EPA 600/R-10/074 | June 2010 | www.epa.gov/ord
United States
Environmental Protection
Agency
A Research Study to Investigate
PCBs in School Buildings
Final Research Plan
-------
-------
EPA/600/R-10/074 June 2010 www.epa.gov/ord
A Research Study To Investigate PCBs in
School Buildings
Final Research Plan
June 16, 2010
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
-------
Disclaimer
The information in this document has been funded wholly by the United States Environmental Protection Agency
(EPA). It has been subjected to the Agency's peer and administrative review and has been approved for publication
as an EPA document. Mention of trade names or commercial products does not constitute endorsement or
recommendation by EPA for use.
-------
Table of Contents
List of Tables iv
1. Background 1
2. Research Objectives 4
3. Study Summary 5
3.1 Study Overview 5
3.2 Investigators 5
3.3 Study Limitations 5
4. School Recruitment and Engagement 6
4.1 Number and Types of Schools 6
4.2 Identifying and Recruiting Schools 6
4.3 Development of Individual School Sampling Plans 6
4.4 Information about the School 6
4.5 Time Needed at Each School 7
5. Sample Collection 8
5.1 Types of Samples 8
5.2 Sampling Locations 9
5.3 Numbers of Samples 10
5.4 Sample Collection Methods 10
5.5 Sample Transport, Storage, and Custody 13
6. Sample Analysis 14
6.1 Background Information 14
6.2 Analysis of Air Samples 14
6.3 Analysis of Wipe Samples 15
6.4 Analysis of Dust, Soil, Caulk, and Material Samples 15
6.5 Congener-Specific Analysis for a Subset of Samples 15
7. Quality Assurance and Quality Control 16
8. Data Analysis, Modeling, and Reporting 17
8.1 Data Receipt and Compilation 17
8.2 Descriptive Statistics 17
8.3 Analysis of Within- and Between-School Variability 17
8.4 Indoor Modeling: PCB Source Evaluations and Relationships with Environmental Levels 17
8.5 SHEDS Model Development and Application 18
8.6 Assessment of Potential Exposures and Exposure Routes 19
9. Reporting Results 20
9.1 Reporting Analysis Results to Schools 20
9.2 EPA Reporting 20
10. Proposed Study Timeline 21
11. References 22
Appendix 25
in
-------
List of Tables
Table 1. PCB Vapor Pressure Ranges: Examples from Three Reports 2
Table 2. Summary of Sample Types 8
Table 3. Locations for Collection of Environmental Samples and Caulk 9
Table 4. Example Location Template for Collection of Other Material Samples 10
Table 5. Estimated Number of Samples To Be Collected for PCB Analysis 11
Table 6. Estimated Number of Samples To Be Collected for Laboratory Chamber Testing 11
Table 7. Air Sample Analytes and Target Detection Limits 15
Table 8. Wipe Sample Analytes and Target Detection Limits 15
Table 9. Material Sample Analytes and Target Detection Limits 15
IV
-------
1. Background
Polychlorinated biphenyls (PCBs) are synthetic
chemicals that were manufactured in the United States
between about 1930 and 1977 for use in various
industrial and commercial applications because of their
nonflammability, chemical stability, high boiling point,
and electrical insulation properties (ATSDR, 2000).
PCBs were used in numerous products and processes,
including electrical, heat transfer, and hydraulic
equipment; as plasticizers in various products; in paints
and finishes; in pigments, dyes, and carbonless copy
paper; and in other industrial and commercial
applications. There are no known natural sources of
PCBs. PCBs are either oily liquids or solids that are
colorless to light yellow. PCBs are mixtures of up to 209
individual chlorinated compounds (known as
congeners).
Most of the PCB mixtures manufactured for
commercial use in the United States are known by the
trade name Aroclor. Each specific Aroclor contained
mixtures of some of the 209 congeners, with chlorine
contents of the different Aroclors ranging from 21% to
68%. Between 1957 and 1971, 12 types of Aroclors
were produced (ATSDR, 2000). During this time, PCBs
were used in completely closed systems (such as
transformers and capacitors), nominally closed systems
(such as hydraulic systems and vacuum pumps), and
open systems (such as plasticizers and paints). In
1970, the manufacturer discontinued use of Aroclors in
open products and uses that could lead to direct
transfer into the environment (Erickson, 1997).
Manufacture of PCBs was banned in the United
States by Congress, and their use was phased out,
except for certain limited uses, by 1978 because of
evidence they are persistent in the environment and
can cause harmful health effects. PCBs have been
shown to cause cancer in animals, and, in chronic
animal studies, PCBs have been shown to cause
effects on the immune, reproductive, nervous, and
endocrine systems. In some studies, exposure to PCBs
has been associated with adverse health effects in
humans. Because of potential neurotoxic and endocrine
effects, there is concern regarding children's exposures
to PCBs.
PCBs are highly persistent in the environment.
As such, they are still present in soils and sediments in
many locations and may be found at low levels in
ambient air and water. PCBs can be released into the
environment from hazardous waste sites, illegal or
improper disposal of industrial wastes and consumer
products, leaks from old electrical transformers and
capacitors containing PCBs, and burning of some
wastes in incinerators. PCBs undergo bioaccumulation
and may eventually enter foods that people consume.
Foods with the highest PCB levels are typically fish,
meat, and dairy products. Dietary consumption of
contaminated foods is believed to be a primary route of
exposure to PCBs for people in the United States.
Additional exposure to PCBs may occur for
people who spend time in buildings where PCB-
containing materials are present. A number of products
manufactured before the mid-1970s contained PCBs.
Products that may contain PCBs include
• dielectric fluid in transformers and capacitors;
• other electrical equipment, including voltage
regulators, switches, circuit breakers, reclosers,
bushings, and electromagnets;
• oil used in motors and hydraulic systems;
• old electrical devices or appliances containing
capacitors with PCBs;
• fluorescent light ballasts;
• cable insulation;
• thermal insulation material, including fiberglass, felt,
foam, and cork;
• adhesives and tapes;
• oil-based paints;
• caulk and window glazing;
• plastics;
• carbonless copy paper; and
• floor finish.
Some of these materials can still be found in buildings,
particularly those constructed between 1950 and 1978.
Production of PCBs used as plasticizers by the
Monsanto Industrial Chemicals Company ranged from
approximately 3 million Ibs in 1957 to 19 million Ibs in
1969, decreasing to zero by 1971 (McCarthy, 2009).
Caulk and similar materials that incorporated PCBs as
plasticizers have been examined as a potential source
of exposure to building occupants. Kohleretal. (2005)
reported on concentrations of PCBs in more than 1300
samples of joint sealants collected from buildings in
Switzerland built between 1950 and 1980. Nearly half
(48%) contained PCBs, and levels exceeding 50 ppm
were found in 42% of the samples. Concentrations
exceeding 10,000 ppm were found in 21% of the
samples, whereas concentrations exceeding
100,000 ppm were found in 9.6% of the samples.
Chlorine content was examined in a subset of Swiss
samples, and more than 90% had results consistent
with mixtures of Aroclors 1248, 1254, 1260, and 1262.
PCBs from caulk and other materials containing
PCBs may disperse into the air and dust indoors and
into the soil around older buildings, leading to the
potential for exposures to people using the buildings
and grounds (Hazrati and Harrad, 2006; Herrick et al.,
2004; Kohler et al., 2002, 2005). The different vapor
pressures of the 209 PCB congeners and the effects of
weathering over 30 or more years may affect which
-------
congeners are present and available for exposure from
different environmental media (Harrad et al., 2009). The
extent of exposure to PCBs in indoor environments may
depend on their vaporization into indoor air in
combination with degradation of materials resulting in
contaminated particles available for contact. Many
researchers have measured or estimated vapor
pressures of PCB congeners and mixtures, and results
vary considerably depending on the method (Erickson,
1997). In general, congener vapor pressures decrease
with increasing levels of chlorination (see Table 1).
Within chlorine-number homolog groups, vapor
pressures increase with increasing levels of chlorination
in ortho positions (Faclonerand Bidleman, 1994).
Based on these results, it might be anticipated that
inhalation exposures to PCB vapors will be weighted to
congeners with lower chlorine numbers unless
PCB-contaminated particulate matter becomes
available for airborne dispersal. It also may mean that
PCB congeners measured in indoor air may not match
patterns found in the Aroclor mixtures used in materials
in a building.
Table 1. PCB Vapor Pressure Ranges: Examples from Three Reports
Homolog Series
Monochlorobiphenyls
Dichlorobiphenyls
Trichlorobiphenyls
Tetrachlorobiphenyls
Pentachlorobiphenyls
Hexachlorobiphenyls
Heptachlorobiphenyls
Octachlorobiphenyls
Nonachlorobiphenyls
Decachlorobiphenyl
DelleSite, 1997a
Pa at 25° C
(Number of Congeners)
7.9E-2to2.1E(3)
7.4E-4 to 3.2E-1 (5)
4.8E-3 to 7.6E-2 (5)
1.8E-5to2.2E-2(5)
4.0E-4to2.2E-3(1)
2.9E-6 to 1 .6E-3 (3)
None reported
2.4E-6to3.0E-5(1)
None reported
2.9E-9to1.4E-5(1)
Falconer and
Bidleman, 1994b
Pa at 20° C or 25° C
Not reported
1.0E-1 to2.5E-1
6.3E-3 to 4.0E-2
3.2E-3to 1.6E-2
5.0E-4 to 2.5E-3
2.5E-4 to 7.9E-4
5.0E-5 to 2.5E-4
Not reported
Not reported
Not reported
Holmes et al., 1993°
Pa at 25° C
(Number of Congeners)
3.2E-1 to 9.3E-1 (3)
5.1E-2to4.2E-1 (12)
8.4E-3to 1.7E-1 (24)
1 .4E-3 to 6.6E-2 (42)
2.7E-4to 1.7E-2(46)
5.4E-5 to 6.4E-3 (35)
1 .4E-4 to 1 .6E-3 (24)
3.8E-5to6.2E-4(12)
1 .OE-4 to 1 .3E-4 (3)
2.8E-5 (1)
Delle Site, 1997: Compiled vapor pressures for selected congeners from multiple references using direct, indirect, and
prediction methods
Falconer and Bidleman, 1994: Liquid saturation vapor pressures; range across average vapor pressures of different levels of
othro-substituted chlorines within homolog (estimated from Figure 2)
cHolmes et al., 1993: Originally from Buckhard et al., 1985, ES&T, 22:503-509. Based on liquid or sub-cooled liquid values
rather than from solid phase
Harrad et al. (2009) noted that the combination
of residential indoor air inhalation and dust ingestion
could exceed dietary intake of PCBs in some scenarios.
