United States Region 2 EPA/902/R-93-001a
Environmental Protection 902 January 1993
Agency
oEPA Staten Island/New Jersey
Urban Air Toxics
Assessment Project
Report
Volume I
Summary
-------
ACKNOWLEDGEMENTS
This report is a collaborative effort of the staffs of the
Region II Office of the U.S. Environmental Protection Agency
(EPA), the New Jersey Department of Environmental Protection and
Energy, the New York State Department of Environmental
Conservation, the New York State Department of Health, the
University of Medicine and Dentistry of New Jersey and the
College of Staten Island. The project was undertaken at the
request of elected officials and other representatives of Staten
Island concerned that emissions from neighboring industrial
sources might be responsible for suspected excess cancer
incidences in the area.
Other EPA offices that provided assistance included the
Office of Air Quality Planning and Standards, which provided
contract support and advice; and particularly the Atmospheric
Research and Exposure Assessment Laboratory, which provided
contract support, quality assurance materials, and sampling and
analysis guidance, and participated in the quality assurance
testing that provided a common basis of comparison for the
volatile organic compound analyses. The Region II Office of
Policy and Management and its counterparts in the States of New
York and New Jersey processed the many grants and procurements,
and assisted in routing funding to the project where it was
needed.
The project was conceived and directed by Conrad Simon,
Director of the Air and Waste Management Division, who organized
and obtained the necessary federal funding.
Oversight of the overall project was provided by a
Management Steering Committee and oversight of specific
activities, by a Project Work Group. The members of these groups
are listed in Volume II of the report. The Project Coordinators
for EPA, Robert Kelly, Rudolph K. Kapichak, and Carol Bellizzi,
were responsible for the final preparation of this document and
for editing the materials provided by the project subcommittee
chairs. -William Baker facilitated the coordinators' work.
Drs. Edward Ferrand and, later, Dr. Theo. J. Kneip, working
under contract for EPA, wrote several sections, coordinated
others, and provided a technical review of the work.
The project was made possible by the strong commitment it
received from its inception by Christopher Daggett as Regional
Administrator (RA) for EPA Region II, and by the continuing
support it received from William Muszynski as Acting RA and as
Deputy RA, and from Constantine Sidamon-Eristoff, the current RA.
The project has received considerable support from the other
-------
project organizations via the Management Steering Committee,
whose members are listed in Volume II.
-------
PREFACE - DESCRIPTION OP THE STATEN ISLAND/NEW JERSEY URBAN AIR
TOXICS ASSESSMENT PROJECT REPORT
This report describes a project undertaken by the States of
New York and New Jersey and the United States Environmental
Protection Agency with the assistance of the College of Staten
Island, the University of Medicine and Dentistry of New Jersey
and, as a contractor, the New Jersey Institute of Technology.
Volume I contains the historical basis for the project and a
summary of Volumes II, III, IV, and V of the project report.
Volume II of the report lists the objectives necessary for
achieving the overall purpose of the project, the organizational
structure of the project, and the tasks and responsibilities
assigned to the participants.
Volume III of the report presents the results and discussion
of each portion of the project for ambient air. It includes
monitoring data, the emission inventory, the results of the
source identification analyses, and comparisons of the monitoring
results with the results of other studies. Volume III is divided
into Part A for volatile organic compounds, and Part B for
metals, benzo[a]pyrene (BaP), and formaldehyde. Part B includes
the quality assurance (QA) reports for the metals, BaP, and
formaldehyde.
Volume IV presents the results and discussion for the indoor
air study performed in this project. It contains the QA reports
for the indoor air study, and a paper on the method for sampling
formaldehyde.
Volume V presents the results of the detailed statistical
analysis of the VOCs data, and the exposure and health risk
analyses for the project.
Volume VI, in two parts, consists of information on air
quality in the project area prior to the SI/NJ UATAP; quality
assurance (QA) reports that supplement the QA information in
Volume III, Parts A and B; the detailed workplans and QA plans of
each of the technical subcommittees; the QA reports prepared by
the organizations that analyzed the VOC samples; descriptions of
the sampling sites; assessment of the meteorological sites; and a
paper on emissions inventory development for publicly-owned
treatment works.
The AIRS database is the resource for recovery of the daily
data for the project. The quarterly summary reports from the
sampling organizations are available on a computer diskette from
the National Technical Information Service.
111
-------
STATEN ISLAND/MEW JERSEY
URBAN AIR TOXICS ASSESSMENT PROJECT
VOLUME I. SUMMARY EPA/902/R-93-00la
TABLE OF CONTENTS
1. EXECUTIVE SUMMARY 1
2. INTRODUCTION 5
2.1 ORGANIZATION OF THE STATEN ISLAND/NEW JERSEY
URBAN AIR TOXICS ASSESSMENT PROJECT REPORT .... 5
2.2 HISTORICAL BACKGROUND 6
2.3 THE STATEN ISLAND/NEW JERSEY URBAN AIR TOXICS
ASSESSMENT PROJECT, THE STATEN ISLAND CITIZEN'S
ODOR NETWORK, AND THE ROLE OF THE AGENCY FOR TOXIC
SUBSTANCES AND DISEASE REGISTRY 10
3. SUMMARY DESCRIPTION OF THE STATEN ISLAND/NEW JERSEY
URBAN AIR TOXICS ASSESSMENT PROJECT 13
3.1 PROJECT OBJECTIVES 13
3.2 SUBCOMMITTEE CONTRIBUTIONS 14
3.2.1 Quality Assurance Subcommittee 14
3.2.2 Ambient Monitoring Subcommittee 14
3.2.3 Data Management Subcommittee 15
3.2.4 Emission Inventory Subcommittee 15
3.2.5 Modeling and Source Identification
Subcommittee 15
3.2.6 Indoor Air Subcommittee 15
3.2.7 Exposure and Health Risk Assessment
Subcommittee 16
4. CONCLUSIONS 17
4.1 DATA SETS 17
4.2 DATA ANALYSES 19
4.3 PATTERNS AND CORRELATIONS 20
4.4 POTENTIAL SOURCES 21
4.5 INDOOR AIR 22
4.6 EXPOSURE AND HEALTH RISK ESTIMATES 23
5. ADDRESSING THE PROJECT OBJECTIVES 24
5.1 OBJECTIVE 1 24
5.2 OBJECTIVE 2 26
5.3 OBJECTIVE 3 26
5.4 OBJECTIVE 4 27
5.5 OBJECTIVE 5 27
5.6 OBJECTIVE 6 28
5.7 OBJECTIVE 7 28
5.8 OBJECTIVE 8 28
IV
-------
5.9 OBJECTIVE 9 29
6. REFERENCES 31
MAP 1-1 32
TABLE 1-1 33
-------
1. EXECUTIVE SUMMARY
The Staten Island/New Jersey Urban Air Toxics Assessment
Project (SI/NJ UATAP) was a program of ambient air monitoring and
meteorological data collection conducted from October 1987
through September 1989, and indoor air sampling conducted from
July 1990 to March 1991. An emission inventory was developed in
support of risk assessment and source identification for the
study area, which consisted of Staten Island and nearby New
Jersey, directly across the Arthur Kill from Staten Island. (See
Map 1-1.)
The project was a cooperative undertaking by the U.S.
Environmental Protection Agency Region II, the States of New York
and New Jersey, the College of Staten Island, and the University
of Medicine and Dentistry of New Jersey.
Quantitation of 40 pollutants—22 volatile organic compounds
(VOCs), 16 metals, benzo[a]pyrene (BaP), and formaldehyde—was
pursued using sorbent samplers for VOCs at 13 sites, hi-vol
samplers for particulates at 5 sites, and aldehyde-specific
samplers for formaldehyde at 5 sites.
At the outset of the project in 1986, it was known that the
sampling and analytical procedures available for determination of
VOCs in ambient air were complex, difficult to perform,
essentially research techniques rather than standardized
monitoring methods.
