& EPA
United States
Environmental Protection
Agency
Office Of Air Quality
Planning And Standards
Research Triangle Park, NC 27711
EPA-453/D-99-001
November 1999
Air
TRIM
Total Risk Integrated Methodology
TRIM.Expo
TECHNICAL SUPPORT DOCUMENT
EXTERNAL REVIEW DRAFT
Environmental Fate,
Transport, & Ecological
Exposure Module
(TRIM.FaTE>
Risk Characterization
Module
(TRlM.Risk)
Exposure-Event Module
1 (TRIMixpo)
-------
EPA-453/D-99-001
TRIM
Total Risk Integrated Methodology
TRIM.Expo TECHNICAL SUPPORT DOCUMENT
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 WV;t Jackson Bpulevard, 12th Floor
Chicago, IL 60604-3590
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
External Review Draft
November 1999
-------
[This page intentionally left blank.]
-------
Disclaimer
This document is an external review draft. It has not been formally released by the U.S.
Environmental Protection Agency and should not at this stage be construed to represent Agency
policy. It is being circulated for comments on its technical merit and policy implications, and
does not constitute Agency policy. Mention of trade names or commercial products is not
intended to constitute endorsement or recommendation for use.
NOVEMBER 1999 i TRIM.EXPO TSD (DRAFT)
-------
1
J
[This page intentionally left blank]
-------
ACKNOWLEDGMENTS
Acknowledgments
As described in this report, the Office of Air Quality Planning and Standards (OAQPS) of
the U.S. Environmental Protection Agency is developing the Total Risk Integrated Methodology.
The principal individuals and organizations in the TRIM.Expo development effort and in the
preparation of this report are listed below. Additionally, valuable technical support for report
development was provided by ICF Consulting.
Robert G. Hetes, EPA, Office of Air Quality Planning and Standards
Deirdre L. Murphy, EPA, Office of Air Quality Planning and Standards
Ted Palma, EPA, Office of Air Quality Planning and Standards
Harvey M. Richmond, EPA, Office of Air Quality Planning and Standards
Amy B. Vasu, EPA, Office of Air Quality Planning and Standards
Thomas E. McKone, Lawrence Berkeley National Laboratory & University of California, Berkeley
Michael P. Zelenka, ICF Consulting
NOVEMBER 1999 iii TRIM.ExpoTSD (DRAFT)
-------
ACKNOWLEDGMENTS
The following EPA individuals reviewed a previous draft of this document.
EPA Models 2000 TRIM Review Team
Robert F. Carousel
National Exposure Research Laboratory
Office of Research and Development
*S. Steven Chang
Office of Emergency and Remedial
Response
Office of Solid Waste and Emergency
Response
Ellen Cooler
National Exposure Research Laboratory
Office of Research and Development
Stan Durkee
Office of Science Policy
Office of Research and Development
Harvey Holm
National Exposure Research Laboratory
Office of Research and Development
John S. Irwin
Office of Air Quality Planning and
Standards
Office of Air and Radiation
Linda Kirkland
National Center for Environmental Research
and Quality Assurance
Office of Research and Development
Matthew Lorber
National Center for Environmental
Assessment
Office of Research and Development
Haluk Ozkaynak
National Exposure Research Laboratory
Office of Research and Development
William Petersen
National Exposure Research Laboratory
Office of Research and Development
Ted W. Simon
Region 4
Amina Wilkins
National Center for Environmental
Assessment
Office of Research and Development
* Team Leader
Review by Other Program Offices
Pam Brodowicz, Office of Air and Radiation, Office of Mobile Sources
William R. Effland, Office of Pesticide Programs
John Girman, Office of Air and Radiation, Office of Radiation and Indoor Air
Steven M. Hassur, Office of Pollution Prevention and Toxics
Terry J. Keating, Office of Air and Radiation, Office of Policy Analysis and Review
Russell Kinerson, Office of Water
Stephen Kroner, Office of Solid Waste
David J. Miller, Office of Pesticide Programs
NOVEMBER 1999
IV
TRJM.EXPO TSD (DRAFT)
-------
PREFACE
PREFACE
This draft document, the TRIM. Expo Technical Support Document, is part of a series of
documentation for the overall Total Risk Integrated Methodology (TRIM) modeling system. The
detailed documentation of TRIM's logic, assumptions, algorithms, equations, and input
parameters is provided in comprehensive Technical Support Documents (TSDs) for each of the
TRIM modules. The purpose of the TSDs is to provide full documentation of how TRIM works
and of the rationale for key development decisions that were made. This report documents the
Exposure-Event module of TRIM (TRIM. Expo).
To date, EPA has issued draft TSDs for the Environmental Fate, Transport, and
Ecological Exposure module (TRIM.FaTE TSD, U.S. EPA 1999a,b) and the TRIM. Expo (this
report). When the Risk Characterization module (TRIM.Risk) is developed, EPA plans to issue a
TSD for it. The TSDs will be updated as needed to reflect future changes to the TRIM modules.
The EPA has also issued the 1999 Total Risk Integrated Methodology (TRIM) Status
Report (U.S. EPA 1999c). The purpose of that report is to provide a summary of the status of
TRIM and all of its major components, with particular focus on the progress in TRIM
development since the 1998 TRIM Status Report (U.S. EPA 1998a). The EPA plans to issue
status reports on an annual basis while TRIM is under development.
In addition to status reports and TSDs, EPA intends to develop detailed user guidance for
the TRIM computer system. The purpose of such guidance will be to define appropriate (and
inappropriate) uses of TRIM and to assist users in applying TRIM to assess exposures and risks
in a variety of air quality situations.
Comments and suggestions are welcomed. The OAQPS TRIM team members, with their
individual roles and addresses, are provided below.
TRIM Coordination Deirdre L. Murphy
REAG/ESD/OAQPS
MD-13
RTP,NC27711
[murphy.deirdre@epa.gov]
TRIM.FaTE Amy B. Vasu
REAG/ESD/OAQPS
MD-13
RTP,NC27711
[vasu.amy@epa.gov]
NOVEMBER 1999 v TRIM.Expo TSD (DRAFT)
-------
PREFACE
TRIM.Expo Ted Palma Harvey M. Richmond
REAG/ESD/OAQPS HEEG/AQSSD/OAQPS
MD-13 MD-15
RTF, NC 27711 RTF, NC 27711
[palma.ted@epa.gov] [richmond.harvey@epa.gov]
TRIM.Risk Robert G. Hetes
REAG/ESD/OAQPS
MD-13
RTF, NC 27711
[hetes.bob@epa.gov]
NOVEMBER 1999 vi TRIM.Expo TSD (DRAFT)
-------
ACRONYMS
ACRONYMS
ACH Air Exchange Rates
ADD Average Daily Dose
AirPEX Air Pollution Exposure Model
AIRS Aerometric Information Retrieval System
AMEM ADL Migration Exposure Model
APEX Air Pollutant Exposure Model
ARB Air Resources Board
ASPEN Assessment System for Population Exposure Nationwide
BEADS Benzene Exposure and Absorbed Dose Simulation
BEAM Benzene Exposure Assessment Model
BM Body mass
BOC Bureau of Census
BW Body weight
CAA Clean Air Act
CAAA Clean Air Act Amendments
CalTOX California Total Exposure Model for Hazardous Waste Sites
CHAD Comprehensive Human Activity Data
CMAQ Community Multi-scale Air Quality
CO Carbon monoxide
CONSEXPO Consumer Product Exposure Model
CPIEM California Population Indoor Exposure Model
DEPM Dietary Exposure Potential Model
DERM Dermal Exposure Reduction Model
DOE U.S. Department of Energy
ECF Energy conversion factor
EDMAS Exposure and Dose Modeling and Analysis System
EE Energy expenditure
EML Exposure Models Library
EPA U.S. Environmental Protection Agency
ETS Environmental tobacco smoke
GEMS Graphical Exposure Modeling System
GIS Geographic information system
GUI Graphical User Interface
HAP Hazardous air pollutant
HAPEM4 Hazardous Air Pollutant Exposure Model, Version 4
HEM Human Exposure Model
HPI Hazard Potential Index
HVAC Heating, ventilation, and air conditioning
IAQM Indoor Air Quality Model
IEM Indirect Exposure Methodology Model
IMES Integrated Exposure Models Evaluation System
INTOXX Integrated Toxic Expected Exceedance
ISC Industrial Source Complex
ISCLT Industrial Source Complex, Long-term
NOVEMBER 1999
vii TRIM.EXPO TSD (DRAFT)
-------
ACRONYMS
ISCST Industrial Source Complex, Short-term
ISMCM Integrated Spatial Multimedia Compartmental Model
LADD Lifetime Average Daily Dose
LSODE Livermore Solver for Ordinary Differential Equations
MAVRIQ Model for Analysis of Volatiles and Residential Indoor Air Quality
MCCEM Multi-Chamber Concentration and Exposure Model
MENTOR Modeling Environment for Total Risk Studies
MEPAS Multimedia Environmental Pollutant Assessment System
MET Metabolic equivalent of work
MIMS Multimedia Integrated Modeling System
MMSOILS Multimedia Contaminant Fate, Transport, and Exposure Model
MPE Multiple Pathways of Exposure
MSA Metropolitan Statistical Area
NAAQS National Ambient Air Quality Standard
NAMS National Air Monitoring Station
NAS National Academy of Sciences
NASQAN National Stream Quality Accounting Network
NATA National Air Toxics Assessment
NCC National Computing Center
NCEA National Center for Environmental Assessment
NCHS National Center for Health Statistics
NEM NAAQS Exposure Model
NHAPS National Human Activity Pattern Survey
NHIS National Health Interview Survey
NIST National Institute of Standards and Technology
NOPES Non-occupational Pesticides Exposure Study
NRC National Research Council
OAQPS EPA Office of Air Quality Planning and Standards
OMS EPA Office of Mobile Sources
ORD EPA Office of Research and Development
OW EPA Office of Water
PBPK Physiologically-based pharmacokinetic
PC Personal computer
PDF Probability density function
PEC Predicted environmental concentration
PEM Personal exposure monitor
pHAP Probabilistic Hazardous Air Pollutant Exposure Model
PM2 5 Paniculate matter with aerodynamic size diameter of 2.5/zm or less
PM10 Particulate matter with aerodynamic size diameter of lOyum or less
pNEM Probabilistic National Ambient Air Quality Standards Exposure Models.
PNL Pacific Northwest Laboratory
PTEAM Particle Total Exposure Assessment Methodology
RESRAD Residual Radiation
RIA Regulatory impact analysis
RMR Resting metabolic rate
SAB EPA's Science Advisory Board
NOVEMBER 1999
Vlll
TRIM.EXPO ISO (DRAFT)
-------
ACRONYMS
SCIES
SCREAM2
SHAPE
SHEDS
SLAMS
STAR
STORE!