Exposures to PCBs in air and dust from contaminated
nonresidential buildings could increase exposures for
some people. For example, levels of PCBs in the indoor
air of some office buildings with PCB-contaminated
sealants (up to 6000 ng/m3 [Kohler et al., 2005]) exceed
the levels reported in residential indoor air (maximum of
14 ng/m3 total PCBs [Harrad et al., 2009] and 35 ng/m3
for congeners 52, 105, and 153 in a contaminated
home [Rudeletal., 2008]).
Of particular concern is the potential for school
children's exposures to PCBs in older schools. Schools
constructed between 1950 and 1978 may contain caulk
that incorporated PCBs as a plasticizer. Caulk
containing PCBs may have been used around exterior
windows and doors, exterior building joints, and in
interior locations. PCBs may vaporize from the caulk
and become airborne. PCBs may be absorbed onto
(or into) other surfaces, materials, or dust. Caulk may
degrade or suffer abrasion wear that can create PCB-
containing dust that is then available for transport in
indoor areas. PCBs from exterior caulk may be
deposited into soils near the school building.
There has been no systematic effort to
characterize PCB sources and environmental levels at
schools across the United States. Measurements of
PCBs in caulk have been made for several college
buildings and a number of primary and secondary
schools. Environmental samples, including air, dust,
wipe, and soil samples, also have been collected at a
number of school buildings. Measurement results for
PCBs in schools in the United States have not been
widely published in the scientific literature, although
Herrick et al. (2007) reported on PCBs in soil collected
near buildings with PCB-contaminated caulk or joint
material. Measurement results have been reported in
presentation and report formats (e.g., Coghlan et al.,
2002; Sullivan, 2008; TRC, 2006, 2008, 2009), and
results from individual schools have been compiled on
the internet (pcbsinschools.org). Total PCBs in caulk in
U.S. college and school buildings have ranged from not
detected to over 200,000 ppm. Concentrations in school
indoor air have ranged up to approximately
1000 ng/m3. Total PCB levels in dust have ranged up to
approximately 80 ppm. Wipe samples have ranged up
to approximately 1 ug/cm2 total PCBs. Soil
concentrations in samples collected next to buildings
have ranged up to approximately 80 ppm of total PCBs.
-------
Although caulk is believed to be a primary
source of PCBs in some older schools, there is still
considerable uncertainty regarding the extent to which
PCBs in other materials used in schools might
contribute to exposures (Coghlan et al., 2002). Other
primary sources (materials manufactured with and
containing PCBs) or secondary sources may be present
in some schools. For example, window glazing has
been found in several locations to contain levels of
PCBs greater than 50 ppm. Secondary sources might
include surfaces, materials, and dust that have been
contaminated through transport of PCBs from caulk.
An in-depth investigation in a high school found PCBs
in numerous materials, including but not limited to
laminate adhesive, mastics, paint, gasket, carpet, foam
padding, and bulk dust (TRC, 2008; Sullivan, 2008;
TRC, 2009). Some of these materials had
concentrations exceeding 10 ppm, ranging up to more
than 250 ppm. If other primary and secondary PCB
sources are present in schools, they could contribute to
exposures to children. To make sound decisions
regarding reducing exposures to PCBs in schools, it is
important to understand the range of potential sources
of PCBs in schools; their contributions to PCBs in air,
dust, and soil; and the magnitude of potential exposures
to children, teachers, and staff in school environments.
There is very limited information on PCBs in
schools in the United States. Neither the sources,
including PCB-contaminated caulk, nor the routes of
exposure have been well characterized in schools.
Indoor models describing the emission, transport,
accumulation, and disappearance of PCBs in buildings
are lacking. As such, there remains considerable
uncertainty regarding the extent to which children and
staff members may exposed to PCBs in school
environments. Research is needed to help fill these
information gaps to improve our understanding of
exposure to PCBs in schools.
-------
2. Research Objectives
Research on sources of PCBs and levels in
school environments is needed to improve risk
management decisions. To better understand the
significance of PCB-contaminated caulk as a source of
exposures to children, teachers, and staff in school
buildings, the U.S. Environment Protection Agency's
(EPA's) Office of Research and Development (ORD)
plans to
• characterize PCB-contaminated caulk and other
potential sources of PCBs in schools;
• investigate relationships between PCB
concentrations in air, on surfaces, and in dust and
soil with potential sources in school buildings;
• evaluate which routes of exposure (e.g., inhalation,
contact with surfaces or dust) are likely to be most
important;
• improve exposure assessment models for school-
related exposures and examine the feasibility for
development of an indoor model for PCBs; and
• provide samples, data, and other information to assist
in developing risk management practices for reducing
exposure to PCBs in schools.
To meet these research objectives, the ORD
National Exposure Research Laboratory (NERL) plans
to conduct a measurement study in up to nine schools
in the United States. The research described in this
study design is being coordinated with research efforts
in the ORD National Risk Management Research
Laboratory (NRMRL) that are aimed at evaluating PCB
emission rates, transport, and exposure mitigation
methods.
-------
3. Study Summary
3.1 Study Overview
A brief overview of the research study is
outlined below. Details regarding each element are
provided in subsequent sections of this research plan.
ORD research will consist of both field and
laboratory components. This study design addresses
the school measurement field study and also includes
the collection of materials from schools that will be used
in NRMRL laboratory investigations of PCB emissions
and transport. The field measurement study will involve
recruitment of a limited number of schools (up to nine)
with PCB-contaminated caulk to participate in the study.
Limiting the number of schools will allow a more
intensive characterization of PCB sources and
environmental measurements that can be used to better
understand the relationships between sources,
environmental concentrations in selected media (dust
and air), and potential exposure. At each school, PCBs
will be measured in indoor and outdoor air, soil adjacent
to the building, and on surfaces or in dust at multiple
locations within the school building. Caulk and other
materials that may be primary or secondary PCB
sources will be collected for PCB analysis. Some
materials also will be collected for subsequent chamber
tests to characterize PCB emissions and transport.
Information will be collected at each school regarding
building characteristics, building materials, ventilation
systems, school operations, and other factors that may
affect the distribution of PCBs in the school building and
the potential for exposure. All samples will be analyzed
for selected Aroclors or homologs to determine total
PCB levels, and a subset of samples will be analyzed
for specific PCB congeners. The measurements and
other study information will be used as input data into
ORD's Stochastic Human Exposure and Dose
Simulation (SHEDS) model to improve the model's
ability to predict exposure distributions for school-age
children, teachers, and other workers under selected
scenarios. In combination with laboratory test results,
the measurement of source materials and
environmental media also may lead to the development
of indoor PCB models. Study results will be provided to
the schools and to Agency Offices and Regions for use
in improving tools for exposure assessment, risk
assessment, and risk management.
3.2 Investigators
Researchers and staff members for this
research study include the following individuals.
• Kent Thomas, NERL, Principal Investigator and
Study Manager
• Ron Williams, NERL, Co-principal Investigator
• Roy Fortmann, NERL, HEASD Division Director and
Co-investigator
• Don Whitaker, NERL, EMAB Branch Chief and
Co-investigator
• Paul Jones, NERL, Co-investigator (statistician)
• Jiaping Xue, NERL, Co-investigator (SHEDS
modeling)
• Carry Croghan, NERL, Co-investigator (data
management)
• Zhishi Guo, NRMRL, Collaborating Investigator
• Contractor, TBD, Sample Collection
• Contractor, TBD, Sample Analysis
3.3 Study Limitations
A primary objective of the school PCB
measurement research study is to characterize
potential PCB sources and environmental levels of
PCBs in multiple locations in schools. These types of
data are needed to understand sources and
relationships between sources and environmental levels
of PCBs to make the most informed risk reduction
decisions. Because of the intensive nature of this
research requiring a large number of measurements
within schools, it will not be feasible to select and enroll
a representative sample of schools for making
inferences about the presence and levels of PCBs in
schools either nationally or within a region. A simple
survey based on only one or two measurements in a
large number of schools would not address the primary
study goal of providing information to inform risk
management decisions in schools. It is also not clear, at
this time, how many and what type of samples, would
need to be collected at each school in a larger
prevalence survey. Results from this study can be used
to help address that question.
Although the study is designed to collect
samples of numerous materials in school buildings,
including those thought likely to contain PCBs, it will not
be possible to collect and analyze all materials present
inside and outside of the buildings. It is possible that
some potentially important primary or secondary PCB
sources will be missed.
The proposed research will not collect or
analyze blood samples from students, teachers, or
school staff. Therefore, it will not be possible to
examine the associations or relationships between
environmental concentrations in schools and PCB body
burden. Although this would be important information, it
would require additional resources to develop control
groups and assess non-school-related sources of PCB
exposures that would limit the primary research goal of
this study.
-------
4. School Recruitment and Engagement
4.1 Number and Types of Schools
It is beyond the scope of the current research
and available resources to understand the extent of
what is likely to be a nationwide issue. Also, at this time,
it is not clear exactly which types of measures and how
many measures per school would be required to
implement a large-scale prevalence survey. The
primary goal of this research effort is to collect a large
number of measurements in schools to help understand
within-school variability and, most importantly,
relationships between PCBs in materials and levels in
air, on surfaces, and in dust and soil. This information is
important for better understanding potential exposures,
informing risk management decisionmaking and
mitigation approaches, and how to better evaluate
schools for PCBs in the future. The research study will
be an intensive monitoring effort performed in up to nine
schools. Elementary, middle, and high schools are of
most interest for this research study. In the event that
insufficient schools agree to participate, college
buildings may be considered for selection. Schools that
were constructed between 1950 and 1978 are most
likely to have PCB-containing caulk, so schools that
were constructed or significantly renovated during that
period will be monitored. Because the study focuses on
PCB source assessment and environmental levels
associated with PCB-containing materials, no control
school buildings constructed after 1978 will be
monitored. In the absence of control school buildings,
appropriate quality control (QC) materials and
procedures will be used to ensure that PCBs measured
in school environments and materials are not the result
of background contamination of materials used for
sample collection and analysis.
4.2 Identifying and Recruiting Schools
NERL will work with EPA regions and State and
local governments as needed to locate schools
interested in having monitoring performed. An
information sheet that provides a summary of the
research goals and study approach has been prepared
as a communication tool and is included in the
appendix. We anticipate that schools that already have
found elevated PCBs in caulk or those that are planning
renovations are most likely to be monitored. In the
event that work is considered in a school where it is not
known whether PCBs are present in the caulk, it may
be necessary to perform a screening sample collection
to determine whether the school is eligible for
participation in the full measurement study.
