United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-451/R-92-002
September 1992
Air/Superfund
AIR/SUPERFUND
NATIONAL TECHNICAL
GUIDANCE STUDY SERIES
Assessing Potential
Indoor Air Impacts
for Superfund Sites
-------
ASSESSING POTENTIAL INDOOR
AIR IMPACTS FOR SUPERFUND SITES
Contract No. 68D00124
Work Assignment 1-76
Prepared for:
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
September 1992
Submitted by:
Pacific Environmental Services, Inc.
560 Herndon Parkway, Suite 200
Herndon, Virginia 22070
-------
DISCLAIMER
This report was prepared for the U.S. Environmental Protection
Agency by Pacific Environmental Services, Inc., Herndon, VA, under
Contract No. 68-00-0124, Work Assignment No. 1-76. The contents are
reproduced herein as received from the contractor. The mention of
product names or trademarks are not intended as endorsements of the
products or their use. The opinions, findings, and conclusions ex-
pressed are those of the authors and do not necessarily reflect those of
the U.S. Environmental Protection Agency.
n
-------
TABLE OF CONTENTS
DISCLAIMER ii
TABLES iv
FIGURES iv
1.0 INTRODUCTION 1-1
1.1 BACKGROUND 1-1
1.2 PURPOSE 1-3
1.3 SCOPE 1-3
2.0 INDOOR AIR IMPACT ASSESSMENT PROCEDURES f -. .. 2-1
2.1 GENERAL CONCEPTS 2-1
2.2 SCREENING LEVEL MODELING 2-3
2.2.1 SIMPLE CONSERVATIVE MODELING 2-3
2.2.2 SCREENING LEVEL 2 MODELING 2-5
2.3 ON-SITE EXTERIOR MONITORING 2-7
2.3.1 DESIGNING THE MONITORING PLAN 2-8
2.3.2 BUILDING INSPECTION 2-11
2.3.3 ESTIMATION OF INDOOR AIR CONCENTRATIONS 2-12
2.4 INDOOR MONITORING 2-13
2.4.1 SCREENING LEVEL MONITORING 2-14
2.4.2 REFINED MONITORING FOR INDOOR AIR
IMPACTS OF SITES 2-16
2.4.3 MAKING LONG-TERM ESTIMATES FROM
SHORT-TERM MONITORING DATA ..... 2-26
3.0 HAZARD ANALYSIS 3-1
3.1 INCREMENTAL INDOOR AIR RISKS 3-1
3.2 RISK COMMUNICATION 3-3
3.3 OVERVIEW OF RISK ASSESSMENT GUIDANCE 3-7
3.3.1 EXPOSURE 3-8
3.3.2 TOXICITY ASSESSMENT 3-9
3.3.3 SOURCES OF TOXICITY VALUES 3-10
3.4.4 QUANTIFYING RISKS 3-11
3.3.5 DATA LIMITATIONS AND INTERPRETATIONS 3-13
APPENDICES
A - PREDICTIVE SCREENING TECHNIQUES
B - MONITORING METHODS
C - CASE STUDIES
in
-------
LIST OF TABLES
Table Page
2-1 Hypothetical Monitoring Data 2-25
2-2 Hypothetical Data Corrected for Ambient Air 2-25
2-3 Hypothetical Data Corrected for Soil Gas Ratio 2-25
LIST OF FIGURES
Figure
3-1 Illustrative Decision Tree ....
iv
-------
1.0 INTRODUCTION
1.1 BACKGROUND
The Comprehensive Environmental Response, Compensation, and
Liability Act of 1980 (CERCLA or "Superfund") and its reauthorization in
the Superfund Amendments and Reauthorization Act (SARA) of 1986 estab-
lishes a national program for responding to releases of hazardous
substances into the environment. The overarching mandate of the
Superfund program is to protect human health and the environment from
current and potential threats posed by these releases.
Occupants of buildings proximate to a site where such releases
have occurred may potentially be exposed to the released chemicals from
their transport into the indoor environment. Typical pathways for entry
of site contaminants into a building include transport through the
ambient air, the use of contaminated groundwater, seeps of non-aqueous
liquids and groundwater through building exteriors, and intrusion of
contaminated soil gases. In some cases, the resultant indoor air
concentrations of site related chemicals may be high enough to warrant
immediate corrective action, while in others they may be inconsequen-
tial.
Responses taken to assess whether or not an indoor air impact has
occurred and the significance of that impact have varied from predictive
modeling to on-site monitoring of the indoor air quality. Both the
modeling and monitoring techniques used have been quite varied.
Predictive modeling techniques used have ranged from quite simple
screening tools to complex approaches. Although guidance exists for
selection of air emission and dispersion models and information docu-
ments are available for indoor air modeling, no established guidance
exists to assist with proper selection of predictive models for assess-
ing potential indoor air impacts. Likewise, monitoring has varied from
collecting grab samples to complex multi-building comparative studies.
Although guidance exists for selection and use of monitoring techniques
1-1
-------
for ambient air and soil gases and information documents are available
for indoor air monitoring, no established guidance exists to assist with
proper selection of methods to assess the potential site related indoor
air impact.
The indoor environment, even for buildings not impacted by outdoor
pollutants, is recognized to be highly polluted. Many indoor studies
have shown that the normal indoor environment contains a wide variety of
pollutants at higher concentrations than found outdoors. The sources of
these pollutants are believed to be primarily consumer products. Indoor
pollution can occur from cleaners and waxes, paints, pesticides,
adhesives, cosmetic and personal care products, hobby supplies, furnish-
ings and clothing, building materials, and heating and cooling systems,
among others.
Many of the chemicals emitted by the materials are the same as
chemicals released from Superfund and hazardous waste sites which
severely complicates any attempt to determine the amount of any indoor
air pollutant that is directly attributable to external sources.
Further complicating the situation is the fact that buildings interact
with the outdoor environment. The rate at which ambient air infiltrates
a building is a function of several factors including wind speed,
indoor-outdoor temperature differences, height of the building, and
leaks in the building shell, as well as vented equipment, such as
bathroom and kitchen fans or oil and gas furnaces and fireplaces. Air
pressures below ambient can develop in the lower stories of a building.
Although these pressure differences are small, typically between 1 and
10 pascals, they can result in the building literally sucking in soil
gases through cracks and openings in below grade walls.
Site contaminant location in relation to building and contaminant
source strength may also affect the rate of chemical entry into the
building. Contaminant may arrive in the ambient air, in groundwater
leaking into the building, from gases migrating laterally from the site,
from nearby contaminated soils, and from contaminated groundwater plumes
passing near or beneath the building. The amount of chemical arriving
at the building then becomes a complex function of meteorological
conditions and soil properties.
1-2
-------
Determining the impact a nearby contaminated site may be having on
the indoor air quality of a specific building can, thus, be seen to be a
difficult task.
1.2 PURPOSE
This manual for assessing indoor air impacts was developed to be
used for buildings proximate to Superfund sites. The procedures and
methods may also be applied to other assessments of hazardous wastes and
hazardous materials. The purpose is to provide information on methods
and techniques that can be used to assess the potential or actual impact
a Superfund site may have on the indoor air quality of buildings
proximate to the site in order that the risk to the occupants from this
source may be estimated. The manual is designed to provide the tools
needed to assist in the identification and resolution of indoor air
quality concerns at Superfund site.
1.3 SCOPE
This manual focuses on contaminant entry into the indoor environ-
ment via subsurface gases and ambient air. It does not address contami-
nant entry via direct intrusion of contaminated groundwater or indoor
uses of contaminated groundwater such as showering. It also does not
address radiologically contaminated sites, which, although an important
consideration for some sites, is believed to be adequately covered by
existing guidance for radon.
It provides summary level information on predictive modeling, on-
site monitoring, and risk characterization. It contains information
from indoor air studies that have been conducted using both modeling and
monitoring approaches. This information was utilized to provide general
information on the selection and use of models, monitoring methods, and
assumptions useful in assessing site impacts on indoor air quality at
levels ranging from scoping to detailed monitoring.
Individuals having different levels of scientific training and
experience are likely to use this manual. Because assumptions and
judgement are required in many parts of the assessment, the individuals
conducting the evaluations are key elements in the process. The manual
1-3
-------
is not intended to instruct non-technical persons how to perform
technical evaluations, nor to allow professionals trained in one
discipline to perform work in another.
The manual cannot address all circumstances. Users must exercise
technical and management judgement, and should consult with appropriate
regional and headquarters staff when encountering unusual or particular-
ly complex technical issues. The procedures described should be viewed
as flexible and can and should be tailored to specific circumstances and
information needs of individual sites, and not as a rigid approach that
must be conducted at every site.
1-4
-------
2.0 INDOOR AIR IMPACT ASSESSMENT PROCEDURES
This Chapter presents general procedures for assessing potential
Superfund site impact on indoor air in structures proximate to the site.
It is recognized that the contaminant source type, location, proximity
to the structure(s) and transport pathways, as well as the geology,
hydrogeology, structural characteristics, and use of the structure(s)
differ widely for all potential situation that may be encountered.
Thus, this discussion, of necessity, is generic. Although the proce-
dures should be generally applicable, they do not constitute a "how-to"
or "cookbook" set of instructions. It is expected that proper applica-
tion will require substantial expertise and professional .judgement.
2.1 GENERAL CONCEPTS
The objective is to assess the potential exposure, and resultant
risk, for occupants of buildings proximate to a Superfund sites to site
contaminants via the indoor air pathway. To accomplish this objective,
various techniques ranging from simple screening models to complex
monitoring may be used. The confidence level in the results increases
with increasing sophistication.
In recognition that conducting monitoring for buildings of a
sophistication level sufficient to segregate site impacts from other
potential sources can be expensive, the approach outlined is based on a
graduated approach to scope the possible extent and magnitude of the
possible impact before on-site detailed monitoring is executed. The
approach follows these general steps:
1. Conduct simple conservative modeling to estimate the poten-
tial magnitude of the exposure.
2. If step 1 suggests a potential problem, conduct more sophis-
ticated modeling to provide more realistic estimates of the
potential exposure.
2-1
-------
3. If Step 2 indicates a significant exposure potential
exists, conduct exterior monitoring at the site to confirm
model predictions and make better estimates of indoor expo-
sures. This step may be preceded by screening level indoor
monitoring.
4. When deemed necessary, conduct monitoring at the building
site at a level adequate to provide the best estimate of
site impact on the indoor air quality and collect data to
estimate exposure and resultant risk over longer time peri-
ods.
Step 1, simple conservative modeling, is suggested in virtually
all cases because it provides overall perspective and can reduce the
potential for conducting unnecessary monitoring and reduce the possibil-
ity of focusing efforts based on complaints which may or may not
represent the locations of most concern. Step 2, sophisticated model-
ing, is suggested as a follow-up to Step 1 because simple models
typically over predict exposures. Executing Step 2 reduces the likel-
ihood of conducting monitoring when, in fact, no exposure of concern may
be occurring.
Step 3, exterior monitoring at a specific building location, is
suggested as the primary on-site technique for assessing the site
related indoor air impacts on specific structures. These techniques
allow reasonable estimates of indoor air concentration without the
necessity of dealing with building specific air flow patterns and indoor
pollutant sources. If it is considered desirable to first conduct
indoor monitoring at the screening level, it should be recognized the
indoor data are useful for order of magnitude estimates only. This
monitoring cannot, except in unusual cases, provide pollutant concen-
trations due to site impacts because of the variability in types and
strengths of indoor sources.
Step 4, detailed building monitoring, is suggested only when it is
necessary to obtain actual indoor air site impact data for a specific
building. It is expected that this monitoring will typically only
provide marginal improvements in the exposure estimates obtainable by
2-2
-------
following Step 3. Indoor air monitoring itself is complex and requires
a high level of professional expertise to obtain reasonable results.
Except in unusual cases, site impacts cannot be estimated unless
exterior monitoring is executed simultaneously.
The above general procedures are applicable principally to indoor
air impacts associated with transport of organics (volatile and semi-
volatile) by soil gases and ambient air. Impacts of particulates
transported to the building may, however, be best resolved with direct
indoor monitoring. This is suggested because, once inside the struc-
ture, dusts tend to settle and then be resuspended by traffic and
cleaning activities. Potential risks from dusts may then be a combina-
tion of inhalation, ingestion, and dermal contact paths.
2.2 SCREENING LEVEL MODELING
The use of screening models corresponds to Step 1 and 2 above.
They are used to obtain order-of-magnitude estimates of the potential
indoor air concentrations. The simple modeling approach described in
Section 2.2.1 should be viewed only as a scoping tool. The modeling
tools discussed in Section 2.2.2 can be, and have been, used to estimate
the risks due to the indoor air pathway in remedial investigations.
2.2.1 Simple Conservative Modeling
Modeling may be used to examine the possibility that site contami-
nants may be transported to buildings in sufficient concentration and
rate to significantly impact the indoor air. This is a particularly
useful step when more than one structure or type of structure may be
impacted due to a fairly large contaminated zone, such as a groundwater
plume or dispersion in ambient air. Information may already be avail-
able to determine whether or not a particular transport mechanism is of
concern and no modeling would be needed for these mechanisms.
A minimal amount of site specific information is needed to run the
simple models. For emissions from contaminated groundwater, the
chemical concentration in the groundwater and depth to the top of the
water table will suffice. Similar emission related information is
needed for contaminants transported through ambient air. Because these
2-3
-------
modeling results should be only used to determine whether or not more
sophisticated modeling is needed, missing data may be supplied using
realistically conservative estimates.
It is recommended that the air quality models SCREEN and/or
TSCREEN be used for estimating the ambient air concentrations at the
building(s) being considered. These models will calculate the short-
term maximum concentrations at various distances from the source.
Because buildings typically have air exchanges rates of 0.5 to 1 per
hour, it is recommended that averaging periods less than 3 hours not be
used. The use of the 24-hour maximum concentration should provide a
conservative estimate even for developmental toxicants. Professional
judgement should be used when estimating possible impacts for very
short-term releases that could occur from sudden releases or remedial
actions.
It is recommended that the simple Fickian diffusion model (Appen-
dix A, Section A.1.1) be used to estimate the contaminant flux from
contaminated groundwater and contaminated soil. The important variables
are chemical concentration and distance from the building floor to the
contamination because diffusivities in air and Henry's Law constants
each vary by less than an order of magnitude for VOCs. Assume 100
percent of the soil gas rising under the building footprint enters the
building, for all building types, and the building air exchange rate is
0.5 per hour.
A first cut approximation can be made without even running the
model by assuming diffusivities in air = 10"5, Henry's Law constants =
10"1 (dimensionless), air filled and total porosities both equal 0.4,
building air changes per hour = 0.5, and 100 percent of the soil gas
rising under the footprint of the building enters. Under these assump-
tions,
2-4
-------
Where
IAC = Indoor air concentrations, M9/m3
CL = Concentration in groundwater, /jg/L
L = Depth to water table, M.
Note that this is the upperbound concentration for the most volatile
component in very porous completely dry soil and should be expected to
yield excessively high results for less volatile compounds and for
tightly packed or wet soils. The equation is of little utility when the
pollutant source is within about 2 meters of the building or when direct
intrusion of contaminated groundwater is occurring.
For contaminated soil, the relationship given in Equation 11 of
Appendix A, Section A.1.3 may be used to estimate the soil gas concen-
tration. For contaminated soil immediately adjacent (ie, within 1
meter) to the building, the relationships of Appendix A Section A.3.1
are recommended to calculate the indoor air concentration.
Note in particular that it is not recommended to estimate a
reduction in soil gas flux entering the building based on the ratio of
assumed area of floor cracks to floor area. Based on current under-
standing, this is a questionable assumption (See Appendix A, Section
A.3).
2.2.2 Screening Level 2 Modeling
Indoor air concentrations calculated using the more sophisticated
modeling techniques provide the best estimate that can currently be made
without sampling at the structure(s) in question. However, use of these
models requires both professional modeling expertise and site-specific
information. The better the site-specific information, the more likely
are the model predictions to accurately represent the concentrations in
the structure. Of all the terms in the various equations of Appendix A,
the effective diffusivity (DTeff) and soil permeabilities (KJ are likely
to have the most uncertainty. This is because not only are soil
properties quite variable vertically and horizontally over regions as
small as a single house, soil moisture contents (and thus air filled
porosity) can vary widely. It is not unusual for Kv to vary by 3 orders
2-5
-------
of magnitude across a site the size of a typical residential lot. While
the use of "average" or "typical" values for these parameters may yield
fairly accurate results, it should not be expected that any current
model will exactly represent the real world.
As stated in Appendix A, Section A.2, air modeling should be
conducted in accordance with the guidance referenced. For emissions
from groundwater and contaminated soils, any of the models of Appendix A
Sections A.I and A.3.3 may be used provided that appropriate consider-
ation is given to the contaminant concentration in the soil and site
conditions affecting the effective diffusivity.
Specifically, the model of Section A.1.1 may be used for groundwa-
ter when the refinement of Section A.1.2 is made. The model may also be
used for contaminated soil when corrected for soil adsorption (Section
A.1.3). Both uses of the model should be corrected using the consider-
ations of Section A.1.6 when neat liquids (non-aqueous phase), or soil
concentrations exceeding the level that calculated soil gas concentra-
tions are greater than compound vapor pressures are present.
The model presented in section A.1.4 has a wide range of applica-
bility for contaminated soils and groundwater. It is the preferred
model for calculating soil gas flux for most applications. However, the
model is not suitable for situations in which non-aqueous phase liquids
or high chemical concentrations, as explained above, are present. For
these applications, the model presented in Section A.1.6 is most
appropriate.
All of the Models of Appendix A, Section A.I, estimate only the
contaminant flux (mass of contaminant impacting on the structure per
unit area of below surface walls per unit time). They do not, there-
fore, directly yield an estimate of indoor air concentration. To obtain
the indoor air concentration, divide the calculated flux by the air
exfiltration rate for the building (see Section A.3.2). It is explicit-
ly recommended that the calculation be based on 100 percent of the flux
entering the building (ie., no correction made for building resistance
to intrusion, such as the fraction of floor area cracked).
The Johnson and Ettinger models presented in Appendix A, Section
A.3.3, appear to provide a good representation of both contaminant
2-6
-------
transport and the effects of building characteristics on soil gas entry
into a building. Properly coupled with appropriate calculations of the
contaminant source gas phase concentration and equations for building
air exchange rates, the models show promise of becoming the best
available short of detailed numerical simulations. The use of this
model is considered appropriate for indoor air impact assessments.
2.3 ON-SITE EXTERIOR MONITORING
In this section the value of, ajid approach to, the use of monitor-
ing in the immediate vicinity of a potentially impacted building and how
this data may be combined with models to yield an estimate of the indoor
impacts from Superfund chemicals is discussed. It is expected that this
on-site monitoring would be conducted
as a consequence of the modeling
described above indicating the potential for significant impact. It is
recognized that on-site monitoring may also be indicated by other
considerations.
The accuracy of the values calculated from modeling is a function
of the quality of input data. Generally, several assumptions will have
been made in these calculations that affect the accuracy. The principal
assumptions are:
1. Soil gas concentration at the soil-building interface is
zero.
2. The effective vertical diffusivity is accurately represented
by the data input for total and air filled porosity.
3. Effective vertical diffusivities are the same throughout the
entire soil column extending an infinite distance from the
building.
4. Soil gas concentrations at the contaminant source are accu-
rately represented by the equilibrium relationships given in
Appendix A for chemicals dissolved in groundwater or ad-
sorbed to soil.
5. No adsorption or biological reaction occurs as the gas dif-
fuses to the building.
None of these are likely to be strictly true. For example, if
assumption 1 were strictly true,- no chemicals would enter the building.
2-7
-------
Exterior monitoring provides the data needed for the best estimate of
chemical concentrations entering the building. These data are necessary
if a reasonable estimate of Superfund site chemicals impacting the
building are to be made. Note that because the typical indoor air
contains such a large number of chemicals that are also found at
Superfund sites, simply knowing the indoor air concentrations usually
only establishes the upper bound for site impact but does not establish
that the chemicals are actually from the Superfund site. Reasonable
exceptions to th'is might be for chemicals not found in non-impacted
buildings (eg., vinyl chloride) or chemical concentration exceeding the
upper end of the typical range.
Since assumption 3 is typically not true, especially in the region
near the building, it should be expected that soil gas concentrations,
and possibly compositions, will vary from location to location around
the building. Thus, the monitoring plan must be designed to take this
into account in so far as practical.
The building itself exerts an influence on the surrounding soil.
Because of the slight negative pressures (ie., 1 to 10 Pa) developed due
primarily to thermal gradients and wind effects, the building may draw
in gases from the surface as well as from several meters laterally from
the building walls. Thus, in permeable soils, soil gas concentrations
adjacent to basement walls, in particular, may be lowered due to
dilution from surface gas. (Note that this also implies that volatile
chemicals applied adjacent to the building, such as pesticides applied
to shrubbery, may be rapidly drawn into the building.)
2.3.1 Designing The Monitoring Plan
This section assumes that impacts on the building indoor air occur
from both transport through ambient air and intrusion of soil gas. The
plan should be developed considering the likely magnitude of the impact
from each source. The procedures for ambient air would quantify
chemical concentrations at the building arising from the Superfund site
as well as any other upwind emission sources. If ambient air transport
is of concern and other potential emission sources exist, upwind-
2-8
-------
downwind sampling may be necessary. For guidance see Volume II of the
Air/Superfund NTGS series (EPA-4501-89-002a).
Ambient Air
Monitoring ambient air for Superfund site impacts is only valid
when the wind direction is from the site toward the building being
monitored. Monitoring equipment should be placed in a location on the
windward side of the building as free as practical from localized
sources of emissions, such as lawn mowers or automobiles, and wind
shields, such as out-buildings or shrubbery. Monitor intakes should
normally be about five feet above ground level (about midpoint of
building ground floor) and about 5 to 15 feet from the building wall.
Duplicate monitors are recommended. Air samples should be collected
over a period of not less than one-hour and preferably over an eight
hour period. Wind speed and direction should be recorded continuously
or at, at least, 30 minute intervals.
The sampling and analytical methods to be used are dependent on
the known or suspected emissions from the Superfund site. For most
applications, EPA methods TO-1, TO-2, or TO-14 will be adequate for
organics. Particulates (inhalable) may be collected using Method IP-10.
Support in selecting the methodologies should be obtained from the
appropriate EPA Division.
Soil Gases
As stated previously, it is to be expected that soil gas concen-
trations will vary with location around the building due to, among
several effects, soil inhomogeneity. Thus, it is necessary to measure
the soil gases at various points around the building. It is recommended
that soil gas probes be used to measure soil gas concentrations. The
use of flux chambers is not recommended. [Flux chambers located within
one to two meters of the building may give significantly negatively
biased results if building underpressurization is exerting an effect on
soil gas flow rates and flow directions. Low permeability zones near
the surface, frozen ground, or wet surface soils (eg., from recent
rains), may also result in low flux chamber results. Flux chambers
located further from the building may or may not realistically represent
2-9
-------
the diffusive flux reaching the region of the building subfloor due to
inhomogeneity of soil properties around the building.]
It is recommended that at least two soil probes be installed on
each side of the building. Statistical considerations presented in
Appendix C of Volume II of the Air/Superfund NTGS series indicate this
would be adequate to determine the average concentration within 20
percent at a 95 percent confidence level.
The objectives of soil gas monitoring are to determine the soil
gas concentrations and the soil permeabilities in the zone very near the
bottom of the building foundation. The probes should be installed so
that the probe tips are between one-half and one meter of the building
floor or basement wall. For slab-on-grade and crawl-space type con-
structions, the probes should be installed on an angle to go under the
building.
It is recommended that slightly undersized pilot holes for the
probes be made with an auger. At this time a sample of soil from near
the bottom of the pilot hole should be obtained and retained for
analysis needed to estimate the effective diffusivity of soil near the
building. (Analyses would include bulk density, moisture content, and
soil characterization, eg., sandy loam).
A volume of soil gas equal to at least two probe volumes should be
drawn through the probes before samples are taken. Gas flow rates
should be low (eg., 10 to 100 cm3/min.) to reduce the possibility of
establishing unwanted pressure gradients.
The exact sampling and analytical methods used will depend on the
contaminants expected from the Superfund site. Assistance should be
obtained from the appropriate EPA Division. The following should be
considered when selecting these methods.
• The indoor air concentrations will likely never exceed five
percent of the soil gas concentration. Thus, very low detection limits
are not required.
• A portable GC will be useful to determine the magnitude of
soil gas concentrations. This or other on-site measurements will likely
be necessary if adsorption tube techniques (eg., T01 and T02) are used
to ensure the capacity of the tubes is not exceeded.
2-10
-------
• The SUMMA canister technique of TO-14 may likely be the best
technique where prior data on chemical compositions and concentrations
are not available.
After soil gas samples are collected, the probe should be used to
determine the soil permeability to gas flow. The method is quite
simple. In brief, a regulated flow of compressed air is passed through
the probe into the soil while measuring the flow rate and probe pres-
sure. Pressure gauges (0-125Pa) and flow meters with range capabilities
of 5 to 450 cm3/nrin are needed. The permeability, which assumes Darcy
flow, is calculated from:
K-
0V-
where,
Kv = Permeability, M2
Q = air flow rate, M3/s
H = viscosity of air, 1.83 x 10s Kg/m-s
r = internal radius of probe, m
Pa = pressure in pascals
Determining the soil permeability is important as it is an indicator as
to whether soil gas intrusion is by diffusion or convection.
2.3.2 Building Inspection
A general inspection of the building should be conducted. The
purpose of the inspection is to assess building construction details
that can be used to judge reasonableness of parameters used to calculate
final indoor air concentrations. Thus, of particular interest are
building size (area of building footprint and area of below grade
walls), construction type (slab-on-grade, crawl-space, or basement,
basement wall construction type - poured concrete or hollow block - ,and
number of stories), presence of obvious cracks in floors or walls in
contact with soil, and details useful for estimating building air
exchange rates (see Appendix A, Section A.3). The inspection forms
given in EPA 400/1-91/-033 and EPA 400/3-91/003 are recommended.
2-11
-------
2.3.3 Estimation of Indoor Air Concentrations
It can be assumed that the soil gas concentrations near the
building are at steady-state. That is, soil gas is diffusing to the
zone of building influence at the same rate that it is entering the
building. Based on the empirical evidence for radon intrusion and
predictions of the Johnson and Ettinger Model (see Appendix A, Section
A.3), this should result in an estimate of contaminant entry that is
conservative, but not strongly so.
Data analysis should begin with a review of soil permeabilities.
In regions where permeabilities are greater than 10~8 cm2, soil gas
flows into the building primarily by convection and at a rate directly
proportional to the permeability. At permeabilities much less than 10"8
cm2, intrusion is primarily by diffusion and is independent of the
permeability. Thus, for permeabilities of 10"8 cm2, or greater, the
empirical relationship for radon presented in Appendix A, Section A.3.1,
may be used to estimate the indoor air concentration. A more precise
estimate for any permeability may be obtained using the Johnson and
Ettinger model.
When using this model in this context, it is important to remember
that the "source" is at the location soil-gas samples were collected
(ie, LT = 0.5m) and the effective diffusivity is for the region between
sampling points and the building walls. Also, the entire area of the
building in contact with soil (floor plus below grade walls) should be
used in the calculation.
It is possible to perform internal consistency checks based
on indoor air concentrations calculated by either method. Using the
calculated indoor concentrations, building air exfiltration rate, and
area of exposed walls, calculate the estimated flux. Calculate the soil
gas concentrations at the actual source (eg. top of water table for
groundwater sources) using the relationships of Appendix A, Section A.I.
Using the measured soil gas concentrations and these data, calculate the
effective diffusivity using the appropriate emission model. This
calculated diffusivity should be reasonable when compared to the
diffusivity calculated from known or reasonable estimates of soil
properties.
2-12
-------
2.4 INDOOR MONITORING
There are various levels at which indoor air monitoring to assess
Superfund site impact can be conducted. It should be clearly understood
at the outset, however, what the utility of each level is and what
conclusions may be reasonably drawn from the data.
The most simplistic approach is to conduct monitoring only inside
the structure. Unless monitoring is being conducted for a specific
compound from the Superfund site that is not found in non-impacted
indoor environments (eg., vinyl chloride), it matters little what
monitoring technique is used so long as the technique has a detection
limit for the target compounds in the mid-to upper-end of the typical
range for non-impacted buildings. This is because only two outcomes are
possible. Either the concentrations are above the maximum reported
concentrations for non-impacted structures, in which case it can be
reasonably concluded that some part (unknown) of that concentration is
due to the Superfund site, or the concentrations are below this maximum,
in which case all that can reasonably be concluded is that if Superfund
site chemicals are entering the structure, they are not causing the
concentrations to exceed the typical range. In the latter case, it
specifically cannot be concluded that none of the target chemicals
identified are from the Superfund site. Thus, monitoring only the
indoor air can, at best, demonstrate that the Superfund site has a major
or minor impact on the indoor air quality. At worst, it provides
information that the site may be contaminating the structure but not
enough information to prove the impact is or is not occurring.
Another indoor only monitoring technique that has been used is to
monitor at the suspected impacted structure and at "control" structures
known or suspected not to be impacted. The objective is to obtain
"typical" concentrations for non-impacted structures for use as a
"background" correction. Because the number of structures typically
monitored does not provide a good statistical sample of the population,
data obtained this way generally provide marginal, if any, improvement
over using the means and ranges from larger studies.
Combining indoor air monitoring with monitoring the potential
pathway(s) (eg., soil gases and ambient air) for Superfund site impacts
2-13
-------
is the preferred method to assess the impact of the site on the indoor
air quality of specific structures when it is necessary to quantitate
the impact through indoor air monitoring. The procedures, discussed in
more detail below, allow refined estimates of the instantaneous concen-
trations and data which can be used to make estimates of the long-term
concentration averages for use in risk assessment.
2.4.1 Screening Level Monitoring
Screening level monitoring can be used to ascertain whether or not
very high levels (relative to typical) of indoor air pollutants exist
and to provide estimates of the indoor air concentrations to guide
design of more refined indoor monitoring. Before this type monitoring
is conducted, decisions should be made as to the appropriate follow-up
action to be taken depending on the concentration levels determined by
monitoring and this information communicated to the building occupants
or other appropriate people (see Chapter 3). It is extremely important
that everyone understand the actions to be taken if the data prove to be
inconclusive with regards to site impact (ie., when concentrations are
within the "typical" range for indoor pollutants).
The use of a portable gas chromatograph (EPA Method IP-1A) is
suggested in this application for organic compounds. The instrument can
be field calibrated and has detection limits adequate for screening
level results. It has the advantage that preliminary results are
immediately available. Alternatively, samples may be collected using
EPA Method IP-1A or TO-14 (SUMMA canisters). The canister method offers
the advantages of positive compound identification, lower detection
limits, and a wider range of compounds identified, with the disadvantag-
es of higher costs and delayed analytical results. The use of adsorp-
tion tubes (eg; EPA TO-1, TO-2, IP-IB) are not recommended for initial
screening unless a rapid response instrument, such as a portable GC, is
available to determine approximate concentrations. Adsorptive capaci-
ties of the tubes varies with different compounds and it is easy to
underload or overload the tubes in unknown environments.
When particulate matter from the site is considered a possible
route for indoor air exposure, samples of both airborne and settled dust
2-14
-------
should be collected. The latter is needed because the larger particles
deposit throughout the structure and may be intermittently resuspended
by traffic or cleaning. Airborne dust may be monitored using an
impactor with filter pack assembly or continuous monitor (EPA method 1P-
10). EPA does not currently have guidance for collecting deposited
dusts. Procedures developed by OSHA can be used [29CFR 1910.132 (a)].
Air samples should be collected over a period of at least one hour
in the living area of each story of the building and the basement, if
applicable. Samples should be collected at between three and six feet
above the floor, near the center of the room, and at least one foot from
any object. Deposited dust samples should be collected from such areas
as refrigerator tops and window sills.
If soil gas intrusion is suspected, it is preferred that the
building windows and doors be kept closed for the preceding 24 hours to
allow establishment of normal pressure gradients. If particulate
monitors are used, it is preferred that indoor smoking be discouraged
for several hours before and during monitoring because of the high
particle count from smoke. Any obvious potential sources of VOCs should
be noted. These include spilled furnace oil, gasoline storage cans and
power equipment in attached garages, paint cans, etc. Inquiries should
be made to determine how much VOC containing products, such as furniture
polishes and hairsprays, have been used in the past 24 hours.
If intrusion by soil gas is of primary concern and there is reason
to suspect ambient air in the vicinity of the building may contain
significant levels of non-site related pollutants of concern, ambient
air monitoring should be conducted concurrent with indoor monitoring.
If indoor and outdoor monitoring are both conducted over at least a 4-
hour period, it is reasonable to subtract out the outdoor concentrations
to obtain an estimate of indoor concentrations due to indoor sources and
potential site chemical impact.
If the only route of concern is infiltration of site contaminants
in the ambient air, indoor monitoring is not required. The average
outdoor concentration near the building may be used as a good approxima-
tion of the indoor concentrations related to site emissions.
2-15
-------
2.4.2 Refined Monitoring For Indoor Air Impacts Of Sites
The objective of the monitoring discussed below is to determine
with high accuracy the actual impact from a Superfund site on the indoor
air of a specific structure. The discussion is predicated on the
assumption that site impacts are from soil gas intrusion and organic and
particulate matter transported via the air pathway. With only slight
modification in the procedures, the impacts of groundwater or non-
aqueous phase liquids seeping directly into the structure may be
determined. However, in these cases, it is expected that indoor air
concentrations would be sufficiently high that decisions could be made
using screening techniques discussed in the previous section.
The procedures allow estimation of concentrations in all stories
of the building and effective air exchange rates in the case of poorly
mixed buildings. Explicitly not included is short-circuiting - pollut-
ants exiting directly to outside without mixing with indoor air. Short-
circuiting can occur when fireplaces or forced air combustion devices
(eg.,oil and gas furnaces) are in operation on the lowest floor of the
building. Additional causes may be clothes dryers or bathroom fans
vented to outside.
The procedures are not prescriptive nor detailed enough to
encompass all possible cases. Professional expertise and judgement are
required especially in such areas as potential pathways to consider,
target compound selection, sampling and analysis methods, and exact
positioning of the monitors.
The procedures also include collecting data useful for making
long-term predictions of indoor concentrations based on short-term
testing. [Professional judgement is needed in deciding if this informa-
tion is needed for the particular assessment]. This is not an exact
science and, although the analysis should provide reasonable results,
the accuracy will depend on how close conditions are on the test day(s)
to average conditions for the structure under test. It is suggested
that monitoring be conducted on several different days under conditions
that approximate the range of normal meteorological and building operat-
ing conditions for the particular building site.