TAP
TEAM
THERdbASE
TOXLT
TRIM
TRIM.Expo
TRIM.FaTe
TSD
USES
USGS
VOC
Screening Consumer Inhalation Exposure Software
South Coast Risk and Exposure Assessment Model, Version 2
Simulation of Human Activities and Pollutant Exposure
Stochastic Human Exposure and Dose Simulation
State and Local Air Monitoring Stations
STability ARray
Storage and Retrieval
Time Activity Patterns
Total Exposure Assessment Methodology
Total Human Exposure Risk database and Advance Simulation Environment
Toxic Modeling System, Long-term
Total Risk Integrated Methodology
TRIM Exposure-Event module
TRIM Environmental Fate, Transport, and Ecological Exposure module
Technical Support Document
Unified System for the Evaluation of Substances
U.S. Geological Survey
Volatile organic compound
NOVEMBER 1999
IX
TRIM.Expo TSD (DRAFT)
-------
TABLE OF CONTENTS
TABLE OF CONTENTS
Disclaimer i
Acknowledgments iii
Preface v
Acronyms vii
Table of Contents xi
1. Introduction 1-1
1 1 Goals and Objectives for TRIM 1-2
1.2 TRIM Design 1-4
1.2.1 Description of TRTM.FaTE 1-6
1.2.2 Description of TRIM.Expo . . . .1-7
1 2 3 Description of TRIM.Risk 1-8
1 3 TRIM Development 1-9
1 3 1 Initial Development Activities 1-9
1.3 2 Recent Activities 1-10
1.3.3 Future Activities 1-11
1.4 Phasing TRIM into OAQPS' Set of Modeling Tools 1-12
2. TRIM.Expo: General Overview and Background 2-1
21 Rationale and Need for TRIM.Expo 2-1
2 2 Important Definitions 2-5
222 Basic Definitions Related to Dose ... 2-6
2.2.3 Other Important TRIM Expo Definitions .... 2-8
2 3 Approach Used in Developing TRIM.Expo 2-9
24 Taxonomy of Exposure Attributes for Multimedia Pollutants 2-10
24 1 Exposure Characterization Process and Exposure Attributes 2-10
242 Dimensions of the Exposure Assessment Problem ... 2-12
2 5 TRIM Exposure-event Concept 2-13
3. Summary Review of Existing Exposure Models and Rationale for Developing
TRIM.Expo 3-1
3.1 Rationale for Developing TRIM.Expo 3-1
3.2 Required Attributes of the TRIM.Expo Module 3-1
3.3 Overview of Current Models and Modeling Approaches 3-2
3.3.1 Inhalation Exposure Models 3-6
3.3.2 Multimedia Exposure Models 3-7
3 4 Strengths and Limitations of Existing Models 3-8
4. Design Framework and Conceptualization of TRIM.Expo 4-1
4.1 Exposure-event Module Structure 4-2
4.1.1 Basic Exposure-event Function 4-3
4.1.2 Exposure or Potential Dose Profiles 4-4
4.1.3 Average Exposure Concentration 4-7
4.1.4 Intake-adjusted Average Exposure Concentration 4-8
NOVEMBER 1999 ~ TRIM.Expo TSD
xi (DRAFT)
-------
TABLE OF CONTENTS
4 1 5 Intermedia Transfer Factor 4-8
4 1 6 Average Daily Potential Dose 4-9
4 2 Defining the Model Components for a TRIM.Expo Application 4-9
4 2 1 Define Study Area, Exposure Districts, and Environmental Media ... 4-11
4211 Ambient Air 4-11
4.2.1 2 Vegetation 4-11
4.2.1 3 Surface Soil 4-11
4.2 1.4 Root Zone Soil 4-12
42.1.5 Vadose Zone Soil 4-12
4 2.1.6 Ground Water 4-12
4.2 1.7 Surface Water 4-12
422 Define Exposure Media and Microenvironments 4-12
423 Define Relevant Intermedia Transfers 4-14
424 Divide Population into Appropriate Sets of Cohorts 4-19
425 Develop an Exposure-event Sequence for Each Cohort 4-20
426 Determine Exposure Media Concentrations and Contact in Each
Microenvironment 4-21
427 Estimate an Intake Rate for Each Dose Event 4-21
428 Extrapolate the Cohort Exposures to the Populations of Interest . . . 4-22
4 2.9 Functional Attributes 4-25
4.2 9 1 Inclusion of Indoor and Outdoor Environments and Their Emission
Sources 4-25
4292 Flexible, Modular, and Portable Algorithms . 4-26
4293 Explicit Treatment of Uncertainty and Variability . . 4-27
4 3 Data Input Requirements 4-30
43.1 Environmental Media Concentrations 4-31
432 Concentrations of Pollutants in Microenvironments 4-32
4321 Indoor Versus Outdoor Concentrations . . . . 4-32
4322 Mass Balance Model Approach 4-33
4 3.2.3 Empirical Indoor/outdoor Ratios Approach 4-34
433 Activity Pattern Data 4-35
434 Demographic and At-risk Population Data 4-36
5. Inhalation 5-1
5.1 Overview of the Approach 5-1
5.1.1 Selection of Study Area 5-2
5 1 2 Selection of Populations of Interest 5-2
5.1.3 Definition of Population Cohorts 5-2
5.1.4 Develop an Inhalation Exposure-event Sequence for Each Cohort . . 5-3
5.1.5 Estimate Pollutant Concentration and Ventilation Rate Associated with
Each Exposure Event 5-4
5.1.6 Extrapolate the Cohort Inhalation Exposures to the Populations of Interest
52 Presentation of the Model Algorithms by Microenvironmental Location 5-7
52.1 Microenvironmental Locations Specific to Indoor Air and Inside Vehicles-7
5.2.2 Microenvironmental Locations Specific to Ambient Air 5-10
5.3 Integration of Exposure Across Multiple Locations and Times 5-12
NOVEMBER 1999 TRJM.Expo TSD
xii (DRAFT)
-------
TABLE OF CONTENTS
5 4 Summary of Inputs and Values 5-13
5 4.1 Data Inputs for the Mass Balance Model 5-13
5.4.2 Data Inputs for Time/activity Patterns 5-15
543 Data Inputs for Ventilation Rate 5-16
6. Ingestion 6-1
6 1 Overview of the Approach 6-1
6.1.1 Selection of Population Cohorts 6-4
6.1.2 Time Resolution of Exposure Events 6-5
6.1.3 Exposure Media Considered 6-5
6.1.3.1 Ingested Water 6-6
6 1.3.2 Food 6-7
6.1 3.3 Soil and Dust 6-8
6 1.4 Exposure Locations . 6-8
61.41 Residential Exposure Locations ... . . 6-8
6.1 4.2 Other Exposure Locations . . 6-9
6.2 Presentation of the Model Algorithms by Exposure Media . .... 6-9
6.2 1 Ingested Water 6-9
6 2.2 Ingestion of Soil and House Dust . . 6-11
6.2.21 Soil Ingestion (Outdoors) ... . 6-12
6222 Dust Ingestion (Indoors) 6-13
6.2 3 Ingestion of Pollutants in Home-grown Produce or Home-bred Animals
6-14
62.3.1 Vegetables, Fruits, and Grains . 6-15
623.2 Dairy Products . 6-17
6233 Eggs 6-18
6.2 3.4 Meat and Poultry 6-19
6.24 Locally-grown Commercial Foods .6-19
6 2.4 1 Vegetables, Fruits, and Grains 6-20
6.2.4.2 Dairy Products 6-21
6.24.3 Eggs 6-21
6.2.4.4 Meat and Poultry 6-21
6245 Fish (Commercial, Subsistence, and Recreational) . . . 6-22
6.2.5 Recreational Sport Meat (Hunting) 6-22
6 3 Integration of Exposures Across Multiple Ingestion Media 6-23
6.4 Discussion of Algorithm Inputs and Values 6-24
7. References 7-1
Appendices
A. Glossary
B. Comparison/Critique of Exposure Models
C. List of TRIM.Expo Input Parameters
NOVEMBER 1999 TRTMExpo TSD
xiii (DRAFT)
-------
[This page intentionally left blank.]
-------
CHAPTER 1
INTRODUCTION
1. INTRODUCTION
The Office of Air Quality Planning and Standards (OAQPS) of the U.S. Environmental
Protection Agency (EPA, or the Agency) has the responsibility for the hazardous and criteria air
pollutant programs described by sections 112 and 108 of the Clean Air Act (CAA). Several
aspects of these programs require evaluation of the health risks and environmental effects
associated with exposure to these pollutants.1 In response to these risk-related mandates of the
CAA, and the scientific recommendations of the National Academy of Sciences (NAS) (NRC
1994), the Presidential/Congressional Commission on Risk Assessment and Risk Management
(CRARM) (CRARM 1997), as well as EPA guidelines and policies, OAQPS recognized the need
for improved fate and transport, exposure, and risk modeling tools that
Have multimedia assessment capabilities;
Have human health and ecological exposure and risk assessment capabilities,
Can perform multiple pollutant assessments (e.g., ability to assess mixtures of pollutants,
ability to track chemical transformations);
Can explicitly address uncertainty and variability;
Have the ability to easily perform analyses iteratively, moving from the use of simpler
assumptions and scenarios to more detailed assessments; and
Are readily available and user-friendly, so that they can be used by EPA, as well as by a
variety of Agency stakeholders
In 1996, OAQPS embarked on a multi-year effort to develop the Total Risk Integrated
Methodology (TRIM), a time series modeling system with multimedia capabilities for assessing
human health and ecological risks from hazardous and criteria air pollutants
The main purpose of the TRIM Status Report (U.S. EPA 1999c) is to summarize the
work performed during the second developmental phase of TRIM. The first phase, which
included the conceptualization of TRIM and implementation of the TRIM conceptual approach
through development of a prototype of the first TRIM module, TRIM.FaTE (U.S EPA 1998a),
was reviewed by EPA's Science Advisory Board (SAB) in May 1998 (U.S EPA 1998b). The
second developmental phase has included refining TRIM.FaTE and developing a model evaluation
plan, initiating development of the second module (TRIM Expo), and conceptualizing the third
module (TRIM Risk) In addition, progress has been made on developing overarching aspects,
such as the computer framework and an approach to uncertainty and variability. Consistent with
the integral role of peer review in the TRIM development plan, the current Status Report and
1 Hazardous air pollutants (HAPs) include any air pollutant listed under CAA section 112(b); currently,
there are 188 air pollutants designated as HAPs. Criteria air pollutants are air pollutants for which national
ambient air quality standards (NAAQS) have been established under the CAA; at present, the six criteria air
pollutants are paniculate matter, ozone, carbon monoxide, nitrogen oxides, sulfur dioxide, and lead.
NOVEMBER 1999 M TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
Technical Support Documents (TSDs) were subjected to review by representatives from the
major program offices at EPA and an EPA Models 20002 review team prior to this SAB advisory.
This TSD describes the development of TRIM.Expo, detailing the work completed to date
toward developing the first TRIM.Expo prototypes. More specifically, the report addresses the
following areas
OAQPS' modeling needs and the intended goals for TRIM,
Design of the TRIM modeling system;
TRIM Expo's relation to the TRIM modeling system,
Purpose and ongoing development of TRIM Expo,
Conceptual framework of TRIM.Expo, in the context of the general approach to exposure
assessment and modeling,
Approach used in TRIM.Expo for calculating inhalation and ingestion exposures,
Comparative overview of existing exposure models and modeling approaches, addressing
the strengths and limitations of some of the more commonly used exposure models,
Plans for developing and evaluating the TRIM Expo prototypes,
Glossary of terms and definitions, and
Listing of examples of input parameters for TRIM Expo
1.1 GOALS AND OBJECTIVES FOR TRIM
The TRIM modeling system is intended to represent the next generation of human and
environmental exposure and risk models for OAQPS. For example, TRIM is expected to be a
useful tool for performing exposure and/or risk assessments for the following CAA programs, the
Residual Risk Program (CAA section 112[f]), the Integrated Urban Air Toxics Strategy (CAA
section 112[k]), studies of deposition to water bodies and mercury emissions from utilities (CAA
2 Following the report of the Agency Task Force on Environmental Regulatory Modeling (U.S. EPA
1994a), the Agency conducted the Models 2000 Conference in December 1997. This conference has led to
renewed emphasis on Agency -wide coordination of model development and the proposal for the implementation of
a Council on Regulatory Environmental Modeling (CREM) to facilitate and promote scientifically-based,
defensible regulatory computer models. The charter for CREM has been reviewed by SAB and is being updated for
implementation by the Agency.
NOVEMBER 1999 T3 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
sections 112[m] and 112[n]), petitions to delist individual HAPs and/or source categories (CAA
sections 112[b][3] and 112[c][9]); review and setting of the national ambient air quality standards
(NAAQS) (CAA section 109), and regulatory impact analyses (RIA).
The goal in developing TRIM is to create a modeling system, and the components of that
system, for use in characterizing human health and ecological exposure and risk in support of
hazardous and criteria air pollutant programs under the CAA The goal in designing TRIM is to
develop a modeling system that is. (1) scientifically defensible, (2) flexible, and (3) user-friendly
(1) Characteristics of the TRIM components important to their scientific defensibility include
the following.
