United States
          Environmental Protection
          Agency
Air and Radiation
(6603J)
93550-61
EPA 540-R-96-004
PB96-963303
January 1996
          Three Multimedia Models
          Used at Hazardous and
          Radioactive Waste Sites
*****
                                Printed on Recycled Paper

-------
THREE MULTIMEDIA MODELS USED AT HAZARDOUS
         AND RADIOACTIVE WASTE SITES
              A Cooperative Effort by

          Office of Radiation and Indoor Air
   Office of Solid Waste and Emergency Response
        U.S. Environmental Protection Agency
             Washington, D.C 20460

         Office of Environmental Restoration
            U.S. Department of Energy
             Washington, D.C. 20585

   Office of Nuclear Material Safety and Safeguards
          Nuclear Regulatory Commission
             Washington, D.C. 20555

-------
                                        FOREWORD

        This report was produced by the Interagency Environmental Pathway Modeling Working Group.
The Working Group includes representatives of the U.S. Environmental Protection Agency - Offices of
Radiation and Indoor Air, and Solid Waste and Emergency Response, the U.S. Department of Energy -
Office of Environmental Restoration, and the U.S. Nuclear Regulatory Commission - Office of Nuclear
Material Safety and Safeguards.  The purpose  of the Working Group is to promote the appropriate and
consistent use of mathematical models in the remediation and restoration process at sites containing - or
contaminated with - radioactive and/or mixed  waste materials.  This report provides an approach for
evaluating and critically reviewing the capabilities of three specific multimedia models: MEPAS Version
3.0, MMSOILS Version 2.2, and PRESTO-EPA-CPG Version 2.0. These  models are  being used by the
sponsoring Offices to support  cleanup  decision-making at various waste sites, and are of  technical
interest to them. The approach for model review advocated in this report is intended  to  assist technical
staff responsible for identifying and implementing multimedia models in support of cleanup decisions at
radioactive and hazardous waste  sites.  It is  hoped that information in this  report will enhance the
understanding of these three models within the context of specific media components, human exposure
and dose, and how they report uncertainty.

       This document is one of several being developed by the Working Group to bring  a uniform
approach to solving environmental modeling problems common to all Federal agencies.  The following
are other reports prepared by this Interagency Working Group:

    •  Computer Models  Used to Support Cleanup Decision-Making at Hazardous and Radioactive
       Waste Sites, EPA 402-R-93-005, March 1993.

    •  Environmental Characteristics of EPA, NRC, and DOE Sites Contaminated with Radioactive
       Substances, EPA 402-R-93-011, March 1993.

    •  Environmental Pathway Models -  Ground Water Modeling in Support of Remedial Decision-
       Making at Sites Contaminated with Radioactive Material, EPA 402-R-93-009, March 1993.

    •  A  Technical Guide to Ground Water Model Selection at Sites Contaminated with Radioactive
       Substances, EPA402-R-94-012, June 1994.

    •  Evaluating Technical Capabilities of Ground Water Models Used to Support the Cleanup of Low-
       Level Radioactive Waste Sites:  A Critique of Three Representative Models, Draft, March 1994

    •  Documenting Ground Water Modeling at Sites Contaminated with Radioactive  Substances, EPA
       540-R-96-003. January 1996.

       The project Officers of the Interagency  Working Group (Beverly Itla - EPA, Paul Beam - DOE.
Sam  Nalluswami  - NRC)  acknowledge the cooperation and insight  of  many  staff  in  preparing this
document from organizations including EPA/Environmental Research Laboratory, Athens Georgia. EPA
Office of Radiation and Indoor Air Criteria and Standards Division, Washington, D.C ; and Batelle/Pacific
Northwest Laboratories, Richland Washington.  We would also like to thank all those from EPA Regions
II, III. IV, V, VI, and VIII; EPA Office of Emergency and Remedial Response; EPA Office of Radiation
Programs/Las  Vegas; EPA National  Air  and  Radiation Environmental  Laboratory; DOE  Office  of
Environmental Restoration, and NRC Office of  Material Safety  and Safeguards, who graciously agreed
to provide review and comment.  We also thank their managers who permitted them the time to provide
us with valuable input.

       This report was prepared under IAG DW89934985, Paul Moskowitz, Project Officer, Brookhaven
National Laboratory.

                                             iii

-------
           KEY ABBREVIATIONS USED IN THIS REPORT
Ci    Curie
cm    centimeter
d     day
g     gram
kg    kilogram
km    kilometer
L     liter
m    meter
mg    milligram
min   minute
pCi    pico-Curie
s     sec
SI    System Internationale
Sv    Sieved
yr    year
                              IV

-------
                            EXECUTIVE SUMMARY

 Purpose

      Multimedia models are used commonly in the initial phases of the remediation
 process where technical interest is focused on determining the relative importance of
 various exposure pathways.  This report provides an approach  for evaluating and
 critically reviewing the capabilities  of multimedia models. This study focused on three
 specific models:  MEPAS Version 3.0, MMSOILS Version 2.2, and PRESTO-EPA-CPG
 Version 2.0.  These models evaluate the transport and  fate of  contaminants  from
 source to receptor through more  than a single pathway.  They have been  used to
 support cleanup decision-making at various sites  and are  of technical interest to the
 sponsoring organizations.  The  approach to model  review advocated in this  study is
 directed to technical staff responsible for identifying,  selecting and applying multimedia
 models for use at sites containing radioactive and hazardous materials  The presence
 of radioactive and mixed wastes  at a site poses special  problems.  Hence, in this
 report, restrictions associated with the selection and application of multimedia models
 for sites contaminated with radioactive and mixed wastes are highlighted   It is hoped
 that information in this report will  enhance the understanding of these three models
 within the context of specific media components, human exposure and dose, and how
 they report uncertainty.
Report Structure

      This report  begins with  a brief  introduction to the concept  of  multimedia
modeling, followed by  an overview  of the three models.  The remaining chapters
present more technical discussions of the issues associated with each compartment
and their direct application to the specific models.  In these analyses, the following
components are discussed:

            •  Source Term
            •  Air Transport
            •  Ground Water Transport
            •  Overland Flow, Runoff, and Surface Water Transport
            •  Food Chain Modeling
            •  Exposure Assessment
            •  Dosimetry/Risk Assessment
            •  Uncertainty
            •  Default Parameters

      The report concludes with a description of evolving updates to the model: these
descriptions were provided by the model developers.

-------
Model Selection

      There are many multimedia models in use for simulating the transport, fate and
effects of contaminants  present at waste sites.   All  of these  models could  not  be
reviewed as part of this effort.  Thus, the sponsoring agencies requested that MEPAS,
MMSOILS, and PRESTO-EPA-CPG be reviewed because of their technical interest in
these specific models.  This should not be interpreted to mean that any of the agencies
endorse any of these models or the specific findings presented in this report

      MEPAS, whose development was sponsored by  the DOE,  has a broad coverage
of pathways  and  scenarios  for  radioactive  and  chemical   hazardous  materials.
MMSOILS and PRESTO-EPA-CPG were developed by the EPA   MMSOILS is meant to
be used for the screening and comparison of hazardous sites contaminated with toxic
chemicals that are released from  underground storage tanks,  impoundments,  waste
piles, landfills, and injection wells.  PRESTO-EPA-CPG was designed  specifically to
provide annual committed dose equivalent estimates from the release of radionuclides
from low-level radioactive waste sites.
Model Components

Source Term

      MEPAS is the most versatile of the three models,  with the greatest ability to
handle a variety of different types of source terms.  Although the PRESTO-EPA-CPG
model handles source terms only from near-surface trenches, its family of models can
handle a variety of source terms, including contaminated soil, waste piles, deep-well
injection, and underground tanks.  MMSOILS is the only one of the  three models
reviewed here that performs a mass balance  for the air and ground water pathways
separately,  relative to the initial  source term.  The generation of leachate  into the
ground water is estimated by different, but similar, means in all three models.

Atmospheric Pathway

      A sector-averaged Gaussian plume algorithm is  used by all three  models to
simulate the atmospheric transport of contaminants. The PRESTO-EPA-CPG model
includes the effects of wet deposition and radioactive decay.  The current version of the
model does not include a volatile source term generation from a storage lagoon or  lake.
While the overall modeling capabilities of MMSOILS are similar to those of MEPAS,
MMSOILS is  less sophisticated  in modeling the atmospheric pathway   It does not
describe  complex terrain, calms, depletion  of the plume  by wet deposition,  and
contaminant decay. MMSOILS cannot model releases from vents or stacks.  However,
this option is  required only for modeling emissions from waste cleanup facilities, not
from hazardous waste sites  The paniculate  emission models which are included in
both MEPAS and MMSOILS are particles from wind erosion, vehicle traffic, and soil-

                                     vi

-------
 spreading operations.  MMSOILS has a model of loading and unloading operations, but
 MEPAS does  not.  Finally, MEPAS has some capabilities which none of the other
 models have, namely, air sources  (i.e., defining a source by ambient concentrations),
 calm meteorological conditions, and complex terrain.

 Ground Water

       Overall, MMSOILS has the most complex ground water pathway, since it alone
 uses  a finite-element  model for  the unsaturated  zone  that  incorporates  layered
 heterogeneity.   Unfortunately, there  are  two problems with  simulating radioactive
 contaminants:  (i) MMSOILS assumes that the contaminant does not decay while it is
 sorbed onto soil, and (ii) MMSOILS only models nonradioactive substances and does
 not explicitly simulate the ingrowth of progeny.  Conceivably, the first problem  could be
 overcome by entering a radioactive decay rate that is multiplied by the K^. The second
 problem  cannot  be avoided  easily,  especially for short-lived contaminants.   The
 PRESTO-EPA model employs a simple, one-dimensional model.   MEPAS simulates
 ground water transport using  a  3-D algorithm, but assumes that radioactive  progeny
 have the same Kd as the  parent. This can introduce error  into the source and down-
 gradient concentration estimations.

 Surface Water Transport

      All  three models take  a  rather similar, simplistic approach to  modeling the
 surface water pathway.  Both MEPAS and MMSOILS link  ground  water and surface
 water media by converting the ground  water flux feeding into the surface water into an
 input flux to the surface water medium.  The PRESTO-EPA-CPG model  uses system
 equations representing the surface  water, subsurface water, and atmospheric  diffusion
 systems to calculate the rate of deep ground infiltration flow,  overland flow, and the rate
 of evapotranspiration.   If conditions allow, the overland flow may combine  with the
 overflowing leachate.  Then,  this combined flow (with or without  leachate overflow)
 interacts with the contaminated soil. The combined overland flow is programmed also
 to simulate the leaching of the contaminant out of the soil and transport into the nearby
 surface water to  be pumped for  human drinking, irrigation,  and  cattle feed.  All three
 models employ the Universal Soil Loss Equation to estimate soil erosion across a site.

 Food Chain Modeling

      MEPAS includes  food chains as  an integral  part of  its exposure-dose
 component  Food chains are considered  separately in MMSOILS and  in PRESTO-
 EPA-CPG as agricultural data supporting exposure estimates.  Although MEPAS can
 be used to model acute toxic atmospheric releases, all three models were designed
 primarily to handle long-term,  chronic exposures.  Each of the  three models employ
 comparable, standard methods for estimating exposure to environmental contaminants
 through the food  chain and other pathways. Therefore, the  same limitations that exist
for all exposure and risk assessments exist for these models. For example, food intake
 is subject to behavioral variations. Both the quantity and type of foods eaten vary from
                                     vii

-------
(Table S.1 cont'd. Summary of model features. I
	 	

Bio-accumulation
Animals
Terrestrial plants
Foliar deposition
Aquatic organisms
Spatial definition
Site Data Required
Contaminant Selection
Hazardous chemical waste
Radioactive waste
Intakes from Ingestion of
Drinking Water
Shower Water
Swimming Water
Leafy Vegetables
Other Produce
Meat
Milk
Finfish
Shellfish
Special Food
Shoreline Sediment
Soil
Intakes from Inhalation
While Showering
Of Air
Of Re-suspended Soil
Intakes from Contact
While Showering
While Swimming
With Shoreline Sediment
With Soil
With Volatiles in Air
External Exposures.
While Swimming
While Boating
From Air
With Soil
With Shoreline Sediment
From Direct Radiation.
MEPAS

yes
yes
yes
yes
2-D
Extensive

yes
yes

yes
yes
yes,,
yes;1
yes'3
yes
yes
yes
yes
yes
yes
yes

yes
yes
yes

yes
yes
yes
yes4
yes

yes
yes
yes
yes
yes
yes
MMSOILS

yes
yes
yes
yes
2-D
Moderate

yes
no

yes
no
yes
yes
no
yes
yes
yes
no
no
no
yes

no
yes
yes

no
no
no
yes
no

no
no
no
no
no
no
PRESTO-EPA-CPG

yes
yes
yes
no
2-D
Moderate

no
yes

yes
no
no
yes
yes
yes
yes
yes
yes
eggs
yes
yes

no
yes
yes6

no
no
no
no
no

no
no
yes
yes
no
no
1 This component not available in 1993 version.
2 From air deposition on crops.
3 From irrigation of crops.
4 Estimations based on either measured concentrations or on calculated accumulations
in soil after atmospheric deposition.
5 In the version modified for cleanup scenarios.
6 On-site scenario only
7 MMSOILS only considers ground surface roughness in wind erosion of particulates.
XII

-------
                          CONTENTS




FOREWORD	 iii




EXECUTIVE SUMMARY	 v




 1.   INTRODUCTION	 1




 2.   MULTIMEDIA PATHWAY ANALYSIS	 3




 3    MODEL OVERVIEW	 9




 4.   SOURCE TERM	17




 5.   AIRTRANSPORT	25




 6    GROUND WATER TRANSPORT	33




 7    RUNOFF, EROSION, AND SURFACE WATER TRANSPORT	43




 8.   FOOD CHAIN MODELING	49




 9    EXPOSURE ASSESSMENT	53




10   DOSIMETRY/RISK ASSESSMENT	57




11    UNCERTAINTY ANALYSIS	63




12   PARAMETER ESTIMATION AND DEFAULT PARAMETERS	65




13   DISCUSSION/CONCLUSIONS	71




14.   REFERENCES	77
                             X1H

-------
                                  TABLES

 S.1  Summary of model features	xi
 2 1  Typical exposure pathways by medium for residential
       and commercial land uses	  7
 2.2  Summary of major intermedia transport routes	  7
 3.1  Some other representative multimedia models	12
 3.2  Examples of model usage of other models	13
 3.3  Evaluations of some other representative multimedia models	15
 4.1  Model source term capabilities	17
 42  Source term of volatilization scenarios	20
 5.1  Atmospheric pathway. Comparison of capabilities	29
 6.1  Ground water pathway: Comparison of capabilities	36
 6.2  Data requirements for MMSOILS	40
 6.3  Data requirements for MEPAS	41
 64  Data requirements for PRESTO-EPA-CPG	42
 7.1  Surface water pathway: Comparison of capabilities	45
 7.2  Data input requirements of the surface water pathway	48
 8.1  Summary of food chain features in models	50
 82  Details of food chain levels in the models	51
 9.1  Summary of exposure features in models	54
12.1  Input parameters for arsenic	68
12.2  Input parameters for benzene	69
13.1  Summary of model features	75
                                  FIGURES

 2.1   General air pathways to humans	 6
 22   General liquid pathways to humans	 6
 23   Relative relationships between input-data quality, output uncertainty,
      and types of problems addressed by each level of assessment	 8
 5 1   Simplified pathways between radioactive materials released to
      atmosphere and humans	25
 5.2   Diagram of pathway interactions	26
 6.1   Conceptual diagram of the ground water pathway	34
 7.1   Pathways for surface water exposure	43
 8.1   Basic food chains as depicted in risk assessments	49
                                     XIV

-------
                              1.  INTRODUCTION

      Significant  efforts are  being made  to  remediate waste  sites  containing
radioactive materials and hazardous wastes.  The remediation goals may be based on
reducing potential chemical or radiation doses  to  the  public from  all significant
scenarios, media,  and exposure pathways. Remediation strategies are typically based
on an assessment founded on the use of computer models because of the complexities
of sites and of the characteristics of the contaminants.  These computer models use
sets of  mathematical  equations incorporating many factors that cause or affect the
movement  of contaminants and radionuclides through various media including their
intake by humans. Computer models are used now routinely by the U.S  Environmental
Protection Agency (EPA) and others [e.g.,  Nuclear Regulatory Commission (NRC)] for
setting standards and regulations, and by  the U.S. Department of Energy (DOE) and
others for determining the priorities and benefits of alternative cleanup options.

      The  EPA Office of Radiation and Indoor Air and  Office of  Solid Waste and
Emergency  Response,  and the  DOE  Office of  Environmental  Restoration are
attempting  to develop a uniform  approach  to solving  their  common problems in
environmental modeling for site remediation and restoration.  As part of this effort, this
report reviews in  detail  three multimedia models  used by these agencies - MEPAS
version 3.0, MMSOILS version 2.2, and PRESTO-EPA-CPG version 20

      The results of the analyses reported here are not intended as an endorsement
of any of the models reviewed.  Rather, the intention is to provide the reader with both
an approach for evaluating mathematical  models  as well as an evaluation of each
model's capabilities and limitations.

      This  report  begins  with  a  brief introduction  to  the concept of multimedia
modeling, followed by an  overview of the three  models.   The remaining chapters
present  more technical reviews of the sub-components of the models.  Each chapter
discusses first  briefly the specific media  component, then how each model handles
radionuclide transport within that compartment, and finally describes evolving updates
to the model.

      In these analyses, the following pathway and risk assessment components are
discussed:
                  •  Source Term
                  •  Air Transport
                  •  Ground Water Transport
                  •  Surface Water Transport
                  •  Food Chain Modeling
                  •  Exposure Assessment
                  •  Dosimetry/Risk Assessment
                  •  Uncertainty
                  •  Default Parameters

                                      i

-------
      The reviews are based on the following sources of information.

   •  Documentation from model developers, including Users' and Methods Manuals,
      Guides, Appendices, Revisions, and Addenda which explain or clarify the use or
      basis for each model.

   •  Reviews of models contained in the peer-reviewed literature and the results of
      formal review programs.

   •  Personal  interviews with the models' developers to answer specific  questions
      about features that are  not discussed in  the  documentation, and to learn of
      proposed or expected developments for new versions of the models.

   •  Computer databases and expert systems,  like the Integrated Model Evaluation
      System (USEPA, 1993b) and Exposure Models Library (USEPA, 1994a), that
      were developed to aid in  selecting models appropriate for different applications.

      The  documentation available does not always describe upgrades or planned
modifications  because  model  building  is an  ongoing process   To  this end,  the
developers  of each model reviewed in  this report were contacted and asked to provide
a  letter-report  to update information  contained  in  the  model documentation.
Improvements were summarized in the individual chapters.

-------
                     2. MULTIMEDIA PATHWAY ANALYSIS

2.1   Media-Based Analyses

      Multimedia modeling begins with  a  source  of contamination and ends with a
calculation of risk for the final assessment.  Ideally,  such a model would evaluate every
possible pathway by which a contaminant is carried through every potential media from
source to humans.  Given a known or assumed concentration of a contaminant at a
source and, from that, computing a risk is an extremely complex procedure.  As stated
by Seigneur et al. (1992):

      "A  comprehensive treatment of all these  processes  would require a
      multimedia transport model with fine spatial and temporal resolutions in all
      media, two-way inter-media transport, treatment of population dynamics
      with resolution  of population cohorts according to activity patterns,  age
      groups and physiological status: description  of population  exposure  in a
      variety of micro-environments; and the development of accurate dose-
      response relationships "

      Multimedia models may neither consider every potential pathway with the same
thoroughness, nor every pathway between media.   Even when some inter-media
pathways  are included, a model may not account  accurately  for the fate of material
transported from one media to another  However, as Ryan (1993) pointed out, people
may be exposed to contaminants indirectly through inter-media transport,  as well  as
directly.

      Several  documents   incorporating  pathway  analyses  including  estimating
radiation  dose   in the environment  were prepared  for  a variety  of well-defined
conditions.   The  Environmental  Impact Statement  (EIS) in support  of licensing
requirements for shallow land burial of low-level radioactive waste (10 CFR 61) is one
example  Regulatory Guide 1.109 issued  by the Nuclear Regulatory Commission is
designed to be used for any release to the environment from effluent streams for any
nuclear  power plant.  This document recommends how a generalized pathway analysis
can be structured for a given effluent medium (air, water) in a particular environmental
setting.

      Figures 2.1 and 2.2  adapted from Dugan  et al. (1990)  illustrate  simplified
pathways for the release of radioactive and hazardous materials to the atmosphere and
water, and their routes of exposure to humans.  In  these suggested transport models,
air is contaminated by re-suspension and volatilization of radionuclides and  chemicals.
The roots  of the  plants take up the material  They  are contaminated externally by the
deposition of suspended particles. Herbivores take in radionuclides from the ingestion
of soil, by  grazing on contaminated vegetation, by drinking water, and by inhaling dust.
The maximally exposed person is someone who lives in. and obtains their food from,
the contaminated area, inhales  contaminated  air, drinks contaminated  water, and
ingests contaminated vegetation, meat, and dairy products from animals raised there

-------
      People can be exposed to radioactive and hazardous contaminants  present in
the soil  at National Priority List (NPL) and other sites through four basic environmental
media  (i) the atmosphere (ii) surface water, (iii) soil and ground water; and (iv) biota
The exposure pathways shown in Table 2.1 are considered by EPA to be typical of
those to be included in any evaluation of human health conducted at a SUPERFUND
site (Office of Emergency and Remedial Response, 1991).

      Among these  different media, a variety of  inter-media  transport mechanisms
exist  (Table 2.2).  Choosing  among the ones that should be included in any model
requires   balancing   several    competing   concepts:   modeling     objectives,
simplicity/complexity  of the model; scenario/site complexity;  data availability; and. the
value of the information provided

22   Level of Analysis

      Practically, three levels of  multimedia analysis can be identified (Whelan, et al  ,
1994).

      •  Screening-level (ranking)
      •  Analytical (prioritization and preliminary assessments)
      •  Numerical (detailed)

      Early  in the process,  screening  models are  used  to identify environmental
concerns  These models are based often on  a structured-value approach  They are
designed to be used with regional/representative information. Models such as the EPA
Hazard  Ranking System (MRS) (USEPA,  1988b;  USEPA, 1990b) divide the site and
release  characteristics into pre-determined categories that are assigned a point value
based on answers to questions.   The score from such systems is useful to determine if
a situation is a problem, but not to provide a risk-based relative ranking of problems

      Detailed analyses require a highly specialized assessment of potential impacts
Methodologies  such  as the Chemical  Migration  Risk  Assessment (CMRA)  are
composite-coupled  approaches  that use  numerically  based  models that are  not
physically linked and represent single-medium models, implemented independently in
series  This approach  is reserved  usually for the most  complex models, is data-
intensive, and relies on the expertise of the analyst. These detailed models are used to
determine the levels  of risk associated with  relatively well-defined, complex  problems,
and tend to focus on special sets  of problems and special types of situations. Although
such tools are appropriate for their intended application, extension beyond site-specific
applications is often either difficult or cost-prohibitive.

