v>EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-451/R-96-001
November 1995
Air/Superfund
AIR/SUPERFUND
NATIONAL TECHNICAL
GUIDANCE STUDY SERIES
Guideline for Predictive Baseline
Emissions Estimation for
Superfund Sites
Interim Final
-------
GUIDELINE FOR PREDICTIVE
BASELINE EMISSIONS ESTIMATION
V FOR SUPERFUND SITES
> ASF-21
by
Environmental Quality Management, Inc.
Cedar Terrace Office Park, Suite 250
3325 Chapel Hill Boulevard
Durham, North Carolina 27707
Contract No. 68-D3-0032
Task Order No. 6
PN 5094-3
Patricia Flores, Work Assignment Manager
U.S. ENVIRONMENTAL PROTECTION AGENCY
REGION III
841 CHESTNUT BUILDING
PHILADELPHIA, PENNSYLVANIA 19107
1QQ
-------
PREFACE
This document was developed for the U.S. Environmental Protection Agency,
Air/Superfund Program and offers technical guidance for use by EPA Air and Superfund
staff, remedial and removal contractors, and potentially responsible parties. Because
assumptions and judgments are required in many parts of the analysis, the user of this
manual needs a strong technical background in emissions and atmospheric dispersion
modeling.
It is envisioned that this manual will be periodically updated to incorporate new
data and information on air pathway analysis procedures. The Agency reserves the right
to act at variance with these procedures and to change them as new information
becomes available without formal public notice.
-------
DISCLAIMER
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use by the U.S. Environmental Protection Agency. The contents
of this report are reproduced herein as received from the contractor. The opinions,
findings, and conclusions expressed are those of the authors and are not necessarily
those of the U.S. Environmental Protection Agency.
-------
CONTENTS
Preface ii
Disclaimer iii
Figures vi
Tables vi
Acknowledgment vii
Objective 1
Background and Purpose 1
Screening Versus Refined Estimates 2
Model Sampling Requirements 3
How to Use These Procedures 5
How the Document is Organized 5
Major Steps in the Air Pathway Analysis 6
Summary 6
Major Steps
Step I - Review Site Background and Gather Data Necessary
to Conduct the Baseline Emissions Estimate 8
Step II - Develop a Site Modeling Plan 10
Step III - Estimate Emission Rates of Each Applicable
Site Contaminant 13
Part 1 - Vapor-Phase Emissions From Surface and
Subsurface Soils Without Landfill Gas
Generation 14
Part 2 - Vapor-Phase Emissions From Subsurface
Soils With Landfill Gas Generation 36
Part 3 - Gaseous Emissions From Nonaerated Surface
Impoundments, Open Top Wastewater Tanks and
Containers, and Aqueous-Phase Contaminants
Pooled at Soil Surfaces 40
Part 4 - Nonvolatile and Semivolatiles Emitted as
Particulate Matter 44
IV
-------
CONTENTS (continued)
Step IV - Estimate Ambient Air Concentrations at
Receptor Locations of Interest 56
Step V - Organize Exposure Point Concentrations for Input to the
Baseline Risk Assessment 61
Appendices
A. Recharge Estimates by Hydrogeologic Settings A-1
B. Example Calculations B-1
v
-------
FIGURES
Number
1
1a
1b
2
3
Wind erosion model decision flow chart
Threshold friction velocity versus aggregate size distribution
Increase in threshold friction velocity with L,.
Function curve used in "unlimited reservoir" model
Thornthwaite's precipitation - evaporation index (PE)
for State climatological divisions
45
47
49
51
53
TABLES
Number
1
2
3
Parameter Estimates for Calculating Average Soil Moisture
Content (0W)
Viscosity of Water at Temperature (f7w)
Additive Volume Increments for Calculating the LeBas
Molar Volume (VB)
16
23
24
VI
-------
ACKNOWLEDGMENT
This manual was prepared for the U.S. Environmental Protection Agency by
Environmental Quality Management, Inc. under Contract No. 68-D3-0032, Task Order No.
6. Mr. Craig Mann (Project Manager) and Ms. Patricia Flores (Work Assignment
Manager) managed the project. The principal author was Mr. Craig Mann of
Environmental Quality Management, Inc.
VII
-------
GUIDELINE FOR PREDICTIVE BASELINE EMISSIONS ESTIMATION
FOR SUPERFUND SITES
OBJECTIVE:
The objective of the following guidance is to provide standardized procedures as
well as acceptable fate and transport models for estimating air pathway exposure point
concentrations when applied to the baseline risk assessment.
BACKGROUND AND PURPOSE:
Part of the human health evaluation, the baseline risk assessment is an analysis
of the potential adverse health effects (current and future) caused by hazardous
substance releases from a site in the absence of any actions to control or mitigate these
releases. An integral part of the baseline risk assessment is the exposure assessment,
whereby estimates are made of the magnitude of actual and/or potential human
exposures, the frequency and duration of these exposures, and the pathways by which
humans are potentially exposed.
Exposure point concentrations are estimated using monitoring data and/or
chemical transport and environmental fate models. Modeling may be used to estimate
future chemical concentrations in media that are currently contaminated or that may
become contaminated, and current concentrations in media and/or locations for which
there are no monitoring data. In some instances, it may not be appropriate to use
monitoring data alone, and models may be required to estimate concentrations. Specific
instances where monitoring data alone may not be adequate are as follows:
0 Where exposure points are spatially separate from monitoring or sampling
points (e.g., air dispersion)
0 Where temporal distribution of data is lacking (e.g., potential future
exposure)
0 Where monitoring data are restricted by the limit of quantitation.
Under circumstances such as these, monitoring data must be supplemented with
modeling estimates or modeling substituted for monitoring to derive exposure point
concentrations.
-------
The purpose of this document is to provide guidance on applying chemical fate
and transport models for the air pathway. Unlike other media, the air pathway is
characterized by short migration times, relatively large exposure areas, and a virtual
inability to mitigate the consequences of a release after the contaminants enter the
atmosphere. Exposure durations may range from only a few minutes to many years and
exposure point concentrations may vary widely due to the fluidity of atmospheric
processes. In addition, the air pathway exposure point concentration associated with a
cancer risk of 1 in 1,000,000 for some chemicals is below current detection limits in
ambient air. For these reasons, the use of predictive models may be advantageous in
estimating exposure point concentrations.
SCREENING VERSUS REFINED ESTIMATES:
Screening estimates of exposure point concentrations may be used in the baseline
risk assessment during the identification of exposure pathways. During this phase of the
baseline risk assessment, screening-level air pathway analyses my be used to help select
those contaminants that will be evaluated further in the detailed analysis or to eliminate
completely a specific air pathway (e.g., inhalation of fugitive dust). Such a justification
could be based on one of the following:
0 The exposure resulting from the pathway is much less than that from
another pathway involving the same medium at the same exposure point
0 The potential magnitude of the risk is low (< 1 x 10* for carcinogens and/or
< a hazard quotient of 1 for noncarcinogens)
0 The probability of the exposure occurring is very low and the risks
associated with the occurrence are low
Use of professional judgement and experience must be exercised to make these
decisions. Before such a decision is made, the site RPM must be consulted.
If a screening-level approach suggests a potential health concern, the estimates
of exposure should be modified to reflect more probable exposure conditions. While
screening-level analyses typically involve the use of conservative models and conservative
default model variables and assumptions, refined estimates are made using more
sophisticated models and site-specific data as input variables. This necessarily requires
a larger site characterization data set.
Sophisticated modeling required to achieve refined exposure point concentration
estimates typically requires that the three dimensional media-specific concentration of
contaminants within the site volume be determined. In addition, other model parameters
will require site-specific data collection (e.g., soil organic carbon content, temperature,
and meteorological data).
-------
Because screening-level analyses typically require less data, they may be
performed earlier in the process and require less resources than do refined analyses.
Screening-level procedures are designed to provide bounding estimates of exposure
which are compared to point-of-departure human health benchmarks. Therefore, all air
pathways and air contaminants for which screening-level modeling exceed a risk-level of
1 x 106 for carcinogens and meet or exceed a hazard quotient of 1 for noncarcinogens,
warrant refined analyses.
MODEL SAMPLING REQUIREMENTS:
As discussed previously, contaminant release, transport, and fate models are often
needed to supplement monitoring data or in lieu of monitoring data when estimating
exposure point concentrations. Therefore, a preliminary site modeling strategy should be
developed during the scoping phase of the Remedial Investigation (RI)/Feasibility Study
(FS) to allow model input data requirements to be incorporated into site characterization
sampling plans.
Both screening and refined models require physical data about the media
incorporating the contamination as well as data about the transporting media. In addition,
physical and chemical property data are required for each contaminant exhibiting a
potential for air release. Finally, the extent of contamination within the incorporating media
must be determined to varying degrees depending upon the assumptions and limitations
of the model(s) selected. For example, screening-level models which estimate emissions
of volatile contaminants from soils typically assume an infinitely deep source of
homogenous contamination, while refined models estimate emissions from a finite source
of contamination. Soil sampling requirements for the screening-level model will therefore
necessitate less extensive sampling. The site modeling strategy should be integrated with
the site sampling plan such that sampling data are sufficient for each modeling phase.
It must be recognized that in general, fate and transport models are inherently
conservative because they attempt to predict inter-media mass transfer of contaminants
based on theoretical processes and/or empirical relationships. In addition, some vectors
of contaminant loss (e.g., transformation) may not be taken into account because such
mechanisms are not completely understood. Model validation is often limited to bench
or pilot-scale studies, and to a lesser extent, to field-scale studies. Therefore, model
predications of exposure point concentrations should be accompanied by an assessment
of model uncertainties. To reduce the relative error and uncertainty of model predictions,
values of critical model input parameters should be as accurate as resources and time
constraints allow. Where possible, model predictions should be compared with
monitoring and/or sampling data to reduce overall uncertainty.
The degree of sampling required to adequately determine model input parameter
values will depend on the modeling objectives and the heterogeneity of the media
incorporating the contaminants. For example, sampling for contaminants in surface and
-------
subsurface soils presents potential difficulties in that the soil medium is not a well mixed
compartment. Typically, contaminant concentrations as well as soil properties are not
homogeneously distributed within the three-dimensional site volume. Therefore, the site
sampling plan should be designed to estimate the spatial and statistical distribution of
contaminants within specified limits.
The recommend approach for developing the site sampling plan is to use the Data
Quality Objectives (DQO) process. The DQO process is a systematic data collection
planning process developed by EPA to ensure that the right type, quality, and quantity
of data are collected to support Agency decisionmaking. The DQO process generates
quantitative and qualitative statements (DQOs) that clarify the purpose of the data
collection effort; define the most appropriate type of data and the conditions under which
the data should be collected; and specify quantitative performance criteria for using the
data. This process is based on the scientific method and usually results in a statistical
sampling plan that allows the site manager to draw valid inferences about contamination
levels over areas of the site.
The DQO process employs statistical concepts for developing either a probalistic
or nonprobalistic sampling plan. Consequently, the DQO process is most successful
when the investigating team includes a member who is knowledgeable in statistics. This
will help ensure that existing data and other information about the site are used most
effectively in designing the sampling plan.
The following EPA documents are useful for developing an adequate site sampling
plan in support of the site modeling plan:
Data Quality Objectives for Superfund: Interim Final Guidance, U.S.
Environmental Protection Agency, Office of Emergency and Remedial Response,
Washington, D.C. EPA/540/R-93/071, Publication 9255.9-01, NTIS PB94-963203,
1993.
Guidance for Data Usability in Risk Assessment (Part A), U.S. Environmental
Protection Agency, Office of Emergency and Remedial Response, Washington,
D.C. Publication 9285.7-09A, NTIS PB94-963203, 1993.
Supplemental Guidance to RAGS: Calculating the Concentration Term, U.S.
Environmental Protection Agency, Office of Emergency and Remedial Response,
Washington, D.C. Publication 9285.7-08I, NTIS PB92-963373, 1992.
Technical Background Document for Soil Screening Guidance: Review Draft.
U.S. Environmental Protection Agency, Office of Emergency and Remedial
Response, Washington, D.C. EPA/540/R-94/102, Publication 9355.4-14-1, NTIS
PB95-963530, 1994.
-------
Guidance for Conducting Remedial Investigations and Feasibility Studies Under
CERCLA, U.S. Environmental Protection Agency, Office of Emergency and
Remedial Response, Washington, D.C. EPA-540/G-89-004, OSWER Directive
9355.3-01, NTIS PB89-184626, 1989.
HOW TO USE THESE PROCEDURES:
The intent of this document is to provide the sequential series of steps necessary
to estimate exposure point concentrations for use in the baseline risk assessment. The
models herein have been extracted from various references; and therefore as a guideline,
all of the relevant information concerning the applicability, assumptions, and limitations of
each model are not necessarily included. To properly use these procedures, the user
must thoroughly understand all relevant information from the original references cited
throughout this document.
Use of these procedures does not preclude the use of techniques for measuring
actual emission rates or ambient air concentrations of airborne contaminants. Where site-
specific conditions do not lend themselves to the use of predictive models, more rigorous
procedures involving measurement techniques may be required. Emissions measurement
and ambient air sampling techniques as well as other modeling techniques which may be
more suited to site-specific conditions can be found in the EPA Air/Superfund Technical
Guidance Study Series. These guidance documents cover a wide range of air issues
associated with the characterization and remediation of Superfund sites. Copies of these
documents are available through the National Technical Information Service (NTIS) in Port
Royal, Virginia.
Finally, it should be understood that these procedures are hierarchical in nature,
building upon preceding steps. Mistakes or inaccurate data in individual steps will cause
the final estimated values to exhibit considerable relative error. If problems or questions
arise, contact your EPA Regional Air/Superfund Coordinator for assistance.
HOW THE DOCUMENT IS ORGANIZED:
This guideline is organized around the major steps required to conduct the air
pathway analysis for the baseline risk assessment, with emphasis placed on the modeling
procedures appropriate for the exposure assessment. The user should be familiar with
the EPA procedures and guidance described in the Risk Assessment Guidance for
Superfund, Human Health Evaluation Manual, Part A, July 1989 and associated
documents.
The body of this document is divided into two main sections: 1) use of emission
models, and 2) use of atmospheric dispersion models. The emission modeling section
is further organized into subsections describing how to estimate emissions from specific
media (e.g., volatile emissions from lagoons or subsurface soils). Each subsection
-------
provides both screening-level and refined modeling procedures. Like the emission
modeling section, the atmospheric dispersion modeling section also provides both
screening-level and refined procedures.