4.3 Development of Individual School
Sampling Plans
The general approach for the proposed
sampling is described in later sections. However, each
school will have unique characteristics, operations, and
requirements that will necessitate development of
school-specific plans. We will work collaboratively with
the school administration to develop a plan for each
school that identifies access times and procedures,
sampling locations, the specific materials to be
collected, and the appropriate school contacts. Part of
this collaborative effort will be an on-site scouting visit
to examine the school space and discuss logistical,
sample collection, and scheduling issues. Research
staff will work with school administrators and staff to
minimize the disruption of school activities and
operations. The school plans also may identify
procedures for reporting study results to the school.
4.4 Information about the School
Some selected information about the school is
needed for two reasons. First, information regarding the
layout, history, and operations is needed to plan the
sample collection locations and to identify materials for
sample collection. Second, selected information also is
needed as inputs for the SHEDS PCB modeling effort.
Two structured forms for recording information about
the school will be developed. The first will be used
during a premonitoring scouting visit to the school to
plan monitoring activities with regard to sampling dates
and times, sampling locations, and identification of the
materials that are both available for sampling and for
which the school provides approval to collect at each
location. This information also will be used later during
data analysis to facilitate examination of correlations
and source relationships for environmental samples.
A detailed walk-through of the school with appropriate
school staff will be needed during this scouting visit.
Results from this scouting visit will be incorporated in
the school-specific study plan. Information of interest
includes
• general school layout (copies of floor plans would be
helpful if available);
• heating, ventilation, and air conditioning (HVAC)
systems information (type of HVAC, areas served,
and use during the year and during the sample
collection period);
• identifying specific locations inside and outside the
school for sample collection;
• materials present in the building at locations to be
monitored;
-------
• materials that the school will allow to be collected for
PCB measurements;
• condition of materials (i.e., caulk that is intact or has
visible signs of deterioration); and
• scheduling information regarding access to the
school for sample collection.
The second form will be used to collect
additional information to be used as input to the SHEDS
model. This form can be provided to the appropriate
school staff for completion and return to the research
team either prior to or following the sampling visit. In
general, the types of information of interest include
• school construction and renovation dates and history,
• school operation (start and end dates and times,
class period times, and other uses),
• general enrollment information (total students, total
number in each grade, and general class sizes),
• information on locations monitored (number of
classes per day by grade level), and
• cleaning information (cleaning methods and
frequency of cleaning for each method, including
monitored areas).
Study staff will need access to school staff who
can provide this SHEDS model input information. It will
be preferable to collect this information during the
scouting visit.
The research described in this study plan is not
human subjects research. No personal information will
be collected from or about any student, teacher, or staff
member. The study will not involve any measurements
of PCBs in the children or school staff, nor do we
anticipate that changes to school activities will be
needed. The research study will not alter the school
environment or any individual's activities.
4.5 Time Needed at Each School
If a school is being considered for participation
in the research study, and it is not known if caulk in the
school contains PCBs, it may be necessary to perform
a screening sample collection visit. At this visit, a small
sample of caulk would be collected from several
locations at the building. The samples would be
analyzed to determine whether PCBs are present, and
their approximate concentrations. If elevated PCB
levels are found, then the school would be eligible to
participate in the full study, with visits as described
below. At this time, there is no predetermined
concentration that would determine potential eligibility;
this likely will be based on a relative basis that would
consider PCB concentrations, the extent of the caulk in
the building, and levels found in other schools.
We anticipate that 3 days will be required to
complete monitoring activities at each school, including
a day for a scouting visit by study staff and a 2-day
period for sample collection. A scouting visit by the
study staff will be performed on 1 day several days or
weeks prior to collecting samples. Study staff will meet
with school administrators and staff to collect the types
of information described above. Part of this visit will be
a walk-through of the school to identify sample
collection locations and the materials of interest to be
collected as part of the PCB source assessment. Study
staff also will discuss logistical planning issues at this
time, including sample collection dates and times,
access and school staff assistance for access, and any
sampling safety and security issues. This initial scouting
visit will take several hours to complete. Information
from this scouting visit and from any other discussions
with school administrators will be used to develop a
brief written sampling plan for the school. The plan will
be provided to the school administrators and EPA
managers for approval prior to the sample collection
visit.
Sample collection will be conducted over a
2-day period. This length of time is needed to allow
collection of air samples with acceptable detection limits
for PCBs. Air sample collection will be initiated on 1 day
and completed approximately 24 h later on the following
day. The remainder of the wipe, dust, caulk, soil, and
material samples also will be collected during these
2 days. It is preferable to collect the air samples over a
period in which any HVAC system is running under
normal conditions and when students are present at the
school because these activities may affect the level of
dust and PCBs in the air. However, research study staff
will work with school administrators to determine the
most suitable sample collection times and dates. It is
acceptable to collect air samples during periods when
students are not present. The remainder of the
environmental and material samples should be
collected during times when students are not present.
-------
5. Sample Collection
5.1 Types of Samples
Screening samples of caulk (interior and
exterior) and window glazing may be collected at
candidate schools to assess whether elevated levels of
PCBs are present. The collection of screening samples
is intended to identify buildings in which PCB-containing
caulk is present to make decisions regarding the
inclusion of a school in the full PCBs in Schools
research study. Screening assessments will be
performed to determine whether PCBs are present in
the caulk and window glazing materials used most
widely in a building because reports to date suggest
that these materials are likely to have the potential for
the highest PCB concentrations. A research operating
procedure, "Screening Collection of Caulk and Window
Glazing from Buildings for PCB Analysis" has been
developed for collection of screening samples (U.S.
EPA, 2010). Sample collection will focus on locations
where students are most likely to spend time, including
classrooms, libraries, multipurpose rooms,
gymnasiums, cafeterias, hallways, and outdoors. It may
not be possible to identify and collect every different
kind of caulk and window glazing in a building during
the screening process. The screening approach may
not rule out the presence of PCB-containing caulk in
every location in a building, nor will it determine whether
there may be other primary or secondary sources of
PCBs in a building. We have not developed any specific
inclusion or exclusion selection criteria based on
screening results. A decision to include or not include a
school may depend on several factors, including the
concentration of PCBs in caulk, how widespread PCB-
containing caulk is in the school, and the overall
availability of schools for recruitment and selection. It is
not known how many schools might need to be
screened to identify up to nine schools that have PCBs
in caulk that would subsequently be selected for full
monitoring.
For schools selected for full monitoring,
samples will be collected in each school to characterize
school environmental levels in air, on surfaces, and in
dust and soil; to measure levels of PCBs in caulks; and
to evaluate other potential primary and secondary
sources of PCBs in other materials. A summary of the
types of samples to be collected is shown in Table 2.
Materials other than caulk that are of potential interest
also are shown in Table 2. These materials are included
based on reports of PCB-containing materials in older
schools or other buildings and the potential of the
materials to contribute to exposures based on
prevalence and area. However, not all of these
materials will be available in each school and school
Table 2. Summary of Sample Types
Sample Type
School Environment
Indoor air
Surface wipe - desk/table surfaces3
Surface wipe - floor/wall/windowsill
surfaces'3
Bulk dust
Outdoor air
Exterior soil
Caulks
Caulk from exterior window/door
Interior caulk
Exterior building joint caulk
Other Materials of Interest0
Material adjacent to caulk
Floor tile mastic
Floor tile
Cove/molding mastic
Cove/molding
Wall paint
Ceiling tile
Gasket
Vent coating
Window glazing
Foam padding
Floor finish
Carpet and pad
Light ballast wipes
Others, to be determined (TBD)
Sample
Type Code
IA
ST
SB
BD
OA
OS
EC
1C
JC
AC
FM
FT
CM
CV
WP
CT
GK
VC
WG
FP
FF
CA
LB
OT
Two wipe samples, collected from a student desk and
classroom table in selected rooms
bUp to three wipe samples, collected from the floor, a
windowsill, and a wall
cMaterials to be collected will depend on their presence and
availability at each school and school location. Specific plans
will be developed for each school.
location, and school permission will be needed to
collect samples of materials when they are present.
Other materials of interest may be identified
based on school scouting visits.
The types and locations of samples in this
research study are intended to support the research
goals identified in Section 2. The research focus is on
developing a better understanding of potential
exposures and relationships between sources and
environmental levels. The sampling approach in this
research effort is not intended to provide data for
characterization of sampling for PCB remediation waste
or for verifying completion of cleanup and disposal.
-------
5.2 Sampling Locations
It will not be possible to characterize PCB levels
in all school building locations. The focus of this
research is on areas where children and staff spend the
most time during the school day and during after-school
programs. As such, samples will not be collected in
basements (unless classrooms or other student activity
areas are present at that level) or mechanical
equipment rooms, even though there may be PCB
sources present. A proposed list of sample collection
locations for the environmental and caulk samples is
shown in Table 3. One study goal is to collect samples
in as many interior locations as is feasible and
affordable, so seven rooms and the HVAC system
(when one is present) have been identified, resources
permitting.
Table 3. Locations for Collection of Environmental Samples and Caulk
Indoor Locations3
Classroom 1
Classroom 2
Classroom 3
Hallway
Cafeteria
Gymnasium
Library
HVAC system
Outdoor Locations3
First side of building
Second side of building
Play area
Environmental Samples
IA
O
O
O
O
O
O
O
ST
O
O
O
O
O
SB
O
O
O
O
O
O
O
BD
O
O
O
O
O
O
OA
O
OS
O
O
O
Caulk Samples
EC
O
O
O
O
O
O
O
1C
O
O
O
O
O
O
O
O
JC
O
O
Locations will depend on their presence and availability at each school; specific plans will be developed for each school.
Rooms with primary sources of PCBs, such as
caulk or window glazing, are likely to have the highest
levels of PCBs in the air and dust and on surfaces.
Because a goal of this research is to better understand
relationships between sources and environmental
levels, rooms with known sources of PCBs are of most
interest, and their selection for monitoring would take
precedence.