      Analytical/Semi-Analytical/Empirical-Based  multimedia models  (designated as
analytical  models) can be used  for prioritization or preliminary assessments   Most
often, they are employed between initial screening and highly  specialized  numerical
models  These models are the most versatile as they do not have the data constraints
                                       4

-------
of the numerical  models,  but are physics-based,  unlike the structured-value models.
The analytical models may  contain some numerical  computations  (hence the  semi-
analytical designation).   The analytical models  provide  environmental  evaluations
through a wide range of applications. These models are fully coupled approaches that
use analytically, numerically, and empirically based algorithms that are combined  into a
single code to describe each environmental medium.

      Figure 2.3  illustrates the value of analytical models in the waste-site evaluation
process.  They can be used in a detailed (i.e., numerical) or an initial-screening (i.e.,
ranking/prioritization)  assessment, where data and circumstances warrant   Figure 2.3
illustrates also the relative relationships between input-data quality, output uncertainty,
and types of problems at each level of assessment.   The computational requirements
tend to be less at the earlier stages of an assessment when there are fewer available
data and, correspondingly, the uncertainty with the output results tends to be greater.
As   the  assessment   progresses,    improved   site-characterization   data   and
conceptualization of the problem increase, thereby reducing the overall uncertainty  in
risk estimates.

      The  analytical  multimedia  models  integrate  standard approaches  into  a
consistent, yet powerful, tool.   The multimedia models incorporate medium-specific,
transport-pathway,  and exposure-route  codes that  are based on standard,  well-
accepted algorithms.   For example,  these multimedia models  employ  analytical
solutions to  the advective-dispersive equations that describe transport in the ground
water environment.  When coupled, the models allow the analyst to immediately assess
the entire process of contaminant release, transport, exposure, and risk at one sitting.
Some models give the user the freedom to by-pass the transport components and use
only the exposure/risk components. The value of a coupled model is exemplified by an
order-of-magnitude reduction in assessment time, as compared to an uncoupled model

      Multimedia models assess  concurrently  multiple waste sites with  multiple
constituents  to include baseline (at time = 0 yrs),  no action (at time > 0  yrs),  during-
remediation,  and  residual (post-remediation) assessments,  including changing  land-
use patterns (e.g.,  agricultural,  residential,   recreational, and   industrial).   The
multimedia models can describe the environmental concentrations within each medium
at  locations  surrounding  the  waste.   Spatially  distributed,  three-dimensional,
concentration isopleths can be constructed detailing the level of contamination within
each  environment.  Three-dimensional risk isopleths can be developed by coupling
land-use patterns  with the environmental concentrations.

-------
           Human*
Figure 2.1  General air pathways to humans (after Dugan et al., 1990)
        Radioactive
        Materials
        ijt  :j  •-.••..  v
        ::  v.. ^•:-..-•. V.....
                                   Aquatic
                                 Plants & Animals
         Ground Water
                                                                 Humans
Figure 2.2  General liquid pathways to humans (after Dugan et al.. 1990)

                                         6

-------
I Table 2.1. Typical ex

MEDIUM
Ground Water



Surface Water




Soil





posure pathways by medium for residential and commercial land uses.
Exposure Pathways, Assuming
RESIDENTIAL/RECREATIONAL LAND
USE
Ingestion from drinking
Inhalation of volatiles
Dermal absorption from bathing
Immersion-external
Ingestion from drinking
Inhalation of volatiles
Ingestion during swimming
Ingestion of contaminated fish
Immersion - external
Ingestion
Inhalation of volatiles & particles
Direct external exposure
Ingestion via plant uptake
Ingestion of meat, milk and other animal
products
Dermal absorption from gardening
COMMERCIAL/INDUSTRIAL LAND
USE
Ingestion from drinking
Inhalation of volatiles
Dermal absorption
Irrigation
Ingestion from drinking
Inhalation of volatiles
Irrigation


Ingestion
Inhalation of volatiles & particles
Direct external exposure



Table 2.2. Summary of major intermedia transport routes.
    Transport from the Atmosphere to Land and Water
       Dry deposition of particulate and reactive gaseous pollutants
       Precipitation scavenging of gases and aerosols
       Adsorption onto particulate matter and subsequent dry and wet deposition
    Transport from Water to Atmosphere, Sediments, and Organisms
       Volatilization
       Aerosol formation at the air/water interface
       Sorption by sediment and suspended solids
       Sedimentation and re-suspension of solids
       Uptake and release by biota
    Transport from Soil to Water, Sediment, Atmosphere, or Biota
       Dissolution in rain water
       Adsorption on soil particles and transport by runoff or wind erosion
       Volatilization from soil and vegetation
       Leaching into ground water
       Re-suspension of contaminated soil particles by wind
       Uptake by microorganisms, plants and animals.

-------
                       •LEVEL OF ANALYSIS.
                                                    numerical
            representative
                                regional
                                                   site-specific
            least-
•INPUT DATA QUALITY
Figure 2.3. Relative relationships between input-data quality, output uncertainty, and
types of problems addressed by each level of assessment (after Whelan et at., 1994.)

-------
                             3.  MODEL OVERVIEW

 3.1    Introduction

       A principal objective of modeling by the EPA, DOE. and NRC is to characterize
 the risks to human health presented by soil contamination present  at sites containing
 radioactive materials, and the benefits to health derived from their cleanup. Estimates
 of the time-varying and time averaged radionuclide concentrations in air, surface water,
 ground water,  soil and food  items, intake of these materials  by humans,  and
 subsequent health risk are needed to fulfill this objective.

       In the  context of  the exposure pathways  identified in Chapter 2, the following
 end-points and processes are of particular importance.

   •   Individual and population doses and risks as a function of time

   •   External radiation exposures to radionuclides on the ground and in the air

   •   Radioactive decay and daughter ingrowth

   •   Indoor radon exposures

      Through  analysis of survey data (Moskowitz et al. 1993, Mills and Vogt, 1983,
 Case et al  1989, USEPA 1989,  USEPA 1990a),  application of the EPA Integrated
 Model  Evaluation System and the Environmental Models Library (USEPA 1993), review
 of scientific and vendor literature, and discussions with project staff, a list of models
 was  identified that  have been or could  be used in a multimedia radiological  risk
 assessment project. In developing this list, there was no attempt to determine initially
 whether these (Table 3.1) models could be appropriately applied to sites contaminated
 with radioactive substances.

      Tables 3.2 and 3.3 give a brief description of representative ways in which each
 of the listed models has been used. The  sponsoring Agencies requested that MEPAS,
 MMSOILS, and  PRESTO-EPA-CPG be  reviewed  in  greater detail.   Each of these
 models was  developed primarily for screening-type use or  for comparisons between
 sites conducted  for the  purpose of ranking  relative risk.  These models were never
 intended for non-screening uses such as remedial design,  etc.   In addition to being
 designed primarily for screening, the three models are generic models; i.e., they are
 meant to be used at a wide variety of sites.

 32   MEPAS

      MEPAS (the  Multimedia  Environmental Pollutant Assessment  System)  is an
analytical model designed by Pacific Northwest Laboratories, which was developed for,
but is not limited to, use at CERCLA, Clean Air Act and Clean Water Act sites (Droppo
et al., 1989,  Volume 1).  MEPAS  is an enhancement of the Remedial Action Priority
                                       9

-------
System (RAPS, Whelan et al., 1987), using the same mathematical formulations and
algorithms plus new additions to the methodology (Droppo, Whelan,  et al., 1989).
MEPAS was  developed for DOE applications, but its development was specifically
designed to address general problems at any hazardous waste site

      MEPAS  develops   an  integrated,  site-specific,  multimedia  environmental
assessment.  It can simulate the transport and distribution of contaminants (chemical
and radiological) over time and space within air, water, soil, and food chain pathways
It estimates  long-term health effects  at receptor locations,  from exposures  over 70
years,  as well as normalized maximum hourly concentrations for  determining acute
effects.  Contaminated media and exposure pathways include air, ground water wells,
water intakes from surface waters, recreation parks along surface  water, on-site soil
ingestion, and direct radiation  Receptors are evaluated as members of population and
agriculture centers within  an 80 km radius of the release unit Currently, the model's
database contains 576 referenced organic and inorganic chemicals  and radionuclides.
The database is updated and expanded continually  A "user-friendly" shell is provided
with  MEPAS  which  aids the user  in  defining the problem,  entering input data, and
executing the model run

3.3   MMSOILS

      MMSOILS (the Multimedia Contaminant Fate, Transport, and Exposure Model)
was developed by the EPA Office of Research and Development as a "screening tool"
for the "relative comparison" of hazardous waste (especially RCRA) sites (U.S. EPA,
1989a).  It was designed  specifically to simulate the release of toxic chemicals from
underground  storage tanks, surface impoundments, waste piles,  and landfills   It can
model the fate and transport of chemicals only, and calculates human exposure and
health risk, as well as concentration in all important media.  MMSOILS has a database
for 240 chemicals  Like MEPAS, MMSOILS is provided with a "user-friendly" shell that
aids the user  in defining the problem, executing the model, and evaluating the output

34   PRESTO-EPA-CPG

      PRESTO-EPA-CPG (CPG for Critical Population Groups) belongs to a family of
EPA exposure-assessment models which includes PRESTO-EPA-POP, PRESTO-EPA-
DEEP,  PRESTO-EPA-BRC, and PATHRAE-EPA (U.S. EPA, 1987)  The PRESTO-EPA
family of analytical  models was designed specifically  for radionuclide transport  via
natural processes, a consideration that dominates the overall structure and operation of
their codes.   PRESTO-EPA-POP (PRESTO for regional Populations) was the first in
the series,  and is  the basis for each of the other models.  PRESTO-EPA-CPG is
designed to calculate the  annual committed dose equivalent to members of a critical
population group resulting from the disposal of low-level radioactive wastes (LLW) by a
near-surface disposal method. The model identifies also the maximum effective dose
equivalent and year of occurrence.  POP models incidental, fatal, and  genetic health
effects to local and regional populations stemming from LLW deposited in shallow land
burial sites also   This model can handle several different wasteforms  within those
                                      10

-------
shallow-land disposal facilities,  absorbing materials,  activated metals, trash, solidified
waste, and  incinerated waste.  It incorporates a database of 40  radionuclides.  The
PRESTO-EPA models are being modified currently to assess the health effects from a
cleanup site by adding contamination scenarios not  already  included, namely:  radon
emission, soil and fish ingestion, and farming on site without protective cover
                                       11

-------
Table 3.1. some otner representative multimedia models.
MODEL NAME
ARCL
DECHEM
DITTY
DOSES
DOSTOMAN
GENII
CEOTOX
GWSCREEN
HHEM
HRS-1
IMPACTS (PART 61)
MEPAS
MILDOS
MILDOS-AREA
MMSOILS
MULTMED
NUREG5512
NUTRAN
ONSITE/MAXI1
PATH1
PATHRAE (-EPA)
PC GEMS
PRESTO-EPA
PRESTO-EPA-BRC
PRESTO-EPA-CPG
PRESTO-E PA-POP
PRESTO-EPA-DEEP
PRESTO-II
RESRAD
RISKPRO
SARAH2
UDAD
UTM-TOX
PRIMARY REFERENCE
Napier and Piepel, 1988
Radiological Assessments
Corporation
Napier etai., 1986
Oak Ridge National
Laboratory
Root, 1 981; King etal., 1985
Napier etal., 1988
McKoneetal., 1983
Root 1991
USEPA, 1991
Stenner et al., 1986
Oztunali and Roles, 1986;
Oztunali et al., 1986
Doctor etal., I990a,b,c;
Droppo et al., 1989; Whelan
etal., 1987
Strenge and Bander, 1981
Yuan etal, 1989
USEPA, 1992
Salhorta et al., 1990
Kennedy and Strenge, 1992
Ross etal., 1980
Napier et al., 1984; Kennedy
etal, 1986; Kennedy etal,
1987
Helton and Kaestner, 1981;
Campbell etal. , 1981
Rogers and Hung, 19873
General Sciences
Corporation, 1982
USEPA 1983
Rogers and Hung, I987b
Rogers and Hung, I987d
Fields etal, 19873, 1987b
Rogers and Hung, 1987C
Fields etal, 1986
Gilbert, 1988
General Sciences Corp, 1992
vandergrift and Ambrose,
1988
Momeni etal, 1979
Browmanetal, 1982
SPONSORING
AGENCY
DOE
DOE
DOE
ORNL
DOE
DOE/NRC

DOE
EPA
DOE
NRC
DOE
NRC
DOE
EPA
EPA
NRC
Atomic Energy
of Canada, Ltd.
NRC
NRC
EPA
EPA
EPA
EPA
EPA
EPA
EPA

DOE
EPA
EPA
NRC

HARDWARE
PLATFORM

MSDOS-PC



MSDOS-PC

MSDOS-PC
Not yet
implemented
MSDOS-PC
MSDOS-PC
MSDOS-PC

MSDOS-PC
MSDOS-PC
MSDOS-PC
Not Yet
Available
IBM Main
Frame
MSDOS-PC

MSDOS-PC
MSDOS-PC


MSDOS-PC
MSDOS-PC
MSDOS-PC

MSDOS-PC
MSDOS-PC
MSDOS-PC


12

-------
Table 3.2 - Examples of usage of other models.
MODEL NAME
ARCL
DECHEM
DITTY
DOSES
GENII
GEOTOX
GWSCREEN
HHEM
HRS-1
IMPACTS
(PART 61)
MEPAS
MILDOS
MILDOS-AREA
MMSOILS
MULTIMED
NUREG 0707
NUREG 5512
NUTRAN
ONSITE/MAXI1
PATH
PATH1
PATHRAE (-EPA)
PC GEMS
DESCRIPTION OF REPRESENTATIVE USAGE
Evaluate decommissioning alternatives by using a site-specific
radiation scenario/exposure pathway analysis to determine the
acceptable levels of residual radioactive contaminants that remain.
Determine acceptable levels of chemicals in soil after clean-up of
Uranium Mill Tailings Remedial Action Project Sites.
Determine the collective dose from long term nuclear waste disposal
sites resulting from ground water pathways.
Estimates of long-term dose to man from buried waste.
Internal dosimetry from chronic and acute radiation exposure.
Evaluated health risks presented from the presence of TNT, RDX and
benzene present in military explosives residuals.
Developed for assessment of ground water pathway from leaching of
radioactive and nonradioactive substances from surface or buried
sources.
Assist RPMs to develop preliminary remediation goals at CERCLA
sites.
Hazard ranking for SUPERFUND site assessment.
Estimates radiological impacts for a given combination of waste
streams and processing options, disposal technology alternatives, and
disposal site environmental settings. Used during the development of
10CFR Part 61 rule.
A risk computation system developed for hazard ranking applications.
Computes environmental radiation doses from uranium recovery
operations
The MILDOS-AREA code provides improved capability for handling
large area sources and updates the dosimetry calculations
Multimedia landfill model.
EPA Toxicity Characteristic Final Rule.
Estimates site-specific limits for allowable residual contamination
Provides generic and site-specific guidance of radiation doses for
exposures to residual radioactive contamination after the
decommissioning of facilities licensed by the NRC.
Calculates the consequences of ground water releases of radioactivity
from a waste repository.
NRC review of license applications for onsite disposal of radioactive
wastes.
Used to implement residual radioactive material guidelines during
decommissioning.
Models the physical and biological processes that result in the
transport of radionuclides through the Earth's surface environment
and eventual human exposure to these radionuclides.
Maximum annual effective dose equivalent to a critical population
group and to offsite populations at risk from the land disposal of
radioactive wastes.
Used to evaluate the spread of toxic chemicals released to air, soil,
surface water and ground water.
13

-------
Table 3.2 cont'd - Examples of model usage.
PRESTO-EPA
PRESTO-EPA-
BRC
PRESTO-EPA-
CPG
PRESTO-EPA-
POP
PRESTO-EPA-DEEP
PRESTO-II
RESRAD
RISKPRO
SARAH2
UDAD
UTM-TOX
Simulates transport of low-level radioactive waste material from a
shallow trench site and assesses human risks associated with such
transport. This model was modified and added to create the PRESTO
family of models.
This is a modified version of PRESTO-EPA-POP. Additions to this
model include estimation of radionuclide transport and exposure to
workers and visitors, population exposures from incinerator releases,
worker and visitor gamma exposures, and onsite farming.
Maximum whole body dose to critical population groups from land
disposal of low-level radioactive waste by shallow and deep methods.
Cumulative population health effects to local and regional basin
populations from low-level waste disposal by shallow land methods.
Cum. population health effects to local and regional populations from
land disposal of low-level radioactive wastes by deep methods.
Evaluation of possible health effects from shallow-land and waste
disposal trenches.
An analytical methodology recommended by the Department of
Energy in its guidelines for allowable concentrations of residual
radioactive material in soil encompassed by the Formerly Utilized
Sites Remedial Action Program (FUSRAP) and Surplus Facilities
Management Program.
Used to evaluate the spread of toxic chemicals released to air, soil,
surface water and ground water. RISKPRO was adapted from
PCGEMS.
Core equations were developed in support of the EPA Land Disposal
Banning Rule."
UDAD provides estimates of potential radiation exposure to
individuals and to the general population in the vicinity of a uranium
processing facility.
A multimedia model which links an atmospheric transport model with
a surface water model.
14

-------
Table 3.3 Evaluations of some other representative multimedia models.


Model Name
DECHEM
GENII
GWSCREEN
MEPAS
MMSOILS
MULTIMED
NUREG 5512
ONSITE/MAXI1
PATHRAE
PRESTO-EPA-BRC
PRESTO-EPA-CPG
PRESTO-EPA-POP
PRESTO-II
RESRAD
RISKRPO
SARAH2
Exposure Pathways

External
Exposure

X

X


X
X
X
X
X
X
X
X
X


Soil
Ingestion



X


X



X
X

X
X


Plant,
Meat, Milk
Ingestion
X
X
X
X
X

X
X
X

X
X
X
X


Inhalation
Particulates

X
X
X
X
X
X
X
X

X
X
X
X
X

Radon

X
X
X




X
X



X



Ground
Water
X
X
X
X
X
X
X
X
X
X
X
X
X
X

X
Administrative Issues

Current/
Planned
Use
X
X



X

X
X

X
X
X
X


Validated/
Peer-
Reviewed

X

X

X

X


X
X
X
X


Site Data
Required

Moderate

Extensive

X
Moderate

Moderate

Moderate
Moderate
Moderate
Moderate


Available
Computer
Code
X
X
X
X
X
X

X
X
X
X
X
X
X
X
X
15

-------
                               4. SOURCE TERM
4.1    Introduction
       This chapter deals with the identification and estimation of the source term.  The
 source of emissions and its physical characteristics must be identified before using any
 model.  Source term estimates are provided either by the user, calculated by the model,
 or  back-calculated from  measured  concentrations at the receptor   Source term
 conservation entails mass balance  calculations to ensure that mass is  conserved
 among multiple release pathways.

 4.2    Model Comparisons

 4.2.1  Source type

       Table  4.1  summarizes possible source types  MEPAS  has the most  varied
 selection of source terms, including  simulation of injection wells, underground  tanks,
 landfills, lagoons, direct  subsurface  injection of wastes from  tanks or wells,  and
 trenches with caps. Furthermore, MEPAS is the only one of the three models reviewed
 that allows the user to specify  any mass-flux, time-varying distribution  of the source
 term   MMSOILS has a variety  of source terms also, but  not surface impoundments,
 direct injection to wells, or trenches with caps.  PRESTO-EPA-CPG has the capability of
 modeling  a  variety of  source  terms  including waste  burial   in capped trenches,
 contaminated soil  with and without cover, landfill, and waste piles   None of these
 models account adequately for the  presence of free phase  or residually-saturated
 material  in the source terms.   Neither do any of these  models have the ability to
 consider facilitated  transport,  an especially  important  factor  in  the  transport of
 radioactive species.
TABLE 4.1 Source term capabilities of the models.

Contaminant Source
contaminated soil
injection well
landfill
surface impoundment
trench with cap
underground storage tank
waste pile

MEPAS
yes
yes
yes
yes
yes
yes
yes
Capability
MMSOILS
yes
no
yes
yes
no
yes
yes

PRESTO
yes
yes"
yes
no
yes
no
yes
1. PRESTO-EPA-PILCPG considers a source term in a pile
2. PRESTO-EPA-DEEP
4.22 Estimation of the source term

      Estimates of the source  term are provided either by the user, or are calculated
internally by the  computer.  For example, the release rate of a contaminant spilled on
the ground is calculated from the contaminant's vapor pressure, soil/vapor partitioning.
                                       17

-------
and molecular diffusion in air.  The PRESTO-EPA-CPG model does not consider the
vaporization of radionuclides, and assumes that all contaminants will  be transported
either through atmospheric pathways in absorbed form or through water pathways in
dissolved form.  The fate of release through water pathways is calculated internally
from  the  mass-balance  equation  using  the  inventory   and  leaching/solubility
characteristics of the given radionuclide and the internally calculated stream flow  In
applications to radioactive waste disposal, this approach will  give conservative results
for  health  risk assessments always, especially when the  chemical  forms  of the
contaminant are unknown.

      For the atmospheric component only, MEPAS has the option of back-calculating
release rates from measured concentrations at the receptor.  This requires data on (i)
air concentration, (ii) soil  concentration, or (iii) both  This option  is not available in the
other two models

4.2.3 Conservation of the  source term

      Since the individual pathway models in MEPAS, MMSOILS, and  PRESTO-EPA-
CPG are linked implicitly, verifications of mass  conservation are needed  to prevent
multiple  accounting  of  the  same mass of contaminant in different  media.  MEPAS
accounts for depletion  of the source via  a  link  between the source's  inventory and
release rate and duration that ensures that a release is over. This option addresses
the theoretical need  for mass conservation, but is useful only when the inventory of the
source is known.  Sometimes there is more certainty about the duration of the release
and the  inventory than about the rate of release.  When the  inventory  of waste is
uncertain,  great care  is needed  in  using  the  source depletion  option properly,
especially for long-term simulations (Peer Review Committee report, 1994).

      MMSOILS includes calculations of mass balance annually to ensure that mass is
conserved  in waste management units that have multiple release pathways.  These
calculations compute the  accumulation and depletion in landfills, impoundments, and
waste piles. For each unit,  the mass that is removed from each  pathway is accounted
for annually, to satisfy overall mass conservation.