2-16
-------
Data To Be Collected
The following data needs listing is provided to give orientation
to types of information and equipment needed.
• Outdoor air concentrations for all target species. This
includes upwind-downwind monitoring of the Superfund site in
some cases.
• Indoor air concentrations for all target species on all
building levels.
• Surface sampling for deposited particulates if particulate
transport from the site is considered.
• Soil gas concentrations for all target volatile compounds
• Indoor and outdoor temperatures
• Complete building inspection
• Physical volume of each level of the building
• Effective air exchange rate for the building
• Wind speed and direction
• Indoor-outdoor pressure difference
• Barometric pressure
• Soil permeability to vapor flow
.The last 6 items are required only when long-term estimates of
indoor air concentrations are needed. Monitoring should be conducted
over a period not less than four hours and preferably at least eight
hours.
Soil Gas Monitoring
When Superfund site impact via transport of pollutants through
soil is of concern, soil gas monitoring at the structure being investi-
gated is required. The general procedures were described in a previous
section. In this case, however, although the actual concentration is
2-17
-------
important, the concentrations of compounds relative to each other.
rather than actual concentrations, are of utmost importance. (The
reasons for this will be discussed under Data Analysis).
Although soil gas concentrations should not vary significantly
over periods as short as one day, it is considered good practice to
collect an integrated sample over the same time period as ambient air
and indoor air sampling. This is particularly important if measurable
rainfall or significant changes in barometric pressure occur within 12
hours before or during sampling. Rainfall and decreasing barometric
pressure may increase soil gas flow into the building. Increasing
barometric pressure may decrease soil gas intrusion.
Measuring soil permeabilities to vapor flow is not essential to
data analysis. However, it provides extra data that may be useful. If
soil gas relative compositions vary significantly at different loca-
tions, soil permeabilities may be used to estimate relative contribu-
tions from each location.
Ambient Air Monitoring
When Superfund site impact via transport through ambient air is of
concern, ambient air monitoring is required. Monitoring at the poten-
tially impacted structure should, however, be conducted in all cases to
correct for target species from all sources that may enter by this
pathway.
Determining the site related fraction of the total amount of
chemicals measured at a monitoring station adjacent to a potentially
impacted structure can range from quite simple to complex. The simplest
case is when no other significant source of target species exists
between the structure and the site and beyond the site. The most
complex is the opposite situation, but little real difference is
presented if the other source is beyond the Superfund site. In these
complex cases, monitoring upwind and downwind (between the site and the
secondary pollutant source and relative to the direction toward the
structure being tested) of the Superfund site as well as at the struc-
ture would be required. Significant sources of target species between
the structure being tested and the Superfund site would require a second
2-18
-------
monitoring station between the Superfund site and the secondary source.
Resolution of these cases would require significant professional
expertise in both monitoring and dispersion modeling. A general guide
to the requirements can be found in Section 4 of Volume II of the
Air/Superfund NTGS Series. Although these cases may not be typical,
special circumstances may warrant their evaluation so as not to overes-
timate the contribution of this pathway to the indoor air concentra-
tions.
For the ambient air mode of transport, determining the actual
concentrations of target species is important. And, because the
concentrations will, in general, be low, methods with low detection
limits are needed. In general, methods such as TO-1 or TO-14 will
suffice for determining the organics but it is recommended that assis-
tance be obtained from the appropriate EPA Division. Particulate
methods used should be capable of collecting integrated samples in
various size ranges. It is preferred that the same monitoring methods
be used outdoors and indoors.
It is preferred that ambient air sampling begin at least one hour
and preferably two hours before indoor air monitoring begins and
continue until at least 30 minutes before indoor monitoring is complete.
Because most buildings have hourly air exchange rates in the 0.5 to 1
range, air entering the building in the period before indoor sampling
remains in the building for a substantial time. Conversely, outdoor air
concentration fluctuations in the final 30 minutes, unless very large,
have virtually no effect on the average indoor air concentration
measured. Because indoor air is reasonably well mixed, concentrations
inside rise and fall in an exponential relationship (Introduction to
Indoor Air Quality: A Reference Manual, EPA/400/3-91/003, page 7):
C,.T - Co +
-------
Ambient air monitors should be located between 5 and 15 feet from
the building with sampler intakes about five to six feet above ground
level. The location should be such that the effects of windshields
(eg., shrubbery) is minimized. Local sources, such as lawn mowers,
should be moved to a downwind location if possible. Wind speed and wind
direction should be recorded continuously during monitoring.
Ambient air monitoring to assess the air transport pathway should
obviously only be conducted when the prevailing wind is from the
Superfund site toward the monitoring location. Calm winds or signifi-
cant wind direction fluctuations can result in significant negative bias
in the results. These are relatively unimportant if this pathway is not
of concern. Wind speed and direction should still be monitored for use
in air exchange rate and building under-pressure calculations.
It should be noted that in the special case where Superfund site
impact is only from the air transport pathway, indoor monitoring is not
necessary (except for deposited particulates). This is because the
average indoor air concentrations of target species attributable to the
Superfund site will be equal to their average concentration in the
outside air. If short-term effects, such as from remedial actions, are
being evaluated, the building air exchange rate may need to be deter-
mined and indoor air concentrations evaluated using the equation above
(or suitable modification dependent on fluctuations in the outdoor
concentrations).
Indoor Air Monitoring
Indoor air monitoring is conducted to obtain the total concentra-
tions of target species arising from ambient air infiltration, soil gas
intrusion, and indoor sources. Information should be collected that
allows characterization of building dynamics such that reasonable
estimates can be made under conditions different from those existing
during monitoring. This will allow more realistic risk assessment
estimates to be made.
If soil gas intrusion is suspected, best monitoring conditions
exist when the indoor temperature is at least 10°F higher than outside
and windspeeds are steady and exceed about five miles per hour. Under
2-20
-------
these conditions reasonable building air exchange rates and under-
pressurizations develop. The worst condition for monitoring is when the
indoor temperature is lower than the outdoor temperatures and winds are
calm. Under these conditions, the lowest level(s) of the building have
a slight over-pressurization and soil gas entry may be restricted or
eliminated. Monitoring should be avoided if significant precipitation
or barometric pressure fluctuations have occurred in the preceding 12
hours.
Because air circulation patterns in the building vary with forced
air circulation rates, outside wind speed and direction, indoor-outdoor
temperature differences, and indoor thermal gradients, several activi-
ties should be completed before locating monitors. Estimate the volume
of the structure by measuring outside dimensions, as well as each room
in the building. Set up equipment to measure the effective air exchange
rate. It is highly recommended that this be done using a tracer gas
method. Procedures for this are given in EPA Method IP-4. In this
procedure, a known quantity of tracer, such as sulfur hexafloride, is
released into the building, well mixed, and the concentration decrease
measured as a function of time. It is preferable to monitor the tracer
simultaneously in as many rooms of the building as possible. Using the
equation presented in the above section on Ambient Air monitoring and
assuming the concentration of tracer gas in the outdoor air is zero,
CI,T = Cj0 e ~V1 ; T = time since tracer release.
But Cio is the mass of tracer released divided by the effective
air volume of the building and v is the infiltration rate, Q, divided by
the effective air volume, V. Thus,
_ msrSF, -JZ
Li.T — e
2-21
-------
Thus, a plot of Ln Cf T versus T yields a straight line with an inter-
cept at T = 0 (equal to the first term on the right side of the equa-
tion) from which the effective air volume of the building can be
estimated and slope (equal to -Q/V, building air exchange rate) from
which the infiltration flow rate can be estimated. Furthermore, since
Q = L [AAT + B/i2] °-5 from EPA 400/3-9/003, page 8, the informa-
tion collected may be used to estimate infiltrations under various
meteorological conditions. In the above equation,
L = leakage area of building
A = Stack coefficient
AT = Indoor-outdoor temperature difference
B = wind coefficient
/i = average wind speed
The above EPA document provides the appropriate values for A and
B. Thus, L, the leakage area may be estimated. Obviously, indoor and
outdoor temperatures and wind speed should be monitored during this
testing. The relative tracer decay rates in different rooms of the
building are indicative of air movement and may be used to locate the
samplers.
If this level of sophistication is considered unnecessary for the
particular application, the effective leakage area can be determined
using the fan pressurization-depressurization method and the infiltra-
tion rate calculated from the above equation. Although the method is
less accurate, primarily due to uncertainty in the actual air volume of
the building, it is simpler to conduct. Air exchange rates should not
simply be estimated for indoor air monitoring of the level discussed
here.
It is also important to measure the pressure difference between
the inside and outside of the building. This pressure difference will
likely be in the range of zero to ten pascals. It will vary with height
in the building. Building under-pressurization is effected by indoor-
2-22
-------
outdoor temperature difference and wind speed. Thus, it is important
that these variables also be measured. The indoor pressure monitoring
location should be between three and six feet above floor level in the
lowest level of the building. (See Appendix A, Section A.3.3, Equation
24.)
Monitors for target species should be located on each level of the
building. Preferred locations are in the living spaces (living rooms,
dens, bedrooms). They should be located in the breathing zone, three to
six feet above floor level, and approximately in the center of the room.
Monitors should not be located near obstructions or obvious sources of
pollutants.
Selection of monitoring methods depends on the target species
selected. In most cases, indoor air methods IP-1A (canisters) or IP-IB
will suffice for organics although specific methods may have to be used
for some species. Airborne particlates should be collected using EPA
method IP-9 when airborne transport of Superfund site particulates is of
concern. In this case, deposited dusts should also be collected [OSHA
procedure, 29CFR 1910.132(a)].
Indoor air samples should be collected over a period of at least
four hours; eight hours of sampling is preferred. During this period,
minimal indoor activity should occur. Pollutant generating activities
such as housecleaning, furniture polishing, and indoor smoking should be
discouraged. Windows and exterior doors should remain closed to the
maximum extent practical.
Data Analysis
A hypothetical case is presented in Tables 2.1, 2.2, and 2.3 to
assist with following the text below. For this example, monitoring data
are given for the ambient air adjacent to the building, upwind back-
ground ambient air, soil gas, and the indoor air at two levels in the
building. The monitoring data collected for the three target compounds
are tabulated in Table 2.1.
The first step should be to consider the impact due to pollutants
in ambient air. It can be assumed that the average target species
concentration in the indoor air attributable to this pathway is the same
2-23
-------
as the average target species concentration in the ambient air.
Subtract the total average concentrations of target species in ambient
air from the average concentrations measured at each indoor monitoring
location (in Table 2.1,column 1 is subtracted from columns 5 and 6 to
yield the results in columns 5 and 6 of Table 2.2). The resulting
indoor concentrations are the sum of target species concentrations
resulting from indoor sources and soil gas intrusion.
If Superfund site impact via the ambient air pathway is of
concern, determine the average target species concentrations in the
ambient air from the upwind monitoring and modeling results and tabulate
these as "Indoor air concentrations due to Superfund Site Target Species
in Ambient Air" (column 3 is subtracted from column 2, both in Table
2.1, to yield column 3 of Table 2.2) . Note that if this pathway is of
concern and appropriate corrections are not made for non-site sources,
the total target species concentration measured in the ambient air must
be considered to be from the Superfund site. These concentrations will
be the same for all indoor monitoring locations.
Using data from the soil gas monitoring and the indoor air
concentration residuals obtained by subtracting the total average
ambient air target species concentrations (column 2 of either Table 2.1
or 2.2) from the total average concentration measured at each monitoring
location, the site contribution due to soil gas intrusion can be
calculated. Because neither the soil gas flow rate nor the source
strength of all indoor sources is known, the calculation is based on the
relative concentrations of target species in soil gas and the indoor air
residuals above. If it is assumed that soil gas components are not
differentially absorbed indoors, then they should appear at all loca-
tions in the same ratios.
Thus, first divide the soil gas concentrations of all target
species by the highest measured concentration in the soil gas (all
concentrations are, thus, expressed as a fraction between zero and one).
In this example, all soil gas concentrations in column 3 of Table 2.1
are divided by 100 to yield the ratios in column 2 of Table 2.3.
2-24
-------
C
o
re
o
o
Q.
CM
0)
.0
nS
CM
ra S-
c
0 -r-
Cf~
fc_
o o
o z:
n) S-
c
O -r-
C C
O 0
o z:
C V)
-i— ra
. ^
O f—
C "-
o o
O OO
S-
•i- -a
< c
•M O
C S-
eu DI
^ u
<: ca
S- 01
-o
-4-> 1
C -i-
C
nj
c
•r™
E
nj
4-9
c:
o
in
CM
o
^jl
CD
O
t-H
in
o
o
o
f-H
0
o
LU
0
t—
in
~
o
CM
O
in
l-H
O
o
o
CM
o
0
o
LU
0
O-
UD
•^
«5
^1
O
tp
in
o
CD
l-H
•
O
cu
c
CO
IM
C
0)
CQ
C
at
s_
o
O)
4->
O
tu
s_
o
o
rs
-I-)
nj
Q
(0
O
O
Q.
CM
CM
O)
jQ
a)
CM
ns S-
O
O •«-
C. C.
o o
o z:
t
nj S-
O
O -r-
0 O
o z:
C
"~
O • i—
O O t»
T3 C
c o u_
i— i O OO
S- CD
•r- C
-o
+-> 1 —
C •!-
CU 3
•r- CQ
is
C
«J
C
•f—
E
nj
.4-3
C
o
o
en
^[
CM
en
en
CO
o
0
in
o
o
•
o
i-H
0
o
LU
O
t—
00
en
^
,_!
00
en
en
f-H
O
in
l-H
O
0
•
o
CM
O
0
o
LU
t_l
Q_
in
^
in
^
o
vo
in
o
*
o
l-H
•
O
0)
C
cu
C
cu
CQ
CJ
O
OO
S-
o
-o
a>
s-
o
ra
o
O
•f—
-t->
CO
o
a.
co
CM
O)
JD
OS
m
C\J
,—
*f~ ^_
0 O
OO +J
X C
rO O
z: z:
VI
ea
1—1
.f— ^.
0 O
00 4->
•f—
X C
(T9 O
z: z:
CM
S-
o
+-> o
C 4->
O re
z: o:
l-H
S-
o
•«-» o
C -t->
O n)
z: a:
0)
O
c/i nj
ra C£.
t n
•
i— O
•r- C
O 0
00 O
C
re
C
•r—
E
(O
4^
C
0
o
en
*!•
CM
en
en
•
CO
en
•*
CM
en
en
CO
0
o
«
rp-4
LU
O
1—
in
^j-
CM
l-H
in
en
en
*
t-H
vo
en
en
CM
to
en
en
CO
0
in
•
o
LU
O
O_
^^
en
^»"
^
^
en
co
•
CM
in
i^
in
f>I
<*^
VO
*
o
CO
C
CO
N
C
CO
CO
2-25
-------
Now divide the appropriate indoor air residual concentrations at
each monitoring location by the soil gas fraction for that compound. In
this example, for monitor location 1, the 3.99 for TCE was divided by
1.00, the 1.998 for PCE was divided by 0.5, and the 4.5 for benzene was
divided by 0.6 to yield the results in column 3 of Table 2.3.
Identify the lowest non-zero quotient from this operation for each
monitoring location and multiply all soil gas fractions by these
numbers. In this example, 3.99 for TCE is the lowest ratio for monitor
location 1. Thus, the results of this step, given in column 5 of Table
2.3, are obtained by multiplying 3.99 by 1 for TCE, 0.5 for PCE, and 0.6
for benzene. The results are the maximum potential concentrations at
each sampling location from soil gases.
These concentrations should be added to the values calculated
above for site impact via the ambient air pathway (column 3 of Table
2.2) to obtain the total site related concentrations to which occupants
may be exposed (summation not shown).
2.4.3 Making Long-term Estimates from Short-term Monitoring Data
The data above could be used for a preliminary risk characteriza-
tion. However, it is preferred that prediction of longer-term average
concentrations be made for risk evaluation purposes. General procedures
to make this estimate are presented below.
First estimate the building air infiltration rate and air exchange
rate under the appropriate long-term meteorological conditions. Using
the building leakage area estimated from the tracer gas testing (or the
fan depressurization test), estimate the long-term average air infiltra-
tion rate from:
Q = L[AAT + B/u2]0-5
using appropriate long-term average values for wind speed and indoor-
outdoor temperature differences. Unless long-term meteorological
conditions are similar to those existing during monitoring, divide Q by
the measured physical volume of the building (rather than the estimated
effective volume) to obtain average air exchanges per unit time.
Secondly, estimate ambient air concentrations at the building for
Superfund site related chemicals due to the air transport mechanism.
2-26
-------
The ambient air concentrations determined during monitoring can be used
to calibrate an air dispersion model for use in this estimation.
Thirdly, estimate the soil gas flow rate into the building under
the long-term meteorological conditions. Estimate the soil gas flow
rate into the building during the on-site monitoring from:
^soil ~ " i/ soil
where,
Q = estimated building air infiltration rate
Cf = calculated indoor air average concentration of chemicals
in soil gas
Csojt = chemical concentration in soil gas
Now calculate K from
^soii = K Pa
where,
Pa = the indoor-outdoor pressure difference measured
during monitoring
.K = a constant comprised of building specific terms
If soil gas intrusion is primarily by diffusion, the Qsoil calcu-
lated can be used for the long-term average. However, if soil gas
intrusion is primarily by convection, Qsoil is proportional (See Johnson
and Ettinger, Environmental Science and Technology, Vol. 25, No. 8,
1991, page 1449) to the building under-pressurization, as indicated by
the above equation for K. Thus, by using Pa calculated for the building
under-pressurization from equation 24 of Appendix A, Qsoil under differ-
ent meteorological conditions can be estimated from the above rela-
tionship. The measured building under-pressurization during monitoring
should be used to calibrate Equation 24 for site specific conditions.
This implicitly assumes soil gas concentrations remain constant. This
may need adjustment if the source strength varies with periods short
compared to the averaging time.
The above procedure allows reasonable estimates to be made of
long-term indoor concentrations from short term data. Using the risk
assessment procedures outlined in Section 3, reasonable assessments of
the indoor air impact of a site can be made.
2-27
-------
3.0 HAZARD ANALYSIS
Inhalation of contaminants in indoor air is but one of the
exposure pathways considered in risk assessments. Explicit guidance for
the conduct of risk assessments is given in "Risk Assessment Guidance
for Superfund: Volume 1 Human Health Evaluation Manual". The Manual has
three parts:
(1) baseline risk assessment given in Part A, EPA/540/1-89/002
(2) development of preliminary remediation goals given in Part
B, Publication 9285.7-01B
(3) remedial alternatives risk evaluation given in Part C,
publication 9285.7-01C
This chapter considers the potential impacts on indoor air to.be
from soil gas, chemical volatilization from contaminated groundwater use
and ambient air infiltration. The procedures are useful for exposures
occurring during baseline, remedial actions, and post clean-up.
Information is provided to estimate risks for cancer and non-cancer
effects and for short-term and long-term impacts. It is explicitly
assumed that adequate information has been developed on indoor air
chemical concentrations and the likely exposed population.
3.1 INCREMENTAL INDOOR AIR RISKS
The risk assessment procedures detailed in RAGS/HHEM are predicat-
ed on evaluating the incremental risks due to contaminants originating
at the Superfund site. Modeling predications can be, and are, used to
estimate emissions from specific site sources and potential chemical
concentrations at various on-site and off-site locations. These models
do not incorporate chemical releases from non-site sources, although
such releases may be modeled to provide comparative background informa-
tion. Likewise, monitoring data are adjusted for background to assess
incremental site impacts. For example, ambient air monitoring is
conducted up-wind and down-wind of the site, up-gradient and down-
gradient wells are used for releases to ground water, and contaminant
3-1
-------
concentrations in background soils are used to determine site contribu-
tions to the total.
For consistency it is necessary that equivalent procedures be
followed when assessing potential site impacts on indoor air quality.
Since it has been well established that indoor air typically contains a
wide variety of chemicals found at Superfund sites, and, in many cases,
in concentrations high enough to be of concern, risk estimates based on
the total measured indoor air concentrations may reasonably be expected
to frequently result in substantial overestimates of site impacts. Risk
estimates based on predictive modeling would not be expected to include
non-site impacts and therefore are in agreement with current EPA
guidance.
Risk estimates based on combinations of exterior monitoring (eg.,
ambient air and soil gases) may contain some extraneous risk due to non-
site related chemicals. For example, ambient air measurements made
outside the building being evaluated may contain chemicals transported
from non-site locations such as industrial operations, auto exhaust,
roadway paving, and pesticide applications. Soil gases may also contain
chemicals from materials deposited on or in the ground within a few
meters of the structure such as pesticide application or fuel oil
spills.
Risk estimates based only on indoor air monitoring would be a
combination of possible site related impacts and probable non-site
related impacts from the many potential indoor sources. As such, they
would not follow the basic guidelines (ie, assess site related impacts).
To be useful, these data must be corrected for background in so far as
practical. EPA guidance provides that if background risk might be a
concern, it should be calculated separately from site-related risk
(RAGS/HHEM Part A, Section 5.7).
Current EPA guidance, therefore, is that efforts to distinguish
between site-related and non-site related impacts on indoor air would be
required. The study design must, therefore, provide a way to distinguish
among site related indoor impacts, pollutants from indoor sources, and
background concentrations in ambient air. Only the site related
potential risks need to be calculated (non-site related risks may be
3-2
-------
calculated but this is not necessarily required). No current EPA
guidance provides for summing the potential risks from site related and
non-site related impacts.
3.2 RISK COMMUNICATION
Section 8.6 of RAGS/HHEM Part A states:
"The results of the baseline evaluation should not be taken as a
characterization of absolute risk. An important use of the risk and
hazard index estimates is to highlight potential sources of risk at a
site so that they may be dealt with effectively in the remedial process.
It is the responsibility of the risk assessment team to develop conclu-
sions about the magnitude and kinds of risk at the site and the major
uncertainties affecting the risk estimates. It is not the responsibility
of the risk assessment team to evaluate the significance of the risk in
a program context, or whether and how the risk should be addressed,
which are risk management decisions."
The ultimate user of the risk characterization results will be the
RPM or other risk manager for the site.
It is important that the indoor air impact assessment contain a
discussion of:
• confidence that the key site-related contaminants were
identified and discussion of contaminant concentrations
relative to background concentration ranges;
• level of confidence in the quantitative toxicity information
used to estimate risks and presentation of qualitative
information on the toxicity of substances not included in
the quantitative assessment;
• level of confidence in the exposure estimates for key expo-
sure pathways and related exposure parameter assumptions;
• the magnitude of the cancer risks and non-cancer hazard
indices relative to the Superfund site remediation goals in
the NCP (e.g., the cancer risk range of 10"4 to 10"7 and non
cancer hazard index of 1.0);
• the major factors driving the site risks (e.g., substances,
pathways, and pathway combinations);
3-3
-------
• the major factors reducing the certainty in the results and
the significance of these uncertainties (e.g., adding risks
over several substances and pathways);
Assessment of potential risks due to the indoor air pathway for
buildings located proximate to Superfund sites may be initiated as a
result of individual property owner requests, community requests,
preliminary assessments, baseline investigations, or assessment of
remedial alternatives. In all cases, it is important to bear in mind
that good communication with the affected people is critical to their
acceptance of the results and recommendations. RPMs are encouraged to
work with the risk assessor and community relations coordinator to
develop the appropriate means for communication.
It is important in estimating risks due to exposure to indoor air
to recognize that pollutant sources other than Superfund sites normally
exist in buildings. In fact, prior studies indicate the cancer risk
factor for private residences not proximate to Superfund sites may be
10"4 or higher. While this fact may be communicated to building occu-
pants, as well as other general risk information, it must be understood
that, in general, people view and accept voluntary and involuntary risks
differently. Thus, the potentially affected population may express
different perceptions and acceptabilities for equivalent risks arising
from exposure to chemicals used in routine household activities and the
same chemicals originating from a Superfund site.
Experience has shown that it is important to develop a management
plan prior to initiation of an indoor air impact assessment. While the
plan will vary with site-specific considerations, it should be adequate
to communicate the sequence of actions to be taken to evaluate exposures
and risks, the methodology that will be employed, the levels of risks
upon which decisions will be made, and actions to be taken based upon
those decisions. An illustrative decision tree is presented in Figure
3-1.
In this illustration, potential exposures, and resultant risks,
are first estimated based on the use of reasonable predictive models.
If the simple screening model predicts significant exposures, then a
more sophisticated model is run to give a better, and more realistic,
3-4
-------
Figure 3-1. Illustrative Decision Tree
Conduct Modeling to E*tIm*U Exposure
1. Conduct corw«rvctlv« tcf*«nlng mod«llno
2. If modal pf«dlcts »Kjmfic«nt impact, th*n conduct
•opni*tlc*t*d modeling
1
Model Predict* Expowre
Risk Above X
Communjctte to Reiident* No
Additional Risk Projected
Results Indicate risks are:
High
Moderate
Low
Confirm Remits by Monitoring Inside
and Outside Building
1
Confirm Model Prediction! by Ambient Air
•nd/or Soil Gas Monitoring
i
Monitoring Results Confirm Risk Exists
Inform Residents of Actions to be Taken
3-5
-------
estimate of the indoor concentrations likely to exist. If the simple
screening model, which is conservative, does not indicate concentrations
could be high enough to cause concern, there is no reason to run the
more sophisticated models, which would only reduce the estimated concen-
trations. If the modeling predicts exposures which would pose risks
above a predetermined value, on-site monitoring, varying with the level
of risk estimated, is conducted. If the on-site monitoring confirms the
model predictions, then predetermined actions may be taken based on the
estimated risk level.
Whatever decision tree is developed for a particular site should
be communicated to the potentially affected public before the plan is
executed. The meaning of the cancer risk factors (eg., 1 xlO"4) and
hazard indices should also be communicated.
Consider, for example, the following case which was developed for
a specific chemical at a specific site. Information developed during
the screening phase indicated the potential for significant indoor
exposures to a chemical with estimated excess lifetime cancer risks to
humans as detailed below.
Apportioned Lifetime Risk from a Four-year Exposure
AGE
0 to 5
6 to 9
10 to 13
14 to 17
18 to 21
100 ppb/v
2.3 E-2
5.8 E-3
4.6 E-3
3.5 E-3
2.3 E-3
10 ppb/v
2.3 E-3
5.8 E-4
4.6 E-4
3.5 E-4
2.3 E-4
1 ppb/v
2.3 E-4
5.8 E-5
4.6 E-5
3.5 E-5
2.3 E-5
0.2 ppb/v
4.6 E-5
1.2 E-5
9.3 E-6
7.0 E-6
4.6 E-6
0.1 ppb/v
2.3 E-5
5.8 E-6
4.6 E-6
3.5 E-6
2.3 E-6
These data suggest that children, especially very young children and
neonates, may be unusually sensitive to the chemical. In this case no
children resided in the buildings on a full-time basis but were present
on a regular basis as frequent visitors or for daycare purposes. Thus,
3-6
-------
it would be relatively simple for parents to take measures to reduce
exposure time for their children.
Before on-site monitoring took place, it was determined that
temporary corrective actions could be taken, if needed, and effective-
ness confirmed within twelve days after the monitoring data was avail-
able. It was also determined that permanent remedial action, if needed,
could be taken, and its effectiveness determined within 60 days.
The following decision scheme was then designed for actions to be
taken depending on the actual concentrations determined by on-site
monitoring:
0.2 to 10
ppb/v
If the concentration were to be found to be within
this range, permanent remedial action would be
taken to bring the levels below 0.2 ppb/v. The
action would be instituted and effectiveness con-
firmed within sixty days.
10 to 100
ppb/v
If the concentration were to be found to be within
this range, temporary corrective actions would be
instituted to bring the levels below 10 ppb/v.
This action would be instituted and confirmed with-
in twelve days. The temporary action would be
followed by a permanent remedial action to bring
the levels below 0.2 ppb/v. The permanent action
would be instituted and effectiveness confirmed
within sixty days.
3.3 OVERVIEW OF RISK ASSESSMENT GUIDANCE
The purpose of the balance of this chapter is to provide an
overview of the Superfund guidance (RAGS Part A) for the indoor air
pathway. As such, it must be understood that it does not contain the
detail necessary for untrained personnel to conduct a risk assessment.
Risk assessment data, even in cases where accepted cancer or non-cancer
3-7
-------
risk factors are available, should be reviewed with a trained profes-
sional for site specific adjustments that might be necessary.
3.3.1 Exposure
The first step considered is exposure of the affected population.
The fact that certain concentration of pollutants exist in the indoor
air does not in and of itself constitute a risk. The magnitude,
frequency, and duration of that exposure for the affected individuals
(or grouping such as adults and children) must be defined. The average
exposure can be expressed as a function of time or of body weight.
Exposure normalized for time and body weight is termed "intake" and is
expressed in units of mg chemical/Kg body weight-day. Thus, there are
three categories of variables for intake:
• chemical concentrations
• population (contact rate, exposure frequency and duration,
and body weight)
• averaging time
The intake variable values should be selected so that the combina-
tion of all intake variables results in an estimate of the reasonable
maximum exposure for the pathway. "Reasonable" is not based solely on
quantitative information and requires professional judgement. EPA
guidelines for "reasonable" generally use the 95th or 90th percentile
values for contact rate, exposure frequency, and duration. If statis-
tical data are not available, use reasonable conservative estimates.
Professional judgement is required to estimate the exposure time and
exposure duration for exposure to indoor air. Both can vary with
building type (home, school, workplace) and age.
Some chemicals can produce an effect after a single or very short-
term exposure to relatively low concentrations. These chemicals include
acute toxicants such as skin irritants and neurological poisons, and
developmental toxicants. At sites where these types of chemicals are
present, it is important to assess exposure for the shortest time period
that could result in an effect. For acute toxicants this is usually a
single exposure event or a day, although multiple exposures over several
days also could result in an effect. For developmental toxicants, the
3-8
-------
time period of concern is the exposure event. This is based on the
assumption that a single exposure at the critical time in development is
sufficient to produce an adverse effect. It should be noted that the
critical time referred to can occur in almost any segment of the human
population (i.e., fertile men and women, the conceptus, and the child up
to the age of sexual maturation).
When evaluating longer-term exposure to non-carcinogenic toxi-
cants, intakes are calculated by averaging intakes over the period of
exposure (i.e., subchronic or chronic daily intakes). For carcinogens,
intakes are calculated by prorating the total cumulative dose over a
lifetime. This distinction relates to the currently held scientific
opinion that the mechanism of action for each category is different.
The approach for carcinogens is based on the assumption that a high dose
received over a short period of time is equivalent to a corresponding
low dose spread over a lifetime (EPA 1986b). This approach becomes
problematic as the exposures in question become more intense but less
frequent, especially when there is evidence that the agent has shown
dose-rate related carcinogenic effects. It is necessary to consult a
toxicologist to assess the level of uncertainty associated with the
exposure assessment for carcinogens.
3.3.2 Toxicity Assessment
Health impacts of indoor air chemicals may be divided into two
broad classes: non-carcinogenic effects and carcinogenic effects.
However, there are additional considerations within each broad class.
A reference dose, RfD, is the toxicity value most often used for
evaluating non-carcinogenic effects resulting from exposures at Superf-
und sites. Note that carcinogens may exhibit non-carcinogenic effects
and that RfDs for these chemicals should also be sought. Various types
of RfDs are available. These include route of exposure (e.i., oral and
inhalation), critical effect (i.e., developmental and others), and
length of exposure event (chronic, subchronic, or single event). A
chronic RfD is an estimate of a daily exposure level for the human
population, including sensitive subpopulations, that is unlikely to
result in an appreciable risk of deleterious effects during a lifetime.
3-9
-------
Chronic RfDs should be used to evaluate potential non-carcinogenic
effects associated with exposure of from seven years to lifetime.
Subchronic RfDs (RfDs) should be used to evaluate effects from exposure
periods of two weeks to seven years. Developmental RfDs (RfDdT) are
used to evaluate the potential effects on a developing organism follow-
ing a single exposure event. General use of the RfDs assumes a thresh-
old type mechanism in which no adverse effect is observed below the RfD.
However, RfDs are generally considered to have an uncertainty spanning
an order of magnitude or more and are not viewed as a strict scientific
demarcation between toxic and non-toxic levels.
Carcinogenic effects are considered to have no threshold and,
thus, calculations are based on the presumption that any exposure
carries a finite probability of a carcinogenic response. For Superfund
purposes, the relationship between dose and response is considered to be
linear in the low-dose region and represented by the slope factor. This
factor generally represents the upper 95th percent confidence limit on
the probability of a response per unit intake of a chemical over a
lifetime. Most carcinogenic slope factors (CSF) are based on adminis-
tered dose, however, some data are based on the absorbed dose. Care
should be exercised in the selection and application of these two
different slope factors. Toxicity values for carcinogenic effects may
also be expressed in terms of risk per unit concentration of the
chemical in the medium where human contact occurs.
3.3.3 Sources of Toxicity Values
EPA's Integrated Risk Information System (IRIS) is a data base
containing up-to-date health risk and regulatory information for
numerous chemicals. It contains only those RfDs and slope factors that
have been verified by appropriate work groups. Information in IRIS
supersedes all other sources. Only if information is not in IRIS should
other sources be consulted. IRIS is only available on-line. For
information on how to use the database, contact IRIS User Support
(513/569-7254).
The Health Effects Summary Tables (HEAST) summarize interim, and
some verified, RfDs and slope factors and contains toxicity information
3-10
-------
for specific chemicals. HEAST also provides references to supporting
toxicity information. HEAST is published quarterly and can be obtained
from the Superfund Docket (202/382-3046)
The Agency for Toxic Substances and Disease Registry (ATSDR) is
developing toxicological profiles for 275 hazardous substances identi-
fied at Superfund sites. The profiles contain general toxicity informa-
tion and levels of exposure associated with various endpoints. Health
effects in humans are discussed by exposure route (e.g., oral, inhala-
tion, and dermal) and duration (eg. acute, intermediate and chronic).
Information on the status of a particular profile can be obtained by
contacting ATSDR or the National Technical Information Service (NTIS) at
703/487-4650 or 800/336-4700.
EPA's Environmental Criteria and Assessment Office (ECAO) may be
contacted at 513/569-7300 for general toxicological information as well
as for technical guidance on the evaluation of chemicals without
toxicity values. ECAO will respond to contractor requests only upon
identification of the RPM or regional risk assessment contact. ECAO
should be contacted before using references other than those cited in
IRIS or HEAST to see if more current information is available. Any
derivation of toxicity values should be done only in conjunction with
the regional risk assessment contact, who will submit the derivation to
ECAO for approval.
