xvEPA
United States
Environmental Protection
Agency
Off ice of
Solid Waste and
Emergency Response
Publication 9355.4-U-1
EPA-540/R-94/018
PB 94-963503
March 1994
Superfund
Technical Background
Document for Draft Soil
Screening Level Guidance
-------
Technical Background Document for
Draft Soil Screening Level Guidance
Office of Emergency and Remedial Response
U.S. Environmental Protection Agency
March 1994
U.S. Environmental Protection Agency,
Rtgion 5.Library (PL-12J)
11 W§st Jackson Boulevard, 12th floor
1L 60604-3590 —
-------
TABLE OF CONTENTS
Page
List of Figures v
List of Tables v
Preface vii
Acknowledgments viii
Part 1 - Definition and Application
1.1 Background 1
1.2 Purpose of SSLs 1
1.3 Summary of SSL Attributes 2
1.4 Limitations of SSLs 3
1.5 How To Use SSLs 4
1.6 Determining the Dilution/Attenuation Factor (Ground Water Pathway) 7
1.7 Application of Tiers 2 Through 4 (Ground Water Pathway) 8
1.8 Measuring Soil Levels 10
1.8.1 Exposure/Averaging Area 10
1.8.2 Sample Pattern 11
1.8.3 Number of Samples 11
1.8.4 Depth 11
1.8.5 Sampling for Background Contamination 12
1.8.6 Additional Sampling Needed for Ground Water Tier 2 12
1.8.7 Geostatistics 12
1.8.8 Sample Analysis 12
Part 2 - Technical Background
2.1 Human Health Basis 13
2.2 Direct Ingestion 16
2.3 Inhalation of Volatiles and Fugitive Dusts 18
2.3.1 Screening Level Equations for Direct Inhalation 18
2.3.2 Volatilization Factor 20
2.3.3 Paniculate Emission Factor 25
2.4 Soil Screening Levels for Protection of Ground Water 25
2.4.1 Conceptual Framework and Organization 26
2.4.2 Development of Soil Partitioning Equation 27
2.4.3 Organic Compounds—Partition Theory 29
2.4.4 Inorganics (Metals)—Partition Theory 31
2.4.5 Assumptions for Soil/Water Partition Theory 35
2.4.6 Dilution/Attenuation Factor Development 37
2.4.7 Soil Screening Levels 40
2.4.8 Site-Specific Adjustments to Partitioning Calculations
(Tiers 2-4) 41
Part 3 - References
References 45
in
-------
Appendices
Page
A Synthetic Precipitation Leaching Procedure (SPLP) (SW-846 Method-1312) A-1
B Limited Validation of the Hwang and Falco Model for Emissions of
Soil-Incorporated Volatile Organic Compounds B-l
C Evaluation of the Dispersion Equations in the Risk Assessment Guidance for Superfund
(RAGS): Volume I - Human Health Evaluation Manual (Part B, Development of
Preliminary Remediation Goals) C-l
D CoUected K^ Values D-l
IV
-------
List of Figures
Number Page
1 Pathways addressed by soil screening 2
2 Exposure scenarios not addressed by Soil Screening Levels 3
3 Tiered approach, ground water pathway 9
4 K^/K^ correlation plot 33
5 Measured and predicted Koc 33
6 PCP and TeCP: log (KJ vs. pH 34
7 Effect of pH on metal Kd values 36
8 Effect of site area on DAF 40
9 Soil to ground water pathway—calculating the DAF 41
List of Tables
1 Superfund Proposed Soil Screening Levels (SSLs) 5
2 Proposed Ground Water Pathway SSLs for Inorganics and
Pentachlorophenol, as a Function of pH 7
3 Effect of Facility Area on Dilution/Attenuation Factor (DAF) 8
4 SSL Target GW Level, Drinking Water Regulations, Health Advisories,
Soil Quantitation Limits 15
5 Chemical-Specific Input Parameters for Ingestion and Inhalation Pathways 19
6 K^ and Kow Values: SSL Organic Chemicals 32
7 Pentachlorophenol K^ as a Function of pH 34
8 Metal Kd Values as a Function of pH 36
9 Soil Screening Levels for Ground Water Protection: Organic Chemicals 42
10 Soil Screening Levels for Ground Water Protection: Inorganic Chemicals 43
-------
VI
-------
PREFACE
This document provides technical details of the derivation of the September 30,1993, draft Soil Screening
Levels (SSLs) Guidance for Superfund (U.S. EPA, 1993a). An expanded^pdated version of this
background document will be developed in support of the revised draft SSL guidance that will appear in
the Federal Register for public comment. Because the guidance is still in draft form, it and this document
should be implemented only in the context of demonstration pilots being overseen by the U.S.
Environmental Protection Agency (EPA). The methods used in the draft guidance and in this document
will undergo rigorous technical review and public comment before these documents are finalized. The
final documents will include SSLs for approximately 60 additional chemicals.
This document is presented in two sections. Section 1 defines SSLs and provides background information
on the development of SSLs and their application and implementation at Superfund sites, including
sampling schemes for measuring SSL attainment. It also provides draft SSLs developed for 30 chemicals.
Section 2 provides the technical basis for the development of SSLs addressing direct ingestion of soil,
inhalation of volatiles and fugitive dust, and the soil-to-ground-water exposure pathway, including the
assumptions and theories used in their development.
vn
-------
ACKNOWLEDGMENTS
This technical background document was prepared by Research Triangle Institute (RTI) under EPA
Contract 68-W1-0021, Work Assignment D2-24, for the Office of Emergency and Remedial Response
(OERR), U.S. Environmental Protection Agency (EPA). Janine Dinan, of OERR's Hazardous Site
Evaluation Division, and Loren Henning, of OERR's Hazardous Site Control Division, are the EPA Work
Assignment Managers for this effort and are also technical contributors to the document. Robert Truesdale
is the RTI Work Assignment Leader. RTI technical contributors include Mr. Truesdale, Steve Beaulieu,
Nancy Jones, and Anne Crook. Dr. Zubair Saleem of EPA's Office of Solid Waste conducted the
EPACMTP modeling effort and provided the discussion on the use of this model for DAF development
(Section 2.4.6).
vin
-------
Technical Background Document for
Draft Soil Screening Level Guidance
Part 1: DEFINITION AND APPLICATION
1.1 Background
On June 19, 1991, the Administrator of the U.S. Environmental Protection Agency (EPA) charged the
Office of Emergency and Remedial Response (OERR) with conducting a 30-day study to outline options
for accelerating the rate of cleanups at National Priorities List (NPL) sites. The study found that the
current investigation/remedy selection process takes over 3 years to complete because each site is treated
as a unique problem, requiring the preparation of site-specific risk assessments, cleanup levels, and
technical solutions. The study proposed that standardizing the remedial planning and remedy selection
process would significantly reduce the time it takes to start cleanups and would improve consistency in
the approach to site remediation. One of the specific proposals was for OERR to "examine the means to
develop standards or guidelines for contaminated soils."
On June 23, 1993, EPA announced the development of Soil Trigger Levels as one of the Administrative
Improvements to the Superfund program. On September 28, 1993, EPA completed a draft Fact Sheet
presenting Soil Screening Levels (SSLs) (formerly known as trigger levels) for 30 chemicals (U.S. EPA,
1993a), representing OERR's first step toward standardizing the evaluation and cleanup of contaminated
soils under the Comprehensive Environmental Response Compensation and Liability Act (CERCLA). This
background document is intended to support this Fact Sheet by providing an overview of the SSLs and
information on their technical basis.
An SSL is a chemical concentration in soil that represents a level of contamination above which there is
sufficient concern to warrant further site-specific study. Concentrations in soil above this screening level
would not automatically designate a site as "dirty," nor trigger a response action; SSLs are not national
cleanup levels or standards. Rather, they suggest the need for further evaluation of the potential risks that
may be posed by site contaminants. Generally, if contaminant concentrations in soil fall below the
screening level and the site meets specific residential use conditions, no further study or action is
warranted for that area under CERCLA (Superfund). However, some States have developed screening
values that are more stringent than the SSLs. Further study may be warranted under such State programs.
1.2 Purpose of SSLs
The primary purpose of the SSLs is to accelerate decisionmaking for contaminated soils. Initially, SSLs
will be used to focus remedial investigations by targeting site areas that warrant further study under
CERCLA. In fostering prompt identification of the contaminants and exposure areas of concern, the SSLs
may also help simplify or accelerate the baseline risk assessment and may serve as Preliminary
Remediation Goals (PRGs) under specified conditions. EPA will explore other potential applications as
this guidance is refined and expanded. Such applications may include removal response actions, site
assessment/NPL listing, voluntary cleanups, and Resource Conservation and Recovery Act (RCRA)
Corrective Actions.
-------
1.3 Summary of SSL Attributes
The SSLs presented in this document have been developed using exposure assumptions for residential land
use and considering three pathways of exposure to the contaminants (see Figure 1):
• Ingestion of soil
• Inhalation of volatiles and fugitive dusts
• Migration of contaminants through soil to an underlying potable aquifer.
These pathways have proven to be the most
common routes of human exposure to contam-
inants in the residential setting at hazardous
waste sites evaluated by EPA. Also, substan-
tial effort has been made to develop widely
acceptable methods to model these particular
pathways.
Key SSL attributes include the following:
• SSLs calculated for the ingestion
and inhalation pathways are based
on standard equations modified
from RAGS Part B (U.S. EPA,
1991).
• SSLs for migration to ground water
pathways are based on a
partitioning equation coupled with
a dilution and attenuation factor
(DAP).
Direct Ingestion
of Ground
Water and Soil
Blowing
Dust and
Volatilization
Figure 1. Pathways addressed by soil screening.
• Conservative default values were used to calculate levels protective of "high-end" individual
exposures.
• SSLs are generally based on a 10~6 risk for carcinogens or a hazard quotient of 1 for
noncarcinogens: SSLs for protection of ground water are based on nonzero maximum
contaminant level goals (MCLGs). If these are not available, maximum contaminant levels
(MCLs) are used; if MCLs are not available, the risk-based targets above are used.
« SSLs are calculated for individual exposure pathways.
Details of these attributes and their technical basis are presented in Section 2 of this document.
The models and assumptions used to develop the SSLs are representative of a "reasonable maximum
exposure" (RME) in the residential setting. RAGS Part A (U.S. EPA, 1989b) outlined the Superfund
program's approach to calculating an RME. Since that time, the EPA (U.S. EPA, 1991) has coined a new
term that corresponds to the definition of RME: "high-end individual exposure." The Superfund
program's method of estimating the high end (outlined in U.S. EPA, 1989b) is to combine an arithmetic
average value for site concentration with high-end values for intake and duration. The estimate of high-
end exposure is then compared to chemical-specific toxicity criteria approved by the Agency, such as those
found in the Integrated Risk Information System (IRIS) and Health Effects Assessment Summary Tables
(HEAST). The method used to set SSLs combines high-end default values for the intake and duration
-------
parameters with EPA toxicity criteria to backcalculate to a screening level in soil. Because Superfund
requires an estimate of the arithmetic average concentration when calculating a high-end exposure,
attainment of SSLs also should be based on an arithmetic average of site-specific samples (Section 1.8).
For the ground water pathway only, SSLs are part of a four-tiered approach to evaluating soil
contaminants that may leach to ground water. The tiers reflect increasing levels of site specificity and cost
but generally decreasing levels of conservatism. The first-tier SSLs rely heavily on concentration levels
derived from mathematical models and nationally based, generic assumptions. If contaminant levels at
a site do not exceed the first-tier SSLs and other site exposure pathways are accounted for in the
assumptions used to derive the SSLs, then the ground water pathway for the area or site is no longer of
concern under CERCLA remedial authority. If contaminant levels at a site equal or exceed the first-tier
SSLs, or other pathways of concern are present, a higher tier screening analysis may be considered or a
full site investigation may be initiated. The other three tiers are distinguished by their approach to
evaluating the soil-to-ground-water pathway. Tier 2 allows site-specific values to replace the generic
defaults in the Tier 1 partitioning equation, Tier 3 uses a leach test, and Tier 4 involves full-scale site-
specific modeling (Section 1.7).
1.4 Limitations of SSLs
SSLs do not automatically trigger the need for
response actions or define "unacceptable" levels of
contaminants in soil. In addition, the levels are
not necessarily protective of all known human " Dermal absorption
exposure pathways, reasonable land uses, or ' lndoor exP°sure to volatlles from so" and
ecological threats. Other exposure scenarios may water
be associated with significant risks to humans or
ecological receptors (Figure 2). OERR will
Ecological effects
Consumption of fish, beef, or dairy
products
continue to seek consensus on the appropriate ' Land uses other than residential
methods required to quantify additional exposure
scenarios on a national basis. The results of these
efforts may be included in the final guidance. Figure 2. Exposure scenarios not addressed
by Soil Screening Levels.
SSLs were not developed as nationwide cleanup levels or standards. They are risk-based levels that
have not been modified by the Superfund remedy selection criteria designed to tailor final cleanup levels
to site-specific conditions (NCP Section 300.430 (3)(2)(i)(A)).
However, SSLs can serve as PRGs in the following cases:
• Where site conditions mimic the model assumptions underlying the SSLs (i.e., all pathways
of concern at a given site match those accounted for in the SSLs), or
• Where the site manager or owner decides not to incur costs of additional site-specific study
to arrive at less conservative but still protective levels.
The primary condition for use of the SSLs is that exposure pathways of concern and site conditions must
match those addressed by the levels. Thus, at all sites it will be necessary to develop a simple conceptual
site model to identify likely source areas, exposure pathways, and potential receptors. This site-specific
conceptual model can then be compared with the SSL methodology to determine the extent to which the
SSLs can serve as PRGs at a site (see Section 1.5). In addition, the following questions should always
be considered before applying the SSLs:
-------
• Are the risks to ecological receptors significant at the site?
• Is there potential for land use other than residential?
• Are there other likely human exposure pathways that were not considered in developing the
SSLs (e.g., dermal contact, indoor exposure, indirect exposure through local fish consumption
or consumption of locally raised beef, dairy, or other livestock)?
• Are there site conditions that are significantly different from the ones modeled (e.g.,
unusually large area of contamination or unusually high fugitive dust levels)?
If any of these four conditions exist, then SSLs cannot be used to screen out sites or portions of sites from
further evaluation. In addition, SSLs should not be viewed independently of either natural background
concentrations or anthropogenic "background" concentrations from other sources. Where natural
background levels are higher than SSLs, generally the SSLs will be of little value since it is inappropriate
to conduct further study or action to address contaminants below background. Similarly, when
anthropogenic "background" levels exceed the SSLs, EPA does not encourage additional study or action
without first attempting to coordinate such action with the authority responsible for managing the more
broadly contaminated area. In either case, the Agency highly recommends the collection of site-specific
data to verify (1) that background concentrations are above the SSLs and (2) that onsite soils are at or
below background.
A mass balance modeling approach was not used in determining the draft SSLs. Instead, SSLs were
derived for each individual pathway, and the lowest applicable value serves as the SSL. Since the SSLs
were developed as conservative screening levels, the environmental fate parameters (e.g., soil type,
fraction organic carbon, and soil porosity) were set at high-end values that favored the pathway for which
the SSL was developed. However, OERR will consider the mass balance approach pending their analysis
of (1) the selection of a finite source for the site and (2) whether fractionating risk between pathways is
appropriate for screening levels (see Section 2.1).
The potential for exposure to multiple contaminants via the same route of exposure has not been included
in the SSL framework due to: (1) different endpoints of toxicity and mechanisms of action for SSL
chemicals (see Section 2.1) and (2) an inability to predict the number and type of contaminants that may
be present at a site. Therefore, the potential for additive effects of multiple contaminants must be
evaluated on a site-specific basis.
Finally, the acute hazard due to "hot spots" (i.e., highly contaminated soils in small, well-defined areas)
has not been included in the SSL framework since the SSLs were designed to address long-term exposure
to residual levels of contaminants. If there is reason to believe that short-term exposures may be
significant, the site manager should consider the potential for acute health effects. For example, the
presence of extremely toxic contaminants (e.g., cyanide) or highly contaminated pockets suggests that
acute exposures are of potential concern (see Section 2.1). Over the next few months, OERR will be
investigating how the spatial and temporal aspects of acute exposures at Superfund sites may impact
human health. The results of this investigation will be contrasted with the SSLs to determine if "acute
SSLs" should replace the SSLs developed in this document for certain chemicals.
1.5 How To Use SSLs
Table 1 contains draft SSLs for 30 chemicals. The first column to the right of the chemical name presents
values based on soil ingestion. The second column presents the lower of two values derived to protect
for either inhalation of volatiles or soil particulates. The third column presents the lowest number of the
first two columns and may be used as the SSL for surface soils under most residential circumstances. For
-------
Table 1. Superfund Proposed Soil Screening Levels (SSLs)a
Pathway-specific values
for surface soils (mg/kg)
Chemical
oc-BHC
Benzene
Benzo(a)pyrene
Carbon tetrachloride
Chlordane
Chlorobenzene
Chloroform
Chrysene
DDT
1 ,4-Dichlorobenzene
1,1-Dichloroethane
1,1-Dichloroethylene
Dieldrin
Ethylbenzene
Methylene chloride
Naphthalene
PCB-1260
Pentachlorophenol
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1-Trichloroethane
Trichloroethylene
Vinyl chloride
Xylenes (mixed)
Arsenic
Cadmium
Chromium(VI)
Mercury
Nickel
Ingestion
0.1 d
22 d
0.11 d
4.9 d
0.49 d
1,600f
100 d
110d
1.9d
27 d
7,800 f
1.1 d
0.04 d
7,800 f
85 d
3,100 '
V
5.3d
12d
1 6,000 f
780 f
h
58 d
0.34 d
1 60,000 f
0.37 d
39 f
390 f
23 f
1,600f
a Screening levels based on human health
b Surface soil SSLs represent
the lower of
Inhalation
1.0d
2.5 d
13.39
1.5d
0.6 g
170 g
1.1 d
0.38 9
3.9s
80 9
450 9
0.17 d
5.1 9
58 9
44 d
52 9
h
h
41 d
150 9
93s
420 8
13d
0.02 d
97 9
2,600 d
6,200 d
930 d
41 f
47,000 d
criteria only.
SurfacG
soil SSLs
(mg/kg)b
0.1 d
2.5 d
0.11 d
1.5d
0.49 d
170 9
1.1 d
0.38 9
1.9d
27 d
450 g
0.17 d
0.04 d
58 9
44 d
52 9
h
h
12d
150s
93 g
420 9
13d
0.02 d
97 9
0.37 d
39 '
390 f
23 '
1,600f
Ground water pathway
Tier 1 (mg/kg)
Unadjusted
0.0001 e
0.001 e
0.71 d
0.003 e
0.2 d
0.05
0.02
0.04
0.23
0.08 e
0.62
0.002 e
0.0001 e
0.33
0.001 e
2.5
0.82
0.001 9li
0.003 e
0.36
0.23s
0.07
0.001 8
0.0002 e
5.7
1.4'
0.81 '
1.9'
0.3'
8.2'
With 10
DAFC
0.001 d
0.01 d
7.1
0.03
2
0.5
0.2
0.4
2.3
0.8
6.2
0.02
0.001 e
3.3
0.007 e
25
8.2
0.009 ej
0.03
3.6
2.3
0.7
0.01 e
0.002 8
57
14'
8.1 '
19'
3'
82 '
levels,
With 100
DAF°
0.01 d
0.1
71
0.3
20
5
2
4
23
8
62
0.2
0.01
33
0.07
250
82
0.09 '
0.3
36
23
7
0.1
0.02
570
140'
81 '
190'
30'
820'
ingestion and inhalation values.
0 DAF = Dilution and attenuation factor.
d P.al^nlatoH woliioc onrracnnnH tn a ranr.ar rick louot nf 1 in 1 OHO OOO
e Level is at or below Contract Laboratory Program required quantitation limit for Regular Analytical Services (RAS).
' Calculated values correspond to a noncancer hazard quotient of 1.
9 Soil saturation concentration (Csat).
h No toxicity criteria available for that route of exposure.
' A preliminary remediation goal of 1 ppm has been set for PCBs based on U.S. EPA (1990) and on Agency-wide efforts
to manage PCB contamination
1 SSLs for pH of 6 8.
-------
sites where migration to ground water is a pathway of concern, SSL values for the ground water pathway
apply.
Three different SSLs address migration of contaminants to ground water; the selection of an appropriate
SSL for this pathway depends on several site-specific conditions. The first column of ground water values
reflects the levels calculated using the partitioning equation with no correction factor added for dilution
and attenuation in the subsurface (unadjusted). The next two columns reflect the levels adjusted by factors
of 10 and 100, respectively (10 and 100 DAF), to account for such dilution and attenuation. Selection
of an appropriate DAF is discussed in Section 1.6.
Metal partitioning in the soil/water system is significantly affected by a variet} of soil conditions, the most
significant of which is pH. For this reason, SSLs for metals were developed for three pH conditions: 4.9,
6.8, and 8.0, representative of national ground water conditions (see Section 2.4.4). Table 1 contains SSLs
for inorganics corresponding to a pH of 6.8. Table 2 presents inorganic SSLs corresponding to pH values
of 4.9 and 8.0. (These preliminary numbers for metals are currently under review and revision and will
be updated in the final SSL guidance.) If pH conditions at a site are not known, the SSL corresponding
to a pH of 6.8 should be used. Table 2 also includes SSLs for pentachlorophenol (PCP), whose
partitioning behavior is also highly pH dependent, as described in Section 2.4.3.
As mentioned in Section 1.4, the first step in applying the SSL guidance is to develop a simple conceptual
model of the site based on available site sampling data, historical records, aerial photographs, and site
hydrogeologic information. This model will establish a site-specific hypothesis about possible contaminant
sources, contaminant fate and transport, potential exposure pathways, and human or environmental
receptors. If the conceptual model indicates that potential exposure pathways and receptors are fully
accounted for in the SSL methodology, the SSLs may be directly applied to the site. However, if the
model indicates that the site is either very large or complex or that there are exposure pathways not
accounted for in the SSL methodology, SSLs will not be suitable for full evaluation of a site. They can
still be used, however, in the site evaluation since SSLs have been derived on a pathway-specific basis.
Thus, it will only be necessary to evaluate those exposure pathways that are not already considered in the
SSL methodology.
The second step in applying SSLs is to collect a representative sample set for each exposure area. (See
Section 1.8 for more detailed guidance on number of samples and locations.) An exposure area is defined
as that geographic area in which an individual may be exposed to contamination regularly. It may involve
the entire site, portions of a site, or a residential lot. The site manager should work to limit total trips to
the site and minimize the number of samples collected and their locations. To maximize efficiency, data
collection should be coordinated with other early sampling efforts to gain a better understanding of basic
site hydrogeology, ecological threats, or the potential for application of various treatment technologies.
For example, the decision may be made early on to collect data for site-specific modeling purposes at a
particular site. The conceptual site model should be compared with site data as they are collected and
revised or updated as appropriate.
The third step is to compare site-specific data with the SSLs in Table 1. Generally, this comparison will
result in one of three outcomes:
1. Site-measured values indicate that an area falls well below any SSL in the table. These areas
of the site can be eliminated from further evaluation.
2. Site-measured data indicate that one or more SSLs have clearly been exceeded by a wide
margin. In this case, the SSLs have helped to identify contaminants and exposure pathways
of concern on which to focus further analysis or data-gathering efforts.
-------
Table 2. Proposed Ground Water Pathway SSLs for Inorganics and
Pentachlorophenol, as a Function of pHa
Proposed
Unadjusted
Chemical pH
Arsenic
Cadmium
Chromium(VI)
Mercury
Nickel
Pentachlorophenol
4.9
1.2
0.006
3.1
0.0002
0.32
0.017
8.0
1.6
10.0
1.4
0.42
15.7
0.0009b
ground water pathway SSLs (mg/kg)
With 10 DAF
4.9
12.5
0.08
31.4
0.002
3.2
0.17
8.0
15.7
100
13.6
4.2
157
0.009b
With 100 DAF
4.9
125
0.81
314
0.02
31.7
1.7
8.0
157
1,001
136
42.2
1,573
0.09
SSLs = Soil screening levels.
DAF > Dilution and attenuation factor.
a Screening levels based on human health criteria only.
b Level at or below Contract Laboratory Program required quantitation limit for Regular Analytical Services
(RAS).
3. A site-measured value exceeds one pathway-specific value but not the others. In this case, it
is reasonable to collect additional data to better characterize risk from that pathway at that site.
When an exceedence is marginally significant, a closer look at site-specific conditions and
exposures may result in the area being eliminated from further study. If there is a marginal
exceedence for the ground water pathway, a manager may choose to collect data specified in
the next higher tier(s) (Section 1.7).
For an NPL site at which SSLs are exceeded, a quick analysis can determine whether the cumulative risks
posed by the site exceed the 10"4 risk level for carcinogens (or hazard index of 1 for noncarcinogens),
which generally are the triggers for remedial action under Superfund. Where the basis for response action
exists and exposure pathways of concern are addressed by the SSLs, the SSLs become PRGs, as defined
in RAGS Part B (U.S. EPA, 1991).
In accordance with the National Contingency Plan (NCP), the decisionmaker will need to consider a
variety of factors in determining whether any modification of the SSLs (used as PRGs) is appropriate in
setting final cleanup levels (NCP Section 300.430(e)(2)(i)(A)). Ultimately, final cleanup levels are set
based on an evaluation of the NCP's nine criteria, including cost, long-term effectiveness, and
implementability. If ground water is the driving pathway, even at this final stage, other SSL tiers can be
used to identify final cleanup levels, as described in Section 1.7.
1.6 Determining the Dilution/Attenuation Factor
(Ground Water Pathway)
A simple linear soil/water partitioning equation is used to calculate the unadjusted SSLs for the ground
water pathway (see Section 2.4). This equation relates contaminant concentrations in soil adsorbed to soil
organic carbon (or in solid form) to soil leachate (pore water) contaminant concentrations in the
unsaturated zone. Migration of soil leachate through the unsaturated zone to the water table generally
reduces contaminant concentrations by attenuation processes such as adsorption and degradation. Ground
water transport in the saturated zone further reduces concentrations through attenuation and dilution.
Generally, to account for those mechanisms in the subsurface environment, a correction factor, or
-------
Table 3. Effect of Facility Area on Dilution/Attenuation Factor (DAF)
DAF: 10 100
Protectlveness Level: 90% 95% 90% 95%
Site Size (acres) <10 <0.5 <1 <0.1
dilution/attenuation factor, can be applied to the unadjusted SSL. Using EPA's Composite Model for
leachate migration with Transformation Products (EPACMTP) (U.S. EPA, 1993c), the Agency has
identified a DAF of 10 as an appropriate correction factor to be applied to the partitioning value in most
cases. Background information on the modeling approach used for SSL DAF development is provided
in Section 2.4.6 of this document.
There are circumstances in which a higher DAF may be appropriate. Based on a number of sensitivity
analyses conducted on EPACMTP (see Section 2.4.6), the Agency has determined that the size of the
contaminated area is the most appropriate parameter on which to base adjustments to the SSL DAF. From
these sensitivity analyses, EPA has developed DAFs corresponding to varying site sizes and levels of
protectiveness (Table 3). The largest allowable areas corresponding to DAFs of 10 and 100 at the 90th
percentile protection level are approximately 10 and 1 acre, respectively. Therefore, for sites of up to 10
acres, a DAF of 10 should be applied to the unadjusted SSLs, while for sites at or below 1 acre, a DAF
of 100 should be applied to the unadjusted SSLs. If a 95th percentile protectiveness level is used, a DAF
of 10 is protective for areas under 0.5 acre and a DAF of 100 is protective for areas less than 0.1 acre.
OERR is considering whether the 90th or 95th percentile protectiveness level should be used in the final
guidance and also is investigating appropriate DAFs for sites larger than 10 acres.
However, there are specific circumstances under which use of a DAF is not recommended, such as in
areas of very shallow ground water or karst topography. When sites are located in areas with water tables,
within 5 feet of the surface, the unadjusted SSLs should be used. In this scenario, contamination is
located in or directly above the saturated zone; therefore, any attenuation processes within the unsaturated
zone would be negligible.
1.7 Application of Tiers 2 Through 4 (Ground Water Pathway)
The assumptions factored into the Tier 1 levels are conservative, rendering the SSLs fairly stringent (see
Section 2.4). Although the generic assumptions used in Tier 1 are not considered overly conservative for
national application, EPA recognizes that site-specific conditions may differ significantly from the generic
assumptions used in developing the Tier 1 SSLs. Therefore, for the ground water pathway, subsequent
tiers of the SSLs allow for the substitution of some of the generic fate and transport assumptions with site-
specific data to derive alternative "screening levels" that are more site-specific (Figure 3). The tiered
framework for migration to ground water represents a sliding scale of increasing site-specificity and
decreasing conservatism. The framework allows the user the flexibility to move away from the
conservative Tier 1 level by incorporating increasing levels of site empirical data.
If site concentrations do not exceed the Tier 1 SSLs multiplied by the appropriate DAF, then the pathway
is excluded from further investigation. However, if site concentrations do exceed the Tier 1 SSLs, they
may be used as PRGs (when appropriate), or a Tier 2, 3, or 4 investigation may be conducted. Site
managers or owners of small, relatively uncomplicated sites may benefit from the Tier 1 levels by
bypassing the additional costs associated with collecting additional data to conduct further investigations.
-------
Increasing
conservatism
Tier 1 Screening Levels
• Partitioning equation
• DAF of 1,10,100
Tier 2 Screening Levels
• Site-specific partitioning equation
• DAF of 1,10,100
Tier 3 Evaluation
• SPLP, DAF of 1,10,100
Tier 4 Evaluation
• Use of fate and transport model
in site-specific application
Increasing
site-specificity
DAF = Dilution/attenuation factor.
SPLP = Synthetic Precipitation Leaching Procedure.
Figure 3. Tiered approach, ground water pathway.
However, it can be in the interest of site managers or owners of large and complex sites to conduct a more
site-specific investigation to develop remediation goals that are tailored to site-specific conditions. Each
tier requires more site-specific information but may lead to a less stringent "screening" concentration.
Note, however, that one purpose of the SSLs is to define a level in soil below which no further study or
action would be required. Alternative levels using site-specific data, although less conservative, must still
be protective of "high-end" individual exposures. Therefore, data inputs and assumptions used for Tiers
2 through 4 should be selected to represent high-end or conservative values for the particular site in
question.
The Tier 2 levels represent a minimal increase in site-specificity and perhaps less conservative SSLs. The
partitioning equation used in the Tier 1 calculation (Equation 2-22, Section 2.4) remains the basis for the
Tier 2 levels along with the same DAP (either 1, 10, or 100). However, site-measured values of organic
carbon, soil porosity, fraction water content, and soil bulk density are substituted into the equation to
calculate SSLs more tailored to site characteristics. If site concentrations do not exceed the Tier 2 SSLs
(which are tailored more to site conditions than Tier 1), then the ground water pathway can be excluded
from further investigation or concern. If site concentrations exceed the Tier 2 SSLs, the user has the
option of conducting a Tier 3 or 4 investigation, recognizing that the increase in site-specificity is
associated with additional costs for collecting site-specific data.
The Tier 3 investigation involves conducting a specific leach test, the Synthetic Precipitation Leaching
Procedure (SPLP) (Appendix A; U.S. EPA, 1992f). If the leach test results divided by the DAF of 10
exceed the acceptable ground water limit (e.g., nonzero MCLG, MCL, 10"6 risk-based values), then further
investigation is warranted. The user is advised to use discretion when applying the SPLP because it may
not be applicable to all contaminated soils (e.g., oily types of waste do not yield suitable results).
Tier 4 represents the highest level of site-specificity in evaluating the migration to ground water pathway.
In this investigation, site-specific data are collected and used in a fate and transport model to confirm the
threat to ground water and further determine site-specific cleanup goals as in a remedial
investigation/feasibility study (RI/FS). A DAF is not used in this tier because the model would account
-------
for fate and transport mechanisms in the subsurface. The advantage of this approach is that it accounts
for site hydrogeologic, climatologic, and contaminant source characteristics and may result in fully
protective but less stringent remediation goals. As in Tiers 2 and 3, the additional cost of collecting the
data required to apply the model should be factored into the decision to conduct a Tier 4 investigation.
Additional guidance on applying Tiers 2 through 4 is provided in Section 2.4.8.
1.8 Measuring Soil Levels
As described in the supplemental guidance to RAGS (U.S. EPA, 1992e), exposure to site contaminants
over a long (chronic) period of time is best represented by an arithmetic average concentration. Because
an individual is assumed to move randomly across an exposure area over time, the spatially averaged soil
concentration can be used to estimate the true average concentration contacted over time. In some cases,
one area of a site may be more attractive to an individual than another area of the site, leading to
nonrandom exposure. In these cases, the risk assessor should consider weighting time spent on different
areas. However, for most current use and all future use scenarios, this type of information will not be
available. Therefore, the assumption of random exposure appears to be the most reasonable alternative.
Given the assumption of random movement across an exposure area, the Agency has decided that
attainment of SSLs should be based on an arithmetic average as well. The issue then becomes the number
of samples required to adequately estimate the mean and the area over which the samples should be
averaged.
1.8.1 Exposure/Averaging Area. In support of RAGS Part A, EPA's Exposure Assessment Group
(BAG) conducted a number of simulation studies using large sets of site-specific data. The goals of the
studies were to examine the use of data sets of varying size to: (1) estimate the true site mean and (2)
compare the upper confidence limit on the arithmetic mean to the true site mean. The studies indicated
that the means from data sets with 10, 20, and even 30 samples often underestimated the true site average
concentration. These results formed the basis for the policy of using the 95 percent upper confidence limit
(UCL95) on the arithmetic mean or, where necessary, the highest measured value of site sample data as
the exposure point concentration in Superfund exposure/risk assessments.
A potential drawback of the EAG study was that the subsets of sample data (either 10, 20, or 30 samples)
were randomly selected and included data from all areas of the site. The random selection of samples
from widely different areas contributed to the variance within each subset of data. For measuring
attainment with the SSLs, OERR wants to test whether the need for analyzing a large number of samples,
or reliance on the UCLg5, could be overcome by limiting the size of the averaging area.
The exposure/averaging area is the area over which an individual can reasonably be expected to move
over a period of time and be exposed to contaminants through certain pathways. An appropriate
exposure/averaging area can vary in size, depending on site-specific conditions. At some sites, this may
be the entire site; at others, the exposure area may be only a portion of the site. For the purposes of this
SSL guidance, the Agency believes that the size of a typical residential lot (0.25 acre) is an appropriate
averaging area. Thus, 0.25 acre should be used as the default for the exposure area, if no other reasonable
exposure areas can be determined.
For large sites that could be divided into many 0.25-acre plots, the number of samples needed to
adequately characterize the site for the purpose of screening out areas from further consideration under
CERCLA becomes quite high. This, coupled with the costs of analytical services for each sample, could
make the sampling costs onerous. In an effort to balance the analytical costs with achieving statistical
10
-------
confidence in determining the mean concentration of contaminants in each exposure area, OERR
recommends the guidance presented in the following subsections for measuring contaminant levels in
surface soils at NPL sites for the purpose of using SSLs.
1.8.2 Sample Pattern. A grid pattern such as a triangular or square/rectangular grid is recommended
to establish sample locations for each exposure area (U.S. EPA, 1987). If there are areas of suspected
contamination or stained soils, then these areas must be sampled and evaluated (i.e., compared to SSLs)
separately from those areas of systematic sampling. This alleviates the potential to miss areas of elevated
contamination during the systematic sampling exercise.
1.8.3 Number of Samples. As mentioned, it is necessary to balance the need to achieve statistical
confidence in determining a meaningful arithmetic mean concentration of contaminants in each exposure
area with the cost of obtaining and analyzing the 20 to 30 samples recommended by the Office of
Research and Development (ORD). For nonvolatile chemicals, compositing of discrete samples is one
option since EPA is interested in determining the arithmetic mean of the contaminant concentration(s).
Twenty discrete samples can be composited down to four or five composite samples, while maintaining
confidence that the area average is not grossly underestimated. For instance, the discrete samples may
be combined into composite samples by: (1) collecting the 20 discrete samples from a systematic grid
across the entire exposure area, compositing all of the samples, and then drawing four subsamples from
the homogenized material; or (2) dividing the exposure area into quadrants and, within each quadrant,
compositing the five discrete samples. The five discrete samples can be collected randomly or by using
a systematic pattern of one in each comer and one in the center of the quadrant. Both approaches are
under evaluation.
Compositing may mask contaminant levels that are slightly higher than the SSL, but areas of high
contamination will still be detected. Compositing is both a reasonable approach and an efficient use of
resources, since, for most compounds, Superfund is interested in average exposure over time. However,
none of the composite samples should exceed the prescribed SSL for any contaminant.
For volatile organic compounds (VOCs), compositing is not appropriate because compounds may volatilize
during homogenization of the samples and contaminant concentrations will be underestimated (U.S. EPA,
1989a, 1992d). OERR advocates that 10 discrete samples be taken per exposure area for VOCs. The
results of each sample should be compared to the SSL(s), and no contaminant level can exceed the
screening level(s). OERR feels that it is not appropriate to average the results of the 10 discrete samples;
as EAG's simulation studies showed that the average of subsets with 10 samples frequently underestimated
the true mean. However, OERR will test this assumption in pilot projects to ascertain if the assumption
holds true for samples collected within a finite area.
1.8.4 Depth. When measuring soil levels at the surface for the inhalation and ingestion pathways,
samples should be taken at a depth of 6 inches. Additional sampling beyond 6 inches may be appropriate
in areas where soil disturbances are reasonably expected as a result of construction practices. For
example, in the Northeast, soil may be excavated to 15 feet before the foundation of a home can be laid,
and this soil usually is used as fill material around the property. As a result, contaminants that were at
depth can be near the surface. Thus, it is important to be cognizant of construction practices in the area
around a site when determining the depth of sampling.
For the ground water pathway, the entire soil column, from the surface to the top of the aquifer, should
be sampled to gather information on contaminant distribution with depth. For the evaluation of vertical
stratification, samples should not be averaged over depth (i.e., the soil core should not be composited over
depth), but rather individual samples should be evaluated at appropriate depth intervals. One soil core per
exposure area may be sufficient. This is not meant to imply that one soil core per exposure area is
11
-------
adequate to characterize the threat to ground water at the site. For instance, where dense nonaqueous
phase liquids (DNAPLs) are suspected, many cores must be taken to determine the presence and/or extent
of contamination.
1.8.5 Sampling for Background Contamination. For metals, background sampling is necessary
to be certain that EPA is not defining levels below natural or anthropogenic background as of regulatory
concern. If a statistical comparison of background concentration and site samples indicates that
background metals concentrations are significantly above the SSLs, use of the SSLs will be of limited
value, as discussed in Section 1.4.
1.8.6 Additional Sampling Needed for Ground Water Tier 2. To implement ground water Tier
2, site-specific soil characteristics must be determined by sampling. Parameters to measure include bulk
density, porosity, organic carbon content, and water content (see Section 2.4.8). Samples should represent
the entire soil column and should be taken along with samples to measure contaminant levels.
1.8.7 Geostatistics. For large areas where the data are not widely scattered, geostatistical approaches,
such as kriging, can be used to estimate sample concentration trends across the exposure area (U.S. EPA,
1989a).
1.8.8 Sample Analysis. Both the discrete VOC samples and the composites must be analyzed by
Contract Laboratory Program (CLP) (or equivalent) methods. (NOTE: Seven of the 30 contaminant SSLs
for the ground water migration pathway at a DAF of 10 are below CLP Regular Analytical Services (RAS)
or CLP-equivalent detection limits. For these contaminants, special analytical services should be requested
for recalibration of the instruments. For example, to measure low levels of VOCs, the gas
chromatograph/mass spectrometer (GC/MS) can be recalibrated to detect at 1, 2, 5, 10, and 25 ppb.)
Where available and appropriate, field methods (soil gas surveys, immunoassays, X-ray fluorescence) can
be used. Again, for compounds other than VOCs, compositing samples is acceptable as long as it is
consistent with the field methodology. If any sample concentration exceeds an SSL, further site study is
required. In addition, 10 to 20 percent of field samples must be sent to a CLP (or equivalent) laboratory
for confirmatory analysis (U.S. EPA, 1992d). Please note that field methods must be capable of achieving
appropriate detection limits for most ground water SSLs.
12
-------
Technical Background Document for
Draft Soil Screening Level Guidance
Part 2: TECHNICAL BACKGROUND
Analysis of Records of Decision (ROD) from Superfund sites indicates that the three most common
pathways of human exposure to soil contaminants in the residential setting are direct ingestion of soil,
inhalation of volatiles and fugitive dusts, and migration of contaminants through soil to an underlying
potable aquifer (soil-to-ground-water pathway). The 30 proposed Soil Screening Levels (SSLs) presented
in this document have been developed considering these three pathways using "reasonable maximum
exposure" (RME) assumptions (see Section 1.3) for a residential land-use scenario. The methology used
to develop these SSLs, including the human health basis and models, assumptions, and input parameters,
is described for each pathway in the following sections.
2.1 Human Health Basis
For soil ingestion and inhalation of volatiles and fugitive dusts, the SSLs correspond to a 10"6 risk level
for carcinogens and a hazard quotient (HQ) of 1 for noncarcinogens. For carcinogens, EPA believes that
setting a 10~6 risk level for individual chemicals and pathways will generally ensure that the cumulative
risks are within the 10~4 to 10~6 risk range for all chemical/pathway combinations typically found at
Superfund sites.
For noncarcinogens, the issue of cumulative risk is much more complex, because risk is evaluated based
on the theory that a threshold exists for noncancer effects. The threshold level, below which adverse
effects are not expected to occur, is the basis for the Agency's Reference Dose (RfD) and Reference
Concentration (RfC). Since adverse effects are not expected to occur at the RfD or RfC and the SSLs
were derived by setting the potential exposure dose equal to the RfD or RfC (i.e., an HQ equal to 1), it
is difficult to address the risk of exposure to multiple chemicals at levels where the individual chemicals
are not expected to cause any harmful effect.
The U.S. Environmental Protection Agency (EPA) believes, and the Science Advisory Board agrees (U.S.
EPA, 1993b), that HQs should be added only for those chemicals with the same toxic endpoint and/or
mechanism of action. The chemicals that have SSLs based on noncarcinogenic effects have RfDs/RfCs
that are based on different endpoints of toxicity. It will be necessary to divide the SSL values listed in
Table 1 to account for additivity, where cocontaminants at a site have RfDs/RfCs based on the same
endpoint of toxicity. Therefore, the potential for additive effects must be evaluated at every site.
EPA's Office of Emergency Remedial Response (OERR) is evaluating the current SSL approach for
noncarcinogens in light of two related issues: apportionment and fractionation. Apportionment is typically
used as the percentage of a regulatory health-based level that is allocated to the source/pathway being
regulated (e.g., 20 percent of the RfD for the contaminated ground water pathway). Apportioning risk
assumes that the applied dose from the source, in this case contaminated soils, is only one portion of the
total applied dose received by the receptor. Traditionally, OERR has focused on quantifying exposures
to a receptor that are clearly site-related and has not included exposures from other sources such as
commercially available household products or workplace exposures. Depending on the assumptions
regarding other source contributions, apportionment may result in more conservative regulatory levels
(i.e., levels that are below an HQ of 1). In contrast to apportionment, fractionation of risk may lead to
less conservative regulatory levels since it assumes that some fraction of the contaminant does not reach
the receptor due to partitioning into another medium.
13
-------
For example, if only one-fifth of the source is assumed to be available to the ground water pathway, and
the remaining four-fifths is assumed to be released to air or remain in the soil, an SSL for the ground
water pathway would be set at five times the HQ of 1 due to the decrease in exposure (since only one-fifth
of the possible contaminant is available to the pathway). However, the SSLs are, by definition,
conservative screening, or "walk-away," levels for remedial action, and available data do not support
chemical-specific assumptions necessary for partitioning at a national level. Therefore, soil contaminants
were not fractionated and the lowest chemical-specific SSL among each of the pathways was selected as
the SSL for that particular chemical (i.e., the SSL was selected from the pathway that drives the risk).*
Finally, the hazard due to acute exposures to contaminants has not been included in this SSL framework,
although the draft SSL guidance instructs site managers to consider the potential for acute exposures.
There are two major impediments to establishing national acute SSLs given the available data on
contaminants and potential scenarios for short-term exposures. First, although data are available on
chronic exposures (i.e., RfDs, RfCs, cancer slope factors), there is a paucity of data relating the potential
for acute effects to the concentrations that are considered protective for chronic exposures. Specifically,
there is no scale to evaluate the severity of acute effects (e.g., eye irritation vs. dermatitis), no consensus
on how to incorporate the body's recovery mechanisms following acute exposures, and no toxicity
benchmarks to apply for short-term exposures (e.g., a 7-day RfD for a critical endpoint). Second, the
inclusion of acute SSLs would require the development of acute exposure scenarios that would be
acceptable and applicable on a national basis. Exposure assumptions would be required for short-term
intake and duration (e.g., pica behavior, maternal exposures) in a variety of settings and climates (e.g.,
urban residential, rural residential, or industrial). Simply put, the methodology and data necessary to
address acute exposures in a manner analogous to chronic exposures have not been developed. Over the
next few months, OERR will be investigating the potential for acute effects at the chronic SSLs.
For the soil-to-ground-water exposure pathway, SSLs have been developed to reflect levels below which
concentration limits in ground water will not be exceeded (Section 2.4.7). The methodology uses nonzero
drinking water maximum contaminant level goals (MCLGs) as the target ground water concentration limits
for each contaminant. If nonzero MCLGs are not available, maximum contaminant levels (MCLs) are
used, and if MCLs are not available, risk-specific concentrations are derived using Agency toxicity criteria,
a target cancer risk of 10"6, and/or a noncancer hazard quotient of 1. These levels were calculated as
specified in U.S. EPA (1991), except that the inhalation exposure pathway was not included in the
calculations.
Table 4 lists the target ground water concentration values (Cw) used for each SSL chemical, along with
drinking water standards (MCLs), 10"6 and 10~4 excess cancer risk (ECR) levels, oral cancer slope factors,
oral, noncancer reference doses (RfD), and Contract Laboratory Program (CLP) quantitation limits (CRQL)
for the subject chemicals. The human health benchmarks (i.e., slope factors, RfDs) were obtained from
the Integrated Risk Information System (IRIS). For 25 of the 30 chemicals addressed in the September
30 draft guidance (U.S. EPA, 1993a), the nonzero MCLGs are the same as the MCLs. Risk-specific target
ground water concentrations were calculated for the remaining five chemicals.
*For seven noncarcinogenic volatiles (chlorobenzene, 1,1-dichoroethane, ethylbenzene, toluene, 1,2,4-
trichlorobenzene, 1,1,1-trichloroethane, and xylenes), SSLs were based on the saturation limit (C^,) (see Section
2.3.2).
14
-------
Table 4. SSL Target GW Level, Drinking Water Regulations, Health Advisories,
Soil Quantltation Limits
Chemical
a-BHC
Benzene
Benzo(a)pyrene
(3-BHC
Carbon tetrachloride
Chlordane
Chlorobenzene
Chloroform
Chrysene
DDT
1 ,4-Dichlorobenzene
1,1-Dichloroethane
1 ,1-Dichloroethylene
Dieldrin
Ethylbenzene
y-BHC (lindane)
Methylene chloride
Naphthalene
PCBs
Pentachlorophenol
Tetrachloroethylene
Toluene
1 ,2,4 -Trichlorobenzene
1,1,1-Trichloroethane
Trichloroethylene
Vinyl chloride
Xylenes
Arsenic
Cadmium
Chromium (VI)
Mercury
Nickel
SSL
Target
GW
Level
(mg/L)
1.35E-05
0.005
0.0002
4.73E-05
0.005
0.002
0.1
0.1
0.0002
0.0003
0.075
3.7
0.007
5.32E-06
0.7
0.0002
0.005
1.5
0.0005
0.001
0.005
1
0.07
0.2
0.005
0.002
10
0.05
0.005
0.1
0.002
0.1
MCL
(RID
PRG)
(mg/L)
0.005
0.0002
0.005
0.002
0.1
0.1
0.0002 P
(0.02)
0.075
(3.7) L
0.007
(0.002)
0.7
0.0002
0.005
(1.5)
0.0005
0.001
0.005
1
0.07
0.2
0.005
0.002
10
0.05
0.005
0.1
0.002
0.1
mg/L at
10E-4
Cancer
Risk
0.001
0.3
0.001
0.005
0.07
0.007
1.4
0.001
0.03
0.4
0.01
0.0005
0.007
1.1
0.001
0.07
0.16
0.77
0.004
0.003
Oral
mg/L at Cancer
10E-6 Slope Oral
Cancer Factor RfD
Risk (mg/kg/d) (mg/kg/d)
1 .4E-05
0.003
1 .5E-05
4.7E-05
0.0007
6.6E-05
0.01
1.5E-05
0.0003
0.004
0.0001
5.3E-06
6.6E-05
0.01
1.1E-05
0.0007
0.002
0.008
4.5E-05
6.3
0.029
7.3
1.8
0.13
1.3
0.0061
7.3E-03
0.34
0.024
0.6
16
1.3
0.0075
7.7
0.12
0.052
0.011
1.9
1.8
0.0007
6.0E-05
0.02
0.01
0.0005
0.1
0.009
5.0E-05
0.1
0.0003
0.06
0.04
0.03
0.01
0.2
0.01
2
0.0003
0.0005
0.005
0.0003
0.02
Soil
CRQL
(mg/kg)
0.008
0.005
0.33
0.008
0.005
0.080
0.005
0.005
0.33
0.016
0.33
0.005
0.005
0.016
0.005
0.008
0.005
0.33
0.033
0.016
0.005
0.005
0.33
0.005
0.005
0.005
0.005
0.01 (CRDL)
0.005 (CRDL)
0.01 (CRDL)
0.0002
(CRDL)
0.04 (CRDL)
CRDL, CRQL = CLP quantitation limits.
GW = Ground water.
L = Listed for regulation.
MCL = Maximum contaminant level.
Sources:
IRIS, 1993.
U.S. EPA, 1992b, d, and e.
PRG = Preliminary remediation goals.
P = Proposed.
RfD = Reference dose.
15
-------
2.2 Direct Ingestion
Calculation of SSLs for direct ingestion of soil is based on the methodology presented for residential land
use in RAGS Part B (U.S. EPA, 1991). Briefly, this methodology backcalculates a soil concentration level
from a target risk (for carcinogens) or hazard quotient (for noncarcinogens). A number of studies have
shown that inadvertent ingestion of soil is common among children 6 and younger (Calabrese et al., 1989;
Davis et al., 1990; Van Wijnen et al., 1990). Therefore, the approach uses an age-adjusted soil ingestion
factor that takes into account the difference in daily soil ingestion rates, body weights, and exposure
duration for children from 1 to 6 years old and others from 7 to 31 years old. The higher intake rate of
soil by children and their lower body weight lead to a lower, or more conservative, risk-based
concentration compared to an adult-only assumption. RAGS Part B uses this age-adjusted approach for
both noncarcinogens and carcinogens.
For noncarcinogens, the definition of an RfD (Section 2.1) has led to debates concerning the comparison
of less-than-lifetime estimates of exposure to the RfD. Specifically, it is often asked whether the
comparison of a 6-year exposure, estimated for children via soil ingestion, to the chronic RfD is
unnecessarily conservative.
The RAGS Part B guidance (U.S. EPA, 1991) links the use of chronic toxicity criteria with a 30-year
exposure, whereas in the proposed Hazardous Waste Identification Rule (57 FR 21450), EPA's Office of
Solid Waste (OSW) set the Concentration Based Exemption Criteria for noncarcinogens by comparing the
6-year childhood exposure with chronic toxicity criteria. EPA's OERR has asserted that the chronic RfD
is protective of sensitive individuals such as children and that combining a more conservative, shorter-term
exposure scenario with chronic toxicity criteria is overly protective. OSW has held the view that the
chronic RfD should not be exceeded during childhood when the intake of potentially contaminated soil
could be much higher than for adults (or for the time-weighted average of intake rates used in RAGS
PartB).
In their analysis of the issue, the Science Advisory Board (SAB) indicates that, for most chemicals, the
approach of combining the higher 6-year exposure for children with chronic toxicity criteria is overly
protective (U.S. EPA, 1993b). However, they noted that there are instances when the chronic RfD may
be based on endpoints of toxicity that are specific to children (e.g., fluoride and nitrates) or when the dose-
response curve is steep (i.e., the difference between the no observed adverse effects level [NOAEL] and
an adverse effects level is small). Thus, depending on the contaminant, exceeding the RfD (i.e., the
"acceptable" daily level) over a short period of time may be cause for concern. For example, if there is
reason to believe that exposure to soil may be higher at a particular stage of an individual's lifetime, one
would want to protect for that shorter period of high exposure.
Although RAGS Part B calculates soil concentrations based on child exposure for 6 years and adult
exposure for 24 years (for a total exposure duration of 30 years), OERR set the SSLs for noncarcinogens
at concentrations that are protective of the increased exposure during childhood. In essence, this method
ensures that the chronic reference dose is not exceeded during this shorter (6-year) time period (Equation
2-1).
Equation 2-1: Screening Level Equation for Ingestion of Noncarclnogenlc Contaminants In
Residential Soil
c . T . , . , THQ x BW x AT x 365 d/yr
Screening Level (mg/kg) = ^ JL
l/RfD0 x 10"* kg/mg x EF x ED x IR
16
-------
Parameter/Definition (units)
Default
Source
THQ/target hazard quotient (unitless)
BW/body weight (kg)
AT/averaging time (yr)
RfD0 /oral reference dose (mg/kg-d)
EF/exposure frequency (d/yr)
ED/exposure duration (yr)
IR/soil ingestion rate (mg/d)
1
15
6a
Chemical-specific
350
6
200
RAGS Part B (U.S. EPA, 1991)
RAGS Part B (U.S. EPA, 1991)
RAGS Part B (U.S. EPA, 1991)
RAGS Part B (U.S. EPA, 1991)
RAGS Part B (U.S. EPA, 1991)
RAGS Part B (U.S. EPA, 1991)
1 For noncarcinogens, Averaging Time is equal to Exposure
Duration.
For carcinogens, both the magnitude and duration of exposure are important. Duration is critical because
the toxicity criteria are based on "lifetime average daily dose." Therefore, the total dose received, whether
it be over 5 years or 50 years, is averaged over a lifetime of 70 years. To be protective of exposures to
carcinogens in the residential setting, RAGS Part B (U.S. EPA, 1991) and OERR focus on exposures to
individuals who may live in the same residence for a "high-end" period of time (i.e., 30 years). As
mentioned above, exposure to soil is higher during childhood and decreases with age. Thus, Equation 2-2
uses the RAGS Part B time-weighted average soil ingestion rate for children and adults; the derivation of
this factor is shown in Equation 2-3.
Equation 2-2: Screening Level Equation for Ingestion of Carcinogenic Contaminants In
Residential Soil
Screening Level (mg/kg) =
TR x AT x 365 d/yr
SF0 x 10"6 kg/mg x EF x IFS .
soil/adj
Parameter/Definition (units)
TR/target cancer risk (unitless)
AT/averaging time (yr)
SF0/oral slope factor (mg/kg-d)"1
EF/exposure frequency (d/yr)
IFsoil/adj /age-adjusted soil ingestion
factor (mg-yr/kg-d)
Default
10'6
70
Chemical-specific
350
114
Source: RAGS Part B (U.S. EPA, 1991).