      PRESTO-EPA-CPG includes mass-balance calculations also to insure that mass
is conserved in waste management units.  This equation calculates the  rate of release
and the mass remaining in the waste units. Then, the radionuclide mass is adjusted for
the radioactive decay at the end of each year

4.2.4 Air pathway

42.4.1   Air source

      MEPAS  has  an   option  allowing  the user to  specify  a uniform ambient
concentration of the  contaminant as a source term when there  are measurements of the
ambient concentration of the contaminant at a waste site  A similar option is provided
                                      18

-------
in  PRESTO  models.   A  user-assigned  strength  of  background  radionuclide
concentration in the atmosphere above the contaminated area can be added to the re-
suspended  radionuclide  concentration  for  assessing  the  combined  health effects.
MMSOILS does not have this option.

4.2.4.2  Radioactive and chemical decay

      MEPAS  can handle both first-order radioactive and chemical decay,  while
MMSOILS  is limited  to chemical  decay.   Progeny formation is not calculated  in
MMSOILS. MEPAS handles progeny formation in ground water, surface water, surface
soil, and  deposited  contaminants  from  wet and  dry  deposition, but not while the
contaminant is moving in the air; it handles it after the contaminant has been deposited.
The PRESTO-EPA-CPG model calculates radiological effects for the progeny produced
by up to a four-member decay chain.

4.2.4.3 Volatilization from soil or spill

      In both MEPAS and MMSOILS, volatilization is calculated by either steady-state
or transient equations, depending on the  scenario of release (Table 4.2).  Steady state
equations are used in scenarios of  landfill release and sediment-controlled  emissions,
whereas time-averaged solutions of transient diffusion equations are used for releases
from spills, contaminated soil,  and land treatment facilities. The steady-state equations
assume a very large source, so that emission does not deplete the source during the
time considered

      In the MEPAS scenarios of releases from spills and land treatment facilities, the
volatilization flux  is  calculated by accounting  for the decrease over  time  of  the
concentration of the  chemical in the soil. A dry-out period  is computed, after  which
emissions stop. MMSOILS  does not define explicitly a similar mechanism  of tracking
and depletion.

       PRESTO-EPA-CPG  assumes that all volatile radionuclides are released in a
water-soluble form and contribute  to the ground water and  surface water pathways.
PRESTO-EPA-CPG was  designed primarily for use  at low-level radioactive  waste
disposal sites, and assumes that the health effects due to  the volatile radionuclides
(primarily 14C and 3H) are negligibly small.  PRESTO-EPA-CPG is not included in  Table
4.2. for this reason.
                                      19

-------
Table 4.2 Source term of volatilization scenarios.

No
1
2
3
4
5
6
Submodel1
Description
Landfill, without
gas generation/
(Farmer's Equation
Covered Sites)
Landfill,
with gas generation
(Municipal Waste)
New spill
Old spill
(Covered Sites.
Adsorbed Phase)
Soil Contaminated up to
the Surface/
(Uncovered Sites,
Adsorbed Phase)
Contaminated Soil
covered with a layer of
Clean Soil/
(Covered Sites,
Adsorbed Phase)
Assumptions
MEPAS
•Steady state
•Very large source (emission does not deplete
source during time frame considered)
•Steady state
•Very large source (emission does not deplete
source during time considered)
•Time-averaged form of transient solution,
emission rates decrease with time
•Emissions occur from liquid above spilled
surface
•A dry-out period is computed, after which
emissions stop
•Time-averaged form of transient solution;
emission rate decreases with time
•Contaminant concentration in cover soil initially
is zero; uniform concentration underneath
cover to finite depth
•Vapor concentration at soil surface is
maintained at zero
•Release rate controlled by soil/vapor
partitioning and molecular diffusion in air
•A dry-out period is computed, after which
emissions stop.
•Time-averaged form of transient solution;
emission rate decreases with increasing time
•Release rate controlled by soil/vapor
partitioning and molecular diffusion in air
•Same assumptions and equations as submodel
4 above
MMSOILS
Same as MEPAS, but
steady-state flux is limited
by mass inventory
NA
NA
•Release rate controlled
by soil/vapor partitioning
and molecular diffusion
in soil gas
(i.e.. controlled by
diffusion of vapor in a
porous medium, which is
1-2 orders of magnitude
less than diffusion in air.
Same as MEPAS
Same as MEPAS
1   Submodel descriptions in /fa//cs refer to MMSOILS.
                                            20

-------
Table 4.2 cont'd Source term of volatilization scenarios.
     Submodel
                                             Assumptions
No
Description
MEPAS
MMSOILS
     Land Treatment
     Facilities/

     (Landfarming Equation)
                       •Time-dependent release rate
                       •Release rate controlled by liquid-phase
                       concentration of contaminant in soil
                       •Contaminant concentration is constant until all
                       its mass vaporizes from liquid-phase	
                                           Same as MEPAS
                                           (slightly different equation
                                           for release rate)
     Sediment-Controlled
     Emissions
                       •Steady state
                       •Includes both sediment-to-water and water-to-
                       air transfer; mass transfer coefficients control
                       diffusion in the two media
                                                                  Same as MEPAS
     Surface impoundments,
     e.g. ponds, lagoons,
     small lakes./
     (Volatilization from a
     Contaminated Water
     Body)	
                       •Two-layer resistance model; a gas and a liquid
                       film across the air-water interface form the
                       dominant resistance to mass transfer
                                           Same as MEPAS (model
                                           not yet in code)'
     General
                             •Emission rates of low-volatility contaminants
                             are constant during time considered
                             •Emission rates of highly volatile contaminants
                             decrease significantly with time; thus, ambient
                             concentrations are computed mainly as a
                             function of total amount of released material
                             rather than emission rate.  User determination of
                             contaminant's total inventory is crucial for highly
                             volatile materials
  1.  Submodel descriptions in italics refer to MMSOILS.
  2.  A model of volatilization from contaminated water is described in MMSOILS manual, but is not
     included in the 1993 computer code. See Model Developer's Comments - Section 4.3.2.
  4.2.4.4 Air-borne depletion due to deposition

         All models account for airborne contaminant depletion via dry deposition.  Wet
  deposition and the associated source  depletion  is included  in MEPAS and PRESTO-
  EPA-CPG, but not in MMSOILS.
  4.2.5 Ground Water Pathway

  4.2.5.1  Generation of leachate

         Contaminants are  introduced  into the  ground water pathway  from  leachate
  originating  in  a waste management  unit.   Leachate migrates vertically through the
  unsaturated zone  and discharges finally into the saturated ground water system.  The
  way in which leachate is generated is similar in the three models.
                                             21

-------
      MEPAS contains the most sophisticated source term for leachate of the three
codes.  MEPAS was set up specifically to permit the user to define the source term if it
is known. The user specifies the total inventory of waste in the unit and the leaching
rate.  The code has a mass-balance check to see that there is enough inventory to
match the amount of material leached from the source.

      If the user cannot define the source term, MEPAS will calculate it based on a
combination of:  (i)  solubility limit on concentration, (ii) equilibrium  partitioning with
contaminated soils, (iii) steady-state concentration of  leachate supplied by user, and
(iv) transient or time-varying releases as specified by user.

      MEPAS provides the user with three source term options:

      Option 1:  the user supplies the  source term  concentration,  the
      code supplies the rate of deep-drainage percolation; then, the code
      calculates the time-varying mass-flux rate. This  information can be
      supplied for a point source,  line source (accounts for both  the  x
      and y directions), or area source The source can be a ponded site
      or a contaminated-soil site. The movement of the contaminant can
      be released directly to the vadose  zone  and then, to the saturated
      zone,  or  it  can  be released directly  to  the  saturated  zone
      Operational releases and  non-operational  releases  (i.e.,  past-
      practice sites) are considered.

      Option 2:  the user supplies time  varying mass-flux rate from  the
      source and  the rate of deep-drainage percolation; then the code
      calculates the initial source term concentration.  This  information
      can be supplied for a point source, line  source  (accounts for both
      the x and y  directions),  or area  source.  The  source can  be a
      ponded site or  a  contaminated-soil site. The  movement of  the
      contaminant can be released directly to the vadose zone and then,
      to the saturated zone, or it can be released directly to the saturated
      zone. This includes direct discharge (e.g., injection well, pipe to a
      river) also.  Operational   releases and  non-operational releases
      (i.e., past-practice sites) are considered.

      Option 3:  a combination of Options 1 and 2.

      MMSOILS  was  designed specifically to address leaching  from landfills and
waste piles. The leachate can be generated from soil, landfills, waste piles, surface
impoundments, and underground  storage  tanks  (USTs).    In  the first three,  the
contaminants are dissolved in infiltrating recharge water derived from precipitation.   By
definition, surface  impoundments contain pre-mixed leachate that infiltrates into  the
unsaturated zone.  MMSOILS uses a continuously mixed reactor model for this  source
type.
                                       22

-------
       Like MEPAS, MMSOILS has several different options for generating leachate for
landfills and waste piles:  (i) solubility limit on concentration of leachate, (ii) equilibrium
partitioning with contaminated soil, (Hi) completely mixed reactor, and (iv) steady-state
concentration  of leachate specified by the user   Thus, the user can choose from  a
variety of options, depending upon how much data are available for the site.  If the data
are  limited, the steady-state option allows  the  user to  specify a  concentration of
leachate.

       PRESTO-EPA-CPG can model  leachate  source  terms from ground surface
contaminated soil, waste trenches with cover,  and waste piles  It has similar options for
generating leachate to those of MMSOILS:   (i) solubility  limit on the concentration of
leachate, (ii) equilibrium partitioning with contaminated waste mixed with soil, and (iii)
release fraction  The latter is similar to the steady-state option of MMSOILS.

      Since landfill leachate is known to contain high concentrations of colloids that
are likely to facilitate the transport of radionuclides. it is especially significant that none
of these models can consider such facilitated transport.

4.2.5.2 Other source terms

      MEPAS and another version of PRESTO-EPA (-DEEP) have additional source
terms for ground water in the form of  injection  wells.  The user  specifies both  the
concentration of contaminant in the injected water and the flow rate  MMSOILS does
not have this source term.

4.2.6 Surface Water Pathway

4.26.1 Soil erosion

      MMSOILS and  MEPAS allow erosion  of the contaminated soil from  the waste
unit. Then the  soil is transported into the surface water where it continues to act as a
source of dissolved contaminants.  The contaminants are dissolved into surface water
using an equilibrium partitioning  approach.   In  the scenario of the source term in
shallow trenches with cover, the PRESTO-EPA-CPG model assumes that the cover is
constructed with clean soil. Therefore, the eroded soil will not contain contaminants as
long as  the cover remains effective {i.e.,  as  soon  as  the cover is eroded)  the
contaminants would begin to dissolve into the surface runoff water and be transported
away from the  contaminated unit   When there  is no cover,  the PRESTO-EPA-CPG
model assumes that the contaminants will be transported into the surface water body
as soon as erosion begins. Thus, the surface  water body will be contaminated from the
beginning of the simulation  An equilibrium partitioning model using a linear sorption-
desorption relationship is used also in calculating the rate of transport of contaminants.

-------
4.2.6.2  Runoff

      All three  codes allow rainfall to leave the unit as runoff.  They assume that the
runoff water is  in  chemical  equilibrium with the contaminated soil,  using  a simple
partitioning model  The assumption that runoff water will be in chemical equilibrium is
an oversimplification   The degree to which runoff water achieves equilibrium will rely
on the partitioning coefficients and the residence time of surface water in contact with
the contaminated surface soils.  In most scenarios, there is likely to be far  less time
than is  required to achieve  chemical equilibrium  As  such, this assumption will be
overly conservative

4263  Ground water inflow

      All three  codes allow  the interception of the contaminated ground water by a
surface  stream   A  complete  mix (also called a "completely mixed") model  is  assumed
by MMSOILS and PRESTO-EPA-CPG as the ground water enters the surface   MEPAS
does not assume a completely stirred tank reactor (a CSTR, operationally the same as
a complete mix  model), but uses a plug-flow with dispersion model  (i.e., a solution to
the advective-dispersive equation)  Such a model accounts for plume  migration in the
lateral direction  from the bank where the source enters the stream.

43   Developer Updates - Source Term

4.3.1  MEPAS

      Droppo (1994) reports that MEPAS Version 3.0 has the capability of direct input
of waterborne  monitoring data in computing  risk values.    A  module to include
geochemistry in the environmental release component of MEPAS is being developed
also The new source term code for MEPAS will provide a coupled contaminant source
term that is partitioned to the different environmental media for transport and exposure.
Two- and three-dimensional,  spatially-varied concentrations for any designated period
will be implemented in MEPAS.  Thus, MEPAS will be able to calculate the contribution
(o downgradient sites from multiple  waste sites.

432  MMSOILS

      The most recent version  of MMSOILS contains a  two-layer resistance model for
air-water interface transfer (see  item 9 Table 4.2).

433  PRESTO-EPA-CPG

      Future versions  of  PRESTO-EPA-CPG  will contain an improved infiltration
submodel for handling uncovered contaminated soil  (Hung, 1994).
                                      24

-------
                               5. AIRTRANSPORT
5.1
Introduction
        The  air-transport  pathway  is  one  of the  principal  pathways  whereby
  radionuclides released from waste sites may reach living organisms   Radionuclides
  may be  discharged to the atmosphere through particulate suspension, venting from
  containers, and volatilization from contaminated water and soil  Once airborne, they
  will disperse downwind and deposit on ground surfaces in a pattern dependent on the
  local  meteorology,  the  location of  the  point of release,  the nature of  the  terrain
  downwind of the release, and the physical and chemical characteristics of the emission
  Exposure to  humans can  occur via direct  radiation,  inhalation,  or  consumption of
  contaminated water, crops, and animals (Figure 5.1).

                                 Direct Radiation
      RADIOACTIVE
       MATERIALS
                                                               HUMANS
 Figure 5.1. Simplified pathways between radioactive materials released to atmosphere
          and humans (after ICRP, 1979).

      The objective of atmospheric transport modeling is to predict the concentration of
 radionuclides at specific locations surrounding the source.  The basic types of data
 required to run these  models include the release  rate of each  radionuclide, physical
 characteristics of the source (e.g., stack height, area, or  release), and meteorological
 data (e.g.,  stability class, wind speed,  precipitation)  For environmental radiological
 assessments, models  should be able to simulate plumes from point sources (e.g.,
 containment leaks) and area sources (e.g., contaminated ponds), for several minutes
 up to several years, and up to about 80 km from the source. Also, these models should
                                       25

-------
include volatilization from soil and water, and particulate emissions from wind erosion
and mechanical operations  As the plume travels downwind, it disperses in the air and
is  depleted also by deposition to ground  surfaces,  radioactive decay, and chemical
decay   Such depletion processes are important for radiological  health assessment.
The  outputs  from  these models  include  the concentration of  air  and  surface
contaminants which can be used in assessing the inhalation and ingestion components
of the exposure  The surface contaminant levels are used also as input to the overland
transport pathways Figure 5.2 shows these interactions.
               (Wind ^Md w* tfrvdton,
                 ctmMpftwtc ftoMty;
                  umpmfcn, «to.)
MrttmkMnto, wrta
MMT. TttJflhMM. MC.)
                          SUSPENSitWEMISSMN
                            •F CONTAMINANTS
                      Surf»c*V»nl
                                           Sur«K*/V*nt
                      Onttmttirt
                     EmtattonftMM
                     _   ATMSSPHEWCinANSPtflT   !
                             AN» MSreftSMW       I
                        Ptum*
                   L
      il
                                 ThSufen
Figure 52 Diagram of pathway interactions (after Droppo et al., 1993)
5.2  Comparisons of models

     The same sector-averaged Gaussian plume equation for air transport is used in all
three models  Most of the volatilization algorithms in MEPAS and MMSOILS are also
the same    However,  MMSOILS and PRESTO-EPA-CPG do not describe complex
                                      26

-------
terrain effects, calms, wet deposition, and contaminant decay.  In addition, PRESTO-
EPA-CPG  does  not  include  several  components  such  as  area  releases  and
volatilization from lagoons, lakes, ponds, and rivers, and the suspension of particles
due to loading and unloading operations.

5.2.1  Radioactive and chemical decay

       Depletion of radionuclides in the plume by radioactive decay may be significant
when the decay is fast (e.g.,  emissions of cesium, iodine, manganese, radon,  and
ruthenium) and transport is slow.  This option is available in PRESTO-EPA-CPG, but
not in  the current version of MEPAS or MMSOILS.  MMSOILS can model first-order
chemical decay once the  contaminant  is  deposited  on soil,  but not radionuclides
explicitly.   None of these  models describes the creation  of progeny  within the air
pathway.

5.2.2  Wet deposition

       Ground deposition can  result from  wet and  dry  processes,  and for many
locations the magnitude of these processes in depleting  airborne radionuclides  is
roughly the same.  Wet deposition is caused by rain scavenging the contaminant and
by deposition of cloud droplets which have absorbed the contaminant.  Dry deposition
is the  direct deposition of  the airborne  contaminant onto a surface by gravitational
settling, or impacting.   Ground deposition  is necessary for linking  the  air-pathway
models with the water and food chain models. Wet deposition  is included by MEPAS
and PRESTO-EPA-CPG,  but  not  by  MMSOILS.     Therefore,  MMSOILS  will
underestimate ground concentrations and  overestimate air concentrations,  with the
error increasing with the distance from the source.

5.2.3  Air source

      Chapter 4, section 4.2.3.1 discusses the air source term for the three models.

5.2.4  Calm conditions

      Calm conditions can be extremely important in  assessing  health impacts to
populations near the source (e.g., up to 10 km). In some locations, the prevailing winds
blow from one direction, and calms (e.g., wind speeds <1m/s) from another direction.  If
calms occur often, they can cause much higher concentrations  at near-field receptors
than predicted by the wind-rose data. The effect of calm conditions is more important
in determining acute effects than long-term ones as such conditions may change over
long periods.  MEPAS can distribute calms as a function of direction, and models them
with a  wind speed of 0.5 m/s.  Calm conditions are not described by the other two
codes.
                                      27

-------
5 2.5 Complex terrain

      The atmospheric pathway in MEPAS takes into account local site influences in a
highly  simplified  manner and  describes  complex  terrain  characteristics such  as
channeling in  a valley, and intersection with hills around a  release.  PRESTO-EPA-
CPG and  MMSOILS account only for different roughness of a flat terrain.   In general,
complex terrain adjustments have more effect on maximum individual exposures than
on average population exposures.  However, in sites where the flow of contaminants
towards surrounding receptors is either interrupted or concentrated by hills or valleys,
these topographical features can affect average population exposures significantly.

526 Acute effects

      MEPAS calculates maximum (hourly)  air concentration and its  location in each
direction to determine acute effects;  the other two models do not.  All three atmospheric
pathway models use annual averages, and predict annual average concentrations and
subsequent exposures.   Average annual  exposures might not represent adequately
strongly  seasonal  (e.g.,  calm  conditions)  or  event-driven (e.g.,  large storms)
environmental transport.   MEPAS  includes  equations  necessary  to describe such
variations, but its present structure is limited to the calculation of annual estimates.

      The features of these  models and their fundamental assumptions are further
described  in Table 5.1.

53   Developer Updates - Air Transport

      As  reported by Droppo (1994), a planned update  of MEPAS will compute a
single  mass  budget  for airborne-waterborne releases rather than  separate  mass
budgets
                                      28

-------
Table 5.1 . Atmospheric pathway: Comparison of capabilities.

No
1



2





3





Submodel
Description
Point Releases
Ground
Elevated
Plume rise
Area Releases
Soil
Landfills
Lagoons
Lakes, ponds, rivers
Multi-point regional
Suspension of particles
Wind speed
Surface roughness (z)
Mechanical disturbance
Loading & Unloading
Soil spreading operations
Availability & Fundamental Assumptions
MEPAS
yes
Sector-averaged Gaussian
Sector-averaged Gaussian
Briggs, 1975
Approximation with point source
at virtual distance and sector
averaged Gaussian plume
yes
yes
yes
yes
no
Sehmel, 1976; Cowherd et al,
1989
yes
0.1
-------
32

-------
                       6. GROUND WATER TRANSPORT

6.1    Introduction

       Often, ground water is an important pathway for wastes found below the land
surface.  Contaminants leach from the waste, move downward through the unsaturated
zone to the water table, and then  migrate in the saturated ground water system. The
contaminants may discharge ultimately either to a drinking water well or to  a surface
stream (Figure 6.1)  Humans are exposed to radioactive and other contaminants by
using  well water or surface water,  and  by  eating organisms  living in the surface
streams

       The ground water component of multimedia models predicts the  concentration
over time at wells and surface discharge areas.  Usually, these calculations are broken
down into three linked sub-pathways:  (i) leaching of contaminants from the waste unit,
(ii) vertical  movement  of  the dissolved contaminant downward to the water table
through  the unsaturated zone,  and  (iii)  migration of the contaminant  in  saturated
ground water to the receptor point. Separate models simulate these three processes,
with the  preceding model supplying a source of contaminated water to  the  next one.
Thus, the leachate generation model  supplies a source of contaminated water  to the
unsaturated zone model, which passes the contaminated water  subsequently  to the
saturated ground water  model at the water table.

       Ideally,  these models would be three-dimensional,  capable of incorporating all
our knowledge of the subsurface, and  of simulating the complex chemical  reactions that
occur  as  the contaminant  migrates   through  the  soil  and  aquifer  materials.
Unfortunately,  even the most sophisticated ground water  models cannot address  all
these  issues.   Since multimedia  models are used often as screening tools  or  for
comparing different sites, each pathway is simplified to incorporate only the most basic
features.  Furthermore, these features must be described with limited data.  For the
ground water pathway, most multimedia models simplify the unsaturated zone to a one-
dimensional (1D)  model which assumes that the contaminant migrates only vertically
from the waste source  to the underlying water table.  In most cases, this is a valid
assumption because the scale of transport in the unsaturated zone tends to be orders
of magnitude smaller than that in saturated ground water.

       Further  simplifications are made for the saturated ground water model.  The
most common assumptions are that  ground water moves at a  uniform rate  and is
unaffected by pumping  wells, changes in  recharge, or other systems stresses  These
assumptions are much  less realistic than the simplifications made to the unsaturated
zone models, but for screening they are adequate as long  as the user understands the
degree of  uncertainty  in the model's results.  The following  disclaimer  from  the
MMSOILS manual is a good synopsis of the problems inherent in this approach:
                                      33

-------
      "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."

62   Comparisons of Models

      Table  6.1  outlines  a  range  of capabilities  for modeling the ground water
pathways. The approach to modeling the ground water pathway in MEPAS (Whelan et
al. 1987) and MMSOILS (US. EPA,  1992)  is similar.  However, PRESTO-EPA-CPG
(U.S. EPA 1987) differs from the other two codes in many ways which are enumerated
below.
  Figure 6.1. Conceptual diagram of the ground water pathway (Whelan et al. 1987).
                                     34

-------
6.2 1  Radioactive decay/progeny ingrowth

      Each model incorporates a first-order decay coefficient to simulate decay. While
MMSOILS  is  not  specifically  designed to  consider  radioactive  decay,  it  does
incorporate a decay term for non-radioactive chemicals  However. MMSOILS limits the
decay to dissolved chemical contaminants; which when adsorbed onto the soil material
do  not undergo decay. While this is a conservative assumption which would lead to
persistence  of the chemical contaminant  in the environment, it is overly conservative
since many radionuclides are strongly sorbed onto most soils.