At least three classrooms will be selected for
sample collection. The three classrooms will be used for
the most intensive measurements, where the greatest
numbers and types of environmental and material
samples will be collected to enable the most complete
assessment of potential relationships between PCB
sources and indoor environmental levels. Depending on
the school layout, classrooms will be selected for as
much diversity as possible based on characteristics
such as construction material differences, location
(different wings or floors), and different HVAC zones
(if present). When multifloor buildings are monitored,
precedence will be to first select rooms on different
floors and, then, to select on other factors. For buildings
with three floors used for instruction, one classroom will
be selected on each floor. For schools with two floors
used for instruction, two of the classrooms will be
chosen from the ground floor with the remaining
classroom from the second floor. A decision may be
made to collect samples in an additional classroom if a
library or separate cafeteria and gymnasium are not
present.
One air sample will be collected outside of the
school building at the same time indoor samples are
collected. When present, caulk used in exterior building
joints will be collected from up to three locations outside
of the school building. Caulk also will be collected from
exterior window and doorframes, when present. Soil
samples will be collected from up to two areas outside
and adjacent to the school building. When exterior joint
caulk is present, a soil sample will be collected at one of
those locations. A soil sample may be collected below
windows with exterior caulk, particularly if PCBs are
present. Another soil sample will be collected in one
area (i.e., playground, athletic playing field) where
children spend time, provided that such an area is
present on school grounds.
Up to 34 samples of materials, in addition to
caulk, may be collected from interior or exterior
locations in each school. One focus for material
collection will be the three rooms with the most
intensive monitoring. In these rooms, attempts will be
made to collect bulk dust; any materials that potentially
could contain PCBs; and samples of materials, such as
window glazing, floor tile, tile mastic, ceiling tile, and
wall paint, that make up the bulk of the room's surface
area. Other potential primary PCB source materials
would be scheduled for collection from other rooms
where air, wipe, and caulk samples are collected when
such materials can be identified.
Some "hidden" materials such as insulation also
may serve as sources or sinks for PCBs. Although
these materials are of interest from an indoor modeling
perspective, it may be difficult to collect such materials.
Some caulks may be present in inaccessible locations
between windows and buildings or between building
-------
structural components. Also, although collection of
some "subsystems" or groups of materials could be of
interest, it is unlikely that such systems easily could be
collected and analyzed. Samples of hidden materials
will be considered on a case-by-case basis, but no
destructive methods will be used to gain access to such
materials.
The selection of materials to be collected will be made
during the scouting visit to the school and will depend
on both the availability of the material and the school's
permission for collection of a sample of sufficient size
for PCB analysis. Identification of suitable materials will
be a collaborative effort by research staff and school
staff. An example of what materials might be collected
is shown in Table 4.
Table 4. Exam
Indoor Locations
Classroom 1
Classroom 2
Classroom 3
Gymnasium
Library
HVAC system
V
C
o
A
C
O
o
o
pie Location Template for Collection of Other Material Samples3
F
M
O
O
O
F
T
O
O
O
C
M
O
O
O
C
V
o
o
o
w
p
o
o
o
w
G
O
o
o
o
L
B
O
O
O
C
T
O
o
o
L
A
L
M
C
A
O
G
K
O
F
P
O
O
F
F
O
0
T
Materials to be collected will depend on their presence and availability at each school and school location. Specific plans will be
developed for each school and location. Up to 34 noncaulk material samples will be collected from each school. Where feasible,
multiple materials will be collected from three rooms to aid source assessment analyses.
5.3 Numbers of Samples
The number of caulk and window glazing
samples that might need to be collected to identify
eligible schools is not known. It is anticipated that up to
10 to 15 samples and QC samples may need to be
collected at each school, but this will vary on the
presence and extent of materials of interest in locations
where students spend the most time.
The proposed number of samples to be
collected for PCB analysis at each school selected for
full monitoring and the total number estimated for
collection across the study is shown in Table 5. The
estimated total number of samples is based on sample
collection in nine schools and collection of samples in
the locations indicated in Tables 3 and 4. The actual
final number of samples will be based on the specific
measurement plans developed for each school that
take into account the availability of space and materials.
Some samples will be collected for laboratory
chamber emissions testing by NRMRL, in addition to
samples being collected for PCB analysis. The numbers
and types of samples proposed for collection to support
chamber testing are shown in Table 6. Collection of
these materials will depend on the availability of
sufficient amounts to support chamber measurements.
Collection of these materials has been proposed to be
completed with the first five schools in the
measurement study to enable sufficient time for the
laboratory testing.
5.4 Sample Collection Methods
Sample collection methods are described
below. General guidance is provided regarding
collection locations in some cases; however, it is
recognized that final decisions regarding collection
locations will be based on the presence and availability
of the location or material.
5.4.1 Indoor and Outdoor Air Samples
U.S. EPA Method TO1OA will be used to collect
total PCBs in air, using a low-volume sampling
approach to minimize the size and noise of the pumps
to be used in school buildings. Sample filters (ORBO
1000 or similar glass tubes), which are 30-mm x 70-mm
tubes filled with precleaned open-cell polyurethane
foam (PUF), will be used. Collection will occur without
use of the optional XAD (highly absorbent resin)
sandwich. The optional total suspended particle (TSP;
quartz) filter collection apparatus will be used as part of
the sample filter assembly; however, the filter and the
PUF will be extracted and analyzed together as a single
sample. Separate particle- and gas-phase air
concentrations will not be obtained. Collection of inline
backup filters will be performed for 5% of the samples
to assess whether breakthrough may be occurring
using this sampling approach.
Sample collection tubes will be situated with the
inlet in a downward facing position at a height of 1 m
from ground or floor level. A sampling stand or similar
apparatus will be used to secure the sampler, as well as
to provide weather proofing for the outdoor monitor. An
active air-flow pump, capable of unattended 24-h
operation (battery or AC) will be used to provide a flow
of 3.5 to 5.0 L/min through the PUF. The highest flow
capacity pump that is available will be used to maximize
the volume of air sampled. However, if samples are
collected during school days, it may be necessary to
minimize pump noise and protect systems from contact.
Flow measurements will be performed and recorded at
the initiation of sampling and then again at the
completion of the nominal 24-h sampling period. Start
10
-------
Table 5. Estimated Number of Samples To Be Collected for PCB Analysis
Sample Type
School Environment
Indoor air
Surface wipe - desk/table
surfaces
Surface wipe -
floor/wall/windowsill
Bulk dust
Outdoor air
Exterior soil - adjacent to building
Exterior soil - from play area or
field
Caulks
Caulk from exterior windows and
doors
Interior caulk
Exterior building joint caulk
Material Adjoining Caulk
Material directly adjoining
Material at increasing distance
Other Materials of Interest0
Total Estimate
Samples
Collected
per School
7
10
17
6
1
6
1
7
8
2
3
3
34
106
Number of
Schools3
9
9
9
9
9
9
9
9
9
9
9
9
9
9
Estimated
Total
School
Samples
63
90
153
54
9
54
9
63
72
18
27
27
306
954
Estimated
QA/QC
Samples'1
18
9
9
9
0
18
0
6
6
6
0
0
0
81
Total
Estimated
Samples
81
99
162
63
9
72
9
69
78
24
27
27
306
1026
Up to nine schools will be monitored.
bQC samples may include field blanks, spiked field controls, and duplicate or side-by-side samples. These do not include
laboratory QC and quality assurance (QA) analyses that are part of a standard laboratory QA program.
cMaterials to be collected will depend on their presence and availability at each school and school location. Specific plans will be
developed for each school and location. Up to 34 non-caulk-material samples will be collected from each school. Where
feasible, multiple materials will be collected from three rooms to aid source assessment analyses.
Table 6. Estimated Number of Samples To Be Collected for Laboratory Chamber Testing3
Sample Type
School Environment
Bulk dust
Caulks
Caulk from exterior window/door
Interior caulk
Other Materials of Interest3
Total Estimate
Number of
Schools
5
5
5
5
Samples
Collected
per School
1
2
1
1
5
Estimated
QA/QC
Samples
0
0
0
0
Total Estimated
Samples
5
10
5
5
25
Materials are to be collected from the first five schools monitored. Materials to be collected will depend on their presence and
availability at each school in amounts sufficient to support chamber emissions testing.
and stop times will be recorded, so that a cumulative
amount of time at an average flow rate can be
calculated to yield a volume of air sample. It is
anticipated that approximately 5 to 7 m3 of air will be
sampled over the monitoring event, depending on the
pump flow rate and actual sampling duration.
Devices to record the interior and exterior
building temperature during sample collection will be
deployed along with the air monitors. Ambient
atmospheric pressure readings will be collected from
the nearest official recording station for the area.
5.4.2 Wipe Samples
Two types of wipe samples will be collected.
The first type will be individual samples collected from
two surfaces that may be contacted routinely by a
student. In classrooms, this may be a student desk and
a classroom table. The second type will be samples
collected from building surfaces. In classrooms, this will
be a floor wipe sample, a wall wipe sample, and a
sample collected from a window sill. Window sill
samples would not be collected in rooms other than
classrooms. Thus, up to five wipe samples could be
11
-------
collected from classrooms, and four samples in libraries
and cafeterias. Two samples (wall and floor) will be
collected from hallways or gymnasiums that do not
have tables or desks.
Wipe samples will be collected based on the
wipe sample collection procedure described in ASTM
Method D6661-01 [2006], "Standard Practice for Field
Collection of Organic Compounds from Surfaces Using
Wipe Sampling." PCB-free gauze wipes will be wetted
with hexane to collect surface wipe samples. Each wipe
sample will be collected in a 100-cm2 area. Transfer of
PCBs from the surface will occur through physical
wiping of the defined surface area while applying
moderate pressure. Wipes will be handled using
appropriate chemical-resistant gloves, followed by
storage in a clean amber glass jar having a Teflon-lined
cap.
5.4.3 Dust Samples
There is the potential for loose (accumulated
and visible) dust to be evident in schoolrooms or in the
HVAC system (if present). If a whole-building HVAC
system is not present, dust will be collected from up to
two single-room heating/cooling systems if they are
present. Where sufficient quantities of dust are
available in several predefined locations, the dust will
be collected using precleaned stainless steel
implements or other appropriate means and transferred
into a clean amber vial with Teflon-lined cap. A sample
size of 2 g for each sample is preferred, whereas 1 g is
the minimum sample size. Vacuum devices using an
EPA-approved collection media (filter or chamber)
represent an alternative method for dust collection. Any
vacuum device being reused will have separate
collection manifold inlets (trapping media), so that
cross-contamination effects are minimized through use
of precleaned inlets between sampling sites. Dust
samples will be collected following the completion of air
sample collection.
In a subset of locations, a dust sample will be
collected, when available, to support laboratory
emissions testing research. To the extent feasible,
samples of 10 g or more of dust are preferred to
support this effort.
