provide
some help in running maps for the areas that use this methodology.
3. Are the partnership and community participation aspects of the screening exercise in the case
study and in the lessons learned section appropriate to achieve Goal C?...
The lessons learned section is wonderful and right on the money. The improvement needs to be in
the mind set which begins a case study like this. EPA can go into one of these with the approach of trying
to help answer the community's questions, the sort of approach that's embodied by the statement, "I don't
know the answer to that, but I'll find somebody that does, or find out what is known about that issue," and
then follow through. Contrast that approach to one which says, "I have a tool here, but it can't answer all
your questions. Let's see which ones it can shed light on or answer." The former is a real partner, while the
latter is a helpful salesman. If partnership and community participation is a goal, it must be approached
\vith the partnership attitude. A helpful salesman may be appreciated, but will never, and can never achieve
the goal of being a full partner, with all the positive benefits that implies. A salesman, even a helpful one,
will never quite be trusted completely.
L-27
-------
B. Specific Charges:
1. Emissions inventory.
I was somewhat disappointed at the "winnowing down" methodology which modeled a collection
of sources which represented 95% of the pounds of emissions. I think there should be a way to model all
the sources that contribute. This will avoid questions about "what was left out of the analysis?" later. The
smaller facilities won't add much, but the more complete analysis will be much more satisfying to the
community. As far as emissions data, the sources will vary by state. TRI is universally suspect as to
accuracy, but it's the best there is in many places. Most states have a database of facilities which includes
smaller facilities not required to report to TRI (MDE had such a database here). At the very least, these two
sources of data should be investigated in any case study. Local monitoring data and other local sources
should be investigated on a case-by-case basis with help from the community and local government. As a
footnote, // is absolutely imperative that before modeling, the lat/long locations of the facilities be ground-
checked. TRI is notoriously bad for having inaccurate lat/long information, and a drive-around with a
global positioning system (GPS) locator can save a lot of embarrassment later.
2. Initial Screen.
The Generic Turner Method essentially calculates an average concentration of a theoretical place
100 meters from a 3 meter high continuous release (essentially as a fugitive release at this height). If this is
to be a bounding estimate (as it appears) to eliminate all the chemical-facility combinations that would not
in themselves be problematic, the use of the 25% factor to lower the concentration at the 100 meter point
by a factor of four seems to defeat the purpose. It would seem better to just assume the wind blows the
same way all the time, and if the chemical-facility combination could not get above the benchmark criteria
as a bounding estimate, then it would be eliminated from further consideration.
As for appropriateness of the criteria, I think that this will be a very conservative calculation, and
should be labeled a bounding estimate. It will eliminate only those chemicals which should be quite a bit
below the risk levels represented by the screening criteria, when more realistic exposure parameters are
used.
I still feel that this step will eventually prove unnecessary and counterproductive, as discussed a
few paragraphs above.
L-28
-------
Secondary and final screen.
My opinion is that the several weeks of computer time needed to run ISC-LT3 was an unnecessary
luxury for this screening exercise. The hourly and daily values calculated by the mode! (that chew up
computer time like crazy) are just not needed. I suggest that either ISC-LT2 or some modification of ISC-
LT3 that runs more efficiently be used. This would allow modeling of all the sources, rather than the
methodology having artifacts like only modeling facilities which account for 95% of the load (which is a
direct result of having a model that takes forever to run). I think the statement (end of 3rd paragraph, page
36) that, "Professional judgment was used to verify that omitted facilities would not affect the analysis"is
silly. Either an analysis was done to verify that the omitted facilities didn't matter, or it was judgment,
which of course doesn't verify anything. I think using ISC-LT3, in its current configuration, for this
analysis is a big drawback. The additional accuracy of ISC-LT3 over ISC-LT2 may be more than eaten up
b> not having all the sources in the model. This could be checked fairly easily before this methodology is
sent on to another case studv.
The reason for the more restrictive criteria of the secondary and final screens (50% of the Region
III criteria) was never explained satisfactorily, other than it was a group decision. This is another artifact of
having a slow model, since with a faster model, you wouldn't have to exclude chemicals and would not
find yourself explaining why you changed criteria - it would never come up.
The issue of screening with health-based values is a real problem here, and it is one that is not
really taken head-on in the methodology. People in the community have health-based concerns and
questions, EPA does an extremely conservative first screen, and yet EPA can say nothing about the relative
safety of the air people are breathing? I know scientists are loathe to make such statements, but EPA's
policy makers, if no one else, need to think about what can be said to the community, or EPA will forever
be the (helpful) salesman and never the partner. Being "only" the salesman means that this methodology,
no matter how many technical bells and whistles are grafted onto it, will ultimately fail to be embraced by
communities. Having health-based criteria, and then punting at the end, is too confusing and looks like a
hidden agenda to sweep potential problems under the rug to many in the community.
4 Does the draft Committee Report adequately and accurately describe the screening exercise and its
results?
The draft Committee Report is quite well written and describes the project in some cases better
than the full report. 1 have several comments on it. I like the sentence under #5, paragraph 2 that says, "A
screening value is an air concentration that the Committee is confident does not pose a significant health
risk." This is about as close as it gets to saying "a safe level." It would be helpful to note here that there
uere a couple of dozen other chemicals that were found or modeled that fell below this level. Later in the
L-29
-------
same paragraph, it might be useful to point out (top of next page, sentence ending "...cannot be directly
compared.") that the State standards may also be levels that do not pose a significant health risk, but the
Committee chose its screening levels so that the committee could make the above definitive statement.
Under #10, it states that volunteers are needed, but doesn't say how one might volunteer or to
whom. The first set of figures (Figs 1-4) have no units.
One unsettling aspect of the report is that it leaves one huge question unasked and unanswered.
Why did the modeling results show essentially no chemicals above the criteria, yet the monitoring
data showed four of them? Does this mean that for individual neighborhoods where models were run and
nothing found, if monitoring data were taken there, toxic pollutants above the "safe" criteria levels would
be found? This is an important question that goes directly to the credibility of the report with the public.
Somebody out there will ask this question!
5. Is the methodology sufficiently protective of sensitive populations?
As far as I can tell, no effort was made to address this question at all in the study. It isn't the
methodology that's "protective" anyway, it's the health-based screening criteria. The way the screening
criteria were selected leads me to believe "the methodology" would allow any new committee for a new
case study to select any criteria they wish (after all, that's how it was done here!). Without some limits, this
question can't be answered.
If the question means, "Are the screening criteria as used here protective of sensitive
populations?", that's a different question, but it still can't be answered without doing the homework
necessary to come to a reasonable conclusion. This report shows no evidence of such homework, nor does
it even get into much discussion about why the criteria values themselves were selected. Without some
record of the logic used, I would have to conclude, "not necessarily."
L-30
-------
Peer Review Comments
Gail Charnley, Ph.D.
HealthRisk Strategies
L-31
-------
Comments on EPA's Baltimore Community Environmental Partnership
Air Committee Technical Report
Gail Charnley
19 November 1999
General Charges
I. Achieving goals A and B
The screening methodology, as applied in Baltimore, partially achieved goal A and has not yet achieved
goal B. Goal B involves making recommendations to improve air quality in the study area, but that issue is
not addressed in the technical report.
The screening methodology indicates that the contaminant sources evaluated do not exceed threshold risk
values. Given the conservative (health-protective) nature of the assumptions underlying the methodology,
the conclusion that those sources do not contribute to adverse health effects is likely to be correct. The
results of the project were limited by its focus only on air toxics from point and area sources, however,
which are fairly extensively regulated. Focusing on air toxics while ignoring important sources of the more
prevalent criteria air pollutants yielded an incomplete picture. Thus it is possible that poor air quality does
contribute to public health problems, but by failing to look at the whole picture, the study could not answer
the question. The report readily admits that not evaluating mobile sources is a problem. As mobile sources
appear to be major contributors to air pollution in the study area and in urban areas in general, it is
important that future efforts attempt to include them.
2. Technical improvements
The list of needed improvements is excellent. I'm not sure that including indoor air risks in the
methodology itself would be useful or practical, however. Comparing ambient air risks to some general
estimates of indoor air risks might be more useful and practical. The only improvement I might add is to
consider using a professional facilitator for future efforts. There is a growing literature suggesting that
professional facilitation by someone who is experienced in community stakeholder-type efforts is fairly
critical for success.
L-32
-------
I think that the priority of the improvements matches the order in which they are listed.
3. Achieving goal C
The partnership aspect of the project was clearly troubled. To some extent, it seems that the partnership
aspect was doomed from the start. By focusing on the question of what the risks are from air toxics, the
project was based on the implicit assumption made by the community that air toxics play a role in their
health problems. The community clearly started with an assumption that poor air quality in Baltimore
poses an unacceptable risk to their health and when that assumption was not verified, withdrew from and
condemned the project and its outcome. The project thus only partly achieved goal C. By not asking the
question—What factors contribute to health problems in the community?—and then finding that air toxics
do not contribute to health problems in the community—the project was left in the uncomfortable position
of being unable to recommend solutions to the real problem. Building community capacity to take
responsibility for their environment and their economy was thus only partly achieved. The contribution of
air quality to public health problems should have been addressed within the framework of the larger
question being addressed by the community health committee.