      Only  MEPAS and PRESTO-EPA-CPG simulate  ingrowth of progeny during
decay; this may be important if travel  times in ground water approach the half-life of the
contaminant, or if the progenies are particularly toxic, or if the progeny is in a different
physical state (e.g., radium to radon).

      MEPAS assumes that the progenies have the same distribution coefficient (Kd)
as the parent.  This is seldom the case and should be considered when reviewing the
output from MEPAS  for scenarios where progeny ingrowth is important   This drawback
is only significant when the parent nuclide has a short half-life or when the travel time is
long relative to the parent's half-life.   The NRC  NUREG-0868 states  that -  the
assumption  of  equal  transport  speeds  makes  a relatively  small  difference to  the
calculation of concentrations of the most  important components, and is conservative "
The user always has the option of modeling the decay products as parents instead of,
or in addition to, the actual parent radionuclides, thereby bounding the problem without
making any assumption about
      There are two versions of PRESTO-EPA-CPG available for simulating progeny
ingrowth: a research model and a screening model. The screening version assumes
that progeny have the same K
-------
Table 6.1 Ground water pathway: Comparison of capabilities.
Capability
Unsaturated Zone
Miscible Transport
Model Type
Advection
Dispersion
Diffusion
Physical/Chemical Processes
Decay
Progeny Ingrowth
Sorption
Other Chemical Reactions
Immiscible Transport
Vapor Transport
Saturated Zone
Miscible Transport
Model Type
Advection
Dispersion
Diffusion
Physical/Chemical Processes
Decay
Progeny Ingrowth
Sorption
Other Chemical Reactions
Immiscible Transport
Density-Dependent Flow
Fractured Zone
MEPAS

yes
1 D semi-
analytical
yes
yes
yes

yes
yes'
linear

no
yesj

yes
3D semi-
analytical
1D
3D
3D

yes
yes'
linear
no
no
no
no
MMSOILS

yes
1D finite-
element
yes
yes
no

yes1
no
linear

no
no

yes
3D semi-
analytical
1D
3D
no

yes1
no
linear
no
no
no
no
PRESTO

yes
1D
analytical
yes
no
no

yes
yes
linear

no
no

yes
1D
analytical
1D
2D
no

yes
yes
linear
no
no
no
no
1.  No decay occurs while contaminant is sorted onto soil.
2.  Progenies have the same adsorption coefficients (K
-------
 same  model (VADOFT) as that  used in Rustic (USEPA,  1989).  a more complex
 subsurface simulator.

       MEPAS uses a one-dimensional, semi-analytical transport model that assumes a
 constant vertical velocity for each layer in the vadose zone. The user may describe the
 vadose zone with multiple layers.  Therefore, the user can account for heterogeneity in
 the vadose zone  by  modeling multiple vadose-zone  layers with velocity variability
 between layers, and infiltration rates that change with time. MMSOILS. with its layered
 finite-element model computes a  non-uniform vertical  velocity based upon changing
 soil properties.

       PRESTO-EPA-CPG  uses  an  empirical  formula developed  by  Clapp  and
 Hornberger (1978) to calculate the average degree of saturation which then is used to
 obtain the unsaturated water  velocity and  retardation factor.   Then, the  rate of
 radionuclide  transport  is calculated from a  steady-state, one-dimensional transport
 equation.

 6.2.3  Saturated Zone Model

       MMSOILS  and  MEPAS take similar  approaches to  simulating  transport of
 contaminants in the  saturated  ground water flow, using a  three-dimensional, semi-
 analytical transport model.  The three-dimensional adjective  is somewhat misleading.
 however, because only  dispersion is considered  in three dimensions  Ground water
 velocities are assumed to be uniform and horizontal,  and thus one-dimensional   The
 codes used in these semi-analytical  transport models are similar to the AT123D code
 (Yeh,  1981).

       Both  the  MEPAS  and  MMSOILS  mathematical   formulations   employ  a
 convolution integral that distributes a transient release from the source to the receptor.
 thereby avoiding problems  with convergence and instability since each time step is
 calculated independently of its predecessor.  Short (i.e., 1 year) or long (i.e.. 10 million
 year)  simulations are possible without the risk of problems of instability or convergence
 because the user-defined time  steps are independent.  A potential problem  may occur
with this integration scheme if the upper and lower bounds and limits of integration are
 not properly selected.

      The PRESTO-EPA-CPG model employs a simple one-dimensional model (Hung.
 1986) to achieve the  goal of analyzing the annual rate of radionuclide transport for
 10,000 years.  Although an analytical model,  PRESTO-EPA-CPG can adopt any form
of boundary conditions and can consider lateral and longitudinal effects  It is designed
as a  screening-type  model, aiming  at an accurate health-effects evaluation with  a
minimum of numerical calculations. To achieve these goals, a correction factor  called
Hung's correction factor compensates for the rate of radionuclide transport obtained
from the analytical model without lateral dispersion effects (Hung, 1986)  Since Hung's
correction factor is derived by matching the total mass of radionuclides  passing through
a particular layer, theoretically,  the estimated cumulative health effects should agree
                                      37

-------
with the results obtained from a numerical approach   As to  the  maximum dose
analysis, the solution using Hung's model and that using a numerical  approach are
fairly close, in general, to each other as long as the transport number is less than 4 and
the Peclet number is  greater than 2 (Hung 1986).  In a general waste disposal site risk
assessment application, the transport number and the Peclet number are usually within
the domain described above   Therefore  the PRESTO-EPA-CPG model can calculate
the maximum  individual  dose with minimum  error compared to the  exact solution
obtained from a one-dimensional model

6.2.4  Mixing-zone calculations

      The results of  the unsaturated zone model are a flux of contaminants that serve
as a source term for  the saturated zone model  The volume of aquifer over which the
contaminant flux is diluted when it first enters the aquifer is called the mixing zone
Each  model treats the mixing  zone differently  The least conservative approach is to
mix the contaminants over the entire saturated thickness of the aquifer,  as  is done in
PRESTO-EPA-CPG.  That approach causes maximum dilution, resulting in  the lowest
possible concentration of contaminants  in the saturated zone beneath the source
However, the unsaturated zone concentration can be concentrated rather than diluted.
as occurs  when the water flow from the unsaturated zone into the aquifer exceeds the
ground water flow out of the source

      MMSOILS computes the vertical  dispersion coefficient and  a mixing depth,
called the depth of penetration,  using the ratio of the horizontal rate of ground water
flow to the vertical flow-rate of water  from the unsaturated zone  As the  amount of
vertical  flow and the vertical dispersivity increase, the thickness of the mixing zone
increases

      MEPAS assumes that the flux of contaminants occurs at the water table and the
depth of mixing depends directly upon the vertical dispersivity value.

      Both MMSOILS and MEPAS assume that the pumping of a well does not affect
the background flow  of ground water   This assumption is  applicable only when the
unit-width  flow strength is much greater  than  the  pumping rate   However, a waste
disposal or cleanup site may be situated  in a region of low or moderate ground water
flow, a situation that would disqualify this  assumption.  In such a region,  a well screen
would be  set near the bottom of the  aquifer and pumped  with several times higher
capacity than the daily demand  In most cases, this would create a free surface draw-
down  in the  vicinity of the well, so that the contaminant plume will be pulled down and
mixed with the bottom layer of  clean water  Although this mixing may occur only in the
vicinity of the well, the quality  of water being pumped out would be close to complete
mixing.  For these reasons, the assumptions of the  PRESTO-EPA-CPG model may be
reasonable

-------
6.2.5 Adsorption of contaminants

      All three models assume a linear, fully reversible, adsorption model in both the
unsaturated and saturated zones.  The only difference between them is that MMSOILS
assumes that contaminants do not decay while sorbed to the soil matrix

626 Complex processes

      None of the models simulate complex transport processes, such as vapor phase
transport,  fracture flow, or  immiscible phase transport.   At a particular  site,  these
processes  may be important.  For example, transport of radon  in soil gas may be
important for exposure assessments, if radium is one of the contaminants of concern
627 Requirements for data input

      The requirements for data input in MMSOILS (Table 6.2). MEPAS (Table 6.3).
and PRESTO-EPA-CPG (Table 6.4) differ significantly.  MEPAS requires the most data,
especially for the soil and unsaturated zones.  A data-input guide is provided that
explains each parameter and gives suggestions on selecting a value for a parameter
One problem with this guide is that metric and English units are mixed   For example,
hydraulic  conductivity is entered in units of ft/d,  while bulk density is  expressed  in
g/cm3.   According to the  developers, the  default units employed  by MEPAS were
chosen to conform with those  most likely to be found in data source documents  The
data-input  shell  of  MEPAS can be  redefined to employ  any  units  (consistent  or
otherwise) desired by the user, since this model can convert any set of units used

      Modeling the unsaturated zone within MMSOILS  could require significantly more
data input than the other two  models, if numerous soil layers are incorporated in the
model (10 data elements are  entered for each soil layer).  System International (SI)
units are  used consistently throughout MMSOILS  except for time  units which are
expressed in years for the decay coefficient,  and hours for hydraulic conductivity

      PRESTO-EPA-CPG  requires very little data for  the  ground water pathway
consistent with its low sophistication in the subsurface transport models  The units are
internally consistent in PRESTO unlike MMSOILS and MEPAS.

6.3   Developer Updates - Ground water Transport

      Improved capabilities for ground  water  transport of light,  non-aqueous phase
liquids  (LNAPLS)  and dense,  non-aqueous phase  liquids (DNAPLS) are planned for
future versions of MEPAS according to Droppo (1994). Version 30 of MEPAS includes
also  an  improved calculation for waterborne transport that  incorporates  double-
precision    mathematical    routines    to    address   problems    that    would

-------
arise because of round-off errors.  New versions of MEPAS will account for curvilinear
flow in the saturated zone also.
 Table 6.2 Data requirements for MMSOILS.
Unsaturated Zone
       Climatic Data
               pan factor for converting pan evaporation to potential evapotranspirat
               latitude of site
               precipitation data (12 months)
               number of days with precip. (12 months)
               average temperature (12 months)
               pan evaporation (12 months)
               starting and ending month  for growing season

       Soil Property Data
               curve number of surface soil
               Held capacity of soil
               wilting point of soil
               depth of root zone (cm)
               number of soil layers (entering the following for each layer)
                      saturated hydraulic conductivity (cm/hr)
                      saturated water content
                      bulk density (gm/cm3)
                      exponent "b" for moisture curve
                      percent organic matter
                      percent clay
                      percent silt
                      percent sand
                      ratio of first-order decay (1/yr)
                      thickness of layer (cm)
Saturated Zone
       hydraulic gradient (dimensionless)
       hydraulic conductivity (m/yr)
       porosity (dimensionless)
       bulk density of aquifer material (gm/cm1)
       fraction of organic carbon in aquifer
       dispersivity (x, y, and z-directions) (m)
       aquifer thickness (m)
       1st order decay rate (yr1)
       time (yr)
       x,y,z locations (meters)
                                              40

-------
Table 6.3  Data requirements for MEPAS.
Unsaturated Zone
        depth of waste site in unsaturated zone (ft)
        length of site (ft)
        width of site (ft)
        waste liquid infiltration rate (ft/d)

        Top Soil Data
                moisture content of soil
                bulk density of soil (g/cm3)
                soil textural classification
                percent sand, silt, and clay
                percent organic content
                percent iron  and aluminum
                pH
                percent vegetative cover
               water capacity
               SCS curve number

        Partially Saturated Zone Data
               soil textural classification
               percent sand, silt, and clay
               percent organic matter
               percent iron and aluminum
               PH
               thickness of partially saturated zone (ft)
               bulk density (g/cm3)
               total porosity
               field capacity
               dispersivity (ft)
               saturated hydraulic conductivity (fVd)

Saturated Zone
               soil textural classification
               percent sand, silt, and clay
               percent organic matter
               percent iron and aluminum
               pH of pore water
               total porosity
               effective porosity
               contaminant velocity  (ft/d)
               thickness of saturated zone (ft)
               bulk density of saturated zone (g/cm3)
               travel distance in saturated zone
  	longitudinal, transverse,  and vertical dispersivity (ft)
                                                41

-------
Table 6.4 Data requirements for PRESTO-EPA-CPG.
Unsaturated Zone
       length of contaminated zone (m)
       average slope of contaminated zone
       component of porosity for gravity water
       component of porosity for pedicular water
       thickness of the top layer (m)
       equivalent diffusivity (m2/hr)
       maximum day length for each month (hr)
       mean daily temperature (*C)
       hourly rainfall (0.1  mm/hr)
       percentage of top layer failure
       annual infiltration rate (m/yr)*
       fraction of residual saturation
       distance from bottom of trench to nominal depth of aquifer (m)
       porosity
       hydraulic  conductivity (m/yr)
       bulk density (g/m3)
Saturated Zone
       distance between well and stream (m)
       distance from trench to well (m)
       ground water velocity (m/yr)
       thickness of aquifer (m)
       dispersion angle (radians)
       bulk density (g/rrn
       longitudinal dispersivity (m)
       porosity    	
• internally calculated, output from transient zone calculation.
                                              42

-------
        7. RUNOFF, EROSION, AND SURFACE WATER TRANSPORT

7.1   Introduction

      Surface  water  may  be an  important pathway  by which  contaminants are
transported from a waste disposal site; such pathways include lakes, streams, rivers,
and the rainfall-runoff process. Contaminants present on the surface  of a waste site
may become entrained or dissolved in surface runoff and be transported to adjacent
bodies of surface water.  Contaminated soil particles detached by the impact of rain
drops or eroded by surface runoff may be transported to surface bodies also. Exposure
to humans can occur then through drinking and using contaminated surface water, or
by eating organisms living in these water bodies (Figure 7.1)
Figure 7.1 Pathways for surface water exposure.

      The pathways for surface water from a contaminant source to drinking water (or
immersion exposure) can be sub-divided into three major components:

      •  Infiltration into the subsurface and  eventual discharge of ground water into
         surface water bodies
      •  Runoff across the land surface transporting dissolved, suspended, and bed
         load as overland flow
      •  Mixing of the contaminated surface  or infiltrating ground water with a surface
         water body.

      The magnitude of each of these components can be predicted using empirical or
physically-based equations that model the transport of contaminants through overland
and ground water flow to a surface water body.  These equations need to consider the
following classes of variables:
                                     43

-------
      •  The quantity and intensity of precipitation
      •  The physical characteristics of the soil that control the soil's ability to absorb
         and transport  water  and  contaminants,  including  antecedent  moisture
         conditions
      •  The nature of vegetative cover
      •  Topographic features of the landscape, such as slope and the presence of
         depressions
      •  The physical characteristics of the surface water body into which runoff flows
         and ground water infiltrates

      The  mechanisms by which  different radionuclides  are transported  to surface
water bodies usually vary with their  respective geochemical properties  Radioactive
metals in solution tend to bind to soil components,  leaving only a small  fraction in
solution  Conversely, some radioactive constituents such as tritium tend to remain in
solution with only a small fraction binding to soil materials.  Consequently,  radioactive
metals in streams  are related to surface runoff (event  water) and soil  erodibility
generally,  while  more  soluble constituents  may  be  related more to  infiltration,
dissolution, and ground water discharge.  It  is essential to model  processes (flow
pathways),  in addition to flow amounts, in order to analyze behavior of these and other
contaminants  in  natural systems.   This conclusion would imply that physically-based
models should produce more  accurate predictions for the surface water components
than empirically-based models.

      Methods for  simulating  surface water transport fall into two broad categories -
empirically  and physically based  Two  examples of these are  the Morton and  the
variable-source-area (VSA) theories   Morton developed his theory for the infiltration
process in the early 1930s. The Morton model  (and several later variations) are based
on the concept of the soil's surface as a barrier to vertical  flow  The VSA  theory was
presented first by Hewlett (1961) and by Hewlett and Hibbert (1967)  in response to
phenomena unexplained by traditional hydrologic theories.

7.2     Model Comparisons

      MEPAS, MMSOILS,  and  PRESTO-EPA-CPG  each  take  slightly  different
approaches to simulating overland flow and the surface water pathway  However, their
difference may not  significantly affect constituent concentrations at points-of-concern.
Each model  incorporates a surface  contaminant source term,  a runoff  model,  an
erosion  model, and a  surface water mixing  component.  Table  71  compares  the
capability of the models

7.2.1  Rainfall-runoff

      A significant percentage of the water associated with rainfall-runoff is present in
the subsurface flow system before a  storm begins (antecedent moisture)   In addition,
simple hydrologic abstractions  based  upon conditions  at the soil surface cannot predict
these contributions,  nor their origin within a watershed, even though proper  application
                                       44

-------
of these techniques will  quantify rainfall-runoff  accurately.  The complexity of the
rainfall-runoff process, the importance  of  surface-subsurface  interactions,  and the
modeling  of surface water flow  pathways are not fully  implemented into  MEPAS,
MMSOILS, or PRESTO-EPA-CPG; no subsurface contributions to surface water bodies
are implemented  into their rainfall-runoff components.   Subsurface contributions to
surface water bodies are  computed solely  by the ground water  submodel.  PRESTO-
EPA-CPG  uses a physically-based  approach  to estimate runoff  by evaluating the
vertical movement of water in the vadose zone beneath the site.  Modifying  this
procedure  to evaluate vadose  and water table conditions  throughout the  catchment
would generate runoff computations consistent  with VSA theory. However,  MMSOILS
and MEPAS (both of which are based upon an Hortonian approach  to the rainfall-runoff
process) cannot simulate rainfall-runoff flow pathways.
Table 7.1 Surface water pathway: Comparison of capabilities.

Runoff Calculations
Erosion
Transport
Miscible
Adsorbed on Soil
Stream Mixing
MEPAS
SCS-CN'
MUSLE4

yes
yes
coupled,
2D, steady-state
advection-dispersion
MMSOILS
SCS-CN1
USLE2

yes
yes
uncoupled,
1D
complete mix
PRESTO-EPA-CPG
Physically Based
Deterministic Approach
USLE: 3

yes
no
uncoupled
2D, steady-state
advection-dispersion
1. Soil Conservation Service Curve Number method
2. Universal Soil Loss Equation
3. Eroded material is removed from the model and not transported off-site
4. MEPAS uses the Modified USLE - MUSLE (Onstad and Foster 1975; Onstad et al . 1976;
Mitchell and Bubenzer, 1980; Novotny and Chesters, 1981)
yes or note = option is available
no = option is not available
      The U.S. Soil Conservation Service Curve Number (SCS-CN) method (a variant
of the Morton method) is used to compute runoff in the MMSOILS and MEPAS models
This empirically-based procedure was designed to be used on a storm-by-storm basis
The SCS-CN method employs a series of curves that relate runoff volume (Q)  to
precipitation (P) and watershed storage (S):

                        Q = (P-0.2S)2
                            P + 0.8S
                                     45

-------
This approach is used widely, and accurately predicts runoff volumes on the scale of a
watershed. MMSOILS modified the interpretation of a storm in implementing the SCS-
CN procedure.  The total monthly depth of precipitation is broken into individual storms
of equal magnitude by dividing the monthly value by the total number of days per month
with precipitation  The antecedent moisture conditions are estimated from the number
of days it rained in a particular month  The authors of MMSOILS recognized that these
assumptions would give an approximation of runoff volumes using the SCS-CN method,
but that this procedure was consistent with the available data.

      The rainfall-runoff component  in MEPAS uses the  same procedure and is based
upon the same assumptions as made in MMSOILS.  In the rainfall-runoff module in
PRESTO-EPA-CPG, overland flow  is modeled using  a  modification  of the one-
dimensional momentum and continuity equations, and infiltration rates are based upon
subsurface conditions  The equation for surface water, soil moisture,  and evaporation
each is solved on the time-scale of minutes and hours over the course of an individual
storm   However, a peer review subcommittee (USEPA,  1984)  that evaluated  the
rainfall-runoff sub model of PRESTO-EPA-CPG  was concerned that:  ••  (i) the rate of
infiltration is insensitive to variation  in the slope of the trench cap:  and (ii) the results
are highly dependent on the initial water storage value which  can not be initialized by
the user"  Hydrologists have recognized for a  long time that these two factors  are
important  for estimating  runoff.   Nevertheless,  since the  solution of the  equations
employed by PRESTO-EPA-CPG is  an initial value problem, the error in  the assumed
initial value can be eliminated by ignoring the initial period of  analysis.  Based on  the
extensive trial  runs that EPA performed, the maximum length  of simulation has  never
exceeded 5 years, and normally was 3 years,  before the initial water storage value
reaches equilibrium   Since the PRESTO-EPA-CPG model has had this adjustment to
the initial  water-storage value built-in to the model, the difficulty of initializing  the
storage value is virtually solved.  As to the insensitivity of the rate of  infiltration to  the
slope of trench cap, this is caused by the relatively large numerical error.  This error
can be suppressed by selecting smaller space and time steps.

      Input Data requirements for the three models are summarized in  Table 7.2.
PRESTO-EPA-CPG  has extensive requirements for climatological data, and hourly
precipitation data are required for each day of the year to run this model   Rainfall-
runoff  and the associated erosion  and  transport of  contamination  are  short-term
transient  phenomena controlled by  individual storms.  A single, short-duration high-
intensity thunderstorm will cause  more erosion  and transport of constituents than a
long-duration,  low-intensity  storm  of equal volume   Detailed climatological data  are
required to simulate  these processes.

7 2.2 Erosion and transport components

      All three models use the universal soil loss equation (USLE) to approximate soil
erosion across  the site (Wischmeier and Smith,  1978)   This  equation is based on a
number of empirical relationships (factors) that followed  from the analysis of years of
accumulated rainfall and erosion data  The equation that embodies this approach is:
                                      46

-------
                             A=R.K»SL«V«P

where the average soil loss (A) is the product of the rainfall (R). soil credibility (K),
slope  (SL),  vegetation  (V),  and  management/conservation  practice  (P)  factors
Representative values for these factors for areas within the 48 contiguous states are
given in Wischmeier and Smith (1978).
      MEPAS uses a modified form of the USLE (MUSLE) recommended by Foster
(1976) for use at sites where storm events are  to be analyzed   In MMSOILS and
MEPAS contaminant transport  is linked  with erosion.   These two models have
components that simulate constituents in the runoff which are in solution and  sorbed
onto the soil being eroded off the site.  PRESTO-EPA-CPG accounts for the erosion of
the  cap  at  a site,  but  does not transport the  eroded  material and  associated
contaminants; it is simply removed from the system. This is because the standardized
PRESTO-EPA-CPG model assumes that the material used for trench-cap construction
is clean soil. The eroded soil would not contain contaminants as  long as the protective
cover is functioning effectively. However, when the protective cover is totally eroded
away, the soil would begin  to contain  contaminants.   However, PRESTO-EPA-CPG
simulates  the transport of dissolved constituents in  surface runoff   Contaminants
adsorbed  onto  soils  exposed to  precipitation and runoff are  partitioned off,  and
considered to be in equilibrium with constituent concentrations  in the water.  These
transport  components in each of the models represent distributed source terms (non-
point sources) to down-slope surface water bodies.