3.3.4 Quantifying Risks
Following the above procedures, chemical intakes and appropriate
RfDs and slope factors will have been obtained. Before proceeding with
calculations, it is highly recommended that the information be consoli-
dated into tabular format to ensure the proper data are used in the
calculations. Specific checks should be made to ensure:
• All RfDs and CSFs are based on inhalation as the route of
exposure
• All RfDs and CSFs are expressed in the same units as used
for chemical intake. Toxicity values obtained from IRIS are
generally expressed as ambient air concentrations (ie.,
mg/m3) instead of administered dose (ie., mg/Kg-day).
3-11
-------
• Non-cancer effects are considered for carcinogens.
• The proper RfDs are used for the exposure scenarios. Do not
use chronic RfDs for short-term exposures (except as initial
screening values) and do not convert short-term exposures to
equivalent lifetime values to compare with chronic RfDs.
Use only subchronic or shorter-term toxicity values that
compare well with the estimated exposure duration.
• All intakes and toxicity values are expressed as adminis-
tered dose. A few chemicals listed in IRIS (eg., trichloro-
ethylene) have the CSF expressed as absorbed dose.
• Averaging times (AT) for non-carcinogenic risks are equal to
the exposure duration (ED). ATs for carcinogenic effects
are always 70 years.
The table must have the data separated such that carcinogenic and
non-carcinogenic risks can be summed separately. (It is best to present
these in separate tables). Also, the table must have the data for non-
carcinogenic risks presented such that short-term, subchronic, chronic,
and developmental risks, as appropriate for the scenario being consid-
ered, can be summed individually.
For carcinogenic effects, multiply the appropriate calculated
intakes by the slope factors for the chemicals. For non-carcinogenic
effects, divide the appropriate calculated intakes by the RfDs for the
chemicals. Note that this procedures gives, for carcinogens only, an
upper-bound estimate of the lifetime incremental risk of developing
cancer. For non-carcinogens, this yields a non-cancer hazard quotient
(HQ), the value of which is not a measure of statistical probability of
non-cancer effects.
For carcinogenic effects, sum only the individual risks calculated
using the slope factors. This sum represents the total upper-bound
incremental lifetime cancer risk.
For non-carcinogenic effects, several separate summations can be
made. If more than one exposure duration (ie, chronic, subchronic,
short-term, or developmental) was considered, each must be summed
separately. Do not sum all non-cancer HQs either by summing individual
data or summing totals for chronic, subchronic, etc. Furthermore, sum
3-12
-------
only those Hqs corresponding to exposures that will be occurring
simultaneously. These sums are designated Hazard Indices (HI).
At this point, one or more tables will exist which contain
individual and total cancer risk factors and non-cancer hazard quotients
and hazard indices. Understanding the meaning of and limitations of
these results is discussed in the following section.
3.3.5 Data Limitations and Interpretations
3.3.5.1 Carcinogenic Effects
It was assumed in the preceding that the cancer risk could be
calculated by multiplying intake by the slope factor. This is valid in
the low-dose portion of the dose response curve. For estimated risks
above 0.01, this assumption may no longer be valid and risk may be
calculated using the one-hit equation:
Risk = 1 - exp [(I)(AF)]
where all terms are as previously defined.
Unless the total risk number calculated is dominated by the risks
from one or two chemicals, the total cancer risk may be overstated.
This can occur because slope factors are upper 95th percentile estimates
of potency and upper 95th percentiles of probability distributions are
not strictly additive.
The total risk summation procedures gave equal weight to class A,
B, and C carcinogens as well as to slope factors derived from human or
animal data. The calculation, therefore, may overestimate the risk from
some chemicals.
The summation procedure explicitly assumes the actions of the
carcinogens are independent. This assumption ignores possible synergis-
tic or antagonistic effects among chemicals and assumes similarity in
mechanisms and metabolism. These assumptions are made because data to
quantitatively assess mixture interactions are generally not available.
3-13
-------
3.3.5.2 Non-Carcinogenic Effects
The hazard quotient calculated assumes no effect below HQ = 1.
The level of concern does not increase linearly as HQ approaches or
exceeds unity. The RfDs are not based on the same severity of toxic
effects and the slope of the dose-response curve can vary widely among
different chemicals. However, for HQ >1, there is concern for potential
non-cancer effects which increases as HQ increases.
The hazard indices (HI) calculated assume that simultaneous
subthreshold (ie, HQ< 1) exposures to several chemicals could result in
an adverse health effect and that the magnitude of the concern is
proportional to the sum of the HQs. Although an HI exceeding unity may
be of concern, the level of concern is higher if individual Hqs exceed
unity. Since the His are derived from Hqs for compounds that may not be
expected to induce the same type of effect or that do not act by the
same mechanism, they can overestimate the potential for effects. If the
HI is greater than unity as a consequence of summing several Hqs of
similar value, it would be appropriate to segregate the compounds by
effect and by mechanism of action and to derive separate his for each
group. This analysis is not simple and should be performed by a
toxicologist.
3-14
-------
APPENDIX A. PREDICTIVE SCREENING TECHNIQUES
-------
TABLE OF CONTENTS
APPENDIX A. PREDICTIVE SCREENING TECHNIQUES
A.I Models to Estimate Contaminant Flux A-l
A. 1.1 Farmer Model A-2
A.1.2 Correction for Transport Resistance of
Capillary Fringe A-4
A.1.3 Correction for Adsorption to Soil . . . A-7
A. 1.4 FAVN Model A-8
A.1.5 Emissions From Contaminated Soil .... A-9
A.1.6 Emissions from Soils Containing High
Chemical Concentrations A-12
A.2 Airborne Pollutants A-14
A.3 Relating Emissions to Indoor Air
Concentrations A-16
A.3.1 Empirical Relationship Based A-16
A.3.2 Typical Simplified Assumptions Used . . A-18
A.3.3 Incorporating Building Impacts on
Infiltration Rate A-20
A.3.4 Impact of Contaminated Outdoor Air . . . A-29
A.4 Sources of Data for Model Parameters A-31
-------
APPENDIX A. PREDICTIVE SCREENING TECHNIQUES
Conducting sampling of the indoor environment at each and every
potentially impacted structure proximate to all sites would be an
enormous undertaking. Furthermore, simply having indoor air quality
(IAQ) data for one day or even a week would likely be of limited
usefulness since the IAQ at other times would not be known. It is
reasonable to expect that chemical intrusion into a building would be
different at different times of the year and would also be effected by
occupant activities such as heating or cooling the building among
others. Thus, single point data alone gives little indication as to
whether the instant IAQ was typical for that particular structure.
Modeling provides an opportunity to estimate the IAQ for a variety
of site conditions and structure types. It should be recognized,
however, that model results inevetiably depend on the quality of input
data available. Seldom is all site specific data needed by the models
actually available. The simplifying assumptions used to substitute for
this data likely will not be exact for any particular site. Thus, while
model results can provide reasonable estimates, one should not expect
highly accurate results. Because modeling can give estimates for a
range of scenarios, it can provide information to assess the likely
level of site impact on the structures and provide an informed basis
upon which to select areas for actual monitoring.
Presented here are some of the models available for estimating the
rate at which subsurface chemicals and chemicals in ambient air can be
transported into the region near a building, models to estimate the
infiltration of those chemicals into the building, and models to
estimate the concentration of the chemicals in the indoor environment.
A.I MODELS TO ESTIMATE CONTAMINANT FLUX
Except in the case of direct groundwater intrusion into a build-
ing, chemicals in the nearby and underlying soils must volatilize into
the soil gas and migrate to the below grade walls before they can
infiltrate a building. Models in this section describe several ap-
proaches, varying in complexity, for estimating this source strength.
A-l
-------
A.1.1 Farmer Model
The Farmer model (1) is a fairly simple screening tool. It was
originally developed for estimating emission rates from covered land-
fills without internal gas generation. It is the basis for this type
model currently recommended in EPA's Superfund Exposure Assessment
Manual (SEAM). The SEAMS model differs from the Farmer model in that
the SEAM model assumes completely dry soil, contains an explicit term
for surface area, and an explicit term for estimating soil gas concen-
tration from waste composition.
The Farmer model can be used for calculating emissions from chemi-
cals dissolved in groundwater and for contaminated soils. It explicitly
assumes that the source pollutant concentration (ie., chemical concen-
tration in groundwater or soil) is not decreased by transport of the
chemical to the surface and the depth to the top of the pollutant source
remains constant. The model considers the flux rate of chemicals to be
a result of Fickian diffusion of the vapor through the soil matrix.
By assumming complete equilibrium is established between chemicals
in the soil gas and the groundwater and soil and ignoring all attenuat-
ing factors such as biodegradation, it is possible to estimate the
chemical concentration in each phase. For the typically low chemical
concentrations in groundwater or soil (the high concentration case is
discussed later), the vapor concentration of the chemical in the soil
pore vapor can be estimated assuming the chemical equilibrium follows
Henry's Law and is unaffected by other components of the system.
The soil gas concentration at the contaminant source (eg. water
table surface) is calculated from:
where,
Cg = Chemical concentration in vapor phase (g/cm3)
Ct = Liquid phase concentration of chemical (g/cm3)
H = Henry's Law constant (atm - m3/mole)
R = Universal gas constant (8.2 x 10"5 atm - m3/mole-°K)
T = Soil temperature ( °K)
A-2
-------
[Note: Some tabulations of the Henry's Law constants use different
units. Dimensionless constants (H/RT) are common. It is important that
proper units be used.]
Once in the vapor phase, the chemical diffuses through the soil at
a rate dependent on the soil porosity, pore space geometry, the chem-
ical's air diffusion coefficient, and concentration gradient between the
source and point of exit from the soil. The effective diffusion
coefficient (Ds) is calculated from:
D P10/3
D. = -2fi5i- (2)
PT
where,
DA = Vapor phase diffusion coefficient in air
(cm2/sec)
Pa = Air filled soil porosity (unitless)
PT = Total soil porosity (unitless)
The steady-state flux (J) is then calculated from:
where,
C2 = Vapor phase concentration at point of soil exit
(9/cm3)
L = Distance from source to point of exit (cm)
Typically, C2 is set to zero to maximize the flux.
Expressed in complete form, the modeling equation is:
Using the above units, J is calculated in grams per second per
square centimeter at the point of exit (ground surface or building
floor). This only provides an estimate of the rate at which chemicals
A-3
-------
are reaching the building. Estimates of the amount of this flux
actually entering the building are needed to estimate indoor air
concentrations. This is considered later in this Appendix.
Because the equation ignores all possible attenuating factors, it
is likely this model overpredicts the contaminant flux. However, due to
its simplicity, this approach provides a simple method to estimate, at a
screening level, the likely maximum rate at which chemicals would be
transported to a building.
A.1.2 Correction for Transport Resistance of Capillary Fringe 2
There is not a clean separation between saturated soil in the
groundwater and air filled soil above. There exists a capillary fringe
above the top of the groundwater table in which the soil pores are also
filled with water (in the upper region of the fringe, the soil moisture
content is below saturation but still elevated in comparison to the soil
above the fringe). Because water in the fringe does not move laterally
with the bulk groundwater, chemical concentrations in the fringe are not
necessarily the same as in the bulk groundwater. Contaminants In the
bulk groundwater must diffuse through this fringe before diffusing
through the air-filled fraction of the soil pores in the unsaturated
zone. Because molecular diffusivities in water are characteristically
four orders of magnitude smaller than through air (eg., air filled soil
pores), the concentration of volatile components can be expected to be
lower at the top of the fringe as compared to the concentration at the
fringe-groundwater interface. The presence of even a thin capillary
fringe should impede transport of volatile components from the groundwa-
ter. Note that the Farmer model essentially assumes this fringe is
small compared to the distance from the top of the groundwater table and
that the volatile component concentration is the same at the top of the
fringe as in the bulk groundwater. This is a conservative assumption
that maximizes the emission flux.
The depth of a capillary fringe is dependent primarily on the size
of soil particles. As the particle size decreases, the depth of the
fringe increases. Thus, for sandy soils a small fringe would be
expected, whereas for a clayey soil a substantial capillary fringe is
likely. The correction would normally be considered only in cases where
A-4
-------
a significant capillary fringe is expected and when estimates made using
the Farmer model indicate a moderately high impact on indoor air. If
the Farmer model predicts a low impact, there is no reason to make this
correction (except to estimate even lower potential impacts). If the
Farmer model predicted high potential impacts, application of this
correction would likely result in a substantially reduced estimated
impact but would likely still indicate on-site soil gas monitoring was
needed.
If the fringe is considered to provide transport resistance, the
total effective diffusivity is represented by:
D- = L
L
/-
dX
(5)
where,
DT = Total effective diffusivity
L = Distance from water table - capillary fringe
interface to point of surface exit of volatile
component (eg., ground surface or basement
floor)
X = A vertical coordinate
ia/3 D (P (x)10/3)
De(X) = DA T * - + ^-^^- - }- (6)
Dw = Diffusivity of contaminant in water
H = Dimensionless Henry's law constant (concentra-
tion in air per concentration in water)
Note that equation 6 is similar to equation 2, but explicitly
corrects for the variable water-and air-filled porosities across the
capillary fringe.
Above the capillary fringe, equation 5 is easily solved since PT-
PH(x) from equation 6 becomes the same Pa from equation 2 and the second
term on the right side of equation 6 can be assumed to be very small in
comparison to the first term.
The problem then can be reduced further to:
A-5
-------
D
-i
T _
D
(v:
cap
where,
D,
cap
and:
a.
Height of capillary fringe
Effective diffusivity through capillary fringe,
f
J .
0
dx
D6(x)
-i
'1>T
IDO(X)
, f
i
dx
9eU>
-i
(8)
where,
LT = Height of the saturated portion of the capillary
fringe
This is readily solved if high resolution spatial data are
available for moisture profile within the capillary fringe. Field
measurements of soil moisture content with, for example, a piezometer
would be required for accurate determinations. For this modeling
approach, it is essential that the thickness of the saturated portion of
the fringe (LT) be determined fairly accurately since this is the region
where it is anticipated that most of the resistance occurs.
Using equation 6,
4
LTH
A.U)
(9)
One approach to solving the last integral in equation 8 when
adequate water-filled porosity data are not available for the region LT
to Lc, is to assume constant water-filled porosity equal to the average
in the saturated and unsaturated zones.
A-6
-------
The calculated value of DT should be substituted for Ds in equa-
tion 3 to apply the correction to the Farmer model for transport from
groundwater.
A.1.3 Correction for Adsorption to Soil
When making estimates of emission flux based on chemical concen-
trations in soils, it is necessary to consider their adsorption to the
soil particles and how the strength of that adsorption affects the
concentration in the soil gas. In this case, the compound will be
distributed among soil particles, soil pore moisture, and soil gas.
Equilibrium between the solid and liquid phases for the compound
can be expressed by: 3/4
C0 = K,Cr (10)
where,
Cs = Concentration in soil
CL = Concentration in soil moisture
Kd = Slope of adsorption isotherm
Kd primarily represents adsorption to organic matter and can be
represented by:
K, = K,
oc
where,
K.
'OC
oc
Organic carbon in soil to water partition coef-
ficient
Fraction of organic carbon in soil
From equation 1, Cg = HCL (using the dimensionless Henry's Law con-
stant).
Thus,
C =
_
k
&.
(11)
OC OC
A-7
-------
This correction is typically not included in models based on
pollutant concentrations in groundwater because the concentration in the
liquid phase can be directly determined and related to the concentration
in the soil gas as previously described. It is used when modeling
emissions using bulk soil concentrations. The correction is only
applicable to cases where the chemical concentration in the soil is low
(defined as a concentration for which the soil gas concentration
calculated from Equation 11 is less than that calculated from the vapor
pressure of the pure compound).
This correction would normally be used only in cases where the
soil has a high organic matter content and when the simple screening
model (Farmer) does not provide clear indication of significant poten-
tial for indoor air impacts. In these cases, application of the more
complex model may well show that there is a low potential for impacts
and on-site monitoring is not required.
A.1.4 FAVN Model *
Modeling presented in the preceeding parts of this Appendix is
used to estimate instaneous surface chemical flux based on instaneous
chemical concentrations in the soil or groundwater. The FAVN model can
be used to calculate the long-term average vapor flux of a volatile
compound from contaminated soil or groundwater when the time to com-
pletely evaporate the compound exceeds the exposure period for the
calculation. The equation upon which it is based was obtained from
EPA's Superfund Exposure Assessment Manual5 (equation 2-19).
FAVN
longteim
_Cg
~T
2tDTHCv
1/2
(12)
where,
FAVN
iongterm
Time-average vapor flux to the surface
over period t (mg/m2-day)
Bulk contaminant concentration in soil
(mg/m3)
Averaging period (days)
A-8
-------
L = Distance from basement bottom or ground
surface to top of contaminated zone (m)
DT = Effective diffusivity of contaminant
through soil pores (m2/d)
H = Dimensionless Henry's Law coefficient
CH = Contaminant concentration in pore water
(mg/m3)
When calculating the flux from groundwater, with a clean soil
layer between the top of the water table and the surface, DT should be
calculated as given in Section 3.1.2 to correct for resistance in the
capillary fringe.
The following relationships may be of use in application of the
model:
• from equation 10,
from equation 1,
CB
from equation 2,
D P 10/3
DT = Ds = A * - , in the unsaturated zone
A. 1.5 Emissions From Contaminated Soil 6
For a soil column with an initial uniform vertical chemical
distribution, the top layers of soil are depleted of volatile chemicals
by the process of volatilization and diffusion, forming a decontaminated
zone at the top. Over time, the size of the uncontaminated zone
increases and the size of the contaminated zone decreases. Hence, the
instantaneous emission rate, which is proportional to the depth of the
contaminated zone, decreases with time. The modeling equations in this
section are particularly suited to cases in which the soil contaminant
A-9
-------
concentrations are low, the contaminant(s) have moderate to low volatil-
ities, and average emission rates over long time periods are desired.
The model cannot be used if chemical concentrations are so high that the
soil gas concentration calculated from Equation 11 exceeds that calcu-
lated from the vapor pressure of the pure compound (see next section for
that case).
Several authors7' 8' &9 have solved second-order differential
equations to simulate the transport of vapors through soil and waste.
The analytical solutions to these equations generally differ due to the
initial and boundary conditions that are applied to the equations.
These conditions are designated according to the specific situation
being modeled, i.e., the nature and distribution of contamination at the
site of interest, and the variation in source concentrations with time.
The following assumptions are made in the derivation of the following
equations:
The chemicals are uniformly distributed throughout the soil
column; no transport of chemicals by water movement occurs; the
total porosity is equal to the sum of air-filled and moisture-
filled porosity; soil properties are constant in both time and
space; diffusion in air is the rate-controlling step, with all
other partitioning occurring instantly; adsorption is reversible;
and the fraction of organic carbon in soil is constant in time and
space.
In a one-dimensional, homogeneous, porous medium, the generalized
mass conservation equation without convection for chemicals undergoing a
first-order biodegradation can be expressed as7:
at
where,
Ca = Chemical concentrations in soil pores, g/m3;
/i = Net degradation rate, day;
z = Soil depth below the surface, m;
t = Time, sec; and
A-10
-------
a = The effective diffusion coefficient, m/sec, given as:
a. =
H H
Where,
DG = Effective air diffusion coefficient, m2/s;
DL = Effective water diffusion coefficient, m2/s;
Da, = Air diffusivity, m2/s;
Water diffusivity, m2/s;
9 = Total soil porosity, 9 = 9a + 0m, cm3/cm3;
9a = Air-filled soil porosity, cm3/cm3;
T = Exposure interval, sec.;
9m = Moisture-filled soil porosity, 6m = wpb, cm3/cm3;
w = Moisture content, g water/g soil;
p = Soil particle density, g/cm3;
pb = Dry bulk density of soil, pb = (l-9)p, g/cm3;
Kd = Distribution coefficient, cm3/g; and
H = Henry's Law constant, dimensionless.
The initial and boundary conditions used here are kept the same as
those used by the USEPA8 in the development of advisory levels for PCB
cleanup:
Initial Concentration: Ca = (H/Kd)Cso 9 t = 0, z > 0
Boundary Condition 1: Ca = (H/Kd)Cso @ t > 0, z - «°
Boundary Condition 2: Ca = 0 @ t > 0, z = 0
Cso = concentration in bulk soil
The solution to Equation 13 using the stated conditions is:
rr S 7 1
C — I \ /"" * or ft ( H t-\ 1/2 fl-|*£ I "\ A \
a ~ \^^I cso * erj-\~^ I« c; o 114;
if-f
L 2
where erf (n) is the error function of the argument n.
Under the stated boundary conditions, Equation 14 can be used to
estimate the instantaneous mass flux of chemical vapors at the soil-air
interface, Na, as a function of time:
A-ll
-------
ac
Using Equation 15, the time-average mass flux, Na, over the
interval t has been estimated to be (USEPA 1986):
N~ = _ G*° _ . e-^ (16)
a
If this general equation is simplified by assuming that contami-
nants move predominantly in the vapor phase(ie., unsaturated soil -
second term in a disappears); the chemical biodegradation rate is
negligible (last term in Equation 16 = 1); the total soil porosity is
equal to the air-filled soil porosity and, thus, the effective diffusiv-
ity, DG, is represented by 6a4/3Dai, Equation 16 becomes identical to the
solution developed by the USEPA8.
It should be noted that the solution presented above applies to
cases with an infinite depth source of constant concentration (see
stated boundary conditions). Thus, although the zone of contamination
is constantly receeding from the initial upper boundary, some chemical
remains at all times. This differs from the simpler models which assume
a constant depth to the contamination. However, applying these condi-
tions to chemical sources of finite depth, may results in an overesti-
mate of soil concentrations, emission rates, and ambient air concentra-
tions, depending on averaging times used, and is therefore conservative
but not overly so.
A.1.6 Emissions from Soils Containing High Chemical Concentrations6
This model is presented for use when the chemical concentration in
the soil is sufficiently high that the soil gas concentration calculated
from equation 11 exceeds that calculated from the vapor pressure of the
compound.
The vapor concentration in the soil pores cannot exceed the
saturation vapor concentration. Therefore,
A-12
-------
- pMW
where:
Csv = Saturation vapor concentration (g/m3);
p = Vapor pressure (atm);
MW = Molecular weight (g/gmole);
R = Gas constant (atm-m3/gnu>le-°K); and
T = Temperature (°K).
The chemical concentration in soil corresponding to the saturation point
vapor concentration can be estimated from the knowledge of vapor
pressure and air-soil partitioning in accordance with Equation 11:
c = —C (17)
C ^ \-i-' >
where:
Css » soil concentration corresponding to the saturation
vapor concentration (ng/g)
For soil concentrations greater than Css, chemical vapor emission rates
can be estimated by assuming that at any time the soil and chemical
approach their steady-state concentrations, (H/Kd)(Css/z):
ff C
H -SS. (18)
dz Kd
and a linear soil concentration profile exists in the layer of chemical-
depleted soil (USEPA 1986). Based on the assumed linear soil concentra-
tion profile and a simple mass balance, the layer of chemical-depleted
soil is:
. 2£*
: 12 T-T, (19)
— ' "^-C - C ) - —C6 - —-^-C
j ss so' «} L at ~> v ss
1 44 AJ
A-13
-------
Substituting Equation 18 into Equation 19 and integrating the resulting
equation over the time interval 0 to t, gives:
z =
-£-ea) + cLHBa} + Pb(cso - css)
The chemical vapor emission rate under saturation condition is:
dc. HCSO
HA = D--^ = DG 2£
A G dz G K^z
or
/ n • /** \ / O « /"* _L A /"• A ^* _i_Oy^*t7«-N*/^ 1
\DG Cas> ^2PbCso + VaCss -fT + ^mCLH Pb Css>
—^ , at z =0 (20)
If 9m = 0, 9T = 8a, and DG = 9a4/3Dai, Equation 20 is identical to the
solution developed by the USEPA (1986). The current equation allows
estimates based on actual soil moisture information rather than assuming
completely dry soil as would be the case using the above assumptions.
A.2 AIRBORNE POLLUTANTS
Under appropriate meteorological conditions, airborne pollutants
from many sources, including a Superfund site, may simultaneously impact
the indoor air quality of structures proximate to a superfund site.
Considerable guidance currently exists for the selection and use of
appropriate emission and dispersion models. Table A-l lists the
principal sources of information. Additional assistance may be obtained
from the EPA Office of Air Quality Planning and Standards and from
modelers in each of the 10 EPA Regional offices.
A-14
-------
c_D
Z
i— •
_J
LU
O
O
J£
TABLE A-1
ES AIRBORNE POLLUTANT
INFORMATION SOURC
LU
O
Z
LU
rr
LU
LL
LU
CC
!•• '; DESCRIPTION
'•'.:• .
'- -.":- •
-":" -V
:';';; :
.
•"-- •
LU
H
r-
^_
o
o
en
CO
A
O
in
•«•
<
Q.
UJ
CO
O
tho conduct of Air Pathway'' Analys
1
CD
£
1
o
o
a
2
o
c
o
o>
CO
§
:c
«
O
CO
11
CO 'fS
a. a
u. O
< ^
NTGS Series:
Volume 1, Application of
Analysis for Superfund Ap
CM
o
o
en
CO
A
0
in
^l-
<
CL
UJ
CO
«
"co
T3
=
C
o
0
c
3
C
Q
£
o
ring and modeling techniques for c
Provides descriptions of monito
w
c
UJ
w
<
«
c
"5
CO
CD
m
NTGS Series:
Volume II, Estimation of
sions at Superfund Sites
CO
o
0
en
op
A
o
in
•*
<
Q.
UJ
c
Q
ing air emissions from remedial act
Provides procedures for estimat
CO
o
ir Emissions
uperfund Sit
< to
NTGS Series:
Volume III, Estimation of
from Cleanup Activities at
,_
o
o
co-
co
,1
o
t
in
<
Q.
111
m
o
*co
"O
i
g techniques for common uncontro
Provides description of modelini
nent Manua
CO
|| Superfund Exposure Asses
pr
r~
CM
CD
CO
r^
CM
O
in
-
Screening Procedures for E
Quality Impact of Stationa
CO .£>
'
S °°
UJ rn
•T "?
< «
0. O
uj in
. . . . ^
r- a CM 1.
_ CM _ <
0 O ° uj
> 0 > *"
"3
•o
o
O
W
1
•*-
Provides a detailed discussion o
c
o
'co
o
0.
CO
s
0
CO
Industrial Source Complex
Model User's Guide
n
0
° §
en *
'J- •*
0 "
in ^
•- <
•*— /y
^ i\
o
>w
•i E
s e
a e
!*
*" -o
CO CO
Tiaximum short-term concentration
sources of toxic releases. Downio
PC-based model for estimating i
from Superfund sites and other
below).
•—
ning Toxic /
o
01
TSCREEN:A Model for Scr
Pollutant Concentration
fy
CM
O
CN
CO O
co ¥•
en *
o o
«— m
? 6
< en
Q_ CQ
UJ O-
o "
-c E
** 0
w 3
E o
» n —
— CL «
•9 •- f
Poo
°- » E
C? §0
'•S S 52
•32 —
0 ~ X
I I
J»5
S) — O
0 0 0
"•5. t: 3
S- o o
i model. The FDM is particularly a(
spears to be a better predictor of p
ind deposition than the Industrial si
S (see below).
Provides guidance on use of the
involve particle deposition. It fl|
involving fugitive dust sources «
Downloadable from SCRAM BB!
o
•o
o
to
Q
o
>
'&
£
o
jz
o
o
•o
°5
0
JO
o
CO
3
o
CM
O
en co
co in
4 ^
^- CN
o •*
in .-
T 6
< CO
Q_ CD
UJ O.
•o
C
rt-terrn area source models (FDM,
models (ISCLT, COM, VALLEY) >
Analyzes and compares five she
three long-term (rector average)
.2 CO
CO fi
E co
ll
O 3
m •*—
Area Source Dispersion Al(
|| Emission Sources at Super
CO
•— {«^
"O O ^
t: Z *"**
S »H>
00 S S
— O ^
•£ •£ s
0 °- rt\
P qi O)
"5 "o !T
CQ H S
-o >
O :=
C
-------
A.3 RELATING EMISSIONS TO INDOOR AIR CONCENTRATIONS
In the case studies presented in Appendix C, a wide range of
assumptions were made for estimating the percentage of the contaminant
flux that actually entered the building. In this section we present
estimating techniques based on both empirical relationships developed
from monitoring data and models. The models presented here are analytic
rather than numerical. Numeric models are also available and have been
used for modeling radon intrusion.lt should also be noted that inade-
quate field data currently exists to validate the modeling connecting
soil gas flux rate to indoor air concentrations.
A.3.1 Empirical Relationship Based on Radon Data<2)
Soil is believed to be the dominant source of radon in indoor air.
In the United States, the mean indoor radon concentration in the living
space of single-family dwellings is believed to be about
55 Bq/m3.<10'11'12) This conclusion is based on measurements in 1,270
homes in which there was no basis for expecting elevated indoor concen-
trations.
The mean concentration of radon in basements is about twice the
mean value for above-ground living spaces.(13'U) Alter and Oswald(13>
reported 9,000 long-term (1-month to 1-year) measurements in basements
yielding an average concentration of 520 Bq/m3 and 12,000 nonbasement
indoor measurements yielding an average concentration of 260 Bq/m3.
Cohen (14> reported an average concentration of 300Bq/m3 from 10,000
short-term (2 days to 1 week) measurements in basements and an average
concentration of 135 Bq/m3 from 34,000 short-term measurements in living
spaces. Although these data show higher indoor concentrations than are
believed to exist for the population as a whole because of the large
number of measurements made in houses with suspected elevated indoor
concentrations <12), the data should be reliable as an indicator of
relative basement/living space concentrations.
Cohen<14> reported the following data from annual average indoor
radon measurements in the living spaces of houses across the country:
A-16
-------
Substructure Type
Basement
Slab-on-grada
Crawl space
Number of Houses
266
85
84
Average"2Rn cone.
In Living Space
-------
It is important to understand that the soil gas concentration
referred to here is that found or calculated for the immediate vicinity
of the building (defined by Nazaroff, Jury, and Rogers (2) as a volume
extending 2 meters laterally from the basement walls and 1 meter below
the floor). It is not the soil gas concentration that would be calcu-
lated from Equation 1 of this Appendix.
A.3.2 Typical Simplified Assumptions Used
Estimations of indoor air concentrations are typically based on:
where,
Cjn = indoor air concentration
E = contaminant infiltration rate
Q = building ventilation rate
The assumptions used to derive both E and Q are important in
arriving at a realistic value of Cin. Note, for example, that the
equation predicts that as the building ventilation rate approaches zero
(minimum is the soil gas intrusion rate unless the building is complete-
ly sealed in which case the expression is indeterminant), the indoor air
concentration approaches the concentration in the soil gas.
As a first approximation, some modelers assume that soil gas
enters only by diffusion into the building and that
E = J»A*F (22)
where,
J = Contaminant flux estimated from source model (g/m2s)
(See Section 3.1)
A = Area of building floor (m2)
F = Fraction of floor through which soil gas can enter.
and
A-18
-------
Q = ACH .y (23;
3600
where,
ACH = building air changes per hour
V = volume of building (m3)
Mueller et al(15> reports that typical ACH for single family
residences range from 0.5 to 1.5, with new or retrofitted energy-
efficient structures generally ranging from 0.5 to 0.8 but with some as
low as 0.2. For general case application, choosing any reasonable value
in this range has only a modest effect on the final predicted indoor
contaminant concentration and is easily within the likely error for the
contaminant flux rate.
As a first assumption for E, it is frequently assumed that for
contaminant entry into buildings with dirt floors or ventilated crawl
spaces, a reasonable conservative assumption is that F = 1 (ie, all soil
gas enters the building). Measured data for radon(16) indicates 0.7 < F
> 1.0 for single family homes with ventilated crawl spaces.
For buildings with slab floors, a typical assumption mode is that
F = 0.001. Data used to support this is obtained from Carlos et al (17>,
who reports that the average California home has an open area of 2- to
10-cm2 per m2 of floor space and Grumund et al <18> who concluded that
the area of cracks to total floor space is 0.01 to 0.1 percent. Some
modelers assume soil gas enters only through an assumed 0.5 cm-wide
crack around the building perimeter at the slab/wall interface (note
that this approach results in decreasing F as the floor area increases).
The technical literature does not support the use of the assumptions
based on percent cracked area of the floor for the attenuation and their
use is not recommended. For screening purposes, it is recommended that
F be set to 1 in all cases.
With these assumptions, it is a simple matter to calculate Cjn
using the flux rate calculated by any method from Section A.I. Models
A-19
-------
for the more complex case in which pressure coupling to the building
must be considered are given in the following section.
A.3.3 Incorporating Building Impacts on Infiltration Rate
It is well recognized that buildings can develop negative pres-
sures (relative to ambient pressure) as a result of temperature differ-
ences and wind effects. The American society of Heating, Refrigerating,
and Air-Conditioning Engineering (ASHRAE) adopted equation 24
calculate this pressure difference, AP:
(19)
to
AP =
AP,
(24)
where,
PV
w' L
N
and P.,
where,
V
P
h
*K
static pressure over the building, Pa
pressure difference due to thermal gradient, Pa
area of building on windward and leeward side,
m2
empirical exponent = 0.65 (Wadden & Scheff,
1983}
0.6008 V2, V = wind velocity in meters per
second;
0.0342 Ph (Tr TJ/V,
wind velocity, M/S
atmospheric pressure, Pa
distance from neutral pressure plane, m
outdoor and indoor temperatures, respectively,
Evaluation indicates that pressure differences due to thermal
gradients are generally small (eg. about 1.5 Pa for a 25'F difference)
and that wind effects dominate in equation 24. Typical values of AP are
A-20
-------
1 to 10 Pa. It should also be recognized that there are many other
potential sources in building that can lead to negative (and, in some
cases, positive) pressures. Among those resulting in negative pressures
are bathroom and kitchen fans, attic fans, forced air combustion devices
(eg., oil and gas fired furnaces), fireplaces, and leaking air supply
ducts located in crawl spaces. Systems resulting in positive pressures
are typically found in certain types of commercial buildings, however,
it should be recognized that leaking air return ducts located in crawl
spaces or attics may cause the indoor pressure to be higher than
Equation 24 would predict.
Soil-gas entry due to depressurization of basements and crawl -
spaces has been experimentally demonstrated by a number of re-
searchers16'18'19'20. The current level of understanding is that both
diffusion and convection contribute to vapor intrusion and specific site
characteristics determine the significance of each. A number of
researchers <16.18«19.2°.21> nave attempted to model the transport for
radon. Johnson and Ettinger22 have adapted this work and extended it to
the case of chemical vapors. This model is described in this section.
For details of the derivation one should consult the original paper.
There are three basic derived equations for this model. They
correspond to:
1) Contaminant source is infinite (with respect to modeling
time of interest) and vapor infiltration is through
cracks/opening in the foundation
2) Soil gas transport into the building is substantially higher
through relatively permeable walls (eg.,concrete block
construction below grade) than through foundation cracks and
openings.
