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

-------