I believe that the screening method could be used in other communities to help understand the role that air
toxics may or may not play in public health, but it should not be used by other communities unless it is part
of a larger project looking at both other sources of air pollution and other potential contributors to public
health problems. While it was not a complete risk assessment, the method provided enough information to
draw conclusions about the likely role of some kinds of air pollution in public health and is a good basis for
priority-setting and for evaluating potential cumulative effects.
It might be helpful to make it very clear at the start what the project can and cannot accomplish because,
while it did answer the narrow question being asked, it did not answer the broader concerns of the
community.
The report should comment on how the members of the technical committee were chosen. Did the
nontechnical community members and environmental advocates participate in the selection? Trust in the
outcome might have been improved by allowing all participants to take part in selecting those who
conducted the actual screening efforts.
It might also be interesting to know how the nontechnical community members reacted to the screening
concept. I often worry that a big risk communication challenge is presented by identifying a list of
chemicals of potential concern in an early screen and then eliminating them by further screens. (Just
L-33
-------
kidding! They weren't toxic after all!) I think this problem was recognized by the air committee, but some
elaboration on their concerns and how they were addressed might be instructive.
Specific Charges
Emissions inventory. As noted above, future projects should include mobile sources.
2. Initial screen, (a) The screening methods were pretty crude, but that's why they call it an initial screen, I
guess. The methods were justified by science policy more than by science, (b) The screening criteria were
appropriate. They were certainly health-protective, but not so extreme that all chemicals were tagged as
being of concern for the next tier.
3. Subsequent screens, (a) I am not technically qualified to comment on the exposure assessment methods.
(b) The screening criteria were appropriate, for the same reason as above, (c) The assumptions underlying
the Region III risk-based concentrations are okay for a screening exercise, which this was, but not for
performing risk assessments. Some additional explanation regarding the choice of RfDs instead of RfCs
would be useful.
4. Committee report. The draft committee report accurately describes the screening exercise and its
results, but I agree with the authors that it is probably not very accessible for nontechnical community
members. The extra efforts being made to make it so are a good idea.
5. Sensitive populations. Due to the very conservative, precautionary-principle-based assumptions
underlying the screening methods, they are sufficiently protective of sensitive subpopulations. In
particular, the toxicity estimates are designed to be very health-protective.
Extraneous Comment
The box on page 23 that addresses risks and hazards perpetuates the false "carcinogens are
nonthreshold/noncarcinogens have thresholds" dichotomy. A qualifier along the lines of "For regulatory
purposes it has been assumed that. " should be added, along with the information that current scientific
evidence indicates that some carcinogens have thresholds and some noncarcinogens do not.
L-34
-------
Peer Review Comments
Douglas Crawford-Brown, Ph.D.
Department of Environmental Science and Engineering
University of North Carolina at Chapel Hill
L-35
-------
Review of Baltimore Community Environmental Partnership Air Committee Technical Report.
Douglas Crawford-Brown
Professor
Director, Office of Environmental Academic Programs
Chair, Environmental Sciences and Studies
University of North Carolina at Chapel Hill
A. General Charges
Question 1. I am somewhat divided on the answer to this question. Let me first say that the risk assessment
methods used in the report are generally of sufficient quality, and certainly go beyond those normally used
in community risk assessments. The Committee should be commended for the effort shown in this report.
The risk assessment methodology will provide conservative estimates of risk under most circumstances
and. therefore, provide a sufficient basis for claims that health is being protected. In this light, therefore, I
believe the assessment will meet the stated goals.
Still, I am always concerned when screening approaches are used to select out a set of chemicals for more
refined study. I realize the need to try to narrow the number of chemicals for more refined assessments,
especially since the final screen involves data collection on individual chemicals that can require
significant time (and would delay risk management decisions). The first screening level is presumed to
produce highly conservative results. The presumption is that the final level of screening, if it were
performed on those chemicals excluded after the first screen, would always produce risk estimates that are
lower than the values in the first screening calculations. If this is the case, the purpose of the first screening
will have been satisfied (i.e. it will have excluded chemicals that would have been shown to pose no
appreciable risk in the final screening, thereby saving resources and time).
But I see no explicit demonstration that the presumption above has been satisfied here. I SUSPECT it was
satisfied, since it usually is satisfied in my own experience, but there also are cases in which it is not
satisfied. One wav to check this would be to ensure that, for the chemicals passing all the way to the final
screen, the risk estimates under the final screen are, in fact, less than those estimated in the first screen.
This would provide greater confidence that the chemicals excluded by the first screen were not likely to
pose a greater risk under the assumptions of the final screen.
In addition, by removing chemicals at the first screen (in fact, by removing a large fraction of the
chemicals), the Committee raises the possibility that the cumulative effect of these excluded chemicals
might be appreciable even if the individual effect falls below a screening risk value. This is always a
potential problem
L-36
-------
with screening approaches and, again, I am sympathetic with the need to narrow the list of chemicals to
allow timely completion of the final screen, but the potential effect of excluding chemicals early in the
process of considering cumulative risk should be mentioned.
I also worry a bit that a formal variability and uncertainty analysis was not performed. The goal of such an
analysis would be to determine if there might be susceptible and/or sensitive individuals whose risk is
larger, and to determine the confidence with which it may be stated that risk goals have been met.
Presumably, the Committee is assuming that use of the RBCs and somewhat conservative models already
addresses these issues. This may or may not be true. An explicit statement to that effect, with supporting
evidence, would improve the assessment and give greater confidence that the public health is being
protected. The issue of variability is particularly germane given the recent EPA focus on risk to children
(initially under the FQPA and SDWA, which do not apply to air releases, but increasingly in all program
offices). The report should state whether risks to sensitive subpopulations, including children, have been
modeled adequately.
Finally, I raise an issue with Figure 5 on page 49. In that figure, it appears to me that 88% of the benzene
measured at the FMC monitoring station is unaccounted for. I am not sure what this means, and the report
is not clear. Does it mean that the measurement is a factor of almost 10 below the measured concentration?
That is how I interpret the results. If that is the case, might this suggest that the model in general is
underpredictive. and that the degree of underprediction for other chemicals might be similarly large? If that
is the case, some chemicals may have been screened out inappropriately. I am not saying this is the case,
only that the report does not provide me the information needed to determine if this is the case. Something
should be added to the report to address this concern.
Question 2. I will address the parts of this question in separate paragraphs in the order in which they appear
in the charge.
Mobile source modeling would be desirable scientifically, but it is a very difficult form of modeling.
Collecting the data bases, separating emissions by time of day and season, estimating route patterns,
estimating length of time a vehicle has been running (which affects emission rate), etc, is a daunting task,
especially when it is placed on top of the task of estimating concentrations from stationary sources. Still, it
would improve aggregate and cumulative risk estimates, and would help identify other risk management
options. Of the 8 additions listed, I rate this addition 5 (on a scale of 1 being lowest priority and 8 being
highest).
1 feel the sources identified are an adequate representation of the total sources. I believe it is unlikely
additional sources will change the risk results appreciably. I rate this addition 2.
The IRIS and HEAST databases are the appropriate ones for such information. The OSW is considering an
expedited review process for assigning toxicity (RFD/RFC and CSF) for chemicals not currently in IRIS or
HEAST as part of their HWIR project. The Committee might consider contacting that office and seeing
L-37
-------
where this process stands. I rate this addition 6 since it might cause some chemicals to enter final screening
that currently are not included due to missing toxicology information.
1 doubt that short term acute effects would be missed by the screening methodology. It is rare that these
drive the risk assessment and risk management decisions (although there are exceptions). The more
common case is that risk management decisions based on protection against more chronic effects is also
protective against short term, acute, effects. The exception tends to be when a facility is short-lived, or
emits very sporadically, but I see no evidence that these cases apply to this study. I rate this addition 3.
I do believe this is an issue, as discussed in my comments previously. The current hope in developing
RfD/RfC values, and CSFs, is that all sensitive subpopulations are included within the uncertainty factors
employed. While this may be true, it is a controversial claim at present and so the EPA has been sent back
to the issue by Congress in the FQPA and SDWA, with an explicit charge to consider if it is true for
children. At the least, the study should include consideration of the issue by determining whether any of the
chemicals which just barely missed the screen (i.e. were marginally excluded from the final assessment)
might be likely to pose risks to sensitive subpopulations not captured by the current uncertainty factors. I
personally believe the current RfD/RfC and CSF values do protect even the most sensitive subpopulations,
but it would be best to consider the issue explicitly in the study. I rate this addition 7.
I believe this is an important issue, if not in the first screen at least in the intermediate or second screen.
Cumulative exposures can now be estimated fairly routinely with existing models (such as the models used
in developing the RBCs), and may show very different results. A particular problem with considering only
the inhalation pathway (as in the first two screening levels, unless I misunderstood what was done in these
screens) is that ingestion can be a significant contributor to risk for many products of combustion. Mercury,
for example, can show a dominant pathway from seafood consumption, and dioxins can be dominated by
beef ingestion. The cumulative assessment could easily show that some chemicals excluded at the lower
screens should have been carried forward into the higher screens. I rate this addition 8.