The project was highly successful in collecting air quality
data over a period of 24 months. These data have been used to
characterize the distribution of air toxics spatially and
temporally over the area and to perform risk assessments for the
ambient air pathway.
Most of the annual averages for individual VOCs at the study
monitoring sites fall within a range of a factor of two. Some of
the intersite differences are almost ten-fold, but such large
differences occur for only a few compounds at a few sites. The
concentrations of the VOCs measured at the sites in the SI/NJ
UATAP were quite uniform. No single monitoring site was
consistently associated with the highest concentrations.
For the year October 1988 through September 1989,
tetrachloroethylene and toluene were the compounds consistently
found at the highest concentrations for the chlorinated and
aromatic groups respectively. The concentrations of the aromatic
compounds toluene, benzene, and the xylenes were higher in the
period from January to March, and lower in the period from April
-------
to June. The chlorinated compounds did not exhibit readily
apparent seasonal trends.
The annual average concentrations for the SI/NJ UATAP sites
were in the same range as those for other urban areas nationwide,
a conclusion based on comparison to the results of the EPA Urban
Air Toxics Monitoring Program (U.S. EPA, 1989; U.S. EPA, 1990).
The SI/NJ UATAP inventory showed for the VOCs monitored in
this project that the emission rate for toluene (primarily from
point sources) was the highest. Among the pollutants with high
cancer unit risk factors, emission rates were highest for benzene
(from mobile sources) and dichlcromethane (methylene chloride)
(from point sources).
Although several sites showed significant variations in
concentrations with wind direction, there were few cases in which
individual point sources could be associated with air quality at
the site. In several cases, it was clear that total loading of
upwind point sources had a strong impact on ambient air quality.
The most important finding, however, was that localized sources,
both mobile and area sources, had the greatest impact on air
quality monitored at a site.
The relatively high annual average concentration for
tetrachloroethene (tetrachloroethylene) found at one site appears
to be attributable to releases from two nearby dry cleaners.
Relatively high concentrations of tetrachloroethylene at the
Staten Island Mall site (also called the Pump Station) may be
attributable to the pumping station of a publicly-owned treatment
works (POTW). Mobile sources (autos and trucks), refineries,
and, to some extent, gasoline stations were found to be the major
contributors to the highest concentrations of benzene at the
project monitors. The case for toluene was similar, but with
some input from other industrial sources and from POTWs. POTWs,
industrial sources, and area sources (dry cleaners) were the
primary sources of the highest concentrations of chlorinated
hydrocarbons at the project monitors.
The annual average concentrations for the metals, BaP, and
formaldehyde in the SI/NJ UATAP were generally in the same
concentration ranges as those for a number of sites in the EPA
UATMP program during the same time period. In some cases for
which the SI/NJ UATAP concentrations appear to be high, as for
cadmium, vanadium (New York sites only, no valid data for New
Jersey sites), and nickel, there is uncertainty regarding
accuracy of the reported SI/NJ UATAP results. Chromium
concentrations were generally higher at the New Jersey sites than
at most of the UATMP sites; no valid data were available for the
New York sites.
-------
A limited study of indoor air showed that concentrations of
13 VOCs in several homes in the study area were generally similar
to concentrations found in several other data bases for indoor
air. 1,1,1-trichloroethane was frequently detected in NJ homes
only. Tetrachloromethane was never detected indoors. Toluene,
benzene, m- and p_-xylenes, o-xylene, ethylbenzene, hexane,
trichloromethane, and tetrachloroethylene were usually or always
found at higher concentrations indoors than outdoors. The
highest concentrations of benzene and toluene in ambient air were
associated with mobile sources and petroleum refineries; yet
indoor concentrations of these chemicals were higher than outdoor
concentrations.
Quantitative estimates of increased lifetime cancer risks
for individual pollutants were in the range of 0.4 to 80 per
million. The Hazard Quotients, which are a measure of the
likelihood of adverse noncancer health effects, were below one
for all pollutants except benzene, chromium, and nickel. The
estimated risks for chromium and nickel are believed to be
conservative, i.e., err in the direction of overestimating risk;
since the chemical species of chromium and nickel in the ambient
air samples were not determined, uncertainty remains regarding
how conservative the estimates are.
The additive risk assessment for noncancer toxicity by
target organ, and for cancer for all pollutants combined assumed
continuous lifetime exposure to the median annual average ambient
air concentrations of nine VOCs, nine metals, BaP and
formaldehyde for the year October 1988 through September 1989.
It yielded a maximum Hazard Index (the sum of the Hazard
Quotients for a given target organ) of 2 (blood formation effects
and respiratory tract irritation) for noncancer toxicity. The
cumulative cancer risk estimate was 96 or 123 per million,
depending on the assumptions about ambient air concentrations of
chromium VI used in the estimates.
The estimated cancer and noncancer toxicity risks associated
with benzene were consistently higher than those estimated for
the other pollutants addressed in the risk assessments. The next
highest estimated risks for ambient air exposure were associated
with nickel, chromium, arsenic, and tetrachloromethane.
Statistically significant site-to-site differences were
found in mean ambient air concentrations for several VOCs. The
risk estimates were not sensitive to the differences, however.
The results of this project are being used by the U.S. EPA
towards fulfilling the mandates of the Urban Area Source Program,
S112(k) of Title III of the Clean Air Act Amendments of 1990: to
list not less than 30 hazardous air pollutants (HAPS)—pollutants
that are or will be listed pursuant to SH2(b) of Title III—
presenting the greatest threat to public health in the largest
-------
number of urban areas; to identify and regulate subject to
standards pursuant to §112(d) of Title III the area source
categories accounting for 90% or more of the aggregate emissions
of each of the 30 identified HAPS; and to take specific action to
reduce the risks posed by the identified HAPS, including
achieving a reduction of not less than 75% in the incidence of
cancer attributable to HAPs emitted by stationary sources.
Because the project did not show a dominant role for any
specific major point source in creating an air quality problem in
the study area, no basis was found for abatement actions directed
at any specific major source.
-------
2. INTRODUCTION
2.1 ORGANIZATION OF THE STATEN ISLAND/NEW JERSEY URBAN AIR
TOXICS ASSESSMENT PROJECT REPORT
The report for the SI/NJ UATAP has been organized into six
volumes. Volume I, this volume, provides a summary of the
overall program and a description of the contents of the
remaining volumes.
Volume II provides the organization and functioning of the
project, brief descriptions of the methods of sampling and
analysis, and the formats of the data reports. The
organizational structure consisted of a Management/Steering
Committee, a Working Group, Technical Subcommittees and an
Advisory Group. The detailed workplans of the technical
subcommittees are provided in the appendices in Volume VI.
Volume III, divided into Part A for the volatile organic
compounds (VOCs) and Part B for the particulates and
formaldehyde, reports the results obtained from the analysis of
two years of ambient air samples. The data are presented in
tabular and graphical forms. The findings are discussed in terms
of their significance with regard to the objectives of the
program.
The results of the eight-month indoor air study, initiated
near the end of the two-year ambient air sampling program, are
presented in Volume IV.
The ambient air VOC concentrations were analyzed for
statistical significance of apparent intersite differences. A
health risk assessment was prepared using the results of the
ambient air monitoring and indoor air monitoring, and statistical
analysis inputs. The results are presented in Volume V.
Volume VI is a compilation of the detailed workplans and
Quality Assurance (QA) plans of the subcommittees, the QA reports
of the sampling and analytical organizations and the QA
Subcommittee, descriptions of the sampling sites, and a reference
paper on air emissions from publicly-owned treatment works
(POTWs). While this material is not required for an
understanding of the data analyses and interpretations, it
provides the basis for a more thorough understanding of the
project.