3) Contaminant is located near the building and decreases with
time (ie, this case provides an average when the contaminant
source would become depleted over the averaging period
used).
The model equation corresponding to case 1 above (section A.3.3.1)
is probably the most useful for general application. In this case it
would be assumed that the distance from the source to the building does
A-21
-------
not change with time (ie., Lt in Equation 25 is a constant) and does not
change in composition over the time of interest for the calculation.
The equation would be used for structures with crawl spaces and slab
floor construction with solid (eg., poured concrete) below grade walls.
The modeling equation corresponding to case 2 above (section
A.3.3.2) is applicable to construction which uses hollow concrete block
construction below grade (including cases where the large voids are
filled with poured concrete). This is a fairly common construction
technique for houses with basements. Soil gas entry into homes with
this type construction is discussed in EPA's guidance for radon reduc-
tion techniques'24'. The model formulation presented assumes an infinite
source at constant depth as does the first equation.
The model equation corresponding to case 3 above (section A.3.3.3)
is applicable to cases where a long term average concentration is needed
and it is reasonably expected that significant changes will occur in the
mass of chemical in the soil (and, therefore, the source to building
distance will change significantly) over the time period of interest.
The model should not be applied when the contaminant is near (within 2
meters) and intrusion into the building is dominated by convective flow.
One of the other two model equations should be used to make estimates
for the time this condition exists. This third model equation may then
be applied for the time period that intrusive flow is diffusion dominat-
ed.
The models below require an estimate of the source vapor concen-
tration, Csource< because what is calculated is the ratio of the concen-
tration inside the building to the soil gas concentration at the source
(ie the "«" in the equations). Two main approaches are used in vapor
transport modeling; in the first Csource is assumed to be proportional to
the residual level in the soil, and in the second Csource is independent
of the residual level, but is a function of composition. The former is
applicable in the limit of "low" residual levels where compounds are
sorbed to the soil, dissolved in the soil moisture, and present in the
vapor space; the latter is applicable for "high" residual levels where
free-phase liquid or precipitate is trapped in the soil interstices. It
is important to note that if one chooses an incorrect model for predict-
A-22
-------
ing Csource, then it is possible to over- or under-predict the actual
Csource va^ue by orders of magnitude. Section A.I provides a limited
discussion of models to estimate Csource for both cases.
In the following equations, the symbol "«" (alpha) is the ratio of
the contaminant concentration in the indoor air to the soil gas concen-
tration at the source (ie., oc = C
building' source
A.3.3.1 Model for Infinite Source, Vapor Infiltrating Through Cracks
The derived equation is:
x exp
| QHlllLcitC*
exp
«i^) . \ *"*• \ . [j^l Lp (_£«&-)_
:"*A««eJ [ObalUi^rl (Q.oilLT\ [ (DC"C*Acr«*)
(25)
where,
a =
DTeff -
AD
building
LT
Qsoil
crack
Dcrack
bui Iding' source'
overall effective diffusion coefficient, cm2/s
cross-sectional area through which contaminants
may pass (can be approximated by area of floor
and below grade walls), cm2
building ventilation rate, cm3/s
distance from contaminant source to building
foundation, cm
volumetric flow rate of soil gas into the build-
ing
thickness of foundation, cm
effective vapor-pressure diffusion coefficient
through the crack, cm2/s
area of cracks/openings through which vapors can
pass, cm2
Interesting features of the equation are:
A-23
-------
(a) (Qsoil Lcrack/DCraCk Acrack) " *• In this 1imit COnVBCtion is
the dominant transport mechanism through the basement (building) floor
and walls.
then,
neffA
L>T AB
effA
T Ag
If the source lies directly beneath the foundation (LT -* 0), then «-*
Qsoii/Qbuiiding' which 1S the proper result for convection-dominated
transport of a vapor stream with concentration Csource. If the source is
"far" from the basement (i.e., DTeff AB/QSOJI LT -*0), then transport is
limited by diffusion from the source to foundation, and « -* (DTEffAB/-
Qbuiiding ^T)- Note that these results are independent of the cracked area
of the floor and walls. This is because contaminant vapors are swept
into the building as fast as they are transported to the soil adjacent
to the floor and walls.
(b) (Qsoii Lcrack/DCrack Acrack) -0- In this limit diffusion 1s the
dominant transport mechanism through the basement floor and walls. When
Dreff AB Lcrack/DCrack Acrack LT * *» then diffusion through the floor and
walls is the rate limiting mechanism, and there is a vapor concentration
"buildup" below the building or basement.
(c) Q building "*^. This limit corresponds to a perfectly sealed
(no ambient air exchange) basement, Qsoil must also approach zero, and
the model predicts that « -»•!; this is, the indoor contaminant vapor
concentration approaches the contaminant vapor concentration in the soil
gas.
Sample calculations were presented for this equation using the
following parameters:
AB = 7m x 10m + 2(2m x 10m) = 138 m2 = 138 x 104 cm2
Lcrack - 6 in. = 15 cm
A-24
-------
^building = 7mxlOmx3mx0.5 volume exchanges/ h = 105 m3/h
= 2.9 x 104 cm3/s
Dair = 0.087 cm2/s (benzene)
DH2° = 0.087 1.0 x 10"5 cm2/s
H = 0.18 cm3 of H20/cm3 of air (benzene)
6m = 0.07 g of H20/cm3 of soil
£T = 0.38 cm3/cm3 of soil
Pb = 1.7 g/cm3
AP = 1.0 Pa = 10 g/cm-s2
and estimating Q soil from
Q.
soil
Hln[2ZcracJt/rcraeJt]
which is an analytical solution for flow to a cylinder of length Xcrack
and radius rcrack located a depth Zcrack below ground surface; this is an
idealized model for soil gas flow to cracks located at floor/wall seams.
ky is the soil permeability to vapor flow (cm2) and /t is the vapor
viscosity (g/cm-s). For this sample problem, Zcrack = 2 m, Xcrack is
taken to be the total floor/wall seam perimeter distance (34 m), and for
consistency rcrack is given by
rcrack ~ ^"B/*crack
where the ratio ;;= Acrack/AB, so that 0 < TJ< 1. For reference, T]= 0.01
corresponds to rcrack = 4.1 cm for the values of AB and Xcrack given above;
rcrack = 1 cm corresponds to rj = 0.0025.
The results in graphical form are:
A-25
-------
10 l« 10 10 10 10 10
Ifl".
Pcnne*bility(an:l
10"
LT =o
#"" W"* I04 M* 104
PtnnubQity [an1]
LT = 100 cm
n • O-Oi
H.CLOOI
to4 io4 W w4 w*
Penne»Wiiy[cnizJ
Ly = 1000 CHI
For a contaminant source adjacent to the building (LT =0) , « is
proportional to the soil permeability to vapor flow .K at ^ »10"*cm2
"7
'6
(permeable soils). In this example, 0.001 < « <0.01 for IO" < K^ tlO
which happens to fall in the range of values typically reported for
radon studies (see section A. 3.1). The results also predict that, for
practical purposes, it can be concluded that the effect of crack size on
contaminant vapor intrusion rates will be relatively insignificant in
the limit of convective-dominated transport.
For LT - 100 cm and 1000 cm, the dependence of « on ^ is a
sigmoidal -shaped curve, where « becomes independent of ^ for both
"large" and "small" soil permeabilities. For less permeable soils, soil
gas flow rates are so low that vapor intrusion is governed entirely by
the relative rates of diffusion through the soil and foundation. As the
soil becomes more permeable, the "sweeping" of contaminant vapors into
the building by soil gas flow increases the intrusion rate. At some
point, however, the ability of the contaminant to diffuse from the
contaminant vapor source to the region of soil gas flow limits the rate
of contaminant vapor transport into the building. For highly permeable
A-26
-------
soils, therefore, ex becomes independent of the soil permeability to
vapor flow and only weakly dependent on foundation properties. « and
the intrusion rate become less dependent on foundation properties as the
distance to the contaminant source increases. This can be seen from the
convergence of the curves at both high and low values of the soil
permeability, and the 77 = 0.01 and 77 = 0.001 predictions as LT increas-
es. That is, the further the source is from the building, the less it
matters what the building foundation is. In the limit of infinite
separation between the source and the building, diffusion through soil
becomes the limiting transport mechanism and the building substructure
does not matter at all.
A.3.3.2 Modeling Using Relatively Permeable Foundation Walls
Garbesi and Sextro(20) conducted a field study of a building with
basement walls constructed of hollow concrete blocks which were subse-
quently backfilled with concrete and coated with asphalt sealant on the
exterior. No evidence of cracks at the wall-floor interface was ob-
served. Their experimental work showed good pressure coupling between
the basement and soil which they interpreted to imply the entire wall
area had a measurable permeability and soil gas could infiltrate over
this entire area.
The Johnson and Ettinger model for this case is:
a =
exp
D
'ff
exp
exp -
-l
(26)
This equation is similar to equation 25 except Acrac(c is replaced
by the area of the basement walls and floor, AB, Dcrack is replaced by
Df, the effective diffusion coefficient through the porous foundation
crack
floor and walls, and Lc is replaced by the foundation/wall thickness,
LF. While eqs 26 and 25 appear similar, they can predict quite differ-
ent results. Equation 26 is independent of the area of cracks/openings
A-27
-------
because intrusion is assumed to occur uniformly over the floor/wall
area. For a given Qsoil, therefore, the soil gas velocity through the
floor/walls is lower for the permeable floor/wall case. The impact of
this is that eq 26 may predict that transport through the foundation is
diffusion dominated, while for the same conditions equation 25 would
predict that it is convection dominated.
A.3.3.3 Model for Source Depletion Over Time
Equation 25 provides a screening estimate of indoor vapor concen-
trations, but does not account for depletion of the contaminant vapor
source. This is reasonable when short-term exposures are being estimat-
ed and it provides a conservative (upper bound) estimate for long-term
exposures. It is not appropriate, however, when more realistic long-
term exposure estimates are desired and it is unlikely that the source
will remain constant for a long period of time. This model formulation
considers the depletion due to transport.
In this approach it is recognized that the separation between
contaminant source and the building increases with time due to source
depletion. It is assumed that the rate at which a steady-state vapor
concentration profile is established is much greater that the rate at
which depletion occurs. Implicit in this approach is the assumption
that depletion occurs first from the layers of contaminant closest to
the building floor and walls, and the mass of contaminant incorporated
in the soil disappears, beginning at the edge closest to the building.
This is a reasonable assumption for diffusion-dominated transport to the
building-soil interface, but not valid for convection-dominated trans-
port from contaminated soil adjacent to a building floor.
For time periods less than the time required for total depletion
of the contaminant, the average emission rate is given by:
[(p2
The corresponding long-term average attenuation coefficient <«> is then
A-28
-------
(a) = H^AtfeAfl I £j_ [(p2 + 2t|rt)i/2 _ p] (28)
^building^source^ \ ^"cj
While this approach is more sophisticated than equation 25,
increasing sophistication usually increases the amount of site-specific
information required. More sophisticated screening models are usually
also based on additional assumptions, and one must be careful to ensure
that these assumptions are valid for specific site characteristics.
For equation 27 and 28, the following definitions apply:
pb = soil density, g/cm3
CR = average residual contaminant level in soil, g/g
AHC = thickness of the contaminant layer, cm
T = time
L°T = initial contaminant-building floor separation, cm
= DgC.
source'
P =
exp -
A.3.4 Impact of Contaminated Outdoor Air
All of the preceding discussion in this chapter has assumed that
the outdoor air is contaminant free. To be conservative, we have also
ignored other building characteristics that would affect the estimated
indoor pollutant concentration. In the following, we attempt to remedy
both assumptions. This discussion is based on modeling equations
presented in EPA/400/3-91/003, "Introduction to Indoor Air Quality - A
Reference Manual." (This manual is a rich source of information.)
A-29
-------
The generalized mass balance equation for indoor air is:
dC,. = (1 - Fb) i/C0 + S - im/C,. - \_ - flFC,. (29)
dt kV kV kV
where,
C,. = indoor concentration (mass/volume);
Fb = fraction of outdoor concentration intercepted by the build-
ing envelope and not mixing with indoor air (dimensionless
fraction);
v = air exchange rate (I/time);
C0 = outdoor concentration (mass/volume);
S = indoor source generation rate (mass/time);
V = actual indoor volume;
kV = effective indoor volume where k is a dimensionless fraction;
A = decay rate (mass/time);
q = flow rate through air cleaning device (volume/time);
F = efficiency of the air cleaning device (dimensionless frac-
tion);
T = time; and
m = mixing factor.
Fb is included because not all outdoor air contaminants that move
into a structure reach the inside. Some fraction, Fb, is intercepted by
the cracks and crevices in the building envelope which decreases the
amount that actually reaches the indoor air.
KV, the effective indoor volume, is included because the actual
indoor volume available for contaminants dispersal depends on the degree
of air circulation. K is less than one when there is no forced mixing
and the degree of circulation depends on thermal gradients indoors.
X, the decay rate, is included to allow accounting for losses due
to indoor chemical reactions, and adsorption of contaminants on indoor
surfaces. Removal of contaminants by mechanical means through air
cleaning devices is dependent on flow through the device and F, the
fractional efficiency for removal of the particular contaminant.
A-30
-------
The mixing factor m is the ratio of the concentration of the
exiting air to the concentration of the indoor air. When the two
concentrations are the same, the air is said to be completely mixed, and
m equals one. When contaminants are exhausted directly from their
source, m will be >1. The complete mixing assumption is often used to
estimate average concentrations over time periods greater than 1 hour.
Equation 29 was developed for ambient air infiltration and
additional contaminant sources and sinks inside the structure. However,
it appears that, with little error, since soil gas comprises a very
small fraction of the indoor air, the source term S could be replaced
with the equivalent term from soil gas modeling.
For the simple case where X, F, and Fb equal zero and m = 1,
equation 29 reduces to :
dt
- c.)
kV
and the equilibrium concentration (when dC,./dt = 0) would be given by:
C, = C0 + S .
ki/V (30)
then if C0 = 0 and K = 1, equation 30 is identical to equation 21.
A.4 SOURCES OF DATA FOR MODEL PARAMETERS
The models presented in this Appendix require a variety of input
parameters for chemical and soil properties. Listed below are sources of
information for many of the more important parameters.
Model Parameter
Henry's Law constant
Diffusion Coefficient
Soil Porosity
Reference Number(s)
1, 25, 26, 27, 28
1, 5, 25, 29, 31
5, 30
A-31
-------
REFERENCES
1 Air/Superfund National Technical Guidance Study Series, Volume
II - Estimation of Baseline Air Emissions at Superfund Sites.
EPA-450/l-89-002a, August 1990.
2 Nazaroff, W.W., W.A. Jury, V.C. Rogers, "Enclosed-Space Vapor
Models - Technical Panel Report," 14 February 1991. Prepared for
Department of the Army as a review of the Rocky Mountain Arsenal
Human Health Exposure Assessment prepared by Ebasco Services, Inc.
under contract DAAA15-88-D-024, September, 1990.
3 Karickhoff, S.W., D.S. Brown and T.A. Scott. 1979. Sorption of
Hydrophobic Pollutants on Natural Sediments and Soils. Water
Research 13:241-248.
4 Karickhoff, S.W. 1981. Semi-Empirical Estimation of Sorption of
Hydrophobic Pollutants on Natural Sediments and Soils. Chemo-
sphere 10:833-846.
5 Superfund Exposure Assessment Manual, EPA 540/1-88/001, U.S. EPA
Office of Remedial Response, Washington, D.C.. April 1988.
6 The section is adapted from Appendix X of the Draft Remedial
Investigation Report for the Redwing Carriers, Inc-Saraland
Apartments Site, Alabama, February 1992.
7 Jury, W.A., W.F. Spencer, and W.J. Farmer. 1983. Behavior assess-
ment model for trace organics in soil. I. Model description. J.
Environ. Qual. 12(4):558-564.
8 U.S. Environmental Protection Agency (USEPA). Office of Research
and Development. Exposure Assessment Group. 1986. Development of
advisory levels for poly-chlorinated biphenyl (PCB) cleanup.
Washington, D.C. EPA/600/6-86-002.
9 Bomberger, D.C., J.L. Gwinn, W.R. Mabey, D. Tuse and T.W. Chou.
1982. Environmental Fate and Transport at the Terrestrial-Atmo-
spheric Interface. Symposium on models for predicting fate of
chemicals in the environment. American Chemical Society, Division
of Pesticide Chemistry 184th National Meeting, Kansas City, MO.
10 Nero, A.V., M.B. Schwehr, W.W. Nazaroff, and K.L. Revzan (1986).
Distribution of Airborne Radon-222 Concentrations in U.S. Homes,
Science. 234, 992-997.
11. Cohen, B.L. (1986). A national survey of 222Rn in U.S. homes and
correlating factors, Health Physics. 51, 175-183.
A-32
-------
12. Nero, A.J. Gadgil, W.W. Nazaroff, and K.L. Revzan (1990). "Indoor
Radon and Decay Products: Concentrations, Causes, and Control
Strategies", Report DOE/ER-0480P, U.S. Department of Energy,
Office of Health and Environmental Research, Washington, D.C.,
November 1990.
13. Alter, H.W. and R.A. Oswald (1987). Nationwide distribution of
indoor radon measurements: A preliminary data base, Journal of
the Air Pollution Control Association. 37, 227-231.
14. Cohen, B.L. (1987). "Radon: A Homeowner's Guide to Detection and
Control", Consumer Reports Books, Consumers Union, Mount Vernon,
New York, p. 61.
15 Mueller Associates, Syscon Co., and Brookhaven National Laborato-
ry, "Handbook of Radon in Buildings," Hemisphere Publishing Co.,
New York, NY., 1988.
16 Nazaroff, W. W. and S. M. Doyle, "Radon Entry into Homes Having a
Crawl space," Health Physics. 48: 265-281.
17 American Society of Heating, Refrigerating and Air Conditioning
Engineering (ASHRAE), ASHRAE Handbook: 1981 Fundamentals. New
York, 1981.
18 Turk, B. H..; Prill, R. J.; Fisk, W. J.; Grimsrud, D. T.; Moed,
B. A.; Sextro, R. G. In Proceedings of the 79th Annual Meeting of
the Air Pollution Control Association; Minneapolis, MN; Air
Pollution Control Association: Pittsburgh, PA, 1986; Paper No.
86-43.2.
19 Nazaroff, W. W.; Lewis, S. R.; Doyle, S. M.; Moed, B. A.; Nero,
A. V., Environmental Science and Technology. 1987, 21, 459-466.
20 Garbesi, Karina, and R. G. Sextro, "Modeling and Field Evidence of
Pressure-Driven Entry of Soil Gas into a House through Permeable
Below-Grade Walls," Environmental Science and Technology. 23:
1481-1487.
21 Loureiro, C. 0.; Abriola, L. M.; Martin, J. E.; Sextro, R. G.
Environmental Science and Technology. 1990, 24: 1338-1348.
22 Johnson, P. C., and R. A. Ettinger, "Heuristic Model for Predict-
ing the Intrusion Rate of Contaminant Vapors Into Buildings,"
Environmental Science and Technology. 1991, 25: 1445-1452.
23 U. S. EPA, "Introduction to Indoor Air Quality - A Reference
Manual," EPA/400/3-91/003, July 1991.
24 U. S. EPA,"Radon Reduction Techniques for Detached Houses,
Technical Guidance," EPA/625/5-87/019, January 1988.
A-33
-------
25 U. S. EPA, "Hazardous Waste Treatment, And Disposal Facilities
(TSDF) -- Air Emission Models. EPA-450/3/87-026.
26 Thibodeau, L.J., and S.T. Hwang, "Landfarming of Petroleum
Wastes - Modeling the Air Emission Problem. Environmental
Progress, l(l):42-46, 1982.
27 Macay, D., and P.J. LeinOnen, "Rate of Evaporation of Low-
Solubility Contaminants from Water Bodies to Atmosphere.
Environmental Science and Technology, 9(13): 1178-1180,
1975.
28 Mackay, D. and W.Y. Shiu, "A review of Henry's Law Constants
for Chemicals of Environmental Interest". Journal of Physi-
cal Chemistry Reference Data,10(4):1175-1199, 1981.
29 Perry, R.H., and C.H. Chilton, Chemical Engineer's Handbook.
Sixth Edition, McGraw-Hill Book Company, Inc. New York, NY.
30 Brady, N. C. The Nature and Properties Of Soils. Eighth
Edition, McMillian Publishing Company, Inc., New York.
31 Lyman, W.J., Reehl, W. F., Rosenblatt, D. H., Handbook of
Chemical Property Estimation Methods. McGraw-Hill Book
Company, Inc. New York. 1982
A-34
-------
APPENDIX B. MONITORING METHODS
-------
TABLE OF CONTENTS
APPENDIX B. MONITORING METHODS
B.I Methods for Use Outside Buildings B-l
B.I.I Organics in Ambient Air B-l
B.I.2 Particulars B-4
B.I.3 Organics in Soil Gases B-14
B.I.4 Soil Permeability to Gas Flow B-16
B.2 Methods for Use Inside Buildings B-17
B.2.1 Pollutant Measurement Methods B-17
B.2.2 Air Infiltration Rates B-19
B.3 Indoor Air Pollutants B-21
-------
APPENDIX B. MONITORING METHODS
In this section information is provided on monitoring methods
relevant to an indoor air impact assessment and types of compounds
typically found in indoor environments. These methods include those
applicable to measuring pollutants outside the building which may enter,
as well as indoor monitoring methods. The general applicability and
procedures are described. References are provided for details of the
methods.
The information provides only a summary of the relevant methods.
Successful application requires skilled professionals and rigorous QA/QC
programs.
B.I METHODS FOR USE OUTSIDE BUILDINGS
Pollutants outside a building may enter through many openings in
the structure. This includes organics and particulates in the ambient
air and gases in the soil in the immediate vicinity of the building.
Outdoor measurements needed to obtain data required in the various
approaches of Section 2 include:
• Temperature
• Wind velocity
• Barometric pressure
• Organics concentrations in ambient air
• Particulate matter in ambient air
• Chemical concentrations in soil gases
• Soil permeability to gas flow
Methods for the latter four are described here. It should be noted
that in some cases it may be preferable to use an indoor method outdoors
to achieve consistency in measurements.
B.I.I Organics in Ambient Air
Because buildings have a relatively high rate of air exchange with
the outside air (air exchange rates of 0.5 to 1 building volume per hour
are typical for a building with windows and doors closed), pollutant
concentrations in ambient air can be significant in an assessment.
B-l
-------
Ambient air may contain many pollutants not associated with the nearby
Superfund site (e.g., auto exhaust, lawn chemicals, industrial emis-
sions, etc.). It is important to recognize that, unless an assessment
is to be made for background risks also, the sampling method and
analytical finish need only address known or suspected site related
pollutants.
Whichever of the following sampling methods is selected, the
equipment should be located so as to minimize collection of pollutants
from nearby sources such as automobiles, gasoline powered equipment, and
oil storage tanks. It should be located on the windward side of the
building away from windshields such as trees and bushes. The intake
should be about 5 feet above ground level (approximately at the midpoint
of the ground level story of the building) and 5 to 15 feet away from
the building.
Ambient air sampling should begin between 1- and 2-hours before
indoor air sampling begins. This is recommended because concentration
fluctuations during this period will be reflected inside the building
for several hours since the building effectively acts as an equalization
chamber. Likewise, for monitoring periods exceeding 4 hours, consider-
ation may be given to terminating ambient air sampling up to 30 minutes
before terminating indoor sampling.
Very complex monitoring situations can occur if ambient air
transport of Superfund site emissions is of concern. In these cases
additional sources emitting compounds of interest may exist upwind of
the Superfund site as well as between it and the building being investi-
gated. In these cases multiple monitoring stations and site specific
modeling could be required to distinguish the impact of the site from
the other sources. Information on various monitoring techniques for
these complex situations can be found in Section 4.3 of EPA 450/1-89-
002a (volume II of the Air/Superfund NTGS Series).
As a special case, it should be recognized that if the ambient air
pathway is the only pathway of concern for Superfund site impact from
non-particulates on the indoor air quality, there is no reason to
conduct indoor monitoring. That is because the average long-term
concentration of Superfund site related chemicals (except possibly
B-2
-------
participates) would be the same as the outdoor average at the building
location.
B.I. 1.1 Methods
In general EPA Methods TO-1, TO-14, IP-1A, or IP-IB will encompass
the types of organics normally associated with a Superfund site.
However, certain specific types of compounds (e.g., pesticides) may
require specific methods. All "TO-x" methods can be found in "Compendi-
um of Methods for the Determination of Toxic Organic Compounds in
Ambient Air," EPA 600/4-84-041 and supplements EPA 600/4-87-006 and EPA
600/4-89-018. All "IP-x" methods can be found in "Compendium of Methods
for the Determination of Air Pollutants in Indoor Air," EPA 600/4-90-
010. It should be noted that the methods in the latter reference are
not currently certified and should not be regarded as officially
recommended or endorsed by EPA. One source of these and other methods
is the U.S. EPA Air Methods Database prepared by the EPA Environmental
Response Team. Information about this PC-based software package can be
obtained by calling 1-800-999-6990.
TO-14 and IP-1A Methods
Both of these methods use pre-evacuated SUMMA® passivated stainless
steel containers to collect whole air samples. The canisters can be -
used to collect grab samples or time integrated samples. For most
efforts related to indoor air monitoring, time integrated samples are
needed.
For time integrated samples a flow restricting device must be used
to control sampling rate. A mass flow controller is preferred for this
use because a critical orifice flow restrictor results in decreasing
flow rate into the canister as the canister pressure approaches atmo-
spheric. Greater sampling volumes can be obtained by using a metal
bellows type pump to boost final canister pressures to as high as 30
psi.
This method allows convenient integration of samples over a
specified time period, remote sampling and central analysis, ease of
shipping and storing samples, unattended sample collection, and collec-
tion of sufficient volumes to allow assessment of measurement precision
and/or analysis of samples by several analytical systems. Contamination
B-3
-------
is a critical issue with canister-based systems because the canister is
the last element in the sampling train. Thus, as with all sampling
systems, care must be used in the selection, cleaning, and handling of
both canisters and associated sampling apparatus.
Although the method has been used for many VOCs and SVOCs (i.e.,
compounds with vapor pressures greater than 10"7 mm Hg at 25'C), it has
been validated for less than 40 compounds (See Table B.I). Furthermore,
minimal documentation is currently available demonstrating VOC stability
in subatmospheric pressure canisters.
One of the primary advantages of the canister methods is that the
sample volume that can be collected is not limited by the "break
through" capacity as are methods using solid sorbents. This is particu-
larly advantageous when sampling in areas of unknown concentrations or
where concentrations are varying during sampling.
Canister samples are analyzed using a high resolution gas chromato-
graph coupled to one or more appropriate detectors. Although non-
specific detectors such as flame ionization (FID), electron capture
(ECO), or photoionization (PID) may be less expensive per analysis and
in some cases more sensitive than specific detectors, they suffer from
variable specificity and sensitivity, non-positive compound identifica-
tions, and lack of resolution of co-eluting compounds. It is strongly
recommended that specific detectors be used for positive identification
and primary quantification. The recommended detectors are mass spec-
trometer operating in either the selected ion monitoring mode (for use
when a specific set of compounds are to be determined) or the SCAN mode
for identification of all compounds.
Methods TO-1 and IP-IB
These methods are based on the collection of VOCs on Tenax® solid
adsorbent. Integrated sampling over periods up to 12 hours can be
performed. The air to be sampled is drawn through the Tenax® loaded
cartridge by small, portable pumps. As with any adsorption method,
knowledge of flow rate and volume sampled are important data.
B-4
-------
TABLE B.I
VOLATILE ORGANIC COMPOUND DATA SHEET8
COMPOUND (SYNONYM)
Freon 12 (Oichlorodifluororaethane)
Methyl chloride {Chloromethane)
Freon 114 (1.2-Oichloro-l.l .2,2-
tetrafluoroethane)
Vinyl chloride (Chloroethylene)
Methyl bromide (Bromomethane)
Ethyl chloride (Cnloroethane)
Freon 11 (Trichlorofluoromethane)
Vtnylldene chloride (1,1-Oichloroethene)
Dichlororaethane (Methylene chloride)
Freon 113 (1.1.2-TrJchloro-1.2,2-
trifluoroethane)'
1,1-Olchloroethane (Ethylidene chloride)
cis-l,2-0ich1oroethylene
Chloroform (Trichloronethane)
1,2-Olchloroethane (Ethylene dichloride)
Methyl chloroform (1,1.1-Trichloroethane)
Benzene (Cyclohexatriene)
Carbon tetrachloride (Tetrachlorcmethane)
1,2-Oichloropropane (Propylene
dichloride)
Trichloroe thy lene (Trichloroethene)
cl s-l,3-01chloropropene (cis-1,3-
dichloropropylene)
trans-l.3-D1chloropropene (cls-1,3-
Oichloropropylene)
1,1.2-Trichloroethane (Vinyl trichloride)
Toluene (Methyl benzene)
1,2-Olbromoethane (Ethylene dibromide)
Tetrachloroe thy lene (Perch loroe thy lene)
Chlorobenzene (Phenyl chloride)
Ethyl benzene
n-Xylene (1,3-Oimetnylbenzene)
•p-Xylene (1,4-Oimethylxylene)
Styrene (Vinyl benzene)
•1,1,2,2-Tetrachloroethane
o-Xylene (1.2-Oiraethylbenzene)
.1.3,5-Trimethylbenzene (Mesitylene)
1,2,4-Trinethylbenzene (Pseudocumene)
.Bi-01 Chlorobenzene (1,3-Oichlorobenzene)
Benzyl chloride («-Chlorotoluene)
o-01chlorobenzene (1,2-Dichlorobenzene)
p-Oichlorobenzene (1,4-Olchlorobenzene)
1,2,4-Trichlorobenzene
Hexachlorobutadiene (1.1,2,3.4,4-
Hexach1oro-l,3-butadiene)
FORMULA
C12CF2
CH3C1
C1CF2CC1F2
CH2-CHC1
CH38r
CH3CH2C1
CC13F
C2H2C12
CH2C12
CF2C1CC12F
CH3CHC12
CHC1-CHC1
CHC13
C1CH2CH2C1
CH3CC13
CC14
CH3CHC1CH2CI
C1CH«CC12
CH3CC1-CHC1
C1CH2CH-CHC1
CH2C1CHC12
BrCH2CH2Br
C12C-CC12
C6HSC1
1.3-(CH3)2C6H4
1,4-(CH3)2C6H4
CHC12CHC12
1.3.S-(CH3)3C6H6
1.2,4-(CH3)3C6H6
1.3-C12C6H4
1,2-C12C6H4
1.4-C12C6H4
1,2.4-C13C6H3
MOLECULAR
WEIGHT
120.91
50.49
170.93
62.50
94.94
64.52
137.38
96.95
84.94
187:38
98.96
96.94
119.38
98.96
133.41
78.12
153.82
112.99
131.29
110.97
BOILING
POINT (*C)
-29.8
-24.2
4.1
-13.4
3.6
12.3
23.7
31.7
39.8
47.7
57.3
60.3
61.7
83.5
74.1
80.1
76.5
96.4
87
76
KLTING
POINT CC)
-158.0
-97.1
-94.0
-1538.0
-93.6
-136.4
-111.0
-122.5
-95.1
-36.4
-97.0
-80.5
-63.5
-35.3
-30.4
5.5
-23.0
-100.4
-73.0
CAS
NUMBER
74-87-3
75-01-4
74-83-9
75-00-3
75-35-4
75-09-2
74-34-3
67-66-3
107-06-2
71-55-6
71-43-2
56-23-5
78-87-5
79-01-6
110.97
112.0
133.41
92.15
187.88
165.83
112.56
106.17
106.17
106.17
104.16
167.85
106.17
120.20
120.20
147.01
126.59
147.01
147.01
181.45
113.8
110.6
131.3
121.1
132.0
136.2
139.1
138.3
145.2
146.2
144.4
164.7
169.3
173.0
173.3
180.5
174.0
213.5
•36.5
•95.0
9.8
•19.0
•45.6
•95.0
•47.9
13.3
•30.6
36.0 '
•25.2
•44.7
•43.8
24.7
39.0
17.0
53.1
17.0
79-00-5
108-88-3
106-93-4
127-18-4
108-90-7
100-41-4
100-42-5
79-34-5
108-67-8
95-63-6
541-73-1
100-44-7
95-50-1
106-46-7
120-82-1
a - EPA 600/4-90-010
B-5
-------
The method offers some advantages over the canister methods but has
significant limitations and overall is complex and difficult to use.
The primary, if not only, advantage is that the method has been validat-
ed for more compounds than the canister methods (See Table B.2). Note
that although small, portable (even personal sampling) pumps may be
used, mass flow controllers are recommended to ensure flow stability. A
10% deviation in initial and final flow rates may invalidate data
collected using cartridges with that pump.
Limitations of the method include:
• Breakthrough volumes of some compounds, such as vinyl chloride,
are very low. The lowest breakthrough volume limits the volume
of sample that can be collected. (See Table B.3).
• High benzene backgrounds from Tenax® are common
• Sensitive to high humidity
• Contamination with compounds of interest (e.g., benzene,
toluene, chloroform, methylene chloride, etc.) commonly encoun-
tered problem
• Variable desorption rates for compounds adsorbed on adsorbent
• Necessary to know approximate range of contaminate concentra-
tions
• Overall accuracy of method is unknown.
The analytical finish for this method is similar to that for the
canister method above. The primary differences are that the compounds
must be thermally desorbed from the Tenax® before injecting it into the
gas chromatograph and that each cartridge can be analyzed only once. It
is strongly recommended, therefore, that the mass spectrometer be
operated in the SCAN mode.