I am not sure what is meant by this. One possibility is that it refers to the fact that pollutants in the ambient
air may enter the house, and then result in exposures that are higher than those estimated when only
ambient air is considered (the methodology used in the study does not seem to consider such a possibility).
If that is what is meant, the issue is somewhat important but not likely to significantly change the results of
the assessment. Another possibility is that indoor exposures are to be estimated based on emissions in the
home itself, as a means to provide a comparative risk assessment. It is increasingly clear that overall risks
to health may be driven more by indoor exposures than by exposures to ambient air. These indoor
exposures are caused, however, by activities under the control of individuals. My understanding of the
current study is that it was intended to identify significant sources of pollution in the ambient air, which is a
common good rather than an individual good. So, while such an assessment may help to place risks in
context, it probably would not change the overall conclusions on public health protection. I rate this
addition 1.
I am not sure what is meant by this issue. GIS is useful not simply as a communication tool, but also in
estimating risk. With respect to communication, GIS provides no more information than a well-drawn map
(in fact, the GIS data base often is obtained from such a map). So, I do not believe GIS would improve
communication, except in the sense of facilitating the production of maps that can be overlain to display
regions of highest pollution, regions where subpopulations are located, regions of sources, etc. With respect
to estimating risk, I had been presuming that the final screening used something akin to GIS to locate
L-38
-------
subpopulations for purposes of estimating exposure. If it was not used in that way, it should be considered
but not given high priority unless (1) the focus shifts from individual risk to population risk and (2) the
inhomogeneity of exposure is large even in small regions (where the additional information on location of
individuals within a grid block might significantly change the risk estimates). Still, people moving about
during the day usually obscures the additional information provided by GIS. I rate this addition 4.
Question 3. I liked the partnership and community participation displayed in this study. It is a commendable
effort and should be continued. Without it, siting, regulatory and other decisions are likely to remain more
contentious than they currently are. Having said this, it is still not clear historically whether such efforts
really improve the decision process and make it less litigious. The danger is that a lot of effort goes into
such a process, everyone participates until the final report is released, and then parties who do not like the
conclusions still sue. But at least everyone has a common point of comparison and no one can claim they
were not present when the risks were estimated. So, I am hopeful and recommend extending this method to
other communities. We are now in the position scientifically, and with respect to computation and
visualization resources, to make models available to such groups that will remove the formerly high barrier
of technical expertise needed to produce risk assessments.
B. Specific Charges
1. I believe the source inventory was adequate for this exercise. I believe it is unlikely that additional
significant sources will be identified by any more detailed collection scheme.
2. The initial screen was appropriate if the inhalation pathway dominates. The Turner concentrations
provide an adequately protective screening tool (I compared them against the results of the plume model in
the course of this review and they compared quite favorably to the highest values in the plume). I am
worried, however, that chemicals for which the inhalation pathway does not dominate will be excluded
incorrectly at this early stage. This is particularly worrisome since it has been my experience that non-
inhalation pathways are the dominant risk pathways even for combustion sources, where inhalation risks
are most likely to be significant. I believe the Committee should consider this point more carefully. A
possibility is to adjust the initial screen by multiplying the inhalation risk by a factor (above 1) that is the
highest ratio of total risk to inhalation risk under some prescribed scenario where the full pathway model
has been run.
3. The final screen was completely appropriate. I do not believe the secondary screen was really needed,
unless it was felt that the time needed to conduct the final screen on 22 chemicals was too large to be of use
in decision-making. I continue to worry about the fact that the secondary screen (as in the case of the initial
screen) does not consider aggregate risk.
4. Yes, this is a well written report that is simple to follow.
5. I believe it is, but there should be some review in the report of the reason for the Children's Health
Initiative, the FQPA and the SDWA amendments, and the implications for this study. At the least, the
report should include a discussion as to why current uncertainty factors used in developing RfD/RfC values
do or do include the sensitive individuals.
L-39
-------
Peer Review Comments
AmyD. Kyle, Ph.D.
School of Public Health
University of California, Berkeley
L-40
-------
Review of Baltimore Community Environmental Partnership Air Committee Technical Report
Draft Document prepared by the US Environmental Protection Agency
Office of Pollution Prevention and Toxics, November 5, 1999
By Amy D. Kyle, PhD MPH
Research Scientist and Lecturer
School of Public Health
University of California, Berkeley
Charges to reviewers
A. General charges
1. Did the screening methodology, as applied in Baltimore, achieve goals A and B, which were to
determine if the current aggregate levels of toxics in the air in the Partnership neighborhoods
resulting from the multiple industrial, commercial and waste facilities in and around the
Partnership may adversely affect community health and (B) to recommend actions to improve
air quality in the Partnership neighborhoods (recommendations to be based on the information
on risk-based priorities provided by the screening exercise. (C) To build the long-term capacity
of the community, including residents and businesses, to take responsibility for their
environment and economy.
The screening methodology did not address the fundamental challenge of how to consider and assess the
health significance of the aggregate burden of pollution. Instead, it winnowed down the list of chemicals
emitted through a screening process that treated each chemical, and, to some extent, each source,
separately. This does not seem to achieve the first goal of the project. There is little integration of
hazardous air pollutants and criteria pollutants.
The recommendations in the document for improvements in air quality are limited. They do not address
reduction in the overall burden of air pollution but rather focus on the four chemicals identified as being of
greatest concern individually. This approach might be more accurately described as addressing the
"worst" hazardous air pollutants rather than the aggregate burden of pollution.
L-41
-------
It is difficult to assess from a document like this whether gains in community capacity- were achieved.
Given the ultimate withdrawal of some of the original participants and lack of participation in the screening
process, it would appear that there are questions about this.
2. The report identified various technical improvements to the screening methodology. Could the
methodology, as modified with the identified improvements, help other communities seeking to
understand and improve air quality? Comment on the appropriateness of the improvements
listed below and their priority. Are there other improvements that should be considered?
As noted, this methodology does not address the fundamental question of how to consider the aggregate
burden of pollution for a community. It relies on a chemical-by-chemical assessment paradigm. This does
not appear to be responsive to the basic questions being asked by the community. Addressing the
improvements recommended by the committee, though they may be advisable, will not solve this basic
problem.
Specifically, it is extremely important to include mobile sources when assessing hazardous air pollutants.
Also, area sources, as typically defined by EPA, should be added.
With regard to the "best source" for toxicity data, the problem is not so much identifying the "best" source
but rather identifying "any" credible source for relevant toxicity data for many chemicals, especially for
inhalation exposure. The fact is that existing sources are simply not adequate. This problem needs to be
rectified for assessments like this to truly reflect health significance of pollutants. At this point, it is not
responsible to represent the toxicity database as sufficiently complete to allow for full assessment of the
likely health significance of hazardous air pollutants, even if the emissions and modeling approaches were
impeccable. An assessment based on the current toxicity databases should be represented as a likely under-
estimate.
It would be a simple change to include short-term acute effects, though unlikely to lead to important
differences in the results.
With regard to the protection of sensitive and urban populations, the issue is not simply the screening
calculations but rather that the toxicity data base does not exist for the protection of infants and children
from effects of toxic substances. With regard to urban populations, the key issue is the significance of
cumulative exposures to multiple pollutants. This methodology, as noted elsewhere, does not
fundamentally address this.
L-42
-------
With regard to adding indoor air risks, it would seem that there are sufficient issues to address for outdoor
air risks. Adding another suite of issues would not seem to be a high priority.
GIS mapping would improve the document and presentation.
Specific Charges.
Comment on the following, given goals A and B.
1. Emissions inventory - were the inventory of sources and the release and monitoring data used in
the Baltimore screening exercise sufficient and appropriate to reach the goals of the committee.'
Should additional sources be included in a source inventory to expand the scope of the
methodology for use in other communities?
As noted elsewhere, it is critical to include mobile sources when providing assessments of air pollution.
Mobile sources should be included in any future assessment project.
It is not entirely clear from the document whether what are usually known as "area" sources are included in
this assessment. This analysis appears to use a definition of area sources that is different from what is
usually meant by this term. This is rather unfortunate, as this will be confusing to any but the most careful
readers of the document. The analysis appears to consider area sources that are like impoundments or
lagoons that provide releases of air pollutants over a space that is better represented as an area than a point,
in contrast to stationary sources. However, the normal definition of area sources includes many small
sources, most of which will not have these characteristics. Area sources may be of particular importance in
cases where people live in close proximity. Future such projects should incorporate all the important
sources of air pollutants - stationary, area, and mobile sources.
Some pollutants may be present in the environment due to historical releases or may have significant
background concentrations. Carbon tetrachloride is an important example of a compound that is no longer
widely released but which remains present in the environment. To gain a complete picture of air pollutants,
background sources should be considered in addition to current releases.