Equation 2-3: Equation for Age-Adjusted Soil Ingestion Factor, IFsoll/adj
IRsoiI/agel-6 x EDagel-6
IR™,
•soil/age?-31
x ED
age7-31
(mg-yr/kg-d)
BW.
agel-6
BW.
age?-31
17
-------
Parameter/Definition (units)
/age-adjusted soil ingestion factor (mg-yr/kg-d)
/ingestion rate of soil age 1-6 (mg/day)
ED t_e/exposure duration during ages 1-6 (yr)
^soii/agev-si /ingestion rate of soil age 7-31 (mg/day)
ED 7.31 /exposure duration during ages 7-31 (yr)
BWagel_6 /average body weight from ages 1-6 (kg)
BW 7.31 /average body weight from ages 7-31 (kg)
Default
114
200
6
100
24
15
70
Source: RAGS Part B (U.S. EPA, 1991).
Table 5 lists the chemical-specific parameters used to calculate direct ingestion SSLs using
Equations 2-1 through 2-3.
2.3 Inhalation of Volatiles and Fugitive Dusts
Agency toxicity criteria indicate that risks from exposure to some chemicals via inhalation far outweigh
the risks via ingestion; therefore, the SSLs have been designed to address this pathway as well. The
models and assumptions used to calculate SSLs for inhalation of volatiles are updates of risk assessment
methods presented in RAGS Part B (U.S. EPA, 1991). RAGS Part B evaluated the contribution to risk
from the inhalation and ingestion pathways simultaneously. Because toxicity criteria for oral exposures
are presented as administered doses (in mg/kg-d) and criteria for inhalation exposures are presented as
concentrations in air (in ug/m3), conversion of air concentrations was required to estimate an administered
dose comparable to the oral route. However, EPA's Office of Research and Development (ORD) now
believes that, due to portal-of-entry effects and differences in absorption in the gut versus the lungs, the
conversion from concentration in air to internal dose is not always appropriate and suggests evaluating
these exposure routes separately.
The models and assumptions used to calculate SSLs for the inhalation pathway are presented in Equations
2-4 through 2-11. Particular attention is given to the volatilization factor (VF), saturation limit (Csat), and
the dispersion portion of the VF and particulate emission factor (PEF) equations, all of which have been
revised since originally presented in RAGS Part B (U.S. EPA, 1991). Chemical-specific input parameters
required by these equations are shown in Table 5.
2.3.1 Screening Level Equations for Direct Inhalation. The equations used to calculate the SSL
for the inhalation of carcinogenic and noncarcinogenic contaminants are presented in Equations 2-4 and
2-5, respectively. These equations are unchanged from RAGS Part B and, therefore, the equations and
input parameters are displayed without discussion. The derivations of VF and PEF have been updated
since RAGS Part B was published and are discussed fully in Sections 2.3.2 and 2.3.3.
Equation 2-4: Screening Level Equation for Inhalation of Carcinogenic Contaminants In
Residential Soil
Screening Level
(mg/kg)
TR x AT x 365 d/yr
URF x 1000 ug/mg x EF x ED x _1_ + 1
VF PEF
18
-------
Table 5. Chemical-Specific Input Parameters for Ingestion and Inhalation Pathways
a-BHC
Arsenic
Benzo(a)pyrene
Benzene
Cadmium
Carbon tetrachloride
Chlordane
Chloro benzene
Chloroform
Chromium(VI)
Chrysene
DDT
1,4-Dichlorobenzene
1,1-Dichloroe thane
1 , 1 -Dichloroethylene
Dieldrin
Ethylbenzene
Mercury
Methylene chloride
Naphthalene
Nickel
PCB-1260
Pentachlorophenol
Perchloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
Trichloroethylene
Vinyl chloride
Xylene
SFo
6.30E+00
1.80E+00
7.3E+00
2.90E-02
1.30E-01
1.30E+00
6.10E-03
7.3E-03
3.40E-01
2.40E-02
6.00E-01
1.60E+01
7.50E-03
7.70E+00
1.20E-01
5.20E-02
1.10E-02
1.90E+00
URF
1.80E-03
4.30E-03
1.70E-03
8.30E-06
1.80E-03
1.50E-05
3.70E-04
2.30E-05
1.20E-02
1.70E-06
9.70E-05
5.00E-05
4.60E-03
4.70E-07
2.40E-04
5.80E-07
1.70E-06
8.40E-05
RfD
3.00E-04
5.00E-04
7.00E-04
6.00E-05
2.00E-02
1.00E-02
5.00E-03
5.00E-04
1.00E-01
9.00E-03
5.00E-05
1.00E-01
3.00E-04
6.00E-02
4.00E-02
2.00E-02
3.00E-02
1.00E-02
2.00E-01
1.00E-02
2.00E+00
RfC
2.00E-02
8.00E-01
5.00E-01
1.00E+00
3.00E-04
3.00E+00
4.00E-01
9.00E-03
1.00E+00
H
4.1E-05
1.5E-06
5.6E-03
2.4E-02
1.9E-03
3.7E-03
2.9E-03
9.5E-07
5.1E-04
2.9E-03
4.3E-03
3.4E-02
4.6E-07
6.4E-03
1.1E-02
2.0E-03
4.6E-04
1.7E-04
2.8E-06
2.6E-02
6.4E-03
2.3E-03
1.4E-02
9.1E-03
8.2E-02
7.0E-03
Kd
3.6E+01
3.5E+04
1.0E+00
4.6E+00
9.9E+02
3.5E+00
9.2E-01
1.9E+03
7.7E+03
1.0E+01
6.8E-01
1.3E+00
2.7E+02
3.7E+00
1.5E+02
4.6E-01
1.6E+01
1.6E+04
8.2E+00
5.3E+00
2.6E+00
3.1E+01
2.6E+00
1.7E+00
2.2E-01
4.7E+00
KM
4.7E-05
1.8E-09
2.3E-01
2.1E-01
7.9E-05
4.3E-02
1.3E-01
2.1E-08
2.7E-06
1.2E-02
2.6E-01
1.1E+00
7.0E-08
7.1E-02
3.1E-03
1.8E-01
1.2E-03
4.4E-07
1.4E-05
2.0E-01
1.0E-01
3.0E-03
2.2E-01
2.2E-01
1.5E+01
6.1E-02
D.,
4.2E-03
3.5E-03
7.0E-03
6.4E-03
3.4E-03
6.2E-03
7.1E-03
3.7E-03
3.5E-03
5.7E-03
7.2E-03
7.3E-03
3.4E-03
5.8E-03
2.2E-03
8.6E-03
5.1E-03
3.6E-03
4.6E-03
6.0E-03
6.3E-03
5.3E-03
6.4E-03
6.5E-03
8.6E-03
5.8E-03
a
2.9E-08
9.0E-13
2.3E-04
1.9E-04
3.9E-08
3.9E-05
1.3E-04
1.1E-11
1.4E-09
9.9E-06
2.6E-04
9.9E-04
3.5E-11
6.0E-05
9.9E-07
2.2E-04
8.8E-07
2.3E-10
9.5E-09
1.7E-04
9.2E-05
2.4E-06
2.0E-04
2.0E-04
5.9E-03
5.2E-05
VF
7.8E+05
O.OE+00
1.4E+08
8.5E+03
9.2E+03
6.7E+05
2.1E+04
1.1E+04
4.0E+07
3.6E+06
4.2E+04
7.9E+03
3.6E+03
2.3E+07
1.7E+04
1.3E+05
8.7E+03
1.4E+05
8.8E+06
1.4E+06
9.8E+03
1.4E+04
8.6E+04
9.1E+03
9.0E+03
5.3E+02
1.8E+04
Solub
4.8E+00
3.8E-03
1.8E+03
7.6E+02
6.0E-03
4.7E+02
8.2E+03
2.0E-03
5.0E-03
7.9E+01
5.5E+03
2.3E+03
1.9E-01
1.5E+02
2.0E+04
3.2E+01
1.4E-02
1.4E+01
1.5E+02
5.4E+02
3.0E+01
1.5E+03
1.1E+03
2.7E+03
2.0E+02
cw
4.8E-01
O.OE+00
3.8E-04
1.8E+02
O.OE+00
7.6E+01
6.0E-04
4.7E+01
8.2E+02
O.OE+00
2.0E-04
5.0E-04
7.9E+00
5.5E+02
2.3E+02
1.9E-02
1.5E+01
O.OE+00
2.0E+03
3.2E+00
O.OE+00
1.4E-03
1.4E+00
1.5E+01
5.4E+01
3.0E+00
1.5E+02
1.1E+02
2.7E+02
2.0E+01
H'
1.7E-03
O.OE+00
6.2E-05
2.3E-01
O.OE+00
9.8E-01
7.8E-02
1.5E-01
1.2E-01
O.OE+00
3.9E-05
2.1E-02
1.2E-01
1.8E-01
1.4E+00
1.9E-05
2.6E-01
4.7E-01
8.2E-02
1.9E-02
O.OE+00
7.0E-03
1.1E-04
1.1E+00
2.6E-01
9.4E-02
5.7E-01
3.7E-01
3.4E+00
2.9E-01
Ci«t
17.3
0.0
13.3
205.7
0.0
371.2
0.6
170.5
854.6
0.0
0.4
3.9
80.0
447.1
381.8
5.1
57.7
0.0
150.6
51.5
0.0
22.4
11.6
84.0
148.4
93.4
421.1
205.7
255.8
97.1
Note: See Equations 2-1 through 2-7 for units.
-------
Parameter/Definition (units)
TR/target cancer risk (unitless)
AT/averaging time (yr)
URF/inhalation unit risk factor
(pg/m3)-1
EF/exposure frequency (d/yr)
ED/exposure duration (yr)
VF/soil-to-air volatilization factor
(m3/kg)
PEF/particulate emission factor
(m3/kg)
Default
10'6
70
Chemical-specific
350
30
Chemical-specific
4.51 x 109
Source: RAGS Part B (U.S. EPA, 1991).
Equation 2-5: Screening Level Equation for Inhalation of Noncarclnogenlc Contaminants in
Residential Soil
Screening Level
(mg/kg)
THQ x AT x 365 d/yr
EF x ED x
LL
|_RfC
1
1
VF PEF
Parameter/Definition (units)
THQ/target hazard quotient (unitless)
AT/averaging time (yr)
EF/exposure frequency (d/yr)
ED/exposure duration (yr)
RfC/inhalation reference
concentration (mg/m3)
VF/soil-to-air volatilization factor
(m3/kg)
PEF/particulate emission factor
(m3/kg) (Equation 2-10)
Default
1
30
350
30
Chemical-specific
Chemical-specific
4.51 x 109
Source: RAGS Part B (U.S. EPA, 1991).
2.3.2 Volatilization Factor. To calculate inhalation SSLs, the volatilization factor must be calculated
from Equation 2-6, using chemical-specific physical and chemical properties (see Table 5 on page 19).
The soil-to-air VF is used to define the relationship between the concentration of the contaminant in soil
and the flux of the volatilized contaminant to air. The VF equation can be broken into two separate
models: a volatilization model to estimate the emissions of the contaminant from the soil, and a dispersion
model to simulate the dispersion of the contaminant in the atmosphere.
20
-------
Equation 2-6: Derivation of the Volatilization Factor
VF (m 3/kg) = (Q/C) x (3-14 x a x T)1/2
(2 x Dei x Pa x KJ
where
a =
Dei ><
Pa + (ps) (1 - Pa)/Ka
Parameter/Definition (units)
Default
Source
VF/volatilization factor (m3/kg)
(Q/C)/inverse of the mean cone, at
the center of a 0.5-acre square
source (g/m2-s per kg/m3)
T/exposure interval (s)
Dei/effective diffusivity (cm2/s)
Pa/air-filled soil porosity (unitless)
Pt/total soil porosity (unitless)
0/soU moisture content
(cm3-water/g-soil)
p/soil bulk density (g/cm3)
Ps/true soil density or particle density
(g/cm3)
K^soil-air partition coefficient
(g-soil/cm3-air)
D/diffusivity in air (cm2/s)
H/Henry's law constant
(atm-m3/mol)
Kd/soil-water partition coefficient
(cm3/g)
K^/organic carbon partition
coefficient (cm3/g)
OC/organic carbon content of soil
(fraction)
101.8
7.9 x 108 s
1-(P/PS)
0.1 (10%)
1.5
2.65
(H/Kd) x 41 (41 is a
conversion factor)
Chemical-specific
Chemical-specific
K^xOC
Chemical-specific
0.02 (2%)
EQM, 1993
U.S. EPA, 1991
EQM, 1992
EQM, 1992
EQM, 1992
EQM, 1993
EQM, 1993
U.S. EPA, 1991
U.S. EPA, 1991
U.S. EPA, 1991
U.S. EPA, 1991
Volatilization Model. The Hwang and Falco (1986) model was used as the basis for the VF equation
presented in RAGS Part B. This model was derived from methods presented by Farmer and Letey (1974)
and Farmer et al. (1980). Farmer et al. presented the empirical equation together with experimental data
involving the volatilization of pesticides from a soil surface. To simplify the calculations, the effective
diffusivity in soil, Dei, which accounts for tortuosity effects (i.e., physical impediments to vapor transport
associated with moisture and soil particles) in porous media, was later approximated by Hwang and Falco
using the effective porosity in dry soil, as shown in Equation 2-7.
21
-------
Equation 2-7: RAGS Part B D,, Equation for Dry Soil
Dei=DixE
0.33
where
Dei = effective diffusivity (cm2/s)
Dj = diffusivity in air (cm2/s)
E = true soil porosity (unitless).
During the reevaluation of RAGS Part B, OERR sponsored a study (Appendix B; EQM, 1992) to validate
the VF equation by comparing the modeled results with data from (1) a bench-scale pesticide study
(Farmer and Letey, 1974) and (2) a pilot-scale study measuring the rate of loss of benzene, toluene,
xylenes, and ethylbenzene from soils using an isolation flux chamber (Radian, 1989). The results of the
study verified the need to modify the VF equation in Part B to take into account the decrease in the rate
of flux due to the effect of soil moisture on effective diffusivity (Dei).
As used in the SSL calculations of VF, Dei is again used to account for the tortuosity effects in porous
media. However, the equation used to describe the decreased flux is taken from Millington and Quirk
(1961) (Equation 2-8). The flux calculation is refined to give a decreased flux rate due to reduced air-
filled porosity and, therefore, accounts for the effect of soil moisture on tortuosity. Use of the Millington
and Quirk expression reduces effective diffusivity and, as a result, reduces emissions.
Equation 2-8: SSL De, Equation with Soil Moisture Effects
Dei=Dix(Pa333/Pt2)
where
Dei = effective diffusivity (cm2/s)
Dj = diffusivity in air (cm2/s)
Pa = air-filled soil porosity (unitless)
P, = total soil porosity (unitless).
In revising the Dei equation, EQM (1992) cited two other studies in addition to Millington and Quirk
(1961). Farmer et al. (1980) and Hartley (1964) defined the soil solution/soil air partition coefficient as
the ratio of the solubility of the contaminant in water to the saturation vapor concentration of the
contaminant and used this ratio to estimate the diffusion between the vapor and nonvapor phase. Farmer
et al. postulated that, although air-filled porosity is found to be a major factor controlling volatilization
flux through the soil-water-air system, the apparent vapor diffusion coefficient does not depend solely on
the amount of air-filled pore space. The presence of liquid films on the solid surfaces not only reduces
the porosity but also modifies the pore geometry and the length of the gas passage.
The assumptions used to estimate Dei with Equation 2-8 are largely the same as those presented in RAGS
Part B (U.S. EPA, 1991). For example, Equations 2-7 and 2-8 indicate that vapor phase diffusion is
regarded as the only transport mechanism for both methods (e.g., no transport takes place via nonvapor
phase diffusion and there is no mass flow due to capillary action). The net result of the change is a less
conservative estimate of the volatilization, due to consideration of the effect of soil moisture content on
the amount and tortuosity of air-filled pore space.
22
-------
Dispersion Model. The box model in RAGS Part B has been replaced with a regression equation
derived from a modeling exercise using meteorological data from 29 locations across the United States
(Appendix C; EQM, 1993). The dispersion model used in the Part B guidance is based on the assumption
that emissions into a hypothetical box will be distributed uniformly throughout the box. To arrive at the
volume within the box (the cubic meters portion of an air concentration expressed as micrograms per cubic
meter), it is necessary to assign values to the length, width, and height of the box. The length (LS) was
the length of a side of a contaminated site with a default value of 45 m; the width was based on the wind
speed in the mixing zone (V) with a default value of 2.25 m (based on a wind speed of 2.25 m/s); and
the height was the diffusion height (DH) with a default value of 2 m.
However, the assumptions and mathematical treatment of dispersion used in the box model may not be
applicable to a broad range of site types and meteorology and do not utilize state-of-the-art techniques
developed for regulatory dispersion modeling. OERR was very concerned about the defensibility of the
box model and contracted Environmental Quality Management to help develop an updated and more
defensible dispersion model that could be used as a replacement in the Part B guidance. The updated
model is presented in EQM (1993) (Appendix C of this report).
The updated model used in the September 1993 SSL draft guidance has the following characteristics:
• It models dispersion from a ground-level area source.
• The receptor is onsite.
• The exposure point concentration represents a long-term/annual average.
• Algorithms for calculating the exposure point concentration are provided for area sources of
different sizes and shapes.
The result of the EQM (1993) analysis is the Q/C term in the VF and PEF equations. As with the output
of the box model in Part B, Q/C is the inverse of the mean concentration at the center of a 0.5-acre-square
source.
Several Q/C algorithms were derived by plotting the results of a number of runs using the Industrial
Source Complex Model, Version 2, in the short-term mode (ISCST2) with full-year meteorologic data
from 29 locations across the United States. The modeling runs included four areas, ranging from 0.5 acres
to 500 acres, and three possible site shapes: a square or a 1:3 or a 1:5 rectangular shape. The analysis
indicated that the configuration of the site did not impact the results significantly, but the 1:3 rectangular
shape provided the most conservative results.
The dispersion modeling was used to derive a regression equation for each site shape relating modeled
air concentration to site area. By plotting the natural log of the normalized air concentration versus the
natural log of the area, a regression equation was developed to approximate the 95% upper confidence
limit of the mean (95% UCLM) of the air concentrations expected at sites. Plugging the default 0.5-acre
site size into the regression equation (Equation 2-9) results in the Q/C value of 101.8 g/m2-s per kg/m3.
Equation 2-9: Regression Equation Used To Estimate Dispersion (EQM, 1993)
Q/c,
mean
2
\
g/m-s
kg/m
3
= [exp(0.1005x-5.3880)]
-i
23
-------
where
x = In (area of contamination in m ).
Soil Saturation Limit. Because of its reliance on Henry's law, the VF model is applicable only when
the contaminant concentration in soil water is at or below saturation (i.e., there .is no free-phase
contaminant present). This corresponds to the contaminant concentration in soil at which the adsorptive
limits of the soil particles and the solubility limits of the available soil moisture have been reached.
Above this point, pure liquid-phase contaminant is expected in the soil. Under such conditions, the partial
pressure of the pure contaminant and the partial pressure of the air in the interstitial pore spaces cannot
be calculated without first knowing the mole fraction of the contaminant in the soil. Therefore, the SSL
cannot be accurately calculated with the VF model. Because of this limitation, the chemical concentration
in soil (SSL) calculated using VF must be compared with the soil concentration at which the soil pore
water is saturated (Csat). If the SSL calculated using VF is greater than Csat, the SSL is set equal to Csat.
The updated equation for deriving Csat is presented in Equation 2-10. This equation takes into account
the amount of contaminant that is in the vapor phase in the pore spaces of the soil in addition to the
amount that is dissolved in the liquid and sorbed to the soil particles.
Equation 2-10: Derivation of the Soil Saturation Limit
= (Kd x Cw x p) + (Cw x Pw) + (Cw x H' x Pa)
-= p
Parameter/Definition (units)
Default
Source
C^/soil saturation concentration
(mg/kg)
Kj/soil-water partition coefficient
(L/kg)
Inorganic carbon partition
coefficient (L/kg)
OC/organic carbon content of soil
(fraction)
Cw/upper-limit of free moisture in
soil (mg/L-water)
S/solubility in water (mg/L-water)
0m/soil moisture content
(kg-water/kg-soil)
p/soil bulk density (kg/L)
Pa/air-filled soil porosity (unitless)
P^/water-filled soil porosity
(unitless)
P/total soil porosity (unitless)
H'/Henry's law constant (unitless)
H/Henry's law constant
(atm-m3/mol)
0/soil moisture content
(L-water/kg-soil)
Ps/true soil density or particle density
(kg/L)
Chemical-specific
0.02 (2%)
Chemical-specific
10% or 0.1
1.5
P - P
rt ra
H x 41, where 41 is
a conversion factor
Chemical-specific
0.1 (10%)
2.65
U.S. EPA, 1991
U.S. EPA, 1991
U.S. EPA, 1991
EQM, 1993
EQM, 1993
EQM, 1992
EQM, 1992
EQM, 1993
EQM, 1992
U.S. EPA, 1991
EQM, 1993
U.S. EPA, 1991
24
-------
2.3.3 Particulate Emission Factor. The paniculate emission factor relates the contaminant
concentration in soil with the concentration of respirable particles (PM10) in the air due to fugitive dust
emissions. Dust generated from open sources is termed "fugitive" because it is not discharged into the
atmosphere in a confined flow stream. Common sources of fugitive dust include unpaved roads,
agricultural tilling operations, aggregate storage piles, and heavy construction operations.
Although the emissions portion of the equation has not been revised since RAGS Part B (i.e., wind-blown
emissions across the site), the dispersion portion formerly derived using the box model has been updated
with the results of an equation derived from the regression analysis discussed in Section 2.3.2. Since it
is based upon site assumptions and not chemical-specific properties, the paniculate emission factor is
calculated using Equation 2-11 and is constant at 4.51 x 109 m3/kg .
The paniculate emission factor derived by using the default values in Equation 2-11 corresponds to a
receptor point concentration of approximately 0.2 pg/m3. This represents an annual average emission rate
estimate that is not appropriate for estimating acute effects. Over the next few months, OSWER will be
investigating the impact of acute exposure estimates on the SSLs.
Equation 2-11: Derivation of the Particulate Emission Factor
PEF(m3/kg) =(Q/C)x
3,600 s/h
0.036 x (1-G) x (U,,/Ut)3 x F(x)
Parameter/Definition (units)
Default
Source
PEF/particulate emission factor
(m3/kg)
(Q/C)/inverse of the mean cone, at
the center of a 0.5-acre-square
source (g/m2-s per kg/m3)
0.036/respirable fraction (unitless)
G/fraction of vegetative cover
(unitless)
Um/mean annual wind speed (m/s)
lyequivalent threshold value of wind
speed at 10 m (m/s)
F(x)/function dependent on Um/Ut
derived using Cowherd (U.S. EPA,
1985a) (unitless)
4.51 x 109
101.8
0.036
0
4.5
12.8
0.0497
EQM, 1993
U.S. EPA, 1991
U.S. EPA, 1991
U.S. EPA, 1991
U.S. EPA, 1991
U.S. EPA, 1991
2.4 Soil Screening Levels for Protection of Ground Water
Potentially toxic chemicals in subsurface soils can pose a threat to human health through the ground water
exposure pathway. Soil screening levels for protection of ground water were developed to identify
chemical concentrations that have the potential to contaminate ground water. The following attributes
were considered important during the development of SSLs for the ground water pathway:
• To ensure early-stage applicability, the method for developing and applying SSLs needs to be
simple, requiring as little site-specific information as possible.
25
-------
• Because of the large nationwide variability in ground water vulnerability, ground water SSLs
should be flexible, allowing adjustments for site-specific conditions if adequate information
is available.
• SSLs should be derived in a manner that is consistent with current understanding of subsurface
processes.
• To the extent possible, the process of deriving and applying SSLs should generate information
that can be used and built upon as the site assessment progresses.
These factors required development of a very simple approach that was nevertheless based on sound
geochemical and hydrogeologic principles and practices.
SSLs for the ground water pathway were developed to be generic (i.e., to be applied nationally to all
sites), by selecting generic assumptions to replace site-specific data, and to be simple to apply and to
modify through the tiered approach described in Section 1.7 of this document. Tier 1 SSLs are designed
for use at the early stages of a site assessment when information about subsurface conditions may be
limited. Because of these constraints, Tier 1 SSLs are based on very conservative simplifying assumptions
about the release and fate and transport of contaminants in the subsurface. As a result, their use as
cleanup levels for some sites would result in unnecessary expenditures of time and money on cleaning up
soils that may not pose a risk to ground water.
SSLs are theoretically and operationally consistent with the more sophisticated modeling efforts that are
conducted to develop the site-specific soil cleanup goals and final soil cleanup levels for protection of
ground water. They thus can be viewed as evolving risk-based levels that may change during the site-
assessment process as more site-specific information becomes available. Through the tiered approach
described in Section 1.7, SSLs can be adjusted if adequate site-specific information is (or becomes)
available. The tiers enable SSLs to be changed as the site assessment continues. Less (or more)
conservative values may be derived if site conditions warrant.
The early use of SSLs at a site will help to focus site investigation efforts on areas of true concern with
respect to ground water quality and will provide information that can be refined during later stages of the
assessment. For example, the methodology used to develop unadjusted Tier 1 SSLs can be used to
develop soil leachate concentration inputs for Tier 4 contaminant fate and transport modeling through the
vadose and saturated zones. In addition, one of the parameters used to develop the Tier 1 SSLs for
organic compounds, K^., is a necessary input to site-specific contaminant fate-and-transport modeling that
may be performed during Tier 4. This document presents valid K^ values for the organic chemicals that
most commonly occur at Superfund sites.
2.4.1 Conceptual Framework and Organization. Contaminant migration through the soil-to-
ground-water exposure pathway can be envisioned as a two-stage process: (1) release of the contaminant
in soil leachate, and (2) transport of the contaminant by water through the unsaturated (vadose) and
saturated (aquifer) zones. In order for a methodology to accurately estimate soil concentration thresholds
that will not result in contamination at the well, it must consider both processes. Thus, to evaluate
potential risks that contaminants pose to human health, one must determine (1) the potential for the
chemical to be released from the soil through leaching and (2) the potential for the chemical to migrate
through the unsaturated zone into ground water and through the aquifer to the receptor's drinking water
well.
The methodology used to calculate Tier 1 soil screening levels for protection of ground water estimates
contaminant release using an equilibrium soil-water partition equation to relate soil leachate (pore water)
26
-------
contaminant concentrations to the concentrations of contaminants sorbed to the soil mass. This equation
allows calculation of a soil concentration (in this case, SSL) in equilibrium with a specified target soil
leachate concentration such as an MCL. These levels can then be used as very conservative indicators
of ground water contamination potential or may be adjusted using a dilution/attenuation factor (DAF)
appropriate for the site (see Section 2.4.6). Section 2.4.2 describes the development of the partitioning
methodology. Sections 2.4.3 and 2.4.4 explain how SSLs are calculated for organics and inorganics,
respectively. Section 2.4.5 lists the assumptions inherent in the soil/water partition equation. These
assumptions should be read and understood before applying SSLs.
Transport of contaminants through the unsaturated and saturated zones usually results in a reduction in
soil leachate concentrations by dilution and through attenuation processes such as adsorption and
degradation. DAFs are used to adjust soil leachate concentrations to account for concentration reduction
by these processes during transport in the subsurface. Section 2.4.6 describes the development and use
of DAFs for adjusting SSLs.
Section 2.4.7 presents the SSLs calculated for protection of ground water. Section 2.4.8 provides guidance
on using Tiers 2 through 4 for adjusting SSLs to site-specific conditions.
2.4.2 Development of Soil Partitioning Equation. The methodology used to develop SSLs
unadjusted for dilution and attenuation is based on the Freundlich equation, which was developed to model
sorption from liquids to solids. The basic Freundlich equation applied to the soil/water system is:
Kd= Cs/Cwn (2-12)
where
Kd = Freundlich adsorption constant (L/kg)
Cs = concentration sorbed on soil (mg/kg)
Cw = solution concentration (mg/L)
n = Freundlich exponent (dimensionless).
Assuming that adsorption is linear with respect to concentration (n=l)* and rearranging:
Cs=(Kd)Cw . (2-13)
For SSL calculation, Cw is the target ground water concentration (e.g., MCLG or MCL).
Adjusting Total Soil Concentrations to Sorbed Concentrations. To specify the action level for
soil samples taken at Superfund sites, one must relate the sorbed concentration (Cs) to the total
concentration measured in a soil sample (C,). Assuming soil-gas concentrations to be insignificant with
respect to total sample concentration (in addition, soil-gas is often lost during sampling), contaminants can
be associated with both the solid soil materials and the soil water (Feenstra et al., 1991). That is:
M, = Ms + Mw (2-14)
*The linear assumption will tend to overestimate sorption and underestimate desorption for most organics at
higher concentrations (i.e., above 10"5 M for organics) (Piwoni and Banerjee, 1989).
27
-------
where
Mj = total contaminant mass in sample (mg)
Ms = contaminant mass sorbed on soil materials (mg)
Mw = contaminant mass in soil water (mg).
Furthermore:
where
BD = dry soil bulk density (kg/L)
Vsp = sample volume (L)
and
Ms=CsBDVsp (2-15)
= Cw 6W Vsp (2-16)
where
6W = water-filled porosity (void fraction)
and
Mt = qBDVsp . (2-17)
Substituting into Equation 2-14:
C, = (Cs BD + Cw 9W) / BD (2-18)
or
Cs = q- C^QJBD . (2-19)
Substituting into Equation 2-13 and rearranging:
q = Cw(Kd + 6,/BD) . (2-20)
The water-filled soil porosity (8W) is related to total soil porosity (0) by:
6W = 6 S (2-21)
where
S = fraction water content (LwateApore)-
Thus, Equation 2-20 becomes:
q = Cw [Kd + (9S/BD)] . (2-22)
Equation 2-22 is used to calculate SSLs (Q) for the ground water pathway, where Cw is the target ground
water concentration for the contaminant in question (see Section 2.1). Kd varies by chemical and soil
type. Because of different influences on Kd values, derivations of Kd values for organic compounds and
metals were treated separately.
28
-------
2.4.3 Organic Compounds— Partition Theory. Past research has demonstrated that, for hydrophobia
organic chemicals, soil organic matter is the dominant sorbing component in soil and that K<, is linear with
respect to soil organic carbon content (OC) as long as OC is above a critical level. Thus, Kd can be
normalized with respect to soil organic carbon to K^, a chemical-specific partitioning coefficient that is
independent of soil type, as follows:
(2-23)
where
= organic carbon partition coefficient (L/kg)
= fraction of organic carbon in soil (mg/mg).
Substituting into Equation 2-22:
. (2-24)
The critical organic carbon content, f^,* , represents OC below which sorption to mineral surfaces begins
to be significant. This level is likely to be variable and to depend on both the properties of the soil and
of the chemical sorbate (Curtis et al., 1986). Attempts to quantitatively relate f^* to such properties (see
McCarty et al., 1981) have been made, but at this time there is no reliable method for estimating f^* for
specific chemicals and soils. Nevertheless, research has demonstrated that, for volatile halogenated
hydrocarbons, f^* is about 0.001, or 0.1 percent OC, for many low-carbon soils and aquifer materials
(Schwarzenbach and Westall, 1981; Piwoni and Banerjee, 1989).
If soil OC is below this critical level, Equation 2-24 should be used with caution. This is especially true
if soils contain significant quantities of fine-grained minerals with high sorptive properties (e.g., clays).
If sorption to minerals is significant, Equation 2-24 will underpredict sorption and overpredict contaminant
concentrations in soil pore water. However, this f^* level is by no means the case for all soils; Abdul
et al. (1987) found that, for a variety of organic compounds and aquifer materials, sorption was linear and
could be adequately modeled down to f^ = 0.0003 by considering K^, alone.
For soils with significant inorganic and organic sorption (i.e., soils with f^ < 0.001), the following
equation has been developed (McCarty et al., 1981; Karickhoff, 1984):
Kd = (K^ fj + (Ki0 fio) (2-25)
where
fio = fraction of inorganic material
fio + foe = 1
K^ = soil inorganic partition coefficient.
Although this equation is considered conceptually valid, KJO values are not available for the subject
chemicals. Attempts to estimate KJO values by relating sorption on low-carbon materials to properties such
as clay-size fraction, clay mineralogy, surface area, or iron-oxide content have not revealed any consistent
correlations and semiquantitative methods are probably years away (Piwoni and Banerjee, 1989). How-
ever, Piwoni and Banerjee developed the following empirical correlation (by linear regression, r2 = 0.85)
that can be used to estimate Kd values for hydrophobic organic chemicals from Kow for low-carbon soils:
log Kd = 1.01 log Kow - 0.36 (2-26)
29
-------
where
Kow = octanol/water partition coefficient.
The authors indicate that this equation should provide a Kd estimate that is within a factor of 2 or 3 of
the actual value for nonpolar sorbates with log Kow < 3.7. If sorption to inorganics is not considered for
low-carbon soils where it is significant, Equation 2-24 will underpredict sorption and overpredict
contaminant concentrations in soil pore water (i.e., it will provide a conservative estimate).
Equation 2-24 is valid only for hydrophobic, nonionizing organic chemicals. One of the organic chemicals
of concern, pentachlorophenol (PCP), ionizes in the soil environment, existing in both neutral
(pentachlorophenol) and ionized (pentachlorophenate) forms within the normal ground water pH range.
The relative amounts of the ionized and neutral species are a function of pH. Because the sorptive
properties of these two forms differ, it is important to consider the relative amounts of the neutral and
ionized species when determining K^. values at a particular pH. Lee et al. (1990) developed a theoreti-
cally based algorithm, developed from thermodynamic equilibrium equations, and demonstrated that the
equation adequately predicts laboratory-measured K^ values for PCP.
The equation assumes that sorbent organic carbon determines the extent of sorption for both the ionized
and neutral species and predicts the overall sorption of a weak organic acid (Koc p ) as follows:
KOC,P = K^O^ K^d-*,,) (2-27)
where
Koc,n> Kocj = sorption coefficients for the neutral and ionized species (L/kg)
-------
Once collected, the values were reviewed. It was not possible to systematically evaluate each source for
accuracy or consistency or to analyze sources of variability because of the wide variations in soil and
sediment properties and experimental and analytical methods and how these were reported in each
reference. This, and the limited number of K^ values for many compounds, prevented any meaningful
statistical analysis to eliminate outliers. The values were qualitatively reviewed, however, and some values
were excluded. Values measured for low-carbon content sorbents were rejected in most cases. Some
references produced consistently high or low values and were eliminated. The final values used are
presented in Appendix C, along with their reference sources.
The range, geometric mean, and number of K^, values for each chemical are presented in Table 6. The
geometric mean was used because it provides a better estimate of central tendency for parameters with
wide variability. Published values were not found for vinyl chloride, chrysene, and three of the Arochlors
(PCBs). To estimate K^ values for these compounds, a linear regression was conducted using the data
for other compounds to obtain the following relationship between K^ and Kow:
K = 0.88 K - 0.114 (1^ = 0.96) . (2-29)
••oc
Kow values used in this correlation (Table 6) were collected in the same way that K^ values were
collected. Figure 4 shows the regression line and the values used to generate it. Figure 5 compares the
measured and predicted K^, values for the subject chemicals. Equation 2-29 may be used to estimate K^.
values for other hydrophobic, nonionizing organic chemicals. This is preferable because of potential
variability in literature and database values, unless a thorough review of published measured values is
possible.
As explained in Section 2.4.3, PCP K^ varies with pH and can be estimated for a particular pH using
Equation 2-27. Lee et al. (1991) derive K^ values from literature measurements of PCP K^ at pH < 3
(19,950 mL/g) where PCP is essentially 100 percent neutral. Similarly, K^ is derived from
measurements of PCP K^ at pH > 9 (400 mL/g) where PCP is essentially 100 percent ionized. Figure 6
plots PCP KO,. values as a function of pH, derived from Equation 2-27 and these values. Table 7 shows
the K^ values for pH values of 5 to 8.2. PCP log K^ values at average and 90th percentile ground water
pH conditions (6.8 and 8.0)* are 2.76 and 2.61 (K^ = 571 L/kg and 407 L/kg), respectively. The 407-
L/kg value was used as a conservative value to calculate the PCP SSLs in this document. However, the
user is encouraged to select and use a more appropriate value K^ if soil pH conditions are known.
2.4.4 Inorganics (Metals)—Partition Theory. The derivation of Kd values is much more
complicated for metals than for organic compounds. Unlike organic compounds, for which Kj values are
largely controlled by a single parameter (soil organic carbon), Kd values for metals are significantly
affected by a variety of soil conditions. The most significant parameters are pH, oxidation-reduction
conditions, iron oxide content, soil organic matter content, cation exchange capacity, and major ion
chemistry. The number of significant influencing parameters, their variability in the field, and differences
in experimental methods result in a wide range of Kd values for individual metals reported in the literature
(over five orders of magnitude). Thus, it is much more difficult to derive generic Kd values for metals
than for organics.
The Kj values used to generate metal SSLs (Table 7) for Cd, Cr3"1", Hg, and Ni were taken from values
generated by EPA using an equilibrium geochemical speciation model (MINTEQ2) for possible use in the
Resource Conservation and Recovery Act (RCRA) Hazardous Waste Identification Rule (HWIR) (U.S.
EPA, 1992a). The values for As and Cr6* were taken from empirical, pH-dependent adsorption
*STORET database values from U.S. EPA (1992a).
31
-------
Table 6. K^ and Kow Values: SSL Organic Chemicals
KOW Koc (L/kg)
Chemical
a-BHC
Benzene
Benzo(a)pyrene
P-BHC
BHC (all isomers)
Carbon tetrachloride
Chlordane
Chlorobenzene
Chloroform
Chrysene
DDT
1 ,4-Dichlorobenzene
1,1-Dichloroethane
1,1-Dichloroethylene
Dieldrin
Ethylbenzene
y-BHC (lindane)
Methylene chloride
Naphthalene
PCBs
Arochlor 1016
Arochlor 1221
Arochlor 1232
Arochlor 1242
Arochlor 1254
Arochlor 1260
Pentachlorophenol
Tetrachloroethylene
1,2,4 -Trichlorobenzene
1,1,1 -Trichloroethane
Trichloroethylene
Toluene
Vinyl chloride
Xylenes
Geomean
6,183
138
1,424,943
7,007
6,074
503
1,128,496
678
90
616,595
1 ,740,940
2,757
61
64
327,529
1,306
4,935
19
2,313
758,578
22,387
30,200
1,000,000
1,972,423
7,161,434
119,536
860
10,715
233
253
549
20
1,099
No.
5
8
3
4
12
6
4
8
4
1
8
9
2
3
4
3
3
2
12
1
2
1
1
2
2
4
9
9
5
6
8
2
12
Calculated
1,661
59
198,785
1,855
1,636
183
161,917
238
40
95,153
237,077
817
29
30
54,548
423
1,363
10
700
114,177
5,152
6,703
145,586
264,589
822,422
NA
293
2,695
93
100
198
11
364
Mm
1,615
19
478,947
1,313
242
123
41,686
83
28
NA
150,000
273
30
NA
8,366
132
242
9
368
54,167
NA
NA
52,083
108,667
NA
NA
79
864
77
29
38
NA
129
Geomean
1,820
50
1,773,449
2,481
1,504
230
49,544
173
46
NA
386,977
511
34
65
13,400
187
1,208
23
792
107,285
NA
NA
169,340
431,380
NA
409*
264
1,561
128
87
129
NA
234
Max.
2,050
96
5,069,757
3,914
3,914
439
58,884
330
78
NA
2,290,868
900
42
NA
25,604
255
2,983
48
1,333
171,250
NA
NA
1,681,808
1 ,262,779
NA
NA
437
2,800
198
174
304
NA
413
No.
2
11
4
8
30
3
2
7
6
0
14
12
5
1
6
5
20
3
18
4
0
0
5
3
0
NA
17
10
8
23
11
0
13
Bold indicates values used for SSL calculations.
Italics indicates values calculated by Koc - (0.88 K0J - 0.114.
NA = Not applicable.
'Value at pH of 6.8
32
-------
7 ••
6
5 -
oo
£ 3 -•
2 -
1 -
0--
0
Measured K^
K^ = (08821^)-0119
(r = 0.98)
Figure 4. KOC/KOW correlation plot.
/.uu n
6 no •
5 00 •
3 4.00 •
r? ^ DO •
200 •
1.00 t
o no •
'• •
•' • •
-•
.
•b'-n-'g-^
: : :
i i i
— r
J— C
i-
T — fc
J .
j
J L
J t
U ° - oa "• (2
S
I s
1
. • ,
•
3' c
c
r „- •
3 ^
r
i c
.... . . ., . .- • . - fc- - ,- - -
1
|
j
[
3 C
1
W g Jd w OQ U O
* J § 13 2 CQ CO
^ -*S j; • C ,C
">> £• 1 -S-
f -S § w
(2 °*
• Measured n Predicted
7 i
• -c
j
__4_c
,....
'
1
• !
3 n n. .Q L
? .
..
. . j . .
.
*
,
•
'~"UCQ c^CrJ i Q, H ^
M M ! H 8 8 "
i S 3 s | | g
< < <
Figure 5. Measured and predicted K
oc-
33
-------
-•neutral-species
123456
Source: Lee et al. (1990; 1991) PH
10 11
Figure 6. PCP and TeCP: log (Koc) vs. pH.
Table 7. Pentachlorophenol Koc as a Function of pH
PH
5.0
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
6.0
6.1
6.2
6.3
6.4
6.5
6.6
KOC
7,436
6,436
5,519
4,698
3,975
3,350
2,818
2,371
1,998
1,691
1,439
1,234
1,068
934
826
740
670
log (K.J PH
3.87
3.81
3.74
3.67
3.60
3.53
3.45
3.37
3.30
3.23
3.16
3.09
3.03
2.97
2.92
2.87
2.83
6.7
6.8
6.9
7.0
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
8.0
8.1
8.2
KOC
615
571
536
507
485
467
453
442
433
426
420
416
412
409
407
405
log (KO,.)
2.79
2.76
2.73
2.71
2.69
2.67
2.66
2.65
2.64
2.63
2.62
2.62
2.61
2.61
2.61
2.61
34
-------
relationships developed for the same effort. Because they are taken from a draft document, these numbers
should be regarded as preliminary and must be used with caution. The values for mercury are especially
suspect; Hg+2 is the only mercury species considered in these model runs. However, they are probably
the best generic values that exist at this time.
The HWIR MINTEQ2 analyses were conducted under a variety of geochemical conditions and metal
concentrations. For the important variables (pH, solid iron oxide [FeOx], dissolved organic matter, and
f^), MINTEQ2 runs were conducted at high, medium, and low values. Kd values selected for preliminary
SSL concentrations were modeled under the following conditions: f^ = 0.001, solid iron oxide = 0.31
percent, and dissolved organic matter = 105 mg/L. Metal concentrations were selected to be close to the
target ground water concentrations (Cw) used to calculate SSLs. All other inputs (e.g., major cations and
anion concentrations) were set at typical or average ground water conditions (U.S. EPA, 1992a). Since
the most significant variable affecting Kd is pH (and this parameter is relatively easy to measure in the
field), metal Kd values for three pH conditions were used to develop SSL levels (see Table 8): 4.9, 6.8,
and 8.0. These values represent 10th, 50th, and 90th percentile pH conditions for U.S. ground waters
(U.S. EPA, 1992a). Figure 7 shows the effect of pH on metal Kd values for the conditions described
above.
The MINTEQ2 model has been revised and improved since the 1992 HWIR analyses, mainly through
updates of its thermodynamic database. OERR is currently developing new Kd values for metals with the
updated model. Thus, the metal Kd values in the draft guidance, and the SSLs derived from them, should
be regarded as preliminary and subject to change in the final SSL guidance.
2.4.5 Assumptions for Soil/Water Partition Theory. The following assumptions are implicit in
the SSL partitioning methodology. These assumptions and their implications for SSL accuracy should be
read and understood before applying SSLs.
1. The concentration in the pore water immediately surrounding the contaminated soil particles
is equivalent to that at the receptor point (downgradient well). This ignores all subsurface
processes that operate to reduce dissolved contaminant concentrations as they are transported
through the unsaturated and saturated zones. These processes include:
• volatilization
• biodegradation
• chemical degradation
• adsorption to downgradient soil or aquifer materials
• dilution due to percolation (recharge) in the unsaturated zone or hydrodynamic dispersion and
mixing in the saturated zone.
This assumption is conservative. The use of a dilution/attenuation factor to account for the latter
two dilution/attenuation processes through a simple multiplier is described in the following
section. DAFs are developed through contaminant transport modeling (site-specific or generic).
2. There is no facilitated transport. This ignores processes such as colloidal transport, transport
via solvents other than water, and transport via dissolved organic matter (DOM). These processes
have greater impact as Kow (and hence, K^) increases.
This assumption is not conservative. However, the second process is only operative if certain
site-specific conditions are present. Transport by DOM and colloids has been shown to be
potentially significant under certain conditions in laboratory and field studies. Although much
35
-------
Table 8. Metal Kd Values as a Function of pH
Metal
Concentration
Metal (mg/L)a
Arsenic
Cadmium
Chromium(VI)
Mercury
Nickel
NA
0.005
NA
0.002
0.1
Metal Kd (L/kg)
pH: 4.9
25
1.6
31
0.1
3.2
6.8
29
162
19
152
82
8.0
31
2,002
14
211
157
NA = Not applicable.
a Concentration at which Kd values were calculated.
Source: U.S. EPA, 1992a.
4 T
— a— - .
— o - •
As
Cd
Cr+3
Cr+6
Hg+2
Ni
Figure 7. Effect of pH on metal Kd values.
36
-------
research is in progress on these processes, the current state of knowledge is not adequate to allow
for their consideration in SSL calculations.
If there is the potential for the presence of nonaqueous phase liquids (NAPLs) in soils at the site
or site area in question, SSLs should not be used for this area. See U.S. EPA (1992b) for
guidance on how to determine the likelihood of dense nonaqueous phase liquid (DNAPL) occur-
rence in the subsurface.
3. Adsorption is linear with concentration. The methodology assumes that adsorption is
independent of concentration (i.e., the Freundlich exponent = 1). This has been reported to be true
for various halogenated hydrocarbons, polynuclear aromatic hydrocarbons, benzene, and
chlorinated benzenes. In addition, this assumption is valid at low concentrations (e.g., at levels
close to the MCL) for most chemicals. As concentrations increase, however, the adsorption
isotherm can depart from the linear.
Studies on trichloroethane (TCE) and chlorobenzene indicate that departure from linear is in the
nonconservative direction, with adsorbed concentrations being lower than predicted by a linear
isotherm. However, adequate information is not available to establish nonlinear adsorption
isotherms for the chemicals of interest. Furthermore, since the SSLs are derived at relatively low
target ground water concentrations (e.g., MCLs), departures from the linear at high concentrations
do not significantly influence the accuracy of the results.
4. The system is at equilibrium with respect to adsorption. This ignores adsorption/desorption
kinetics by assuming that the soil and pore water concentrations are at equilibrium levels. In other
words, the pore-water residence time is assumed to be longer than the time it takes for the system
to reach equilibrium conditions.
This assumption is conservative. If equilibrium conditions are not met, the concentration in the
pore water will be less than that predicted by the methodology. The kinetics of adsorption are
not sufficiently understood for a sufficient number of chemicals and site conditions to consider
equilibrium kinetics in the methodology.
5. Adsorption is reversible. The methodology assumes that desorption processes operate in the
same way as adsorption processes, since most of the K^ values are measured by adsorption
experiments rather than by desorption experiments. In actuality, desorption is slower to some
degree than adsorption and, in some cases, organics can be irreversibly bound to the soil matrix.
In general, the significance of this effect increases with Kow.
This assumption is conservative. Tailing and irreversible sorption will result in lower pore-water
concentrations than predicted by the methodology. Again, the level of knowledge on desorption
processes is not sufficient to consider desorption kinetics and degree of reversibility for all of the
subject chemicals.
2.4.6 Dilution/Attenuation Factor Development. SSLs can be adjusted to account for contaminant
dilution and attenuation during transport through the vadose and saturated zones to a compliance point or
receptor well. Such adjusted values will more accurately reflect a contaminant's threat to ground water
resources. For the Tier 1 SSLs, a DAF of 10 is generally recommended, with some adjustment allowed
according to facility size. These DAF values were developed as follows.
As contaminants move through the soil and ground water, they are subjected to a number of physical,
chemical, and biological processes that affect the eventual contaminant concentration level at receptor
37
-------
points. These processes include, but are not limited to, attenuation due to sorption of contaminants onto
soil and aquifer grains, chemical transformation (e.g., hydrolysis, redox reactions, precipitation), biological
degradation, and dilution due to mixing of soil leachate with ambient ground water. The contaminant
concentration arriving at a receptor point is therefore generally lower than the original contaminant
concentration in the leachate leaving the site.
The reduction in concentration can be expressed succinctly by the DAF, defined as the ratio of the original
concentration of a contaminant in soil leachate to the concentration of the same contaminant at the receptor
point. The lowest possible value of DAF is therefore 1, corresponding to the situation where there is no
dilution or attenuation of a contaminant at all; i.e., the concentration at the receptor point is the same as
that at the source. High DAF values on the other hand correspond to a high degree of dilution and
attenuation of the contaminant from the source to the receptor point.
The Agency has developed subsurface fate and transport models to assess the impact on ground water
quality due to migration of contaminants from wastes on land. Specifically, these models predict the DAF
for a potential site of a domestic drinking water receptor well that may withdraw water from the saturated
zone under, or downgradient of, a contaminated area. EPACML (EPA's Composite Model for Landfills)
is the fate and transport model that EPA has used for the Toxicity Characteristic (TC) Rule (55 FR 11798,
March 1990) and for the RCRA Delisting Program (56 FR 67197, December 30, 1991). The TC Rule
finalized only those chemicals that are nondegraders and can be represented by an infinite source. The
EPACML is appropriate when there is no significant mounding of water table under the waste sites
because it accounts for only horizontal uniform flow.
The model used to develop DAFs for this guidance is EPA's Composite Model for Leachate Migration
with Transformation Products (EPACMTP, U.S. EPA, 1993c). The EPACMTP is an enhanced version
of the EPACML and was first proposed for use in HWIR (57 FR 21450). EPACMTP has a three-
dimensional module to simulate ground water flow that can account for mounding under waste sites. The
model also has a three-dimensional transport module and both linear and nonlinear adsorption in the
unsaturated and saturated zones and can simulate chain decay, thus allowing the simulation of the
formation and the fate and transport of daughter (transformation) products of degrading chemicals. The
model can also be used to simulate the finite source scenario.
EPACMTP is comprised of three main interconnected modules:
• An unsaturated zone flow and contaminant fate and transport module
• A saturated zone ground water flow and contaminant fate and transport module
• A Monte Carlo driver module, which generates model parameters from nationwide probability
distributions.
The unsaturated and saturated zone modules simulate the migration of contaminants from initial release
into the soil to a downgradient receptor well. The Agency has extensively verified both the unsaturated
and saturated zone modules against other available analytical and numerical models to ensure accuracy
and efficiency. Both the unsaturated zone and the saturated zone modules of the EPACMTP, used for the
calculation of DAFs for the SSLs, have been reviewed by the EPA Science Advisory Board and found
to be suitable for generic applications such as the derivation of nationwide DAFs.
Modeling Procedure. For nationwide Monte Carlo model applications, the input to the model is in the
form of probability distributions of each of the model input parameters. The output from the model
consists of the probability distribution of DAF values, representing the likelihood that the DAF will not
38
-------
be less than a certain value. For instance, an 85th percentile DAF of 100 means that the DAF will be 100
or higher in at least 85 percent of the cases.
For each model input parameter, a probability distribution is provided, describing the nationwide likelihood
that the parameter has a certain value. The parameters are divided into four main groups:
• Source-specific parameters, e.g., area of the waste unit, infiltration rate
• Chemical-specific parameters, e.g., hydrolysis constants, organic carbon partition coefficient
• Unsaturated zone-specific parameters, e.g., depth to water table, soil hydraulic conductivity
• Saturated zone-specific parameters, e.g., saturated zone thickness, ambient ground water flow
rate, location of nearest receptor well.
Probability distributions for each parameter used in the model have been derived from nationwide surveys
of waste sites, such as EPA's landfill survey (53 FR 28692). During the Monte Carlo simulation, values
for each model parameter are randomly drawn from their respective probability distributions. In the
calculation of the DAFs for the SSLs, site data from over 1,300 sites were used to define parameter ranges
and distributions. Each combination of randomly drawn parameter values represents one out of a
practically infinite universe of possible waste sites. The fate and transport modules are executed for the
specific set of model parameters, yielding a corresponding DAF value. This procedure is repeated,
typically on the order of several thousand times, to ensure that the entire universe of possible parameter
combinations (waste sites) is adequately sampled. In the derivation of DAFs for the SSLs, the model
simulations were repeated 15,000 times for each scenario investigated. At the conclusion of the analysis,
a cumulative frequency distribution of DAF values is constructed and plotted.
EPA assumed an infinite waste source of fixed surface area for the SSL modeling scenario. This
assumption is relatively conservative because EPA is considering no further investigation at sites where
soil contaminant concentrations are below the SSLs. However, the Agency is considering the use of the
finite source scenario for chemicals that degrade rapidly and/or are observed in ground water only at short
distances from the source.
For the SSL modeling scenario, the Agency performed a number of sensitivity analyses consisting of
fixing one parameter at a time to determine the parameters that have the greatest impact on DAFs. The
results of the sensitivity analyses indicate that the climate (net precipitation), soil types, and size of the
contaminated area have the greatest effect on the DAFs. The EPA feels that the size of the contaminated
area lends itself most readily to practical application of the SSLs. Figure 8 shows DAFs corresponding
to selected source areas for three percentiles.
To calculate the DAF for the SSLs, the receptor point was taken to be a domestic drinking water well
located 25 feet downgradient from the edge of the contaminated area. The location of the intake point
(receptor well screen) was assumed to vary between 15 and 300 feet below the water table (these values
are based on empirical data reflecting a national sample distribution of depth of residential drinking water
wells). The sensitivity analyses indicated that the placement of the well 25 feet downgradient of the
contaminated area is more conservative than allowing the well to be located directly beneath the
contaminated area. The location of the intake point allows for mixing within the aquifer. The Agency
believes that this is a reasonable assumption because there will always be some dilution attributed to the
pumping of water for residential use from an aquifer. The placement of the well was assumed to vary
uniformly within the lateral boundaries of the plume. Degradation and retardation of contaminants were
not considered in this analysis. Figure 9 shows a schematic of the compliance point location.
39
-------
1000
u_
<
O
0.1
AREA (acres)
85 TH PERC. -*- 90 TH PERC. -*- 95 TH PERC.
Figure 8. Effect of site area on DAF.
Parameters:
• X (distance from source to well) = 25 ft
• Y (transverse well location) = Monte Carlo within
width of plume
• Z (well intake point below water table) = Monte
Carlo, range 15 •+ 300 ft
• Rainfall = Monte Carlo
• Soil type = Monte Carlo
• Depth to aquifer = Monte Carlo
• Assumes infinite source term
Figure 9. Soil to ground water pathway—calculating the DAF.
40
-------
The results of these analyses, indicate that the largest allowable areas corresponding to DAFs of 10 and
100 at the 90th percentile protection level are approximately 10 and 1 acre, respectively. Therefore, for
sites of up to 10 acres, a DAF of 10 should be applied to the unadjusted SSLs, while for sites at or below
1 acre, a DAF of 100 should be applied to the unadjusted SSLs. If a 95th percentile protectiveness level
is used, a DAF of 10 is protective for areas less than 0.5 acre and a DAF of 100 is protective for areas
less than 0.1 acre. EPA is considering whether the 90th or 95th percentile protectiveness level should be
used in the final guidance.