-------
2.2 HISTORICAL BACKGROUND
The SI/NJ UATAP is a study of the ambient levels of selected
volatile organic compounds and particulate matter species in the
county of Richmond (Staten Island), New York, and in neighboring
counties (Middlesex, Union, and Essex) of New Jersey to determine
the exposures (and associated risk) of residents of the area to a
variety of toxic air pollutants. (See Map 1-1.) The study was
undertaken in response to concerns of these residents that their
health may be at serious risk due to exposure to toxic air
pollutants emitted routinely by industrial sources in the area,
as well as by episodic releases often characterized by
disagreeable odors. Furthermore, a number of studies had
concluded that residents of Staten Island had experienced a
higher incidence of cancers than other communities of similar
socioeconomic status.1 Reflecting the concerns of their
constituents, elected officials and other representatives of
Staten Island asked state and federal officials to investigate
the causes of recurrent odor episodes, and to determine whether
or not emissions from neighboring industrial sources might be
responsible for suspected excess cancer incidences in the area.
Because of Staten Island's low population density relative
to other parts of New York City, it has generally experienced
lower concentrations of the criteria air pollutants than those
other areas of New York City. However, the Island is bordered on
the west by a complex of major industries including
pharmaceutical plants, oil refineries, and chemical storage
facilities. Other potential sources of toxic and/or odorous
organic compounds include sewage treatment plants and the 1400-
acre Fresh Kills Landfill, the world's largest landfill.
Therefore, many of the residents have developed a high level of
concern about the toxicity of the ambient air.
According to a 1985 series of articles in a local newspaper,
the Staten Island Advance,. Staten Island residents had been
concerned about pollution from New Jersey for over 100 years. An
1882 report of the New York State Board of Health stated, "Most
of the buildings on the North Shore of Staten Island are private
residences, occupied by families long residing on the Island, and
from the causes here named, and for the first time, their homes
have been made uncomfortable and in the case of many of their
inmates, unhealthy, from causes beyond their reach, but wholly
under the control of a neighboring state and people."
1 Section 2.3 of this volume describes ATSDR (Agency for Toxics
Substances and Disease Registry) reviews of three of these
studies, and the findings that the studies were flawed and not
supportive of the asserted association between cancer
incidence and air pollution.
-------
In 1928, J. Meyers, writing in "The New York State Journal
of Medicine," said that in the period from 1911 to 1920, Staten
Island was ranked first in New York City in terms of cancer
deaths with a rate of 92.5 in 100,000 people. The identified
cause was that "much of [Staten Island's] northern shore had
suffered for many years from smoke, fumes and vapors from the
great oil refineries, and chemical, metal and other works
situated on Constable Hook, Bayonne and adjacent territory."
In 1967, a study published in "Archives of Environment
and Health" concluded that respiratory cancer rates for Staten
Islanders exposed to the highest amount of air pollution from New
Jersey were higher than for Islanders in low pollution areas.
In 1983, the New York City Department of Health (NYCDOH)
found in a study on the trends in New York City's respiratory
cancer deaths that Staten Island's death rate was the highest of
the boroughs of the City between 1960 and 1980. The rate had
risen from 27 out of every 100,000 people to 42.3 out of every
100,000 people, an increase of more than 57 percent in the 20-
year period compared to a citywide death rate increase of 35
percent.
In May 1985, Staten Island Congressman Guy Molinari called a
special town meeting featuring a panel of scientists including a
toxicologist from the University of Medicine and Dentistry of New
Jersey, a pulmonary specialist from the VA Medical Center in
Brooklyn, an industrial hygienist from Mt. Sinai Hospital, a
pulmonary specialist (physician) in private practice on Staten
Island, and a health effects researcher from the College of
Staten Island. The meeting was convened because many people in
the community had expressed increasing concern that toxic
contaminants in the Staten Island air were the cause of unusually
high respiratory cancer rates on Staten Island. The panelists
shared a common position that the residents were at considerable
risk due to emissions from the petrochemical complex in the
nearby New Jersey area. One of the panelists also asserted that
children raised on Staten Island and in New York City's other
boroughs exhibit disproportionately high incidences of pediatric
asthma.
In the same year (1985), a study entitled "111 Winds,"
conducted by the staff of Congressman Molinari, used published
census data, cancer statistics, and prevailing wind pattern data
to demonstrate that in the United States, counties such as Staten
Island, located downwind from petrochemical plants, have a higher
incidence of respiratory cancer than those upwind.
In an effort to be responsive to these concerns, federal,
state, and local officials met from time to time during the early
1980's to determine what appropriate actions they might take to
-------
address the concerns that had been expressed about the frequent
odor episodes, as well as the unscheduled or accidental release
of chemicals by industries bordering the Arthur Kill (the river
separating Staten Island from New Jersey). When, in a period of
months from October 1984 to January 1985, fifteen major chemical
release incidents occurred, officials again called for special
federal investigations. Among the officials who pressed for
these investigations, were the Borough President of Staten Island
and the Congressional representative for the area. Both made
personal appeals to the EPA Administrator to undertake the
necessary studies. These meetings and consultations led to the
undertaking of a number of specific initiatives.
In Karch of 19B4, the EPA Region II Administrator sent the
EPA National Emergency Response Team (ERT) based in Edison, New
Jersey, into the field in an attempt to document the presence of
toxic substances in the ambient atmosphere on Staten Island and
neighboring New Jersey. The ERT performed a one-week
investigation of ambient air concentrations in areas using state-
of-the-art measurement techniques. It identified the presence of
about 30 toxic chemicals near many of the sources suspected of
causing odor and toxics problems. However, it could not quantify
the contaminant concentrations nor conclusively link the
identified chemicals with emissions or odors from any specific.
source. Gusty wind conditions prevailed during the monitoring
period and no serious odor events occurred during the time.
In a related investigation later that year (September 1984),
the EPA's National Enforcement Investigations Center (NEIC)
agents visited locations identified as possible sources of odors.
They identified liquid effluent from a sewage treatment plant as
the possible source of the so-described cat-urine odor that had
often been the basis for coraplaints. From this effluent, it was
possible to trace the origin of the offending substance to a
nearby pharmaceutical plant. The discharge of the offending
liquid into the sewage system was discontinued as a result of
this investigation.
In the same year (1984), the EPA released a report referred
to as the six-Month study, which documented that significant
amounts of toxic substances existed in the air over large,
densely-populated urban areas. This provided further incentive
for the EPA Regional Office to design and conduct an ambient air
quality monitoring and assessment project. Lacking sufficient
funding and resources to conduct an independent study, the
Regional Office set out to undertake a cooperative effort with.
other units of government, with industry, and with environmental
and academic institutions.
Through its inquiries, the Regional Office discovered
that a number of agencies and organizations had themselves
independently planned to undertake some form of ambient air
-------
monitoring activity in the Staten Island/Northern New Jersey
Area.
0 The New York State Department of Environmental
Conservation (NYSDEC) had decided to undertake a $5
million statewide enhancement of its ongoing sampling
program with approximately $1 million of the total to be
used for airborne toxics throughout the state. NYSDEC
planned to set up monitoring sites for air toxics at four
locations in Staten Island.
0 The New Jersey Department of Environmental Protection
(NJDEP, now NJDEPE) was about to undertake an ambient air
monitoring program for a variety of volatile organic
compounds at two sites in northern New Jersey with technical
support from the New Jersey Institute of Technology (NJIT).
0 The College of Staten Island (CSI) was about to undertake an
Island-wide ambient air monitoring program for volatile
organic compounds using funds provided by the Governor of
New York State. CSI also planned to undertake a health
effects study of the area.
0 The New York City Department of Environmental Protection
(NYCDEP) was planning to conduct ambient air monitoring
activities in the Staten Island area, but had not yet
formulated specific plans.
0 The Interstate Sanitation Commission (ISC) which had over
the years received and responded to citizens' complaints
concerning interstate odors and pollution transport, was
interested in participating in the study.