The monitoring data available for this study were limited to observations from a single site. However,
these data were influential in identifying several pollutants that were not predicted to be present in
important amounts by the modeling. This might raise a red flag. It may be that, for methodology of this
type to be accepted, field confirmation of the predictions is needed. In this case, despite the representations
of
L-43
-------
conservative and health protective methods, the modeling predicts substantially lower concentrations ot all
measured pollutants than are actually found. This leads to doubts about whether the predictions are correct.
This discrepancy should be discussed in the document. If there is a reasonable explanation for the
differences, it should be presented. If there is not, then perhaps future such projects should establish and
operate monitors for the periods that are to be monitored to provide a reality check for the modeling.
It ma> be appropriate to evaluate other models that can accept a broader range of data and better
characterize pollution from sources other than stationary sources. The ASPEN model used in the EPA
cumulative exposure project appears to have achieved better correlation with monitoring data than the
approach used here. A description of this is included in a manuscript been accepted for publication. '
Additional information is posted at the EPA website on this project
(http://www.epa.gov/ttn/uatw/cep/paper.html.)
It would seem appropriate to use some verification for the estimates of releases included in this document.
These are based on permit conditions and self-reported results. Some field verification of at least some of
these estimates would inspire more confidence in the results.
When reviewing monitored data for hazardous air pollutants, it is critically important to determine the
detection limit for the methods used. Because there are not standardized methods for hazardous air
pollutants, as there are for the criteria pollutants, states may use different methods. Some methods used by
some states have detection limits that are higher than health benchmarks. It would be important to
determine whether this was the case here and, if so, how any values reported as being below detection were
handled.
It is not entirely clear that the area selected for analysis would include all sources contributing to pollution
in the target area. The document did not discuss how the geographic area was selected. For some
pollutants, transport can be important. If this methodology is to be developed for use in other situations, it
would be important to analyze carefully the spatial area that needs to be considered to capture all sources of
pollutants that might affect a neighborhood.
Rosenbaum A. Axelrad D. Woodruff TJ. Wei Y. Ligocki M. Cohen J National estimates of outdoor air toxics concentrations
Journal ot the Air and Waste Management Association 1999.49 1 74-1 85 11 do not have the published paper as yet to send to you
L-44
-------
2. Initial screen - Were the methods for calculating airborne concentrations, potential dose, and risk
appropriate and scientifically justified7 Were the screening criteria that were applied to identify'
chemicals for further analysis appropriate?
The methods used for the initial screen do not seem to be consistent with the overall goals of the project,
nor with the methods used later in the project. The document recognizes that it might have been better to
use some of the methods used in the later screen for the earlier screen.
The goal of the project was to assess the aggregate levels of toxics in the air in the partnership
neighborhoods. Yet, the first step in the project was to use a strategy of treating each chemical or
contaminant separately and screening out those not found, by themselves, to exceed a benchmark hazard
index or cancer risk estimate. This approach would appear to be at odds with the overall goals of the
project. If you want to assess the significance of aggregate pollution levels, then you need to consider the
aggregate burden of pollution and to use methods that would reflect this.
Within the approach adopted, it does not seem to make sense to use a more health protective approach to
screening at a later step in the assessment and to use a less health protective approach at an earlier stage in
the screening. Specifically, the first screening step calculates a cancer risk of the modeled concentrations
in the target area and compares it to a one in a million risk level. It also compares the dose resulting from a
modeled concentration to a reference dose. Yet, at later stages, the approach is to compare the modeled
concentrations (or monitored concentrations) to half of similar benchmarks. This does not make sense.
It does not appear that the analysis considered the question of the persistence of chemicals in the
environment at any stage. This could be important, as ambient concentrations will reflect both the input to
the area and the time that a contaminant remains resident.
The document switches back and forth between the use of the term reference dose and reference
concentration. It appears that the approach used is to calculate the equivalent of a reference dose based on
reference concentrations. This would be a per body weight dose, but derived from studies and analyses
relevant to inhalation exposure. This usage is rather confusing, as in most cases, the term reference dose is
used to refer to toxicity through routes other than inhalation, particularly ingestion, while the term
reference concentration is based on the toxicity resulting from exposure through the inhalation route.
While the approach used here may make sense, it again leads to confusion. Perhaps another term could be
selected.
A critical element in the analysis is the selection of the toxicity values used as points of comparison. It
would be most helpful if these could be clearly identified at some point in the document. The values used
for the initial screening do not seem to appear at all. Only some of those used for the second and third
L-45
-------
rounds of screening are included in the materials supplied by Region 3. It would be most helpful to pull
out the chemicals reviewed here and compile the various reference values that were used. It is very
difficult to answer this question without better information about what was used.
3. The secondary and final screens - Was the modeling approach for developing estimates of
neighborhood concentrations from multiple sources technically sound? Were the screening criteria
that were applied appropriate9 Were the assumptions built into the Region III risk-based
concentrations appropriate9
Though modeling of air pollution is not my area of expertise, it would appear from a comparison of the
modeled estimates and the monitored data that the modeling was not accurate. This suggests that it was not
technically sound.
It is not entirely clear what assumptions are being referred to here, with regard to the Region III risk-base
concentrations.
4. Does the draft Committee Report in Appendix J adequately and accurately describe the screening
exercise and its results?
The draft committee report is somewhat difficult to follow and would benefit from the addition of graphics.
That said, accuracy could be improved with regard to the issues identified below.
First, Appendix J implies that the modeling captures all of the facilities that are contributing pollutants to
the area. Facilities are included only if their emissions exceed a screening level. This means that the
modeling will under-predict the overall concentrations.
Second, the appendix does not reveal the discrepancies between the model predictions and the monitoring
results. These cast doubt on the accuracy of the modeling. This should be disclosed and discussed.
Third, the appendix does not fully describe the sources that not included in the exercise.
Fourth, the discussion of the screening levels does not explain that each chemical was compared separately
to the cancer screening concentrations. The overall cancer risk that might result from combining exposures
to many chemicals, each of which is below the screening target, was not assessed. This seems to be
obscured in the report.
Fifth, the descriptions of the limitations of the study seem to point to issues that are less relevant than the
genuine limitations of this analysis. This appears to suggest that the principal limitation is a lack of data on
time and activity patterns. However, there is nothing in the charge to the group suggesting that people
expected this kind of detailed information. It appears that they expected an assessment of outdoor
concentrations overall. This might be seen as a lower bound on the exposures that individuals might
experience, because concentrations are often higher indoors than outdoors. It would be more fair to this
process to point out the limitations of the study to answer the initial questions of the people in the
L-46
-------
community rather than to point to additional research questions not initially included. Similarly,
explanations that emphasize the significance of diet and heredity seem quite beside the point of this
analysis, which is supposed to focus on air pollution.
Sixth, the document does not provide the best available estimates of outdoor concentrations of these
chemicals, but only of certain of the chemicals that passed a screening process.
5. Is the screening methodology as used in Baltimore sufficiently protective of sensitive populations*
Please suggest any improvements of this aspect of the screening methodology.
See previous comments.
Page-specific comments.
Page 5. Given the erosion of participation in this project, the sponsors might consider whether it is
consistent with the initial design to move forward with a report.
Page 13. The potential for violation of permit conditions is not addressed in this methodology.
Page 16. To reach conclusions from an analysis such as this, it would be important to include all pollution
sources, including those noted at the bottom of page 16 as being excluded.
Page 23, first full paragraph. It would seem to be important to have community representation during the
selection of screening levels. The lack of representation is troubling.
Page 23-25. A table of values used should be included here. An assessment of the data gaps in the
underlying toxicity database should also be included.
Page 26: calculation of the air concentration and potential dose. This method appears to compare the
estimated concentration of each chemical at each facility to a screening value. If this is the correct
interpretation of the text, it is difficult to determine how this would integrate exposures from multiple
sources. If each often sources of a chemical each produced a concentration below the screening level, it
would be excluded. Yet, taken together, they might result in a concentration of concern, even for a single
chemical.
Page 27. For air pollutants, the assumptions of exposure duration of 24 hours per day, 365 days per year,
may not be particularly conservative for urban populations. Pollution concentrations are fairly consistent
in urban areas; there are not many places people can go to reduce their exposures.
L-47
-------
Page 29. table at the top of the page. It is not clear from this table whether the entries represent what might
come from one facility or from all of the facilities for the chemicals identified.
Page 30. Should include a summary of the monitoring results, with all chemicals and annual mean values.
Page 30, box. The reasons for excluding these chemicals should be further developed. Some of these
chemicals can also have area sources and should not be quickly excluded. Having a committee use
"professional judgment" to exclude chemicals without clear explanation is not a transparent process.
Page 30, last paragraph. It would be important to address aggregate exposure at the initial screening step.
Otherwise, sources and chemicals have already been excluded. The results described here should be
demonstrated in the report.
Page 3 1, first full paragraph. Several of the criteria pollutants are mentioned here as being included, but
the methodology does not seem to address these pollutants.
Page 37, first paragraph. The alternate definition of an area source is given here. This is very confusing.
It also appears that those sources usually defined as area sources are not included in this analysis.