7.2.3 Surface water mixing

      MEPAS, MMSOILS, and PRESTO-EPA-CPG use a similar approach to assess
radionuclide  impacts within surface water bodies.   Constituents  may be  delivered via
the rainfall-runoff process, seepage through the movement of ground  water, or as direct
discharge from the  waste site. MEPAS uses a plug-flow-with-dispersion approach for
the surface waterborne  component (Whelan et al., 1987).  MMSOILS uses a simple.
complete-mix model for assessing  impacts   MMSOILS includes stream and surface
water body (lake or pond) submodels, while MEPAS includes a stream and  a wetlands
component.  The wetlands component in MEPAS has sufficient flexibility to simulate a
surface water body also. The stream component is a one-dimensional approximation in
MMSOILS, while  MEPAS includes a two-dimensional,  vertically averaged stream-flow
model  with  unidirectional advection in the flow  direction  and dispersion  in the
transverse direction (Whelan et al., 1987).   PRESTO-EPA-CPG  considers  flow within
surface water bodies to  be one-dimensional.  However, transport is  evaluated using a
two-dimensional  (lateral and transverse)  steady-state approximation     Additional
surface water pathway components can be incorporated also  into the PRESTO-EPA-
CPG  and  MMSOILS  models  because  within these two models the surface water
components are uncoupled from the other submodels.

                                     47

-------
Table 7.2 Data input requirements of the surface water pathway.

Climatological Data
Precipitation
Storms
Site Parameters
Vegetative Cover
Land surface Slope
Antecedent
Moisture
Waste
Concentrations
Soil Parameters
Bulk density
Porosity
MEPAS
MMSOILS
PRESTO

Monthly Summary
Index on storm type
Monthly Summary
Average Number of
Storms/Month
Hourly Data
—

yes
yes
SCS curve number
yes
yes
yes
Estimated from storm
frequency
yes
yes
yes
Computed from soil
conditions in the
vadose zone
yes

yes
Calculated from
specific weight and
bulk density
yes
yes
yes
yes
7.3   Developer Updates - Surface Water Pathway

7.3.1  MEPAS

      Droppo  (1994) reports that efforts are underway  to add to  MEPAS specific
environmental transport capabilities for wetlands, lakes, estuaries, and river sediment
interactions.

732  PRESTO-EPA-CPG

      The modified  version of the PRESTO-EPA-CPG model (designed for cleanup
scenarios) will  be  able to simulate contaminants in the surface  runoff from the
beginning of a simulation,  if the scenario of contaminated soil  without  soil  cover  is
selected.
                                     48

-------
                          8. FOOD CHAIN MODELING

8.1 Introduction

   Food chains  are  biospheric pathways  through  which humans  are  exposed  to
environmental  contaminants.    They  are represented  by  bioaccumulations  of
contaminants  in the edible portions  of animals and  plants that are affected  by the
facility.  Food chains consist of one or more trophic levels (steps) between the physical
environment and human intake of contaminants (Figure 8.1).
        ATMOSPHERIC DEPOSITION (wet or dry)  	» p|ants  	> Animal Products
                                                             Humans
                                 irrigation
        WATER (surface or ground)                 >  Plants 	> Animal Products
                                                          >  Humans

        SOIL	> Plants  	*• Animal Products
                              Humans
       SURFACE WATER	+ Fin-or Shellfish  	* Humans


       AQUATIC SEDIMENT        >  Fin-or Shellfish           » Humans
           Figure 8.1.  Basic food chains as depicted in risk assessments

8.2   Comparisons of Models

      Food  chains are  described  as   part  of  the  exposure-dose  models   in
documentation for MEPAS (Whelan, 1993), separately in MMSOILS  (USEPA.  1992).
and  as  agricultural data  supporting  exposure  estimates  in  PRESTO-EPA-CPG
(USEPA, 1987).  MEPAS documentation has detailed formulas to describe each aspect
of food chain modeling.  As  part of its equations for animal feeds, leafy  vegetables.
and other produce, MEPAS separates descriptions by the origin of plant contamination
from wet or dry atmospheric deposition, irrigation, or accumulations in soil   Both foliar
and root uptake  are described   Table 8.1  summarizes the descriptions  of the food
chain in the models, and Table 82 gives more detailed descriptions of each level in the
food chain.

                                      49

-------
      MEPAS, MMSOILS,  and  PRESTO-EPA-CPG  use  inputs  of time-weighted
average  concentrations of contaminants.   Equations are  given  for estimating the
average  concentrations of contaminants at the  location  of exposure resulting from
transport in media and food chains.  Measured concentrations can be used instead of
those calculated in MEPAS
Table 8.1 Summary of food chain features in models.

Selection of Contaminant
Hazardous Chemical Waste
Radioactive Waste
Transfer to Plants
From Air
From Irrigation
From Soil
Transfer to Animal Products
Meat and Milk
From Air Deposition or» Feed Crops
From Irrigation of Feed Crops"
From Soil Consumption
From Drinking Water
Finfish
Shellfish
MEPAS
yes
yes
yes
yes
yes
yes
yes
yes
no
yes
yes
yes
MMSOILS
yes
no
yes
yes
yes
yes1
yes
yes
yes
yes
yes
yes3
no
PRESTO-EPA-CPG
no
yes
yes
yes
yes
yes
yes
yes
yes4
yes
yes4
no
1  Combines intake from forage, soil and water.
2  Includes pasture grass, stored feed and leafy vegetables.
3.  Calculations based on either water or sediment concentrations.
4.  Included in the modified model adaptable to cleanup scenahos.

      The  irrigation pathway to food chain transport is not well described in MMSOILS,
and uptake into plants does not account for edible fractions. For vegetation, MMSOILS
uses  a  general food  chain equation  to  describe   atmospheric  deposition  and
bioconcentration from contaminated soil, with a specific description in root crops of the
bioconcentration of contaminants from soil.  For plant materials, the use of measured
concentrations in the three models negates the need for such distinctions.  MEPAS and
PRESTO-EPA-CPG distinguish  leafy vegetables from  other types of produce.  The
number of days in a growing  season that vegetable matter including pasture grass and
feed is exposed to radionuclides and the fraction of animal feed that is contaminated is
considered in PRESTO-EPA-CPG estimates.   MEPAS includes  the  length  of  the
growing season in vegetable and feed-crop production  also. All feed for milk cows is
assumed to be contaminated forage. All feed for meat animals is assumed to be grain-
type plants.  If some of the feed is uncontaminated, then the user may modify the intake
rates of animal feed to reflect the proportion of contaminated intake.
                                       50

-------
Table 8.2 Details of food chain levels in the models.

Plant
Meat
Milk and
Milk Products
Fish
MEPAS
Feed crops for
animals
Leafy vegetables
such as lettuce,
cabbage, and spinach
Other vegetables
including grains, root
crops, and food not
generally exposed to
depositional material
beef
pork
bovine, poultry
cow
freshwater finfish and
shellfish
MMSOILS
Pasture grass and hay
Above around
includes leafy (lettuce). and
silage (com), exposed
produce (non citrus fruits,
berries, field crops —
cucumber, tomato,
squash, eggplant)
Below around
root crops
beef
cow
freshwater finfish
PRESTO-EPA-CPG
Grass
Leafy vegetation
Produce
including grains and root crops
beef
goat
cow
goat
freshwater finfish and shellfish
      A single formula in MMSOILS describes transfers to animal products (meat and
milk) by combining the chemical concentrations  in,  and the intake rates of, animal
feeds,  soil, and drinking water.  The transfer factors in  MMSOILS  either take into
account the fraction of food that is fat and bioconcentration in the organism, or use
partition coefficients for transfers from soil to meat and/or to milk. In MEPAS, transfers
to animal products account for direct transfers from media, and transfers from media to
forage  plants and feed. At present, MMSOILS does and  MEPAS does not describe
transfers to animal products from accumulations in ingested soil, but MEPAS  includes
animal ingestion of contaminated water.   PRESTO-EPA-CPG includes atmospheric,
irrigation,  and  drinking water  sources of contamination  of vegetation  and animal
products.

      Whole-body concentrations  of  chemical and radiological contaminants are
considered in both finfish and shellfish by MEPAS and PRESTO.  MMSOILS describes
only chemical concentrations in "fish" (assumed from the documentation to be finfish),
but can base its calculations  on concentrations  in water and/or sediments.    Bio-
concentration factors can be selected for fish in general, or for particular species.  The
bioconcentration factors for contaminants  in the water column are adjustable also for
the lipid content (%) of fish. MMSOILS uses a sediment-to-fish partition coefficient for
chemicals in sediments. MEPAS does not relate concentrations in aquatic animals to
those in the sediments.

8.3   Developer Updates Food Chain

      None of the models plan future updates that would  have any direct impact on
assessing transport and exposure via the food chain.
                                      51

-------
                         9. EXPOSURE ASSESSMENT

9.1    Introduction

       Often, exposures and  doses are confused.  Exposures are quantities of toxic or
carcinogenic agents that are potentially taken into the human body, or (as described by
Ruttenber,  1993a,b) "...the  contact  between  an  organism  and  its environment"
Exposures (e.g., mg, g, pCi) to a contaminant are estimated usually by multiplying the
intakes of  environmental media or  foods  (e.g.,  L,  m3,  kg)  by  the  respective
concentrations of the agent in the media and/or foods (e.g., mg per L, g per m3, pCi per
kg) and summing their products. Exposure rates include specified periods, such as mg
per d or pCi per yr.  Exposures to chemicals are generally expressed in metric units of
mass,  while exposures to radionuclides are expressed in standard units of picocuries
(pCi) or in System Internationale (SI) units of Becquerels (Bq)

       Exposure pathways are the last  stage of transport modeling,  and include those
parts of the transport directly  related to the behavior and characteristics of those at risk.
Exposure estimates are affected by uncertainties in pathway exposure factors (PEFs -
McKone, 1990).  PEFs are terms that translate unit concentrations in media (e.g.,  pCi
per L) and food chain components into exposures per unit time (exposure rates). PEFs
use  information on human physiology  and behavior, and environmental transport for
specific media (McKone,  1990).   These uncertainties arise from  biological  and
behavioral variations, as well  as the accuracy and precision of the estimated  values for
each parameter. Many characteristics important to personal exposure are constant, or
have a near-constant distribution nationwide (e.g., breathing rates).  Others (such as
intake  of local foods) can be  significantly different, depending on local production and
exports.

9.2    Comparison of Models

      The  scope and  complexity of the food  chain  and exposure calculations for
multimedia pathways vary with each of the three models. MEPAS covers a wide variety
of sources  of movement in  the food  chain and human exposure to chemical  and
radioactive  contaminants  (Table 9.1)   This coverage includes on-site  and off-site
pathways.   MMSOILS analyzes on- and off-site exposures to chemicals buried in  the
soil.  PRESTO-EPA-CPG analyzes on-site and off-site food chain transport and human
exposures for buried, low-level radioactive wastes.

      MEPAS,  MMSOILS,  and  PRESTO-EPA-CPG  use  inputs  of  time-weighted
average concentrations of contaminants. The  exposure modules provide users  with
equations for estimating the average concentrations of  contaminants at the location of
exposure resulting from transport in media and food chains. Measured concentrations
can be used instead of those that are calculated in MEPAS and PRESTO-EPA-CPG,
but not in MMSOILS.

-------
Table 9.1 Summary of exposure features in models.

Contaminant Selection
Hazardous Chemical Waste
Radioactive Waste
Intakes from Inqestion of
Drinking Water
Shower Water
Swimming Water
Leafy Vegetables
Other Produce
Meat
Milk
Finfish
Shellfish
Special Food
Shoreline Sediment
Soil
Intakes from Inhalation
While Showering
Of Air
Of Re-suspended Soil
Intakes from Contact
While Showering
While Swimming
With Shoreline Sediment
With Soil
With Volatiles in Air
External Exposures from Radiation: while
Swimming and Boating; in air, Soil and
Shoreline Sediment; or from
Measurements of Direct Radiation.
MEPAS

yes
yes

yes
yes
yes
yes' \
yes12
yes
yes
yes
yes
yes
yes
yes3

yes
yes
yes

yes
yes
yes
yes3
yes
yes



MMSOILS

yes
no

yes
no
no
yes
yes
yes
yes
yes
no
no
no
yes

no
yes
yes

yes
yes
no
yes
no
no



PRESTO-EPA-CPG

no
yes

yes
no
no
yes
yes
yes
yes
yes
yes
eggs
yes
yes4

no
yes
yes

no
no
no
no
no
air immersion, ground
surface


1  From air deposition on crops.
2  From irrigation of crops.
3  Estimations based on either measured concentrations or on calculated
  accumulations in soil after atmospheric deposition.
4  Included in the modified model adaptable to cleanup scenarios.
5  Root crops.
      MEPAS and MMSOILS (as described below) incorporate exposure rates in their
formulas for doses from chemical contaminants.  Both MEPAS and PRESTO-EPA-CPG
use  additional  special algorithms  for MC
analysis (see Chapter 10).
and  H  in  vegetable  and animal product
      Although  MEPAS  and MMSOILS  describe  human exposures  from  ingesting
water during swimming, only MEPAS describes ingestion during showering  In MEPAS.
exposure rates for ingestion of contaminants in vegetable matter,  in animal  products.
and in special foods are described separately for each transport medium
                                       54

-------
      MEPAS separately calculates the rates of ingestion of contaminants in shoreline
sediments and the soil contaminants.  Soil contaminants can be described either by
measured concentrations, or by calculated accumulations from atmospheric deposition.
Similarly, uptakes from contact with shower water, swimming, shoreline sediments,  and
soils are described separately.

      MEPAS calculates exposure rates  separately for inhaling  contaminants in air,
showers, and re-suspended soil.  Also, it is the only  one of  the three models  that
distinguishes indoor from outdoor inhalation, which  is  an important  procedure in
estimating exposures to volatile organic chemicals and radon

      For radioactive materials, MEPAS models external exposure from irradiation by
radionuclides in air, soil, shoreline sediments, and from radionuclides in water during
swimming, and boating. Measurements of direct radiation can be used in the exposure
calculations also.

      In MMSOILS, exposures to non-carcinogenic toxic chemicals are compared to
available Reference Dose  (RfD) or  Health  Advisory  (HA) values for comparable
intervals. Average daily intake rates (mg per kg per d) are time-weighted for 1 day, 10
days, longer periods, and for sub-chronic  and chronic periods.  HA values  (exposure
rates in mg per day) are modified to RfD units (dose rates) by adjusting for body weight.

      For radionuclide exposures, PRESTO-EPA-CPG estimates include the period of
the growing season that vegetable matter is exposed, and fractions of human intake of
water and animal  products that are contaminated. Animal products are distinguished
as meat and milk from cattle and/or goats.

9.3   Developer Updates - Exposure Assessment

      Future versions of MEPAS will include ingrowth of progeny in the  exposure
assessment  component of the model (Whelan,  personal communication, 1994).   No
updates are  planned for MMSOILS and PRESTO-EPA-CPG that would have a direct
impact on the calculation of exposures for any of these models
                                      55

-------
                      10. DOSIMETRY/RISK ASSESSMENT

 10.1   Introduction

       Dose-assessment codes integrate two components into the assessment of risk:
 one component calculates the transport mechanism of the released contaminant, and
 the other calculates the human  uptake and dose associated with the exposure   This
 chapter explores how MEPAS, MMSOILS, and PRESTO-EPA-CPG translate exposures
 from various sources  into numerical estimates of dose and risk (e.g., the number of
 excess fatal cancers).  The estimation of radiation dose is discussed in some  detail
 below  because dose estimation for radioisotopes must be treated differently than that
 for chemicals.

       In contrast to chemicals, the effect of radioactive contaminants on humans is a
 function of the nature of the energy released during their radioactive decay  Radiation
 is emitted by a  radioisotope when it  transforms (disintegrates) into another isotope
 (e.g., 226Ra to 222Rn).  Each radioisotope is unique, and the decay rate, energy, and the
 type of radiation differ. The term activity expresses the  number of disintegrations of a
 radioisotope per unit time. The fundamental unit of activity was the Curie (Ci,  or 3.7 x
 1010 disintegrations per second, equal to the activity of 1 g of 226Ra), but the current SI
 unit is  the Bq (i.e., 1 disintegration per second).  The most common types of ionizing
 radiation are alpha (a) and beta (B) particles, and gamma (g)  rays (photons).   Each of
 these three  types can be emitted over a  range of energies,  expressed commonly in
 units of thousands of  electron volts (KeV), or millions of electron volts (MeV)  Each
 radionuclide has its own characteristic radiation type or types  and range of associated
 energy levels. Alpha particles can travel only about 2  or 3 cm in air and no more than
 about  0.01 mm in body tissue because they have mass.  On the other  hand,  beta
 particles can travel much further in air and tissue, yet have about the same kinetic
 energy. A 1  MeV beta particle can travel approximately a meter in air and can likely
 penetrate the thickness of human skin,  but not much beyond that   The minimum
 energy required for skin penetration for alpha and beta radiations are 7.5 MeV and 70
 KeV, respectively.  External alpha radiation is not often of concern for the purpose of
 radiation protection since few alpha decays achieve energies in that range.   On the
 other  hand,   a  weightless  1  MeV gamma-ray can penetrate  a sheet of paper  or
 aluminum foil easily, and could pass entirely through  the  human body.   Therefore,
 gamma radiation is the principal source of concern for external  radiation exposures

       Doses for chemicals are the amounts  of toxic or carcinogenic agents  (or their
 metabolically activated products) that reach a tissue or organ  within the body. Doses
 are expressed in mass of agent accumulated per unit mass of organ or body weight.
 Dose rates include specified periods in the expression  (e.g., g per kg per d, pCi per kg
 per d).   Dose estimates are derived generally from exposure estimates by  using dose
conversion factors, which  usually represent the fraction of exposure that is delivered to
a target organ (or target organs), with adjustments for temporal factors,  retention, and
conversions  to  active  progeny.   Dose  factors have been  defined extensively for
radionuclides, but there is a great deal of uncertainty  for chemical  doses.  Therefore,
                                      57

-------
risk factors are expressed often in  reference to  exposures because  it  is easier to
quantify exposures (e.g., EPA's HEAST and IRIS expressions of risks per unit ingested
or inhaled).

      Radiation dose has been expressed in units of rads (radiation absorbed dose),
and rem (roentgen equivalent, man).  The rad is a measure of the energy per unit mass
(100 erg/gm) delivered to any mass when a given amount of radiation is absorbed. The
rem is a special unit of dose equivalent and depends on the type of radiation absorbed
(ICRP, 1990)   The dose equivalent rem is  numerically equal to the product  of an
absorbed dose in rads and a quality factor (also termed the radiation weighting  factor
[ICRP, 1991]).   For example,  neutrons and alpha particles  deliver  energy in high
density packets, while X- and gamma rays deliver the same amount of energy at lower
density.  Quality factors for  radiation are  represented usually by the term relative
biological  effectiveness  (RBE), which  is a ratio of the magnitude  of  a particular
biological effect of one type of radiation to the magnitude of the same effect of another
type of radiation.  The SI units for rad and rem are the Gray (Gy) and Seivert (Sv),
respectively.  One Gy (1 Joule/kg) is equal to 100 rads, while one Sv is equal to 100
rem.

      One of the key parameters in deriving the dose to an individual or population is
the dose conversion factor. This factor relates a given intake of radioactive material to
a  radiation  dose.    In  general,  the  dose  conversion  factors  are  derived  from
recommendations made by  the International Commission on  Radiation Protection
(ICRP)    The ICRP recommendations  on dose  limits  are the primary guidance
documents used by international and national organizations for estimating effects of
ionizing radiation exposure on radiation workers and  members of the public (ICRP,
1977, 1979, 1986, 1988, 1989, 1990).  Within the United States, the methodology for
dose conversion is based on  guidance from five different organizations: the ICRP, the
National Council on Radiation Protection and Measurement (NCRP, 1993), DOE (DOE,
1988a, 1988b), EPA (USEPA, 1988a, 1993), and NRC.

102  Review of Models

      Outlined below are the methods used by PRESTO-EPA-CPG and MEPAS to
calculate doses from radiation exposures.  MMSOILS does not calculate a radiation-
based dose; it provides comprehensive analyses of pathways and receptors for on- and
off-site exposure, and it is limited to the transport of non-radiogenic chemical materials.
Using MMSOILS to  predict radiation dose from exposures  requires additional health-
physics analyses to convert the intake dose and correct for decay.  For this reason, we
discuss only chemical and not radiation dose for MMSOILS in the text that follows.
                                      58

-------
 10.2.1 Overall dose

 10.2.1.1     Chemical Dose

      Intakes of chemicals are referred to as administered doses (AD) to conform with
 RfD values, and MMSOILS assumes that the absorbed fraction of  intake  is 100% in
 humans  ADs are calculated as mg per day for water and food ingestion  (vegetable,
 meat, milk and fish), and inhalation or mg per occurrence for soil ingestion  The AD for
 soil  contact are  expressed as  mg  per visit to  the contaminated area,  and  the
 calculations may  require modifications to adjust for alterations of chemicals during
 cutaneous transfer. Although MMSOILS does not specify ingestion of, nor contact with,
 shoreline sediment, it is assumed that  these exposures  can be  described  by  the
 comparable equations for soil ingestion and contact

      Carcinogenic chemical  intakes, as 70-year  lifetime average  daily  doses,  are
 compared to EPA's potency factors (now supplanted by slope factors — USEPA, 1989)
 There is a marked inconsistency in the use of 75-year lifetimes for toxicity and 70-year
 lifetimes for carcinogenicity.

 10.2.12     Radiation Dose

      The  current version  of PRESTO-EPA-CPG  employs dose  conversion  factors
developed by Eckerman (USEPA, 1994)  to calculate dose from internal and external
exposures (Rogers and Hung, 1987).  The dose conversion factors are extracted from
the RADRISK data file (Dunning et al., 1980) and the weighting factors are consistent
with the definitions used  in ICRP Publications 26 and 30 (ICRP, 1977 and 1979).  The
 effective dose equivalent is the weighted sum of the 50-year committed dose equivalent
to the specified and remainder organs.  The cancer risk coefficients are calculated from
the radiation risk models based on  1980  U.S. vital  statistics.  On the other hand,  the
 radionuclide genetic risk coefficients for serious heritable disorders to all generations
are calculated from the product of the average absorbed dose to the ovaries and testes
up to age 30 per unit intake before that age.  Genetic Risk Coefficients of  2 60 x 10~2
and 69 x 10~2 Gy"1 for low-LET and high-LET radiation are used to calculate the risk
conversion factors, respectively (USEPA, 1994b).