Conservation of pollutant mass. The modeled pollutant(s)' mass will be conserved
within the system being assessed, wherever appropriate and feasible, including during
intermedia transfers For pollutants where transformation is modeled, the mass of the
core substance (e.g., mercury for methylmercury as well as divalent mercury) within the
modeling simulation will be preserved
Ability to characterize parameter uncertainty and variability. For critical
parameters, the impacts of parameter uncertainty and variability on model outputs will be
tracked and, where feasible, differentiated
Capability for multiple pollutant, multiple media, multiple exposure pathway
assessment. The TRIM modeling system is being designed to facilitate assessment of
risks posed by aggregate exposures to single or multiple chemicals from multiple sources
and via multiple exposure pathways
(2) To ensure flexibility, the features of TRIM include the following
Modular design. Major components of TRIM will be independent and can be used
individually, with outside information or models, or in combination Only those model
components necessary for evaluating the particular pollutants, pathways, and/or effect
endpoints of interest need be employed in an assessment
Flexibility in temporal and spatial scale. Exposure and risk assessments will be possible
for a wide range of temporal and spatial scales, including hourly to daily or yearly time
steps, and from local (10 kilometers (km) or less) to greater spatial scales (depending on
the module)
Ability to assess human and ecological endpoints. Impacts to humans and/or biota can
be assessed
(3) To ensure that TRIM will be user-friendly for a variety of groups, including EPA, state
and local agencies, and other stakeholders, TRIM will have the following characteristics.
NOVEMBER 1999 1-3 TRIM. EXPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
Easily accessible. The TRIM modeling system will be accessible for use with a personal
computer (PC) The system may be available for download from the Internet and
accessible through an Agency model system framework (e.g., Models-3 (U.S. EPA
1999d))
Well-documented. Guidance materials for use of the TRIM modeling system will be
provided through a user's guide, with a focus on the modular aspects of the modeling
system, limitations of the modeling system, and appropriate uses, user responsibilities, and
user options
Clear and transparent. The graphical user interface of the TRIM computer framework
will provide transparency and clarity in the functioning of the TRIM modules, and output
from the risk characterization module will document modeling assumptions, limitations,
and uncertainties
1.2 TRIM DESIGN
The current TRIM design (Figure 1-1) includes three individual modules The
Environmental Fate, Transport, and Ecological Exposure module, TRIM.FaTE, accounts for
movement of a chemical through a comprehensive system of discrete compartments (e.g., media,
biota) that represent possible locations of the chemical in the physical and biological environments
of the modeled ecosystem and provides an inventory, over time, of a chemical throughout the
entire system In addition to providing exposure estimates relevant to ecological risk assessment,
TRIM FaTE generates media concentrations relevant to human pollutant exposures that can be
used as input to the Exposure-Event module, TRJM.Expo. In TRIM Expo, human exposures are
evaluated by tracking population groups referred to as "cohorts" and their inhalation and
ingestion through time and space In the Risk Characterization module, TRIM.Risk, estimates of
human exposures or doses are characterized with regard to potential risk using the corresponding
exposure- or dose-response relationships The TRIM Risk module is also being designed to
characterize ecological risks from multimedia exposures The output from TRIM.Risk will
include documentation of the input data, assumptions in the analysis, and measures of uncertainty,
as well as the results of risk calculations and exposure analysis
An overarching feature of the TRIM design is the analysis of uncertainty and variability
A two-stage approach for providing this feature to the user is being developed3. The first stage
includes sensitivity analyses that are useful in identifying critical parameters, while more detailed
uncertainty and variability analyses using Monte Carlo methods (e.g., for refined assessment of
the impact of the critical parameters) are available in the second stage The uncertainty and
variability feature augments the TRIM capability for performing iterative analyses. For example,
the user may perform assessments varying from simple deterministic screening analyses using
conservative default parameters to refined and complex risk assessments where the impacts of
parameter uncertainty and variability are assessed for critical parameters.
3 This approach is being developed for the overall TRIM system. However, it has only been implemented
to date for the TRIM.FaTE module
NOVEMBER 1999 T4 TRTM.EXPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
Air Quality Models
(eg,ISC3, AERMODl
Other Multimedia Models
(eg,MEND-TOX)
Figure 1-1
Conceptual Design of TRIM
Environmental Fate,
Transport, and
Ecological Exposure
(TRIM.FaTE)
Inputs
eg,
Physical, Chemical
Properties
Site-specific Data
CIS Data
Monitoring Data]
Temporal and
Spatial Distributio^
of Pollutant
Concentrations
Inputs
eg,
Activity Data
(eg, CHAD)
Population Data
(eg,1990BOC)
Indoor/Outdoor 7
Concentration Ratios.
[media concentrations
relevant to human
exposures]
Exposure Event
(TRIM.Expo)
Dosimetry Models
(eg,CO, Pb models)
Temporal and Spatial
Distribution of
Exposures within
Exposed Human
Population
Temporal and Spati;
Distribution of Dose:
within Exposed
Population
Human Health
Dose-response
Assessment
(e.g.RfC, URE)
Risk
Characterization
(TRIM.Risk)
[media and biota
concentrations and biota
pollutant intake rates
relevant to ecological
exposures]
Ecological Effects /
Assessment I
(e g , endpomts, critenaA
Documentation of assumptions and input data
Quantitative risk and exposure
characterization (human and ecological)
Measures of uncertainty and variability
Description of limitations (graphicalAabular/
GIS presentation)
NOVEMBER 1999
1-5
TRIM EXPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
Additionally, the modular design of TRIM allows for flexibility in both its development
and application Modules can be developed in a phased approach, with refinements being made as
scientific information and tools become available. Furthermore, the user may select any one or
more of these modules for an assessment depending on the user's needs. For example, when
performing a human health risk assessment for an air pollutant for which multimedia distribution is
not significant, TRIM.Expo may be applied using ambient concentration data or the output from
an air quality model external to TRIM; the output from TRIM Expo may then be used as input to
TRIM.Risk to perform the desired risk analyses. In the case of a multimedia air pollutant, such as
mercury, the user may choose to run all three TRIM modules to assess both human and ecological
risks posed by multipathway exposures from multiple media.
Overview descriptions of the TRIM modules are provided in Sections 1.2 1 through 1 2 3,
the status and plans for development are presented in Section 1 3, and plans for application appear
in Section 1.4 A summary of the previous SAB comments and OAQPS responses is presented in
Chapter 2 of the TRIM Status Report The approach for handling uncertainty and variability in
TRIM is described in Chapter 3 of the TRIM Status Report. Certain aspects of the TRIM.FaTE
module are addressed in greater detail in Chapters 4 through 7, and additional details on
TRIM Expo and TRIM.Risk are provided in Chapters 8 and 9, respectively, of the TRIM Status
Report Chapter 10 of the TRIM Status Report discusses the computer framework that is being
implemented for the TRIM system In addition, the TRIM FaTE TSD (U.S. EPA 1999a,b)
provides more detail on TRIM FaTE.
1.2.1 DESCRIPTION OF TRIM.FaTE
The first TRIM module to be developed, TRIM.FaTE, is a spatial compartmental mass
balance model that describes the movement and transformation of pollutants over time, through a
user-defined, bounded system that includes both biotic and abiotic components (compartments)
The TRIM FaTE module predicts pollutant concentrations in multiple environmental media and in
biota and pollutant intakes for biota, all of which provide both temporal and spatial exposure
estimates for ecological receptors (i.e., plants and animals) The output concentrations from
TRIM FaTE also can be used as inputs to a human exposure model, such as TRIM Expo, to
estimate human exposures
Significant features of TRIM.FaTE include. (1) the implementation of a truly coupled
multimedia model, (2) the flexibility to define a variety of scenarios, in terms of the links among
compartments as well as the number and types of compartments, as appropriate for the desired
spatial and temporal scale of assessment, (3) the use of a transparent approach to chemical mass
transfer and transformation based on an algorithm library that allows the user to change how
environmental processes are modeled, (4) an accounting for all of the pollutant as it moves among
the environmental compartments; (5) an embedded procedure to characterize uncertainty and
variability; and (6) the capability to provide exposure estimates for ecological receptors The
TRIM FaTE module is the most fully developed of the TRIM modules at this time, and this
development has produced a library of algorithms that account for transfer of chemical mass
throughout an environmental system, a database of the information needed to initialize these
algorithms for a test site, and a working computer model.
NOVEMBER 1999 1-6 TRTM.Expo TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
1.2.2 DESCRIPTION OF TRIM.Expo
The TRIM.Expo module, similar to most human exposure assessment models, provides an
analysis of the relationships between various chemical concentrations in the environment and
exposure levels of humans Because multiple sources of environmental contamination can lead to
multiple contaminated media, including air, water, soil, food, and indoor air, it is useful to focus
on the contaminated environmental media with which a human population will come into contact.
These media typically include the envelope of air surrounding an individual, the water and food
ingested by an individual, and the layer of soil and/or water that contacts the surface of an
individual. The magnitude and relative contribution of each exposure pathway must be
considered to assess total exposure to a particular chemical Currently, the focus of TRIM.Expo
development is on inhalation and ingestion exposure, however, dermal exposure will be added
later
The exposure analysis process
consists of relating chemical
concentrations in environmental media
(e.g., air, surface soil, root zone soil,
surface water) to chemical
concentrations in the exposure media
with which a human or population has
contact (e.g., air, tap water, foods,
household dusts, and soils). The initial
prototype for TRIM.Expo will predict
exposure by tracking the movement of a
population cohort through locations
where chemical exposure can occur
according to a specific activity pattern
In a typical application, TRIM FaTE
could be used to provide an inventory of
chemical concentrations across the
ecosystem at selected time intervals (e.g.,
days, hours) For chemicals that are not
persistent and/or bioaccumulative,
processed air monitoring data or air
dispersion modeling results can be
substituted for TRIM.FaTE output data.
The TRIM Expo module would then use
these chemical concentration data,
combined with the activity patterns of
the cohorts, to estimate exposures. The ^^^^^^^^^^^^^^^^^^^^^^^^^^"
movements are defined as an exposure-
event sequence that can be related to time periods for which exposure media concentrations are
available (e.g., from TRIM FaTE, ambient data, and/or dispersion modeling results). Each
exposure event places the population cohort in contact with one or more environmental media
within a specified microenvironment (e.g., inside a home, along a road, inside a vehicle) in an
exposure district for a specified time interval. In addition to the location assignments, the
TRIM.Expo KEY TERMS
Cohort - A group of people within a population with
the same demographic variables who are assumed
to have similar exposures.
Activity pattern - A series of discrete events of
varying time intervals describing information about
an individual's lifestyle and routine. The information
contained in an activity pattern typically includes the
locations that the individual visited (usually
described in terms of microenvironments), the
amount of time spent in those locations, and a
description of what the individual was doing in each
location (e.g., sleeping, eating, exercising).
Microenvironment - A defined space in which
human contact with an environmental pollutant takes
place and which can be treated as a well-
characterized, relatively homogeneous location with
respect to pollutant concentrations for a specified
time period.
Exposure district - A geographic location within a
defined physical or political region where there is
potential contact between an organism and a
pollutant and for which environmental media
concentrations have been estimated either through
modeling or measurement.
NOVEMBER 1999
1-7
TRIM EXPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
exposure event would provide information relating to the potential for pollutant uptake, such as
respiration rate and quantity of water consumed The TRIM.Expo module is intended to
contribute to a number of health-related assessments, including risk assessments and status and
trends analyses.
1.2.3 DESCRIPTION OF TRIM.Risk
Risk characterization is the final step in risk assessment and is primarily used to integrate
the information from the other three key steps (i.e., hazard identification, dose-response
assessment, exposure assessment) Within the TRIM framework, TRIM.Risk, the risk
characterization module, will be used to integrate the information on exposure (human or
ecological receptor) with that on dose-response or hazard and for providing quantitative
descriptions of risk and some of the attendant uncertainties. The TRIM Risk module will provide
decision makers and the public with information for use in developing, evaluating, and selecting
appropriate air quality standards and risk management strategies. The purpose of TRIM Risk is
to integrate information from other TRIM modules and to facilitate the preparation of a risk
characterization The TRIM.Risk module will, therefore, be able to summarize or highlight the
major points from each of the analyses conducted in the other TRIM modules. Where possible,
the TRIM.Risk module will do so in an automated manner In general, TRIM.Risk will (1)
document assumptions and input data, (2) conduct risk calculations and data analysis, and (3)
present results and supporting information
Current and proposed EPA guidance on risk characterization will guide the development
of TRIM.Risk. The TRIM.Risk module will be developed in a phased approach similar to other
TRIM modules Ideally, TRIM.Risk will provide all of the information required to prepare a full
risk characterization. However, the type and variability of information needed for this purpose
are vast Therefore, the type of information generated by TRIM.Risk will evolve over time as the
Agency gains experience and has the resources to implement more flexibility. For example, early
versions of TRIM Risk will be limited to preparing summaries of input data and results, without
supporting text However, as the Agency gains experience, it may be possible to incorporate
generic language to more fully describe the information required for a full risk characterization.