B-6
-------
TABLE B.2
COMPOUNDS IDENTIFIED AND QUANTIFIED BY AUTOMATED
GC-MS-DS PROCEDURE WITH TYPICAL DETECTION LIMITS IN
FULL SCAN MODE8
Quantitatlon Detection Limits
Compound Mass fm/z) fnq)
perfluorotoluene (internal standard)* 217 0.3
benzene* ' 78 2.6
methyl benzene* 91 2.0
1,2-dimethylbenzene* 106 0.5
1,3,5-trimethylbenzene 120 2.5
ethylbenzene* 91 1.6
ethylbenzene* 104 1.7
(1-methylethyl) benzene 105 1.1
butylbenzene 91 0.7
l-methyl-4-(l-methylethyl) benzene 119 4.0
chlorobenzene* 112 1.7
bromobenzene 156 14.1
1,2-dichlorobenzene* 146 12.4
l-ethenyl-4-chlorobenzene 138 2.0
trichloromethane . 83 2.7
tetrachloromethane* 82 2.1
bromochloromethane* 130 2.1
bromotrichloromethane* 163 1.6
dibromomethane* 174 4.5
tribromomethane* 171 8.5
1,1-dichloroethane* 63 5.7
1,2-dlchloroethane 62 3.8
1,1,1-trichloroethane* 99 1.7
1,1,2-trichloroethane* 85 2.1
1,1,1,2-tetrachloroethane 31 0.9
1,1,2,2-tetrachloroethane 83 6.5
pentachloroethane* 167 1.8
1,1-dichloroethane* 961 6.9
trichloroethene* 132 0.8
tetrachloroethene 166 2.6
bromoethane* .108 7.8
1,2-dibromoethane* 107 3.3
1-chloropropane* 42 1.7
2-chloropropane* 43 3.4
1,2-dichloropropane 63 4.0
1,3-dichloropropane 76 9.6
1,2,3-trichloropropane 753 4.7
l-bromo-3-chloropropane 158 1.6
3-chloro-l-propene 41 1.6
B-7
-------
Compound
1,2-dibromopropane*
2-chlorobutane
1,3-dichlorobutane
1,4-dichlorobutane
2-3-dichlorobutane*
l,4-dichloro-2-butane (cis)
3,4-dichloro-l-butane
tetrahydrofuran
1,4-dioxane
l-ch!oro-2,3-epoxypropane
2-chloroethoxyethene
benzaldehyde*
acetophenone
benzonitrile
ISOMER GROUPS
TABLE B.2 -- continued
Quantitation
Mass (m/z)
121
57
55
55
90
752
75
72
88
71
631
77
105
103
1,3- & OR 1,4-dimethylbenzene
1,2- & OR 1,3-dichlorobenzene*
2- & OR 3- & OR 4-chloro-l-
methylbenzene*
SURROGATE GROUPS AND INTERNAL STANDARDS
4-bromofluorobenzene
chlorobenzene-ds
1,4-dichlorobenzene
1,4-di f1uorobenzene
(BFB)
106
146
126
95
117
150
114
Detection Limits
(ng)
14.4
3.5
0.5
8.2
5.1
1.9
6.5
1.2
3.9
8.1
8.2
5.9
2.9
1.3
0.5
1.3
0.5
* Compounds used to calibrate GC-MS-DS on a daily basis either by direct
injection or on spiked adsorbent tubes.
a - EPA 600/4-90-010
B-8
-------
TABLE8 B.3
BREAKTHROUGH VOLUMES6 AND SAFE SAMPLING VOLUMES6 FOR
TENAX-GC AND TENAX-TA
Acetaldehyde
Acrolein
Acrylonitrile
Ally! chloride
Benzene
Benzyl chloride
Bromobenzene
Carbon tetrachloride
Chlorobenzene
Chloroform
Chloroprene
Cresol
p-Dichlorobenzene
1,4-Dioxane
Ethylene dibromide
Ethylene dichloride
Ethylene oxide
Formaldehyde
Hexachlorocyclo-
pentadiene
Methyl bromide
Methyl chloroform
Methylene chloride
Nitrobenzene
Perch!oroethylene
Phenol
Propylene oxide
Trichloroethylene
Vinyl chloride
Vinylidene chloride
Xylene
Tenax-GC
Tenax-TA
breakthrough breakthrough
volume6 volume6
38*C
0.6
4
-
-
19
300
300
8
150
8
-
440
510
-
60
-
-
-
-
0.8
-
3
-
-
-
3
21
0.6
-
200
20*C
0.6
5
8
8
36
440
"27
184
13
26
570
820
58
77
29
0.5
0.6
2000
0.8
9
5
520
100
300
3
45
.06
4
177
35'C
0
2
3
3
15
200
13
75
5
12
240
330
24
35
12
0.3
0.2
900
0.4
4
2
240
45
140
1
17
.03
2
79
Tenax-GC
safe sampling
volumec
38'C
0.3
1.7
8.2
130
130
3.5
6.5
4
191
221
87
26
0.4
1.5
1.5
8.5
.03
89
Tenax-TA
safe sampling
volume0
20'C 351C
2
3
3
14
175
11
5
5
10
230
290
23
30
12
800
3
2
200
40
120
1
18
2
70
1
1
6
80
5
2
2
5
95
130
10
14
5
360
2
95
18
55
7
32
"EPA 600/4-90/010
Breakthrough volumes expressed as liters/gram of sorbent.
°Safe sampling volume = {[Breakthrough volume (L/g)]/1.5) x 0.65 grams of
sorbent.
dBreakthrough volumes for other chemicals can be extrapolated on the basis of
boiling points for chemicals in the same chemical class.
B-9
-------
Other Methods for Specific Organic Classes
Other site related organic compounds of potential interest include
aldehydes and ketones, pesticides, and polynuclear aromatic hydrocar-
bons.
Aldehyde and ketone concentrations may be determined by either EPA
Methods TO-5 or IP-6A. The methods have different sampling techniques
but similar analysis techniques (both'use high pressure liquid chroma-
tography (HPLC)). In Method TO-5, ambient air is drawn through a midget
impinger sampling train containing 10 ml of a 2N HC1/0.05% 2,4-dini-
trophenylhydrozine (DNPH) reagent. Aldehydes and ketones react with the
DNPH to form stable derivatives which can be separated using HPLC.
Method IP-6A uses a prepackaged silica gel cartridge coated with
acidified DNPH. The relatively high pressure drop across the cartridges
(»30 inches of water) limits flow rates to about 1.5 liters/minute.
Some experimentation with HPLC operating conditions is necessary for
good compound separation. Table B.4 gives sensitivities versus sampled
volume.
Pesticides may be determined using EPA Methods 608 (EPA 600/4-82-
057) or Method IP-8. The latter is preferred for the current purpose
and is briefly described. Ambient air is drawn through a cartridge
filled with polyurethane foam (PUF) at a rate of 1 to 5 liters/minute.
Concentrations of 0.01 to 50 /ig/m3 can be determined with sampling
periods of 4 to 24 hours. The PUF is recovered and extracted with an
ether/hexane mixture. The extract is concentrated and analyzed using
gas-liquid chromatography with electron capture detector. Compounds
listed in Table B.5 have been determined with this method.
Polynuclear aromatic hydrocarbons (PAHs) may be determined using
EPA Method TO-13 or IP-7. Method TO-13 is a high volume method applica-
ble only to outdoor environments. IP-7 is a low volume method but has
detection capabilities adequate for indoor air impact assessments. The
method is applicable to PAHs adsorbed on particulates as well as in the
vapor phase. It also can be adapted for use with a PM-10 cut-point
sampling nozzle.
In Method IP-7, air is drawn through a filter backed by either an
XAD-2 or PUF filled cartridge. Up to 30m3 can be sampled at a rate of
B-10
-------
TABLE B.4
SENSITIVITY (ppb, v/v) OF SAMPLING/ANALYSIS USING
ADSORBENT CARTRIDGE FOLLOWED BY HPLC
Sample Volume, I
Compound
Formaldehyde
Acetaldehyde
Acrolein
Acetone
Propionaldehyde
Crotonaldehyde
Butyraldehyde
Benzaldehyde
Isovaleraldehyde
Valeraldehyde
o-tolualdehyde
tn-tolualdehyde
p-tolualdehyde
Hexanaldehyde
2.5-dimethyIbenzaIdehyde
10
20
30
40
50
100 200
300
400 500 1000
1.45
1.36
1.29
1.28
1.28
1.22
1.21
1.07
1.15
1.15
1.02
1.02
1.02
1.09
0.97
0.73
0.68
0.65
0.64
0.64
0.61
0.61
0.53
0.57
0.57
0.51
0.51
0.51
0.55
0.49
0.48
0.45
0.43
0.43
0.43
0.41
0.40
0.36
0.38
0.38
0.34
0.34
0.34
0.36
0.32
0.36
0.34
0.32
0.32
0.32
0.31
0.30
0.27
0.29
0.29
0.25-
0.25
0.25
0.27
0.24
0.29
0.27
0.26
0.26
0.26
0.24
0.24
0.21
0.23
0.23 •
0.20
0.20
0.20
0.22
0.19
0.15
0.14
0.13
0.13
0.13
0.12
0.12
0.11
0.11
0.11
0.10
0.10
0.10
0.11
0.10
0.07
0.07
0.06
0.06
0.06
0.06
0.06
0.05
0.06
0.06
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.03
0.03
0.03
0.04
0.03
0.04
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.02
0.03
0.03
0.03
0.03
0.03
0.02
0.02
0.02
*0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Note: ppb values are measured at 1 atra and 25'C; sample cartridge Is eluted with 5 ml acetonitrile,
and 25 ml are injected onto HPLC column.
Note: Maximum sampling flow through a DNPH-coated Sep-PAK* cartridge is about 1.5 L per minute.
B-ll
-------
TABLE B.5
PESTICIDES DETERMINED BY GAS CHROMATOGRAPHY/
ELECTRON CAPTURE DETECTOR (GC-ECD)
Aldrin
BHC (a-and /3-Hexa-
chlorocyclohexanes)
Captan
Chlordane, technical
Chlorothalonil
Chlorpyrifos
2,4,-D esters
fi.fi,-DDT
fi.fi,-DDE
Dieldrin
Dichlorvos (DDVP)
Dicofol
2,4,5-Tri chlorophenol
Folpet
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Lindane (r-BHC)
Methoxychlor
Mexacarbate
Hirex
trans-Nonachlor
Oxychlordane
Pentachlorobenzeiie
Pentachlorophenol
Ronnel
B-12
-------
up to 20 liters per minute without significant breakthrough. Because
some of the PAH collected by the filter may volatilize and be collected
on the adsorbent, the two should not be analyzed separately. The filter
and adsorbent are extracted in Soxhlet extractors, cleaned up with
silica gel column chromatography and analyzed using gas chromatography
(with FID or MS detector) or by HPLC. In general, MS operated in SCAN
mode is preferred because of the positive compound identifications.
B.I.2 Particulates
Methods are available to determine total suspended participates and
respirable (<10/xm) particles. In many cases only the latter will be of
concern. However, in some cases consideration may be given to the
larger particulate. Those particles that penetrate the building
envelope will rapidly deposit on indoor surfaces. Exposure to these
particulates may occur through dermal contact, ingestion, or inhalation
if resuspended by occupant traffic or during such normal activities as
cleaning.
Total suspended particles may be determined by the High-Volume
Method given in Section 2.2 of "Quality Assurance Handbook for Air
Pollution Measurement System: Volume II - Ambient Air Specific Meth-
ods," EPA 600/4-77-027a. Ambient air is drawn into a covered housing
through a filter (nominal 8 x 10 inch), with a collection efficiency of
at least 99 percent for particle size 0.3pi, at 1.1 to 1.7 m3/min.
Particles up to 50 jwn (aerodynamic diameter) collect on the filter
surface. Air containing up to 750 #g/m3 may be sampled for up to 24
hours at this rate. Particulate concentration is determined by dividing
air volume sampled into the weight of particulate collected by the
filter. If should be noted that because of the high air volume sampled,
the method is not applicable to indoor air sampling.
Method IP-10A can be used to determine the concentration of
particulate in both the <2.5/im and 2.5 to 10/wn ranges. The method
contains both a fixed site monitor and a personal sampler. Only the
fixed site monitor is recommended here for outdoor ambient air monitor-
ing. The monitor operates on the principal of impaction. The horizon-
tal slotted inlet prevents very large particles from entering. Either a
one-stage or two-stage impactor assembly can be used to trap particles
B-13
-------
above 10/im or 2.5/zm, respectively. Participate passing the impactor
section is collected on a filter for subsequent weighing. Note that
particles in both size ranges cannot be collected simultaneously with a
single monitor. It is possible to connect two sampling heads to the
same pump assembly to simultaneously collect in both size ranges. The
limit of detection for a 12 hour sampling period at the specified
sampling rate of 4 liters per minute is 4/ig/m3.
The sampler does not use a critical orifice to control flow rate.
Thus, a feedback control from a mass flow meter is required to vary pump
speed in order to maintain flow constant with ±5 percent.
B.I.3 Organics in Soil Gases
Knowledge of the chemical concentrations in soil gases near a
building will be important when site impact via the underground pathway
is of concern. These situations can arise when soil gases migrate
directly from the site (e.g., landfills), the nearby soil is contaminat-
ed, or the underlying groundwater is contaminated. As discussed in
Section 2, it is important in these cases to determine concentrations in
soil gas very near the building (i.e., in the region between 0.5 and 1
meter from the underground walls and floor). Gases in this soil region
are most likely to diffuse through building cracks or be drawn into the
building as a result of building underpressurization.
Because soil properties are inhomogeneous, both vertically and
horizontally, it is expected that concentrations will vary around the
building. Therefore, soil gas concentrations must be measured at
several locations around the building. Following the approach used in
Appendix C of Volume II of the Air/Superfund NTGS Series (EPA-450/1-89-
002a), it is estimated that a minimum of two sampling points should be
used on each side of the building (typical residential building). It is
preferred that one additional sample be collected from directly under
the building (approximate center). It may not be practical to collect
this sample for slab-on-grade or basement constructions because a hole
would have to be made in the floor. Efforts to collect this sample
should be made, if possible. Buildings of this type generally have a
very permeable zone (1 to 4 inches in depth) immediately below the slab
B-14
-------
due either to the use of a gravel bed underlying the slab or gaps formed
by soil not making continuous contact with the slab.
The preferred technique for collecting soil gas samples is the soil
vapor (ground) probe. The installation of probes is simple. A small
diameter pipe (=0.5 inch OD) is inserted into the ground to the desired
depth either by hammering or driving it down a slightly undersized pre-
augured hole. The use of pre-augured holes is recommended if soil
permeabilities, described below, are to be measured. Retractable
centering probe tips should be used to keep soil out of the probe. The
probe should be withdrawn a minimum of two inches to expose soil at the
probe tip.
A minimum of two probe volumes of soil gas should be extracted
before samples are taken. Because soil gas concentrations must be
fairly elevated if indoor air impact is to be of significant concern,
low detection limits are not typically required. (Soil gas intrusion
rates into a building are typically less than 5 percent of ambient air
infiltration rates.) However, positive compound identification and
accurate concentrations relative to other measured compounds are needed.
It is preferred that integrated samples be collected over at least
one hour when indoor air monitoring is not being also conducted. If
indoor air monitoring is also being conducted, integrated samples over
the same time period should be collected. (Although soil gas concentra-
tions should not vary significantly over this time period, some varia-
tion is possible and it is best to eliminate this uncertainty.)
In most cases, canister sample collection and analysis procedures
such as those given in EPA Methods TO-14 or IP-1A are preferred because
prior knowledge of actual concentrations is not required and concerns
over breakthrough volumes are eliminated. However, professional
judgement is required in the selection of technique based on known or
suspected contamination at the site. For example, it may be possible to
adapt IP-7 for pesticides or IP-8 for PAHs if soil contamination in the
immediate vicinity of the building is known or suspected. Sampling flow
rates, and, thus, sample volumes, would need to be kept low to reduce
concern about drawing in surface air.
B-15
-------
B.I.4 Soil Permeability to Gas Flow
The permeability of soil to gas flow can be an important indicator
of the rate at which soil gas can enter a building. If the permeability
is low (i.e., below 10~8cm2), it is unlikely building underpressures
have much effect on the rate of intrusion and gases will enter primarily
by diffusion. In this case, soil gas concentrations below the building
may be higher than estimated from modeling equations. As the permeabil-
ity increases, however, building pressure becomes more important and may
result in gases several meters from the building being drawn inside.
This effect reduces the soil gas concentration (relative to model
assumptions) at the soil-building interface but increases the intrusion
rate thus resulting in higher indoor concentrations. In essence, this
is the reason predictive models assume a soil gas concentration of zero
at the basement floor-soil interface.
Permeabilities are easy to measure using the soil gas probes
discussed above. Permeabilities should not be measured until after soil
gas samples have been obtained because the injected gas will distort
concentration measurements. Permeabilities are determined by measuring
the gas pressure in the probe as a metered flow of air is passed through
the probe into the soil. A cylinder of compressed air, flow meters with
ranges from about 5 to 500 cm3/min., and a differential pressure gauge
with a range of 0 to 125 pascals are needed. Probe pressure should be
measured at three or more flow rates to obtain a good average. The
permeability is calculated, assuming Darcy flow, from
KV = Q u
4wr2Pa
where,
KY = Permeability, m2
Q = Air Flow rate, m3/s
H = Viscosity of air, 1.83 x 10s Kg/m-s
r = Internal radius of probe, M
Pa = Probe pressure in pascals
The probe tip internal diameter should be measured to within ±0.005
inches before inserting into the soil and after extraction. Method
reference is contained in DMSA Action LTD., "Review of Existing Informa-
tion and evaluation for possibilities of research and development of
B-16
-------
instrumentation to determine future levels of radon ^at a proposed
building site.", Report INFO-0096, Atomic Energy Control Board, Ottawa,
Canada (1985).
B.2 METHODS FOR USE INSIDE BUILDINGS
Critical measurements to be made indoors include those to determine
pollutant concentrations and those to determine air exchange rates.
Both are discussed here.
B.2.1 Pollutant Measurement Methods
The air inside a typical building may contain hundreds of compounds
(over 3800 compounds have been identified in tobacco smoke alone). To
assess the potential impacts from a nearby contaminated site, it is only
necessary to monitor for those compounds identified with the site.
Because these compounds are also common in building materials and
consumer products, their detection in the indoor air does not necessari-
ly indicate their origin.
Discussion of appropriate sampler locations is included in Section
2. Principal considerations in locating samplers are:
• Air exchange rates in the area should be large compared to
sampling volume.
• Samplers should be located in the normal breathing zone (i.e.,
3 to 5 feet above floor level).
• At least one sampling location should be established on each
level of the building.
• Samplers should not be located near obstructions, air supply or
return registers, or local pollutant sources.
• Exterior windows and doors should be kept closed as much as
possible. All indoor doors should be open.
• Sampling is best conducted under conditions that yield building
underpressurizations of several pascals. Low rates of air
infiltration and soil gas convective intrusion may result from
low building underpressurizations.
In general it is recommended that indoor air monitoring use methods
listed in the "Compendium of Methods for the Determination of Air
B-17
-------
Pollutants in Indoor Air," EPA-600/4-90/010. In particular it is
recommended that:
• Method IP-1A, Appendix B, use of portable gas chromatograph, be
used for initial screening investigations and to assist in
calculating proper sample size for methods based on use of
solid adsorbents.
• Method IP-1A, canister, be used for collection of most target
VOCs. Analytical finish by GC-MS in SIM or SCAN mode.
• Method IP-IB, Tenax® cartridge, be used only for compounds not
quantifiable by IP-1A. Analytical finish by GC-MS in SCAN
mode.
• Method IP-6A, solid adsorbent, be used for aldehydes/ketones.
Method TO-5 may also be used.
• Method IP-7, PUF/XAD solid adsorbents, be used for PAHs. Note
that if the PUF adsorbent is used, the method may also be used
to determine pesticides. GC-MS in SIM or SCAN mode is the
preferred analytical finish.
• Method IP-8, PUF adsorbent, be used for pesticides. The low
sample volume collected in this method may be inadequate for
PAH quantitation in indoor and ambient air. GC-MS in SIM or
SCAN mode is the preferred analytical finish.
• Method IP-lOa, size specific impactor, be used for airborne
particulate matter. The single stage impactor collecting
particles up to 10/fm aerodynamic diameter is preferred. The
stationary microenvironmental exposure monitor (MEM) is pre-
ferred. Collected particulates may be subjected to additional
analysis if desired.
Specific circumstances may also warrant collection of deposited
dusts/chemicals. The OSHA Method, described in 29 CFR 1910.132(a), may
be used to collect deposited material for characterization. The method
cannot be used for air inhalation estimates. Collection procedure is
simple. A Whatman 42 filter is moistened with an appropriate solvent
and used to wipe a known area (generally about 100 cm2). The filter is
folded, with exposed side in, and folded again to form a 90-degree angle
at the center of the filter. The filter is placed in a clean vial for
transport to the laboratory. A blank filter that is also moistened with
B-18
-------
the solvent should be submitted in a separate vial. Clean gloves,
impervious to the solvent used and potential contaminants, should be
worn during all phases of collection.
B.2.2 Air Infiltration Rates
Knowledge of air infiltration rates is not necessary to determine
site related impacts on indoor air quality for the time period over
which sampling occurred. However, if any estimate is to be made of
likely concentrations at other times, knowledge of the infiltration rate
and building air changes per unit time during the sampling period are
required.
Preferred techniques to make these measurements are IP-4A and IP-4B
which are both tracer gas methods. Indirect measurements can be made
using the fan pressurization/depressurization method (ASTM-779).
Because the ASTM method significantly affects building air infiltration
rates, it cannot be used immediately before (=12 hours) or during indoor
monitoring.
Method IP-4A can be used to obtain the integrated air exchange rate
over the monitoring period. This method uses constant emitting perfluo-
carbon (PFT) sources which must be placed in the building at least 8
hours before monitoring begins. One source is required for about every
500 ft2 of living area. They are temperature sensitive and care must be
used in placement. Samples are collected on activated charcoal
spherules inside capillary adsorption tubes over the entire monitoring
period. The tubes are analyzed by GC-ECD. The method includes cleanup
for other compounds that respond to the ECD detector.
The method is applicable to small and large buildings and can be
used to determine interzonal flows as well as exchange rates. The
method assumes the effective exchange volume of the building is equal to
the physical volume. Poor mixing within the building may effect this
assumption.
Method IP-4B is a tracer concentration decay method. In this
method, a tracer, such as SF6, is injected into the structure, thorough-
ly mixed and its concentration measured over time. The concentration
decreases due to air exchange with outside air. Tracer concentrations
can be measured on-site with a GC-ECD or SF6 specific detectors. Grab
B-19
-------
samples or time-integrated samples for off-site analysis can also be
obtained.
Adequate mixing in the indoor environment is critical to the
success of this method. Local fans and/or heating system fans should be
used to obtain tracer concentration throughout the structure within 5
percent of the average. It is preferred that measurements be made in
several locations on each floor of the building. Tracer injections can
be made at about one-hour intervals to obtain a series of infiltration
measurements. If sufficient difference exists in meteorological condi-
tions during these measurements, it is possible to characterize the
effects of temperature, wind, and structural factors on the infiltration
rates.
Tracer concentration is assumed to follow
C = C0 exp (-IT)
where,
C = Concentration at time T
C0 = Concentration at T = o
I = Air exchange rate
T = Time
The air exchange rate can be obtained from the slope of a LnC versus
time plot.
When using either Method IP-4A or 4B, measurements should be made
of indoor and outdoor temperatures, wind velocity, barometric pressure,
and building underpressurization.
In the ASTM Method, a large fan is mounted on a door connected to
the outside. The fan is used to draw air out of and force air into the
building at several differential pressures. Flow rates are determined
from a fan calibration curve. The effective leakage area is calculated
from
L - Q/(2APp)-°-5
where,
L = Leakage area
Q = Air flow rate
AP = Pressure difference across building shell
p = Density of air
B-20
-------
Infiltration rates and air exchange rates can then be estimated
using
Q = L(AAT + B/i2)0"5 from page 8 of EPA 400/3-91/003
where,
Q = Air infiltration rate, CFM
L = Effective leakage area, in2
AT = Indoor-outdoor temperature difference, °F
/x = Wind speed, mph
A = Stack coefficient
B = Wind coefficient
Values for A and B are in the above document.
B.3 INDOOR AIR POLLUTANTS
Literally hundreds of compounds have been detected in indoor air.
Over 3,800 compounds have been detected in tobacco smoke alone. The
sources of the compounds are many and include building materials,
furnishings, cleaners and waxes, paints, pesticides, hobby supplies,
combustion devices, and personal care products. It has been estimated
(Tancrede et al, "The Carcinogenic Risk of Some Organic Vapors Indoors:
A Theoretical Study", Atmospheric Environment, Vol. 21, No. 10, 1987)
that cancer risks from exposure to indoor air probably well exceeds 1 x
lO'4.
Many of the compounds typically associated with a contaminated site
are also present in typical non-impacted indoor environments. This
significantly complicates monitoring efforts to differentiate site
contributions to indoor air concentrations. The data currently avail-
able indicates that compounds detected and concentrations of those
compounds is a function of many factors including building type (e.g.,
office, residence, school, etc.), age, and location.
Although many studies of indoor air exposures have been made (most
notable is The Total Exposure Assessment Methodology (TEAM) Study - see
EPA 600/6-87-002Q for summary and analysis), it does not appear to be
feasible at this time to tabulate the averages and ranges for individual
compounds in all types of potentially impacted structures in a way that
would be useful for determining site impacts from indoor air concentra-
tions. The TEAM studies do, however, have a substantial amount of
B-21
-------
information on averages and ranges for many compounds in a number of
different building types. This may be considered the best currently
available source of information on typical concentrations.
Examples of some specific compound measurements are contained in
EPA 400/3-91-003, "Introduction to Indoor Air Quality: A Reference
Manual." EPA 600-R-92-02, "Indoor Air Quality Data Base for Organic
Compounds," provides the best current summary of data on organics from
studies over the last decade. This data base contains information on
over 220 organic compounds. The compounds most frequently reported in
the studies reviewed were: formaldehyde, tetrachloroethylene, 1,1,1-
trichloroethane, trichloroethylene, benzene, p-dichlorobenzene, toluene,
ethylbenzene, xylene, decane, and undecane. It should be noted that not
every study monitored for the same compounds or used similar methods.
B-22
-------
APPENDIX C. CASE STUDIES
-------
TABLE OF CONTENTS
APPENDIX C. CASE STUDIES
C.I Buildings on Landfill Sites C-l
C.I.I Quality Assurance/Qua!ity Control . . . C-5
C.I.2 Results C-7
C.I.3 Conclusion Reached C-9
C.I.4 Comments on The Study C-9
C.2 Buildings Near Landfill C-ll
C.2.1 Nature and Extent of the Problem .... C-13
C.2.2 Prevention of Gas in Structures .... C-20
C.2.3 Comments on the Study C-20
C.3 Buildings Over Groundwater Plume (A) C-21
C.3.1 Nature and Extent of the Problem .... C-22
C.3.2 Indoor Air Modeling Results C-30
C.3.3 Comments on Case Study C-30
C.4 Buildings Over Groundwater Plume (B) C-35
C.4.1 Nature and Extent of Problem C-37
C.4.2 Conclusions C-41
C.4.3 Comments on Case Study C-42
C.5 Buildings Over Groundwater Plume (C) C-44
C.5.1 Nature and Extent of the Problem .... C-44
C.5.2 Results C-48
C.5.3 Conclusion C-50
C.5.4 Comments on Study C-51
C.6 Buildings Near Sludge Disposal Pit C-53
C.6.1 Nature and Extent of the Problem .... C-53
C.6.2 Residential Air Monitoring C-57
C.6.3 Comments on Indoor Air Effort C-58
C.7 Fugitive Dust Case Study C-60
C.7.1 Nature and Extent of the Problem .... C-60
C.7.2 Comments on the Study C-66
C.8 Buildings Over Waste Oil Pit C-67
C.8.1 Nature and Extent of Problem C-69
-------
APPENDIX C
CASE STUDIES
The potential impacts of Superfund sites on the indoor air quality
of nearby buildings have been investigated in a number of cases. This
Appendix documents a sample of such cases. The examples cover a range
of site conditions, potential types of indoor air impacts, phases in the
Superfund program, and judgements of investigating personnel.
It should be noted that, in most cases, additional work beyond that
presented here has been conducted. Therefore, this Appendix should be
read in the context that it presents partial results of investigations
and should not be construed as the final analysis of the potential site
impact. The purpose is to provide an overview of the modeling and
monitoring techniques that have been used and the relative success of
these techniques.
It is not intended as a revie of all possible case types. It is
also not intended as documentation that indoor air impacts have been
shown to occur. It is hoped that sufficient detail is provided to allow
site investigations to avoid some of the major pitfalls in assessing
indoor air impacts.
C.I BUILDINGS ON LANDFILL SITES
The Landfill extends over approximately ten acres of land. It
received both municipal and industrial wastes from about 1910 to 1967.
The wastes included drummed industrial wastes and uncontained liquid
wastes. Prior to 1964, open burning took place at the site. After
closure, when the wastes were covered by approximately two feet of soil,
the property was subdivided and sold for both residential and commercial
development. Based on verbal reports of methane buildup in buildings
currently located on the site, a study was conducted in three commercial
buildings in late 1985. The results of the study indicated the presence
C-l
-------
of methane in one of the buildings at levels well below the lower
explosive limit for methane in air, even in the floor cracks where
levels were most highly concentrated. However, the concern existed that
the methane entering buildings acts as a carrier gas for toxic species.
Preliminary soil gas analysis conducted by the regulated party's
contractor utilizing a portable chromatograph indicated the presence of
eleven toxic chemical species.
Therefore, indoor air quality sampling for volatile organic
chemicals was performed at the site. It took place between approximate-
ly 9 AM and 5 PM on a day in September, 1990.
The purpose of this study was to collect data of a quality suffi-
cient to document the presence of any toxic volatile organic chemicals
in buildings constructed on the former landfill site which may pose a
health risk to their occupants.
The buildings sampled were selected on the basis of an on-site
reconnaissance visit conducted jointly by EPA and State personnel as
well as existing data. They are single-family residences referred to
here as Building A and Building B. Both homes are of split-foyer design
and have garages adjacent to finished portions of their lower levels.
Aside from automobile storage, miscellaneous small gasoline engines (and
the fuel for such) are commonly kept in such indoor areas. The garage in
Building A, in particular, smelled strongly of fuels and solvents (HNu
readings were 2 ppm). Virtually all of the tentatively identified
compounds detected are known components of gasoline. Both homes are
inhabited by cigarette smokers. Although no smoking took place during
the sampling period, information on the prior 24 hours was not obtained.
Sampling and analytical methodologies were tailored for purposes of
identifying and quantifying the compounds listed in Table C-l.
Compounds of interest were selected based on preliminary data developed
during the soil gas sampling conducted in November and March and April
1989. EPA personnel performed both the field sampling activities and
the analytical work.
C-2
-------
TABLE C-l
Targeted Compounds
A - Tenax
Benzene
Ethylbenzene
Tetrachlorocthylene
Toluene
Trichloroethylene
1,1, 1-trichloroethane
m, p-xylene
o-xylene
B - Spherocarb
Trans-1, 2-dichloroethylene
Vinyl Chloride
Sampling and analysis of indoor air conformed to Method TO-1,
Method for the Determination of Volatile Organic Compounds in Ambient
Air using Tenax Adsorption and Gas Chromatography/Mass Spectrometry
(GC/MS) and Method TO-2, Method for the Determination of Volatile
Organic Compounds in Ambient Air by Carbon Molecular Sieve Adsorption
and GC/MS, from the Compendium of Methods for the Determination of Toxic
Organic Compounds in Ambient Air, EPA-600/4-84-041, May 1987.
One sampling station was located on each of the lower two floors of
each building. Each station on the lower floors consisted of primary
and secondary tubes in series plus a duplicate (of each type of tube).
Each station on the upper floors, consisting of single tubes of each
type, was located upstairs on the main living level in the kitchen. The
lower-level station in each building was expected to produce the highest
readings of any of the stations. Because, the possibility of break-
through of the species through the cartridges, with resultant loss of
accuracy, was greatest at this point, the series tube configurations
were utilized there to gauge breakthrough. A single cartridge of each
C-3
-------
type was also set up outdoors to monitor ambient concentrations, for
comparison with those measured indoors.
Samples were collected over an 8-hour period, utilizing personal
constant-flow pumps to draw ambient air through cartridges containing
adsorbents. Certain nonpolar volatile organic compounds having boiling
points in the range of approximately 25 to 95 *C were captured on Tenax
GC (poly 2,6-Diphenyl phenylene oxide). Other more volatile compounds
having boiling points ranging from -26 to +50'C (which will pass through
Tenax adsorbent) were collected on a carbon molecular sieve (CMS)
adsorbent, Spherocarb.
A sample volume requirement of approximately 10 liters dictated a
flow rate of approximately 21 ml/min over the 8-hour sampling period.
Sampling cartridges were positioned in the "breathing zone" (approxi-
mately three feet above floor level).
Occupants were requested to eliminate, insofar as is possible, the
opening of any doors or operation of ventilation/exhaust fans (clothes
driers act as exhaust fans) for a minimum of 24 hours preceding the
sampling period. Windows remained closed for 24 hours preceding
sampling. Door openings were kept to a minimum during the sampling
period.
During the eight hour period while the pumps were running, methane
levels were measured utilizing an Organic Vapor Analyzer (OVA) in the
gas chromatograph mode, with the output going to a strip-chart recorder.
An Hnu (which does not respond to methane) was used for total non
methane readings. The highest methane readings obtained were 8 ppm
(directly above a cat litter pan).
Temperature, barometric pressure and relative humidity (RH) were
measured periodically inside the buildings throughout the eight-hour
sampling period. Temperature readings were obtained with an Ertco
thermometer, pressure with a Taylor aneroid barometer and relative
humidity with a Bacharach Instrument sling psychrometer. Inside
temperatures were fairly constant at 22-23'C and RH was constant at 42-
44 percent. The test report did not give barometric pressure.
C-4
-------
C.I.I Quality Assurance/Quality Control
Cartridge Conditioning
Cartridges were thermally conditioned in a Tekmar Model 5100
Thermal Conditioner two days prior to sampling with a nitrogen purge
rate of approximately 100 ml/min. Tenax cartridges were conditioned for
14 hours at 250°C. Spherocarb cartridges were conditioned for 15 hours
at 399° C.
After the Tenax cartridges were conditioned, they were injected
with 5 micro!iters of a surrogate solution, composed of deuterated
Benzene (d6), Toluene (d8) and bromofluorobenzene in methanol, using the
flash vaporization technique. The cartridges were then refrigerated
prior to field sampling activities.
Pump Calibration
Prior to sampling, the pumps were calibrated using a Mini-Buck
commercial primary standard flow meter with optical sensing of bubble
passage. Following sample collection, calibrations were again checked
with the Mini-Buck. The criterion for the difference between the
initial and final flow rates is no more than 15%. Where the difference
exceeds 15%, the data is reported as an estimated value. Four pumps
exhibited a flow rate difference of more than 15%.
Sample Storage and Transportation
- Cartridges were stored in individual containers following
conditioning for transportation and storage. The individual containers
were sealed in jars containing granulated charocal to minimize the
possibility of cross-contamination.
- Cartridges were stored in a clean laboratory freezer maintained
at approximately -20 to 0* C following conditioning, and were returned
to said freezer following sample collection while awaiting analysis.