0 The Arthur Kill Industrial Business Association (AKIBA), a
consortium of businesses, expressed an interest in joining a
cooperative effort provided that a major emphasis was placed
on odor tracking.
At the request of the EPA Region II Administrator
Christopher Daggett, representatives of these organizations and
agencies met on several occasions in 1984 to determine what kind
of cooperative project could be put together using the pooled
resources of these organizations. It was agreed that an AKIBA-
-------
developed odor tracking project2 should proceed on its own with
whatever assistance might be provided by the ISC because of the
common interest of both organizations in odor problems and
because of their experiences in addressing both odor episodes and
episode response. It was decided, as well, that those agencies
in a position to contribute resources and expertise for
performing air quality sampling and analysis using advanced
techniques would join together under the leadership of the EPA
Region II Office to develop an ambient air monitoring project.
The group decided to invite the University of Medicine and
Dentistry of New Jersey (UMDNJ) and the New York State Department
of Health (NYSDOH) to participate in order to provide needed
expertise in risk assessment. The NYCDEP and the ISC failed to
obtain the necessary resources to join in the project.
2.3 THE STATEN ISLAND/NEW JERSEY URBAN AIR TOXICS ASSESSMENT
PROJECT, THE STATEN ISLAND CITIZEN'S ODOR NETWORK, AND
THE ROLE OF ATSDR
During 1985 and 1986, the effort that became known as the
Staten Island/New Jersey Urban Air Toxics Assessment Project -
(SI/NJ UATAP) organized a series of committees and began to
develop plans for conducting monitoring, collecting other
information, and interpreting the results. In October 1986, the
project's Steering Committee formulated the objectives for the
project, listing nine specific objectives. Ambient air
monitoring would be the most expensive and extensive project
activity, and would be used to address most of the objectives.
In addition, the Committee agreed to pursue indoor monitoring,
emission inventories, various levels of data assessment and
The AKIBA study was conducted by The Research Corporation of
New England (TRC) using meteorological data and odor reports
to develop a methodology for tracking the sources of odors
during odor incidents. Its final report (released in 1989)
concluded that municipal facilities (sewage treatment plants
and landfills) were the most frequent and the most intense
sources of odors in the Arthur Kill region, often responsible
for adverse impacts on nearby communities. The report
specifically pointed to the Linden-Roselle sewage treatment
plant at Tremley Point and the Fresh Kills Landfill as making
major contributions to the regions's odor problems. As a
follow-up to the study, AKIBA installed an odor hotline for
use during odor episodes to alert industries to check their
facilities for malfunctions and potential unauthorized
releases.
10
-------
interpretation, and exposure and health risk assessment, along
with quality assurance and data handling for all of the other
phases of the project.
In 1987, the ambient air monitoring phase of the project was
initiated. Monitoring activities and other field work continued
until 1989. The ambient monitoring phase of the project included
15 sites (see Map 1-1) at which the following parameters were
measured:
volatile organic compounds at 13 sites,
metals at 5 sites,
formaldehyde at 5 sites, and
meteorological data at 4 sites.
The indoor air monitoring phase of the project began in July
1990 and concluded in March 1991, encompassing four indoor sites
and two associated outdoor sites. The emission inventory portion
of the project spanned the period from October 1987 to December
1991. Once the monitoring and inventory data began to appear,
the data handling, data interpretation, and exposure and health
risk assessment phases were initiated.
In 1990, EPA also undertook an ancillary study, called the
Staten Island Citizen's Odor Network, to further address the
concerns of the Staten Islanders about air quality during odor
events. EPA supplied canister devices similar to those utilized
in the ambient monitoring phase of the project to six Staten
Island homeowners and asked them to activate the devices when
they detected odor^s of concern. There were few occasions in
which odor episodes triggered the use of these samplers. On no
occasions were unusually high concentrations of air toxics found
to correlate with odor episodes.
In an effort to address the health effects issues, EPA asked
the Agency for Toxic Substances and Disease Registry (ATSDR) to
review the 1979 NYCDOH and the 1984 CSI cancer incidence studies,
and the 1985 111 Winds study developed by the staff of
Congressman Molinari. ATSDR determined that all three studies
were flawed in design, in the handling of statistical
information, and in the conclusions reached.3 Based on this, the
conclusions that there were links between reported cancer
incidence and air pollution could not be supported. CSI
expressed an interest in conducting further health-based studies.
It is important to realize that the incidence of cancer and
other diseases in a population is determined by the combined
effects of many genetic, socioeconomic and environmental factors
in addition to air contaminant exposure. Studies of other
ATSDR, 1988a; ATSDR, 1988b;, ATSDR 1988C.
11
-------
communities with a concentration of petroleum refineries and
other industries, and cancer rates higher than in other nearby
communities, have failed to show statistically significant
correlations of cancer incidence and air contaminant levels. In
a study in Contra Costa County, California,4 for example, the
only variable identified as a significant factor in lung cancer
was smoking. Thus, the assignment of possible sources of cancer
may be difficult without a carefully planned epidemiological
study.
In studies of air pollution directed toward an understanding
of population exposures, risks, and health effects, it is
important to recognize (a) limitations in approaches to and,
hence, in results associated with estimation of lifetime
exposure; (b) the relative contributions of indoor air and
ambient air to total inhalation exposure; and (c) the relative
contribution of inhalation exposure to total exposure via all
routes. Studies (Wallace et al., 1987) have shown that personal
exposures to VOCs—that is, integrated, measured concentrations
for 24-hour personal air samples—are usually more closely
related to indoor air concentrations than to ambient air
concentrations. Nevertheless, the sources of concern in this
study were industrial and non-point sources whose impact would be
assessed by evaluating ambient air quality.
Personal communication of J. Wesolowski of the California State
Department of Health to T.J. Kneip in 1991 concerning the
results of an unpublished report on a study in Contra Costa
County.
12
-------
3. SUMMARY DESCRIPTION OF THE STATEN ISLAND/NEW JERSEY URBAN AIR
TOXICS ASSESSMENT PROJECT
The project was organized in a committee/subcommittee style
along lines suggested by the U.S. EPA Region II Air and Waste
Management Division. A Management/Steering Committee was
established to determine the objectives of the program and to
provide ongoing guidance on the operation as it evolved.
Subcommittees were set up and a Project Work Group organized.
Each subcommittee established a workplan and QA plan; the
organizations responsible for the actual work were designated by
agreement (such as a commitment by a state laboratory) or by
contract to the EPA or a state agency. The Working Group was
responsible for the technical details of the work in conformance
to the objectives and the workplans of the subcommittees. The
membership of these groups and many other details of the
organization of the program are reported in Volume II.
3.1 PROJECT OBJECTIVES
The objectives established for the program were as follows:
1. Characterize air quality for selected volatile organic
compound (VOCs) for the purpose of doing an exposure
assessment for various population, commercial and
industrial interfaces.
2. Characterize air quality for the parameters identified by
EPA as high-risk urban toxics for the purpose of using
exposure assessment for comparison with other studies.
3. Characterize indoor air quality for selected VOCs for the
purpose of doing exposure assessment for various types of
commercial facilities and residences.
4. Evaluate indoor/outdoor concentration relationships for
selected VOCs.
5. Perform emission source inventory (including point, area and
mobile sources), so as to formulate hypotheses linking major
contaminants to potential sources.
6. Obtain air quality data for the purpose of identifying
potential sources using meteorological modeling.
7. Evaluate indoor air quality data to identify possible
sources.
13
-------
8. Evaluate episodic odor occurrences and relate such episodes
to air quality data.
9. Evaluate some general abatement strategies.
3.2 SUBCOMMITTEE CONTRIBUTIONS
3.2.1 Quality Assurance Subcommittee
This subcommittee provided guidance on the development of
the QA plans of the other subcommittees and reviewed the overall
sampling plans including site selections. It reviewed the QA
operations of the sampling and analytical organizations, examined
quarterly data reports and QA reports, field-audited the
operations, and examined the final data sets submitted in order
to establish the validity of the data sets reported in these
volumes and used to meet the objectives of the program. The
subcommittee implemented a quality assurance program
establishing a basis for interorganization comparisons, so that
any differences in reported concentrations of a compound at
different sites could be assessed for statistical significance.