Many EPA risk assessments will be expected to address or provide descriptions of (1) individual
risk,4 including the central tendency and high-end portions of the risk distribution, (2) population
risk, and (3) risk to important subgroups of the population such as highly exposed or highly
susceptible groups or individuals, if known Some form of these three types of descriptors will be
developed within TRIM.Risk and presented to support risk characterization. Because people
process information differently, it is appropriate to provide more than one format for presenting
the same information. Therefore, TRIM Risk will be designed so that the output can be presented
in various ways in an automated manner (e.g., Chart Wizard in Microsoftฎ Excel), allowing the
user to select a preferred format.
4 The phrase individual risk as used here does not refer to a risk estimate developed specifically for a
single member of a population Rather, it refers to the estimate of risk for a subgroup of a population that is
presented as an estimate of the risk faced by a person rather than by the population as a whole.
NOVEMBER 1999 Ti TRIM.Expo TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
1.3 TRIM DEVELOPMENT
In the development of TRIM, existing models and tools are being relied upon where
possible Adopting or incorporating existing models or model components into a tool that meets
OAQPS' needs is preferable as it is usually the most cost-effective approach. Consequently,
review of existing models and consideration of other current modeling efforts is an important part
of TRIM development activities. Reviews of relevant models existing at the initiation of
development activities for each module are described in this document and in the TRUVI.FaTE
TSD Additionally, OAQPS is closely following several current activities as they relate to TRIM.
Current Agency model development activities relevant to TRIM development include the
recently published updated guidance on assessing health risks associated with indirect exposure to
combustor emissions (U.S. EPA 1999h). This guidance, previously referred to as the Indirect
Exposure Methodology (IEM), is now called the Multiple Pathways of Exposure (MPE) method
In addition, the multimedia model, FRAMES-HWIR, has recently been developed by the Agency
to support a specific risk assessment need regarding hazardous chemicals released from land-
based waste management units The FRAMES-HWIR model has been developed as part of a
focused fast-track (two-year) effort to support a risk-based regulation regarding disposal of
hazardous waste (FIWIR99) 5 Another model of interest for multimedia pollutants is the
Stochastic Human Exposure and Dose Simulation (SHEDS) model (e.g., Ozkaynak et al 1999)
The OAQPS will be carefully considering the various aspects of MPE, FRAMES-HWIR, and
SHEDS with regard to OAQPS needs, as well as compatibility with or future improvements or
evaluations of TRIM. As TRIM is intended to be a dynamic method, developmental activities will
consider and respond as appropriate to newly available methods and scientific information
A current major Agency research project involves the design and development of a flexible
software system to simplify the development and use of air quality models and other
environmental decision support tools This system, called Models-3, is designed for applications
ranging from regulatory and policy analysis to understanding the complex interactions of
atmospheric chemistry and physics (U.S. EPA 1999d) The June 1999 release of Models-3
contains a Community Multi-Scale Air Quality (CMAQ) modeling system for urban- to regional-
scale air quality simulation of tropospheric ozone, acid deposition, visibility, and fine particles
The long-term goal is to extend the system to handle integrated cross-media assessments and
serve as a platform for community development of complex environmental models In recognition
of the availability of Models-3 over the longer term, OAQPS has designed and is developing the
TRIM computer framework to be compatible with the Models-3 system
1.3.1 INITIAL DEVELOPMENT ACTIVITIES
The first phase of TRIM development included the conceptualization of TRIM and the
implementation of the TRIM conceptual approach through the development of a prototype of the
first TRIM module, TRIM.FaTE (U.S. EPA 1998a). The progress on TRIM.FaTE included the
development of (1) a conceptual design for the module; (2) a library of algorithms that account
5 The FRAMES-HWIR documentation is scheduled for public release in fall 1999.
NOVEMBER 1999 1^9 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
for chemical mass transfer throughout the ecosystem, (3) a database to initialize the algorithms for
a test site, and (4) a working prototype in spreadsheet format.
Consistent with Agency peer review policy (U.S EPA 1998c) and the 1994 Agency Task
Force on Environmental Regulatory Modeling (U.S EPA 1994a), internal and external peer
review are an integral part of the TRIM development plan. Following the first phase of TRIM
development, OAQPS submitted TRIM to SAB under their advisory method of review (U S. EPA
1998b) In May 1998 in Washington, DC, the Environmental Models Subcommittee
(Subcommittee) of the Executive Committee of SAB reviewed the TRIM project The SAB
Subcommittee was charged with assessing the overall conceptual approach of TRIM and the
specific approach of TRIM FaTE
The SAB Subcommittee reported that the development of TRIM and the TRIM.FaTE
module was conceptually sound and scientifically based (U S. EPA 1998b) The SAB
Subcommittee provided specific recommendations related to six specific charge questions The
SAB recommendations are detailed in Chapter 2 of the TRIM Status Report along with brief
responses, and changes to TRIM FaTE based in part on the SAB recommendations are
highlighted in Chapter 4 of the TRIM Status Report
1.3.2 RECENT ACTIVITIES
During the most recent developmental phase of TRIM, progress has been made in many
areas, including a change to the overall modular design of TRIM. As shown in Figure 1-1, the
TRIM design now includes three modules TRIM FaTE, TRIM Expo, and TRIM Risk The
design presented to SAB in May 1998 included three other modules (Pollutant Uptake,
Biokinetics, and Dose/Response) In recognition of the flexibility of the TRIM design, which
provides an ability to rely on a variety of input data and outside models, OAQPS decided not to
include the development of these modules in the TRIM design at this time.
In consideration of SAB comments, TRIM.FaTE was refined, including the development
of new and updated capabilities, as well as the development and limited testing of methodologies
for model set-up, uncertainty and variability analysis, and evaluation. In addition, OAQPS
developed a conceptual plan for TRIM Expo, initiated work on a prototype of TRIM Expo
(initially focusing on inhalation), and developed a conceptual design for TRIM Risk
Furthermore, the overall computer framework for TRIM was designed and implemented in a PC-
based platform, and substantial progress was made in installing TRIM.FaTE into this framework.
Changes and additions to TRIM.FaTE are discussed in more detail in Chapter 4 of the TRIM
Status Report The development of TRIM Expo is discussed in Chapter 8, and the conceptual
plan for TRIM Risk is described in Chapter 9 of the TRIM Status Report. In addition, the
TRIM FaTE TSD provides more details on TRIM.FaTE
The current TRIM documentation has gone through internal Agency peer review, which
involved reviewers across the Agency, including the major program offices, the Office or
Research and Development, and staff involved in the Agency's Models 2000 efforts. The current
SAB advisory will be the second on TRIM development activities.
NOVEMBER 1999 1-10 TRIM.ExPO TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
1.3.3 FUTURE ACTIVITIES
Following the 1999 SAB advisory, improvements will be made to the uncertainty and
variability approach, TRIM.Expo prototype, and TRIM.Risk conceptual plan These revisions are
scheduled to be completed in 2000. As needed, refinements will be made to the TRIM.FaTE
evaluation plan, and completion of the bulk of those activities are also scheduled for 2000. The
Agency has planned for a substantial amount of progress on each of the TRIM modules for 2000
and 2001, as described below
TRIM.FaTE. Future work on TRIM.FaTE will include model evaluation activities and
additional development of the module to accommodate additional chemicals The
TRIM FaTE module is expected to be available for limited external use late in 2000 and to
be publicly released in 2001.
TRIM.Expo. Future work on TRIM.Expo in 2000 will include the further development
of ingestion algorithms, incorporation of EPA's Air Pollutant Exposure Model (APEX)
coding into the TRIM platform followed by adjustments to APEX to include ingestion
algorithms, a test case of the inhalation pathway, and a test case of inhalation and
ingestion pathways Over the longer term, addition of the dermal pathway to the module
will be initiated
TRIM.Risk. Development of TRIM Risk will begin after SAB comments are received on
the conceptual design Module development will include identification of data needs and
formatting of data outputs Programming for a TRIM Risk prototype is expected to be
completed in 2000
TRIM computer framework. Further development of the TRIM computer framework,
including incorporation of the TRIM.Expo (inhalation) module, will take place during
2000. Features to be refined during this time frame include limited geographic information
system (GIS) or mapping capabilities Additionally, long-range comprehensive GIS
planning will occur Development of user guidance materials is planned for 2000 (see text
box)
In addition to consulting with Agency scientists during future TRIM development (i.e.,
peer involvement), in late 2000 or early 2001, OAQPS will seek both internal and external peer
review of new aspects following the next phase of TRIM development. In addition to the SAB,
which provides the Agency with reviews, advisories, and consultations, other external peer review
mechanisms consistent with Agency policy (U.S. EPA 1998c) include the use of a group of
independent experts from outside the Agency (e.g., a letter review by outside scientists), an ad
hoc panel of independent experts, and peer review workshops The OAQPS intends to seek the
peer review mechanism appropriate to the importance, nature, and complexity of the material for
review.
NOVEMBER 1999 1-11 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION-
GUIDANCE
Development of the TRIM user's guide is scheduled to begin in 2000, along with a plan for
training activities. The OAQPS recognizes the importance of developing detailed user guidance that
will assist users in defining, for a particular modeling application, the spatial and temporal resolution,
compartments and linkages, and parameters and initial conditions. For example, the TRIM.FaTE
guidance will likely emphasize the value of performing several different preliminary simulations in
verifying the adequacy of the parcel and compartment specifications for the desired application.
Similarly, detailed user guidance will be developed for TRIM.Expo to assist users in defining cohorts,
study areas, exposure districts, and microenvironments, as well as various parameters and exposure
factors.
It also will be important for the guidance to note the responsibility of the user in defining the
simulation as appropriate to the application. For example, in TRIM.FaTE, default values will likely
be made available with the model for a variety of parameters ranging from physiological
characteristics of various biota to physical characteristics of abiotic media; the user will need to
consider appropriateness of these values or others (e.g., site-specific data) for their application.
While the TRIM modules are intended to provide valuable tools for risk assessment, and their
documentation and guidance will identify, as feasible, uncertainties and limitations associated with
their application, the guidance will emphasize that their appropriate use and the characterization of
uncertainties and limitations surrounding the results are the responsibility of the user.
1.4 PHASING TRIM INTO OAQPS' SET OF MODELING TOOLS
As mentioned earlier, TRIM is intended to support assessment activities for both the
criteria and hazardous air pollutant programs of OAQPS As a result of the greater level of effort
expended by the Agency on assessment activities for criteria air pollutants, these activities are
generally more widely known To improve the public understanding of the hazardous air
pollutant (or air toxics) program, the Agency published an overview of the air toxics program in
July 1999 (U S EPA 1999e) Air toxics assessment activities (National Air Toxics Assessment,
NOVEMBER 1999
1-12
TRlM.Expo TSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
or NAT A) are described as one of the program's key components.6 The NATA includes both
national- and local-scale activities. The TRIM system is intended to provide tools in support of
local-scale assessment activities, including multimedia analyses.
One of the Agency's most immediate
needs for TRIM comes in the Residual Risk
Program, in which there are statutory
deadlines within the next two to nine years for
risk-based emissions standards decisions. As
described in the Residual Risk Report to
Congress (U.S EPA 1999f), TRIM is
intended to improve upon the Agency's ability
to perform multipathway human health risk
,,.,., . associated with air emissions of a criteria
assessments and ecological risk assessments air pollutant (e.g.. ozone) or one or several
for HAPs with the potential for multimedia
environmental distribution Another
important upcoming use for TRIM is in
exposure assessment in support of the review TRIM R'
of the ozone NAAQS The TRIM.Expo and
EXAMPLES OF TRIM APPLICATIONS
A human health or ecological assessment
of multimedia, multipathway risks
associated with mercury emissions from
one or several local sources could be
performed using all three modules in the
TRIM system.