- Cartridges were transported to and from the field in an ice-
filled cooler.
C-5
-------
Blanks
One cartridge of each type used in the sampling process accompanied
the sampling cartridges to the field and was handled in precisely the
same manner as the sample cartridges, except that they remained in their
containers to serve as field blanks.
A laboratory blank of each type used remained in the freezer and
was subject to the same criteria as the field blanks.
The results of the analysis of the Tenax trip blank showed it to be
contaminated with two compounds, 1,1,1-trichloroethane and benzene. The
Tenax laboratory blank was found to be contaminated with four compounds,
1,1,1-trichloroethane, benzene, toluene and 1, 2, 3-trichlorobenzene.
All of the sample cartridge data for benzene, with two exceptions, met
the criterion and were considered valid from the standpoint of blank
contamination. All of the Tenax cartridge data for toluene met the
criterion for blank contamination.
No contamination was found on either the trip or laboratory
Spherocarb cartridges.
Breakthrough
Breakthrough of compounds through the adsorption medium was
assessed by a comparison of the levels found on the primary-secondary
cartridge pairs. Secondary, or backup cartridges must meet the criteri-
on of containing less than 20% of the amount of each species found on
the primary cartridge of each pair. At Building B, two compounds failed
the criterion. At Building A, all the components failed the criterion
with the exception of benzene. None of the Spherocarb primary-secondary
pairs exhibited any breakthrough.
Precision
Each sampling event consisted of the collection of an additional
set of parallel samples collected simultaneously at different flow
rates. Agreement between parallel samples should generally be within ±
25% if the concentration is at least 10 times greater than the minimum
detection level. None of the targeted Tenax compounds detected at this
level failed the criterion. Methylene chloride was the only compound
detected on Spherocarb which failed the criterion.
C-6
-------
C.I.2 Results
Much of the chemical concentration data from this study were
qualified because of:
• Four pumps not meeting flow-rate acceptance criterion
• Poor recovery on two cartridges
• Blank contamination for 1,1,1-Trichloroethane and benzene
• Breakthrough on primary cartridges using Tenax (precision
criteria were met however).
Summarized results are in Table C-2.
C-7
-------
TABLE C.2
Maximum Concentrations of Targeted Compounds Found (1) (PPBV/V)
A - Tenax
benzene
ethylbenzene
tetrachloroethylene
toluene
trichloroethylene
1,1, 1-trichloroethane
m,p-xylene
Building B
5.9
2.6
0.34
19*
0.04*
3*
8.4®
Building A
17*
7.7*
0.52
56*
ND
5.2*
25®*
Ambient
ND(3)
ND
ND
ND
ND
ND
ND
* - Estimated Value
B - Spherocarb
trans- 1 ,2-dichloroethylene
vinyl chloride
ND
ND
ND
ND
ND
ND
(1) - 18 other compounds were detected; the highest was 15 ppb V/V for isopropylbenzene
(2) - Reported as total o,m,p-xylenes
(3) - ND - not detected in any sample
C-8
-------
C.I.3 CONCLUSION REACHED
The test report stated,
"In summary, all eight of the targeted compounds detected in the
homes studied here have potential sources apart from the underlying
landfill and are, in fact, not uncommon in a household environment.
Therefore, it is impossible to determine with complete certainty which
compounds are actually migrating into the homes from past landfill ing
and which are present merely as a result of typical household activi-
ties.
Finally, whether or not these levels constitute a health hazard
must be determined by a risk analysis."
The complete test report was submitted to the residents along with
a cover letter that states, in part:
For Building A;
"The levels of chemicals detected in your home were comparable to
those found in the average home.
At this time we do not know the origin of these chemicals. As we
are planning to complete the Remedial Investigation of the landfill in
the summer of 1992, we hope, at that time to determine the source of
these indoor pollutants."
For Building B;
"The levels of chemicals detected in your home were found to be
slightly higher than those normally found in an average home. However,
the levels detected in your home do not pose an immediate health threat.
At this time we do not know the origin of these chemicals. As we
are planning to complete the Remedial Investigation of the landfill in
the summer of 1992, we hope, at that time to determine the source of
these indoor pollutants.
The chemicals identified in your home at slightly elevated levels
are: benzene, ethylbenzene, toluene, and xylenes. Apart from the
landfill, possible sources of these indoor air pollutants are numerous."
C.I.4 Comments On The Study
This case provides a good example of some of the pitfalls of indoor
air monitoring. The study met its primary objective which was to
"document the presence of any toxic volatile organic chemicals in
C-9
-------
buildings which may pose a health risk to their occupants." However,
because the indoor air concentrations of chemicals of the same identity
as those associated with the Superfund site were similar to typical
indoor concentrations, no conclusive statements about the source of the
chemicals could be made. This would be true even if all sampling and
analysis had gone perfectly.
In retrospect, because the study design was such the contribution
of site chemicals to the indoor air concentrations could not be deter-
mined, it would have been preferable to state the purpose as: Determine
whether or not indoor air contaminants were of types and concentrations
to suggest significant intrusion of site related chemicals. With this
as the purpose, it would have been easier to develop a risk communica-
tion plan that had specific follow-up steps depending on the contaminant
concentrations determined. Building occupants could, then, have been
advised before monitoring took place what actions would be taken
depending on the concentrations found in this screening study.
C-10
-------
C.2 BUILDINGS NEAR LANDFILL
The Landfill is in South King County, Washington. Puget Sound is
slightly more than a mile to the west. Residential areas surround the
site, with the exception of a commercial strip along Highway 99 to the
west and a stand of second-growth mixed conifer-hardwood forest on the
north. Two elementary schools and a community college are within one
mile of the site. Interstate 5 (1-5) borders the site on the east.
Approximately one mile east of 1-5 is the Green River, which meanders
north, becomes the Duwamish River, and enters Puget Sound. Figure C-l
shows the location of the landfill and the landmarks in its vicinity.
From 1945 to 1966, the site was operated as a gravel pit. The pit
originally was adjacent to a peat bog lake located northeast of the
center of the present landfill. As the pit was mined, water was drawn
from the lake to wash silt and clay from the gravel and sand, then the
water was returned to the lake. Silt and clay built up on the lake
bottom. Near the end of the operation of the gravel mine the barrier
between the lake and the gravel pit was broken, allowing the silty lake
water to flow into the gravel pit. As a result, a clay/silt layer
underlies much, but not all, of the landfill.
In January 1966, the 60-acre site was leased and used as a landfill
for nonputrescible waste, which includes organic material that decompos-
es slowly, such as demolition debris and wood wastes. The landfill
received demolition debris from commercial haulers and wood wastes and
yard trimmings. Records beginning in 1980 indicate that some industrial
wastes also were deposited. Information included in EPA's Emergency and
Remedial Response Information System (ERRIS) files indicates that the
landfill may have received industrial liquid and sludge wastes before
1980.
C-ll
-------
LANDFILL
SCALE w FEET
500 1000
FIGURE C-1
Location Map
C-12
-------
During the course of operations at the landfill, an estimated 3
million cubic yards of solid waste were deposited covering approximately
40 acres up to about 130 feet deep in places. The east side of the
landfill rises above the adjacent property; the landfill surface slopes
downward to the northwest corner. The landfill was closed in October
1983. The entire site was covered with 6 to 24 inches of a silt/sand
material when operations ceased; it now appears as an open grassy area
with scattered shrubs and a few areas of exposed soil. Approximately
102 acres of land east of 1-5 currently drain directly into the solid
waste.
C.2.1 Nature And Extent Of The Problem
Potential problems related to solid waste landfills similar to the
Landfill mainly derive from the combustible gas that is a byproduct of
decomposition, and leachate, i.e., water or other liquid that has come
into contact with the waste. Methane, carbon dioxide, and small amounts
of hydrogen sulfide are normally generated by the decomposition of
landfill wastes. Small amounts of toxic volatile organics are commonly
found as well. These gases can be released to the atmosphere or migrate
offsite by means of underground pathways. Leachate can flow or seep
offsite to contaminate surface water or it can infiltrate the groundwa-
ter underlying the landfill if no artificial or natural barriers exist.
The Landfill is a natural drainage basin from which no surface water
exits. However, depending on subsurface conditions, groundwater
contamination is possible.
The landfill was closed in the fall of 1983 and extensive testing
of gas and water in the landfill and its vicinity began. Samples of
leachate and groundwater from monitoring wells in and around the
landfill and gas samples from gas probes indicated the presence of
organic and inorganic contaminants with a high potential for offsite
migration. In May 1986, the EPA placed the site on its National
Priority List for cleanup and a remedial investigation was initiated.
Initial remedial measures were taken to control the offsite
migration of gas and prevent possible emissions from the landfill
C-13
-------
surface; gas extraction wells were installed offsite and gas migration
control wells were installed on the perimeter of the landfill.
Combustible Gas
Combustible gas, primarily methane (CH4), was detected at concen-
trations up to 10,000 ppm in the basements of homes near the Landfill in
1985. A series of gas probes were installed around the site perimeter
to monitor gas concentrations. Gas was found in many of the probes.
Measures were taken to seal cracks in building foundations. Numerous
shallow and some deep probes were installed offsite. A program to
monitor homes for gas was established. Data from the offsite probes and
the monitoring program indicated that the gas had migrated up to
approximately 2,600 feet from the landfill and was seeping into homes at
that distance. Additional probes were installed farther from the site.
The sealing programs within homes and businesses was expanded, and some
families were evacuated.
In September 1985, and again in September 1986, a series of onsite
gas migration control wells around the perimeter of the landfill were
installed as initial remedial measures. Designed to prevent offsite gas
migration by withdrawing and venting the gas, these wells are connected
by manifold piping to motor blowers and flares. As part of final
landfill closure, portions of this system will be replaced by permanent
equipment that will remain in operation for the indefinite future.
Leachate
It is estimated that over 50 million gallons per year of surface
water (drainage and direct precipitation) enter the landfill. Measure-
ments of water levels in leachate monitoring wells indicate that
stormwater discharge from drainage pipes produces rapid and significant
increases in water levels within the solid waste. Since there is no
surface runoff from the landfill, leachate must eventually enter the
groundwater system if it does not remain in the landfill. Leachate
flows to the southeast corner of the landfill and then downward and then
east and west away from a mound located beneath the Sand Aquifer sink.
Based on chloride concentrations, leachate is calculated to be indistin-
guishable from background groundwater at a maximum of 3,000 feet from
C-14
-------
the landfill boundary to the east and southeast and 1,500 feet from the
landfill boundary to the west.
Offsite Gas Monitoring
Data recorded from the monitoring probes were computerized and used
to generate gas concentration isopleths (maps showing gradients of gas
concentrations) for the areas surrounding the Landfill.
The isopleths for February 1986, shown in Figures C-2 and C-3,
represent the status of landfill gas migration just after the Phase I
onsite gas migration control system was put into continuous operation.
Methane was detected northwest of the landfill in concentrations over
75%, north in concentrations over 25%, and east and southwest in
concentrations over 40%. Methane was detected in the deep soil zone in
all directions from the landfill and in concentrations over 90% in some
areas.
C-15
-------
LANDFILL
A* M« Of protttl (no
»*»» may not M shown
0 SCO 1000
FIGURE C-2
Shallow Gas Concentration
Isopleth, February, 1986
C-16
-------
LANDFILL
Note AJ s«ts of (xo«$ and
wefts may not be shown
SCALE IN FEET
0 500 1000
FIGURE C-3
Deep Gas Concentration
Isopleth, February, 1986
C-17
-------
Field measurements were taken using the following instruments:
• Hnu PI-101 photoionization detector (total non-methane volatile
organic compounds)
• MSA 361 portable detector (combustible gas, hydrogen sulfide,
and oxygen)
• Fyrite carbon dioxide analyzer
Gas samples were collected in Tedlar bags and shipped. Gas samples were
analyzed for Hazardous Substances List volatile organic compounds (HSL
VOC) by gas chromatography/mass spectrometry (GC/MS) in accordance with
USEPA Method 624 guidelines.
This analysis found that subsurface gas collected from the onsite
gas extraction wells and flare manifolds contained a wide variety of
substances, including 23 USEPA HSL VOCs. The compounds found most
frequently and in the highest concentrations in the onsite subsurface
gas included ethylbenzene, vinyl chloride, total xylenes, toluene, and
benzene. The maximum concentrations of these compounds were in the low
parts-per-million (ppm) range.
TABLE C-3
ESTIMATED ONSITE/OFFSITE ATTENUATION OF PRIMARY
HSL VOCS DETECTED IN SUBSURFACE GAS SAMPLES
CAS#
100-41-4
1 33O-20-7
108-88-3
71-43-2
1 00-42-5
75-01-4
Target
Compound
Ethylbenzene
Xylenet (total)
Toluene
Benzene
Styrene
Vinyl Chloride
MAXIMUM CONCENTRATION
On-Site
(ppb)
16,610
29.195
24,044
1,384
508
31,215
Off-
Site
(ppb)
127
106
68
185
134
275
Atten.
(%)
99.2
99.6
99.7
86.6
73.6
99.1
MEAN CONCENTRATION
On-Site
(ppb)
2.825
3.419
1.920
318
41
2.607
Off-
Site
(ppb)
44
19
18
28
18
35
Atten.
(%)
98.4
99.4
99.1
91.2
56.1
98.8
Attenuation (%) = Onsite Concen. (Dob) - Offsite Concen. (ppb) x 100
Oniite Concen. (ppb)
C-18
-------
Vinyl chloride was frequently associated with the BTX-group
compounds in onsite subsurface gas, but during sampling offsite, it was
found at only two of the gas monitor probe locations, both south of the
landfill.
The analysis concluded that offsite migration of at least some of
the contaminants consistently identified in landfill gas has occurred,
possibly in all directions away from the landfill. The basis for this
conclusion is the presence of the BTX-group compounds in gases found
onsite and offsite. However, offsite concentrations of most BTX-group
compounds were attenuated by more than 98 percent (Table C-3). Further-
more, no BTX compounds were found beyond 2,300 feet from the landfill
perimeter. Because BTX-group compounds are also present in gas from
other sources, their presence in a particular sample of offsite gas is
not conclusive evidence that the gas originated from the landfill; it is
only further evidence that the gas may have originated in the landfill.
Mechanisms for Gas Movement
Where gas migration potential exists, the inherent slight pressure
of gas being generated by decomposition, fluctuations in barometric
pressure, and fluctuations in water table may create mechanisms for gas
movement.
The low gas generation pressures in the landfill initially move the
gas away from the points of origin through the paths of least resis-
tance. The generation pressures are estimated to be in the range of two
inches of water column, based upon measurements in other landfills.
Changes in atmospheric pressure of -9.5 inches of water column are
common during the winter months as high pressure ridges give way to low
pressures during storms. This can create a relatively strong pressure
differential between the air or gas within the soil or landfill and the
air at the ground surface.
When barometric pressure is falling, the air or gas within the soil
or landfill tends to flow toward the ground surface through the paths of
least resistance. When barometric pressure is rising, air from the
ground surface tends to move down into the soil both vertically and
C-19
-------
horizontally. The combined effect of the generation pressures and the
barometric changes is a pumping action within the soils.
C.2.2 Prevention Of Gas In Structures
Several residences adjacent to control well installations typically
recorded levels of methane above 10,000 ppm despite repeated efforts to
seal the foundations of those structures until the control wells were
turned on. According to gas monitoring data, all gas was effectively
removed from these structures within one day of the startup of the
offsite control wells. Further, gas has remained out of structures in
the vicinity of the landfill since the implementation of the offsite
control wells. Combustible gas above 100 parts per million has not been
recorded in a structure in the vicinity since November 1986.
C.2.3 Comments On The Study
The study demonstrated that landfill gas migrated from the site
independently from contaminated groundwater. It demonstrates the diffi-
culty in preventing soil gas intrusion by retrofit patching of existing
buildings and the effectiveness of soil gas extraction wells. The data
suggests that vinyl chloride is migrating in the direction of groundwa-
ter flow rather than with near surface gases.
Because of the high concentrations of methane detected in the
initial phase of the investigation, few data were obtained related to
other gases that might also be migrating. This initial oversight, while
understandable, significantly reduced the data needed for an accurate
baseline assessment.
C-20
-------
C.3 BUILDINGS OVER GROUNDWATER PLUME (A)
The site is located at the southern end of San Francisco Bay. In
1982 and 1983, it was discovered that three facilities (identified here
as A, B, and C) in the city had leaking underground tanks contaminating
soil and groundwater. All three were placed on the National Priorities
List (NPL). Facility C has since been removed from the NPL and is being
regulated under the Resource Concentration and Recovery Act (RCRA).
Although each site has its own source of pollution, the off-site
groundwater pollution areas have merged and the sites and are being
treated as one unit.
The area is heavily populated with commercial, light industrial,
and residential use. According to the 1980 Census information, there
are approximately 4,000 people living in the general area of the off-
site groundwater contamination:
The off-site area is determined by the extent of the groundwater
contamination of volatile organic compounds. The plume of contaminated
groundwater extends approximately one and three-quarters mile north of
the site. This area is composed of light industry and residential
housing. The former junior high school is located west to northwest of
the site.
The closest residential neighborhood begins about 1/4 mile north of
the site. According to the 1980 Census, of the approximately 1,500
housing units, 54% are single family homes and 24% have 10 or more
units. The area consists of modest single family homes interspersed
with small apartment complexes. Within 3/4 mile of the site is a mobile
home park which is primarily occupied by senior citizens.
A former elementary school houses approximately 200 children per
day. These children attend either a state child development program, a
state preschool, or a YMCA day-care program and a Head Start Program,
all operating at the school facility. There is at least a half acre of
grassy land on the side of the facility that is not included in the
fenced off playground. This is used as a neighborhood playground.
C-21
-------
C.3.1 Nature and Extent of Problem
Chlorinated volatile organic compounds (VOCs) are the major
contaminants tested for and found in soils and groundwater. Highest
concentrations were found around the leaking underground tanks. The
tanks and much of the contaminated soil were removed from the sites in
1983 and 1984. Soils with a total VOC concentration as high as 15,700
parts per million (PPM) were excavated. Soil boring samples were
obtained in 1988 at the A and B Facilities. The data are in Table C-4.
Groundwater
The VOC contamination in groundwater is currently monitored by 29
wells on the A site and 30 wells on the B site, in addition to 83 off-
site wells. There are 7 extraction wells on the A site and 6 extraction
wells on the B site. There are 23 off-site extraction wells. The on-
site extraction wells have been in operation since 1985, and the off-
site wells began operation in 1986 and 1988.
Three major water-bearing zones (aquifers) - defined as the A-,
B-, and C-aquifer zones - exist at the site. The A-aquifer is the
shallowest and the C-aquifer is the deepest of these three zones. The
B-aquifer is further divided into the Bl through B5 zones. The approxi-
mate depths (below ground level) at which these zones occur at the A
site are as follows: A: 10-28 ft; Bl: 28-50 ft; B3: 70-90 ft; B4: 90-
110 ft; and B5: 110-123 ft. A regional aquitard, the B-C aquitard, is
reportedly located at the depth range from 100-150 feet below ground
level. The C-aquifer, which supplies most of the municipal water in the
region, is located below the regional aquitard.
The aquifer zones appear to consist of mostly discontinuous layers
and lenses of fine to coarse sand, gravels, and often a substantial
proportion of clay and silt. These predominantly discontinuous layers
and lenses are separated and/or isolated by low-permeability clays and
silts (aquitards).
C-22
-------
o
CO
CO
en
o
ts>
00
o
o
<
I-H
t— 1
o
2
"§
a
c
o
re
4->
0)
o
o
U
'6
w
o
c
'5
m
a>
TJ
"to
c.
0
C
jo
0
T>
1
o
k.
TO
'5
CT
U
re
t
M
3
W
Aquitard
o
o
epnsqns
TJ
C
TJ
CN
CO
TJ
C
TJ
c
0>
TJ
J5
6
_c
•*
10
CO
CO
TJ
C
CO
CN
1
0
o
d
„.
0
o
d
TJ
c
•Dichloro-
e
31
u o>
TJ
C
CO
LO
O
d
CN
CO
o
TJ
CN
CO
CN
O
d
•oethylene
o
^
_o
'C
1-
c
CO
d
TJ
c
T>
C
0
c
0)
c
loroethyle
.c
u
re
4-»
0)
I-
n
c
CO
TJ
TJ
C
CO
CN
TJ
CN
"^
CO
o
d
V
®
richloroethan
i-
i
*
-
CD
0
0
O
TJ
C
8
o
TJ
C
05
CN
r—
TJ
C
C
«
c
^
,2-Dichloroett
U)
C
re
4-*
LO
O
cp
TJ
C
(^
CN
T—
d
i
TJ
C
O
d
CO
TJ
CO
CO
1
o
d
V
oethylene
•r
^
(j
H
if)
CN
0
O
0
TJ
C
^.
8
d
i
TJ
C
O
in
CO
TJ
c
loroethylene
JZ
u
k.
V
4-»
4^
C
CN
1
TJ
C
hlorobenzene
o
0
1
s
c
o
•a
o
•8
S -o
S £
o «
•
C-23
-------
Based on the well monitoring results at the site, it appears the
bulk of the VOC contamination is confined to the A-, Bl- B2-, and B3-
aquifers. However, low concentrations of VOCs have been detected in the
B4-aquifer. On-site VOC concentrations in the fourth quarter of 1990 at
the A and B Facilities are given in Table C-5 and C-6.
Off-site VOC concentration in groundwater for the fourth quarter
of 1990 are given in Table C-7. The data were from predetermined wells
that were not necessarily in the middle of the plume. The maximum
concentration of trichloroethylene, 1, 2-dichloroethylene, 1, 1-dichlor-
oethane, and tetrachloroethylene exceeded drinking water standards.
Soil Gas
On March 12, 1991, the soil surface emission of organic vapors was
analyzed using a flux chamber at off-site locations just to the north of
Facility A and near the former Junior High School. The soil moisture
content and ambient temperature during this sampling period were not
typical for this area and would tend to under-represent typical organic
vapor emission from the soil. Another sampling round was conducted on
April 28, 1991. One sampling location north of Facility A was repeated
while the rest of the samples were taken on the former elementary school
property. The data are given in Table C-8. [Note: The ATSDR draft
report on this site states that data collected using the neutral
pressure isolation flu chamber may be inappropriate for estimating flux
into buildings.]
Indoor Air Monitoring
Indoor air monitoring has been conducted only at Facility B. On
May 16, 1991, a crude analysis for organic compounds was made using a
hand-held detector. All readings were below the limit of detection (0.5
to 1.0 ppm). On May 18,1991, sampling was conducted inside and outside
the building. One-hour samples were collected at each location five
feet above ground. The results are in Table C-9.
C-24
-------
U
LU
_1
OQ
cc
UJ
Q
Z
D
O
CC
(D
Z
CO
Z
o
<
cc
UJ
o
Z
o
o
o
o
UJ
t
CO
I
O
O
ration
c
c
o
o
o
o
'5
CT
S
w
00
m
0}
? I- "°
.= C o
•^ 5 -5
5*1
w
q
c-i
8
01
i
g
i
u>
00
i
§
| Dichloromethane
'
?
i
o
i
R
i
1
o
c
a
Dichloroflourome
o
i-
?
in
0
e
1 ,2-Dichloroethai
'
-------
(O
I
U
LU
_l
m
cc
LU
<
Q
Z
D
O
CC.
CO
•z.
g
i-
cc
LU
O
•z.
o
o
o
o
LU
CO
I
o
CO
U
2
ncentration (ppbl
0
o
Q>
'5
CT
m
5
<
M
III
a *
CO
•0
c
o
1
TD
O
CM
•o
C
o
o
CM
,1,1 -Trichloroethane
r-
0
in
to
i
c
O
CM
•6
c
•O
6
o
CM
r—
CO
T—
o
V
k>»
U.
c
o
o
CO
c
CO
CO
c
in
d
inyl chloride
>
CD
•6
C
0
O
q
CN
in
•6
c
o
CO
OJ
•6
o
T) •
C-26
-------
UJ
_i
CD
f
cc.
UJ
<
Q
Z
D
O
CC
05
Z
O
<
cc
LU
U
Z
o
o
U
o
111
u.
LL
o
c
o
'+-•
2
*->
CO
u
c
o
0
CO
*»•
cr
<
3
CD
2
CN
CO
»—
CO
<
V)
0 w "S
111
i« ^i n
Q > 2
u
^f
r»'
i
TJ
C
TJ
C
O
«—
TJ
C
O3
op*
TJ
C
9
•o
b
t
omethane
c
_o
u
0
^t
o
TJ
C
CO
o
TJ
C
IT)
CO
TJ
C
CO
CN
TJ
|
b
O
f—
§
o
s-
o
o
sz
o
TJ
C
TJ
C
TJ
C
0
*—
i
TJ
C
CO
tN
TJ
jj
tn
hloroethane
o
5
*
r-
*t
cp
TJ
C
TJ
C
fc^
r^*
i
•D
C
0
o
r^
TJ
c
(^
TJ
U
O
o
CN
richloroethane
H-
r—
LT>
? —
TJ
O
O
CN
-ri
W
TJ
C
O
CO
1
TJ
C
CO
CN
TJ
C
b
o
CN
T"
n
c
0
(U
til
CO
CN
TJ
C
TJ
C
TJ
C
tn
CM
TJ
c
^
«
TJ
u
CO
hloroethylene
o
b
T—
't
T—
TJ
C
TJ
C
T)
C
O
CO
1
T)
C
co
TJ
ja
CO
hloroethylene
u
b
i
CNJ
00
CO
TJ
C
o
o
»r
CN
TJ
C
O
o
t
CN
TJ
C
O
o
*f
CM
TJ
C
O
CO
up
TJ
C
U1
oethylene
U-
O
JT
o
*c
(-
TJ
C
TJ
TJ
C
T)
C
CN
TJ
C
u
UD
loroethylene
^
u
TO
*-»
^
>
0
>
G
s.
8
*
0)
o
i
c
re
i:
*•"
w
*~ - j
eg o
2 c a)
*•; 5 .c
Jn — ~
« JO
0) -C O
TJ 'C •*-
5^d d
2S5 5
ro re «
s - 1 1
3-1 .2-j £
« " -5 " -"5
5 S o 5 o
a ffl TJ « TJ
o £ SI £ S
c o o o o
„ S: ca *: Q.
11 « 2 « S
0 O Q.CJ Q.
Z « ^ U -Q
C-27
-------
cc
o
0_
op
6
LU
_l
m
LU
H
op
LL
LL
O
(O
y
LU
I
U
y
z
DC
O
c
o
o
a
CO
u.
co
a
c
CO
£
CO
V)
o
2
c
0)
u
c
o
u
en
o
CD
m
|~
'o
re
LL
£
0)
X
er Junior
School
E
0
LL
CO
£
'o
re
LL
C
c
c
c
ichloromethane
Q
c
c
c
c
o
ichlorodiflouromethan
Q
c
c
•o
c
c
,1-Dichloroethane
t—
00
co
o
00
d
CO
ID
d
CO
r-
d
•6
CO
CO*
6
,1,1 -Trichloroethane
i—
iD
co
CO
d
in
en
d
CO
TJ
O)
CD
d
d
CO
o
c
c
c
•a
,1-Dichloroethane
r-
C
T3
C
co
U)
d
•o
Q)
s-1 ,2-Dichloroethylen
'o
c
•o
c
•o
•O
0)
C
ans-1 ,2-Dichloroethyl
4-1
O
CO
o
TJ
C
•D
•a
richloroethylene
1—
c
c
•a
c
c
etrachloroethylene
H-
c
c
en
d
•6
•o
c
,2-Dichlorobenzene
T-
2
E
c
g
o
4-<
T3
(D
= not detected abovi
= not tested
•0 4-
c c
C-28
-------
CD
1
0
LU
i
m
^
^~
CJ
Lt_
LU
I
H
^
QC
1-
LU
CD
2
^
Z
(/)
^
O
2
LU
I
O
u
z
(3
OC
0
c
o
ss
5
Q)
Q.
U)
C
ro
a.
c
a>
'&
0)
ro
U)
c
o
4^
2
4->
C
0)
c
0
u
^
a
c
'J5
'5
CO
0)
2
<2
D
O
a
c
'•6
(i
c
•o
"Z
(O
LU
LU
at
c
13
u
c
u.
or
u
c
a>
£
o
it
O
Q>
a>
2
o
4-<
CO
C
o
a
a
3
CO
c
0
1
TJ
8
CL
O
o
oo
c
c
c
H
s
c
Dichloromethane
i
c
4-<
4-*
C
€
c
Dichlorodiflouro-
methane
o
CO
c
c
4-*
C.
c
c
c
1 ,2-Dichloroethane
IT)
in
CO
d
CO
o
CO
o
CM
CO
0
CO
6
1,1,1-Trichloro-
ethane
i
c
4-»
C
4^
4->
«
CO
o
0)
LL
IT)
T3
C
C
C
•o
c
•o
c
•o
c
Vinyl chloride
,
•o
c
•a
c
T>
•o
C
•o
•o
1 ,2-Dichloroethylene
0
CO
CM
CM
O
O>
O
0
6
tn
o
6
V
Trichloroethylene
OT
T3
CM
O
0
T5
1
r—
o
Tetrachloroethylene
o
o
CO
•o
c
•a
c
•o
c
c
•o
•o
c
Dichlorobenzene
13
E
c
o
u
M
V
•o
0)
8
C
J "O
D Q)
U W
•^ 09
3 o
c C
II II
3 4^
4)
c
0)
N
C
0)
8
5
^:
u
'•5
*-«'?4
u
T3
a>
a.
a
=re£
O .
C-29
-------
C.3.2 Indoor Air Modeling Results
Estimates of indoor air concentrations were made, using models,
for current residences, the former elementary school, and possible
future apartment complexes. The Farmer model (see Appendix A of this
report) was used to estimate emission flux at ground surface. Based on
assumptions about building construction (area of the flooring through
which gas could infiltrate is the dominant factor) and building air
exchange rates, the indoor air concentrations were then calculated. The
modeling results are given in Tables C-10 through C-12. [The tables show
that identical groundwater concentrations and contaminant vapor phase
concentrations were used for the average and plausible worst cases.]
Interpretation of Results
The Agency for Toxic Substances and Disease Registery (ATSDR)
reviewed the above data (as well as other data not presented here) and
estimated cancer and non-cancer risks based on their standard methodolo-
gy. The analysis indicated that only current off-site residences (not
including the former elementary school) have any increased risk from
exposure to chemicals volatilizing from the groundwater plume. The
upperbound lifetime excess cancer risk was estimated to be 1 x 10"4 (low
increased risk) for average case conditions, and 2 x 10~3 (moderate
increased risk) for the plausible maximum conditions. The summary of
the ATSDR report states that the site is an indeterminate public health
hazard because of potential human exposure to concentrations of chemi-
cals that upon long-term exposure could cause adverse health effects.
C.3.3 Comments on Case Study
1. As noted by ATSDR, use of the flux chamber to check emission
flux predictions of the model at ground surface may be valuable, but
flux chamber data may be inappropriate as flux source data for infiltra-
tion into a building due to negative pressure in the building.
2. The modeling report states that convective flow due to
building vacuum (ie. stack effect) is generally more important than
simple diffusion into the building. Only diffusion into the buildings
C-30
-------
U
LU
CQ
U
o
LL
O
Z
g
^
cc
\-
z
LU
o
z
o
0
cc
J<
cc
o
o
o
z
LL
o
z
o
h-
LU
CO
0
cc
LL
r-
LU
Q
CC
(D
Z
O
Q
CO
LU
U
Z
LU
Q
CO
LU
CC
(J
z
Q 0
UJ CJ
5 <
P £E
9
UJ
r»
CO
27E-01
CM
O
UJ
CO
o>
CM
LU
r»
n
9
Ul
CO
97E-03
^
-309
LO
CO
9
UJ
LO
N
0)
9
UJ
CO
00
9
UJ
n
37E-02
t
o
9
UJ
CM
CO
P)
9
UJ
O
09
03
O
CM
O
CM
richloro-
thylana
H- «
9
UJ
CD
n
n
CM
9
UJ
S3
9
UJ
LO
cn
CM
9SE-03
CM
9
UJ
8
"
P)
9
UJ
o
LO
O
^
--
ans-1,2-Di-
hloroethylene
^ o
CM
9
UJ
r~
CO
68E-01
*
CO
9
UJ
o
o
09
9
LU
0
35E-01
P)
o
9
Ul
8
0)
9
UJ
O
0)
ri
CM
CM
P)
i
LL
§
UJ
p^
n.
n
9
UJ
CO
09
9
Ul
LO
^
0
UJ
CD
CM
n
9
LU
CM
P)
9
UJ
CM
fN
O
9
Ul
O
03
^
CM
9
Ul
o
03
CM
P>
<6
n
o
,1,1-Trl-
hloroathana
•- o
C-31
-------
O
LU
CO
f
O
o
U-
O
Z
o
ENTRATI
O
Z
0
o
cc
cc
0
0
Q
z
LL
0
Z
o
\-
F
LU
1
SCHOOl
^.
CC
^
^—
Z
UJ
UJ
UJ
cc
UJ
cc
o
LL
d
Q 0
UJ U
t—
ESTIW
INDOOR >>
UJ
i
x
u.
<
I
O
o
o
u
Vl
VAPOR PH
O
z
o
u
cc
UJ
GROUNOW/
S E
£ ? *2
lit
n
IK
= 2
? vJ
5 I
< 3
In
il«
II
5 >-5
Q •£
- =s
* « ^
1 -^ o E
o*
E
1 I
I11
_J
O
2
Ul
U
9
UJ
q
r>
eo-308
^
.58E-O9
ID
CO
9
Ul
V
fM
9
Ul
(S|
2.71E-02
in
9
UJ
O
*~
9
UJ
to
*~
2
r)
o
«- j:
*" *
n
1.61E-0
18E-O4
"*
O
Ul
CT)
CO
(SI
_
UJ
JJ
"'
n
9
UJ
r»
a>
1.97E-03
9
UJ
CD
CO
(Si
9
Ul
n
4.37E-02
0
9
UJ
(N
CO
P)
9
UJ
O
CO
00
o
CN
O
(S|
richloro-
thylene
1- 0
9
UJ
CD
CD
6OE-O4
r*
O
UJ
*~
UJ
ID
0)
9
UJ
(SI
CO
ui
6.82E-O4
c
o
0
u.
o
9
UJ
CSI
03
61E-03
^"
.15E-09
"
O
UJ
CD
O)
CM
r)
9
Ul
(SI
7.21E-O3
CD
9
UJ
8
N
(S
9
UJ
O
CD
(SI
r>
CD'
r>
CD
I
,1,1-Trl-
iloroetha
•- o
C-32
-------
CN
i
u
LU
CO
1-
CJ
o
LL
O
Z
0
r-
^
CC
f—
Z
LU
O
Z
o
u
cc
<
cc
o
o
Q
Z
LL
O
Z
o
h-
co
LU
U
Z
LU
Q
(/O
LU
CC
LU
r-
w
1
Z
0
LU
CC
D
H
D
LL
LU
o
Q 0
uj u
<£
5 <
w o
UJ O
Z
p
^
cc
X
u.