5.4.4 Outdoor Soil Samples
Soil samples will be collected at two locations
adjacent to the building and below joint material or
exterior caulk. These samples will be collected where
PCB-contaminated caulk or joint material is known to be
present or may be present if the PCB content of the
caulk or joint material is unknown. At each of the two
locations, soil samples will be collected at three
distances from the building: 0.5 ft, 3 ft, and 8 ft. (Note:
Samples at the two greater distances may be analyzed
only if the closest sample concentration exceeds 1 ppm
total PCBs). Each soil sample will be collected to a 2-in
depth. The sample will include soil from the root zone if
vegetation is present but will not include vegetation.
A total sample size of approximately 100 g will be
collected at each sampling location. Soil sampling will
take place using a precleaned stainless scoop and
deposited into a clean amber glass container having a
Teflon-lined cap.
Another soil sample will be collected from an
area where children spend time, such as a playground
or athletic playing field. In this case, a composite
sample will be collected. This sample is not intended to
fully characterize a school yard but is intended to
assess whether there is a potential for soil exposures in
play areas that would need to be considered in
exposure modeling efforts. An area 10 ft x 10 ft will be
identified. Soil samples will be collected from the 0 to
2-in depth line at five locations within the designated
area, and equal amounts of soil from each location will
be combined into a single container. A total sample size
of approximately 100 g will be collected.
5.4.5 Caulk Samples
Caulk samples will be collected from rooms
selected for sampling in a building. Samples of interior
and exterior caulk will be collected, when present.
Samples of known PCB-containing caulk have the
highest priority for collection. Samples of other caulks
will be collected if their PCB content is not known. The
exact number and locations for collection will be
decided on a room-by-room basis.
Sample collection will generally follow the
standard operating procedure, "Sampling Porous
Surfaces for Polychlorinated Biphenyls (PCBs)" (U.S.
EPA, 2008a). However, only one sample of each
selected type of caulk will be collected, rather than
three, with duplicate samples collected in designated
rooms to examine variability. Also, a sample size of 2 to
4 g will be collected, rather than 10 g. Caulk samples
will be collected by physically removing sections of
caulk using clean knives, scalpels, and tweezers (or
other clean implements, as needed). A piece or pieces
of caulk of at least 2 g will be removed from the site of
interest and placed in a single precleaned amber glass
jar having a Teflon-lined cap. Caulk samples will be
collected as they exist in the selected rooms; no
emphasis will be made on collecting samples of
deteriorating caulk because intact caulk potentially may
contain higher levels of PCBs.
In a subset of locations, a second caulk sample
will be collected, when available, to support laboratory
emissions testing research. For this sample subset, a
10-in section of caulk will be removed. To the extent
feasible, it is preferred that the section be kept as intact
as possible. However, it is understood that it may not be
possible to remove intact 10-in sections. Collection
efforts should focus on removing the largest intact
pieces possible to support the laboratory emissions
testing research.
In addition to room location, additional
descriptive information will be collected regarding the
caulk collected for analysis. Information on its location,
use, current condition, and the presence of any paints
or coatings will be collected.
12
-------
5.4.6 Caulk-Adjacent Material Samples
Materials adjoining PCB-contaminated caulk
may absorb PCBs over long periods of time. In some
situations, remediation practices may call for removal of
the contaminated adjoining material, as well as the
contaminated caulk. However, there is uncertainty
regarding how much material would have to be
removed for remediation. To address this uncertainty,
samples of porous materials directly adjoining caulk will
be collected, where feasible. And, in a subset of
locations (one set of samples per school), up to four
samples will be collected perpendicularly from caulk at
the distances of 0, 1.0, 2.0, and 4.0 in from the caulk.
It is preferred that a range of different types of materials
be assessed in this way across the study, including
wood, brick, block and other masonry, and drywall or
plaster. Material samples will be collected only after air
sample collection has been completed.
Sample collection generally will follow the
standard operating procedure, "Sampling Porous
Surfaces for Polychlorinated Biphenyls (PCBs)" (U.S.
EPA, 2008a). However, multiple-depth sampling will not
be used. Samples will be collected from the first 0.5 in
from the surface that is available for contact. It is likely
that power tools will be required to collect these
samples, and steps must be taken to minimize
contamination of school areas with dust. Different
approaches will be considered, depending on the
material. The method should minimize the potential for
alteration of PCB content via heating or loss of fine
material. Appropriate methods will be discussed with
the school managers to ensure that collection of the
material is acceptable, and that the method used
therefore is acceptable. Alternative approaches that
produce less dust in the school environment may be
considered.
5.4.7 Other Material Samples
As described in earlier sections, selected
materials other than caulk will be collected, when
available and accessible, as potential primary or
secondary sources of PCBs in school buildings. Each of
these materials may require different collection
methods. Where possible, 10 g of material is the
preferred material sample size, with a minimum sample
size of 2 g. Samples will be collected using instruments
such as a scalpel, razor, spatula, utility knife, putty
knife, or other hand tool, as needed. All tools will be
precleaned to prevent cross-contamination of PCBs.
Other approaches, including some limited use of power
tools, may be considered for collection of materials;
however, these approaches will need to be discussed
and agreed to by all parties prior to implementation, and
steps be taken to avoid contamination of school areas.
It is not certain how fluorescent light ballast samples will
be collected. One approach may be to use wipe
material to wipe the ballast if accessible or an area
around the ballast if this can be done safely. Materials
collected for PCB analysis will be stored in clean amber
glass containers with Teflon-lined lids. Material samples
will collected following the completion of air sample
collection.
Several materials may be scheduled for
collection for laboratory emissions testing. Although
obtaining intact material for chamber testing is
preferred, it is recognized that it may not be possible to
collect and store large intact pieces of material. In these
cases, the largest material pieces consistent with
availability and adequate sample storage will be
collected.
5.5 Sample Transport, Storage, and Custody
All samples will be transported and stored for
analysis under conditions appropriate for minimizing
contamination by PCBs or losses of PCBs. In general, it
is anticipated that transport at refrigerator temperatures
(approximately 4° C) and storage at general freezer
temperatures (approximately -20° C) will be adequate.
Samples with potentially high levels of PCBs (such as
caulks or other primary PCB source materials) must be
stored and transported in a way that will prevent cross-
contamination of samples with low levels of PCBs (such
as air and wipe samples). Chain-of-custody procedures
will be followed to document the custodian and transfer
of samples from sample collection through sample
analysis.
13
-------
6. Sample Analysis
6.1 Background Information
Decisions regarding sample analysis methods
are complicated for PCBs, in general, and for this study,
in particular, given the large number of congeners, the
large range in congener vapor pressures, the variety of
Aroclors that were manufactured and used in different
products, the age and weathering of materials in older
schools, and the differences in toxicity and potential
mechanisms of action for different congeners.
Information from the extensive monitoring at New
Bedford High School, MA, may be relevant for selecting
analysis procedures.
• Aroclor 1254 appears to be the primary Aroclor found
in caulk, but other Aroclors were found in other bulk
materials.
• Aroclors 1242, 1254, and 1260 were found most
often in the many bulk material samples from New
Bedford High School; however, laboratory reports
state that most samples exhibited altered PCB
patterns for the Aroclors, suggesting potential
weathering over the long time span since
construction.
• Aroclors 1248 and 1260 were reported most often for
surface wipe samples collected from New Bedford
High School; however, laboratory reports indicated
that Aroclor 1248 was not actually present but was
used to represent the PCB congeners observed;
Aroclor 1260 exhibited an altered PCB pattern when
it was reported.
• Wipe samples collected in New Bedford High School
often were reported with no PCBs above the
detection limit.
• Air samples collected in New Bedford High School
were measured as homologs; most of the detectable
PCBs came in the di-, tri-, and tetrachlorobiphenyl
homolog series, suggesting greater emissions of
PCBs with higher vapor pressures and the relative
absence of PCBs with lower vapor pressures.
Assessing relative contributions of different
materials to the PCBs found in indoor air is difficult
because, while Aroclor 1242 has much higher
percentages of the di-, tri-, and tetrachlorobiphenyls
than do Aroclors 1254 and 1260, there could be large
amounts of caulk containing high levels of Aroclor 1254
in some buildings, so even minor fractions of the di-, tri-,
and tetrachlorobiphenyls in Aroclor 1254 could be an
important contributor to air levels. It also appears that
PCBs found on surfaces may not correspond well to
standard Aroclor patterns.
Under ideal circumstances, a high-resolution
gas chromatography/high-resolution mass spectrometry
method (HRGC/HRMS; such as EPA Method 1668B)
would be used to quantify as many of the individual
PCB congeners as possible in every sample collected
in this study. Congener-specific data would
facilitate assessments of the contribution of sources to
environmental levels and potentially could provide
information regarding specific congeners, based on
their relative toxicity. However, HRGC/HRMS methods
are very costly to perform.
The following were important considerations in
selecting analytical methods and PCB reporting for this
study.
• The resources available for analyzing more than
1000 samples
• The ability to assess relationships between sources
and PCB levels in environmental media
• The need to obtain measurable results for as many
samples as possible to perform source assessments
• The need to obtain total PCB measurements for
comparisons to standards or recommended levels
and to inform school decisionmaking
• A desire to obtain some congener-specific
information regarding the PCBs present in school
environments and for source evaluations
The proposed analysis approach for this study
is similar to the approach used in the New Bedford High
School measurement studies; analysis of Aroclors in
caulk, wipe, dust, and other material samples and the
analysis of chlorine number homologs in air samples.
Total PCBs will be reported for every analysis to
facilitate comparisons to other samples and to relevant
regulatory or action levels. To better evaluate these
measurements, a small number of samples from each
school also will be selected for additional congener-
specific analysis. More specific information is provided
below.
6.2 Analysis of Air Samples
All air samples will be analyzed for the PCB homolog
series and total PCBs shown in Table 7, using EPA
Method 8082A (GC with electron capture or electrolytic
conductivity detectors) if appropriate and applicable.
EPA Method 680 may be considered if Method 8082A
is not appropriate. Air samples are solvent extracted
and analyzed by GC with an appropriate detector. The
method must be sufficiently sensitive to measure PCBs
in ambient outdoor air and to ensure measurable results
at sufficiently low concentrations for a majority of the
samples to be collected indoors. Proposed quantifiable
limits are shown in Table 7. Final detection limits
selected for the analyses will be based on laboratory
capabilities and resource considerations. To better
understand how quantitation for homolog series
compares with the three-Aroclor quantitation
approaches as described in Method 10A, up to 10% of
the air sample extracts (those with the highest
14
-------
measurable results) will be quantified using both the
homolog and Aroclor approaches.