The subcommittee prepared periodic memoranda pointing out
problems as they arose and proposed corrective actions. It
certified the satisfactory performance of the QA plans and
formally accepted the data sets that met the QA objectives.
The QA was not so extensive for the particulates as for the
VOCs due to the perspective that, unlike VOCs sampling, sampling
and analytical methods for particulates were well-established and
capable of delivering results of acceptable quality.
3.2.2 Ambient Monitoring Subcommittee
This subcommittee was organized to define the sampling and
analysis strategies for the substances that were chosen for
monitoring. In conjunction with the Working Group and the
Management/Steering Committee, the subcommittee developed the
list of substances likely to fit the project objectives. The
final list of compounds was defined by estimating the likelihood
that each substance would be present at concentrations measurable
with the sampling and analytical methods available at the time.
The overall process of selecting the compounds was designed to
assure the maximum likelihood of obtaining data with good
accuracy and precision. Sampling was conducted with the goal of
quantitating 40 chemicals in ambient air. VOCs were sampled
14
-------
using sorbents; particulates, using hi-vol samplers; and
formaldehyde, using aldehyde-specific samplers. The subcommittee
also took primary responsibility for selection of the sampling
sites. Subsequent to collection of the concentrations data, this
subcommittee prepared the initial data analysis for the report.
3.2.3 Data Management Subcommittee
This subcommittee was initially established to develop a
uniform format for data submissions, to collect the data from the
analytical laboratories, to circulate the assembled results on a
periodic basis for use in the project, and to establish the
project data base. During these operations, it became clear that
this subcommittee needed to review the submitted results for
discrepancies and other problems prior to circulation of the
assembled results. The subcommittee compiled data tables for
calculation and presentation of quarterly and annual averages,
prepared summary graphs of selected data, and provided a basis
for the initial site-to-site and interlaboratory comparisons.
3.2.4 Emission Inventory Subcommittee
This subcommittee was established to develop a description
of the major point, area, and mobile sources near each sampling
site. The vast amount of information developed was a product of
new approaches to the integration of a number of data resources,
as well as field work at each site.
3.2.5 Modeling and Source Identification Subcommittee
This subcommittee took on the tasks of producing pollution
roses for the many sites and pollutants, and back- trajectories
for selected pollutants on particular dates at several sites.
Identification of the sources of distinctly high concentrations
of VOCs at the monitoring sites was the goal of this work.
Results of the data management and emissions inventory efforts
served as inputs for the source identification.
3.2.6 Indoor Air Subcommittee
This subcommittee undertook the design and execution of a
limited indoor air sampling and analysis program focusing on 12
VOCs. Canister samplers were used for VQCs, and an aldehyde-
specific samplers was used, albeit unsuccessfully, for
15
-------
formaldehyde. The eight-month period of the study, July 1990 to
March 1991, followed the ambient air program. Ambient air
samples from sites near the houses selected for the indoor air
sampling were taken simultaneously with the indoor air sampling.
3.2.7 Exposure and Health Risk Assessment Subcommittee
This subcommittee was established to select the substances
that represented potential health threats at the concentrations
measured, and produce estimates of population exposures and
health risks. The inhalation exposure and associated health risk
assessments were generated using ambient air and indoor air
average concentration data, toxicological information, and the
results of the statistical analyses of the VOCs data.
16
-------
4. CONCLUSIONS
4.1 DATA SETS
At the outset of the project in 1986, it was known that the
sampling and analytical procedures available for determination of
VOCs in ambient air were complex, difficult to perform, and
essentially research techniques rather than standardized
monitoring methods. It was anticipated that the precision of an
individual measurement might be no better than a factor of two,
and that site-to-site differences might be difficult to
demonstrate.
The VOC data were evaluated using the considerable quantity
of intralaboratory and interlaboratory comparison data generated
for the study. Almost all of the VOCs data met the stringent
conditions stated in the quality assurance plans at the outset of
the project. These conditions were embodied in a two-level
approach. The first level was the obligation of each laboratory
and included the establishment of good laboratory practices by
each organization, including the following:
proper calibration,
use of analytical standards,
sampling and analytical blanks,
duplicate analyses,
participation in inter-laboratory sampling and analytical
comparisons, and
data review.
The second level of QA, project-level, consisted of the
following:
verifying that the QA/QC procedures of each organization
were appropriate and were implemented;
requiring written QA plans from each organization, reviewing
the plans and performing QA audits to confirm performance
of'the work;
coordinating submission of Performance Evaluation samples
reviewing the results;
coordinating the performance of collocation sampling
experiments and reviewing the analytical results; and
reviewing periodically the monitoring data and quarterly QA
reports of each organization, and recommending and
requiring corrective action as appropriate.
The results of the QA analysis provided comparisons of
paired analyses for absorber and canister samples taken
17
-------
simultaneously at the same sites by each of the participating
organizations. The data for the canister samples were accepted
as reference values for each sampling period. The number of
sample pairs ranged from 4 to 130 depending on the organization
and compound compared. The data are provided in tables in
Section 2 of Volume III, Part A. A statistical evaluation of the
results for absorption tube and canister sample comparisons and
the results of the collocation sampling experiments was provided.
The accuracy of the results for the VOCs is represented by
the difference between the canister reference result and the
absorber analysis for each compound. For most of the compounds
with measurable concentrations, the concentrations reported for
the sorbents were within +30 to 40% of those reported for the
canister reference, with all cases falling within +63% of the
reference. There was no contamination problem, therefore the
results for the project provide satisfactory accuracy. The
precision for duplicate samples taken by all organizations is in
the range of 10 to 30%, which is excellent for concentrations of
the magnitudes measured in this program.
The single Tenax absorber was found to have excessive
breakthrough for dichloromethane; however collection efficiencies
for all other compounds analyzed were satisfactory . The
trisorbent tubes used by the New York State Department of
Environmental Conservation were found to have satisfactory
collection efficiencies for all compounds.
The graphs in Section 3 of Volume III, Part A, show quite
clearly that the sample-to-sample variations are actual changes
in the measured concentrations. Work with the meteorological
analyses demonstrated that patterns in the data in some cases are
related to wind direction and air parcel trajectories.
The data sets for the VOCs are clearly of very high quality.
Where problems were found during the QA reviews, data sets were
withdrawn or caveated as supported by less than completely
satisfactory precision or accuracy data. In the latter case
caution is recommended in the use of the data.
For the metals, data were omitted from the project data base
for numerous chemicals due to insufficient accuracy or
insufficient data to establish data quality. Other data were
caveated due to low recoveries or, in the case of formaldehyde,
an interference with the method. For still other chemicals,
standard reference materials were unavailable to gauge accuracy
of the reported results. While measurement of the VOCs were the
focus of the QA effort of this project, the particulates do
figure importantly in the risk assessment. In fact, two of the
clear risk-related recommendations from this project relate to
the metals data.
18
-------
The QA results for this project demonstrate the need for
attention to interlaboratory differences in data quality,
especially when comparing concentration data from different
studies, e.g., when analyzing data from a national data base.
4.2 DATA ANALYSES
Most of the annual averages for individual VOCs at the study
monitoring sites fall within a range of only a factor of two.
Some of the intersite differences are almost ten-fold, but such
large differences occur for only a few compounds at a few sites.
When compared to the Urban Air Toxics Monitoring Program results
for vocs (U.S. EPA, 1989; U.S. EPA, 1990) for many sites in the
United States, the annual averages for the SI/NJ UATAP sites are
generally in the same range as those for other urban areas
nationwide. The data for the SI/NJ UATAP sites fall in the
middle to low end of the ranges for annual average comparisons
with the exception of the tetrachloroethylene results for the
Dongan Hills and Pump Station sites. These averages are on the
high end of the range for the UATMP study.
The concentrations of the VOCs measured at the sites in the
SI/NJ program are quite uniform. No single monitoring site
consistently had the highest concentrations; however, some sites
consistently had higher annual average concentrations for a
number of compounds. For example, the annual average
concentrations for the Dongan Hills, Elizabeth, Port Richmond and
Eltingville sites were typically >1.4 ppb for benzene and >4 ppb
for toluene. The'Carteret and Bayley Seton sites exceeded these
levels in one of the two years. Both the Dongan Hills and the
Staten Island Mall sites had high average concentrations for some
compounds.
The Dongan Hills site had the highest annual average
concentration for tetrachloroethene (tetrachloroethylene). This
appears to be attributable to releases from two dry cleaners
identified in the microinventory within 100 meters of the
monitor, and from two more about 250 and 400 meters away. Dry
cleaners are known sources of this compound; other studies have
shown a relationship between distance from a dry cleaner and
exposure measurements (Upton et al., 1989). The 14 gas stations
within 1,000 meters of the monitor, the fire house at the
monitor, and the nearby streets were all possible sources of
aromatic hydrocarbons at the Dongan Hills monitor; however, the
statistical analysis did not find the apparently higher
concentrations of benzene and toluene at this site significantly
higher than at all other sites in the study.
19
-------
Annual average concentrations of tetrachloroethylene and
other chlorinated hydrocarbons were relatively high at the Staten
Island Mall site (also called the Pump Station), which is located
adjacent to a sewage pumping station. Publicly-owned treatment
works (POTWs) are known sources of VOCs, particularly
trichloroethylene; sewage collection systems are believed to be
large sources as well. The possibility that the pumping station
is a source of these compounds may be investigated should the
detailed exposure-health risk analysis indicate the need.
The data analyses demonstrate the influence of local sources
in determining concentrations that may be used subsequently as
typical and representative of an area. So when comparing
concentrations reported for different areas, examination of
siting is important. When attributing observed high
concentrations to sources, field work was important in this study
for supplementing available data bases. Area sources (e.g.,
small dry cleaners and gas stations) and non-industrial sources
(e.g., POTWs) were important in explaining concentrations
measured at the monitoring sites.
The annual average concentrations for toxic metals, BaP, and
formaldehyde in the SI/NJ UATAP were in the same concentration
ranges as those for a number of sites in the EPA UATMP program
during the same time period. Where the SI/NJ UATAP data appear
to be high, as is the case for cadmium, vanadium (New York sites
only, no valid data for New Jersey sites), and nickel, there is a
lack of certainty regarding accuracy of the reported SI/NJ UATAP
results. Chromium concentrations were generally higher at the
New Jersey sites than at most of the UATMP sites; no valid data
were available for the New York sites.
4.3 PATTERNS AND CORRELATIONS
For the year October 1988 through September 1989,
tetrachloroethylene and toluene were the compounds consistently
found at the highest concentrations for the chlorinated and
aromatic groups respectively.
The concentrations of the aromatic compounds toluene,
benzene, and the xylenes were higher in the period from January
to.March, and lower in the period from April to June. The
chlorinated compounds did not exhibit readily-apparent seasonal
trends.
Patterns appear to exist in the VOCs data for some of the
individual compounds in time series plots (graphs of
concentration versus sampling data for a given site and
pollutant), or in comparisons of seasonal and annual averages.
20
-------
Data for some of the monitoring sites indicate that some
concentrations are site-related, and in some cases may be related
to specific sources.
The aromatic compounds toluene, benzene, and the xylenes are
known to be emitted by petroleum refineries, automobiles, and gas
stations. Several of the chlorinated hydrocarbons are widely
used industrial solvents; tetrachloroethylene is the most widely-
used dry-cleaning solvent. Almost all of the compounds in these
groups were found at levels exceeding background concentrations
(i.e., concentrations at the background sites for the project),
but concentrations of several of the chlorinated compounds were
at or near background levels, indicating that local or regional
sources had a minimal impact on the concentrations in air. A
number of site-related concentration differences were further
evaluated for identification of possible source or spatial
relationships.
There are a number of interesting and potentially useful
temporal and spatial patterns in the particulates data, with some
substances showing differences between the sampling sites in the
two states, but not between sites within the states.
Since an ozone interference negatively biased the
formaldehyde results, little information is derived from the
apparent site-to-site differences for this compound.
4.4 POTENTIAL SOURCES
The emission microinventory data for the thousand-meter
circle around each monitoring site aided in the evaluation of the
potential site-related concentration differences as noted for the
high tetrachloroethylene concentrations at the Dongan Hills site.
The point source inventories provide compound-by-compound
estimates of annual emissions at specific source locations; they
supported efforts to determine sources of the VOCs.
The data analyses using pollution rose calculations and
back-trajectory modeling demonstrated the usefulness of these
approaches in achieving a full understanding of the complex data
sets. The data have been shown to have sufficient accuracy and
precision for the requirements of the computer programs used in
the pollution rose and back-trajectory analyses. The time
resolution of the data sets is also compatible with the
meteorological data used in these programs. These methods assist
in defining spatial and temporal variations and in obtaining
possible identities of sources through the back-trajectory method
of tracing air parcel histories.
21
-------
Mobile sources (autos and trucks) and refineries were found
to be the major contributors to the highest concentrations of
benzene and toluene at the project monitors. POTWs, industrial
sources, and area sources (dry cleaners) were the primary sources
of the highest concentrations of chlorinated hydrocarbons at the
project monitors.
The results of the statistical analyses provided input
regarding possible exposure and health risk differences due to
the spatial variations in the concentrations.
4.5 INDOOR AIR
The New York State Department of Health performed a limited
study of indoor air pollution as an extension of the SI/NJ UATAP.
The study was carried out in the year following completion of the
two-year ambient air sampling program. One ambient monitor was
operated together with the sampling carried out in two homes in
New York in the Travis neighborhood on Staten Island, and one
ambient monitor was operated with sampling in two homes in
Carteret, New Jersey. Samples were taken over a 24-hour interval
(two sequential 12-hr samples) every twelve days over the period
from July 10, 1990, to March 19, 1991. The spring period was not
sampled; however this is not a critical problem since both spring
and fall are transition periods between the summer and winter
extremes of indoor ventilation characteristics.
The results were found to be in generally good agreement
with the data from the ambient air study of the previous two
years, as well as with several other data bases for indoor air
concentrations.
The VOCs frequently detected (found in 75% or more of the
samples) in indoor air in the homes of both New Jersey and New
York include chloromethane, dichloromethane, hexane, benzene,
toluene, ethylbenzene, m- and p.-xylenes, and o-xylene. 1,1,1-
trichloroethane was frequently detected in NJ homes only.
Several other chlorinated hydrocarbons were less often detected
indoors, and tetrachloromethane was never detected indoors.
Toluene, benzene, m- and p.-xylenes, p_-xylene, ethylbenzene,
hexane, trichloromethane, and tetrachloroethylene were usually or
always found at higher concentrations indoors than outdoors. It
is noteworthy that source identification results associated the
highest concentrations of benzene and toluene in ambient air with
mobile sources and petroleum refineries; yet on the basis of
average concentrations, indoor concentrations were higher than
outdoor concentrations for these chemicals.
22
-------
4.6 EXPOSURE AND HEALTH RISK ESTIMATES
Quantitative estimates of increased lifetime cancer risks
for individual pollutants were in the range of 0.4 to 61 per
million for the Level 1 analysis (ambient air), and 1 to 80 per
million for the Level 2 analysis (ambient and indoor air, VOCs
only). The Hazard Quotients for non-carcinogenic effects were
below one for all pollutants except benzene, chromium, and
nickel. The estimated risks for chromium and nickel are believed
to be conservative, i.e., err in the direction of overestimating
risk, since the specific chemical species of chromium and nickel
in the ambient air samples were not measured.