An assessment of human health risks
volatile HAPs in a metropolitan area could
be developed using an external air model or
ambient concentration data from fixed-site
monitors coupled with TRIM.Expo and
TRIM Risk modules augmented with external
air quality monitoring data and models are
intended to support this type of criteria pollutant assessment as well as risk assessments for non-
multimedia HAPs
Consistent with the phased plan of TRIM development, the application of TRIM will also
be initiated in a phased approach With the further development of the TRIM modules in 2000
and 2001, EPA will begin to use the modules to contribute to or support CAA exposure and risk
assessments These initial applications also will contribute to model evaluation The earliest
TRIM activities are expected to include the use of TRIM FaTE side-by-side (at a comparable
level of detail) with the existing multimedia methodology7 in risk assessments of certain
multimedia HAPs (e.g., mercury) under the Residual Risk Program. As TRIM Expo is developed
6 Within the air toxics program, these activities are intended to help EPA identify areas of concern (e g.,
pollutants, locations, or sources), characterize nsks, and track progress toward meeting the Agency's overall air
toxics program goals, as well as the risk-based goals of the various activities and initiatives within the program,
such as residual nsk assessments and the Integrated Urban Air Toxics Strategy More specifically, NATA activities
include expansion of air toxics monitoring, improvements and periodic updates to emissions inventories, national-
and local-scale air quality modeling, multimedia and exposure modeling (including modeling that considers
stationary and mobile sources), continued research on health effects of and exposures to both ambient and indoor
air. and use and improvement of exposure and assessment tools. These activities are intended to provide the
Agena with improved characterizations of air toxics risk and of risk reductions resulting from emissions control
standards and initiatives for both stationary and mobile source programs.
" In support of the Mercury Report to Congress (U.S. EPA 1991 a) and the Study of Hazardous A ir
Pollutant Emissions from Electric Utility Steam Generating Units Final Report to Congress (U.S. EPA 1998d),
the Agency relied upon the Indirect Exposure Methodology, which has recently been updated and is now termed
the Multiple Pathways of Exposure methodology (U.S. EPA 1999h). This methodology is being used in initial
assessment activities for the Residual Risk Program (U.S. EPA 1999f).
NOVEMBER 1999 Ti3 TRIM.ExpoTSD (DRAFT)
-------
CHAPTER 1
INTRODUCTION
to accommodate inhalation modeling of HAPs and after it has undergone testing, OAQPS plans to
initially run it side-by-side (at a comparable level of detail) with EPA's existing inhalation
exposure model, HEM (Human Exposure Model (U S. EPA 1986)). When TRIM.Risk has been
completed, it will be used, as appropriate, in both criteria and hazardous air pollutant risk
assessments
In later years, OAQPS intends to use TRIM and the TRIM modules in a variety of
activities including (1) residual risk assessments using TRIM.FaTE, TRIM.Expo, and TRIM.Risk,
in combinations appropriate to the environmental distribution characteristics of the HAPs being
assessed, (2) urban scale assessments on case study cities as part of the Integrated Urban Air
Toxics Strategy, and (3) exposure and risk assessments of criteria air pollutants (e.g., ozone,
carbon monoxide) in support of NAAQS reviews
NOVEMBER 1999 1-14 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM.Expo: GENERAL OVERVIEW AND BACKGROUND
2. TRIM.Expo: GENERAL OVERVIEW AND BACKGROUND
Human exposures to pollutants can result from contact with contaminated air, water, soils,
and food, as well as with drugs and consumer products. Exposures may be dominated by contact
with a single medium, or concurrent contacts with multiple media may be significant The nature
and extent of such exposures depend largely on two things: (1) human factors and (2) the
concentrations of a pollutant in the exposure media. Human factors include all behavioral,
sociological, and physiological characteristics of an individual or cohort (i.e., a group of people
within a population that can be aggregated because the variation in exposure within the group is
much less than the group-to-group variation across the population) that directly or indirectly
affect a person's contact with the substances of concern. Important behavioral factors are contact
rates with food, air, water, and soils. Activity patterns, which are defined by an individual's or
cohort's allocation of time spent in different activities at various locations, are also significant
because they directly affect the magnitude of exposures to substances present in different indoor
and outdoor environments Information on activity
patterns are taken from measured data collected
during field and telephone surveys of individuals'
daily activities, the amount of time spent engaged in
those activities, and the locations where the
activities occur
EXPOSURE
The contact between a target organism
and a pollutant at the outer boundary of
the organism. Exposure may be
quantified as the amount of pollutant
available at the boundary of the receptor
organism per specified time period.
From an exposure assessment standpoint,
the principal goal is to estimate or measure
exposure as a function of both the relevant human
factors and the measured or estimated pollutant concentrations in the contact, or exposure, media.
2.1 RATIONALE AND NEED FOR TRIM.Expo
The models currently being used by OAQPS for estimating human exposure to criteria air
pollutants and HAPs do not include multimedia exposures Furthermore, the models that estimate
exposures to HAPs do not adequately estimate the spatial and temporal patterns of exposures for
all of the HAPs listed in section 112(b) of the CAA In addition, a review of currently available
exposure models and modeling systems revealed that no single model or modeling framework
meets the needs of OAQPS, or could function effectively by itself as part of the TRIM modeling
system for estimating multimedia exposures for a population. Most models are constrained by the
types of media and environmental processes that can be addressed. Furthermore, no model was
identified that addresses the broad range of pollutants and environmental fate and transport
processes that are anticipated to exist for HAPs and criteria air pollutants. Therefore, OAQPS
concluded that the currently available exposure models and modeling frameworks are not
sufficiently integrated multimedia systems that can provide the temporal and spatial resolution
needed for estimating human exposures.
NOVEMBER 1999 2-1 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo GENERAL OVERVIEW AND BACKGROUND
EXPOSURE ASSESSMENT
Measurement or estimation of the magnitude, frequency, duration,
and route of exposure of biological organisms to pollutants in the
environment for a specified time period. An exposure assessment
also describes the nature of exposure and the size and nature of
the exposed populations.
The TRIM exposure
assessment process relates
pollutant concentrations in
the larger environmental
media to pollutant
concentrations in the
immediate exposure media
with which a human
population has direct contact TRIM Expo simulates the movement of an individual and/or
cohorts according to activity patterns through locations (called microenvironments) in a defined
physical or political region (i.e., exposure districts). The movement of individuals or cohorts
coincides with pollutants at varying concentrations. This creates the potential for contact between
individuals or cohorts and pollutants, thus allowing the estimation of exposures of various
individuals and cohorts within the population to the pollutants of interest
TRIM'S EXPOSURE ASSESSMENT PROCESS
TRIM.Expo relates pollutant concentrations in the larger
environmental media to pollutant concentrations in the
immediate exposure media with which a human population has
direct contact.
While OAQPS
supports and recognizes the
value of collecting pollutant
monitoring data for a variety
of tasks (e.g., assessment of
trends, determination of
attainment of standards for
criteria pollutants, evaluation
of exposure models), exposure modeling is clearly needed for several reasons. First, it is very
difficult to monitor the exposure of humans to low concentration mixtures of a large number of
environmental pollutants Direct monitoring of exposure (i.e., personal monitoring) has been
carried out for a number of airborne pollutants, but only limited direct monitoring studies have
been used to assess ingestion and dermal exposures Furthermore, direct monitoring is a resource
intensive process This is especially true for large populations, where the planning, conduct, and
evaluation of such direct monitoring studies can be a lengthy and costly process
In addition, the use of modeling is required or preferred to address the following
regulatory assessment needs
Consideration of hypothetical situations. Modeling can address situations such as
impacts of planned facilities, proposed controls on existing facilities, exposure upon
attainment of ambient standards, and accidental releases
Temporal flexibility. Modeling can be performed for a future time period
Source attribution. Individual sources and/or source categories can be tracked
throughout the modeling process to yield estimates of the relative contribution or
importance of each source or source category to overall exposures.
Inclusion of more chemical species. Personal monitoring techniques do not exist for all
pollutants at the present time
NOVEMBER 1999
2-2
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 2
TRJM.Expo GENERAL OVERVIEW AND BACKGROUND
Representation of long-term conditions. Modeling can address long-term time scales
needed for assessment of chronic exposures. Personal monitoring studies are generally
short-term studies, since wearing the monitoring equipment is too burdensome for
long-term studies to be practical
The extent to which modeled exposure estimates would differ between the currently
available models and a truly coupled source-to-dose, multimedia modeling system, such as TRIM,
is unknown However, models that are not fully coupled have long been considered to lack
scientific credibility. Therefore, OAQPS has determined that it is necessary to undertake efforts
to develop a truly coupled multimedia exposure modeling framework. Figure 2-1 illustrates the
selected modeling features of TRIM Expo Because TRIM Expo will be developed using a
phased approach, future development of the model may include model components in addition to
those shown in Figure 2-1
NOVEMBER 1999 2-3 TRIM.Expo TSD (DRAFT)
-------
o
c.
H
CM
O
ฃ
.8f
u
e
o
U
c
z
I
o
CD
D
w
a:
u
tu
O
M
4^
a
+j
O
Q
en
2
X
s
f-
ซ, ,
c
'ro
0
to
.t:
2
sgetables/Fi
>
o
0 1
O o>
O 0)
u. ฃ
home
commercia
ro
^
o
rish
- recreati
c/>
airy product
o
"ro
o
- comme
o
1
ฃ
D
O
Q.
ro
commercia
o
E
to
CO
ff
JO
"ro
o
- comme
ฃ o>
I |
S.lsis
*-2coipajfc ,_5c
30)งซงt;ฃ ajฐi
Q- ฃ I & I & O t S 1 "S
_ O ' ' < S w
o:
w
CQ
5"
i
C
-------
CHAPTER 2
TRIM.Expo- GENERAL OVERVIEW AND BACKGROUND
2.2 IMPORTANT DEFINITIONS
To enhance understanding throughout this document, the following discussions define
various terms relevant to exposure modeling. These definitions are based primarily on the
guidelines published by EPA (U.S. EPA 1992a).
2.2.1 BASIC DEFINITIONS RELATED TO EXPOSURE
BASIC DEFINITIONS RELATED TO EXPOSURE
Environmental media Components of the physical environment that carry a pollutant, and through
which pollutants can move and reach organisms (e.g., ambient air, ground
water, surface water, surface soil, root zone soil, vadose zone soil, and
several classes of vegetation).
Exposure
Exposure district
Exposure factor
Exposure medium
Exposure pathway
Exposure route
The contact between a target organism and a substance at the outer
boundary of an organism. Exposure may be quantified as the amount of
substance available at the boundary of the receptor organism per specified
time period. For inhalation, exposure over a period of time can be
represented by a time-dependent profile of the exposure concentration.
A geographic location within a defined physical or political region, where
there is potential contact between an organism and a pollutant, and for which
environmental media concentrations have been estimated either through
modeling or measurement.
A normalizing or standardizing factor used in an exposure assessment as a
surrogate for specific information that is not available for a particular subject,
cohort, or demographic group. These factors often are drawn from a
distribution or range of data; see, for example, EPA's Exposure Factors
Handbook (U.S EPA1997b).
The part of the physical environment that surrounds or contacts organisms at
the time of an exposure. The exposure media in TRIM.Expo include: outdoor
air, indoor air (multiple microenvironments), tap water, home-grown food,
locally-produced food, prepared food, breast milk, house dust, soil, swimming
pools, and other recreational surface water.
The physical course a pollutant takes from the source to the organism
exposed. An exposure pathway describes a unique mechanism by which an
individual or population is exposed to pollutants or physical agents at, or
originating from, a site. Each exposure pathway includes a source, or
release from a source, an exposure point, and an exposure route. If the
exposure point differs from the source, a transport/exposure medium (such
as air) or media (in cases of intermedia transport, such as water to air) is also
included
The way a pollutant or physical agent comes in contact with an organism
(e.g., inhalation, ingestion, dermal contact).