Page 37, second paragraph. It would strengthen the analysis to demonstrate the actual emissions are indeed
below permitted levels. Compliance or other data might be available to allow this.
Page 40, last paragraph. It would seem appropriate also to consider overall cancer risk.
Page 49. This chart requires some explanation. Again, it would appear to demonstrate that the modeling
was not technically sound.
Appendix D. Should include the Region III table, the ATSDR MRLs and the IRIS values.
Appendix I. Should be highlighted. The contrast between the predicted and measured values is striking.
Check the detection limit for vinyl chloride.
L-48
-------
Peer Review Comments
Kenneth L. Mitchell, Ph.D.
U.S. EPA Region 4
L-49
-------
Kenneth L. Mitchell, Ph.D.
L'SEPA Region 4
November 29, 1999
GENERAL CHARGES
5.Did the screening methodology, as applied in Baltimore, achieve goals A and B?
In a broad sense, the Study did achieve the goals outlined in A and B. The analysis did lead to an
assessment of the levels of toxics in the Partnership neighborhoods that may adversely affect
community health. And the report does include specific actions to improve the air quality in the
area. However, as discussed below, the efficacy of the methodology used to accurately reflect the
potential health impacts of air pollutants can be improved upon.
6.The report identifies various technical improvements to the screening methodology. These are
listed below. Could the methodology (emissions inventory, initial screen, secondary screen, final
screen), as modified with the improvements identified below, help other communities seeking to
understand and improve air quality? Please comment on both the appropriateness of he
improvements listed below and their priority. Are their other improvements that should be
considered?
The methodology, as modified with the improvements identified below, could help other
communities seeking to understand and improve air quality.
l.Add mobile source modeling: The Baltimore exercise focused on stationary and area
sources. This task will expand the capacity of the methodology to include mobile source modeling.
The methodology would benefit strongly from the inclusion of mobile source emissions
and an evaluation of their impact on the overall concentrations of toxic chemicals in
ambient air. (Indeed, the document would also benefit from some analysis of the impact of
all criteria pollutants as well.) It is clear from recent modeling exercises (USEPA, 1999)2
that mobile sources can have a very significant impact on the overall quality of air,
particularly in urban areas. An appraisal of these sources will lead to a much better
understanding of the problem at hand as well as more effective strategies for protecting
public health. (High priority)
:See http:\\www.epa.gov\cumulativeexposre
L-50
-------
2.Review and improve the source inventory review: Review existing source inventories to
identify additional sources of emissions to ensure that all significant sources are included.
A well developed emissions inventory is crucial to the success of a screening process.
However, there is a point of diminishing returns where tracking down every small release
may provide little additional information (unless there is some reason to believe that there
are so many small sources that, in toto. they would prove a significant source). Based on
m> more limited knowledge of building source inventories, the level of detail identified in
this document for the development of a source inventory seems appropriate and should
suffice to meet the goals of the project. (Medium to low priority)
3.Identify the best source for toxicity data: Compare available toxicity data bases to identify
the most accessible and complete source(s) of data for community screening exercises.
It is crucial that toxicity values which have been peer reviewed by persons knowledgeable
in the field of toxicology and epidemiology be used to evaluate potential health impacts for
toxic air pollutants. Given that a number of such values may exist for any given chemical,
it is also crucial for trained scientists to review the available literature and select toxicity
values that are scientifically supportable.
A complication in the toxicity factor selection process is that a number of science policy
decisions must be made. For example, if a particular chemical is generally considered to
be a potential human carcinogen, but there is disagreement over the published findings in
the toxicological or epidemiological literature about its relative potency, a decision must be
made as to whether and how far one will go in developing a carcinogenic potency slope
factor. Any number of other "science policy" scenarios can be mentioned which affect
almost any health assessment (including the one described in this document).
It is crucial, therefore, that before a study begins, the stakeholders identify a hierarchy of
toxicity data sources as well as decisions on how they will address the numerous science
policy issues that will come up during the assessment. The assessors should then apply
these decisions consistently throughout the entire process. For example, in this document,
step 1 apparently relies on IRIS and HEAST toxicity values only. Step 2, however, uses
the Region 3 RBC methodology (which relies on IRIS, HEAST, and several other sources
of toxicity information). Unless there is good reason (e.g., updated toxicity studies),
"changing course in mid-stream" on toxicity issues or science policy determinations can
seriously compromise the overall supportability of an assessment.
This is not to say that the process cannot include flexibility. Indeed, stakeholders may wish
to delve into the literature in their search for a supportable toxicity value. Nevertheless, a
L-51
-------
process for performing such evaluations should be established at the outset of the
assessment, with a clear understanding of when such an analysis will be undertaken and by
whom. (High priority)
4.Expand the Baltimore methodology to include short-term acute effects.
Acute toxiciry is clearly an important issue for communities and should generally be
addressed by the methodology. One issue with acute toxicity evaluations is the lack of
consistently derived toxicity values appropriate for the types of exposures that would be of
concern in such evaluations (i.e., acute toxicity values protective of the general public
under routine exposure conditions).
Similar to chronic toxicity information, it will be crucial for any acute exposure evaluation
to clearly define the rationale for the selection of the toxicity values used in an assessment.
For example, the use of occupational values divided by some uncertainty factor would need
clearly stated and supportable evidence that such a methodology would result in screening
values appropriate for the exposures at hand.
One recent attempt at deriving acute toxicity values protective of the general public under
routine exposure conditions was undertaken by the California EPA. We suggest reviewing
their methodology for developing Acute Reference Exposure Levels3 if acute assessments
are to be included in a later edition of this methodology. (High priority)
S.Review the screening calculations to determine if they are appropriate for and protective of
sensitive and urban populations.
As noted elsewhere in this document, the current screening calculations should be
reviewed with an eye towards establishing and documenting the logic behind the screening
process as well and the numerous technical details that form the basis for the methodology.
In its current state, there are technical flaws which call into question the appropriateness of
this methodology for evaluating impacts to sensitive and urban populations. (High
priority)
California EPA (1999), Technical Support Document for The Determination of Acute
Reference Exposure Levels for Airborne Toxicants as part of the Air Toxics "Hot Spots"
Program Risk Assessment Guidelines, Office of Environmental Health Assessment, March
(http://oehha.ca.gov/scientific/acuterel.htm).
L-52
-------
6.Develop a method to screen for cumulative exposures in the Initial Screening Step.
The initial screening step should take into account the potential for aggregate risks and
hazards from contemporaneous exposures to multiple carcinogens and noncarcinogens.
One way to do this is to use the maximum concentration found or estimated within the
study area and to compare it to an individual chemical concentration that is set at a level
which, in and of itself, accounts for the potential for multiple chemical exposures. For
example, carcinogenic screening numbers could be set at a level of 1E-06 and noncancer
screening numbers could be set at a hazard quotient of 0.1. These values are selected for
the following reasons:
Carcinogens: The level of 1E-06 is selected since it would take simultaneous exposure to
20 chemicals all present at a level of 1E-06 to collectively reach a cancer risk of 1E-04, the
commonly accepted upper end of acceptable risk. Since this would be an unlikely
situation, the screening level of 1E-06 is a reasonable and conservative starting point for
the screening process.
Noncarcinogens: The hazard quotient of 0.1 is selected since it would take a simultaneous
exposure to 10 chemicals all present at a hazard quotient of 0.1 to collectively reach a
hazard index of 1. the commonly accepted upper bound for noncarcinogenic chemical
exposures. Since the toxic effects of noncarcinogens range widely across a variety of
metabolic mechanisms and target organs, it is unlikely that one would be
contemporaneously exposed to 10 chemicals all present at a hazard quotient of 0.1 and all
exerting the same toxic effect. As such, the screening level of 0.1 for an individual hazard
quotient is a reasonable and conservative starting point for the screening process. Similar
to the screening of carcinogenic chemicals, the maximum concentration found or estimated
should be compared to the screening value in this first screening step. (High priority)
7.Expand the methodology to include indoor air risks to provide a more comprehensive
picture of air risks.
Whether or not to include indoor air risk is very dependent on the goals of the project. If a
goal is to provide a more comprehensive picture of overall air risks, the stakeholders must
understand from the outset that the sources and types of indoor air contaminants can be
very different from those in ambient outdoor air. In addition, stakeholders must also
understand that indoor air across a geographic region can be highly variable, making it
difficult to assess
1 We presume that the authors mean "cumulative" here to be the sum total of contemporaneous toxic exposures
to carcinogens and noncarcinogens by the inhalation pathway. We suggest avoiding the use of this term since EPA
is currently evaluating the concept of "cumulative risk" to include multiple pathways. Cumulative risk, in that
sense, means a more holistic evaluation of risk than that posed by just one pathway.
L-53
-------
in a representative fashion for inclusion in a comprehensive risk-based screening
assessment. This is not to say that any assessment should not at least discuss the
prevalence and effects of common indoor air pollutants (e.g., second hand smoke).
(Medium to Low priority)
S.Incorporate GIS mapping to enhance the communication of the modeling and screening
results.