When sites are located in areas of unusually shallow water table, within 5 feet of surface, unadjusted SSLs
should be used. In this scenario, contamination is located in or directly above the saturated zone;
therefore, any dilution and attenuation processes within the unsaturated zone would be negligible.
2.4.7 Soil Screening Levels. SSLs were calculated for the subject chemicals using Equation 2-22.
The target ground water concentration, Cw, used for each chemical was the nonzero drinking water MCLG,
the MCL, or the 10"6 ECR level (see Section 2.1).
SSLs—Organics. Equation 2-24 was used to calculate SSLs for the listed organic chemicals (Table 9).
Several inputs were necessary for this calculation. K^ values were obtained as described in Section 2.3.3.
Fractional soil organic carbon content (f^) was set at 0.002 (0.2 percent OC) representative of typical
organic carbon content of vadose zone materials below the root zone (Carsel et al., 1988). Default values
obtained from U.S. EPA (1985b) were used for soil porosity (0.5), fraction water content (0.3), and bulk
density (1.5 kg/L). Target ground water concentrations are given in Section 2.1.
SSLs—Inorganics. Equation 2-22 was used to calculate preliminary SSLs for the listed inorganic
chemicals (Table 10). The Kd values necessary to make these calculations were derived as described in
Section 2.3.4. Three sets of inorganic SSLs are presented, one for each of the following pH conditions:
4.9, 6.8, and 8.0. The user is encouraged to use the value that most closely approximates site pH
conditions or to select an appropriate value from Figure 7. If site pH values are not known, the SSLs for
pH of 6.8 should be used. Further site-specific adjustment of metal partitioning is more complicated;
research is under way to develop adjustable levels for metals. As stated previously, the metal KjS and
SSLs in this document are preliminary. The Agency is conducting studies to develop revised metal Kd
values that will be used in the final guidance.
Adjustment for Dilution and Attenuation. Tables 9 and 10 also present SSLs adjusted to account
for contaminant dilution and attenuation during transport through the vadose and saturated zones to a
receptor well. Two DAFs are presented: 10 and 100. The appropriate DAF should be selected based on
size of the contaminated area and site characteristics as described in Section 1.6.
2.4.8 Site-Specific Adjustments to Partitioning Calculations (Tiers 2-4). This section addresses
adjustments to the SSLs under the tiered approach described in Section 1.7. The Tier 2 levels represent
a minimal increase in site-specificity and perhaps less conservative screening levels. For organics, the
partitioning equation used in the Tier 1 calculation (Equation 2-24) remains the basis for the Tier 2 levels
along with the same DAF (either 1, 10, or 100). However, site-measured values of organic carbon, soil
porosity, fraction water content, and soil bulk density are substituted into the equation to calculate
screening levels more tailored to site characteristics. If site concentrations do not exceed the Tier 2 SSLs,
then the pathway is excluded from further investigation or concern. The rationale behind this decision
is that, because Tier 2 incorporates site-specific information, the levels are more representative of actual
site conditions than Tier 1. If site concentrations exceed the Tier 2 SSLs, the user has the option of
41
-------
Table 9. Soil Screening Levels for Ground Water Protection:
Organic Chemicals
Inputs
Chemical
a-BHC
Benzene
Benzo(a)pyrene
Carbon tetrachloride
Chlordane
Chlorobenzene
Chloroform
Chrysene
DDT
1 ,4-Dichlorobenzene
1,1-Dichloroethane
1 , 1 -Dichloroethylene
Dieldrin
Ethylbenzene
Methylene chloride
Naphthalene
PCB 1260
Pentachlorophenol8
Tetrachloroethylene
Toluene
1 ,2,4 -Trichlorobenzene
1,1,1-Trichloroethane
Trichloroethylene
Vinyl chloride
Xylenes (mixed)
KOC
(L/kg)
1,820
50
1 ,773,449
230
49,544
173
46
95,153
386,977
511
34
65
13,400
187
23
792
822,422
409
264
129
1,561
128
87
11
234
Target
GW Level
Cw
(mg/L)
1.4E-05
0.005
0.0002
0.005
0.002
0.1
0.1
0.0002
0.0003
0.075
3.65
0.007
5.3E-06
0.7
0.005
1.46
0.0005
0.001
0.005
1
0.07
0.2
0.005
0.002
10
Soil Screening Levels (mg/kg)
Unadjusted
DAF: 1
fj': 0.002
0.0001
0.001
0.71
0.003
0.2
0.05
0.02
0.04
0.23
0.08
0.62
0.002
o.ooor
0.33
o.oor
2.5
0.82
O.OOr
0.003
0.36
0.2
0.07
o.oor
0.0002
5.7
DAF-adjusted
10
0.002
0.007
0.01
7.1
0.03
2.0
0.5
0.2
0.4
2.3
0.8
6.2
0.02
0.007
3.3
0.007
25
8.2
0.009
0.03
3.6
2.3
0.7
0.01
0.002
57
100
0.002
0.1
1.0
71
0.3
20
5
2
4
23
8
62
0.2
o.or
33
0.07
250
82
0.09
0.3
36
23
7
0.1
0.02
570
DAF = Dilution and attenuation factor.
GW = Ground water.
Italics indicate values below quantitation limit for Contract Laboratory Program Routine Analytical
Services.
a pH of 6.8 (may be adjusted using site-specific pH values during Tier 2 analysis).
b Fraction organic carbon (mg carbon/mg soil) may be adjusted using site-specific values during Tier 2
analysis.
42
-------
Table 10. Soil Screening Levels for Ground Water Protection: Inorganic Chemicals
Inputs
Kd (L/Kg)
Chemical
Arsenic
Cadmium
Chromium(VI)
Mercury
Nickel
pH: 4.9
25
2
31
0.1
3
6.8
29
162
19
152
82
8.0
31
2,002
14
211
157
Preliminary
Target
GW
Level
Cw
(mg/L)
0.05
0.005
0.1
0.002
0.1
Soil Screening
Unadjusted
DAF: 1
pH": 4.9
1.2
0.008
3.1
0.0002
0.3
1
6.8
1.4
0.8
1.9
0.3
8.2
1
8.0
1.6
10
1.4
0.42
15.7
10
4.9
12.5
0.08
31
0.002
3.2
10
6.8
14
8
19
3
82
Levels (mg/kg)
DAF-adjusted
10
8.0
15.7
100
13.6
4.2
157
100
4.9
125
0.81
314
0.02
32
100
6.8
140
81
190
30
820
100
8.0
157
1,000
136
42
1,570
a Use of 6.8 unless site pH conditions are known.
DAF = Dilution and attenuation factor (select and use only with consideration of site characteristics).
GW = Ground water.
-------
conducting a Tier 3 or 4 investigation, realizing that increased costs are associated with collecting
additional site data.
The SSLs in this document are based on an f,^ of 0.002 (0.2 percent OC), typical of subsurface soils. If
the actual f,^ of a site's soils is known, the value(s) may be substituted into Equation 2-24. Organic
carbon analyses of soil are inexpensive and relatively easy to conduct (see Nelson and Sommers, 1982).
However, because of variability in f^, values, especially with depth, it is important that the f^ values
obtained be representative of the soils where the contaminants reside. Consider splitting samples taken
for contaminant analysis and reserving the splits for analysis of f^. Alternatively, the f^ samples should
be taken from the same general location and same depth as the contaminant samples. This will help to
ensure that f^ results are representative of the contaminated soils.
If a site's f^ values are below 0.001 (0.1 percent OC), chemical sorption to soil mineral matter becomes
significant and Equation 2-24 should be used with caution. Use of this equation under such conditions
will tend to underpredict sorption and hence overpredict contaminant concentrations in soil pore water (i.e.,
it is overly conservative). Unfortunately there is not a simple relationship between mineral matter and
sorption as there is between sorption and soil organics. However, Piwoni and Banerjee (1989) developed
an empirical correlation that can be used to estimate Kd values for hydrophobic organic chemicals from
Kow under such conditions (see Equation 2-26, Section 2.4.3). The Kd values estimated using Equation
2-26 should be used in Equation 2-22.
The Tier 3 investigation involves conducting a specific leach test, the Synthetic Precipitation Leaching
Procedure (SPLP) (Appendix A; U.S. EPA, 1992f). If the leach test results divided by the DAF of 10
exceed the acceptable ground water limit (e.g., nonzero MCLG, MCL, 10"6 risk-based values), then further
investigation would be warranted. The SPLP may not be applicable to all contaminated soils (e.g., oily
types of waste do not yield suitable results). Therefore the user is advised to use discretion when applying
the SPLP. [Additional guidance on the use and limitations of the SPLP will be provided in a future draft
of this document].
Tier 4 represents the highest level of site-specificity in evaluating the migration to ground water pathway.
In this investigation, site-specific data are collected and used in a fate and transport model to confirm the
threat to ground water and further determine site-specific cleanup goals as would typically be done for the
remedial investigation/feasibility study (RI/FS). A fixed, generic DAF is not used in this tier because the
model would account for fate and transport mechanisms in the subsurface. The advantage of this approach
is that it accounts for site hydrogeologic, climatologic, and contaminant source characteristics and may
result in fully protective but less stringent remediation goals. However, the additional cost of collecting
the data required to apply the model should be factored into the decision to conduct a Tier 4 investigation.
Site-specific ground water models should be used to develop Tier 4 levels by backcalculating soil leachate
concentrations from the target ground water concentrations. The DAF is the soil leachate concentration
divided by the target ground water concentration. For example, if the modeling estimates that a
concentration of 10 mg/L at the receptor well results from a soil leachate concentration of 100 mg/L, the
DAF is 10.
Ground water models should be selected and used only by professionals trained and experienced in
subsurface science. These professionals should be familiar with the application and use of the particular
model to be used. The selection of a ground water model should reflect the amount of information that
is available about a site's subsurface conditions. For example, a simple analytical model is usually all
that is appropriate at the early stages of a site assessment. Separate model runs will be necessary to
develop a DAF for each chemical of concern at a site. The Koc values in this document are appropriate
44
-------
for use in the modeling effort, and the soil/water partition equation can be used to estimate soil leachate
concentrations for model inputs.
[An evaluation of 10 fate and transport models for potential use in the Tier 4 evaluation will be included
in a future guidance document].
45
-------
Technical Background Document for
Draft Soil Screening Level Guidance
Part 3: REFERENCES
Abdul, A.S., T.L. Gibson, and D.N. Rai. 1987. Statistical correlations for predicting the partition
coefficient for nonpolar organic contaminants between aquifer carbon and water. Hazardous
Waste and Hazardous Materials 4(3):211-223.
Calabrese, E.J., H. Pastides, R. Barnes, et al. 1989. How much soil do young children ingest: an
epidemiologic study. In: Petroleum Contaminated Soils, Vol. 2. E.J. Calabrese and P.T.
Kostecki, eds. pp. 363-417. Chelsea, MI, Lewis Publishers.
Carsel, R.F., R.S. Parrish, R.L. Jones, J.L. Hansen, and R.L. Lamb. 1988. Characterizing the uncertainty
of pesticide leaching in agricultural soils. /. of Contam. Hyd. 2:111-124.
Curtis, G. P., P. V. Roberts, and M. Reinhard. 1986. A natural gradient experiment on solute transport
in a sand aquifer. Water Resources Research 22(13):2059-2067.
Davis, S., P. Waller, R. Buschom, J. Ballou, and P. White. 1990. Quantitative estimates of soil ingestion
in normal children between the ages of 2 and 7 years: population-based estimates using Al, Si,
and Ti as soil tracer elements. Archives of Environmental Health, 45:112-122.
Dragun, J. 1988. The Soil Chemistry of Hazardous Materials. Hazardous Materials Control Research
Institute, Silver Spring, MD.
EQM (Environmental Quality Management). 1992. Limited Validation of the Hwang and Falco Model
for Emissions of Soil-Incorporated Volatile Organic Compounds. Contract No. 68-D8-0111.
Prepared for Office of Solid Waste and Emergency Response, U.S. Environmental Protection
Agency. Washington, DC.
EQM (Environmental Quality Management). 1993. Evaluation of the Dispersion Equations in the Risk
Assessment Guidance for Superfund (RAGS): Volume I - Human Health Evaluation Manual (Part
B, Development of Preliminary Remediation Goals). Contract No. 68-02-D 120. Prepared for
Office of Emergency and Remedial Response, U.S. Environmental Protection Agency.
Washington, DC.
Farmer, W. J., and J. Letey. 1974. Volatilization Losses of Pesticides from Soils. EPA-660/2-74/054.
Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC.
Farmer, W. J., M. S. Yang, J. Letey, and W. F. Spencer. 1980. Land Disposal Of Hexachlorobenzene
Wastes. EPA-600/2-80/119. Office of Research and Development, U.S. Environmental Protection
Agency, Washington, DC.
Feenstra, S., D.M. MacKay, and J.A. Cherry. 1991. A method for assessing residual NAPL based on
organic chemical concentrations in soil samples. Ground water Monitoring Rev. 11(2): 128-136.
57 FR 21450. May 20, 1992. Hazardous Waste Management System: Identification and Listing of
Hazardous Waste. Proposed Rule.
46
-------
53 FR 28692, August 1, 1988.
Hartley, G. S. 1964. Herbicide behavior in the soil. In: The Physiology and Biochemistry of Herbicides,
L. J. Aldus, Ed. Academic Press, New York.
Hwang, S. T., and J. W. Falco. 1986. Estimation of Multimedia Exposure Related to Hazardous Waste
Facilities. Y. Cohen, Ed. Plenum Publishing.
IRIS (Integrated Risk Information System). 1993. U.S. Environmental Protection Agency, Duluth, MN.
Karickhoff, S.W. 1984. Organic pollutant sorption in aquatic systems. /. Hydraul. Eng. 110(6):707-735.
Lee, L. S., P. S. C. Rao, P. Nkedi-Kizza, and J. J. Delfino. 1990. Influence of solvent and sorbent
characteristics on distribution of pentachlorophenol in octanol-water and soil-water systems.
Environ. Sci. Tech. 24:654-661.
Lee, L. S., P. S. C. Rao, and M. L. Brusseau. 1991. Nonequilibrium sorption and transport of neutral
and ionized compounds. Environ. Sci. Tech. 25:722-729.
Lyman, W. J., W. F. Reehl, and D. H. Rosenblatt. 1982. Handbook of Chemical Property Estimation
Methods. McGraw-Hill, New York.
McCarty, P.L., M. Reinhard, and B.E. Rittmann. 1981. Trace organics in ground water. Environ. Sci.
Technol. 15:40-51.
Millington, R. J., and J. M. Quirk. 1961. Permeability of porous soils. Trans. Faraday Soc. 57:1200-
1207.
Nelson, D. W., and L. E. Sommers. 1982. Total Carbon, Organic Carbon, and Organic Matter. Chapter
29 in: Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties, 2nd Edition,
Page, A.L., R. H. Miller, and D. R. Keeney (eds.). American Society of Agronomy, Soil Science
Society of America, Madison, WI.
Piwoni, M. D., and P. Banerjee. 1989. Sorption of volatile organic solvents from aqueous solution onto
subsurface solids. /. Contam. Hydrol. 4(2): 163-179.
Radian. 1989. Short-term Fate and Persistence of Motor Fuels in Soils. Report to the American
Petroleum Institute, Washington, DC.
Schwarzenbach, R. P., and J. C. Westall. 1981. Transport of non-polar organic compounds from surface
water to ground water. Environ. Sci. Tech. 15(11): 1360-1367.
U.S. EPA (Environmental Protection Agency). 1985a. Rapid Assessment of Exposure to Paniculate
Emissions from Surface Contamination Sites. EPA/600/8-85/002. Office of Health and
Environmental Assessment, Washington, DC.
U.S. EPA (Environmental Protection Agency). 1985b. Water Quality Assessment: A Screening
Procedure for Toxic and Conventional Pollutants. WPA/600/6-85/002b.
47
-------
U.S. EPA (Environmental Protection Agency). 1987. Data Quality Objectives for Remedial Response
Activities: Development Process (Volume 1). EPA/540/G-87/003. 9355.0-07B. Office of
Emergency and Remedial Response. PB88-131370/CCE. NTIS. Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1989a. Methods for Evaluating the Attainment of Soil
Cleanup Standards. Volume 1. EPA 230/02-89-042. Office of Policy, Planning and Evaluation.
Washington, DC.
U.S. EPA (Environmental Protection Agency). 1989b. Risk Assessment Guidance for Superfund, Volume
1: Human Health Evaluation Manual, Part A, Interim Final. EPA/540/1-89/002. Publication
9285.7-01B. Office of Emergency and Remedial Response. PB90-155581/CCE. NTIS.
Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1990. Guidance on Remedial Actions for Superfund Sites
with PCB Contamination. EPA 540/G-90/007. Office of Emergency and Remedial Response,
Washington, DC. PB91-921206/CCE. NTIS. Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1991. Risk Assessment Guidance for Superfund: Volume
1 - Human Health Evaluation Manual (Part B, Development of Risk-Based Preliminary
Remediation Goals). Interim. Office of Emergency and Remedial Response. PB92-963333.
NTIS, Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1992a. Background Document for Finite Source
Methodology for Wastes Containing Metals. Office of Solid Waste, Washington, DC.
U.S. EPA (Environmental Protection Agency). 1992b. Drinking Water Regulations and Advisories.
Office of Water, Washington, DC.
U.S. EPA (Environmental Protection Agency). 1992c. Estimating the Potential for Occurrence of
DNAPL at Superfund Sites. Pub. 9355.4-07FS. Office of Solid Waste and Emergency
Response/R.S. Kerr Environmental Research Laboratory, Ada, OK. PB92-963358. NTIS,
Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1992d. Guidance for Data Useability in Risk Assessment
(Part A). Office of Emergency and Remedial Response. PB92-963356. NTIS, Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1992e. Guidance for Data Useability in Risk Assessment
(PartB). Office of Emergency and Remedial Response. PB92-963362. NTIS, Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1992f. Synthetic precipitation leaching procedure (SPLP),
Method 1312. In: Test Methods for Evaluating Solid Waste, Physical/Chemical Methods. EPA
Publication SW-846. Third Edition (September 1986), as amended by Update I (July).
U.S. EPA (Environmental Protection Agency). 1993a. Draft Soil Screening Level Guidance. Quick
Reference Fact Sheet. Office of Emergency and Remedial Response. PB93-963508. NTIS,
Springfield, VA.
U.S. EPA (Environmental Protection Agency). 1993b. Science Advisory Board Review of the Office of
Solid Waste and Emergency Response draft Risk Assessment Guidance for Superfund (RAGS),
Human Health Evaluation Manual (HHEM). EPA-SAB-EHC-93-007.
48
-------
U.S. EPA (Environmental Protection Agency). 1993c. Background Document for EPA's Composite
Model for Leachate Migration with Transformation Products, EPACMTP. Office of Solid Waste,
Washington, DC. July.
Van Wijnen, J.H., P. Clausing, and B. Brunekreef. 1990. Estimated soil ingestion by children.
Environmental Research, 51:147-162.
49
-------
-------
APPENDIX A
SYNTHETIC PRECIPITATION LEACHING PROCEDURE (SPLP)
(SW-846 METHOD 1312)
-------
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE
1.0 SCOPE AND APPLICATION
1.1 Method 1312 is designed to determine the mobility of both organic
and inorganic analytes present in samples of soils and wastes.
2.0 SUMMARY OF METHOD
2.1 For liquid samples (i.e.. those containing less than 0.5 % dry
solid material), the sample, after filtration through a 0.6 to 0.8 pm glass
fiber filter, is defined as the 1312 extract.
2.2 For samples containing greater than 0.5 % solids, the liquid phase,
if any, is separated from the solid phase and stored for later analysis; the
particle size of the solid phase is reduced, if necessary. The solid phase is
extracted with an amount of extraction fluid equal to 20 times the weight of the
solid phase. The extraction fluid employed is a function of the region of the
country where the sample site is located if the sample is a soil. If the sample
is a waste or wastewater, the extraction fluid employed is a pH 4.2 solution.
A special extractor vessel is used when testing for volatile analytes (see Table
1 for a list of volatile compounds). Following extraction, the liquid extract
is separated from the sample by 0.6 to 0.8 urn glass fiber filter.
2.3 If compatible (i.e.. multiple phases will not form on combination),
the initial liquid phase of the waste is added to the liquid extract, and these
are analyzed together. If incompatible, the liquids are analyzed separately and
the results are mathematically combined to yield a volume-weighted average
concentration.
3.0 INTERFERENCES
3.1 Potential interferences that may be encountered during analysis are
discussed in the individual analytical methods.
4.0 APPARATUS AND MATERIALS
4.1 Agitation apparatus: The agitation apparatus must be capable of
rotating the extraction vessel in an end-over-end fashion (see Figure 1) at 30
± 2 rpm. Suitable devices known to EPA are identified in Table 2.
4.2 Extraction Vessels
4.2.1 Zero Headspace Extraction Vessel (ZHE). This device is for
use only when the sample is being tested for the mobility of volatile
analytes (i.e.. those listed in Table 1). The ZHE (depicted in Figure 2)
allows for liquid/solid separation within the device and effectively
precludes headspace. This type of vessel allows for initial liquid/solid
separation, extraction, and final extract filtration without opening the
vessel (see Step 4.3.1). These vessels shall have an internal volume of
1312 - 1 Revision 0
November 1992
-------
500-600 ml and be equipped to accommodate a 90-110 mm filter. The devices
contain VITON* 0-rlngs which should be replaced frequently. Suitable ZHE
devices known to EPA are identified in Table 3.
For the ZHE to be acceptable for use, the piston within the ZHE
should be able to be moved with approximately 15 psi or less. If it takes
more pressure to move the piston, the 0-rings in the device should be
replaced. If this does not solve the problem, the ZHE is unacceptable for
1312 analyses and the manufacturer should be contacted.
The ZHE should be checked for leaks after every extraction. If the
device contains a built-in pressure gauge, pressurize the device to 50
psi, allow it to stand unattended for 1 hour, and recheck the pressure.
If the device does not have a built-in pressure gauge, pressurize the
device to 50 psi, submerge it in water, and check for the presence of air
bubbles escaping from any of the fittings. If pressure is lost, check all
fittings and inspect and replace 0-rings, if necessary. Retest the
device. If leakage problems cannot be solved, the manufacturer should be
contacted.
Some ZHEs use gas pressure to actuate the ZHE piston, while others
use mechanical pressure (see Table 3). Whereas the volatiles procedure
(see Step 7.3) refers to pounds-per-square-inch (psi), for the
mechanically actuated piston, the pressure applied is measured in torque-
inch-pounds. Refer to the manufacturer's instructions as to the proper
conversion.
4.2.2 Bottle Extraction Vessel. When the sample is being
evaluated using the nonvolatile extraction, a jar with sufficient capacity
to hold the sample and the extraction fluid is needed. Headspace is
allowed in this vessel.
The extraction bottles may be constructed from various materials,
depending on the analytes to be analyzed and the nature of the waste (see
Step 4.3.3). It is recommended that borosilicate glass bottles be used
instead of other types of glass, especially when inorganics are of
concern. Plastic bottles, other than polytetrafluoroethylene, shall not
be used if organics are to be investigated. Bottles are available from a
number of laboratory suppliers. When this type of extraction vessel is
used, the filtration device discussed in Step 4.3.2 is used for initial
liquid/solid separation and final extract filtration.
4.3 Filtration Devices: It is recommended that all filtrations be
performed in a hood.
4.3.1 Zero-Headspace Extraction Vessel (ZHE): When the sample
is evaluated for volatiles, the zero-headspace extraction vessel described
in Step 4.2.1 is used for filtration. The device shall be capable of
supporting and keeping in place the glass fiber filter and be able to
withstand the pressure needed to accomplish separation (50 psi).
'VITON® is a trademark of Du Pont.
1312 - 2 Revision 0
November 1992
-------
NOTE: When it 1s suspected that the glass fiber filter has been ruptured, an
In-line glass fiber filter may be used to filter the material within the
ZHE.
4.3.2 Filter Holder: When the sample is evaluated for other than
volatile analytes, a filter holder capable of supporting a glass fiber
filter and able to withstand the pressure needed to accomplish separation
may be used. Suitable filter holders range from simple vacuum units to
relatively complex systems capable of exerting pressures of up to 50 psi
or more. The type of filter holder used depends on the properties of the
material to be filtered (see Step 4.3.3). These devices shall have a
minimum internal volume of 300 ml and be equipped to accommodate a minimum
filter size of 47 mm (filter holders having an internal capacity of 1.5 L
or greater, and equipped to accommodate a 142 mm diameter filter, are
recommended). Vacuum filtration can only be used for wastes with low
solids content (<10 %) and for highly granular, liquid-containing wastes.
All other types of wastes should be filtered using positive pressure
filtration. Suitable filter holders known to EPA are shown in Table 4.
4.3.3 Materials of Construction: Extraction vessels and
filtration devices shall be made of inert materials which will not leach
or absorb sample components. Glass, polytetrafluoroethylene (PTFE), or
type 316 stainless steel equipment may be used when evaluating the
mobility of both organic and inorganic components. Devices made of high-
density polyethylene (HOPE), polypropylene (PP), or polyvinyl chloride
(PVC) may be used only when evaluating the mobility of metals.
Borosilicate glass bottles are recommended for use over other types of
glass bottles, especially when inorganics are analytes of concern.
4.4 Filters: Filters shall be made of borosilicate glass fiber, shall
contain no binder materials, and shall have an effective pore size of 0.6 to
0.8-/im or equivalent. Filters known to EPA which meet these specifications are
identified in Table 5. Pre-filters must not be used. When evaluating the
mobility of metals, filters shall be acid-washed prior to use by rinsing with IN
nitric acid followed by three consecutive rinses with deionized distilled water
(a minimum of 1-L per rinse is recommended). Glass fiber filters are fragile and
should be handled with care.
4.5 pH Meters: The meter should be accurate to ± 0.05 units at 25'C.
4.6 ZHE Extract Collection Devices: TEDLAR*2 bags or glass, stainless
steel or PTFE gas-tight syringes are used to collect the initial liquid phase and
the final extract when using the ZHE device. These devices listed are
recommended for use under the following conditions:
4.6.1 If a waste contains an aqueous liquid phase or if a waste
does not contain a significant amount of nonaqueous liquid (i.e.. <1 % of
total waste), the TEDLAR* bag or a 600 ml syringe should be used to collect
and combine the initial liquid and solid extract.
2TEDLAR* is a registered trademark of Du Pont.
1312 - 3 Revision 0
November 1992
-------
4.6.2 If a waste contains a significant amount of nonaqueous
liquid in the Initial liquid phase (i.e.. >1 % of total waste), the
syringe or the TEDLAR bag may be used for both the initial solid/liquid
separation and the final extract filtration. However, analysts should use
one or the other, not both.
4.6.3 If the waste contains no Initial liquid phase (1s 100 X
solidj^or has no significant solid phase (is 100 % liquid), either the
TEDLAR bag or the syringe may be used. If the syringe is used, discard
the first 5 mi of liquid expressed from the device. The remaining
aliquots are used for analysis.
4.7 ZHE Extraction Fluid Transfer Devices: Any device capable of
transferring the extraction fluid into the ZHE without changing the nature of the
extraction fluid is acceptable (e.g.. a positive displacement or peristaltic
pump, a gas-tight syringe, pressure filtration unit (see Step 4.3.2), or other
ZHE device).
4.8 Laboratory Balance: Any laboratory balance accurate to within ±
0.01 grams may be used (all weight measurements are to be within + 0.1 grams).
4.9 Beaker or Erlenmeyer flask, glass, 500 ml.
4.10 Watchglass, appropriate diameter to cover beaker or Erlenmeyer
flask.
4.11 Magnetic stirrer.
5.0 REAGENTS
5.1 Reagent grade chemicals shall be used in all tests. Unless
otherwise indicated, it is intended that all reagents shall conform to the
specifications of the Committee on Analytical Reagents of the American Chemical
Society, where such specifications are available. Other grades may be used,
provided it is first ascertained that the reagent is of sufficiently high purity
to permit its use without lessening the accuracy of the determination.
5.2 Reagent Water. Reagent water is defined as water in which an
interferant is not observed at or above the method's detection limit of the
analyte(s) of interest. For nonvolatile extractions, ASTM Type II water or
equivalent meets the definition of reagent water. For volatile extractions, it
1s recommended that reagent water be generated by any of the following methods.
Reagent water should be monitored periodically for impurities.
5.2.1 Reagent water for volatile extractions may be generated
by passing tap water through a carbon filter bed containing about 500
grams of activated carbon (Calgon Corp., Filtrasorb-300 or equivalent).
5.2.2 A water purification system (Millipore Super-Q or
equivalent) may also be used to generate reagent water for volatile
extractions.
1312 - 4 Revision 0
November 1992
-------
5.2.3 Reagent water for volatile extractions may also be prepared
by boiling water for 15 minutes. Subsequently, while maintaining the
water temperature at 90 ± 5 degrees C, bubble a contaminant-free inert gas
(e.g. nitrogen) through the water for 1 hour. While still hot, transfer
the water to a narrow mouth screw-cap bottle under zero-headspace and seal
with a Teflon-lined septum and cap.
5.3 Sulfuric acid/nitric acid (60/40 weight percent mixture) H2S04/HN03.
Cautiously mix 60 g of concentrated sulfuric acid with 40 g of concentrated
nitric acid.
5.4 Extraction fluids.
5.4.1 Extraction fluid #1: This fluid is made by adding the
60/40 weight percent mixture of sulfuric and nitric acids to reagent water
(Step 5.2) until the pH is 4.20 + 0.05. The fluid is used to determine
the Teachability of soil from a site that is east of the Mississippi
River, and the Teachability of wastes and wastewaters.
NOTE: Solutions are unbuffered and exact pH may not be attained.
5.4.2 Extraction fluid #2: This fluid is made by adding the
60/40 weight percent mixture of sulfuric and nitric acids to reagent water
(Step 5.2) until the pH is 5.00 ± 0.05. The fluid is used to determine
the leachability of soil from a site that is west of the Mississippi
River.
5.4.3 Extraction fluid #3: This fluid is reagent water (Step
5.2) and is used to determine cyanide and volatiles leachability.
NOTE: These extraction fluids should be monitored frequently for impurities.
The pH should be checked prior to use to ensure that these fluids are made
up accurately. If impurities are found or the pH is not within the above
specifications, the fluid shall be discarded and fresh extraction fluid
prepared.
5.5 Analytical standards shall be prepared according to the appropriate
analytical method.
6.0 SAMPLE COLLECTION, PRESERVATION, AND HANDLING
6.1 All samples shall be collected using an appropriate sampling plan.
6.2 There may be requirements on the minimal size of the field sample
depending upon the physical state or states of the waste and the analytes of
concern. An aliquot is needed for the preliminary evaluations of the percent
solids and the particle size. An aliquot may be needed to conduct the
nonvolatile analyte extraction procedure (see Step 1.4 concerning the use of this
extract for volatile organics). If volatile organics are of concern, another
aliquot may be needed. Quality control measures may require additional aliquots.
Further, it is always wise to collect more sample just in case something goes
wrong with the initial attempt to conduct the test.
1312 - 5 Revision 0
November 1992
-------
6.3 Preservatives shall not be added to samples before extraction.
6.4 Samples may be refrigerated unless refrigeration results in
irreversible physical change to the waste. If precipitation occurs, the entire
sample (including precipitate) should be extracted.
6.5 When the sample is to be evaluated for volatile analytes, care
shall be taken to minimize the loss of volatiles. Samples shall be collected and
stored in a manner intended to prevent the loss of volatile analytes (e.g..
samples should be collected in Teflon-lined septum capped vials and stored at
4*C. Samples should be opened only immediately prior to extraction).
6.6 1312 extracts should be prepared for analysis and analyzed as soon
as possible following extraction. Extracts or portions of extracts for metallic
analyte determinations must be acidified with nitric acid to a pH < 2, unless
precipitation occurs (see Step 7.2.14 if precipitation occurs). Extracts should
be preserved for other analytes according to the guidance given in the individual
analysis methods. Extracts or portions of extracts for organic analyte
determinations shall not be allowed to come into contact with the atmosphere
(i.e.. no headspace) to prevent losses. See Section 8.0 (Quality Control) for
acceptable sample and extract holding times.
7.0 PROCEDURE
7.1 Preliminary Evaluations
Perform preliminary 1312 evaluations on a minimum 100 gram aliquot of
sample. This aliquot may not actually undergo 1312 extraction. These
preliminary evaluations include: (1) determination of the percent solids (Step
7.1.1); (2) determination of whether the waste contains insignificant solids and
is, therefore, its own extract after filtration (Step 7.1.2); and (3)
determination of whether the solid portion of the waste requires particle size
reduction (Section 7.1.3).
7.1.1 Preliminary determination of percent solids: Percent
solids is defined as that fraction of a waste sample (as a percentage of
the total sample) from which no liquid may be forced out by an applied
pressure, as described below.
7.1.1.1 If the sample will obviously yield no free
liquid when subjected to pressure filtration (i.e.. is 100%
solids), weigh out a representative subsample (100 g minimum) and
proceed to Step 7.1.3.
7.1.1.2 If the sample is liquid or multiphasic,
liquid/solid separation to make a preliminary determination of
percent solids is required. This involves the filtration device
discussed in Step 4.3.2, and is outlined in Steps 7.1.1.3 through
7.1.1.9.
7.1.1.3 Pre-weigh the filter and the container that will
receive the filtrate.
1312 - 6 Revision 0
November 1992
-------
7.1.1.4 Assemble filter holder and filter following the
manufacturer's instructions. Place the filter on the support
screen and secure.
7.1.1.5 Weigh out a subsample of the waste (100 gram
minimum) and record the weight.
7.1.1.6 Allow slurries to stand to permit the solid phase
to settle. Samples that settle slowly may be centrifuged prior to
filtration. Centrifugation is to be used only as an aid to
filtration. If used, the liquid should be decanted and filtered
followed by filtration of the solid portion of the waste through
the same filtration system.
7.1.1.7 Quantitatively transfer the sample to the filter
holder (liquid and solid phases). Spread the sample evenly over
the surface of the filter. If filtration of the waste at 4*C
reduces the amount of expressed liquid over what would be expressed
at room temperature, then allow the sample to warm up to room
temperature in the device before filtering.
NOTE: If sample material (>1 % of original sample weight) has obviously adhered
to the container used to transfer the sample to the filtration apparatus,
determine the weight of this residue and subtract it from the sample
weight determined in Step 7.1.1.5 to determine the weight of the sample
that will be filtered.
Gradually apply vacuum or gentle pressure of 1-10 psi, until air
or pressurizing gas moves through the filter. If this point is not
reached under 10 psi, and if no additional liquid has passed through the
filter in any 2-minute interval, slowly increase the pressure in 10 psi
increments to a maximum of 50 psi. After each incremental increase of 10
psi, if the pressurizing gas has not moved through the filter, and if no
additional liquid has passed through the filter in any 2-minute interval,
proceed to the next 10-psi increment. When the pressurizing gas begins to
move through the filter, or when liquid flow has ceased at 50 psi (i.e..
filtration does not result in any additional filtrate within any 2-minute
period), stop the filtration.
NOTE: Instantaneous application of high pressure can degrade the glass fiber
filter and may cause premature plugging.
7.1.1.8 The material in the filter holder is defined as
the solid phase of the sample, and the filtrate is defined as the
liquid phase.
NOTE: Some samples, such as oily wastes and some paint wastes, will obviously
contain some material that appears to be a liquid, but even after applying
vacuum or pressure filtration, as outlined in Step 7.1.1.7, this material
may not filter. If this is the case, the material within the filtration
device is defined as a solid. Do not replace the original filter with a
fresh filter under any circumstances. Use only one filter.
1312 - 7 Revision 0
November 1992
-------
7.1.1.9 Determine the weight of the liquid phase by
subtracting the weight of the filtrate container (see Step 7.1.1.3)
from the total weight of the filtrate-filled container. Determine
the weight of the solid phase of the sample by subtracting the
weight of the liquid phase from the weight of the total sample, as
determined in Step 7.1.1.5 or 7.1.1.7.
Record the weight of the liquid and solid phases.
Calculate the percent solids as follows:
Weight of solid (Step 7.1.1.9)
Percent solids - x 100
Total weight of waste (Step 7.1.1.5 or 7.1.1.7)
7.1.2 If the percent solids determined in Step 7.1.1.9 is equal
to or greater than 0.5%, then proceed either to Step 7.1.3 to determine
whether the solid material requires particle size reduction or to Step
7.1.2.1 if it is noticed that a small amount of the filtrate is entrained
in wetting of the filter. If the percent solids determined in Step
7.1.1.9 is less than 0.5%, then proceed to Step 7.2.9 if the nonvolatile
1312 analysis is to be performed, and to Section 7.3 with a fresh portion
of the waste if the volatile 1312 analysis is to be performed.
7.1.2.1 Remove the solid phase and filter from the
filtration apparatus.
7.1.2.2 Dry the filter and solid phase at 100 ± 20'C
until two successive weighings yield the same value within ± 1 %.
Record the final weight.
Note: Caution should be taken to ensure that the subject solid will not flash
upon heating. It is recommended that the drying oven be vented to a hood
or other appropriate device.
7.1.2.3 Calculate the percent dry solids as follows:
Percent (Weight of dry sample + filter) - tared weight of filter
dry solids - x 100
Initial weight of sample (Step 7.1.1.5 or 7.1.1.7)
7.1.2.4 If the percent dry solids is less than 0.5%,
then proceed to Step 7.2.9 if the nonvolatile 1312 analysis is to
be performed, and to Step 7.3 if the volatile 1312 analysis is to
be performed. If the percent dry solids is greater than or equal
to 0.5%, and if the nonvolatile 1312 analysis is to be performed,
return to the beginning of this Section (7.1) and, with a fresh
portion of sample, determine whether particle size reduction is
necessary (Step 7.1.3).
1312 - 8 Revision 0
November 1992
-------
7.1.3 Determination of whether the sample requires particle-size
reduction (particle-size Is reduced during this step): Using the solid
portion of the sample, evaluate the solid for particle size. Particle-
size reduction is required, unless the solid has a surface area per gram
of material equal to or greater than 3.1 cm, or is smaller than 1 cm in
its narrowest dimension (i.e.. is capable of passing through a 9.5 mm
(0.375 inch) standard sieve). If the surface area is smaller or the
particle size larger than described above, prepare the solid portion of
the sample for extraction by crushing, cutting, or grinding the waste to
a surface area or particle size as described above. If the solids are
prepared for organic volatiles extraction, special precautions must be
taken (see Step 7.3.6).
Note: Surface area criteria are meant for filamentous (e.g.. paper, cloth, and
similar) waste materials. Actual measurement of surface area is not
required, nor is it recommended. For materials that do not obviously meet
the criteria, sample-specific methods would need to be developed and
employed to measure the surface area. Such methodology is currently not
available.
7.1.4 Determination of appropriate extraction fluid:
7.1.4.1 For soils, if the sample is from a site that is
east of the Mississippi River, extraction fluid II should be used.
If the sample is from a site that is west of the Mississippi River,
extraction fluid 12 should be used.
7.1.4.2 For wastes and wastewater, extraction fluid #1
should be used.
7.1.4.3 For cyanide-containing wastes and/or soils,
extraction fluid 13 (reagent water) must be used because leaching
of cyanide-containing samples under acidic conditions may result
in the formation of hydrogen cyanide gas.
7.1.5 If the aliquot of the sample used for the preliminary
evaluation (Steps 7.1.1 - 7.1.4) was determined to be 100% solid at Step
7.1.1.1, then it can be used for the Section 7.2 extraction (assuming at
least 100 grams remain), and the Section 7.3 extraction (assuming at least
25 grams remain). If the aliquot was subjected to the procedure in Step
7.1.1.7, then another aliquot shall be used for the volatile extraction
procedure in Section 7.3. The aliquot of the waste subjected to the
procedure in Step 7.1.1.7 might be appropriate for use for the Section 7.2
extraction if an adequate amount of solid (as determined by Step 7.1.1.9)
was obtained. The amount of solid necessary is dependent upon whether a
sufficient amount of extract will be produced to support the analyses. If
an adequate amount of solid remains, proceed to Step 7.2.10 of the
nonvolatile 1312 extraction.
7.2 Procedure when Volatiles are not Involved
A minimum sample size of 100 grams (solid and liquid phases) is
recommended. In some cases, a larger sample size may be appropriate, depending
on the solids content of the waste sample (percent solids, See Step 7.1.1),
1312 - 9 Revision 0
November 1992
-------
whether the Initial liquid phase of the waste will be mlsdble with the aqueous
extract of the solid, and whether Inorganics, semi volatile organics, pesticides,
and herbicides are all analytes of concern. Enough solids should be generated
for extraction such that the volume of 1312 extract will be sufficient to support
all of the analyses required. If the amount of extract generated by a single
1312 extraction will not be sufficient to perform all of the analyses, more than
one extraction may be performed and the extracts from each combined and allquoted
for analysis.
7.2.1 If the sample will obviously yield no liquid when subjected
to pressure filtration (I.e.. is 100 % solid, see Step 7.1.1), weigh out
a subsample of the sample (100 gram minimum) and proceed to Step 7.2.9.
7.2.2 If the sample 1s liquid or multlphaslc, liquid/solid
separation 1s required. This Involves the filtration device described 1n
Step 4.3.2 and 1s outlined in Steps 7.2.3 to 7.2.8.
7.2.3 Pre-welgh the container that will receive the filtrate.
7.2.4 Assemble the filter holder and filter following the
manufacturer's instructions. Place the filter on the support screen and
secure. Acid wash the filter if evaluating the mobility of metals (see
Step 4.4).
Note: Acid washed filters may be used for all nonvolatile extractions even when
metals are not of concern.
7.2.5 Weigh out a subsample of the sample (100 gram minimum) and
record the weight. If the waste contains <0.5 % dry solids (Step 7.1.2),
the liquid portion of the waste, after filtration, is defined as the 1312
extract. Therefore, enough of the sample should be filtered so that the
amount of filtered liquid will support all of the analyses required of the
1312 extract. For wastes containing >0.5 % dry solids (Steps 7.1.1 or
7.1.2), use the percent solids information obtained in Step 7.1.1 to
determine the optimum sample size (100 gram minimum) for filtration.
Enough solids should be generated by filtration to support the analyses to
be performed on the 1312 extract.
7.2.6 Allow slurries to stand to permit the solid phase to settle.
Samples that settle slowly may be centrifuged prior to filtration. Use
centrifugation only as an aid to filtration. If the sample is
centrifuged, the liquid should be decanted and filtered followed by
filtration of the solid portion of the waste through the same filtration
system.
7.2.7 Quantitatively transfer the sample (liquid and solid phases)
to the filter holder (see Step 4.3.2). Spread the waste sample evenly
over the surface of the filter. If filtration of the waste at 4*C reduces
the amount of expressed liquid over what would be expressed at room
temperature, then allow the sample to warm up to room temperature in the
device before filtering.
1312 - 10 Revision 0
November 1992
-------
NOTE: If waste material (>1 % of the original sample weight) has obviously
adhered to the container used to transfer the sample to the filtration
apparatus, determine the weight of this residue and subtract it from the
sample weight determined in Step 7.2.5, to determine the weight of the
waste sample that will be filtered.
Gradually apply vacuum or gentle pressure of 1-10 psi, until air
or pressurizing gas moves through the filter. If this point if not
reached under 10 psi, and if no additional liquid has passed through the
filter in any 2-minute interval, slowly increase the pressure in 10-psi
increments to maximum of 50 psi. After each incremental increase of 10
psi, if the pressurizing gas has not moved through the filter, and if no
additional liquid has passed through the filter in any 2-minute interval,
proceed to the next 10-psi increment. When the pressurizing gas begins to
move through the filter, or when the liquid flow has ceased at 50 psi
(i.e.. filtration does not result in any additional filtrate within a
2-minute period), stop the filtration.
NOTE: Instantaneous application of high pressure can degrade the glass fiber
filter and may cause premature plugging.
7.2.8 The material in the filter holder is defined as the solid
phase of the sample, and the filtrate is defined as the liquid phase.
Weigh the filtrate. The liquid phase may now be either analyzed (see
Steps 7.2.12) or stored at 4'C yntil time of analysis.
NOTE: Some wastes, such as oily wastes and some paint wastes, will obviously
contain some material which appears to be a liquid. Even after applying
vacuum or pressure filtration, as outlined in Step 7.2.7, this material
may not filter. If this is the case, the material within the filtration
device is defined as a solid, and is carried through the extraction as a
solid. Do not replace the original filter with a fresh filter under any
circumstances. Use only one filter.
7.2.9 If the sample contains <0.5% dry solids (see Step 7.1.2),
proceed to Step 7.2.13. If the sample contains >0.5 % dry solids (see
Step 7.1.1 or 7.1.2), and if particle-size reduction of the solid was
needed in Step 7.1.3, proceed to Step 7.2.10. If the sample as received
passes a 9.5 mm sieve, quantitatively transfer the solid material into the
extractor bottle along with the filter used to separate the initial liquid
from the solid phase, and proceed to Step 7.2.11.
7.2.10 Prepare the solid portion of the sample for extraction by
crushing, cutting, or grinding the waste to a surface area or particle-
size as described in Step 7.1.3. When the surface area or particle-size
has been appropriately altered, quantitatively transfer the solid material
into an extractor bottle. Include the filter used to separate the initial
liquid from the solid phase.
NOTE: Sieving of the waste is not normally required. Surface area requirements
are meant for filamentous (e.g.. paper, cloth) and similar waste
materials. Actual measurement of surface area is not recommended. If
1312 - 11 Revision 0
November 1992
-------
sieving is necessary, a Teflon-coated sieve should be used to avoid
contamination of the sample.
7.2.11 Determine the amount of extraction fluid to add to the
extractor vessel as follows:
20 x % solids (Step 7.1.1) x weight of waste
filtered (Step 7.2.5 or 7.2.7)
Weight of -
extraction fluid
100
Slowly add this amount of appropriate extraction fluid (see Step
7.1.4) to the extractor vessel. Close the extractor bottle tightly (it is
recommended that Teflon tape be used to ensure a tight seal), secure in
rotary extractor device, and rotate at 30 ± 2 rpm for 18 ± 2 hours.
Ambient temperature (i.e.. temperature of room in which extraction takes
place) shall be maintained at 23 ± 2*C during the extraction period.
NOTE: As agitation continues, pressure may build up within the extractor bottle
for some types of sample (e.g.. limed or calcium carbonate-containing
sample may evolve gases such as carbon dioxide). To relieve excess
pressure, the extractor bottle may be periodically opened (e.g.. after 15
minutes, 30 minutes, and 1 hour) and vented into a hood.
7.2.12 Following the 18 + 2 hour extraction, separate the material
in the extractor vessel into its component liquid and solid phases by
filtering through a new glass fiber filter, as outlined in Step 7.2.7.
For final filtration of the 1312 extract, the glass fiber filter may be
changed, if necessary, to facilitate filtration. Filter(s) shall be
acid-washed (see Step 4.4) if evaluating the mobility of metals.
7.2.13 Prepare the 1312 extract as follows:
7.2.13.1 If the sample contained no initial liquid phase,
the filtered liquid material obtained from Step 7.2.12 is defined
as the 1312 extract. Proceed to Step 7.2.14.
7.2.13.21 If compatible (e.g.. multiple phases will not
result on combination), combine the filtered liquid resulting from
Step 7.2.12 with the initial liquid phase of the sample obtained
in Step 7.2.7. This combined liquid is defined as the 1312
extract. Proceed to Step 7.2.14.
7.2.13.3 If the initial liquid phase of the waste, as
obtained from Step 7.2.7, is not or may not be compatible with the
filtered liquid resulting from Step 7.2.12, do not combine these
liquids. Analyze these liquids, collectively defined as the 1312
extract, and combine the results mathematically, as described in
Step 7.2.14.
7.2.14 Following collection of the 1312 extract, the pH of the
extract should be recorded. Immediately aliquot and preserve the extract
1312 - 12 Revision 0
November 1992
-------
for analysis. Netals allquots must be acidified with nitric acid to pH <
2. If precipitation 1s observed upon addition of nitric acid to a small
aliquot of the extract, then the remaining portion of the extract for
metals analyses shall not be acidified and the extract shall be analyzed
as soon as possible. All other allquots must be stored under
refrigeration (4-C) until analyzed. The 1312 extract shall be prepared
and analyzed according to appropriate analytical methods. 1312 extracts
to be analyzed for metals shall be add digested except 1n those Instances
where digestion causes loss of metallic analytes. If an analysis of the
undigested extract shows that the concentration of any regulated metallic
analyte exceeds the regulatory level, then the waste Is hazardous and
digestion of the extract is not necessary. However, data on undigested
extracts alone cannot be used to demonstrate that the waste Is not
hazardous. If the Individual phases are to be analyzed separately,
determine the volume of the Individual phases (to + 0.5 %), conduct the
appropriate analyses, and combine the results mathematically by using a
simple volume-weighted average:
(V,) (C,)'+ (V2) (C2)
Final Analyte Concentration -
vi + 'V2
where:
V, - The volume of the first phase (L).
C, - The concentration of the analyte of concern in the first phase (mg/L).
V2 « The volume of the second phase (L).
C2 - The concentration of the analyte of concern in the second phase
(mg/L).
7.2.15 Compare the analyte concentrations in the 1312 extract with
the levels identified in the appropriate regulations. Refer to Section
8.0 for quality assurance requirements.
7.3 Procedure when Volatiles are Involved
Use the ZHE device to obtain 1312 extract for analysis of volatile
compounds only. Extract resulting from the use of the ZHE shall not be used to
evaluate the mobility of non-volatile analytes (e.g.. metals, pesticides, etc.).
The ZHE device has approximately a 500 ml internal capacity. The ZHE can
thus accommodate a maximum of 25 grams of solid (defined as that fraction of a
sample from which no additional liquid may be forced out by an applied pressure
of 50 psi), due to the need to add an amount of extraction fluid equal to 20
times the weight of the solid phase.
Charge the ZHE with sample only once and do not open the device until the
final extract (of the solid) has been collected. Repeated filling of the ZHE to
obtain 25 grams of solid is not permitted.
Do not allow the sample, the initial liquid phase, or the extract to be
exposed to the atmosphere for any more time than is absolutely necessary. Any
1312 - 13 Revision 0
November 1992
-------
manipulation of these materials should be done when cold (4'C) to minimize loss
of volatiles.
7.3.1 Pre-weigh the (evacuated) filtrate collection container
(see Step 4.6) and set aside. If using a TEDLAR bag, express all liquid
from the ZHE device into the bag, whether for the initial or final
liquid/solid separation, and take an aliquot from the liquid in the bag
for analysis. The containers listed in Step 4.6 are recommended for use
under the conditions stated in Steps 4.6.1-4.6.3.
7.3.2 Place the ZHE piston within the body of the ZHE (it may be
helpful first to moisten the piston 0-rings slightly with extraction
fluid). Adjust the piston within the ZHE body to a height that will
minimize the distance the piston will have to move once the ZHE is charged
with sample (based upon sample size requirements determined from Step 7.3,
Step 7.1.1 and/or 7.1.2). Secure the gas inlet/outlet flange (bottom
flange) onto the ZHE body in accordance with the manufacturer's
instructions. Secure the glass fiber filter between the support screens
and set aside. Set liquid inlet/outlet flange (top flange) aside.
7.3.3 If the sample is 100% solid (see Step 7.1.1), weigh out
a subsample (25 gram maximum) of the waste, record weight, and proceed to
Step 7.3.5.
7.3.4 If the sample contains <0.5% dry solids (Step 7.1.2), the
liquid portion of waste, after filtration, is defined as the 1312 extract.
Filter enough of the sample so that the amount of filtered liquid will
support all of the volatile analyses required. For samples containing
>0.5% dry solids (Steps 7.1.1 and/or 7.1.2), use the percent solids
information obtained in Step 7.1.1 to determine the optimum sample size to
charge into the ZHE. The recommended sample size is as follows:
7.3.4.1 For samples containing <5% solids (see Step
7.1.1), weigh out a 500 gram subsample of waste and record the
weight.
7.3.4.2 For wastes containing >5% solids (see Step
7.1.1), determine the amount of waste to charge into the ZHE as
follows:
25
Weight of waste to charge ZHE = x 100
percent solids (Step 7.1.1)
Weigh out a subsample of the waste of the appropriate size and
record the weight.
7.3.5 If particle-size reduction of the solid portion of the
sample was required in Step 7.1.3, proceed to Step 7.3.6. If particle-
size reduction was not required in Step 7.1.3, proceed to Step 7.3.7.
7.3.6 Prepare the sample for extraction by crushing, cutting, or
grinding the solid portion of the waste to a surface area or particle size
1312 - 14 Revision 0
November 1992
-------
as described in Step 7.1.3.1. Wastes and appropriate reduction equipment
should be refrigerated, if possible, to 4*C prior to particle-size
reduction. The means used to effect particle-size reduction must not
generate heat in and of itself. If reduction of the solid phase of the
waste is necessary, exposure of the waste to the atmosphere should be
avoided to the extent possible.
NOTE: Sieving of the waste is not recommended due to the possibility that
volatile* may be lost. The use of an appropriately graduated ruler is
recommended as an acceptable alternative. Surface area requirements are
meant for filamentous (e.g.. paper, cloth) and similar waste materials.
Actual measurement of surface area is not recommended.
When the surface area or particle-size has been appropriately
altered, proceed to Step 7.3.7.
7.3.7 Waste slurries need not be allowed to stand to permit the
solid phase to settle. Do not centrifuge samples prior to filtration.
7.3.8 Quantitatively transfer the entire sample (liquid and solid
phases) quickly to the ZHE. Secure the filter and support screens into
the top flange of the device and secure the top flange to the ZHE body in
accordance with the manufacturer's instructions. Tighten all ZHE fittings
and place the device in the vertical position (gas inlet/outlet flange on
the bottom). Do not attach the extraction collection device to the top
plate.
Note: If sample material (>1% of original sample weight) has obviously adhered
to the container used to transfer the sample to the ZHE, determine the
weight of this residue and subtract it from the sample weight determined
in Step 7.3.4 to determine the weight of the waste sample that will be
filtered.
Attach a gas line to the gas inlet/outlet valve (bottom flange)
and, with the liquid inlet/outlet valve (top flange) open, begin applying
gentle pressure of 1-10 psi (or more if necessary) to force all headspace
slowly out of the ZHE device into a hood. At the first appearance of
liquid from the liquid inlet/outlet valve, quickly close the valve and
discontinue pressure. If filtration of the waste at 4*C reduces the
amount of expressed liquid over what would be expressed at room
temperature, then allow the sample to warm up to room temperature in the
device before filtering. If the waste is 100 % solid (see Step 7.1.1),
slowly increase the pressure to a maximum of 50 psi to force most of the
headspace out of the device and proceed to Step 7.3.12.
7.3.9 Attach the evacuated pre-weighed filtrate collection
container to the liquid inlet/outlet valve and open the valve. Begin
applying gentle pressure of 1-10 psi to force the liquid phase of the
sample into the filtrate collection container. If no additional liquid
has passed through the filter in any 2-minute interval, slowly increase
the pressure in 10-psi increments to a maximum of 50 psi. After each
incremental increase of 10 psi, if no additional liquid has passed through
the filter in any 2-minute interval, proceed to the next 10-psi increment.
1312 - 15 Revision 0
November 1992
-------
When liquid flow has ceased such that continued pressure filtration at 50
ps1 does not result 1n any additional filtrate within a 2-mlnute period,
stop the filtration. Close the liquid inlet/outlet valve, discontinue
pressure to the piston, and disconnect and weigh the filtrate collection
container.
NOTE: Instantaneous application of high pressure can degrade the glass fiber
filter and may cause premature plugging.
7.3.10 The material in the ZHE is defined as the solid phase of
the sample and the filtrate is defined as the liquid phase.