Table 7. Air Sample Analytes and
Target Detection Limits
Homolog Series
Monochlorobiphenyls
Dichlorobiphenyls
Trichlorobiphenyls
Tetrachlorobiphenyls
Pentachlorobiphenyls
Hexachlorobiphenyls
Heptachlorobiphenyls
Octachlorobiphenyls
Nonachlorobiphenyls
Decachlorobiphenyls
Total PCBs
Proposed
Quantifiable Limit3
0.5 ng/nf
0.5 ng/nf*
0.5 ng/mj
0.5 ng/nf*
1 ng/mj
1 ng/mj
1 ng/mj
2 ng/mj
2 ng/mj
2 ng/mj
—
Assuming a sample volume in the range of 5 to 7 m . Actual
detection limit will be based on the volume collected for each
sample and laboratory method and capability.
6.3 Analysis of Wipe Samples
All wipe samples will be analyzed for the Aroclors and
total PCBs listed in Table 8 using the EPASW-846
extraction method EPA Method 3540C (soxhlet
extraction) and analysis method EPA Method 8082A.
Obtaining measurable results (values above a detection
limit) for a large percentage of the samples will be
challenging, but it is an important goal of this study to
inform source assessment analyses and to provide
information for SHEDS model development. Proposed
quantifiable limits for the Aroclors are shown in Table 8.
Final detection limits selected for the analyses will be
based on laboratory capabilities and resource
considerations.
Table 8. Wipe Sample Analytes and Target
Detection Limits
Aroclor
1016
1221
1232
1242
1248
1254
1260
Total PCBs
Proposed Quantifiable Limit3
0.1 ug/IOOcm^
0.1 ug/100cm2
0.1 ug/100cm2
0.1 ug/100cm2
0.1 ug/100cm2
0.1 ug/100cm2
0.1 ug/100cm2
—
Actual detection limit will be based on the area collected and
analyzed for each sample and laboratory method and
capability.
6.4 Analysis of Dust, Soil, Caulk, and
Material Samples
All dust, caulk, and material samples will be
analyzed for the Aroclors and total PCBs listed in Table
9 using the EPA SW-846 extraction method EPA
Method 3540C (soxhlet extraction) with analysis
method EPA Method 8082A. Proposed quantifiable
limits for the Aroclors are shown in Table 9. Final
detection limits selected for the analyses will be based
on laboratory capabilities and resource
considerations. Quantifiable limits may vary among
materials and samples because of matrix effects and
interferences.
Table 9. Material Sample Analytes and
Target Detection Limits
Aroclor
1016
1221
1232
1242
1248
1254
1260
Total PCBs
Proposed Quantifiable Limit3
0.1 ug/g
0.1 ug/g
0.1 ug/g
0.1 ug/g
0.1 ug/g
0.1 ug/g
0.1 ug/g
—
Assuming a minimum sample size of 2 g for each sample.
Actual detection limit will be based on the amount of material
analyzed and laboratory method and capability.
6.5 Congener-Specific Analysis for a
Subset of Samples
It will not be possible to analyze all samples
collected in this study using congener-specific analysis
methods. However, congener-specific analysis can
provide important information on source relationships
with environmental levels, about aging and weathering
impact on sources, and, possibly, for limited toxicity
assessments by risk assessors. Therefore, a subset of
samples will be selected for high-resolution congener-
specific analysis, using EPA Method 1668B or
equivalent (HRGC/HRMS) following the original sample
analysis. This approach may depend on the ability to
utilize the original sample extract for the congener-
specific analysis. Otherwise, collection of duplicate
samples for this analysis will be needed at the time of
the original sample collection.
It is anticipated that no more than 40 congener-specific
analyses can be supported. The most preferable
approach will be to select an air, surface wipe
(desk/table), dust, and caulk sample extract from one
room in each of the nine schools for congener-specific
analysis. The sample set will be selected based on
results from the original Aroclor and air homolog series
analyses. The sample set with the highest values in the
air, wipe, and dust samples will be selected to improve
the ability to detect minor congeners. However, this
selection approach may need to be modified if it is not
possible to obtain all four samples from a single location
with PCB levels sufficiently high to support congener
analysis. A decision also may be made to include
another type of material extract in place of or in addition
to the dust sample. This approach may miss important
congeners in other locations and from other sources.
It is also important to recognize that a number of
congeners may still co-elute when using a single-
column high-resolution analysis method.
15
-------
7. Quality Assurance and Quality Control
The overall QA/QC objective is to generate
measurement data of high and known quality. A QA
project plan (QAPP) will be developed for this research
study and will be approved prior to the collection of
samples. The QAPP will define organizational
responsibilities, data quality goals, assessment
procedures, and review and auditing requirements. The
plan also will describe sample coding and chain of
custody requirements. Key QC elements will include
• defining organizational responsibilities;
• setting target goals for accuracy, precision,
completion rates, and detection limits;
• accuracy assessment using spiked field controls;
• precision assessment using duplicate samples and
replicate sample analyses;
• background assessment using field and laboratory
blank materials and analyses;
• detection limit goals and procedures; and
• analytical control plans and evaluations.
Some of the key QA elements will include
• accuracy assessment using independent analytical
standards,
• field and laboratory audits,
• data review procedures and data reviews, and
• developing and maintaining adequate sample
custody procedures.
Field sample collection and laboratory analyses
are likely to involve contract organizations. These
organizations will have certification or accreditation for
PCB sample collection and analysis or will be able to
demonstrate experience and performance.
A number of different types of QC samples will
be generated or collected as part of this research study.
Ideally, at least one set consisting of a field blank, a
spiked field control, and a duplicate sample would be
collected for each type of environmental sample at each
school. It will be possible to meet this goal for air
samples. For some sample media (surface wipes, soil,
and caulk), the duplicate sample will represent not only
the sample collection and analysis precision but also
will incorporate any nonhomogeneity in surface loading
or content. For some materials, such as caulk and other
building materials and for soil, it may not be feasible to
prepare appropriate blank or spiked control materials.
Laboratory QA and QC analyses will be incorporated
into the study as defined in the QAPP.
16
-------
8. Data Analysis and Modeling
8.1 Data Receipt and Compilation
The field study contractor will provide sample
collection and school information to EPA. The sample
analysis contractor will provide sample analysis, QC,
and QA data to EPA. Copies of sample custody, data
collection, and sample analysis report forms will be
provided by the contractors and maintained by the
investigator. Electronic data files will be organized and
compiled by the EPA data manager. All electronic data
files will be maintained on an access-controlled
networked server with daily backup.
8.2 Descriptive Statistics
Simple descriptive statistics will be generated
for each sample type collected at an individual school
and for each sample type collected across all schools in
the study. Statistics will include central measures (both
the mean and standard deviation and the geometric
means and geometric standard deviation) and a range
(minimum and maximum). If warranted with sufficient
measurable data, box plot summaries will be prepared
by sample type to show the distributions within and
across schools in this study. The distributional
parameters will apply only to the schools in this study
and will not be used for making inferences to other
schools on a regional or national basis. The results for
air, wipe, soil, and caulk measurements for the schools
in this study may be used for comparison to
recommended or regulated concentration levels.
8.3 Analysis of Within- and Between-School
Variability
Defining the variability in environmental levels
of PCBs in schools in this study will help guide
development of future school monitoring plans with
regard to the numbers and types of samples to collect.
It is also important to characterize within- and between-
school variability for subsequent model development.
In addition to generating media-specific standard
deviations, an assessment of within- and between-
school measurement results will be performed using an
appropriate statistical approach (e.g., covariance
parameter estimation).
8.4 Indoor Modeling: PCB Source Evaluations
and Relationships with Environmental Levels
Indoor models are used to describe and predict
relationships between various sources of pollutants and
their distributions and movements in indoor
environments. Indoor models are valuable for
understanding how different kinds of pollutant sources
and conditions in the building relate to the amount of
pollutant that may become available for contact and
exposure. Models are also important to help understand
which risk mitigation actions are likely to be most
effective in reducing the potential for exposures.
To the best of our knowledge, no peer-reviewed
indoor model has been published for PCBs in buildings
such as schools. Developing and evaluating such
models is complex because there are potentially
multiple sources of PCBs in indoor environments, both
vapor and particle phases must be considered with
regard to emissions and transfers, and PCBs are found
as mixtures of congeners often having a wide range in
vapor pressures. To construct an indoor model,
information would be needed regarding
• emission rates of PCBs from primary source
materials into the air;
• indoor air concentrations of PCBs;
• absorption rates of PCBs from the air onto or into
other materials and dust (secondary sources);
• emission rates from secondary sources;
• estimation of source surface areas;
• particle generation rates from primary and secondary
sources;
• levels of PCBs in dust;
• particle transfer or distribution rates within a room or
building;
• operational parameters and usage of HVAC systems;
• temperature and humidity data during sample
collection, as well as information on the relationships
between those factors and PCB emission and
transfer rates;
• air exchange rates between building zones and with
outdoor air; and
• removal rates of particles and dust from cleaning,
filtration, or other mechanisms.
Some of the necessary information will be
collected in this measurement study, particularly with
regard to concentrations of PCBs in different media,
and, potentially, the identification of primary and
secondary source materials. The laboratory chamber
material PCB emission and transfer experiments to be
performed by NRMRL will provide some additional
information. However, some of the information for
development of a full indoor PCB model will be difficult
or impossible to collect. For example, the scope of this
research plan does not include measurement of air
exchange rates. Such measurements are difficult to
perform accurately in large buildings, particularly those
with no or multiple forced air handling systems. Also, it
will not be feasible to assess PCB source particle
generation and distribution factors in-situ. Thus, the
development of an indoor model will rely on estimations
and assumptions for some factors.
Although recognized as important for indoor
pollutant modeling, accurate measurement of ventilation
conditions (indoor-outdoor air exchange and interzonal
17
-------
flows) in large old buildings (and for specific rooms
within those buildings) is very difficult. Among the
difficulties for implementation in this research study are
that
• many schools of this age do not have central HVAC
systems;
• the actual operating conditions of existing HVAC
systems may not be to design specifications, and
those specifications may not be known;
• it will not be feasible to sufficiently instrument a
building to collect flow and pressure data because of
cost and time constraints;
• it will not be possible to release tracer gases to
characterize ventilation;
• the assumption of a well-mixed single zone required
for simple ventilation assessments is unlikely to be
met for many schools; and
• ventilation is likely to be quite variable and highly
dependent on short-term changes in which doors and
windows are open.