      PRESTO-EPA-CPG employs  the DARTAB  submodel to estimate both exposure
and annual committed doses (mrem per y) for as many as 40 radionuclides, by adding
the weighted doses to each organ  from  all  pathways of exposure.  Exposures from
ingested intakes of radionuclides are expressed as person pCi per year   External
irradiation  exposures from air or contaminated ground surfaces are.  expressed as
person-pCi  per m3 (volume), and  person-pCi per  m2 (surface) respectively  DARTAB
combines estimates of radionuclide exposure with dosimetric and health effects data to
generate predicted impacts.

      MEPAS  uses  the methodology in  Federal  Guidance  Report -11  (FGR-11
[USEPA, 1988]) to calculate dose from internally deposited radionuclides For external
                                      59

-------
exposures, MEPAS obtains dose using the Eckerman (USEPA, 1994) dose conversion
factors.   In contrast to PRESTO-EPA-CPG which calculates and  reports a  dose.
MEPAS calculates dose rate, as rem per unit time, but only reports fatal cancer risks
These risks  (5.0 x  10"2 fata! cancers  Sv"1) are consistent with  the  latest  ICRP
recommendation (ICRP, Publication 60, 1991) for radiation workers.  If MEPAS is used
to calculate risk for the whole population, the suggested risk factor is 7.3 x 10"2 fatal
cancers Sv"1  (NCRP, 1993);  i.e.. the sum of the fatal cancer, non-fatal  cancer, and
severe genetic effects. These cancer effects represent a stochastic outcome. MEPAS
does  not  calculate an equivalent dose for any single organ or tissue for evaluating
deterministic  health  effects  like those associated with  organ-seeking nuclides (e g ;
iodine for thyroid, and strontium for bone).

      MEPAS and MMSOILS estimate doses (e.g., mg per kg per d; pCi per kg per d)
on the basis of average  daily exposures. This may not determine risks  accurately, for
example, when exposures vary widely in intensity and duration.  PRESTO-EPA-CPG
uses annual exposures because it estimates maximum doses on a year-by-year basis
(up to 10,000 years) but, overall, this is similar to the assessment of radionuclides in
MEPAS.   On  the other hand, PRESTO-EPA-POP calculates the fatal  cancer and
serious  genetic risks for the local and downstream basin population  However the
uncertainty of the overall risk assessment tends to increase with time because of the
increasing  uncertainty of demographic predictions  Since  the  PRESTO-EPA-CPG
model uses a hypothetical scenario in which the critical population resides in a user-
assigned location and  calculates  the  dose for each person, the uncertainty  in the
demographic  distribution will not be the concern of this analysis.  In PRESTO-EPA-
POP,  the  uncertainty in the  demographic  distribution will  affect  significantly the
uncertainty of the number of the fatal cancer and serious genetic effects  - a situation
that applies to all models.

10.2.2 Inhalation dose

      There is a significant relationship between particle size, lung retention, and dose
for estimating dose due to inhalation.  The physical association  of  radioactivity with
micron size and sub-micron size particles is crucial for the inhalation pathway   The
smaller particles penetrate deeper into the  lung,  are more efficiently  deposited there.
and are cleared inefficiently as long as they are insoluble

      All assumptions associated with the intake-dose constants of FGR-11 (USEPA,
1988) become default parameters within MEPAS.  MEPAS uses a 1 micron  particle size
for aerosols based on FGR-11.  Furthermore, MEPAS uses the ICRP-30 dynamic lung
model and its compartment  parameter  values,  the so-called  D-W-Y lung  clearance
class  for chemical transportability  (ICRP,  1979).   In this case,  D refers to  those
materials that can be  removed from the lung in days, W  in weeks,  and Y in  years,
respectively.   The PRESTO-EPA-CPG organ dose rates are calculated  from pre-
calculated conversion factors, derived using the default value of 1  micron  particle size
for aerosols and the lowest solubility D-W-Y lung class (class Y is used in most cases)
These default values are recommended by ICRP.
                                      60

-------
 10.2.3 Radon/Thoron progeny and long-lived fission products

      The modified  PRESTO-EPA-CPG  version  adaptable  to  cleanup  scenarios
 includes a radon inhalation pathway.  MEPAS does not derive dose calculations for
 natural radon/thoron progeny (RTP), but allows the user to specify a radon emission
 rate from which atmospheric transport of radon and  progeny is evaluated using "mock
 radon" radionuclides.  This representation allows an evaluation of dose from radon.  It
 includes an equilibrium amount of short-lived progeny  and the maximum amount of
 long-lived progeny, thus providing  a conservative estimate of radiation risk.  222Rn, and
 220Rn are a concern at sites contaminated with 226Ra or 224Ra, respectively.  Since many
 sites have high levels of Ra, failure to account for dose due to 222Rn,220Rn would be a
 significant deficiency.

      Long-lived fission  products  and  nuclear  materials   cannot  be  neglected
 necessarily because of their low specific activity,  especially  those that have been
 classified as "most hazardous materials" such  as neptunium (Np), plutonium (Pu), and
 americium (Am).  On the  other hand,  those radioisotopes with short half-lives (e.g.,
 isotopes of Np, Pu,  and Am other  than 237Np, 239Pu,  and 241Am) are of less concern to
 any health risk  assessment of  long-term waste disposal.  However,  the  release of
 contamination that includes long-lived isotopes like those of 237Np, 241Am, and 239Pu is
 of the utmost  concern for any environmental safety and health study.

 10.2.4 Period of exposure

      PRESTO-EPA-CPG model calculates the rate of dose commitment (mrem/yr) by
 multiplying the exposure rate (pCi/yr) with the dose-conversion coefficient (mrem/pCi).
 A 50 year dose commitment factor is used for  this calculation.  The value of the dose
 commitment factor is obtained from FGR-11 for each radionuclide.  MEPAS bases its
 dose analysis on intake for a user-defined exposure duration (70 year default).

 10.2.5 Individual vs. population-based dose

      MEPAS calculates the dose to an individual located at  some place in time and
 both individual and population risks.  Early versions of MEPAS calculated a  Hazard
 Potential Index (HPI) which is based on the population exposure.  Neither the MEPAS
 or MMSOILS models calculate onsite exposures.  The  original version of PRESTO-
 EPA-CPG model calculated the annual and maximum dose to the on-site farmers from
 drinking,  irrigation, and cattle-feed pathways.  The estimation  of the cumulative fatal
 and genetic health  effects does not include normally the effects to on-site farmers.
 Since the  PRESTO-EPA-CPG model calculates the  total population health effects by
adding the local population health effects with downstream population health effects,
the user may combine the on-site farmers into the  local population.   The results of
analysis  include  the  fatal and genetic  effects  from   the on-site  farmer and  the
downstream population.

                                      61

-------
10.3  Developer Updates - Dose/Risk Assessment

10.3.1  MEPAS

      Planned updates for MEPAS include modules to calculate ecological risk, acute
human risk, and worker risk {Droppo. 1994).

      MEPAS  is being incorporated into the  Remedial Action Assessment System
(RAAS), a screening tool for cleanup remedies. MEPAS' baseline risk assessment is
used  by RAAS as  a starting point for estimating residual risk  to  evaluate  the
effectiveness of alternative remedies

1032  PRESTO-EPA-CPG

      PRESTO-EPA-CPG is  being  modified to include dose from  progeny nuclide
ingrowth and for the soil-ingestion pathway (Hung, 1994).  Version  2.1 of  PRESTO-
EPA-CPG contains also revised computational models, procedures, and data for dose
conversion to conform with ICRP 30. SI units are adopted also for  radioactivity and
dose equivalent.   The modified PRESTO-EPA-CPG  model  will improve the on-site
residence  to cover  all applicable  scenarios that could be  adapted to  a cleanup
scenario.
                                     62

-------
                         11. UNCERTAINTY ANALYSIS

 11.1   Introduction

       Uncertainties  in risk  assessment are important in making risk  management
 decisions.  In a risk assessment, descriptions of uncertainties may indicate the quality
 of information,  range of knowledge, and  level of confidence  available.   Extensive
 sensitivity and uncertainty analyses were made to develop  EPA's Low-Level Waste
 Environmental Standards.  Several factors contribute to these uncertainties, including:
 limitations  in the  data  that  characterize  sites and  source terms,  uncertainties  in
 scenarios  and  choices of  parameters to fit  different  scenarios,  uncertainties  in
 formulating the transport model and in physical parameters used  as input to the models
 (e.g., diffusion coefficients),  exposure parameters,  and dose-response relationships.
 The word parameter is used  in this document as a component (= property or variable)
 that can  be  characterized  either quantitatively  or  qualitatively.   Some  factors
 contributing to the uncertainty in final risk estimates are more important than others.

       Uncertainty  arises from  combinations of heterogeneity  (variability), errors  in
 measurement, and lack of knowledge.

       •     Heterogeneity is the variability within a parameter, such as  the
         variability in the characteristics of a  population.   For example,  it is
         relatively easy to determine the amount  of water that an  individual
         drinks  daily, but the amount  will  vary from  day to day  and among
         individuals in a population.

       •     Error in measurement arises from  inadequacy of  sampling,
         sampling biases, errors in the measurements,  and imprecision

       •     Lack  of knowledge can involve  parameters that  are  expressed
         quantitatively and components of a risk assessment that do not have
         numerical values.  Major sources of  uncertainty include inadequate
         knowledge of physical processes, such as  environmental transport
         mechanisms, and gaps in qualitative knowledge, such  as future land-
         use scenarios. Parameters and their ranges of values can be affected
         profoundly  by choices among these components of a risk assessment,
         in turn affecting the overall uncertainties of the risk estimates.

       Most parameters used in risk assessments contain elements of heterogeneity,
errors  in measurement, and lack of knowledge.  For example, the amount of water that
is imbibed daily is  heterogeneous across a  population, but each sample is  subject to
errors in measurement and sampling bias.

      As part of a risk assessment, uncertainty analyses  should  be performed to
determine which parameters exert a significant influence on the overall risk estimates.

                                      63

-------
      Sensitivity  analyses are  used  to  assess which parameters  are the  most
important contributors to the magnitude of an overall  risk estimate, and they frequently
are undertaken as part of a screening-level assessment.  These analyses compare all
parameters in an assessment for the overall effects of  a specific degree of change
(e g . a 20% variation) in each parameter (Morgan and Henrion, 1990)

      Uncertainty analyses estimate the contribution of uncertainty associated with
each variable to the overall uncertainly of a  risk estimate (Rish, 1988)   A sensitivity
analysis can be performed as part of an uncertainty analysis to identify the parameters
that contribute the most to the variance of the final risk estimates.  In other words, the
analysis quantifies the sensitivity of uncertainty of a risk estimate to a changed range or
assumed type of distribution of a single variable.

11 2  Comparison of Models

      The three models use deterministic (single) values for parameters   It is difficult
in  such deterministic assessments to sort out the contributions of individual parameters
to  the overall uncertainty of the risk estimates, because the calculations can combine
high (90th or 95th  percentile) parameter values with lower (50th percentile or average)
values (Burmaster and Harris, 1993)

      The accuracy of the value of the parameters in  all three models is difficult to
verify  Although such difficulties are balanced by the user's ability to input alternative
values, no specific instructions are provided for  performing uncertainty analyses as a
way of estimating the adequacy and precision of assumptions   The "Documentation
and Users  Manual for MMSOILS"  (USEPA,  1992)  expresses the clearest concern
about the overall uncertainty of the risk estimates, and of specific parameters  MEPAS
and MMSOILS suggest and  allow input of  site-specific and  region-specific data to
reduce uncertainty for food chain and exposure parameters, and to provide alternative
choices for doing  some calculations when the  concentrations  of contaminants  have
been measured at specific sites. MEPAS also gives some regional data in its reference
tables  MEPAS supplies a report, based on site-specific sensitivity analysis, that gives
the user information on sensitive parameters in each of the codes comprising the model
(Doctor et al., 1990). A second MEPAS report (Droppo et al., 1990) provides a system-
wide uncertainty analysis for representative sites  and  constituents.

113  Developer Updates - Uncertainty Analysis

      An operational version of a sensitivity/uncertainty module for MEPAS is being
tested at several  sites (Droppo,  1994).  Besides user-input parameters,  a version is
planned that  will  allow sensitivity analysis via Monte-Carlo analysis on the physical,
chemical, and toxicity  parameters associated with health  impacts  from  the  MEPAS
database   The addition  of the capability  to  perform uncertainty analysis within the
existing PRESTO-EPA-CPG model is planned
                                       64

-------
             12. PARAMETER ESTIMATION AND DEFAULT VALUES

 12.1   Introduction

       Inaccurate values for input parameters can be a major source of error in health
 risk assessments.  Input data can be faulty because of poor judgment in estimating the
 parameter and over-reliance upon default values which may not be  applicable to  a
 given  scenario.   The data from which parameter values were derived  may not be
 relevant to the  specific set of conditions to  be addressed  by the model.   The
 determination of a parameter  value  itself carries some inherent  uncertainty  since
 processes within environmental models have a large  natural variability  in time and
 space, and such values are based often on experimental data that refer to only  a few
 discrete points in time and space.

       In  this chapter,  the  three models are  compared  in terms  of their  data
 requirements, availability  of default  values,  and  guidance  to  select  or  estimate
 parameter values.  This chapter  assesses also the relative importance of different
 parameters in the given models, based on information in the manuals and in previous
 reviews.  The parameters of greatest importance are the ones that have  the greatest
 potential  impact  on  the  outputs for the  particular  assessment  (i.e.,  long-term
 environmental and public  health risks  for  environmental  restoration  and  waste
 management activities).  In this  context,  four categories  of model parameters are
 considered following the approach outlined by IAEA (1989): (i) source parameters (e.g.,
 rate, time, and duration of a release, nuclide speciation, source strength, chemical and
 physical  form of the  release radionuclides), (ii) environmental transport parameters
 (e.g.,  wind speed, precipitation height, porosity, sorption, partition coefficients, soil
 hydrogeological  properties), (iii)  bioaccumulation  parameters,  and  (iv) dose  and
 exposure parameters (e.g., living and consumption habits, health standards).

 12.2   Comparison of Models

 12.2.1  Description of the data input requirements for models and their default values

       MEPAS has data on more than 576 chemicals and radionuclides in the chemical
 and sorption Kd databases.  The former includes information on a) physical properties,
 b) environmental decay, c) environmental transfer,  d)  radiological dosimetry, and e)
 chemical toxicity.  The sorption K
-------
      Sorption K^s for organic constituents are computed within MEPAS using
values, organic matter content, and the soil's proportion of sand, silt, and clay  The
inorganic and  organic Kd values determined by MEPAS are provided to the  user as
suggested values for each constituent of the soil layer; as such they are not presented
as "correct" values  Rather, they are meant to represent typical values  that might be
found for the  constituents of  concern  The user can take the  suggested  value  or
replace it with site-specific information.

      MMSOILS has a  database with chemical,  transport,  decay, and chemical  or
radiological dosimetry characteristics for 240 pollutants.  The code requires several
pathway-specific pieces  of information (e.g.. atmospheric  pathway, surface  water
pathway, ground  water  pathway,  infiltration leaching and recharge, food  chain
bioaccumulation pathway). One hundred and seventy-seven input values are required,
many of which are  provided in tables or suggested in the users manual.  The manual
gives guidance for selecting or estimating the parameter values,  but that guidance is
not as detailed as that of MEPAS.

      PRESTO-EPA-CPG requires input data on site-specific and radionuclide source
terms, hydrogeologic and meteorological conditions, radiological dosimetry, and health
effects  The manual describes all the required parameters,  but includes default values
and/or guidance to select/estimate values for  only  six of them   Since the model
provided input  data sets for the humid permeable, the humid impermeable, and the arid
permeable sites in the United  States, it gives  a  wide spectrum of sites across the
United States.   Furthermore, since each of the submodels  within PRESTO-EPA-CPG
is  designed  to be  as dynamic  as  possible,  most of the  required input parameters
represent system characteristics and are measurable.  That  is, the PRESTO-EPA-CPG
model uses less dependent variables that should be calculated, in theory, by the model
as user-assigned input parameters.  Therefore, the model requires less guidance  to
assign input parameters that would require extensive empirical data bases.

12 2.2 Comparison  of values from MMSOILS and MEPAS databases

      Table 12.1 compares the values from the databases of MEPAS and MMSOILS
for a relatively immobile chemical, arsenic.  Table  12.2  shows the same  data for
benzene, a relatively mobile chemical. PRESTO-EPA-CPG was excluded because (i) it
includes only radionuclides, not  chemicals, and (ii) the documentation  provides only a
few default values   These comparison tables provide examples of the range  in input
parameters that can exist between MEPAS and MMSOILS for two typical contaminants.

12.2.3 Source term  parameters

      Problems related  to  quantifying the  parameters of the  source  term  include
sparse or inaccurate information about  identifying  the  types  of wastes  present,
determining the quantities of waste,  and estimating waste  distribution.  Release and
solubility parameters define the  mass of leachate to be released to the subsurface from
a waste management unit. Some site-specific parameters are difficult to  ascertain, yet
                                      66

-------
they can have a significant impact on the results. For example, one such parameter is
the rate  of suspension of particles  caused  by vehicles disturbing the site.   MEPAS
accounts for  the silt content of  the  road,  vehicle  weight,  length, frequency  of
mechanical disturbances, and other such parameters.  Different analysts could choose
easily different values of these parameters for the same site.


12.2.4 Environmental transport and decay parameters

      An important parameter that  defines the distribution of the source term among
different  media is the  Kd.  K
-------
       Although  these comments may be correct for  a specific problem that is being
assessed, they are not universally true.  For example, if the problem is temporally far-
field, the rate of release has no effect on the results; only a change in the source  term
inventory affects the results  For a temporally near-field case, the  rate of release
determines the risk. If the soil type is radically altered, then the risk will be influenced
substantially by this change.
Table 12.1 Input Parameters for Arsenic.

Molecular Weight (g/mol)
Vapor Pressure (mm Hg)
Henry's Law Constant (atm mj/mol)
Solubility In Water (mg/L)
Organic Carbon Partition Coefficient (Koc) (mUg)
Octanol-Water Partition Coefficient (Kow) (mUg)
Partition Coefficient, Kd, (ml/gm) - at pH=
Kd for clay+organics <10% of total soil (mL/g)
Kd for clay+organics 10-30% of total soil (mUg)
Kd for clay+organics >30% of total soil (mL/g)
Kd for aquifer (VnL/g;
Environmental Half-Life In Air (days)
Environmental Half-Life In Water (days)
Environmental Half-Life In Soil (days)
Ground Water 15' Decay (1/yr)
Unsaturated Sediments 1ST Order Decay (1/yr)
Chemical Decay In The WMU ( 1/yr)
Chemical Decay Constant In Off-Site Field (1/yr)
1SI Order Decay In Stream (1/yr)
Fish Bio-Accumulation Factor (1/kg)
Bio-concentration In Fish (mg/kg fish)/(mg/L water)
Sediment/fish Partition Coef. (mg/kg fish)/(mg/ kg soil)
Shellfish Bio-Accumulation Factor (1/kg)
Soil-To-Plant Uptake
Soil-To-Meat Partition Coef. (mg/kg beef)/(mg/kg soil)
Feed-To-Meat Coefficient (d/kg)
Feed-to-Cow Milk Coefficient (g/L)
Transfer Factor For Cattle (kg/kg)
Transfer Factor For Milk (mg/kg milk)/(mg/kg intake)
Soil To Milk Partition Coef. (mg/kg milk)/(mg/kg so;/)
Uptake From Soil To Plant (mg/kg plant)/(mg/kg soil)
Soil Moist. To Root Factor (mg/kg root)/(mg//kg solute)
Water Purification Factor
Deposition Velocity (m/s)
Atmospheric Deposition Class
Inhalation Cancer Potency Factor (kg-d/mg)
Ingestion Cancer Potency Factor (d/mg)
MEPAS
75
0.0
NA
0.0
NA
0.0
>9 5-9 <5
0.6 5.86 5.86
2.0 19.4 19.2
2.0 19.4 21.5
NV.6.9E+07'
NV1, 6.9E+07'
NV\ 6.9E+07"
NA
NA
NA
NA
NA
100'. 1.0*
NA
NA
40
0.0015' 0.01
NA
0.002
0.002
NA
NA
NA
NA
NA
0.7
0.001
1
5.0E+01
1.5
MMSOILS
75
0.0001
1.E-08
100000
NV
NV
354
354
3.54
354
NA
NA
NA
0.0
0.0
0.0
0.0
0.0
NA
1.00
0.00
NA
NA
0.00
NA
NA
0.002
000006
0.00
0.04
0.00
NA
NA
NA
NA
NA
NV  = no value is listed in electronic database   NA = not applicable
1.   = value in electronic database of MEPAS version 3.x
2.   = value in manual (Chemical Databases; Strenge, Peterson, and Sager, 1989)  Note that the
    values in subsequent versions of the manual are continuously updated to conform
    with the values in the electronic database.
                                          68

-------
Table 122 Input Parameters for Benzene.