For ease of use, the procedures are organized in outline form. Within each
subsection, the appropriate model solution is given including a legend of model
parameters. Solutions to model parameters which must be calculated are given either
within the legend or as separate equations. Notes on model application are also provided
after the equations as well as model references. It is strongly recommended that every
effort be made to obtain cited references to ensure that each model is used properly and
that assumptions and limitations are understood. The references cited are also valuable
for the analysis of uncertainty required in the baseline risk assessment.
Finally, Appendix B to this document provides example calculations for each model.
In application, these models are typically executed within a computer spreadsheet or
similar program. The appendix solutions can be used to perform quality assurance
checks on spreadsheet calculations.
MAJOR STEPS IN THE AIR PATHWAY ANALYSIS:
I. Review site background information and gather data necessary to conduct
the baseline emissions estimate.
II. Develop a site modeling plan.
III. Estimate emission rates of each applicable site contaminant.
IV. Estimate ambient air concentrations at receptor locations of interest.
V. Organize exposure point concentrations for input to the baseline risk
assessment.
SUMMARY:
As with most fate and transport models, the refined model algorithms provided in
this guidance are intrinsically conservative, and when properly applied, may be said to
exhibit a relative error of perhaps one order of magnitude. The screening-level models
provided herein are even more conservative because of model assumptions and because
the default model parameter values provided represent high-end values of known or
theoretical distributions. Screening-level models, therefore, produce bounding estimates
which may be used to eliminate contaminants and/or pathways from further evaluation.
With site-specific model parameter values and with adequate site characterization data
(i.e., contaminant concentrations and distributions), the refined models may be used to
estimate exposure point concentrations consistent with reasonable maximum exposure
-------
(RME) assumptions. To be consistent with present EPA guidance, the refined modeling
procedures predict long-term average exposure point concentrations. These
concentrations are subsequently used with high-end chronic exposures (i.e., intake and
duration) to derive RME for both current and future land-use assumptions.
7
-------
STEP I. REVIEW SITE BACKGROUND AND GATHER DATA NECESSARY TO
CONDUCT THE BASELINE EMISSIONS ESTIMATE
1. Review the nature of contamination at the site and identify potential air exposure
pathways and receptors (e.g., inhalation of volatile organics by an adult resident
100 m from the fence line, etc.).
2. Assemble all relevant site data. This information may include but is not limited to:
0 Site configuration and features (maps)
0 List of identified chemical contaminants
0 Sampling concentration data for each media (soil, water, and/or air)
0 Spatial coordinates of each sample in three dimensions
0 Soils data (porosity, bulk density, and/or particle density, moisture
content, etc.)
0 Presence of soil crust and crust thickness, friability and soil
aggregate size distribution
0 Location and distance to receptors of interest
0 Local meteorological data (annual average temperature, windspeed
and prevailing direction, if applicable)
0 Extent of surface vegetation and/or surface coverings.
3. Assemble chemical property data for all site contaminants. Data requirements will
vary depending on the type of compound and specific emission rate equation (s)
used in Step III. Chemical properties required for modeling may include:
0 Molecular weight
0 Vapor pressure
0 Henry's law constant
0 Diffusion coefficient in air
0 Diffusion coefficient in water
0 Organic carbon partition coefficient
0 Solubility in water.
8
-------
References for Physical and Chemical Properties of Contaminants:
Air/Superfund National Technical Guidance Study (NTGS) Series, Volume II -
Estimation of Baseline Air Emissions at Superfund Sites. Appendices F and G,
U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards, Research Triangle Park, North Carolina, EPA-450/1-89-002a, NTIS
PB90-270588, August 1990.
CHEMDAT8 Data Base of Compound Chemical and Physical Properties. U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards
Technology Transfer Network (TTN), CHIEF Bulletin Board, Research Triangle
Park, North Carolina, TTN Data Number: (919) 541-1447, SYSOP Number: (919)
541-4814.
Environmental Fate Constants for Organic Chemicals Under Consideration for
EPA's Hazardous Waste Identification Projects. U.S. Environmental Protection
Agency, Office of Research and Development, Athens, Georgia, EPA/600/R-
93/132, NTIS PB93-221646, August 1993.
Estimation Programs Interface (EPI) PC-Program or individual estimation
modules. Syracuse Research Corp., Syracuse, New York, EPI and modules have
been assessed by the U.S. Environmental Protection Agency, Office of Pollution
Prevention and Toxics. Syracuse Research Corp.
Technical Background Document for Soil Screening Guidance, Review Draft.
U.S. Environmental Protection Agency, Office of Emergency and Remedial
Response, Washington, D.C., EPA/540/R-94/102, NTIS PB95-963530, November
1994.
-------
STEP II. DEVELOP A SITE MODELING PLAN
1. Based on process knowledge, historical records, or prior sampling,
subdivide the site into regions where the contaminant variability is likely to
be similar within each region.
2. Further subdivide each site region identified in No. 1 above, into square air
pathway exposure areas (EAs). An air pathway EA is the smallest area over
which an individual can reasonably be expected to move over a period of
time and be exposed to air contaminants.
A. For residential future land-use, EPA has determined that each EA
should be no larger than 0.5 acres.
B. For commercial/industrial or other future land-uses, determine the
high-end EA (approximately the 10th percentile area) using historic
land-use data for the site geographic region.
C. For ground-level air release from surface water bodies, the future
land-use EA will normally be equivalent to the soil EA(s) adjacent to
the source of emissions.
3. For surface soils (< 6 inches deep), develop a site sampling plan to
determine the 95% upper confidence limit (UCL) of the arithmetic mean
concentration of each contaminant within each EA.
4. Develop a site sampling plan for subsurface soils:
A. Determine the 95% UCL of the arithmetic mean concentration of
each contaminant within each EA soil volume (i.e., in three
dimensions) using the DQO process,
OR
B. Drill at least two to three boreholes within each EA and sample in 2-
foot intervals from 6 inches below the ground surface until no further
contamination or the ground water table is encountered.
1. Average the analysis results of the discrete samples (volatiles)
or combine discrete samples into a composite sample
(semivolatiles and nonvolatiles) for each borehole. If
composite sampling is used for semivolatiles and nonvolatiles,
only samples within the zone of contamination should be used
(i.e., do not mix clean soil with contaminated soil).
10
-------
2. With discrete sampling, use the highest depth-weighted
average borehole concentration within each EA to represent
the mean concentration of each contaminant.
Note: The number and location of subsurface soil sampling (i.e., borehole)
locations within each EA should be based on knowledge of likely soil
contamination patterns and subsurface conditions. This usually means that
core samples be taken directly beneath areas of high surface
contamination. Thus, surface soil samples and field measurements (e.g.,
soil gas surveys) will provide information to help locate core sampling. Note
that there may be sources buried in subsurface soils that are not discernible
at the surface.
If time and resource constraints prevent a determination of the 95% UCL of
the arithmetic mean concentrations to depth (i.e., Step 4.A.), the alternative
sampling strategy presented in Step 4.B. is designed to provide
conservative estimates of subsurface soil contaminant concentrations within
each EA. This prescribed subsurface soil sampling procedure will not be
sufficient to fully characterize the extent of contamination and the mean
concentration of subsurface contaminants because it does not account for
contaminant variability. For this reason, it is conservatively assumed that
the highest average borehole concentration underlies the entire exposure
area.
Composite sampling of surface and subsurface soils may be appropriate for
nonvolatile and semivolatile contaminants. Discrete sampling is required for
volatile contaminants to avoid loss of the analytes. Volatile contaminants
may be said to have a vapor pressure > 0.1 mm Hg and boiling points <
120°C.
5. For aqueous systems, develop a sampling plan to determine the 95% UCL
of the arithmetic mean concentration of each contaminant.
6. After sampling is complete, determine which air pathways or air
contaminants are suitable for screening-level estimates of exposure point
concentrations.
Note: One option is to perform screening-level analyses for all air
pathways and contaminants. If a screening-level analysis is
not performed, a refined analysis must be substituted.
7. Perform screening-level analyses, as appropriate, and estimate risks for
each air contaminant and air pathway.
11
-------
8. Document the results of the screening-level analyses and prepare an
assessment of model uncertainties.
9. Perform a refined analysis for all air pathways and air contaminants for
which screening-level analyses exceed a risk-level of 1 x 1CT6 for
carcinogens and meet or exceed a hazard quotient of 1 for noncarcinogens.
10. Document the results of the refined analyses and prepare an assessment
of model uncertainties.
12
-------
STEP III. ESTIMATE EMISSION RATES OF EACH APPLICABLE SITE
CONTAMINANT
This section includes both screening-level and refined emissions modeling
procedures for the following emission scenarios:
1. Vapor-phase emissions from surface and subsurface soils without landfill
gas generation
2. Vapor-phase emissions from subsurface soils with landfill gas generation
3. Vapor-phase emissions from aqueous-phase contaminants in
impoundments, open-top tanks, and pooled at soil surfaces
4. Semivolatile and nonvolatile contaminants emitted as particulate matter due
to wind erosion.
Model predictions of emissions for these four scenarios are to open spaces. If
vapor-phase intrusion of site contaminants into indoor spaces via subsurface soils is likely
for either current or future land-use, the following references should be consulted to
assess any associated risks:
Air/Superfund National Technical Guidance Study (NTGS) Series, Assessing
Potential Indoor Air Impacts for Superfund Sites. U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Research Triangle Park,
North Carolina, EPA-451/R-92-002, NTIS PB93-122257, September 1992.
Johnson, P. C., and R. A. Ettinger. 1991. Heuristic Model for Predicting the
Intrusion Rate of Contaminant Vapors Into Buildings. Environ. Sci. Technol.,
25(8): 1445-1452.
Technical Background Document for Soil Screening Guidance, Review Draft.
Appendix B. U.S. Environmental Protection Agency, Office of Emergency and
Remedial Response, Washington, D.C., EPA/540/R-94/102, NTIS PB95-963530,
November 1994.
Little, J. C., J. M. Daisey, and W. W. Nazaroff. 1992. Transport of Subsurface
Contaminants Into Buildings, An Exposure Pathway for Volatile Organics.
Environ. Sci. Technol., 26(11):2058-2066.
13
-------
STEP III. PART 1 - VAPOR-PHASE EMISSIONS FROM SURFACE AND
SUBSURFACE SOILS WITHOUT LANDFILL GAS GENERATION
Vapor-phase emissions from surface and subsurface soils without the generation
of landfill gas are a function of the rate of diffusion through the soil matrix and the initial
contaminant soil concentration at the time of sampling. Emissions may be estimated
using either the initial vapor-phase concentration (i.e., soil gas analysis) or the initial bulk
concentration which includes all phases within the soil matrix. Soil gas analysis, however,
can only account for the vapor-phase soil concentration at the time of sampling and
cannot account for the amount of contamination sorbed to soil particles, in solution with
soil moisture, or in residual phase within the soil matrix. Therefore, emission estimates
using soil gas analysis incorporate steady-state equilibrium assumptions (i.e., an infinite
source of emissions). Soil gas analysis is used for screening-level emission estimates,
while bulk samples may be used for screening-level or refined estimates.
Estimating Emissions Using Bulk Soil Concentrations:
1. If soil bulk concentrations are to be used to calculate emission rates, estimate the
saturation concentration (Csat) of each contaminant in the vadose zone. Csat for
each contaminant is the concentration at which the adsorptive limit of the soil plus
the theoretical solubility limit of the liquid-phase contaminant in the available soil
moisture have been reached. Concentrations > Csat indicate potential
nonaqueous-phase liquids (NAPLs) or solids within the soil matrix.
(1)
where Csati = Soil saturation concentration of component i, mg/kg - soil
Ky i = Soil/water partition coefficient of i, L/kg (Equation 3)
H;' = Henry's law constant of i, unitless (= ^ x 41, where 41 is a units
conversion factor at standard conditions)
H, = Henry's law constant of i, atm - rrf/mol
0a = Air-filled soil porosity, unitless (=0t - Gw)
0W = Average long-term volumetric soil moisture content, L-water/
L-soil (Equation 2)
14
-------
0, = Total soil porosity, unitless (=1 - l3>//os)
ps = True soil or particle density, kg/L-soil (default = 2.65 kg/L)
§ = Solubility in water of i, mg/L-water
B = Average soil dry bulk density, kg/L-soil.
Estimation of average long-term volumetric soil moisture content (0W):
3) (2)
where 0W = Average long-term volumetric soil moisture content,
L-water/L-soil
0, = Total soil porosity, unitless (Equation 1 legend)
I = Average water infiltration rate, m/yr (Appendix A)
K. = Soil saturated hydraulic conductivity, m/yr (Table 1)
1/(2b+3) = Soil-specific exponential parameter, unitless (Table 1).
Values for Kg and the exponential term 1/(2b+3) are shown in Table 1 by soil
texture class. Average long-term soil moisture content should be approximated as the
average from the soil surface to the bottom of contamination.
15
-------
TABLE 1. PARAMETER ESTIMATES FOR CALCULATING AVERAGE SOIL MOISTURE
CONTENT (0J
Soil texture
Sand
Loamy sand
Sandy loam
Silt loam
Loam
Sandy clay loam
Silt clay loam
Clay loam
Sandy clay
Silt clay
Clay
K;, m/yr
1,830
540
230
120
60
40
13
20
10
8
5
1/(2b + 3)
0.090
0.085
0.080
0.074
0.073
0.058
0.054
0.050
0.042
0.042
0.039
The average water infiltration rate (I) may be estimated using hydrogeologic settings by
assuming the infiltration rate is equivalent to the recharge rate. Appendix A gives the
estimated recharge rates by hydrogeologic setting for the continental United States.
Reference for Equation Nos. 1 and 2. Table 1. and Appendix A: Technical
Background Document for Draft Soil Screening Level Guidance, Review Draft.
U.S. Environmental Protection Agency, Office of Emergency and Remedial
Response, Washington, D.C., EPA/540/R-94/102, Publication 9355.4-14-1, NTIS
PB95-963530, November 1994.