The additive risk assessment for noncancer toxicity by
target organ, and for cancer for all pollutants combined, yielded
a maximum Hazard Index of 2 (hematopoietic effects and
respiratory tract irritation), and a cumulative cancer risk of 96
or 123 per million, depending on the reference concentrations and
chromium VI assumptions used in the estimates.
The estimated cancer and noncancer toxicity risks associated
with benzene were consistently higher than those estimated for
the other pollutants addressed in the Level 1 and Level 2
analyses. The next highest estimated risks in the Level 1
analysis were associated with nickel, chromium,
tetrachloromethane, and arsenic; while the next highest estimated
risks in the Level 2 analysis (VOCs only) were from
trichloromethane and tetrachloroethene.
Using the 1988 and 1989 Urban Air Toxics Monitoring Program
(UATMP) studies as the basis for comparison, the risk estimates
for tetrachloroethene, nickel, and chromium were higher for the
study area than for other urban areas nationwide. An assessment
of the significance of the magnitudes of the differences is not
offered; some of the differences might be attributable to
differences in sampling and chemical analysis for the studies.
23
-------
5. ADDRESSING THE PROJECT OBJECTIVES
This section presents the project results in terms of the
goals established for the project.
5.1 OBJECTIVE 1
Characterize air quality for selected volatile organic
compounds (VOCs) for the purpose of doing an exposure
assessment for various population, commercial/ and
industrial interfaces.
This project was successful in characterizing air quality
for selected VOCs at 13 sites across the project area over a two-
year period. Yearly average concentrations of the target
compounds at each monitoring site were produced by the various
monitoring organizations, and validated according to the quality
assurance (QA) program for the project.
In addition to developing the desired air quality data, the
project was able to investigate and demonstrate the relative
merits of several techniques for measuring airborne VOCs. At the
start of the project, the state of the art for measuring VOCs was
as follows:
*> sampling through sorbent tubes (particularly tubes filled
with Tenax) with analysis by gas chromatograph/mass
spectrometer (GC/MS) was the most accepted technique;
*• sampling into specially polished stainless steel canisters,
with analysis by GC/MS, was being promoted; and
*• sampling directly into field-based analytical instruments
(particularly gas chromatographs (GCs)) was in early
development.
Some members of the project had experience in using the
first two methods, and some were interested in developing
expertise in all three. Some EPA offices (Region II, the Office
of Air Quality Planning and Standards, and the Office of Research
and Development) were interested in encouraging this development
and in comparing the results of the various techniques in real-
world settings.
In order to further all of these goals and to fall within
the financial framework of the project, it was agreed that two-
tube distributed-volume sorbent sampling would be the primary
24
-------
monitoring technique, to be performed every six days at every
site. Canisters would be used as a QA frame of reference,
rotating from site to site, except that some sites would be
sampled regularly by both methods. The real-time GCs would be
fit in, as development warranted (although, in fact, this never
came to pass). Special QA procedures would then be implemented
to investigate method-to-method and organization-to-organization
comparisons.
The various collocations, performance evaluation samples,
and on-site audits indicated the following:
>• Intermethod comparisons. Sorbent tube and canister sampling
methods are able to produce comparable results in terms of
both detection limits and variability. A given project or
circumstance may be more amenable to one method or the
other. For example, a project that will require long sample
holding times would be more suited to canister sampling,
while a project that requires ultra-low detection limits may
benefit from sample-concentration capabilities of sorbent
tubes.
*> Interorganization comparisons. Differences between
monitoring organizations are more important than differences
between methods.
* Viability of a multi-organization, multi-method project.
Extensive project-level QA activities, on top of the normal
QA program within the various organizations, are able to
account for and minimize variability. In this case, data
quality was not compromised by the use of multiple methods,
and the inter-organization variability was certainly less
than it would have been if they had all used the same method
but had not participated in extensive project-level QA
activities.
> Usefulness of two-tube sorbent sampling. Two-tube sorbent
sampling can be very effective when the concentrations for
the target contaminants are only slightly above background
and do not vary significantly, and long-term averages are
desired. The more standard four-tube system may still be
needed when individual sample values are critical, or when
ambient concentrations may vary widely, as in the vicinity
of a major source.
25
-------
5.2 OBJECTIVE 2
Characterize air quality for the parameters identified by
EPA as high-risk urban toxics for the purpose of using
exposure assessment for comparison with other studies.
The project was highly successful in capturing air quality
data for selected parameters at 13 sites in the project area over
a period of 24 months. These data have been used to characterize
the distribution of air toxics spatially and temporally over the
area and to perform exposure assessments for the ambient air
pathway. For most of the 10 VOCs compared, the 1989 data (the
year of data most complete with respect to number of sites and
number of samples) from this project show a striking similarity
to the data obtained from the 14 sites in the 1989 Urban Toxics
Monitoring Program (UATMP). One SI/NJ UATAP site experienced
higher mean concentrations for tetrachloroethylene than any 1988
or 1989 UATMP site. Several UATMP sites (Pensacola, FL; Dallas,
TX; Wichita, KS) often experienced lower concentrations than the
SI/NJ UATAP sites. Several UATMP sites (Chicago, IL; Sauget, IL;
Miami, FL) often experienced higher concentrations of the VOCs
than the SI/NJ UATAP sites.
5.3 OBJECTIVE 3
Characterize indoor air quality for selected VOCs for the
purpose of doing exposure assessment for various types of
commercial facilities and residences.
Over an eight-month period in 1990, indoor sampling was
conducted at four residences using sampling and analysis
techniques consistent with those used in the ambient air sampling
phase of the project. No sampling was done in commercial
facilities. Concentrations of some of the 12 VOCs quantitated
were found to be more variable indoors than outdoors. The
results have been incorporated into the risk assessments.
26
-------
5.4 OBJECTIVE 4
Evaluate indoor/outdoor concentration relationships for
selected VOCs.
Studies found in the literature have shown -that for some
pollutants in some settings, indoor pollution levels are
influenced or dictated by ambient air levels outside the
buildings. These reports show that the relationships between
concentrations in indoor air and outdoor air vary from site to
site. At many sites the indoor levels of some toxic air
pollutants far exceed outdoor levels. Comparing the
concentrations found in the four selected residences with the
concurrent outdoor measurements made near the residences, it was
found that the indoor concentrations of the eight VOCs frequently
detected indoors usually or always exceeded their outdoor
concentrations. Indoor/outdoor concentration ratios were similar
to those found in TEAM (Total Exposure Assessment Methodology)
studies.
5.5 OBJECTIVE 5
Perform emission source inventory (including point, area,
and mobile sources), so as to formulate hypotheses linking
major contaminants to potential sources.
A primary purpose for this study was to determine whether or
not large industrial sources in New Jersey were the origin of
toxic air pollutants which might be adversely impacting the
health of the residents of Staten Island and neighboring areas of
New Jersey. Since the majority of pollutants selected for study
are commonly derived from both point, area and mobile sources, it
was important to document the location and distribution of
sources around monitoring sites very carefully in order to
identify source-receptor relationships.
An inventory was assembled for the various classes of
sources (point, area, and mobile). Several innovations were
necessary to derive emission estimates for some sources such as
publicly-owned sewage treatment works. A microinventory was
performed in the circle one kilometer in radius around each
sampling site to supplement source information in the existing
data bases.
For the VOCs monitored in this project, the inventory showed
that estimated emissions for toluene (primarily from point
sources) were the highest. Among the pollutants with high cancer
unit risk factors, benzene (from mobile sources) and
27
-------
dichloromethane (methylene chloride) (from point sources) were
the VOCs with the highest emissions.
5.6 OBJECTIVE 6
Obtain air quality data for the purpose of identifying
potential sources using meteorological modeling.