O
5
UJ
X
0
(J
Z
o
o
Ul
CC
O
Q.
>
O
O
(J
cc
UJ
0
Z
o
cc
o
i E -
S 3 CO
2 g >
C 5 3
01 CO
s 1
*E?
"S c CM
5 x E
1? * C»
CL 2 —
* "S
ft f^
« E
< S
1 i co
si£
«I«
1 co
1 1
< ^
i •*
-• Si
t j a E o
I " 1 'e g
i o S
E
i 1
•g
U!
o
0
UJ
I
o
o
UJ
CO
CD
,_
CM
O
Ul
•9
"*
§
Ul
CO
CM
*-
o>
o
Ul
0)
co
CM
O
Ul
03
en
in
S
Ul
CD
cn
8
Ul
0
r-
O
UJ
in
in
cri
in
en
1,1-Dichloro-
ethylena
CM
o
Ul
in
fx
CM
CO
o
Ul
CM
^
8
Ul
in
CD
•-
o
Ul
01
*
s
111
CM
CM
8
UJ
CM
CN
8
Ul
o
ls-
CM
O
UJ
0
CM
2
CO
Tetrachloro
ethylene
o
Ul
CO
o
in
O
Ul
O
•~
8
Ul
03
co
8
Ul
in
rx
CO
r-
O
Ul
S
CM
O
Ul
3
8
Ul
CM
CO
8
Ul
o
0)
03
O
CO
in
o
CD
in
Trichloro-
ethylena
o
Ul
0
CM
CM
8
Ul
o
in
8
Ul
CD
*-
O
Ul
CO
03
CO
8
Ul
co'
8
Ul
CD
8
Ul
fx,
0
*•
O
Ul
o
en
CD
O
CM
O
CM
Vinyl chloride
0
LU
in
-------
was considered. No explanation for not considering convection was
given.
3. Whereas gas was considered to infiltrate buildings through
the entire floor area for current private residences, infiltration
through only a 0.5 cm crack around the perimeter was allowed for the
former elementary school and future residences. For example, for the
school, this assumes only 0.08 percent of the soil gas flux enters. For
future residences the assumption is 0.2 percent flux entry. The
assumption that gas entry rates are a direct function of the percent
cracked area is not supportable and could result in significant under-
predictions of indoor air levels.
4. The modelling approach made no attempt to correct for the
capillary fringe above the groundwater. While this is appropriate for
an initial simple screening approach, it should be recognized that this
could lead to large overestimations of contaminant flux to the surface.
5. Incorrect Henry's Law Constants were used leading to a
potential overestimation of flux by a factor of about 50. (Non-dimens-
ional values used in a dimensional equation).
Because of the opposing influence of item 2 through 5, the
relevance of the model predictions is uncertain. Recognizing that the
initial modeling approach was simplistic, in retrospect it would have
been preferred to conduct more refined modeling before releasing the
data or conducting on-site monitoring.
C-34
-------
C.4 BUILDINGS OVER GROUNDWATER PLUME (B)
The site (Figure C-4), a former printed circuit board manufactur-
ing facility, is located at the southern end of San Francisco bay. The
facility operated from 1970 to 1985. The U.S. Environmental Protection
Agency (EPA) placed the site on the National Priorities List (NPL) on
February 1, 1990.
The primary site of both wet and dry manufacturing processes was
the "wet floor" building where waste waters containing heavy metals and
organic compounds were released to the "wet floor" and then discharged
to the neutralization sump. Other solvents and wastes were placed in
drums and disposed of off-site. It is still unclear exactly how the
chemicals discharged to the "wet floor" or the neutralization sump were
released to the soil and then the ground water.
In 1985 manufacturing stopped. As part of the plant closure
process, soil and ground-water investigations were initiated. The
investigation showed chlorinated volatile organic compounds (VOCs) and
metals (copper, lead and nickel) are the major contaminants found in the
soil and ground-water samples. The highest concentration of organic
contaminants in the soils was found around the sump and the "wet floor".
The sump and surrounding contaminated soil and the contaminated soil
beneath the "wet floor" were removed.
Seven wells on the site and 24 off-site wells (Figure C-4)
currently monitor the VOC contamination in the groundwater. The ground
water monitoring indicates that the plume extends north underneath
Interstate 101 (1/8 mile) and west of the site approximately a quarter
mile. Remediation of the groundwater began around 1986. Currently
there are 3 extraction wells on the site and 4 extraction wells off-
site.
The site consists of a number of high-ceiling one-story buildings
now occupied by other businesses. Light industry and businesses also
rent the other four buildings that were once part of the manufacturing
complex.
C-35
-------
KEY
• Monliorinf well
A Extraction weU
Conumintted pound wttcr pturac
Figure 1. Contaminated ground-water plume.
FIGURE C-4
C-36
-------
According to 1990 census information, approximately 1000 people
live in the 10 block area surrounding the plume of groundwater contami-
nation connected with the site. The bulk of the population resides on
the south side of Highway 101, with 1990 census information showing only
10 people living on the north side of Interstate Highway 101.
Land in the area surrounding the site has been used for residen-
tial, commercial, and agricultural purposes for many years. The areas
surrounding the site are zoned for a mixture of multiple family and two
family residential units, and general and limited industrial use.
Within the four block vicinity of the site, there are 409 housing units
with many of these being 14 unit complexes.
The area north of Highway 101 is zoned light industrial and
planned community designation. This type of zoning is designed for
research and development, office space, and light manufacturing. No new
housing is allowed under this zoning. Within this area approximately 20
acres of land are used for commercial farming. Two schools are located
northwest within a half mile of the site. The nearest city park is one-
half mile southwest of the site.
Four major water-bearing zones (aquifers)-defined as the A-, B-,
intermediate - and C- zones-exist at the site. The A-aquifer is the
shallowest and the C-aquifer is the deepest of these three zones. The
approximate depths (below ground level) at which these zones occur at
the site are as follows: A: 10-20 feet; B: 30-40 feet; and intermediate:
60-75 feet. A regional impermeable zone, the B-C aquitard, is reported-
ly located at a depth range from 100-150 feet below ground level. The
C-aquifer, which supplies most of the municipal water in the region, is
located below the regional aquitard.
C.4.1 Nature and Extent of Problem
Between August 1985 and August 1986, the neutralization sump and
some surrounding soil (70 cubic feet) was excavated and sequentially
backfilled to grade level with pea gravel, well-compacted clay, and
asphalt. In September 1986, about 255 cubic yards of soil were excavat-
ed under the "wet floor".
In May 1987, soil samples were taken at five feet depth from
borings at the sump and soil around it and analyzed for organic contami-
C-37
-------
nants {"wet floor" soil was not analyzed). Results are shown in Table
C-13. Results of the post-remediation analyses showed the presence and
levels of various organic contaminants not originally found in the
neutralization sump area.
TABLE C-13
ORGANIC CONTAMINANTS IN ON-SITE SUBSURFACE SOIL
1, 1-Dichloroethane
1 , 1 , 1-Trichloroethane
1 , 1-Dichloroethylene
t-l,2-Dichloroethylene
Trichloroethylene
Sump
Before
(4/85)
nd
10
7.3
nd
nd
Sump
After
(5/87)
nd-24
nd-69
nd
nd-110
41-220
" Wet "Wet
Floor" Floor"
Before After
(10/86)
na
na
na
na
nd-100
na
na
na
na
na
nd = not detected above detection limits
na = not analyzed
Since 1986, 7 on-site monitoring wells have been installed to
characterize the vertical and horizontal extent of contamination. The
most contaminated wells are located near to or downgradient, north to
northwest, from the sump excavation. The migration of contaminants
through the major water-bearing aquifers is being monitored. The two
shallowest aquifers (A and Bl) are contaminated with dichloroethane,
1,1,1-trichloroethane, 1,1-dichloroethylene, trans-1, 2-d-
ichloroethylene, and trichloroethylene at concentrations exceeding state
or federal drinking water standards (Table C-14).
Three extraction wells were installed on-site: two draw from the
A-aquifer and one draws from the B-aquifer. The untreated ground water
is discharged by permit into the sanitary sewer system.
C-38
-------
TABLE C-14
ORGANIC CONTAMINANTS IN ON-SITE MONITORING WELLS
Data are taken from the quarterly monitoring well report, fourth
quarter, 1990. The historical range of contaminant concentrations for
each aquifer is given with the highest concentration currently measured
in that aquifer given in parenthesis.
1,1- & 1,2-Dichloroethane
1,1,1 -Trichloroethane
1 ,1-Dichloreothyiene
t-1 ,2-Dichloroethyiene
Trichloroethylene
Drinking Water
Standards
0.5'
200'-"
6"
10C
5'-b
Aquifer Concentration (ppb)
A B Intermediate
nd- 13 (4)
nd- 45 (6.2)
nd- 17(1.4)
nd- 36(7.03)
nd-1 22(20)
nd- 20(nd)
nd-1 3,000(1. 8)
nd- 1 ,000 (nd)
nd- 14(4.3)
1.3-7,500(26)
nd
nd
nd
nd
nd
nd = not detected above detection limits
•California MCL
"Federal MCL
'proposed California MCL
Starting in late 1986, 24 off-site monitoring wells were installed to
assess the extent and severity of the migration of contaminants from the
site. The two shallowest aquifers (A and B) were found to be contami-
nated with organic compounds. Six chlorinated organic compounds
identical to those found at the site, have been detected in the ground
water (Table C-15). Concentrations of dichloroethane, 1,1,1-trichloro-
ethane, 1,1-dichlorethylene, trans-l,2-dichloroethylene, and trichloro-
ethylene, exceed drinking water standards.
Four extraction wells were installed off-site: two draw from the
A-aquifer and two draw from the B-aquifer. The untreated ground water
is discharged by permit into the City sanitary sewer system.
C-39
-------
TABLE C-15
ORGANIC CONTAMINANTS IN OFF-SITE MONITORING WELLS
Data are taken from the quarterly monitoring well report, fourth
quarter, 1990 (10). The historical range of contaminant concentrations
for each aquifer is given with the highest concentration currently
measured in that aquifer given in parenthesis.
1,1- & 1,2-Dichloroethane
1,1,1 -Trichloroethane
1 , 1 -Dichloreothylene
t, 1 ,2-Dichloroethyien8
Trichloroethylene
Drinking Water
Standards
0.5'
200'*
6'
10e
5'-k
Aquifer Concentration (ppb)
A B intermediate
nd- 330(64)
nd- 2500(1 40)
nd- 420(47)
nd- 580(580)
nd- 1599(520)
nd- 310(310)
nd- 1 100(12)
nd- 350(140)
nd- 550(550)
nd- 320(200)
nd
nd-1.0(nd)
nd
nd
nd
nd = not detected above detection limits
•California MCL
"Federal MCL
"proposed California MCL
The concentrations of site-related contaminants in ground water
are at levels that would be of potential public health concern if
domestic use of the contaminated ground water occurred. No municipal
wells currently exist near the site. There are several private wells in
the area or directly within the contaminated ground-water plume, but
most of these are not active. Of those wells that are still active, the
water pumped from these wells is not currently used for domestic
purposes. At this time, sufficient water for municipal use is available
from aquifers having higher quality water and water yield. Additional-
ly, regulatory barriers exist to prohibit installation of shallow
private wells.
A theoretical model (Farmer-See Appendix A) was used to estimate
the contaminant soil gas flux and the air concentrations within single-
family residences located above the contaminated plume. Air concentra-
tions for an "average case scenario", "plausible maximum scenario", and
a "most plausible case" were calculated. These scenarios use different
C-40
-------
assumptions for area of infiltration (crawlspace or crack around
perimeter of building), the fraction of air that infiltrates from the
crawlspace, and/or the air exchange rate of the home or business.
Based on the concentrations of organic contaminants accumulating
in a house as developed by this model, the risk of developing cancer
from inhaling volatilized organic contaminants for current residents
living above the contaminated ground water was calculated. The exposure
was assumed to be 24 hours per day, 365 days per year for 40 years .for
residents. The exposure was assumed to be 8 hours per day, 260 days per
year for 40 years off-site workers.
The lifetime excess cancer risk from the inhalation of volatiliz-
ing organic contaminants from the shallow ground water was estimated
from the addition of the individual risks to the three potential
carcinogens, 1,1-dichloroethylene, 1,1-dichloroethane and trichloroeth-
ylene.
Lifetime excess cancer risk for off-site residents ranged from 1.8
x 10"8 ("no increased risk") for the least conservative estimate and 2.9
x 10"5 ("no apparent increased risk") for upperbound worst-case condi-
tions. Lifetime excess cancer risk for off-site workers ranged from 2.1
x 10"9 ("no increased risk") for the least conservative estimate and 3.5
x 10"6 ("no increased risk") for upperbound worst-case conditions.
Potential risks were also assessed for the noncarcinogens (trans-
1,2-dichloroethylene and 1,1,1-trichloroethane) for residents above the
contaminated ground-water plume. The analyses indicated that the
inhalation of organic contaminants from off-site ground water would not
result in adverse noncarcinogenic effects.
C.4.2 Conclusions
Based on information reviewed, ATSDR and CDHS concluded that the
site is not an apparent public health hazard. As noted in the human
exposure pathways section above, off-site resident and worker exposure
is predicted by an air-model but the exposure is at a level below that
of health concern.
Future significant exposure to ground-water contaminants is
unlikely if the ground-water extraction and treatment system reduces
concentrations of site related contaminants to below levels of health
C-41
-------
concern, no wells currently in place are used for domestic purposes, and
future drinking water wells are not placed in areas of known contamina-
tion if ground-water remediation does not clean up contaminants to
drinking water standards.
C.4.3 Comments on Case Study
As we reviewed this case study, we were struck by the similarity
to the preceding case study. The two sites are both located at the
southern end of San Francisco Bay and are only a few miles apart. They
appear to have similar hydrology and geology. Based on the reported
chemical concentrations in the ground water plumes for the two sites,
the groundwater appears to be substantially more contaminated at the
site discussed in this section. Based on the relative lifetime excess
cancer risks for current off-site residents for the two cases, it
appears that for this second case indoor air model calculation used a
diffusion only method and a flux attenuation factor based on an infil-
tration area represented by a 0.5 cm crack around the building perime-
ters. We note that this is extremely more restrictive (yields lower
predicted indoor concentrations) than the assumptions used for current
residents in the preceding example but is similar to that used for the
former elementary school and future residences in that case example.
This assumption is likely overly restrictive and insupportable in both
case examples.
For illustration purposes only, we have calculated indoor air
concentrations for off-site structures above the ground water plume for
the current site using the same parameters used for current off-site
residences in the previous case example. We also ignore corrections for
the capillary fringe (as did the study authors) because only a screening
level result is desired. [This correction should be made if a better
estimate is needed.] Because we did not have access to sufficient data
to calculate the geometric mean concentrations in the ground water, as
was used in the previous case, we have used one-half the maximum values
C-42
-------
TABLE C-16
HYPOTHETICAL CALCULATIONS FOR INDOOR AIR
IN STRUCTURES ABOVE THE GROUND WATER PLUME
Compound
1 , 1 -Dichloroethylene
Trichloroethylene
1 , 1 -Dichloroethane
Trans 1 ,2-Dichloroethylene
1,1,1 -Trichloroethane
Sunnyvale
Off-site cone.
(max) in
ground water
4.3
120
2.5
11
6.3
Mountain
Viow"1
Off-site Cone.
(max) in
ground water
47
520
64
580
140
Modeled Indoor Air Concentrations (ug/m3)
Sunnyvale
AVG
4.24
5.50
0.087
0.46
0.87
Plausible
Max
16.4
21.3
0.34
1.77
3.37
Mountain View
AVG
23
12
1.1
12
9.7
Plausible
Max
90
46
4.3
47
37
(1) - For calculation purposes, one-half this value was used in model
C-43
-------
given in Table C-15. The results are in Table C-16. These calculated
results indicate that under these modeling assumptions, indoor air near
the site would be substantially more contaminated than in the previous
case. Thus, it is likely these estimates are much too high.
C.5 BUILDINGS OVER GROUNDWATER PLUME (C)
The site is located in a suburb of Casper, Wyoming. A residential
subdivision is located approximately one-quarter mile northeast of an
industrial park. An oil refinery is located west-southwest of the
subdivision on the opposite side of the North Platte River.
In the mid-1980's, a contaminated water plume was discovered
underlying part of the residential subdivision. Because the residences
all use wellwater, an indoor air study comparing periods with and
without showers operating was conducted.
Air sampling was performed in five residences with contaminated
wells, five homes with uncontaminated wells and five Casper area homes
as controls. Each home was sampled during a 5-hour period with the
shower operated at least 10 minutes and on a subsequent day, another 5-
hour period without shower operation. In addition, 5-hour basement and
5-hour outdoor air samples were collected. Samples were obtained with
low-flow air pumps and Tenax collectors and were analyzed for volatile
organics. The study was designed to compare shower vs. non-shower
concentrations and to compare results among the three study groups to
see if significant differences existed in indoor concentrations.
C.5.1 Nature and Extent of Problem
Organic contaminants found in groundwater samples were used to
divide the residents into two groups, contaminated (Group 1) and
uncontaminated (Group 2). In the residential air study, these two
groups were compared to background homes in the Casper area (Group 3).
Residential sampling included indoor air with shower (IAS), indoor
air without shower (IA), basement air (BA), and outdoor air (OA)
collected in the backyard of each home. In addition, a questionnaire
was administered at each home to identify potential alternative sources
of airborne contaminants.
C-44
-------
The air/Tenax analysis was designed for quantitation of HSL
volatile compounds. Tenax tubes were used for sample collection. The
Tenax tubes can retain most volatile HSL compounds (lowest boiling,
i.e., chloroethane excluded), but do not adequately retain many light
hydrocarbons and low boiling compounds, i.e., saturated hydrocarbons
-------
O3
C
cc
-z.
UJ
O
•z.
o
o
r-
U
UJ
_i
CO
o
U
y
z
CC
o
UJ
z
cc
o
CQ
CC
O
Z
o
cc.
t
o
UJ
UJ
CD
Z
a
o
0
a
cS
Outdoor
c
E
o
V)
CD
S
C 0
O .C
W yj
*
c S
o o
« £
f -
°- i
0
TJ
3
O
Basement
o
c o
o ^
2 to
o
Q.
C $
0 0
111 -C
£-
°- $
<*J
co" *
CO —
3
1 I
m
^
^
rZ «
Benzene
CO
• in
1 I
t 1
1°
! 1
i !
. .
1 i
o
•o
•c
z
Carbon Tetrac
I i
1 I
1 I
1 1
1 I
in
d *•;
3
: i
; i
0
c
.c
1,1-Dichloroe
1 I
1 1
1 1
1 1
1 1
§3
l i
i :
Ethylbenzene
I i
I I
1 I
1 !
1 I
1 1
! i
*~
2-Hexanone
6 •*:
7 o
3
I !
1 1
1 1
1 1
3
co •-;
3
i°"
Methylene
Chloride
I i
I !
v—
^2
3
O <">
3
1 1
23
f J-
3
in" >"
n —
3
0)
3
Tetrachloro-
ethylene
: ;
(N
0 CO
3
i°'
! 1
0)
"° 9
3
CO
' ui
3
CO
"9 «•>
3
Trichloro-
olhylene
?5
3
i :
f*s
in «
o> ro
3
en
6 °\
3
«2
in
in
iri PI
O) »-
CO t
'
1 1
(*^
!*• "?
3
P>
r-.'
CM
en i;
CO (si
O
to
d
r^
en ^;
^ *~
in en
3
Toluene
C-46
-------
T3
03
C
O
o
O
LLJ
CO
cn ^~
(j M
C
O
•£ ^
'•6 ^j
3
(0
Q.
3
O
<5
o
o
•5
3
0
1
isement
m
$
c o
o - to
d
(O
4 oi
Benzene
1 I
d "
3
9 d
3
° d
3
3
O
T3
•c
O
Carbon Tetrach
! 1
I 1
I 1
1 1
: :
i ;
c
03
1,1-Dichloroeth
1 i
1 I
1 I
1 I
! !
1 1
Ethylbenzene
1 i
1 I
1 i
CO ^j
3 *~
i !
cn ^_
3 *~
2-Hexanone
m "
n
in •
3 *~
: i
i i
CO
r^" ^
CO 01
3
in
co q
3
Methylene
Chloride
I I
I I
D "J
i0'
in
6
-------
In order to calculate airborne concentration, the flow rates were
adjusted to conditions of standard temperature and pressure. The
compound weights were blank corrected by subtracting mean weights of all
field blank analytes from the reported field sample value prior to
dividing by standard liters of air to calculate the flow rate.
Descriptive statistics, including mean and range of concentration
were calculated for each sample. When calculating mean values, one-half
the instrument detection limit was used for non-reported values. The
data were lognormally distributed and samples were randomly selected.
Data were analyzed statistically using the Mann-Whitney (MW) U-
Test for independent groups. The data were then logarithmically
transformed and analyzed using the more powerful t-test for basement
samples and Analysis of Variance (ANOVA) for indoor and outdoor samples.
One tailed probability values (p) are reported as an indication of
statistical significance.
The MW test converts the original data to ranks and compares two
independent groups of data. Ranks are used due to the fact that the
small size and high variability of the data do not met the assumptions
of normality or homogeneity of variance necessary to use parametric
statistical analysis. The MW test closely approximates the parametric
t-test. It is less powerful than the t-test in finding a difference
between two means if one exists.
A questionnaire was completed by each participant to identify the
following potential sources of airborne organic contaminants: cigarette
smoking, use of water operated appliances, use of organic chemicals,
house cleaning, and open windows during sampling. The results indicated
that Group 2 had more smokers, greater use of appliances, more house
cleaning, and more open windows during sampling. Participants were
asked not to use chemicals during the sampling period.
C.5.2 RESULTS
Shower Compared to Non-Shower Exposure
Samples collected with and without a shower from homes in Group 1
were compared to determine whether groundwater contaminants entered
indoor air via showering. As expected, PCE, TCE, and TCA levels were
C-48
-------
slightly higher in IAS samples compared to IA samples. A statistical
comparison of the data indicated that the results were not significant
at the (ANOVA) p<0.10 level. The highest significance level was found
for toluene at (MW) p<0.34. IA and IAS samples also contained virtually
the same amounts of contaminants when comparisons were made for Groups 2
and 3.
Airborne Contaminant with Shower Exposure in the Three Groups
Airborne contaminant exposure in IAS samples were assessed in the
three study groups to determine the contribution of contaminants from
drinking water to residential air. Mean values were higher in Group 1
homes, compared to Groups 2 and 3, for PCE, TCE, and TCA. However, the
values were not found to be statistically significant at the (ANOVA)
p<0.10 level. The highest level of significance was found for TCE
between Groups 1 and 2 (MW) (p<0.20). Significance levels for other
comparisons ranged from (MW) p<0.27 to (MW) p<0.42. The lowest levels
of PCE and TCA were found in Group 3 homes. TCE levels were very
similar in Groups 2 and 3.
Basement Air Samples
None of Group 2 homes had basements, consequently, comparisons
were made between Groups 1 and 3. PCE concentration was higher in the
Group 1 samples at a significance level of (t-test) p<0.05. Mean TCE
concentration was higher in Group 1, but not significantly (t-test
p<0.17). The other contaminants were not tested statistically due to
the small difference between mean values.
Ambient Air Samples
Outdoor air concentrations of all contaminants were consistently
lower in Group 1 compared to Group 3 indicating that they did not
contribute to indoor levels in Group 1. TCE and PCE were not found in
Group 1 outdoor samples, but were found in Group 3 outdoor samples.
Toluene and benzene levels were lower in Group 1 samples at significance
levels of (MW) p<0.26 and (MW) p<0.20, respectively. TCA levels were
slightly lower in Group 1 samples compared to Group 3 samples.
C-49
-------
C.5.3 Conclusion
Contaminants found in previous well water samples from homes in
Group 1 were also present in indoor air. Mean levels of PCE, TCE, and
TCA were higher in Group 1 homes compared to Group 2 and 3 homes (not
statistically significant at p<0.1 level). The fact that outdoor
concentrations of these contaminants were relatively low, even though
the primary wind direction was found to be from the southwest emanating
from the industrial park toward the subdivision, suggests an indoor
source of contaminants. Showering was investigated as a potential
source. TCE, PCE, and TCA mean concentrations were found to be slight-
ly, but not significantly higher in IAS as compared to IA samples
collected in Group 1 homes.
PCE concentrations in basement samples were significantly higher
(t-test; p<0.05) in Group 1 homes compared to Group 3 homes. The source
of PCE in basements may have been vapor directly from contaminated
groundwater or from appliances that use water, such as hot water heaters
or washing machines, located in the basement. If groundwater were the
source, TCE levels would also be expected to be elevated. TCE values
were higher in Group 1 homes, but not at as significant a level (t-test
p<0.17) as PCE. A possible explanation for this result is that PCE
partitions from water to air more readily than TCE and, consequently,
may volatilize from groundwater and accumulate in residential basements
at a higher rate than TCE. The air/water partitioning of the two
contaminants was determined using Henry's Law constants at constant
atmosphere: PCE, 1.1 x 103; TCE, 5.5 x 102. The larger of the Henry's
Law constant, the greater the equilibrium concentration of the compound
in air compared to its concentration in water. It should also be noted
that the sample size for the basement sample comparison is quite small,
three homes per group. Very small sample size makes a statistical
comparison of the data difficult and consequently, makes the data
comparison inconclusive.
The influence of other potential sources of airborne contaminants
were investigated. Prior to sampling, all participants agreed to avoid
use of household or other chemical during the sampling period. Group 2
homes had increased numbers of cigarette smokers, use of water operated
C-50
-------
appliances, use of organic chemicals and house cleaning activity during
sampling. These factors would be expected to increase the airborne
levels of the same volatile organic contaminants found in well water, as
well as increased volatile organics from cigarette smoke and organic
cleaners and solvents. However, the measured indoor air contaminant
levels were not consistently elevated in Group 2 homes compared to
Groups 1 and 3.
Cigarette smoking may not have been an important factor in
contributing to the airborne contaminant levels presented in Table 2-17.
According to the Surgeon General Report, 1981 (Wynder, E. and Hoffman,
D.), none of the airborne contaminants identified in the study were
found as "major toxic agents" in cigarette smoke. The report does say
that the list presented was incomplete and adds that cigarette smoke may
contain "such carcinogens as volatile chlorinated olefins". This group
may contain PCE and TCE.
The fact that more windows were open during sampling in Group 2
homes may have lowered airborne indoor contaminant concentration. This
would negate the effects of the previously discussed factors that tend
to increase contaminant levels.
The levels of contaminants found in homes in all three study
groups were typical of indoor concentrations found in enclosed living
spaces.
Outdoor mean contaminant levels in the three groups and 24-hour
samples were also found to be lower than those in the urban areas of
four major U.S. cities. (Houston, St. Louis, Denver, and Riverside).
C.5.4 Comments on Study
This study was fairly complex in that it monitored at three type
homes, with and without showers operating, as well as ambient air. It
comes to the conclusion that the indoor air quality in all cases is
within the "typical" range for residences. It should be noted that
because none of the Group 2 homes had basements, whereas group 1 and
Group 3 did, and the possibility of soil gas intrusion from the contami-
nated groundwater plume existed for Group 1 homes, the Group 2 homes
would not appear to be an adequate control group. However, the design
C-51
-------
has left open the question whether or not soil gas is infiltrating the
homes even at these low levels by presenting an inconclusive statistical
result indicating a significant difference from the controls.
Extraction wells were later installed at this site to control
groundwater migration.
C-52
-------
C.6 BUILDING NEAR SLUDGE DISPOSAL PIT
The site is located in a rural section of Brunswick County, North
Carolina (Figure C-5).
The site was used by sludge hauling and oil spill cleanup compa-
nies for the disposal of septic tank sludge, oil sludge, and other waste
materials from 1969 to 1976. These wastes were disposed in shallow
(approximately two-to eight-feet deep) unlined pits or directly on the
land surface at the site. A residential subdivision now exists in the
area of the former disposal areas. Typical dwellings in the subdivision
consist of manufactured homes situated on one or two-acre lots, each
having a private domestic well. There are approximately 150 residential
lots in the area, 70 of which were occupied in 1990.
In May 1976, the owner was informed that an oil disposal pit
violated North Carolina statutes and must be cleaned up immediately.
This pit was approximately 60 feet long, 20 feet wide, and two to four
feet deep. At the time, it was estimated that approximately 2,000 to
3,000 gallons of black oil were contained within the pit. The owner
pumped the oil from the pit and then covered the pit with soil.
Documentation pertaining to the chemical composition of materials
disposed in the pit, the fate of the liquid removed from the pit, and
the quantities and characteristics of the material buried on site is not
available.
C.6.1 Nature and Extent of Problem
In August 1976, the failure of an earthen berm allowed approxi-
mately 20,000 gallons of black oil to escape from an unlined pit on the
property and flow into an on-site creek. The oil remaining in the pit
was pumped and transferred to tank trucks and hauled away. The bottom
sludges and some oil-stained soils were excavated and disposed in the
County Landfill. The remaining thicker sludges, which could not be
C-53
-------
SITE LOCATION MAP
SOURCE: USCS OUAtXUNGLE ACME.
NOflTH CAROCXA 1514.
CONTOURS AND ELEVATIONS
IN METERS
SCALE
FIGURE C-5
C-54
-------
pumped, were mixed with sand and buried on site; however, the burial
location was not documented.
In 1980, the property changed hands through foreclosure and
subdivided for residential development. Family A purchased two lots in
the subdivision in 1982 and 1983 and in July 1983, uncovered creosote,
oil spill residue, tank bottom sludges, and septic tank sludges in the
yard.
State of North Carolina personnel sampled and analyzed soils at
the site in response to the discovery. The investigation confirmed the
presence of these materials in soil close to the surface at locations on
all sides of the house. The well was contaminated with phenols (100
mg/1), dimethyl benzene, methyl phenol, one unknown organic compound,
creosote, and septic tank sludge. The shallow well was subsequently
condemned and the house was connected to a neighbor's well system.
In September 1983, EPA and the Region IV Field Investigation Team
(FIT) conducted an investigation of the property. The investigation
included an electromagnetic survey of the area; initial ambient air
monitoring under the home; and collection and laboratory analysis of
surface water and sediment samples, five subsurface soil samples, and
six groundwater samples from off-site and on-site wells. All readings
of ambient air were negative. Based on surface water and sediment
sampling and analysis, no contamination of Chinnis Branch was detected.
Groundwater analysis for inorganic compounds indicated the presence of
seven priority pollutant constituents. Groundwater analyses for organic
compounds (purgeable, extractable, pesticide/PCBs, and other chlorinated
compounds) indicated the presence of 81 different compounds, many of
which are commonly associated with creosote or coal-tar derivatives.
Soils analyses identified a total of 24 inorganic constituents in the
subsurface soil samples, 10 of which are target compound list (TCL)
parameters. Soils analyses also indicated the presence of 46 organic
compounds in the subsurface samples, 14 of which are TCL parameters, and
many of which are polynuclear aromatics associated with asphaltic and
coal-tar derivatives.
In February 1984, EPA used ground penetrating radar (GPR) to
further delineate the site boundaries. The GPR survey revealed two
C-55
-------
anomalous areas that roughly correspond to the location of two surface
impoundments illustrated on historical maps. The pits were estimated to
be six feet deep toward the center and three feet deep toward the edges.
In March 1984, an Immediate Removal Action (IRA) at the site was
requested by the EPA. The IRA at the property consisted of the excava-
tion and removal of approximately 1,770 tons of oil sludge and soils
which were transported to a hazardous waste landfill in Pinewood, South
Carolina. Soil cleanup activities were completed in April 1984.
In May 1984, EPA installed nine groundwater monitoring wells at
the site and conducted sampling and analysis of groundwater samples for
volatile organic compounds. Relatively high concentrations (in the ppm
range) of benzene, ethyl-benzene, toluene, and xylenes (BETX) were
detected in the groundwater samples from the shallow aquifer. Groundwa-
ter sampling indicated that the deep aquifer had not been affected by
the contamination in the shallow zone. Neither the upgradient nor down-
gradient deep wells contained detectable volatile organic compounds.
EPA recommended that these monitoring wells be regularly monitored.
In 1988, the groundwater monitoring wells were resampled and
samples analyzed for purgeable organics and some inorganic parameters
(metals, nutrients, etc.). As was the case with the EPA 1984 data, the
1988 monitoring well data indicated that gasoline or waste petroleum
product (especially benzene, toluene, and xylenes) were still prevalent
at the site. However, the concentration of purgeable organics detected
during the two sampling episodes varied significantly. In addition, the
1988 data indicated the possibility of low level benzene, ethyl-benzene,
and xylenes in a deep well which would indicate that an aquifer underly-
ing the surficial aquifer had now been affected.
In 1989, EPA determined that the level and extent of on-site
contamination warranted a more thorough investigation and assessment.
Consequently, a Remedial Investigation (RI) was undertaken through a
CERCLA action. The purpose of the initial RI was to assess the nature
and distribution of contaminants at the site and to provide the data
necessary for developing a Feasibility Study (FS) and ultimately
conducting a Remedial Action (RA).
C-56
-------
The field investigation of the initial RI was conducted from
January 1990 through April 1990, and included soil gas surveys, surface
and subsurface soil sampling, monitor well installation, groundwater
sampling (of monitor wells and residential wells), hydraulic conductivi-
ty testing, and identification of potential ecological and human
receptors. Three separate study areas were addressed during the initial
RI. Area 1 was comprised of the Family A Property, which is the
original site. Area 2, located approximately 1.5 miles from Area 1 in a
similar rural residential area, was identified during the document
review. Historical records indicated that a waste oil disposal pit
existed within Area 2. Area 3, located adjacent to Area 1, was identi-
fied in historical records as a potential waste disposal site based on
historical aerial photographs of the area. Area 2 was not well docu-
mented and could not be located even after extensive interviewing and a
thorough reconnaissance of the area. Therefore, no field investigation
was initiated. A majority of the field investigation activities
subsequently centered in Areas 1 and 3.
The initial RI report concluded that the extent of contamination
which poses a risk to human health or potential ecological receptors is
limited to the area in the vicinity of the Family A residence. Contami-
nation has impacted surface and subsurface soils, the shallow aquifer,
and creek sediments in this area. The predominant media and compounds
contributing to public health risk include: benzene and lead in ground-
water, and carcinogenic polynuclear aromatic compounds (PAHs) and lead
in surface soils.