NOTE: Some samples, such as oily wastes and some paint wastes, will obviously
contain some material which appears to be a liquid. Even after applying
pressure filtration, this material will not filter. If this 1s the case,
the material within the filtration device is defined as a solid, and is
carried through the 1312 extraction as a solid.
If the original waste contained <0.5 % dry solids (see Step 7.1.2),
this filtrate 1s defined as the 1312 extract and is analyzed directly.
Proceed to Step 7.3.15.
7.3.11 The liquid phase may now be either analyzed immediately
(see Steps 7.3.13 through 7.3.15) or stored at 4*C under minimal headspace
conditions until time of analysis. Determine the weight of extraction
fluid #3 to add to the ZHE as follows:
20 x % solids (Step 7.1.1) x weight
of waste filtered (Step 7.3.4 or 7.3.8)
Weight of extraction fluid «
100
7.3.12 The following steps detail how to add the appropriate
amount of extraction fluid to the solid material within the ZHE and
agitation of the ZHE vessel. Extraction fluid #3 is used in all cases
(see Step 5.7).
7.3.12.1 With the ZHE in the vertical position, attach a
line from the extraction fluid reservoir to the liquid inlet/outlet
valve. The line used shall contain fresh extraction fluid and
should be preflushed with fluid to eliminate any air pockets in the
line. Release gas pressure on the ZHE piston (from the gas
inlet/outlet valve), open the liquid inlet/outlet valve, and begin
transferring extraction fluid (by pumping or similar means) into
the ZHE. Continue pumping extraction fluid into the ZHE until the
appropriate amount of fluid has been introduced into the device.
7.3.12.2 After the extraction fluid has been added,
immediately close the liquid inlet/outlet valve and disconnect the
extraction fluid line. Check the ZHE to ensure that all valves are
in their closed positions. Manually rotate the device in an
end-over-end fashion 2 or 3 times. Reposition the ZHE in the
1312 - 16 Revision 0
November 1992
-------
vertical position with the liquid inlet/outlet valve on top.
Pressurize the ZHE to 5-10 psi (if necessary) and slowly open the
liquid inlet/outlet valve to bleed out any headspace (into a hood)
that may have been introduced due to the addition of extraction
fluid. This bleeding shall be done quickly and shall be stopped
at the first appearance of liquid from the valve. Re-pressurize
the ZHE with 5-10 psi and check all ZHE fittings to ensure that
they are closed.
7.3.12.3 Place the ZHE in the rotary extractor apparatus
(if it is not already there) and rotate at 30 ± 2 rpm for 18 ± 2
hours. Ambient temperature (i.e.. temperature of room in which
extraction occurs) shall be maintained at 23 ± 2'C during
agitation.
7.3.13 Following the 18 ± 2 hour agitation period, check the
pressure behind the ZHE piston by quickly opening and closing the gas
inlet/outlet valve and noting the escape of gas. If the pressure has not
been maintained (i.e.. no gas release observed), the ZHE is leaking.
Check the ZHE for leaking as specified in Step 4.2.1, and perform the
extraction again with a new sample of waste. If the pressure within the
device has been maintained, the material in the extractor vessel is once
again separated into its component liquid and solid phases. If the waste
contained an initial liquid phase, the liquid may be filtered directly
into the same filtrate collection container (i.e.. TEDLAR* bag) holding the
initial liquid phase of the waste. A separate filtrate collection
container must be used if combining would create multiple phases, or there
is not enough volume left within the filtrate collection container.
Filter through the glass fiber filter, using the ZHE device as discussed
in Step 7.3.9. All extracts shall be filtered and collected if the TEDLAR*
bag is used, if the extract is multiphasic, or if the waste contained an
initial liquid phase (see Steps 4.6 and 7.3.1).
NOTE: An in-line glass fiber filter may be used to filter the material within
the ZHE if it is suspected that the glass fiber filter has been ruptured
7.3.14 If the original sample contained no initial liquid phase,
the filtered liquid material obtained from Step 7.3.13 is defined as the
1312 extract. If the sample contained an initial liquid phase, the
filtered liquid material obtained from Step 7.3.13 and the initial liquid
phase (Step 7.3.9) are collectively defined as the 1312 extract.
7.3.15 Following collection of the 1312 extract, immediately
prepare the extract for analysis and store with minimal headspace at 4*C
until analyzed. Analyze the 1312 extract according to the appropriate
analytical methods. If the individual phases are to be analyzed
separately (i.e.. are not miscible), determine the volume of the
individual phases (to 0.5%), conduct the appropriate analyses, and combine
the results mathematically by using a simple volume- weighted average:
(V,) (C,) + (V2) (C2)
Final Analyte
Concentration V, + V2
1312 - 17 Revision 0
November 1992
-------
where:
V, - The volume of the first phases (L).
C, - The concentration of the analyte of concern in the first phase (mg/L).
V2 * The volume of the second phase (L).
C, - The concentration of the analyte of concern in the second phase
(mg/L).
7.3.16 Compare the analyte concentrations in the 1312 extract with
the levels identified in the appropriate regulations. Refer to Section
8.0 for quality assurance requirements.
8.0 QUALITY CONTROL
8.1 A minimum of one blank (using the same extraction fluid as used for
the samples) for every 20 extractions that have been conducted in an extraction
vessel.
8.2 A matrix spike shall be performed for each waste type (e.g..
wastewater treatment sludge, contaminated soil, etc.) unless the result exceeds
the regulatory level and the data is being used solely to demonstrate that the
waste property exceeds the regulatory level. A minimum of one matrix spike must
be analyzed for each analytical batch. As a minimum, follow the matrix spike
addition guidance provided in each analytical method.
8.2.1 Matrix spikes are to be added after filtration of the 1312
extract and before preservation. Matrix spikes should not be added prior
to 1312 extraction of the sample.
8.2.2 In most cases, matrix spike levels should be added at a
concentration equivalent to the corresponding regulatory level. If the
analyte concentration is less than one half the regulatory level, the
spike concentration may be as low as one half of the analyte
concentration, but may not be less than five times the method detection
limit. In order to avoid differences in matrix effects, the matrix spikes
must be added to the same nominal volume of 1312 extract as that which was
analyzed for the unspiked sample.
8.2.3 The purpose of the matrix spike is to monitor the
performance of the analytical methods used, and to determine whether
matrix interferences exist. Use of other internal calibration methods,
modification of the analytical methods, or use of alternate analytical
methods may be needed to accurately measure the analyte concentration in
the 1312 extract when the recovery of the matrix spike is below the
expected analytical method performance.
8.2.4 Matrix spike recoveries are calculated by the following
formula:
%R (% Recovery) = 100 (Xs - Xu) / K
where:
Xs = measured value for the spiked sample
Xu = measured value for the unspiked sample, and
K" = known value of the spike in the sample.
1312 - 18 Revision 0
November 1992
-------
K - known value of the spike in the sample.
8.3 All quality control measures described in the appropriate analytical
methods shall be followed.
8.4 The use of internal calibration quantitation methods shall be
employed for a metallic contaminant if: (1) Recovery of the contaminant from the
1312 extract is not at least 50% and the concentration does not exceed the
appropriate regulatory level, and (2) The concentration of the contaminant
measured in the extract is within 20% of the appropriate regulatory level.
8.4.1. The method of standard additions shall be employed as the
internal calibration quantitation method for each metallic contaminant.
8.4.2 The method of standard additions requires preparing
calibration standards in the sample matrix rather than reagent water or
blank solution. It requires taking four identical aliquots of the
solution and adding known amounts of standard to three of these aliquots.
The forth aliquot is the unknown. Preferably, the first addition should
be prepared so that the resulting concentration is approximately 50% of
the expected concentration of the sample. The second and third additions
should be prepared so that the concentrations are approximately 100% and
150% of the expected concentration of the sample. All four aliquots are
maintained at the same final volume by adding reagent water or a blank
solution, and may need dilution adjustment to maintain the signals in the
linear range of the instrument technique. All four aliquots are analyzed.
8.4.3 Prepare a plot, or subject data to linear regression, of
instrument signals or external-calibration-derived concentrations as the
dependant variable (y-axis) versus concentrations of the additions of
standards as the independent variable (x-axis). Solve for the intercept
of the abscissa (the independent variable, x-axis) which is the concentra-
tion in the unknown.
8.4.4 Alternately, subtract the instrumental signal or external-
calibration-derived concentration of the unknown (unspiked) sample from
the instrumental signals or external-calibration-derived concentrations of
the standard additions. Plot or subject to linear regression of the
corrected instrument signals or external-calibration-derived concentra-
tions as the dependant variable versus the independent variable. Derive
concentrations for the unknowns using the internal calibration curve as if
it were an external calibration curve.
8.5 Samples must undergo 1312 extraction within the following time
periods:
1312 - 19 Revision 0
November 1992
-------
SAMPLE MAXIMUM HOLDING TIMES (days)
Volatiles
Semi-
volatiles
Mercury
Metals,
except
mercury_
From: Field
Collec-
tion
To: 1312
extrac-
tion
14
14
28
180
From: 1312
extrac-
tion
To: Prepara-
tive
extrac-
tion
NA
7
NA
NA
From: Prepara-
tive
extrac-
tion
To: determi-
native
analysis
14
40
28
180
Total
Elapsed
Time
28
61
56
360
NA - Not Applicable
If sample holding times are exceeded, the values obtained will be considered
minimal concentrations. Exceeding the holding time is not acceptable in
establishing that a waste does not exceed the regulatory level. Exceeding the
holding time will not invalidate characterization if the waste exceeds the
regulatory level.
9.0 METHOD PERFORMANCE
9.1 Precision results for semi-volatiles and metals: An eastern soil
with high organic content and a western soil with low organic content were used
for the semi-volatile and metal leaching experiments. Both types of soil were
analyzed prior to contaminant spiking. The results are shown in Table 6. The
concentrations of contaminants leached from the soils were consistently
reproducible, as shown by the low relative standard deviations (RSDs) of the
recoveries (generally less than 10 % for most of the compounds).
9.2 Precision results for volatiles: Four different soils were spiked
and tested for the extraction of volatiles. Soils One and Two were from western
and eastern Superfund sites. Soils Three and Four were mixtures of a western
soil with low organic content and two different municipal sludges. The results
are shown in Table 7. Extract concentrations of volatile organics from the
eastern soil were lower than from the western soil. Replicate leachings of Soils
Three and Four showed lower precision than the leachates from the Superfund
soils.
1312 - 20
Revision 0
November 1992
-------
10.0 REFERENCES
1.0 Environmental Monitoring Systems Laboratory, "Performance Testing of
Method 1312; QA Support for RCRA Testing: Project Report". EPA/600/4-
89/022. EPA Contract 68-03-3249 to Lockheed Engineering and Sciences
Company, June 1989.
2.0 Research Triangle Institute, "Interlaboratory Comparison of Methods 1310,
1311, and 1312 for Lead in Soil". U.S. EPA Contract 68-01-7075, November
1988.
1312 - 21 Revision 0
November 1992
-------
Table 1. Volatile Analytes1
Compound CAS No.
Acetone 67-64-1
Benzene 71-43-2
n-Butyl alcohol 71-36-3
Carbon disulfide 75-15-0
Carbon tetrachloride 56-23-5
Chlorobenzene 108-90-7
Chloroform 67-66-3
1,2-Dichloroethane 107-06-2
1,1-Dichloroethylene 75-35-4
Ethyl acetate 141-78-6
Ethyl benzene 100-41-4
Ethyl ether 60-29-7
Isobutanol 78-83-1
Methanol 67-56-1
Methylene chloride 75-09-2
Methyl ethyl ketone 78-93-3
Methyl isobutyl ketone 108-10-1
Tetrachloroethylene 127-18-4
Toluene 108-88-3
1,1,1,-Trichloroethane 71-55-6
Trichl oroethylene 79-01-6
Trichlorofluoromethane 75-69-4
l,l,2-Trichloro-l,2,2-trifluoroethane 76-13-1
Vinyl chloride 75-01-4
Xylene 1330-20-7
1 When testing for any or all of these analytes, the zero-headspace extractor
vessel shall be used instead of the bottle extractor.
1312 - 22 Revision 0
November 1992
-------
Table 2. Suitable Rotary Agitation Apparatus1
Company
Location
Model No.
Analytical Testing and
Consulting Services,
Inc.
Associated Design and
Manufacturing Company
Environmental Machine and
Design, Inc.
IRA Machine Shop and
Laboratory
Lars Lande Manufacturing
Millipore Corp.
Warrington, PA
(215) 343-4490
Alexandria, VA
(703) 549-5999
Lynchburg, VA
(804) 845-6424
Santurce, PR
(809) 752-4004
4-vessel extractor (DC20S);
8-vessel extractor (DC20);
12-vessel extractor (DC20B)
2-vessel
4-vessel
6-vessel
8-vessel
12-vessel
24-vessel
(3740-2);
(3740-4);
(3740-6);
(3740-8);
(3740-12);
(3740-24)
8-vessel (08-00-00)
4-vessel (04-00-00)
8-vessel (011001)
Whitmore Lake, MI 10-vessel (10VRE)
(313) 449-4116 5-vessel (5VRE)
Bedford, MA
(800) 225-3384
4-ZHE or
4 1-liter
bottle extractor
(YT300RAHW)
1 Any device that rotates the extraction vessel in an end-over-end fashion at 30
±2 rpm is acceptable.
1312 - 23
Revision 0
November 1992
-------
Table 3. Suitable Zero-Headspace Extractor Vessels1
Company
Location
Model No.
Analytical Testing &
Consulting Services, Inc.
Associated Design and
Manufacturing Company
Lars Lande Manufacturing2
Millipore Corporation
Environmental Machine
and Design, Inc.
Warrington, PA
(215) 343-4490
Alexandria, VA
(703) 549-5999
C102, Mechanical
Pressure Device
3745-ZHE, Gas
Pressure Device
Whitmore Lake, MI ZHE-11, Gas
(313) 449-4116 Pressure Device
Bedford, MA
(800) 225-3384
Lynchburg, VA
(804) 845-6424
YT30090HW, Gas
Pressure Device
VOLA-TOX1, Gas
Pressure Device
1 Any device that meets the specifications listed in Step 4.2.1 of the method is
suitable.
2 This device uses a 110 mm filter.
1312 - 24
Revision 0
November 1992
-------
Table 4. Suitable Filter Holders1
Company
Nucleopore Corporation
Micro Filtration
Systems
Millipore Corporation
Location
Pleasanton, CA
(800) 882-7711
Dublin, CA
(800) 334-7132
(415) 828-6010
Bedford, MA
(800) 225-3384
Model/
Catalogue #
425910
410400
302400
311400
YT30142HW
XX1004700
Size
142 mm
47 mm
142 mm
47 mm
142 mm
47 mm
1 Any device capable of separating the liquid from the solid phase of the waste
is suitable, providing that it is chemically compatible with the waste and the
constituents to be analyzed. Plastic devices (not listed above) may be used when
only inorganic analytes are of concern. The 142 mm size filter holder is
recommended.
Table 5. Suitable Filter Media1
Company
Millipore Corporation
Nucleopore Corporation
Whatman Laboratory
Products, Inc.
Micro Filtration
Systems
Location Model
Bedford, MA AP40
(800) 225-3384
Pleasanton, CA 211625
(415) 463-2530
Clifton, NJ GFF
(201) 773-5800
Dublin, CA GF75
(800) 334-7132
(415) 828-6010
Pore
Size
(Mm)
0.7
0.7
0.7
0.7
1 Any filter that meets the specifications in Step 4.4 of the Method is suitable.
1312 - 25 Revision 0
November 1992
-------
TABLE 6 - METHOD 1312 PRECISION RESULTS FOR SEMI-VOLATILES AND METALS
Eastern Soil foH 4.2)
FORTIFIED ANALYTES
bis(2-chloroethyl)-
ether
2-Chlorophenol
1 , A -Dichlorobenzene
1,2- Dichlorobenzene
2-Methylphenol
Nitrobenzene
2 ,4-Dimethylphenol
Hexachlorobutadiene
Acenaphthene
2 . 4 - Dinitrophenol
2 ,4-Dinitrotoluene
Hexachlorobenzene
gamma BHC (Lindane)
beta BHC
METALS
Lead
Cadmium
Amount
Spiked
(Mg)
1040
1620
2000
8920
3940
1010
1460
6300
3640
1300
1900
1840
7440
640
5000
1000
Amount
Recovered*
(Mg)
834
1010
344
1010
1860
812
200
95
210
896**
1150
3.7
230
35
70
387
% RSD
12.5
6.8
12.3
8.0
7.7
10.0
18.4
12.9
8.1
6.1
5.4
12.0
16.3
13.3
4.3
2.3
Western Soil (oH 5.0)
Amount
Recovered*
(Mg)
616
525
272
1520
1130
457
18
280
310**
23**
585
10
1240
65.3
10
91
% RSD
14.2
54.9
34.6
28.4
32.6
21.3
87.6
22.8
7.7
15.7
54.4
173.2
55.2
51.7
51.7
71.3
* - Triplicate analyses.
** - Duplicate analyses; one value was rejected as an outlier at the 90%
confidence level using the Dixon Q test.
1312 - 26
Revision 0
November 1992
-------
TABLE 7 - METHOD 1312 PRECISION RESULTS FOR VOLATILES
Soil
No.
1
(Western)
Avg.
Compound Name
Acetone
Acryloni.tr lie
Benzene
n- Butyl Alcohol
(1-Butanol)
Carbon disulfide
Carbon tetrachloride
Chlorobenzene
Chlorofonn
1 , 2-Dichloroethane
1,1- Dichlor oe thane
Ethyl acetate
Ethylbenzene
Ethyl ether
Isobutanol (4 -Methyl
-1-propanol)
Methylene chloride
%Rec.* %RSD
44.0
52.5
47.8
55.5
21.4
40.6
64.4
61.3
73.4
31.4
76.4
56.2
48.0
0.0
47.5
12
68
8
2
16
18
6
8
4
14
9
9
16
ND
30
.4
.4
.29
.91
.4
.6
.76
.04
.59
.5
.65
.22
.4
.3
Soil
No.
2
(Eastern)
Avg.
%Rec.
43.8
50.5
34.8
49.2
12.9
22.3
41.5
54.8
68.7
22.9
75.4
23.2
55.1
0.0
42.2
* %RSD
2
70
16
14
49
29
13
'16
11
39
4
11
9
ND
42
.25
.0
.3
.6
.5
.1
.1
.4
.3
.3
.02
.5
.72
.9
Soil No. 3
(Western and
Sludge)
Avg.
%Rec.** %RSD
116.0
49.3
49.8
65.5
36.5
36.2
44.2
61.8
58.3
32.0
23.0
37.5
37.3
61.8
52.0
11
44
36
37
51
41
32
29
33
54
119
36
31
37
37
.5
.9
.7
.2
.5
.4
.0
.1
.3
.4
.8
.1
.2
.7
.4
Soil No. 4
(Western and
Sludge)
Avg.
%Rec.
21.3
51.8
33.4
73.0
21.3
24.0
33.0
45.8
41.2
16.8
11.0
27.2
42.0
76.0
37.3
*** %RSD
71.4
4.6
41.1
13.9
31.5
34.0
24.9
38.6
37.8
26.4
115.5
28.6
17.6
12.2
16.6
Methyl ethyl ketone
(2 -Butanone)
Methyl isobutyl
ketone
1,1,1,2-Tetrachloro-
e thane
1,1,2,2-Tetrachloro-
ethane
Tetrachloroethene
Toluene
1.1,1-Trichloro-
ethane
1,1,2-Trichloro-
e thane
Trichloroethene
Trichloro-
fluoromethane
1,1,2-Trichloro-
trifluoroe thane
Vinyl chloride
56.7
81.1
69.0
85.3
45.1
59.2
47.2
76.2
54.5
5
10
6
7
12
8
16
5
11
.94
.3
.73
.04
.7
.06
.0
.72
.1
61
88
41
58
15
49
33
67
39
.9
.9
.1
.9
.2
.3
.8
.3
.4
3
2
11
4
17
10
22
8
19
.94
.99
.3
.15
.4
.5
.8
.43
.5
73.
58.
50.
64.
26.
45.
40.
61.
38.
7
3
8
0
2
7
7
7
8
31.3
32.6
31.5
25.7
44.0
35.2
40.6
28.0
40.9
40.6
39.8
36.8
53.6
18.6
31.4
26.2
46,4
25.6
39.0
40.3
23.8
15.8
24.2
37.2
38.8
25.4
34.1
20.7 24.5
18.1 26.7
10.2 20.3
12.6
6.95
7.17
60.1
58.0
72.8
28.5
21.5
25.0
34.0
67.8
61.0
19.8 33.9
15.3
11.8
24.8
25.4
* Triplicate analyses
** Six replicate analyses
*** Five replicate analyses
1312 - 27
Revision 0
November 1992
-------
Motor
(30 ± 2 rpm
Extraction Vessel Holder
Figure 1. Rotary Agitation Apparatus
1312 - 28
Revision 0
November 1992
-------
Liquid Inlet/Outlet Valve
Top Flange
^MHMH
Support ScreenS*
7
Support Screen'
Viton O-Rings
Bottom Flange
Pressurized Gas
Inlet/Outlet Valve
Pressure
Gauge
Figure 2. Zero-Headspace Extractor (ZHE)
1312 - 29
Revision 0
November 199
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE
7 3 A«*e«ible filter
ho Ider weigh out
•ubaampl•. al1ov
• olid* to jettle.
transfer subsanple
to filter holder.
filter. determine %
•olid*
3-D Begin again
with larger
• ufaaatnple
7 4 Dry filter and
•olid pha*e. record
•eight, calculate %
dry »olida
7 4 S Begin again
MIth nev *ub»ample
No
— ^
744 Diacard
phaaea . will uae
new 1 i quid phaae a*
e« tract
1312 - 30
Revision 0
November 1992
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE (continued)
7 S 3«l*raina if
tub****I* r«qu*f••
ojrlicla-•!!•
7(21 »!!»• i«li<
ph«i« oI t««»t« <••
i«u:« onor t*
f i i'. ration
1312 - 31
Revision 0
November 1992
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE (continued)
7622
Quantitatively
Iranxfer
appropriate amount
a f *««pl• to filter
ho ld*r apply
prasture to {titor
jntil liquid flo-
eea»e«
7623 H.igh
filtrate, analyte
f 111 ra Le? now or
•tare until ti»e of
••tract analp*i«
7 6 2 5 P.rfor.
par ticlv- me
reduction.
quantitativel*
.n.f.r ,«Ud. and
filter to ••tractor
of ••traction fluid
to ••tractor.
••tract for 18
hour*
? 6 4 After
••traction, filter
Uquiri and *olid
pha»e*
1312 - 32
Revision 0
November 1992
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE (continued)
7 6 5 Fii '.ered
material is aef.ned
as extract
765 Filtered
1iqu^d f r om Stepi
7 6 * and 7623
a r * ::• f ined a *
••tract
1 S 6 7 6 7
R*cord pH of
••tract pr*f«rv«.
analyze by
appropriate method*
768 Compare
contaminant
en tract to
appr oprla t«
threshold*
STOP
1312 - 33
Revision 0
November 1992
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE (continued)
7 7 4 D«t«rniin«
optimum taapl* fii«
1 7 Ai.nbU 2HE
d.vic. for volatili
analytit
>0 5*
7 7 } H.i9h
7 7 6 Cool ta.pl*
particle ti»
without generating
h»t
? 7 ? Do net
canlrifug« *••!•*
prior to filtration
1312 - 34
7 7 4 fiUrala it
d.fi
Revision 0
November 1992
-------
METHOD 1312
SYNTHETIC PRECIPITATION LEACHING PROCEDURE (continued)
7 7 3 Tr.nif.r
i«mp1• to ZHC
^ 7 9 Attach
filtrate eel 1eelion
container apply
1 iQuia f1o* ceates
7 7 10 filtrate 11
a.fin.a •» o.lracl
7 7 11 D«t«roin*
weight of
•xtraction fluid
to add to 2HE
7 7 13 Pu.p
•traction fluid
into ZHC
7 7 M Re
head>P
•eve any
*c« .
7
far
7 IS Rotate ZHC
18 hour* • l 22C
7 7 U S.p.r.t.
phaioo
7 7 ;? Define
• •t racl
? ? 18 Anal?t« by
«pproprill*
mathod* combin*
ratult* tf
7 7 19 C««p«c«
eontamnant
appropriate
thraohoIdo
STOP
1312 - 35
7 7 .6
Revision 0
November 1992
-------
-------
APPENDIX B
LIMITED VALIDATION OF THE HWANG
AND FALCO MODEL FOR EMISSIONS
OF SOIL-INCORPORATED VOLATILE
ORGANIC COMPOUNDS
-------
-------
LIMITED VALIDATION OF THE HWANG
AND FALCO MODEL FOR EMISSIONS
OF SOIL-INCORPORATED VOLATILE
ORGANIC COMPOUNDS
by
Environmental Quality Management, Inc.
3109 University Drive, Suite B
Durham, North Carolina 27707
Contract No. 68-D8-0111
Work Assignment No. 92-10
Subcontract No. 0111-EQ1
PN5055
Janine Dinan, Work Assignment Manager
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE
WASHINGTON, D.C. 20460
September 1992
-------
DISCLAIMER
This project has been performed under contract to The Cadmus Group, Inc. It
was funded with Federal funds from the U.S. Environmental Protection Agency under
Contract No. 68-D8-0111. The content of this publication does not necessarily reflect
the views or policies of the U.S. Environmental Protection Agency nor does mention of
trade names, commercial products, or organizations imply endorsement by the U.S.
Government.
-------
CONTENTS
Figures iv
Tables y
Acknowledgment vi
1. Introduction 1
Soil PRGs 2
Project objectives 2
Technical approach 3
2. Review of the RAGS/HHEM, Part B Volatilization Model 4
Model derivation 6
Summary of model assumptions and limitations 11
3. Model Validation 14
Bench scale validation 15
Pilot scale validation 25
4. Parametric Analysis of the Hwang and Falco Model 38
Affects of soil parameters 38
Affects of nonsoil parameters 41
5. Conclusions 44
References 47
Appendices
A. Bench-Scale Model Validation Data A-1
B. Pilot-Scale Model Validation Data B-1
iii
-------
FIGURES
Number
1 Predicted and measured emission rates of lindane versus time 18
2 Comparison of modeled and measured emission rates of lindane 19
3 Predicted and measured emission rates of lindane versus time
employing the Millington and Quirk expression of D^ 21
4 Predicted and measured emission rates of dieldrin versus time 22
5 Predicted and measured emission rates of dieldrin versus time
employing the Millington and Quirk expression of D.J 23
6 Comparison of modeled and measured emission rates of dieldrin 24
7 Predicted and measured emission rates of benzene versus time 29
8 Predicted and measured emission rates of toluene versus time 30
9 Predicted and measured emission rates of ethylbenzene versus time 31
10 Comparison of modeled and measured emission rates of benzene,
toluene, and ethylbenzene 33
11 Predicted and measured emission rates of benzene versus time
employing the Millington and Quirk expression of D.J 35
12 Predicted and measured emission rates of toluene versus time
employing the Millington and Quirk expression of D^ 36
13 Predicted and measured emission rates of ethylbenzene versus time
employing the Millington and Quirk expression of D.J 37
IV
-------
TABLES
Number Page
1 Summary of Statistical Analysis of Bench-Scale Validation 26
2 Summary of Statistical Analysis of Pilot-Scale Validation Using
the Default Hwang and Falco Effective Diffusion Coefficient 32
3 Summary of Statistical Analysis of Pilot-Scale Validation Using
the Millington and Quirk Expression for Effective Diffusivity 34
-------
ACKNOWLEDGMENT
This report was prepared for the U.S. Environmental Protection Agency by
Environmental Quality Management, Inc. of Durham, North Carolina under contract to
The Cadmus Group, Inc. Dr. Joanne Wyman with The Cadmus Group, Inc. served as
the project technical monitor and Craig Mann with Environmental Quality Management,
Inc. managed the project and was author of the report. Janine Dinan of the U.S.
Environmental Protection Agency's Toxics Integration Branch provided overall project
direction and served as the Work Assignment Manager.
VI
-------
SECTION 1
INTRODUCTION
The Risk Assessment Guidance for Superfund: Volume I - Human Health
Evaluation Manual (RAGS/HHEM) Part B provides guidance on using the U.S.
Environmental Protection Agency (EPA) toxicity values and exposure information to
derive risk-based preliminary remediation goals (PRGs). In general, PRGs provide
remedial design staff with long-term cleanup level targets to use during analysis and
selection of remedial alternatives.
The National Contingency Plan which is found in 40 CFR, Part 300, mandates
that the selected remedial alternative meet all applicable or relevant and appropriate
requirements (ARARs), and provide protection of human health and the environment.
PRGs are developed to quantify the standards that remedial alternatives must meet in
order to achieve these two "threshold criteria." Of major concern in establishing PRGs
are "long-term effectiveness and permanence" of the remedy. These balancing criteria
for remedy selection are used to establish the risk posed to the community once the
remediation is complete. Risk-based PRGs quantify the degree of residual risk after
cleanup has been completed. If ARARs do not exist for the contaminant of concern or
for the media of concern, risk-based PRGs are developed to protect human health.
PRGs are typically developed during the scoping phase or concurrent with initial
phases of the Remedial Investigation/Feasibility Study (RI/FS). Risk-based PRGs are
considered initial guidelines developed with readily available information and can be
modified as additional site data are obtained. A risk-based concentration is considered
a final remediation level only after appropriate analysis in the RI/FS and in the Record
of Decision (ROD).
-------
1.1 SOIL PRGs
PRGs for the soil medium are calculated for carcinogenic and noncarcinogenic
contaminants from standard residential and commercial/industrial land-use equations
given in RAGS/HHEM, Part B. Integral to these equations is the soil-to-air volatilization
factor (VF) which defines the relationship between the concentration of 6ontaminants in
soil and the volatilized contaminants in the air. The VF (m3/kg) is calculated as the
inverse of the ambient air concentration at the center of a ground-level, nonbouyant
area source of emissions. The RAGS/HHEM, Part B equation for calculating the VF
consists of two parts: 1) a volatilization model, and 2) a dispersion model.
The volatilization model mathematically predicts volatilization of contaminants fully
incorporated in soils as a diffusion-controlled process. The basic assumption in the
mathematical treatment of the movement of volatile contaminants in soils under a
concentration gradient is the applicability of the diffusion laws. The changes in
contaminant concentration within the soil as well as the loss of contaminant at the soil
surface by volatilization can then be predicted by solving the diffusion equation for
different boundary conditions.
This document reports on several studies in which volatilization of contaminants
from soils was directly measured and from which data were obtained necessary to
calculate emissions of contaminants using the RAGS/HHEM, Part B volatilization model.
These data are then compared and analyzed by statistical methods to determine the
relative accuracy cf the model.
1.2 PROJECT OBJECTIVES
The primary objective of this project was to assess the relative accuracy of the
RAGS/HHEM, Part B volatilization model using experimental emission flux data from
previous studies as a reference data base.
-------
1.3 TECHNICAL APPROACH
The following series of tasks comprised the technical approach for achieving the
project objective:
1. Review the theoretical basis and development of the RAGS/HHEM, Part B
volatilization model to verify the applicable model boundan/conditions and
variables, and to document the assumptions and limitations of the model.
2. Perform a literature search and survey (not to exceed nine contacts) for
the purpose of determining the availability of acceptable emission flux data
from experimental and field-scale measurement studies of volatile organic
compound (VOC) emissions from soils. Acceptable data must have
undergone proper quality assurance/quality control (QA/QC) procedures.
3. Determine if the emission flux measurement studies referred to in Task
No. 2 also provided sufficient site data as input variables to the
volatilization model. Again, acceptable variable input data must have
undergone proper QA/QC procedures.
4. Review, collate, and normalize emission flux measurement data and
volatilization model variable data, and compute chemical-specific emission
rates for comparison to respective measured emission rates.
5. Perform statistical analysis of the results of Task No. 4 to establish the
extent of correlation between measured and modeled values and perform
parametric analysis of key model variables.
-------
SECTION 2
•
REVIEW OF THE RAGS/HHEM, PART B VOLATILIZATION MODEL
The soil-to-air volatilization factor as calculated from Equation No. 8 of the
RAGS/HHEM, Part B incorporates the Hwang and Falco (1986) model for volatilization
of polychlorinated byphenyls (PCBs) incorporated in soils as developed by EPA's
Exposure Assessment Group (EPA, 1986a). The model calculates the instantaneous
emission flux at time, t, as:
A/ = ' C*
A = (n a t)°5 ' 7Q ' 5
where NA = Instantaneous emission flux, g/cm2-s
E = Soil porosity, dimensionless
D.J = Effective diffusivity of component i in soil, cm2/s (= D, • E033)
D, = Diffusivity of component i in air, cm2/s
n = 3.1416
t = Time from soil sampling, s
H = Henry's Law constant, dimensionless form
Ky = Soil/water partition coefficient, cm3-water/g-soil
Cso - Initial contaminant soil concentration , g/g-soil
-------
and,
where P. is the true density of soil (i.e., particle density, p).
The model assumes that the surface of the contaminated soil column is exposed
to the atmosphere. The initial and boundary conditions are:
1. Initial condition Q = (H/K^)C.0, at t = 0, L ^ 0
2. Boundary condition Q = (H/K;)C.0, at L = «>, t > 0
3. Boundary condition Q = 0, at L = 0, t > 0
where Q = concentration of component i in the vapor phase in the air-filled soil pore
spaces, and L = depth of contaminated soil column.
The average flux rate, N., over exposure interval, T, is time-averaged as:
/v; (T) = 2/V.m (3)
or
/v. = 2/V.
To calculate the total average emission rate, Q(g/s), the average flux rate is multiplied
by the area of contaminated soil:
-------
Q = A(N.)
(4)
2.1 MODEL DERIVATION
The Hwang and Falco model is derived from the methods presented by Farmer
and Letey (1974) and Farmer, et al. (1980). Farmer, et al. considered a system where
pesticide is uniformly mixed with a layer of soil and volatilization occurs at the soil
surface. If diffusion is the only mechanism supplying pesticide to the surface of an
isotropic soil column, and if the diffusion coefficient, D, is assumed to be constant, the
general diffusion equation is:
*L£ - 1 l£ = 0 (5)
dx2 D dt
where c = Soil concentration, g/cm3 total volume
x = Distance measured normal to soil surface, cm
D = Apparent diffusion coefficient in soil, crr^/s
t = Time, s.
If the pesticide is rapidly removed by volatilization from the soil surface and is
maintained at a zero concentration, the initial and boundary conditions are:
c = C0 at t = 0, 0 £ x 22 L
c = 0 at x = 0 and t > 0
3c/3x = 0 at x = L
where L is the depth of the contaminated soil column (cm) and C0 is the initial
concentration in soil (g/cm3).
-------
Recognizing the analogy between the heat transfer equation (Fourier's Law) and
the transfer of matter under a concentration gradient (Pick's Law), Farmer, et al.
employed the heat transfer equations of Carslaw and Jaeger (1959, page 97, Equation
No. 8) to solve the diffusion equation given these boundary and initial conditions for the
emission flux, f(g/cm2-s), as:
oo
2 (-1)" exp(-/72/.2/0f)
(6)
where the following analogies to the Carslaw and Jaeger solution are assumed:
v (temperature) = c
V0 (initial temperature) = C0
k (thermal diffusivity), and
K (thermal conductivity) = D.
The summation term in Equation No. 6 decreases with increasing L and
decreasing D and t. If this term is small enough to be negligible, Equation No. 6 for L
= oo, reduces to:
/ = D CJ(n Of)0-5 (7)
The concentration for the semi-infinite case is given by Crank (1985) as:
C = C0 erf [x/2(Df)° 5] (8)
-------
Equation Nos. 7 and 8 are also applicable to a finite system (0 < x
-------
or
(11)
dt
where
a » L (12)
[£ + (P,)C\-
In this manner, Hwang and Falco redefined the general diffusion equation given by
Farmer, et al. for vapor phase diffusion and soil adsorption.
It should be noted that both the Farmer, et al. equation and the Hwang and
Falco equation have many of the same assumptions. Both equations assume that
vapor phase diffusion is the only transport mechanism moving contaminants from the
soil column to the soil surface. This assumes no transport via nonvapor phase
diffusion or mass flow due to capillary action within the soil column. These
assumptions were shared due to the relative insolubilities of hexachlorobenzene
(Farmer, et al. 1980) and PCBs (Hwang and Falco 1986a) which comprised the
contaminants of concern for the respective studies. In addition, the experimental
conditions of Farmer, et al. and the theoretical concerns of Hwang and Falco excluded
mass flow from consideration. Farmer, et al. and Hartley (1964), however, suggest that
a soil solution/soil air partition coefficient defined as the ratio of the solubility of the
contaminant in water to the saturation vapor concentration of the contaminant may be
used to estimate the major mode of diffusion between the vapor and nonvapor phase.
Because the vapor phase diffusion coefficient is approximately 10* larger than the
solution (nonvapor) phase diffusion coefficient, a partition coefficient of 10* may be
considered as a transition point for determining when vapor diffusion or solution
diffusion becomes dominant. Chemicals with partition coefficients much smaller than
10* will diffuse mainly in the vapor phase while those with partition coefficients much
-------
greater than 10* will diffuse mainly in the solution phase. Therefore, chemicals which
diffuse mainly in the solution phase are not considered applicable to either the Farmer,
et al. or Hwang and Falco procedures.
In the Hwang and Falco model, the contaminant concentrations in soil and in
interstitial vapors are assumed to be in local equilibrium and related by the following
equation:
C = \"\C. (13)
•m
Hwang and Falco thus assume that the concentration of interstitial vapors is a
function of the ratio of the Henry's Law constant and its soil/water partition coefficient.
This relationship holds true below the soil concentration at which the adsorptive
properties of the soil and the theoretical dissolution limit of the contaminant in the
available free soil moisture have been reached. The term (H/K^) is thus defined by
Hwang and Falco as the soil/air partition coefficient (K^.).
Finally, Hwang and Falco substitute the effective diffusivity is soil, D^, for the
Farmer, et al. apparent diffusion coefficient in soil, D. The effective diffusivity in soil, D.,,
is used to account for tortuosity effects in porous media; however, the effective porosity
for dry soil (D, • E° 33) is used for simplicity. Hwang and Falco, however, do reference
the Farmer, et al. 098°) procedures for incorporating the effect of soil geometry and
soil moisture in the porosity term. Farmer, et al. (1980) refined the flux calculations to
give a decreased flux rate due to reduced air-filled porosity taking into account the
effect of soil moisture on tortuosity. The equation from Millington and Quirk (1961) is:
where D.J = Effective diffusivity in soil, cm2/s
10
-------
P. = Air-filled soil porosity, dimensionless
P, = Total soil porosity, dimensionless.
Therefore, the Hwang and Falco solution for instantaneous emission flux at the
soil surface (Equation No.1) a is direct derivation of the Farmer, et al. reduced equation
for diffusion-controlled emission flux (Equation No.7). The Hwang and Falco equation,
however, attempts to establish the relationship between the concentration in air and the
concentration adsorbed to soil through the use of the air/soil partition coefficient, f^,.
This is equivalent to the Farmer, et al. isotherm coefficient, R,,, and is approximated in
moist soils as the ratio of the Henry's Law constant to the soil/water partition
coefficient, f^. As an equivalent expression to the Farmer, et al. reduced equation, the
Hwang and Falco equation is valid until time t = L2/14.4o (Equation No. 9).
2.2 SUMMARY OF MODEL ASSUMPTIONS AND LIMITATIONS
The Hwang and Falco volatilization model is analogous to the mathematical
solution for heat flow in an infinite solid (Carslaw and Jaeger, 1959), and as such it's
applicability to diffusion processes is limited to the initial and boundary conditions upon
which the model is derived. The model assumes that there are no other vectors for
movement or loss of contaminant other than vapor phase diffusion from the soil column
to the soil surface (diffusion-controlled). Other vectors such as mass flow due to
capillary action, redistribution of contaminants due to rain events, loss of contaminants
at the lower boundary due to leaching, nonvapor phase or solution diffusion,
biodegradation, photolysis, and possible codistillation at the soil surface are not
considered.
The model also assumes an infinitesimal layer of noncontaminated soil at the
soil-air-interface once local equilibrium has been reached (Q = 0 at L = 0, t > 0). This
implies adequate wind velocity at the boundary layer to completely volatilize
contaminants at the soil surface rendering the process diffusion-controlled. The
boundary conditions also specify that the depth of the contaminated soil column, L, is
11
-------
infinite, thus assuming zero vertical movement or loss from the lower boundary over an
infinite time period. In this manner, the model is analogous to the Farmer, et al.
reduced flux solution (Equation No. 7). The validity of this argument becomes
questionable with decreasing depth of the actual contaminated soil column, L, and
increasing time. .
The soil/air partition coefficient, K^., assumes that under equilibrium conditions,
contaminants are partitioned only between the aqueous phase, interstitial vapors, and
that adsorbed to soil. The Henry's Law constant employed to derive K^ is the ratio of
the contaminant's vapor pressure and solubility in water. Thus, K^, assumes an excess
of water at equilibrium. In this manner, back-calculation of the total concentration in soil
(i.e., the PRG) employing the Hwang and Falco model is invalid for soil concentrations
at and above which the saturation vapor concentration is achieved. This concentration
is estimated in the RAGS/HHEM, Part B as the saturation concentration, C..,:
where C,,,, = Saturation concentration, mg/kg
KU = Soil/water partition coefficient, L/kg
s = Solubility in water, mg/L
r\n = Soil moisture content, kg-water/kg-soil
0m = Soil moisture content, L-water/kg-soil.
As defined in Hwang and Falco (1986), the soil/water partition coefficients for
PCBs, K,, were experimental values reported in the scientific literature. In the
RAGS/HHEM, Part B equation, f^ is defined as the product of the octanol/water
partition coefficient, K^, and the soil organic carbon fraction. This theoretically derived
value of Kj does not account for the adsorptive properties of the clay, silt, or dissolved
12
-------
organic content of the soil which may increase the actual adsorptive soil properties,
thereby effecting the time for the flux to arrive at steady-state conditions.
The effective diffusion coefficient, D^, used in the Hwang and Falco model
assumes dry soil, thus maximum diffusion conditions. Use of the Millington and Quirk
expression (Equation No. 14) will reduce the effective diffusivity. Although air-filled
porosity is found be to the major factor controlling volatilization flux through the soil-
water-air system, the apparent vapor diffusion coefficient does not depend only on the
amount of air-filled pore space (Farmer, et at., 1980). The presence of liquid films on
the solid surfaces not only reduce the porosity, but also modifies the pore geometry
and the length of the gas passage. Farmer, et al. thus recommend use of the
Millington and Quirk expression to derive the effective diffusion coefficient when an
estimate of long-term soil water contents can be made.
In general, the Hwang and Falco model describes the desorption of
contaminants from soil and the vapor phase diffusion of the contaminants to the soil
surface to replace that lost by volatilization to the atmosphere. The model assumes an
exponential decay curve over time once steady-state conditions are achieved. In
actuality, there is a high initial flux rate from the soil as surface concentrations are
depleted. The lower flux rate characteristics of the later portion of the decay curve is
thus determined by the rate at which contaminants diffuse upward. This type of
desorption curve has been well documented in the literature. It isHmportant to note that
the Hwang and Falco model does not account for the high initial rate of volatilization
before equilibrium is attained and will tend to underpredict emissions during this period.
Finally, the Hwang and Falco model is applicable only to single chemical compounds
fully incorporated into isotropic soils. Effective solubilities and activity coefficients in
multicomponent systems are not addressed in the determination of the air/soil partition
coefficient nor is the effect of nonlinear soil adsorption and desorption isotherms.
However, because of the complexities involved with theoretical solutions to these
effects, their contribution to the relative accuracy of the model is difficult to predict,
especially in multicomponent systems.
13
-------
SECTION 3
MODEL VALIDATION
To achieve the project objective, Environmental Quality Management, Inc. (EQ)
executed a literature search and a survey of professional environmental
investigation/research firms as well as regulatory agencies to obtain experimental and
field data suitable for comparing modeled emissions with actual emiss-c^s. The
literature search uncovered several papers and bench-scale experimental studies
concerned with the volatilization and vapor density of pesticides and chlorinated
organics incorporated in soils (Farmer, et al., 1972, 1974, and 1980; Spencer and
Cliath, 1969 and 1970; and Spencer, 1970).
From the literature search, one study was found which met the data
requirements for this project (Farmer, et al., 1972 and 1974). This study reports the
experimental emissions of lindane (1,2,3,4,5,6-hexachlorocyclohexane, gamma isomer)
and dieldrin (1,2,3,4,10,10-hexachloro-6,7-epoxy-1 ^^a.S.S^.S.Sa-octahydro-l ,4-endo,
exo-5, 8-dimethanonapthalene) incorporated in Gila silt loam. Suitable data for input to
the Hwang and Falco model were also available from this study.
The objective of the survey of professional firms and regulatory agencies was to
find pilot-scale or field-scale studies of volatilization of organic compounds using the
EPA emission isolation flux chamber. The candidate flux chamber studies must also
have provided adequate data for input to the Hwang and Falco model.
Flux chamber studies were chosen to provide pilot-scale or field-scale
measurement data needed for model validation. Flux chambers have been widely used
to measure flux rates of VOCs and inorganic gaseous pollutants from a wide variety of
sources. The flux chamber was originally developed by soil scientists to measure
biogenic emissions of inorganic gases and their use dates back at least two decades
14
-------
(Hill, et al., 1978). In the early 1980's, EPA became interested in this technique for
estimating emission rates from hazardous wastes and funded a series of projects to
develop and evaluate the flux chamber method. The initial work involved the
development of a design and approach for measuring flux rates from land surfaces. A
test cell was constructed and parametric tests performed to assess chamber design
and operation (Kienbusch and Ranum, 1986 and Kienbusch, et al., 1986). A series of
field tests was performed to evaluate the method under field conditions (Radian
Corporation, 1984 and Balfour, et al., 1984). A user's guide was subsequently
prepared summarizing guidance on the design, construction, and operation of the EPA
recommended flux chamber (Keinbusch, 1985). The emission isolation flux chamber Is
presently considered the preferred in-depth direct measurement technique for
emissions of VOCs from land surfaces (EPA, 1990).
EQ contacted several environmental consulting firms as well as State and local
agencies. In addition, the EPA data base of emission flux measurement data was
reviewed (EPA, 1991). Although several flux measurement studies were reviewed, only
one applicable study was identified with adequate QA/QC documentation and with the
necessary input data for the Hwang and Falco model (Radian Corporation, 1989).
3.1 BENCH-SCALE VALIDATION
From Farmer, et al. (1972 and 1974) the influence of pesticide vapor pressure on
volatilization was measured by comparing the volatilization from Gila silt loam of dieldrin
with that of lindane. Volatilization of dieldrin and lindane was measured in a closed air-
flow system by collecting the volatilized insecticides in ethylene glycol traps. Ten grams
of soil were treated with either 5 or 10>g/g of C-14 tagged insecticide in hexane. The
hexane was evaporated by placing the soils in a fume hood overnight. Sufficient water
was then added to bring the initial soil water content to 10 percent. For the
volatilization studies, the treated soil was placed in an aluminum pan 5 mm deep, 29
mm wide, and 95 mm long. This produced a bulk density of 0.75 g/cm3. The
aluminum pan was then introduced into a 250 mL bottle which served as the
15
-------
volatilization chamber. A relative humidity of 100 percent was maintained in the
incoming air stream to prevent water evaporation from the soil surface. Air flow was
maintained at 8 mL/s equivalent to approximately 0.018 miles per hour. The
temperature was maintained at 30° C. The soil was a Gila silt loam, which contained
0.58 percent organic matter.
The volatilized insecticides were trapped in 25 mL of ethylene glycol.
Insecticides were extracted into hexane and anhydrous sodium sulfate was added to
the hexane extract to remove water. Aliquots of the dried hexane were analyzed for
lindane and dieldrin using liquid scintillation. The extraction efficiencies for lindane and
dieldrin were 100 and 95 percent, respectively. The concentrations of volatilized
compounds were checked using gas-liquid chromatography. All experiments were r> m
in duplicate.
From the experimental data, variables for input to the Hwang and Falco
volatilization model were:
0 Soil porosity (E) = 1 - Vjp
0 Bulk density (F;) = 0.75 g/cm3
0 Effective diffusion coefficient (D.) = D, • E°'33
0 Initial soil concentration (C.J = 5 ppm and 10 ppm.
The diffusion coefficient in air, D,, is from Equation No. 2-4 of EPA (1988). The
soil/water partition coefficient, K^, was set equal to the product of the soil organic
carbon content fraction, 0.0058, and the organic carbon partition coefficient, K^, from
EPA (1986). All data were input to a spreadsheet and emission flux rates were
subsequently computed. Emission rates, Q, were computed as the product of the flux
rates and the soil surface area (27.55 cm3). Initial soil concentrations were compared
against calculated values of the saturation concentration, C,.t, using Equation No. 15.
Water solubilities used in Equation No. 15 are from EPA (1986); soil water content was
10 percent (w/w) from Farmer (1972 and 1974). Initial soil concentrations of 5 and 10
ppm for dieldrin, and 10 ppm for lindane were above their respective theoretical
16
-------
saturation concentrations by less than or equal to a factor of 50. Appendix A contains
the spreadsheet data for both lindane and dieldrin at initial soil concentrations of 5 ppm
and 10 ppm.
The instantaneous emission rate values predicted by the model for each period
corresponding to measurements of volatilization flux were compared, ffigure 1 shows
the comparisons of predicted and measured values for lindane at an initial
concentration of 10 ppm. A best curve was fit to both the measured and predicted
values. As expected, both curves indicate an exponential decrease in emission rate
with time.
The ratio of the modeled emission rate to the measured emission rate was
determined as a measure of the relative difference between the two emission rates
(Rgure 2). The natural log of this ratio was then analyzed by using a standard paired
Student's t-test. This analysis is equivalent to assuming a lognormal distribution for the
emission rates and analyzing the log-transformed data for differences between modeled
and measured emission rates.
The data were also analyzed by using standard linear regression techniques.
Again, the data were assumed to follow a lognormal distribution. A simple linear
regression model was fit to the log-transformed data and the Pearson correlation
coefficient was determined. The Pearson correlation coefficient is a measure of the
strength of the linear association between the two variables.
From a limited population of 4 observations, the correlation coefficient was
calculated to be 0.998 with a mean ratio of modeled-to-measured values of 5.6. The
actual significance level or p-value of the paired Student's t-test was 0.0163, which
indicates some degree of statistical difference between modeled and measured values
at the 95 percent confidence level. The lower and upper confidence limits were
calculated to be 1.8 and 17, respectively. This indicates that at the 95 percent
confidence limit, the modeled emission rates are 1.8 to 17 times higher than the
measured emission rates. At an initial concentration of 5 ppm, the correlation
17
-------
oo
CD
9
8
7-
6-
c LJJ 5-
o o 01
CO
cn
0)
03
T3
C
^•^
E
4-
3-
2-
1-
CURVEFIT: Soil Concentration = 10 ppm
y (predicted) = 4.06189E-09 * x~ -0.4998
y (measured) = 2.72935E-11 * e ~ (((x -192.2279) ^ 2)/10871.8546)
0 20 40 60 80 100 120 140 160 180 200
Time From Sampling (hrs)
Measured Predicted
n
Figure 1. Predicted and measured emission rates of lindane versus time.
-------
CO
-20.9-
-21.0-
-21.1 -
-21.2-
-21.3-
« -21.4^
rr
C -21.5-
,§ -21.7 ^
-2
-22.0-
-22.1
-22.2-
-22.3
-22.4-
-22.5:
-22.8-
-26
-25
-24 -23
Log Measured Emission Rate
-22
-21
10ppm
•o- -»-«
5ppm
Figure 2. Comparison of modeled and measured emission rates of lindane.
-------
coefficient was 0.99 with a mean ratio of 6.2, p-value of 0.0414, and a 95 percent
confidence interval of 2.0 to 19.
In an attempt to determine the relative significance of the Hwang and Falco
assumption of dry soil on the effective diffusion coefficient and the subsequent effect on
the emission rate, the predicted values were recalculated using the Millington and Quirk
expression for effective diffusivity (Equation No. 14). Figure 3 shows the results of this
comparison. As expected, the relative difference between the modeled and measured
emission rates decreased. Statistical analysis at an initial concentration of 10 ppm
indicates a correlation coefficient of 0.99, and mean ratio of 3.7, a p-value of 0.0333,
and a resultant decrease in the 95 percent confidence interval to 1.2 to 11. Pooling
data between measured and modeled values for initial concentrations of 5 ppm and 10
ppm, reduced the correlation coefficient and p-value to 0.93 and 0.0006, respectively;
but the increase in the population to 8 observations decreased the 95 percent
confidence interval to 2.3 to 6.7.
For dieldrin, Rgure 4 shows the comparison of modeled to measured values
using the default Hwang and Falco assumption of dry soil, while Figure 5 shows the
comparison using the Millington and Quirk expression for effective diffusivity. As can be
seen, the relative difference between modeled and measured values in the case of
dieldrin is significantly reduced over that of lindane. In addition, the predicted values
using the Millington and Quirk expression are initially lower than the measured values
and are subsequently higher once equilibrium is achieved. The initially higher
measured emission rates are to be expected as surface contaminants are depleted and
the model boundary conditions are achieved. The relative magnitude between the
difference of modeled and measured emission rates for lindane and dieldrin may be
due to analytical precision and accuracy differences between the two compounds.
Figure 6 shows the comparison of the log-transformed data for the modeled and
measured emission rates of dieldrin at an initial concentration at both 5 ppm and 10
ppm. The modeled data in Figure 6 were calculated using the standard Hwang and
Falco assumption of dry soil. At an initial concentration of 10 ppm, the correlation
20
-------
N>
I
(1)
I f
C LLJ
O O
C CD
i§ E
I
03
5-
4-
3-
2-
1-
0
CURVE FIT: Soil Concentration = 10 ppm
y (predicted) = 2.71067E-09 * x~ -0.5003
y (measured) = 2.72935E-11 * e ~ (((x -192.2279) ~ 2)/10871.8546)
0 20 40 60 80 100 120 140 160 180 200
Time From Sampling (hrs)
Measured —- Predicted
Figure 3. Predicted and measured emission rates of llndane versus time employing the Mlllington and Quirk expression of D.,
-------
ro
1.6
1.4
~ 1.2H
(D ^
"(0 O
cc -7
C HI 1-
O O
"(0 *~
(0 W
E g 0.8H
LU .=
1 0.6H
Q>
Q
0.4H
0.2
0
Soil Concentration = 10 ppm
CURVE FIT:
y (predicted) = 7.37639E-10 * (x^-0.4989)
y (measured) = 2.63479E-11 * e ~ (((x-254.(.880) ~ 2)732642.6542)
-Q-
50 100 150 200 250
Time From Sampling (hrs)
300
Measured — Predicted
Figure 4. Predicted and measured emission rates of dieldrin versus time.
-------
1.4
,co
u*
0)
to o
a: -7
c LU
o o
8 «
1 |
LU
Q)
1-
0.6-
0.4^
0.2
0
Soil Concentration = 10 ppm
CURVE FIT:
y (predicted) = 5.29339E-10 * 1.1777~ (1/x) * x~ -0.4994
y (measured) = 2.63479E-11 * e/v(((x-254.6880)/v2)/32642.6542)
50
100 150 200
Time From Sampling (hrs)
250
300
Measured — Predicted
Figure 5. Predicted and measured emission rates of dieldrin versus time employing the Millington and Quirk expression of D(l
-------
-22.6-
-22.7-
-22.8-
-22.9-
-23.0-
O -23-1
| -23.2-
c -23.3
•| -23.4-
'£ -23.5
m -23.8
TJ
J -23.7-
•§ -23.8
5 -23.9
O) -24.0-
~' -24.1-
-24.2-
-24.3-
-24.4-
-24.5
-24.8-
-26
x' D
,'D
-25
-24
-23
-22
Log Measured Emission Rate
10ppm
•O--B-O
5ppm
Figure 6. Comparison of modeled and measured emission rates of dieldrin.
-------
coefficient was calculated to be 0.98, the mean ratio 1.5, the p-value 0.0023, and the
95% confidence limit 1.2 to 1.8.
After applying the Millington and Quirk expression, pooled data (5 ppm and 10
ppm initial concentrations) indicate a correlation coefficient of 0.97. The p-value of the
paired Student's t-test was calculated to be 0.4133 indicating no statistically significant
difference between modeled and measured values (i.e., a mean ratio of 1).
Table 1 summarizes statistical analysis for the bench-scale comparative validation
of the Hwang and Falco model. As can be seen from these data, a larger population of
observations in conjunction with use of the Millington and Quirk expression of effective
diffusivity reduced the 95 percent confidence interval in the case of lindane and resulted
in no significant statistical difference in the average between modeled and measured
values in the case of dieldrin.
3.2 PILOT-SCALE VALIDATION
From Radian Corporation (1989), a study was designed to determine how
different treatment practices affect the rate of loss of benzene, toluene, xylenes, and
ethylbenzene (BTEX) from soils. The experiment called for construction of four piles of
loamy sand soil, each with a volume of approximately 4 cubic yards (7900 pounds ), a
surface area of 8 square meters, and a depth of 0.91 meters. Each test cell was lined
with an impermeable membrane and the soil in each cell was sifted to remove particles
larger than three-eighth inch in diameter. The contaminated soil for each pile was
prepared in batches using 55-gallon drums. In the "high level" study, each soil batch
was brought to 5 percent moisture content and 6 liters of gasoline added. Additional
water was then added to bring the soil to 10 percent moisture by weight. The drums
were capped and sat undisturbed overnight. The drums were then opened the next
day and shoveled into the test cell platform. Twenty-two soil batches were prepared for
each soil pile. Each batch consisted of 360 pounds of soil and 6.0 liters of fuel.
Therefore, each soil pile contained 7900 pounds of soil and 132 liters of gasoline. Each
soil pile was then subjected to one of the following management practices:
25
-------
TABLE 1. SUMMARY OF STATISTICAL ANALYSIS OF BENCH-SCALE VALIDATION
Chemical
Lindane (10 ppm)
LJndane (5 ppm)
Lindane (10 ppm,
Millington-Quirk)
Lindane (5 ppm,
Millington-Quirk)
LJndane (pooled data,
Millington-Quirk)
Dieldrin 10 ppm
Dieldrin 5 ppm
Dieldrin (10 ppm,
Millington-Quirk)
Dieldrin (5 ppm,
Millington-Quirk)
Dieldrin (pooled data,
Millington-Quirk
N
4
4
4
4
8
7
7
7
7
14
Correlation
coefficient
0.998
0.99
0.99
0.99
0.93
0.98
0.99
0.98
0.99
0.97
Mean ratio:
Modeled-to-
measured
5.6
6.2
3.7
4.1
3.9
1.5
1.4
1.1
1.0
1.0
p-value
0.0163
0.0414
0.0333
0.0277
0.0006
0.0023
0.0004
0.8367
0.4327
0.4133
95%
confidence
interval
(1.8, 17)
(2.0, 19)
(1-2,11)
(1.3, 13)
(2.3, 6.7)
(1.2, 1.8)
(1.2, 1.6)
a
a
a
'p-value > 0.05 indicates no
measured values; therefore,
significant difference between the average of modeled and
no 95% confidence interval is calculated.
0 A control pile that was not moved or treated
0 An "aerated" or "mechanically mixed" pile
0 A soil pile simulating soil venting or vacuum extraction
0 A soil pile heated to 38° C.
Losses due to volatilization during the mixing process reduced the residual BTEX
in soil. For the purpose of this validation study, however, these losses ensured that
initial soil concentrations of benzene, toluene, and ethylbenzene would be below or
26
-------
within a factor of two of their respective saturation concentrations. Because the mixed
pile, vented pile, and heated pile were subject to mechanical disturbances or thermal
treatment, only the control pile data were used in this study.
In general, the test schedule called for collection of soil samples and air emission
loss measurements during the first, sixth, and seventh weeks. Soil samples were
»
collected randomly within specified grid areas by composite core collection to the
maximum depth of the pile. Emission losses were measured similarly using an
emission isolation flux chamber as specified in Kienbusch (1985). Only data for which
soil samples and flux chamber measurements were taken on the same day were used
for this study.
In addition to total soil concentrations of BTEX, soils were analyzed for moisture
content and bulk density. Over the duration of the experiment, soil moisture, 0m,
averaged 10 percent by weight; bulk density, P., was measured at 1.5 g/cm3.
Default values given by the RAGS/HHEM, Part B were used in the Hwang and
Falco model for those variables which either could not be calculated from site-specific
data or for which site data were not collected. Particle density, p, was thus set equal to
2.65 g/cm3; total porosity, E, was calculated as E = 1-P7/p; air-filled porosity, P., was
calculated as P. = E- QmPi, and the soil organic carbon content fraction was set equal
to 0.02. The Henry's Law constant and solubility in water of each compound are from
EPA (1986), and the soil/water partition coefficient for each compound, Kj, was
calculated as the product of the soil organic carbon content fraction and the organic
carbon partition coefficient, K^, from EPA (1986). Diffusivity in air, Q, for each
compound is from Equation No. 2-4 of EPA (1988).
Analysis of BTEX in soil samples was accomplished by employing the EPA 5030
extraction method and the EPA 8020 analytical method. The BTEX method was
modified to reduce the sample hold time to one day in an effort to improve the
accuracy of the method. Five soil samples were submitted in duplicate. The relative
percent differences (RPD) ranged from 8.0 to 48.9 percent. The average RPD for the
five samples was 26.8 percent. In addition, EPA QC sample analysis indicated average
percent recoveries ranging from 89 percent for m-xylene to 119 percent for toluene.
27
-------
The pooled coefficient of variation (CV) for all the BTEX analysis was 10.5 percent.
Spiked sample recoveries (eight samples) ranged from 75 percent for m-xylene to 168
percent for toluene. The average spike recoveries ranged from 108 percent for
benzene to 146 percent for toluene. Finally, both system blanks and reagent blanks
indicated no contamination was found in the analytical system.
It should be noted that the standard method used for BTEX analysis was
observed to have contributed to the variabilities in soil concentrations. The EPA
acceptance criteria based on 95 percent confidence intervals from laboratory studies
are roughly 30 to 160 percent for the BTEX compounds during analysis of water
samples. The necessary extraction step for soil samples would increase this already
large variability.
Analysis of vapor phase organic compounds via the emission isolation flux
chamber was accomplished using a gas chromatograph (GC). Gas samples were
collected from the flux chamber in 100 ml, gas-tight syringes and analyzed by the GC
in laboratory facilities adjacent to the test site. During the study, a multicomponent
standard was analyzed daily to assess the precision and daily replication of the
analytical system. The results of the analysis indicated a good degree of reproducibility
with coefficients of variation ranging from 5.1 to 16.3 percent.
From these data, instantaneous emission rates were calculated for benzene,
toluene, and ethylbenzene corresponding to each time period at which flux chamber
measurements were made. Appendix B contains the spreadsheet data for benzene,
toluene, and ethylbenzene at initial soil concentrations of 110 ppm, 880 ppm, and 310
ppm, respectively.
Employing the default assumption of dry soil (i-e., D, = D, • E°33), Figures 7, 8,
and 9 show the comparison of modeled and measured emission rates for benzene,
toluene, and ethylbenzene, respectively. The Radian Corporation study noted that the
second measured value in each figure represented a data outlier, possibly due to the
formation of a soil fissure, reducing the soil path resistance and increasing the emission
flux.
28
-------
0)
I to
c y
o o
Q)
§
N
C
0)
CD
1.8-
1.6-
1.4-
1.2-
(/)(/) '
•F
-------
CO
C
(D
8H
7
6
o
To r
rr V
c w *
o o 5
CO
0) 4.
LU
l —
3
2
•H
0
Soil Concentration = 880 ppm
CURVE FIT:
y (predicted) = 4.135E-04 * x~ -0.4990
y (measured) = 0.1198E-03 * 0.9976 ~ x * y ~ -0.4163
-CD
0 100 200 300 400 500 600 700 800 900 10001100
Time From Sampling (hrs)
Measured Predicted
D
Figure 8. Predicted and measured emission rates of toluene versus time.
-------
1.6
CO
rr
§
co
E
g
co 0.8
LU 0)
o E
g E 0.6
N
0.4-
LU
0.2-
0
0
Soil Concentration = 310 ppm
CURVE FIT:
y (predicted) = 7.24790E-39 * e~ ((lnx-308.0597) ^2/1212.2352)
y (measured) = 0.0007 * 3.24201 E-13 ~ (1 /x) * x ~ -1.1464
-Q-
100 200 300 400 500
Time From Sampling (hrs)
600
700
Measured Predicted
D
Figure 9. Predicted and measured emission rates of ethylbenzene versus time.
-------
Table 2 presents the results of the statistical analysis of the comparison of
modeled and measured values using the default Hwang and Falco effective diffusion
coefficient. For both benzene and ethylbenzene, measured values were below the
detection limits after the fifth observation; measured values for toluene were below the
detection limit after the seventh observation.
TABLE 2. SUMMARY OF STATISTICAL ANALYSIS OF PILOT-SCALE VALIDATION
USING THE DEFAULT HWANG AND FALCO EFFECTIVE DIFFUSION COEFFICIENT
Chemical
Benzene (110 ppm)
Ethylbenzene (310 ppm)
Toluene (880 ppm)
Pooled data
N
5
5
7
17
Correlation
coefficient
0.81
0.98
0.95
0.80
Mean ratio:
Modeled-to-
measured
3.0
3.6
2.9
4.3
p-value
0.0580
0.0050
0.0039
0.0001
95%
confidence
level
(0.94, 9.8)
(1.9,6.9)
(2.3, 16)
(2.7, 6.8)
Figure 10 shows the comparison of the log-transformed data for the modeled
and measured emission rates of the three compounds indicating some degree of
variability from the expected straight-line exponential decay. As can be seen from
Table 2, correlation coefficients ranged from.XD.81 for benzene to 0.98 for ethylbenzene,
while p-values and 95 percent confidence intervals indicate a significant statistical
difference between modeled and measured values. Pooled data indicate a correlation
coefficient of 0.80, a mean ratio of 4.3, a p-value of 0.0001, and a 95 percent
confidence interval of 2.7 to 6.8.
By applying the Millington and Quirk expression of effective diffusivity to the
Hwang and Falco model, however, statistical analysis indicates no significant difference
between the average of the modeled and measured emission rates. Table 3 presents
the results of the statistical analysis after applying the Millington and Quirk expression
to the modeled emission rates. As can be seen by these data, p-values of the paired
32
-------
Log Modelled Emission Rate
c
A
O
O
i
A
I
I
A
(A
5"
<•
a
&
A
I
o_
«
A
O"
A
i
l
^k
o
I
-------
TABLE 3. SUMMARY OF STATISTICAL ANALYSIS OF PILOT-SCALE VALIDATION
USING THE MILLINGTON AND QUIRK EXPRESSION FOR EFFECTIVE DIFFUSIVITY
Chemical
Bezene (110 ppm)
Ethylbezene (310
ppm)
Toluene (880 ppm)
Pooled data
N
5
5
7
17
Correlation
coefficient
0.81
0.98
0.95
0.80
Mean ratio:
Model-to-
measured
1.02
1.22
2.04
1.43
p-value
0.9596 '
0.4342
0.1215
0.1164
95%
confidence
interval
a
a
a
a
'p-value >0.05 indicates no significant difference between the average of modeled and
measured values; therefore, no 95% confidence interval is calculated.
Student's t-test for individual compounds and for pooled data are greater than 0.05,
indicating no significant statistical difference.
Figures 11, 12, and 13 show the comparisons of modeled emission rates
employing the Millington and Quirk expression for effective diffusivity and measured
emission rates for benzene, toluene, and ethylbenzene, respectively. Note that in the
case of toluene and ethylbenzene the modeled and measured emission rate curves
intersect as steady-state conditions are achieved; while the curves for the more volatile
benzene indicate that steady-state conditions have been achieved.
34
-------
CO
en
ITT
52
0)
"co
DC
C
p
'co
CO
E
LU
CD
C
N
C
CD
CD
27
1.8-
1.6-
1.4-
x— «^
2 i-2-
o
*~ 1-
co '
CD
| 0.8-
t
0.6-
0.4-
0.2-
0-
"^ - outiyer Soil Concentration = 110 ppm
CURVE
FIT:
y (predicted) = 3.27984E-05 * x" -0.50
y (measured) = 1.05949E-06 + (-8.5871 1E-10) *x + Q.WQMx
3\
A
\ V
\ X.
X. Q~~~~- — _
^-*— — _
I
' • _
1 I 1
0 100 200 300 400
aV ~" ~ t~
~^mm~~~^&— __•
1 I
500 600 700
Time From Sampling (hrs)
Measured Predicted
• D
Figure 11. Predicted and measured emission rates of benzene versus time employing the Millington and Quirk expression of D.,
-------
3.5-
.CO
0
*-J
co ^s
1 k
1 8
Outlyer
Soil Concentration = 880 ppm
E E
LJLJ ;=
0)
c
CD
_3
o
E 1-5-
CURVE FIT:
y (predicted) = 1 .397E-04 * x ~ -0.4999
y (measured) = 0.1198E-03 * 0.9976 ~x *
-0.41 63
0 100 200 300 400 500 600 700 800 900 1000 1100
Time From Sampling (hrs)
Measured Predicted
Figure 12. Predicted and measured emission rates of toluene versus time employing the Mllllngton and Quirk expression of
-------
CO
-si
3
0)
«
cc
5-
4-
1 w 3^
LLJ 0)
Q) E
c F
o Cx 0
N £•
C
CD
1-
0
0
Soil Concentration = 310 ppm
CURVE FIT:
y'(predicted) = 2.37669E-05 * x~ -0.4995
y (measured) = 0.0007 * 3.24201 E-13 ~(1/x) *x/v-1.1464
Outlyer
100
200 300 400 500
Time From Sampling (hrs)
600
700
Measured ------ Predicted
Figure 13. Predicted and measured emission rates of ethylbenzene versus time employing the Millington and Quirk expression of D.,
-------
SECTION 4
PARAMETRIC ANALYSIS OF THE HWANG AND FALCO MODEL
This section presents the results of parametric analysis of the key variables of
the Hwang and Falco volatilization model (Equation No. 1). Because the Hwang and
Falco model is a direct derivation of the model presented in Farmer, et al. (1980), the
parametric observation: of Farmer, et al. are also directly applicable. Hwang and Falco
redefined the general diffusion equation of Farmer, et al. establishing the relationship
between vapor phase diffusion and soil phase adsorption. For this reason, other model
variables related directly to vapor density in interstitial pore spaces and soil adsorption
capacity were further analyzed for this study.
From Farmer, et al., air-filled porosity was found to be the most significant soil
parameter affecting the final steady-state flux through soil. The volumetric water
content of the soil and the soil bulk density determine the air-filled porosity. Other soil
parameters such as soil organic matter content and soil texture were found to have
affected the time for the flux to arrive at steady-state conditions but did not effect the
magnitude of the final flux except as they influenced soil bulk density.
In addition it was found that soil temperature increased volatilization exponentially
due to the effect on the vapor pressure. Farmer, et al. also noted that the chemical
stability and resistance to microbial degradation of the contaminant were directly related
to the maximum rate of flux and the persistence of the contaminant in soil.
4.1 AFFECTS OF SOIL PARAMETERS
In this section, the experimental results of Farmer, et al. (1980) are discussed as
they relate to the effect of soil water content, soil bulk density, air-filled soil porosity,
and temperature on the vapor phase diffusion in soil.
38
-------
So/7 Moisture Content
Farmer, et al. indicates that the effect of soil moisture content on the volatilization
flux of contaminants through soils is exponential. Increasing soil water content
decreases the pore spaces available for vapor diffusion and will decrease volatilization
flux. In contrast, increasing soil water content has also been shown to increase the
volatility of pesticides in soil under certain conditions (Gray, et al., 1965; and Spencer
and Cliath, 1969 and 1970). In essence, the soil water content affects the pesticide
adsorption capacity by competing for soil adsorption sites. Under these conditions, an
increase in soil moisture above a certain point will tend to desorb contaminants,
increasing the flux dependent on the relative water and contaminant adsorption
isotherms.
Bulk Density
Soil compaction or bulk density also determines the porosity of soil and thus
affects the vapor phase diffusion through the soil. Experimental results from Farmer, et
al. indicate that soil bulk density also has an exponential effect on volatilization flux
through the soil. From previous considerations of the effect of soil water content, a
higher bulk density will have similar effects to that of an increased soil moisture content
The higher the soil bulk density, the smaller the steady-state flux.
So/7 Air-Filled Porosity
The effects of soil water content and soil bulk density on volatilization through
soil can be contributed to their effect on the air-filled porosity, which in turn is the major
soil factor controlling volatilization through the soil. The effect of air-filled porosity is
manifested in the expression of the effective diffusion coefficient. The effective diffusion
coefficient, however, does not depend only on the amount of air-filled pore space. The
presence of liquid film on the solid surfaces not only reduces porosity, but also modifies
the pore geometry increasing tortuosity and the length of the gas passage. Use of the
Millington and Quirk expression of the effective diffusion coefficient better accounts for
39
-------
these effects than does the assumption of dry soil. This assertion is verified by this
study.
So/7 Temperature
The effect of soil temperature on the volatilization flux through soil is
«
multifunctional. The diffusion in air, D,, is theoretically related to temperature, T, and the
collision integral, O, in the following manner (Lyman, et at., 1990):
y-05
D (variation of) (16)
Q(f)
The exponential coefficient for temperature varies from 1.5 to 2 over a wide range of
temperatures. Barr and Watts (1972) found that 1.75 gave the best values for gaseous
diffusion. Farmer, et al. (1980) estimates the effective diffusion coefficient at
temperature T2 as :
D2 = D, (Vrj05 (17)
where D2 = Diffusion coefficient at T2
D, = Diffusion coefficient at T,
T = Absolute temperature.
Finally, a temperature increase will effect the vapor pressure function of the
Henry's Law constant used to define the soil/air partition coefficient, K... which causes
an increase in the vapor concentration gradient across the soil layer. In addition, any
additional heat generated inside a landfill due to aerobic decomposition of organic
wastes will have a short-term effect on the temperature of the soil. In actual fact,
temperature gradients will exist across the soil due primarily to seasonal variations.
40
-------
Vapor diffusion is influenced by such gradients; however, these effects of fluctuating
soil temperatures will tend to cancel one another over time. The overall effect of
temperature on volatilization flux can be approximated by use of Equation No. 17 and
an average soil temperature.
4.2 AFFECTS OF NONSOIL PARAMETERS
The remaining variables in the Hwang and Falco model that are not related to
soil properties are the initial soil concentration, C.0, the time from sampling, t, the
soil/water partition coefficient, K^, and the summation expression in the original Farmer,
et al. flux equation (Equation No. 6).
Initial Soil Concentration
The effect of change in the initial soil concentration term in the Hwang and Falco
model is linear; i.e., an increase in C,0 of 100 percent causes an increase in the
emission rate of 100 percent. Probably the greatest degree of uncertainty in the value
of Cso is likely to be either insufficient soil sampling to adequately characterize site soil
concentrations, or the variability in percent recovery of contaminants as it applies to
existing sampling and analysis methods for organic compounds in soils. Typically,
present extraction and analysis method recovery variability increases the likelihood of
under prediction of the emission rate (i.e., more contaminant is present in the soil than
is reported by sampling and analysis methods).
Time From Sampling
The time variable, t, is an exponential rate operative in the Hwang and Falco
model. When emission rates computed from the model are plotted against the
reciprocal of t °5, a straight line is obtained; therefore, when the emission rates are
plotted against t, an asymptotic curve results. In this manner, the steady-state
volatilization rate remains constant while the soil concentration is reduced exponentially
(i.e., never reaching zero).
41
-------
Soil/Water Partition Coefficient
The RAGS/HHEM, Part B defines the soil/water partition coefficient, 1^, in the
Hwang and Falco model as the product of the organic carbon partition coefficient, K^,
and the soil organic carbon fraction, OC. By basing the soil/water partition coefficient
«
on soil organic carbon rather than on total mass, most of the variation in sorption
coefficients between different soils is eliminated. The remaining variation may be due to
other characteristics of soil such as clay content and surface area, cation exchange
capacity, pH, soil moisture salinity, concentration of dissolved organic matter in soil
water, and nonlinear adsorption isotherms (Lyman, et al., 1990). In the Hwang and
Falco model, the effect of a change in the value of ^ on the emission rate is identical
to that of a change in time, t, such that by holding all other variables constant,
increases in the value of KU will generate the same asymptotic emission rate curve as
increases in the value of t. For this reason, the model is very sensitive to variations in
the value of K^.
Farmer, et al., Summation Expression
Finally, the initial diffusion model (Equation No. 6) from Farmer, et al. (1974)
includes a summation expression which is eliminated by Hwang and Falco by assuming
L = oo. in actuality, however, the depth of the contaminated soil column is finite.
Farmer, et al. explained that the reduced equation diffusion model boundary conditions
were valid until t > L2/14.4 D (Equation No. 9). The same limitation applies to the
Hwang and Falco model at t > L2/14.4 a. Therefore, by derivation, the Hwang and
Falco model equivalent to the Farmer, et al., original solution is:
N. =
a 0°
H_
oo
exp(-n2L2/a f)
(18)
42
-------
If the value of the summation term in Equation No. 18 is negligible in comparison to 1,
the equation can be reduced. The summation term will be small if the expression in the
exponential, n2L2/at, js large, say on the order of 10 or more. The expression rrV/at
is largely dependent on the value of L. For example, from the lindane experiments of
Farmer, et al. (1972), the boundary conditions of the Hwang and Falco model are
violated for a soil depth, L, of 0.5 cm at t = 11.6 hours (Appendix A, "t Max").
However, from the BTEX experiments of Radian Corporation (1989), the boundary
conditions of the model are not violated for ethylbenzene at L = 91 cm until t = 19,204
hours or 2.19 years (see Appendix B, "t Max").
If the Hwang and Falco model is applied in its reduced form to ethylbenzene
under the conditions specified in Radian Corporation (1989), the instantaneous
emission rate for the default exposure interval given in RAGS/HHEM, Part B (7.9 x 108
seconds or 30 years) is 2.26 x 10 '8 g/s. In contrast, if the non reduced solution
(Equation No. 18) is applied, the instantaneous emission rate for the same exposure
interval is 1.00 x 10 "8 g/s, reducing the original solution by a factor of 0.444. It can
then be assumed that over the default exposure period, use of the reduced equation
will tend to overpredict the instantaneous and average emission rates by a factor of
approximately two.
43
-------
SECTION 5
CONCLUSIONS
From the results of this study, rt can be concluded that for the compounds
included in the experimental data, there are no statistically significant differences
between the measured emission rates and those predicted by the Hwang and Falco
volatilization model employing the Millington and Quirk expression of effective diffusivity.
These results are valid if the initial soil concentration is within a factor of 10 or below the
theoretical soil saturation concentration, C,.,, and if the initial and boundary conditions
of the model are reasonably attained under bench-scale and pilot-scale conditions.
The initial and boundary conditions of the model specify an infinitesimal layer of
uncontaminated soil at the soil/air interface (Q = 0, at L = 0, t > 0) and an infinite
contaminated soil column (Q = H/I^)C.0, at L = «, t > 0). In this manner, the model
will tend to underpredict emissions before steady-state conditions are achieved as the
contaminants at the soil surface are depleted. By assuming an infinite contaminated
soil column, the model employs the reduced form of the original Farmer, et al. solution.
With a finite contaminated soil column of approximately 100 cm in depth, however, the
model will tend to overpredict the average emission rate over extended periods of time
(e.g., 30 years) by a factor of approximately two.
As applied in the RAGS/HHEM, Part B, the Hwang and Falco volatilization model
employs the conservative assumption of dry soil when calculating effective diffusivity in
soil (D.j = D, • E:3'33). The results of this study indicate that this assumption tends to
overpredict average emissions by approximately one order of magnitude. On the other
hand, use of the Millington and Quirk expression of effective diffusivity
[D(. = D, (P.333/Ff)] indicates no statistically significant difference between modeled
and measured emission rates.
44
-------
In its present form, the Hwang and Falco volatilization model indicates good
agreement between predicted and measured emission rates under controlled
conditions. The following assumptions and simplications, however, have been made:
1. Q and C,0 are related by Henry's Law precluding the presence of
nonaqueous phase contaminants (i.e., free phase contaminants).
2. The soil/air partition coefficient (K^. = H/Ky) does not take into account
the effects of the relative difference of the Henry's law constant, H, at
temperatures for which values of H are experimentally derived and those
of in situ soils. In addition, the effects of multicomponent systems on
effective solubilities of contaminants is not addressed in the value of the
Henry's Law constant.
3. The soii/air partition coefficient, J^., does not account for the effects of
other soil and soil water properties (e.g., clay/silt content, water salinity,
pH, etc.) on the value of the soil/water partition coefficient, fV Because
K^ is the product of the organic carbon partition coefficient, K^, and the
soil organic carbon content, OC, the environmental system in which ^ is
to be used must not differ significantly from the one implied by the
experimental conditions in which the values of K,,,, are derived.
4. The model assumes no mass flow of contaminants due to water
movement in the soil. Mass flow due to capillary action or redistribution
of contaminates due to rain events may be significant if applicable to site-
specific condition.
5. The model assumes that diffusion is vapor-phase controlled. Compounds
for which solution-phase diffusion is significant are not addressed.
6. Chemical stability and biodegradation are not considered in the model.
Compounds that are chemically unstable or are subject to accelerated
microbial degradation will exhibit markedly reduced steady-state emission
rates than predicted values.
7. The model is valid only if the effective diffusion coefficient in soil is
constant. This assumes isotropic soils and completely homogeneous
incorporation of contaminants.
8. The model assumes sufficient air velocity at the soil surface to completely
volatilize all contaminants. Soil surfaces in sheltered areas may be
subject to limited air velocity and turbulence resulting in violation of the
model's boundary conditions.
45
-------
9. The model assumes no movement of contaminants occur across the
lower boundary (e.g., leaching).
Emission rates predicted by the Hwang and Falco volatilization model indicate
good correlation to measured emission rates under controlled condition, but predicted
values for field conditions would be subject to error because the boundary conditions
and environmental conditions such as wind velocity, incorporation depth, and water
movement are not as well defined as they are in the laboratory or under pilot-scale
conditions. Nonetheless, results of this study indicate that the model should make
reasonable estimates of loss by vapor-phase diffusion.
46
-------
REFERENCES
Balfour, W. D., B. M. Eklund, and S. J. Williamson. Measurement of Volatile Organic
Emissions from Subsurface Contaminants. In Proceedings of the National Conference
on Management of Uncontrolled Hazardous Waste Sites. September 1984, pp. 77-81.
Hazardous Materials Control Research Institute, Silver-Springs, Maryland.
Barr, R. F. and H. F. Watts. 1972. Diffusion of Some Organic and Inorganic
Compounds in Air. J. Chem. Eng. Data 17:45-46.
Carslaw, H. S., and J. C. Jaeger. 1959. Conduction of Heat in Solids. 2* Edition
Oxford University Press, Oxford.
Crank, J. 1985. The Mathematics of Diffusion. Oxford University Press, New York.
U.S. Environmental Protection Agency. 1986. Superfund Public Health Evaluation
Manual. Office of Emergency and Remedial Response. EPA-540/1-86-060.
U.S. Environmental Protection Agency. 1986a. Development of Advisory Levels for
Potychlorinated Biphenyls (PCBs) Cleanup. Office of Health and Environmental
Assessment. EPA-600/6-86-002.
U.S. Environmental Protection Agency. 1988. Superfund Exposure Assessment
Manual. Office of Emergency and Remedial Response. EPA-540/1-88-001.
U.S. Environmental Protection Agency. 1990. Procedures for Conducting Air Pathway
Analyses for Superfund Activities, Interim Final Documents: Volume 2 - Estimation of
Baseline Air Emissions at Superfund Sites. Office of Air Quality Planning and
Standards. EPA-450/1-89-002a.
U.S. Environmental Protection Agency. 1991. Database of Emission Rate Measurement
Projects - Technical Note. Office of Air Quality Planning and Standards. EPA-450/1-
91-003.
Farmer, W. J., K. Igue, W. F. Spencer, and J. P. Martin. 1972. Volatility of
Organochlorine Insecticides from Soil. I and II Effects. So/7 Sci. Soc. Amer. Proc.
36:443-450.
47
-------
Farmer, W. J., and J. Letey. 1974. Volatilization Losses of Pesticides From Soils.
Office of Research and Development. EPA-660/2-74/054.
Farmer, W. J., M. S. Yang, J. Letey, and W. F. Spencer. 1980. Land Disposal of
Hexachlorobenzene Wasfes. Office of Research and Development. EPA-600/2-80/119.
Gray, R. A., and A. J. Weierch. 1965. Factors Affecting the Vapor Los$ of EPIC from
Soil. Weeds 13:141-147.
Hartley, G. S. 1964. Herbicide Behavior in the Soil. I. Physical Factors and Action
Through the Soil. P. 111-161. In LJ. Audus (ed.) The Physiology and Biochemistry of
Herbicides. Academic Press, London and New York.
Hill, F. B., V. P. Aneja, and R. M. Felder. 1978. A Technique for Measurement of
Biogenic Sulfur Emission Fluxes. J. Env. Sci. Health AIB (3), pp. 199-225.
Hwang, S. T., and J. W. Falco. 1986. Estimation of Multimedia Exposure Related to
Hazardous Waste Facilities. Cohen, Y. (ed). Plenum Publishing Corp.
Kienbusch, M. Measurement of Gaseous Emission Rates from Land Surfaces Using an
Emission Isolation Flux Chamber • User's Guide. Report to EPA-EMSL, Las Vegas
under EPA Contract No. 68-02-3889, Work Assignment No. 18, December 1985.
Kienbusch, M. and D. Ranum. 1986. Validation of Flux Chamber Emission
Measurements On a Soil Surface - Draft Report to EPA-EMSL, Las Vegas, Nevada.
Kienbusch, M., W. D. Balfour, and S. Williamson. The Development of an Operations
Protocol for Emission Isolation Flux Chamber Measurements on Soil Surfaces.
Presented at the 79th Annual Meeting of the Air Pollution Control Association (Paper
86-20.1), Minneapolis, Minnesota, June 22-27, 1986.
Lyman, W. J., W. F. Reehl, and D. H. Rosenblatt. 1990. Handbook of Chemical
Property Estimation Methods. American Chemical Society, Washington, D.C.
Millington, R. J., and J. M. Quirk. 1961. Permeability of Porous Solids. Trans. Faraday
Soc. 57:1200-1207.
Radian Corporation. So/7 Gas Sampling Techniques of Chemicals for Exposure
Assessment - Data Volume. Report to EPA-EMSL, Las Vegas under EPA Contract No.
68-02-3513, Work Assignment No. 32, March 1984.
Radian Corporation. Short-term Fate and Persistence of Motor Fuels in Soils. Report to
the American Petroleum Institute, Washington, D.C. July 1989.
48
-------
Spencer, W. F., M. Cliath, and W. J. Farmer. 1969. Vapor Density of Soil Applied
HEOD as Related to Soil Water Content, Temperature, and HEOD Concentration. Soil
Sci. Soc. Amer. Proc. 33:509-511.
Spencer, W. F., and M. Cliath. 1970. Vapor Density and Apparent Vapor Pressure of
Lindane (x-BHC). J. Agr. Food Chem. 18:529-530.
Spencer, W. F. 1970. Distribution of Pesticides Between Soil, Water and Air. In
Pesticides in the Soil: Ecology, Degradation and Movement. A symposium, February
25-27, 1970. Michigan State University.
49
-------
-------
APPENDIX A
BENCH-SCALE MODEL VALIDATION DATA
A-1
-------
-------
Comparison of Farmer et al. (1 972 and
Chemical
Lindane
Lindane
Lindane
Lindane
Lindane
Lindane
Lindane
Lindane
1974) Measured Emission Rates to Hwang and Falco Model Predicted Emission Rates
Contaminated
Cso
(ug/g)
10
10
10
10
5
5
5
5
Emitting
Area
(cm2)
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
Soil
Depth
(m)
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
Soil
Type
Qila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Bulk
Density
(g/cm3)
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
Particle
Density
(g/cm.3)
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
Soil
Moisture
(Wt. %)
10
10
10
10
10
10
10
10
Solubility,
s
(g/cm3)
7.80
7.80
7.80
7.80
7.80
7.80
7.80
7.80
Organic
Carbon,
OC
(fraction)
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
Saturation,
Csat
(ug/g)
5.7
5.7
5.7
5.7
5.7
5.7
5.7
5.7
Cso = initial soil concentration
Koc = organic carbon partition coefficient
Kd = soil/water partition coefficient
Di = diffusivity in air
H = Henry's Law constant
t = time from sampling
Pt = total soil porosity
Pa = air-filled soil porosity
Del » effective diffusion coefficient assuming dry soil
Dei (MQ) = effective diffusion coefficient using the Millington and Quirk expression
alpha = (Dei x E)/[E + Ps x (1-E) x Kd/H)]
alpha (MQ) = alpha using the Millington and Quirk expression of Dei
Predicted Emission Rate = modeled values assuming dry soil
Predicted Emission Rate, (MQ) = modeled values using the Millington and Quirk expression of Dei
tMAX= L~ 2/14.4 x alpha
-------
Measured
Emission
Flux
(ng/cm2-day)
1160
320
140
90
500
160
60
40
KOC
(cm3/g)
1080
1080
1080
1080
1080
1080
1080
1080
Kd
(cm3/g)
6.26
6.26
6.26
6.26
6.26
6.26
6.26
6.26
Kas
(g/cm3)
0.000051
0.000051
0.000051
0.000051
0.000051
0.000051
0.000051
0.000051
Di
(cm2/s)
5.43E-02
5.43E-02
5.43E-02
5.43E-02
5.43E-02
5.43E-02
5.43E-02
5.43E-02
Dei
(cm2/s)
4.69E-02
4.69E-02
4.69E-02
4.69E-02
4.69E-02
4.69E-02
4.69E-02
4.69E-02
Dei
(MQ)
(cm2/s)
2.08E-02
2.08E-02
2.08E-02
2.08E-02
2.08E-02
2.08E-02
2.08E-02
2.08E-02
t
t
H Cumulative Cumulative Pt
(atm-m3/mol)
7.85E-06
7.85E-06
7.85E-06
7.85E-06
7.85E-06
7.85E-06
7.85E-06
7.85E-06
(sec)
86400
259200
432000
604800
86400
259200
432000
604800
(hrs)
24
72
120
168
24
72
120
168
(unitless)
0.717
0.717
0.717
0.717
0.717
0.717
0.717
0717
-------
Pa
(unitless)
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
alpha
(cm2/s)
9.73E-07
9.73E-07
9.73E-07
9.73E-07
9.73E-07
9.73E-07
9.73E-07
9.73E-07
alpha
(MQ)
(cm2/s)
4.32E-07
4.32E-07
4.32E-07
4.32E-07
4.32E-07
4.32E-07
4.32E-07,
4.32E-07
Predicted
Emission
Rate
(9/s)
8.30E-10
4.79E-10
3.71 E-10
3.14E-10
4.15E-10
2.39E-10
1.86E-10
1.57E-10
Measured
Emission
Rate
(9/s)
3.70E-10
1.02E-10
4.46E-11
2.87E-1 1
1.59E-10
5.10E-11
1.91E-11
1 .28E-1 1
Predicted
Emission
Rate, (MQ)
(g/s)
5.53E-10
3.19E-10
2.47E-10
2.09E-10
2.76E-10
1.60E-10
1.24E-10
1.04E-10
Cso>Csat
(Yes/No)
Yes
Yes
Yes
Yes
No
No
No
No
tMAX
(hrs)
11.16
11.16
11.16
11.16
11.16
11.16
11.16
11.16
t>tMAX
(Yes/No)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
-------
Comparison of Farmer et al. (1972 and 1974) Measured Emission
Chemical
Dieldrin
Dieldrin
Dieidrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Dieldrin
Cso
(ug/g)
10
10
10
10
10
10
10
5
5
5
5
5
5
5
Contaminated
Emitting Soil
Area Depth Soil
(cm2) (m) Type
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
27.55
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Giia silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Gila silt loam
Rates to
Bulk
Density
(9/cm3)
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
Hwang and Faico Model Predicted Emission
Particle
Density
(g/cm3)
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
Soil
Moisture
(Wt. %)
10
10
10
10
10
10
10
10
10
10
10
10
10
10
Solubility,
s
(g/cm3)
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
0.195
Organic
Carbon,
OC
(fraction)
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
0.0058
Rates
Saturation,
Csat
(ug/g)
02
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Cso = initial soil concentration
Koc = organic carbon partition coefficient
Kd = soil/water partition coefficient
Di = diffusivity in air
H = Henry's Law constant
t = time from sampling
R = total soil porosity
Pa = air-filled soil porosity
Dei = effective diffusion coefficient assuming dry soil
Dei (MQ) = effective diffusion coefficient using the Millington and Quirk expression
alpha = (Dei x E)/[E + Ps x (1-E) x Kd/H)]
alpha (MQ) = alpha using the Millington and Quirk expression of Dei
Predicted Emission Rate = modeled values assuming dry soil
Predicted Emission Rate, (MQ) = modeled values using the Millington and Quirk expression of Dei
tMAX = L~ 2/14.4 x alpha
-------
Pa
(unitless)
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
alpha
(cm2/s)
3.24E-08
3.24E-08
3.24E-08
3.24R-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-03
3.24E-08
3.24E-08
alpha
(MQ)
(cm2/s)
1 .67E-08
1 .67E-08
1.67E-08
1.67E-08
1 .67E-08
1.67E-08
1 .67E-08
1.67E-08
1 .67E-08
1.67E-08
1 .67E-08
1 .67E-08
1 .67E-08
1 .67E-08
Predicted
Emission
Rate
(9/s)
1.51E-10
8.74E-1 1
6.77E-11
6.18E-11
5.72E-11
5.05E-1 1
4.37E-11
7.57E-11
4.37E-11
3.39E-11
3.09E-11
2.86E-11
2.52E-1 1
2.19E-11
Measured
Emission
Rate
(9/s)
1.28E-10
8.29E-11
4.46E-11
3.51 E-11
3.35E-1 1
2.87E-1 1
2.71 E-11
6.38E-11
3.67E-11
2.39E-11
2.07E-11
1.91 E-11
1.75E-11
1.28E-11
Predicted
Emission
Rate, (MQ)
(9/s)
1.09E-10
6.27E-1 1
4.85E-11
4.43E-11
4.10E-11
3.62E-11
3.13E-11
5.43E-1 1
3.13E-11
2.43E-1 1
2.22E-11
2.05E-1 1
1.81 E-11
1 .57E-1 1
CsoCsat
(Yes/No)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
tMax
(his)
290
290
290
290
290
290
290
290
290
290
290
290
290
290
t>tMAX
(Yes/No)
No
No
No
No
No
No
No
No
No
No
No
No
No
No
-------
Measured
Emission
Flux Koc
(ng/cm2-day) (cm3/g)
400
260
140
110
105
90
85
200
115
75
65
60
55
40
1700
1700
1700
1700
1700
1700
1700
1700
1700
1700
1700
1700
1700
1700
Kd
(cm3/g)
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
9.86
Kas
(g/cm3)
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
0.000002
Di
(cm2/s)
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
4.88E-02
Dei
(cm2/s)
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
4.22E-02
Dei
(MQ) H
(cm2/s) (atm-m3/mol)
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
2.17E-02
4.
4.
4.
58E-07
58E-07
58E-07
4.58E-07
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
58E-07
58E-07
58E-07
58E-07
58E-07
58E-07
58E-07
58E-07
58E-07
58E-07
t t
Cumulative Cumulative Pt
(sec) (hrs) (unitless)
86400
259200
432000
518400
604800
777600
1036800
86400
259200
432000
518400
604800
777600
1036800
24
72
120
144
168
216
288
24
72
120
144
168
216
288
0.717
0.717
0.717
0.717
0.717
0.717
0717
0.717
0.717
0.717
0.717
0.717
0.717
0.717
-------
Pa
(unitless)
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
0.642
alpha
(cm2/s)
3.24E-08
3.24E-08
3.24E-08
3.24R-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-08
3.24E-03
3.24E-08
3.24E-08
alpha
(MQ)
(cm2/s)
1.67E-08
1.67E-08
1.67E-08
1.67E-08
1 .67E-08
1 .67E-08
1 .67E-08
1 .67E-08
1.67E-08
1.67E-08
1.67E-08
1 .67E-08
1 .67E-08
1 .67E-08
Predicted
Emission
Rate
(9/s)
1.51E-10
8.74E-11
6.77E-11
6.18E-11
5.72E-1 1
5.05E-11
4.37E-11
7.57E-11
4.37E-11
3.39E-11
3.09E-11
2.86E-11
2.52E-11
2.19E-11
Measured
Emission
Rate
(9/s)
1.28E-10
8.29E-11
4.46E-11
3.51 E-11
3.35E-1 1
2.87E-11
2.71 E-11
6.38E-11
3.67E-11
2.39E-11
2.07E-11
1.91 E-11
1 .75E-1 1
1 .28E-1 1
Predicted
Emission
Rate, (MQ)
(g/s)
1.09E-10
6.27E-1 1
4.85E-11
4.43E-11
4.10E-11
3.62E-11
3.13E-11
5.43E-1 1
3.13E-11
2.43E-11
2.22E-11
2.05E-1 1
1.81 E-11
1 .57E-1 1
Cso>Csat
(Yes/No)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
tMax
(hrs)
290
290
290
290
290
290
290
290
290
290
290
290
290
290
t>tMAX
(Yes/No)
No
No
No
No
No
No
No
No
No
No
No
No
No
No
-------
-------
APPENDIX B
PILOT-SCALE MODEL VALIDATION DATA
B-1
-------
-------
Comparison of Radian Corporation (1989) Measured Emission Rates to Hwang and Falco Model
Chemical
Ethylbenzene
Ethylbenzene
Ethylbenzene
Ethylbenzene
Ethylbenzene
Ethylbenzene
Benzene
Benzene
Benzene
Benzene
Benzene
Benzene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
Sample
Point
3
4
5
6
7
8
3
4
5
6
7
8
3
4
5
6
7
8
9
Cso
("9/9)
310
310
310
310
310
310
110
110
110
110
110
110
880
880
880
880
880
880
880
Flux Contaminated
Chamber Soil
Area Depth Soil
(m2) (m) Type
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.13
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Loamy sand
Bulk
Density
(9/cm3)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
Particle
Density
(9/cm3)
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
2.65
Predicted Emission Rates
Soil
Moisture
(Wt. %)
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
Solubility,
s
(g/cm3)
1 52.000
152.000
1 52.000
152.000
1 52.000
1 52.000
1750.000
1750.000
1750.000
1750.000
1750.000
1750.000
535.000
535.000
535.000
535.000
535.000
535.000
535.000
Organic
Carbon,
OC
(fraction)
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Cso = initial soil concentration
Koc = organic carbon partition coefficient
Kd = soil/water partition coefficient
Di = diffusivity in air
H = Henry's Law constant
t = time from sampling
Pt •» total soil porosity
Pa = air-filled soil porosity
Dei = effective diffusion coefficient assuming dry soil
Dei (MQ) = effective diffusion coefficient using the Millington and Quirk expression
alpha = (Dei x E)/[E + Ps x (1 -E) x Kd/H)]
alpha (MQ) = alpha using the Millington and Quirk expression of Dei
Predicted Emission Rate = modeled values assuming dry soil
Predicted Emission Rate, (MQ) = modeled values using the Millington and Quirk expression of Dei
tMAX = L " 2/14.4 x alpha
-------
Saturation,
Csat
(ug/g)
349.6
349.6
349.6
349.6
349.6
349.6
465.5
465.5
465.5
465.5
465.5
465.5
374.5
374.5
374.5
374.5
374.5
374.5
374.5
Measured
Emission
Flux
(ug/m2-min)
2640
1700
1080
250
180
0
2760
9000
910
400
290
0
14800
17300
4910
1340
830
340
260
KOC
(cm3/g)
1100
1100
1100
1100
1100
1100
83
83
83
83
83
83
300
300
300
300
300
300
300
Kd
(cm3/g)
22.00
22.00
22.00
22.00
22.00
22.00
1.66
1.66
1.66
1.66
1.66
1.66
6.00
6.00
6.00
6.00
6.00
6.00
6.00
Kas
(g/cm3)
0.011983
0.011983
0.011983
0.011983
0.011983
0.011983
0.138066
0.138066
0.138066
0.138066
0.138066
0.138066
0.043528
0.043528
0.043528
0.043528
0.043528
0.043528
0.043528
Di
(cm2/s)
0.0667
0.0667
0.0667
0.0667
0.0667
0.0667
0.0871
0.0871
0.0871
0.0871
0.0871
0.0871
0.0783
0.0783
0.0783
0.0783
0.0783
0.0783
0.0783
Dei
(cm2/s)
0.0472
0.0472
0.0472
0.0472
0.0472
0.0472
0.0616
0.0616
0.0616
0.0616
0.0616
0.0616
0.0554
0.0554
0.0554
0.0554
0.0554
0.0554
0.0554
Dei
(MQ) H
(cm2/s) (atm-m3/mol)
0.005333
0.005333
0.005333
0.005333
0.005333
0.005333
0.006964
0.006964
0.006964
0.006964
0.006964
0.006964
0.00626
0.00626
0.00626'
0.00626
0.00626
0.00626
0.00626
0.00643
0.00643
0.00643
0.00643
0.00643
0.00643
0.00559
0.00559
0.00559
0.00559
0.00559
0.00559
0.00637
0.00637
0.00637
0.00637
0.00637
0.00637
0.00637
t
Cumulative
(sec)
86640
266100
422640
1816200
2506380
3099000
86640
266100
422640
1816200
2506380
3099000
86640
266100
422640
1816200
2506380
3099000
3689400
t
Cumulative
(hrs)
24
74
117
505
696
861
24
74
117
505
696
861
24
74
117
505
696
861
1025
-------
Pt
(unitless)
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
0.434
Pa
(unitless)
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
0.284
alpha
(cm2/s)
7.36E-05
7.36E-05
7.36E-05
7.36E-05
7.36E-05
7.36E-05
1.11E-03
1.11E-03
1.11E-03
1.11E-03
1.11E-03
1.11E-03
3.14E-04
3.14E-04
3.14E-04
3.14E-04
3.14E-04
3.14E-04
3.14E-04
alpha
(MQ)
(cm2/s)
8.
8.
8
32E-06
32E-06
32E-06
8.32E-06
8.32E-06
8.32E-06
1.25E-04
1 .25E-04
1.25E-04
1 .25E-04
1.25E-04
1
3
3
3
3
3
3
3
.25E-04
.55E-05
.55E-05
.55E-05
.55E-05
.55E-05
.55E-05
.55E-05
Predicted
Emission
Rate
(g/s)
1 .45E-05
8.25E-06
6.54E-06
3.16E-06
2.69E-06
2.42E-06
1 .99E-05
1.14E-05
9.01 E-06
4.35E-06
3.70E-06
3.33E-06
8.47E-05
4.83E-05
3.84E-05
1 .85E-05
1 .58E-05
1 .42E-05
1 .30E-05
Measured
Emission
Rate
(9/s)
5.72E-06
3.68E-06
2.34E-06
5.42E-07
3.90E-07
O.OOE+00
5.98E-06
1.95E-05
1 .97E-06
8.67E-07
6.28E-07
O.OOE+00
3.21 E-05
3.75E-05
1 .06E-05
2.90E-06
1 .80E-06
7.37E-07
5.63E-07
Predicted
Emission
Rate.(MQ) Cso>Csat
(g/s) (Yes/No)
4.
2.
2.
1.
9.
8.
86E-06
77E-06
20E-06
06E-06
04E-07
13E-07
6.69E-06
3.
3
1.
82E-06
03E-06
46E-06
1.24E-06
1.12E-06
2.85E-05
1 .63E-05
1 .29E-05
6
22E-06
5.30E-06
4.76E-06
4
37E-06
No
No
No
No
No
No
No
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
tMAX
(hrs)
19204
19204
19204
19204
19204
19204
1276
1276
1276
1276
1276
1276
4504
4504
4504
4504
4504
4504
4504
t>tMAX
(Yes/No)
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
-------
-------
APPENDIX C
EVALUATION OF THE DISPERSION EQUATIONS
IN THE RISK ASSESSMENT GUIDANCE FOR
SUPERFUND (RAGS): VOLUME I - HUMAN
HEALTH EVALUATION MANUAL (PART B,
DEVELOPMENT OF PRELIMINARY REMEDIATION GOALS)
-------
-------
EVALUATION OF THE DISPERSION EQUATIONS
IN THE RISK ASSESSMENT GUIDANCE FOR
SUPERFUND (RAGS): VOLUME I - HUMAN
HEALTH EVALUATION MANUAL (PART B,
DEVELOPMENT OF PRELIMINARY
REMEDIATION GOALS)
by
Environmental Quality Management, Inc.
Cedar Terrace Office Park, Suite 250
3325 Chapel Hill Boulevard
Durham, North Carolina 27707
and
E.H. Pechan & Associates
5537 Hempstead Way
Springfield, Virginia 22151
Contract No. 68-02-D120
Work Assignment No. II-86
PN 5046-2
Janine Dinan, Work Assignment Manager
Richard Atherton, Project Officer
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF EMERGENCY AND REMEDIAL RESPONSE
TOXICS INTEGRATION BRANCH (5024G)
401 M STREET, S.W.
WASHINGTON, D.C. 20460
April 1993
-------
DISCLAIMER
This report was prepared for the U.S. Environmental Protection Agency by
Environmental Quality Management, Inc., Durham, North Carolina, under E.H. Pechan's
Contract No. 68-02-D120, Work Assignment No. II-86. The contents are reproduced
herein as received from the contractor. The opinions, findings, and conclusions
expressed are those of the authors and not necessarily those of the U.S. Environmental
Protection Agency.
-------
CONTENTS
Figures vi
Tables vii
Acknowledgment vjjj
1. Introduction 1
Executive summary 1
Purpose of study 2
Description of report 4
2. Project Objectives 6
Project objectives 6
Technical objectives 7
3. Technical Approach and Results 8
Introduction 8
Model techniques examined 9
Selected model and method of application 11
Methodology protocol and investigation 12
Dispersion modeling results 32
Statistical analysis 32
4. Comparison of Results and Conclusions 48
Comparison of results 48
Conclusions 53
References 55
in
-------
CONTENTS (continued)
Appendices
A. Area Source Coordinates for Each Shape and Size Used
in the Study A-1
B. Distribution of National Weather Service Surface
Stations by Climatological Region B-1
C. Example Printout of ISCST2 Model Input Options for
Each Area Shape C-1
D. Corrections to RAGS - Part B Sections 3.3.1 and
3.3.2 D-1
E. Summary Statistics E-1
F. Example of Statistical Calculations F-1
IV
-------
FIGURES
Number
3-1 Subdivision of a 0.5 acre area source into smaller area source
components near a centrally located receptor (Case
No. 8) 14
3-2 Further subdivision of a 0.5 acre area source
(Case No. 16) 15
3-3 Area subdivision for a receptor shifted to the northeast of center 19
3-4 Selected climatic zones 26
3-5 Selected meteorological sites and applicable prevailing wind
direction (0-99 in whole degrees toward, 100-359 in ten
of degrees) 30
3-6 Average concentrations due to various shapes and sizes 37
3-7 Regression analysis of normalized concentrations for a square
area source 39
3-8 Regression analysis of normalized concentrations for a rectangle
1:3 area source 41
3-9 Regression analysis of normalized concentrations for a rectangle
1:5 area source . 43
4-1 Benzene PRG versus source shape and size using the inverse of
the mean normalized concentration 51
4-2 Benzene PRG versus source shape and size using the inverse of
the 95% UCL of the mean normalized concentration 52
-------
TABLES
Number
3-1 Maximum Unit Concentrations at Various Receptors
for a 0.5 Acre Site at Various Levels of Area
Source Subdivision 16
3-2 Percent Difference Between Concentrations for Various
0.5 Acre Area Source Cases 17
3-3 Concentration Estimates for Various Receptor Locations on the
0.5 Acre Site 20
3-4a Square Area Sources Characteristics 21
3-4b Percentage of Square Area Near Receptor That is Missed 21
3-5a 1:3 Rectangular Area Source Characteristics 22
3-5b Percentage of 1:3 Rectangular Area Source Near Receptor
That is Missed 22
3-6a 1:5 Rectangular Area Source Characteristics 23
3-6b Percentage of 1:5 Rectangular Area Source Near Receptor
That is Missed 23
3-7 Selected Surface and Upper Air Mixing Height Data by Climatic
Region 27
3-8 Prevailing Wind Directions for Each of the 29 Meteorological Sites 28
3-9 Options Selected in the ISCST2 Modeling 31
3-10 Normalized Maximum Concentrations for the 0.5 Acre Sites
for Each Shape and Size by Each Meteorological Site 33
VI
-------
TABLES (continued)
Number
3-11 Maximum Normalized Concentrations, kg/m3, for All 29
Meteorological Stations for 5.0 Acres 34
3-12 Maximum Normalized Concentrations, kg/m3, for All 29
Meteorological Stations for 50.0 Acres 35
3-13 Maximum Normalized Concentrations, kg/m3, for All 29
Meteorological Stations for 500.0 Acres 36
4-1 Comparison of Existing Versus Project Methodology for Benzene 50
VII
-------
ACKNOWLEDGMENT
This report was prepared for the U.S. Environmental Protection Agency, Office of
Emergency and Remedial Response, Toxics Integration Branch, by Environmental
Quality Management, Inc., Durham, North Carolina. This project was directed by Mr.
David R. Dunbar and managed by Mr. Craig Mann. The principal authors were Messrs.
Craig Mann, George Schewe, and Ron Freyberg. The U.S. Environmental Protection
Agency Work Assignment Manager was Ms. Janine Dinan.
VIII
-------
SECTION 1
INTRODUCTION
1.1 EXECUTIVE SUMMARY
The objective of this project was to develop a methodology for estimating via
dispersion modeling the maximum long-term ambient air concentration at the most
exposed individual (MEI) from uniform, nonbuoyant emissions of gaseous pollutants
and particulate matter nominally 10 microns and less (PM,0) from postn-emediation
Superfund sites. The final product is a methodology that is to be substituted for the
existing box dispersion model presently used in the Risk Assessment Guidance for
Superfund (RAGS) - Part B equations for estimating the volatilization factor (VF) and the
particulate emission factor (PEF). The VF and the PEF are integral in accounting for
the air pathway inhalation contribution to the risk-based preliminary remediation goal
(PRG) for soils.
This study used the most applicable U.S. Environmental Protection Agency
(EPA) regulatory dispersion model and a Nationally representative sample of U.S.
meteorological data to statistically estimate the normalized ambient air concentration
(concentration per unit emission rate) at the MEI for a range of area source sizes and
shapes. From the modeling results, regression analyses were performed on the data
to generate equations for estimating the mean and upper bound concentrations as a
function of source size and shape. The applicable equations were then rearranged to
conform with the requirements for substitution in the RAGS - Part B VF and PEF
equations.
The findings of the study indicate that the project methodology provides a
theoretically accurate estimate of the maximum ambient air concentration at the MEI
within the source configurations analyzed, while the existing box model does not
-------
account for an increase in concentration as a function of an increase in source size or
shape.
This report provides the project objectives, technical approach, results, and
conclusions of the study as well as the supporting data and suggested revisions to the
RAGS - Part B VF and PEF equations.
1.2 PURPOSE OF STUDY
On October 13 and 14, 1992, the EPA Superfund Program sponsored a meeting
with a number of State Agency representatives through the Association of State and
Territorial Solid Waste Management Officials (ASTSWMO) to discuss State issues
surrounding the cleanup of hazardous waste sites. A majority of State participants said
they wanted EPA to provide standardized exposure models, especially for the
dispersion of volatiles and fugitive dusts from waste sites.
Currently, Superfund uses a dispersion equation in the RAGS: Volume I -
Human Health Evaluation Manual (Part B, Development of Preliminary Remediation
Goals) (EPA 1991) to account for atmospheric dispersion of contaminants within the
area of contamination. The assumptions and mathematical treatment of dispersion
used in the equation, however, may not be applicable to a broad range of site types
and meteorology; nor does the equation represent state-of-the-art techniques using
regulatory dispersion models.
The purpose of this project is to prepare a method for estimating the dispersion
of volatile contaminants and inhalable fugitive dust (PM10) and the resultant upper
bound and average exposure point concentrations in air for the MEI residing on or near
the area of contamination.
The RAGS - Part B provides guidance on using EPA toxicity values and exposure
information to derive risk-based PRGs. In general, PRGs provide remedial design staff
with long-term cleanup level targets to use during analysis and selection of remedial
alternatives.
-------
The National Contingency Plan (NCR) which is found in 40 CFR, Part 300,
mandates that the selected remedial alternative meet all applicable or relevant and
appropriate requirements (ARARs), and provide protection of human health and the
environment. PRGs are developed to quantify the standards that remedial alternatives
must meet in order to achieve these two "threshold criteria." The major concerns in
establishing PRGs are "long-term effectiveness and permanence" of the remedy. These
balancing criteria for remedy selection are used to establish the risk posed to the
community once the remediation is complete. Risk-based PRGs quantify the degree of
residual risk after cleanup has been completed. If ARARs do not exist for the
contaminant of concern or for the media of concern, risk-based PRGs are developed to
protect human health.
PRGs are typically developed during the scoping phase or concurrent with initial
phases of the Remedial Investigation/Feasibility Study (RI/FS). Risk-based PRGs are
considered initial guidelines developed with readily available information and can be
modified as additional site data are obtained. A risk-based concentration is considered
a final remediation level only after appropriate analysis in the RI/FS and in the Record
of Decision (ROD).
PRGs for the soil medium are calculated for carcinogenic and noncarcinogenic
contaminants from standard residential and commercial/industrial land-use equations
given in RAGS - Part B. Integral to these equations is the soil-to-air VF which defines
the relationship between the concentration of contaminants in soil and the volatilized
contaminants in the air. The VF (m3/kg) is calculated as the inverse of the ambient air
concentration at the center of a ground-level, nonbuoyant area source of emissions.
The RAGS - Part B equation for calculating the VF consists of two parts: 1) a
volatilization model, and 2) a dispersion model.
The dispersion model is a simple box model from EPA 1986, and consists of the
inverse of the product of length of side of the contaminated areas (LS), the windspeed
in the mixing zone (V), and the diffusion height (DH). When the box model (m3/s)'1 is
multiplied by the emission rate (e.g.,//g/s), the product is the ambient air concentration
at the center of the "box" at a height of 2 meters.
-------
Similarly, this box model is also used in RAGS - Part B to calculate the PEF,
which is the inverse of the ambient air concentration of particles with an aerodynamic
diameter of 10 microns or less (PM10). As such, the box model treats PM^ as a
gaseous pollutant (i.e., complete reentrainment).
Because this type of box dispersion model represents a rather simplistic
approach to estimation of gaseous dispersion for on-site exposure receptors, EPA
commissioned this study to review the most recent and accepted dispersion model
tools and techniques for the purpose of constructing a methodology for estimating the
dispersion of gaseous and PM10 pollutants from post-remediation Superfund sites. This
methodology is to be applicable to a broad range of sites throughout the United States
and include an analysis of the range of site sizes and shapes and meteorological
conditions that are typical of closed Superfund sites.
The product of this effort is a methodology that can be used to estimate the
maximum long-term average and upper bound ambient air concentrations at the MEI.
The MEI is the point of maximum long-term average ambient air concentration resulting
from continuous and uniform site emissions of gaseous or PM,0 pollutants. This
methodology is to be substituted for the box model in the RAGS - Part B equations for
estimating the VF and PEF.
1.3 DESCRIPTION OF REPORT
Section 1 provides an executive summary of this report as well as the
background and history for this project.
Section 2 provides the overall project objectives and the technical objectives for
this study.
Section 3 details the overall technical approach including the options explored
and the methodology chosen as well as the technical protocol employed to achieve the
project objectives.
-------
Section 4 presents a comparison of results between the existing RAGS - Part B
box dispersion model and the project methodology as well as the conclusions drawn
from the project analyses.
Finally, the appendices provide the analytical data and the analyses employed to
derive the methodology specified in the project objectives. In addition, the appendices
provide example calculations and suggested revisions to the RAGS - Part B VF and
PEF equations.
-------
SECTION 2
PROJECT OBJECTIVES
The following provides both the overall objectives of this project and the
technical objectives to be attained in support of the project objectives.
2.1 PROJECT OBJECTIVES
The primary objective of this project was to replace the existing box dispersion
model used in the RAGS-Part B VF and PEF equations with a more accurate and better
documented approach for estimating the inverse of the maximum annual average
ambient air concentration at the MEI at a post-remediation Superfund site. This new
approach is to be based on empirical relationships between critical source terms (e.g.,
size, shape, etc.), meteorology, and the inverse of the source strength term as
estimated by either the Hwang and Faico (1986) model for volatilization from soils or
the Cowherd (1985) unlimited reservoir model for emissions of paniculate matter due to
wind erosion of soils.
In this regard, the following are the specific project objectives:
1. To develop a methodology for determining a National estimate of the
maximum annual average ambient air concentration of gaseous and PM,,,
pollutants emitted from mechanically undisturbed soil.
2. The methodology will accurately consider: 1) the typical range of sizes
and shapes of nonbuoyant, ground-level sources of emissions (area
sources) from post-remediation Superfund sites, 2) the expected range of
site topography, and 3) the range of annual meteorological conditions
across the United States.
-------
3. The methodology will be developed such that the results may be
substituted for the existing box dispersion model in the RAGS-Part B
equations for determining both the VF and PEF with minimal revision.
4. The methodology will accurately estimate the inverse of the maximum
annual average concentration based on methods employing the most
applicable preferred EPA refined regulatory air dispersion model.
2.2 TECHNICAL OBJECTIVES
The following represent the technical objectives of this project as intermediate
milestones for achieving the project objectives:
1. Select for this analysis the most appropriate EPA refined regulatory
dispersion model algorithm considering the requirement to estimate the
maximum annual average concentration from area sources at close-in and
on-site receptors.
2. Modify, as required, the area source term to accurately account for close-
in and on-site source/receptor geometry.
3. Review and select for this analysis the most conservative dispersion
coefficients for this analysis (i.e., urban or rural).
4. Review and select for this analysis a range of area source sizes and
shapes typical of post-remediation Superfund sites.
5. Review and select for this analysis a statistically representative sample
population of data from surface meteorological stations and their
respective upper air meteorological stations from the total population of
National Weather Service stations.
6. Using the selected dispersion model and the data obtained from
Technical Objective Nos. 1 through 5, establish the empirical relationship
between the combination of variables analyzed and the predicted
maximum annual average concentration such that the mean concentration
and 95 percent upper confidence limit of the mean concentration may be
estimated for given values of area source size and shape.
-------
SECTION 3
TECHNICAL APPROACH AND RESULTS
3.1 INTRODUCTION
As discussed in Section 2, the project objectives included using a more refined
modeling approach than the current Superfund box model approach for estimating
worst-case onsite concentrations. The generation of such a technique is valuable in
that the concentration estimates are more in line with other EPA model
recommendations and therefore, more defensible in terms of the ability to provide air
quality impacts of closed Superfund site emissions.
At the sacrifice of oversimplicity, the existing box model is a quick and easy-to-
use technique. A technique was selected in this study to maintain this same degree of
user compatibility and yet use dispersion modeling that considers the details of
sources, meteorology, and dispersion tools currently available for analyses. The
technique combined the ease of a "look-up" type of analysis (using tables, curves, or
regression fits) with the merits and results of a more sophisticated modeling analysis.
This was achieved by running the selected model multiple times for area sources
representative of those expected to generate fugitive emissions at closed Superfund
sites. A number of source shapes and sizes are possible, ranging from very small (0.5
acre) to large (500 acres), and include shapes such as square, circular, rectangular,
and polygonal.
Detailed model simulations also included several locations whereby the
geographical and meteorological differences were considered. This selection process
for the modeling techniques, the sources of emissions, and the meteorological sites are
presented in the following subsections.
8
-------
Of additional concern is the characterization of the receptors, i.e., those locations
where a person may be exposed to airborne emissions from a closed Superfund site.
For the current RAGS techniques, as well as for these updated more detailed model-
based techniques, the receptor can refer to both offsite as well as onsite impacts. The
selection of the worst-case location for a person on or near a former Superfund area is
discussed and documented herein.
3.2 MODEL TECHNIQUES EXAMINED
This documentation is not intended to be an in-depth review of the various area
source algorithms that are available such as the virtual point source, the point source
array, the finite line source, or the line source integration techniques. This type of
review can be found in "Review and Evaluation of Area Source Dispersion-Algorithms
for Emission Sources at Superfund Sites" (EPA 1989), and "Sensitivity Analysis of the
Revised Area Source Algorithm for the ISC2 Short-Term (ISCST2) Model" (EPA 1992a).
These reports, as well as the suggested screening analysis for a simple area source
algorithm (Wilson 1991), provide a base of area source applications and review
analyses that assists in selecting the appropriate modeling technique for this study. At
former Superfund sites, the potential exposure of concern is that which may occur to
the general public and assumes no implementation of institutional controls after closure.
Therefore, standard residential or commercial/industrial exposure scenarios are
applicable. This presents a modeling issue that is somewhat different than those for
which most recommended dispersion models were designed, namely, offsite receptor
locations.
The modeling approach -was required to consider several criteria for model
selection and application for this analysis. These include:
0 The ability of the model to handle area sources
0 The ability of the model to provide reasonable and representative
concentration estimates for onsite receptors
0 The need for annual (long-term) averaging periods
9
-------
0 The capability of the model to use full-year meteorological data sets
0 The availability of yearly meteorological data for multiple sites.
The order of the above criteria reflects the order of importance assigned to each
of the criteria with the ability to model area sources the primary feature required. The
second priority was the ability to model area sources for receptor locations both off and
onsite. The EPA "Guideline on Air Quality Models" (EPA 1986) specifies that several
models may be appropriate for the consideration of fugitive emissions from areawide
type sources. These models are referred to directly in the Guideline as the Industrial
Source Complex Model (EPA 1992), the Point-Area-Line (PAL) Model (Peterson 1987)
or through electronic notification on the Support Center for .Regulatory Air .Models
(SCRAM) Bulletin Board, the Fugitive Dust Model (FDM) (TRC 1990). Of these models,
the PAL and FDM models fit the area/receptor criteria, but PAL cannot model long-term
periods and FDM uses a sector averaging technique for lateral dispersion, which was
judged inadequate for this study. At first, the ISCST2 Model (EPA 1992) did not appear
to meet the onsite receptor criteria (the ISCLT2 Model was not considered for reasons
similar to FDM) although it met all other selection criteria. This is because the ISC2
User's Manual suggested restrictions on its use for area sources recognizing the
potential inaccuracies in concentration estimates. These inaccuracies are attributable to
the finite line emission source approach of the current ISCST2 model which limits
receptor-to-area source distances to one source side length. If a receptor is desired
closer to a given area source, however, the source could be subdivided, which reduces
the side length of any individual subdivision, maintaining the integrity of the overall
emissions and allowing a receptor to be placed somewhat closer to the area with a
decrease in potential calculational error. Pushing this subdivision technique to the
extreme (smaller and smaller areas) allows the receptor to be placed closer and closer
to the original source edge. For an onsite receptor location, only those small
subdivisions that are immediately adjacent to the receptor are subject to error in
calculating the concentration estimates (as long as other sources are of appropriate
size to meet the one side length criterion or greater receptor-to-area source criterion).
10
-------
Therefore, considering the ISCST2 Model area source capabilities when used with
subdivisions, the ISCST2 Model was selected for use.
3.3 SELECTED MODEL AND METHOD OF APPLICATION
The methodology selected for this analysis was the Industrial Source Complex
Model, Version 2, in its short-term mode (ISCST2). In this regard, ISCST2 calculates
concentration estimates for each hour of a meteorological data set to estimate an
annual average concentration. Other models such as PAL and FDM were not entirely
appropriate because they could not process a full year of meteorological data sets or
because sector averaging was used, which would tend to average concentrations. The
ISCST2 full year of meteorological data sets or sector averaging was used. The
ISCST2 model was run for the three area source sizes and shapes (Section 3.4.1) for
29 National Weather Service sites such that a representative annual average
concentration for a national mix of sites was generated. The ISCST2 Model is a steady-
state Gaussian plume model that uses hourly meteorological data along with source
emissions and other source characteristics to calculate ambient air concentrations at
user selected receptor locations. The ISCST2 Model uses a finite line emission source
approach to model area sources. Using ISCST2, an area source may be modeled as
one square area or subdivided into a series of smaller adjacent square areas that
improves the accuracy of the results for nearby receptors. Testing of the ISCST2
algorithm for area sources revealed that the model provides the best estimates of area
source concentrations when the receptor is located at least one area side length from
the area source (EPA 1992a). To allow the use of ISCST2 in this analysis where the
estimation of onsite concentrations is required, a methodology was devised to allow
accurate onsite concentrations to be estimated.
The methodology used the principle of subdividing the area source into smaller
square area sources. Where a receptor was located within the area source boundary,
the subdivisions were made very small near the receptor. Further subdivision was
carried out near the receptor and out away from the receptor until the concentration
11
-------
estimates for subsequent calculations converged within approximately three percent of
the previous model/source scenario concentrations. When this occurred, no further
subdivision was performed as the model results were deemed representative of the
overall area source impacts on the receptor. The process of selecting the appropriate
receptor locations, meteorological data sets, and area source subdivision configurations
is further explained in subsequent subsections.
For the three source shapes and four sizes selected, and for the selected
meteorological data sets, the ISCST2 Model was used to generate maximum annual
average concentrations for a unit emission rate of 1.0 g/m2-s. Using these maximum
concentrations, regression analyses were performed over all meteorological data sets
and the 95 percent upper confidence limits were calculated for each area source size
and shape. These equations are presented in Section 3.6, and presented for
substitution into the Superfund RAGS - Part B equations in Appendix D.
3.4 METHODOLOGY PROTOCOL AND INVESTIGATION
3.4.1 Source Selection
Review of several sources of information, including the Records of Decision
(RODs) data base, past studies of active Superfund sites, and ongoing cleanup efforts
allowed the selection of area sources for this study. The criteria for this selection was
based primarily on representativeness of both size and shape of the potential areas of
emissions and the desire to cover as many in the detailed modeling as possible. The
most predominant shape was a square which was also judged to be representative of a
circular source of the same area. To allow for areas not square or circular in shape,
two additional shapes were selected, namely, rectangles with side-to-length ratios of 1
to 3 and 1 to 5.
Area source sizes were selected to bracket the range of sizes found in the
literature or encountered in previous Superfund studies. For each of the three shapes
given above, sizes of 0.5, 5.0, 50.0, and 500.0 acres were modeled.
12
-------
3.4.2 Source Subdivision
The project methodology is similar to that recommended in the ISCST2 Model
User's Manual (EPA 1992) which states that if the receptor is closer than one area
source side length to the emission source, the area should be subdivided into smaller
sources. In testing this theory, it was found that a greater number of sources tended to
refine the estimated concentration. To minimize the errors associated with subdividing
the different size areas that were analyzed in this study and to minimize the potential
number of sources that were required to represent each area, a scheme was devised
that combined subdividing the area with concentrating the smaller area source
subdivisions to the near receptor source components. Figure 3-1 illustrates this
concept and provides a basis for subdividing the area sources to provide concentration
estimates that are representative of the impacts of the emission sources at receptors
onsite and near the edge of the area. With this resolution, and given the three percent
convergence criterion, this technique of applying the ISCST2 Model was thought to
provide concentrations that would concur with other applicable EPA models such as
FDM or PAL
The final number of area sources required for a 0.5 acre site (45 m by 45 m,
2025 m2) was resolved after multiple applications of finer and redistributed area source
subdivisions using the ISCST2 Model. Eighteen scenarios were reviewed starting with
the simplest which consisted of one square area source. This source was subdivided
into 4, 16, 36, 64, 81, and 100 equal areas (as illustrated for the 36 area source set in
Figure 3-2) and continued with further subdivisions close to each centrally located
receptor. In addition, the onsite receptor location was varied, i.e., moved both
diagonally and horizontally, to test if prevailing winds and source subdivision and
orientation would yield different results. One meteorological data set for Trenton, New
Jersey from 1989 was used in this analysis.
Table 3-1 shows the resulting concentration estimates for the variable area
source subdivisions for a centrally located receptor as well as for receptors located
around the edge of the property.
13
-------
45.00
10m
5m
2.5m
45.00
Figure 3-1. Subdivision of a 0.5 acre area source into smaller area source components
near a centrally located receptor (Case No. 8).
14
-------
Figure 3-2. Further subdivision of a 0.5 acre area source (Case No. 16).
15
-------
TABLE 3-1. MAXIMUM UNIT CONCENTRATIONS AT VARIOUS RECEPTORS FOR A 0.5 ACRE SITE AT VARIOUS
LEVELS OF AREA SOURCE SUBDIVISION
Run
Identification
1
2
3
4
5
6
7
8
15
16
17
Case number of
area sources
1(1x1)
4(2x2)
16(4x4)
36(6x6)
64(8x8)
81(9x9)
100(10x10)
80
68
92
104
Receptor locations, x,y*
0.0
0
2960726
4946168
6147752
4500075
652892
5462401
8537251
6890590
8599995
8599995
0.25
0
2403004
3179206
3523934
4217779
3910758
3883353
3681026
3657128
3687329
3688852
25,0
0
2351337
3130088
3471754
4171871
3718460
3785313
3464095
3444188
3469130
3470379
25,25
1304385
1842145
2247864
2427864
2960231
2569402
2591204
2416078
2397076
2420914
2422036
-25,0
0
1496938
1936937
2102594
2304048
2170491
2267750
2525232
2519996
2526573
2526903
0.-25
0
1990201
2628039
2922216
3147034
1288721
3195456
3085376
3063832
2090621
3091918
-25.-2S
828500
1159197
1392480
1491735
1643043
1565611
1576843
1697947
1690047
1699902
1700364
0)
'The center of the area is at 0,0 m.
-------
Comparison of the concentrations in Table 3-1 between various levels of area
source subdivision indicates that the desired three percent convergence of
concentration occurred between 80 and 104 subdivisions when subdivided as in Figure
3-1. Convergence percentage is calculated by dividing the difference between two
estimates by the original estimate and multiplying by 100. Further subdivision to
smaller components did not result in significantly better concentration estimates for the
maximum receptor. Table 3-2 shows comparisons of some of the pertinent
convergence levels of concentrations for the center and edge receptors. As can be
seen, the percent difference between 80 subdivisions (Case 8) and 92 subdivisions
(Case 16) were less than one percent at the center receptor and other receptors as
well.
TABLE 3-2. PERCENT DIFFERENCE BETWEEN CONCENTRATIONS FOR VARIOUS
0.5 ACRE AREA SOURCE CASES*
Case
numbers
compared
6 and 7
8 and 15
8 and 16
16 and 17
Receptor location, x(m), y(m)
0,0
16.3
19.3
-0.7
0.00
0,25
0.7
0.7
-0.2
-0.04
25,0
-1.8
0.6
-0.1
-0.04
25,25
-0.8
0.8
-0.2
-0.05
-25,0
-4.5
0.2
-0.1
-0.01
0.-25
-0.2
0.7
-0.2
-0.04
-25,-25
-0.7
0.5
-0.1
-0.03
Percent difference in concentration
'Case 6:
Case 7:
Case 8:
Case 15:
Case 16:
Case 17:
81 square, equal sources
100 square, equal sources
80 graduated sources
68 graduated sources
92 graduated sources
104 graduated sources.
Further testing was performed for the Case No. 8 square scenario (80
subdivisions) by moving the receptor to various locations within the area source. The
17
-------
area subdivision scheme was moved along with the receptor location to maintain the
resolution and accuracy found in the subdivision technique. Figure 3-3 provides one
illustration of how the area subdivisions were moved to the northeast maintaining the
receptor at the middle of the subdivisions. This test was performed to identify any
effects that prevailing winds might have had on the location of the maximum onsite
concentration. This analysis was performed for two meteorological data sets, Casper,
Wyoming, with a high frequency of prevailing winds, and Newark, New Jersey with a
lower frequency of prevailing winds. In all cases, the maximum concentrations
occurred at the center of the square area source. Using the same subdivision theory
for the 1:3 and 1:5 rectangular areas and receptors at various onsite locations, the
maximum concentration was found to also occur at the center of the rectangular areas.
Table 3-3 shows the concentration results of the receptor location/area subdivision
shifts. The maximum concentration may shift for a given scenario but the maximum
always occurred at the central receptor when comparing all the scenarios (Case No. 8
greater than any other receptor/area source case).
An additional factor tested for the rectangles was to compare the concentrations
for a north-south aligned source (long side aligned north and south) and one aligned
with the prevailing wind directions. The analysis estimated slightly higher
concentrations when aligned with the prevailing wind direction, and thus for each
meteorological site, the 1:3 and 1:5 rectangular areas were aligned with the prevailing
winds.
18
-------
45.00 m
10m
2.5m
5m
45.00 m
Figure 3-3. Area subdivisions for a receptor shifted to the northeast of center.
19
-------
TABLE 3-3. CONCENTRATION ESTIMATES FOR VARIOUS RECEPTOR
LOCATIONS ON THE 0.5 ACRE SITE
Case
identification
No. 8
No. 9
No. 10
No. 11A
No. 11B
No. 12
No. 13
No. 14
Primary
receptor
location
Center
Northeast
Northeast
Northeast
Northeast
East
East
East
Maximum
concentration at
0,m; Q,m,tjg/ir?
8537251
6096166
6095600
5983768
6082262
6347859
6153920
5994692
Maximum
concentration,
^9/m3
8537251
8074374
6971161
5983768
6082262
7887059
7759499
5994692
Maximum receptor
location
x,m
0
12.5
17.5
0
0
12.5 ^
17.5
0
y,m
0
12.05
17.5
0
0
. 0
0
0
The resulting source shapes and sizes used in this analysis are presented in
Tables 3-4a through 3-6b. The source sizes of 0.5, 5.0, 50.0, and 500.0 acres were
approximated in units of meters to allow a more easily defined source array and
therefore they did not exactly equal the desired acreage. This did not present a
problem, however, as the actual areas shown in Tables 3-4a through 3-6a were used in
all subsequent calculations. Also shown in each table are the number of sources used
to represent the area and the percentage of the area that the configuration did not
consider. It should be noted that the nearest subdivided small areas (Figure 3-1) were
not modeled because they are adjacent to the receptor and the area edge-to-downwind
receptor distance did not meet the one side length minimum distance criteria. As can
be seen in Tables 3-4b to 3-6b, the percentage of missing area in each analysis (and
therefore unaccounted for emissions) was less than 0.5 percent in all cases. Appendix
20
-------
TABLE 3-4a. SQUARE AREA SOURCES CHARACTERISTICS
Desired area
size, acres
0.5
5
50
500
Desired
area, m2
2025
20250
202500
2025000
Actual
side length, m
45x45
142.2x142.3
450 x 450
1423x1423
Modeled side length,
m
45
140
450
1400
Modeled area,
m2
2025
19600
202500
1960000
Modeled
acreage
0.50
4.84
50.00
483.95
Percentage of
desired size
1.00
0.97
1.00
0.97
TABLE 3-4b. PERCENTAGE OF SQUARE AREA NEAR RECEPTOR THAT IS MISSED
Small box selection, less than 5% of overall area side length
Acres
0.5
5
50
500
Overall side length, m
45
140
150
1400
Smallest box side, m
1.61
5.00
16.07
50.00
No. of area sources
96
96
96
, 96
Percent area missing"
0.26
0.12
0.26
0.12
'Very small area percentage near the central receptor.
-------
TABLE 3-5a. 1:3 RECTANGULAR AREA SOURCE CHARACTERISTICS
Desired area
size, acres
0.5
5
50
500
Desired
area, m2
2025
20250
202500
2025000
Actual
side length, m
25.98 x 77.94
82.158 X 246.475
259.8 x 779.4
821.58 x 2464.75
Modeled side length,
m
25x75
80 x 240
250x750
800 x 2400
Modeled
area, m2
1875
19200
187500
1920000
Modeled
acreage
0.46
4.74
46.30
474.07
Percentage of
desired size
0.93
0.95
0.93
0.95
TABLE 3-5b. PERCENTAGE OF 1:3 RECTANGULAR AREA SOURCE NEAR RECEPTOR THAT IS MISSED
Small box selection, less than 5% of overall area side length
Acres
0.5
5
50
500
Overall side length, m
25x75
80 x 240
250 x 750
800 x 2400
Smallest box side, m
1.25
4.00
12.50
40.00
No. of area sources
128
128
128
128
Percent area missing*
0.15
0.08
0.15
0.08
'Very small area percentage near the central receptor.
-------
TABLE 3-6a. 1:5 RECTANGULAR AREA SOURCE CHARACTERISTICS
Desired area
size, acres
0.5
5
50
500
Desired
area, m2
2025
20250
202500
2025000
Actual
side length, m
20.1246 x 100.623
63.696 x 318.198
201.246x1006.23
636.396x3181.98
Modeled side length,
m
20 x 100
64 x 320
200x1000
640 x 3200
Modeled
area, m2
2000
20480
200000
204800
Modeled
acreage
0.49
5.06
49.38
505.67
Percentage of
desired size
0.99
1.01
0.99
1.01
CO
TABLE 3-6b. PERCENTAGE OF 1:5 RECTANGULAR AREA SOURCE NEAR RECEPTOR THAT IS MISSED
Small box selection, less than 5% of overall area side length
Acres
0.5
5
50
500
Overall side length, m
20x100
64x320
200x1000
640 X 3200
Smallest box side, m
2.00
6.40
20.000
64.00
No. of areas sources
106
106
106
106
Percent area missing"
0.39
0.20
0.39
0.20
"Very small area percentage near the central receptor.
-------
A presents the coordinates and side lengths for each area subdivision for each of the
three source sizes and shapes.
3.4.3 Meteorological Data
Given the objectives of this analysis, i.e., to obtain a set of annual average
concentrations for specified area shapes and sizes for estimating the worst-case fugitive
emission impacts, a range of available meteorological data was examined. The
meteorological data used as a data base set was that found on the SCRAM bulletin
board under the surface and upper air meteorology site information. There are 230
potential surface sites identified in the contiguous U.S. The total population of surface
sites was too great to consider for full scale dispersion modeling given the extent of the
available resources. Therefore, a portion of the data was selected on the basis of a
simple statistical analysis such that the smaller data set represented the overall data
set.
The first step of the sampling process was to determine the number of sites
required to represent the overall population of 230 sites. Because the overall objective
was to estimate the population mean concentration, the number of sites to be included
in the analysis was based on estimating the mean to within a specified margin of error,
with a high degree of confidence (i.e., 95%). Data from selected sites were used to
obtain preliminary estimates of the expected magnitude and variability of the
concentrations from different sites. Based on this information, a sample of 29 sites was
expected to provide a sufficient number of measurements with which to estimate the
overall population concentration with a high degree of confidence.
Broad variations in both extent and geographical representativeness, however,
across the U.S. did not allow simple random selection from the overall data set. The
U.S. is typified by broad variations due to land-sea interactions, proximity to terrain,
location with regard to pressure systems and frontal movements, latitude, and broad
climatological locations with regard to air masses (e.g., cold, dry continental versus
warm, humid marine). Broad atmospheric zones can be defined on the basis of
general meteorological conditions that affect the transport and dispersion of pollutants.
24
-------
These zones were reviewed in terms of the primary characteristics of each area's air
pollution climatology, topographical features, inversion conditions, windspeeds and
wind directions, mixing heights, and overall climate. Nine distinct and broad zones
were selected to represent the range of atmospheric conditions across the U.S. The
zones were defined as:
Zone I: North Pacific Coastal
Zone II: South Pacific Coastal
Zone III: Southwest
Zone IV: Northwest Mountains
Zone V: Central Plains
Zone VI: Southeast
Zone VII: Midwest
Zone VIII: Northern Atlantic
Zone IX: South Florida.
Figure 3-4 shows the approximate boundaries of each air pollution dispersion climatic
zone used in this analysis. Appendix B lists each of the 230 sites considered in the
data base and arranges them by location in the appropriate climatic region. Data from
Appendix B and the sample population of 29 sites were combined to determine how
many sites should be selected for each climatic zone. First thought was to allocate the
29 sites to the zones by the ratio of the number in each zone to the total of 230 sites.
This procedure, however, did not seem appropriate, because the density of sites was
more a function of population than geographical coverage. The methodology used to
select the appropriate number of sites to represent each zone was based on
meteorological representativeness and variability across each zone. Large scale
regional average conditions were used to select the sites. The number of sites selected
for each zone is as follows:
Zone I - 3
Zone II -2
Zone III - 3
Zone IV - 5
Zone V - 3
Zone VI - 5
25
-------
Figure 3-4. Selected climatic zones.
-------
Zone VII - 4
Zone VIII - 3
Zone IX- 1.
Table 3-7 lists the sites selected for this analysis, the zone to which each site was
assigned, the year of data used, and the associated upper air site selected to
characterize the mixing heights.
Another meteorological condition that was required to be characterized for the
modeling analysis was the prevailing wind direction. This was used to differentiate each
site and to orient each 1:3 and 1:5 rectangular area source to maximize the potential
impact of the source at each location. The prevailing wind directions were found in the
Climatic Atlas of the U.S. (NOAA 1974) for each of the 29 sites modeled. These values
are presented in Table 3-8 for each site. The ISCST2 Model was adjusted to account
for prevailing wind direction by using the feature of the model that allows the wind
directions to be corrected to the north (for example, where magnetic north was used to
install a monitor). For this analysis, the area sources were left in a north-south
orientation and the wind directions reoriented with respect to the long side of each
source. Figure 3-5 shows each selected site and the applicable prevailing wind
direction.
One final meteorological variable is the consideration of rural versus urban
geographical setting and how it affects the selection of dispersion coefficients in the
ISCST2 Model. Sensitivity testing resulted in more conservative estimates of
concentration using rural dispersion coefficients. For this reason, rural coefficients were
used for all model runs in this study.
3.4.4 Model Application Options
The latest version of the short-term Industrial Source Complex Model (ISCST2)
(EPA 1992) available from EPA was used in this analysis. The present ISCST2 Model is
similar to its earlier versions, but has been reprogrammed and has a new format for
inputting data in batch mode using keywords to .indicate specific data components.
The ISCST2 Model used in this analysis included all model options and algorithms
27
-------
TABLE 3-7. SELECTED SURFACE AND UPPER AIR MIXING HEIGHT DATA BY CLIMATIC
REGION
Site
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Surface data
location
Albuquerque
Atlanta
Bismarck
Boise
Casper
Charleston
Chicago
Cleveland
Denver
Fresno
Harrisburg
Hartford
Houston
Huntington
Las Vegas
Lincoln
Little Rock
Los Angeles
Miami
Minneapolis
Philadelphia
Phoenix
Portland
Raleigh-Durham
Salem
Salt Lake City
San Francisco
Seattle
Winnemucca
Surface
station No.
23050
13874
24011
24131
24089
13880
94846
14820
23062
93193
14751
14740
12960
3860
23169
14939
13963
23174
12839
14922
13739
23183
14764
13722
24232
24127
23234
24233
24128
Year of
data
1989
1989
1989
1989
1989
1989
1989
1989
1989
1989
1987
1989
1988
1989
1989
1989
1988
1989
1989
1989
1987
1989
1989
1989
1989
1989
1989
1989
1985
Upper air level
data location
Albuquerque
Athens
Bismarck
Boise
Lander
Charleston
Peoria
Buffalo
Denver
Oakland
Sterling
Albany
Victoria
Huntington
Desert Rock
Omaha
Little Rock
Oakland
West Palm Beach
St. Cloud
Sterling
Tuscon
Portland
Greensboro
Salem
Salt Lake City
Oakland
Quillayute
Winnemucca
Associated upper
air station No.
23050
13873
24011
24131
24021
13880
14942
14733
23062
23230
93734
14735
12912
3860
3160
94918
13963
23230
12844
14926
93734
23160
14764
13723
24232
24127
23230
94240
24128
Year of
data
1989
1989
1989
1989
1989
1989
1989
1989
1989
1989
1987
1989
1988
1989
1989
1989
1988
1989
1989
1989
1987
1989
1989
1989
1989
1989
1989
1989
1985
28
-------
TABLE 3-8. PREVAILING WIND DIRECTIONS FOR EACH OF THE 29 METEOROLOGICAL
SITES
Site
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Surface data
location
Albuquerque
Atlanta
Bismarck
Boise
Casper
Charleston
Chicago
Cleveland
Denver
Fresno
Harrisburg
Hartford
Houston
Huntingdon
Las Vegas
Lincoln
LJttle Rock
Los Angeles
Miami
Minneapolis
Philadelphia
Phoenix
Portland
Raleigh-Durham
Salem
Salt Lake City
San Francisco
Seattle
Winnemucca
Surface
station No.
23050
13874
24011
24131
24089
13880
94846
14820
23062
93193
14751
14740 .
12960
3860
23169
14939
13963
23174
12839
14922
13739
23183
14764
13722
24232
24127
23234
24233
24128
Year of
data
1989
1989
1989
1989
1989
1989
1989
1989
1989
1989
1987
1989
1988
1989
1989
1989
1988
1989
1989
1989
1987
1989
1989
1989
1989
1989
1989
1989
1985
Prevailing wind
direction, degrees
300
140
110
310
45
45
20
0
0
140
95
180
320
45
40
0
45
90
270
140
140
270
0
45
0
340
120
90
90
Rotating angle
degrees
-60
140
110
-50
45
45
20
0
0
140
95
180
-40
45
40
0
45
90
-90
140
140
-90
0
45
0
-20
120
90
90
29
-------
Figure 3-5. Selected meteorological sites and applicable prevailing wind direction
(0-99 In whole degrees toward, 100-359 in ten of degrees).
-------
of the former versions and was operated on a personal computer (486/33 with
extended RAM memory). The steps taken to generate the normalized concentration
estimates for each area size and shape for each meteorological site were very specific
and straightforward. These steps have been summarized in the preceding sections
with regard to model selection, site characterization, meteorological data, and
receptors. The ISCST2 Model has a number of additional or related options that are
summarized in Table 3-9 relating to model applications and results. Appendix C
presents an example of the input variables for the ISCST2 Model for one source size
and one meteorological site for each source shape to illustrate the form of the ISCST2
Model variables.
TABLE 3-9. OPTIONS SELECTED IN THE ISCST2 MODELING
Concentrations in micrograms/cubic meter were computed.
Rectangular coordinates for the discrete receptor locations were used.
Terrain elevation was not considered for sources and receptors.
The Rural Mode option was selected.
Default wind profile exponent values were used.
Default vertical potential temperature gradient values were used.
The downwind distance plume rise option was used for all sources.
Buoyancy-induced dispersion was used.
The wind system measurement height was set equal to 10 meters for all sites.
No building aerodynamic downwash was considered for the analysis.
Program control parameters, receptors, and source input data were output for
preliminary runs but not all subsequent runs.
Concentrations during calm windspeed hours were set to zero using the ISCST2
internal processing.
Any annual averaging times were selected consistent with the requirements of the
analysis.
The units default for concentration output was set equal to 0.1 x 102 to convert the
input units of g/nf-s to kg/m3.
31
-------
3.5 DISPERSION MODELING RESULTS
Normalized concentrations were generated for each area source size (4) and
shape (3) for each meteorological data set (29) at the center of each area source
(additional receptors were maintained around each area source for comparison, but the
center was always the highest value). Tables 3-10 through 3-13 provide the maximum
concentration estimates for each size area by meteorological data set and shape.
Concentrations were generally within approximately 50 percent from the lowest to the
highest for a given shape. Differences between maximum concentrations for the three
sizes at a given meteorological site are relatively small (generally less than about 10%),
which indicates that shape may not have been as significant as originally thought.
The concentrations in Tables 3-10 through 3-13 were used to generate all
regression analyses. Figure 3-6 presents a histogram summary of the average
concentrations for each shape and size showing an increase in normalized
concentrations with size.
3.6 STATISTICAL ANALYSIS
Summary statistics were calculated for each of the nine climatic zones and each
of the area sources. Tables E-1, E-2, and E-3 in Appendix E present the summary
statistics for the square area source, the 1:3 rectangle area source, and the 1:5
rectangle area source, respectively. The summary statistics include the total number of
meteorological stations within the zone and the number that were part of the sample.
In addition, for each of the four area sizes, the minimum, maximum, and mean
normalized concentrations for each zone are presented.
Overall, there are 29 concentrations for each shape and size configuration.
Regression analysis was used to evaluate the relationship between the normalized
concentrations and the area size for each of the three shapes. The resultant
regression equations were then used to determine the mean normalized concentration
and the upper 95% confidence limit for the mean normalized concentration.
32
-------
TABLE 3-10. NORMALIZED MAXIMUM CONCENTRATIONS FOR THE 0.5 ACRE
SITES FOR EACH SHAPE AND SIZE BY EACH METEOROLOGICAL SITE
Met Station
Albuquerque
Atlanta
Bismarck
Boise
Casper
Charleston
Chicago
Cleveland
Denver
Fresno
Harrisburg
Hartford
Houston
Huntington
Las Vegas
Lincoln
Little Rock
Los Angeles
Miami
Minneapolis
Philadelphia
Phoenix
Portland
Raleigh
Salem
Salt Lake City
San Francisco
Seattle
Winnemucca
Square
(0.5 acres)
0.00889
0.00968
0.00895
0.01092
0.00743
0.01000
0.00765
0.00901
0.00986
0.01214
0.00899
0.01044
0.00945
0.01417
0.00772
0.00917
0.01010
0.01054
0.00869
0.00828
0.00824
0.01 149
0.00999
0.00967
0.01011
0.00961
0.00836
0.00886
0.01071
1:3 Rectangle
(0.5 acres)
0.00881
0.00968
0.00895
0.01178
0.00775
0.01015
0.00761
0.00897
0.01042
0.01281
0.00930
0.01102
0.00922
0.01413
0.00777
0.00993
0.01035
0.01189
0.00852
0.00824
0.00802
0.01280
0.00990
0.00958
0.01097
0.01043
0.00836
0.00982
0.01017
1:5 Rectangle
(0.5 acres)
0.00883
0.00974
0.00898
0.01238
0.00790
0.01030
0.00762
0.00897
0.01068
0.01321
0.00940
0.01133
0.00921
0.01420
0.00775
0.01030
0.01041
0.01218
0.00849
0.00831
0.00785
0.01326
0.00988
0.00963
0.01134
0.01073
0.00850
0.01009
0.00999
33
-------
TABLE 3-11. MAXIMUM NORMAUZED CONCENTRATIONS, kg/m3, FOR ALL 29
METEOROLOGICAL STATIONS FOR 5.0 ACRES
Met Station
Albuquerque
Atlantq
Bismarck
Boise
Casper
Charleston
Chicago
Cleveland
Denver
Fresno
Harrisburg
Hartford
Houston
Huntingdon
Las Vegas
Lincoln
Little Rock
Los Angeles
Miami
Minneapolis
Philadelphia
Phoenix
Portland
Raleigh
Salem
Salt Lake City
San Francisco
Seattle
Winnemucca
Square
(5 acres)
0.01207
0.01314
0.01217
0.01470
0.01000
0.01357
0.01034
0.01211
0.01345
0.01635
0.01221
0.01426
0.01283
0.01922
0.01046
0.01262
0.01371
0.01451
0.01189
0.01114
0.01119
0.01583
0.01367
0.01307
0.01376
0.01300
0.01133
0.01203
0.01459
1:3 Rectangle
(5 acres)
0.01241
0.01359
0.01262
0.01632
0.01077
0.01424
0.01065
0.01260
0.01457
0.01782
0.01299
0.01539
0.01301
0.01986
0.01087
0.01391
0.01452
0.01653
0.01207
0.01153
0.01 122
0.01785
0.01401
0.01343
0.01521
0.01444
0.01173
0.01354
0.01443
1:5 Rectangle
(5 acres)
0.01053
0.01161
0.01073
0.01481
0.00937
0.01232
0.00904
0.01161
0.01276
0.01579
0.01115
0.01350
0.01098
0.01696
0.00916
0.01238
0.01244
0.01459
0.01018
0.00988
0.00933
0.01592
0.01182
0.01146
0.01345
0.01280
0.01013
0.01195
0.01192
34
-------
TABLE 3-12. MAXIMUM NORMALIZED CONCENTRATIONS, kg/m3, FOR ALL 29
METEOROLOGICAL STATIONS FOR 50.0 ACRES
Met Station
Bismarck
Boise
Casper
Charleston
Chicago
Cleveland
Denver
Fresno
Harrisburg
Hartford
Houston
Huntington
Las Vegas
Lincoln
Little Rock
Los Angeles
Miami
Minneapolis
Philadelphia
Phoenix
Portland
Raleigh
Salem
Salt Lake City
San Francisco
Seattle
Winnemucca
Square
(50 acres)
0.01462
0.01765
0.01192
0.01629
0.01233
0.01443
0.01610
0.01963
0.01451
0.01704
0.01536
0.02306
0.01242
0.01523
0.01642
0.01746
0.01433
0.01330
0.01335
0.01907
0.01652
0.01564
0.01638
0.01557
0.01358
0.01429
0.01749
1:3 Rectangle
(50 acres)
0.01509
0.01952
0.01277
0.01700
0.01263
0.01496
0.01740
0.02132
0.01540
0.01832
0.01550
0.02371
0.01285
0.01673
0.01734
0.01982
0.01448
0.01371
0.01333
0.01804
0.01680
0.01599
0.01604
0.01723
0.01401
0.02143
0.01721
1:5 Rectangle
(50 acres)
0.01285
0.01776
0.01112
0.01474
0.01073
0.01266
0.01526
0.01579
0.01324
0.01609
0.01310
0.02027
0.01083
0.01491
0.01487
0.01751
0.01222
0.01177
0.01109
0.01915
0.01419
0.01367
0.01598
0.01529
0.01213
0.01418
0.01421
35
-------
TABLE 3-13. MAXIMUM NORMALIZED CONCENTRATIONS, kg/m3, FOR ALL 29
METEOROLOGICAL STATIONS FOR 500.0 ACRES
Met Station
Albuquerque
Atlanta
Bismarck
Boise
Casper
Charleston
Chicago
Cleveland
Denver
Fresno
Harrisburg
Hartford
Houston
Huntingdon
Las Vegas
Lincoln
Little Rock
Los Angeles
Miami
Minneapolis
Philadelphia
Phoenix
Portland
Raleigh
Salem
Salt Lake City
San Francisco
Seattle
Winnemucca
Square
(500 acres)
0.01796
0.01951
0.01845
0.02176
0.01502
0.02079
0.01540
0.01811
0.01995
0.02446
0.01788
0.02130
0.01950
0.02891
0.01524
0.01919
0.02058
0.02203
0.01799
0.01644
0.01653
0.02391
0.02086
0.01986
0.02018
0.01917
0.01692
0.01792
0.02189
1:3 Rectangle
(500 acres)
0.01795
0.01967
0.01847
0.02394
0.01549
0.02074
0.01531
0.01812
0.02121
0.02607
0.01863
0.02228
0.01882
0.02889
0.01547
0.02069
0.02113
0.02437
0.01775
0.01666
0.01615
0.02628
0.02073
0.01949
0.02186
0.02107
0.01720
0.01949
0.02088
1:5 Rectangle
(500 acres)
0.01532
0.01687
0.01577
0.02196
0.01354
0.01806
0.01306
0.01536
0.01872
0.02328
0.01608
0.01967
0.01594
0.02479
0.01307
0.01854
0.01819
0.02165
0.01503
0.01436
0.01345
0.02363
0.01756
0.01675
0.01946
0.01879
0.01504
0.01733
0.01727
36
-------
Concentration vs Shape of Areas
0.03000 -i
0.02500 -
Concentration,
ug/m3
0.5 5.0 50.0 500.0
Acres
1:3 Rectangle • 1:5 Rectangle
Figure 3-6. Average concentrations due to various shapes and sizes.
37
-------
3.6.1 Regression Analysis
Simple linear regression analysis was used to evaluate the nature of the
relationship between the normalized concentration and the size of the area. Each of
the three shapes was considered separately. Preliminary plots of the data illustrated
that the relationship between the normalized concentrations and the size of the area
was exponential. Therefore, the relationship was linearized by taking the natural
logarithms (In) of each variable. The regression analysis was then performed on the In-
transformed data. The Pearson correlation coefficient, which indicates the strength of
the linear association between the two variables, was calculated for each regression
analysis.
Square Area Source-The linear relationship between the In-normalized
concentration (kg/m3 per g/mf-s on original scale) and the In of the area (m2 on
original scale) is illustrated in Figure 3-7. Each of the 29 concentrations at each of the
four sizes are plotted along with the regression line. The Pearson correlation coefficient
was 0.87, which indicates a strong linear relationship between the two variables. The
slope of the estimated linear regression equation was significantly greater than zero
(p=0.0001), which also indicates a significant linear association. The linear regression
equation for estimating the mean In-normalized concentration for a given size area is
presented below:
Y = -5.3880 + 0.1005 ln(X)
where
Y is the estimated mean In-normalized concentration
X is the size of the area in m2.
The exponential of Y is then calculated to estimate the normalized concentration
on the original scale (i.e., kg/m3 per g/rrf-s). For example, the estimated mean
normalized concentration for an area of 2025 m2 is exp(-4.6229) = 0.0098 kg/m3 per
g/nf-s. A detailed example of the calculation is presented in Appendix F.
38
-------
CO
CD
-3.5:
-3.8 :
£ -17
I « -3-8
1 --3.9^
§ I -4.0
00
•Q ° -4.1
!•-"
o J -4.3
"& ^ -4.4
If-41
-4.8
-4.7
-4.8
-4.9
-5.0
Y = -5.3880 + 0.1005 ln(X)
[ I I I I I M 1 I | I 1 I I I I II I [ I M I I I I M | I 1 I I M I I I | 1 I 1 I I 1 1 I I | I 1 I I 1 1 I I I [I I 1 I 1
7 8 9 10 11 I 12 13
Natural Logarithm of Area (X), m* In original units
I | I I I I I I I I 1 1'
14 15
Figure 3-7. Regression analysis of normalized concentrations for a square area source.
-------
The following equation is then used to estimate the upper 95% confidence limit
(UCL) for the mean normalized concentration at X=\:
95% UCL = exp[Yh + 2.92 s(Yh)] kg/m3 per g/nf-s
where
Yh = -5.3880 + 0.1005 Info)
stfh) = The standard error of Yh = 0.0213 [1/4 + (Info) - 11.0509)2/26.3608]
For example, the estimated 95% UCL for the normalized concentration for an
area of 2025 m2 is exp(-4.6229 + 2.92 x 0.0149) = 0.0103 kg/m3 per g/nf-s. A
detailed example of the calculation is presented in Appendix F.
1:3 Rectangular Area Soarce-The linear relationship between the In-normalized
concentration (kg/m3 per g/nf-s on original scale) and the In of the area (m2 on
original scale) is illustrated in Figure 3-8. Each of the 29 concentrations at each of the
four sizes are plotted along with the regression line. The Pearson correlation coefficient
was 0.85, which indicates a strong linear relationship between the two variables. The
slope of the estimated linear regression equation was significantly greater than zero
(p=0.0001), which also indicates a significant linear association. The linear regression
equation for estimating the mean In-normalized concentration for a given size area is
presented below:
Y = -5.3466 + 0.1004 In (X)
where
Y is the estimated mean In-normalized concentration
X is the size of the area in m2.
The exponential of Y is then calculated to estimate the normalized concentration
on the original scale (i.e., kg/m3 per g/rrf-s). For example, the estimated mean
40
-------
-3-5:
-16-
> -17-
-------
normalized concentration for an area of 2025 m2 is exp(-4.5822) = 0.0102 kg/m3 per
g/mf-s. A detailed example of the calculation is presented in Appendix F.
The following equation is then used to estimate the upper 95% confidence limit
(UCL) for the mean normalized concentration at X=\:
95% UCL = exp[?h + 2.92 s(YJ] kg/m3 per g/rrf-s
where
Yh = -5.3466 + 0.1004 In (X,)
= the standard error of th = 0.0269 [1/4 + (ln(XJ - 11.0509)2/26.3608]
For example, the estimated 95% UCL for the normalized concentration for an
area of 2025 m2 is exp(-4.5822 + 2.92 x 0.0187) = 0.0108 kg/m3 per g/rrf-s. A
detailed example of the calculation is presented in Appendix F.
1:5 Rectangular Area Soarce-The linear relationship between the In-normalized
concentration (kg/m3 per g/mf-s on original scale) and the In of the area (m2 on
original scale) is illustrated in Figure 3-9. Each of the 29 concentrations at each of the
four sizes are plotted along with the regression line. The Pearson correlation coefficient
was 0.77, which indicates a strong linear relationship between the two variables. The
slope of the estimated linear regression equation was significantly greater than zero
(p =0.0001), which also indicates a significant linear association. The estimated linear
regression equation for estimating the mean In-normalized concentration for a given
size area is presented below:
? = -5.2263 + 0.0798 ln(X)
where
¥ is the estimated mean In-normalized concentration
X is the size of the area in m2.
42
-------
-3.6:
~ ~37 ~-
I -3.M
-3.9:
-4.1
-4.2
§
2 E
»»
E i
I
•4.6
-4.7
-4.8
•4.9
Y = -5.2263 + 0.0798 ln(X)
I I I I I I I I I I I I I I I I I I I I I I I I I
i i i i [ i i i i i i i i i | i i i i i i i i i [ i i i
10 11 12
I
Natural Logarithm of Area ( X ), m2 on original scale
13
14
15
Figure 9. Regression analysis of normalized concentrations for a rectangle 1:5 area source.
-------
The exponential of Y is then calculated to estimate the normalized concentration
on the original scale fl.e., kg/m3 per g/rrf-s). For example, the estimated mean
normalized concentration for an area of 2025 m2 is exp(-4.6188) = 0.0099 kg/m3 per
g/mf-s. A detailed example of the calculation is presented in Appendix F.
The following equation is then used to estimate the upper 95% confidence limit
(UCL) for the mean normalized concentration at X=\:
95% UCL = exp[Yh + 2.92 s(Yh)] kg/m3 per g/m2-s
where
Yh = -5.2263 + 0.0798
s(yh) = the standard error of Yh = 0.0287 [1/4 + (Info) - 1 1 .0509)2/26.3608]
For example, the estimated 95% UCL for the normalized concentration for an
area of 2025 m2 is exp(-4.6188 + 2.92 x 0.0200) = 0.0105 kg/m3 per g/rrf-s. A
detailed example of the calculation is presented in Appendix F.
3.7 SUBSTITUTION OF METHODOLOGY IN RAGS - PART B EQUATIONS
From the regression analyses performed on the modeled data, a series of
regression equations were constructed to represent the mean of the normalized
concentration (C/Q)^. The 95% UCL of the mean (C/Q)<,S% UCL for each regression
was also calculated. In order to derive the VF and PEF in the RAGS - Part B Equation
Nos. 8 and 9, the inverse of the mean normalized concentration (Q/C),^ was
calculated as well as the inverse of the 95% UCL of the mean concentration
(Q/C)95%UCL-
To calculate (Q/C)^ as a function of the area of the applicable source shape
and size, the natural logarithm of the source area is entered into the applicable
regression equation. For example, given a square source with an area of 2,205 m2 the
solution is:
44
-------
- [exp (0.1005X - 5.3880)]
-1
where:
therefore,
X = 1n (2025) = 7.613
(QlO)meM = [exp (0.1005(7.613) - 5.3880)]
-1
101-78
UCL as a function of the area of the applicable source shape and size, is
therefore calculated for the same square source as:
[Yh - 2.92
-i
where:
Y
X
Y
X
0.1005X - 5.3880
natural log of the square area in m2
s(Y,) = 0.02128
I.0509)2
(26.3608)
since:
= 7.613 =
45
-------
then,
Y =0.1005 (7.613) - 5.3880
Y = -4.623 = Yh
and,
s(Yf) = 0.02128
025 (7-613 - 11.0509)*
26.3608
s(Yh) = 0.01486
therefore,
(Q/C)95%ucL = (exp [-4.623 + 2.92 (0.01486)])'1
(Q/C)95%UCL = 97.46 (g/m2 - s per kg/m3)
Once (Q/C) is determined, its value need only be multiplied by the inverse of the
emission flux in units of g/m2 - s to derive the inverse of the maximum annual average
concentration at the center of the area source (i.e., the term VF).
3.7.1 Application of (Q/C) to the RAGS - Part B VF Equation
In order to apply either (Q/C)^^ or (Q/C^ UCL to the RAGS - Part B VF
equation, the existing conversion factor of 103 kg/g was deleted because this
conversion was incorporated into the value of (Q/C). In addition, because the Hwang
and Falco model predicts the inverse of the emission flux in units of cm2 - s/g, a
conversion factor of 10"4 nf/cm2 was added to change these units to m2 - s/g in order
to maintain the correct units when multiplied by (Q/C).
Therefore the revised VF equation is:
46
-------
* -14"
"ei x "a X
3.7.2 Application of (Q/C) to the RAGS - Part B PEF Equation
In order to apply the same methodology to the RAGS - Part B PEF equation, the
existing conversion factor of 1000 g/kg was deleted because this conversion was
incorporated into the value of (Q/C). In addition, because the Cowherd model predicts
the inverse of the emission flux in units of m2 - h/g, the existing conversion factor of
3600 s/h was retained in order to maintain the proper units when multiplied by (Q/C).
Therefore, the revised PEF equation is:
PEF (m*lkg) = (QIC)
0.036 x (1 -G) x (UJUf x F(x)
47
-------
SECTION 4
COMPARISON OF RESULTS AND CONCLUSIONS
4.1 COMPARISON OF RESULTS
In order to compare the results of the project methodology to that of the existing
box dispersion model, the inhalation PRG for benzene was calculated using both
methods for the three source shapes at areas of 0.5, 5, 50 and 500 acres.
The emission flux calculated by the Hwang and Falco model used the following
variables:
& =1.5 g/cm3
p. = 2.65 g/cm3
0 =0.1 cm3-water/g-soil
D, = 0.0871 cmVs
H = 0.00559 atm-m3/mol
K.C =83 cm3/g
OC = 0.02
T = 7.9 x 108 s.
The mean of the normalized concentration (C/Q)man and the 95% UCL of the
mean normalized concentration (C/Q)g5% UCL were calculated for each source shape
and each area size using the appropriate regression equations along with their
respective inverse values [(Q/C^ and (Q/C),,5% UCJ. The values of the VF for
benzene were then calculated using the existing box model, (Q/CXnean, and
(Q/C)g5% UCL. From these data, the PRG was calculated using all three methods. The
inhalation PRG for benzene was calculated using the standard commerciaJ/industriaJ
scenario as:
48
-------
pRG _ 77? x AT x 365 dayslyear
(PfttW) ~
where:
PRQ = preliminary remediation goal (ppmw)
TR = target risk (1 x 1fJe)
AT = averaging time (30 years)
URFj = inhalation unit risk factor (8.3 x 10fl //g/m3 for benzene)
EF = exposure frequency (250 days/year)
ED = exposure duration (25 years)
VF = volatilization factor (m3/kg).
Table 4-1 shows the results of the comparison. Rgure 4-1 illustrates the relative
change in the PRG calculated using (Q/C)^, while Rgure 4-2 shows the same
relationships when the PRG is calculated using (Q/C)a6%UCL. As can be seen from both
Figures 4-1 and 4-2, the box model yields the same PRG regardless of source size (i.e.,
a straight line). This is because both the length of side of the contaminated area (LS)
and the area of contamination (A) change proportionally, while the diffusion height (HD)
remains constant. Given this relationship, the only way to change the value of VF, and
thus the PRG given a constant emission flux, is to change the value of the windspeed in
the mixing zone (V).
Figures 4-1 and 4-2 and the VF values calculated by the project methodology in
Table 4-1 show that both the values of the VF and the PRG decrease with an increase
in source size. This is to be expected in that the MEI is exposed to a larger upwind
emission contribution by an increasingly larger area source. Therefore, as the ambient
concentration at the MEI increases with source size, the allowed soil concentration
(PRG) decreases proportionally at the same target risk. This is not accounted for in the
box model.
49
-------
TABLE 4-1. COMPARISON OF EXISTING VERSUS PROJECT METHODOLOGY FOR BENZENE
Area
Contaminant (acres)
SQUARE
Benzene 0.5
Benzene 5
Benzene 50
Benzene 500
1:3 RECTANGLE
Benzene 0.5
Benzene 5
Benzene 50
Benzene 500
1.5 RECTANGLE
Benzene 0.5
Benzene 5
Benzene 50
Benzene 500
Invert* of
emission flux
(cm2-«/kg)
1.06E+09
1. QBE +09
1.08E+09
1.08E+09
1.08E+09
1.08E+09
1 OBE+09
1.08E+09
1 08E+09
1.08E+09
1.08E+09
1.06E+09
(C/Q)
Mean
(kg/m3/g/m2-»)
0.0098
0.0124
0.0156
0.0197
0.0102
0.0129
0.0102
0.0205
0.0099
0.0119
0.0142
0.0171
(C/Q)
95%UCL
(kg/m3/g/m2-i)
0.0103
0.0126
0.0159
0.0206
0.0106
0.0132
0.0166
00216
0.0105
0.0122
0.0146
0.0162
(QIC)
Mean
(g/m2-s/kg/m3)
101.78
60.76
64.07
50.64
97.73
77.56
61.55
48.65
101.37
84.35
70.19
58.41
(QIC)
95NUCL
(g/m2-t/kg/m3)
97.46
79.27
62.88
46.65
92.52
75.76
60.11
46.21
95.61
82.27
68.44
55.05
VF using
box model
(m3/kg)
10833
10833
10833
10833
10833
10833
10833
10833
10833
10833
10833
10833
VF using
(QIC)
Mean
(m3/kg)
11026
8748
6941
5507
10587
8402
6668
5291
10981
9138
7604
6328
VF using
(QIC)
95%UCL
(m3/kg)
10558
8587
6812
5270
10023
8207
6512
5006
10357
8912
7414
5964
PRG using
box model
(ppm)
5.34
5.34
5.34
5.34
5.34
5.34
5.34
5.34
5.34
5.34
5.34
5.34
PRG using
(QIC)
Mean
(ppm)
5.43
4.31
3.42
2.71
5.21
4.14
3.28
2.61
5.41
4.50
3.75
3.12
PRG using
(0/C)
95%UCL
(ppm)
5.20
4.23
3.36
2.60
4.94
4.04
3.21
2.47
5.10
4.39
3.65
2.94
-------
10
Q.
CD
oc
Q_
1
1E+03
1E+04 1E+05 1E+06
Area of Source (m2)
1E+07
Box Model -•- Square -S- 1:3 Rect.
1:5 Rect.
Figure 4-1. Benzene PRG versus source shape and size using the inverse of the
mean normalized concentration.
51
-------
10
Q.
a.
CD
o:
o_
1E+03' ' "VE+041 ' "1E+051 "'lE+Oe' ' "lE+07
Area of Source (m2)
Box Model -•- Square -&- 1:3 Rect. -s- 1:5 Rect.
Figure 4-2. Benzene PRG versus source shape and size using the inverse of the
95% UCL mean normalized concentration.
52
-------
As can be seen from the data, the difference in values of the PRG as a function
of source shape is relatively small and ranges from approximately 5 percent at an area
of 0.5 acres to approximately 16 percent at 500 acres. The reason for this relatively
small difference is due primarily to the effect of calculating an annual (long-term)
average maximum concentration which tends to average any particular worst-case
meteorological conditions. This tends to reduce significantly the effects of wind
direction and windspeed at the center of the area source regardless of source shape.
4.2 CONCLUSIONS
The conclusions drawn from this study are listed below.
1. The existing box model does not account for the increase in the ambient
concentration at the MEI due to the increased upwind emission
contribution associated with progressively larger area sources.
2. The project methodology adequately accounts for all sources of
continuous, uniform emissions as the maximum annual average
concentration at the MEI.
3. The project methodology determined that the MEI (point of maximum
impact) for annual average ambient concentrations occurs at the center of
the three source shapes included in the analysis.
4. The project methodology provides a theoretically reasonable National
estimate of the maximum annual average concentration at the MEI within
the range of source sizes and shapes included in the study.
5. The project methodology is not applicable for certain
topographical/meteorological conditions (e.g., coastal sea breeze
phenomenon, complex terrain, etc.).
6. The relative maximum difference between the PRGs calculated for the
three source shapes for each given area was approximately 16 percent.
7. The relative maximum difference between the PRGs calculated using the
inverse of the mean normalized concentration and the inverse of the 95%
UCL of the mean normalized concentration was approximately 6 percent.
53
-------
8. Use of the applicable regression equation and associated calculation of
the 95% UCL is a reasonable substitute for the existing box model for
calculating the RAGS - Part B volatilization factor and particulate emission
factor.
Given the relatively small maximum percent difference between the PRGs
calculated for the three source shapes (\.e., approximately 16 percent), the final
methodology incorporates the regression and 95% UCL equations for the 1:3 aspect
ratio rectangular source; this source shape yields the most conservative (lowest) PRG.
The revisions to the RAGS - Part B VF and PEF equations found in Appendix D have
incorporated this approach.
Finally, Appendix D also includes two versions each of the revised VF and PEF
equations. Figures D-1 and D-3 calculate (Q/C) using the 95% UCL of the mean
normalized concentration, while Figures D-2 and D-3 calculate (Q/C) for the-VF and the
PEF using the mean of the normalized concentration. Employing the regression and
95% UCL equations for the 1:3 rectangle, this will produce a maximum relative
difference between the PRGs so calculated of approximately 6 percent.
54
-------
REFERENCES
1. Cowherd, C., Muleski, G., Engelhart, P., and Gillette, D. Rapid Assessment of
Exposure to Particulate Emissions from Surface Contamination Sites. EPA-
600/8-85-002. U.S. Environmental Protection Agency. Prepared for the Office of
Health and Environmental Assessment, Washington, D.C.
2. U.S. Environmental Protection Agency. Guideline on Air Quality Models. Office
of Air Quality Planning and Standards, Research Triangle Park, N.C. EPA-
450/2-78-027R, 1986.
3. U.S. Environmental Protection Agency. Review and Evaluation of Area Source
Dispersion Algorithms for Emission Sources at Superfund Sites. Office of Air
Quality Planning and Standards, Research Triangle Park, N.C. EPA-450/4-89-
020, 1989.
4. U.S. Environmental Protection Agency. Risk Assessment Guidance for
Superfund: Volume I - Human Health Evaluation Manual (Part B, Development
of Risk-Based Preliminary Remediation Goals). Publication No. 9285.7-01 B.
Office of Emergency and Remedial Response, Washington, D.C. 1991.
5. U.S. Environmental Protection Agency. User's Guide for the Industrial Source
Complex (ISC2) Dispersion Models. Office of Air Quality Planning and
Standards, Research Triangle Park, N.C. EPA-450/4-92-008a, 1992.
6. U.S. Environmental Protection Agency. Sensitivity Analysis of the Revised Area
Source Algorithm for the ISC2 Short-Term (ISCST2) Model, Draft. Office of Air
Quality Planning and Standards, Research Triangle Park, N.C. 1992a.
7. Hwang, S. T., and FaJco, J. W. Estimation of Multimedia Exposure Related to
Hazardous Waste Facilities. Cohen, Y. (Ed), Plenum Publishing Corp. 1986.
8. Peterson, W. B., and Rumsey, W. B. User's Guide for PAL 2.0 - A Gaussian -
Plume Algorithm for Point, Area, and Line Sources. EPA-600/8-87-009. U.S.
Environmental Protection Agency, Research Triangle Park, N.C. 1987.
55
-------
9. TRC Environmental Consultants. User's Guide for the Fugitive Dust Model
(FDM), (Revised). EPA-910/9-88-202R. U.S. Environmental Protection Agency,
Region X, Seattle, Washington. 1990.
10. Wilson, R. B. Simple Area Source Algorithm Revisited. Memorandum to Pat
Cirone, Chief of Health and Environmental Assessment Section. U.S.
Environmental Protection Agency, Region X, Seattle, Washington. 1991.
11. National Oceanic and Atmospheric Administration. Climatic Atlas of the United
States. National Climatic Center. Asheville, North Carolina.
56
-------
APPENDIX A
AREA SOURCE COORDINATES FOR EACH SHAPE
AND SIZE USED IN THE STUDY
A-1
-------
Sailor*
45 x 45 m 0.5 Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
x-sw
m
22.500000
22.500000
22.500000
22.500000
22.500000
16.071430
16.071430
12.857140
12.857140
12.857140
-9.642860
12.857140
12.857140
12.857140
-9.642860
12.857140
-9.642860
12.857140
12.857140
-6.428570
-6.428570
-6.428570
-6.428570
-6.428570
-6.428570
-6.428570
-6.428570
-3.214290
-3.214290
-3.214290
-3.214290
-3.214290
-3.214290
-3.214290
-3.214290
-3.214290
-3.214290
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
3.214290
3.214290
3.214290
3.214290
3.214290
3.214290
3.214290
Y-SW
m
22.500000
12.857140
-3.214290
3.214290
12.857140
-3.214290
0.000000
22.600000
12.857140
-6.428570
-6.428570
-3.214290
-3.214290
0.000000
0.000000
3.214290
3.214290
6.428570
12.857140
12.857140
-9.642860
-6.428570
-3.214290
0.000000
3.214290
6.428570
9.642860
-22.50000C
-16.07143C
-12.857140
-9.642860
-6.428570
3.214290
6.428570
9.642860
1 2.857 14C
16.07143C
-16.07143C
-12.857141
-9.64286C
-6.42857C
3.214290
6.428570
9.642860
12.85714C
-12.857140
-9.64286C
-6.42857C
-3.21 429C
O.OOOOOC
3.21 429C
6.42857C
Sid* Length
m
9.642857
9.642857
6.428571
9.642857
9.642857
3.214286
3.214286
9.642857
6.428571
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
6.428571
9.642857
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
6.428571
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
6.428571
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
-------
Square
45 x 45 m 0.5 Acres
ID
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
X-SW
m
3.214290
3.214290
6.428570
6.428570
9.642860
6.428570
9.642860
6.428570
9.642860
6.428570
9.642860
6.428570
3.214290
12.85714O
12.857140
16.071430
12.857140
12.857140
12.857140
12.857140
-3.214290
-3.214290
-3.214290
-3.214290
-1.607140
-1.607140
0.000000
0.000000
1.607140
1.607140
1.607140
1.607140
-1.607140
-1.607140
-1.607140
-1.607140
-0.803570
-0.803570
0.000000
0.000000
0.803570
0.803570
0.803570
0.803570
Y-SW
m
9.642860
-22.500000
-12.857140
-6.428570
-6.428570
-3.214290
-3.214290
0.000000
0.000000
3.214290
3.214290
6.428570
12.857140
-22.500000
-12.857140
-3.214290
-3.214290
0.000000
3.214290
12.857140
-3.214290
-1.607140
0.000000
1.607140
-3.214290
1.607140
-3.214290
1.607140
-3.214290
-1.607140
0.000000
1.607140
-1.607140
-0.803570
0.000000
0.803570
-1.607140
0.803570
-1.607140
0.803570
-1.607140
-0.803570
0.000000
0.803570
Ski* Length
m
3.214286
9.642857
6.428571
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
3.214286
6.428571
9.642857
9.642857
9.642857
6.428571
3.214286
3.214286
9.642857
9.642857
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
1.607143
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
0.803571
-------
Square
140 X 140 m S Acres
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
x-sw
m
-70.00
-70.00
-70.00
-70.00
-70.00
-50.00
-50.00
-40.00
-40.00
-40.00
-30.00
-40.00
-40.00
-40.00
-30.00
-40.00
-30.00
-40.00
-40.00
-20.00
-20.00
-20.00
-20.00
-20.00
-20.00
-20.00
-20.00
-10.00
-10.00
-10.00
-10.00
-10.00
-10.00
-10.00
-10.00
-10.00
-10.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
Y-SW
m
-70.00
-40.00
-10.00
10.00
40.00
-10.00
0.00
-70.00
-40.00
-20.00
-20.00
-10.00
-10.00
0.00
0.00
10.00
10.00
20.00
40.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
30.00
-70.00
-50.00
-40.00
-30.00
-20.00
10.00
20.00
30.00
40.00
50.00
-50.00
-40.00
-30.00
-20.00
10.00
20.00
30.00
40.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
Side Length
m
30.00
30.00
20.00
30.00
30.00
10.00
10.00
30.00
20.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
20.00
30.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
20.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
20.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
-------
Square
140 x 140m 5 Acres
ID
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
x-sw
m
10.00
10.00
20.00
20.00
30.00
20.00
30.00
20.00
30.00
20.00
30.00
20.00
10.00
40.00
40.00
50.00
40.00
40.00
40.00
40.00
-10.00
-10.00
-10.00
-10.00
-5.00
-5.00
0.00
0.00
5.00
5.00
5.00
5.00
-5.00
-5.00
-5.00
-5.00
-2.50
-2.50
0.00
0.00
2.50
2.50
2.50
2.50
Y-SW
m
30.00
-70.00
-40.00
-20.00
-20.00
-10.00
-10.00
0.00
0.00
10.00
10.00
20.00
40.00
-70.00
-40.00
-10.00
-10.00
0.00
10.00
40.00
-10.00
-5.00
0.00
5.00
-10.00
5.00
-10.00
5.00
-10.00
-5.00
0.00
5.00
-5.00
-2.50
0.00
2.50
-5.00
2.50
-5.00
2.50
-5.00
-2.50
0.00
2.50
Side Length
m
10.00
30.00
20.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
20.00
30.00
30.00
30.00
20.00
10.00
10.00
30.00
30.00
5.00
5.00
5.00
5.00
5.00
5.00
5.00
5.X
5.00
5.00
5.00
5.00
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.50
2.50
-------
Square
450 x 450 m SO Acres
ID
1
2
3
4
5
6
7
8
9
10
11
. 12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
x-sw
m
-225.00000
-225.00000
-225.00000
-225.00000
-225.00000
-160.71430
-160.71430
-128.57140
-128.57140
-128.57140
-96.42860
-128.57140
-128.57140
-128.57140
-96.42860
-128.57140
-96.42860
-128.57140
-128.57140
-64.28570
HS4.28570
-64.28570
-64.28570
-64.28570
-64.28570
-64.28570
-64.28570
-32.14290
-32.14290
-32.14290
-32.14290
-32.14290
-32.14290
-32.14290
-32.14290
-32.14290
-32.14290
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
32.14290
32.14290
32.14290
32.14290
32.14290
32.14290
32.14290
Y-SW
m
-225.00000
-128.57140
-32.14290
32.14290
128.57140
-32.14290
0.00000
-225.00000
-128.57140
-64.28570
64.28570
-32.14290
-32.14290
0.00000
0.00000
32.14290
32.14290
64.28570
128.57140
-128.57140
-96.42860
-64.28570
-32.14290
0.00000
32.14290
64.28570
96.42860
-225.00000
-160.71430
-128.57140
-96.42860
^4.28570
32.14290
64.28570
96.42860
128.57140
160.71430
160.71430
128.57140
96.42860
64.28570
2.14290
4.28570
6.42860
28.57140
128.57140
96.42860
64.28570
32.14290
0.00000
2.14290
4.28570
Side Length
m
96.42857
96.42857
64.28571
96.42857
96.42857
32.14286
32.14286
96.42857
64.28571
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
6458571
96.42857
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
64.28571
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
6458571
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
-------
Square
450 x 450 m SO Acres
ID
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
x-sw
m
32.14290
32.14290
64.28570
64.28570
96.42860
64.28570
96.42860
64.28570
96.42860
64.28570
96.42860
64.28570
32.14290
128.57140
128.57140
160.71430
128.57140
128.57140
128.57140
128.57140
•32.14290
-32.14290
-32.14290
-32.14290
•16.07140
-16.07140
0.00000
0.00000
16.07140
16.07140
16.07140
16.07140
-16.07140
-16.07140
-16.07140
-16.07140
-8.03570
-8.03570
0.00000
0.00000
8.03570
8.03570
8.03570
8.03570
Y-SW
m
6.42860
225.00000
128.57140
64.28570
64.28570
32.14290
32.14290
0.00000
0.00000
2.14290
2.14290
4.28570
28.57140
-225.00000
-128.57140
-32.14290
-32.14290
0.00000
32.14290
128.57140
-32.14290
-16.07140
0.00000
16.07140
-32.14290
16.07140
32.14290
6.07140
32.14290
16.07140
0.00000
6.07140
-16.07140
-8.03570
0.00000
8.03570
-16.07140
8.03570
16.07140
8.03570
16.07140
8.03570
0.00000
8.03570
Ski* Length
m
32.14286
96.42857
64.28571
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
32.14286
64.28571
96.42857
96.42857
96.42857
64.28571
32.14286
32.14286
96.42857
96.42857
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
16.07143
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
8.03571
-------
Square
1400 x 1400 m SOOAcres
ID
1
2
3
4
5
6
7
6
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
x-sw
m
-700.00
-700.00
-700.00
-700.00
-700.00
-500.00
-500.00
-400.00
-400.00
-400.00
-300.00
-400.00
-400.00
-400.00
-300.00
-400.00
-300.00
-400.00
-400.00
-200.00
-200.00
-200.00
-200.00
-200.00
-200.00
-200.00
-200.00
-100.00
-100.00
-100.00
-100.00
-100.00
-100.00
-100.00
-100.00
-100.00
-100.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
Y-SW
m
-700.00
-400.00
-100.00
100.00
400.00
-100.00
0.00
-700.00
-400.00
-200.00
-200.00
-100.00
-100.00
0.00
0.00
100.00
100.00
200.00
400.00
-400.00
-300.00
-200.00
-100.00
0.00
100.00
200.00
300.00
-700.00
-500.00
-400.00
-300.00
-200.00
100.00
200.00
300.00
400.00
500.00
-500.00
-400.00
-300.00
-200.00
100.00
200.00
300.00
400.00
-400.00
-300.00
-200.00
-100.00
0.00
100.00
200.00
Side Length
m
300.00
300.00
200.00
300.00
300.00
100.00
100.00
300.00
200.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
200.00
300.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
200.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
200.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
-------
Square
1400 x 1AQQ m
ID
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
x-sw
m
100.00
100.00
200.00
200.00
300.00
200.00
300.00
200.00
300.00
200.00
300.00
200.00
100.00
400.00
400.00
500.00
400,00
400.00
400.00
400.00
-100.00
-100.00
-100.00
-100.00
-50.00
-50.00
0.00
0.00
50.00
50.00
5.00
50.00
-50.00
-50.00
-50.00
-50.00
-25.00
-25.00
0.00
0.00
25.00
25.00
25.00
25.00
Y-SW
m
300.00
-700.00
-400.00
-200.00
-200.00
-100.00
-100.00
0.00
0.00
100.00
100.00
200.00
400.00
-700.00
-400.00
-100.00
-100.00
0.00
100.00
400.00
-100.00
-50.00
0.00
50.00
-100.00
50.00
-100.00
50.00
-100.00
-50.00
0.00
50.00
-50.00
-25.00
0.00
25.00
-50.00
25.00
-50.00
25.00
-50.00
-25.00
0.00
25.00
Side Length
m
100.00
300.00
200.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
200.00
300.00
300.00
300.00
200.00
100.00
100.00
300.00
300.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
-------
1: 3 Rectangle
25 x 75 m 0.5 Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-12.500
-12.500
0.000
0.000
-12.500
-12.500
-12.500
-12.500
-6.250
-6.250
-6.250
-6.250
0.000
0.000
0.000
0.000
6.250
6.250
6.250
6.250
-12.500
-12.500
-12.500
-12.500
-12.500
-12.500
-12.500
-12.500
-9.375
-9.375
-9.375
-9.375
-9.375
-9.375
-9.375
-9.375
-6.250
-6.250
-6.250
-6.250
-3.125
-3.125
-3.125
-3.125
0.000
0.000
Y-SW
m
-37.500
25.000
-37.500
25.000
-25.000
-18.750
12.500
18.750
-25.000
-18.750
12.500
18.750
-25.000
-18.750
12.500
18.750
-25.000
-18.750
12.500
18.750
-12.500
-9.375
-6.250
-3.125
0.000
3.125
6.250
9.375
-12.500
-9.375
-6.250
-3.125
0.000
3.125
6.250
9.375
-12.500
-9.375
6.250
9.375
-12.500
-9.375
6.250
9.375
-12.500
-9.375
Side Length
m
12.5
12.5
12.5
12.5
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
-------
1: 3 Rectangle
25 x 75 m 0.5 Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
0.000
0.000
3.125
3.125
3.125
3.125
6.250
6.250
6.250
6.250
6.250
6.250
6.250
6.250
9.375
9.375
9.375
9.375
9.375
9.375
9.375
9.375
-6.250
-6.250
-6.250
-6.250
-6.250
-3.750
-3.750
-1.250
-1.250
1.250
1.250
3.750
3.750
3.750
3.750
3.750
-3.750
-3.750
-3.750
-3.750
-3.750
-3.750
-2.500
-2.500
Y-SW
m
6.250
9.375
-12.500
-9.375
6.250
9.375
-12.500
-9.375
-6.250
-3.125
0.000
3.125
6.250
9.375
-12.500
-9.375
-6.250
-3.125
0.000
3.125
6.250
9.375
-6.250
-3.750
-1.250
1.250
3.750
-6.250
3.750
-6.250
3.750
-6.250
3.750
-6.250
-3.750
-1.250
1.250
3.750
-3.750
-2.500
-1.250
0.000
1.250
2.500
-3.750
-2.500
Side Length
m
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
3.125
, 3.125
3.125
3.125
3.125
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
-------
1: 3 Rectangle
25 x 75 m 0.5 Acret
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
no
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
x-sw
m
-2.500
-2.500
-2.500
-2.500
-1.250
-1.250
-1.250
-1.250
0.000
0.000
0.000
0.000
1.250
1.250
1.250
1.250
1.250
1.250
2.500
2.500
2.500
2.500
2.500
2.500
-1.250
-1.250
-1.250
-1.250
-0.625
-0.625
O.COO
0.000
0.625
0.625
0.625
0.625
Y-SW
m
-1.250
0.000
1.250
2.500
-3.750
-2.500
1.250
2.500
-3.750
-2.500
1.250
2.500
-3.750
-2.500
-1.250
0.000
1.250
2.500
-3.750
-2.500
-1.250
0.000
1.250
2.500
-1.250
-0.625
0.000
0.625
-1.250
0.625
-1.250
0.625
-1.250
-0.625
0.000
0.625
Side Length
m
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
.25
.25
.25
.25
.25
.25
.25
.25
1.25
1.25
1.25
1.25
1.25
1.25
1.25
0.625
0.625
0.625
0.625
0.625
0.625
0.625
0.625
0.625
0.625
0.625
0.625
-------
1: 3 Rectangle
80 x 240 m 5 Acres
ID
1
2
3
4
5
6
7
8
9
10
n
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-40.00
-40.00
0.00
0.00
-40.00
-40.00
-40.00
-40.00
-20.00
-20.00
-20.00
-20.00
0.00
0.00
0.00
0.00
20.00
20.00
20.00
20.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-30.00
-30.00
-30.00
-30.00
-30.00
-30.00
-30.00
-30.00
-20.00
-20.00
-20.00
-20.00
-10.00
-10.00
-10.00
-10.00
0.00
0.00
Y-SW
m
-120.00
80.00
-120.00
80.00
-80.00
-60.00
40.00
60.00
-80.00
-60.00
40.00
60.00
-80.00
-60.00
40.00
60.00
-80.00
-60.00
40.00
60.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
30.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
30.00
-40.00
-30.00
20.00
30.00
-40.00
-30.00
20.00
30.00
-40.00
-30.00
Side Length
m
40.00
40.00
40.00
40.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
-------
1:3 Rectangle^
80 x 240 m 5 Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
0.00
0.00
10.00
10.00
10.00
10.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
30.00
-20.00
-20.00
-20.00
-20.00
-20.00
-12.00
-12.00
-4.00
-4.00
4.00
4.00
12.00
12.00
12.00
12.00
12.00
-12.00
-12.00
-12.00
-12.00
-12.00
-12.00
-8.00
-8.00
Y-SW
m
20.00
30.00
-40.00
-30.00
20.00
30.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
30.00
-40.00
-30.00
-20.00
-10.00
0.00
10.00
20.00
30.00
-20.00
-12.00
-4.00
4.00
12.00
-20.00
12.00
-20.00
12.00
-20.00
12.00
-20.00
-12.00
-4.00
4.00
12.00
-12.00
-8.00
-4.00
0.00
4.00
8.00
-12.00
-8.00
Side Lenath
m
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
8.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
-------
1:3 Rectangle
60 x 240 m S Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
x-sw
m
-8.00
-8.00
-8.00
-8.00
-4.00
-4.00
-4.00
-4.00
0.00
0.00
0.00
0.00
4.00
4.00
4.00
4.00
4.00
4.00
8.00
8.00
8.00
8.00
8.00
8.00
-4.00
-4.00
-4.00
-4.00
-2.00
-2.00
0.00
0.00
2.00
2.00
2.00
2.00
Y-SW
m
-4.00
0.00
4.00
8.00
-12.00
-8.00
4.00
8.00
-12.00
-8.00
4.00
8.00
-12.00
-8.00
-4.00
0.00
4.00
8.00
-12.00
-8.00
-4.00
0.00
4.00
8.00
-4.00
-2.00
0.00
2.00
-4.00
2.00
-4.00
2.00
-4.00
-2.00
0.00
2.00
Side Length
m
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
-------
-------
1:3 Rectangle
250 x 750 m SO Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-125.00
-125.00
0.00
0.00
-125.00
-125.00
-125.00
-125.00
-62.50
-62.50
-62.50
-62.50
0.00
0.00
0.00
0.00
62.50
62.50
62.50
62.50
-125.00
-125.00
-125.00
-125.00
-125.00
-125.00
-125.00
-125.00
-93.75
-93.75
-93.75
-93.75
-93.75
-93.75
-93.75
-93.75
-62.50
-62.50
-62.50
-62.50
-31.25
-31.25
-31.25
-31.25
0.00
0.00
Y-SW
m
-375.00
250.00
-375.00
250.00
-250.00
-187.50
125.00
187.50
-250.00
-187.50
125.00
187.50
-250.00
-187.50
125.00
187.50
-250.00
-187.50
125.00
187.50
-125.00
-93.75
-62.50
-31.25
0.00
31.25
62.50
93.75
-125.00
-93.75
-62.50
-31.25
0.00
31.25
62.50
93.75
-125.00
-93.75
62.50
93.75
-125.00
-93.75
62.50
93.75
-125.00
-93.75
Side Length
m
125.00
125.00
125.00
125.00
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
-------
1:3 Rectangle
250 x 750 m SO Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
0.00
0.00
31.25
31.25
31.25
31.25
62.50
62.50
62.50
62.50
62.50
62.50
62.50
62.50
93.75
93.75
93.75
93.75
93.75
93.75
93.75
93.75
-62.50
-62.50
-62.50
-62.50
-62.50
-37.50
-37.50
-12.50
-12.50
12.50
12.50
37.50
37.50
37.50
37.50
37.50
-37.50
-37.50
-37.50
-37.50
-37.50
-37.50
-25.00
-25.00
Y-SW
m
62.50
93.75
-125.00
-93.75
62.50
93.75
-125.00
-93.75
-62.50
-31.25
0.00
31.25
62.50
93.75
-125.00
-93.75
-62.50
-31.25
0.00
31.25
62.50
93.75
-62.50
-37.50
-12.50
12.50
37.50
-62.50
37.50
-62.50
37.50
-62.50
37.50
-62.50
-37.50
-12.50
12.50
37.50
-37.50
-25.00
-12.50
0.00
12.50
25.00
-37.50
-25.00
Side Length
m
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
31.25
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
25.00
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
-------
1: 3 Rectangle
250 x 750 m SO Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
no
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
x-sw
m
-25.00
-25.00
-25.00
-25.00
-12.50
-12.50
-12.50
-12.50
0.00
0.00
0.00
0.00
12.50
12.50
12.50
12.50
12.50
12.50
25.00
25.00
25.00
25.00
25.00
25.00
-12.50
-12.50
-12.50
-12.50
-6.25
-6.25
0.00
0.00
6.25
6.25
6.25
6.25
Y-SW
m
-12.50
0.00
12.50
25.00
-37.50
-25.00
12.50
25.00
-37.50
-25.00
12.50
25.00
-37.50
-25.00
-12.50
0.00
12.50
25.00
-37.50
-25.00
-12.50
0.00
12.50
25.00
-12.50
-6.25
0.00
6.25
-12.50
6.25
-12.50
6.25
-12.50
-6.25
0.00
6.25
Side Length
m
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
12.50
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
6.25
-------
1: 3 Rectangle
800 x 2400 m 500 Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-400.00
-400.00
0.00
0.00
-400.00
-400.00
-400.00
-400.00
-200.00
-200.00
-200.00
-200.00
0.00
0.00
0.00
0.00
200.00
200.00
200.00
200.00
-400.00
-400.00
-400.00
-400.00
-400.00
-400.00
-400.00
-400.00
-300.00
-300.00
-300.00
-300.00
-300.00
-300.00
-300.00
-300.00
-200.00
-200.00
*00.00
-200.00
-100.00
-100.00
-100.00
-100.00
0.00
0.00
Y-SW
m
-1200.00
800.00
-1200.00
800.00
-800.00
-600.00
400.00
600.00
-800.00
-600.00
400.00
600.00
-800.00
-600.00
400.00
600.00
-800.00
-600.00
400.00
600.00
-400.00
-300.00
-200.00
-100.00
0.00
100.00
200.00
300.00
-400.00
-300.00
-200.00
-100.00
0.00
100.00
200.00
300.00
-400.00
-300.00
200.00
300.00
-400.00
-300.00
200.00
300.00
-400.00
-300.00
Side Length
m
400.00
400.00
400.00
400.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
-------
1:3 Rectangle
8QQ x 2400 m 500 Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
0.00
0.00
100.00
100.00
100.00
100.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
200.00
300.00
300.00
300.00
300.00
300.00
300.00
300.00
300.00
-200.00
-200.00
-200.00
-200.00
-200.00
-120.00
-120.00
-40.00
-40.00
40.00
40.00
120.00
120.00
120.00
120.00
120.00
-120.00
-120.00
-120.00
-120.00
-120.00
-120.00
-80.00
-80.00
Y-SW
m
200.00
300.00
-400.00
-300.00
200.00
300.00
-400.00
-300.00
-200.00
-100.00
0.00
100.00
200.00
300.00
-400.00
-300.00
-200.00
-100.00
0.00
100.00
200.00
300.00
-200.00
-120.00
-40.00
40.00
120.00
-200.00
120.00
-200.00
120.00
-200.00
120.00
-200.00
-120.00
-40.00
40.00
120.00
-120.00
-80.00
-40.00
0.00
40.00
80.00
-120.00
-80.00
Side Length
m
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
80.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
-------
1: 3
fiOQx 2400 m 500 Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
x-sw
m
-80.00
-80.00
-80.00
-80.00
-40.00
-40.00
-40.00
-40.00
0.00
0.00
0.00
0.00
40.00
40.00
40.00
40.00
40.00
40.00
80.00
80.00
80.00
80.00
80.00
80.00
-40.00
-40.00
-40.00
-40.00
-20.00
-20.00
0.00
0.00
20.00
20.00
20.00
20.00
Y-SW
m
-40.00
0.00
40.00
80.00
-120.00
-80.00
40.00
80.00
-120.00
-80.00
40.00
80.00
-120.00
-80.00
-40.00
0.00
40.00
80.00
-120.00
-80.00
-40.00
0.00
40.00
80.00
-40.00
-20.00
0.00
20.00
-40.00
20.00
-40.00
20.00
-40.00
-20.00
0.00
20.00
Side Length
m
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
-------
1 : 5
2Qx IQQm Q.S Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-10.00
-10.00
-10.00
-10.00
-10.00
-10.00
0.00
0.00
0.00
0.00
0.00
0.00
-10.00
-10.00
-10.00
-10.00
-5.00
-5.00
-5.00
-5.00
0.00
0.00
0.00
0.00
5.00
5.00
5.00
5.00
-10.00
-10.00
-10.00
-10.00
-10.00
6.00
6.00
6.00
6.00
6.00
-6.00
-6.00
-6.00
-6.00
-6.00
-6.00
-6.00
-6.00
Y-SW
m
-50.00
-40.00
-30.00
20.00
30.00
40.00
-50.00
-40.00
-30,00
20.00
30.X
40.00
-20.00
-15.00
10.00
15.00
-20.00
-15.00
10.00
15.00
-20.00
-15.00
10.00
15.00
-20.00
-15.00
10.00
15.00
-10.00
•6.00
-2.00
2.X
6.X
-10.X
-6.X
-2.X
2.X
6.X
-10.X
-8.X
-6.X
-4.X
-2.X
O.X
2.X
4.X
Side Length
m
10.X
10.X
10.X
10.X
10.X
10.X
10.X
10.X
10.X
10.X
10.X
10.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
5.X
4.X
4.X
4.X
4.X
4.X
4.X
4.X
4.X
4.X
4.X
2.X
2.X
2.X
2.X
2.X
2.X
2.X
2.X
-------
1 : S Rectangle
20 x 100 m Q.S Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
-6.00
-6.00
-4.00
-4.00
-4.00
-4.00
-4.00
-4.00
-4.00
-4.00
-4.00
-4.00
-2.00
-2.00
-2.00
-2.00
-2.00
-2.00
-2.00
-2.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
4.00
Y-SW
m
6.00
8.00
-10.00
-8.00
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
-10.00
-8.00
-6.00
-4.00
2.00
4.00
6.00
8.00
-10.00
-8.00
-6.00
-4.00
2.00
4.00
6.00
8.00
-10.00
-8.00
-6.00
-4.00
-2.00
0.00
2.00
4.00
6.00
8.00
-10.00
-8.00
-6.00
-4.00
-2.00
0.00
2.00
4.00
Side Length
m
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
-------
1 : S Rectangle
20 x 100 m 0.5 Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
106
106
x-sw
m
4.00
4.00
-2.00
-2.00
-2.00
-2.00
-1.00
-1.00
0.00
0.00
1.00
1.00
1.00
1.00
Y-SW
m
6.00
8.00
-2.00
-1.00
0.00
1.00
-2.00
1.00
-2.00
1.00
-2.00
-1.00
0.00
1.00
Side Length
m
2.00
2.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-------
1 :5 Rectangle
64 x 320 m 5 Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-32.00
-32.00
-32.00
-32.00
-32.00
-32.00
0.00
0.00
0.00
0.00
0.00
0.00
-32.00
-32.00
-32.00
-32.00
-16.00
-16.00
-16.00
-16.00
0.00
0.00
0.00
0.00
16.00
16.00
16.00
16.00
-32.00
-32.00
-32.00
-32.00
-32.00
19.20
19.20
19.20
19.20
19.20
-19.20
-19.20
-19.20
-19.20
-19.20
-19.20
-19.20
-19.20
Y-SW
m
-160.00
-128.00
-96.00
64.00
96.00
128.00
-160.00
-128.00
-96.00
64.00
96.00
128.00
-64.00
-48.00
32.00
48.00
-64.00
-48.00
32.00
48.00
-64.00
-48.00
32.00
48.00
-64.00
-48.00
32.00
48.00
-32.00
-19.20
-6.40
6.40
19.20
-32.00
-19.20
-6.40
6.40
19.20
-32.00
-25.60
-19.20
-12.80
-6.40
0.00
6.40
12.80
Side Length
m
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
16.00
12.80
12.80
12.80
12.80
12.80
12.80
12.80
12.80
12.80
12.80
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
-------
1 :5 Rectangle
64 x 320 m 5 Acres
ID
47
48
49
50
51
52
53
54
55
56
-57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
-19.20
-19.20
-12.80
-12.80
-12.80
-12.80
-12.80
-12.80
-12.80
-12.80
-12.80
-12.80
-6.40
-6.40
-6.40
-6.40
-6.40
-6.40
-6.40
-6.40
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
12.80
12.80
12.80
12.80
12.80
12.80
12.80
12.80
Y-SW
m
19.20
25.60
-32.00
-25.60
-19.20
-12.80
-6.40
0.00
6.40
12.80
19.20
25.60
-32.00
-25.60
-19.20
-12.80
6.40
12.80
19.20
25.60
-32.00
-25.60
-19.20
-12.80
6.40
12.80
19.20
25.60
-32.00
-25.60
-19.20
-12.80
-6.40
0.00
6.40
12.80
19.80
25.60
-32.00
-25.60
-19.20
-12.80
-6.40
0.00
6.40
12.80
Side Length 1
m
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40'
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
6.40
-------
1 : S Rectangle
64 x 320 m 5 Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
x-sw
m
12.80
12.80
-6.40
-6.40
-6.40
-6.40
-3.20
-3.20
0.00
0.00
3.20
3.20
3.20
3.20
Y-SW
m
19.20
25.60
-6.40
-3.20
0.00
3.20
-6.40
3.20
-4.40
3.20
-6.40
-3.20
0.00
3.20
Side Length
m
6.40
6.40
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
3.20
-------
1 :5 Rectangle
200 x 1000m SO Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-100.00
-100.00
-100.00
-100.00
-100.00
-100.00
0.00
0.00
0.00
0.00
0.00
0.00
-100.00
-100.00
-100.00
-100.00
-50.00
-50.00
-50.00
-50.00
0.00
0.00
0.00
0.00
50.00
50.00
50.00
50.00
-100.00
-100.00
-100.00
-100.00
-100.00
60.00
60.00
60.00
60.00
60.00
-60.00
-60.00
-60.00
-60.00
-60.00
-60.00
-60.00
-60.00
Y-SW
m
-500.00
-400.00
-300.00
200.00
300.00
400.00
-500.00
-400.00
-300.00
200.00
300.00
400.00
-200.00
-150.00
100.00
150.00
-200.00
-150.00
100.00
150.00
-200.00
-150.00
100.00
150.00
-200.00
-150.00
100.00
150.00
-100.00
-60.00
-20.00
20.00
60.00
-100.00
-60.00
-20.00
20.00
60.00
-100.00
-80.00
-60.00
-40.00
-20.00
0.00
20.00
40.00
Side Length
m
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
50.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
-------
1 ; 5 Rectangle
200 x 1000 m 50 Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
-60.00
-60.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-40.00
-20.00
-20.00
-20.00
-20.00
-20.00
-20.00
-20.00
-20.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
Y-SW
m
60.00
80.00
-100.00
-80.00
-60.00
-40.00
-20.00
0.00
20.00
40.00
60.00
80.00
-100.00
-80.00
-60.00
-40.00
20.00
40.00
60.00
80.00
-100.00
-80.00
-60.00
-40.00
20.00
40.00
60.00
80.00
-100.00
-80.00
-60.00
-40.00
-20.00
0.00
20.00
40.00
60.00
80.00
-100.00
-80.00
-60.00
-40.00
-20.00
0.00
20.00
40.00
Side Length
m
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
' 20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
20.00
-------
1 : S Rectangle
200 Jt 1000 m 50 Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
X-SW
m
40.00
40.00
-20.00
-20.00
-20.00
-20.00
-10.00
-10.00
0.00
0.00
10.00
10.00
10.00
10.00
Y-SW
m
60.00
80.00
-20.00
-10.00
0.00
10.00
-20.00
10.00
-20.00
10.00
-20.00
-10.00
0.00
10.00
Side Length
m
20.00
20.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
10.00
-------
1 :5 Rectangle
640 x 3200m 500 Acres
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
x-sw
m
-320.00
-320.00
-320.00
-320.00
-320.00
-320.00
0.00
0.00
0.00
0.00
0.00
0.00
-320.00
-320.00
-320.00
-320.00
-160.00
-160.00
-160.00
-160.00
0.00
0.00
0.00
0.00
160.00
160.00
160.00
160.00
-320.00
-320.00
-320.00
-320.00
-320.00
192.00
192.00
192.00
192.00
192.00
-192.00
-192.00
-192.00
-192.00
-192.00
-192.00
-192.00
-192.00
Y-SW
m
-1600.00
-1280.00
-960.00
640.00
960.00
1280.00
-1600.00
-1280.00
-960.00
640.00
960.00
1280.00
-640.00
-480.00
320.00
480.00
-640.00
-480.00
320.00
480.00
-640.00
-480.00
320.00
480.00
-640.00
-480.00
320.00
480.00
-320.00
-192.00
-64.00
64.00
192.00
-320.00
-192.00
-64.00
64.00
192.00
-320.00
-256.00
-192.00
-128.00
-64.00
0.00
64.00
128.00
Side Length
m
320.00
320.00
320.00
320.00
320.00
320.00
320.00
320.00
320.00
320.00
320.00
320.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
160.00
128.00
128.00
128.00
128.00
128.00
128.00
128.00
128.00
128.00
128.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
-------
1 :5 Rectangle
640 x 3200 m 500 Acres
ID
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
x-sw
m
-192.00
-192.00
-128.00
-128.00
-128.00
-128.00
-128.00
-128.00
-128.00
-128.00
-128.00
-128.00
-64.00
-64.00
-64.00
-64.00
-64.00
-64.00
-64.00
-64.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
128.00
128.00
128.00
128.00
128.00
128.00
128.00
128.00
Y-SW
m
192.00
256.00
-320.00
-256.00
-192.00
-128.00
-64.00
0.00
64.00
128.00
192.00
256.00
-320.00
-256.00
-192.00
-128.00
64.00
128.00
192.00
256.00
-320.00
-256.00
-192.00
-128.00
64.00
128.00
192.00
256.00
-320.00
-256.00
-192.00
-128.00
-64.00
aoo
64.00
128.00
198.00
256.00
-320.00
-256.00
-192.00
-128.00
-64.00
0.00
64.00
128.00
Side Length
m
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
64.00
-------
1 : S Rectangle
MO x 3200m 500 Acres
ID
93
94
95
96
97
98
99
100
101
102
103
104
105
106
x-sw
m
128.00
128.00
-64.00
-64.00
-64.00
-64.00
-32.00
-32.00
0.00
0.00
32.00
32.00
32.00
32.00
Y-SW
m
192.00
256.00
-64.00
-32.00
0.00
32.00
-64.00
32.00
-64.00
32.00
-64.00
-32.00
0.00
32.00
Side Length
m
64.00
64.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
-------
APPENDIX B
DISTRIBUTION OF NATIONAL WEATHER SERVICE
SURFACE STATIONS BY CLIMATOLOGICAL REGION
B-1
-------
DISTRIBUTION OF NATIONAL WEATHER SERVICE SURFACE STATIONS BY CUMATOLOGICAL REGION
Region
II
II
II
II
II
III
III
III
III
III
III
III
III
III
III
III
III
III
III
III
III
III
III
III
IV
IV
IV
IV
State
California
California
California
California
California
Oregon
Oregon
Oregon
Oregon
Oregon
Oregon
Washington
Washington
California
California
California
California
California
Arizona
Arizona
Arizona
Nevada
Nevada
New Mexico
New Mexico
New Mexico
Oklahoma
Oklahoma
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Colorado
Colorado
Colorado
Idaho
Running Count
Per Region
1
2
3
4
5
6
7
8
9
10
11
12
13
1
2
3
4
5
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
1
2
3
4
NWS Site
Arcata/Airport
Red Bluff/Municipal/Airport
Reddino/AAF
Sacramento/Executive Airport
San Francisco/I nt'l Airport
Astoria/Clatsop County Airport
Eugene/Mahlon Swet Airport
Medford/Jackson County Airport
North Bend/FAA Airport
Portland/lnt'l Airport
Salem/McNary Field
Olympia/Airport
Seattle/Seattle-Tacoma Int'l Airport
)aggett/FAA Airport
Fresno/Air Terminal
Los Angeles/lnt'l Airport
San Diego/Lindbergh Field
Santa Barbara/FAA Airport
Phoenix/Sky Harbor Int'l Airport
Prescott/Municipal
Tucson/lnt'l Airport
.as Vegas/McCarran Int'l Airport
Tonapah/FAA Airport
Albuquerque/lnt'l Airport
Gallup/FAA Airport
Roswell/lndustrial Air Park
Oklahoma City/Will Rogers Wor.
Tulsa/lnt'l Airport
Abilene/Municipal Airport
Amarillo/lnt'l Airport
Dallas/Fort Worth/Regional Airport
El Paso/lnt'l Airport
Lubbock/Regional Airport
Midland/Regional Air Terminal
San Angelo/WSO Airport
Stephenville
Wichita Falls/Municipal Airport 1
1
Colorado Springs/Municipal Airport
Denver/Stapleton Int'l Airport
Grand Junction/Walker Field
Boise/Air Terminal
B-2
-------
DISTRIBUTION OF NATIONAL WEATHER SERVICE SURFACE STATIONS BY CUMATOLOGICAL REGION
Region
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
State
Idaho
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Montana
Nevada
Nevada
Nevada
Nevada
Nevada
Nevada
Oregon
Oregon
Utah
Utah
Washington
Washington
Wyoming
Wyoming
Wyoming
Wyoming
Wyoming
Arkansas
Iowa
Iowa
Iowa
Iowa
Kansas
Kansas
Kansas
Kansas
Kansas
Kansas
Minnesota
Minnesota
Minnesota
Minnesota
Missouri
Missouri
Running Count
Per Region
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
NWS Site
Pocatello/Municipal Airport
Billings/Logan Int'l Airport
Glasgow/I nt'l Airport
Great Fa I Is/1 nt'l Airport
Helena/Airport
Kalispell/Glacier Pk Int'l Airport
Lewistown/FAA Airport
Miles City/Municipal Airport
Missoula/Johnson-Bell Fid.
Desert Rock
Elko/Municipal Airport
Ely/Yelland Field
Lovelock/FAA Airport
Reno/Cannon Infl Airport
Winnemucca/WSO Airport
Pendleton/Municipal Airport —
Redmond/FAA Airport
Cedar City/FAA Airport
Salt Lake City/lnt'l Airport
Spokane/I nt'l Airport
Yakima/Air Terminal
Casper/Natrona Co. Int'l Airport
Cheyenne/Municipal Airport
Lander/Hunt Field
Rock Springs/FAA Airport
Sheridan/County Airport
Fort Smith/Municipal Airport
Des Moines/lnt'l Airport
Mason City/FAA Airport
Sioux City/Municipal Airport
Waterloo/Municipal Airport
Concordia/Blosser Municipal Airport
Dodge City/Municipal Airport
Goodland/Renner Field
Russell/FAA Airport
Topeka/Municipal Airport
Wichita/Mid-Continent Airport
Duluth/lnt'l Airport
International Falls/lnt'l Airport
Minneapolis-St. Paul/1 nt'l Airport
Rochester/Municipal Airport
Kansas City/FAA Airport
Kansas City/lnt'l Airport
B-3
-------
DISTRIBUTION OF NATIONAL WEATHER SERVICE SURFACE STATIONS BY CUMATOLOGICAL REGION
Region
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
State
Missouri
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
Nebraska
North Dakota
North Dakota
North Dakota
North Dakota
South Dakota
South Dakota
South Dakota
South Dakota
Wisconsin
Wisconsin
Alabama
Alabama
Alabama
Alabama
Alabama
Arkansas
Florida
Florida
Florida
Florida
Florida
Florida
Florida
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Georgia
Louisiana
Louisiana
Louisiana
Louisiana
Mississippi
Running Count
Per Region
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
NWS Site
Springfield/Regional Airport
Grand Island/ Airport
Lincoln/Municipal Airport
Norfolk/Karl Stefan Mem. Airport
North Omaha/NWSFO Airport
North Platte/Lee Bird Fid.
Omaha/Eppley Airfield
Scottsbluff/County Airport
Bismarck/Municipal Airport
Fargo/Hector Field
Minot/FAA Airport
Williston/Sloulin Int'l Airport
Huron/Regional Airport
Pierre/FAA Airport
Rapid City/Regional Airport
Sioux Falls/Foss Field —
Eau Claire/FAA Airport
La Crosse/Municipal Airport
Birmingham/Municipal Airport
Centreville/WSMO
Huntsville/Madison County Jet
Mobile/WSO Airport
Montqomerv/WSO Airport
Little Rock/Adams Field
Apalachicola/Municipal Airport
Daytona Beach/Regional Airport
Gainesville/Municipal Airport
Jacksonville/lnt'l Airport
Orlando/I nt'l Airport
Pensacola/Regional Airport
Tallahassee/Municipal Airport
Athens/Municipal Airport
Atlanta/Atlanta-Hartsfield Int'l Airport
Augusta/Bush Field
Columbus/Metropolitan Airport
Macon/Lewis B. Wilson Airport
Savannah/Municipal Airport
Waycross/WSMO
Baton Rouge/Ryan Airport
Lake Charles/Municipal Airport
New Orleans/lnt'l Airport
Shreveport/Regional Airport
Jackson/Thompson Field
B-4
-------
DISTRIBUTION OF NATIONAL WEATHER SERVICE SURFACE STATIONS BY CUMATOLOGICAL REGION
Region
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
State
Mississippi
Mississippi
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
North Carolina
South Carolina
South Carolina
South Carolina
Tennessee
Tennessee
Tennessee
Tennessee
Tennessee
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Texas
Virginia
Illinois
Illinois
Illinois
Illinois
Illinois
Indiana
ndiana
ndiana
ndiana
Kentucky
Kentucky
Kentucky
Kentucky
Kentucky
Michigan
Michigan
Running Coun
Per Region
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
NWS Site
Meridian/Key Field
Tupelo
Ashevi He/Regional Airport
Cape Hatteras/WSO
Charlotte/Douglas Int'l Airport
Greensboro, High Point/Winsto.
Raleigh/Raleigh-Durham Airport
Wilmington/New Hanover County
Charleston/lnt'l Airport
Columbia/Metro Airport
Greer/Greenville-Spartanburg Airport
Bristol/Tri City Airport
Chattanooga/Lovell Field
Knoxville/MC Ghee Tyson Airport
Memphis/lnt'l Airport
Nashville/Metro Airport ~
Austin/Municipal Airport
Brownsville/I nt'l Airport
Corpus Christi/lnt'l Airport
Hondo/WSMO Airport
Houston/Intercontinental Airport
.ufkin/FAA Airport
Port Arthur/Jefferson County
San Antonio/WSFO
Victoria/WSO Airport
WACO/Madison-Cooper Airport
Richmond/R.E. Byrd Infl Airport
Chicago/O'Hare Int'l Airport
rioline/Quad-City Airport
Peoria/Greater Peoria Airport
Rockford/Greater Rockford Airport
Springfield/Capital Airport
Evansville/Dress Regional Airport
Fort Wayne/Baer Field
ndianapolis/lnt'l Airport
South Bend/Michiana Regional
Covington/Greater Cincinnati
Jackson/Julian Carroll Airport
Lexington/Bluegrass Field
Louisville/Standiford Field
Paducah/WSO Airport
Alpena/Phelps Collins Airport
Detroit/City Airport
B-5
-------
DISTRIBUTION OF NATIONAL WEATHER SERVICE SURFACE STATIONS BY CUMATOLOGICAL REGION
Region
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
State
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Missouri
Missouri
Missouri
New York
New York
New York
New York
New York
Ohio
Ohio
Ohio
Ohio
Ohio
Ohio
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Pennsylvania
Virginia
West Virginia
West Virginia
West Virginia
Wisconsin
Wisconsin
Wisconsin
Connecticut
Delaware
Maine
Maine
Maryland
Massachusetts
New Hampshire
New Jersey
New Jersey
New York
Running Count
Per Region
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
1
2
3
4
5
6
7
8
9
10
NWS Site
Detroit/Metropolitan Airport
Flint/Bishop Airport
Grand Rapids/Kent Co. Int'l Airport
Lansing/Capital City Airport
Muskegon/County Airport
Sault Ste Marie/NWSO
Traverse City/FAA Airport
Columbia/Regional Airport
St. Louis/Lambert Int'l Airport
St. Louis/Spirit of St. Louis
Binghamton/Edwin A. Link Field
Buffalo/Greater Buffalo Int'l Airport
Massena/FAA Airport
Rochester/Rochester-Monroe Co.
Syracuse/Hancock Infl Airport
Akron/Akron-Canton Regional ~
Cleveland/Hopkins Int'l Airport
Columbus/Port Columbus Int'l Airport
Dayton/I nt'l Airport
Toledo/Express Airport
Youngstown/Municipal Airport
Bradford/FAA Airport
Erie/I nfl Airport
Harrisburq/Capital City Airport
Dittsburgh/WSCOM 2 Airport
Williamsport-Lycom ing/County
Roanoke/Woodrum Airport
3eckley/Raleigh Co. Memorial Airport
Charleston/Kanawha Airport
Huntington/Tri-State Airport
Green Bay/Austin Straubel Field
Madison/Dane Co. Regional Airport
Milwaukee/General Mitchell Field
Hartford/Bradley Int'l Airport
Wilmington/Greater Wilmington
Bangor/FAA Airport
Portland/lnt'l Jetport
Baltimore/BLT-Washinaton Int'l Airport
Boston/Logan Int'l Airport
Concord/Municipal Airport
Atlantic Citv/Airport NAFEC
Newark/lnt'l Airport
Albany/County Airport
B-6
-------
DISTRIBUTION OF NATIONAL WEATHER SERVICE SURFACE STATIONS BY CUMATOLOQICAL REGION
Region
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
VIII
IX
IX
IX
IX
IX
IX
State
New York
New York
New York
Pennsylvania
Pennsylvania
Pennsylvania
Rhode Island
Vermont
Virginia
Virginia
Virginia
Florida
Florida
Florida
Florida
Florida
Florida
Running Count
Per Region
1 1
12
13
14
15
16
17
18
19
20
21
1
2
3
4
5
6
NWS Site
ISUP
New York/J. F. Kennedy Int'l Airport
New York/Laguardia Airport
Allentown/Bethlehem-Easton Airport
Philadelphia/lnt'l Airport
Wilkes-Barre/WB-Scranton WSO
Providence/T. F. Green State Airport
Burlington/lnt'l Airport
Norfolk/lnt'l Airport
Washington D.CVDulles Int'l Airport
Washington D.C./National Airport
Fort Myers/Page Field
Key West/lnt'l Airport
Miami/lnt'l Airport
Tampa/lnt'l Airport
Vero Beach/Municipal Airport
West Palm Beach/lnfl Airport
B-7
-------
-------
APPENDIX C
EXAMPLE PRINTOUT OF ISCST2 MODEL INPUT
OPTIONS FOR EACH AREA SHAPE
C-1
-------
-------
ISCST2 - VERSION 92021 *** *** 45m x 45m 0.5 Acres (96 subdivisions) LITTLE ROCK/LITTLE ROCK *** 4/15/93
REVISED DRAFT *** 1SH5LTL.DAT *** 16:21:38
PAGE 1
MODELING OPTIONS USED: CONC RURAL FLAT DEFAULT
*** MODEL SETUP OPTIONS SUMMARY ***
Model is Set Up for Calculation of Average Concentration Values.
Model Uses Rural Dispersion.
Model Uses Regulatory DEFAULT Options:
1. Final Plume Rise.
2. Stack-Tip Downwash.
3. Buoyancy-induced Dispersion.
4. Use Calms Processing Routine.
5. Not Use Missing Data Processing Routine.
6. Default Wind Profile Exponents.
7. Default Vertical Potential Temperature Gradients.
8. "Upper Bound" Values for Supersquat Buildings.
9. No Exponential Decay for RURAL Mode.
Model Assumes Receptors on FLAT Terrain.
Model Assumes No FLAGPOLE Receptor Heights.
Model Calculates PERIOD Averages Only.
This Run Includes: 96 Source(s); 1 Source Group(s); and 17 Receptor(s)
This Model Assumes a Pollutant Type of: ANYPOL
Model Set to Continue RUNning After the Setup Testing.
Output Options Selected:
Model Outputs Tables of PERIOD Averages by Receptor
Model Outputs Tables of Highest Short Term Values by Receptor
(RECTABLE Keyword)
Note: The Following Flags May Appear Following CONC Values:
c for Calm Hours
m for Missing Hours
b for Both Calm and Missing Hours
Misc. Inputs: Anem. Hgt. (m) = 10.00; Decay Coef. = 0.0000; Rot.
Angle =0.0, Emission Units = (GRAMS/SEC); Emission Rate Unit Factor =
0.10000E-02, Output Units = (KILOGRAMS/CUBIC-METER)
Input Runstream File: Ish51tl.dat; Output Print File: IshSltl.out
Detailed Error/Message File: ERROR.OUT
-------
ISCST2 - VERSION 92021 *** *** 25m x 75m 0.5 Acres (128 subdivisions) FRSMO/OAKLAND *** 4/15/93
REVISED DRAFT *** 1RH5FRS.DAT 1:3 RECTANGLE *** 18:17:10
PAGE 1
MODELING OPTIONS USED: CONC RURAL FLAT DEFAULT
*** MODEL SETUP OPTIONS SUMMARY ***
Model Is Set Up for Calculation of Average Concentration Values.
Model Uses Rural Dispersion.
Model Uses Regulatory DEFAULT Options:
1. Final Plume Rise.
2. Stack-tip Downwash.
3. Buoyancy-induced Dispersion.
4. Use Calms Processing Routine.
5. Not Use Missing Data Processing Routine.
6. Default Wind Profile Exponents.
7. Default Vertical Potential Temperature Gradients.
8. "Upper Bound" Values for Supersquat Buildings.
9. No Exponential Decay for RURAL Mode.
Model Assumes Receptors on FLAT Terrain.
Model Assumes No FLAGPOLE Receptor Heights.
Model Calculates PERIOD Averages Only.
This Run Includes: 128 Source(s); 1 Source Group(s); and 17
Receptor(s)
The Model Assumes a Pollutant Type of: ANYPOL
Model Set To Continue RUNning After the Setup Testing.
Output Options Selected:
Model Outputs Tables of PERIOD Averages by Receptor
Model Outputs Tables of Highest Short Term Values by Receptor
(RECTABLE Keyword)
Note: The Following Flags May Appear Following CONC Values:
c for Calm Hours
m for Missing Hours
b for Both Calm and Missing Hours
Misc. Inputs: Anem. Hgt. (m) = 10.00; Decay Coef. = 0.0000; Rot.
Angle = 140.0, Emission Units = (GRAMS/SEC); Emission Rate Unit Factor
= 0.10000E-02, Output Units = (KILOGRAMS/CUBIC-METER)
Input Runstream File: 1RH5FRS.DAT; Output Print File: 1RH5FRS.OUT
Detailed Error/Message File: ERROR.OUT
-------
ISCST2 - VERSION 92021 *** *** 20m x 100m 0.5 Acres (106 subdivisions) FRESNO/FRESNO *** 4/15/93
REVISED DRAFT *** 2RH5FRS.DAT 1:5 RECTANGLE *** 10:58:54
PAGE 1
MODELING OPTIONS USED: CONC RURAL FLAT DEFAULT
*** MODEL SETUP OPTIONS SUMMARY ***
Model Is Set Up for Calculation of Average Concentration Values.
Model Uses RURAL Dispersion.
Model Uses Regulatory Default Options:
1. Final Plume Rise.
2. Stack-tip Downwash.
3. Buoyancy-induced Dispersion.
4. Use Calms Processing Routine.
5. Not Use Missing Data Processing Routine.
6. Default Wind Profile Exponents.
7. Default Vertical Potential Temperature Gradients.
8. "Upper Bound" Values for Supersquat Buildings.
9. No Exponential Decay for RURAL Mode.
Model Assumes Receptors on FLAT Terrain.
Model Assumes No FLAGPOLE Receptor Heights.
Model Calculates PERIOD Averages Only.
This Run Includes: 106 source(s); 1 source(s); and 17 receptor(s)
This Model Assumes a Pollutant Type of: ANYPOL
Model Set To Continue RUNning After the Setup Testing.
Output Options Selected:
Model Outputs Tables of PERIOD Averages by Receptor
Model Outputs Tables of Highest Short Term Values by Receptor
(RECTABLE keyword)
Note: The Following Flags May Appear Following CONC Values:
c for Calm Hours
m for Missing Hours
b for Both Calm and Missing Hours
Misc. Inputs: Anem. Hgt.(m) = 10.00; Ddecay Coef. = 0.0000; Rot.
Angle= 140.0, Emission Units = (GRAMS/SEC); Emission Rate Unit Factor=
0.10000E-02, Output Units = (KILOGRAMS/CUBIC-METER)
Input Runstream File: 2RH5FRS.DAT; Output Print File: 2RH5FRS.OUT
Detailed Error/Message File: ERROR.OUT
-------
-------
APPENDIX D
CORRECTIONS TO RAGS - PART B
SECTIONS 3.3.1 AND 3.3.2
D-1
-------
Figure D-1 represents suggested revisions to the RAGS - Part B Section 3.3.1
"Soil-to-Air Volatilization Factor" incorporating the 95% UCL of the mean normalized
concentration as determined by regression analysis of the modeled data. This
methodology uses the most conservative results of the three source shapes modeled
(1:3 aspect ratio rectangle).
D-2
-------
Figure D-1. Soil-to-air volatilization factor (95% UCL of mean normalized concentration).
VF(m*lkg) = QIC x 'x^X * ™~* m2/cm2 (8)
(2 x DM x P. x Ktt)
where:
QIC (g/m2-s/kg/m3) = (exp [/„ + 2.925(^)1)
-i
given: Yh = Y
X, =X
Y = O.1004X - 5.3466
X = natural logarithm of the contiguous area of contamination in m2
[»•
»
and
a(cm2/s) - Dei * P>
P. * (P.)d-
Standard default parameter values that can be used to reduce Equation (8) are listed
below. The value of Q/C is the inverse of the normalized concentration from EQ 1993
and represents the emission flux (g/rrf-s) per unit concentration (kg/m3). Typical"
values for OC and p. are from EPA 1984, EPA 1988b, and EPA 1988f. Site-specific
data should be substituted for the default values listed below wherever possible.
Standard values for chemical-specific D,, H, and K^ can be obtained by calling the
Superfund Health Risk Technical Support Center.
D-3
-------
Parameter
VF
0
&
P.
T
P
H
oc
Definition, un'rts
volatilization factor
(m3/kg)
effective diffusivity (cnfs)
total soil porosity (unitless)
air-filled soil porosity
(unitless)
average soil moisture
content (cm3-water/g-soil)
soil bulk density (g/cm3)
true soil density or particle
density (g/cm3)
soil-air partition coefficient
(g-soil/cm3-air)
exposure interval (s)
diffusivity in air (cnf/s)
Henry's law constant
(atm-m3/mol)
soil-water partition
coefficient (cm3/g)
organic carbon partition
coefficient (cm3/g)
organic carbon content of
soil (fraction)
Default
P(P.333/Pt2)
1-flJ/O
site-specific
site-specific
2.65 g/cm3
(H/K,)x41, where 41 is a
units conversion factor
7.9 x 109 s
chemical-specific
chemical-specific
chemical-specific, or K,,c x
OC
chemical-specific
site-specific, or 0.02
D-4
-------
Figure D-2 represents suggested revisions to the RAGS - Part B Section 3.3.1
"Soil-to-Air Volatilization Factor" incorporating the mean of the normalized concentration
as determined by regression analysis of the modeled data. Again, this methodology
uses the most conservative results of the three source shapes (1:3 aspect ratio
rectangle).
D-5
-------
Figure D-2. Soil-to-air volatilization factor (mean of normalized concentration).
where:
QIC (glm2-s/kglm3} = [exp (0.1004X - 5.3466)1
-i
given: X = natural logarithm of the contiguous area of contamination in m2
and,
D* ^ P.
a(cm2ls)
P. + (P.)(1-P.)/K.
D-6
-------
Parameter
VF
P,
P.
0
p.
*.
T
H
K.C
OC
Definition, units
volatilization factor
(m3/kg)
effective diffusivity (cnfs)
total soil porosity (unitless)
air-filled soil porosity
(unitless)
average soil moisture
content (cm3-water/g-soil)
soil bulk density (g/cm3)
true soil density or particle
density (g/cm3)
soil-air partition coefficient
(g-soil/cm3-air)
exposure interval (s)
diffusivity in air (cnf/s)
Henry's law constant
(atm-m3/mol)
soil-water partition
coefficient (cm3/g)
organic carbon partition
coefficient (cm3/g)
organic carbon content of
soil (fraction)
Default
D,(P.333/Pt2)
site-specific
site-specific
2.65 g/cm3
x 41, where 41 is a
units conversion factor
7.9 x 108 s
chemical-specific
chemical-specific
chemical-specific, or K,,,. x
OC
chemical-specific
site-specific, or 0.02
D-7
-------
Figure D-3 represents suggested revisions to the RAGS - Part B Section 3.3.2
"Particulate Emission Factor" incorporating the 95% UCL of the mean normalized
concentration as determined by regression analysis. This methodology uses the most
conservative results of the three source shapes (1:3 aspect ratio rectangle).
D-8
-------
Figure D-3. Paniculate emission factor (95% UCL of mean normalized concentration).
PEF (m*/kg) = QIC x 360° s/h (9)
* 0.036 x (1 -G) x (UJUt)3 x F(x)
where:
QIC (g/m2-sfkg/m3) = (exp(Yh + 2.92 s(Yh)])
-1
given: Yh = Y
\ =X
Yh = 0.1004X-5.3466
X = natural logarithm of the contiguous area of contamination in m2
= 0.02685
0.25
(Xk - 11.0509)
26.3608
Parameter Definition (units) Default
0.036 respirable fraction (g/m2- 0.036 g/nf-h
h)
G fraction of vegetative 0
cover (unitless)
Um mean annual windspeed 4.5 m/s
(m/s)
14 equivalent threshold value 12.8 m/s
of windspeed at 10 m
(m/s)
F(x) function dependent on 0.0497 (determined using
Um/U, (unitless) Cowherd 1985)
D-9
-------
Figure D-4 represents suggested revisions to the RAGS - Part B Section 3.3.2
"Particulate Emission Factor" incorporating the mean of the normalized concentration as
determined by regression analysis of the modeled data. This methodology uses the
most conservative results of the three source shapes (1:3 aspect ratio rectangle).
D-10
-------
Figure D-4. Particulate emission factor (mean of normalized concentration).
PEF (m3lkg) * QIC x 360° slh (9)
0.036 x (1 -G) x (UJU,)3 x F(x)
where:
QIC (glm2-slkglm3) = [exp (0.1004X - 5.3466)]'1
given: X = natural logarithm of the contiguous area of contamination in m2
Parameter Definition (units) Default
0.036 respirable fraction (g/mf-h) 0.036 g/rrP-h
G fraction of vegetative cover (unrtless) 0
Um mean annual windspeed (m/s) 4.5 m/s
U, equivalent threshold value of 12.8 m/s
windspeed at 10 m (m/s)
F(x) function dependent on Um/U, 0.0497 (determined using
(unitless) Cowherd 1985)
D-11
-------
-------
APPENDIX E
SUMMARY STATISTICS
E-1
-------
-------
TABLE E-1. SUMMARY STATISTICS FOR SQUARE AREA SOURCE
Region
I
II
III
IV
Total number of
met stations
13
5
19
30
Number of sampled
met stations
4
2
2
6
Area, m2
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500 |
1960000
Normalized concentration, Kg/m3
per g/m2-s
Mean
0.00933
0.01270
0.01519
0.01897
0.01134
0.01543
0.01855
0.02325
0.01019
0.01395
0.01676
0.02094
0.00938
0.01270
0.01519
0.01884
Minimum
0.00836
0.01133
0.01358
0.01692
0.01054
0.01451
0.01746
0.02203
0.00889
0.01207
0.01444
0.01796
0.00743
0.01000
0.01192
0.01502
Maximum
0.01011
0.01376
0.01652
0.02086
0.01214
0.01635
0.01963
0.02446
0.01149
0.01583
0.01907
0.02391
0.01092
0.01470
0.01765
0.02189
-------
Table E-1 (continued)
Region
V
VI
VII
VIII
Total number of
met stations
35
52
49
21
Number of sampled
met stations
3
5
4
2
Area, m2
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
Normalized concentration, Kg/m3
per g/mf-s
Mean
0.00880
0.01198
0.01438
0.01803
0.00978
0.01326
0.01587
0.02005
0.00996
0.01347
0.01608
0.02001
0.00934
0.01273
0.01520
0.01892
Minimum
0.00828
0.01114
0.01330
0.01644
0.00945
0.01283
0.01536
0.01950
0.00765
0.01034
0.01233
0.01540
0.00824
0.01119
0.01335
0.01653
Maximum
0.00917
0.01262
0.01523
0.01919
0.01010
0.01371
0.01642
0.02079
0.01417
0.01922
0.02306
0.02891
0.01044
0.01426
0.01704
0.02130
-------
Table E-1 (continued)
Region
IX
Total number of
met stations
6
Number of sampled
met stations
1
Area, m2
2025
19600
202500
1960000
Normalized concentration, Kg/m3
per g/rrf-s
Mean
0.00869
0.01189
0.01433
0.01799
Minimum
0.00869
0.01189
0.01433
0.01799
Maximum
0.00869
0.01189
0.01433
0.01799
-------
-------
TABLE E-2. SUMMARY STATISTICS FOR 1:3 RECTANGLE AREA SOURCE
Region
I
II
III
IV
Total number
of met stations
13
5
19
30
Number of
sampled met
stations
4
2
2
6
Area, m2
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
Normalized concentration, Kg/m3 per
g/m2-s
Mean
0.00976
0.01362
0.01707
0.01982
0.01235
0.01718
0.02057
0.02522
0.01081
0.01513
0.01641
0.02212
0.00972
0.01357
I 0.01616
0.01968
Minimum
0.00836
0.01173
0.01401
0.01720
0.01189
0.01653
0.01982
0.02437
0.00881
0.01241
0.01478
0.01795
0.00775
0.01077
0.01277
0.01547
Maximum
0.01097
0.01521
0.02143
0.02186
0.01281
0.01782
0.02132
0.02607
0.01280
0.01785
0.01804
0.02628
0.01178
0.01632
0.01952
0.02394
-------
Table E-2 (continued)
Region
V
VI
VII
VIII
Total number
of met stations
35
52
49
21
Number of
sampled met
stations
3
5
4
2
Area, m2
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
Normalized concentration, Kg/m3 per
g/nrf-s
Mean
0.00904
0.01269
0.01518
0.01861
0.00980
0.01376
0.01640
0.01997
0.01000
0.01403
0.01668
0.02024
0.00952
0.01331
1 0.01583
0.01922
Minimum
0.00824
0.01153
0.01371
0.01666
0.00922
0.01301
0.01550
0.01882
0.00761
. 0.01065
0.01263
0.01531
0.00802
0.01122
0.01333
0.01615
Maximum
0.00993
0.01391
0.01673
0.02069
0.01035
0.01452
0.01734
0.02113
0.01413
0.01986
0.02371
0.02889
0.01102
0.01539
0.01832
0.02228
-------
Table E-2 (continued)
Region
IX
Total number
of met stations
6
Number of
sampled met
stations
1
Area, m2
2025
19600
202500
1960000
Normalized concentration, Kg/m3 per
g/irf-s
Mean
0.00852
0.01207
0.01448
0.01775
Minimum
0.00852
0.01207
0.01448
0.01775
Maximum
0.00852
0.01207
0.01448
0.01775
-------
-------
TABLE E-3. SUMMARY STATISTICS FOR 1:5 RECTANGLE AREA SOURCE
Region
I
II
III
IV
Total number
of met stations
13
5
19
30
Number of
sampled met
stations
4
2
2
6
Area, m2
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
Normalized concentration, Kg/m3 per
g/nf-s
Mean
0.00995
0.01184
0.01412
0.01735
0.01270
0.01519
0.01665
0.02247
0.01105
0.01323
0.01586
0.01948
0.00991
0.01180
I 0.01408
0.01723
Minimum
0.00850
0.01013
0.01213
0.01504
0.01218
0.01459
0.01579
0.02165
0.00883
0.01053
0.01257
0.01532
0.00775
0.00916
0.01083
0.01307
Maximum
0.01134
0.01345
0.01598
0.01946
0.01321
0.01579
0.01751
0.02328
0.01326
0.01592
0.01915
0.02363
0.01238
0.01481
0.01776
0.02196
-------
Table E-3 (continued)
Region
V
VI
VII
VIII
Total number
of met stations
35
52
49
21
Number of
sampled met
stations
3
5
4
2
Area, m2
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
2025
19600
202500
1960000
Normalized concentration, Kg/m3 per
g/rrf-s
Mean
0.00920
0.01100
0.01318
0.01622
0.00986
0.01176
0.01404
0.01716
0.01005
0.01219
0.01423
0.01732
0.00959
0.01442
1 0.01359
0.01656
Minimum
0.00831
0.00988
0.01177
0.01436
0.00921
0.01098
0.01310
0.01594
0.00762
0.00904
0.01073
0.01306
0.00785
0.00933
0.01109
0.01345
Maximum
0.01030
0.01238
0.01491
0.01854
0.01041
0.01244
0.01487
0.01819
0.01420
0.01696
0.02027
0.02479
0.01133
0.01350
0.01609
0.01967
-------
Table E-3 (continued)
Region
IX
Total number
of met stations
6
Number of
sampled met
stations
1
Area, m2
2025
19600
202500
1960000
Normalized concentration, Kg/m3 per
g/m2-s
Mean
0.00849
0.01018
0.01222
0.01503
Minimum
0.00849
0.01018
0.01222
0.01503
Maximum
0.00849
0.01018
0.01222
0.01503
-------
-------
APPENDIX F
EXAMPLE OF STATISTICAL CALCULATIONS
F-1
-------
Estimation of Mean Normalized Concentration
The linear regression equation for estimating the mean In-normalized
concentration for a square area source at a given size area is presented below:
Y = -5.3880 + 0.1005 ln(X) (1)
where
Y is the estimated mean In-normalized concentration; and
X is the size of the area in m2.
To estimate the mean normalized concentration for an area of 2025 m2, Equation
(1) is first used to calculate the mean normalized concentration on the In-scale as
follows:
Y = -5.3880 + 0.1005 ln(2025)
= -5.3880 + 0.1005 x 7.6133
= -5.3880 + 0.7651
= -4.6229
To estimate the mean normalized concentration on the original scale (i.e., kg/m3
per g/rrf-s), the exponential of Y is calculated as follows:
4.6229
exp(-4.6229), or equivalently, e"
where e = 2.7183.
Therefore, the estimated mean normalized concentration for a square area
source of 2025 m2 would be:
exp(-4.6229) = 0.0098 kg/m3 per g/m2-s
F-2
-------
Estimation of the 95% UCL for the Mean Normalized Concentration
To estimate the upper 95% confidence limit (UCL) for the mean normalized
concentration for a square area source of X=)^, m2, the following equation is used:
95% UCL = exp[Yh + 2.92 s(Vh)] kg/m3 per g/m2-s (2)
where
Yh = -5.3880 + 0.1005 Info) (3)
= the standard error of Yh = 0.0213 [1/4 + (Info) -.11.0509)2/26.3608I4)
To estimate the 95% UCL for the mean normalized concentration for a square
area source of \= 2025 m2, equation (3) is first used to estimate the mean normalized
concentration on the log scale as follows:
Yh = -5.3880 + 0.1005 In (2025)
= -5.3880 + 0.1005 x 7.6133
= -5.3880 + 0.7651
= -4.6229
Equation (4) is then used to calculate the standard error of Yh as follows:
5(9; ) = 0.0213 [1/4 + (Info) - 11.0509)2/26.3608]
= 0.0213 [0.25 -i- (ln(2025) - 11.0509)2)/26.3608]
= 0.0213 [0.25 + (7.6133 - 11.0509)2/26.3608]
= 0.0213 [0.25 + 11.8171/26.3608]
= 0.0213 [0.25 + 0.4483]
= 0.0213 0.6983
= 0.0149
Finally, Equation (2) is used to calculate the 95% UCL as follows:
95% UCL = expFYh + 2.92 s(Yh)] kg/m3 per g/m2-s
F-3
-------
exp[-4.6229 + 2.92 x 0.0149]
exp[-4.6229 + 0.0434]
exp[-4.5795]
0.0103 kg/m3 per g/nf-s
F-4
-------
APPENDIX D
COLLECTED Koc VALUES
-------
-------
Chemical Koc Source
Chloroform
Average
Median
Geom. Mean
Ylethylene chloride
Average
Median
Geom. Mean
Carbon tetrachloride
Average
Median
Geom. Mean
1,1-Dichloroethane
1,2-Dichloroethane
1 ,2-Dichloroethane
1 ,2-Dichloroethane
1 ,2-Dichloroethane
Average
Median
Geom. Mean
1,1,1-Trichloroethane
Average
Median
Geom. Mean
45
42
78
28
40
58
49
44
46
9
28
48
28
28
23
123
224
439
262
224
230
32
36
42
30
32
34
32
34
129
198
152
179
129
100
77
107
134
129
128
Gerstl (1990) (secondary)
Wilson etal. (1981)
Koch (1983)
Grathwohl (1990)
Hutzler etal. (1983)
Loch et al (1986) (column studies - average of 3 values)
Schwille(1988) (secondary)
Gerstl (1990) (secondary)
Koch (1983)
Koch (1983)
Abdul et al. (1987) (secondary - MS thesis)
Schwille (1988) (secondary)
Chiou etal. (1979)
Hassettetal. (1980)
Wilson etal. (1981)
Schwille (1988) (secondary)
Karickhoff(1981)
Gerstl (1990) from Sabljic (1987)
Hassettetal. (1980)
Schwille (1988) (secondary)
Chiou etal. (1979)
Hodson& Williams (1988)
Loch et al. (1986) (column studies - average of 2 values)
Seip et al. (1986) (average of 3 values)
Friesel et al. (1984) (average of 17 values)
D-l
-------
Chemical fZy-" "" • : KOC Source . --;•:.-. ^-. ^-^y-.:
Vinyl chloride
1,1-Dichloroethylene
Trichloroethylene
Average
Median
Geom. Mean
65
92
117
135
87
148
174
65
126
52
92
84
46
89
109
41
107
62
110
103
69
100
123
29
94
92
87
Schwille(1988) (secondary)
Gerstl (1990) from Webster et al. (1985) (avg. of 5 values)
Gerstl (1990) from Hutzler et al. (1986) (avg. of 3 values)
Gerstl (1990) from Crittenden et al. (1986)
Gerstl (1990) (secondary)
Gerstl (1990) (secondary)
Gerstl (1990) from Sabljic (1987)
Abdul etal. (1987)
Schwille (1988) (secondary)
Stauffer & Maclntyre (1986)
Wilson etal. (1981)
Piwoni & Banerjee (1989)
Paviostathis & Mathavan (1992) (desorption)
Pignatello(1990a)(24h)
Pignatello (1990a) (72h)
Pignatello (1990b) (7 day)
Gabarini and Lion (1986)
Reinhard etal. (1991)
Loch et al. (1986) (column studies - average of 2 values)
Seip et al. (1986) (average of 3 values)
Brusseau et al. (1991)
Friesel et al. (1984) (average of 18 values)
Grathwohl (1990)
Lee etal. (1989)
2/18/94
D-2
-------
|CHemical Koc Source
Tetrachloroethylene
Average
Median
Geom. Mean
Chlorobenzene
Average
Median
Geom. Mean
1,4-Dichlorobenzene (para-)
Average
Median
Geom. Mean
363
263
225
237
269
238
325
204
373
346
428
269
279
79
191
248
437
281
269
264
83
330
117
255
126
210
217
191
210
173
437
643
273
900
398
598
700
300
486
832
528
429
544
507
511
Chiouetal. (1979)
Abdul etal. (1987)
Wilson etal. (1981)
Piwoni & Banerjee (1989) (avg. of 8 values)
Pignatello (1990a) (average of 2 values)
Friesel et al. (1984) (average of 18 values)
Burris etal. (1991)
Schwarzenbach & Westall (1981)
Schwarzenbach & Westall (1981)
Gerstl (1990) from Schwarzenbach & Giger (1982)
Roy etal. (1987)
(2)
Brusseau et al. (1991)
Loch et al. (1986) (column studies - avg. of 3 values)
Lee et al. (1989) (average of 2 values)
Seip et al. (1986) (average of 2 values)
Bedient etal. (1983)
Grathwohl (1990)
Chiouetal. (1983)
Schwille(1988) (secondary)
Wilson etal. (1981)
Schwarzenbach & Westall (1981)
Gerstl (1990) (secondary)
Gerstl (1990) from Schwarzenbach & Giger (1982)
Koch (1983)
(2)
Koch (1983)
Gerstl (1990) from Schwarzenbach & Giger (1982)
Chiouetal. (1983)
(2)
Schwarzenbach & Westall (1981)
Wilson etal. (1981)
Southworth & Keller (1986) (average of 3 values)
Hutzler et al. (1983) (batch)
Hutzler et al. (1983) (column)
Reinhard etal. (1991)
Chin and Weber (1989)
Loch et al. (1986) (column studies)
Friesel etal. (1984)
D-3
-------
B®!*ili^ii^ffiH- •••••• : '-•-'••' :
1,2,4-Trichlorobenzene
Average
Median
Geom. Mean
Benzene
Average
Median
Geom. Mean
Toluene
Average
Median
Geom. Mean
/'-"'••••Koc-'"^^
2,283
864
1,428
2,800
1,389
1,416
1,300
1,441
1,288
2,300
1,651
1,422
1,561
31
26
83
79
96
49
60
63
45
19
65
56
60
50
115
99
150
300
98
151
82
38
95
304
247
153
115
129
Source '^V-'V-'V':':7':*
Gerstl (1990) from Schwarzenbach & Giger (1982) (2)
Chiouetal. (1983)
South worth & Keller (1986) (average of 3 values)
Schwarzenbach & Westall (1981)
Wilson et al. (1981) (average of 2 values)
Lee et al. (1989) (average of 2 values)
Banerjee et al. (1985) (average of 21 values)
Frieseletal. (1984)
Chin and Weber (1989)
Reinhard et al. (1991)
Chiouetal. (1983)
Vowles and Mantoura (1987)
Karickhoffetal. (1979)
Koch (1983)
Rogers etal. (1980)
Abdul etal. (1987)
Karickhoff(1981)
Piwoni & Banerjee (1989) (0.19% OC)
Seip et al. (1986) (average of 3 values)
Lee etal. (1989)
Mabey etal. (1982)
Abdul etal. (1987)
Vowles and Mantoura (1987)
Wilson etal. (1981)
Paviostathis & Mathavan (1992) from US EPA (1986)
Paviostathis & Mathavan (1992) (desorption)
Gabarini and Lion (1986)
Gerstl (1990) from Nathwani & Phillips (1977) (3 values)
Lee et al. (1989) (average of 2 values)
Seip et al. (1986) (average of 3 values)
Bedient etal. (1983)
Schwarzenbach & Westall (1981)
D-4
-------
Chemical •>;• ,';""•' •
Xylenes
o
m
P
0
P
o
o
Average
Median
Geom. Mean
Ethylbenzene
Average
Median
Geom. Mean
Naphthalene
Average
Median
Geom. Mean
Koc
129
204
166
222
260
276
240
258
178
316
413
192
333
245
240
234
165
170
240
255
132
192
170
187
584
1,300
550
897
870
813
787
414
368
515
1,122
843
912
594
939
1,096
1,300
1,333
846
857
792
Source . ;--..';... v •:..'>. :ilW?8K
Abdul etal. (1987)
Abdul etal. (1987)
Abdul etal. (1987)
Vowles and Mantoura (1987)
Vowles and Mantoura (1987)
Briggs(1981)
Paviostathis & Mathavan (1982) from US EPA (1986)
Reinhard etal. (1991)
Meylan et al. (1992) (compilation avg.)
Gerstl (1990) from Reinert & Rogers (1987)
Gerstl (1990) from Schwarzenbach & Giger (1982) (2)
Seip etal. (1986)
Schwarzenbach & Westall (1981)
Chiou et al. (1983)
Meylan et al. (1992) (compilation average)
Hodson& Williams (1988)
Vowles and Mantoura (1987)
Lee et al. (1989) (average of 4 values)
Stauffer & Maclntyre (1986)
Karickhoff(1979)
Abdul (1987)
Kan & Tomson (1990) (average of 2 values)
Karickhoff(1981)
Schwarzenbach (1981)
Southworth & Keller (1986) (average of 3 values)
Briggs (1981)
Reinhard etal. (1991)
Gerstl (1990) from Wauchope et al. (1983) (avg. of 2)
Meylan et al. (1992) (compilation average)
Vowles and Mantoura (1987)
Hodson & Williams (1988)
Kishi et al. (1990) (average of 5 values)
Lokke (1984) (average of 8 values)
McCarthy & Jiminez (1985)
Bedient etal. (1983)
Karickhoff(1982)
D-5
-------
^heraicar;s|?SH':s;^;:;:K-':r,v • ' - Koc : Source •.; - ••:':^-:<>;.--;:;T=X \:i :S:; !?:*y=a:s |
Benz-a-pyrene
Average
Median
Geom. Mean
Chrysene
PCBs
Arochlor 1016
Average
Median
Geom. Mean
Arochlor 1242
Average
Median
Geom. Mean
Arochlor 1254
Average
Median
Geom. Mean
5,069,757
4,570,882
891,251
478,947
2,752,709
2,731,067
1,773,449
NA
171,250
92,143
54,167
155,000
118,140
123,572
107,285
1,681,808
135,000
151,250
77,857
52,083
419,600
135,000
169,340
585,000
1,262,779
108,667
652,149
585,000
431,380
Smith et al. (1977) (average of 3
values)
Eadieetal. (1990)
Landrumetal. (1984)
DiToro (1985) from Smith et al.
(1978)
Verschueren (1983)
Verschueren (1983)
Verschueren (1983)
Verschueren (1983)
Witkowski et al. (1988)
Verschueren (1983)
Verschueren (1983)
Verschueren (1983)
Verschueren (1983)
Voice & Weber (1985) (average
of 2 values)
Weber et al. (1983) (average of 8 values)
Wildish et al. (1980)
2/18/94
D-6
-------
Chemical •:>:;\: ':'•• ::..;
2,2',4,4',5,5'-PCB
2,2',4,4',5,5'-PCB
2,2',4,4',5,5'-PCB
2,2',4,4',5,5'-PCB
2,2',4,4',5,5'-PCB
2,2',4,41,5,5'-PCB
2,2',4,41,5,5'-PCB
2,21,4,4',5,5'-PCB
2,2',4,4',6,6'-PCB
2,2',4,4',6,61-PCB
2,2',4,4',6,6'-PCB
2,2',5,5'-PCB
2,2',5-PCB
2,2',5-PCB
2,2',5-PCB
2,2',5-PCB
2,2',4-PCB
2,2'-PCB
2,3,6,2',3',6'-PCB
2,4'-PCB
2,4'-PCB
2,4,4-PCB
2,4,5,2',4',5'-PCB
2,4,5,2',4',5'-PCB
2,5,2',5'-PCB
2,6,2',6'-PCB
2-PCB
4,4-PCB
Pentachlorophenol
Average
Median
Geom. Mean
Chlordane
Average
Median
Geom. Mean
Koc
1,200,000
420,000
123,000
5,623,413
319,351
466,145
1,424,130
53,390,000
1,200,000
48,180,000
36,030,000
562,341
25,361
29,411
104,705
37,153
69,183
8,250
480,000
13,383
13,792
41,356
379,280
213,796
81,028
430,000
2,930
450,000
23,988
35,481
38,905
900
24,819
29,735
13,139
58,884
41,686
50,285
50,285
49,544
Source '": '"• : 7 x."' r: '. '•"".. ::VI:'';
Kenaga & Goring (1980)
Karickhoff(1981)
Dunnivant et al. (1992)
DiToroetal. (1991)
Voice & Weber (1985)
Voice & Weber (1985)
Voice & Weber (1985)
Coates & Elzerman (1986)
Karickhoff(1981)
Coates & Elzerman (1986)
Coates & Elzerman (1986)
DiToroetal. (1991)
Voice & Weber (1985)
Voice & Weber (1985)
Voice & Weber (1985)
Meylan et al. (1992) (compilation avg.)
Meylan et al. (1992) (compilation avg.)
Chiou etal. (1983)
Voice & Weber (1985)
Chiou etal. (1983)
Chiou et al. (1979) from Haque & Schnedding (1976)
Chiou etal. (1983)
Chiou et al. (1979) from Haque & Schnedding (1976)
Schwarzenbach & Westall (1981)
Chiou et al. (1979) from Haque & Schnedding (1976)
Coates & Elzerman (1986)
Chiou etal. (1983)
Coates & Elzerman (1986)
Schellenberg et al. (1984)
Schellenberg etal. (1984)
Schellenberg etal. (1984)
Dragun (1988) from Kenaga & Goring (1980)
Chin and Weber (1989)
Johnson-Logan et al. (1992) (average of 14 values)
2/18/94
D-7
-------
Chemical
DDT
Average
Median
Geom. Mean
Dieldrin
Average
Median
Geom. Mean
alpha-BHC
Average
Median
Geom. Mean
beta-BHC
Average
Median
Geom. Mean
Koc
150,000
443,137
238,000
243,000
243,000
229,000
2,290,868
275,422
343,983
377,170
1,300,947
407,380
426,580
426,580
528,219
360,577
386,977
8,366
12,780
12,000
12,589
14,000
25,604
14,223
12,685
13,400
2,050
1,615
1,833
1,833
1,820
1,995
3,914
3,473
2,274
1,313
2,171
2,581
3,162
2,610
2,427
2,481
Source |
Dragun (1988) from McCall et al. (1980)
Mingelgrin & Gerstl (1983) from Shin et al. (1970)
Kenaga& Goring (1980)
Reinbold et al. (1979) from Hamaker (1974)
Hamaker & Thompson (1972) from Shin et al. (1970) (2)
Hamaker & Thompson (1972) from Goring (1962)
Eadieetal. (1990)
Landrumetal. (1984)
Koch (1983)
Chiou(1989)
DiToro (1985) from Pierce et al. (1977) (avg. of 7 values)
Chin and Weber (1989)
Hodson & Williams (1988)
Gerstl (1990) (secondary - avg)
Felsot & Wilson (1980) (average of 7 values)
Briggs(1981)
Jury et al. (1989) from Jury et al. (1987)
Meylan et al. (1992) (compilation avg.)
Kishietal. (1990)
Sharom et al. (1980) (average of 3 values)
DiToro (1985) from Wahid & Sethunathan (1979)
DiToro (1985) from Wahid & Sethunathan (1979)
Karickhoff (1981) from Dragun (1988)
Reinbold et al. (1979) (average of 2 values)
Hamaker & Thompson (1972) from Mills & Biggar (1968)
Hamaker & Thompson (1972) from Lotse et al. (1968)
Hamaker & Thompson (1972) from Spencer & Cliath (1970)
Gerstl (1990) from Wahid & Sethunathan (1979)
DiToro (1985) from Wahid & Sethunathan (1979) (2)
Meylan et al. (1992) (compilation avg.)
3.39457943
D-8
-------
| Chemical
gamma-BHC (lindane)
(6)
(2)
Average
Median
Geom. Mean
BHC (all isomers)
Average
Median
Geom. Mean
Koc
\
1,995
1,080
735
242
2,274
1,946
1,021
2,983
911
1,081
965
1,342
1,446
1,300
1,453
1,517
1,000
1,701
982
955
1,346
1,191
1,208
1,716
1,485
1,504
Source
Karickhoff(1981)
Lyman "Adsorbtion Coefficient for Soils..."
McCalletal. (1980)
Kanazawa(1989)
Hamaker & Thompson (1972) from Spencer & Cliath (1970)
Hamaker & Thompson (1972) from Mills & Biggar (1969)
Hamaker & Thompson (1972) from Kay & Elrick (1967)
Chiou et al. (1979)
Kenaga & Goring (1980)
Rao & Davidson (1979)
Mingelgrin & Gerstl (1983) (average of 2 values)
Reinbold et al. (1979) (average of 2 values)
Miller & Weber (1986) (average of 4 values)
Jury & Ghodrati (1979) from Jury et al. (1987)
Reinhard et al. (1991)
Sharom et al. (1980) (average of 3 values)
Hodson & Williams (1988) from Rippen et al. (1982)
DiToro (1985) from Wahid & Sethunathan (1979) (4)
Kishi et al. (1990) (average of 5 values)
Gerstl (1990) (secondary average)
D-9
-------
K^Bibliography
Abdul, A.S., T.L. Gibson, and D.N. Rai. 1987. Statistical correlations for predicting the partition
coefficient for nonpolar organic contaminants between aquifer organic carbon and water.
Hazardous Waste & Hazardous Materials. 4(3):211-222.
Banerjee, P., M.D. Piwoni, and K. Ebeid. 1985. Sorption of organic contaminants to a low
carbon subsurface core. Chemosphere. 14(8): 1057-1067.
Bedient, P.B. et al. 1982. In Modeling the Fate of Chemicals in the Aquatic Environment. Ann
Arbor Science, Ann Arbor, MI: 215-246.
Bedient, P.B., A.M. ASCE, N.K. Springer, E. Baca, T.C. Bouvette, S.R. Hutchins and M.B.
Thompson. 1983. Ground-water transport from wastewater infiltration. J. Environ. Eng. New
York. 109:485:499.
Briggs, G.G. 1981. Adsorption of pesticides by some Australian soils. Aust. J. Soil Res.
19:61-68.
Brusseau, M.L. and P.S.C. Rao. 1991. Influence of sorbate structure on nonequilibrium sorption
of organic compounds. Environ. Sci. Technol. 25(8): 1501-1506.
Brusseau, M.L., A.L. Wood, and P.S.C. Rao. 1991. Influence of Organic Cosolvents on the
Sorption Kinetics of Hydrophobic Organic Chemicals. Environ. Sci. Technol. 25(5):903-910.
Bums, D.R. 1991. Flumic acid modified silica as a model aquifer material. Environ. Toxicol.
Chem. 10:433-440.
Chin, Y.P and W.J. Weber, Jr. 1989. Estimating the Effects of Dispersed Organic Polymers on
the Sorption of Contaminants by Natural Solids. 1. A Predictive Thermodynamic Humic
Substance-Organic Solute Interaction Model. Environ. Sci. Technol. 23(8):978-984.
Chiou, C.T., P.E. Porter, and D.W. Schmedding. 1983. Partition equilibria of nonionic organic
compounds between soil organic matter and water. Environ. Sci. Technol. 17(4):227-231.
Chiou, C.T. 1989. Theoretical considerations of the partition uptake of nonionic organic
compounds by soil organic matter. Reactions and Movement of Organic Chemicals in Soils,
SSSA Special Publication no. 22:1-29.
Chiou, C.T., L.J. Peters, and V.H. Freed. 1979. A physical concept of soil-water equilibria for
nonionic organic compounds. Sci. 206:831-832.
D-10
-------
Coates, J.T., and A.W. Elzerman. 1986. Desorption kinetics for selected PCB congeners from
river sediments. J. Contam. Hydrol. 1:191-210.
Crittenden, J.C., NJ. Hutzler, and D.G. Geyer. 1986. Water Res. 22:271-284.
Di Toro, D.M. 1985. A particle interaction model of reversible organic chemical sorption.
Chemosphere. 14(10):1503-1538.
Di Toro, D.M. et al. 1991. Technical basis for establishing sediment quality criteria for nonionic
organic chemicals using equilibrium partitioning. Environ. Toxicol. Chem. 10:1541-1585.
Dragun, J. 1988. The Soil Chemistry of Hazardous Material. HMCRI. Silver Spring, MD.
Dunnivant, P.M., P.M. Jardine, D.L. Taylor, and J.F. McCarthy. 1992. Cotransport of cadmium
and hexachlorobiphenyl by dissolved organic carbon through columns containing aquifer material.
Environ. Sci. Technol. 26(2):360-368.
Eadie, B.J., N.R. Morehead, and P.P. Landrum. 1990. Three-phase partitioning of hydrophobic
organic compounds in Great Lakes waters. Chemosphere. 20(1-2):161-178.
Felsot, A. and J. Wilson. 1980. Adsorption of carbofuran and movement on soil thin layers.
Bull. Environ. Contam. Toxicol. 24:778-782.
Friesel, P., G. Milde, and B. Steiner. 1984. Interactions of halogenated hydrocarbons with soils.
Fresenius Z. Anal. Chem. 319:160-164.
Garbarini, D.R. and L.W. Lion. 1986. Influence of the nature of soil organics on the sorption
of toluene and trichloroethylene. Environ. Sci. Technol. 20(12): 1263-1269.
Gerstl, Z. 1990. Final Report: Evaluating the Groundwater Pollution Hazard of Toxic Organics
by Molecular Connectivity. Submitted to the Israel National Council for Research and
Development and the Commission of the European Community, General Directorate of Science,
Research and Development.
Goring, C.A.I. 1962. Theory and principles of soil fumigation. Advances in Pest Control
Research. 5:47-84.
Grathwohl, P. 1990. Influence of Organic Matter from Soils and Sediments from Various Origins
on the Sorption of Some Chlorinated Aliphatic Hydrocarbons: Implications on Koc Correlations.
Environ. Sci. Technol. 24(11):1687-1693.
Hamaker, J.W. 1974. The interpretation of soil leaching experiments in Environmental
Dynamics of Pesticides. R. Haque and V.H. Freed, Eds., Plenum Press, New York, NY.
D-ll
-------
Hamaker, J.M., and Thompson, J.M. 1972. Adsorption. Organic Chemicals in the Soil
Environment. 1:49-143.
Haque, R. and D.W. Schmedding. 1976. J. Environ. Sci. Health. Bll(2):129.
Haque, R., D.W. Schmedding, and V.H. Freed. 1974. Environ. Sci. Technol. 8:139-142.
Hassett, J.J., J.C. Means, W.L. Banwart, and S.G. Wood. 1980. Sorption Properties of
Sediments and Energy-Related Pollutants. EPA 600/3-80-041. U.S. EPA. Athens, GA.
Hodson, J. and N.A. Williams. 1988. The estimation of the adsorption coefficient (Koc) for
soils by high performance liquid chromatography. Chemosphere. 17(l):67-77.
Hutzler, N.J., J.C. Crittenden, and J.S. Gierke. 1986. Water Res. 22:285-295.
Hutzler, N.J., J.C. Crittenden, J.L. Oravitz and P.A. Schaepe. 1983. Groundwater transport of
chlorinated organic compounds. American Chemical Society, Div. Environ. Chem., 186th
National Meeting. 23:499-502.
Johnson-Logan, L.R., R.E. Broshears, and S.J. Klaine. 1992. Partitioning Behavior and the
Mobility of Chlordane in Groundwater. Environ. Sci. Technol. 26(ll):2234-2239.
Jury, W.A. and M. Ghodrati. 1989. Overview of organic chemical environmental fate and
transport modeling approaches. Reactions and Movement of Organic Chemicals in Soils, SSSA
Special Publication no. 22:271-304.
Jury, W.A., A.M. Winer, W.F. Spencer, and D.D. Focht. 1987b. Transport and transformations
of organic chemicals in the soil-air-water ecosystem. Rev. Environ. Contam. Toxicol. 99:120-
164.
Kan, A.T., and M.B. Tomson. 1990. Ground water transport of hydrophobic organic compounds
in the presence of dissolved organic matter. Environ. Toxicol. Chem. 9:253-263.
Kanazawa, J. 1989. Relationship between the soil sorption constants for pesticides and their
physiochemical properties. Environ. Toxicol. Chem. 8:477-484.
Karickhoff, S.W. 1981. Semi-empirical estimation of sorption of hydrophobic pollutants on
natural sediments and soils. Chemosphere. 10(8):833-846.
Karickhoff, S.W. 1982. Sorption kinetics of hydrophobic pollutants in natural sediments.
Contaminants and Sediments. 2:193-205.
Karickhoff, S.W., D.S. Brown, and T.A. Scott. 1979. Sorption of hydrophobic pollutants on
natural sediments. Water Res. 13:241-248.
D-12
-------
Kay, B.D. and D.E. Elrick. 1967. Soil Sci. 104:314.
Kenaga, E.E., and C.A.I. Goring. 1980. Relationship between water solubility, soil sorption,
octanol-water partitioning, and concentration of chemicals in biota. Aquatic Toxic, pp. 78-115.
Kishi, H., N. Kogure, and Y. Hashimoto. 1990. Contribution of soil constituents in adsorption
coefficient of aromatic compounds, halogenated alicyclic and aromatic compounds to soil.
Chemosphere. 21(7):867-876.
Koch, R. 1983. Molecular connectivity index for assessing ecotoxicological behaviour of
organic compounds. Toxicol. Environm. Chem. 6:87-96.
Landrum, P.P., S.R. Nihart, BJ. Eadie, and W.S. Gardner. 1984. Reverse-Phase Separation
Method for Determining Pollutant Binding to Aldrich Humic Acid and Dissolved Organic Carbon
of Natural Waters. Environ. Sci. Technol. 18(3): 187-192.
Lee, J.F., J.R. Crum, and S.A. Boyd. 1989. Enhances Retention of Organic Contaminants by Soils
Exchanged with Organic Cations. Environ. Sci. Technol. 23(11):1365-1372.
Loch, J.P.G., H.J. Kool, P. Lagas, and J.H.A.M. Verheul. 1986. Removal and retention of
volatile chlorinated hydrocarbons in the soils' unsaturated zone. Contaminated Soil. 63-77.
Lokke, H. 1984. Sorption of selected organic pollutants in Danish soils. Ecotoxicol. Environ.
Safety. 8:395-409.
Lotse, E.J., D.A. Graetz, G. Chesters, G.B. Lee, and L.W. Newland. 1968. Environ. Sci.
Technol. 2:353.
Mabey, W.R., J.R. Smith, R.T. Podoll, H.J. Johnson, T. Mill, T.W. Chiou, J. Gates, I. Waight
Partridge, H. Jaber, and D. Vanderberg, "Aquatic Fate Process Data for Organic Priority
Pollutants," EPA Report No. 440/4-81-014, 1982.
McCall, P.J., R.L. Swann, D.A. Laskowski, S.M. Unger, S.A. Vrona, and H.J. Dishburger. 1980.
Estimation of chemical mobility in soil from liquid chromatographic retention times. Bull.
Environ. Contam. Toxicol. 24:190-195.
McCarthy, J.F. and B.D. Jimenez. 1985. Interactions between polycyclic aromatic hydrocarbons
and dissolved humic material: binding and dissociation. Environ. Sci. Technol. 19(11): 1072-
1076.
Meylan, W. and P.H. Howard. 1992. Molecular Topology/Fragment Contribution Method for
Predicting Soil Sorption Coefficients. Environ. Sci. Technol. 26(8): 1580-1587.
D-13
-------
Miller, C.T. and W.J. Weber, Jr. 1986. Sorption of hydrophobic organic pollutants in saturated
soil systems. J. Cont. Hvdrol. l(l/2):243-261.
Mills, A.C. and J.W. Biggar. 1969. J. Aer. Food Chem. 17:604.
Mills, A.C. and J.W. Biggar. 1969. Soil Sci. Soc. Am. Proc. 33:210.
Mingelgrin, U. and Z. Gerstl. 1983. Reevaluation of partitioning as a mechanism of nonionic
chemicals adsorption in soils. J. Environ. Qual. 12(1):1-11.
Nathwani, J.S. and C.R. Phillips. 1977. Chemosphere. 4:157-162.
Paviostathis, S.G. and G.N. Mathavan. 1992. Desorption kinetics of selected volatile organic
compounds from field contaminated soils. Environ. Sci. Technol. 26(3):532-538.
Pierce, R.H., Jr., C.E. Olney, and G.T. Felbeck, Jr. 1974. pp'-DDT adsorption to suspended
paniculate matter in sea water. Geochimica et Cosmochimica Acta. 38:1061-1073.
Pignatello, J.J. 1990a. Slowly reversible sorption of aliphatic halocarbons in soils. I. Formation
of residual fractions. Environ. Toxicol. Chem. 9:1107-1115.
Pignatello, J.J. 1990b. Slowly reversible sorption of aliphatic halocarbons in soils. II.
Mechanic aspects. Environ. Toxicol. Chem. 9:1117-1126.
Piwoni, M.D. and P. Banerjee. 1989. Sorption of volatile organic solvents from aqueous
solution onto subsurface solids. J. Contam. Hvdrol. 4(2):163-179.
Rao, P.S.C. and J.M. Davidson. 1979. Water Res. 13:375.
Reinbold, K.A., J.J. Hassett, J.C. Means, and W.L. Banwart. 1979. Adsorption of Energy-
Related Organic Pollutants: A Literature Review. EPA 600/3-79-086. U.S. EPA. Washington,
DC.
Reinert, K.H. and J.H. Rogers. 1987. In: Rev. Environ. Contam. and Toxicol. 98:1-41.
Rippen, G., M. Ilgenstein, W. Klopffer, and H.J. Poremski. 1982. Ecotoxicol. Environ. Safety.
6:236-245.
Rogers, R.D., J.C. McFarlane, and A.J. Cross. 1980. Adsorption and desorption of benzene in
two soils and montmorillonite clay. Am. Chem. Soc. 14(4):457-460.
Roy, W.R., et al. 1987. Batch-type adsorption procedures for estimating soil alternation of
chemicals. Technical Recourse Document. Illinois State Geol. Library.
D-14
-------
Sabljic, A. 1987. Environ. Sci. Technol. 21:358-366.
Schellenberg, K., C. Leuenberger, and R.P. Schwarzenbach. 1984. Sorption of chlorinated
phenols by natural sediments and aquifer materials. Environ. Sci. Technol. 18(9):652-657.
Schwarzenbach, R.P. and W. Giger. 1982. In: 446-471.
Schwarzenbach, R.P., and J. Westall. 1981. Transport of nonpolar organic compounds from
surface water to groundwater. Laboratory sorption studies. Environ. Sci. Technol. 15(11): 1360-
1367.
Schwille, F. 1988. Dense Chlorinated Solvents in Porous and Fractured Media. Lewis
Publishers, Inc. Chelsea, MI.
Seip, H.M., J. Alstad, G.E. Carlberg, K. Martinsen, and R. Skanne. 1986. Measurement of
mobility of organic compounds in soils. The Science of the Total Environment. 50:87-101.
Sharom, M.S., J.R.W. Miles, C.R. Harris, and F.L. McEwen. 1980. Behaviour of 12 insecticides
in soil and aqueous suspensions of soil and sediment. Water Res. 14:1095-1100.
Shin, Y., J.J. Chodan, and A.R. Walcott. 1970. Adsorption of DDT by soils, soil fractions and
biological materials. J. Agr. Food Chem. 18:1129-1133.
Smith, J.H., W.R. Mabey, N. Bohonos, B.R. Holt, S.S. Lee, T.-W. Chou, D.C. Venberger, and
T. Mill. 1977. Environmental Pathways of Selected Chemicals in Freshwater Systems. Part 1:
Background and Experimental Procedures. U.S. EPA: Athens, GA. EPA-600/7-77-113.
Smith, J.H., W.R. Mabey, N. Bohonos, B.R. Holt, S.S. Lee, T.-W. Chou, D.C. Venberger, and
T. Mill. 1978. Environmental Pathways of Selected Chemicals in Freshwater Systems. Part 2:
Laboratory Studies. U.S. EPA: Athens, GA. EPA-600/7-78-074.
Southworth, G.R., and J.L. Keller. 1986. Hydrophobic sorption of polar organics by low organic
carbon soils. Water. Air, and Soil Pollution. 28:239-248.
Spencer, W.F. and M.M. Cliath. 1970. Soil Sci. Soc. Am. Proc. 34:574.
Stauffer, T.B., and W.G. Maclntyre. 1986. Sorption of low-polarity organic compounds on
oxide minerals and aquifer material. Environ. Toxicol. Chem. 5:949-955.
Verschueren, K. 1983. Handbook of Environmental Data on Organic Chemicals. 2nd ed., Van
Nostrand Reinhold: New York, NY.
Voice, T.C., and WJ. Weber, Jr. 1985. Sorbent concentration effects in liquid/solid partitioning.
Environ. Sci. Technol. 19(9):789-796.
D-15
-------
Vowles, P.O. and R.F.C. Mantoura. 1987. Sediment-water partition coefficients and HPLC
retention factors of aromatic hydrocarbons. Chemosphere. 16(1):109-116.
Wahid, P.A. and N. Sethunathan. Sept./Oct. 1979. Sorption-desorption of alpha, beta and
gamma isomers of hexachlorocyclohexane in soils. J. Ag. and Food Chem. 27(5): 1050.
Wauchope, R.D., K.E. Savage, and W.C. Koskinen. 1983. Weed Sci. 31:744-751.
Weber, W.J., Jr., T.C. Voice, M. Pirbazari, G.E. Hunt, and D.M. Ulanoff. 1983. Sorption of
hydrophobic compounds by sediments, soils and suspended solids - II. Water Res. 17(10): 1443-
1452.
Webster, G.R.B., KJ. Friesen, L.P. Sarna, and D.C.G. Muir. 1985. Chemosphere. 14:609-622.
Wildish, D.J., C.D. Metcalfe, H.M. Akagi, and D.W. McLeese. 1980. Flux of aroclor 1254
between estuarine sediments and water. Bull. Environ. Contam. Toxicol. 24:20-26.
Wilson, J.T., C.G. Enfield, W.J. Dunlap, R.L. Cosby, D.A. Foster, and L.B. Baskin. 1981.
Transport and fate of selected organic pollutants in a sandy soil. J. Environ. Qual. 10(4):501-
506. w- C
Witkowski, P.J., P.R. Jaffe, and R.A. Ferrara. 1988. Sorption and desorption dynamics of
aroclor 1242 to natural sediment. J. Contam. Hydrol. 2:249-269.
D-16
•ft U S GOVEflNMENT POINTING OFFICE. 1994— 300-642 /12658
------- |