This is an excellent suggestion and every effort should be made at the outset of a project to
incorporate this vital tool in not only the analysis of data, but also its presentation.
However, a note of caution is appropriate. It is very easy to put environmental and public
health data on a map and draw conclusions. It is more challenging to put environmental
and public health data on a map correctly and come to the correct conclusions. Factors as
simple as the scale chosen for mapping data can have a strong influence on the ultimate
interpretation. Extreme care must therefore be taken when deciding to map data using GIS.
Ultimately, stakeholders must understand the limitations of GIS, the level of data that will
be needed to draw supportable conclusions, and the high level of resource requirements
(including necessary specialized technical expertise) before committing to using this tool.
(Medium to High priority)
9. A re the partnership and community participation aspects of the screening exercise
described in the case study and in the lessons learned section appropriate to achieve goal C? Could
this screening exercise be used in other geographic areas to reach this goal? Can you identify any
improvements or changes in the screening exercise that would help accomplish this goal?
The technical document and lessons learned section of this document do a reasonably good job of
describing the process of identifying and including appropriate stakeholders in setting up, running,
interpreting, and communicating a screening evaluation and results. While these activities are the
important foundation for Project Goal C, this Project Goal is more prospective in scope. In other
words, Project Goal C is really geared towards how to use the results of a properly carried out
screening project to take action, not simply how to get people together to do a screening project.
In that sense, this document does not meet the needs of Goal C, nor could it be used as an example
for other communities attempting to meet this goal.
To achieve Project Goal C, stakeholders must all agree up-front to a plan of action that is
dependent, in part, on the outcome of the screening evaluation. This is commonly done by
developing a "Risk Management Plan" prior to performing any screening level work. The contents
of such a plan can include information on acceptable risk levels, guidelines for voluntary pollution
prevention activities, funding and education to enhance stakeholder involvement in carrying out
these actions, and strategies for sustainable development that meet the need to maintain a health
environment. The Plan may even go as far as to envision changes in existing statutory or
regulatory authorities to effect environmentally beneficial results. Ultimately, the plan can say
anything the stakeholders want. However, having such a plan and obtaining buy-in from all
affected parties prior to beginning
L-54
-------
the screening process will form the basis for Project Goal C to be achieved. The current document
appears to include very little of what could be described as a risk management plan. (High priority)
SPECIFIC CHARGES
1. The Emissions Inventory. Were the inventory of sources and the release and monitoring data
used in the Baltimore screening exercise sufficient and appropriate to reach the goals of the
committee? Should additional sources be included in a source inventory to expand the scope
of the methodology for use in other communities?
See responses to 2a and 2b under General Charges above.
2. The initial screen: (a) Were the methods for calculating airborne concentrations, potential
dose, and risk appropriate and scientifically justified? (b) Were the screening criteria that
applied to identify chemicals for further analysis appropriate?
a. The method selected appears to be reasonable for calculating airborne concentrations,
potential dose from a predicted concentration, and risk/hazard. However, comments given
elsewhere in this review should be taken into account to refine the method to make it more
justifiably conservative as a first step in a tiered screening approach. For example,
noncancer doses should be compared to a HQ of 0.1, not 1.
b. As noted elsewhere, a modification of the screening criteria would make this initial step
more conservative and more appropriate.
In addition, there are several troubling statements in the document regarding the addition or
deletion of chemicals based on "professional judgment" (see pp. 30-31). Such decisions
must be thoroughly documented so that anyone may see the precise logic behind the
decision. For example, consider the phrase (p. 30) "Aldrin, acrylamide were not
selected for further evaluation... because the professional judgment of the Committee
determined that the chemicals did not present a risk to the community." A stakeholder not
involved in this decision would be quite justified in questioning this statement (given the
lack of supporting documentation). Also, while there is some logic to including chemicals
for which there is no toxicity data, one could also make the argument that refining their
airborne concentrations by modeling is an extraneous exercise since one still does not
know what such refined concentrations mean lexicologically. The document should
discuss this uncertainty.
3. The secondary' and the final screen: (a) Was the modeling approach for developing estimates
of neighborhood concentrations from multiple sources technically sound? (b) Were the
L-55
-------
screening criteria that were applied appropriate? (c) Were the assumptions built into the
Region III risk-based concentrations appropriate?
a. Based on my more limited knowledge of modeling, the approach appears to be technically
sound with the caveat that the document is extremely ambiguous on how and why the
receptor grid system and selected receptors were selected. For example, why was a coarse
grid system even contemplated (since it was not subsequently used) and how were the
receptors points that represent the four neighborhoods selected (are they located at census
tract population centroids, near sensitive subpopulations, etc.?). Also, are the modeled
concentrations used in the screening at a grid receptor the aggregate concentrations from
all sources? What was compared to the screening level (the maximum annual aggregate
concentration at a receptor)? Where is the monitoring station on the receptor grid and was
this also selected as a modeling receptor point?
b. The screening criteria could be appropriate had they not been juxtaposed with a different
set of screening criteria in Step 1 (different toxicity values, etc.). For example, the Region
3 RBC values are commonly used for screening contaminant levels in environmental media
and are appropriately used in this evaluation. However, they include a set of presumptions
about exposure that are logically inconsistent with the screening criteria used in Step 1
(presumably the most conservative step). Specifically, Stepl presumes an adult exposed
for a lifetime. The RBC values, on the other hand presume (for carcinogens) a person
exposed for only a portion of a lifetime (30 years), part of which is exposure as a child and
part as an adult. Apparently the Committee intended to deal with this inconsistency by
dividing the RBC values in half. While dividing a carcinogenic RBC value in half gives a
value approximately that of assuming an adult exposed for a lifetime, for noncarcinogens
the same operation gives a screening concentration that is half that of the Step 1 screening
values. This is because, for noncarcinogens, the exposure duration term cancels out of the
hazard equation (i.e., the length of exposure is irrelevant). Thus, the Committee has
selected, for noncarcinogens in Step 2, screening values that are twice as conservative as
those of Step 1. And Step 1, by definition, is supposed to be the most conservative step.
One way to correct this inconsistency would be to reconstruct the overall screening process
as follows:
(i) Select a conservative set of screening values (e.g., the Region 3 RBC values).
(ii) Use these values at a level of 1E-06 for carcinogens and one-tenth their value for
noncarcinogens (to account for possible contemporaneous exposure to multiple
noncarcinogens that have the same mechanism of action or affect the same target organ).
(iii) Calculate concentrations as described in Step 1 (i.e., using Turner's method) and
compare the MAXIMUM concentration found or estimated in any airshed to the screening
level. Keep only those chemicals that fail the screen.
L-56
-------
(iv) Perform modeling as in Step 2 on those chemicals that failed the initial screen.
Compare concentrations at selected receptor points and monitoring stations to the SAME
screening levels used in the initial screen. Keep only those chemicals that fail the
secondary screen for any given airshed.
(v) Use refined modeling to compare the failing chemicals to the SAME screening levels
used in the initial screen. The chemicals that continue to fail are then the ones targeted for
reductions.
Ultimately, such a screening methodology maintains a consistent set of toxicological values
to derive screening levels at a set level of risk or hazard (all conservative since this is still
only a screening method - not a risk assessment). One simply refines the actual
concentrations in air from conservative to more realistic. In addition, one may also build in
the option to use modeling results at a monitoring position, rather than the monitored
values themselves, depending on site specific circumstances (e.g., problems with the
credibility or age of the monitoring data).
c. The assumptions build into the Region 3 RBC values are reasonably conservative and
generally appropriate for screening programs such as the one described in this document.
However, the values should be reevaluated as we learn more about exposure patterns and
responses, or have reason to believe that the exposures presumed by the RBC methodology
are not protective for a particular site. For example, the RBC table presumes an exposure
duration of 30 years (based on residency evaluations). If a particular population is known
to be less mobile than that presumed by the RBC methodology, alterations to that
methodology (i.e., to derive more strict screening values) would be in order.
4. Does the draft Committee Report (see Appendix J) adequately and accurately describe the
screening exercise and its results?
With a few exceptions, the Committee Report and the technical document are consistent. However,
we suggest addressing the following points:
a. Appendix J indicates that only carcinogenic screening values were used in the screening
process. This was not the case.
b. Appendix J also tends to give details not present in the technical document. If anything,
the technical document should include everything in Appendix J. For example. Section 5 of
Appendix J indicates that the model was used to determine chemical specific aggregate
concentrations at grid receptors. The technical document is more ambiguous on this point.
Likewise, Appendix J goes into details about what is being done, say, on the national level
about air emissions, whereas the technical document provides less detail on this point.
L-57
-------
5. Is the screening methodology as used in Baltimore sufficiently protective of sensitive
populations? Please suggest any improvements of this aspect of the screening methodology.
With the modifications suggested elsewhere in this comment document, the Baltimore evaluation
could be sufficiently protective of sensitive populations. For example, the modeling efforts should
much more clearly define why grid receptors were chosen where they were. If these grid receptors
do not include the locations of sensitive subpopulations, any new evaluation should be augmented
to include the locations of such populations located in the study area (i.e., day care facilities,
schools, nursing homes, and hospitals).