Molecular Weight (g/mol)
Vapor Pressure (mm Hg)
Henry's Law Constant (atm mj/mol)
Solubility In Water (mg/L)
Organic Carbon Partition Coefficient (Koc) (mUg)
Octanol-Water Partition Coefficient (Kow) (mUg)
Partition Coefficient (Kd) For Waste - at pH=
Kd for clay+organics <10% of total soil (mUg)
Kd for clay+organics 10-30% of total soil (mUg)
Kd for clay+organics >30% of total soil (mUg)
Kd for aquifer (mL/g)
Environmental Half-Life In Air (days)
Environmental Half-Life In Water (days)
Environmental Half-Life In Soil (days)
Ground Water 1st decay (1/yr)
Unsaturated Sediments 1&: Order Decay (1/yr)
Chemical Decay In The WMU (1/yr)
Chemical Decay Constant In Off-Site Field ( 1/yr)
1S1 Order Decay In Stream (1/yr)
Finfish Bio-Accumulation Factor (1/kg)
Bio-Concentration In Fish (mg/kg fish)/(mg/L water)
Shellfish Bio-Accumulation Factor (1/kg)
Sediment/Fish Partition Coef. (mg/kg fish)/(mg/ kg soil)
Soil-To-Edible Plant
Soil-To-Meat Partition Coef (mg/kg beef)/(mg/kg soil)
Beef Uptake (d/kg)
Feed-To-Cow Milk Coefficient (g/L)
Feed-To-Meat Coefficient (d/kg)
Milk Update (d/L)
Transfer Factor For Cattle (kg/kg)
Transfer Factor For Milk (mg/kg milk)/(mg/kg intake)
Soil To Milk Partition Coef. (mg/kg milk)/(mg/kg soil)
Uptake From Soil To Plant (mg/kg plant)/(mg/kg soil)
Soil Moist. To Root Factor (mg/kg root)/(mg//kg solute)
Water Purification Factor
Deposition Velocity (m/s)
Atmospheric Deposition
Inhalation Cancer Potency Factor (kg-d/mg)
Ingestion Cancer Potency Factor (d/mg)
MEPAS
78
95
5.6E-03
1750
83
1.32
computed
NV1, 6.9E+07-
NV\6.9E+07-
NV1, 6.9E+07'
NA
NA
NA
NA
NA
0\24'
NA
0\3.9'
NA
o'.o.ss-
NA
0',3.36E-06'
0.002
0.002
1 .05E-06
NA
NA
NA
NA
NA
1.00
1 .2E-06
1
2.9E-02
2.9E-02
MMSOILS
78.11
94.2
5.7E-03
1690
31
NA
0.0
0.0
0.0
0.0
NA
NA
NA
0.0
0.0
0.0
00
0.0
NA
24.48
NA
0.00
NA
0.00

NA
NA
NA
0002
0.00000107
0.00000000
14.5606
2.12987
NA
NV
NV
NV
NV
NV = no value is listed in electronic database  NA = not applicable
1.   = value in electronic database of MEPAS version 3.x
2.   = value in manual (Chemical Databases; Strenge, Peterson, and Sager, 1989). Note that the
    values in subsequent versions of the manual are continuously updated to conform
    with the values in the electronic database
12.2.5 Parameters for bioaccumulation
       MEPAS lists some different values of bioaccumulation in its electronic and paper
database   For example, benzene's feed-to-meat  coefficient  has a zero value in the
                                          69

-------
electronic database, and 3.36E-06 value in the paper database.  The model user needs
to be aware that MEPAS' electronic database  is constantly being updated,  and the
paper database serves only as an example of representative values.  In addition, the
electronic database of MEPAS contains zero values for those parameters (e.g., finfish
bioaccumulation  parameter)  that are  evaluated at run time  by  MEPAS' exposure
assessment component using correlation relationships.  Correlation analysis is not
used for non-zero values in the database.

1226 Parameters of dose and exposure

      When source terms and estimates of contaminant flux along different exposure
pathways are determined, the next set of data required  includes the more site-specific
parameters that determine exposure and dose.

      In  MMSOILS, the "action levels" for air-borne releases of several contaminants
are incorrect.  For example,  the action  level for 1,1,2 trichloroethane is listed as 0.6
ng/m3, whereas the PEL (Permissible Exposure Level, based on 40 hour work-week) is
45 mg/m3.  The action  level for selenium is listed as 3.5 |ig/m3, whereas the  PEL for
selenium is 0.2 mg/m3.  A default value of 99.9 mg/m3 is assigned to several other
chemicals (e.g.,  mercury, acetone, dioxane), while their PELs range from 0.1-2400
mg/m3.

      In  addition, MEPAS Version 3.0 allows the user to  input  certain uptake and
exposure parameters.  Documentation is provided for the inclusion  of such parameters.
There are procedures for modifying MEPAS  default parameter  files contained in files
that can be edited, rather than the previously hard-wired versions.

123  Developer Updates - Default Parameters

      The  number of contaminants  in the MEPAS database is under expansion to
include new parent radionuclides,  and organic and inorganic chemicals (Droppo, 1994).
Radioactive decay chains are being expanded  also.  Whereas the current   MEPAS
contains  approximately  500 chemicals, the  September  1994  release  contains
approximately 700.
                                      70

-------
                        13. DISCUSSION/CONCLUSIONS
 13.1 Source Term

      Hazardous waste sites may leak contaminants into the environment and pose a
 threat to human health.  At many sites, in addition to the general characteristics of the
 site, the only fact that is known is that hazardous materials are being released   The
 specific cause of the release of contaminants from that site source may be undefined
 In addition, the strength and nature of hazardous releases may vary with time, and the
 impact may shift among the pathways affected by those releases.  These uncertainties
 challenge the multimedia modeler to arrive  at some  reasonable  approximation  for a
 source term that will simulate closely the actual site release of hazardous waste.

      The three models can simulate a variety of source terms; MEPAS has the most
 varied capabilities for source term modeling of the three models reviewed.  None of the
 models  can deal with the presence of free or residually-saturated, non-aqueous liquid,
 within the source term.  Each model bases leachate production on either a steady-state
 value or on leachate solubility or equilibrium-partitioning.

 13.2 Air Transport

      Volatile hazardous waste components leaving  a site through  the atmospheric
 pathway are subject to (he vagaries of weather at the surface of the earth  Air-transport
 modeling assumes commonly that transport through the atmosphere occurs as a  more
 or  less  weakly  organized  plume whose direction,  dimensions, and contaminant
 concentrations are controlled by the speed  and direction of the wind, the  rate and
 quantity of precipitation, and the extent of fallout from that plume.

      All three models use a standard, sector-averaged, Gaussian-plume approach for
 air-transport and model annual-average concentrations and exposures.  MEPAS and
 MMSOILS use the same volatilization algorithm.  Only PRESTO-EPA-CPG models
 radioactive decay   Only  MEPAS  can   model atmospheric calm  conditions  with
 accompanying channeling and complex terrain characteristics.

 13.3 Ground Water Transport

      The movement of ground water is  most  often  very slow, and that movement
occurs over an area and volume that  is often  much larger than the area and volume of
the source.   Once  in the ground water,  contaminants are isolated  and difficult to
remove.  As an essential human  nutrient, almost every human  being is exposed in
some way to contaminants that originate in ground water, some people much more than
others.   All of these factors imply that the ground water pathway can pose  a serious
threat to public health at hazardous waste sites.
                                      71

-------
      Therefore, perhaps the most serious limitation of screening-type models like
MEPAS, MMSOILS and PRESTO-EPA-CPG is the relative simplicity of their ground
water transport components.  While each model  includes a ground water component
that is capable of modeling both unsaturated and saturated transport, that component
is  simulated using classical  assumptions - uniform,  linear flow in a  homogeneous,
layered medium with  equilibrium, and  reversible adsorption of miscible contaminants.
The transport of materials through the ground water system is complex, both in terms of
the effect  of non-uniform flow regimes and because of chemical reactions  between
ground water and matrix.  The  properties of radioactive materials themselves  pose
special problems   Only MEPAS and PRESTO-EPA-CPG  consider  the fate  and
transport   of  radioactive progeny: although  both  assume  the  same  adsorption
characteristics  for progeny as for  parent.   MMSOILS does not model the decay of
adsorbed contaminants.

134 Erosion, Overland Flow, Runoff and Surface  Water Transport

      Multimedia models are very often employed for use at sites where radioactive
contamination  originates at or near the land's surface.  Therefore,  source materials
may be not only subject to atmospheric conditions  or leached into the ground water, but
also possibly removed by surface transport processes.

      As with  the ground water component, MEPAS,  MMSOILS and  PRESTO-EPA-
CPG employ similar,  simplistic models for runoff, erosion, and mixing within surface
water bodies   These simplistic approaches are based, for the most part, on empirical
equations that may have little or no physical  basis. The same limitations for simulating
the transport of  radionuclides that exist for simple  ground water models pertain to
simple surface water models:   accurate flow paths  can  be  very important when
modeling  constituents that  decay,  and  simple equilibrium  partitioning  between
dissolved  contaminants and sediments within surface  water bodies may not be
sufficient in environments where,  for example, resuspension can be important.  None of
these models considers volatilization from surface water bodies into the atmosphere.
MEPAS  and  MMSOILS have  separate  subroutines  for modeling transport within
wetlands and lakes, respectively.

13.5 Food Chain Modeling and Exposure Assessment

      The assessment of human health risk is the primary objective of most transport
and exposure assessment modeling.  The penultimate step before the final  calculation
of  human exposure and risk  is to estimate the concentration of contaminants in food
and drink  to which humans  will be exposed.  MEPAS includes food  chains as an
integral part of its exposure-dose component.  Food chains are considered separately
in  MMSOILS  and in  PRESTO-EPA-CPG  as agricultural data supporting  exposure
estimates  Although MEPAS  can be used to model acute toxic atmospheric releases,
all three models were designed primarily to handle long-term, chronic exposures
                                      72

-------
      Each of the three models employ comparable, standard methods for estimating
exposure to environmental contaminants through the food chain and other pathways
Therefore,  the same limitations that exist for all exposure and risk assessments exist
for these models. For example, food intake is subject to behavioral variations  Both
the quantity and type of foods eaten vary from location to location  In addition, the poor
state of knowledge of the combined effects of radionuclides and toxic chemicals may
have additional implications for the accuracy of exposure calculations at mixed-waste
sites

136 Dosimetry/Risk Assessment

      Human  health risk is a function of the  actual  impact that  an  environmental
contaminant has on an  individual or  group.  The quantity of a toxic substance that
produces an adverse response in an organism is known as a  dose.  Dose can refer to
individual organs or a whole body; it can be either internal or external: and it can be
either acute or chronic.  In addition,  the model may be designed to calculate either
individual or population dose on and/or off a contaminated site.

      Only  MEPAS  and  PRESTO-EPA-CPG  can  calculate  human  dose  from
radionuclides.  MMSOILS is designed to consider only toxic chemicals. Both MEPAS
and PRESTO-EPA-CPG  use methods defined by Eckerman to calculate external dose.
The chronic dose calculated by MEPAS is based on a default lifetime exposure of 70
years, while PRESTO-EPA-CPG uses 50 years for its dose period.  MEPAS calculates
individual dose and  individual  and population risk.  Both  MEPAS and MMSOILS were
designed for the estimation  of off-site exposures and dose, while  PRESTO-EPA-CPG
can calculate both on- and off-site dose by including the on-site population as part of
the local" population.

137 Uncertainty Analysis

      The accuracy of a predicted outcome cannot be better  than the accuracy of the
data input to the model.  In addition,  no matter how accurate the input data  is, if the
algorithms used by the model  do not closely mirror "real-world" processes, the output
generated by the simulation will be of little use.  Not all input data and not every model
algorithm affect the accuracy of the outcome of the simulation equally. Uncertainty and
sensitivity analyses  are procedures that the model user can follow in  an attempt  to
quantify the impact of parameter (input) and/or structural (algorithm) error

      Uncertainty analyses  can  and  should be performed with any model  as long as
the user can vary input  parameters and observe the results of the simulation  The
capabilities  for performing uncertainty  analyses with the three models reviewed in this
report are a function of the number  and diversity of input parameters  In this  way,
PRESTO-EPA-CPG  has fewer environmental input parameters than either MEPAS or
MMSOILS.  MEPAS  and MMSOILS discuss methods  for estimating  more accurate
values for site-specific parameters when site-monitoring data are available.

                                      73

-------
138 Parameter Estimation and Default Values

      The three models use deterministic (single) values for parameters.  It is difficult
in such deterministic assessments to sort out the contributions of individual parameters
to the overall uncertainty of the risk estimates,  because the calculations can combine
high (90th or 95th percentile) parameter values with lower (50th  percentile or average)
values  The accuracy of the value of the parameters in all three models  is difficult to
verify  Although such difficulties are balanced by the user's ability to input alternative
values, no specific instructions are provided for performing uncertainty analyses as a
way of estimating the adequacy and precision of assumptions.

139 Overall Summary

      Table 13.1 presents a summary of the  features contained  within  each of the
three models    For the purpose of  simulating the transport, fate  and  effects of
radioactive contaminants through more than one pathway, both MEPAS and PRESTO-
EPA-CPG are  adequate for screening  studies; MMSOILS only handles  nonradioactive
substances and must be modified before it can be used in these  same applications  Of
the  three models, MEPAS is the most  versatile,  especially if the user needs to model
the  transport fate and effects of hazardous and radioactive contaminants
                                       74

-------
Table 13.1 Summary of model features.

Contaminant Selection
Hazardous chemical waste
Radioactive waste
Air Pathway
Gas/Vapor emissions
Paniculate emissions
Point/Area/Air sources
Volatilization
Plume rise
Plume reflections on
ground/lid
Calm conditions
Complex terrain
Ground roughness
Dry deposition
Wet deposition
Radioactive decay
Chemical decay
Re-suspension
Inhalation
Indoor exposure
Onsite exposure
Short time exposure
Spatial definition
Surface Water Pathway
Overland flow (runoff)
Overland sediment
Suspended solids
Sediment
Volatilization
Spatial definition
Ground Water Pathway
Spatial definition
Time dependence
Soil Pathway
Volatilization
Infiltration
Ground water loss
Degradation
Soil ingestion
Spatial definition
Time dependence
MEPAS

yes
yes

yes
yes
yes/yes/yes
yes
yes

yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes1
yes
yes
2-D

yes
yes
yes
yes
yes
2-D

3-D
yes

yes
yes
yes
yes
yes
1-D
yes
MMSOILS

yes
no

yes
yes
no/yes/no
yes
no

no
no
no
no7
yes
no
no
no
yes
yes
no
yes
no
2-D

yes
yes
no
no
yes
1-D

3-D
yes

yes
yes
yes
yes
no
1-D
yes
PRESTO-EPA-CPG

no
yes

yes
yes
yes/yes/no
no
no

yes
no
no
yes
yes
no
yes
no
yes
yes
no
c.
yes"
yes
2-D

yes
yes
no
no
yes
1-D

1-D
yes

no
yes
yes
yes
yes
no
yes
75

-------
Table 13.1 cont'd. Summary of model features.

Bio-accumulation
Animals
Terrestrial plants
Foliar deposition
Aquatic organisms
Spatial definition
Site Data Required
Contaminant Selection
Hazardous chemical waste
Radioactive waste
Intakes from Ingestion of
Drinking Water
Shower Water
Swimming Water
Leafy Vegetables
Other Produce
Meat
Milk
Finfish
Shellfish
Special Food
Shoreline Sediment
Soil
Intakes from Inhalation
While Showering
Of Air
Of Re-suspended Soil
Intakes from Contact
While Showering
While Swimming
With Shoreline Sediment
With Soil
With Volatiles in Air
External Exposures:
While Swimming
While Boating
From Air
With Soil
With Shoreline Sediment
From Direct Radiation.
MEPAS

yes
yes
yes
yes
2-D
Extensive

yes
yes

yes
yes
yes
yes;'
yes23
yes
yes
yes
yes
yes
yes
yes

yes
yes
yes

yes
yes
yes
yes4
yes

yes
yes
yes
yes
yes
yes
MMSOILS

yes
yes
yes
yes
2-D
Moderate

yes
no

yes
no
yes
yes
yes
yes
yes
yes
no
no
no
yes

no
yes
yes

no
no
no
yes
no

no
no
no
no
no
no
PRESTO-EPA-CPG

yes
yes
yes
no
2-D
Moderate

no
yes

yes
no
no
yes
yes
yes
yes
yes
yes
eggs
yes
yes

no
yes
yes-

no
no
no
no
no

no
no
yes
yes
no
no
1 This component not available in 1993 version.
2. From air deposition on crops.
3. From irrigation of crops.
4. Estimations based on either measured concentrations or on calculated accumulations
in soil after atmospheric deposition.
5. In the version modified for cleanup scenarios.
6. On-site scenario only.
7. MMSOILS only considers ground surface roughness in wind erosion of participates
76

-------
                              14. REFERENCES

Abdul, A.S. and R.W. Gillham, 1984.  Laboratory studies of the effects of the capillary fringe
      on streamflow generation. Water Resources Research, vol. 20. no.4, p. 691-698.
Bemier, P.Y., 1985.  Variable source areas and storm generation, an update of the concept
      and a simulation effort. Journal of Hydrology, vol. 79, p. 195-213.
Browman, M.G.,  M.R.  Patterson,   and T.J.  Sworski,  1982.    Formulation of the
      physicochemical processes in the ORNL Unified Transport Model of Toxicants.
      ORNL/TM-8013, Oak Ridge National Laboratory, Oak Ridge, TN
Burkholder, H. C. and Rosinger, E.  L., 1980. A model for the transport of radionuclides
      and their decay products through geological media.  Nuclear  Technology, vol.
      49, pp. 150-158.
Burley, G., 1990a,   Transuranium Elements,  Volume 1:   Elements Of  Radiation
      Protection.  EPA-520/1-90-015, USEPA, Washington, DC.
Burley, G., 1990b,  Transuranium Elements. Volume 2.  Technical Basis For Remedial
      Actions. EPA-520/1-90-016,  USEPA, Washington, DC.
Burmaster, D.E., and R.H. Harris, 1993.  Perspective:  the magnitude of compounding
      conservatisms in Superfund risk assessments.  Risk Analysis,  vol. 13, p.  131-
      134.
Campbell, J.E., D.E. Longsine,  and R.M.  Cranwell, 1981.   Risk  Methodology  for
      Geologic Disposal  of Radioactive Waste:  The NWFT/DVM  Computer  Code
      User's Manual.  NUREG/CR-2081, Sandia National Laboratory, Albuquerque,
      NM.
Case, M.J., S.J. Maheras,  M.D. Otis, and R.B. Baca, 1989.  A Review and Selection of
      Computer Codes for Establishment of the Performance  Assessment  Center.
      DOE/LLW-83,  National  Low-Level Radioactive Waste Management Program,
      Idaho Fall, ID.
Cheng Y.H., 1989.  User's Guide for the SYSCPG Program - A PC Version of the
      Presto-EPA-CPG Operation System.  EPA 520/1-89-017, Washington, DC.

Cheng Y.H., 1992.  User's Guide for the SYSPOP Program - A PC Version of the
      Presto-EPA-POP Operation System.  EPA 400R 92003, Washington, DC.

Childs,  E.G., 1960.  The non-steady  state of the water  table in drained land.  Journal of
      Geophysical Research, vol. 65, no.  2, p. 780-782.
Clapp, R.B., and G. M. Hornberger,  1978. Empirical equations for some soil hydraulic
      properties. Water Resources Research, 14(4):601-604.
Cohen, Y., 1986. Inter-media transport modeling in multimedia systems.  Pollutants in
      a Multimedia Environment. Y. Cohen, ed.,  Plenum, NY.
                                     77

-------
Doctor, P.O., T.M. Miley and C.E. Cowan, 1990.  Multimedia Environmental Pollutant
      Assessment  System  (MEPAS)  Sensitivity  Analysis  of  Computer  Codes
      Prepared for the U.S. Department of Energy.  PNL-7296. Pacific Northwest
      Laboratory, Richland, WA.

DOE,  1988a.  Internal Dose Conversion Factors for Calculation of Dose to the Public
      DOE/EH-0071, U.S. Department of Energy, Washington. DC
DOE,  1988b. External Dose-Rate Conversion Factors For Calculation Of Dose To The
      Public.  DOE/EH-0070, U.S. Department of Energy. Washington. DC.
Droppo, J.G., Jr., D.L. Strenge, J.W. Buck, B.L. Hoopes, R.D Brockhaus. MB. Walter
      and.  G.  Whelan,  1989a.   Multimedia  Environmental Pollutant  Assessment
      System (MEPAS) Application Guidance Volume  1 -  User's Guide  PNL-7216.
      Pacific Northwest Laboratory, Richland, WA.
Droppo, J.G., Jr., D.L. Strenge, J.W. Buck, B.L. Hoopes, R.D. Brockhaus, MB Walter
      and G.  Whelan, 1989b.   Multimedia  Environmental Pollutant  Assessment
      System (MEPAS) Application Guidance Volume 2 - Guidelines for Evaluating
      MEPAS Parameters.  PNL-7216 Pacific Northwest Laboratory  Richland, WA
Droppo, J.G., Jr., G. Whelan, J.W. Buck, D.L. Strenge, B L  Hoopes. and M.B. Walter,
      1989c.     Supplemental  Mathematical  Formulations:     The   Multimedia
      Environmental Pollutant Assessment System  (MEPAS)   PNL-7201    Pacific
      Northwest Laboratory, Richland. WA
Droppo, J.G., Jr., J W. Buck, D.L. Strenge, and B.L. Hoopes,  1993  Risk computation
      for environmental restoration activities.  Journal of Hazardous Material, vol. 35.
      p. 341-352.
Dugan,  T.A., G.L. Gels, J.S.  Oberjohn, and  L.K  Rogers.  1990.   Feed  Materials
      Production Center Annual  Environmental Report for Calendar  Year  1989.
      Environmental Management Department, Westinghouse Materials  Company of
      Ohio, FMPC-2200 Special UC-707.
Dunning, D. E., Jr.,  R. W.  Leggett,  and M.G. Yalcintas.  1980.   A  Combined
      Methodology for Estimating Dose Rate and Health  Effects from Exposures to
      Radioactive Pollutants.   ORNL/TM-7105, Oak  Ridge National  Laboratory. Oak
      Ridge, TN.
Fields,  D.E.,   C.A.  Little,  F.  Parraga,  V.  Rogers,  and  C  Y   Hung.  1987a
      PRESTO-EPA-POP: A Low-Level Radioactive  Waste Environmental Transport
      and  Risk Assessment  Code,   Volume  1.  Methodology  Manual.    EPA
      521/1-87-024-1, Washington, DC.
Fields,  D.E.,   C.A.  Little,  F.  Parraga,  V.  Rogers,  and  C  Y   Hung,  1987b
      PRESTO-EPA-POP:  A Low-Level Radioactive Waste Environmental Transport
      and Risk Assessment Code. Volume 2, User's Manual  USEPA 521/1-87-024-2.
      Washington, DC.
                                    78

-------
Fields.  D.E.,  C.J.  Emerson,  R.O.  Chester,  C A.  Little  and  H.  Hiromoto,  1986
      PRESTO-II: A Low-Level Waste Environmental Transport and Risk Assessment
      Code.  ORNL-5970, Oak Ridge National Laboratory, Oak Ridge, TN.

Foster, G.R., 1976.  "Sediments,  General: Reporter's Comments." In:  Proceedings of
      the National Symposium on Urban Hydrology, Hydraulics, and Sediment Control.
      pp. 129-139, University Kentucky, Lexington, Kentucky, July 26-29, 1976.