The use of occupant-generated carbon dioxide (CO2)
has been used to provide information regarding
ventilation, either in a decay mode, when people leave
a room or the building, or in equilibrium mode, when the
CO2 respiration can be estimated from the number of
people present, the size of the people, and the room
volume. Background information on this approach and
the conditions that must be met to do this have been
described (Persily, 1997), and an ASTM method has
been developed ("Standard Guide for Using Indoor
Carbon Dioxide Concentrations to Evaluate Indoor Air
Quality and Ventilation," ASTM D6245-07 [2007]).
A central tenet of the procedure for producing accurate
ventilation rate information (whether using decay or
equilibrium procedures) is that the space to be
evaluated is a single zone (an indoor space or spaces
wherein the CO2 concentration is uniform, and that
exchanges air only with the outdoors). By definition, the
CO2 concentration must not differ by more than 10%
from the average across all locations. It will be difficult
to meet the single-zone and concentration uniformity
requirements in buildings with large spatial variability in
air flow and occupancy. For spaces that do not meet
the 10% criterion, it must be demonstrated that there is
no significant airflow from other building spaces
(hallways or other rooms) into the test space. There are
literature reports of CO2-based ventilation
measurements in schoolrooms (e.g., Godwin and
Batterman, 2007; Scheff et al., 2000). However, it is not
clear from these reports whether a single well-mixed
zone was present, and the relative contributions of
outdoor air and indoor air from other school spaces to
the ventilation of a single room is not clear. In a recent
report (Nazaroff et al., 2010), the authors describe
CO2-based ventilation measurements made in
individual classrooms. In this work, ventilation rates
were measured under several different unoccupied
sessions in which conditions were manipulated and,
then, while the building was occupied, using an
observer to monitor specific room and occupancy
conditions throughout the measurement. The authors
note that many classrooms are likely to be coupled by
means of airflow to other occupied spaces, that the
coupling is unlikely to be strong enough to produce a
larger well-mixed space, and that the coupling could be
too large to be ignored. Bartlett et al. (2004) describe
some of the issues associated with the transient nature
of occupancy and door openings on the ability to
accurately assess ventilation. There is a very large
difference in ventilation rates in response to changes in
air handler settings, opening classroom windows, and
with the classroom door open or closed. For modeling
purposes, it would be necessary to assess ventilation
under multiple conditions and determine or estimate the
amount of time each condition applies (again assuming
or determining that the ventilation is from outdoor air
and not indoor air from other spaces). This research
study will not include direct ventilation measurements
because accurate measurements will not be feasible
within the resources and likely availability of time within
the schools when occupied, and also because
ventilation and air PCB concentration measurements
would be required under different school conditions to
fully characterize relationships. A range of air exchange
rates will be included in the modeling effort to include
both the high and low ventilation rates, as suggested by
literature.
Analysis results for PCBs in materials will be
compiled for each school location and across each
school to examine relative differences in magnitude.
Correlation analyses will be performed between
environmental media. Correlation analyses will be
performed between material PCB levels and the levels
of total PCBs in air, surface wipes, and dust inside
rooms and across a school. Where warranted with
sufficient measurement data, potential source
contributions will be considered with regard to surface
areas and estimated or measured emission factors and
school conditions. Multivariate regression modeling may
be feasible, using data from multiple locations with
intensive source and environmental data. Alternatively,
cluster analysis may be applied for assessing multiple
potential source materials with regard to relationships to
environmental levels. Analyses will need to consider
effects of differences in vapor pressure and weathering
with regard to the distribution of different congeners or
sets of congeners. Some source assessment analyses
may be done best, or may only be achievable, with
congener-specific data, which will be limited in this
study.
8.5 SHEDS Model Development and
Application
Models are important for predicting the potential
exposures to PCBs in school environments. Results
from modeling can be used to estimate the overall
exposure and the relative routes of exposure. Such
information can inform the need for remediation actions
18
-------
by schools, as well as the most effective ways to reduce
exposures. The SHEDS-Multimedia Version 3 is ORD's
state-of-the-science probabilistic aggregate residential
exposure model (More et al., 2006; Zartarian et al.,
2000). SHEDS-Multimedia is a physically based,
probabilistic model that predicts, for user-specified
population cohorts, exposures incurred via inhaling
contaminated air, touching contaminated surface
residues, and ingesting residues from hand- or object-
to-mouth activities (U.S. EPA, 2009). Inputs include
chemical usage (optional); environmental concentration
data; and exposure factor information, including
activities and time spent in specific microenvironments.
Outputs include population and individual outputs for
various exposure or dose metrics and key factors and
pathways.
PCB measurement data and school operations
information will be used as inputs into the SHEDS-
Multimedia model to improve the model's ability to
predict exposure distributions for school-age children
under selected scenarios. Exposure determinants, such
as PCB concentrations in air, in dust, and on surfaces
students contact most frequently, are important model
inputs, and covariance of these determinants with PCB
sources and school operation factors, such as cleaning
methods and frequencies and HVAC conditions, will be
used to develop SHEDS factors and subsequent
exposure distributions. Data from various media from
the study will be fitted for the distribution and its
parameters. Results will be compared with those of
other studies. SHEDS can be used to estimate the main
contributions from various pathways, such as inhalation,
dermal, and indirect ingestion. Sensitivity analyses will
identify key inputs, and this analysis may help guide risk
management decisions for schools.
8.6 Assessment of Potential Exposures and
Exposure Routes
School environmental measurement data and
the models described above will be used to develop
estimates of potential exposures and exposure ranges.
The relative exposures by exposure route (inhalation,
dermal contact, and ingestion) will be examined.
Different conditions of PCB sources, the school
environment, and activities will be assessed for their
impact on exposures. This information can be used to
assess whether exposure levels exceed levels of
concern and to help determine what risk management
or remediation efforts will be most effective in reducing
exposures.
19
-------
9. Reporting Results
9.1 Reporting Analysis Results to Schools
Measurement data collected for each school
will be provided to the designated representative for
each school or school district. Measurement results will
be provided in tabular format. Results will be provided
immediately following QA review and approval by the
designated EPA QA manager and approval for release
by EPA management. Measurement data will not be
withheld pending completion of data analysis and
reporting by EPA scientists. Reports will note
measurements that exceed EPA recommended values.
9.2 EPA Reporting
An EPA research report will be prepared to
more completely describe the study, compile
measurement results across all schools, and report on
data analysis results. Recommendations may be
included regarding approaches and methods for future
school assessments and effective ways to manage and
reduce PCB exposures of children and school staff. It is
anticipated that one or more science journal articles will
be prepared based on this research effort. The
information and study results will be provided to the
EPA Administrator, regions, offices, and other
stakeholders to improve school PCB exposure
assessments, source characterization, and risk
management or reduction efforts.
20
-------
10. Proposed Study Timeline
The overall objective is to complete this
research study in a 12- to 18-mo time period. To meet
this objective, the following timeline is proposed.
- External peer review of study plan, March 2010
- School recruitment, winter/spring/summer 2010
- Initiate PCB screening in candidate schools,
spring 2010
- Final research plan, June 2010
- QA project plan prepared, June 2010
- Field study and sample analysis contracts in
place, summer 2010
- Initiate sample collection in study schools,
summer 2010
- Sample analysis, summer/fall 2010
- Data receipt and QA review, fall 2010
- Provide measurement data to schools, fall/early
winter 2010
- Data analysis and modeling, winter 2010/2011
- Draft report, February 2011
- Final report, March 2011
The timeline and schedule depend on the ability
to identify and reach agreement with schools to
participate in the research.
21
-------
11. References
ASTM. 2006. ASTM Method D6661-01 -Standard Practice
for Field Collection of Organic Compounds from
Surfaces Using Wipe Sampling. ASTM International.
ASTM. 2007. ASTM Method D6245-07 - Standard Guide for
Using Indoor Carbon Dioxide Concentrations to Evaluate
Indoor Air Quality and Ventilation. ASTM International.
ATSDR. 2000. Toxicological Profile for Polychlorinated
Biphenyls (PCBs). U.S. Department of Health and
Human Services, Public Health Service, Agency for
Toxic Substances and Disease Registry. November,
2000.
Bartlett KH, Marinez M, and Bert J. 2004. Modeling of
Occupant-Generated CO2 Dynamics in Naturally
Ventilated Classrooms, Journal of Occupational and
Environmental Hygiene, 1:139-148.
Coghlan KM, Chang MP, Jessup DS, Fragala MA, McCrillis K,
and LockhartTM. 2002. Characterization of
Polychlorinated Biphenyls in Building Materials and
Exposures in the Indoor Environment. Proceedings:
Indoor Air 2002. The 9th International Conference on
Indoor Air Quality and Climate, Monterey, CA, USA,
June 30-July 5, 2002.
Delle Site A. 1997. The Vapor Pressure of Environmentally
Significant Chemicals: A Review of Methods and Data at
Ambient Temperature. J Phys Chem Ref Data,
26(1): 157-193.
Erickson MD. 1997. Analytical Chemistry of PCBs, Second
Edition. CRC Press, Boca Raton, FL, USA.
Falconer RL, and Bidleman TF. 1994. Vapor Pressures and
Predicted Particle/Gas Distributions of Polychlorinated
Biphenyl Congeners as Functions of Temperature and
Ortho-Chlorine Substitution. Atmos Environ, 28(3):547-
554.
Godwin C, and Batterman S. 2007. Indoor Air Quality in
Michigan Schools. Indoor Air, 17:109-121.
Harrad S, Ibarra C, Robson M, Melymuk L, Zhang X,
Diamond M, and Douwes J. 2009. Polychlorinated
Biphenyls in Domestic Dust from Canada, New Zealand,
United Kingdom and United States: Implications for
Human Exposure. Chemosphere, 76:232-238.
Hazrati S, and Harrad S. 2006. Causes and Variability in
Concentrations of Polychlorinated Biphenyls and
Polybrominated Diphenyl Ethers in Indoor Air. Environ
SciTechnol, 40:7584-7589.
Herrick RF, McClean MD, Meeker JD, Baxter LK, and
Weymouht GA. 2004. An Unrecognized Source of PCS
Contamination in Schools and Other Buildings. Environ
Health Perspect, 112(10):1051-1053.
Herrick RF, Lefkowitz DJ, and Weymouth GA. 2007. Soil
Contamination from PCB-Containing Buildings. Environ
Health Perspect, 115(2): 173-175.