NOVEMBER 1999
2-5
TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo. GENERAL OVERVIEW AND BACKGROUND
2.2.2 BASIC DEFINITIONS RELATED TO DOSE
TRIM.Expo will focus on the estimation of exposure only; estimation of dose, where
appropriate, will need to be addressed by dosimetry modeling. Nevertheless, it is useful to
review some concepts related to dose to create a context for understanding how the outputs of
TRIM.Expo are likely to be used. This section contains definitions for several of these terms and
clarifies the terminology used in this document. Figure 2-2 illustrates the relationships among
many of these terms for the inhalation, ingestion, and dermal contact routes.
Absorbed dose
Applied dose
Average Daily Dose
(ADD)
Intake
Lifetime Average
Daily Dose (LADD)
Potential dose
Uptake
BASIC DEFINITIONS RELATED TO DOSE
The amount of a pollutant crossing the exchange boundaries of an organism
after contact, usually expressed as the mass of pollutant absorbed into the
body per unit body mass per unit time, such as mg/kg/d. Absorbed dose is
calculated from the intake and absorption efficiency. For inhalation
exposure, absorbed dose is the amount of material that passes from the gas
volume of the lung into the blood. For ingestion exposure, absorbed dose is
the quantity of pollutant that passes from the gastrointestinal tract across the
gut wall and into the blood stream. For dermal exposure, absorbed dose is
the quantity of material that passes through the stratum corneum into the
living cells of the epidermis and dermis and then into the blood stream. In
some cases, the absorbed dose is referred to as the internal dose.
The amount of a pollutant given in mg/kg/d that comes in contact with the
living tissue of an organism by entering into the lungs, by entering the
gastrointestinal tract, and/or by crossing the stratum corneum into the living
cells of the epidermis. In some experimental designs, the applied dose is
referred to as the administered dose.
Dose rate within a population averaged over body weight and an averaging
time and typically expressed in terms of mg/kg/d.
The process by which a pollutant is physically moved through an opening in
the outer boundary (usually the mouth or nose) of the human body, typically
via ingestion or inhalation.
The average daily dose within a population when the averaging time is the
expected individual lifetime. The LADD is usually expressed in terms of
mg/kg/d. The LADD is used for compounds with carcinogenic or chronic
effects.
An approximation of applied dose that is simply the amount of pollutant in the
food or water ingested, air inhaled, or material applied to skin. The potential
dose for ingestion and inhalation is analogous to the administered dose in a
dose-response experiment. For the dermal route, the potential dose is the
amount of pollutant applied or the amount of pollutant in the medium applied
to skin.
The process by which a substance crosses an absorption barrier and is absorbed
into the body. Although the chemical is often contained in a carrier medium, the
medium itself is typically not absorbed at the same rate as the chemical.
NOVEMBER 1999
2-6
TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo GENERAL OVERVIEW AND BACKGROUND
Figure 2-2
Illustration of the Relationships among Exposure and Dose
for the Inhalation, Ingestion, and Dermal Contact Routes
A. Ingestion and inhalation routes
Mnuปh nr Pharmacokinetic
Mouth or Biomembrane orocesses
nose processes
Exposure
media (air,
water, food, Intake
etc.)
Potential ^- ApP"
dose | do
Distribution,
! metabolism,
Uptake accumulation f^ ^\
ed ! i 1 Organ(s) I
se ! Absorbed or V^ "/
| internal dose ^-. ^^
Biologically-
Epithelial lining of the effective dose
lungs or Gl tract
B. Dermal contact route
Biomemb
n
Exposure j
medium
(soil, water)
I^m
Outer layer
Pharmacokinetic
rane processes
Distribution,
Uptake metabolism, and ^, -^^
accumulation /^ ^\
Absorbed or V Organ(s) J
internal dose N. ^/
Biologically-
, . . effective dose
of skin
(stratum corneum)
NOVEMBER 1999
2-7
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo GENERAL OVERVIEW AND BACKGROUND
2.2.3 OTHER IMPORTANT TRIM.Expo DEFINITIONS
Figure 2-3 illustrates the interconnected nature of the relationships among location
(districts), environmental media, microenvironments, intermedia transfers, exposure media, and
cohorts as defined in this section.
Cohort
Intermedia transfer
Microenvironment
OTHER IMPORTANT TRIM.Expo DEFINITIONS
A group of people within a population with the same demographic variables
who are assumed to have similar exposures. Individuals within a cohort can
be aggregated for exposure assessment purposes because the variation in
exposure within the cohort is much less than the cohort-to-cohort variation
across the broader population.
An algorithm for "linking" the environmental media with the
microenvironmental media that exposed individuals occupy (e.g., air
compartments) or the exposure media with which they come in contact (e.g.,
air, water, food, soil). An intermedia transfer algorithm relates the pollutant
concentration in a microenvironmental medium to the concentration in an
ambient environmental medium that provides pollutant inputs to that
microenvironment
A defined space in which human contact with an environmental pollutant
takes place and which can be treated as a well-characterized, relatively
homogeneous location with respect to pollutant concentrations for a
specified time period. Microenvironments include spaces within buildings
(e.g., rooms, household volumes, restaurants, schools), spaces inside
vehicles (e.g., cars, buses, trains), and other spaces within which humans
have contact with environmental pollutants (e.g., swimming pools, bathtubs,
lakes, rivers)
NOVEMBER 1999
2-8
TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM.Expo GENERAL OVERVIEW AND BACKGROUND
Figure 2-3
Relationships among Districts, Environmental Media, Microenvironments,
Intermedia Transfers, Exposure Media, and Cohorts
2.3 APPROACH USED IN DEVELOPING TRIM.Expo
Based on a review of currently available exposure models and modeling systems
(discussed in Chapter 3 of this report), no single model was identified that would exclusively
satisfy all of the modeling features and requirements that TRIM.Expo will satisfy. However,
several components of existing models were found to meet certain requirements of TRIM.Expo,
and these components will be adapted for use in TRIM.Expo. For example, EPA is developing
the Air Pollutant Exposure Model (APEX), which is based on EPA's human exposure models,
probabilistic NAAQS Exposure Models (pNEM), and Hazardous Air Pollutant Exposure Model
(HAPEM4). The APEX framework will be adapted to address both short- and long-term
exposures for the inhalation pathways modeled by TRIM.Expo. It also includes the
NOVEMBER 1999 2^9TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo GENERAL OVERVIEW AND BACKGROUND
incorporation of activity patterns to track cohorts and individuals as they move among exposure
media. The pNEM model also includes a mass balance treatment of the relationship between
environmental medium (i.e., outdoor air) and exposure medium (i.e., indoor air), as well as the
characterization of uncertainty and variability.
For the ingestion pathways, exposure algorithms from the California Total Exposure
Model for Hazardous Waste Sites (CalTOX) will be incorporated into TRIM. Expo. The
CalTOX framework is capable of modeling multimedia transport and transformation of
pollutants, and multipathway exposure for humans.
For the first prototype of TRIM.Expo, the exposure routes will be limited to inhalation
and ingestion. However, TRIM.Expo will eventually be capable of modeling exposures from all
three main routes of environmental exposure: inhalation, ingestion, and dermal contact.
2.4 TAXONOMY OF EXPOSURE ATTRIBUTES FOR MULTIMEDIA
POLLUTANTS
Despite the large amount of data and the numerous variables used in an exposure
assessment, a relatively small subset of these variables actually significantly influences the
estimated exposures (Morgan and Henrion 1990). However, little is known about how to isolate
this defining set of variables and data. One potential method of sorting and organizing the
complex web of exposure information that is collected during a modeling study is to develop a
taxonomy of exposure-related questions that an exposure model is expected to answer. The
purpose of this taxonomy is to first identify the relevant characteristics or properties as building
blocks, and then rank the attributes of exposure that are important to both analysts and decision-
makers. The taxonomy of exposure-related questions derived during TRIM.Expo's development
is presented in the following sections.
2.4.1 EXPOSURE CHARACTERIZATION PROCESS AND EXPOSURE
ATTRIBUTES
Most of the inhalation pollutant exposures addressed by TRIM.Expo are location-specific
and the individual's or cohort's locations and activities need to be tracked over time and space.
Accounting for time and space is an important factor in determining exposure because of the
spatial and temporal differences in pollutant concentrations among different exposure media. A
log of the time- and activity-specific locations of individuals or cohorts and the time-specific
concentrations of pollutants in the relevant exposure media is needed to create such an exposure
characterization process. A critical issue in the exposure characterization process is to identify
appropriate and transparent methods to combine pollutant concentration data with activity
tracking information to assess both short- and long-term exposures. To develop the exposure
characterization process for TRIM.Expo, the following attributes that define an exposure event
(i.e., human activities that bring people in contact with one or more pollutants in a
microenvironment) were identified:
Route of exposure;
Temporal and spatial scale of the pollutant concentration;
NOVEMBER 1999 2-10 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM.Expo GENERAL OVERVIEW AND BACKGROUND
The seasonality of human activities and pollutant concentrations;
Time scale of the health effects;
Duration of the exposure event;
Contributing environmental media;
Exposure media; and
Demographic characteristics of the exposed individual/cohort (e.g., age, gender).
As shown in Figure 2-2, the route of exposure refers to the way the pollutant can enter the
receptor during the exposure event (i.e., inhalation, ingestion, dermal contact) and lead to actual
absorption of the pollutant into the body (referred to as the route of potential uptake, i.e., the
process by which a pollutant crosses an absorption barrier and is absorbed into the body). The
health effects resulting from an exposure may vary significantly among these three routes of
exposure. Moreover, both the exposure media and exposure activity tend to be strongly
associated with the route of potential uptake. For example, air is associated with the inhalation
route, and the inhalation rate varies significantly with the type of activity. Water, food, and soil
are associated with the ingestion route and with eating and hand-to-mouth related activities.
Pollutant concentrations over time and space are needed to construct an exposure event.
If a pollutant shows little spatial variation in concentration over a large region, even if there is
time variation in that region, there is little need for large numbers of separate geographic regions
in an assessment. Similarly, for pollutant concentrations that do not vary significantly in time,
even if they show large spatial variation, using longer time steps may be possible than that
needed for a pollutant whose concentrations vary more greatly with time. TRIM.Expo will be
flexible with regard to level of spatial resolution.
The interplay between the periodicity of human activities and pollutant levels is an
important attribute for tracking by the exposure-event model. The often cyclic nature of human
activities and pollutant levels is typically expressed in terms of seasonality in exposure modeling.
Seasonality also affects people's habits such as whether they leave windows open or closed.
This has a strong influence on the ability of outdoor pollutants to penetrate indoors. Another
important attribute related to time that can affect exposure levels is the difference in activity
patterns generally observed between weekdays and weekends. A person's activities, particularly
those who work, are usually significantly different on weekends than they are on weekdays. It is
important for an exposure model to distinguish between weekdays and weekends when
constructing exposure-event sequences. TRIM.Expo will retain this information so that it is
available to an analyst performing an exposure assessment. It should also be noted that in
addition to a weekday/weekend effect in activity patterns, the emissions of many pollutants have
a discernable weekday/weekend difference. The combination of effects caused by the day of the
week, particularly weekdays versus weekends, makes this an important factor in exposure
assessments.
The time scale of exposure associated with health effects for a particular pollutant also
strongly affects the temporal resolution required of the exposure event-model. For some
pollutants, including most of the criteria air pollutants and those HAPs associated with health
effects due to short-term exposures, exposures may need to be assessed for periods as short as
one hour or less. For many HAPs, only long-term cumulative exposure may need to be
NOVEMBER 1999 2A\ TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo GENERAL OVERVIEW AND BACKGROUND
characterized. TRIM.Expo will have the capability of aggregating exposures up from small time
increments so that the subsequent time profile of exposure can match the time pattern or period
of concern from a health perspective.
The durations of the exposure events and human activities are also important
considerations in the structure of the exposure-event model. Other factors that affect the
structure of the exposure-event model are the demographic characteristics of an exposed
individual or cohort, such as age or gender, that may influence both the activity pattern and the
health response to exposure. Proximity to particular emission sources, or health status, may also
be important factors that affect the structure of the exposure event.