Gburek, W.J., 1977. The Mahantango Creek Watershed - General hydrology and research
      results.  Watershed Research in  Eastern  North America.  Smithsonian  Institution,
      vol. 1, pp. 227-250.
General Sciences Corporation, 1982.   Graphical Exposure Modeling System (GEMS)
      User's Guide.  Office of Pesticides and Toxic Substances.  USEPA, Washington,
      DC
General Sciences Corporation, 1992. RISKPRO.  Laurel, MD.
Gilbert, T.L.,  C. Yu, Y.C. Yuan, A.J. Zielen, M.J. Jusko,  and A.  Wallo III,  1988.  A
      Manual for Implementing Residual Radioactive Material Guidelines.  ANL/ES-
      160, DOE/CH/8901, Argonne National Laboratory, Argonne,  IL.
Goldman,  S.J., K. Jackson, and T.A. Bursztynsky, 1986. Erosion and Sediment Control
      Handbook.  McGraw-Hill, NY.
Helton, J.C., and  Kaestner,  P.C., 1981.  Risk  Methodology for Geologic Disposal of
      Radioactive Waste; Model Description and User's Manual, the Pathways Model
      NUREG/CR-1636, vol.  1, SAND 78-1711 AN, Albuquerque. NM.
Hewlett, J.D., 1961. Watershed Management.  Southwestern Forest Experimental Station,
      Annual Report. U.S. For. Ser., pp. 61-66.
Hewlett, J.D., and A.R. Hibbert, 1967.  Factors affecting the response of small watersheds
      to precipitation in humid areas, in: W.E. Sopper and H. W. Lull (eds.), Int. Svmp  on
      Forest Hydrology. Pergamon, Oxford, pp. 275-290.
Horton, R.E., 1933. The role of infiltration in the hydrologic cycle.  Transactions  of the
      American Geophysical Union. 14:446-460
Hung, C.Y., 1983. A model to simulate infiltration of rainwater through the cover of a
      radioactive waste trench under saturated and unsaturated conditions.  Role of
      the Unsaturated Zone in Radioactive and Hazardous Waste Disposal,  edited by
      J.W. Mercer, F.S.C. Rao, and I Wendell Marine, Ann Arbor Science.
Hung, C.Y., 1986.  An optimum groundwater transport model for application to the
      assessment  of  health  effects  due to  land  disposal of  radioactive  waste.
      Proceedings of Nuclear and Chemical Waste Management, vol. 6, pp.41-50.
Hung, C.Y., 1989.  User's Guide for the SYSCPG Program  - A PC Version of the
      Presto-EPA-CPG Operation System  USEPA 520/1-89-017, Washington, DC
                                      79

-------
Hung. C.Y..  1992  User's Guide for the SYSPOP Program - A  PC Version of the
      Presto-EPA-POP Operation System.  USEPA 400R 92003, Washington, DC
IAEA. 1989    Evaluating the Reliability of Predictions Made  Using  Environmental
      Transfer Models.  International Atomic Energy Agency, Safety Series No  100,
      Vienna, Austria.
ICRP 1975    Publication 23.  The  Reference  Man:  Anatomical  Physiological and
      Metabolic Characteristics.  Pergamon Press. Oxford.
ICRP, 1977  Publication 26.  Recommendations of ICRP   Annals  of the International
      Commission on Radiation Protection, vol 1, no  3   (Superseded by  ICRP
      Publication 60 )
ICRP, 1979  Publication 29  Radionuclide Release into the Environment - Assessment
      of  Doses  to  Man.   Annals of the International Commission on  Radiation
      Protection, vol. 2, no. 2.
ICRP. 1979  Publication 30.  Limits For The Intake Of Radionuclides By Workers. Part
      1_  Annals of the International  Commission on Radiation Protection, vol. 2. no.
      3/4
ICRP 1985  Publication 43.  Principles Of Monitoring For Their Radiation  Protection
      Of The  Population.   Annals  of the International  Commission on  Radiation
      Protection, vol  15, no.  1
ICRP. 1986.  Publication 46. The Metabolism Of  Plutonium And  Related  Elements.
      Annals of the International Commission on  Radiation Protection, vol. 16, no  2/3.
ICRP, 1987   Publication  51, Data for Use  in Protection Against  External  Radiation.
      International Commission  on Radiation  Units and Measurements, vol.  17. no.
      2/3.
ICRP. 1988.  Publication 54. Individual Monitoring for Intakes of Radionuclides by
      Workers: Design & Interpretation. Annals of the International Commission on
      Radiation Protection, vol 19, nos. 1-3.
ICRP, 1989   Publication  56. Age-Dependent  Doses to Members  of the Public  from
      Intake of Radionuclides:  Part  1.  Annals of  the International Commission on
      Radiation Protection, vol. 20, no. 2.
ICRP, 1990.   Publication 60.   1990  Recommendations of ICRP   Annals of  the
      International Commission on Radiation Protection, vol. 21, no. 1.
Kennedy. WE., Jr., and D.L Strenge, 1992  Residual Radioactive Contamination from
      Decommissioning.  NUREG/CR-5512, Division of Regulatory Applications.  U.S.
      Nuclear Regulatory Commission, Washington, DC
Kennedy, WE. Jr., R A. Peloquin, B.A Napier, and S.M Neuder, 1986  Intruder Dose
      Pathway  Analysis  for the  Onsite  Disposal of  Radioactive Wastes:    The
      ONSITE/MAXI1  Computer Program   NUREG/CR-3620. Supplement 1, 1986,
      Washington, DC.
                                     80

-------
Kennedy, W.E., Jr., R.A. Peloquin, B.A Napier and S.M. Neuder, 1987.  Intruder Dose
      Pathway Analysis  for the Onsite Disposal  of Radioactive  Wastes:   The
      ONSITE/MAXI1  Computer Program. NUREG/CR-3620, Supplement  2,  1987,
      Washington, DC

King, C.M.,  E.L. Wilhite,  R.W. Root, Jr., D.J. Fauth,  K.R. Routt, R H. Emslie, R.R.
      Beckmeyer, R.A Fjeld, G A.  Hutto, and J.A. Vandeven, 1985.  The Savannah
      River Laboratory DOSTOMAN code  - A compartmental  pathways model  of
      contaminant transport CONF-8509121-13, Proceedings of the DOE Low-Level
      Waste Management Program Seventh Annual  Participants Information Meeting,
      Las Vegas, NV.
Makhlouf,  E.M., and Kavanaugh, MC., 1993.   Department  of Energy  PEIS Risk
      Assessment Model  Peer Review.  Prepared by Montgomery Watson, Walnut
      Creek, CA for Oak Ridge National Laboratory, Risk Analysis  Section, Oak
      Ridge, TN.
McKone, T.E. and D.W. Layton, 1986. Exposure and Risk Assessment of Toxic Waste
      in a Multimedia Context.  UCRL-9353, Lawrence  Livermore National  Laboratory,
      Livermore, CA
McKone, T.E.,  1990.  Dermal uptake of organic chemicals from  a  soil matrix.  Risk
      Analysis, vol. 10: 407-419.
McKone, T.E., W.E. Kastenberg, and D.  Okent, 1983. The use of landscape chemical
      cycles for indexing  the health risks of toxic elements and radionuclides.  Risk
      Analysis, vol. 3, 189-205.
Merrel, G.B. , 1994   Modifications To The PRESTO-EPA-CPG Code To Facilitate The
      Analysis Of Soil Contamination Sites.   RAE-9231/6-1, U.S.  EPA, Washington,
      DC.
Mills,  M.  and  D. Vogt, 1983.   A Summary of Computer  Codes for Radiological
      Assessment.     NUREG/CR-3209,  U.S.   Nuclear  Regulatory  Commission,
      Washington, DC.
Mitchell, J.K., and G.D. Bubenzer, 1980  Soil Loss Estimation.  Soil Erosion, eds. M.J.
      Kirby and R.P.C Morgan. Wiley, NY.
Moment, M.H.,  Y. Yuan and A.J. Zielen. 1979.   Uranium Dispersion and Dosimetrv
      (UDAD) Code.  NUREG/CR-0553.  Argonne National Laboratory, IL.
Morgan, M.G., and  M. Henrion,  1990.  Uncertainty.  A  Guide to  Dealing with
      Uncertainty in Quantitative Risk  and  Policy  Analysis.   Cambridge University
      Press, NY
Moskowitz,  P.O., R.  Pardi, M.P. DePhillips,  and A.F. Meinhold, 1993.  Computer
      Models  Used  to   Support  Cleanup  Decision-Making  at  Hazardous  and
      Radioactive  Waste  Sites.  EPA  402-R-93-005,  Office  of Air  and Radiation,
      USEPA, Washington, DC.
                                     81

-------
Moskowitz, P.O., R Pardi, M.P. DePhillips, andA.F. Meinhold, 1992, Computer models
      used to support cleanup decision making at  hazardous and radioactive waste
      sites.  Risk Analysis, vol. 12. no. 4. p. 591-621.
Napier, B.A.,  and  G.F Piepel, 1988.  A Manual for Applying the Allowable Residual
      Contamination Level Method For  Decommissioning Facilities On The Hanford
      Site. PNL-6348/UC602, Pacific Northwest Laboratory, Richland, WA..
Napier, B.A., R.A.  Peloquin, and D.L Strenge, 1986.  DITTY- A Computer Program for
      Calculating  Population Dose Integrated Over Ten Thousand  Years. PNL-4456,
      Pacific Northwest Laboratories, Richland, WA.
Napier,  B.A.,  R A.  Peloquin, D.L.  Strenge and  J.V.  Ramsdell,  1988.    Hanford
      Environmental Dosimetry Upgrade Project.  GENII - The Hanford Environmental
      Radiation Dosimetrv System  3 Volumes.    PNL-6584  Pacific  Northwest
      Laboratories. Richland, WA

Napier, B A , R A  Peloquin, W.E. Kennedy, Jr., and S.M. Neuder, 1984. Intruder Dose
      Pathway  Analysis  for the  Onsite  Disposal  of  Radioactive  Wastes:   The
      ONSITE-/MAXI1 Computer Program.  NUREG/CR-3620 (PNL-4054).
NCRP.  1987   Recommendation on  Limits for Exposure to Ionizing  Radiation:
      Recommendations  of  the  National  Council  on  Radiation  Protection  and
      Measurement. Publication 91, NCRP, Bethesda, MD.
NCRP, 1993.  Limitations of Exposure to Ionizing Radiation:  Recommendations of the
      National  Council on Radiation Protection and Measurement.  Publication  116,
      NCRP  Bethesda, MD
Novotny, V., and G.  Chesters, 1981.  Handbook of Non-point Pollution. Van Nostrand
      Reinhold, NY
O'Brien, A.L.,  1982. Rapid water table rise.  Water Resources Bulletin, vol.  18, no  4,  p.
      713-715.
O'Lougnlin, EM, 1981.  Saturation regions in catchments and their relations to soil and
      topographic properties. Journal of Hydrology, vol 55, 229-246.
Office of Emergency and Remedial Response, 1991.  Risk Assessment Guidance for
      Superfund:  Volume 1 - Human Health  Evaluation Manual  (Part B. Development
      of Risk-based Preliminary Remediation Goals).  Publication  9285.7-01 B,  U.S.
      Environmental Protection Agency, Washington, DC
Onstad, C A , and G.R.  Foster,  1975 "Erosion Modeling over Watersheds." Trans.
      ASAE (Am  Soc. Agric. Eng.). vol 18, no. 2, pp. 288-292.
Onstad, C A.,  R.F.  Priest and K.E.  Saxton, 1976.    Predicting  Sediment Yields.
      National  Symposium on Soil Erosion and Sedimentation  by Water   American
      Society of Agricultural Engineers Proceedings, St. Joseph, Michigan,  December
      1977.
                                     82

-------
Oztunali, O.I., and  G W.  Roles,  1986.   Update  of  Part  61  Impacts Analysis
      Methodology. Methodology Report. Volume 1. NUREG/CR-4370-Vol 1
Oztunali, O.I., W.D. Ron, R. Eng and G.W. Roles,  1986.  Update of Part 61 Impacts
      Analysis  Methodology.     Codes  and  Example  Problems.  Volume  2.
      NUREG/CR-4370-Vol.2

Peer Review Committee  Report,  1994. The Report of the Peer Review Committee for
      Multimedia Models for Use  in The DOE Programmatic Environmental  Impact
      Statement  Risk  Assessment.  Prepared for DOE,  Office  of  Environmental
      Restoration and Waste Management, Washington, DC.

Radiological Assessments Corporation, 1983. PRESTO-EPA: A Low-Level Radioactive
      Waste Environmental Transport and Risk Assessment Code - Methodology and
      User's Manual.  Neeses, SC.
Rish, W.R., 1988.  Approach to  Uncertainty in Risk Analysis. ORNL/TM-10746, Oak
      Ridge National Laboratory, Oak Ridge, TN.
Rogers,  V and C. Hung, 1987a.  PATHRAE-EPA: A Low-Level Radioactive  Waste
      Environmental Transport and Risk Assessment  Code, Methodology and  Users'
      Manual. EPA 520/1-87-028, Office of Radiation  Programs, Washington, DC

Rogers, V., and C.Y. Hung, 1987b.  Low-Level and NARM Radioactive Wastes. Model
      Documentation.  PRESTO-EPA-BRC.  EPA 520/1-87-027, Office of Radiation
      Programs, USEPA, Washington, DC.
Rogers,  V , and C.Y. Hung, 1987c.  PRESTO-EPA-DEEP A Low-Level Radioactive
      Waste Environmental Transport and Risk Assessment Code - Methodology and
      User's Manual.    EPA 520/1-87-025,  Office of Radiation Programs, USEPA,
      Washington, DC.
Rogers.  V. and C.Y. Hung,  1987d.  PRESTO-EPA-CPG:  A Low-Level Radioactive
      Waste Environmental Transport and Risk Assessment Code. Methodology and
      User's Manual.    EPA 520/1-87-026,  Office of Radiation Programs. USEPA,
      Washington, DC.
Root, R.W., 1981  Documentation and  User's Guide  for DOSTOMAN - A Pathways
      Computer Model of Radionuclide  Movement.  DPST-81-549,  E.I  DuPont de
      Nemours & Co.
Root, R.W., 1991   GWSCREEN:  A Semi-Analytical Model for Assessment  of the
      Ground Water Pathway from Surface  or  Buried Contamination: Theory and
      User's Guide. EEG-GEO-10158, Idaho National Engineering Laboratory,  Idaho
      Falls, ID.
Ross,  B., C.M. Koplik, M.S. Giuffre, S.P. Hodgin,  J.J. Duffyu, and J Y Nalbandian,
      1980.  User's Guide to NUTRAN:  A Computer  Analysis System for Long-Term
      Repository Safety. Technical Report AECL-TR-121, Atomic Energy of Canada
      Ltd.

                                    83

-------
Ruttenber, A.J., 1993a. Assessing environmental health risks. Parti  Chemtech. vol.
      23, no. 4, p. 53-59.
Ruttenber, A. J., 1993b. Assessing environmental health risks. Part 2  Chemtech. vol.
      23, no. 5, p. 52-58.
Ryan, P., 1993.  Review of mathematical models for health risk assessment: IV. Inter-media
      chemical transport.  Environmental Software, vol  8, p 157-172
Salhotra, A.M., P.  Mineart, S. Sharp-Hansen and T. Allison, 1990  Multimedia exposure
      assessment model (Multimed) for evaluating land disposal of wastes: Vol 1.; Model
      theory. Prepared for the U.S. Environmental Protection Agency, Athens, GA.
Schnabel, R.R. and W.J. Gburek, 1985. Comments on "Calibration on NPS model loading
      factors".  Trans. ASCE. Journal of Environmental Engineering Division, vol. 111, no.
      1,p. 103-107.
Seigneur,  C.,  1993.   Multimedia  Modeling.   ENSR Consulting  and Engineering,
      Alameda, CA
Seigneur, C., A. Venkatram, D. Galya, P. Anderson, D. Liu, D. Fohart,  R Von Burg, Y.
      Cohen, T  Permutt, and  L. Levin,  1992. Review of mathematical  models for
      health risk assessment, I, overview.  Environmental Software, vol. 7, p. 3-7.
Stenner, R.D.,  R.A.  Peloquin and K.A.  Hawley.  1986.  Modified Hazard  Ranking
      System/Hazard  Ranking  System   for  Sites   With  Mixed  Radioactive  and
      Hazardous Wastes - Software Documentation.  PNL-6066   Pacific Northwest
      Laboratory, Richland, WA.
Strenge, D.L. and T J. Bander, 1981. MILDOS - A Computer Program For Calculating
      Environmental   Radiation  Doses   From  Uranium   Recovery  Operations.
      NUREG/CR-2011 PNL-3767.  Pacific Northwest Laboratory, Richland, WA
Strenge, D.L.,  S.R.  Peterson, and  S. Sager,  1989, Chemical Databases For The
      Multimedia Environmental Pollutant Assessment System (MEPAS): Version 1.
      PNL-7145,  Pacific Northwest Laboratories, Richland, WA.
USEPA, 1983.    PRESTQ-EPA:  A Low-Level  Radioactive  Waste  Environmental
      Transport and Risk  Assessment  Code  - Methodology and  User's  Manual.
      USEPA, Washington, DC.
USEPA, 1984, Summary Report on The Peer Review of PRESTO-EPA (Release 2.4).
      February 7 & 8, 1984, Arlie House, Virginia, USEPA, Washington, DC.
USEPA., 1987   PRESTO-EPA-CPG: A Low-Level Radioactive Waste Environmental
      Transport and Risk  Assessment Code.  Documentation and  User's  Manual.
      USEPA-520/1-87-026, Washington, DC.
USEPA, 1988a.  Limiting Values of Radionuclide Intake and Air Concentration and
      Dose Conversion Factors for  Inhalation.  Submersion, and Ingestion.  Federal
      Guidance Report No  11  EPA-520/1-88, Washington, DC
                                     84

-------
 USEPA, 1988b.  Hazard Ranking System (MRS) for Uncontrolled Hazardous Substance
      Releases. Proposed Rule. 53 Federal Register 51962.
 USEPA,  1989a.  OSWER  Models Study: Promoting Appropriate  Use of Models in
      Hazardous Waste/Superfund Programs.  Information Management Staff, Office
      of Program Management and Technology, Office of Solid Waste and Emergency
      Response, USEPA, Washington, DC.
 USEPA, 1989b.  Risk Assessment Guidance for Superfund. Volume I. Human Health
      Evaluation Manual  (Part A).  Interim Final.    EPA/540/1-89/002.   Office  of
      Emergency and Remedial Responses, USEPA, Washington, DC
 USEPA,  1990a.    Report  on the  Usage  of  Computer  Models  in  Hazardous
      Waste/Superfund Programs: Phase II  Final Report.  Information Management
      Staff, Office of Solid Waste and Emergency Response, Washington. DC
 USEPA, 1990b.  Hazard Ranking System (HRS) for Uncontrolled Hazardous Substance
      Releases. Proposed Rule. 55 Federal Register 51532, December 14. 1990
 USEPA, 1991.  Risk Assessment Guidance for Superfund: Volume  1 - Human Health
      Evaluation Manual (Part B, Development of Risk-based Preliminary Remediation
      Goals). Publication 9285.7-01 B, Office of Emergency and Remedial Response,
      Washington, DC
 USEPA, 1992.   MMSOILS: Multimedia Contaminant Fate. Transport, and Exposure
      Model. Office of Research and Development, Washington, DC.
 USEPA, 1993a.  External Exposure To Radionuclides In Air, Water,  and Soil   Federal
      Guidance Report No.  12. EPA-420-R-93-081  (Oak  Ridge National  Laboratory.
      Oak Ridge, TN; USEPA, Washington, DC.
 USEPA, 1993b.  Integrated Model Evaluation System (IMES).  Prepared by Versar Inc.
      for the United States Environmental Protection Agency, Washington, DC
 USEPA, 1994a.   EPA Exposure Models Library and IMES On CD-ROM Office  of
      Health and Environmental Assessment, United States Environmental Protection
      Agency, Washington, DC.
 USEPA, 1994b.  Estimating Radiogenic Cancer  Risks.  EPA  Report,  402 R-93-076.
      June 1994.
 USNRC,  1972.   Federal Register:  Part IV: Nuclear Regulatory Commission:  10CFR
      Part 20 et al., Standards for Protection Against Radiation: Final Rule.  Nuclear
      Regulatory Commission, Washington, DC.
Vandergrift, S.B.  and RB  Ambrose Jr, 1988.   SARAH2. A  Near  Field Exposure
      Assessment Model for Surface Waters.  EPA/600/3-88/020.
 Ward, R.C., 1984. On the response  to precipitation of headwater streams  in humid areas
      Journal of Hydrology, vol. 74, p. 171-189.
Whelan, G., J.W. Buck, and A  Nazarali. 1994 "Modular Risk Analysis for Assessing
      Multiple  Waste  Sites"  PNL-SA-24239,  In:  Proceedings of  the  US  DOE

                                    85

-------
      Integrated  Planning  Workshop   U.S. Department of  Energy,  Idaho National
      Engineering Laboratory, Idaho Falls, ID.
Whelan, G., J W. Buck, D.L. Strenge, J.G., Droppo Jr., and B.L. Hoopes, 1986. Overview
      of the multimedia environmental pollutant assessment system (MEPAS).  Hazardous
      Waste and Hazardous Materials, vol.  9, no. 2, p. 191-208.
Whelan  G, J G  Droppo Jr,  D.L. Strenge,  MB.  Walter, and  J.W.  Buck, 1989.  A
      Demonstration  of the  Applicability  of Implementing the Enhanced Remedial
      Action Priority System  (RAPS) for Environmental Releases. PNL-7102. Pacific
      Northwest Laboratory. Richland.  WA
Whelan. G , D.L.  Strenge, J.G. Droppo Jr, and B.L. Steelman,  1987  The Remedial
      Action Priority System  (RAPS):  Mathematical Formulations.  PNL-6200. Pacific
      Northwest Laboratory. Richland,  WA.

Whelan, G., 1993.  Multimedia Environmental Assessment System (MEPAS).  Pacific
      Northwest Laboratory draft prepared for USDOE, Pacific Northwest Laboratory,
      Richland, WA.

Wischmeier. W.H , and Smith, D.D., 1978.  Predicting Rainfall Erosion Losses: A Guide
      to  Conservation Planning.   Agricultural Handbook No.  537, United States
      Department of Agriculture, Washington, DC

Woolhiser, DA, 1975.  The watershed approach to understand our environment.  Journal
      of Environmental Quality, vol. 4, p. 17-20.
Yeh, G.T., 1981.  AT123D:  Analytical  Transient One-. Two-,  and Three-Dimensional
      Simulation of Waste  Transport in the Aquifer System.   ORNL-5602,  Oak Ridge
      National Laboratory, Oak Ridge, TN.
Yu. C.,  Zeilen, A.J., Y.C.  Cheng,  et  al.,  1993, Manual  For Implementing Residual
      Radioactive Material Guidelines Using Resrad Version 5.0  Argonne National
      Laboratory, IL.
Yuan, Y.C..  J.H.C. Wang, and A Zielen, 1989  MILDOS-AREA:  An Enhanced Version
      of MILDOS or Large Area Sources.  ANL/ES-161. Argonne National Laboratory,
      IL
Zaltsberg,  E., 1986.  Comment on "Laboratory studies on the effects of the capillary fringe
      on  stream flow generation", by A.S.  Abdul and R.W. Gillham. Water Resources
      Research, vol. 22, no  5. p. 837-838.
Zaslavsky,  D.  and A.S. Rogowski,  1969.  Hydrologic and morphologic implications of
      anisotropy and  infiltration in  soil  profile development.  Proceedings  of the  Soil
      Science Society America, vol  33.  no. 4. p. 594-599.
Zaslavsky,  D. and  G.  Sinai,  1981   Surface hydrology:  Ill-Causes of lateral  flow Proc.
      ASCE. Journal Hydraulics Division, vol. 107, no. 1, p. 37-52.
                                      86

-------