Holmes DA, Harrison BK, and Dolfing H. 1993. Estimation of
Gibbs Free Energies of Formation for Polychlorinated
Biphenyls. Environ Sci Technol, 27:725-731.
Hore P, Zartarian V, Xue J, Ozkaynak H, Wang SW, Yang
YC, Chu PL, Sheldon L, Robson M, Needham L, Barr D,
Freeman N, Georgopoulos P, Lioy PJ. 2006. Children's
Residential Exposure to Chlorpyrifos: Application of
CPPAES Field Measurements of Chlorpyrifos and TCPy
within MENTOR/SHEDS-Pesticides Model. Sci Total
Environ, 366:525-537.
Kohler M, Tremp J, Zennegg M, Seller C, Minder-Kohler S,
Beck M, Lienemann P, Wegmann L, and Schmid P.
2005. Joint Sealants: An Overlooked Diffuse Source of
Polychlorinated Biphenyls in Buildings. Environ Sci
Technol, 39:1967-1973.
Kohler M, Zennegg, and Waeber. 2002. Coplanar
Polychlorinated Bipheynyls (PCBs) in Indoor Air. Environ
Sci Technol, 36:4735-4740.
McCarthy JF. 2009. Managing the Risks of PCBs in Building
Materials. Presentation to the Federal Interagency
Committee on Indoor Air Quality. October 21, 2009.
NazaroffWW, BhangarS., Mullen NA, Hering SV, and
Kreisberg NM. 2010. Ultrafine Particle Concentrations in
Schoolrooms and Homes, Final Report Contract No 05-
305. California Air Resources Board, Sacramento, CA.
PersilyAK. 1997. Evaluating Building IAQ and Ventilation with
Indoor Carbon Dioxide. ASHRAE Transactions,
103(2): 193-204.
Rudel RA, Seryak LM, and Brody JG. 2008. PCB-Containing
Wood Floor Finish is a Likely Source of Elevated PCBs
in Resident's Blood, Household Air and Dust: A Case
Study of Exposure. Environ Health, 7(2).
Scheff PA, Paulius VK, Huang SW, and Conroy LM. 2000.
Indoor Air Quality in a Middle School, Part I: Use of CO2
as a Tracer for Effective Ventilation. Applied
Occupational and Environmental Hygiene, 15(11):824-
834.
Sullivan, D. 2008. Polychlorinated Biphenyls (PCBs) and
Indoor Air: Source Investigation and Remdial Approach
for a Public School Building in New Bedford,
Massachusetts, USA. 28th International Symposium on
Halogenated Organic Pollutants (POPs), Birmingham,
England, UK, August 17-22, 2008.
TRC. 2006. New Bedford High School Indoor Polychlorinated
Biphenyls Sampling. Report of Findings. Prepared by the
TRC Environmental Coroporation. Prepared for the City
of New Bedford, MA. November 17, 2006.
TRC. 2008. New Bedford High School Indoor Polychlorinated
Biphenyls Source/Sink Sampling Program. Report of
Findings. Prepared by the TRC Environmental
Coroporation. Prepared for the City of New Bedford, MA.
October 2008.
TRC. 2009. New Bedford High School Indoor Polychlorinated
Biphenyls Quasi-Random Bulk Material Sampling
Program. Report of Findings Addendum. Prepared by
the TRC Environmental Coroporation. Prepared for the
City of New Bedford, MA. May 2009.
22
-------
U.S. EPA. 1985. Method 680 - Determination of Pesticides
and PCBs in Water and Soil/Sediment by Gas
Chromatography/Mass Spectrometry. November 1985.
U.S. EPA. 1991. Wipe Sampling and Double Wash/Rinse
Cleanup as Recommended by the Environmental
Protection Agency PCS Spill Cleanup Policy. Revised
April 1991.
U.S. EPA. 1996. Method 3540C - Soxhlet Extraction. Test
Methods for Evaluating Solid Waste, Physical/ Chemical
Methods (SW-846). December 1996.
U.S. EPA. 1999. Compendium Method TO-10A-
Determination of Pesticides and Polychlorinated
Biphenyls in Ambient Air Using Low Volume
Polyurethane Foam (PUF) Sampling Followed by Gas
Chromatographic/Multi-Detector Dections (GC/MD).
Compendium of Methods for the Determination of Toxic
Compounds in Ambient Air. January 1999.
U.S. EPA. 2007a. Method 8082A-Polychlorinated Biphenyls
(PCBs) by Gas Chromatography. Test Methods for
Evaluating Solid Waste, Physical/Chemical Methods
(SW-846). February 2007.
U.S. EPA. 2007b. Method 3550C - Ultrasonic Extraction. Test
Methods for Evaluating Solid Waste, Physical/Chemical
Methods (SW-846). February2007.
U.S. EPA. 2007c. Method 3546 - Microwave Extraction. Test
Methods for Evaluating Solid Waste, Physical/Chemical
Methods (SW-846). February 2007.
U.S. EPA. 2008a. Standard Operating Procedure for
Sampling Porous Surfaces for Polychlorinated Biphenyls
(PCBs). EIA_Poroussampling 1, Revision 3, July 2008.
U.S. EPA. 2008b. Method 1668B - Chlorinated Biphenyl
Congeners in Water, Soil, Sediments, Biosolids, and
Tissue by HRGC/HRMS. November 2008.
U.S. EPA. 2009. SHEDS-Multimedia: SHEDS-Multimedia
version 3: ORD/NERL's Model to Estimate Aggregate
and Cumulative Exposures to Chemicals.
http://www.epa.gov/heasd/products/sheds multimedia/s
heds mm.html.
U.S. EPA. 2010. Research Operating Procedure for
Screening Collection of Caulk and Window Glazing From
Buildings for PCS Analysis. EMAB-AP-PSS-01. Office of
Research and Development, National Exposure
Research Laboratory, Research Triangle Park, NC.
March 2010.
Zartarian VG, Ozkaynak H, Burke JM, Zufall MJ, Rigas ML,
and Furtaw Jr. EJ. 2000. A Modeling Framework For
Estimating Children's Residential Exposure and Dose to
Chlorpyrifos Via Dermal Residue Contact and Non-
Dietary Ingestion. Environ Health Perspect, 108(6):505-
514.
23
-------
-------
APPENDIX
Study Summary Information for Schools
25
-------
-------
U.S. EPA Research Study to Investigate PCBs in School Buildings
Need for the Research
Caulk containing polychlorinated biphenyls (PCBs) was used in some buildings, including schools, in the 1950s
through the 1970s. Other sources of PCBs may also be present in some older school buildings. To better
understand the potential problem and improve risk management decisions by schools, EPA is conducting a study
on mitigation methods to reduce exposure to PCBs in caulk, and is planning to conduct an exposure monitoring
study in schools on PCB-contaminated caulk and other potential sources of PCBs. We are soliciting your help in
identifying schools that could participate in this monitoring study.
Research Goals
EPA research on PCBs in schools is designed to identify and evaluate potential sources of PCBs to better
understand exposures to children, teachers and other school workers; and to improve risk management decisions.
The research will:
• characterize potential sources of PCBs in schools (caulk, coatings, light ballasts, etc.);
• investigate the relationship of these sources to PCB concentrations in air, dust, and soil;
• improve PCB exposure assessment models; and
• develop data that can be used to help identify best practices for risk management.
Study Design
Information below describes the general approach for collection of data to address the research goals. The final
design of the study will be prepared in collaboration with staff in the participating EPA Regions, OPPT and other
stakeholders. Measurement plans for each school will be developed in collaboration with responsible parties at the
school. The study will not involve any measurements of PCBs in the children or school staff, nor will there be
changes needed to school activities.
Study Dates
It is anticipated that the sample collection in schools will be able to commence in summer of 2010, contingent on
the successful recruitment of schools.
Number of Schools
Up to nine schools could participate in the research study.
Type of Schools
Primary or secondary schools constructed prior to 1975 and that are in use or in good condition are of most
interest. College buildings may be considered. Schools that have not received extensive renovations are preferred.
Screening Caulk Sample Collection
If it is not known whether a school has PCB-containing caulk, an initial screening walk-through and collection of
caulk samples would be scheduled, requiring 2- 4 hours on one day or evening when students are not present.
Time Needed in the School
If a school participates in the full monitoring study, we expect that EPA contractors will need to have access to the
school on three days, including an initial scouting day, followed some days or weeks later by sample collection on
two consecutive days.
Samples to Be Collected
• Air samples collected over a 24-hour period in up to nine areas in the schools (including classrooms and
other areas) and at one exterior location during the same period.
• Wipe samples collected from surfaces in each of up to nine areas in the school (including classrooms and
areas such gyms, cafeterias, libraries, or halls). Samples would be collected from contact surfaces such as
desks and tables, and others will be from building surfaces such as floors and windowsills.
• Dust bulk dust samples collected in the HVAC system (if present) and other locations.
• Soil samples from up to three locations outside the school building.
• Small samples of caulk around windows/doors, exterior joints, and interior joints in each of the nine areas
sampled.
27
-------
• Small samples of other potential primary and secondary PCB containing materials. These may include
wallboard, masonry, paint, adhesives, tile or other materials from locations in each school.
• A few larger samples of caulk, dust, and materials for laboratory chamber testing.
Other Information Needed
No personal information will be collected from or about any student, teacher or staff member. Some additional
information about the school is needed to help interpret the measurement results and improve models used to
estimate PCB exposures. In general, the types of information include:
• School building history (year built, renovation dates, etc.);
• School dimensions, layout, and materials;
• School operation (start/end dates and times, other uses);
• Enrollment (total students, class sizes, number in each grade, classes per day);
• HVAC information (type of HVAC, areas served, use during year and during sample collection period); and,
• Cleaning information (cleaning methods and frequency of cleaning for each method).
Impact on Students and Staff
Efforts will be made to minimize any impact or disruptions because of the research.
• The scouting can be performed on a day when school is not in session.
• It is preferable, but not required, that air samples be collected while school is in session because activities
and HVAC conditions can affect levels of dust in the air.
• Air sample collection devices will require an electrical outlet. Pumps will be small and efforts will be made
to make the pumps as quiet as possible.
• Other types of samples would be collected after school hours to minimize disruptions.
Some staff time will be needed to facilitate access to and within the school building and to provide information
about the school.
28
-------
-------
xvEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
S300
Recycled/Recyclable Printed on paper inat contains a minimum of
50% postconsumer fiber cantenl processed chlorine free
------- |