C.6.2 Residential Air Monitoring
Air samples were collected in February 1990 in the crawl space and
inside the residences of Family A and Family B. Sampling was conducted
at these two residences since they are situated on or near the former
waste disposal pits. Unlike the Family A residence, the Family B
residence is not located above or adjacent to identified contaminated
areas.
A total of five residential air samples were collected using EPA
method TO-14 from within the crawl spaces and interiors of the two
C-57
-------
homes. The air samples were analyzed for the 34 TCL VOCs listed in
Table 2.18. Only three compounds, chloromethane, methylene chloride,
and 1,1,1-trichloroethane were detected. Methylene chloride was
detected inside the Family B residence at a concentration of 11 ppbv.
Low levels of chloromethane (16 ppbv) and 1,1,1-trichloroethane (1.5
ppbv) were detected in the crawl space beneath the Family B residence.
No VOCs were detected within or beneath the Family A residence.
Of these three VOCs, 1,1,1-trichloroethane was the only compound
detected in sampling in the vicinity of the site and it was only
detected in low//g/kg concentrations in background soil samples. None
of these VOCs were detected in known contaminated areas. These
observations suggest that the source(s) of VOCs in the Family B resi-
dence are not related to the site.
C.6.3 COMMENTS ON INDOOR AIR EFFORT
The fact that no VOCs were detected in or under the Family A resi-
dence is counter to published information on typical residential indoor
air quality. These data are suspect.
Perhaps of more significance is what was not sampled and analyzed
for. The soils analysis demonstrated the site is contaminated with
creosote, polynuclear aromatic hydrocarbons, phenols, and other hazard-
ous compounds. Even low concentrations of some of these can result in
substantially elevated risk numbers. Vapor pressures are high enough to
consider their transport.through soil gases especially considering
contamination was found in the immediate vicinity of the residence.
C-58
-------
TABLE C.18
TARGET COMPOUNDS
COMPOUNDS
CHLOROMETHANE
BROMOMETHANE
VINYL CHLORIDE
CHLOROETHANE
METHYLENE CHLORIDE
1,1-DICHLOROETHENE
1,1-DICHLOROETHANE
CHLOROFORM
1,2-DICHLOROETHANE
1,1,1-TRICHLOROETHANE
1.1.1 -TRICHLOROETH ANE
CARBON TETRACHLORIDE
DIBROMOCHLOROMETHANE
1 ,2-DICHLOROPROPANE
TRANS-1 ,3-DICHLOROPROPENE
TRICHLOROETHENE
BENZENE
DIBROMOCHLOROMETHANE
1 ,1 ,2-TRICHLOROETHANE
CIS-1 ,3-DICHLOROPROPENE
BROMOFORM
1,1,2.2-TETRACHLOROETHANE
TETRACHLOROETHENE
TOLUENE
CHLOROBENZENE
ETHYL BENZENE
TOTAL XYLENES
ACETONE
CARBON DISULFIDE
VINYL ACETATE
METHYL ETHYL KETONE
METHYL ISOBUTYL KETONE
METHYL BUTYL KETONE
STYRENE
1,2-DICHLOROETHENE (TOTAL)
C-59
-------
C.7 FUGITIVE DUST CASE STUDY
Primary zinc smelting operations at this facility (Figure C.6)
since the turn of the century emitted large quantities of zinc, cadmium,
lead, and copper into the atmosphere in the vicinity of the plant.
Significant concentrations of these heavy metals have been measured in
the soil within a large area surrounding the plant. The plant area is
currently a Superfund and a RCRA site. The facility is currently
operating under new ownership as a secondary zinc smelter. A NAMS
reference monitor is actively monitoring the area for attainment of the
lead NAAQS. The facility has been in compliance with the standard for
the past several years.
C.7.1 Nature and Extent of Problem
Investigations were conducted at two homes by the State in
response to citizen requests. Two dust sampling methods used were based
on protocols of the Center for Disease Control, Lead Poisoning Preven-
tion Branch, Division of Environmental Hazards and Health Effects,
Center for Environmental Health and Injury Control (CDC). The first
method used a vacuum pump and filter cassette to pick up dust from a one
square foot area. A plastic template was used to measure the area to be
sampled - except on window sills where the area was estimated. The
second method, based on the Kellogg Dust Protocol Number 2, was wipe
sampling. A Whatman Number 42 filter paper dampened with isopropyl
alcohol was used, although the original method specified alcohol swabs.
Field blanks were taken at each residence by momentarily opening
the cassette. The wipe sample field blank filter was removed from the
transport tube, moistened with alcohol, then immediately returned to the
transport tube. A new disposable latex glove was used to handle each
filter.
C-60
-------
PREVAILING WIND
X = AIR MONITORS
MINE TAILINGS
1 SMELTER
X COMPLEX
LOCATIONAL MAP
FIGURE C-6
C-61
-------
The U.S. Public Health Service/EPA Toxicological Profile Document
for lead uses a value of 300 /jg/g lead in household dust, as a baseline-
value to estimate overall population lead exposure from various media.
The urban Particulate Standard 1648 from the National Institute for
Standards and Technology (NIST), which is a composite urban dust sample,
has values of 6,550 jig/g for lead, 75 /ig/g for cadmium and 4,760 /zg/g
for zinc. The lead values obtained from the household dust in the two
homes were well above the national averages.
An indoor monitoring study was developed to determine the extent
of exposure to lead and heavy metals inside the residents homes.
Twenty-five homes were selected randomly in areas where there is a large
probability of deposition of airborne lead. Atmospheric dispersion
modeling with ISCLT was used to determine these areas. The homes were
vacuumed sampled with mini rams outfitted with HEPA filters.
Results from the indoor study demonstrated that most of the homes
in the area had very high levels of lead. However, since a health
threat had not yet demonstrated to exist due to the high levels of lead
found in the residential homes, ATSDR was contacted to determine whether
the environmental contamination is a public health hazard for the
residents.
ATSDR proposed an exposure study to determine the extent of the
health hazard, if any, in the town. Although the study has been
completed, the results have not yet been reported.
The study:
• Measured blood lead and urinary cadmium levels among a
representative group of residents
• Compared the levels of lead and cadmium found in residents
to levels found in a comparison community while controlling
for other risk factors known to influence exposure to these
heavy metals
• Performed a standardized panel of medical tests on both
groups
A Superfund Removal Action was scheduled to be performed at those
homes in with elevated lead levels. The objective was to clean the
homes and eradicate them of lead, thus reducing the probable health
C-62
-------
threat. However, Superfund is reluctant to perform this Remedial Action
until adequate evidence is available to identify the source of lead dust
currently infiltrating the homes so that recurrence can be prevented.
Data obtained in the 25 home sampling program showed that lead levels
inside the homes were much higher than in the outside soil. Although
this suggests an airborne pathway for the indoor air contamination, the
data did not conclusively establish whether the contamination was from
soil erosion, the smelter, or other sources.
The problem, therefore, has become one of determining the actual
current source of lead dust infiltrating the homes. Based on current
information, it can be demonstrated with reasonable scientific certainty
that:
1. Past primary Zn smelting practices have contributed to the
burden of hazardous substances in soil and dust
2. Present secondary Zn recovery processes are adding to the
burden of hazardous substances in soil and dust
3. Both 1. and 2. are "non-de minimus"
4. Other common anthropogenic sources are responsible for little
of the hazardous substance burden
The problem, thus, becomes one of determining whether the lead dust
is from the old primary smelter operations and is being tracked into
homes or carried in by wind erosion, or whether it is from wind dispen-
sion of the current secondary smelting operations.
There are five basic identification methods that can be used:
geostatistical, elemental composition coupled with multi-variate analy-
sis, chemical speciation, individual particle analysis, and lead isotope
ratios. In the geostatistical approach, the spatial variability of the
substance of interest is determined and displayed on isoplethic plots.
while this provides an easily understandable presentation from which one
can infer source locations, a large number of highly representative
samples are needed.
The elemental composition method requires determination of multiple
elements from both potential sources and receptors. The data are then
analyzed using multi-variate techniques such as ratio discriminant
C-63
-------
functions and factor analysis. While one can potentially learn a lot
from a few samples and infer sources by comparison of source and
receptor patterns, the method is not sensitive for subtle sources and
post-release chemical transformation can pose analysis difficulties.
Also, the method depends on having a fairly complete picture of all
potential sources.
Chemical speciation methods depend on determining the actual
chemical compound or "species" in source and receptors. Very selective
analytical techniques are required. The method suffers significantly if
post release transformations take place. For lead, many source com-
pounds may weather to lead sulfate making source identification diffi-
cult or impossible.
Individual particles may be analyzed microscopically and by various
analytical techniques to determine elemental and chemical composition.
This approach appears to combine the best features of the chemical
speciation and multi-variate methods. For example, particles from lead-
based paint would be associated with carbonates, chromates, titanium,
and zirconium whereas lead particles from batteries would contain
antimony.
The use of lead isotope ratios can potentially be a very powerful
technique, especially for the current problem. Lead ores from different
geological ages contain different ratios of the four lead isotopes
206Pb", 207Pb, 208Pb to the minor isotope 204Pb. Because of this, different
economically important lead ore bodies have different lead isotope
fingerprints as shown in Table 2.19. These ratios can easily be
determined using either thermal ionization mass spectroscopy (MS) or
inductively coupled argon plasma MS.
C-64
-------
TABLE C.19
LEAD ISOTOPE RATIOS IN WORLD-WIDE LEAD ORES
Source
Rosetta, S. Africa
Ivigtut, Greenland
Broken Hill, Australia
Bunker Hill, Idaho
Coeur D'Alene Group
Pine Point, NWT
Tintic, Utah
Casapalca, Peru
Metalline Falls, WA
Missouri Leads
206/204
12.5
14.8
16.0 -
16.2 -
16.1 -
18.2 -
18.5 -
18.6 -
19.5
Ratio
16.2
16.4
16.7
18.6
18.7
19.0
21.0 -24.0
It is suspected that lead ores used during primary smelter
operations were from a limited number of geologically well-defined
sources and the secondary smelting operations process lead from a highly
average group of world-wide sources. Thus, the lead isotope ratios will
be substantially different from these operational periods. If this is
the case, then the isotope ratios in household dusts and collected
human body fluids will provide convincing evidence for the current
source of the lead dust. Furthermore, it is likely the primary Zn
smelter emissions were reflective of sulfide ore feedstocks containing
Zn, Cd Pb, As, Se, Cu, Ga, In, Ag, and the slag was rich in iron and
manganese. Secondary zinc smelter operations, on the other hand, likely
reflect compositions of diverse ferrous alloy feed materials and contain
high levels of elements not associated with primary zinc smelting such
as Be, V, Co, Ni, Mo, Pd, Pt, and lanthanides.
Based on the scientific evidence and available analytical methodol-
ogies, the investigative course of action to be taken is:
1. Review existing site characterization data; identify source areas;
review site history (what ores were smelted). Review other informa-
tion such as RCRA records, Air Enforcement records, OSHA records -
(the smelter has been cited in the past for worker exposure in the
plant and subsequent high blood lead levels).
C-65
-------
2. Develop piggyback sampling strategy for residential areas; obtain
samples of sources (slag, primary smelter emissions, secondary feed
materials, waste piles).
Types of Samples Desired
Sources
1. Primary Zn smelter stack emissions
2. Slag piles
3. Pre-1980 air filters
4. Post-1980 (present) air filters
5. Fugitive sources of dust from secondary Zn plant (piles, bins,
etc.)
6. Raw materials received
7. Products made
8. Present day stack emissions, baghouse dust, etc.
Receptors
1. House dust
2. Soil from yards, public areas, etc.
3. Hillside soil
4. Park, etc. dust
3. Apply a combination of the following approaches:
a. Elemental composition + MVA
b. Lead isotope ratio analysis
c. Individual particle analysis
C.7.2 Comments on Study
The investigative procedures outlined are currently underway. The
value of this case example is that it illustrates that investigative
techniques are frequently available to distinguish impacts from specific
current sources from other current and past emission sources. Although
this is a very specific case involving smelters, the fingerprinting
investigative approach may be applicable to other investigations of
Superfund site impacts on indoor air quality.
C-66
-------
C.8 BUILDINGS OVER WASTE OIL PIT
The Site is located in Mobile County, Alabama. The 5.1 acre site,
about 11 miles north of Mobile, Alabama, is bounded to the east by U.S.
Highway 43, by a gas pipe line easement to the north, an undeveloped lot
to the west, and a residential development to the south.
According to available records, in March 1961 a trucking terminal
was constructed for intrastate and interstate commercial trucking opera-
tions. The Site was used for parking, maintaining, and cleaning trucks
and trailers. According to a salesman at the Site, from May 1966 until
late 1969, tank trailers were regularly and routinely washed at the
Site. The washwater was then discharged in the rear of the terminal
property where the ground level was low.
The salesman states that he was told that in late 1965 (before he
worked at the Site) levees were built around the terminal to contain
materials and an aboveground asphalt tanker on wheels was placed at the
back of the property for storing asphalt products to be returned to
customers.
In February 1971, the terminal operator was encouraged to move its
operations. In March 1971, an offer to buy the property was accepted.
Figure C-7 shows the terminal layout as of May 1972, including the
configuration of a dike built in 1971. In September 1971, six geotechn-
ical borings, were drilled at the Site to depths ranging from 40 feet to
55 feet. The data collected from these borings were used to design the
foundations of the Site. Logs of these borings were compiled on a
drawing that was part of the building architectural plans. These boring
logs show the stratigraphy underlying the Site consists of a sand
stratum, up to 7.5 feet thick, overlying a silty clay stratum, which
attains a thickness of up to 34 feet. Underlying the silty clay is
another sand stratum of which the thickness was not determined by the
borings. To date, a drawing showing the locations of these borings has
not been found.
C-67
-------
C-68
-------
In 1973, the property was sold to the property's current owner
which constructed an apartment complex (see Figure C-8) consisting of 13
buildings on the Site. Approximately 160 people live in the apartment
complex.
C.8.1 Nature and Extent of Problem
In 1977, the owner first became aware of tar-like substances
seeping to the surface. The Department of Housing and Urban Development
(HUD), which inspected the apartments on an annual basis, inspected the
seepage and recommended to the managing agent that the seepage be
excavated to a depth of 1^ feet and the area be filled in with sand.
This was done periodically as the seepage appeared. HUD continued to
inspect the apartments on an annual basis. The managing agent reported
the presence of the tar-like substance to the Department of Health.
They continued to excavate the property and fill with sand when any
seepage occurred. Continued appearance of the tar-like substance
prompted the manager of the apartment complex to report the matter to
the State in the fall of 1984.
The State investigated the Site and reported the matter to the U.S.
Environmental Protection Agency (USEPA) during the same year. The State
identified 18 individual seeps where a tar-like substance was oozing to
the surface and described the material as a tar or asphalt with a
vitreous luster. The material was described as very sticky with a
petroleum odor and at cooler temperatures was very plastic and viscous.
The investigation included six hand augured holes ranging from 1.1 feet
to 5.0 feet in depth. Samples were taken from two holes in the interior
grassy courtyard area of the apartments and one hole in the southwest
corner of the complex in a grassy area inside the paved driveway. One
analysis was taken from these three soil samples although it is unclear
C-69
-------
: - 0
FIGURE
C-8
C-70
-------
if this was an analysis of a composite sample. The sample contained the
following compounds and concentrations:
Compounds Concentration (ug/g)
Naphthalane 135.4
Acenaphthene 101.9
Fluorene 128.6
Phenanthrene 33.9
Anthracene 53.7
Benzene 0.05
Preliminary sampling by the USEPA in April and May of 1985 detected
concentrations of naphthalene, 1,2,4 trichlorobenzene and other com-
pounds in the tar-like material and soil. In February 1990, the USEPA
added the Site to the NPL with an MRS score of 30.83.
Compounds reported to have been detected in the subsurface 4 feet
from the southeast corner of the courtyard area included butyl ate
(20,000 ppb), vernolate (90,000 ppb) and a mixture of compounds normally
found in coal-tar products. Compounds reported to have been detected in
the subsurface 4 feet from the east-central courtyard included butyl ate
(2,000,000 ppb), vernolate (300,000 ppb), eptam (30,000 ppb), and coal-
tar products at a total concentration of (70,000 ppb).
No organic compounds were reported to have been detected in a
composite surface sample of scrapings of white material from nine areas
in the complex. However, this sample was reported to contain 37,000,000
ppb sodium, with a Ph of 8.7.
Samples SA-17, SA-18, and SA-20 were surface samples of tar-like
seep material. Sample SA-17 from the east side of the Site was reported
to contain vernolate (1,000,000 ppb), 1, 2, 4-trichlorobenezene
(9,300,000 ppb), coal tar products (total concentration of 380,000 ppb),
carbon disulfide (5,000-ppb) and two unidentified terpenes (20,000 ppb).
A petroleum based product was also reported to have been detected in
this sample. Sample SA-18 in the southeastern corner of the courtyard
was reported to contain butylate (70,000 ppb) and vernolate (300,000
ppb) along with some coal products (total concentration of 760,000 ppb).
A petroleum based product was also reported to have been detected in
this sample. East of the ditch by the playground, Sample SA-20 was
C-71
-------
reported to contain coal-tar products at a total concentration of
(2,160,000 ppb), xylene (2,000 ppb) and a petroleum product. The USEPA
samples of April and May of 1985 found no contamination at the sampling
points located outside or downstream of the reported position of the
former levee.
A Phase I RI was conducted in February and March of 1991. Phase II
was conducted in September of 1991. The investigation was quite
thorough and reporting all details is beyond the present scope.
Given in Table C-20 is a summary of the number of compounds found
on-site.
TABLE C-20
SUMMARY OF NUMBER OF CONSTITUENTS DETECTED IN EACH MEDIUM
Media Type/
Sample Location
Alluvial Groundwater
Surficial Groundwater
Soils
Ditches
Tar-like Material
Total
Detected
Constituents
41
66
110
68
69
Total
Constituents
with
Unqualified
Detections
27
37
40
20
34
Total
Constituents
with More than
One Unqualified
Detection
15
24
32
13
19
Table C-21 gives organic chemical concentration data from the
monitoring wells at the Site.
C-72
-------
r-
1
O
UJ
03
CD
'I
5
c
Q
II
_
"o
a
c
o
"c
o
5
II
5
5
VI
c
o
CO
£
-Q
"el <
gj =
II
0
o
0}
c
o
3
a
_J
Q
—
c
o
o
03
o
cc
II
UJ
cc
c
n
a
c
0
o
0}
c
o
a
0)
II
f2
u
'5
0
Z
o
J—
ui "J
0^
II) *-
ti U ^J
Q 2 (n
O ~™N
U
O O O O O O «— O ^" O 00 O •— LOroOOOOOOOOT-OOOO^l"O
•— OOOOcoOO*.OOCoOOOOtOOOOOMCScr)
co co ro co J CS «— «~ CJ »— ^-r-CO
--
a.
O
o
•z.
o
m
cc
o
o
cc
a
UJ
x
a.
cc.
O
X
UJ
X
uj X
O uj
iu —
O w
< CQ
CL
O
$5
_i ca
< cc.
x <
H O
X O
a. cc
X X
ui
X
X
a.
X
UJ
x a x
||H|§S
UJ CC Q UJ O O
CM O 7 CM Z I-
^ J " a £ S
ca > <
o w x
< ca u
g_
X
UJ
I
Q «V
V M
•
H X J
£?£
X
X
Q.
UJ
1 fc
i£ 2
: -7 N
X 5 . X S
UJ ', UJ UJ .
X Ci Z X -J
-J •. uj -JO
(E X
3x
5^
O
z
"I I £i fc
"t < " 3
CM u ca ca
o
ui
<
O
u ca
o. uj
UJ >
0
a
a
c
a
o
6
UJ UJ
OL CC
UJ UJ
cc cc
pooppopooqoqpoppooooopcorococococofpco
C-73
-------
•a
0)
c
o
o
o
UJ
_l
QQ
o
0
=
«5
o
o
II
—
o
£
o
c
o
'c
o
5
II
5
M
C
o
s
o
.O
"°l<
ll
.2
Q
II
_J
Q
c
o
i
X
o
ir
II
UJ
CC
0
JD
(0
r
o
o
$
GO
C
o
0
0
w
II
w
5
1
o
0
OOOfOOOOOOOOOOOOOOOOOOOOOOOO
pr-f-;(^qoqoqq'-;'-_oqqoqqqqqqoqqq
O)OU1
T 00
TJ- CQ CO
C-74
-------
Indoor Air Modeling
No actual monitoring was conducted inside the structures. All
estimates are based on modeling. [The modeling approach for soil gas
intrusion is described in Appendix A, Sections A.1.5 and A.1.6].
In this case, the modeling approach to estimate soil gas flux was
based on the assumption that the contaminants are initially uniformly
distributed vertically and horizontally throughout the soil. The model
then allows equilibrium to be established among chemicals adsorbed on
soil, chemicals dissolved in soil water, and chemicals in the soil gas.
The soil gas diffuses to the surface resulting in an ever-expanding zone
of clean soil from the soil surface downward. It appears that the model
assumptions result in an average flux over a ten year period. These
source modeling equations were based on and conform to those given in
EPA/600/6-86-002, "Development of Advisory Levels for Poly-chlorinated
Biphenyl (PCB) Cleanup."
The above calculation yields only the flux (mass of chemical
exiting the soil surface per unit area per unit of time, eg. g/m2-s).
To calculate the rate of entry into the structures, it was necessary to
make estimations of:
• the ratios of diffusive to convective flow of soil gas,
• the area of the structure through which soil gas could enter,
• the air exchange rate for the structure.
The ratios of convective to diffusive flow was calculated from the
Peclet number, a dimensionless parameter, that considers the diffusiv-
ity, soil permeability, and vacuum created by the structure. The
calculation indicated the diffusion mechanism dominated (see comments
section below). The area through which gas could enter the structure
was estimated as 0.1 percent of the floor area, based on literature data
for homes on slabs. The structure air exchange rate was estimated to be
about 0.7 air exchanges per hour, based on average wind speeds and
indoor-outdoor temperature difference.
In addition, outdoor air concentrations can contribute to indoor
air concentrations. Since the buildings are surrounded by the Grassy
Area, indoor air concentrations can be affected by vapors from the
Grassy Area. Thus, the total indoor air concentration was estimated by
C-75
-------
summing the modeled indoor air and 100% of the Grassy Area air concen-
trations (indoor exposure, excluding seeps). Seeps of tar-like material
may also contribute to the indoor air concentrations. Thus, total
indoor air concentration including vapor emissions from seeps was
estimated by summing modeled indoor air concentrations, Grassy Area air
concentrations (99.66%), and emissions from seeps of tar-like material
(0.34%) (indoor exposure, including seeps). This conservatively assumes
that all tar seeps are located in the Grassy Area and contribute 100% to
indoor air. The results, excluding contribution from seeps, are given
in Table C-22.
Data from all sources was used to complete a risk assessment. The
results for the current exposure are summarized in Tables C-23 and C-24.
(Risks for future exposure scenarios, which assume use of on-site wells,
are not included in the table, but are discussed below.
Total current cancer risks including exposure to seeps were
estimated to be between 7 x 10"5 and 2 x 10"4 for the various receptor
populations. The most important pathway contributing to the risk is
indoor inhalation of vapors containing PAHs from seeps of tar-like
materials. In particular, benzo(a)anthracene, benzo(b)fluoranthene, and
chrysene contribute about 90% of the risk from indoor inhalation.
Under the Future I scenario including seeps, the cancer risks range
between 4 x 10"4 and 2 x 10~3. Ingestion of water from the alluvial
aquifer accounts for most of the cancer risk, approximately 75% of the
total risk. Arsenic and beryllium constitute the greatest proportion of
the risk, with aldrin, BEHP and dieldrin also contributing to the total
risk.
The Future II scenario including seeps poses risks ranging from 3 x
10"3 to 1 x 10"2. Inhalation of indoor air vapors during showering
account for approximately 100% of the risk, and the risk is primarily
due to chloroform from the on-site surficial aquifer.
C-76
-------
CO
z
O
h-
cc
l-
z
UJ
o
(N Z
eg o
i °
LLJ £
CO
oc
o
o
Q
Z
Q
LU
_l
UJ
Q
O
c
a
i:
V
u
c
o
u
"o
9 £ i-0
- (3^-2
c —
l}|i
ujS
Oi on
||
a
III
m
a en
_ 3
^
l=\
a "~
si
i!li
u -^.
Itf
- • 1.
Contaminant!
?
O
in
1
s
•
3
ff>
§
CM
O
§
6.248-06
45.50471
6.078-02
S
id
a
ACENAPHTHENE
?
S
CM
1
O
ft
CM
O
e
to
CM
O
O
3
1.248-05
0.130183
8.588-04
o
§
Si
ACETONE
CM
§
n
(D
8
O
8
|
8
S
4.508-06
1
1
5
O
CM
CO
in
ALDRIN
?
to
CM
§
01
CM
O
O
p
8
O
O
s
5.778-06
128.2497
S
CO
S
o
CO
N
ANTHRACENE
^
CO
5
n
c\j
8
O
CM
in
8
i
(vi
O
8
s
in
1811.575
4.830-05
in
o
e
in
8
BENZO(A)ANTHRACENE
CM
£
CO
CM
m
p
CM
O
C
CO
CM
O
O
r-
to
S
c
in
CM
0
c
o
3
0
0
CT
ri
4.806-06
9384 613
3.948-O4
S
CM
0
CHLORDANE. ALPHA
CM
Ol
tn
CO
CM
*
5
fe
o
V
0
o
4.808-06
295.7S52
3.948-04
o
c
p
0
CHLORDANE, GAMMA
o
n
i-
0
I
8
«
8
CO
8
7.306-06
3.14816
9
in
CM
O
§
s
CHLOROBENZENE
?
to
o
1
o
at
q
O
ri
8
Cvi
1.048-05
2.073532
oj
S
CO
CM
O
CHLOROFORM
C-77
-------
n
OJ
D
,C
C
o
O
s^
CJ
«N
1
O
LU
m
t-
o
f-
oc
f— •
2
LU
O
Z
o
o
cc
^
oc
0
o
Q
^
Q
LU
_J
LU
Q
O
C
0
C
tJ
3
- 0 * ~
c *—
1 fl"e
UJ -2
&«
It
•'-3
=? • <
J| ^ W
H S 3.
I"5
a en
3~
f-T
1 c'V
a
•i
, |j
i a
i"f
"•^
If!
j
2
9
u
*-
c
*
^-
&
CM
3
2
8
•
2i
CM
8
X
^
9
fi
n
1811575
1
8
£
CM
§
1 CHRYSENE
2
$
0
CM
*
r~
0
*
CO
8
3
A
*"
?
3
^
1
9
3
2
*
8
P
c
X
"*•
i
5
9
in
0
r*.
••^
3
%
8
8
&
*
3
1
ri
E9GICTO
I
n
0
^
n
«
| DICAMBA
CM
tn
0
$
•^
8
e>
8
4
Q
«
8
5
a
9
r?
"
414.8168
1
3
fi
to
8
V
(M
*
0
1
•n
8
X
Ol
ri
8
8
f?
•*
9
1!
*r
1129.943
3
2
a
8
g
03
9
$
rj
CM
o
£
CM
s
<^
n
72.12009
?
1
CM
S
g
Ol
S
1 FLUORENE
^
i
CM
^
c?
*-
8
rj
8
^
y
•»
8
:?
n
9
1
»
43.51676
8
£
CO
m
s
1
in
5
CC
n
HEPTACHL
-
m
rj
,_
a
8
§
8
*
8
9
5
*
5.174339
|
3
0
K
§
.
Cj
cc
O
I
C-78
-------
TJ
0)
a
.£
•*-*
c
o
O
"^
C\J
OJ
O
LU
m
j_
O
H
oc
H
LU
0
•z.
o
0
DC
DC
O
O
Q
Z
Q
LU
|
LU
Q
O
C
c
o
• c
(5
o
1M||
-
§^_
•
3II-E
SlCji
ii
• "S •=
i ; fi
t- S ~-
t»
a en
^~
I"-*
= e1^
a
=f
= Ii
i-t
|ii
c
c
E
c
,3
OJ
C
C4
(M
4
8
i
8
+
§
*
8
$
*
?
S
in
232625
2.400-04
S
X
o
8
OL
m
CM
CN*
"
O
W7
8
V
8
+
£
•»
8
$
0
§
c
in
SZ9TCZ
3.050-05
8
X
a
m
o
m
csj
to
n
1
r-
8
a
8
O
8
9
tn
23.2625
\
8
2
n
a
(UNOANE)
S
5?
m
r>
f,
"
n
«
8
S
8
+
*
tfl
D
8
9
in
14389.85
2.800-O6
8
*
to
UJ
CDJPYREN
| INDENO(1.2,3-
0
t~-
to
o
i
o
o
V
S
^
?
*
0
(9
*
«
£
a
38.39342
2.280-01
S
&
«">
S
§
1 METHOXYCHt
^
1
CO
,_
i
*
8
*
8
o
8
Si
h-
«
0
to
0243466
1.910-03
0
§
o
R
i
!->
METHYL ETW
o
CT
0
3
8
S
m
9
Z
"
0
1.310-01
8
*
r-
3
UJ
o
1
UJ
S
._
*
r>
_
5
o
«
8
o
o
$
rj
9
§
in
0.764199
0.000057
8
S
1
MOLINATE
ft
i
f-
C?
0
2
n
o
-i*
to
09
S
+
in
•"
CM
O
N
rJ
9
1
in
10.42455
^
S
5
a
|| NAPHTHALEN
ft
(M
O
OJ
„
3
*
g
•*•
a
*~
o
fi
9
•8
in
I
6
g
S
g
PEBULATE
_
£
-------
TJ
Q)
a
_c
•4-*
C
o
O
CM
I
O
UJ
_J
m
cc
h-
2
HI
O
2
O
o
DC
CC
O
O
Q
2
Q
111
_J
HI
Q
O
1
•o
m
Ja
•
u
3
"o
(A
8 - • §•£
15 sf t
- 0 * —
ill!
>->,§£
?f
U
111
311
•
11
fi
II!
a
s!
• ?-s-
^$1*
-51
S5
|-g
• - T
Ift
•t
C
a
E
a
3
9
K
h~
?
2
rs
3
+
S
S
•t-
§
n
8
o
7.009-06
I
S
1.12» + 03
8
UJ
2
a
fc
C-80
-------
TABLE C-23
SUMMARY OF PATHWAY-SPECIFIC CARCINOGENIC RISKS
Exposure Scenario
RME Receptor (Wortt-Caee Scenario)
Sum of Adult
Cancer Rltk
Sum of
9-Year-OW
Cancer Rltk
Sum of
4-Year-OW
Cancer Rltk
Sum of 9-and
4-Year-OW
Cancer Risk
CURRENT EXPOSURE: No known documented cases of exposure at these levels
Target Area E
Ingestion of soil
Dermal contact with soil
Inhalation of vapor
Inhalation of particulates
Grassy Area
Ingestion of soil
Dermal contact with soil
Inhalation of vapors
Total Indoor Exposure
Inhalation of vapors (including seeps)
Inhalation of vapors (excluding seeps)
Ditch
Ingestion of sediments
Dermal contact with sediments
7x10"'
7 X 1 O'6
6 x 10'*
1 x 10"*
1 X 10'7'
1 X 10-"
9 X 10-"
5X 10*
2X10"7
0
0
2 X 108
2 X 10'6
6 X 10'"
1 x 10'9
6X 10"'
7 X 10"'
1 X 1 0"'
5X 10*
2X10'7
2X 10'7
3X 10"s
5 X 108
2 X 1 0'6
7 x 1 0'9
2x10'
5 X 1 0'7
2 X 10'*
1 X 10"
6X 10*
2 X 10'7
6 X 10'7
2 X 20'*
6x10-'
4 x 1 0'6
1 x 1 0'6
3 x.10-'
6 X 1 0'7
3 X 1 0'*
2X10"'
1 X 10-*
3X10'7
8 X 10'7
5X 10'*
EXPOSURE TO SEEPS OF TAR-LIKE MATERIAL: A««ume« no removal of teep* for RME Receptor*
Ingestion of tar-like material
Dermal contact with tar-like material
Inhalation of vapors
TOTAL CURRENT EXPOSURE
(EXCLUDING SEEPS)
TOTAL CURRENT EXPOSURE
(INCLUDING SEEPS)
4X 10'*
7X 107
5 X 10'*
7 X 106
5X 10"'
1 X 10*
5 X 10'*
3 X 1 0'6
9 X 10'6
2X 10'7
2 X 10*
5 X 10'9
4 X 10'6
1 X 10"4
3 X 1 0'7
3X 10-*
1 X 10-5
7 X 10-*
2 X 10"4
C-81
-------
TABLE C-24
SUMMARY OF PATHWAY-SPECIFIC TOTAL HAZARD INDICES
Exposure Scenario
RME Receptor (Worst-Case Scenario)
Adult
Nine-Year Old
Four- Year Old
CURRENT EXPOSURE: No known documented case* of exposure at theee level*
Target Area E
Ingastion of soil
Dermal contact with soil
Inhalation of vapor
Paniculate inhalation
Grassy Area
Ingestion of soil
Dermal contact with soil
Inhalation of vapor
Paniculate inhalation
Northern Ditch
Ingestion of sediments
Dermal contact with sediments
Indoor Exposure
Inhalation of vapor*, total indoor (including seep*)
Inhalation of vapor*, total indoor (excluding *eep*)
4 X 10 J
3 x 10'
2 X 10J
9 X lO*
3 X 10J
6 X 10-3
5 X 10'
0
0
1 X 10*
1 X10J
1 X 10'1
1 X 10°
3 X 10J
1 X 10'
3 X 1 0 3
7X 10-3
9X10''
3 X 10J
1 X 10'
2 X 10*
2X 10J
4x 10'
2 X 10°
4 X 10J
2 X 10 '
2 X 10'
2X 10'1
1 X 10^
7 X 104
1 X 10'1
2 X 10*
3X 10J
EXPOSURE TO SEEPS OF TAR-LIKE MATERIAL: A«ume« no removal of seep* for RME Receptor
Ingestion of tar-like material
Dermal contact with tar-like material
Inhalation of vapor*
TOTAL CURRENT EXPOSURE
(EXCLUDING SEEPS)*
TOTAL CURRENT EXPOSURE
(INCLUDING SEEPS)2
9 X 10'
7X10^
1 X 10J
4X 10'1
4X10'1
2X 10-4
2X 10"3
2X 103
1X10°
1X10°
9 X 10"4
3 X 10J
2X 10-3
2 X 10°
2 X 10°
C-82
------- |