Estimation of Ky if not available in the scientific literature or not estimated using the
references for Step I.3:
x f
(3)
where
Soil/water, partition coefficient of component i, L/kg (or cnf/g)
Organic carbon partition coefficient of i, L/kg (or cm3/g)
16
-------
foc = Fraction of organic carbon in soil, mg/mg (default = 0.006).
Estimation of K^ if not available in the scientific literature:
Use one of the following equations based on the chemical class closest to
the subject contaminant. If the contaminant does not fit any given class,
use Equation No. 4 (based on largest sampling):
Based on a wide variety of contaminants (mostly pesticides):
*1377' ^
Based on aromatics. polynuclear aromatics. triazines. and dinitroaniline
herbicides:
is _ •< n
° 937 logK^ - 0.006 ) (4a)
oc
Based on aromatics or polynuclear aromatics:
Based on s-triazines and dinitroaniline herbicides:
+ 0.02) (4C)
Based on insecticides, herbicides, and fungicides:
aw) -0.1s) (4d)
17
-------
Based on substituted phenylureas and alkyl-N-phenylcarbamates:
KOC
= 1 QM0'524 logK°'} * ° 855'
where K^ = Organic carbon partition coefficient, L/kg (or cnf/g)
= Octanol/water partition coefficient, L/kg (or cm3/g).
Reference for Equation Nos. 3-4e: Superfund Exposure Assessment
Manual, Section 3.5.2.4, Office of Emergency and Remedial Response,
Washington, D.C., EPA/540/1-88/001, April 1988.
Reference for Value of foe in Equation No. 3: Technical Background
Document for Soil Screening Level Guidance, Review Draft. U.S.
Environmental Protection Agency, Office of Emergency and Remedial
Response, Washington, D.C., EPA/540/R-94/102, Publication 9355.4-1 4-1,
NTIS PB95-963530, November 1994.
18
-------
Volatile Emissions from Surface Soils When NAPL is Present
2. With measured bulk concentrations > Csat (Equation 1), calculate the emission flux
from surface soils for each contaminant.
Note: Under this scenario, nonaqueous-phase liquids (NAPLs) or solids may exist
in the unsaturated portion of the vadose zone, usually as an immiscible
layer or discrete film. The following model should be used in the absence
of clean soil above the contamination.
Screening-Level and Refined Procedures:
When nonaqueous-phase liquids or solids are present, the screening-level and
refined procedures are identical. The emission model operates under the assumption of
steady-state conditions (infinite source) over the averaging period.
\1/2
a116
where Fj = Average maximum emission flux of component i from surface soils
over exposure averaging period r, g/nf-s
B = Average soil dry bulk density, g/cnf-soil
Coi = Initial soil concentration of i, g/g-soil
Q,eq = Equilibrium vapor concentration when NAPL is present in the soil,
g/crrf-vapor (Equation 6)
De j = Maximum effective diffusion coefficient of i when NAPL is present,
crrf/d (Equation 7)
T = Exposure averaging period, days (for residential land-use
T = 10,950 days)
0.116 = Factor to convert g/cmP-d to g/nf-s.
Calculation of equilibrium vapor concentration when NAPL is present (Q,eq):
19
-------
where
R
T
v,eq
X; P, MWj
RT
(6)
Q,eq = Equilibrium vapor concentration of i when NAPL is present,
g/cm3-vapor
^ = Mole fraction of i in the residual mixture, g/mol per g/mol
PI = Pure component vapor pressure of i, mm Hg
= Molecular weight of i, g/mol
= Molar gas constant, 62,361 mm Hg-cm3/mol-°K
= Average in situ soil temperature, ° K.
Calculation of maximum effective diffusion coefficient when NAPL is present (Dei):
0
10/3
T- Da,i
e
10/3
w
a
(7)
where
D
e.i
©
D
a,i
e
Dw,,
= Maximum effective diffusion coefficient of component i when NAPL
is present, crrf/d
= Air-filled soil porosity unitless (Equation 1 legend)
= Total soil porosity, unitless (Equation 1 legend)
= Diffusion coefficient of i in air, cnf/d (Equation 9)
= Henry's law constant of i, unitless (Equation 1 legend)
= Water-filled soil porosity, unitless (Equation 2)
= Average soil dry bulk density, g/cm3-soil
= Diffusion coefficient of i in water, crrf/d (Equation 10).
20
-------
Volatile Emissions from Subsurface Soils When NAPL is Present
3. For measured bulk concentrations > Csat (Equation 1), calculate the emission flux
from subsurface soils for each contaminant.
Note: Under this scenario, nonaqueous-phase liquids (NAPLs) or solids may exist
in the unsaturated portion of the vadose zone, usually as an immiscible
layer or discrete film. The following model should be used when a layer of
clean soil is above the contamination.
Screening-Level and Refined Procedures:
When nonaqueous-phase liquids or solids are present, the screening-level and
refined procedures are identical, the emission model operates under the assumption of
steady-state conditions (infinite source) over the averaging period.
•
0.116
where F, = Average maximum emission flux of component i from subsurface
soils over exposure averaging period r, g/rrf-s
& = Average soil dry bulk density, g/cnf-soil
C0, = Initial soil concentration of i, g/g-soil
r = Exposure averaging period, days (for residential land-use
r = 10,950 days)
c-, = Distance from soil surface to top of contamination, cm
Q,eq = Equilibrium vapor concentration when NAPL is present in the soil,
g/cm3-vapor (Equation 6)
De j = Maximum effective diffusion coefficient of i when NAPL is present,
crrf/d (Equation 7)
0.116 = Factor to convert g/cnf-d to g/rrf-s.
21
-------
Estimation of diffusion coefficient of component i in air (Dai) if not available from the
scientific literature or not estimated using the references in Step 1.3:
0.001 71-75
(9)
(8.64 x 104) sec/Jay
IE V, )1'3 + (E V.
where Dai = Diffusion coefficient of component i in air, cnf/d
T = Average temperature, ° K
(; MWa = Molecular weight of component i and air (28.8), respectively,
g/mole
Pab = Absolute pressure, atmospheres
; ZVa = Molecular diffusion volumes of component i and air (20.1),
respectively, cnf/mol. This is the sum of the atomic diffusion
volumes of the compound's atomic constituents.
Atomic diffusion volumes for use in estimating Da -t:
C = 16.5 Cl = 19.5 Aromatic ring = -20.2
H = 1.98 Br = 35.0 Heterocyclic ring = -20.2
O - 5.48 F = 25.0
N = 5.69 S = 17.0
Example of calculating IV, for benzene, QH^:
C = 6 x 16.5 = 99.00
H = 6x1.98 = +11.88
110.88
Aromatic ring = -20.20
90.68 crrf/mol
Note: Equation No. 9 may not be appropriate for polar compounds. Where
possible, values of Dai in the scientific literature should be
used.
22
-------
Estimation of diffusion coefficient of component i in water (Dwi) if not available from
the scientific literature or not estimated using the references in Step 1.3:
13.26 A-1Q-5
1.14
\0.589
(8.64 *104 sec/day)
(10)
where Dwl = Diffusion coefficient of component i in water, crrf/d
/7W = Viscosity of water at average in situ soil temperature, centipoise
(Table 2)
\£i = LeBas molar volume of i, cnf/mol (Table 3).
TABLE 2. VISCOSITY OF WATER AT TEMPERATURE
°c
0
1
2
3
4
5
6
7
8
9
10
/7w(cp)
1.787
1.728
1.671
1.618
1.567
1.519
1.472
1.428
1.386
1.346
1.307
°C
11
12
13
14
15
16
17
18
19
20
^7w(cp)
1.271
1.235
1.202
1.169
1.139
1.109
1.081
1.053
1.027
1.002
°C
21
22
23
24
25
26
27
28
29
30
/7W(CP)
0.9779
0.9548
0.9325
0.9111
0.8904
0.8705
0.8513
0.8327
0.8148
0.7975
23
-------
TABLE 3. ADDITIVE VOLUME INCREMENTS FOR CALCULATING THE LeBas MOLAR
VOLUME (V B)
Atom
C
H
O (except as noted below)
In methyl esters and ethers
In ethyl esters and ethers
In higher esters and ethers
In acids
Joined to S, P, N
N
Double bonded
In primary amines
In secondary amines
Increment,
cm3/mol
14.8
3.7
7.4
9.1
9.9
11.0
12.0
8.3
15.6
10.5
12.0
Atom
Br
Cl
F
I
S
Ring
3-Membered
4-Membered
5-Membered
6-Membered
Naphthalene
Anthracene
Increment,
crrf/mol
27.0
24.6
8.7
37.0
25.6
-6.0
-8.5
-11.5
-15.0
-30.0
-47.5
Example of calculating VB' for aniline,
C = 6x 14.8 =
H = 7x3.7 =
N = 1 x N-primary amine =
6-membered ring
88.8
25.9
10.5
125.2
-15.0
110.2 cm3/mol
Reference for Equation Nos. 5. 6. 7. and 8: Emergency Standard Guide
for Risk-Based Corrective Action Applied at Petroleum Release Sites.
American Society for Testing and Materials, Philadelphia, Pennsylvania.
ASTM ES38-94, December 1994.
Reference for Equation No. 9: Superfund Exposure Assessment Manual,
Section 2.3.2. U.S. Environmental Protection Agency, Office of Emergency
and Remedial Response, Washington, D.C., EPA/540/1-88/001, April 1988.
24
-------
Reference for Equation No 10: Lyman, W. J., W. F. Reehl, and D. H.
Rosenblatt. Handbook of Chemical Property Estimation Methods.
American Chemical Society, Washington, D.C., 1990.
25
-------
Volatile Emissions from Surface Soils When NAPL is Not Present
4. For measured bulk concentrations < Csat (Equation 1), calculate the emission flux
from surface soils for each contaminant.
Note: Under this scenario, all contaminants are assumed to be in equilibrium
between aqueous-phase, sorbed-phase, and vapor-phase (fully incorporated). The
following model should be used in the absence of clean soil above the
contamination.
Screening Procedures:
l^f 0.116 OH
where F, = Average maximum emission flux of component i from surface soils
over exposure period r, g/rrf-s
& - Average soil dry bulk density, g/cnf-soil
Coi = Initial soil concentration of i, g/g-soil
DAi = Apparent diffusion coefficient of i in soil, crrf/d (Equation 12)
n = 3.1416
r = Exposure averaging period, days (for residential land-use
r = 10,950 days)
0.116 = Factor to convert g/cnf-d to g/rrf -s.
Calculation of apparent diffusion coefficient (DAi):
where DAi = Apparent diffusion coefficient of component i in soil, crrf/d
26
-------
0a = Air-filled soil porosity, unitless (Equation 1 legend)
Daj = Diffusion coefficient of i in air, crrf/d (Equation 9)
0W = Water-filled soil porosity, unitless (Equation 2)
Dwi = Diffusion coefficient of i in water, cnf/d (Equation 10)
0, = Total soil porosity, unitless (Equation 1 legend)
ft = Average soil dry bulk density g/cm3-soil
K^ = Soil/water partition coefficient of i, crrf/g (Equation 3)
HJ' = Henry's law constant of i, unitless (Equation 1 legend).
Refined Procedures:
where Fj = Average emission flux of component i over exposure averaging
period r, g/rrf-s
T = Exposure averaging period, days (for residential land-use
7- = 10,950 days)
Fj(t) = Instantaneous emission flux of i at time = t, g/rtf-s
t = Time, days.
Integration of Equation No. 13:
Note: In order to numerically integrate Equation 13, the instantaneous emission
flux must be calculated for a series of time-steps from t=0 to t=r (Equation
15). For acceptable resolution, a minimum of 100 time-steps is required for
extended exposure averaging periods (e.g., 30 years). Therefore:
27
-------
(14)
where r = Exposure averaging period, days (Residential = 30 yr = 10,950
days)
h = Time-step interval for ^ to 1n, days (= r/n)
n = Number of time-step intervals (default = 100)
Fo,i,2. .n = Emission flux of i at time-zero fa) and each succeeding time-step,
g/nf-s (Equation 15)
Time-zero fa) should be set equal to 0.25 days.
Calculation of emission flux from surface soils at each time-step (i.e., contamination
begins at soil surface):
D
'2
A,i
1-exp
0.116
(15)
where Fi(t) = Emission flux of component i from surface soils at each time-step,
g/nf-s
0 = Average soil dry bulk density, g/cnf-soil
Coi = Initial soil concentration of i, g/g-soil
= Apparent diffusion coefficient of i in soil, cnf/d (Equation 12)
= 3.1416.
= Cumulative time at each time-step, days
LC = Depth from soil surface to bottom of contamination, cm
0.116 = Factor to convert g/cnf-d to g/nf-s.
DA.,
n
t
28
-------
IMPORTANT NOTE:
Equation 14 is a numerical evaluation of the integral in Equation 13. Errors may
result for volatilization fluxes from chemicals with large Henry's law constants in
combination with relatively shallow depths of contamination unless extremely small time
intervals are used in the numerical integration. The errors result in overestimation of the
average emission flux such that for an extremely long time period (e.g., 30 years), the
cumulative mass lost through volatilization may exceed the initial mass present in the soil.
To eliminate this potential error, compare the total mass lost using the refined volatilization
model to the total initial mass:
= F A r (16)
, c
MfJ = COJ 0 Ae Lc (17)
where Iv^ = Total mass of component i lost by volatilization over the exposure
averaging period, g
l\4 T = Total initial mass of i, g
F; = Calculated average emission flux of i over the exposure averaging
period r, g/rtf-s
Ac = Area of contamination, nf
T = Exposure averaging period, s
Coi = Initial soil concentration of i, g/Mg-soil (= mg/kg)
G = Average soil dry bulk density, Mg/rrf-soil (= g/cnf)
Lc = Average depth to the bottom of contamination, m.
Note: In the case of buried waste, substitute the thickness of contamination (wc)
for the depth to the bottom of contamination (Lc).
29
-------
If
M-T > MiT ,
The average emission flux over the exposure averaging period (g/rrf-s) is
(18)
i
A
c
Note: See Step III.6 on page 33 for refined modeling computer codes.
30
-------
Volatile Emissions from Subsurface Soils When NAPL is not Present
5. For measured bulk concentration < Csat (Equation 1), calculate the emission flux
from subsurface soils for each contaminant.
Note: Under this scenario, all contaminants are assumed to be in equilibrium
between aqueous-phase, sorbed-phase, and vapor-phase (fully
incorporated). The following model should be used when a layer of clean
soil is above the contamination.
Screening-Level and Refined Procedures:
When the contamination is covered by clean soil, the screening-level and refined
procedures are identical. The average emission flux of component i (g/nf-s) is calculated
using Equation 13 and integrated using Equation 14.
The emission flux from subsurface soils at each time-step is calculated as:
DA.>
n t
1/2
exp
dl
4DA, f
- exp
K + "c)2"
4 DA! t
0.116
(19)
where Fi(t) = Emission flux of component i from subsurface soils at each
time-step, g/nf-s
/? = Average soil dry bulk density, g/cnf-soil
= Initial soil concentration of i, g/g-soil
= Apparent diffusion coefficient of i in soil, cm2/d (Equation 12)
77 = 3.1416
-'o.i
\i
t = Cumulative time at each time-step, days
c-. = Depth from soil surface to top of contamination, cm
wc = Thickness of contaminated soil, cm
0.116 = Factor to convert g/cnf-d to g/rrf-s.
31
-------
See the IMPORTANT NOTE after Equation 15.
Reference for Equation Nos. 11. 12. 15. and 19: A Comparison of Soil
Volatilization Models in Support of Superfund Soil Screening Level Development.
Environmental Quality Management, Inc., developed for the Office of Emergency
and Remedial Response, Washington, D.C. Contract No. 68-D3-0035, Work
Assignment No. 0-25, September 1994.
Jury, W. A., D. Russo, G. Streile, and H. El Abd. 1990. Evaluation of Volatilization
by Organic Chemicals Residing Below the Soil Surface. Water Resources
Research, Vol. 26, No. 1:13-20.
32
-------
Computer Codes for Estimating Volatile Emissions from Surface and Subsurface
Soils When NAPL is Not Present
6. The refined procedures specified in Step 111, Part 1, Sections 4 and 5 account for
a finite source of emissions. These procedures, however, do not take into account
the effects of water advection in the vadose zone (i.e., surface evaporation due to
capillary action or leaching to the water table). Nor do the refined procedures
account for the effects of a boundary layer at the soil-air interface. The effects of
water advection may be significant for highly mobile compounds (e.g., phenol),
and the effect of a soil-air boundary layer will act to reduce emissions for
compounds with relatively low Henry's law constants (< 2.5 x 1CT5).
Two public domain PC software programs are available from the U.S. EPA which
are capable of estimating time-dependent emissions of volatiles from both surface
and subsurface soils and which include solutions for both water advection and a
soil-air boundary layer.
The first program is called thejExposure .Model for Soil-Organic fate and .Transport
(EMSOFT). This program is a menu-driven enhancement of the analytical solution
of Jury et al. (1990) and is capable of calculating time and depth-averaged
emissions and soil concentrations as well as emissions and soil concentrations
versus time. The program can be used with stratified initial soil concentrations, but
soil properties are assumed constant with depth. EMSOFT is available at no
charge from the U.S. EPA Exposure Assessment Group, Office of Health and
Environmental Assessment by sending a FAX request to Amy Wilkins; the FAX
number is (202) 260-1722 or FTS 8-260-1722. Software orders should include the
requestor's name, organization, address, and return telephone and FAX numbers.
The second computer program is called the Vadose Zone Leaching Model
(VLEACH Version 2.2). VLEACH is a numerical code ground water model that is
also capable of calculating time-dependent emissions from surface and subsurface
soils. Model inputs are similar to that of EMSOFT, but VLEACH allows for areal
distribution of soil properties as a series of polygons, each represented by a
vertical stack of cells which allows for stratified contaminant distributions. VLEACH
Version 2.2 is available from the Center for Subsurface Modeling Support (CSMOS)
at EPA's Environmental Research Laboratory in Ada, Oklahoma by calling (405)
436-8656.
Either EMSOFT or VLEACH can be used to estimate time-averaged volatile
emissions from surface or subsurface soils. These models offer the most accurate
emission estimates available and each is relatively user-friendly. Both models have
been validated with bench-scale and, to a lesser extent, with field-scale
experimental data over relatively short time-periods.
33
-------
Estimating Emissions Using Soil Gas Measurements
7. With measured soil gas concentrations, calculate the emission flux of each
contaminant.
Note: Soil gas measurements account only for the vapor-phase concentration of
each contaminant, and thus cannot account for the total initial concentration
(g/cnf - total volume). Therefore, the assumptions under which the
following emission model operates include an infinite source of emissions.
Screening-Level and Refined Procedures:
When using soil gas analyses, the screening-level and refined procedures are
identical.
d
0.116
(20)
where
-max
= Maximum emission flux of component i, g/rrf-s
0,1 = Measured vapor-phase concentration of i immediately above the
vapor source, g/cnf -vapor (Equation 22)
Djev = Effective vapor-phase diffusion coefficient of i in soil, crrf/d
(Equation 21)
d = Depth from surface to immediately above the vapor source, cm
0.116 = Factor to convert g/crrf-d to g/rrf-s.
Calculation of Effective Vapor-phase Diffusion Coefficient (Djev):
D, =
0
10/3
'a,i T"
ef
(21)
34
-------
where D|ev = Effective vapor-phase diffusion coefficient of component i in soil,
crrf/d
Dai = Diffusion coefficient of i in air, cnf/d (Equation 9)
6a = Air-filled soil porosity above the vapor source, unitless (Equation
1 legend)
0, = Total soil porosity above the vapor source, unitless (Equation 1
legend).
Calculation of the Vapor-phase Concentration (0,,) If Soil Gas Measurements are in Units
of Parts-Per-Million by volume:
C=c > (22)
where Q,, = Vapor-phase concentration of component i, g/cm3 -vapor
Csg | = Measured soil gas concentration of i, ppmv
= Molecular weight of i, g/mol.
Reference for Equations 20 and 21: Emergency Standard Guide for Risk-Based
Corrective Action Applied at Petroleum Release Sites. American Society for
Testing and Materials. Philadelphia, Pennsylvania. ASTM ES38-94, December
1994.
35
-------
STEP III. PART 2 - VAPOR-PHASE EMISSIONS FROM SUBSURFACE SOILS
WITH LANDFILL GAS GENERATION
Vapor-phase emissions from subsurface soils with landfill gas generation are a
function of the convective transport rate of the gas. Codisposal sites contain toxic or
hazardous wastes in combination with municipal or sanitary wastes which generate landfill
gases (e.g., methane, carbon dioxide, and hydrogen gas). These "sweep" gases greatly
increase the migration velocity of volatile nonmethane organic compounds (NMOCs) and
their subsequent release to the atmosphere.
Screening Procedures:
The following screening-level emission model may be used for both vented and
unvented landfills and operates under the assumption that the emission flux has reached
steady-state conditions and that the soil column offers no further resistance to vapor flow.
This model does not require quantitative determination of the soil vapor-phase
concentration of each contaminant (i.e., assumes the saturation vapor concentration) but
does require a determination of which NMOCs are present in the soil gas on a speciated
basis.
/=?** = CWi/Vy0.116 (23)
where F^3* = Maximum emission flux of component i, g/rtf-s
Csvi = Saturation vapor concentration of i, g/cm3-vapor (Equation 24)
Vy = Mean landfill gas velocity in the soil pores, 141 cm/d
0.116 = Factor to convert g/crrf-d to g/rrf-s.
Note: For vented landfills, multiply the emission flux by the landfill area to
obtain the emission rate (g/s).
Calculation of Saturation Vapor Concentration (Csvi):
(24)
RT
36
-------
where CSVI = Saturation vapor concentration of component i, g/cm3-vapor
Pi = Pure component vapor pressure of i, mm Hg
MVy = Molecular weight of i, g/mol
R = Molar gas constant, 62,361 mm Hg-cm3/mol-°K
T = Average absolute in situ soil temperature, ° K.
Refined Procedures:
E. =
/ = - f 'EM dt (25)
T Jo <«
where Ej = Average landfill cell emission rate of component i over the
exposure averaging period r, g/s
r = Exposure averaging period, yr (for residential land-use T = 30
years)
E,(t) = Instantaneous emission rate of i at time = t, g/s
t = Time, yr.
Integration of Equation No. 25:
Note: In order to numerically integrate Equation 25, the instantaneous emission
rate must be calculated for a series of time-steps from t = 0 to t = r. For
acceptable resolution, a minimum of 30 time-steps is required. Therefore:
h
^ (E0 + 2£j + 2E2 + ... + 2En_i + En)
where r = Exposure averaging period, yr
h = Time-step interval for ^ to ^, yr (h = 1 yr)
37
(26)
-------
EQ ! 2 n = Emission rate of i at the initial and each succeeding time-step, g/s
n = Number of time-steps.
If the landfill cell year-by-year acceptance rate is known:
Ew = 2 L0 R {exp(-/rc)-exp(-/rf)} Cvj (3.17 x 1(T2) (27)
where E,(t) = Emission rate of component i at the initial and each succeeding
time-step, g/s
Lj = Methane generation potential, nf/Mg (default = 125
m3/Mg)
R = Average annual landfill cell acceptance rate, Mg/yr
k = Methane generation rate constant, 1/yr (default = 0.04/yr)
c = Time since landfill cell closure to each time-step, yr
t = Cumulative time from initial cell refuse placement to each time-step,
yr
Cvi = Measured vapor-phase concentration of i immediately above the
cell vapor source, g/cm3-vapor (Equation 22)
3.17 x 10"2 = Factor to convert m3/yr to crrf/s.
Note: For unvented landfills, divide the emission rate by the landfill cell area to
obtain the emission flux (g/rrf-s).
If year-by-year landfill cell acceptance rate is unknown:
/? = L WZ pr (1.0 x 10-3)(1/TL) (28)
where R = Average annual cell acceptance rate, Mg/yr
L = Landfill cell length, m
38
-------
W = Landfill cell width, m
Z = Landfill cell depth, m
pr = Refuse density, kg/m3 (default = 650 kg/m3)
1.0 x 10"3 = Conversion factor
TL = Total time landfill cell accepted waste, yr.
Reference for Equation 23: Superfund Exposure Assessment Manual, Section
2.3.2, U.S. Environmental Protection Agency, Office of Emergency and Remedial
Response, Washington, D.C., EPA/540/1-88/001, April 1988.
Reference for Equation 27: Compilation of Air Pollutant Emission Factors,
Volume I: Stationary Point and Area Sources (AP-42). Section 2.4. U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina, 1994.
39
-------
STEP III. PART 3 - GASEOUS EMISSIONS FROM NONAERATED SURFACE
IMPOUNDMENTS, OPEN TOP WASTEWATER TANKS AND CON-
TAINERS, AND AQUEOUS-PHASE CONTAMINANTS POOLED AT SOIL
SURFACES
The following emission models are used to estimate vapor-phase emissions from
aqueous wastes. The screening-level model operates under the assumption of steady-
state conditions.
With aqueous-phase concentrations, calculate the emission flux of each
contaminant.
Screening Procedures:
Ft = K,CU(\ A-104) (29)
where F| = Maximum emission flux of component i, g/nf-s
Kj = Overall mass transfer coefficient of i, cm/s (Equation 30)
CLJ = Liquid-phase concentration of i, g/cm3
(1 mg/L = IxlO-6 g/cm3)
1 x 104 = Factor to convert g/cnf-s to g/rrf-s.
Calculation of overall mass transfer coefficient (Kj):
J_ = -1 + JLL (30)
K, k,L H, k,G
where K, = Overall mass transfer coefficient of component i, cm/s
k|L = Liquid-phase mass transfer coefficient of i, cm/s
(Equation 31)
R = Ideal gas constant, 8.2x10"5 atm-m3/mole-°K
T = Average system absolute temperature, ° K
H, = Henry's Law constant of i, atm-nf/mole
40
-------
k,G = Gas-phase mass transfer coefficient of i, cm/s (Equation
32).
Estimation of liquid-phase mass transfer coefficient (k,L):
where KL = Liquid-phase mass transfer coefficient of component i, cm/s
MVy = Molecular weights of oxygen (32.0) and component i, respectively,
g/mol
T = Average system absolute temperature, ° K
,O2 = Liquid-phase mass transfer coefficient of oxygen at 25° C, 0.002 cm/s.
Estimation of gas-phase mass transfer coefficient (kjG):
** =
MWU
\0.335
MW, )
298
1.005
(32)
where kjG = Gas-phase mass transfer coefficient of component i, cm/s
MWH 0; MWj = Molecular weights of water (18.0) and component i, respectively,
2 g/mol
T = Average system absolute temperature, ° K
kg, HjO = Gas-phase mass transfer coefficient of water vapor at 25°C, 0.833
cm/s.
Note: The screening-level procedures are for quiescent systems. If agitated or
dynamic systems are encountered, refer to the refined procedures that
follow.
41
-------
Reference for Equations 29. 30. 31. and 32: Superfund Exposure Assessment
Manual, Section 2.3.2.1, U.S. Environmental Protection Agency, Office of Emergency
and Remedial Response, Washington, D.C., EPA/540/1-88/001, April 1988.
Refined Procedures:
The refined procedures for estimating vapor-phase emissions from aqueous systems
can be found in the following document:
Air Emissions Models for Waste and Wastewater. U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Research Triangle Park, North
Carolina. EPA-453/R-94-080A, November 1994.
The procedures for estimating emissions due to volatilization from aqueous systems
have been incorporated into a PC-based expert system entitled the Wastewater Treatment
Compound Property Processor and Air Emissions Estimator (WATERS). Both the
document referenced above and the WATERS program are available free of charge from
the EPA Office of Air Quality Planning and Standards (OAQPS) Technology Transfer
Network (TTN) Bulletin Board. The TTN data telephone number is (919) 541-1447; the
SYSOP number is (919) 541-4814. The reference document and WATERS can be found
within the Clearinghouse for Inventories and Emission Factors (CHIEF) section of the TTN
under the subsection entitled "AP-42 and File Transfers" followed by the subsection
entitled "Emission Estimation Software."
Of particular application to baseline conditions at Superfund sites, are the sections
of the reference document entitled Disposal Impoundments with Quiescent Surfaces, Oil
Film Surfaces, and Stationary Tank Storage. It should be kept in mind, however, that the
reference document and the WATERS model are most applicable to dynamic aqueous
systems. Care must be taken to ensure that the correct model is applied to actual site
conditions (e.g., steady-state versus plug flow, etc.). In addition, the user should note
that transformation processes (e.g., biodegradation) are included in some models.
Unless site-specific data are available on transformation rates, these processes should
not be included in the model predictions (i.e., transformation rate constants should be set
equal to zero).
Finally, the WATERS model incorporates the chemical and physical property data
base CHEMDAT8, which includes data for over 900 chemicals. If WATERS is used along
with other emission estimation procedures outlined in this document, chemical and
physical properties must be held constant. This may require revising data for specific
chemicals within the CH EM DATS data base for consistency.
42
-------
Nonaqueous-phase Volatile Contaminants Directly Exposed to the Atmosphere:
For any and all nonaqueous-phase volatile contaminants directly exposed to the
atmosphere, in-depth air pathway analysis (APA) is warranted. Source monitoring is
recommended to determine emission rates, supplemented by ambient monitoring and/or
refined modeling. Applicable situations include open drums/containers, fresh spills, etc.
where residual-phase contamination exists.
43
-------
STEP III. PART 4 - NONVOLATILES AND SEMIVOLATILES EMITTED AS
PARTICULATE MATTER
The following procedures are used to estimate the emissions of nonvolatile
contaminants and semivolatile contaminants adsorbed to soils as fugitive dust generated
by wind erosion. Concentrations used in the emission models should be from erodible
surface materials (surface to 6 inches). In the case of screening-level estimates, no mass
balance is performed for semivolatiles emitted as paniculate matter and also emitted in
vapor-phase from soils. For refined estimates, the average residual concentration within
each exposure area due to wind erosion over the exposure period may be calculated and
this concentration used as the initial concentration for estimating vapor-phase emissions.
Screening-Level Procedures:
For screening-level estimates, use Equation 33 from the refined procedures section
which follows. Set the equivalent threshold value of windspeed at a 7 meter anemometer
height (u.7) to 11.32 m/s. This value of u,7 was calculated assuming a corrected threshold
friction velocity (u,*) of 0.625 m/s and a typical roughness height (ZD) for open terrain of
0.5 cm.
Refined Procedures'.
For estimating emissions from wind erosion, either of two emission flux (g/nf-s)
models are used depending on the erodability classification of the site surface material.
These two models are the: 1) "unlimited reservoir," and 2) "limited reservoir." Each site
surface of homogeneous contaminant concentration must be placed into one of these two
classifications. The following decision flow chart (Figure 1) is used to determine: 1)
whether no wind erosion potential exists, or 2) which of the two emission flux models is
applicable for site conditions. The instructions within each box of the flow chart are
detailed in the list of steps below.
It should be noted that the two emission flux models (Equations 33 and 35)
represent average annual emissions. This assumes continuous emissions over time. In
actuality, emissions do not occur except during periods when the windspeed meets or
exceeds the threshold friction velocity for the given soil aggregate size. A continuous
average emission flux is calculated to account for a continuous exposure interval (i.e.,
hours/day x days/year x years).
Detailed Steps for Flow Chart:
No. 1 Continuous Vegetation?
Continuous vegetation means "unbroken" vegetation covering 100 percent of
the emission source area to be analyzed.
44
-------
No.4a
No.4d
No.4e
Figure 1. Wind erosion model decision flow chart.
45
-------
No. 2 Is Crust Present?
Crusted surfaces are regarded as having a "limited reservoir" of erodible
particles. Check for crust thickness/strength during the site inspection.
No. 3a Determine Threshold Friction Velocity
Threshold friction velocity (u't) is that wind velocity at which erodible surface
particles are suspended. To determine u't, the mode of the surface aggregate
size distribution must be determined. The distribution mode is the aggregate
size containing the highest percentage of material from a representative
sample. This can be determined with a field sieving procedure as follows:
1. Prepare a nest of sieves with the following openings: 4 mm, 2 mm, 1
mm, 0.5 mm, and 0.25 mm. Place a collector pan below the bottom
sieve (0.25 mm opening).
2. Collect a sample representing the surface layer of loose particles
(approximately 1 cm in depth for an uncrusted surface), removing any
objects larger than about 1 cm in average physical diameter (nonerodible
material). The area to be sampled should not be less than 30 cm x 30
cm.
3. Pour the sample into the top sieve (4 mm opening), and place a lid on
top.
4. Rotate the covered sieve/pan by hand using broad sweeping arm
motions in the horizontal plane. Complete 20 rotations at a speed just
necessary to achieve some relative horizontal motion between the sieve
and the particles.
5. Inspect the relative quantities of catch within each sieve and determine
where the mode in the aggregate size distribution lies, i.e., between the
opening size of the sieve with the largest catch and the-opening size of
the next largest sieve (e.g., 0.375 mm lies between the 0.5 mm and the
0.25 mm sieve).
With the aggregate size distribution mode, determine the threshold friction velocity
(u't) in cm/s from the relationship in Figure 1a. A conservative default is 50 cm/s.
Note: Soil particle size distribution determined by laboratory methods cannot
be used to determine the in-situ soil aggregate size distribution.
46
-------
\
o
o
§
o
ID
O
CXI
co
CO
O
E
E,
o
•o
o
}
CO
o>
«-•
CO
o>
2
DJ
O)
V)
0°
CM
c!
c
jo
J3
i
w
O)
£
O)
D)
re
(0
I
0>
o
_o
0)
c
o
2
o
£
(0
£
re
T—
s
3
O)
il
o o o
O O o
O 00 to
o
o
o o o
o oo 10
o
CM
\n 'A;pO|aA
47
-------
No. 3b Correct for Nonerodible Elements
Mark off a representative site area 1m x 1m and determine the fraction of total
area, as viewed from directly overhead, that is occupied by nonerodible
elements (e.g., stones, clumps of grass, etc.). Nonerodible elements can be
said to exceed 1 cm in diameter. Correct the overhead fractional area of
nonerodible elements to the equivalent projected frontal area. An example
would be that a spherical stone with an area of 10 cm2 as viewed from
overhead but half-buried in the soil, would have a frontal projected area of 5
cm2. Determine the ratio of the frontal projected area of nonerodible elements
to the total overhead area of the erodible soil. This ratio (LJ is used with the
relationship shown in Figure 1 b to determine the appropriate correction factor.
Multiply u't by the correction factor to obtain the corrected threshold friction
velocity (14*).
Note: If data for determining !_<. is not available, a conservative default
value of 0.003 may be used for nonsmooth soil surfaces. This
results in a correction factor of approximately 1.25.
No. 3c Is Corrected Threshold Friction Velocity >75 cm/s?
75 cm/s is an empirical number determined through observation of actual soil
types.
No. 4a Determine Crust Thickness/Strength
and
No. 4b Crust Easily Crumbled?
If the crust thickness is <0.6 cm or if the crust can be easily crumbled by
hand pressure it exhibits a potential for wind erosion.
No. 4c Loose Material Present?
Determine if there is loose erodible material above any hardened crust.
No. 4d Estimate Size Distribution Mode and Threshold Friction Velocity
Estimate the aggregate size distribution mode of the loose material above the
hardened crust and determine the threshold friction velocity (u't) (Step 3a).
No. 4e Correct for Nonerodible Elements (ut*)
(Step 3b)
48
-------
\
p
o
o
o
o
_0
SI
c
o
2
o
CO
to
a>
o
c
.a
2
2
O)
O
o
o
C,
uo.ipuj
Ajpo|9A
p|oqs9JL|i
49
-------
Using either the "unlimited reservoir" or the "limited reservoir" model as determined
from Figure 1, calculate an annual average emission flux (g/nnf-s) for each contaminant
found in the erodible surface material.
A. Using the "unlimited reservoir" model
Emission flux for inhalable particles _£ 10//m (PM10):
0.036 (1-V)
u.
F(x
1
3600
(33)
where
V
= PM10 annual average emission flux of component i, g/rtf-s
= Fraction of contaminated surface with continuous vegetative cover
(equals 0 for bare soil)
[u] = Mean annual windspeed, m/s (from local climatological data)
u<7 = Equivalent threshold value of windspeed at a 7 m anemometer height,
m/s (Equation 34)
Q = Fraction by weight of component i from bulk samples of surface
material
F(x) = Function obtained from the relationship in Figure 2
(x=0.886 U(/[u], dimensionless ratio).
Calculation of the equivalent threshold value of windspeed at a 7 m
anemometer height (u,7):
ut
(Ml
/100
(34)
where u,7 = Equivalent threshold value of windspeed at a 7 m anemometer
height, m/s
L^* = Threshold friction velocity corrected for nonerodible elements, cm/s
(Step 3b)
50
-------
e.
1.5
'x'
LL
C
o
+-"
g 1
u.
0.5
n
i^
••
••
»
«=
•^
V
^
- F(x) tends to 1.91
. as x tends to zerc
X = 0.886 u7/ [u]
s
1
s
\
V
>
v
\
\
\
l\
\
\
k
\l
\
\
1
\
v
0 0.5 1 1.5 2
X
NOTE: lfx>2,
= 0.18(8x3 + 12x)e'x
Figure 2. Function curve used in "unlimited reservoir" model.
51
-------
2D = Surface roughness height, cm (default = 0.5 cm for flat terrain).
See also the Note after Equation 37.
B. Using the "limited reservoir" model.
Emission flux for inhalable particles _< 10//m (PM10):
F, = 0.5
c,
(35)
' 31,536,000 s/yr
where Fj = PM10 annual average emission flux of component i, g/nf-s
0.5 = Particle size multiplier for PM10
N = Number of surface material disturbances per year (default = 12/yr for
sites with no activity)
Pi = Erosion potential of component i corresponding to the observed or
probable highest windspeed for the ith period between disturbances,
9/m2
Q = Fraction by weight of component i from bulk samples of surface
material
V = Fraction of continuous vegetative cover
PE = Thornthwaite's Precipitation - Evaporation Index used as a measure
of soil moisture content (Figure 3).
Note: Values of the PE from Figure 3 which are < 50 should be set
equal to 50.
Calculation of the erosion potential (P():
P, = 58 (u* - i/;)2 + 25 (i/' - i/;), P, = 0 for u* < u*t (36)
52
-------
v>
O
Q
to
_0
OJ
£
o
•s
E
O
o
*_*
n
4-1
07
L.
£
oT
X
O
TJ
C
C3
>_
O
Q.
re
LU
,0.
o
*^
to
x:
«-*
c
o
£
CO
u.
-------
where u* = Windspeed at surface roughness height 2D of the observed or
probable highest windspeed for the ith period between disturbances,
m/s (Equation 37)
14* = Threshold friction velocity corrected for nonerodible elements, m/s
(Step 3b)
Calculation of u*:
U* = (37)
where u = Observed or probable windspeed at roughness height z^ m/s
U, = Observed or probable windspeed at anemometer height z, m/s
z = Anemometer height, cm
Zj = Surface roughness height, cm
Note: Equation 35 calculates the total yearly erosion potential (g/m2) as a
function of the number of disturbances (N) and the highest windspeed
between disturbances. The erosion potentials of all events are summed
and divided by a one year averaging time in seconds. This calculation
yields an average emission flux (including periods of zero emissions) to
account for continuous exposure. Therefore, the calculated annual
average emission flux and subsequent ambient air concentration are not
appropriate for exposure averaging times less than one year. Actual
short-term concentrations (e.g., minutes to hours) will be considerably
higher.
Equation 35 may also be used for sites with an unlimited reservoir of
erodible particles by assuming N = number of hourly meteorological
windspeed observations per year where the resulting windspeed at the
surface roughness height (Equation 37) is greater than or equal to the
value of the corrected threshold friction velocity
Reference for Equation No. 33: Rapid Assessment of Exposure to Particulate
Emissions from Surface Contamination Sites, Sections 1 - 4.1.2. Office of Health
and Environmental Assessment, Washington, D.C. EPA-600/8-85/002, NTIS PB85-
192219, February 1985.
54
-------
Reference for Equation Nos. 35 and 36: Adapted from Compilation of Air Pollutant
Emission Factors, Volume I: Stationary Point and Area Sources, and Supplements
(AP-42). Office of Air Quality Planning and Standards, Research Triangle Park,
N.C., 1985.
55
-------
STEP IV. ESTIMATE AMBIENT AIR CONCENTRATIONS AT RECEPTOR
LOCATIONS OF INTEREST
Background:
Once emission rates have been calculated for all sources, atmospheric dispersion
models are used to estimate exposure point ambient air concentrations at receptors of
interest. The following dispersion models have been determined to yield the best
estimates of long-term average concentrations given the types of emission sources and
source/receptor geometry typically encountered at Superfund sites under baseline or
undisturbed conditions.
At the time of publication of this document, the screening-level and refined models
specified herein for area sources of emissions are regulatory models. Testing has
confirmed that for area source emissions, these algorithms produce better concentration
estimates for near-field receptors (i.e., within one side-length of the area source). In
addition, these algorithms also allow on-site receptors, critical for most future land-use
scenarios (e.g., future residential or commercial/industrial receptors within the area of
contamination).
Screening-Level Procedures:
The screening-level procedures use the TSCREEN dispersion model (dated 95260)
amended with the new SCREENS model area source algorithm. Due to these revisions
in the TSCREEN model, ambient air impacts for scenarios that use the new revised area
source algorithm are different than those obtained from the previous version of TSCREEN
(dated 94133). The TSCREEN files and associated documentation can be obtained from
the U.S. EPA Office of Air Quality Planning and Standards (OAQPS) Technology Transfer
Network (TTN). See page 42 for details on how to access the OAQPS TTN.
The TSCREEN files can be found within the Support Center for Regulatory Air
Models (SCRAM) section of the TTN. All executable files and associated document files
should be downloaded and uncompressed as necessary using the file PKUNZIP.EXE
which is also available on the TTN.
56
-------
Current Land-Use:
Estimates of risk for current land-use involve estimating long-term average exposure
point air concentrations at contemporaneous receptors. Such receptors might include
actual residences, commercial/industrial properties or other types of receptors near the
site. When more than one receptor exists it may be necessary to estimate concentrations
at each receptor. These concentrations along with the respective exposure assumptions
will determine which receptor controls the risk.
1. On a site plot plan or other map drawn to scale, divide the site into the square
exposure areas (EAs) previously determined for sampling and for emission
estimation.
2. Locate on the map used in Step 1, the location of each potential receptor.
3. Measure and record the distance from the center of each EA to the first
receptor.
4. Use the TSCREEN model to estimate the combined air concentration
contributed by each EA at the first receptor. Run the model for area sources
with the following specifications:
A. Set the "Initial Form of Release" as a "Superfund Release Type."
B. Set the "Superfund Release Type" to "Soil Excavation."
C. Answer Yes to "Is the Emission Rate (QJ known (Y/N)."
D. Enter as the emission rate, the area of one EA in square meters (e.g., an
EA of one-half acre is 2023.5 rrf).
E. Under "Release Parameter," enter as the "Area of the Emitting Source
(A)" the area of one EA in square meters (i.e., the same area as in Step
4.D. above).
F. Specify the "Urban/Rural Classification."
G. Under "Fenceline Distance," enter the shortest distance between the
center of each EA and the receptor as determined in Step 3.
H. "Receptor Height above Ground (Zr)" should remain zero.
I. Answer Yes to "Do you have specific locations where you would like
pollutant concentrations to be calculated (Y/N)."
57
-------
J. Enter the remaining EA-to-receptor distances as determined in Step 3.
If more than 30 distances are required, two or more model runs will be
necessary.
K. After the model has run, record the estimated concentrations for each
discrete receptor distance entered. These are normalized concentrations
based on an emission flux of 1 g/nf-s.
5. For each contaminant, multiply the actual screening-level emission flux
estimates for each EA by the corresponding normalized concentration as
determined in Step 4.K. The product is the corrected concentration
contributed by each EA.
6. For each contaminant, sum the corrected concentrations and multiply each
sum by 0.08 to convert the one-hour TSCREEN estimate to an annual average
concentration.
7. Repeat steps 4 through 6 for each additional receptor.
8. For each contaminant, the highest receptor annual average concentration is
used as the screening-level exposure point concentration for current land-use.
Future Land-Use:
Estimates of risk for future land-use involve estimating long-term average exposure
point air concentrations at potential future receptor locations. Typically, the potential
future receptor that will experience the highest long-term air concentration contributed by
the site will be located directly above the center of the exposure area with the highest
emission flux of each contaminant. In addition to the emissions from this exposure area,
the contribution of the emissions from the remaining exposure areas at this location must
also be considered.
1. On a site plot plan or other map drawn to scale, divide the site into the square
exposure areas (EAs) previously determined for sampling and emission
estimation.
2. Locate on the map used in Step 1, the center of the EA with the highest
emission flux of the first contaminant.
3. Measure and record the distance from the center of each remaining EA to the
center of the EA determined in Step 2.
58
-------
4. Use the TSCREEN model to estimate the combined air concentration as
contributed by each exposure area at the center of the exposure area
determined in Step 2. Run the model for area sources with the following
specifications:
A. Duplicate the model inputs previously specified for the current land-use
scenario (Step 4, A-F).
B. Under "Fenceline Distance," enter 1 meter. This entry is used to estimate
the concentration at the center of the EA with the highest emission flux.
C. "Receptor Height above Ground (Z^)" should remain zero.
D. Answer Yes to "Do you have specific locations where you would like
pollutant concentrations to be calculated (Y/N)."
E. Enter the remaining distances from the center of each EA to the center
of the EA determined in Step 2.
F. After the model has run, record the estimated concentrations for each
discrete receptor distance entered including the 1 meter distance. These
are normalized concentrations based on an emission flux of 1 g/rrf-s.
5. Perform Steps 5 and 6 as specified for the current land-use scenario.
6. Repeat Steps 4 and 5 above for each contaminant.
7. The combined annual average air concentration estimated at the center of the
EA with the highest emission flux of each contaminant is used as the
screening-level exposure point concentration for future land-use.
Note: For emissions from vented storage tanks, use the "Stacks, Vents,
Conventional Point Sources" gaseous release type as the "Initial
Form of Release." Set the exit velocity (Exit V) equal to 0.01 m/s;
set the emission rate (QJ equal to 1 g/s.
For emissions from vented landfills, use the "Municipal Solid Waste
Landfills" gaseous release type as the "Initial Form of Release."
Set the emission rate (QJ equal to 1 g/s; set the release height
above ground equal to the height of the vents.
59
-------
References for Step IV. Screening-Level Procedures: Adapted from A Tiered
Modeling Approach for Assessing the Risks Due to Sources of Hazardous Air
Pollutants. U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards, Research Triangle Park, North Carolina, EPA-450/4-92-001, 1992.
User's Guide to TSCREEN: A Model for Screening Toxic Air Pollutant
Concentrations. U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, North Carolina, EPA-454/B-94-
023, 1994.
Refined Procedures:
Refined dispersion modeling procedures use refined dispersion models to estimate
exposure point concentrations at receptors of interest. Refined dispersion models use
meteorological data representative of site conditions and are capable of simultaneously
estimating contributions from multiple sources of emissions with different emission rates.
For area sources of emissions, the refined procedures specify the use of the Industrial
Source Complex Model Version 3 in the Snort-Term mode (ISCST3). Similar to the
algorithms incorporated into the new TSCREEN model used for screening-level analysis,
the ISCST3 model algorithms have been shown to produce better concentration estimates
for near-field receptors and also allow on-site receptors.
The ISCST3 model can be used with five years of one-hour meteorological data from
the nearest National Weather Service (NWS) station which is representative of site
conditions; or may be used with one year of quality-assured onsite one-hour
meteorological data. Both the ISCST3 model and NWS meteorological data are available
on the Support Center for Regulatory Air Models (SCRAM) bulletin board of the Office of
Air Quality Planning and Standards (OAQPS) Technology Transfer Network (TTN). See
page 42 for details on how to access the OAQPS TTN. Setup and execution of this
model must follow the procedures specified in the references for this section.
As with the screening-level procedures, the refined procedures are used to estimate
exposure point air concentrations for both current and future land-use scenarios. Annual
average air concentrations estimated using the refined emission and dispersion modeling
procedures are used as exposure point concentrations in the baseline risk assessment.
References for Step IV. Refined Procedures: Guideline on Air Quality Models
(Revised). U.S. Environmental Protection Agency, Office of Air Quality Planning
Standards, Research Triangle Park, North Carolina, EPA-450/2-78-027R, 1988.
User's Guide for the Industrial Source Complex (ISC3) Dispersion Models,
Volumes 1 and 2. U.S. Environmental Protection Agency, Office of Air Quality
Planning Standards, Research Triangle Park, North Carolina, EPA-454/B-95-003a
and b, 1995.
60
-------
STEP V. ORGANIZE EXPOSURE POINT CONCENTRATIONS FOR INPUT TO THE
BASELINE RISK ASSESSMENT
1. In tabular form, list the long-term (annual average) ambient air concentrations of
each potential airborne contaminant derived from Step IV. Include these air
concentrations in the summary of exposure point concentrations for all pathways.
Note: For baseline conditions, long-term concentration averages representative
of the reasonable maximum exposure scenario are most applicable for
the baseline risk assessment. If emission potentials are significant,
however, short-term or acute concentration estimates (e.g., 1-h, 3-h, 8-h,
or 24-h average) may be required.
2. List all variables and assumptions used in the emission and dispersion modeling
analyses and discuss the uncertainty associated with each as well as how this
uncertainty may affect the final estimates.
Reference for Step V: Risk Assessment Guidance for Superfund; Volume I, Human
Health Evaluation Manual (Part A), Interim Final. U.S. Environmental Protection
Agency, Office of Emergency and Remedial Response, Washington, D.C.,
EPA/540/1-89/002, December 1989.
61
-------
APPENDIX A
RECHARGE ESTIMATES BY HYDROGEOLOGIC SETTINGS
A-1
-------
Recharge Estimates for DRASTIC Hydrogeologic Settings
fromCNewell et a!., Hydrogeologic Database for Ground Water Modeling, 1989)
Mountain slope and mountain flank settings are not included.
Recharge (m/yr)
Region Setting Min. Max. Avg.
Western Mountain Ranges
Alluvial Mountain Valleys Facing East 0.00 0.05 0.03
Alluvial Mountain Valleys Facing West 0.05 0.10 0.08
Glacial Mountain Valleys 0.10 0.18 0.14
Wide Alluvial Valleys Facing East 0.05 0.10 0.08
Wide Alluvial Valleys Facing West 0.10 0.18 0.14
Coastal Beaches 0.25 0.38 0.32
Swamp/Marsh 0.10 0.18 0.14
Mud Flows 0.18 0.25 0.22
Regional Average: 0.14
Alluvial Basins
Alluvial Mountain Valleys 0.00 0.05 0.03
Alluvial Fans 0.00 0.05 0.03
Alluvial Basins with Internal Drainage 0.00 0.05 0.03
Playa Lakes 0.00 0.05 0.03
Swamp/Marsh 0.05 0.10 0.08
Coastal Lowlands 0.25 0.38 0.32
River Alluvium With Overbank Deposits 0.18 0.25 0.22
River Alluvium Without Overbank Deposits 0.25 0.38 0.32
Mud Flows 0.18 0.25 0.22
Alternating Sandstone and Shale 0.18 0.25 0.22
Continental Deposits 0.00 0.05 0.03
Regional Average: 0.14
Columbia Lava Plateau
Alluvial Mountain Valleys 0.05 0.10 0.08
Lava Flows: Hydraulically Connected 0.05 0.10 0.08
Lava Flows: Not Hydraulically Connected 0.05 0.10 0.08
Alluvial Fans 0.05 0.10 0.08
Swamp/Marsh 0.00 0.05 0.03
River Alluvium 0.10 0.18 0.14
Regional Average: 0.08
Colorado Plateau and Wyoming Basin
Resistant Ridges 0.00 0.05 0.03
Consolidated Sedimentary Rocks 0.00 0.05 0.03
River Alluvium 0.10 0.18 0.14
Alluvium and Dune Sand 0.00 0.05 0.03
Swamp/Marsh 0.18 0.25 0.22
Regional Average: 0.09
High Plains
Ogalalla 0.00 0.05 0.03
Alluvium 0.00 0.05 0.03
Sand Dunes 0.00 0.05 0.03
Playa Lakes 0.00 0.05 0.03
Braided River Deposits 0.10 0.18 0.14
Swamp/Marsh 0.18 0.25 0.22
River Alluvium with Overbank Deposits 0.00 0.05 0.03
River Alluvium without Overbank Deposits 0.00 0.05 0.03
Alternating Sandstone, Limestone, and
Shale Sequences 0.00 0.05 0.03
Regional Average: 0.06
A-2
-------
Recharge Estimates for DRASTIC Hydrogeologic Settings
from(Newell et aL Hydrogeologic Database for Ground Water Modeling, 1989)
Mountain slope and mountain flank settings are not included.
Recharge (m/yr)
Region Setting Min. Max. Avg.
Non-Glaciated Central Region
Alluvial Mountain Valleys 0.10 0.18 0.14
Alternating Beds of Sandstone,
Limestone, or Shale Under Thin Soil 0.10 0.18 0.14
Alternating Beds of Sandstone,
Limestone, or Shale Under Deep Regolith 0.10 0.18 0.14
Solution Limestone 0.25 0.38 0.32
River Alluvium with Overbank Deposits 0.18 0.25 0.22
River Alluvium without Overbank Deposits 0.18 0.25 0.22
Braided River Deposits 0.10 0.18 0.14
Triassic Basins 0.10 0.18 0.14
Swamp/Marsh 0.10 0.18 0.14
Blocks 0.00 • 0.05 0.03
Unconsolidated and Semi-Consolidated
Aquifers 0.00 0.05 0.03
Regional Average : 0.15
Glaciated Central Region
Till Over Bedded Sedimentary Rock 0.10 0.18 0.14
Till Over Outwash 0.10 0.18 0.14
Till Over Solution Limestone 0.10 0.18 0.14
Till Over Sandstone 0.10 0.18 0.14
Till Over Shale 0.10 0.18 0.14
Outwash 0.18 0.25 0.22
Rock 0.25 0.38 0.32
Outwash Over Solution Limestone 0.25 0.38 0.32
Moraine 0.18 0.25 0.22
Buried Valley 0.18 0.25 0.22
River Alluvium with Overbank Deposits 0.10 0.18 0.14
River Alluvium without Overbank Deposits 0.25 0.38 0.32
Glacial Lake Deposits 0.10 0.18 0.14
Thin Till Over Bedded Sedimentary Rock 0.18 0.25 0.22
Beaches, Beach Ridges, and Sand Dunes 0.25 0.38 0.32
Swamp/Marsh 0.10 0.18 0.14
Regional Average: 0.20
Piedmont Blue Ridge Region
Alluvial Mountain Valleys 0.18 0.25 0.22
Regolith 0.10 0.18 0.14
River Alluvium 0.18 0.25 0.22
Mountain Crests 0.00 0.05 0.03
Swamp/Marsh 0.10 0.18 0.14
Regional Average : 0.15
Northeast and Superior Uplands
Alluvial Mountain Valleys 0.18 0.25 0.22
Glacial Till Over Crystalline Bedrock 0.18 0.25 0.22
Glacial Till Over Outwash 0.18 0.25 0.22
Outwash 0.25 0.38 0.32
A-3
-------
Recharge Estimates for DRASTIC Hydrogeologic Settings
fromCNeweil et al., Hydrogeologic Database for Ground Water Modeling. 1989)
Mountain slope and mountain flank settings are not included.
Recharge (m/yr)
Region Setting Min. Max. Avg.
Moraine 0.18 0.25 0.22
River Alluvium with Overbank Deposits 0.18 0.25 0.22
River Alluvium without Overbank Deposits 0.25 0.38 0.32
Swamp/Marsh 0.10 0.18 0.14
Bedrock Uplands 0.10 0.18 0.14
Glacial Lakes/Glacial Marine Deposits 0.10 0.18 0.14
Beaches, Beach Ridges and Sand Dunes 0.25 0.38 0.32
Regional Average: 0.22
Atlantic and Gulf Coast
Regional Aquifers 0.00 0.05 0.03
Unconsolidated and Semi-Consolidated
Shallow Surfacial Aquifers 0.25 0.38 0.32
River Alluvium with Overbank Deposits 0.18 0.25 0.22
River Alluvium without Overbank Deposits 0.25 0.38 0.32
Swamp 0.25 0.38 0.32
Regional Average: 0.24
Southeast Coastal Plain
Solution Limestone and Shallow Surfacial
Aquifers 0.25 0.38 0.32
Coastal Deposits 0.25 0.38 0.32
Swamp 0.25 0.38 0.32
Beaches and Bars 0.25 0.38 0.32
Regional Average: 0.32
A-4
-------
A-5
-------
APPENDIX B
EXAMPLE CALCULATIONS
B-1
-------
EXAMPLE 1
Calculate the Soil Saturation Concentration (Csat) of Benzene
Soil/Site Properties:
0 Site is located in Maine.
0 Hydrogeologic setting is "outwash."
0 Soil type is "sandy loam."
Soil dry bulk density (ff) = 1.5 kg/L.
Soil particle density (os) = 2.65 kg/L (default).
0 Fraction of organic carbon in soil (tJ = 0.006 (default).
Chemical Properties of Benzene:
0 Henry's law constant (HJ = 5.4 E-03 atm-nf/mol.
0 Solubility in water (S,) = 1,780 mg/L
0 Organic carbon partition coefficient (K^,c) = 57 L/kg.
A. Estimate average long-term volumetric soil moisture content (8W) using Equation 2:
0, = 1 -B//os (Equation 1 legend)
0, = 1 - 1.5/2.65
0, = 0.434
I = 0.32 m/yr (Appendix A)
K. = 230 m/yr (Table 1)
1/(2b+3) = 0.080 (Table 1)
Therefore, 0W = 0.434 (0.32/230)0'080
0U, = 0.256
'w
B. Calculate the soil saturation concentration using Equation 1:
B-2
-------
where K^ = l^c x 4C (Equation 3)
K,i =57x0.006
l^j = 0.342
9a = 0, - 0W (Equation 1 legend)
0a = Q.434 - 0.256
0a = 0.178
H|' = H| x 41 (Equation 1 legend)
Hi' =5.4E-03x41
If = 0.2214
Therefore,
1780
Csa,,. = lfu" (0.342 x 1.5 + 0.256 + 0.2214 x 0.178)
1.5
Csati = 959 mg/kg
B-3
-------
EXAMPLE 2
Calculate the Average Emission Flux (F,) of Benzene From Surface Soils When
NAPL is Present
Site, soil, and chemical properties are the same as in Example 1.
Also:
0 Diffusion coefficient of benzene in air (Dai) = 7,517 crrf/d.
0 Diffusion coefficient of benzene in water (D^) = 0.847 crrf/d.
0 Exposure averaging period (r) = 30 yr = 10,950 days.
0 Mole fraction of benzene in the residual mixture (X,) = 0.02.
0 Pure component vapor pressure of benzene (P,) = 95.2 mm Hg at 25°C.
0 Molecular weight of benzene (MVX) = 78 g/mol.
0 Average in situ soil temperature (T) = 52° F = 284.11°K.
0 Initial soil concentration (Coi) = 5,000 mg/kg = 0.005 g/g.
A. Calculate the equilibrium vapor concentration of benzene (Q,_eq) when NAPL is
present using Equation 6:
X, P, MW,
v'eq ~ RT
c = 0.02 x 95.2 x 78
v'eq 62,361 x 284.11
Cveg = 8.38 E-06 g/cm3-vapor
B-4
-------
B. Calculate the maximum effective diffusion coefficient (Dej) when NAPL is present
using Equation 7:
e
10/3
/"n 10/3
°iv
-, [0.17810/3 T_17)
}e/ = 7,517 +
0.4342
0.2214
0.25610/3
0.4342
0.847
De/ = 127.56 cm2/day
C. Calculate the average emission flux from surface soils using Equation 5:
? n n \1/2
"
-------
EXAMPLE 3
Calculate the Average Emission Flux (Fr) of Benzene from Subsurface Soils When
NAPL is Present
Site, soil, and chemical properties are the same as given in Examples 1 and 2.
Also:
0 Distance from soil surface to top of contamination (dj = 100 cm.
A. Calculate the emission flux from subsurface soils using Equation 8:
^ C0i/
T
L 2 C,w D,, r]
\ " °°< \
1/2
-d
uc
0.116
1.5x 0.005]
10,950 J
hoo2 4
2>
' 8
38 I
E-(
1
D6
5
X
X
1 27.56 x
0.005
10
,950
1/2
0.116
= 1.16E-06g//r72-s
B-6
-------
EXAMPLE 4
Estimate the Diffusion Coefficients in Air (Da,) and in Water (Dwi) of Toluene
Given:
0 Chemical structure of toluene (CH shown below.
A.
0 Molecular weight (MVX) of toluene = 92.14 g/mol.
0 Absolute pressure (Pab) = 1 atm.
0 Average temperature (T) = 25° C = 298° K.
Estimate the diffusion coefficient in air using Equation 9:
0.001 r1-75
1
KE v> )1'3 + (E
""• MW* (8.64 x 10<) sec/rfay
0.001 x 298 1-75
D.
\
92.14 28.8
a.1
(20.1 )1
(8.64 x 104)
B-7
-------
DaJ = 7,151 cm2 /day
where: IV calculated as:
C (7x16.5) = 115.50
H (8x1.98) = +15.84
131.34
Aromatic ring = -20.20
111.14
B. Estimate the diffusion coefficient in water using Equation 10:
DWJ - ?;?6 x 10-5. (8-64 x 1°4 sec/^
From Table 2, /7W = 0.8904 at 25° C
From Table 3, V'B is calculated as:
C (7 x 14.8) = 103.6
H (8 x 3.7) = +29.6
133.2
6-Membered ring = -15.0
118.2
Therefore:
D . = 13"26 x 10'5 (8.64 x
"'' 0.8904114 (118.2)0589
Dwl = 0.7866 cm2/day
B-8
-------
EXAMPLE 5
Calculate the Average Maximum Emission Flux (F,) of Benzene from Surface Soils
When NAPL is Not Present (Screening-Level Procedures)
Site and soil conditions are the same as in Examples 1 and 2. Chemical properties are
also the same as in Examples 1 and 2.
Also:
0 Initial soil concentration (Coi) = 50 mg/kg = 5.0 E-05 g/g.
A. Calculate the apparent diffusion coefficient (DAi) using Equation 12:
DAJ - eT D., Hi + el013 D
D = [(0.1781073 x 7,517 x 0.2214 + 0.25610/3 x 0.847)/0.4342]
(1.5 x 0.342 + 0.256 + 0.178 x 0.2214)
DAi = 34.93
B. Calculate the screening-level average maximum emission flux using Equation 11:
(4 a.V*
\—^\ 0.116
V 7C T j
- / A y 04. 00 \1/2
F:= 1.5X5.0E-5 4 x 3^6— 0.116
13.1416 x 10,950]
F, = 5.54 E-07 g/m2-s
B-9
-------
EXAMPLE 6
Calculate the Average Emission Flux (F,) of 1,2,4-Trichlorobenzene From Surface
Soils When NAPL is Not Present (Refined Procedures)
Given:
0 Initial soil concentration (Col) = 50 mg/lkg
0 Area of contamination (A0) = 1,000 m2
0 Depth to the bottom of contamination (LJ = 300 cm
0 Soil dry bulk density (ft) = 1.5 g/crrf-soil
0 Soil water-filled porosity (0W) = 0.256 cm3/cm3
0 Soil air-filled porosity (0a) = 0.178 cm3 /cm3
0 Soil total porosity (9,) = 0.434 cm3/cm3
0 Soil organic carbon fraction (foc) = 0.006
0 Organic carbon partition coefficient (H^,c) = 1,540 cm3/g
0 Diffusion coefficient in air (Da) = 2,592 crrf/d
0 Diffusion coefficient in water (D,,,) = 0.711 crrf/d
Henry's law constant (H1) = 0.0582
0 Exposure averaging period (r) = 30 yr = 10,950 days
0 Number of time-steps (n) = 100
0 Time-zero fa) = 0.25 days.
A. Calculate the time-step interval for ^ to ^:
h = r/n (Equation 14 legend)
h = 10,950/100
h = 109.5 days
B-10
-------
B. Calculate the emission flux of 1,2,4-Trichlorobenzene for 100 time-steps using
Equation 15:
DAJ
rr t
1/2
1 pxn
[ Ll |
4D,,f
0.116
Emission flux at time-zero (tj:
F = 1.5x 5.0 E-05
0.184 1 1/2
3.1416 x 0.25
1-exp -
3002
4x 0.184 x 0.25
0.116
F,0) = 4.21 E-06 g/m2-s
Emission flux at the last time-step (two):
F/(100) = 1.5x 5.0E-05
0.184
3.1416 x 10,950.25
1-exp -
3002
4 x 0.184 x 10,950.25
0.116
F/(100) = 2.01 E-08 g/m2-s
C. Integrate the emissions of 1,2,4-Trichlorobenzene across the 100 time-steps to
derive the average emission flux using Equation 14:
2F, + 2F2+ ... + 2Fn.1
B-11
-------
D. Compare the total mass lost from volatilization (M^) to the initial mass (MjT) and
compute the final average emission flux using Equations 16, 17, and 18.
Note: These calculations are best performed using a computer spreadsheet
program. The following MICROSOFT EXCEL 5.0 computer printout
shows the solution for this example.
B-12
-------
t-
V)
a:
0
if>
1 ^
S
£
o
CC
o
a
o
8
d
S
§
§
9
L
O
CD
9
D
CO
N
9
S
o
o
N
O
Ol
§
£
a
o
s
0
8
O
S
8
§
9
UJ
q>
o
9
5
1
9
s
O
o
s
o
§
f
CC
o
a
o
8
o
s
§
§
L
Q
*-
9
CO
S
9
s
d
0
8
o
O)
1
§
CC
d
S
0
8
o
If
S
§
LJ
Q
(N
9
M
CO*
s
9
00
o
o
O)
in
o
01
§
*
CC
d
CO
o
8
o
in
(£
8
§
L
O
<"
99
LJ UJ
•- <•
Is
9 9
33
o" c
0 C
s?
0 C
CO 0
SJ
S c
CC C
o c
s p
o c
SS
o c
to <
85
§
in
cp
QJ
o
^-
2.09E-08
-
i
5
s
s
o"
o
!o>
iO
o
§
c
8
CC
°
<5>
0
s
8
§
9
UJ
Q
R
207E-08
-4
3
9
N
O
O
S
o
en
§
i
CO
0
s
0
8
o
s
8
§
9
UJ
Q
O
(N
I
S
s
o
o
s
o
O)
§
tt
CO
d
r!
o
8
o
&
8
§
9
LU
O
(N
O
Total mass
2.05E-08
M
•y
O
.84E-01
o
0
Ol
in
o
o>
s
d
CC
°
S
0
8
S
8
§
8
L
in
Q
8
5
lost from
2.04E-08
§
3
s
o
o
O)
in
o
Ol
§
|
CC
d
S
o
8
o
So
8
§
L
in
O
o
inital mass, 1
volatilization,
2.03E-08
N
O
5
UJ
s
N
0
0
s
§
O)
§
d
CC
d
CO
o
8
o
8
§
§
8
L
in
Q
in
8
I
^
9
i
D
CN
b
.84E-01
o
0
a
N
CN
O)
S
d
CO
d
S
o
8
d
s
"
8
1
8
S
5
s
J
0
(N
0950.25
u
1
C)
o
8
N
3
(N
Ol
s
c
CC
0
s
0
s
So
8
§
L
m
o
m
r--
o
in
3
J
O
CM
S
h
UJ
(N
in
iri
5.83E-08
Calculated average flux =
CO
o
i
o
111
s
in
/erage flux =
fo
to
r
8
5
)-(R7+(2*SUM(R8.R106))+
5
5
X
0)
i1
B
io
8
(D
U
5
§
I
1
1
1
R
o
3
3
o
I
A
S
L
nal average flux =
Spreadsheet equation for f
c
s.
u
£
Ul
c
a
t
j
(0
UJ
C
j
from S urface Soils
e
VI
E
Ul
0
1 ,2,4 -Trichlo robe nzen
f
^
O
O
0
o
o
o
o
o
0
CO*
gl
o
"1
1-
o
0
o
0
0
o
CM
90-300' I id-300 I 80-300 1
s-jiu/B *n|-j uoissjujg
-
-
B-13
-------
EXAMPLE 7
Calculate the Average Emission Flux (Fj) of 1,2,4-Trichlorobenzene From
Subsurface Soils When NAPL is not Present (Refined Procedures)
Soil, chemical properties, and initial soil concentration are the same as in Example 6.
Also:
0 Depth to the top of contamination (d) = 30 cm
0 Thickness of contaminated soil (wc) = 100 cm
0 The exposure averaging period (T) = 30 yr = 10,950 days
0 Number of time-steps (n) = 100
0 Time-zero ft,) = 0.25 days
0 The time-step interval = r/n = 109.5 days.
A. Calculate the emission flux of 1,2,4-Trichlorobenzene for 100 time-steps using
Equation 19:
0,1
DAJ
n t
1/2
exp
[ dl |
4 DAJ t
- exp
(dc + wef~
4 DA/ t \
0.116
Emission flux at time-zero (tj:
= 1.5x 5.0 £-05
°-184
1/2
3.1416 x 0.25
exp
302
4x 0.184 x 0.25
- exp -
(30 + 100)2 '
4 x 0.184 x 0.25
0.116
B-14
-------
F,(0) = Og/m2-s
Emission flux at the last time-step (two):
F,nom = 1.5 x 5.0E-05 f 0-184
'(100) 3.1416 x 10,950.25
x Fexpf- - exp- 0.116
|_ I 4x0.184x10,950.25] I 4 x 0.184 x 10,950.25
F,{100) = 1.55E-08g//772-s
B. Integrate the emissions of 1,2,4-Trichlorobenzene across the 100 time-steps to
derive the average emission flux using Equation 14:
PI = ~r \\
C. Compare the total mass lost from volatilization (lv\Tv) to the initial mass (Mi-T) and
compute the final average emission flux using Equations 16, 17, and 18.
Note: These calculations are best performed using a computer spreadsheet
program. The following MICROSOFT EXCEL 5.0 computer printout
shows the solution for this example.
B-15
-------
LU
s
8
LU
a
9
LU
8
8
6
I
a
§
•o
o
I
T3
C
I
9
LU
S
9
LU
8
P
o:
O)
'w
J2
'6
W
a)
o
m
a
2
E
o
S1
o
(A
§
in
-------
EXAMPLE 8
Calculate the Maximum Emission Flux (Fjmax) of Benzene Using Soil Gas
Concentrations
Given:
0 Measured soil gas concentration (Csgi) = 500 ppmv.
0 Soil air-filled porosity (9a) = 0.178.
0 Soil total porosity (0,) = 0.434.
0 Depth to immediately above the vapor source (d) = 220 cm.
0 Chemical properties of benzene are the same as in Examples 1 and 2.
A. Calculate the vapor-phase diffusion coefficient (D,ev) using Equation 21:
fl10/3
» , 7,517 -
0.4342
Dfv = 127.34 cmzld
B. Convert the measured soil gas concentration from ppmv to g/cm3-vapor using
Equation 22:
MW;
2.404 x 1010
B-17
-------
CVJ = 500
78
2.404 x 1010,
Cvi = 1.62 E-06 g/cm2-vapor
C. Calculate the maximum emission flux using Equation 20:
•\ev
0.116
r-max
1.62 E-06 x 127.34
200
0.116
/-max
- 1.20 E-07 g/m2-s
B-18
-------
EXAMPLE 9
Calculate the Maximum Emission Flux (Fjmax) of Benzene from Subsurface Soils
with Landfill Gas Generation (Screening-Level Procedures)
Giver:
The chemical properties of benzene are the same as in Examples 1 and 2.
Temperature (T) = 25° C = 298° K.
A. Calculate the saturation vapor concentration (Csvi) of benzene using Equation 24:
P, MW,.
CSVJ = RJ.
c _ 95.2 x 78
s'w " 62,361 x 298
Csw = 4.00 E-04 g/cm3-vapor
B. Calculate the maximum emission flux using Equation 23:
F™* = CWi/ Vy 0.116
F™x = 4.00 E-04 x 141 x 0.116
F,max = 6.54 E-03 g/m2-s
B-19
-------
EXAMPLE 10
Calculate the Average Landfill Emission Rate (EJ of Benzene with Landfill Gas
Generation (Refined Procedures)
Given:
A single cell codisposal landfill with a maximum capacity of 75,000 Mg.
Methane generation potential (LJ = 125 nf/Mg (default).
Average annual landfill acceptance rate (R) = 5,000 Mg/yr.
Methane generation rate constant (k) = 0.04/yr (default).
Time since landfill closure (c) = 5 yr (i.e., time-zero or ^ = 5 yr).
Time from initial refuse placement in landfill (t) = 20 yr at time-zero.
Measured vapor-phase benzene concentration (Cvi) = 500 ppmv = 1.62 E-06
g/cm3.
0 Exposure averaging period (r) = 30 yr.
0 Time-step interval (h) = 1 yr.
A. Calculate the emission rate of benzene for 30 one-year time-steps using Equation
27:
E/(0 = 2 L0 R exp(-*c)-exp(-tt) Cvi (3.17 x
Emission rate at time-zero (tj:
Ei(0) = 2x 125 x 5,000{exp(-0.04 x 5)-exp(-0.04 x 20)} x 1.62 E-06 (3.17 x 1CT2)
E/(0) = 2.37 E-02 g/s
Emission rate at the last time-step (t30):
B-20
-------
5/(30) = 2x 125 x 5,000{exp(-0.04 x 35)-exp(-0.04 x 50)} x 1.625-05 (3.17 x 10~2)
E,30) = 7.14 E- 03 g/s
B. Integrate the emissions of benzene across the 30 time-steps to derive the average
emission rate using Equation 26:
% (E0 + 2E, + 252 + ...
5n)
Note: These calculations are best performed using a computer spreadsheet
program. The following MICROSOFT EXCEL 5.0 computer printout shows the
solution for this example.
B-21
-------
Calculation of Average Landfill Emission Rate of Benzene with Landfill Gas Generation
Row
Col.
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
A
Measured
vapor-phase
cone.,
Cv,i
o
(g/cm -vapor)
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
1.62E-06
B
Average
annual
^acceptance
rate,
R
(Mg/yr)
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
c
Methane
generation
potential,
Lo
(m3/Mg)
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
125
Spreadsheet equation for average emission rate =
2.50E-02-]
OfYlFJYj
CT
of 1 50E-02
1
•<5 1 rjQE-02 -
UJ
500E-03
r\(r\F-ff\
I
D
Methane
generation
rate
constant,
k
d/yr)
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
E
Time
since
closure,
c
(yr)
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
F
Time from
initial
refuse
"~ placement,
t
(yr)
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
Calculated average emission rate =
G
Emission
rate,
Ei(t)
(9/s)
2.37E-02
2.28E-02
2.
2.
19E-02
10E-02
2.02E-02
1.94E-02
1 .87E-02
1 .79E-02
1.72E-02
1.65E-02
1.59E-02
1 .53E-02
1.47E-02
1.41E-02
1.35E-02
1 .30E-02
1.
25E-02
1.20E-02
1.
1.
15E-02
11E-02
1.07E-02
1.02E-02
9.84E-03
9.45E-03
9.08E-03
8.72E-03
8.38E-03
8.05E-03
7.74E-03
7.43E-03
7.
1.
14E-03
38E-02
(1/30)*(1/2*(G8 + (2*SUM(G9:G37))+G38))
Landfill Emissions of Benzene
t^
"\
"^^^
^^—
5 10 15 20 25 30 35
Time Since Closure, yr
B-22
-------
EXAMPLE 11
Calculate the Emission Flux (Fj) of Benzene from a Quiescent Surface
Impoundment (Screening-Level Procedures)
Given:
0 Chemical properties of benzene are the same as in Examples 1 and 2.
0 Liquid-phase benzene concentration (Cg) = 800 mg/L = 8.0 E-04 g/cm3.
0 Average impoundment temperature (T) = 13°C = 286° K.
A. Calculate the liquid-phase mass transfer coefficient (^L) of benzene using Equation
31:
(MW^f T\
•vi ~ I IA/>
IL MW, J (298) [ L
/OO\0-5 /ODC\
, = 32 286 (Q OQ2)
4 UsJ (298J'
kjL = 1.23 E- 03 cm/s
B. Calculate the gas-phase mass transfer coefficient (k|G) of benzene using Equation
32:
/ ..,., N0.335
| MWHO\ I T V-005
k = ^ _L //f HZ0\
10 MW, l G 2 '
I ^ Q \0.335/ OQC \1 -005
= I8-} f 286 (Q
IG (7Q) (298)
B-23
-------
k.G = 4.89 E-01 cm/s
C. Calculate the overall mass transfer coefficient (K^) of benzene using Equation 30:
± = J_ + R T
Kf k/L HI kiG
1 1 8.2 E-05 x 286
1.23 £-03 5.4 E-03 x 4.89 E-01
-1 = 821.89
K,
K(. = 1.22 E-03 cm/s
D. Calculate the emission flux of benzene using Equation 29:
F, = K,. CL/ (1 x 104)
F, = 1.22 E-03 x 8.0 E-04 x (1 x 104)
F,. = 9.76 E-03 g/m2-s
B-24
-------
EXAMPLE 12
Calculate the PM10 Annual Average Emission Flux (Fr) of Cadmium Due to Wind
Erosion Using the "Unlimited Reservoir" Model
Given:
The mean annual windspeed ([u]) = 5.2 m/s.
Fraction by weight of cadmium in surface soils (Q) = 5.0 E-04.
Fraction of contaminated surface with continuous vegetation (V) = 0.1.
The aggregate size distribution mode of surface material = 0.3 mm.
Surface roughness height (ZD) = 0.5 cm (default).
A. Estimate the threshold friction velocity (i\') from Figure 1a:
0 From Figure 1a, the value of 14' corresponding to an aggregate size
distribution mode of 0.3 mm = 40 cm/s
B. Correct the value of i\' for nonerodible elements:
0 With no other data, assume the default correction factor of 1.25.
0 Calculate the threshold friction velocity corrected for nonerodible elements (i\*)
as:
u,* = 14' x Correction Factor
Ut* = 40 cm/s x 1.25
u,* = 50 cm/s
C. Calculate the equivalent threshold value of windspeed at a 7 m anemometer height
(U[7) using Equation 34:
ILln
/100
B-25
-------
u=
JO infill
0.4 [0.5JJ
/100
u] = 9.06 mis
D. Estimate the value of F(x) using Figure 2:
From Figure 2: x = 0.886 (U//0-0)
x = 0.886 (9.06/5.2)
x = 1.54
From Figure 2: F(x) = 0.83
E. Calculate the average annual emission flux of cadmium using Equation 33:
0.036 (\-V)
iui piv \ c
7 '(.•* 1 °Y
1
3600
F,=
0.036 x (1 -0.1) x -^- 0.83 x 5.0 E-04
9.06
3600
F, = 7.06E-10g//7?2-s
B-26
-------
EXAMPLE 13
Calculate the PM10 Annual Average Emission Flux (F{) of Cadmium Due to Wind
Erosion Using the "Limited Reservoir" Model
Given:
0 Number of site disturbances per year (N) = 12 (default).
0 Fraction by weight of cadmium in surface soils (Q) = 5.0 E-04.
0 Fraction of contaminated surface with continuous vegetation (V) = 0.1.
0 Site is located on the southeastern coast of North Carolina.
0 The threshold friction velocity corrected for nonerrodible elements (u,*) = 50
cm/s = 0.50 m/s.
0 The fastest windspeed between each disturbance (uj = 14.5 m/s at a 10 m
anemometer height (z = 1000 cm).
0 Surface roughness height (zj = 0.5 cm.
A. Calculate the fastest windspeed between disturbances at the surface roughness
height (u*) using Equation 37:
. uz 0-4
u =
14.5x0.4
u =
In (1000/0.5)
u* = 0.76 mis
B. Calculate the erosion potential (Pi) for each period between disturbances using
Equation 36:
PI = 58 (tv* - i/r*)2 + 25 (u* - u;), P, = 0 for u* < u{
Since u > 14*, P, ^ 0, therefore
B-27
-------
P, = 58 (0.76 - 0.50)2 + 25(0.76 - 0.50)
P, = 10.42 g/m<
B. Estimate the Thornthwaite's Precipitation-Evaporation Index (PE) from Figure 3:
PE = 104
C. Calculate the PM10 annual average emission flux of cadmium using Equation 35:
F. = 0.5
,ti (PE/5Q?
' 31,536,000 s/yr
Because the fastest windspeed between each of the 12 disturbances per year is the same
value, calculate the emission flux for the first disturbance and multiply by 12 to derive the
annual average emission flux:
142 (1-0.1)'
(104/50)2
5.0 £-04
1
31,536,000
F, = 1.725-11 g/m2-s
= 1.725-11 g/m2-s x 12
Fi = 2.06E-10g/m2-s
B-28
-------
EXAMPLE 14
Estimate the Current Land-Use Exposure Point Concentration of Benzene Using
the Screening-Level Dispersion Modeling Procedures
Given:
The emissions of benzene are from surface soils over an urban 2 acre area
of contamination.
Dividing the area of contamination into 4-1/2 acre square exposure areas
(EAs), the predicted emission flux from each EA is:
Source No.
Emission flux, g/nf-s
1
2
3
4
2.60 E-08
4.88 E-07
3.75 E-07
1.25 E-07
The site layout and source-to-receptor distances are shown below.
Offsite Receptor
Run the TSCREEN dispersion model as specified for current land-use in Step IV.
The results are shown below.
B-29
-------
Source
No.
2
4
1
3
Distance to
receptor, m
90
160
170
230
TSCREEN
normalized
concentration,
//g/m3 per g/nf-s
0.5983 E + 07
0.2626 E + 07
0.2393 E + 07
0.1483 E + 07
Predicted
emission flux,
g/rrf-s
4.88 E-07
1.25E-07
2.60 E-08
3.75 E-07
Total
Corrected
concentration,
//g/m3
2.92
0.33
0.06
0.56
3.87
Convert the total 1-h concentration to an annual average:
3.87 jjg/m3 x 0.08 = 0.31
The average long-term exposure point concentration for current land-use is
predicted to be 0.31 //g/m3.
B-30
-------
EXAMPLE 15
Estimate the Future Land-Use Exposure Point Concentration of Benzene Using the
Screening-Level Dispersion Modeling Procedures
Given:
Emissions data are the same as in Example 14.
Because emission source No. 2 has the highest emission flux, the future land-use
receptor is located at the center of source No. 2. The emission contribution of all four
sources will therefore be estimated at this receptor.
Site layout and source-to-receptor distances are shown below.
45 m
Run the TSCREEN dispersion model as specified for future land-use in Step IV. The
results are shown below.
B-31
-------
«*
*-
Source
No.
2
4
1
3
Distance to
receptor,
m
1
45
45
64
TSCREEN normalized
concentration, //g/m3 per
g/nf-s
0.3413 E+08
0.1623 E+08
0.1623 E + 08
0.9520 E + 07
Predicted
emission
flux, g/rrf-s
4.88 E-07
1.25E-07
2.60 E-08
3.75 E-07
Total
Corrected
concentration,
//g/m3
16.66
2.03
0.42
3.57
22.68
Convert the total 1-h concentration to an annual average:
22.68//g/m3 x 0.08 = 1.81 jng/m:
The average long-term exposure point concentration for future land-use is predicted to
be 1.81//g/m3.
B-32
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-451/R-96-001
3 RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
Air/Superfund National Technical Guidance Study Series -
Guideline for Predictive Baseline Emissions Estimation for
Superfund Sites
November 1995
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
Craig S. Mann
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Quality Management, Inc.
Cedar Terrace Office Park, Suite 250
3325 Chapel Hill Boulevard
Durham, North Carolina 27707-2646
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D3-0032
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Region III
841 Chestnut Building
Philadelphia, Pennsylvania 19107
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
EPA Work Assignment Manager: Patricia Flores (215) 597-9134
16. ABSTRACT
This report is a compendium of models (equations) for estimating air emissions and exposure point
concentrations from Superfund sites under baseline or undisturbed conditions. Exposure point
concentrations can be used hi the baseline risk assessment to eliminate air pathways or contaminants
from further consideration (i.e., screening analyses) or can be used in support of monitoring during the
exposure assessment to evaluate risk (i.e., refined analyses). The models contained in this compendium
may not accurately predict emissions for all possible scenarios. Where uncertainty exists, these models
and the default input variables have been designed to overpredict emissions.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Toxic Air Emissions
Superfund
Exposure Assessment
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
, 104
20. SECURITY CLASS (Page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
-------
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicago, IL 60604-3590
------- |