Cohorts are subsets of a population grouped so that the variation of exposure within a
cohort is much lower than the variation between/among cohorts. The reason for using this
approach is because the available data are not adequate to estimate the exposure of each
individual in a population. Therefore, information about people who are expected to have similar
exposures is aggregated together to make more efficient use of the limited data. The cohorts are
assumed to contain people with exposures that can be characterized by the same probability
distributions for key characteristics. The demographic variables used to describe a cohort are
selected to minimize the differences between individuals within the cohort. The model selects an
individual from the appropriate cohort and uses that individual's activity pattern data to create an
exposure-event sequence for that day. The exposure event sequence represents the exposure
pattern for the entire cohort. TRIM.Expo account for within cohort variability through multiple
runs of the model for the exposure duration under study. As new statistical techniques are
developed, future versions of TRIM.Expo will be modified to use the best available approaches
for characterizing time/activity data. At the present time, however, the current method of using
cohorts is a useful technique for modeling the exposures of a large population in the absence of
complete time/activity pattern data.
TRIM.Expo is being designed to allow flexibility in the user's ability to select a cohort's
characteristics. The demographic variables (e.g., age, gender, work status) that characterize a
cohort can be modified by a user of TRIM.Expo providing that the appropriate data are available.
The level of stratification in the identification of the cohorts' characteristics depends on the
particular problem being addressed and on the availability of data. Typically, input data for the
demographic variables of a cohort are taken from census data. However, TRIM.Expo can be run
using data supplied by the analyst as long as it is formatted correctly and provides sufficient
information to execute the model. Hence, the cohorts' characteristics can be chosen for
individualized studies on a site-specific or case-specific basis.
2.4.2 DIMENSIONS OF THE EXPOSURE ASSESSMENT PROBLEM
The exposure attributes noted above were organized into a set of key exposure
dimensions. The three most important dimensions of the exposure assessment problem were
determined to be the (1) route of exposure, (2) time scale of an exposure event relevant to the
pollutant's associated effects, and (3) degree of location dependence (i.e., dependence of
exposure on the location of the exposed subject). Addressing these three dimensions has a
NOVEMBER 1999 2-12 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 2
TRIM.Expo. GENERAL OVERVIEW AND BACKGROUND
significant impact on the structure of the exposure model (e.g., on the exposure media included,
the degree of spatial resolution, and the level of temporal and spatial aggregation).
For example, consider a model used to assess inhalation exposures to pollutants with
health effects that depend on the number and duration of contacts above some threshold
concentration versus an exposure model used to assess ingestion contact with a pollutant that has
health effects that depend primarily on a lifetime cumulative intake. The former model requires
a compilation of short-term exposure events and must provide relatively detailed information on
the location of the exposed individual. For the latter model, different temporal and spatial detail
is required about the exposed individuals or cohorts (i.e., a detailed time/location profile of the
exposed cohort or individual is not as crucial as information on the location of the exposed
cohort's/individual's food or water supply and the cumulative intake of food or water from a
specific supply).
The primary time scales for exposure assessment models vary from short-term resolution
(e.g., minutes to hours and days) to long-term resolution (e.g., days to months and years). Short-
term resolution allows one to assess both cumulative intake as well as the number and duration
of peak exposure events. Long-term resolution allows primarily for the assessment of
cumulative intake. The quantitative distinction between short-term and long-term depends to
some extent on the pharmacokinetics (i.e., uptake and distribution) and toxicokinetics of the
pollutants being considered.
Location-dependence specifies the level of detail required for the time-activity budget of
an exposed individual. For example, to address inhalation exposures where pollutant
concentrations vary significantly among several districts in which the exposed cohort or
individual lives and differ strongly between indoor and outdoor microenvironments, location-
dependence is high. However, if the properties of the pollutant are such that concentrations are
similar in almost all microenvironments, then location-dependence is lower. For instance, for
ingestion exposures to a pollutant in ground water that is distributed throughout a region, the
location of the exposed cohort or individual is much less important than the source of the
cohort's drinking water.
The above set of attributes gives rise to a broad range of exposure situations, such as
short-term inhalation exposure with strong location dependence, long-term ingestion exposure
with weak location-dependence, and short-term dermal contact exposure. The general exposure-
event function used in TRIM.Expo can be adapted across the broad range of problems defined by
these attributes. In some situations, combining two or more sets of exposure model attributes
may be necessary (i.e., combining long-term ingestion exposures that are weakly location-
dependent with short-term inhalation exposures that are strongly location-dependent).
TRIM.Expo will be designed to combine exposure model attributes, where possible.
2.5 TRIM EXPOSURE-EVENT CONCEPT
As stated previously, human exposures to pollutants may be dominated by contacts with a
single medium, orjconcurrent contacts with multiple exposure media may be significant. The
nature and extent of such exposures is mainly influenced by the pollutant concentrations in the
NOVEMBER 1999 2^3 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 2
TRIM Expo GENERAL OVERVIEW AND BACKGROUND
exposure media and human factors (i.e., behavioral, sociological, and physiological
characteristics of an individual or cohort that directly or indirectly affect contact with pollutants).
Therefore, from an exposure assessment standpoint the principal challenge is to estimate or
measure the individual or cohort exposure as a function of both the relevant human factors and
the pollutant concentrations (measured and/or estimated) in the exposure media.
TRIM.Expo is built around the concept of simulating a series of exposure events.
Exposure events are human activities that bring people in contact with a contaminated exposure
medium within a specified microenvironment at a given geographic location for a specified
period of time. In TRIM.Expo, exposure of each individual or cohort is determined by a
sequence of exposure events specific to the individual or cohort. The exposure-event sequence is
a chronologically ordered series of events that identifies the locations and amount of time spent
in those locations. Each exposure-event sequence consists of a series of events with durations
ranging from one to 60 minutes. Each exposure event assigns the individual or cohort to a
particular combination of exposure district, microenvironment, and activity. To construct
exposure events, an individual or cohort is linked with a series of time-specific activities and the
exposure districts and microenvironments associated with those activities. The following
important attributes of an exposure event are used to estimate the corresponding exposure
concentrations and potential doses:
Pollutant concentration in an environmental medium (e.g., ambient air, surface water);
Any significant intermedia transfer from environmental media to the exposure medium,
(e.g., from soil to house dust to air in an indoor microenvironment);
Pollutant concentration in an exposure medium (e.g., personal air, tap water);
Duration of contact with the exposure medium; and
Time scale of interest.
NOVEMBER 1999 2-14 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 3
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRJM.Expo
3. SUMMARY REVIEW OF EXISTING EXPOSURE MODELS
AND RATIONALE FOR DEVELOPING TRIM.Expo
This chapter provides a review of current exposure modeling approaches and an overview
of several existing and emerging exposure assessment models. The exposure models and
modeling frameworks described in this chapter are critically compared and their respective
strengths and weaknesses are assessed. A more detailed comparison of the features for each of
the different exposure models identified is provided in Appendix B.
This review revealed that none of the models described here adequately meets the
exposure modeling needs of OAQPS (see Section 1.1 for a discussion of the needs of OAQPS).
The review in this chapter highlights the unique features included in TRIM.Expo for meeting
OAQPS' modeling needs.
3.1 RATIONALE FOR DEVELOPING TRIM.Expo
Current models used by OAQPS for estimating human exposure to criteria and hazardous
air pollutants do not include multimedia exposures. Furthermore, the models currently in use for
estimating exposures to HAPs do not adequately estimate the spatial and temporal patterns of
exposures for all of the HAPs listed in section 112(b) of the CAA. Adopting or integrating
existing models into a framework that meets OAQPS' needs represents the most cost-effective
means for developing the tools needed to support the regulatory decision-making activities
related to hazardous and criteria air pollutants.
Based on the review of currently available exposure models and modeling systems, there
is no single model or modeling framework that meets the needs of OAQPS, nor any that could
function effectively by itself as part of the TRIM modeling system for estimating multimedia
exposures for a population. Most models are limited in the type of media and environmental
processes that they are capable of addressing. No model currently exists that addresses the broad
range of pollutants and environmental fate and transport processes that are anticipated to be
encountered by OAQPS and other stakeholders when evaluating the risks from the multitude of
hazardous and criteria air pollutants.
To summarize, none of the currently available exposure models is a sufficiently
integrated multimedia model that provides the temporal and spatial resolution needed for
estimating human exposures. It is not known to what extent modeled exposure estimates would
differ between the currently available models and a truly integrated multimedia exposure model,
such as TRIM.Expo. However, models that are not fully coupled have long been considered to
lack scientific credibility. Therefore, OAQPS has determined that it is necessary to undertake
efforts to develop a truly coupled multimedia exposure modeling framework.
3.2 REQUIRED ATTRIBUTES OF THE TRIM.Expo MODULE
In addition to the five features that are required of the TRIM Exposure-Event module
(listed in Section 2.1), OAQPS determined that the module must also (1) address varying time-
NOVEMBER 1999 3^1 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 3
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
steps (i.e., one hour or greater) and provide sufficient spatial detail at varying scales; (2) have the
"transparency" needed to be practical for a large and diverse group of users; (3) be modular in
design; and (4) be easily accessible.
A key element in the development of TRIM.Expo is the need for the exposure model
system or framework to be modular in design. A modular design is one that partitions the
various algorithms necessary for evaluating the different aspects of an exposure assessment into
generally discrete packages, or modules, which are able to interact with each other. By creating
an exposure framework that is modular, only those model components necessary for evaluating
particular aspects of the exposure assessment and/or endpoints of interest need to be used for a
particular application.
OAQPS has decided to separately characterize uncertainty and variability on a selective
basis. In concordance with EPA's probabilistic modeling guidance (U.S. EPA 1997c), a staged
approach (as described in Section 4.2.9.3) will be used in characterizing uncertainty and
variability (i.e., rather than attempting to characterize uncertainty and variability for all
parameters, sensitivity analyses will be used to identify a limited number of critical parameters
that most influence the exposure outcomes and, thus, will be subjected to further analysis).
These parameters will be examined in more detail to determine whether it is appropriate to
separately characterize uncertainty and variability based on available information. For some
parameters, such as body v/eight, there are sufficient data to support the explicit characterization
of variability. However, for other parameters where data may be insufficient to support the
separate characterization of uncertainty and variability, a distribution will be defined to reflect
overall parameter uncertainty, including inherent variability.
3.3 OVERVIEW OF CURRENT MODELS AND MODELING
APPROACHES
Exposure modeling approaches have long been based on physical principles. They were
developed using a concise physical interpretation of the factors affecting exposures, which were
determined prior to the development of a particular model. Examples of this type of modeling
approach are the NAAQS Exposure Model (NEM) (U.S. EPA 1983) and several different indoor
air quality mass balance models (Nazaroff and Cass 1986, Ryan et al. 1983, Ozkaynak et al.
1982). A major limitation of these models is that they do not capture the full variability of
people's activities as part of the exposure simulation. In addition, uncertainty in the values of the
parameters used to make the estimates is not included.
To overcome these limitations, subsequent exposure models were developed that used a
stochastic approach. This made it possible for estimates of population exposure to be
characterized as distributions rather than point estimates. One of the first models developed to
use this "probabilistic" approach was the Simulation of Human Activities and PollutantExposure,
or SHAPE, model (Ott 1982, Ott 1984, Ott et al. 1988). Shortly afterward, a probabilistic
version of NEM was developed. The model (there were actually several pollutant-specific
versions) was referred to as the probabilistic NAAQS Exposure Model, or pNEM (McCurdy and
Johnson 1989). While it would be difficult to accurately represent the activities of an individual
NOVEMBER 1999 3^2 TRIM.Expo TSD (DRAFT)
-------
CHAPTERS
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
due to day-to-day variation, the general behavior of groups or subsets of the population can be
well represented using stochastic processes. By explicitly including variability and uncertainty
in the models, the effects of the uncertainty on the modeled exposure values can be evaluated.
Many of the early models stressed exposures to industrial and mobile sources and
attempted to account for the variability in those exposures. Some of the earliest work in
exposure modeling attempted to simulate human exposures to lead and carbon monoxide. With
the periodic review and revision of the NAAQS, models were developed to assess the exposures
of the population to the air pollutants for which ambient standards had been established.