ADDITIONAL COMMENTS
The technical document suffers from a critical lack of detail in both the both the logic of the
selected screening process as well as the scientific basis for the methodology. While a verbatim
recitation of standard technical detail and policy is not necessary, sufficient citations to relevant
texts are, and there are virtually no citations in this document. In short, anyone should be able to
pick up this document and be able to understand exactly how the authors arrived at their
conclusions.
Carol Browner's policy on the development of Agency risk characterization2 intimates that all such
Agency documents must be clear, transparent, reasonable, and consistent. While the Baltimore
methodology does not present a "risk characterization" per se, it should nevertheless meet the spirit
of the risk characterization policy. As such, it is suggested that this document be rewritten with an
eye towards including substantially more detail.
There are a number of examples of risk screening methodologies that have been evaluated and
tested, but which are conspicuously absent from this document. Indeed, there is the appearance of
this methodology having been developed quite de novo. We suggest that the authors review
alternate methodologies and include a thorough discussion of these methods in the text of the
technical document. The purpose of such a discussion would be to show that the developers of this
methodology reviewed and understood the existing literature on the subject of environmental
screening methodologies and adapted it to the specific needs of the Baltimore study. Some
example methodologies that provide insight into the environmental screening processes include:
- Guinnup, David E., A Tiered Modeling Approach for Assessing the Risks due to Sources of
Hazardous Air Pollutants, USEPA Office of Air Quality Planning and Standards (EPA-
450/4-92-001).
2USEPA (1995), Policy for Risk Characterization at the US Environmental Protection
Agency, Office of the Administrator, March 21.
L-58
-------
Standard Guide for Risk-Based Corrective Action Applied at Petroleum Release Sites,
American Society for Testing and Materials (El 739-95el), West Conshohocken, PA, 1999.
Page 1 7, first paragraph states that the partnership area included ZIP codes 21225 and 21226, but
then goes on to include 8 additional ZIP codes. We suggest clarifying the exact boundaries of the
study area and highlighting it on a map.
Page 17, the first paragraph indicates that permitted facilities and TRI facilities were used to make
the final list of master facilities. We suggest describing the types of facilities that require permits
under Maryland law. As written, one is left wondering whether there are numerous unpermitted
facilities, the emissions from which (collectively) could amount to a large portion of the overall
environmental load.
4. We suggest including a table that summarizes the emissions inventories that were queried, the type
of data available (i.e., chemicals reported and type of emissions data such as total pounds released
per year, etc.), the years data was available, the specific data element that was ultimately used in
the screen, and a rationale for inclusion in the analysis. For example, if TRJ data was available for
multiple years, which year was used in the screen and why?
5. The discussion related to the Fairfield monitoring site (page 30, first full paragraph) indicates that 4
years of data have been collected from which annual average, minimum, and maximum
concentrations were available for 41 different chemicals. Which year was used in the screen?
Which value was compared to the screening value? The maximum? The annual average?
(NOTE: The use of the maximum monitored values or estimated value for any source is
particularly important in Step 1 of the screening methodology, since aggregation of source
contributions is not performed.)
6. The text of the technical document often provides a range of years for which data is available, but
for which the analysis apparently focuses on just one year. For example, the first full paragraph on
page 19 indicates that ambient air monitoring data from the five Baltimore sites for 1992-1996
were compared to the monitoring station in the partnership area (in Appendix J). A review of
Appendix J, however, shows that this analysis was for only one year (1996) and only 4 chemicals.
7. We suggest clarifying the text to indicate that the screening value at a grid receptor is the sum total
for a chemical from all modeled sources. This is not clear in the document.
8. Page 27, sentence beginning "A very conservative estimate...," this paragraph indicates that an
inhalation rate of 1 n\3fh is presumed. However, the document then goes on (in the highlighted box
on page 28) to state an inhalation rate of 20 m3/d. The second inhalation rate (i.e., 20 m3/d) is
correct and should be used consistently throughout the analysis for adults.
L-59
-------
Page 40, the section on grouping chemicals according to "similar organs or physiological systems"
needs to be reconsidered for the following reasons:
- Apparently only respiratory and neurological effects were evaluated (with the neurological
evaluation missing from Appendix I). Any analysis of the disaggregation of hazard indices
should consider the full range of mechanistic and target organ effects. There is no
rationale provided for the selection of these two effects or whether these are even the
critical effects for the chemicals evaluated.
The "target organ effect" analysis is generally only used in the determination of whether
hazard indices in a risk assessment should be disaggregated based on mechanism or target
organ effect. What apparently has been done here is to compare modeled concentrations of
chemicals exerting similar toxic effects to screening levels to determine if they exceed (in
aggregate) these screening values. In concept, such a comparison can only be made
comparing doses to toxicity metrics (RfDs) to derive a hazard quotient. The additivity of
the various hazard quotients is an assessment based on mechanism of toxicity or target
organ effect.
Appendix I indicates, however, that comparisons of doses have been made to a variety of
screening levels, some of which are not toxicity metrics (e.g., sulfur dioxide is compared to
the National Ambient Air Quality Standard - NAAQS - for this compound). This results in
the development of hazard quotients and pseudo-hazard quotients which cannot be added
using the hazard index methodology. Adding such values together leads to an entirely
erroneous result. It is suggested that the authors either consult a toxicologist with
demonstrated experience in the application of the principles of the hazard index
methodology or drop the analysis from the document entirely.
The above discussion highlights a related problem that recurs throughout this analysis:
namely, an undocumented selection of toxicity and pseudo-toxicity metrics and the use of
screening values which are not toxicity metrics to quantitate risk or hazard. As noted
previously, NAAQSs are not toxicity metrics and cannot be used as such. Neither are
ACGIH TLVs divided by an uncertainty factor (sulfuric acid), nor ATSDR MRLs.1 We
suggest reevaluating the basis for toxicity metric selection and to apply it consistently
throughout the document.
Please note that none of this is to say that concentrations should not be compared to non-
toxicity metric screening levels. For example, comparison of air concentrations to the
NAAQS is not only permissible, but desirable. The point is that such an analysis cannot be
subsequently used in assessing hazard quotients or additivity of hazard quotients using the
hazard index approach.
1 ATSDR MRLs can theoretically be used, under limited circumstances, as a toxicity metric due
to the similar nature of their development to EPA RfDs. However, a justification must be made
for such a use, and the uncertainties of the analysis documented.
L-60
-------
10. Table 3 on page 46 indicates "NA" for a number of secondary screen emission rates which have
final screen emission rates. How can this be0 If the final screen is a refinement of the secondary
screen, the secondary screen should have emission rates for all of these chemicals.
Table 4 on page 48, we suggest discussing why some of the estimated concentrations in the final
step are higher than estimated concentrations from the secondary screen. One might presume that,
given the supposed increasing conservativeness of the screen steps as one goes from Step 3 to Step
2 to Step 1. that Step 3 estimates might be less than those of Step 2. We also suggest adding the
three monitored chemicals to this table to make it more comprehensive.
12. We suggest making the screening methodology flexible when determining whether to move from
one step to another. Generally, screening methodologies of this sort may or may not complete all
steps, depending on site-specific circumstances. For example, the initial screen might clearly point
to one source as the primary emitter of concern. Spending more time and money on screening
would probably not change that conclusion. In this instance, stakeholders might decide to take
action after the first step and drop any further analysis.
13. Page 62 indicates that one lesson learned would be to verify modeling results with monitoring
results. Performing this analysis should not be a lesson learned for this document. Rather, it is
crucial that this analysis be done for this version of the document since this is the primary way, in
this study, to "ground-truth" the estimates from the model.
14. For the Fairfield monitor, the document should state whether it is a source-oriented monitor or a
community-based monitor. A source-oriented monitor is positioned specifically to determine
whether a particular source is affecting a particular population. A community-oriented monitor is
positioned so as to provide readings suitable for estimating exposure over a larger geographic area
(e.g., a large urban area). This same comment holds for the other Baltimore area monitors
mentioned in Appendix J.
I 5. We suggest reviewing Appendix G for accuracy. For example, cancer slope factors are given as
mg/kg-d rather than (mg/kg-d)"1 We also suggest removing extraneous information that is not used
in the screening process (e.g., the waste minimization prioritization tool - WMPT - information).
16. Appendix I, Table 1, the "Screening Comparison Concentrations" are not one-half of the Region 3
RBC values, as indicated in the text. For example, the screening value for ammonia is given as 100
ug/m3 One-half the RBC value is 50 ug/m3 We suggest revising this Appendix and the text to
match. (Also note that there are not similar screening comparison tables for Steps 1 and 3, but
there should be.)
We reiterate an aforementioned comment here, given how crucial it is to the overall success of the
screening process. Table 1 of Appendix I illustrates that there is little documentation or
justification for the selection of screening levels (for any of Steps 1-3) or how they are applied. We
strongly suggest revisiting this question and revising the methodology accordingly.
L-61
-------
17. Appendix J, Numbers 9 and 13 should be enhanced to more fully describe the Clean Air Act
requirements to address air toxics in urban air, including a more thorough discussion of MACT
standards, the residual risk program, cleaner fuels, etc.
18. Appendix K, the discussion of the modeling provided in this Appendix does not match the text of
the body of the technical report (e.g., the text does not talk about multiple modeling scenarios).
Why were the 5 years for modeling (1987-1992) selected instead of more recent years? Were the
results from these different modeling years evaluated separately or combined in some way?
19. What was the rationale for the selection of the discrete neighborhood receptors (e.g., Cherry Hill at
a given lat/long)?
20. The document should include a much more full description of the airsheds and meteorology of the
area. Basic information such as windroses is missing from the document and should be included to
frame not only the problem, but also for use in developing appropriate solutions.
21. The document should include a thorough analysis of uncertainties associated with the assessment
and their effect on the analysis outputs. Only by including such an analysis can one determine
whether decisions can be made with the current level of analysis or whether additional work must
be performed (to reduce existing uncertainties) before any risk management decisions can be made.
L-62
-------
Peer Review Comments
Ronald E. Wyzga, Sc.D.
Electric Power Research Institute
L-63
-------
Comments on Baltimore Community Environmental Partnership Air Committee Report
Overall Comments
The document presents a tiered approach to evaluating community risk due to modeled levels of air
contaminants in neighborhoods of southeast Baltimore City and contiguous Anne Arundel County,
Maryland. The approach begins with a screening list of chemicals of interest from TRI and state release
inventories, and an inventory' of fixed facilities in the area that might emit those substances to air. Then
three successive air quality modeling and constituent screening exercises are carried out to calculate
potential incremental health risks in the neighborhoods being studied. The resulting calculations showed
that only benzene was identifiable as being both currently emitted from the inventoried fixed sources, and
posing a potential air concentration above the risk-based concentrations used for screening levels of
concern. The remaining three chemicals were identified as not being due to emissions from current fixed
sources (1,3-butadiene, methyl chloride, and carbon tetrachloride).
The general approach taken seems reasonable, although there are significant gaps in the information
provided about the conclusions reached. In particular, no explanation is given for screening or higher level
analyses of chemicals whose primary exposure route of concern is ingestion or other non-inhalation
pathways. Although both dioxins and mercury, for example, are listed as having been selected in Level 1
screening because of risks levels of concern (hazard quotient > 1 or cancer risks > 10"6), these chemicals are
of concern primarily by ingestion routes indirectly through foods. In particular, dioxins are lipophilic, so
are of concern due to ingestion of meats and dairy products, while mercury requires fish ingestion. Yet no
discussion is provided of the manner in which screening risks were calculated for these chemicals. Nor is
any discussion provided of whether these chemicals arise from local sources, or from "ambient" levels
(levels in background media with no attribution to local sources). Thus it is not clear about how such
chemicals can be screened in or out of the subsequent analyses.
Although the approach is reasonable, its limitations make it of limited value. The lack of congruence
between the methodology results and the monitoring data is disturbing. It suggets that the results of the
current methods are of questionable value.
General Charges
1. I don't believe that these goals were met. The greatly limited emissions inventory would not allow any
reasonable assessment of community health impacts. It is imperative to consider mobile sources,
volatile emissions from landfills, etc. and small sources. Each of these has potential to contribute
significantly to community risks. Analyses by EPA (1990) indicate motor vehicles and related
activities (fueling & fuel processing) may account for about 75% of their calculated excess cancer
cases nationally, or 75% of 1,700 to 2,700 cases annually. Since then, the unit risk for 1,3-butadiene
has been re-evaluated and cut by a factor of about 3, but it is likely that other fuel constituents play a
significant role that was unaccounted for in 1990. One reason given for the meaningful divergence
between the monitored values of 1,3-butadiene, methyl chloride, carbon tetrachloride, and benzene is
the potential emissions from wastes sites and landfills. If these emissions are sufficient to be monitored
and trigger risk concerns, they cannot be ignored. Small sources could also be important. I'm not sure
that the current methodology, for example, would capture the impact of a small dry cleaning
establishment whose emissions of perchlorethylene might reach immediate neighbors.
L-64
-------
Without any reasonable characterization of risks, the methodology is of little value in aiding the
development of risk-based priorities.
The most important conclusion that I make from this exercise is the importance of monitoring. The
method did not identify the greatest potential risks; monitoring activities did. I would urge the
expansion of monitoring to include other sites and a full suite of toxics about which there is concern.
This has far higher priority than the extension of a methodology "whose results to date are not validated
by monitoring data.
If the methodology is to be extended, the most important improvement is the development of a
comprehensive emissions inventory. See my comments above. This is not an easy task for the sources
currently missing; perhaps community involvement could help here.
To be consistent with other EPA studies, toxicity data should be from the IRIS database. It should be
recognized and communicated to the community that the unit risks and RfDs/RfCs are conservative
numbers designed to be protective; risks derived from them are upper limits. The IRIS numbers,
however, are based upon a thorough (although sometimes out of date) review of the literature and their
derivation is well-articulated.
Short-term acute effects could be important; their consideration need also includes potential accidents,
which would require all types of probabilistic assumptions. The consideration of acute effects and
exposures would also present modeling problems. I would urge the study group to estimate the chronic
risks correctly before venturing off into an even more difficult area.
The EPA risk assessment guidelines (and the data and methods applied) make provision for sensitive
individuals. Unless there is good reason to suspect that these are not sufficiently protective for the
population under study, I would not revise them.
we have
I would give lower priority to applying GIS mapping systems and cumulative exposures until
far more confidence in the existing results.
I would ignore the indoor environment; this would require too many assumptions and would not be
appropriate for this study. Where the indoor environment would mitigate ambient concentrations may
be of interest, however. For example, SO2 and ozone are both adsorbed on indoor surfaces; hence,
indoor levels of these pollutants are far lower than outdoor concentrations.
It is difficult to evaluate goal C from the materials provided. Clearly there must be scientific
confidence in the results of the screening study. I don't have confidence in these results at present; the
divergence of the results of the screening exercise and the monitoring program do not provide
confidence in the results. The lack of any clear explanation of how to interpret the results of the study
to the community is also a detriment. The study is seriously limited because it ignores many potentially
important sources; on the other hand, it employs a very conservative methodology that will
overestimate risks. Neither of these are clearly communicated in the report. I believe that this is
necessary to obtain the respect of the community.
L-65
-------
Specific Charges
1. See my comments above. I believe that the current inventory was neither sufficient nor appropriate.
The initial screen was reasonable for large sources. I'm not sure if it would have captured the
hypothetical dry cleaning establishment that I mentioned above. These smaller sources may be more
important because they are emitted at ground level.
The above could also apply to the secondary and final screens. There should have been greater
attempts to understand the discrepancy between the results of these screens and the monitoring data.
Appendix J provides an accurate description of the process.
The methodology as applied in Baltimore is of limited value; it ignores potentially very important
sources; it does not provide results which are consistent with monitoring data. It applies very
conservative methods to a few well-defined sources. See some of the specific comments below which
indicate areas where the methodology could be made less conservative and still be protective. Before
this methodology is applied elsewhere, it needs to be improved and shown to agree with the results of
monitoring data in Baltimore.
Specific Comments:
•Pg. 9, second bullet
Pg. 24, last paragraph
Pg. 25, first full paragraph
Pg. 29, table
Pg.40
"The actual risk ..." The use of the term "actual" is imprecise
and nonstandard for risk calculations. The word "actual" is used
throughout the paragraph. It would be more correct to state that
"The site-specific potential risk based on field measurements of
concentrations ... could not be determined."
The definition of a Reference Dose should be expanded a bit to
make it clearer. The units of an RfD need a bit of explanation; it
refers to dose in mg of the substance of interest per kilogram of
subject's body weight per day.
The example given for cancer risk, 6 * 10"4, seems unusually
large when all of the results arrived at later are 2 orders of
magnitude or more lower. Suggest 6* 10'7 as a more relevant
example.
The table carries too many significant figures for a risk
assessment; last two columns should not display more than 1 or,
if it is important to distinguish between outcomes, 2 significant
figures.
The use of a 50% conservative multiplier for the EPA Region 3
risk-based concentrations (RBCs) seems unnecessary. The RBCs
are calculated from EPA RiDs and CSFs, which in themselves
have incorporated uncertainty factors of multiple values of 3 or
10. An additional conservatism in these screening levels appears
superfluous.
L-66
-------
Pg. 4 1 et seq. The speciation of Chromium into Cr"'" vs. Cr*vl is critical for the
inhalation risk assessments. Yet no explanation is offered for the
speciation used. In particular, the use of a 30% Cr'vl fraction for
the BG&E power plants is unexplained. If this is from direct
measurements by BG&E, it should be so noted. EPRI data
indicate that a more appropriate figure in general is about 15%;
EPA in its utility air toxics report to Congress used an 1 1% Cr"vl
fraction for coal-fired power plants. Additionally, the fraction of
Cr"vl seems high for other sources as well.
L-67
------- |