The air quality data for the target VOCs collected during
this project were suitable for use in meteorological modeling for
identifying potential sources.
Although several sites showed significant variations in
concentrations with wind direction, there were few cases in which
individual point sources could be associated with air quality at
the site. In several cases, it was clear that total loading of
upwind point sources had a strong impact on ambient air quality.
The most important finding, however, was that localized sources,
both mobile and area sources, had the greatest impact on air
quality monitored at a site.
5.7 OBJECTIVE 7
Evaluate indoor air quality data to identify possible
sources.
The variability in correlations between indoor and
corresponding outdoor concentrations from day to day and from
house to house often did not support indoor or outdoor sources as
the clear cause for the observed differences in concentrations.
There were some exceptions, however, where consistent indoor
sources, short-duration activities, or local outdoor sources were
implicated. The association of higher indoor concentrations with
specific indoor sources was hampered by the lack of a complete,
detailed inventory of potential sources within each residence—an
asset planned but not realized.
5.8 OBJECTIVE 8
Evaluate episodic odor occurrences and relate such episodes
to air quality data.
In an effort to determine whether or not concentrations of
targeted organic compounds increase during odor episodes, EPA
issued canisters to six residents of the area during the period
October 1989 through August 1990. Participants were asked to
collect ambient air samples over a 30-minute period during each
28
-------
odor episode. They were also asked to record such available
meteorological parameters as wind speed and direction, the time
the sample was collected, the general type of odor which was
present, the strength of the odor, and any activity in the
neighborhood which might be associated with the chemical
concentrations found in the sample. These activities could
include fuel oil deliveries, use of gasoline lawn mowers, and
vehicles idling for an extended period of time.
The Interstate Sanitation Commission's odor complaint log
for the test period showed that in a majority of the 24 instances
when samples were collected, odor complaints were received from
several citizens in the area. This independently confirms that
odors were present when the samplers were activated by the
participating resident. The greatest number of responses by any
one participant was five, the fewest was two. However, it was
not possible to link any odor episodes to any specific event.
The odors were most frequently described as smelling like garbage
or burning garbage and cement mix. This effort was referred to
as the Staten Island Citizen's Odor Network.
The samples were analyzed for 17 VOCs. The vast majority of
concentrations obtained from the odor episode samples were <3.0
ppbv. The maximum concentration during any one odor episode was
19.0 ppbv (toluene). Five chemicals—toluene, o- and m/p.-
xylenes, benzene, and ethylbenzene—were found in all samples.
Methylene chloride, tetrachloroethene, and 1,1,1-trichloroethane
were frequently found. Table 1-1 lists the maximum
concentrations for the 17 chemicals detected in the odor episode
samples. The mean concentrations of these chemicals during the
odor episodes were similar to those observed during the non-
episode periods of the SI/NJ UATAP project. However,
concentrations of chloroform and chloromethane were about 10
times higher in the odor episode samples than in the non-episode
period. Overall, those compounds with the highest measured
concentrations in the non-episode periods tended to have the
highest concentrations during the odor episode periods.
5.9 OBJECTIVE 9
Evaluate some general abatement strategies.
Because the project did not show a dominant role for any
specific major point source in creating an air quality problem in
the study area, no basis was found for abatement actions directed
at any specific major source. Further, much of the potential
problem in the study area appears to be related to area and
mobile sources.
29
-------
The importance of these sources of air toxics emissions in
creating ambient air toxic problems was recognized by Congress
when it passed the 1990 Amendments to the Clean Air Act, which
contain strategies for addressing these emissions. Specifically,
the Act's reformulated gasoline requirements in Title II require
a 15 percent reduction in toxics by the year 1995; this reduction
requirement increases to 20 to 25 percent in the year 2000.
Specific reductions in air toxics emissions from both point and
area sources are mandated by Title III of the Clean Air Act.
30
-------
6. REFERENCES
Agency for Toxic Substances and Disease Registry (1988a). Health
consultation, review of College of Staten Island thesis and
reports concerning Staten Island respiratory cancer mortality
rates: memorandum of March 24, 1988, from Medical Officer, Health
Assessment Coordination Activity, Office of Health Assessment, to
W. Nelson, Public Health Advisor, U.S. Environmental Protection
Agency Region II. Atlanta, GA: Department of Health and Human
Services, Public Health Service. (3 pp.)
Agency for Toxic Substances and Disease Registry (1988b). Health
consultation, review of College of Staten Island thesis and
reports concerning Staten Island respiratory cancer mortality
rates: memorandum of March 24, 1988, from Medical Officer, Health
Assessment Coordination Activity, Office of Health Assessment, to
W. Nelson, Public Health Advisor, U.S. Environmental Protection
Agency Region II. Atlanta, GA: Department of Health and Human
Services, Public Health Service. (8 pp.)
Agency for Toxic Substances and Disease Registry (1988c). Health
consultation, review of The 111 Winds; memorandum of July 14,
1988, from Medical Officer, Health Assessment Coordination
Activity, Office of Health Assessment, to W. Nelson, Public
Health Advisor, U,S. Environmental Protection Agency Region II.
Atlanta, GA: Department of Health and Human Services, Public
Health Service.
U. S. Environmental Protection Agency (1989). Nonmethane
Organic Compound Monitoring Program, Final Report, Volume II:
Urban Air Toxics Monitoring Program, April 1989. Office of Air
Quality Planning and Standards, Research Triangle Park, North
Carolina. (EPA-450/4-89-005).
U. S. Environmental Protection Agency (1989). Urban Air
Toxics Monitoring Program, Draft Final Report. Draft of May
1990, EPA No. 68D80014; received from Radian Corporation.
Upton, A,C.; Kneip, T.J.; Toniolo, P. (1989) Public health
aspects of toxic chemical disposal sites. Ann. Rev. Publ. Hlth.
10: 1-25.
Wallace, L.A. (1987) Project summary, the total exposure
assessment methodology (TEAM) study. Washington, DC:U.S.
Environmental Protection Agency, Office of Acid Deposition,
Environmental Monitoring and Quality Assurance; EPA report no.
EPA/600/S6-87/002.
31
-------
~
S3
MAP 1-1
SI/NJ UATAP
Monitoring Locations
SITES:
1. Westerleigh
2. Travis
3. Annadale
4. Great Kills
5. Port Richmond
6. Dongan Hills
7. Pumping Station
8. Clifton
9. Tottenville
A. Elizabeth
B. Carteret
C. Sevaren
D. Piscatavay
E. Highland Park
-------
Table 1-1: RESULTS OF THE STATEN ISLAND CITIZEN'S ODOR COMPLAINT NETWORK
Summary of analytical results for canister samples taken by
a group of residents during the period from October 1989
through August 1990
# of Samples in which
Concentrations
Compound compound was detected* MAX
BENZENE
CHLOROFORM
CHLOROMETHANE
1,3-BUTADIENE
P-DICHLOROBENZENE
ETHYLBENZENE
HEXANE
METHYLENE CHLORIDE
N-OCTANE
STYRENE
TETRACHLOROETHENE
TOLUENE
1,1,1-TRICHLOROETHANE
TRICHLOROETHENE
O-XYLENE
M/P-XYLENE
PROPYLENE
24
2
7
2
13
24
10
20
6
14
22
24
15
9
24
24
9
5.60
21.10
1.35
0.62
0.60
2.50
2.90
12.71
0.90
1.20
4.16
19.00
17.39
1.40
3.50
15.00
7.23
MIN
0.65
0.09
0.53
0.57
0.07
0.24
0.21
0.58
0.41
0.02
0.29
0.64
0.26
0.03
0.35
0.53
1.06
MEAN
2.2467
0.5950
0.9429
0.5950
0.2239
0.7796
1.4130
2.4090
0.6467
0.3250
0.7118
6.6092
2.2307
0.3511
1.1392
4.0563
2.8289
•24 valid samples
33
------- |