Over time, it has become clear that there are important outdoor sources other than large
industrial facilities and mobile sources, that exposures can occur both indoors and outdoors, and
that the sources of the pollutants can likewise be found both indoors and outdoors. Over the past
decade, increasingly sophisticated methodologies have been developed for modeling and
evaluating exposures. However, there is currently no single exposure model that can estimate all
pollutants, all sources, and all routes of exposure. The models that have emerged are largely still
restricted to single-medium exposure assessments. In recent years, emphasis has been placed on
developing modeling frameworks that can assess multimedia, multipollutant human exposures
within a unified framework. This type of approach is relatively new, and a usable framework for
conducting multipathway exposure assessments has yet to emerge. This chapter provides a
general overview of modeling approaches used to date to estimate exposures.
Table 3-1 identifies numerous air quality and exposure models and modeling systems.
More detailed information is given in Appendix B. that are currently publicly available along
with the agency or group who was its major developer. Some exposure models are pioprietary
(/. e., they are not "public domain") and therefore must be purchased from the developer. To
facilitate public access, EPA has decided that no proprietary models will be considered in the
development of TRIM.Expo. Examples of proprietary models include RISK* ASSISTANTฎ and
LifeLine8.
The EPA's Office of Research and Development (ORD) is currently developing a number
of exposure models and modeling systems. The development is on-going and therefore is not
included in Table 3-1 at this time. However, since these models represent a major effort by
ORD, a description of each has been included in Appendix B. Examples of new and continuing
exposure modeling efforts at ORD include the development of the Stochastic Human Exposure
and Dose Simulation (SHEDS) Model (see Section B.7.7). The SHEDS Model is a probabilistic,
physically-based model that simulates aggregate exposure and dose for population cohorts and
multimedia pollutants of interest. Initial applications of the model have assessed children's
exposures to pesticides (SHEDS-Pesticides) and population exposures to PM (SHEDS-PM).
Another effort within ORD is the development of the Modeling ENvironment for TOtal Risk
(MENTOR) project. The objective of the on-going MENTOR project is to develop, apply
through case studies, and evaluate state-of-the-art computational tools, that will support
multipathway, multiscale source-to-dose studies and exposure assessments for a wide range of
environmental pollutants. More detail about MENTOR can be found in Appendix B. The EPA's
ORD is also engaged in the development of the MODELS-3/Multimedia Integrated Modeling
System (MIMS). The MIMS will have capabilities to represent the transport and fate of nutrients
NOVEMBER 1999 3OTRIM.EXPO TSD (DRAFT)
-------
CHAPTER 3
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
and pollutants over multiple scales. The system will provide a computer-based problem solving
environment for testing our understanding of multimedia (atmosphere, land, water)
environmental problems, such as the movement of pollutants through the hydrologic cycle, or the
response of aquatic ecological systems to land-use change.
Table 3-1
Air Quality and Exposure Models and Modeling Systems and Their Developers
Model
Developer
INDOOR AIR EXPOSURE MODELS
INDOOR, EXPOSURE, and RISK
MAVRIQ (Model for Analysis of Volatiles and
Residential Indoor Air Quality)
AMEM (ADL Migration Exposure Model)
CPIEM (California Population Indoor Exposure
Model)
EPA/Office of Research and Development (ORD)
EPA/ORD
EPA/Office of Pollution Prevention and Toxics (OPPT)
California Air Resources Board (CARB)/lndoor Program
INDOOR / OUTDOOR AIR EXPOSURE MODELS
pNEM (probabilistic NAAQS)
HAPEM4 (Hazardous Air Pollutant Exposure Model)
AirPEx (Air Pollution Exposure Model)
HEM (Human Exposure Model)
SHAPE (Simulation of Human Activities and
Pollutant Exposure)
BEAM (Benzene Exposure Assessment Model)
pHAP (probabilistic HAP Exposure Model)
EPA/OAQPS
EPA/OAQPS
National Institute of Public Health and the Environment
(RIVM) [Netherlands]
EPA/OAQPS
EPA/ORD
EPA/ORD
EPA/OAQPS
CONSUMER PRODUCT EXPOSURE MODELS
CONSEXPO (CONSumer EXPOsure Model)
SCIES (Screening Consumer Inhalation Exposure
Software)
DERMAL
MCCEM (Multi-Chamber Concentration and
Exposure Model)
National Institute of Public Health and the Environment
(RIVM) [Netherlands]
EPA/ OPPT/Economics, Exposure, and Technology
Division (EETD)
EPA/OPPT/EETD
EPA/OPPT; updated by EPA/ORD
DIETARY EXPOSURE MODELS
DEPM (Dietary Exposure Potential Model)
EPA/ORD
MULTIMEDIA EXPOSURE MODELS
The Exposure Commitment Method
National Radiological Protection Board (NRPB) [United
Kingdom]
NOVEMBER 1999
3-4
TRIM.Expo ISO (DRAFT)
-------
CHAPTERS
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
Model
Layton et al. (1992) Indoor/Outdoor Air/Soil
Transport Model
CalTOX (California Total Exposure Model for
Hazardous Waste Sites)
MMSOILS (Multimedia Contaminant Fate, Transport,
and Exposure Model)
RESRAD (RESidual RADiation)
USES (Unified System for the Evaluation of
Substances)
BEADS (The Benzene Exposure and Absorbed
Dose Simulation)
DERM (Dermal Exposure Reduction Model)
SCREAM2 (South Coast Risk and Exposure
Assessment Model, Version 2)
Integrated Spatial Multimedia Compartmental Model
(ISMCM)
Developer
U.S. Department of Energy (DOE)
California Environmental Protection Agency/Department
of Toxic Substances Control (DTSC)
EPA/ORD
DOE and Argonne National Laboratory
National Institute of Public Health and the Environment
(RIVM) [Netherlands]
EPA/ORD
Stanford University/Environmental Engineering and
Science Group
South Coast Air Quality Management District
University of California (Los Angeles)/School of
Engineering and Applied Science
EXPOSURE SIMULATION MODEL SYSTEMS
GEMS (Graphical Exposure Modeling System)
THERdbASE (Total Human Exposure Risk database
and Advanced Simulation Environment)
MEPAS (Multimedia Environmental Pollutant
Assessment System)
EPA/OPPT
EPA/ORD
DOE and Battelle Pacific Northwest Laboratory
In general, the models that most closely meet the design goals for TRIM development are
the focus of this chapter. Generally, these include models that are able to calculate short-term
exposures (i.e., 1 hour or shorter in duration), because they can be adapted to evaluate long-term
exposures as well. They may also be able to explicitly treat variability and uncertainty. Other
desirable model attributes are the utilization of a mass balance approach for estimating indoor air
concentrations and the ability to track potential intake rates concurrent with exposure. For
inhalation, this means providing estimates of the respiration rate (also called ventilation or
breathing rate) for various activities. Additional useful features include accounting for indoor air
emission sources and the ability to include geographic mobility (e.g., commuting) in the
exposure simulation.
One model that has many of the desirable attributes is pNEM/CO (Johnson et al. 1992b,
Johnson et al. 1999). Although this model is for a single medium only (air), it already
incorporates nearly all of the features needed for the inhalation component of TRIM.Expo (see
Appendix B, Table B-l). In addition to the criteria listed above, pNEM/CO is well documented
and is already being used by OAQPS as an input to regulatory decision-making. Furthermore,
the 1992 version of pNEM/CO has undergone review by the Clean Air Scientific Advisory
Committee.
NOVEMBER 1999
3-5
TRIM.Expo TSD (DRAFT)
-------
CHAPTERS
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
For modeling the non-inhalation routes of exposure, the CalTOX model (McKone 1993a,
b, c), developed at the Lawrence Berkeley National Laboratory (LBL), already includes many of
the features needed. CalTOX has the ability to calculate multipathway exposures for organic
chemicals and some metals. In addition, the model is stochastic and can quantify the variability
and uncertainty in the exposure calculations. CalTOX, pNEM/CO, and several other models for
estimating non-inhalation and inhalation exposures are discussed in the following sections.
3.3.1 INHALATION EXPOSURE MODELS
Perhaps the largest number of exposure models have been developed to assess the
relationship between chemical releases to outdoor air and human exposure to these pollutants
both indoors and outdoors. The early "indoor/outdoor" exposure models were the first to use
newly collected information on activity patterns and microenvironmental concentrations. They
simulated the microenvironmental concentrations based on empirical data derived from field
measurements.
The EPA played a major role in developing exposure models that addressed the air
pathway. The original purpose of these models was to assist in setting the ambient air quality
standards by estimating the population exposure to air pollutants when alternative air quality
standards were just met. One of the first models developed for this purpose was the NEM. The
NEM was pollutant-specific and included versions for estimating exposures to ozone, carbon
monoxide, and paniculate matter. Later, a stochastic method for randomly selecting values for
important variables was incorporated into the models. These models were referred to as
"'probabilistic" and hence are known as probabilistic NEM, or pNEM. Groups outside of EPA
have used the NEM approach and developed variations of the NEM for specific applications.
These models include SAI/NEM (Hayes et al. 1984, Hayes and Lundberg 1985), REHEX (Winer
et al. 1989, Lurmami et al. 1990, Lurmann et al. 1992), and the Event Probability Exposure
Model (EPEM) (Johnson et al. 1992a). All three of these models are related to the NEM
approach, although significant variations now exist among the models (McCurdy 1994).
The EPA also developed exposure models for specific sources or types of sources. In
1985, EPA's Office of Mobile Sources (OMS) in conjunction with EPA's Office of Research and
Development (ORD) developed a model for estimating human exposure to non-reactive
pollutants emitted by motor vehicles. This model, named the Hazardous Air Pollutant Exposure
Model for Mobile Sources, or HAPEM-MS, is similar in methodology to the pNEM models
(Johnson 1995). However, it differs from pNEM in the averaging time for exposure
concentrations. Instead of the hourly resolution of pNEM, HAPEM-MS aggregates the hourly
exposure concentrations to 3-month averages, because HAPEM-MS is designed to address
exposures to pollutants with carcinogenic and other long-term effects. Subsequently, HAPEM-
MS has been enhanced and now is able to model exposures to numerous air toxics from different
sources through the use of the air dispersion module of the Assessment System for Population
Exposure Nationwide (ASPEN) model (SAI 1999). Given the model's ability to estimate
exposures from different types of sources (i.e., not just mobile sources), the Mobile Sources, or
"MS," designation has been dropped from the model's name. The latest version of the model is
called HAPEM4.
NOVEMBER 1999 3-6 TRIM.Expo TSD (DRAFT)
-------
CHAPTERS
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM Expo
Agencies in other countries have developed exposure models that are specific to their
population. The Dutch, for example, have developed an inhalation exposure model based on the
pNEM approach which is used for estimating exposures of people in the Netherlands (although
the model may be adapted for any location). The model, called the Air Pollution Exposure
Model, or AirPEx (Freijer et al. 1997), works on a personal computer (PC) using Windowsฎ.
The PC platform enhances the accessibility of the model to various stakeholder groups that do
not have extensive programming expertise. Several new exposure models are being designed to
run on PCs, and some existing models, previously run on large machines, are being modified to
run on PCs and via the Internet.
3.3.2 MULTIMEDIA EXPOSURE MODELS
Ingestion is another important route of exposure; however, modeling ingestion exposures
presents a different set of requirements than does inhalation. For example, exposure to a
particular pollutant from a certain food source can occur in a single location or in many places
over time. The actual location where the exposure takes place may not be the same as where the
contamination of the exposure medium occurred. Another difference is the time period for
exposure due to ingestion. There may be long lags between the contamination of the exposure
medium (e.g., food, water, soil) and the time that exposure occurs. Much of the exposure
modeling that has been done for ingestion pathways has been conducted as part of multimedia
modeling efforts. Hence, ingestion exposures discussed in the context of multimedia exposure
models.
Efforts to assess human exposure from multiple media date back to the 1950s when the
need to assess human exposure to global fallout from nuclear testing led rapidly to a framework
that included transport through and transfers among air, soil, surface water, vegetation, and
various paths of the food chain. