United States
Environmental Protection
Agency
Office of Research and
Development
Washington, D.C. 20460
September 1992
Research and Development
MMSOILS: Multimedia
Contaminant Fate, Transport,
and Exposure Model
Documentation and
User's Manual
92
148
c.2
s
Priniso on Recycled Paper '
. -,!..tVi / ,- - » *«;"
-------
-------
to
iPA/ma/iouui
September 1992
MMSOILS: Multimedia
Contaminant Fate, Transport,
and Exposure Model
Documentation and
User's Manual
**„»
rCM
o
CD
Exposure Assessment Group
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
and
Office of Environmental Processes and Effects Research
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
HEADQUARTERS LIBRARY
ENVSRONMENTAL PROTECTION AGENCY
WASHINGTON, O.C. 20-360
y
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection Agency
policy and approved for publication. Mention of trade names or commercial products does not
constitute official endorsement or recommendation for use. *
11
-------
CONTENTS
TABLES vii
FIGURES .' viu
FOREWORD x
PREFACE xi
ABSTRACT '.' xii
ACKNOWLEDGMENTS xiii
1.0 INTRODUCTION 1-1
1.1 PURPOSE AND SCOPE 1-2
1.2 ORGANIZATION OF THE DOCUMENT 1-5
2.0 APPLICATIONS AND LIMITATIONS OF THE METHODOLOGY 2-1
2.1 APPLICATION OF THE MODEL 2-1
2.1.1 Waste Management Unit Identification 2-2
2.1.2 Define the Transport Pathways 2-3
2.1.3 Define die Exposure Scenarios 2-5
2.2 MODEL LIMITATIONS 2-6
2.3 MODEL PREDICTIONS VERSUS MEASURED DATA 2-7
3.0 WATER AND MASS BALANCE 3-1
3.1 SUMMARY OF WASTE MANAGEMENT UNITS 3-1
3.2 MONTHLY NET RECHARGE 3-2
3.3 WATER BALANCE FOR LANDFILLS AND IMPOUNDMENTS 3-4
3.3.1 Liner and Cap Failures 3-7
3.3.2 Leachate Collection Systems 3-7
3.4 MASS BALANCE 3-8
3.4.1 Impoundment Mass Balance 3-8
3.4.2 Landfill and Waste Pile Mass Balance 3-11
3.5 LEACHATE GENERATION FROM LANDFILLS AND WASTE PILES ..3-12
111
-------
3.5.1 Solubility Limit Approach i . . .
3.5.2 Partitioning Approach . . .
3.5.3 Completely-mixed Reactor Approach
3.5.4 Steady-State Leachate Concentration 3-15
4.0 ATMOSPHERIC PATHWAY 4-1
'£•
4.1 VOLATUJZATION RELEASE MODELS 4-2
4.1.1 Fanner's Equation * Covered Sites, Liquid Phase 4-7
4.1.2 Landfarming Equation - Uncovered Sites, Liquid Phase 4-7
4.1.3 Covered Sites, Adsorbed Phase 4-8
4.1.4 Uncovered Sites, Adsorbed Phase .4-9
4.1.5 Volatilization from a Contaminated Water Body 4-10
4.2 RELEASE MODELS FOR ESTIMATING PARTICIPATE EMISSIONS . . 4-11
4.2.1 Paniculate Emissions due to Wind Erosion •. . 4-13
4.2.2 Participate Emissions due to Vehicle Traffic 4-14
4.2.3 Loading and Unloading Operations 4-14
4.2.4 Soil Spreading Operations 4-15
4.3 ATMOSPHERIC TRANSPORT AND DISPERSION MODEL
4.3.1 Box Model ....".. 4-17
4.3.2 Sector Averaged Gaussian Plume Model 4-17
5.0 GROUNDWATER PATHWAY 5-1
5.1 CALCULATING RECHARGE 5-1
5.2 CALCULATING RUNOFF 5-3
5.3 ESTIMATING EVAPOTRANSPIRATION 5-8
5.3.1 Pan Evaporation Data 5-9
5.3.2 Thornthwaite Method 5-9
5.3.3 Actual Evapotranspiration • • • • 5"10
5.4 TRAVEL THROUGH THE PARTIALLY SATURATED ZONE 5-11
5.5.1 Summary of the Partially Unsaturated Zone Numerical Model 5-13
5.6 MIXING ZONE CALCULATIONS 5-16
IV
-------
5.7 GROUNDWATER FATE AND TRANSPORT MODEL 5-18
6.0 SOIL EROSION PATHWAY 6-1
6.1 EROSION OF CONTAMINATED SOIL 6-1
6.2 OFF-SITE SOIL CONTAMINATION DUE TO EROSION 6-2
6.3 OFF-SITE SOIL CONTAMINATION DUE TO DEPOSITION 6-3
7.0 SURFACE WATER PATHWAY . .7-1
7.1 SOIL EROSION TO A STREAM .7-1
7.2 RUNOFF TO A STREAM 7-4
7.3 GROUNDWATER DISCHARGE TO A STREAM 7-5
7.4 FATE AND TRANSPORT ANALYSIS IN RIVERS 7-5
7.5 CHEMICAL FATE IN A SMALL LAKE 7-7
8.0 FOOD CHAIN BIOACCUMULATION PATHWAY 8-1
8.1 BIOCONCENTRATION IN FISH 8-3
8.2 BIOACCUMULATION IN CATTLE AND DAIRY PRODUCTS 8-7
8.3 BIOACCUMULATION IN PLANTS AND VEGETABLES 8-12
8.3.1 Deposition on Plants 8-13
8.3.2 Plant Uptake from Contaminated Soil 8-14
8,3.3 Plant Uptake by Root Crops 8-16
9.0 HEALTH RISK ASSESSMENT PROCESS 9-1
9.1 RISK EQUATIONS 9-1
9.2 EXPOSURE EQUATIONS 9-3
9.2.1 Time-Weighting of Exposures 9-4
9.2.2 Inhalation Pathway 9-6
9.2.3 Ingestion Pathway 9-6
9.2,4 SoU Contact 9-7
-------
9.3 EXPOSURE PARAMETERS . . 9-
9.3.1 Absorption Assumptions
10.0 USER'S GUIDE 10-1
10.1 SYSTEM REQUIREMENTS AND INSTALLATION PROCEDURES 10-ld
10.2 MODEL DATA FILES 10-4.
10.3 DESCRIPTION OF INPUT DATA FILES 10-6
10.3.1 Chemical Properties Data File (CHEMPRP.###) 10-7
10.3.2 Atmospheric Pathway Data File (ATMSPRM.###) 10-15
10.3.3 Surface Water Pathway Data File (SWTPATH.###) 10-19
10.3.4 Ground-water Transport Pathway Data File
(GWTRANP.###) 10-22
10.3.5 Infiltration Leaching and Recharge Data File
ONFILTR.##^ 10-2?
10.3.6 Food Chain Bioaccumulation Pathway Data
File (FOODCHN.###) 10-30
10.3.7 Control Data File (CONTROL.###) 10-32
10.3.8 Landfill Characteristics Data File (LANDFIL.###)
10.3.9 Surface Impoundment Characteristics Data
File (IMPOUND.###) 10-40
10.3.10 Waste Pile Characteristics Data File (WSTPILE.###) 10-44
10.3.11 Injection Well Characteristics Data File (INJECTW.###) . . 10-44
10.3.12 Tank Release Characteristics Data File fTANKREL.###) . . 10-45
11.0 REFERENCES •)• • !!•!
12.0 LIST OF SYMBOLS
APPENDIX A: VERIFICATION OF PARTIALLY SATURATED ZONE MODEL A-l
APPENDIX B: GROUNDWATER TRANSPORT VERIFICATION B-l
VI
-------
LIST OF TABLES
4-1. Summary of models for estimating volatilization release rate from soils 4-4
4-2. Coefficients used to calculate lateral virtual distances for pasquill-gifford dispersion
rates .4-19
5-1. Runoff curve numbers for hydrologic soil-cover complexes under average conditions
of antecedent moisture 5-5
5-2. Hydrologic soil group descriptions 5-6
5-3. Classification of vegetative covers by their
hydrologic properties 5-7
5-4. Rainfall limits for estimating antecedent moisture conditions 5-8
8-1. Interception fraction for various types of vegetation 8-14
9-1. Inhalation rate 9-9
9-2. Ingestion of drinking water 9-9
9-3. Ingestion of fish, meat, milk, and vegetables 9-10
9-4. Ingestion of soil ..-••--. 9-10
9-5. Dermal contact with soil 9-11
9-6. Additional exposure parameters 9-11
10-1. Default values of percent silt, vehicle weight, vehicle speed and number of wheels 10-17
s
10-2. Hydraulic properties of soils classified by soil texture 10-29
vu
-------
LIST OF FIGURES
1-1. Coupling between source location, environmental transport mechanisms, and human
exposure pathways 1-4
3-1. Flow chart showing sequence of calculations required for groundwater transport ~
depending on type of wmu 3-3
3-2a. Schematic water balance for a layer 3-5
3-2b. Schematic water storage in a layer 3-6
3-3. Release pathways from a WMU 3-9
4-1. Conceptual model of the atmospheric pathway 4-1
4-2. Physical site characteristics and phase of the chemical within the soil column for
volatilization release models 1 and 2 4-5
4-3. Physical site characteristics and phase of the chemical within the soil column for
volatilization release models 3 and 4 4-6
5-1. Conceptual model of contaminant movement in the groundwater pathway 5-
7-1. Conceptual model of surface water pathway 7-2
7-2. Conceptual model of stream contamination due to groundwater inflows 7-3
8-1. Uptake factor for beef versus octanol-water partition coefficient 8-10
8-2. Uptake factor for milk versus octanol-water partition coefficient . 8-11
8-3. Bioconcentration factor (on a dry weight basis) in vegetables versus soil-water partition
coefficient for metals 8-17
8-4. Bioconcentration factor (on a dry weight basis) in vegetables versus octanol-water
partition coefficient for organics 8-18
8-5. Root concentration factors (RCF) versus octanol-water partition coefficient for
organics 8-19 , *"
10-1. Listing of input data files for mmsoils 10-5 \
Vlll
-------
10-2. Variables in the chemical properties data file (CHEMPRP.###) . 10-8
10-3. Variables in the atmospheric pathway data file (ATMSPRM.###) 10-16
10-4. Variables in the surface water pathway data file (SWTPATH.###) . 10-19
10-5. Variables in the ground-water transport pathway data file (GWTRANP.###) 10-23
10-6. Variables in the infiltration leaching and recharge data file (INFILTR.###) 10-27
10-7. Variables in the food chain bioaccumulation pathway data file (FOODCHN.###) . 10-30
10-8. Variables in control data file (CONTROL.###) . 10-33
10-9. Variables in landfill characteristics data file (LANDFIL.###) 10-35
10-10. Variables in surface impoundment characteristics data file (IMPOUND.###) 10-41
10-11. Variables in waste pile characteristics data (WSTPILE.###) 10-44
10-12. Variables in injection well characteristics data file
(INJECTW.###) 10-45
10-13. Variables in tank release characteristics data file (TANKREL.###) 10-46
A-l. Partially saturated zone model comparison for a thin unsaturated zone A-2
A-2. Partially saturated zone model comparison for a thick unsaturated zone A-3
B-2. Comparison between point source solution and MMSOILs for Y=0 at steady state . . B-3
B-3. Comparison between point source solution and MMSOILs for Y=200 at steady state . B-4
B-4, Comparison between transient response of point source model and MMSOILs B-5
IX
-------
FOREWORD
The Exposure Assessment Group (EAG) of EPA's Office of Research and Development has
three main functions: (1) to conduct exposure assessments; (2) to review assessments and related
documents; and (3) to develop guidelines for Agency exposure assessments. The activities under each"
of these functions are supported and respond to the needs of the various EPA program offices. In
relation to the third function, EAG sponsors projects aimed at developing guidelines or refining
techniques used in exposure assessments.
The purpose of this document is to provide guidance on the use of the model MMSOHs:
Multimedia Contaminant Fate, Transport, and Exposure Model for estimating exposure and risk form
soils contaminated by toxic chemicals. This guidance and accompanying model were developed by
combining a multimedia exposure and risk estimation methodology for application to sites where
releases of toxic chemicals have occurred. The methodology described is only applicable as a
screening tool designed to assist EPA in setting priorities for hazardous-waste site management.
Michael A. Callahan,
Director
Exposure Assessment Group
-------
PREFACE
The Exposure Assessment Group (EAG) of the Office of Health and Environmental
Assessment has prepared this model and guidance document to assist individuals in evaluating
exposures to hazardous-waste site contaminants. This document identifies multimedia pathways
related to hazardous-waste site contaminants, presents methodologies for estimating exposures for
each pathway, and provides guidance for applying these methodologies in estimating the human health
risk based on levels of contaminants in soils. The purpose of this model and guidance document are
to serve as a technical resource for estimating potential health risks at sites contaminated by toxic
wastes or spills of toxic chemicals.
XI
-------
ABSTRACT. ^^
This document describes the methodology used by the MMSOILS model for estimating the
human exposure and health risk associated with releases of contamination from hazardous waste sites.
IT
The methodology is a multimedia model addressing the transport of a chemical in ground water,"
surface water, soil erosion, the atmosphere, and accumulation in food. The human exposure •.--
pathways considered in the methodology include: soil ingestion, air inhalation of volatiles and
particulates, dermal contact, ingestion of drinking water, consumption offish, consumption of plants
grown on contaminated soil, and consumption of animals grazing on contaminated pasture. For
multimedia exposures, the methodology provides estimates of human exposure through individual
pathways and combined exposure through all pathways considered. The risk associated with the total
exposure dose is calculated based on chemical-specific toxicity data.
The methodology is intended for use as a screening tool. It is critical that the results are
interpreted in the appropriate framework. The intended use of the exposure assessment tool is
screening and relative comparison of different waste sites, remediation activities, and
evaluation. The methodology can be used to provide an estimate of health risks for a specific site,
but the uncertainty of the estimated risk may be quite large (depending on the site characteristics and
available data) and this uncertainty must be considered in any decision making process.
XII
-------
ACKNOWLEDGMENTS
This document was developed by EPA's Office of Research and Development (ORD) with
assistance provided ICF Incorporated and Versar Inc., in partial fulfillment of EPA Contract No. 68-
DO-0101. Initial MMSOILS model development was conducted by ICF Technology Inc., under EPA
Contract No. 68-C8-00083, with substantial subsequent work conducted under EPA Contract No. 68-
WO-0009. The development of MMSOILS and this documentation and users guide was the result of
the efforts of many individuals mentioned below:
Mr. Malcolm Field (ORD, Office of Health and Environmental Assessment (OHEA), Exposure
Assessment Group (EAG)) served as the EPA Work Assignment Manager and technical editor of both
the text and figures for the final revisions to the MMSOILS model and this documentation and users
guide.
Mr. Gerard Laniak (Environmental Research Laboratory, Athens, GA) facilitated a peer review team
of EPA staff and consultants to assist in finalizing the MMSOILS model.
Mr. Tom McKeon (ICF Technology, Inc.) initiated the design and development of the MMSOILS
model for EPA/ORD/OHEA and managed all development and modification activities represented
in this release version.
Mr. Tom Gherkin (ICF Incorporated) was the ICF Work Assignment Manager responsible for work
performed under EPA Contract Nos. 68-WO-0009 and 68-DO-0101.
Mr. Bob Fares (Versar Inc.) was the Versar Work Assignment Manager responsible for work
performed under EPA Contract No. 68-DO-0101.
The following individuals also contributed substantially to the development of MMSOILS
and/or the model documentation and users guide:
Stephen SchmelUng, Robert S. Kerr Environmental Research Laboratory
Colin Wagoner, ICF Technology
Chris Sawwa, ICF Information Technology
Ravindra Sannareddy, ICF Incorporated
Chris Liro, ICF Incorporated
William Mills, Tetra Tech, Inc.
Robert Johns, Tetra Tech, Inc.
Terry Allison, Computer Sciences Corporation
xiii
-------
-------
1.0 INTRODUCTION
The soils at many hazardous waste sites, including Superfund sites and others, are
contaminated with a variety of chemicals. These contaminated soils pose a potential threat to
human health via transport through various environmental media. The important environmental
transport mechanisms include:
• overland flow to surface water bodies;
* erosion of contaminated soil to off site exposure points;
• . transport, dilution, transformation, and degradation within the surface water body;
• transport of contaminants through the partially saturated (vadose) zone;
• ground-water flow within regional aquifers underlying the site;
* volatilization of contaminants from the soil;
• airborne suspension of contaminated particulate matter due to disturbance of the soil; and
• bioconcentration and movement in the food chain.
The exposure potential and associated health risks of contaminated soils are dependent
not only upon the environmental transport mechanisms but also the pathways by which a human
may be exposed to the transported chemical. The important human exposure pathways include:
* ingestion through consumption of contaminated water;
* inhalation of volatilized contaminants in a gaseous form;
• inhalation of contaminants adsorbed to suspended particulates;
* dermal contact and/or ingestion of contaminated soils; and
• ingestion of contaminants that have moved through the food chain.
More than one environmental transport mechanism may contribute to a human exposure
pathway. As a result, an exposure assessment for a contaminated site must consider the different
environmental transport mechanisms that may contribute to one or more of the human exposure
pathways. A few examples of the coupling between various transport mechanisms and exposure
pathways include:
1-1
/ \
-------
• ingestion of contaminated water which results from contaminants in surface water and/or
ground water;
• ingestion of contaminants through food which results from contaminated irrigation water
applied to the crops and/or contaminated soil in the agricultural area; and
• dermal exposure to soil in an agricultural area which becomes contaminated due to soil
erosion from the site and/or atmospheric deposition.
The process of estimating the environmental concentration at exposure points in various
media, and examining the coupling between different transport mechanisms and human intake
pathways is a complex task. The multimedia methodology described herein is intended for use
as a tool to assist in exposure assessments under such conditions.
1.1 PURPOSE AND SCOPE
This document describes the development of methodologies designed to estimate
exposures and health risks associated with the release and subsequent fate and transport of
chemicals from contaminated soils and various hazardous waste management units (HWMU).
The methodologies have been developed as a tool to assist in exposure assessment studies of the
contaminated soils at inactive hazardous waste disposal facilities and other sites. The four basic
functions of the multimedia methodology are:
1) Estimate chemical release rate from the soil into each environmental media based on
chemical properties and land use at the site.
2) Based on the chemical release rate and the proximity to exposed populations, estimate
the chemical concentration at exposure points in each environmental media considered.
3) Based on the chemical concentration at exposure points and assumptions regarding human
intake levels, estimate the human exposure through inhalation, ingestion and absorption.
1-2
-------
4) Based on the estimated human exposures at exposure points, estimate the potential health
risk based on toxitity data for the specific chemical.
The multimedia methodology uses several mathematical models to represent the important
fate and transport processes. The models are relatively simple analytical models which were
selected to represent the different chemical release and transport processes in water, air and soil.
These models woe selected primarily because the required input parameters are readily
available. The multimedia methodology addresses the coupling between the source location,
chemical release rate, environmental transport mechanisms, and human exposure pathways as
illustrated schematically in Figure 1-1.
The intent of this document is to describe the multimedia methodology for estimating
exposures to soil contaminants. The equations which form the basis of the various model
components are described along with assumptions inherent in the approach.
It is important to be cognizant of the uncertainty inherent in this type of model. Often
the most basic parameters, such as contaminant concentration in soil, vary significantly over a
given site and the distribution may be poorly understood. These uncertainties, coupled with
approximations that were used to streamline the modeling process lead to results that may differ
from reality by orders of magnitude. As such, the user is cautioned to examine the input and
output of the model closely and consider a sensitivity study to evaluate the impact of varying
input parameters on the calculated results.
1-3
-------
FIGURE 1-1. Coupling between source location, environmental
transport mechanisms, and human exposure pathways.
1-4
-------
1.2 ORGANIZATION OF THE DOCUMENT ,
The remaining chapters of this document describe the multimedia exposure assessment
methodology and are organized as follows:
Chapter 1: Introduction
Chapter 2: Applications and Limitations of the Model
Chapter 3: Water and Mass Balance
Chapter 4: Atmospheric Transport Pathway
Chapter 5: Ground-water Transport Pathway
Chapter 6: Soil Erosion Pathway
Chapter 7: Surface Water Transport Pathway
Chapter 8: Food Chain Accumulation Pathway
Chapter 9: Health Risk Assessment Process
Chapter 10: User's Guide
Chapter 11: References
Chapter 12: List of Symbols
Appendix A: Verification of Partially Saturated Zone Model
Appendix B: Ground-water Transport Verification
Chapter 2 describes the intended uses of the methodology along with several important
limitations. Chapter 3 describes the water and mass balance procedures that link different
transport mechanisms and help ensure that physically reliable calculations are made. Chapters
4, 5, 6, 7, and 8 discuss the five principal environmental transport mechanisms addressed in the
methodology: 1) the atmosphere, 2) ground water, 3) soil erosion, 4) surface water and 5)
bioaccumulation in food. For each of the five environmental transport mechanisms, a general
description of the basic processes influencing the chemical release rate and environmental
transport is provided including the mathematical equations which are used to represent the
processes. The description of each of the environmental transport pathways contains the
following components:
1-5
-------
1) One or more release mechanisms representing the mass flux of the chemical entering the
specific environmental media. These release mechanisms are used as the source terms
for the fate and transport models.
2) A fate and transport model used to evaluate the movement of specific chemicals in each
environmental media subject to mixing and/or degradation as a function of distance and
time.
Chapter 9 discusses the health risk assessment process and the human exposure pathways
which include inhalation, ingestion of water, food and soil, and absorption through dermal
contact with contaminated soils. Chapter 10 serves as a user's manual for the computer model
which was designed to assist in the evaluation of each pathway. This chapter includes a
description of the necessary input and output files, and the data necessary for the input files.
Chapters 4, 5, 6, 7, 8, 9, and 10 all contain information on the selection of key input
parameters. Some of the input parameters may be easily estimated, such as the physical
dimensions of the site. Other input parameters, such as the sediment delivery fraction and other
input parameters, are much more difficult to estimate. For these types of parameters, a range
of appropriate input values are identified along with a citation of the original reference where
more information can be obtained.
Chapter 12 provides a summary of the symbols used throughout the document. The
symbols related to specific model components are also listed at the beginning of each section
where they axe used.
1-6
-------
2.0 APPLICATIONS AND LIMITATIONS OF THE METHODOLOGY
In order to apply the multimedia methodology properly, it is essential that the user be
familiar with the general concept of its application, its limitations, and the inherent assumptions.
One example of the model limitations is the conceptual model used to represent the ground-water
pathway (see Chapter 4). The simplified analytical model is based on flow and dispersion
through an idealized homogeneous, uniform porous media. This type of a conceptual model is
not appropriate for representing flow through karst terrains and the user must recognize this
limitation. All of the other pathways have similar limitations that the user must comprehend in
order to utilize this methodology within the intended framework.
Reading this document will help a user understand the methodology, although, just
reading this document will not be sufficient to teach a new user all the information necessary
to properly apply the model, interpret the results, and make critical decisions based, in part, on
the results. The user should have some technical background and skills to help in the many
judgmental decisions that are required for use of the model and interpretation of results. A
thorough review of the pathway components and experience using the methodology will help a
user learn how the various physical, chemical, and biological processes are represented. With
a complete understanding of the methodology and inherent assumptions, a user can make
appropriate decisions in the selection of input parameters, applicability of specific model
components for site specific characteristics, and interpretation of results. The important
assumptions inherent in the individual media specific pathways are described throughout
Chapters 3 through 9.
2.1 APPLICATION OF THE MODEL
The multimedia model is divided into five distinct transport pathways: atmospheric,
surface water, ground water, soil erosion, and food chain bioaccumulation. For each transport
pathway there are a number of potential exposure scenarios which may or may not apply
depending on how the transport pathway has been defined. Not all of the transport pathways
2-1
-------
and exposure scenarios may be required for a specific site. Therefore, it is necessary to define
the specific pathways and scenarios applicable to a given site before applying the model.
The steps that need to be completed before applying the model are: 1) identify the waste
management unit (i.e., the source of contaminant), 2) define the transport pathways, and 3)
identify the appropriate exposure scenarios. The transport pathways and exposure scenarios will
determine which sections of the model need to be run. For example, if all the waste is buried
in a lined landfill with a leachate collection system and capped such that infiltrating rainwater
cannot reach the contaminants, the ground-water pathway may not need to be simulated in the
model. If the ground-water pathway is not simulated, then the ingestion of ground water would
not need to be considered, although ingestion of other sources of water may still need to be
considered. These three preliminary steps for model application are defined in more detail in
the following sections.
2.1.1 Waste Management Unit Identification
The first step that needs to be completed before using the model is to identify the exact
source or waste management unit to be simulated with the model. Two factors need to be
considered: 1) the area of contamination, and 2) the physical characteristics of the area. In the
case of a large landfill, the waste source may be a single cell rather than the entire landfill.
The second feature to consider is the physical characterization of the area. The waste
management' unit should be defined such that it can be simulated with the same transport
scenario(s) and the same exposure scenario(s). For example, if part of a landfill is closed and
covered with a cap, it should be considered separately from another part of a site which is
uncovered and may be subject to different release processes (volatilization, erosion, infiltration,
particular emissions, etc.). In addition, the unit must be sufficiently characterized to support
development of mass flux estimates of a chemical into each transport pathway.
2-2
-------
2.1.2 Define the Transport Pathways
Tlie second step is to define the transport pathways that apply to the contaminant source.
The pathways and the types of issues that must be considered when choosing a specific pathway
are listed below.
Atmospheric Pathway - The atmospheric pathway needs to be simulated if the potential for
airborne releases exists at the site. The types of issues that must be considered include the
following:
1) Is there surface contamination that is potentially susceptible to wind and/or mechanical
suspension?
2) Are there volatile chemicals at the site that may be released into the atmosphere?
3) Has atmospheric contamination been measured at the operable unit?
4) Has atmospheric contamination been measured at exposure points in the vicinity of the
operable unit?
5) Are physical activities at the site, such as vehicle traffic, likely to disturb the soil surface
such that airborne releases are likely?
Surface Water Pathway - The surface water pathway needs to be simulated if there exists a
potential for contaminants to leave the site via runoff into surface water or discharge of affected
ground water. The types of issues that must be considered include the following:
1) Is there surface contamination that is potentially susceptible to erosion?
2-3
-------
2) Are there surface water bodies nearby that could become contaminated by soil erosion?
3) Are there surface water bodies nearby that could be contaminated by the discharge of
contaminated ground water?
4) Are there any direct discharges of contaminants to surface water?
5) Has surface water and/or sediment contamination been measured in water bodies near the
operable unit?
6) Is there any possibility that the sediments in the surface water oody may result in human
exposures?
Ground-water Pathway - The ground-water pathway needs to be simulated if there exists a
potential for contaminants to be transported through the unsaturated and saturated ground-water
systems. The types of issues that must be considered include the following:
1) Is there a potential for infiltration through contaminated soils at the operable unit?
2) Is there a potential for contaminated leachate to enter the unsaturated and saturated
ground-water zones?
3) Has ground-water contamination been measured in the area of the operable unit?
4) Has ground-water contamination been measured at exposure points?
Soil Erosion Pathway - The soil erosion pathway needs to be analyzed if there exists a potential
for contaminated soil to be eroded off site to potential exposure points. The type of issues that
must be considered include the following:
1) Can soils travel off site due to erosion?
2-4
-------
2) Has contamination been measured in soils of nearby off site areas?
3) Is there a potential for human exposure to contaminants in off site soils?
Food Chain Pathway - The food chain bioaccumulation pathway needs to be simulated if there
exists a potential for contaminants, to enter the food chain. The types of issues that must be
considered include the following:
1) Are contaminants or contaminated feed crops ingested by animals which in turn are
consumed by humans?
2) Has contamination been measured in animals or on crops in the area of the operable unit?
3) Has contamination been measured in soils of nearby agricultural areas?
4) Is there a potential for soil erosion from the site to reach nearby agricultural areas?
5) Is there current or potential agricultural use of water (ground water or surface water)
contaminated by the waste site?
2.1.3 Define the Exposure Scenarios
Once the transport pathways have been defined, the last step is to define the applicable
exposure scenarios. As stated above, the model can simulate a number of exposure scenarios,
although not all of these may apply in every case. For each transport pathway defined for a
particular waste management unit, the user must determine if humans could possibly be exposed,
and if so, by what exposure mechanism (i.e., inhalation, ingestion, or dermal contact). The
types of exposure scenarios that should be considered are: 1) inhalation of contaminated air or
soil particles, 2) drinking contaminated water from surface water or from a well, 3) ingestion
of contaminated meat, dairy products and crops, and 4) dermal contact with contaminated soil
or water.
M -. j'
-------
Before a user begins any evaluation and/or modeling of a site, it is useful to prepare a
s
short written description of the scenario to be modeled in order to define the applicable transport
pathways and exposure scenarios. Sketches or maps of the site are very helpful in developing
a quick conceptualization of the site and the problems being simulated. A sketch can quickly
and clearly show the location and dimensions of the waste unit, the transport pathways being
considered, and the location and type of receptors being considered for exposure.
2.2 MODEL UMTTAHONS
The multimedia model is intended for use as a screening tool and it is critical that model
results be interpreted in the appropriate framework. The intended uses of the model include the
following:
1) Relative comparison of human exposures and health risks from different transport
pathways. This type of information may assist in the development of monitoring
strategies and the design of remedial actions.
2) Relative comparison of human exposures and health risks associated with different
chemicals at the site.
3) Relative comparison of the effectiveness of different remedial actions at a site.
4) Screening of required cleanup levels for different sites. Through the choice of
conservative input parameters, and/or suitable factors of safety, cleanup levels for
different sites may be identified such that the health risks associated with the remaining
contaminated soils is reduced to an acceptable level.
<
S) Relative comparison of human exposures and health risks from different sites. This may
be useful for allocating limited resources available for remedial action. It is important
that comparisons between different sites use similar levels of site specific data and/or
screening level data.
2-6
-------
The key .phrases in all of the intended uses are "screening" and "relative comparison".
The model can be used to provide an estimate of health risks for a specific site, but the estimated
risk contains uncertainty and this uncertainty must be incorporated into any decision making
process. The algorithms used to represent the different environmental transport mechanisms are
simple models that cannot represent the detailed heterogeneity and complex environmental
influences affecting the fate and transport of chemicals. These simple models are useful within
the screening framework to identify ranges of probable outcomes based on expected ranges of
the important input parameters. Given these limitations, this model should not be applied in
instances where heterogeneous conditions or complex environmental influences exit. For
example, application of this model to sites with fractured-rock aquifers or karstic aquifers and/or
sites with dense nonaqueous phase liquids (DNAPLs) may yield highly inaccurate results. It is
the users responsibility to use this model appropriately and to correctly interpret the results based
on the site-specific conditions of the hazardous-waste site being modeled.
The results of the model are estimates of chemical concentration at exposure points,
human exposures, and corresponding health risks. At best, the model-predicted concentrations
are order-of-magnitude estimates that are entirely dependent on the conceptualization of the
physical system and the definition of input parameters. While the predicted concentrations and
risk levels are quantitative estimates, they will contain uncertainty making them most useful for
qualitative or relative comparisons.
2.3 MODEL PREDICTIONS VERSUS MEASURED DATA
The majority of this document is devoted to describing methods for estimating chemical
concentration at exposure points in various environmental media. This, involves different
chemical release mechanisms and environmental fate and transport models. The description of
methods in this document should not be interpreted as a preference for the use of fate and
transport models rather than actual measured data of chemical concentrations in water, air, soil,
and/or food at exposure points. If measured data are available that are representative of the
location and averaging period of interest, then measured data would be preferred over modeled
2-7
-------
estimates. However, measurements of present conditions need to be interpreted carefully in
order to estimate chemical concentrations over the entire exposure period.
The mechanisms by which the mass of chemical at a site may be depleted over a long-
term modeling horizon include effective degradation/decay and the various release mechanisms
through which a chemical may enter the different environmental transport pathways. For sites
where the mass of a chemical present is relatively large, measured concentration levels at
exposure points, when available, should be used for some environmental transport pathways.
One extreme example of such a site would be a large mill tailings pile, which may
contain millions of tons of contaminated soil. Under these conditions, measured concentrations
are preferable for the atmospheric, surface water, soil erosion, and food chain pathways. The
time required for contaminants to move through a slow moving ground-water system may be
very long. Measured concentrations under present conditions for the ground-water pathway need
to be evaluated carefully in terms of the distance from the site to exposure points and estimated
chemical travel time through the ground-water system.
Even when contaminant inventories are small, measured chemical concentrations in each
of the environmental media at any off site points can be used in conjunction with a model of the
fate and transport processes to better define a number of the key input parameters. For all of
the transport pathways, measured data may be used to better define the mass of chemical
entering the specific environmental pathway. For the ground-water pathway, any measured data
may also be useful to better define the velocity of contaminant movement. In all cases,
empirical data can be used to calibrate and test the models being applied. Additionally, any
measured concentration levels in soil, water, air, and/or food at exposure points may be
incorporated directly into the exposure assessment (see Chapter 9).
2-8
-------
3.0 WATER AND MASS BALANCE
3.1 SUMMARY OF WASTE MANAGEMENT UNITS
MMSOILs is capable of simulating five types of waste management units (WMUs) that
may be found at RCRA sites: landfills, surface impoundments, waste piles, underground
injection wells and underground storage tanks (USTs). The physical characteristics of each
WMU are different, which necessitates a different series of calculations tailored to those
characteristics. Figure 3-1 is a flow chart depicting the sequence of calculations that are
required for each WMU. In the following sections these computations are described; Figure 3-1
can be used as reference to see whether a particular set of calculations are relevant to a given
WMU. In the following subsections, the calculations used for net recharge, water balance, and
mass balance are described. In Section 3.5, the algorithms for leachate generation from landfills
and waste piles are introduced.
Definition of Parameters
Mg — Initial contaminant mass in the WMU (kg);
M(t) - Contaminant mass in the WMU at the end of year t (kg);
Ma ~ Contaminant mass added to the WMU during the year (kg/yr);
Ma = Total contaminant mass removed from the WMU per year (kg/yr);
Mfec = Contaminant mass removed from the WMU via in-situ decay per year (kg/yr);
Mlw = Contaminant mass removed from the WMU via leachate per year (kg/yr);
M^ * Contaminant mass removed from the WMU via volatilization per year (kg/yr);
AM = Error in estimate of mass loss for impoundments (kg/yr);
Qa = Flow rate of influent to an impoundment (m3/yr);
OH, = Flow rate from WMU to the vadose zone (mVyr);
Ca = Chemical concentration in the influent to an impoundment (kg/m3);
C0 = Initial leachate concentration for CMR (mg/1);
CM = Chemical concentration in the leachate leaving the WMU (mg/m3);
k^ — Chemical decay rate in a WMU (yrl);
3-1
-------
= Empirical desorption parameter (yr1);
= Chemical loss rate from an impoundment due to volatilization (yr1);
V, = Volume of an impoundment (m3);
Vw . = Volume of waste layer (m3);
Of* = Field capacity of waste layer (dimensionless);
a = Chemical loss rate from impoundment through all mechanisms (yr1)
S0 = the initial storage in the WMU (m3);
Qm = the monthly net recharge (mVmonth);
= the monthly flow that is spilled (mVmonth);
= the monthly flow that is collected (nvVmonth); and
AS = the change in storage (m3).
3.2 MONTHLY NET RECHARGE
Net recharge is calculated on a monthly basis for each type of WMU with the exception
of underground injection wells, which are assumed to discharge waste directly to an aquifer.
Net recharge is defined as the precipitation minus the sum of surface runoff and
evapotranspiration. Details of these calculations are more fully described in Chapter 5. For
natural systems, net recharge is the amount of water that enters the soil column and percolates
towards the water table. As such, the magnitude of the net recharge strongly impacts transport
of chemicals from a WMU to the subsurface. For landfills and waste piles, natural recharge is
the only available mechanism used to move water vertically through a WMU. For
impoundments, an influent rate supplements recharge, and for USTs a leakage rate supplements
natural recharge. By design, water balance calculations are not needed for waste piles or USTs
v
because unlike landfills or impoundments, neither has a cap or liner system that is capable of
modifying die natural flow regime.
3-2
-------
FIGURE 3-1. Row chart showing sequence of calculations required
for ground-water transport depending on type of wmu.
CALCULATE IfOMfMLY
RCCHARGC
WATCH OALANCC
MASS BAUNCC
VAOOSC ZONI
I W^^B^^P^^^^^^ I
I
CAlCUUTf MOONO
ZONK IMICXNCSi
CHOUNOWATW
TRANSPORT
CALCULATIONS
3-3
-------
3.3 WATER BALANCE FOR LANDFILLS AND IMPOUNDMENTS
Simulations for landfills and impoundments are different than simulations for other
WMUs because of the inclusion of man-made containment structures (liners, caps and collection
systems) that are capable of modifying the rate of vertical water movement. Both landfills and
impoundments may have several types of liners, and landfills may have one of several types of :
caps. One or more leachate collection systems may be specified, which are mechanisms for
removing water from the WMU. In any case, the liners, caps and the waste are composed of
a number of individual layers that are treated as porous media, with the ability to transmit, store,
or divert water (in the case of a collection system). Consequently, the model requires the
following physical properties for each layer (including the waste layer): thickness, hydraulic
conductivity, field capacity, and saturation. A typical scenario for a landfill might include, from
top to bottom, a vegetative cover, a drainage layer, a synthetic layer over a waste layer, and a
pair of synthetic liners with leachable collection and removal systems below the waste layer.
The water balance is calculated on a monthly interval using Equation 3-1.
Qa - Q^ - Q^ - A5 Eq. (3-1)
Figure 3-2a illustrates Equation 3-1 conceptually. The water balance is calculated using a two-
step process. In the first step, each layer is checked to determine whether it is capable of
passing all of the water that reaches it from the overlying layer. Starting from the top and
working down, the first layer is checked to see whether it is capable of passing the net recharge,
in the case of landfills, or the net recharge plus impounded water plus influent, in the case of
impoundments. The amount of water that a layer can pass is calculated by Darcy's Law using
the amount of hydraulic head on a layer, its hydraulic conductivity, thickness and area (Figure
3-2b). In this way, the most restrictive layer, if any is found and the amount of water that can
be passed through that restrictive layer is established as &«. If none of the layers are restrictive
enough to inhibit the available flow, then gL, is equal to the available water in the system.
3-4
-------
FIGURE 3-2a. Schematic water balance for a layer.
(
(UNOntL)
OHL^C
UNCI
Of*
Y//////,.
3-5
-------
FIGURE 3-2b. Schematic water storage in a layer.
Ota
(AT SATURATION)
STQRAOC (AT riOD CAPACITY)
3-6
-------
In the second step water is moved between the layers and layer storage is changed if
necessary. Hie most restrictive' layer and those above it are filled to saturation, and layers
below the restrictive layer are drained to Meld capacity. The flow of water out of the WMU
(CoJ| can be greater man Qm if water is released from storage from below a restrictive layer.
The water balance is computed on an monthly basis to account for seasonal variations in net
recharge, and the calculated flow values are summed annually for use elsewhere in MMSOILS.
3.3.1 Liner and Cap Failures
Liners and caps are generally constructed from low permeability materials including clays
and synthetic materials that act to restrict the flow of water through a layer. Liners and caps
are known to fail in certain conditions (due to tearing, cracking, seam separation, destruction
by burrowing animals, chemical breakdown and desiccation). MMSOILS can optionally
simulate these liner and cap failures by allowing the user to specify the timing and magnitude
of a series of failures.
Up to five independent failure events may be specified for each layer in the WMU.
These are specified by the year in which the failure occurs and the fraction of available water
that will be passed through the layer beginning in that year and continuing until a new failure
is specified. The magnitude of the failure is specified as a fraction between 0 and 1. A failure
fraction of 0 means that the liner is intact and a failure fraction of 1 indicated that the liner has
failed completely. With this approach, a progressively deteriorating liner can be simulated,
which might pass 5 percent of the available water after 5 years, 25 percent after 10 years and
100 percent after 25 years.
3.3.2 Leachate Collection Systems
The effectiveness of leachate collection systems can also be specified in MMSOILS. This
parameter can be used to mimic different collection system designs. For instance, closely spaced
drainage pipes will generally be more efficient in removing leachate that widely separated
systems. In MMSOILS the collection system may fail catastrophically in a specified year. This
3-7
-------
functionality was included to simulate the situation when a landfill has been permanently closed
and the owner/operator no longer maintains the collection system.
3.4 MASS BALANCE
A mass balance routine is included in MMSOILS to ensure that mass is conserved in
WMUs mat have multiple release pathways (Figure 3-3). The relevant WMUs are landfills,
impoundments and waste piles. For each of these WMUs, the mass that is removed from the
WMU through each pathway is equal to the sum of the mass that has been added to the unit and
the mass in the unit at any given time. Mass balance calculations are made on an annual basis.
Results of the mass balance calculations are written to the output file for each year and the
annual results are accumulated over the lifetime of the WMU and are presented as a summary
table at the end of the annual calculations.
3.4.1 Impoundment Mass Balance
Mass balance within an impoundment is simulated based on a continuously mixed reactor
model. In this model, the chemical mass within the impoundment is in a dynamic equilibrium
between the mass added and the mass lost through volatilization, in situ decay and leakage
through the bottom of the impoundment. The relative importance of each of these loss
mechanisms depends on the volatility of the chemical and depth of the impoundment, a user
specified decay factor, and the leakage rate from the impoundment.
An equation for the mass loss due to each of these mechanisms can be developed from
the statements that the rate of change of mass in the impoundment is equal to the rate of mass
addition minus the rate of mass subtraction or:
Eq.<3-2)
3-8
-------
FIGURE 3-3. Release pathways from a WMU.
MASS AOOED
TO WMU
ATMOSPHERIC
EMISSION
IN SITU
DECAY
MASS REMAINING
IN WMU
TO COLLECTION
SYSTEMS
LEACHATC
GENERATION
3-9
-------
Equation 3-2 can be salved by separation of variables and integration. Applying the initial
condition that the mass in the impoundment at time zero is equal to zero gives:
M(t) - Q. «+ MO e^ Eq. (3-3)
where
o = — + k^ + k^ Eq. (3-4)
A mass balance expression for the impoundment is used to develop an equation for the amount
of contaminant mass left in an impoundment at the end of each year:
M(t) = Me + Ma - M9 Eq. (3-5)
Solving for Af0 and using the expression for Mft) given in Equation 3-4 leads to an expression
for the total mass loss:
_ Eq. (3-6)
Expressions for the loss due to each individual mechanism acting alone can be developed
in an analogous fashion. However, this approach overestimates the individual contributions
since each individual loss mechanism must compete for a finite amount of mass. The sum of
the individual loss mechanisms over-estimates the total loss by an amount AM as follows:
3-10
-------
AM * **, +#*, *'Jf -Af Eq. (3-7)
where the subscripts gw, dec, vol, and a in Equation 3-7 refer to the amount of mass removed
through flow to the vadose zone, decay, volatilization and the sum of the three processes,
respectively. The excess mass (AM) is divided between the different loss mechanisms according
to their relative contribution to the total loss.
Since the magnitude of each of these mechanisms is calculated as the sum of a fraction
of the mass in the impoundment at the beginning of a year and a fraction of the mass added to
the impoundment during the year), it is impossible to deplete more than the available mass.
The chemical that is added to the impoundment is assumed to be in solution and the
solution mixes rapidly with the water that was previously impounded. Since the impoundment
is fully mixed, the concentration of the leachate is equal to the concentration in the
impoundment. When an impoundment is closed and turned into a landfill, the leachate is
generated using one of the methods described below.
3.4.2 Landfill and Waste File Mass Balance .
Mass balance for a landfill is handled in a slightly different manner than for an
impoundment. For landfills, each of the loss mechanisms (in situ decay, volatilization and
leachate generation) is calculated independently and the mass in the landfill is tracked using
Equation 3-8:
M(t) * M. + Ua - M^ - M^ - M^ Eq. (3-8)
The sum of the individual loss mechanisms in Equation 3-8 is checked against the available mass
to ensure that the total loss is less than or equal to the available mass. If the loss exceeds the
available mass then each of the individual losses is scaled accordingly.
3-11
-------
The initial contaminant mass in a WMU is assumed to be zero in year zero and the unit
is constructed for a user specified number of years. In subsequent years, the initial mass is
taken as the mass in the WMU at the end of the previous year. The amount of mass that is
added each year is equal to the incremental waste volume times the bulk density of the waste.
MMSOILS offers the option to deplete the chemical mass from within a landfill or waste
pile. The process is an exponential decay such that the mass left in the WMU at the end of time
/is:
M(t) *M,e-*-' Eq. (3-9)
and the amount of mass that decays in the same time period is M0 - Mft) or
If,-If.(IV-1) Eq.(M0
The volatilization process and the algorithms that are available within MMSOILS are
described in Chapter 4. Another significant difference between landfills and impoundments is
the flexibility in choices for a leachate generation mechanism. The options for simulating
.leachate generation are presented in the following sections.
3.5 LEACHATE GENERATION FROM LANDFILLS AND WASTE PILES
MMSOILS has four leachate generation algorithms that can be used to calculate the
contaminant concentration in the waste leachate. The best mathematical approach for simulating
leachate quality is dependent on the type of waste management unit under consideration. The
leachate generation approaches for calculating the chemical concentration in landfill leachates
are the following: the solubility limit approach, the partitioning approach, the completely-mixed
reactor approach, and the steady-state leachate concentration approach. The impoundment model
uses a continuously-stirred reactor approach to model leachate quality during the impoundment's
3-12
-------
active life as described above. After its closure, the impoundment reverts to a landfill, and one
of the four mathematical approaches for landfills can be used.
As described in the water balance section of this report, MMSOILS is capable of
simulating leachate collection systems. The chemical concentration in the collected water is
calculated using one of the four methods described below. The amount of mass that is captured
in a WMU by a simulated collection system is equal to the product of the leachate concentration
and the flow rate of water into the collection systems.
3.5.1 Solubility Limit Approach
The first leachate generation algorithm makes the assumption that the leachate
concentration is approximated by the solubility limit of the chemical in water. This method is
often convenient when the solubility limit is available but other data (such as a partitioning
coefficient) are not.
3.5.2 Partitioning Approach
The partitioning model estimates leachate strength as a function of the contaminant mass
disposed, the incremental growth of the landfill, the constituent-specific partition coefficient (1Q
and aqueous solubility, and the annual infiltration rate. This approach assumes equilibrium
partitioning between the soil and soil moisture based on the partitioning coefficient and the mass
of chemical'released in the unit each year. The user must input a soil concentration for landfills,
or an influent concentration for impoundments. A mass balance of the waste unit is incremented
on a yearly basis as additional material is added to the unit during its operational life, and as
chemical constituents are removed from the unit through leachate generation.
*
To estimate die mass flux of a contaminant leaching out of the waste layer, the following
equation is used in MMSOILS based on the partitioning method:
3-13
-------
Eq. (3-11)
The mass flux to the vadose zone is equal to the product of the leachate concentration and the
volume of water moving through the bottom of the WMU.
The key parameter in this estimate of leachate concentration is the soil-water partition
coefficient, K^. For organic chemicals, a partition coefficient may be estimated based on the
organic carbon in the soil and the organic carbon partition coefficient, K^, for the specific
chemical. Estimates of K^ mat be available from literature sources or a value may be estimated
based on the octanol-water partition coefficient,
For inorganic contaminants, estimates of K^ are more difficult to obtain due to additional
influences of pH, oxidation state of the element, clay content, and organic matter in the soli
material. An extensive review of measured and estimated partition coefficients for inorganic
materials as a function of pH, oxidation state, clay and organic material content has recently
been compiled by Peterson (1987).
The soil-water partition coefficient is the key parameter in defining whether a chemical
will stay adsorbed to soil or leach from the soil and migrate to ground water. Any method or
literature sources used to estimate this parameter will contain substantial uncertainty that must
be incorporated in the evaluation of the results. A discussion of some of the sources of
uncertainty is provided by Baes, et al., (1984). The discussion in Baes, et al., (1984) deals
primarily with estimated values for inorganic elements although many of the key points also
apply to Kj values for organic chemicals as well. The key source of uncertainty discussed by
Baes include measurement errors, additional errors associated with a value calculated as the ratio
of two measured values, and variation with soil types (i.e., pH, clay content, organic matter
content, free iron and manganous oxide contents, and particle size distributions). Detailer
discussions of uncertainty associated with methods used for estimating soil-water partition
coefficients for organic chemicals are provided by Mingelgrin and Gerstl (1983) and Karickhoff
(1984). All of these authors emphasize that substantial uncertainty exists with both measured
3-14
-------
and estimated K* values. This uncertainty may cover one or more orders of magnitude
depending on the type of chemical and site conditions. The user must use site-specific data
whenever possible and recognize the large degree of uncertainty.
3.5.3 Completely-mixed Reactor Approach
The completely-mixed reactor (CMR) approach was developed for the simulation of
leachate generation from municipal solid waste landfills. It estimates contaminant concentrations
as a function of the initial leachate concentration, the cumulative infiltration into the waste unit,
the total volume of waste in the unit, and the incremental growth of the landfill. It can account
for individual contaminant partitioning where empirical data are available defining a desorption
parameter. MMSOILS adopts the following equation for estimating the leachate concentration
out of the waste unit based on the CMR approach (EPA, 1986):
Eq. (3-12)
3.5.4 Steady-State Leachate Concentration
In some cases, there may be little or no information on waste characteristics other than
an empirical leachate concentration. In such cases, it will be possible to simulate release of a
constant strength leachate from landfills and closed impoundments for the duration of the
modeling period.
3-15
-------
-------
4.0 ATMOSPHERIC PATHWAY
The atmospheric pathway considers the release of contaminants from the site in the form
of vapors and fugitive dust emissions from wind erosion and mechanical disturbances. Once the
contaminant is in the atmosphere it is transported by wind and dispersed due to turbulence in the
flow. This chapter describes the models used to represent the following processes: volatilization
from soils, volatilization from a water body, paniculate emissions due to wind erosion and
mechanical disturbances, atmospheric transport and dispersion, and atmospheric deposition. For
each of these processes, a list of parameters and definitions are given along with the basic
equations which are used to represent the processes. Some of the assumptions and limitations
associated with the use of the different models to represent the actual physical processes are
listed. The conceptual model of the atmospheric pathway is shown in Figure 4-1.
FIGURE 4-1. Conceptual model of the atmospheric pathway.
PREVAIUM6* WIMO
4-1
-------
4.1 VOLATILIZATION RELEASE MODELS
This section describes four different models for estimating the release rate of chemicals
due to volatilization from soils (see Table 4.1) and an additional model for estimating
volatilization release rate from a water body. The model user must select the volatilization
release model which best represents the conditions at the WMU of interest. The different
models represent different physical characteristics of the site (i.e., covered or uncovered), and
phase of a chemical in the soil (i.e., liquid form in soil moisture or adsorbed to the soil
material). The selected models are discussed in more detail in EPA (1987) and EPA (1986a).
Definition of Parameters:
CB = total chemical concentration in soil (mg/kg),
C, - adsorbed chemical concentration in soil (mg/kg),
Ca = liquid-phase chemical concentration in soil (g/cm3),
€„ = saturation vapor concentration of chemical (g/cm3),
D = phase transfer coefficient (cm2/s),
Dt - diffusion coefficient of compound in air (cm2/s),
d = thickness of cover (cm),
4 = depth of dry zone (cm),
ti = Henry's law constant in concentration form (dimensionless),
H = Henry's law constant (atm mVmol),
Kd — soil-water partition coefficient (ml/g),
Ld = total depth of contamination from soil surface (cm),
MW — mole weight of compound (g/mol),
N = mass flux (g/s/cm2),
P, = vapor pressure of compound (mm Hg),
R = gas constant (8.2 x Itf5 [atm m3]/[mol °K]),
Rm = molar gas constant (62,361 [mm Hg cm3]/[mol TCJ),
Sr — vapor phase saturation (unitless),
S, = liquid phase saturation (Sv + Sw = 1) (unitless),
4-2
-------
T = absolute temperature (°K),
t = time (sec),
XF1 = units conversion factor (10s mg/kg per g/g),
a - effective diffusion parameter (cmVs),
0 = porosity
p, = particle density of soil (g/cm3).
p, = bulk density of soil (g/cm3)
C^ — chemical concentration in air (g/cm3),
Cw = chemical concentration in water (g/cm3),
Fa a mass flux to the atmosphere (g/hr/cm2),
H = . Henry's law constant (atm-mVmol),
kL = overall water to air mass transfer coefficient (cm/hr),
k, = liquid-phase mass transfer coefficient (cm/hr),
k'(COi) = liquid-phase mass transfer coefficient for COj (20 cm/hr),
kt = gas-phase mass transfer coefficient (cm/hr),
t
'(Hyp) = gas-phase mass transfer coefficient for H2O (3000 cm/hr),
MW = molecular weight of chemical of interest (g/mol),
= molecular weight of CQj (44 g/mol),
MW(H O) = molecular weight of H2O (18 g/mol),
K - gas constant (8.2 x 10* atm-m3/mol-°K),
T = absolute temperature ( °K).
The first two volatilization release models are for covered and uncovered sites where
the contaminant is in liquid form dissolved in the soil moisture. A schematic illustration of the
physical site characteristics and phase of the chemical within the soil column for these two
release models are shown in Figure 4-2. The third and fourth volatilization release models are
for covered and uncovered sites where the chemical is in an adsorbed phase on the soil material.
A schematic illustration of the physical site characteristics and phase of the chemical within the
soil column for these two release models is shown in Figure 4-3.
-------
I
I
w
s
m
3
It
it
si
I
I
II
•JE
S .5
H
I
IB
(U
e
-------
FIGURE 4-2. Physical site characteristics and phase of the chemical
within the soil column for volatilization release models 1 and 2.
UPUIO
IU 60IL
5££T10UAL
L»kl60VERED
4-5
-------
FIGURE 4-3. Physical site characteristics and phase of the chemical
within the soil column for volatilization release models 3 and 4.
is. -g«
IPi^JOOf* *^r
*!«V*^s5*5r 6O1U i
9OIU
6H6MI6AI.
*G9 4011- MATERIAL
6ITE
50IL
6H&MI6AI
10
"D6FTH
.(50UTAMHJATBO
S1T£
4-6
9
-------
4.1.t Farmer's Equation - Covered Sites, Liquid Phase
The first release model described is referred to as Farmer's equation (EPA 1987).
This model represents the steady state vapor emission rate from a covered site. The model
assumes that the steady state emission rate does not deplete the mass of chemicals in the soil.
This implies mat the mass of chemical is infinite, however, in the context of an exposure
assessment, it simply means a large mass of chemical is present that is not significantly depleted
over the 75-year exposure duration. This model also assumes that the chemical concentration
in the waste is high enough to produce a saturated vapor concentration in the air phase at the
surface of the waste. This implies that the emission rate is independent of the soil and/or liquid
phase concentration, as shown in Equations 4-1 and 4-2 where the chemical concentration in the
waste does not appear. The steady state release rate, N, is:
Eq. (4-1)
where, Cw, the saturated vapor concentration is defined as;
4.1.2 Landfarming Equation - Uncovered Sites, Liquid Phase
The second volatilization release model is for uncovered sites where the contaminant
is in liquid form in the pore space of the soil. This is referred to as the Landfarming equation
(EPA, 1987)- The emission rate, N, is estimated as the average emission rate over a time
interval as:
4-7
•
-------
2DCat Eq. (4-3)
where, the phase transfer coefficient, D, is estimated as;
D * Dt E«3 H' Eq. (4-4)
and the dimensionless form of Henry's Law constant, ti, is estimated as;
U' = JL Eq. (4-5)
R T
and the total soil concentration, CBi is;
CB = C^flS,, + C^flS, * C,pB Eq. (4-6)
4.1.3 Covered Sites, Adsorbed Phase
The third volatilization release model is for covered sites where the contaminant is
adsorbed to the soil (EPA, 1986a). The model is based on a mass balance of the chemical which
is transported by diffusion in the vapor phase. Vapor phase and soil adsorbed phases of the
chemical are assumed to be in equilibrium.
The initial conditions used to derive the model are; the cover is initially clean and
extends to a depth, d, from the soil surface, and the depth of the contaminated soils extends to
a total depth, Ld, below the soil surface where the concentration within the depth, d
-------
uniform. The boundary conditions for this specific solution are that the vapor concentration at
the soil surface is zero (corresponding approximately to continuous removal by wind) and that
the gradient of the concentration at the bottom of the contaminated soil is zero. The emission
rate, N, is calculated as the instantaneous emission rate at a specific time, /, as:
a
a(2rn-l)V t
cos\
2n+l
2
Eq. (4-7)
This analytical solution is based on a Fourier series representation of the diffusion equation. The
series solution specifies the summation of an infinite series, however the series converges fairly
quickly depending on the time of interest. An appropriate number of terms for the summation
may be estimated based on setting the exponential term in Equation 4-6 to a small number (e.g.,
10~") and solving for the number of terms required. This equation is numerically integrated
over a time interval to estimate the average flux over the interval.
4.1.4 Uncovered Sites, Adsorbed Phase
The fourth volatilization release model is for uncovered sites where the contaminant
is adsorbed to the soil (EPA, 1986a). Similar to the third volatilization release model, this
model is based on a mass balance of the chemical which is transported by diffusion in the vapor
phase where the vapor phase and soil.adsorbed phases of the chemical are assumed to be in
equilibrium.
The initial condition used for this specific solution is that the soil concentration is
uniform for an infinite depth and the boundary condition is that the vapor phase concentration
at the soil surface is zero. The emission rate, N, is calculated as the average emission rate over
a time interval, from time equals 0 to r, as:
4-9
-------
hctt Kd
where the effective diffusion parameter is calculated as:
.(4-9)
The validity of the specified initial condition used for this solution (i.e., uniform concentration
for an infinite depth) depends on the effective diffusion parameter, a, the time frame of interest,
and the total depth of contamination. The influence of the specified initial condition at a specific
site may be easily tested through comparison of model predictions using volatilization release
model 3 (Equation 4-6), with the depth of clean cover, d, specified as an arbitrarily small value,
such as 0.1 cm.
4.1.S Volatilization from a Contaminated Water Body
The mass flux of a chemical into the air from contaminated water is estimated using
a two-layer resistance model of the air-water interface. The two-layer resistance model is
described in a variety of sources including Mackay (1985), EPA (1984b) and, Liss and Slater
(1974). The two-layer resistance model is an approximation that is useful for simplifying the
theoretical calculations of gas exchange rates. The two-layer resistance model assumes that two
thin films exist at the air water interface (a gas film and a liquid film) and that these two films
form the dominant resistances to mass transfer across the interface.
Derivation of the two-layer resistance model is provided in Mackay (1985) and Liss
and Slater (1974). The model is represented as:
4-10
-------
Eq. (4-10)
where,
*! =
RT
' #F
Eq. (4-11)
MW
0.5
Eq. (4-12)
AfW
Eq. (4-13)
The chemical concentration in the water is estimated using methods outlined in
Chapter 6 (see Chapter 6 for additional discussion of the gas film transfer coefficient). The gas
concentration is assumed to be zero, corresponding approximately to continuous removal by
wind. The mass flux term in Equation 4-9 is specified over a unit area basis hence, it must be
multiplied by the area of the water body to get the total mass flux to the atmosphere. This vapor
phase chemical mass flux is then combined with volatilization losses from the site in order to
estimate the total mass flux to the atmosphere.
4.2
RELEASE MODELS FOR ESTIMATING PARTICULATE EMISSIONS
This section describes four empirical models for estimating the average release rate
of particulates. The first model can be used to estimate the release rate due to wind erosion,
4-11
-------
a complete description of this model can be found in EPA (1985a) or EPA (1989c). These
particles are assumed to be less than 10 microns in diameter. The second, third and fourth
models, which are for particles less than 30 microns in diameter, can be used to estimate the
release rate due to mechanical disturbances, a complete description of these models can be found
in EPA (1985b).
Definition of parameters:
Af SB area of site (hectares),
C, = contaminant concentration in surface soils (g/g),
E^ = emission factor from an unpaved road per vehicle- kilometer of travel (kg/VKT),
F(x) = plotted function from EPA (1985a) (dimensionless),
PM10 = emission rate for inhalable particulates, less than 10 urn, (g/hr),
PMjo = emission rate for particulates less than 30 /xm, (g/hr),
pe = number of days per year with at least 0.01 inches of precipitation (days/year),
S = mean vehicle speed (km/hr),
s = percent silt content (percent),
uw = mean annual wind speed (m/s),
u, = threshold velocity at 7 m height (m/s),
VG «= fraction of contaminated surface with vegetative cover (dimensionless),
VKT - average vehicle kilometers of travel per hour on unpaved road, (# veh. x km/hr),
W = mean vehicle weight (Mg),
w = mean number of wheels per vehicle (#),
x = dimensionless ratio (x=0.886 u/u),
XF2 = units conversion factor (10* mVha),
Erf = emission factor from loading operations (kg/Mg),
kp = particle size multiplier (dimensionless),
DH = drop height (m),
M = material moisture content (percent),
YD = dumping device capacity (m3),
.= emission factor from spreading operations (kg/ha),
4-12
5
-------
kf = particle size multiplier (dimensionless),
Mw = mass of soil loaded per day (Mg/day),
Ay = spreading area per day (hectares/day).
4.2.1 Participate Emissions due to Wind Erosion
This empirical model is applicable for sites that are not covered by continuous
vegetation, where the surface soils may be classified as having an "unlimited reservoir" of
erodible surface particles. Generally this would be restricted to loose sandy soils which do not
form a surface crust, or sites where the surface soil is regularly disturbed, such as by vehicle
traffic. If the surface is covered by continuous vegetation the wind erosion would be essentially
zero. The equation used to represent the wind erosion of inhalable particulates, PMIO, is:
PM10 =0.036 (1-VG)
u,
F (X) A3XF2 Eq. <4"14)
The threshold velocity at 7 m («,) is based on the threshold friction velocity which may be
estimated based on the mode of the soil aggregate size distribution. A procedure for estimating
this parameter is outlined in EPA (1985a). For this empirical model (i.e., an "unlimited
reservoir surface"), EPA (1985a) suggests that the threshold friction velocity is less than 75
cm/sec, corresponding to a soil aggregate size distribution mode of 1 mm. This value of a
threshold friction velocity (in cm/sec) is then converted to an equivalent wind speed at a 7 meter
height (in m/sec) based on a logarithmic velocity profile (e.g., see EPA, 1985a).
The computer model developed to assist in this multimedia methodology performs this
conversion internally, hence the user only needs to input the threshold friction velocity (in
cm/sec) based on the mode of the soil aggregate size distribution (see Chapter 9). Similarly for
the function F(x) in Equation 4-13, this parameter is calculated by the model based on other
necessary input data. For more details on how these parameters are calculated see EPA (1985a).
4-13
-------
4.2.2. Paniculate Emissions due to Vehicle Traffic
The empirical equation used to represent the emission factor for participates, Em
(kg/VKT), from an unpaved road per vehicle kilometers traveled per hour, VKT, is:
E • • 0.8 (1 7) — • — ——
-" f ' 12 48 2.7
S f W^ar fwl"
365
Eq. (4-15)
Default values for some of these parameters are provided by EPA (1985b) for cases where site
specific information is not available. The default values are shown in Chapter 10. The number
of days with greater than 0.01 inches of precipitation is a local/regional parameter that is
available from weather stations and/or maps presented in various sources including EPA
(1985b).
4.2.3 Loading and Unloading Operations
Emissions during loading and unloading of contaminated soil can be estimated using
an equation given in EPA (1985b). The emission factor equation is:
„„ *kf(0.0009)
u^
2.2
DH
1.5
M
4.6
Eq. (4-16)
Equation 4-16 provides an emission factor for kilograms of paniculate emitted per megagrams
(Mg) of soil unloaded or loaded. The particle size multiplier varies with aerodynamic panicle
size and is given numerical values of 0.73 and 0.77 for batch drop and continuous drop
4-14
-------
operations for particle sizes less than 30 urn. The other parameters in Equation 4-16 must be
defined based on site operation characteristics.
4.2.4 Soil Spreading Operations
It is assumed that soil spreading operations cause emission on the same order as
agricultural tilling. The emission factor derived from EPA (1985b) is used to estimate
emissions:
£.. = 5.38 k s°>6 lq. (4-17)
The particle size multiplier (kj varies with aerodynamic particle size, and is given as 1 for total
particulate and 0.33 for particulates less than 30 pm.
The total emission for particles less than 30 pm is calculated by adding the
contributions from each of the three mechanisms (vehicular traffic, loading and unloading, and
soil spreading).
X = Ct(EUK-VKD
4.3 ATMOSPHERIC TRANSPORT AND DISPERSION MODEL
MMSOILs uses one of two algorithms for estimating ambient atmospheric
concentrations due to emissions from waste management units. The first algorithm uses a box
model to estimate concentrations within 100 meters of a source. The second algorithm
implements a sector-averaged form of a Gaussian plume model to calculate concentrations at
distances greater than 100 meters from the source. The box model is used in proximity to the
4-15
5"
-------
source because the manner in which the Gaussian plume model represents area! sources is
inaccurate near the source. These two models are described in the following sections. A virtual
distance is used as an approximation for representing the differences between an area! source and
a point source. The basic assumptions of the underlying model are: the spread of the plume
follows a Gaussian distribution, the emission rate is uniform and continuous over the source, and
meteorological conditions remain constant between the source and the receptor location. The
basic Gaussian plume model is modified to account for deposition flux onto the ground surface
through the use of a source depletion algorithm described by Horst (1984).
Definition of parameters!
Ay = width of waste management unit (m),
Ca = atmospheric concentration (mg/m3),
dx — source depletion factor due to deposition flux (dimensionless),
fq = frequency of the specific stability array parameters for classification i J (stability
class, wind speed) (dimensionless),
Ma = mass flux of contaminant into the atmosphere (g/s),
uw = average wind speed (m/s),
vd = deposition velocity (m/s),
X = distance in x coordinate direction (parallel to velocity u) from source to point of
interest (m),
xv = virtual distance required for point source plume to spread to width of site (m),
Y = distance in y coordinate direction (perpendicular to velocity u) from source to
point of interest (m),
Z — distance in z coordinate direction (perpendicular to velocity u) from source to
point of interest (m),
ay = mixing coefficient in y direction, standard deviation of Gaussian plume (m),
-------
4.3.1 Box Model
A box model is used to estimate atmospheric concentrations for observation points
within 100 meters of the source. The atmospheric concentration calculated using this model is:
C = .. M' _ Eq. (4-19)
The mixing height Zm is fixed at Z meters in the model.
4.3.2 Sector Averaged Gaussian Plume Model
For points more than 100 meters from the waste management unit, a sector-averaged
Gaussian plume model is used to calculate atmospheric concentrations.
A sector averaged form of the Gaussian plume model is used for estimating
atmospheric concentrations and deposition fluxes at exposure points. The model represents the
average concentration within a directional sector, 22.5 degrees, based on:
1) the mass flux from the source area;
i
2) the wind speed;
t
3) the distance from the virtual source;
4) the vertical standard deviation of the Gaussian plume based on the stability class; and
5) the deposition flux onto the ground surface.
The atmospheric transport and dispersion model for the sector averaged concentration of a
specific sector is (Horst, 1984):
4-17
-------
exp
-Z*
* f
I"* Jy
if «„,
Eq. (4-20)
Hie source depletion factor, dx, is used to represent the net removal of contaminants from the
plume due to deposition as a function of distance from the source. Hie factor d^ is calculated
as (Horst, 1984, PasquUl, 1974):
exp
exp
Eq. (4-21)
The virtual distance, JTV, is used to distribute the mass flux from a point source over a finite area
representing the width of the site. The virtual distance is calculated as the distance required for
the transverse standard deviation of the Gaussian plume, a,, to grow to the half width of the site.
This distance will be different for each stability class and is calculated as:
Eq. (4-22)
where the stability dependent coefficients p and q are given in Table 4-2.
The vertical and transverse mixing lengths, at and oy, are estimated as a function of
the atmospheric stability class and the distance from the source. The relationships are
represented as polynomial functions of the distance from the site for each stability class. The
atmospheric transport parameters are represented by a stability array (STAR) tabulation of the
wind speed and stability class frequencies for each directional sector. For some cases, a
complete STAR tabulation may not be available and a conservative assumption suggested by
4-18
-------
TABLE 4-2. Coefficients Used to Calculate Lateral Virtual Distances
for Pasquill-Gifford Dispersion Rates
PASQUILL STABILITY
CATEGORY
A
B
C
D
E
F
P
209.14
154.46
103.26
68.26
51.06
33.92
q
0.890
0.902
0.917
0.919
0.921
0.919
EPA (1992a) is to use atmospheric stability class E, with the wind blowing directly from the site
to the exposure point 30 percent of the time at 3 meters/second.
4-19
-------
-------
5.0 GROUND-WATER PATHWAY
The ground-water pathway examines the net recharge (precipitation minus
evapotranspiration), leaching of contaminants from the soil, transport through the partially
saturated zone, and contaminant transport/dispersion within an aquifer. This section describes
the mathematical models used to represent each of the different processes and describes some
of the assumptions and limitations associated with the use of the different models to represent
the physical processes. The conceptual model used to represent the ground-water pathway is
shown in Figure 5-1.
/
5.1 CALCULATING RECHARGE
Recharge is the volume of moisture which percolates into the soil and moves downward
to the ground water. In areas not receiving irrigation, recharge is calculated as the difference
of total precipitation (rainfall and snow melt) and total losses. Losses are generally defined as
runoff and evapotranspiration. Runoff, the portion of precipitation that flows over the surface
of the soil rather than infiltrating the soil column, is generally estimated on the basis of soil
type, topography, existing soil moisture content, and stage of growing season.
Calculation of recharge is performed using a water balance approach. This approach is
tantamount to an accounting method wherein inputs (precipitation, irrigation, surface runon and
subsurface lateral flow) are added and outputs (runoff, subsurface lateral flow, and
evapotranspiration) are subtracted leaving a residual for changes in storage (available moisture
in soil) and recharge. Lateral inflow and outflow are assumed to be equal. The calculation then
requires data for precipitation, irrigation, runon/runoff, and evapotranspiration.
Irrigation inputs can be determined from operating and post-closure plans for the site.
Runon is generally excluded from landfills through engineered barriers. If runon is allowed, it
can be added to precipitation for the calculations. Runoff is a function of storm intensity and
duration; thaw conditions; antecedent moisture; permeability and infiltration capacity; slope; and
amount and type of vegetation.
-------
FIGURE 5-1. Conceptual model of contaminant movement in the
ground-water pathway.
&M
PUAKJ view
5-2
-------
5.2 CALCULATING RUNOFF .
The U.S. Soil Conservation Service has developed a method for estimating storm runoff
volumes from small agricultural catchments with various kinds of soil and land use. The
technique is based on a simplified infiltration model of runoff and an empirical approximation.
For each catchment and storm, a curve number is chosen for use in the analysis. The curve
number is an empirical rating of the hydrologic performance of a large number of soils and
vegetative covers throughout the United States.
Definition of parameters:
CN = curve number representing infiltration capacity of soil (I/cm),
IA = initial abstraction representing surface storage (cm),
Pt = depth of precipitation from a storm of 24 hour duration (cm),
Qr = depth of runoff from a storm of 24 hour duration (cm),
Sm = maximum potential retention (cm).
The SCS Curve Number runoff formula is expressed in the following form:
£,.(5-1)
(P.-0.8S,)
This formula is applicable when the precipitation is greater than the initial abstraction and the
initial abstraction is approximated as one fifth the maximum potential storage (i.e., Pf > IA, and
IA = 0.2 SJ. If these conditions are not met, the uncertainty of the estimation will increase.
The curve number (CN) is related to the maximum potential storage as:
5-3
-------
100°
.
Runoff curve numbers for various combinations of soil, cover, and land use practice can be read
from Table 5-1. The hydrologic soil groups and cover types are defined in Tables 5-2 and 5-3.
Major soils of the United States have been classified into the hydrologic groups described in
Table 5-2, and are listed in a U.S. Soil Conservation Service (1972) handbook and in county
Soil Survey Reports. The curve numbers in Table 5rl apply to average soil moisture conditions.
The antecedent moisture levels are classified into three groups in Table 5-4 on the basis of total
precipitation occurring within the preceding 5 days. The curve numbers of Table 5-1 refer to
antecedent moisture condition n in Table 5-4. While the user must input a curve number based
on antecedent moisture condition, the model will automatically adjust the calculation for the
other cases if applicable.
In order to use Equation 5-1 to estimate the amount of runoff, the total monthly
precipitation must be divided into storm events. This is done based on the average number of
days per month with precipitation (reported in the climatological data summary as the number
of days with precipitation greater than 0.01 inches). The total monthly depth of precipitation
is broken into individual storms of equal magnitude. The antecedent moisture condition is
estimated based on the average amount of precipitation occurring during the -month (5 day
average) and the curve number is adjusted for the appropriate antecedent condition. This method
of estimating runoff is certainly an approximation, but it is consistent with the readily available
data (i.e., average monthly depth of precipitation, and number of precipitation events).
5-4
-------
TABLE 5-1. Runoff Curve Numbers for Hydrologic Soil-Cover
Complexes Under Average Conditions of Antecedent Moisture
LAND USE OR COVER
Pasture or Range
Meadow (permanent)
Woodlands (farm woodlots)
^
Farmsteads
Roads, dirt
Roads, hard-surface
TREATMENT
OR
PRACTICE
Contoured
Contoured
Contoured
HYDROLOGI
C
CONDITION
Poor
Fair
Good
Poor
Fair
Good
Good
Poor
Fair
Good
HYDROLOGIC
SOiL GROUP
A
68
49
39
47
25
6
30
45
36
25
59
72
74
B
79
69
61
67
59
35
58
66
60
55
74
82
84
C
86
79
74
81
75
70
71
77
73
70
82
87
90
D
89
84
80
88
83
79
78
83
79
77
86
89
92
Source:
U.S. Soil Conservation Service, 1972
5-5
-------
TABLE 5-2. Hydrologic Soil Group Descriptions
HYDROLOG
1C SOIL
GROUP
SOILS INCLUDED
(Low runoff potential). Soils having high infiltration rates even when
thoroughly wetted, consisting chiefly of sands or gravels that are deep
and well to excessively drained. These soils have a high rate of water
transmission.
B
Soils having moderate infiltration rates when thoroughly wetted, chiefly
moderately deep to deep, moderately well to well drained, with
moderately fine to moderately coarse textures. These soils have a
moderate rate of water transmission.
Soils having slow infiltration rates when thoroughly wetted, chiefly with
a layer that impedes the downward movement of water or of moderately
fine to fine texture and a slow infiltration rate. These soils have a slow
rate of water transmission.
(High runoff potential). Soils having very slow infiltration rates when
thoroughly wetted, chiefly clay soils with a high swelling potential; soils
with a high permanent water table; soils with a clay pan or clay layer at
or near the surface; and shallow soils over nearly impervious materials.
These soils have a very slow rate of water transmission.
5-6
-------
• TABLE 5-3. Classification of Vegetative Covers by Their
Hydrologic Properties
VEGETATIVE
COVER
Crop rotation
Native pasture
Permanent
Meadow
Woodlands
s
HYDROLOGIC
CONDITION
Poor
Good
Poor
Fair
Good
Poor
Fair
Good
SOILS INCLUDED
Contain a high proportion of row crops, small
grains, and Mow.
Contain a high proportion of alfalfa and
grasses.
Heavily grazed or having plant cover on less
than 50 percent of the area.
Moderately grazed; 50-75 percent plant cover.
Lightly grazed; more than 75 percent plant
cover.
100 percent grass cover.
Heavily grazed or regularly burned so that
litter, small trees, and brush are destroyed.
Grazed but not burned; there may be some
litter.
Protected from grazing so that litter and
shrubs cover the soil.
Source: U.S. Soil Conservation Service, 1972
5-7
-------
TABLE 5-4. Rainfall Limits for Estimating
Antecedent Moisture Conditions
ANTECEDENT
MOISTURE CONDITION
CLASS
I
n
m
5-DAY TOTAL ANTECEDENT RAINFALL (INCHES)
DORMANT SEASON
Less than 0.5
0.5-1.1
Over 1.1
GROWING SEASON
Less than 1.4
1.4-2.1
Over 2.1
Source: U.S. Soil Conservation Service, 1972
5.3
ESTIMATING EVAPOTRANSPIRATION
The evapotranspiration term (ET) is used to represent the combined effect of
evaporation of water from the soil and transpiration of water by plants. Estimates of ET are first
calculated in terms of the Potential Evapotranspiration (PET) and then adjusted to estimate the
Actual Evapotranspiration (AET) based on available soil moisture. Two methods have been
incorporated into the multimedia model for estimating the PET based on available data. The
first method, which is preferred, uses measured or estimated pan evaporation data, while the
second uses the Thornthwaite method based on monthly average temperature data to calculate
thePJET. ""
Definition of parameters:
AET
AW
AWC
Ep
ET
actual evapotranspiration (cm/month),
available soil moisture (cm),
available water capacity of the soil (cm),
the monthly pan evapotranspiration (cm/month),
the monthly evapotranspiration (cm/month),
5-8
-------
Ea SB the unadjusted PET (cm/month),
f() = a functional relationship relating AET to PET (dimensionless),
/ = annual heat index (dimensionless),
k = pan correction factor dependent upon vegetation cover and type (dimensionless),
PET = potential evapotranspiration (cm/month),
Rz = depth of root zone (cm),
7^ = mean monthly air temperature (°C),
0 = volumetric water content of soil (dimensionless),
8W = volumetric water content of soil at wilting point (dimensionless),
Bf = volumetric water content of soil at field capacity (dimensionless). This
parameter is a loose definition and cannot be directly measured readily.
5.3.1 Pan Evaporation Data
Evaporation is frequently measured at weather stations and is referred to as Pan
Evaporation (Ep). In addition, maps of annual average Ep have been developed showing
estimated isopleths of constant Ep throughout the United States (e.g., see Dunne and Leopold,
1978, pg. 102).
Local estimates of Ep must then be converted to PET based on:
PET *k-Ep
*
The pan correction factor, k, typically ranges between 0.6 and 0.8 (EPA, 1984a).
5.3.2 Thornthwaite Method
If actual measurements of evaporation are not available from local weather stations,
then an empirical method must be used. The Thornthwaite method is an alternative method for
estimating the PET which consists of the following empirical formulas:
5-9
-------
Ea - 1.6
10 r,
Eq. (5-4)
12
E
i'l
r«
T
1.5
Eq. (5-5)
where i = subscript representing month i, and
a » 0.4P * 0.0179(1) - 0.0000771ft)2 + 0.000000675(I/
The unadjusted PET value are adjusted for the number of days per month and the latitude, which
affects the length of the day. These adjustment factors are based on tables present in the
literature (Dunne & Leopold 1978) and typically range from 0.5 to 1.5.
5.3.3 Actual Evapotranspiration
Given the Potential Evapotranspiration (PET) calculated by one of the two previous
methods, the Actual Evapotranspiration (AET) is estimated using the following relationship (from
Dunne and Leopold 1978):
Eq. (5-7)
The available soil moisture is estimated as the rooting depth of vegetation, Rzt times
the difference between the available soil moisture, 6, and the soil moisture at the wilting point
6-.
5-10
-------
AW = Rt - (6 - $w) Eq- (5-8)
The available water capacity of the soil is estimated as the rooting depth of vegetation
times the difference between the moisture content at field capacity, 9P and at the wilting point
AWC = Rz . • (0, - ew) Eq. (5-9)
Average parameters of various soil properties are presented by Rawls et al. (1982)
for the 11 basic soil textural classes, these values are listed in Chapter 9.
The functional relationship between the available soil moisture and the available water
capacity, f (AW/ AWC), may depend on the vegetative cover and the soil texture and various
relationships have been suggested. The model uses a simple linear relation suggested by
Thornthwaite and Mather (1955) such tha.tf(AW/AWC) = AW/AWC.
5.4 TRAVEL THROUGH THE PARTIALLY SATURATED ZONE
The equations used to represent fluid flow and contaminant transport through the
partially saturated zone are based on an assumption of steady state, one dimensional flow
conditions and a unit hydraulic gradient (i.e. , only the influence of gravity is considered). The
process evaluated in the analysis are advection due to the net annual recharge, retardation due
to contaminant/soil interactions based on the soil-water partition coefficient, degradation
processes represented as first order decay, and dispersion within the partially saturated zone.
The water velocity within the partially saturated zone is assumed to be constant vertically,
therefore steady-state conditions are applied. Parameters controlling water movement are
described next and are followed by a description of the numerical model that is used to simulate
chemical dispersion. The functional relationships for representing the hydraulic properties of
soils are from Campbell (1974), and the empirical parameters for representing various soil
5-11
-------
classifications may be obtained from EPA (1987) and references cited therein (see Chapter 10
for a table of values). .
Definition of parameters!
b = soil specific exponent parameter representing the moisture retention relationship
(dimensionless),
K, = the saturated hydraulic conductivity of the specific soil type (cm/yr),
9, = saturated volumetric water content of the specific soil type (dimensionless),
B = volumetric water content of the specific soil type under steady recharge conditions
(dimensionless),
q — the annual average recharge rate (cm/yr),
qm — monthly recharge rate (m/month),
R
-------
Retardation as used in both the partially saturated zone calculations described in this
section and the ground-water transport model (Section 5.7) is an important concept that deserves
further discussion. It is generally accepted that dissolved chemicals migrating through a porous
medium often travel at a slower velocity than the water in which they are dissolved. This
phenomenon, described as chemical retardation, is controlled by the process of partitioning
between the liquid and solid phases. As a chemical adsorbs to the grains of a porous medium,
its transport in the dissolved phase is delayed. As a simple representation of this partitioning
process, that is valid for rapid, reversible partitioning (relative to water movement), is
K4--L Eq. (5-12)
For chemicals where this approach is valid, the contaminant velocity relative to the
pore water velocity is
Ve = VR4 Eq. (5-13)
where Rd is the retardation factor, which is defined as
^.K, Eq. (5-14)
6
5.5.1 Summary of the Partially Saturated Zone Numerical Model
Contaminant movement through the partially saturated zone is simulated in MMSOILs
in a routine called PSAT using a finite element solution for chemical transport called VADOFT.
VADOFT is a subset of the EPA model RUSTIC (EPA, 1989a) that simulates solute transport
processes including hydrodynamic dispersion, advective flow, linear equilibrium adsorption and
5-13
-------
first-order decay. VADOPT uses the annual time series of leachate releases from the WMU and
generates a delayed time series that represents the leachate flux at the water table.
Hie documentation for RUSTIC suggests that the minimum element size required to r
simulate the unsaturated zone is: i
— * 4 Eq. (5-15)
where
Az = element size (cm)
ap*a — dispersivity (cm)
Hie expression that is used for dispersivity within the partially saturated zone is:
°W - 2 + 0-022L, Eq- (5-16)
where
Lm- = the total thickness of the partially saturated zone (m),
since the number of elements, NP, is:
NP = i- Eq. (5-17)
Combining these three equations gives:
5-14
-------
* Eq. (5-18)
0 H
o + .........
n.4
which converges to approximately 12 as the thickness of the unsaturated zone increases.
The number of nodes per layer is calculated by apportioning NP between the layers
based on the thickness of each layer relative to the total thickness of the partially saturated zone.
A minimum of two nodes are used for any layer.
Time is discretized in PSAT using the Courant number criteria recommended in the
RUSTIC documentation:
At < '-' Eq. (5-19)
The smallest values of Az, 6, S and Rd and qm the largest value of V define the
minimum time step required for a simulation. The minimum time step for the time period when
the WMU is active is set to either one month or three months depending on the Courant number
criteria. This is done to ensure that the time step is sufficiently small to resolve changes in the
leachate flux from the WMU, which can vary annually. After the active WMU period the time
step reverts to the value calculated in Equation 5-18.
Transport calculations in PSAT continue for at least as long as the active WMU
period and then terminate based on the time frame of interest for the ground-water transport
c calculations or the calculated total travel time through the partially saturated zone, whichever
is shorter.
5-15
-------
Chemical decay in the partially saturated zone is only calculated for the fraction of
the travel time that the chemical is in the aqueous phase; no decay is calculated while the'
chemical is adsorbed to the soil. This was implemented by scaling the first order decay
coefficient by the retardation factor. As a consequence, the amount of decay applied in the
i
partially saturated zone is minimized. This is a conservative approach with respect to protecting -
human health and the environment. /•
The output from PSAT is a time series of concentrations and mass fluxes directly
above the water table. These are tabulated on a annual time-step during the active period of the
WMU and on a time-step that depends on the recharge rate for the period after the WMU is
closed. It should be noted that the conditions at the end of the WMU simulation period are
assumed to hold for future releases. Thus, the model simulates steady-state conditions if
leachate is still being released at the end of the simulation period.
5.6 MIXING ZONE CALCULATIONS
The conceptual link between the partially saturated zone (PSAT) and saturated zone
(GWPTH) models is a mixing zone in the aquifer directly under the WMU, where contaminated
recharge mixes with regionally flowing ground water (Figure 5-1). The mixing zone is used as
a volumetric source of contamination for the ground-water calculations.
As contaminated water enters the water aquifer it is diluted by the ambient ground
water in the aquifer. The amount of dilution is controlled by the ratio of the horizontal water
flux to the vertical water flux through the mixing zone:
. K'J-Ay-Az
—^
The cross sectional area through which the vadose zone water moves is equal to the
area of the WMU. The cross sectional area through which ground water leaves the mixing zone
5-16
-7 f
-------
is equal to the width of the WMU times the depth of .the mixing zone. This depth, called the
depth of penetration, is calculated by the model. A derivation of Equation 5-20 can be found
. in Appendix A of EPA, 1990.
MMSOELs uses a method for calculating the depth of penetration based on a peer
reviewed model (CEPACML) developed by EPA's Office of Solid Waste. Two factors are
assumed to contribute to the depth of penetration: an advective component and a dispersive
component:
Az
1 - exp.
l2avAx
Eq. (5-21)
When the pore water velocity in the aquifer is relatively small compared with the
recharge rate, the calculated mixing zone thickness may be insufficient to transport all of the
available water and a chemical could be concentrated within the mixing zone. To avoid this
unrealistic scenario, the thickness of the mixing zone is re-defined as follows:
Az
_ q" I IPO |
-Ax
Eq. (5-22)
If the depth of penetration calculated in this fashion is thicker than the extent of the
aquifer, a further adjustment is required. In this scenario, Az is set equal to the aquifer
thickness and the horizontal extent of the mixing zone is increased to maintain a mass balance.
Spreading the source laterally can be thought of as a mounding effect although the local
hydraulic gradient is not modified.
5-17
-------
5.7 GROUND-WATER FATE AND TRANSPORT MODEL
The model used to represent the fate and transport of a contaminant in an aquifer is
based on a quasi-analytical solution to the advection dispersion equation incorporating retardation
and first order decay. The solution represents the case where a time-variant mass flux into a
three dimensional volume results in a uniform concentration at the volume boundary. The
source is a three dimensional volume whose strength is specified by the mass flux that reaches
the water table after transport through the unsaturated zone. The contaminant is transported
downgradient by the regional flow field and allowed to disperse as a function of the dispersion
coefficients in the x, v, and z coordinate directions.
Definition of parameters:
Ax = length of die volume source in the x direction (m),
Ay = width of the volume source in the y direction (m),
Az = depth of the volume source in the z direction (m),
Cp, = concentration in ground water at exposure point (mg/1),
Dx = dispersion coefficient in the x direction (mVyr),
Dy = dispersion coefficient in the y direction (nrVyr),
Df = dispersion coefficient in the z direction (m2/yr),
Mw = mass flux (mg/yr),
kL = first order decay constant (1/yr),
Rd = retardation factor (dimensionless),
Ty - time (yr),
t = time (dummy integration variable) (yr),
ug = regional pore velocity in the x direction (m/yr),
u = ug/Jf, (m/yr),
9 = porosity (dimensionless),
D, * - DJR, (m'/yr),
Dy = D^ (mVyr),
Z>; = D& (mVyr).
5-18
-------
XSF = scale dependent dispersivity factor,
x a* down gradient distance of observation point (m),
y = lateral distance from plume center of observation point (m),
z = depth of observation point (m),
The processes of mechanical dispersion and molecular diffusivity are lumped in the
ground-water transport model. Based on the results of field studies that have indicated that
dispersion is dependent on the travel distance, scale dependent dispersion has been implemented
in the model (Gelhar, et al., 1985). In this approach, multipliers are used to establish the
amount of dispersion that is applied in the contaminant transport calculations in the x, y, and z
directions independently. For example, the dispersion coefficient in the or direction is calculated
as:
Eq. (5-23)
Dt =
Similar equations apply for dispersion in the y and z directions. The factors used in
the x, y, and z directions are often set to 0.1, 0.02, and 0.01 respectively.
The quasi-analytical solution is of the form (EPA, 1986b):
C (X,Y,Z,Ty) = **">/R" f * e-W-> F(X,Y,Z,Ty) dt Eq. (5-24)
where:
5-19
X/
-------
X + Ax/2 -u' (Ty-t)
^4 D; (Ty-t) \
erf
erf
Y+Ay/2
^4Dy (Ty-t)
Z+Az/2
^4Df' CTyt)
-erf
-erf
-erf
"X -Ax/2 -u' (Ty-t)"
J4D,' (Ty-t)
Y-Ay/2
J4D; (Ty-t)
Z-Az/2
-------
6.0 SOIL EROSION PATHWAY
This pathway is used to evaluate contaminant movement to off site soils through two
mechanisms: soil erosion and atmospheric deposition. While atmospheric deposition is obviously
not related to soil erosion processes, the effect of atmospheric deposition is included at this point
in the methodology since it is a mechanism by which off site soils may become contaminated.
This chapter describes the models used to represent: soil erosion from a site, delivery fraction
of eroded soil and mixing with off site soils, and soil contamination due to atmospheric
deposition.
6.1 EROSION OF CONTAMINATED SOIL
Contaminants adsorbed to surface soil may be transported off site through the process of
soil erosion. In the multimedia model, the erosion loss term is used for estimating the
accumulation of a chemical in off site soils (agricultural and/or residential exposure points) near
the waste site. The erosion loss term is also used for estimating the mass of contaminant
adsorbed to soil, which forms a source term for the steady-state transport analysis in the
receiving water body (see Chapter 7).
Definition of parameters:
Y(s) = sediment yield (metric tons/hectare/year),
RF = the rainfall factor, expressing the erosion potential of average annual rainfall in
the locality (metric tons-meter/ha-hour),
KF = soil credibility factor (metric tons/ha/unit RF),
LS = length - slope factor (dimensionless),
CF = cover factor, 1.0 for bare soil (dimensionless),
PF = erosion control practice factor, 1.0 for uncontrolled waste site (dimensionless).
The Universal Soil Loss Equation (USLE), developed by Wischmeier and Smith
(1965), is used for estimating the annual soil erosion loss rate. The formula for the USLE is:
6-1
-------
Y(s) *2.24'RF'KF-LS*CF'PF Eq. (6-1)
This estimate of annual soil loss, Y(s), is per unit hectare surface area of the site and hence must
be multiplied by the area to get the annual mass of soil loss. This value is the soil loss rate from
the site, not the soil delivery rate to an off site point or some nearby water body. The next
section describes the estimation of mass of soil delivered to an off site point.
6.2 OFF-SITE SOIL CONTAMINATION DUE TO EROSION
Definition of parameters;
C, = 'soil adsorbed contaminant concentration level (mg/kg),
Ca = soil concentration in off site field (mg/kg),
Z>, = soil delivery rate to off site field (kg/year),
Y(s) = sediment yield (metric tons/hectare/year),
A, ~ area of the waste site (hectare),
SDf = soil delivery fraction from waste site to agricultural field
(dimensionless),
XF3 = conversion factor (1000 kg/metric tons).
Ma} = mass of soil in the mixing depth of the field .(mixing depth x area x bulk density)
(kg).
X, = a loss rate coefficient representing chemical degradation and movement
through the root zone (I/year).
If an exposure point (agricultural and/or residential exposure point) is close to a waste
site and downgradient, then erosion of soil from the site may reach the off site point and
represent the major source of soil contamination. Under these circumstances, the fraction of soil
eroded from the waste site that is transported to the off site area (soil delivery fraction, SDf)
must be estimated. However, this parameter is difficult to estimate. Guidance may be found
in EPA (1988a), which suggests a worst case estimate would be a fraction of about 0.5 and
6-2
' . I
-------
more typical case of 0.1 for fields adjacent to and directly down gradient from a waste site.
These estimates are based on technical judgement rather than field data. This parameter is
highly site specific, if channeling and/or drainage patterns between the waste site and the
agricultural field divert all die surface runoff then the soil delivery fraction should be zero.
The soil delivery rate to the agricultural field, Dtt is estimated as:
D, = Y(s)-As> SDf- XF3 Eq. (6-2)
The chemical concentration in the soil of an off site field is estimated based on the
soil delivery rate, the area of the off site field, an assumed mixing depth, the sum of any
degradation/decay mechanisms along with mass loss due to soil erosion from the off site field
(represented by a first order decay coefficient), and the soil concentration at the waste site. The
mixing depth is assumed to be 10 cm which is an approximation used by EPA (1988a). The soil
concentration in the off site field under steady state conditions may then be estimated as:
_?£^ Eq. (6-3)
6.3 OFF-SITE SOIL CONTAMINATION DUE TO DEPOSITION
The chemical concentration in the soil of an off site field, Cat may also be the result
of atmospheric deposition. The chemical concentration in off site soils resulting from
atmospheric deposition is a function of the mass of chemical deposited and net loss rate of
chemical in the off site soil due to erosion, leaching through the root zone, and any
biodegradation. The results of the atmospheric pathway described in Chapter 3 (i.e.,
atmospheric concentration) are used in this section in order to estimate off site soil concentration
due to deposition.
6-3
-------
Definition of parameters: • ..
Ctt = atmospheric concentration (mg/m3),
Cm — soil concentration in agricultural field (mg/kg),
vd = deposition velocity (m/sec),
r — time period of plant exposure (sec),
P = density thickness of the root zone (bulk density of the soil times the
rooting depth) (kg/m2),
\ = a loss rate coefficient representing chemical degradation and movement
through the root zone (I/sec).
If atmospheric deposition is the primary means of soil contamination, the soil
concentration under conditions of a steady long term release where the initial concentration is
zero at time t=0 may be estimated as follows (Moghissi, et al., 1980):
C - Eq.«M)
P\
The soil loss rate coefficient, X,, is the sum of all soil losses due to leaching, soil erosion, and
chemical degradation. The equation only applies when the total period of atmospheric deposition
is long compared to both the length of the growing season and to the soil loss rate coefficient,
X,. This chemical concentration in the agricultural soil may be combined with any concentration
due to soil erosion (from Equation 6-3).
6-4
-------
7.0 SURFACE WATER PATHWAY
The surface water pathway evaluates contaminant entering one of two types of receiving
water bodies, a stream or a small lake. For contaminants entering a small lake, the source term
is the contaminated bed sediments resulting from the erosion of contaminated soil from an
adjacent waste site. The potential source terms incorporated in the model for contaminants
entering a stream include the erosion of contaminants adsorbed to the soil, inflow of runoff
containing dissolved contaminants, and the discharge of contaminated ground water into the
stream. The conceptual model for chemical transport to a stream and downstream to exposure
locations is shown in Figures 7-1 and 7-2.
7. 1 SOIL EROSION TO A STREAM
Contaminants adsorbed to surface soil may be transported to a nearby surface water body
through the process of soil erosion. In the multimedia model, the erosion of contaminant
adsorbed to soil forms a source term for the steady-state transport analysis in the receiving water
body. The erosion loss term is also used for estimating the accumulation of a chemical in off
site soils near the waste site (see Chapter 5).
Definition of parameters:
Y(s) = sediment yield (metric tons/hectare/year),
Sd = sediment delivery ratio (dimensionless),
Dd = die overland distance between the site and the receiving water body (meters),
Mer = contaminant mass loading rate due to erosion (g/year),
16^ = contaminant mass loading rate dissolved in runoff (g/year),
6 = porosity of surficial soil (dimensionless),
Kd - equilibrium partition coefficient (ml/g),
PB = bulk density of surficial soil (g/cm3)
C, = soil adsorbed contaminant concentration level (mg/kg),
A, - surface area of the site (hectares).
7-1
-------
. FIGURE 7-1. Conceptual model of surface water pathway.
7-2
-------
FIGURE 7-2. Conceptual model of stream contamination due to
ground-water inflows.
LAUD
OjjQMM.
VIEW
7-3
-------
The .Universal Soil Loss Equation (USLE), developed by Wischmeier and Smith
(196S), is used for estimating the annual soil erosion loss rate (see Chapter 5). A sediment
delivery ratio, S4, is used to represent the reduction in soil mass delivered to a water body as
the distance from the site to the river becomes larger (i.e., losses due to redeposidon between
the source area and the water body). Sd is estimated with the following equation (from EPA,
1987; converted from feet to meters):
0.77- (Dd)
-a 22
Eq. (7-1)
The estimate of annual average soil loss rate may be combined with a soil
contaminant concentration level to estimate the mass of contaminants entering the receiving
surface water. The annual average contaminant source term for the surface water transport
analysis is calculated as:
M,, *> C,-Y(s)-A,
Eq. (7-2)
7.2
RUNOFF TO A STREAM
The mass loading to a stream from runoff containing dissolved contaminant is
calculated by assuming that water running off the site is in chemical equilibrium with the
surficial soils. An equilibrium partitioning relationships between the soil and the runoff is used
to calculate the chemical concentration in the dissolved phase. The mass of dissolved
contaminant is expressed:
M,
Eq. (7-3)
7-4
-------
Hie sediment delivery ratio is included in equation 7-3 because it is assumed that the
runoff will contact contaminated soil that has been redeposited between the site and the stream.
The rate of flow in the stream is also adjusted by the flow of water from the site to the stream.
7.3 GROUND-WATER DISCHARGE TO A STREAM
The other potential source of stream contamination addressed in the model is the
discharge of contaminated ground water to a stream. The model user must discern whether or
not the ground-water aquifer discharges to the stream. This may be accomplished through the
evaluation of water table elevation data in the vicinity of the stream and water quality data in
the stream if no other significant chemical sources reach the stream. If the ground water
discharges to the stream, the model assumes that the entire contaminant plume is intercepted by
the stream. This may not be true if the source entering the ground water is a long distance from
the stream, however it is conservative in terms of estimating the stream concentration and is
useful for screening purposes.
The mass flux entering the stream through ground-water discharge is the integral of
the ground-water concentration at the stream times the Darcy velocity, integrated over the area
through which the discharge occurs. The ground-water concentration at the stream is calculated
on an annual basis using the computation methods described in Chapter 5. These calculations
allow for a first-order decay process to degrade contaminants in the ground-water environment
if this is appropriate for a particular chemical constituent.
7.4 FATE AND TRANSPORT ANALYSIS IN RIVERS
The two most important processes affecting the fate and transport of chemicals within
a river are the initial dilution and advection. Other attenuation processes that may affect the fate
of a contaminant include volatilization, sorption to suspended sediment, hydrolysis, photolysis,
and biodegradation. The simple fate and transport model represents all of these processes as a
single lumped first order decay process. Descriptions of these attenuation mechanisms along
7-5
-------
with methods for estimating an effective first order decay coefficient are described by Neely and
Blau (1985), EPA (1984b), EPA (1980a), and EPA (1979).
Definition of parameer:
Q
kL
x
u,
= concentration in river water (mg/1),
= total mass flux rate of contaminant entering the stream from soil erosion,
dissolved runoff, and ground-water discharge (g/year),
~ average flow rate within the stream (mVyear),
= lumped first order decay coefficient (I/year),
— downstream distance from the source location to the exposure point (m),
= average stream velocity (m/yr).
The analytical equation used to predict chemical concentration in a river follows:
tw
~Q~
•K)
Eq. (7-4)
This equation accounts for the initial dilution and a lumped first order decay. In the case of
runoff of eroded soil, Equation 7-4 assumes that the entire mass of contaminant entering the
stream in soil adsorbed form will be dissolved into solution. This is a conservative assumption
that may be addressed in more detail if data on the source term and suspended sediment
concentrations are available. The removal of contaminants from the water column due to settling
of sediments may be incorporated into the effective first order decay term (examples are
provided in EPA, 1984b). Equation 7-4 also assumes that the contaminant in the stream is
uniformly mixed across a cross-section.
7-6
-------
7.5 CHEMICAL FATE IN A SMALL LAKE.
The evaluation of organic chemical concentration in a small lake is based on a
sediment to water to air mass transfer model used by Hwang (1987). The model is based on
empirical mass transfer coefficients used to represent the transfer of organic chemicals from the
sediments to water and transfer from water to the air via volatilization.
Definition of parameters:
Cltd - concentration in bottom sediments (mg/kg),
Dtt = sediment dilution ratio, between 0 and 1 (dimensionless),
C, - waste site surface soil concentration (mg/kg),
CA, = dissolved chemical concentration in lake water (mg/1), —
kw = water-phase mass transfer coefficient (cm/hr),
ke = sediment-phase mass transfer coefficient (cm/hr),
kL = overall water to air mass transfer coefficient (cm/hr),
K& = sediment-to-water partition coefficient for lake sediments
(1 water/kg sediment),
CD = drag coefficient (0.00237 for wind speeds 4-12 m/s),
Vw = wind velocity at 10 meters above surface (cm/min),
h = average depth of water (cm),
F = average wind fetch for the lake, assumed to be the half width of the
lake (cm),
pa = . density of air (1.2 x 1048 g/cm3),
pw — density of water (1.0 g/cm3),
Dw = chemical diffusivity in water (cmVsec),
r, = diffusion path length below the water sediment interface (cm),
k, = liquid-phase mass transfer coefficient (cm/hr),
kt = gas-phase mass transfer coefficient (cm/hr),
H = Henry's law constant (atm-mVmol),
R = gas constant (8.2 x 10* atm-m3/mol-°K),
7-7
-------
7 - absolute temperature (°K),
'(COj) = liquid-phase mass transfer coefficient for CO2 (20 cm/hr),
k*(H) = as-phase mass transfer coefficient for H2O (3000 cm/hr),
MW = molecular weight of chemical (g/mol),
= molecular weight of CO2 (44 g/mol),
= molecular weight of H2O (18 g/mol).
Ot = lake sediment porosity (dimensionless),
The model requires estimation of chemical concentration in the lake sediments as the
key input parameter. The worst case would be a small pond entirely surrounded by a waste site
where all soil erosion and sediments entering the pond contained the surface soil concentration
of the waste site. Under this worst case condition, the chemical concentration in the bottom
sediments would approach the concentration in the surface soils. Note that this worst case
scenario deals only with sediment contamination arising from present levels of soil contamination
and does not address other potential sources of chemical entering the lake sediments. Many
cases exist where the concentrations in bottom sediments of a water body are higher than any
adjacent surface soils. This may be due to additional source terms contributing to the water
body, different loss rates for bottom sediments and surface soils, or higher partition coefficients
due to the different nature of the lake sediments (e.g., high organic matter content).
In this model a sediment dilution ratio, Du, is incorporated as a user-defined input
parameter to account for clean sediments entering the lake. A worst case would be a dilution
factor of 1 (i.e., all sediments entering the lake arise from the waste site). Conditions more
likely to be encountered would be where the chemical concentration in lake sediments are diluted
by clean sediments entering the lake from areas other than the waste site.
The user may estimate this sediment dilution fraction based on estimated sediment
loads (i.e., USLE calculations) from different areas on the perimeter of the lake, the relative
fraction of the lake perimeter where the waste site will contribute sediments, or perhaps the
relative fraction of the watershed area where the waste site contributes sediments. All of these
7-8
-------
estimation methods are dearly qualitative rather than quantitative and several orders of
magnitude of uncertainty could be expected unless specific data characterizing sediment
concentrations are available.
The chemical concentration in sediments is estimated as:
. (7-5)
The steady state model, by Hwang (1987), estimates water concentration based on the
sediment concentration as:
Eq. (7-6)
The three mass transfer coefficients are estimated as follows:
Water-phase mass transfer coefficient:
FMW"2 p.
Sediment-phase mass transfer coefficient:
Overall water to air mass transfer coefficient (Liss and Slater, 1974):
k ^3600^L. Eq. (7-8)
7-9 /
60
-------
1
_ _
k( Hkt
Eq. (7-9)
The liquid phase and gas phase mass transfer coefficients for the water air interface
may be estimated as:
*,
MWCO,
~~MW~
Eq. (7-10)
MW
0.5
Eq. (7-11)
The gas film coefficient, kf, calculated using Equation 7-11, states that the coefficient varies with
MW0-5. This relationship has been widely used because of the simplicity (Rathbun and Tai,
1986), however theoretical and experimental considerations indicate that the coefficient varies
with MW^gj (Cohen, 1989). In this methodology, the square root relationship of Equation 7-11
is used to be consistent with the references from which the equation was cited.
7-10
-------
8.0 FOOD CHAIN BIOACCUMULATTON PATHWAY
The food chain bioaccumulation pathway uses the transport of contaminants from the site
via other environmental transport pathways as the source term(s). Examples of environmental
transport pathways which may serve as the source terms for the food chain pathway include
atmospheric transport and deposition, soil erosion, and migration within ground water and
subsequent use for irrigation. Based on these source terms, the food chain pathway examines
the accumulation of a chemical within fish, terrestrial plants, and cattle. Simple models of
bioaccumulation using bioconcentration factors (BCF) and transfer factors (f) are used for this
pathway. The bioconcentration factors are used to represent the partitioning of a chemical
between: 1) water and fish, 2) edible parts of terrestrial plants and soil, and 3) root vegetables
and soil moisture. The transfer factors are used to represent the uptake of chemical by animals
as a function of the mass of chemical ingested in feed and water. —
A considerable amount of BCF data for different chemicals are available for fish and
some empirical correlations based on limited available data have been derived for accumulation
in above ground crops, root vegetables, and animals. All of the empirical correlations are based
on the logarithms of the accumulation parameters and various physical properties of different
chemicals.
In general, the available data exhibit a degree of scatter around the correlation
relationships for the different accumulation parameters. For each of the correlation equations,
the data from which the equations are derived are plotted (where available) along with the curve
representing the empirical estimation. These plots are important for the model user as they
represent a range of parameters that should be evaluated for the different accumulation
parameters. Some of the parameters estimated using the correlation relationships will contain
uncertainty that may be greater than one order of magnitude, and this degree of uncertainty must
be incorporated in the analysis. The available data for plants and particularly for animals are
very limited and in some cases contradictory. The correlation relationships are useful for
screening purposes and should be used to estimate a range of parameter values, which may cover
several orders of magnitude.
8-1
C>
-------
Definition of parameters!
At = area of the waste site (hectare),
BCF0 = bioconcentration factor for species in experimental study (ml/g fish),
BCFf — bioconcentration factor for fish (ml/g fish),
BCFV = bioconcentration factor for species of concern (ml/g fish),
BF = bioconcentration in beef fat [(mg/kg fat)/(mg/kg feed)],
BSF = sediment-to-fish partition coefficient [(mg/kg fish)/(mg/kg soil)],
Bv = soil-to-plant concentration factor (kg soil/kg dry plant),
Ca = total atmospheric concentration, vapor and participates (mg/m3),
Cw = soil concentration in agricultural field (mg/kg),
CJ. = chemical concentration in fish (mg/kg),
Cfl -= chemical concentration in cattle (mg/kg), —
<^2 = chemical concentration in milk (mg/kg),
Cfrf = chemical concentration in feed used for cattle (mg/kg),
Cltd = chemical concentration in sediments (mg/kg),
Cv = chemical concentration in vegetables (mg/kg wet plant),
C^ = chemical concentration on vegetation due to foliar deposition (mg/kg wet plant),
Cw = chemical concentration in vegetation due to uptake from the soil (mg/kg wet
plant),
Cw = chemical concentration in water (mg/1),
D, = soil delivery rate to agricultural field (kg/year),
Ffa ± fraction of fat in a food product (g/g),
Fj = feed-to-meat transfer factor for cattle (day/kg),
F2 = feed-to-milk transfer factor for cattle (day/kg),
f^ =B fraction of wet plant remaining as dry material (g/g),
LC9 - lipid content for species of concern (g/g),
LCa - lipid content for species in experimental study (g/g),
Mf = mass of soil in the mixing depth of the field (mixing depth x area x bulk density)
(kg),
8-2
-------
P = density thickness of the root zone (bulk density of the soil x rooting depth)
(kg/m2),
PI = soil-to-meat partition coefficient [(mg/kg meat)/(mg/kg soil)],
P2 = soil-to-milk partition coefficient [(mg/kg milk)/(mg/kg soil)],
Qf = consumption rate of feed for cattle (kg/day),
Qt = consumption rate of soil for cattle (kg/day),
Qw = consumption rate of water for cattle (I/day),
r = fraction of flux captured by vegetation (dimensionless),
RCF = root concentration factor for vegetables (mL/g wet root),
SDf = soil delivery fraction from waste site to agricultural field (dimensionless),
t = time period of plant exposure (sec),
v4 - deposition velocity (m/sec),
XF3 = conversion factor (1000 kg/metric tons), —•
Y(s) — sediment yield (metric tons/hectare/year),
YV SB vegetative density (kg/m2),
\, = a loss rate coefficient representing chemical degradation and movement through
the root zone (I/sec),
Xw = weathering loss rate coefficient (I/sec).
8.1 BIOCONCENTRATION IN FISH
Bioconcentration of chemicals in fish is thought to be primarily due to intake through
the gills rather than through food, although there is evidence that biomagnification through
aquatic food chains can also be important under certain environmental conditions. Based on
intake through the gills, the concentration of a chemical in fish may be directly related to the
water concentration as follows:
Cf=BCFf*Cw Eq. (8-1)
8-3
-------
Data bases are available which contain measured or estimated values ofBCFs. These
include:
HSDB (Hazardous Substance Data Base): HSDB (sometimes referred
to as TOXNET) is a file maintained and reviewed by the National Library
of Medicine which is stored on the National Library of Medicine's
Toxicology Data Network (TOXNET). The data base is derived primarily
from a core set of monographs and standard texts, although there is some
information derived from government documents, technical reports, and
the primary journal literature. The HSDB is updated monthly; however,
since only a small portion of the data base is derived from journals, the
data base will not necessarily contain the most current information.
CHEMFATE: CHEMFATE is one of a set of data bases maintained by
the Syracuse Research Corporation (Merrill Lane, Syracuse, New York
13210-4080; Telephone: 315-425-5100). This data base was assembled
in the early part of 1986, and has had a modest level of update.
DATALOG: DATALOG is a bibliographic data base which is also
maintained by the Syracuse Research Corporation. Quantitative values are
not given; the citation must be investigated. The advantage of this data
base is that it is updated weekly.
CIS ENVIROFATE: ENVIROFATE is one of three data bases
containing information on bioconcentration factors initiated by the
Environmental Protection Agency and now operated by Chemical
Information Systems (CIS). ENVIROFATE is concerned with chemical
fate and reports BCF data only for chemicals whose production exceeds
one million pounds per year. Stated with the BCF is an evaluation of the
reliability of the study from which the BCF was derived. None of the
three data bases are scheduled for an update in the immediate future,
8-4
-------
J
although CIS has stated that one is being planned. ENVIROFATE was last
updated in July of 1985.
CIS AQUIRE: AQUIRE (AQUatic Information REtrieval) contains
information on chemicals and their interaction with aquatic organisms.
However, data on aquatic mammals, birds, and bacteria are not included.
AQUIRE also evaluates the reliability of the study from which the BCF
was derived. This data base is regularly updated and may be accessed via
EPA's telecommunication network.
CIS CESARS: CESARS is an extremely detailed data base, containing
185 fields, however, it is limited to chemicals found in the Great Lakes.
The data base is primarily concerned with aquatic organisms. The last —
. update occurred in February, 1985.
Kenaga and Goring (1980) note that BCF values from different species of fish vary
depending on percent lipid content. An experimental measurement can be adjusted on the basis
of the percent lipid content in the species of concern:
BCFV = BCF"'LC* Eq. (8-2)
Empirical estimation techniques have been developed based on correlations with
physical properties of various chemicals (i.e., octanol-water partition coefficient, water
solubility). A thorough review of empirical estimation techniques for BCFf is presented by
Bysshe (1982). One of the suggested estimation techniques based on correlations with octanol-
water partition coefficients is:
logBCFj = 0.76-log K^ -0.23 Eq. (8-3)
8-5
-------
Octroi-water partition coefficients may also be found in the HSDB, CHEMFATE,
DATALOG, and CESARS data bases. While experimental data are preferable to calculated
data, it should be remembered that there are also large uncertainties associated with experimental
data. Experimental partition coefficients were found to vary by three orders of magnitude for
one chemical (Garten and Trabalka, 1983). Variation between experimental partition coefficients
of one order of magnitude is not uncommon. Similar if not greater variation can be found
between experimental measurements of BCFs.
In determining the appropriate value to use when there are multiple experimental
determinations of the BCFa, the modeler should carefully examine the different experiments to
see if one is most analogous to the field site under consideration.. Flow-through bioassays,
where water with a constant concentration of a chemical flows continuously through a fish tank,
provide a good determination of relative BCFs of chemicals. All conditions can be controlled
in this type of experiment. The greatest concern with this type of experiment is whether or not
equilibrium between chemical and fish is reached.
Although all experimental parameters can be controlled with a flow-through bioassay,
ambient conditions cannot necessarily be simulated. Model ecosystems have been used to
attempt to remedy this problem. While model ecosystems have the potential of providing an
accurate simulation, they still have the problem of not accurately representing field parameters
such as temperature, pH, percent dissolved oxygen, etc. According to Bysshe (1982), these
experiments are not as conservative as flow-through bioassays.
A third method of determining BCFs is to make experimental measurements in an
actual field situation. The difficulty with this type of determination is that fish do not
necessarily remain in one location where there is one concentration of a given chemical. Factors
such as temperature and total dissolved oxygen are not constant. Of course, these uncertainties
will apply to the situation where the model is being applied, therefore, these studies can be quite
useful if the situation of concern is similar to the study conditions. For instance, if one
examines only laboratory studies, it will become readily apparent that water is a much more
important contributor to bioconcentration than food. According to Bysshe (1982), environmental
8-6
-------
studies have not always borne this conclusion out She points out that very high measured BCFs
in the Great Lakes may well be due to food rather than water concentrations.
In summary, if one is interested in a qualitative determination of the bioconcentration
potential (i.e., determining whether the chemical bioconcentrates or not), flow-through bioassays
should provide an adequate answer. However, if one is interested in the quantitative
bioconcentration potential of a compound, it is important to examine the experimental studies
to determine which ones are most relevant.
This methodology also incorporates an alternative algorithm for estimating
accumulation in fish based on the sediment concentration and a sediment-to-fish partition
coefficient. This algorithm is included as an alternative to Equation 8-1 for cases where water
concentration is not the dominant mechanism for accumulation in fish and data are available far
estimating the sediment concentration and a sediment-to-fish partition coefficient. This
concentration in fish may be estimated as:
Eq. (8-4)
This algorithm is an alternative to Equation 8-1 and only one of the two methods, Equation 8-1
or Equation 8-4, may be used for a given scenario.
8.2 BIOACCUMULATION IN CATTLE AND DAIRY PRODUCTS
The concentration of a chemical in the meat and milk of cattle is estimated using
transfer factors combined with the mass of chemical intake by the cattle based on contaminated
feed, soil, and water. The chemical concentration in the meat, Cfl, and milk, Cflt of cows are
estimated as follows:
8-7
-------
• • . •=•>
Hie transfer coefficients in Equations 8-5 and 8-6 represent the ratio of chemical concentration
in the animal (in meat and milk) to the mass of chemical consumed per day. Transfer
coefficients are critical parameters and should ideally be obtained from controlled animal feeding
studies based on long term exposures. This may not be possible for some studies in which case
all available sources of data should be reviewed to identify appropriate ranges of transfer
coefficients.
In the absence of chemical-specific data characterizing the transfer coefficient, an
empirical method for estimating it may be necessary. Kenaga (1980) presents bioconcentration
factors in beef rat (BF) for a variety of different chemicals. These data are used to develop
correlation relationships between various chemical properties (i.e., water solubility, octanol-
water partition coefficient) and bioconcentration in beef fat.
Modifications of the correlation relationships derived by Kenaga (1980) have been
used in some exposure assessment models (Whelan et al., 1987, and Holton et al., 1984) and
have also received some stiff criticism (Garten and Trabalka, 1983, and EPA 1986c). Once a
bioconcentration factor in beef fat is obtained via review of feeding studies or empirical
estimation, this value must be converted to transfer factors, F, and F2. This is calculated based
on the fraction of rat (FjJ in the food product and the consumption rate of feed by the animal
(Qf). If data are available for bioconcentration in fat (BF) the transfer factors may be estimated
as follows:
BF«F
F, « **~ Eq. (8-7)
8-8
-------
Eq. (g-8)
More recently a somewhat larger data set of chemical accumulation in the meat and
milk of cattle and cows has been compiled by Travis and Arms (1988). In this paper the authors
use some of the data presented by Kenaga (1980) and a variety of additional sources to estimate
correlation relationships for the meat and milk transfer factors, F, and F2, based on the octanol-
water partition coefficient of 36 different organic chemicals. It should be noted that the Travis
and Arms paper is subject to the same sources of criticism as the Kenaga paper. The
bioconcentration factor (BF) is converted to transfer factors based on the assumptions that meat
is 25 percent fat, whole milk is 3.68 percent fat, the dry feed ingestion rate for lactating cows
is 16 kg/day and 8 kg/day for nonlactating cattle and cows. The two correlation relationships,
which have correlation coefficients of 0.81 and 0.74 respectively are:
j m -7.6 + logK^ Iq. (8-9)
logF2 = -8.1 + logK^ Eq. (8-10)
The data used to develop the correlations with the octanol-water partition coefficient and the
derived correlation equations are presented in Figures 8-1 and 8-2. The data indicate that the
correlation relationships provide an estimate of the transfer factors within about two orders of
magnitude for the chemical data set on which it is derived.
The methodology also incorporates an alternative method for estimating chemical
concentration in meat and/or milk based solely on the soil concentration and a soil-to-meat
and/or soil-to-milk partition coefficient. The concentration in meat and milk may be estimated
as follows:
8-9
-------
FIGURE 8-1. Uptake factor for beef versus octanol-water
partition coefficient.
Source: Travis and Arms (1988)
Transfer Factor in Beef (8 kg/day feed, 25s fat)
-1-
-2-
,£
oa
u
3
3-
-4-
-5-
-6-
-7-
-8
T
1
1>
345
LOG (Kow)
8-10
-------
FIGURE 8-2. Uptake factor for milk versus octanol-water
partition coefficient.
Source: Travis and Arms (1988)
Transfer Factor in Milk (16 kg/day feed, 3.68* fat)
-In
-2H
-3H
S
CQ
U
3
-4H
-5H
-6H
-7
• •
T
3
"T
4
LOG (Kow)
6
8-11
-------
C, •Pi-C/ Eq. (8-11)
C^-jyC, Eq. (8-12)
These equations are provided as alternatives to Equations 8-5 and 8-6 (i.e., only one of the two
pairs of equations, Equations 8-5 and 8-6 or Equations 8-11 and 8-12, are used for a given
scenario).
8.3 BIOACCUMULATION IN PLANTS AND VEGETABLES
The accumulation of chemicals in plants and vegetables may be represented through
two mechanisms, atmospheric deposition onto the plant and uptake through the root system. All
vegetables are classified as either an above ground crop (e.g., lettuce, com etc.) or a root crop
(e.g., carrots, potatoes, etc.). The distinction is made based on the available estimation
techniques for predicting the accumulation within different parts of a plant.
An area of current ongoing research is the transfer of organic chemicals from
contaminated soil through volatilization from soil and direct uptake from the atmosphere by
aerial plant parts. This is a new area of research and currently available data are limited. As
more data become available in the future, this is a pathway that should be carefully examined
that is not currently addressed in this methodology.
For above ground crops, two basic algorithms are incorporated for estimating
chemical concentration within the plant. The models used to represent chemical accumulation
in plants are based on algorithms described by Moghissi et al. (1980). Various forms of these
algorithms have been used in different multimedia models by Whelan et al. (1987), Travis et al.
(1986), and Holton et al. (1984). It is important to emphasize that while these food chain
algorithms have been used in several multimedia models and represent the current "state of the
8-12
-------
8.3.1 Deposition on Plants
The concentration in a plant due to deposition, C-t is represented below (Moghissi
etal. 1980):
The weathering loss coefficient, Xw, is used to represent the rate at which particulate matter is
removed from plants due to environmental factors (i.e. , wind and water). Ranges of values for
a half-life due to weathering reported by references in Baes et al. (1984) range from 2.8 to 34
days and a "somewhat arbitrary value" of 14 days is suggested. This corresponds to a weathering
loss coefficient, X», of 5.73 x 10"7 (I/seconds). It should be noted that the studies were
performed using inorganic materials. If there is site contamination resulting from pesticides
formulated to a large degree with organic materials (as an example), then this result for Xw
should be used with caution. The value for the interception fraction r has been estimated by
Baes et al. (1984) for various types of vegetation. For pasture grasses and hay, an empirical
relationship is presented:
r = 1 - e'i88r" EQ- <8-14>
where Yn is the productivity of pasture grass in kg/m2. Pasture grasses are the only type of
vegetation upon which literature references are readily available concerning the interception
fraction. Baes has generated interception fraction values for other types of vegetation based on
"typical" planting patterns. The values suggested by Baes are listed in Table 8.1.
8-13
-------
TABLE 8-1. Interception Fraction for Various Types of Vegetation
(from Baes et aL, 1984)
VEGETATION
Leafy Vegetables (Lettuce, spinach, broccoli, celery
cauliflower, etc.)
Silage (Com, sorghum)
Exposed Produce (Non-citrus fruits, berries, and field crops
such as cucumber, asparagus, tomato, squash, and eggplant)
r
.30
.44
.052
Source: Baes et al., 1984
8.3.2 Plant Uptake from Contaminated Soil -
The chemical concentration in vegetation (on a wet weight basis) due to uptake from
soil, Cw, is represented by a soil-to-plant concentration factor, Bv. The relationship is
represented below:
C -B-C./L Eq.(8-15)
A typical value for the fraction of wet plant remaining as dry plant, f^ is IS percent. Various
empirical estimation techniques have been proposed for predicting the soil-to-plant concentration
factor based on the soil-water partition coefficient (Baes, 1982) and the octanol-water partition
coefficient (Holton et al., 1984 and Travis and Arms, 1988). The relationship presented by Baes
is from data for metals. The regression relationship presented by Holton et al. (1984) is based
on the original work by Baes (1982) and an additional regression relationship relating the soil-
water partition coefficient to the octanol-water partition coefficient. The relationship for metals
by Baes is (converted from natural to base ten logarithms) has a correlation coefficients of 0.88
and is expressed as:
8-14
tfit
-------
logB¥ « 1.54 - LiS'logK, Eq. (8-16)
This relationship is presented for accumulation in plants on a dry weight basis. The data
presented by Baes (1982) are shown in Figure 8-3 along with the correlation relationship of
Equation 8-16. Figure 8-3 illustrates a fair degree of scatter about the correlation estimate, but
over 90 percent of the data on which the correlation was based, fall within one order of
magnitude of the correlation estimate.
For organic chemicals, the data presented by Travis and Arms (1988) appear to be
the largest available compilation of bioconcentration factors for vegetation. The regression
equation relating bioconcentration in vegetation to chemical octanol-water partition coefficient
presented by Travis and Arms (1988) is as follows: ~"
logBv * 1.588 - 0.578•logK^ Eq. (8-17)
This relationship is presented for accumulation in plants on a dry weight basis. The data used
to develop the correlation with the octanol-water partition coefficient and the derived correlation
equation are presented in Figure 8*4. The data indicate that the correlation relationship provides
an estimate of the bioconcentration factor within about two orders of magnitude for the chemical
data set on which it was derived.
The chemical concentration in the soil of an agricultural field, €„, may be the result
of various mechanisms including soil erosion for agricultural areas adjacent to a site and
atmospheric deposition for areas farther removed. The methods described in Chapter 5 are used
to estimate the chemical concentration if off site soils (see Equations 5-1, 5-2, and 5-3 for soil
erosion and Equation 5-4 for atmospheric deposition).
The concentration in an above ground crop, Cv, is then the sum of the concentration
due to deposition, C*, and to uptake from the soil, Cm.
8-15
-------
Eq. (8-18)
8.3.3 Plant Uptake by Root Crops
The concentration in root crops is estimated based on a root concentration factor
(RCF) and the chemical concentration in the soil moisture. A correlation between the RCF and
the chemical octanol-water partition coefficient relationship has been derived by Briggs et al.
(1982). The relationship was derived from experimental results based on uptake of chemical by
barley roots and additionally compared with published data on chemical uptake by roots of corn,
tomatoes, soy bean, and rice. The relationship, which has a correlation coefficient of 0.98,
derived by Briggs follows: —
tog(RCF - 0.82) « 0.77-togK^ - 1.55
Eq. (8-19)
The data presented by Briggs and the correlation relationship is shown in Figure 8-5.
The correlation matches the data for a wide variety of chemicals and plants quite well. The RCF
in Equation 8-19 is defined as the ratio of the root concentration divided by the concentration
in solution [(g chemical/g wet root)/(g chemical/mL solution)]. In order to.estimate uptake from
soil, a soil-water partition coefficient is used to estimate the concentration in solution based on
the soil concentration. For the root crops the uptake factor (Bv in Equation 8-15) is estimated as
follows:
RCF
K,
Eq. (8-20)
Since this uptake factor, Bv, is already based on a wet root weight basis, the fraction of wet plant
weight as dry material,,/*,, used in Equation 8-15 is specified as 1.0 for estimating chemical
accumulation in root crops.
8-16
ill-
-------
FIGURE 8-3. Bioconcentration factor (on a dry weight basis) in
vegetables versus soil-water partition coefficient for metals.
Source: Baes (1982)
Concentration ratio (plant to soil)
!.•
!•• B-il
B-ta
l.t •-•3
!.• B-*
!• lf« lt«t • !•••• !•••••
3oil-watar partition eooTrieiant Kd
!••••••
8-17
-------
FIGURE 8-4. Bioconcentration factor (on a dry weight basis) in vegetables
versus octanol-water partition coefficient for organics.
Source: Travis and Arms (1988)
0-
-1-
-2-
3
-3-
-4-
-5-
n 1 1 1 1 1 r
1234567
LOG (Row)
8 9 10
8-18
-------
FIGURE 8-5 Root concentration factors (rcf) versus octanol-water
partition coefficient for organics.
Source: Briggs et al.(1982)
1-
U
es
o
3
o-
-i-
-2
-2
-1
0
i
1
I
4
LOG (Kow)
8-19
-------
-------
9.0 HEALTH RISK ASSESSMENT PROCESS
The objective of a public health risk assessment is to provide an estimate of the
magnitude of potential adverse health effects due to a release of hazardous substances.
Information derived from the fate and transport assessment, exposure pathways, and points of
exposure analyses serve to identify the exposed population. The combination of exposure point
concentrations and human intake levels permit the calculation of media-related exposure doses
to an individual.
Although the future role of a contaminated site may not be known a priori, the nature of
current and future land uses at a contaminated site will have a substantial effect on the exposure
pathways of contaminants released from the site. Generally, the future land use can be difficult
to predict. In fact, numerous examples exist of residential developments being built on or near
waste sites shortly after the facilities are closed. Consequently the potential for hazard
associated with various hypothetical land use scenarios should be evaluated.
When a closed hazardous waste landfill is located far from population and agricultural
areas, the exposure pathways that need to be considered relate to the use of the affected ground
water as drinking water, bioaccumulation in fish if the contaminants reach a surface water body,
and inhalation of contaminated air on site or in the surrounding area. If the planned future use
of the site includes residential houses, all of the potential pathways associated with the land use
should be investigated.
9.1 RISK EQUATIONS
A quantitative estimate of the hazard from exposure for non-carcinogenic health effects
is performed by comparing exposure concentrations with reference doses (R/Ds). Once known
as the acceptable daily intake (ADI), the RfD is an estimate (with uncertainty spanning perhaps
an order of magnitude) of a daily exposure that is likely to be without deleterious effects during
a lifetime (EPA, 1989a). Thus, as the ratio of the chronic daily intake (CDT) to the RfD
increases, so does the presumed hazard to human health; however, it is not possible at present
9-1
-------
to determine more precisely the relationship between this ratio and probability of harm. This
relationship can be expressed as:
Z=lL > 2 Potential for Adverse Health Effects
RfD
Eq. (9-1)
An important point to note when using Equation 9-1 is that RJD values are reported as
administered dose, hence the CDI value used in Equation 9-1 should also be based on an
administered dose, not an absorbed dose (i.e., human absorption fractions should be specified
as 100 percent). When evaluating the above expression, it must be kept in mind that the
precision of the RfD is no more than one digit, therefore the CDI should also be rounded to one
digit, and the calculated quotient should also be rounded to one digit (EPA 1989a). For
example, if the ratio of the CDI to the RfD is 1.3 the ratio cannot be distinguished from 1 and
should be rounded to 1.
A similar risk assessment for carcinogenic chemicals is derived from the use of
mathematical probability models. The calculation is expressed in the equation (for cases where
the risk level is less than
Risk
70-yr lifetime average daily dose (mg/kg/day)
potency factor (mg/kg/day)
Eq. (9-2)
The cancer potency factor (also called slope factor or unit risk) is a calculated value expressing
the upper bound risk of developing cancer as a mathematical probability value per 70-year
lifetime of daily dose based on a non-threshold dose-response relationship. It is usually based
on a linearized multistage extrapolation model for animal data and general linear modeling for
human data. The cancer potency factors represent upper bound estimates of risk and are thought
to be generally conservative given their upper bound nature. The true risks, which are not
possible to identify, are unlikely to be higher than the upper bound values, and may be
substantially lower.
9-2
-------
Hie Human Health Assessment Group of .the Office of Health and Environmental
Assessment of EPA is responsible for identifying many of the potency factors for chemicals that
have been evaluated as human or potential human carcinogens. Other EPA health hazard
assessment programs also develop toxicity parameters.
Summaries QfRfDs (for noncarcinogens) and cancer potency factors (for carcinogens) are
provided in EPA (1986d) and are available on the most current basis from EPA's IRIS system
(EPA, 1989b). Protection of public health requires that total human exposure, £,-„,, to a
contaminant from all possible routes of exposure should not exceed the RfD for noncarcinogens.
RfD > ETal (mg/kg/day for noncarcinogens) Eq. (9-3)
Since exposures may occur from various routes, the dose from each route must be
summed to get the total exposure, £&,. Similarly, for carcinogens, the total human exposure
from all pathways combined should be used to estimate the upper bound risks for the situation
under investigation.
9.2 EXPOSURE EQUATIONS
The critical aspect of performing risk assessments lies in the calculation of exposure
doses. The level of detail in estimating exposure doses should reflect the level of detail in the
assumptions about the exposure parameters. The various assumptions required will certainly
affect the level of conservativeness and the degree of uncertainty in the screening exercise,
therefore, conservative parameters should be selected unless site specific data indicate otherwise.
The following algorithms incorporate generic assumptions that may be modified for
specific cases and conditions. These algorithms may be simplified based on what is known
about the exposure parameters and the exposed population.
9-3
//f
-------
Definition of parameters:
ADS = Total amount of chemical administered per day (mg/day),
ADV SB Total amount of chemical administered per occurrence (mg),
AS = Fraction of chemical absorbed through the skin (g/g),
BW = Body weight (kg),
Ca = Concentration of chemical in air (mg/m3),
CD/ = Average daily intake of the chemical (mg/kg/day),
C^ = Concentration of chemical in food (mg/kg),
C, = Chemical concentration in soil (mg/kg),
Cw = Concentration of chemical in water (mg/L),
DF = Fraction of day during which exposure occurs (hours/24),
Ia = Inhalation rate of air for human (m'/day), —
lf = Ingestion rate of food for human (kg/day),
/, = Soil ingestion rate (mg soil/day),
lw = Ingestion rate of water for human (L/day),
L = Length of exposure period (years, or any other time units consistent with LF
below),
LF = Length of time during exposure period during which exposure actually occurs
(years, or any other time units consistent with L above),
ND = Number of days during exposure period L (dimensionless),
SC = Amount of soil contacted per visit (g soil/visit),
V7 = Total number of visits to site or total number of days at residence during
exposure time interval (dimensionless),
XF4 = Conversion factor (1 kg/10* mg),
XF5 = Conversion factor (1 kg/1000 g),
9.2.1 Time-Weighting of Exposures
The average daily intake (CDI) is the value compared with the RfD for evaluating the
potential health hazard of noncarcinogenic chemicals. It is critical that the CDI is compared to
9-4
-------
an R/D derived for a similar time period as the selected time-weighting interval. Typical values
include chronic and subchronic /^D values and 1-day, and 10-day Health Advisory (HA) levels
which may be provided in concentration units. Selecting an appropriate time-weighting interval
is important because of the known toxicity that occurs from short term, high-dose exposures to
many chemicals. The Systemic Toxicants Assessment Branch of EPA has suggested the
following stepwise procedure for evaluating different time-weighting intervals:
1) Calculate an average daily dose using a time-weighting period not
exceeding day. Compare this dose with applicable limits for one-day
exposure, such as one-day HA levels.
2) Repeat the same calculation for a 10-day time-weighting interval.
3) Repeat the same calculation for a longer time-weighting interval based
on available longer term HA values.
4) Repeat the same calculation using a subchronic R/D value and an
appropriate time weighting interval.
5) Repeat the same calculation using a chronic RfD value and an
appropriate time weighting interval.
The key point in this stepwise procedure is to ensure that the time-weighting period used in the
analysis does not exceed the period over which doses are received. If at any step in the
procedure the dose exceeds the applicable criteria, then the site conditions may represent a health
hazard.
If an RfD value is provided in IRIS, that value should always be used. However, if
there is no verified RfD but a HA is available, the HA may be used to evaluate potential health
hazards due to drinking water exposure. HA values must be modified to the units of an R/D by
multiplying by the intake level of water and dividing by the body weight.
9-5
-------
For carcinogenic chemicals, the potency factors are based on an average dose over
a 75-year lifetime interval. Therefore, the time period used in Equation 9-5 for carcinogenic
chemicals should be 75 years, and the 75-year average GDI is then used in Equation 9-2 to
calculate risk.
9.2.2 Inhalation Pathway
The inhalation pathway examines the exposure resulting from atmospheric transport
of chemicals. The dose is calculated as an average ventilation rate times the average chemical
concentration and a factor the fraction of time during which exposure occurs. The daily
exposure through the inhalation pathway is represented as:
ADS
C.-I.-DF
Eq. (9-4)
The average chronic daily intake (exposed dose) over any specified time period, L,
(e.g., 1 day, 10 day, 1 year, or 75 year period) is calculated using the formula:
GDI =
ADS'LF
BW-L
Eq. (9-5)
It is critical that the CDI calculated in Equation 9-5 is compared to an RjD derived for a similar
time period as the time-weighting period (L).
9.2.3 Ingestion Pathway
The ingestion pathway examines the exposure resulting from ingestion of chemicals
in water, food, and soil. For exposure through chemicals in water the ingestion pathway is
represented as:
9-6
-------
ADS = CW-IW'DF Eq. (9-6)
For exposure through chemicals in food the ingestion pathway is represented as:
ADS = CM'I,-DF Eq. (9-7)
JOPU J
where the concentration and intake of food, C^ and 1^ are evaluated for fish, meat, milk, and
vegetables. The chronic daily intake for ingestion of water and food is calculated using Equation
9-5.
For exposure through ingestion of contaminated soil, the ingestion pathway calculates
the administered dose per occurrence (ADV) and is represented as:
ADV * Cs'If-XF4 Eq. (9-8)
The chronic daily intake for the ingestion pathway is calculated using Equation 9-10.
CD/ = - Eq. (9-9)
BW'L'ND
9.2.4 Soil Contact
The soil contact pathway is used to represent the exposure resulting from dermal
contact and absorption through the skin. The soil concentration at an off site exposure point
may result from erosion of soil from the waste site and/or atmospheric deposition. The exposure
resulting from soil contact is represented as:
9-7
-------
ADA = C,'SC-XF5-AS
Eq. (9-10)
The average daily intake is then determined using the formula:
CDD =
ADA-VI
Eq. (9-11)
All of the same restrictions regarding the exposure interval, L, and the appropriate
RfD value discussed previously in regards to Equation 9-5 also apply to Equation 9-10 (i.e., for
noncarcinogenic chemicals the RfD and the exposure interval should be for a consistent time
intervals and for carcinogenic chemicals the time interval should be 75 years). —
9.3
EXPOSURE PARAMETERS
In order to estimate the exposure doses and associated risks for the different
pathways, all of the human intake and consumption parameters must be defined. Average values
for the exposure parameters are listed in Tables 9-1 through 9-5. The values listed in these
tables can be modified if more site specific and/or population-specific data are available.
9.3.1 Absorption Assumptions
The toxicity data available from IRIS are reported as administered dose, hence the
use of actual human absorption fractions are not required for estimating health hazards when
appropriate toxicity data are available (i.e., a 100 percent absorption fraction should be used for
comparison with toxicity data from IRIS). If the human exposure route (e.g., dermal exposure)
involves a different efficiency of absorption into the body than the toxicity study on which the
critical toxicity value is based, an adjustment must be made to estimate health hazard.
The remaining parameters for estimating exposure dose, body weight and exposure
duration are listed in Table 9-6.
9-8
-------
TABLE 9-1
Inhalation Rate
INHALATION RATE
Adult
Child (1-6 yis)
Child (6-15 yrs)
Worker (at job)
mVDAY
20
10
15
10*
* per 8-hour work day
Source: EPA, 1986d and 1988b
TABLE 9-2
Ingestion of Drinking Water
* per 8-hour work day
Source: EPA, I986d
INGESTION RATE
Adult
Child (1-10 yrs)
Worker (at job)
liters/day
2
1
1*
9-9
/
-------
TABLE 9-3
Ingestion of Fish, Meat, Milk, and Vegetables
AVERAGE INGESTION RATE
Fish
Meat
Milk
Vegetables and
fruit
kg/day
0.028
0.10
0.30
0.34
Source: EPA, 1988b
TABLE 9-4
Ingestion of Soil
INGESTION RATE
Adult
Child (1-6 yrs)
Child (6-11 yrs)
mg/day
25 (average)*
100 (maximum)*
200 (average)*
800* (maximum)"
50 (average)'
250 (maximum)'
* The maximum value of 800 mg/day soil intake for a child is indicative of an ingestion rate by a child exhibiting
behavior.
Source: « LaGoy, 1987
<"> EPA, 1988b
9-10
-------
TABLE 9-5
Dermal Contact with Soil
CONTACT RATE
Adult or child
g/day
1 (average)
3 (maximum)
Source: EPA, 1986d
TABLE 9-6
Additional Exposure Parameters
BODY WEIGHT
Adult
70
Child (0-1 yr)
10
Child (1-6 yrs)
15
Child (6-11 yrs)
30
Occupational Exposures - 8 hours/day, 240 days/year, up to 45 years.
Residential Exposures - 24 hours/day, 365 days/year, up to 70-year lifetime.
Duration of Lifetime - The Exposure Factors Handbook recommends 75 years as an
average duration, however, slope factors are derived for a 70-year lifetime. A 70-year
lifetime is used throughout this report.
9-11
-------
-------
10.0 USER'S GUIDE
This chapter presents information to assist users of MMSOILS to install and operate the
model. Because MMSOILS has been designed to be an easy to use, menu-driven system with
associated Help screens, this users' guide primarily focuses on installation of the system on a
computer and then provides some guidance on how to select parameter values for the model
inputs.
The user's guide is broken into several sections. First, the computer system requirements are
presented and the procedures for installing MMSOILS are described. Second, the model input
files are described with discussion on data that must be included in the input files along with
potential sources of these data. The final section provides discussion on model outputs.
10.1 SYSTEM REQUIREMENTS AND INSTALLATION PROCEDURES
MMSOILS was developed to operate on an IBM or 100 percent compatible personal
computer system. The minimum system requirements are listed below:
• IBM or compatible personal computer
• 512 KB available random access memory
• One 3-1/2" or 5-1/4" floppy disk drive
• One hard disk drive (2.0 MB storage available)
• DOS 3.x or higher
Because of the extensive computations involved in MMSOILS, it will operate most effectively
on a 386 or 486 series computer equipped with a math coprocessor.
MMSOILS comes equipped with a self-installation program that will automatically create
a subdirectory for the model, reconfigure as necessary the system "config.sys" file, and install
the software. Follow the steps below to install MMSOILS on personal computer.
10-1
-------
Step 1: Turn on your computer system and start DOS
As shown above, your computer system must be equipped with a hard disk drive with at least
2.0 MB of free space available to effectively run MMSOILS. The install program must be run
from DOS (version 3.x or higher), so boot the computer and access the DOS prompt before
moving on to Step 2. '
Step 2: Place the MMSOILS disk in your floppy drive
Place the disk with 3-1/2" or 5-1/4" version of MMSOILS, in your floppy drive and type
"install" at the DOS prompt. The "install" command will access the install program on the
MMSOILS diskette and will create an MMSOILS directory on the default drive of your
computer (the install program will prompt you to specify which drive on which to place the
program). It will also create one subdirectory ("outputs") where the MMSOILS model output
Mies will be created and stored during model execution. MMSOILS will be configured to run
from these default directories.
Upon completing the installation, the system will print the following message to indicate that the
system has been installed successfully on the computer:
"MMSOILS has been successfully installed."
If necessary, the install program will alter the system's "config.sys" file to allow operation of
MMSOILS (which requires a Files = 55 or greater line in the config.sys file).
Step 3: Running MMSOILS
To run MMSOILS, type "MMS" from the directory in which it has been installed. If
MMSOILS has been installed on the "C:\MMS" directory for example, then make "C:\MMS"
your default directory before entering this command:
10-2
-------
C:\MMS\MMS
Upon entering this command, the database files in MMSOILS will be indexed as
necessary and the system will be available. The main menu has six primary functions described
briefly below. To access them, use the cursor arrow keys to move to the selected option and
press the < Biter > key. Instructions for operating MMSOILS are provided on the menu
screen.
• Model inputs: This function allows the user to create or modify inputs for
MMSOILS runs. It uses a look-up screen to allow you to save and select from
existing MMSOILS input sets.
• Run set: This function allows the user to run MMSOILS with an input set(s)
created from the Model Inputs menu option. This menu choice provides the
option for running a single case or running a batch of cases.
• Chemicals: This function allows the user to add, delete, or modify chemicals to
the chemical database in MMSOILS. When adding chemical information when
creating model inputs, MMSOILS allows you to access data from this database.
MMSOILS comes pre-loaded with data on a variety of chemicals of interest for
many contamination problems. This menu option allows the user to update this
default list.
• Results: This function allows the user to view and print model results.
• Utilities: The Utilities menu provides several options related to set-up and
maintenance of the system: file indexing, default directories, and transferring files
to floppy disk.
• Exit: This menu option allows the user to exit from the MMSOILS system,
10-3
-------
In addition, MMSOILS has user Help functions (press the "Fl" key) to assist users in operating
the system. After accessing Help, press the escape key to return to the MMSOILS system.
10.2 MODEL DATA FILES
MMSOILS automatically creates eight input files from the system entry screens in order
to run a single problem. While the MMSOILS entry screens allow the user to create these files
without having to manipulate them directly from DOS, we present the data requirements based
on these files as they are run by MMSOILS. Using the Results function from the MMSOILS
system menu, these input files can be reviewed and/or printed.
The input and output files are identified by fixed seven character names followed by a
three character extension (e.g., CHEMPRP.001) where the three characters of the extension,
001, may be used to define different groups of input and corresponding output files. The three
characters of the extension are defined by the user upon execution of the model. The three
characters may be defined as any letter, A-Z, any number, 0-9, and some of the other characters
on the computer keyboard (see the DOS operating manual for valid characters to use in
filenames). The following description of file names will use the symbols ### as the user defined
characters representing a specific data set (e.g., CHEMPRP.###).
Figure 10-1 summarizes the input data files that must be created or modified by the
model user. This version of MMSOILS requires up to eight input data files. The first six files
(CHEMPRR###, ATMSPRM.###, SWTPATH.###, GWTRANP.###, INFILTR.###, and
FOODCHN.###) define the physical, chemical, and environmental properties of a site as well
as the constituents of concern. Of these six files, only CHEMPRP.### and CONTROL. ### are
required in every model run. The CONTROL.### file designates how a simulation is executed,
such as the selection of a WMU type and the number of chemicals and pathways to be
considered. The CHEMPRP.### file includes the chemical-specific information for the model
run. The last four files (LANDFIL.###, IMPOUND.###, WSTPILE.###, INJECTW.###, and
TANKREL.###) define the input information specific to the type of WMU; only one WMU file
10-4
-------
can be included in each model run.
Example input files ate included in this Chapter and are provided with the model. The
input files are intended to be self explanatory in that they include a comment describing each
input variable, the necessary units of the input variable, and where possible, ranges for the
Figure 10-1
Listing of Input Data Files for MMSOILS.'
No.
1
2
3
4
5
6
7
8
9
10
11
12
File
Name
CHEMPRP
ATMSPRM
SWTPATH
GWTRANP
INFDLTR
FOODCHN
CONTROL
LANDFIL
IMPOUND
WSTPILE
INJECTW
TANKREL
Variables
46
33
25
17
26
19
11
54
58
6
6
4
File Description
Chemical properties data set
Atmospheric pathway data set ""*
Surface water pathway data set
Ground-water transport pathway data set
Infiltration leaching and recharge data set
Food Chain bioaccumulation pathway data set
Control file
I^Atuifill ctumcteristics dfitfl itf
Surface impoundment characteristics data set
Waste pile characteristics data set
Injection well characteristics data set
.Tank release characteristic* data set
a. The file name extension "MH" is ignored here for simplicity.
specific variable. Example input files are shown in Figures 10-2 through 10-13. The
ENVTRNP program (environmental fate and transport model) creates the necessary input data
files for the EXPOSE program (human intake calculations). The model user has the option to
modify values in these files before running the EXPOSE program.
The fate and transport model creates five output files. The first output file is a summary
of all input parameters, internally calculated parameters, and concentration in each
environmental pathway at exposure points. The other output files summarize specific results for
10-5
-------
use as inputs for post-processing of MMSOILS results. The output files from the fate and
transport model are:
1) RESULTS.###: Summary of input parameters, calculated parameters, and exposure
point concentration levels.
2) ACTLEVL.###: Comparisons of results with action level data.
3) SURFOUT.###: Summary of ground-water exposure point concentrations for use in
contouring software package.
4) EXPCONC.###: Concentration at exposure point in water, air, soil, and
food items. _
5) INTAKES.###: Default values for human intake levels and absorption fractions for
water, air, soil and food items.
The EXPCONC.### and INTAKES.### files are the input files for the human intake
calculations. Both of these output files may be easily modified for input to the program for
human exposure calculations. The user may choose to modify these values based on different
exposure assumptions and measured environmental concentration levels at the exposure point.
These files are also essentially self-explanatory with a comment describing each variable and the
proper units.
10.3 DESCRIPTION OF INPUT DATA FILES
Figures 10-2 to 10-13 provide the formats of input data files. Each exhibit shows the
variables' line positions (column 1), variable types (column 2), and a brief description for each
of the variables (column 3). The following sections describe the input parameters on a file-by-
file basis.
10-6
ts
-------
10.3.1 Chemical Properties Data File (CHEMPRP.###)
Hie data set for the chemical properties is shown in Figure 10-2. Most of the data in
this file are chemical specific parameters that can be found in available literature sources. This
file allows the transport simulation for more than one chemical in a model run. All properties
must be entered for each chemical, one chemical at a time. Therefore, the following procedure
is followed once for chemical 1, again for chemical 2, and so on for up to 10 chemicals. The
variables are described below:
Line 1. The first line in the data set is the name of chemical whose parameters are listed
in the subsequent lines.
Lines 2-9. These seven lines describe chemical properties. Sources of data for estimating
these parameters include EPA (1986d), EPA (1987), and Lyman et al. (1982).
Lines 10-12. These lines are site-specific, and defining concentration on soils and concentration
in liquid form. "Concentration on soil" (Line 11) is used by the atmospheric and
erosion pathways. It is also used in the ground-water pathway when the WMU
is a landfill and the leachate concentrations are calculated using the partitioning
method. It can be zeroed out in all other cases. "Concentration in liquid"(Line
12) is only used by one of the atmospheric models for estimating the mass loss
of a volatile or semi-volatile contaminant into the atmosphere from topsoil through
s
volatilization.
Lines 13-18. The next six lines in the data set represent the toxicity data for the specific
chemical. For noncarcinogens, the input parameter is the RfD in mg/kg/day. For
carcinogens, the input parameter is the cancer potency factor in (mg/kg/day)'1.
10-7
-------
Figure 10-2
Variables in the Chemical Properties Data File (CHEMPRP.###).
Line
1
2
*»
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Variable Type*
TITLE
TTTf P
ft ft A MV
R
R
R
R
R
R
R
TITLE
R
R
TITLE
R
R
TITLE
R
R
TITLE
R
R
Int (0/1)
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
Description
Name of Chemical
Chemical •nttti^fftig« H«>« •**
Henry's law coefficient (atm*mVmoI)
. Molecular diffusivity in air (cm«cm/s)
Molecular diffusivity in water (cm*cm/s)
Vapor pressure (mm Hg)
Molecular weight (gm/mol)
Adsorption coefficient on organic caibon, JTM (ml/gm)
Solubility limit in water (mg/1)
Chemical concentration on soil and/or soil moisture
Concentration on soil (mg/kg)
Concentration in liquid (mg/1)
Chemical toxicity data for noncarcinogenf
Reference dose level oral route (mg/kg/day)
Reference dose level inhalation route (mg/kg/day)
Chemical toxicity data for carcinogens —
Carcinogen potency factor oral route ((mg/kg/day)'1)
Carcinogen potency factor inhalation route ((mg/lcg/day)'1)
Dermal absorption of contaminated water
Skin permeability constant (cm/hr)
JCj for ytttTf mfittrittl* (wl/pn)
Flag: 0 = organics; 1 - inorganics (if 1, read 4 more yar.)
Xj for clay+organk* < 10% of total soil mass (ml/gm)
Kd for clay+organics 10 - 30X of total soil mass (ml/gm)
f.t for clay + organic* 30% of total soil mass (ml/gm)
Kf for aquifer (ml/gm)
Ground water first-order decay (1/yr)
Unsttunted sediment* first-order decay (1/yr)
Concentration in leachate (if constant teachatc desired) (mg/1)
Concentration in influent (if impoundment used) (mg/1)
Concentration in injection well (if injection well used) (mg/1)
Chemical decay in waste management unit (1/yr)
Volatilization parameter for impoundments (1/yr)
Bioconcentration in fish ((mg/kg fish)/(mg/l water))
Sediment to fish partition coef. ((mg/kg &sh)/(mg/kg soil))
Transfer factor for cattle ((mg/kg meat)/(mg/kg intake))
Soil to meat partition coefficient ((mg/kg beeO/(mg/kg soil))
Transfer factor for milk ((mg/kg milk)/(mg/kg intake))
Soil lo milk partition coef. ((mg/kg milk)/(mg/kg soil))
Uptake from soil to plant ((mg/kg dry plant)/(mg/kg soil))
Soil moisture to root factor ((mg/kg root wet)/(mg/kg solut.))
Chemical decay constant in off-site field (1/yr)
Partition coeff. In off-site fields (ml/gm)
First-order decay in stream (1/yr)
Kt for lake sediments (for inorganics only) (I/kg)
Action level of contaminat in air 0«g/m')
Action level of contaminat in soil (mg/kg)
Action level of contaminat in ground water (mg/1)
Action level of contaminat in surface water (mg/1)
« R = real number: Int m integer: values in naientheses renresent acceotablc ranee.
10-8
-------
The first parameteris the fyD for the oral intake route and the second is the RfD
for the inhalation intake route. For carcinogens, the first two parameters are
input as 0.0 and potency factors are specified in the next two parameters (oral
followed by inhalation). For noncarcinogens, these last two parameters are input
at 0.0.
Lines 19-20. The next two lines describe the skin permeability constant for dermal absorption
through contact with water, in units of cm/hr. Data sources for this parameter
are rather limited, a table of values for various chemicals is included in EPA
(1987).
Lines 21-26. The next six lines are new parameters. "Kd for waste materials" (Line 21) is a
critical parameter to estimate leaching from a WMU to the ground water. ThTs
is necessary to specify a K* value for wastes for all leachate calculations based on
the partitioning approach. The waste layer Kd indicates the degree to which the
chemical sorbs to the waste material. Chemicals which strongly sorb may have
a range of values from 5 to 20 ml/gnu Chemicals which sorb weakly, and which
will leach out of the system quickly, may have %, values of between 0.05 to 0.1
ml/gm. The values are both chemical specific and site specific, and a range of
Kd's for waste will generate a wide range of leachate concentrations.
Line 22 is used to specify whether the constituent of concern is organic or not.
If Line 22 is "0," the constituent is organic. Otherwise, "1" should be specified
to represent inorganics.
Kj can only be approximated by KK • FK if the constituent is organic with known
Kx value. For inorganics, KM and FK are meaningless, and the user must specify
appropriate Kd values explicitly in CHEMPRP.###. A survey compiled by
DOE's Office of Environment, Safety, and Health (OESH) has demonstrated that
10-9
-------
J^i can be linked to the clay and organic content in the total soil mass ~ the
higher the clay and organic content, the higher the K4 value (i.e., higher sorption
capability for the soil).1
Lines 23-26 are needed only if the organic/inorganic flag line (Line 22) is I.2
The user is required to provide three Kd values for low, medium, and high clay
and organics content in the soil (Lines 23-25). The three levels are defined by
a combined clay and organics content of less than 10 percent, between 10 and 30
percent, and greater than 30 percent, respectively. The model determines the
proper Kd value to use for each soil layer, based on the soil properties defined in
INFILTR.###. The user also needs to specify the Kt for the aquifer in Line 26
if the constituent is inorganic. Otherwise, the Kd for aquifer is calculated by the
model by multiplying "K^* for organic chemicals (Line 8) by "Fa" for the
ground-water pathway (defined in GWTRANP.###).
Lines 27. This parameter should be assumed to be zero unless site specific data indicate
otherwise.
Line 28. For a conservative analysis this parameter, the unsaturated sediments first order
decay coefficient, should be assumed to be zero unless site specific data indicate
otherwise. Many organic chemicals are subject to biodegradation in soils but the
conditions under which this activity will occur are highly variable and difficult to
predict. Hydrolysis may occur whenever the chemical is found in moist or
saturated conditions. The ground-water pathway is very sensitive to degradation
1 Pacific Northwest Laboratory. Multimedia Environmental Pflllutant Assessment
System fMEPASJ. Addendum D: Constituent Data Base. Prepared for U.S. Department of
Energy's Environmental Survey, Office of Environment, Safety, and Health, September 7,
1987.
2 If the flag line is zero, the four lines for K4 values are not needed, and the "ground-
water first-order decay" parameter will become Line 23.
10-10
-------
because travel times through the partially saturated zone may be quite long.
Ranges of half-lives for some chemicals are listed in EPA (1986d). Below the
top soil layer degradation rates should be specified as zero if biodegradatkm is the
principal mechanism, even if degradation is considered in the surface soil. The
units of the decay coefficient are (years)'1.
Lines 29-34. The next six parameters describe site-specific constituent parameters.
"Concentration in leachate" (Line 29) is used if the waste management unit is a
landfill and the leachate calculations are based on the completely-mixed reactor
method with empirical sorption parameter or the assumed constant concentration.
In the former case, the input leachate concentration represents the initial
concentration in the waste.' In the latter case, the input leachate concentration
represents the constant leachate concentration.3 This line is also used if WMtJ
is a tank to input constituent concentration in the release.
"Concentration in influent" (Line 30) is used if the WMU is an impoundment and
the leachate calculations are based on the completely mixed reactor method or a
constant leachate concentration. In the former case, the input influent
concentration represents the initial concentration of the impounded liquid waste.
In the latter case, the input influent concentration represents the average leachate
concentration, both before and after the impoundment is drained of liquid and
becomes a landfill. If the WMU is not an impoundment, enter zero for this line.
"Concentration in injection well" (Line 31) is used if the WMU is an injection
well. The input concentration represents the concentration of the fluid which is
assumed to be injected directly into the saturated zone. If the WMU is not an
injection well, enter zero for this line.
3 The constant leachate concentration method assume an unlimited source of
contaminants and does not conserve mass. This approach can be used if the WMU is a
highly contaminated site with the constituents of concern highly sorbed to the soil media.
10-11
-------
"Chemical decay in waste management unit" (Line 32) is used in all methods for
estimating leachate quality of impoundments, and in the partitioning method for
landfills. As before, if decay is zero, there is no waste decay. "Empirical
chemical-specific desorption parameter" (Line 33) is used in the completely-mixed
*
reactor method for both landfills and impoundments. It is an empirically-based
parameter which is available only for a handful of chemicals. It should be noted -
that in these models, the desorption parameter is tied only to time. In fact,
desorption should also be a function of the volume of water through the system.
As a result, if the empirical value is too large in comparison with the volume of
water flushing the system, the program will stop with a "math overflow" error.
To remove this problem, the user can adjust the waste layer thickness or the
empirical desorption parameter. This parameter should be used with caution.
"Volatilization parameter for impoundments" (Line 34) is used in all
impoundment leachate calculations. It is equal to ln(2)/i(1/2), where ta/2) equals the
half life of a chemical in years.
Lines 35-42. The next ten parameters are chemical specific parameters that represent the
accumulation of chemicals in fish, meat, milk, vegetation, and root vegetables.
They are used only if the food chain pathway is requested as specified in
CONTROL.### file, otherwise they are skipped.
The first parameter is the bioconcentration in fish in units of liters of water per
kg of fish. This parameter can be estimated from various data bases described
in Section 8.1 or using Equation 8-3 in the absence of any other data sources.
The second parameter, a sediment-to-fish partition coefficient (in units of mg/kg
fish / mg/kg soil) may be used if chemical specific data are available and
sediment concentrations are known. This parameter may be specified as 0, in
which case the concentration in fish is estimated based on the water concentration
and BCF factor. If this parameter is greater than 0, the model will then prompt
10-12
-• *,
-------
die user to specify the sediment concentration.
Hie third parameter is the transfer factor for uptake of chemicals by cattle in
units of days/kg. Sources of data for this parameter include Kenaga (1980),
Garten and Trabalka (1983), and Travis and Arms (1988). If available literature
do not provide estimates for a specific chemical, approximate values (within about
2 orders of magnitude) can be estimated using Figure 8.1 or Equation 8-9 located
in Section 8.2.
The fourth input parameter is a soil-to-meat partition coefficient (in units of
mg/kg meat/ mg/kg soil). Data from which this parameter may be estimated are
very limited: a value for 2,3,7,8-TCDD is suggested in EPA (1988a). This
parameter may be specified as 0.0 and the concentration in meat will be estimated
using the transfer factor defined on the previous line. If it is greater than 0.0,
then the soil-to-meat partition coefficient is used to estimate the meat
concentration.
The fifth input parameter is the transfer factor for uptake of chemicals into milk
in units of days/kg. One source of limited data for this parameter is Travis and
Arms (1988). If available literature do not provide estimates for a specific
chemical, approximate values (within about 2 orders of magnitude) can be
estimated using Figure 8.2 or Equation 8-10 located in Section 8.2. The sixth
input parameter is a soil-to-milk partition coefficient (in units of mg/kg milk/
mg/kg soil). Data from which this parameter may be estimated are very limited:
a value for 2,3,7,8-TCDD is suggested in EPA (1988a). This parameter may be
specified as 0.0 and the concentration in milk will be estimated using the transfer
factor defined on the previous line. If it is greater than 0.0, then the soil-to-milk
partition coefficient is used to estimate the milk concentration.
The seventh parameter is the bioconcentration factor for uptake of chemicals in
10-13
-------
vegetation in units of kg soil/kg dry plant. Sources of data for this parameter
include Baes (1982) for inorganics and Travis and Arms (1988) for organic
chemicals. If available literature do not provide estimates for a specific chemical,
approximate values (within about 2 orders of magnitude) can be estimated using
Figure 8.3 or Equation 8-16 for inorganics or Figure 8.4 or Equation 7-17 for
organics (see Section 8.3.
The last parameter in this group is the root concentration factor for uptake of
chemicals in root vegetables in units of kg soil/kg root wet weight. One source
of data for this parameter is Briggs et al. (1982) and references therein. If
available literature do not provide estimates for a specific chemical, approximate
values (within about 1 order of magnitude) can be estimated using Figure 8.5 or
Equation 8-19 from Section 8.3. ~
Lines 43-44. These parameters describe chemical decay and partition in the off-site agricultural
fields. The first parameter is the first order decay coefficient of the specific
chemical in the surface soils of the agricultural field in units of (years)*1. For a
conservative analysis this parameter should be assumed to be equal to zero unless
site specific data indicate otherwise.
**
The next parameter is the chemical specific soil-water partition coefficient for the
agricultural soils in ml/g. This can be estimated based on the fraction organic
/
carbon in the agricultural soil and the organic carbon sorption coefficient, Kx.
Lines 45-46. The last two parameters (first-order decay in stream and Kd for lake sediments)
relate to the surface water transport and are input only if the surface water
pathway is selected.
Line 45 is the "lumped* first order decay coefficient used to represent the
combined effect of all attenuation mechanisms in the stream. Some observed
10-14
-------
valves for various chemicals are listed in EPA (1986d) and EPA (1980a);
estimation techniques are described in EPA (19845).
Line 46 is the sediment-water partition coefficient for lake sediments. This can
be estimated based on the organic carbon sorption coefficient, Kx) and the
fraction organic carbon in the lake sediments. Lake sediments will typically have
a higher organic carbon fraction than surface soils, with values ranging from
about 0.03 to 0.2.
Lines 47-50. These parameters specify action levels that may be compared to exposure
concentrations for screening purposes.
10.3.2 Atmospheric Pathway Data File (ATMSPRM.###) -
The data set for the atmospheric pathway is shown in Figure 10-3. The data in this file
represent the physical characteristics of the site, the types of atmospheric source terms, and
atmospheric transport parameters.
Ones 1-9. The first eight parameters (source type for volatilization, depth of clean cover,
depth of contamination, temperature, site width, and mixing depth of surface soil)
must be selected based on a knowledge of the site. The mixing depth of surface
soil should be selected based on the land use at the site. If the surface soil is
undisturbed, a depth of 5 cm is suggested for the mixing depth. If the surface
soil is disturbed due to mechanical action (digging, vehicle traffic, etc.)* a depth
of 10 to 20 cm is suggested. The deposition velocities for PM10 and PM30 can
be estimated based on particle size and density, typical values range from 0.01
to 0.05 m/s.
10-15
-------
Figure 10-3
Variables in the Atmospheric Pathway Data File (ATMSPRM.###).
Line Variable Type*
Dcscnpdoo
1
3
4
5
6
7
8
9
10
11
12
&A
13
14
IS
16
17
18
19
20
21
22
23
24
25
315
*w
27
28
29
30
31
32
33
34
TITLE
••*» /i j*9nM\
Int (1/2/3/4)
TITLE
R
R
R
R .
R
R
TITLE
R
R
R
TITLE
R
R
R
R
R
R
R
R
R
R
TITLE
Int (1-10)
•**•• \* ***/
TITLE
R
Int (1-36)
R
Int (1-6)
R
F ••••«•• li i t\n_m • mlaSMtiMSiMlvM.M. /I 1 1 JB.H _*1\
Source type lor volatilization (l,2,3,or 4)
Physical, environmental p-imrfrTT for ate
Depth of clean cover (cm)
Depth of contamination (cm)
Temperature (degrees centigrade)
Muting depth of site soil (em)
Deposition velocity of PM10 partkulatcs (m/s)
Deposition velocity 01 HMIV psntnimci (nut)
Wind erosion parameters
Wind velocity (m/s)
Paiticulate emission panunejierB
Silt content of road surface (percent)
Mean vehicle speed (Km/or)
Mean vehicle weight (Mg)
Number of wheels per vehicle
Number of days per year with at least 0.01 inches rain
Vehicle kilometers on unpaved surface (no. vehicles*km/hr)
Drop height for loading operations (m)
Moisture content of loading material (percent)
Capacity of loading machinery (cubic meters)
Daily load of material (Mg)
Atmospheric stability data
Number of points to examine •tmc*pheric concentration
Data set for air transport to point 1
Distance to point 1 (m)
Number of stability classes for point 1 (f)
Wind velocity (m/s)
Stability class (dinxmstontess)
Frequency for point 1 (dimensionless)
a. R - real number, Int » integer, values in parenth
epresent acceptable range.
Lines 10-14. These lines characterize wind erosion parameters, and are described in more
detail in EPA (1985a) and EPA (1985b). The average wind erosion velocity can
be determined based on local meteorological records. The threshold friction
velocity used in the model for the wind erosion source term ranges from about
10-16
-------
25 to 75 cm/s for loose sandy soils. The vegetative cover fraction ranges from
0 to 1, with 0 representing no vegetation (worst case) and 1 being continuous
vegetative cover (which corresponds to no wind erosion). Typical values of the
surface roughness height for bare fields to grassland range from 1 to 3 cm,
respectively.
Lines 15-21. These lines describe participate emission parameters due to physical disturbances.
Guidance for selecting appropriate values for these parameters is provided in
more detail in EPA (1985a) and EPA (1985b). Average values for percent silt,
vehicle weight, vehicle speed, and number of wheels are listed in Table 10-1.
Meteorological data can be used determine the number of days with at least 0.01
inches of rain. The average number of vehicle kilometers traveled per hour on
unpaved surfaces is a site specific parameter. If this parameter is input as zero,
the vehicle participate emission will be zero.
Table 10-1. Default values of percent silt, vehicle weight,
vehicle speed and number of wheels (from EPA 1985a).
Site
Residential/
Rural
Industrial
Percent Silt
Avg. Range
15
8
(5-68)
(2-29)
Speed (km/hr)
Avg. Range
48
24
(40-64)
(8-32)
Weight (Mg)
2
3
#of
Wheels
4
4
Line 22. The drop height (in meters) is used to represent the height from which machinery
drops contaminated soil during loading and unloading operations, typical values
range from about 0.5 to 2 meters.
10-17
-------
Line 23. The moisture content of the material -(in percent) depends on the texture of the
material being loaded and the antecedent rainfall conditions. Very wet material
such as sludge or saturated soils may have a moisture content in the range of 30
to 50 percent. Drier materials in more arid climates may fall in a typical range
of about 3 to 25 percent, depending on the soil texture. The wilting point of the
soil material may be used as a reasonable approximation under dry conditions (see ^
Table 10-2 in Section 10.3.5 and convert volume fraction to percent).
Line 24. The capacity of the loading/unloading machinery represents the volume of
material moved per load, typical values for a front end loader range from about
1 to 3 cubic meters. For a backhoe values range-from about 0.1 to 1 cubic
meters.
Lines 25-26. The daily load of material loaded or unloaded (in mg) and the area of spreading
operations (in ha) must be estimated from typical operations at a site (e.g., the
number of truckloads per day). Both of these values may be specified as 0.0
when this type of emission source does not apply to the site.
Lines 27-34. These lines are input for atmospheric stability data. For each exposure point
examined, the first parameter is the distance from the site to the exposure point
in meters. The second parameter is the number of stability class data sets which
follow in the input file. The stability class data are only input for the directional
sector in which the exposure location exists (e.g., north-northeast). The stability
class data include the wind speed in m/s, the stability classification, and the
frequency of time that conditions are in that specific classification. The stability
classifications are input as numbers where l=class A, 2=class B, 3=class C,
4-class D, 5=class E, 6=class F. For each exposure point, in a specific
directional sector, a total of 36 stability classifications (6 wind speed categories
for each of 6 stability classifications) may be available from a weather station.
If these data are not available, a typical conservative approximation (not an
10-18
-------
absolute worst case) is to assume a wind speed of 3 m/s, a stability class of 5
(class E), and a frequency of 0.30.
Note that only one air transport location is shown in Figure 10-3 for illustration. When Line
27 in the new ATMSPRM.### file is input for more than one location, the last six lines should
be repeated as many times to define the additional locations specified.
10.3.3 Surface Water Pathway Data File (SWTPATH.###)
The data for the surface water pathway are shown in Figure 10-4 (file name
SWTPATH.###). The required data in this file are primarily site specific parameters.
Figure 10-4
Variables in the Surface Water Pathway Data File (SWTPATH.###).
Line
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Variable Type1
TITLE
lot (0/1)
R
R
R
R
TITLE
R
R
R
R
R
TITLE
lot (0/1)
R
R
TITLE
lot (0/1)
R
R
R
R
R
R
R
Description
River data
River calculations flag: 0 = no calc.; 1 = perform calc.
Flow rate in river (nrVyr)
Downstream distance in river to point of use (m)
Average stream velocity (m/yr)
Distance from waste site to river (m)
Erosion parameters (USLE calculations)
USLE Rainfall factor (tonnes-m/ha-hr)
USLE soil credibility factor (tonnes/ha/unit R)
USLE length-slope term (dimensionless)
USLE cover factor (dimensionless)
USLE erosion control factor (dimensionless)
Ground-water discharge to river flag
0 = no discharge of ground water to river; 1 — discharge to river
River concentration output begins this year
River concentration output ends this year
Lake parameters
Lake calculations flag: 0 * no calc.; 1 = perform calc.
Sediment delivery fraction (dimensionless)
Fraction organic carbon in lake sediment* (I/kg)
Wind velocity above lake (cm/hr)
Average water depth (cm)
Surface area of lake (ha)
L*fcy sediment porosity (dimensionless)
Diffusion path length for contaminated sediments (5-10 cm)
a. R = real number; Int = integer; values in parenuu
epresent acceptable range.
10-19
-------
Potential sources of the data include the USGS, SCS; and EPA (1986d) for chemical half-lives
in surface water.
Linel.
Line 2.
Title for the input file
This line inputs a flag for designating whether or not the surface water
pathway to a river should be simulated. If the flag is input as 0, the river
pathway is not used, if the flag is input as 1 the river pathway is
evaluated.
Line 3.
This parameter is the annual average river flow rate in mVyr. This
parameter can be estimated from historical gauging records if available,
or estimated based on the upstream area of the watershed and average flow
rates from nearby streams where gauging records are available. The
USGS maintains the largest network of gauging stations and additional
information may be available from local Flood Control Districts.
Line 4.
This input parameter is the downstream distance to the point of surface
water use; this is a site specific parameter.
Line 5.
This input parameter is the average stream velocity, in m/yr, under the
annual average flow condition. This parameter can be estimated from
gauging records if available (i.e., 9-207 forms from the local USGS
office). If these data are not available, an approximate range for this
parameter is about 1 x 107 to 5 x 107 m/yr. This corresponds to a range
of 1 to 5 ft/s. The sixth parameter is the distance from the site to the
river in meters and is a site specific parameter.
Lines 8-12. This group of parameters are used for the USLE soil erosion calculations.
Sources of data for the necessary parameters include the local SCS office and
10-20
-------
EPA (1987). The first parameter, the USLE rainfall factor, is a measure of the
intensity and erosion potential of local/regional rainfall.
• Regional maps showing the distribution of this parameter are provided in EPA
(1987). The second parameter, the soil credibility factor, typically ranges from
0.1 to 0.4, with 0.4 representing a higher potential for erosion. This value
should be obtained from the local SCS office. The length-slope term is used
to represent the effect of the topographic slope and the length over which
erosion will occur. Figures for estimating this parameter are available in EPA
(1987). The cover factor is a measure of the fraction of surface that is bare and
the type of vegetative cover; tables of values for different types of vegetation
and percentage ground cover are listed in EPA (1987). The last. USLE
parameter, the erosion control practice factor, is related to agricultural practices
used to reduce erosion, such as contouring and irrigated furrows. Generally
this should be assumed to be 1 for waste sites unless specific measures are
taken to reduce erosion.
Lines 13-14. These parameters are related to the river surface water pathway is the flag
designating whether or not contaminated ground-water discharges to the river
(0= no ground-water discharge to the river, 1= ground-water discharge to the
river). Two methods for estimating whether ground-water discharge occurs
include evaluating water table elevation data near the stream relative to the
stream surface, and examining the stream bank for ground-water seepage faces.
Lines 15-16. These lines control the river concentration calculation periods (i.e., starting and
ending years of river concentration calculations).
Lines 17-25. This last group of parameters in this file are related to the lake surface water
pathway. The first parameter is a flag for designating whether or not the
surface water pathway in a lake should be simulated. If 0, the lake pathway is
10-21
-------
not considered, if 1 the lake pathway is simulated. The second parameter in
this group, the sediment delivery fraction, is used to relate sediment
concentrations at the bottom of a lake to surface soil concentrations at the waste
ate. If the lake sediment concentrations are known, this parameter can be
*
calculated as the ratio of the lake sediment concentrations divided by the waste
site soil concentrations. In the absence of lake sediment concentration data, »
approximate methods for estimating this parameter include: 1) USLE
calculations for different areas, including the waste site, which deliver
sediments to the lake or 2) taking the ratio of the entire lake perimeter to the
length of the perimeter which'will deliver contaminated sediments.
Line 20 represents the fraction of organic carbon in lake sediments. If the
constituent of concern is organic, the Kj for lake sediments is calculated By
multiplying Line 20 by the K^ of this constituent. If it is inorganic, the Kd is
input directly as the last line in CHEMPRP.###.
The wind velocity (in cm/hr) over the lake can be estimated from local
meteorological data. The area of the lake (in ha) is used to estimate the mass
flux to the atmosphere due to volatilization losses from the lake and also to
estimate the average wind fetch over the lake. The porosity of lake sediments
is typically greater than surface soils with an approximate range being 0.3 to
0.6. The last parameter, the average diffusion path length below the water
sediment interface, is essentially an empirical resistance parameter and values
between 5 to 10 cm are suggested.
10.3.4 Ground-water Transport Pathway Data File (GWTRANP.###)
The data for the ground-water transport pathway are shown in Figure 10-5, (file name
GWTRANP.###). The required data in this file are primarily site specific parameters. Potential
sources of the data include the U.S. Geological Survey (USGS) and any local drilling companies.
, 10-22
-------
Figure 10-5
Variables in the Ground-water Transport Pathway Data File (GWTRANP.###).
Line
1
2
3
4
5
6
, 7
8
9
10
11
12
13
14
15
16
17
Variable Type1
TITLE
R
R
R
R
R
R
R
R
Int (0/1)
R
bit (1-50)
TITLE
R
R
R
R
Description
Input file for ground-water transport pathway
. Gradient (m/m)
Conductivity (m/yr)
Porosity (dimensionless)
Bulk density (gm/cm3)
FOC (fraction of organic carbon) for aquifer
Diipenivity coefficient in x-direction (m)
Diipenivity coefficient in y-direction (m)
Diipenivity coefficient in {•direction (m)
Flag for gw source: 0 = constant; 1 = first-order decay due to teaching
Depth of aquifer (m)
Number of points and/or times to evaluate concentration in ground water
Data set for time and x-y-z location of ground-water use point
Time (years) —
x location (m)
y location (m)
z location (m)
Line 1. Title for the input file.
Line 2-6. This group of parameters characterize the hydrogeologic properties of the aquifer.
The first parameter is the regional gradient. The only reliable way to estimate
this parameter is to evaluate water level data from at least three observation wells
in the vicinity of the site. Typical values for this parameter range from about
0.0001 to 0.01, although some values will certainly be outside of this range. The
hydraulic conductivity of the aquifer material is best estimated through evaluation
of pump test data from wells in the vicinity. Ranges of hydraulic conductivity for
different types of consolidated soils are described by EPA (1987) and Freeze and
Cherry (1979). Typical values for porosity in aquifers range from about 0.05 to
0.3. Bulk density of aquifer materials range from about 1.3 to 1.8 g/cm3.
10-23
-------
The last parameter in this group is the fraction of organic carbon in the aquifer
material. The soil-water partition coefficient for the aquifer material may be
based on the organic carbon partition coefficient, K^, defined in the chemical
properties data set and an estimated fraction organic carbon in the aquifer
material. Typical values of the fraction organic carbon in deeper soils range from
0.00001 to 0.001. Under these conditions of a very low fraction organic carbon,
some organic chemicals will tend to adsorb to other materials in the soil, such as
clay. As a result, it is very difficult to predict the adsorption and effective
retardation at a site without measured chemical-material specific sorption
coefficients or plume migration data. Fortunately, for a steady state worst case
analysis without degradation in the aquifer, this parameter does not have a large
impact on the results. However, if degradation within the aquifer is considered,
this parameter is of critical importance. ~"
Lines 7-9. The next three parameters, dispersivity coefficients in the x, y and z directions,
define the amount of dilution which occurs between the site and the exposure
point. These dispersivity coefficients are used to calculate scale-dependent
dispersivities. (Dispersivity = dispersivity coefficient • distance to point of
exposure.) In the absence of site data, one rough approximation is that the
dispersivity the direction of flow is equal to one tenth (0.1) the distance to the
exposure point (Gelhar and Axness, 1983) for a typical case. As a worst case, the
longitudinal dispersivity coefficient can be estimated as approximately one
hundredth (0.01) of the distance to the exposure point. Estimates of dispersivity
coefficients in the y and z directions range from about one tenth to one third of
the dispersivity in the x direction, with one tenth being the more conservative
estimate (i.e., worst case).
Line 11. The next parameter is a flag which indicates whether or not the source term for
the ground-water pathway is constant. If the value is 0, the source term is treated
10-24
-------
as a constant. If the value is 1, the source term is approximated as decaying
exponentially where the decay coefficient is estimated based on the original mass
of chemical present and the mass loss rate due to leaching.
Line 12. The thickness of the aquifer may be available from well logs at exposure points
or monitoring wells near the site (if available). If specific data are not available,
a conservative approximation is to use the screened interval of wells at exposure
points or relatively shallow depth, on the order of 2 to 3 meters.
Line 13-17. This group of parameters define the number of and the corresponding times and
locations at which the ground-water concentrations are to be evaluated. This data
refers to time and x,y,z location is used to identify the location of the first
.. exposure point. The user should select an evaluation time appropriate for the
contaminant travel time from the source to the point of ground water use. The
velocity of contaminant movement in the ground water is based on the following
factors: hydraulic gradient, hydraulic conductivity, porosity, bulk density and soil-
water partition coefficient. The contaminant travel time from the source to the
exposure point can be estimated as the distance divided by the velocity. The
velocity of contaminant movement can be calculated as:
where,
VM = velocity of contaminant movement (m/yr),
K — hydraulic conductivity (m/yr),
J = regional gradient (m/m),
E = porosity (cmVcm3),
10-25
-------
p> = bulk density (g/cm3),
Kj - soil-water partition coefficient (ml/g).
An additional factor that should be considered when estimating the time required
for steady state conditions to be reached is the effect of dispersion on the leading
edge of the plume.
The time required for the concentration at the exposure point to reach a steady
state value can then be estimated as:
X + (250-
=
s where,
Tri as time required for a peak concentration to be
reached at an exposure point (yr),
dispx = dispersivity in the x direction (m).
If the user specifies an evaluation time less than the time required for the plume
s to reach a peak or a steady state value at the exposure point of interest, the model
will ask the user if the time required for a peak concentration to be reached should
be used or if the currently specified time is of interest.
The last four lines in the GWTRANP.### should be repeated as many times as indicated
in Line 12 for the number of time/location combinations at which the constituent concentrations
are estimated.
10-26
-------
10.3.S Infiltration Leaching and Recharge Data File (INFILTR.###)
The inputs for the infiltration, recharge and leaching calculations are shown in Figure 10-
6 (file name INFILTR.###). Literature sources for the required data in this file include EPA
(1987), and Rawls et al. (1982). These references, among others, describe average properties
of soils based on textural classification. In order to make use of these references, the soils at
the site must be identified as one of the eleven textural classifications. A county Soil Survey
usually has a map of the different regional soil types.
Figure 10-6
Variables in the Infiltration Leaching and Recharge Data File (INFILTR.###).
Line Variable Type"
Description
1 TITLE
2 R
3 R
4 R
5 Int (1-10)
6 TITLE
7 R
8 R
9 R
10 R
11 R
12 R
13 R
14 R
IS R
16 R
17 TITLE
18 R
19 R
20 R
21 STRING OF 12 REALS
22 STRING OF 12 REALS
23 STRING OP 12 REALS
24 STRING OF 12 REALS
25 Int (1-12)
26 Int (1-12)
Input file for infiltration recharge and leaching calculations
Fidd capacity of toil (vol/vol)
Wilting point of foil (vol/vol)
Depth of root zone (cm)
Number of Mil layers
Data set for foil layer i
Saturated conductivity for soil layer i (cm/hr) • ' •
Saturated water content for soil layer t (vol/vol)
Bulk density for soil layer i (gin/cm5)
Exponent b for moisture curve for soil layer t (dunensionless)
Percentage organic matter in soil for soil layer i (percent)
Percentage clay in soil for soil layer i (percent)
Percentage silt in soil for soil layer t (percent)
Percentage sand in soil for soil layer i (percent)
Ratio of first order decay for soil layer i (1/yr)
Depth of soil layer i (cm)
Precipitaion, temperature, and evaporation data for calculating recharge
Pan factor for converting Ep to PET
Latitude of site
Curve number for surface soil
Monthly precipitation Jan-Dec (cm)
Number of days with precipitation Jan-Dec
Monthly average temperature Jan-Dec (degree C)
Monthly Ep Jan-Dec (cm)
Starting month for growing season (0 for no growing season)
Ending month for growing season (0 for no growing season)
a. R = real number; Int = integer, values in parent
epi
eptable range.
10-27
-------
Line 1. Title for the input file.
Lines 2-3. These input lines describe surface soil characteristics, the first input parameter
Is the field capacity of the surface soil. This parameter typically ranges from a
low value of 0.09 for sand up to about 0.4 for clay. The second input parameter -
is the wilting point of the surface soil. This parameter typically ranges from a
low value of 0.03 for sand up to about 0.27 for clay.
Line 4. This parameter defines the depth of the root zone. If the site is used for
agriculture, specific data for different crops may be available from the Soil
Conservation Service (SCS) or County Extension Service office. If the site is
predominantly covered with grass, a root zone depth of 10 to 20 cm is suggested.
Line 5. This parameter, the number of layers, defines the number of distinct soil horizons
in the soil column. These data may be available from a county Soil Survey.
Lines 6-16. The group of input parameters must be defined for each soil layer specified.
Representative values for the first four parameters, saturated conductivity,
saturated water content, bulk density and exponent b, are described in EPA
(1987). For sand, representative values are 21 cm/hr, .395,1.5 g/cm3 and 4 for
hydraulic conductivity, water content, bulk density, and exponent b, respectively.
For clay, representative values of the same parameters are 0.06 cm/hr, 0.482,1.4
g/cm3, and 11.4. Some of these parameters are listed in Table 10-2.
The next parameter represents the ratio of the first-order decay of a constituent
in a particular soil layer relative to the decay defined in Line 29 of the new
CHEMPRP.### file.) The final parameter input for each soil layer is the
thickness of the soil layer (vertical direction). These data may be available from
a county Soil Survey or well logs in the vicinity of the site.
10-28
-------
unea o-io snouia DC repeated 10 cuii
considered.
Table 10-2. Hydraulic properties
rapona 10 uie numi
of soils classified
?er 01 sou layers
by soil texture (from EPA, 1987).
Soil Texture
Class
Sand
Loamy Sand
Sandy Loam
Loam
Silt Loam
Sand Clay Loam
Clay Loam
Silty Clay Loam
Sandy Clay
Silty Clay
Clay
Exponent
b
4.05
4.38
4.90
5.30
5.39
7.12
7.75
8.52
10.4
10.4
11.4
Field
Capacity
(cnrVcm*)
Of
0.091
0.125
0.207
0.270
0.330
0.255
0.318
0.366
0.339
0.387
0.396
Wilting
Point
(cm3/cms)
Ow
0.033
0.055
0.095
0.117
0.133
0.148
0.197
0.208
0.239
0.250
0.272
Saturated
Water Content
(cnWcm3)
0^
0.395
0.410
0.435
0.485
0.451
0.420
0.477
0.476
0.426
0.492
0.482
Saturated
Conductivity
(cm/hr)
**tit
21.00
6.11
2.59
1.32
0.68
0.43
0.23
0.15
0.12
0.09
0.06
Lines 17-24. This group of data represents the monthly average precipitation, temperature, and
evaporation data. These data should be available from a local weather station.
Typical values for the pan evaporation factor range from about 0.6 to 0.8. The
latitude of the site may be obtained from a map. If pan evaporation data are not
available, the monthly average temperature is used to calculate the Potential
Evapotranspiration (PET) using the Thornthwaite method. If the Pan Factor is
specified as 0.0, the Thornthwaite method is used to estimate PET. In either
case, monthly values of both pan evaporation and average temperature must be
input (e.g., input one or the other as 0.0 for all twelve months).
10-29
-------
Lines 25-26. The last two lines in this input file allow the user to define the length of the
growing season, which affects the runoff curve number for estimating the runoff.
(see Section 5.2 of the model documentation.)
10.3.6 Food Chain Bioaccumulation'Pathway Data File (FOODCHN.###)
The inputs for the food chain pathway are shown in Figure 10-7 (file name FOODCHN.###).
The data in this file describe the necessary parameters for calculating uptake of chemicals from
the soil, water, and air into food items. Additionally, the source of water (i.e., river, lake, or
ground water) for irrigation and cattle is specified in this file.
Figure 10-7
Variables in the Food Chain Bioaccumulation Pathway Data File (FOODCHN.###).
Line
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Variable Type*
TITLE
TITLE
R
R
R
R
R
R
R
TTTLE
R
R
R
R
R
R
R
lot (1/2/3)
Int (1/2/3)
a. R = real number: Int = inteeer
Description
fe*wl
jnpm me lof IOOQ cmm pzvK>Un
Feed and Vegetative growth parameters
Deposition interception fraction (dimensionless)
Future production (kg/m1)
Vegetable production (kg/in2)
Soil intake rate for cattle (kg/day)
Feed rate for cattle (kg/day)
Fraction of cattle feed from pasture (dimensionleu)
Water intake by cattle (I/day)
Soil erosion parameter*
Sediment delivery fraction (SDF) of agr. field (dimensionless)
Area of agricultural field (m*)
SDF to offsite field for human exposure (dimensionless)
Area of field for offsite human exposure (m2)
Fraction of organic carbon in offsite fields (ml/gin )
Bulk density of soil in offsite field (gm/cm')
Depth of applied irrigation in (m/yr)
Flag for source of irrigation water 1 = river; 2 - lake; 3 = ground water
Flag for source of cattle water: 1 = river; 2 = lake; 3 = ground water
: values in oarentheses rcDrescnt acceptable ranee.
10-30
-------
Lines 1-9. This .group of parameters in this file are related to the vegetative production and
intake of feed and water by cattle. The first parameter in this group is the
deposition interception fraction, which represents the fraction of atmospheric
deposition that lands on the vegetation. Baes et al. (1984) and Table 8.1 in
Section 8.3 summarize some of these data.
The second parameter in this group is the productivity of the'pasture used for
grazing in kg/m2. A value of 0.7 kg/m2 is suggested by Baes et al. (1984). The
third parameter in this group is the productivity of agricultural areas used for
vegetables in kg/m2. Specific values of a crop yield may be available from a
County Extension Service Office. Typical values for different types of vegetation
for various areas have been plotted by Baes et al. (1984). For leafy vegetables
and other types of produce the values range from about 0.5 to greater than 3*5
kg/m2 in a few high production areas. For grains, the values range from about
0.1 to 0.5 kg/m2.
The last four parameters in this group are related to the intake of soil, food, and
water by cattle. (Note: If the soil-to-meat and soil-to-milk partition coefficients
are as greater than 0, then the next four values are not used in the calculations,
although but they must still be defined in the input file). The first parameter is
the mass intake of soil by cattle (in kg/day). Estimates of this value range from
about 2 to 15 percent of the dry weight food intake for grazing cattle. For a
feed intake of 8 kg/day this would correspond to a soil intake of 0.16 to 1.2
kg/day. The second parameter is the feed rate for cattle in kg/day dry weight.
A value of 8 kg/day is suggested by Travis and Arms (1988). The second
parameter is the fraction of feed that may be contaminated. If the cattle derive
all of their feed from exposed pasture, then the feed fraction should be 1. If half
the feed is from uncontaminated grain then the feed fraction should be 0.5. The
last parameter in this group is the intake of water by cows in L/day; a value of
about 50 L/day is suggested.
10-31
-------
Lines 10-17. The next group of parameters in this data file represent the soil erosion delivered
to an agricultural field and off site human exposure point The first parameter
is the sediment delivery fraction (dimensionless). This parameter represents the
fraction of soil eroded from the waste site that will reach the field. Typical
values should range from 0.0 to 0.5. This value is a site specific parameter. If
gullies between the waste site and the agricultural field intercept the eroded soil
this value should be 0., while a typical worst case value would be 0.5. The next
parameter is the area of the agricultural field in m2. These two parameters,
sediment delivery fraction and area of field over which the site erosion is spread,
are defined for both the closest agricultural area and the point of off site human
exposure.
Fraction of organic carbon in the off-site field is used to estimate the partition
coefficient of organics. typical values range from 0.01 to 0.03.
The next parameter is the bulk density of the soil in the agricultural field,
typically between 1.3 and 1.8 g/cm3. The last parameter is the annual depth of
applied irrigation in m/yr. Typical values may be available from a County
Extension Service Office.
Lines 18-19. The last two parameters define the source of water for irrigation and cattle
consumption. The flags should be defined as follows; 1 for river water, 2 for
''' lake water, or 3 for ground water. The first flag defines the source of water used
for irrigation, and the second the source of water used for cattle.
10.3.7 Control Data File (CONTROL.###)
The data set for the control file, CONTROL.###, is shown in Figure 10-8. The purpose
of this file is to allow flexible control of the model runs without making extensive changes to
individual input files for each model run. This file designates how a simulation is executed,
10-32
-------
such as the selection of a WMU type and the number of chemicals and pathway(s) to be
considered. For example, if the user wishes to examine each transport pathway and to test a few
constituents of concern just to make sure that the input data files are set up properly, he/she
needs only to modify the CONTROL.### file without changing the contents other input files.
After the user is confident about the input data for the model, he/she can then set up a
CONTROL.### file to simulate all pathways for all chemicals considered. This file also
contains some variables that are not closely associated with any other input data files. Each
variable line in CONTROL.### is briefly described below.
Figure 10-8
Variables in Control Data File (CONTROL.###).
Line
1
2
3
4
5
6
7
8
9
10
11
Variable Type?
TITLE
Int (0/1}
bit (0/1)
Int (0/1)
Int (0/1)
Int (1/2/3)
Int (1/2/3/4)
R
R
R
Int (1-10)
Description
Central file
The atmospheric pathway if considered
The ground-water pathway ii considered
The surface water pathway ii considered
The food chain pathway is considered
Flag for source of human water intake: 1 = river; 2 = lake; 3 =
Flag to identify waste management unit: 1 = injection well; 2
= impoundment; 4 •= waste pile
Side width (m)
Ending time (days)
Starting time (days)
Number of chemicals to be considered
ground water
= landfill; 3
a. R = real number; Int = integer; values in parentheses represent acceptable range.
Line 1.
This is the title line of the file.
Lines 2-5. These four lines control whether a specific pathway should be considered in the
model run. The selection of pathways is not totally independent and is
determined by the relationship between each pathway. A pathway which acts as
a predecessor to another pathway must be included if the.latter pathway is
10-33
-------
considered in a model run. For example, the atmospheric deposition affects the
off-site field soil contamination in the food chain pathway and therefore should
be included if the food chain toggle is on (i.e., Line 5 = 1). On the other hand,
contaminant transport in the ground water is mainly attributed to the leachate
generated within a WMU, therefore, the atmospheric pathway can be turned off '
if only the ground-water pathway is simulated.
Line 6. The user specifies the primary drinking water source for the population near a
contaminated site. The model will use this selection to estimate human exposure
to the contaminated water taken from river, lake, or ground water.
Line 7. This line designates which WMU to simulate in a model run. The corresponding
WMU data file should be defined in a separate file. -
Line 8. This line defines the size of a WMU assuming the unit encompasses a square
area.
Line 9-10. These two variables define the period over which the average concentration of a
contaminant in the surface soil is calculated to be reported in the output file.
Line 11. Each model run can include up to 10 chemicals without exceeding the internal
constraint on the working memory of the model.
10.3.8 Landfill Characteristics Data File (LANDFIL.###)
The data set for the landfill characteristics, LANDFIL.###, is shown in Figure 10-9.
The data in this file represent landfill cover types, landfill liner types, data on layer failures, and
waste characteristics. Each of the variables in the data set is described below. The description
is based on a generic landfill unit with a vegetative cover and a synthetic liner.
10-34
-------
Figure 10-9
Variables in Landfill Characteristics Data File (LANDFIL.###).'
Line Variable Type*
Description
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Int (1-2400)
lot (1-360)
Iota/2/3)
Int (1/2/3/4/5)
R
R
R
R
R
R (0-1.0)
Int (0/1)
R
R
R
R
R
R (0-1.0)
Int (0/1)
R
R
R
R
R
R (0-1.0)
Int (0/1)
R
R
R
R
R
R (0-1.0)
Int (0/1)
Number of months
Month when capture system fails
Landfill cover type
Landfill liner type
[ Layer 1 - Vegetative Cover ]•
Thickneii - layer 1 (m)
Ana of layer (m2)
Hydraulic conductivity (em/hr)
Field capacity (vol/vol)
Saturation limit (vol/vol)
Capture system effectiveness (dimensionless)
Clay or synthetic flag
[ Layer 2 - Drainage Layer ]*
Thickness - layer 2 (m)
Area of layer (m1)
Hydraulic conductivity (cm/hr)
Field capacity (vol/vol)
Saturation limit (vol/vol)
Capture system effectiveness (di
Clay or synthetic flag
isionless)
ijonless)
[ Layer 3 - Waste Layer ]•
Thickness - layer 3 (m)
Area of layer (m*)
Hydraulic conductivity (em/hr)
Field capacity (vol/vol)
Saturation limit (vol/vol)
Capture system effectiveness (di
Clay or synthetic Gag
[ Layer 4 - Drainage Layer ]*
Thickness - layer 4 (m)
Area of layer (m2)
Hydraulic conductivity (cm/hr)
Field capacity (vol/vol)
Saturation limit (vol/vol)
Capture system effectiveness (dimensionless)
Clay or synthetic flag
a. This exhibit shows LANDFIL.MW with a vegetative cover and a synthetic liner. The number of layers is dependent on
the liner/cover design and will be different for other designs.
b. R = real number, Int = integer; value* in parentheses represent acceptable range.
c. The three lines, including the lines above and below this line, are included here only for illustrative purposes. The actual
file should not include these lines.
10-35
-------
Figure 10-9
Variables in Landfill Characteristics Data File (LANDFEL###)» (continued).
Line Variable Type*
Description
[ Layer 5 - Synthetic Liner ]*
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
R
R
R
R
R
R (0-1.0)
Int (0/1)
Int
R (0-1.0)
Int
R (0-1.0)
Int
R (0-1.0)
Int
R (0-1.0)
Int
R (0-1.0)
R
Int (1/2/3)
Int (1-70)
R
R
Thickness - layer 5 (m)
Am of layer (m*)
MvflMtilte «nuiii£ttvitv fem/ttr\
Field capacity (vol/voQ
Saturation limit (vol/vol)
Capture system effectiveness (dimensionleai)
Clay or synthetic flag
Month of failure 1
Degree of failure 1 (dimenaionle**)
Month of failure 2
Degree of failure 2 (dimensionless)
Month of failure 3
Degree of failure 3 (dimensionless)
Month of failure 4 —
Degree of failure 4 (dimenstonleu)
Month of failure 5
Degree of failure 5 (dimcnstonleu)
Waste layer initial moisture content (vol/vol)
End of waste additions (yr)
Incremental area of waste added each year (m1)
Bulk density of the waste (gnu/cm*)
a. This exhibit shows LANDFIL.*## with a vegetative cover and a synthetic liner. The number of layers is dependent on
the liner/cover design and will be different for other designs.
b. R = real number; Int = integer, values in parentheses represent acceptable range.
c. The three lines, including the blank lines above and below this line, are included here only for illustrative purposes. The
actual file should not include these lines.
Line 1. This line specifies the length of time for simulating leachate generation at the
source. The user can specify a length up to 2,400 months (200 years) for this
line.
Line 2. This line specifies the month when the capture system is abandoned or fails and
pumping is discontinued.
10-36
/C
-------
One 3. Hits line specifies the landfill cover type: 1 = vegetation; 2 = clay; and 3 =
RCRA cap. (See Section 2.1 for definitions.) In this example, cover type 1 is
used.
Line 4. This line specifies the landfill liner type: 1 = unlined; 2 = clay liner; 3 -
single synthetic liner; 4 = clay/synthetic double liner; and 5 = MTR
composite/synthetic double liner. (See Section 2.1 for definitions.) In this
example, liner type 3 is used.
At mis point, the total number of layers in the landfill being simulated is the sum
of the number of layers in the cover plus the waste layer plus the number of
layers in the liner. In this example, cover type 1 has 2 associated layers, plus the
waste layer, plus the two layers associated with liner type 3 for a total of five
layers. Each of the layers in the WMU requires the input of several parameters.
These parameters are as follows'and must be defined for each of the layers in the
system.
[ Layer 1 - Vegetative Cover ]
Lines 5-7. These lines represent the thickness, area, and hydraulic conductivity of Layer 1.
Note that the area (Line 6) refers to the areal extent of the landfill upon
completion as opposed to the incremental and continually expanding area during
i
! landfill expansion and operation.
Line 8. This line defines the field capacity of the layer. In cases where either the layer's
initial moisture level or wilting point is known, enter the difference between the
soil's field capacity and either the initial moisture level or wilting point.
- -. Examples are as follows:
c >
v . • Initial moisture level of soil is known, then enter field capacity minus
initial moisture level;
10-37
-------
Initial moisture level of soil is not known, but wilting point of the soil is,
then enter field capacity minus wilting point; and
Neither initial moisture level or wilting point is known, enter field
capacity of soil.
Line 9. This line defines the saturation limit of the layer, which represents the maximum
moisture holding capacity of the layer.
Line 10. This line defines the leachate/recharge capture system's effectiveness. Enter zero
if no capture exists in this layer. Otherwise enter the fraction of leachate
removed. This parameter is created mainly for the drainage layer. (See Layer
4 below for an example.)
Line 11. This parameter tells the program whether or not .this layer is a clay or synthetic.
Zero means the layer is not a clay or synthetic. If it were a clay or synthetic, the
user must enter "1" and sequentially define the failure scenarios. (See examples
for Layer 5 below).
[ Layer 2 - Drainage Layer }
Lines 12-18. These are the parameters describing Layer 2 (see above for definitions).
[ Layer 3 - Waste Layer ]
Lines 19-25. These are the parameters describing the waste layer. For field capacity, enter the
field capacity of the waste layer. Further input will prompt the user for the
actual waste moisture level (Line SO) at a later time.
[ Layer 4 - Drainage Layer ]
Lines 26-32. These are the parameters describing Layer 4. For capture system effectiveness,
enter the fractional effectiveness of the capture system. For example, a value of
0.75 (i.e., 75 percent) indicates that this layer captures 75 percent of the
10-38
-------
recharge/leachate volume, which would have moved downward into Layer 5 had
this layer not existed.
[ Layer 5 - Synthetic Liner ]
Lines 33-39. These are the parameters describing Layer 5 (as defined above). Since Layer 5
is a synthetic layer (as indicated in Line 39), the user must define five different
failure release scenarios by specifying the years and degrees of failure in the
subsequent lines.
Lines 40-49. Five sets of failure scenarios are defined in these ten lines. The first line of each
set specifies the time (in months) of failure, and the line that follows defines the
degree of failure. Note that the degree of failure lets the user define failure
release scenarios by quantitatively assigning release magnitude in much the same
way as defining capture effectiveness. For example, a value of 0.1 represents 10
percent of the leachate "pooled" above the synthetic would be allowed to release
in the month specified in the previous data line. Also note that if less than five
failure release scenarios are desired during the simulation, the remaining failure
scenarios should be set at a month equal to the maximum number of months, or
one month longer than the planned simulation (in this example, 841).
At this point, all five layers have been defined. Care should be taken to correctly
and accurately address each piece of needed information, and to account properly
for the capture systems and failure scenarios.
Line 50. This line defines the waste layer's initial moisture content. This parameter allows
the user to simulate the release of leachate from a waste which was wet to some
degree upon emplacement into the landfill system.
Line 51. The user specifies -the method for calculating leachate concentration: 1 =
partitioning approach; 2 = completely-mixed reactor approach; and 3 - constant
10-39
-------
strength approach.
Line 52. Eater the year in which waste is no longer added to the landfill, i.e., landfill is
closed.
Line 53. Enter the incremental axea of waste added each year. Note that the thickness of :-
the waste layer has been previously defined; therefore, if only the incremental
volume is known on a yearly basis, one can calculate the incremental area by
dividing the incremental volume by the thickness.
Line 54. Enter the bulk density of the waste.
10.3.9 Surface Impoundment Characteristics Data File (IMPOUND.###) ~"
The data set for the surface impoundment characteristics, IMPOUND.###, is shown in
*•
Figure 10-10. The data in this file represents the cover types (after being converted to a
landfill), liner types, layer failures, and waste characteristics. Most of the variables are similar
to the landfill characteristic data set variables, especially the parameters associated with each
liner/cover layers.
A surface impoundment may be closed as a landfill upon termination of its operation.
If that is the case, as shown in this example, the user must define the landfill cover design and
the time at which it is installed. If the impoundment is never converted to a landfill, set the
landfill cover type to "1," define all soil specific layer parameters with zeros or default values
for the two layers associated with cover type 1, and define the year in which conversion occurs
as some year larger than the actual simulation period (e.g., enter 75 years even though the
simulation period is only 70 years). Because of the similar file structures between
IMPOUND.### and LANDFIL.###, the data lines are only briefly described below.
1CMO
-------
Figure 10-10
Variables in Surface Impoundment Characteristics Data File (IMPOUND.###).'
Line Variable Type*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IS
19
20
21
22
23
24
25
26
27
28
29
30
31
32
lot (1-2400)
lot (1-360)
bit (1/2/3)
Int (1/2/3/4/5)
R
R
R
R
R
R (0-1.0)
lot (0/1)
R
R
R
R
R
R (0-1.0)
Int (0/1)
R
R
R
R
R
R (0-1.0)
Int (0/1)
R
R
R
R
R
R (0-1.0)
bit (0/1)
Number of months
Month when capture lyttem fails
Landfill cover type
Landfill liner type
[ Layer 1 - Vegetative Cover ]*
Thickness - layer 1 (m)
Area of layer (m2)
Hydraulic conductivity
Field capacity
Saturation limit
. Capture system efibcthrcaeM
Cky or synthetic flag
[ Layer 2 - Drainage Layer ]'
Thickneii - layer 2
Ana of layer
Hydraulic conductivity
Field capacity
Saturation limit
Capture syitem effectiveness
Clay or synthetic flag
[Layer 3-Waste Layer]'
Thickneii - layer 3
Area of layer
Hydraulic conductivity
Field capacity
Saturation limit
Capture lyttem effectiveness
Clay or synthetic flag
[ Layer 4 - Drainage Layer J*
Thickness - layer 4
Area of layer
Hydraulic conductivity
Field capacity
Saturation limit
Capture system effectiveness
Clay or synthetic flag
a. This exhibit shows LANDFIL.### with a vegetative cover and a synthetic liner. The number of layers is dependent on
the liner/cover design and will be different for other designs.
b. R = real number; Int = integer; values in parentheses represent acceptable range.
c. The three lines, including the blank lines above and below this line, are included here only for illustrative purposes. The
actual file should not include these lines.
1041
-------
Figure 10-10
Variables in Surface Impoundment Characteristics Data File (IMPOUND.###y (continued).
Line Variable Type*
Description
[ Layer 5 - Synthetic Liner )*
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
fn
50
51
52
53
54
55
56
57
58
R
R
R
R
R
R (0-1.0)
Int (0/1)
bit
R (0-1.0)
lot
R (0-1.0)
Int
R (0-1.0)
lot
R (0-1.0)
Int
R (0-1.0)
R
R
R
R
R
R
Int (1/2/3)
R
Thidmen - layer S
Area of layer
Hydraulic conductivity
Field capacity
SfltUEBtlOfl uflixt
Capture system effectiveness
Clay or lynthctic flag
Month of failure 1
Degree of failure 1
Month of failure 2
Degree of failure 2
Month of failure 3
Degree of failure 3
Month of failure 4 —
Degree of failure 4
Month of failure 5
Degree of failure 5
car of conversion to lananii
Initial pond volume
Maximum allowable pond volume
Monthly influx to pond
Area of pond surface
Yearly evaporation
Initial waste moisture level
Mfithwt for calculating tetchi1** conwritn tior "
Bulk density of the waste
a. This exhibit shows LANDFIL.*!f with a vegetative cover and a synthetic liner. The number of layers is dependent on
the liner/cover design and will be different for other designs.
b. R = real number; Int = integer, values in parentheses represent acceptable range.
c. The three lines, including the blank lines above and below this line, are included here only for illustrative purposes. The
actual file should not include these lines.
s
Lines 1-4. These four lines are the same for both the surface impoundment and landfill
units. Consult the definitions and use of these parameters in the previous section
(Section 3.3.2).
10-42
/ft
-------
Lines 5-49. The first 35 lines describe the properties of five layers, and the next 10 lines
describe the five different failure scenarios, similar to landfill characteristic data
set discussed in Section 3.3.2.
Line 50. Enter the year in which the surface impoundment is converted to a landfill. If
the impoundment is never converted to a landfill, specify a year beyond the
modeling period as indicated in Line 1. (Note that the time units are different for
this line (in years) and Line 1 (in months).)
Line 51. Enter the volume of water in cubic meters stored in the surface impoundment at
the beginning of the simulation.
Line 52. Enter die maximum volume of liquid waste in cubic meters which can be stored
in the surface impoundment without any overflow.
Line 53. Enter the influx of liquid waste in cubic meters added to the pond for storage.
Line 54. Enter the area of the pond surface in meters squared.
Line 55. Enter the yearly surface evaporation in centimeters per year from a surface
evaporation atlas for the location of the site being modeled.
Line 56. Enter the moisture level associated with the waste at the time the surface
impoundment is converted to a landfill.
Line 57. Specify the method for calculating leachate concentration. This parameter allows
the user to choose approaches for calculating concentration of the leachate when
the unit is converted to a landfill. The three choices include: 1 = partitioning
approach; 2 = completely-mixed reactor approach; and 3 = constant strength
10-43
-------
approach. •
Line 58. Bulk density of the waste (after converting to a landfill).
10.3.10 Waste Pile Characteristics Data File (WSTPILE.###)
The data set for waste pile characteristics (WSTPILE.###) is shown in Figure 10-11.
The data set characterizing waste piles is simplified, as runoff is estimated within the model
based on the Soil Conservation Service Curve Numbers (the same method used in the earlier
version of MMSOILS). Note that no cover or liner was designed for a waste pile. As the data
in Figure 10-11 is rather self explanatory (after the user is familiar with the LANDFIL.### and
IMPOUND.### files), no further description is provided.
Figure 10-11
Variables in Waste Pile Characteristics Data (WSTPILE.###).
Line
1
2
3
4
5
6
a. R = real nun
Variable Type*
Int (1-2400)
Int (1-360)
Int (1-70)
R
R
R
tiber: Int — inteecr
Description
Number of months
Month capture system is abandoned or failed (pumping stops)
Year waste pile is closed
Incremental surface area nMfff each year pile is opcnuinK (n>)
Bulk density of waste (gm/cm1)
Thickness of waste (no)
: values in narentheses rcnresent accentable ranve.
10.3.11 Injection WeU Characteristics Data FUe (INJECTW.###)
The data set for injection well characteristics, INJECTW.###, is shown in Figure 10-12.
Line 1 is the flag for the type of well failure desired: 0 = grout failure; 1 = well casing
10-44 _
.
-------
failure; and 2 = use the more conservative approach between grout and well casing failures.
Hie next three lines (Lines 2-4) require the user to specify the hydraulics of the injection well,
i.e., pressure head, pump head, and injection rate. Line 5 is the flag for the type of well
considered, which affects the equation used for estimating the waste release rate from grout or
casing failures. A flag of "1" represents the water flood (secondary recovery) type, and "2" for
dedicated disposal type. Finally, the user has the option to choose between the best-estimate
solution (Line 6 = 1) and the worst case solution (Line 6 = 2) to estimate the waste release
rate.
Figure 10-12
Variables in Injection Well Characteristics Data File (INJECTW.###).
Line
1
2
3
4
5
6
Variable Type
Int (0/1/2)
R
R
R
Int (1/2)
bit (1/2)
Description
Flag for type of well failure desired
Injection rate (mVyr)
Pump head (m)
Flag for type of well
Flag for best case or wont case solution
a. R = real number; Int - integer; values in parentheses represent acceptable range.
10.3.12 Tank Release Characteristics Data File (TANKREL.###)
The data set for tank release characteristics, TANKREL.###, is shown in Figure 10-13.
Line 1 indicates number of months used for leachate calculation. Line 2 denotes number of
years of tank release. The following lines will indicate the annual release in mVyr, for as many
number of years indicated in Line 2. Please note that a variable annual release from the tanks
are acceptable to the model simulation.
10-45
-------
r
Figure 10-13
Variables in Tank Release Characteristics Data File (TANKREL.###).
Line Variable Type
Description
1
2
3
4
Int
Int
R
R
Number of months used for tachate generation calculations (months).
Number of yeui of Unk release (years).
Tank release in year 1 (mVyr).
Tank release hi year 2 (m'/yr).
a. R = real number; Int = integer, values in parenthi
epresent acceptable range.
1(M6
-------
11.0
Baes, C.F. 1982. Prediction of radionuclide KJ values from soil-plant concentration ratios.
Trans. Amer. Nuclear Society, 41, pp.53-54.
Baes, C.F., R.D. Sharp, AX. Sjoreen, and R.W. Shor. 1984. Review and Analysis of
Parameters for Assessing Transport of Environmentally Released Radionuclides Through
Agriculture. ORNL-5876, Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Briggs, G.G., R.H. Bromilow, and A.A Evans. 1982. Relationships between lipophilicity and
root uptake and translocation of non-ionized chemicals by barley. Pestic Sci., 13, pp.
495-504. ~
Bysshe, S.E. 1982. Bioconcentration factor in aquatic organisms. In, Handbook of Chemical
Property Estimation Methods, ed. by WJ. Lyman, W.F.Reehl, and D.H. Rosenblatt.
McGraw-Hill, New York.
Campbell, G.S. 1974. A simple model for determining unsaturated conductivity from moisture
retention data. Soil Science, Vol. 177, pp. 311-314.
Cohen, Y. 1989. University of California, Los Angelels. Personnel communication regarding
gas film mass transfer coefficients.
Dunne, T. and L.B. Leopold. 1978. Water in Environmental Planning. W. H. Freeman and
Co., New York, New York.
Freeze, R.A. and J.A. Cherry. 1979. Groundwater. Prentice-Hall, Englewood Cliffs, New
Jersey.
11-1
Si
ll
-------
Garten, C.T. and J.R. Trabalka. 1983. Evaluation of models for predicting terrestrial food chain
behavior of xenobiotics. Environmental Science and Technology. Vol. 10, pp. 590-595.
Gelhar, L.W. and C.L. Axness. 1983. Three-Dimensional Stochastic Analysis of
Macrodispersion in Aquifers. Water Resour. Res., Vol. 19, No. 1, pp. 161-180.
Gelhar, L.W., A. Mantoglou, C. Welty, K.R. Rehfeldt. 1985. A Review of Field-Scale
Physical Solute Transport Processes in Saturated and Unsaturated Porous Media. EPRI
EA-4190, Electric Power Research Institute, Palo Alto, CA.
Heller, P.R., G.W. Gee, and D.A. Myers 1985. Moisture and Textural Variations in
Unsaturated Soils/Sediments Near the Hanford Wye Barricade. PNL-5377. Pacific
Northwest Laboratory, Richland, Washington. ""
Holton, G.A., C.C. Travis, EX. Etnier, C. Cook, F.R. O'Donnell, D.M. Hetrick, and E.
Dixon. 1984. Multiple Pathway Screening-Level Assessment of a Hazardous Waste
Incineration Facility. ORNL/TM-8652. Oak Ridge National Laboratory, Oak Ridge,
Tennessee.
Horst, T.W. 1984. The modification of plume models to account for dry deposition. Boundary-
Layer Meteorology, 30, pp. 413-430.
Hunt, B. 1978. Dispersive Sources in Uniform Groundwater Flow. Jou. of the Hyd. Div.
ASCE, Vol. 104(Hyl), pp. 75-85.
Hwang, S.T. 1987. Multimedia approach to risk assessment for contaminated sediments in a
marine environment. Superfund '87, Proceedings of the Eighth National Conference.
Nov. 1987, Washington, D.C., pp 485-491.
11-2
I '
-------
Kenaga, £.£.,. 1980. Correlation of bioconcentration factors of chemicals in aquatic and
terrestrial organisms with their physical and chemical properties. Environmental Science
and Technology, Vol 14 (5), pp. 553-556.
Lagoy, P.K. 1987. Estimated soil ingestion rates for use in risk assessment. Risk Analysts, 7(3),
355.
Liss, P.S. and P.O. Slater. 1974. Flux of gases across the air-sea interface. Nature, 247, pp.
181-184.
Lyman, W.J., W.F. Reehl, and D.H. Rosenblatt. 1982. Handbook of Chemical Property
Estimation Methods. McGraw-Hill, New York.
Mackay, D. 1985. Air-water exchange coefficients. In, Environmental Exposure from
Chemicals, Vol. 1, edited by, Neely, W.B. and G.E. Blau. CRC Press Inc., Boca
Raton, Florida.
Maybey, W.R. et al. 1982. Aquatic Fate Process Data for Organic Priority Pollutants, SRI
International, U.S. EPA. Contract Nos. 68-01-3867 and 68-033-2981. EPA-44B/4-81-
014.
Moghissi, A.A., R.R. Marland, F.J. Congel, and K.F. Eckerman. 1980. Methodology for
environmental human exposure and health risk assessment. In, Dynamics, Exposure, and
Hazard Assessment of Toxic Chemicals, edited by R. Haque. Ann Arbor Science.
Chapter 31.
Neely, W.B. and G.E. Blau. 1985. Environmental Exposure from Chemicals, Vol. 1 and 2.
CRC Press Inc., Boca Raton, Florida.
Pasquill, F. 1974. Atmospheric Diffusion, John Wiley and Sons, London.
11-3
-------
Peterson, S. 1987. Chemical Data Base, prepared,by ICF-Northwest. Letter report to G.
Whelan, Pacific Northwest Laboratory, Richland, Washington, September 1, 1987.
Pionke, H.B. and R. J. DeAngelis. 1980. Method for distributing pesticide loss in field runoff
between the solution and adsorbed phase. Chapter 19 in CREAMS: A Field Scale
Model for Chemicals, Runoff, and Erosion from Agricultural Management Systems,
edited by Knisel, W.G. U.S. Department of Agriculture, Conservation Research Report
No. 26, Tuscon, Arizona.
Rathbun, R.E. and Tai D. Y., 1986. Gas-film coefficients for the volatilization of Ketones from
water. United States Geological Survey Water-Supply Paper 2286.
Rawls, W.J., D.L. Brakensiek, K.E. Saxton. 1982. "Estimation of Soil Water PropertiesT*
Transactions of the American Society of Agricultural Engineers. Vol. 25, No. 5, pp.
1316-1320 and 1328.
Tetra Tech. 1992. "Addition of VADOFT Transport to MMSOILS (Draft)." Prepared for
U.S. EPA Environmental Research Laboratory, January 1992.
Thomthwaite, C.W. and J.R. Mather. 1955. "The Water Balance." Publications in
Climatology. Vol. Vm, No. 1, Drexel Institute of Technology, Laboratory of
Climatology, Centertron, New Jersey.
Travis, C.C. G.A. Holton, EX. Etnier, C. Cook, F.R. O'Donneli, D.M. Hetrick, and E.
Dixon. 1986. Assessment of inhalation and ingestion population exposures from
incinerated hazardous wastes. Environmental International, Vol. 12, pp. 533-540.
\
Travis, C.C. and A.D. Arms. 1988. Bioconcentration of organics in beef milk and vegetation.
Environmental Science and Technology, Vol. 22(3), pp. 271-274.
11-4
-------
U.S. EPA. 1979. Water-Related Environmental Fate of 129 Priority Pollutants, Vol. I and H.
U.S. Environmental Protection Agency, Washington, D.C. EPA-440/4-79-029 a,b.
U.S. EPA. 1980a. Physical and Chemical Properties of Hazardous Waste Constituents. U.S.
Environmental Protection Agency, Environmental Research Laboratory, Athens, Georgia.
U.S. EPA. 19805. Water quality criteria documents: Availability. 45 Fed. Reg. 79318-79379.
U.S. Environmental Protection Agency, Washington, DC.
U.S. EPA. 1984a. User's manual for the pesticide root zone model (PRZM): Release 1. U.S.
Environmental Protection Agency, Athens, GA, EPA-600/3-84-109.
U.S. EPA. 1984b. Technical Guidance Manual for Performing Wasteload Allocations. Book
n Streams and Rivers. Chapter 3, "Toxic Substances." .U.S. Environmental Protection
Agency, Washington, D.C., EPA-440/4-84-022.
U.S. EPA. 1985a. Rapid Assessment of Exposure to Paniculate Emissions from Surface
Contamination Sites. Office of Health and Environmental Assessment, Exposure
Assessment Group, U.S. Environmental Protection Agency, Washington, D.C.,
EPA/600/8-85/002, NTIS PB85-192219.
U.S. EPA. 1985b. Compilation of Air Pollutant Emission Factors Volume I: Stationary Point
and Area Sources, Fourth Edition. AP-42 U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina.
NTIS PB86-124906
U.S. EPA. 1986a. Development of Advisory Levels for Polychlorinated Biphenyls (PCBS)
Cleanup. Office of Health and Environmental Assessment, Exposure Assessment Group,
U.S. Environmental Protection Agency, Washington, D.C., EPA/600/6-86/002, NTIS
PB86-232774.
11-5
-------
U.S. EPA. 1986b. Mathematical Model Selection Criteria for Performing Exposure
Assessments: Groundwater Contaminants from Hazardous Waste Facilities. Draft
Technical Support Document. Exposure Assessment Group, U.S. EPA, Washington,
D.C. '
U.S. EPA. 1986c. Methods for Assessing Exposure to Chemical Substances. Volume 8
Methods for Assessing Environmental Pathways of Food Contamination. U.S.
Environmental Protection Agency, Washington, D.C., EPA 560/5-85-008, NTIS PB87-
107850.
U.S. EPA. 1986d. Superfund Public Health Evaluation Manual. U.S. Environmental Protection
Agency, Washington, D.C., EPA/540/1-86/060.
U.S. EPA. 1987. Superfund Exposure Assessment Manual. Draft report. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington,
D.C.
U.S. EPA. 1988a. Estimating Exposures to 2,3,7,8-TCDD. External Review Draft. Office
of Health and Environmental Assessment, Exposure Assessment Group, U.S.
Environmental Protection Agency, Washington, D.C., EPA/600/6-88/005a.
U.S. EPA. 1988b. Exposure Factors Handbook, Draft Report. Office of Health and
Environmental Assessment, Exposure Assessment Group, U.S. Environmental Protection
Agency, Washington, D.C.
U.S. EPA. 1989a. Reference Dose (RfD): Description and Use in Health Risk Assessments.
Integrated Risk Information System (IRIS) on line. Inter-Agency Reference Dose (RfD)
Work Group, Office of Health and Environmental Assessment, Criteria and Assessment
Office Cincinnati, OH.
11-6
-------
U.S. EPA. 19896. Integrated Risk Information System (IRIS) on line. Office of Health and
Environmental Assessment, Criteria and Assessment Office Cincinnati, OH.
U.S. EPA. 1989c. Hazardous Waste Treatment, Storage, and Disposal Facilities (TSDF) -
Fugitive Paniculate Matter Air Emissions Guidance Document. EPA-450/3-89-019,
Office of Air Quality Planning and Standards, Research Triangle Park, NC.
U.S. EPA. 1990. Background Document for EPA's Composite Model for Landfills
(EPACML), Office of Solid Waste, February 1990.
U.S. EPA/O.R.D./E.R.L.-Athens. 1992a. Summary of Review and Evaluations of the
Technical Approach to Corrective Action Regulatory Impact Analysis: Fate and
Transport." —
Wischmeier, W.H. and D.D. Smith. 1965. Predicting Rainfall-Erosion Losses from Cropland
East of the Rocky Mountains. Agricultural Research Ser., USDA Handbook No. 282.
Whelan, G., D.L. Strenge, J.G. Droppo, B.L. Steelman and J.W. Buck. 1987. The Remedial
Action Priority System (RAPS): Mathematical Formulations. DOE/RL/87-09, PNL-6200,
UC-11. Pacific Northwest Laboratory, Richland, Washington.
11-7
-------
-------
12.0 LIST OF SYMBOLS
ADS = Total amount of chemical administered per day (mg/day),
ADV = Total amount of chemical administered per occurrence (mg),
AET = actual evapotranspiration (cm),
al = length of the volume source in the x direction (m),
AS = Fraction of chemical absorbed through the skin (dimensionless),
As = surface area of the site (hectares),
AW - available soil moisture (cm),
AWC = available water capacity of the soil (cm),
b = soil specific exponent parameter representing the moisture
retention relationship (dimensionless),
BCF = Bioconcentration factor for species in experimental study (ml/g
€5
fish),
BCFf — Bioconcentration factor for fish (ml/g fish),
BCF = Bioconcentration factor for species of concern (ml/g fish),
sp
bl = width of the volume source in the y direction (m),
BSF = sediment-to-fish partition coefficient (mg/kg fish / mg/kg soil),
Bv = soil-to-plant concentration factor (kg soil/kg dry plant),
BW = Body weight (kg),
Ca = Concentration of chemical in air (mg/m3),
Ca = soil concentration in agricultural field (mg/kg),
CD = drag coefficient (0.00237 for wind speeds 4-12 m/s),
GDI = Average daily intake of the chemical (mg/kg/day),
Ct — chemical concentration in fish (mg/kg),
CF — cover factor, 1.0 for bare soil (dimensionless),
Cfl — chemical concentration in cattle (mg/kg),
Cjz = chemical concentration in milk (mg/kg),
Cjttd - chemical concentration in feed used for cattle (mg/kg),
Cfirf - Concentration of chemical in food (mg/kg),
C,,, = concentration in groundwater at exposure point (mg/1),
12-1
-------
C, = depth of the volume source in the z direction (m),
CL = the concentration of contaminant in the leachate (mg/1),
Cfc, = dissolved chemical concentration lake water (mg/1),
CN = curve number representing infiltration capacity of soil (dimensionless),
C0 = fixed concentration in the volume source (mg/1),
Cn, = concentration in river water (mg/1),
C, = soil adsorbed contaminant concentration level (mg/kg),
Cted - concentration in bottom sediments (mg/kg),
Cti = liquid-phase chemical concentration in soil (g/cm3),
€„ = saturation vapor concentration of chemical (g/cm3),
Cv = chemical concentration in vegetables (mg/kg),
C* = chemical concentration on vegetation due to deposition (mg/kg),
€„ — chemical concentration in vegetation due to uptake from the soil
(mg/kg),
Cw = Concentration of chemical in water (mg/liter),
d = thickness of cover (cm),
D = phase transfer coefficient (cm2/s),
Dd = the overland distance between the site and the receiving water body
(meters),
DP . = Fraction of day during which exposure occurs (hours/24 hours),
DH = drop height (m),
D, = diffusion coefficient of compound in air (cm2/s),
Ds = soil delivery rate to agricultural field (kg/year),
Die = sediment dilution ratio, between .0 and 1 (dimensionless),
Dw = chemical diffusivity in water (cnrVsec),
Dx «= dispersion coefficient in the x direction (mVyr),
c
ds = source depletion factor due to deposition flux (dimensionless),
(mVyr),
Dy = dispersion coefficient in the y direction (nWyr),
D; - D/K, (mVyr),
dz = depth of dry zone (cm),
12-2
-------
Dz =s dispersion coefficient in the z direction (mVyr),
Dt' m D& (nWyr),
E = soil porosity (cm3/cm3),
Emv = emission factor from an unpaved road per vehicle-kilometer of
travel (kg/VX7),
E{ = lake sediment porosity (dimensionless),
Erf = emission factor from loading operations (kg/Mg),
£„, = emission factor from spreading operations (kg/ha),
Ep = monthly pan evapotranspiration (cm/month),
ET = monthly evapotranspiration (cm/month),
Ed = the unadjusted PET (cm/month),
f() = a functional relationship,
F(x) = plotted function in U.S. EPA (198Sa) Fig.4-3 pg 36,
F = average wind fetch, assumed to be the half width of the pond (cm),
F; = feed-to-meat transfer factor for cattle (day/kg),
F2 = feed-to-milk transfer factor for cattle (day/kg),
fit = frequency of the specific stability array parameters for
classification (ij),
ft* - fraction of wet plant remaining as dry material (g/g),
H = Henry's law constant (atm»m3/mol),
ti — Henry's law constant in concentration form (dimensionless),
h , = average depth of water (cm),
/ = annual heat index (dimensionless),
IA = initial abstraction representing surface storage (cm),
Ia = Inhalation rate of air for human (mVday),
If = Ingestion rate of food for human (kg/day),
IN = Soil ingestion rate (mg soil/day),
/„ = Ingestion rate of water for human (liters/day),
K = hydraulic conductivity (m/yr),
k — pan correction factor dependent upon vegetation cover and type
(dimensionless),
12-3
-------
— soil-water partition coefficient (ml/g),
= sediment-to-water partition coefficient for late sediments (1
water/kg sediment),
kt ~ sediment-phase mass transfer coefficient (cm/hr),
KF = soil credibility factor (metric tons/ha/unit R),
' = gas-phase mass transfer coefficient for H2O (3000 cm/hr),
kt = gas-phase mass transfer coefficient (cm/hr),
KL = first decay constant (1/yr),
%
l(COi) = liquid-phase mass transfer coefficient for Ctij (20 cm/hr),
kf - liquid-phase mass transfer coefficient (cm/hr),
kL = overall water to air mass transfer coefficient (cm/hr),
kp = particle size multiplier (dimensionless), • _
K, — the saturated hydraulic conductivity of the specific soil type
(cm/yr),
kw = water-phase mass transfer coefficient (cm/hr),
J = regional gradient (m/m),
L ~ Length of exposure period (years),
LC0 = lipid content for species in experimental study (mg/mg),
LQ = lipid content for species of concern (mg/mg),
Ld = total depth of contamination from soil surface (cm),
LF = Length of time during exposure period during which exposure
occurs (years),
LS = length - slope factor (dimensionless),
M = material moisture content (percent),
Ma = mass flux of contaminant into the atmosphere (g/s),
Mj = mass of soil in the mixing depth of the field (mixing depth x
area x bulk density) (kg), __ '".__
MC,. — contaminant mass loading rate due to erosion (g/year),
Mw = total mass flux rate of contaminant entering the stream from both
soil erosion and groundwater discharge (g/year),
12-4
-------
MW = mole weight of compound (g/mol),
MW
(COj m molecular weight of COj (44 g/mol),
ffj O) - molecular weight of H2O (18 g/mol),
MW — molecular weight of chemical (g/mol),
N = mass flux (g/s/cm2),
ND = Number of days during exposure period L (dimensionless),
P = density thickness of the root zone (bulk density of the soil times
the rooting depth) (kg/m2),
pe = number of days per year with at least 0.01 inches of precipitation
(days),
PET = potential evapotranspiration (cm),
PF = erosion control practice factor, I'.O for uncontrolled waste site
(dimensionless),
P{ — depth of precipitation from a storm of 24 hour duration (cm),
PM10 — emission rate for inhalable particulates, less than 10 pm, (g/hr),
pv — vapor pressure of compound (mm Hg),
Pj = soil-to-meat partition coefficient (mg/kg meat / mg/kg soil),
P2 = soil-to-milk partition coefficient (mg/kg milk / mg/kg soil),
q = the annual average recharge rate (cm/yr),
qn = monthly net recharge (mVmonth),
Q — average flow rate within the stream (mVyear),
£jfe«f = consumption rate of feed for cattle (kg/day),
Qr = depth of runoff from a storm of 24 hour duration (cm),
Qs = consumption rate of soil for cattle (kg/day),
Qw = consumption rate of water for beef (I/day),
Qm = monthly net recharge (m3/ month),
Qsmu - monthly flow that is spilled (mVmonth),
p***
Qcoli = monthly flow that is collected (mVmonth),
CLi « monthly flow out of the bottom of the WMU (mVmonth),
R = gas constant (8.2 x 1048 atm«m3/mol-cK),
r = fraction of flux captured by vegetation (dimensionless),
12-5
•\
i-r-
-------
RCF = root concentration factor for vegetable? (g wet root/ml),
Rd = retardation factor (dimensionless),
RF — the rainfall factor, expressing the erosion potential of average
annual rainfall in the locality (metric tons-meter/ha-hour),
r, = diffusion path length below the water sediment interface (cm), *-
^ = molar gas constant (62,361 [mm Hg cm3]/[mol °K]), . v
Rj = depth of root zone (cm),
5 = percent silt content of road surface (percent),
5 = mean vehicle speed (km/hr),
SC — Amount of soil contacted per visit (g soil/visit),
Sd = sediment delivery ratio (dimensionless),
SDf = soil delivery fraction from waste site to agricultural field
(dimensionless), ~
Sm = maximum potential retention (cm),
S0 = initial storage in the WMU (m3),
AS = change in storage (m3),
T = absolute temperature (°K),
t - time period of plant exposure (time),
T* = mean monthly air temperature (°C),
I> = time (yr),
u = ug/R4 (m/yr),
ug - regional pore velocity in the x direction (m/yr),
u, =^ average stream velocity (m/yr),
u, — threshold velocity at 7 m height (m/s),
uw = average wind speed (m/s),
V7 = Total number of visits to the site or total number of days at
residence (days),
V^ = velocity of contaminant movement through the vadose zone
(cm/yr),
vd = deposition velocity (m/s),
12-6
-------
VG — fraction of contaminated, surface with vegetative cover
(dimensionless),
VKT = average vehicle kilometers of travel per hour on unpaved road (#
veh. x km/nr),
' - V^ = steady pore velocity in the vadose zone (cm/yr),
st 0 Vw - wind velocity at 10 meters above surface (cm/min),
w = mean number of wheels per vehicle (#),
W = mean vehicle weight (Mg),
X = distance in x coordinate direction (parallel to velocity u) from
source to point of interest (m),
x = dimensionless ratio (x =0.886 uVu),
XF1 = units conversion factor 106 mg/kg per g/g),
XF2 * units conversion factor (10* mVha),
XF3 - conversion factor (1000 kg/metric tons),
XF4 = Conversion factor (1 kg/106 mg),
XF5 = Conversion factor (1 kg/lOOOg),
xv = virtual distance required for point source plume to spread to width
of site (m),
7 SB. distance in y coordinate direction (perpendicular to velocity u)
from source to point of interest (m),
YD = dumping device capacity (m3),
yw = sediment yield (metric tons/hectare/year),
YV = vegetative density (kg/m2),
Z — distance in z coordinate direction (perpendicular to velocity u)
from source to point of interest (m),
X = effective diffusion parameter (cnWs),
X, = a loss rate coefficient representing chemical degradation and
movement through the root zone (I/time),
Xw = weathering loss rate coefficient (I/time),
pa - density of air (1.2 x ID40 g/cm3),
PB = bulk density of soil (g/cm3),
12-7
-------
p, — particle density of soil (g/cm3),
pw — density of water (1.0 g/cm3),
oy . = mixing coefficient in y direction, standard deviation of Gaussian
plume (m),
-------
APPENDKA
Verification of Ground water Transport Model A-l Partially Saturated Zone Calculations
Tetra Tech performed several tests of the PSAT routine during the process of
implementing the VADOFT code into MMSOILS (Tetra Tech, 1992). Since that document is
not widely available, their results are reproduced here for completeness.
The PSAT routine was compared with two other solute transport models, ROAM and
MYGRT. ROAM is a finite-difference unsaturated zone transport model with similar boundary
conditions to the MMSOILS/VADOFT model and MYGRT is an analytical groundwater
transport model. The comparisons of these models if for two of the Poppy SWMU 007 cases
with large transport times, in order to test the validity of the new MMSOILS model for these
cases. Only a single layer unsaturated zone is used in these comparisons in order to allow for
the use of the analytical model.
Figure A-l presents these three model results for Poppy SWMU 007 with an unsaturated
zone thickness of 8.3 m and a retardation factor of 224. All input data for the three models was
the same, except that discretization and time steps were also required in the numerical model.
all three models show very similar responses. The two numerical models are essentially the
same, and differ only slightly from the analytical model. These results tend to confirm the
general response of the new MMSOILS model. Figure A-2 shows the same comparison for an
s
unsaturated zone of 103 m and a retardation factor of 224. The results in Figure A-2 confirm
the observations made in Figure A-l, as the numerical model response is shifted slightly to
smaller times.
The deviation between the numerical models and the analytical model is expressed
primarily by a shift of the model response to slightly smaller value of times. It is believed that
the differences arise because of the differing boundary conditions in the models. The no flux
boundary in the numerical model appears to slightly accelerate the concentration response
A-l
-------
FIGURE A-l. Partially saturated .zone model comparison
for a thin unsaturatd zone.
ViOOFT
— RQAMMOd*
--«-- MYGRT Model
i
3000
TliiM (years)
Model Comparison for SWMUQ07 with High Rd/Thln Unsat Zone:
MMSOILS/VADOPT, ROAM, and MYRGT Model Results
A-2
-------
FIGURE A-2. Partially saturated zone model comparison
for a thick unsaturated zone.
I
§
i
I
I
400
300
200
100
—•«-•• MMSOU.S/VAOOFT
-"O- MVGflTMoM
tSOOO 20000 2SOOO 30000 35000 40000 4SOOO SOOOO SSOOO
Tlnw (years)
Model Comparison for SWMUQ07 with High Rd/Thick Unsat Zone:
MMSOILS/VADOFT, ROAM, and MYGRT Modet Results
A-3
-------
relative to the analytical model. However, despite the difference in the concentration response
of the numerical and analytical model, the flux response of these models would not be that
different when the dispersive flux from the analytical model is considered. In effect, the
accelerated concentration response in the no-flux boundary models results in a more favorable
comparison of total flux between both approaches.
These slight differences in the concentration profile, however, are only observed for
systems with a relatively high dispersion coefficient. When dispersion coefficients are small
(typical groundwater cases), the dispersive flux is so small that both the analytical and numerical
approaches generate very similar concentration responses.
A-4
-------
APPENDIX B
Groundwater Transport Verification
The groundwater transport formulation used in MMSOILS is based on mass flux from
a box shaped source located at the water table. This appendix describes a comparison between
this formulation and an analytical solution for groundwater transport based on a point source
solution (Hunt, 1978). The transient solution for concentration is a uniform flow field as a
function of time and space presented by Hunt is:
c (x,y,z,t) * M exp.
xu
~2D.
Eq. (B-l)
exp
-RU
2D.
erfc
R-Ut
exp
RU
erfc
Eq. (B-2)
where
M is the mass flux (g/yr),
U is the pore water velocity (m/yr),
9 is the porosity,
erfc is the complimentary error function,
B-l
-------
Dx is the longitudinal dispersion coefficient (m2/yr),
Dy is the lateral dispersion coefficient (mVyr), and
Dz is the vertical dispersion coefficient (mVyr).
This solution accounts for 3-D advective and dispersive transport of a solute in an infinite
media. For the case of a chemical release into a half-space with the water table representing an
upper, no-flux boundary, the concentration based on Equation B-l must be doubled.
The first test case simulated with MMSOILS and the Hunt solution was designed to
verify the steady-state behavior groundwater transport. For this test, an impoundment was
simulated which released a constant, concentration of a thin vadose zone. The concentration of
the water table reached a constant value after 4 years and remained constant for the duration of
the 150 year simulation period. Both programs were set up to examine concentration as a
function of distance 140 years after the beginning of the simulation period. Figure B-l shows
the response of both models along the x-axis (directly down gradient from the center of the
source) for x values ranging from 1,000 to 6,000 meters. The concentrations differ by about
8 percent and 1,000 meters and the difference is approximately 2 percent at 6,000 meters. The
point source solution consistently predicts a higher concentration, but the difference decreases
away from the source.
Figure B-2 shows the calculated response of each model for the same input parameters
except that the y coordinate has been fixed at 200 meters. This distance is slightly wider than
the width of the impoundment specified for the MMSOILS simulation. The peak concentrations
are approximately 40 percent lower than the results shown on Figure B-l. The differential
between the point source solution and the MMSOILS solution are also smaller, particularly at
B-2
-------
FIGURE B-l. Comparison between point source solution and MMSOILs
for Y=0 at steady state.
0.
0.00
Hun I
msoiLS
1000 ..2000 3080 4000 5000 6000
Ol> lartc* (m)
B-3
„
-------
FIGURE B-2. Comparison between point source solution and MMSOILs
for Y=200 m at steady state.
•=• r.
0. 12-..
0.10-
0.
08H
O
C0.06H
e
•
u
gl
U
0.04-I
0.02-
0.00
HunL
MMSOILS
1000 2000 3000 4000 S000 6000
B-4
-------
100 meters where the difference is approximately 5 percent. At 6,000 meters the concentrations
were within 2 percent as they were in the comparison for y = 0.
A comparison of the transient behavior of each model was undertaken by calculating the
chemical concentration at 3 different fixed points at a number of different times. Figure B-3
illustrates the results of these simulations at points on the flow axis 1,054, 4,054, and 6,054
meters downgradient from the source. Close to the source, the difference between the point
source solution and the MMSOILS solution is approximately 8 percent for times greater that 20
years. The discrepancy between the two methods is much less for all time periods at both of
the more distant observation points.
FIGURE B-3. Comparison between transient response of point
source model and MMSOH-s..
0. 12-]
0.10-
\
o
Z0.08-
5
£0.06-
LJ
0.04-
0.02-
0.00
DISTANCE OT 10S4 I-ETERS FROM SOURCE AT X-9
POINT SOME MMtmcM. SOLUTION
MCM. mlCX SOLUTION
DISTANCE OT <
-------
In summary, the box source used to represent the source of groundwater contamination
in MMSOILS produces results that are similar to results from a point source solution (Hunt,
1978). Differences of up to 8 percent were observed for short times at calculation points close
to the origin. 'For longer times and more distal observation points, the methods converge to a
common solution.
B-6
------- |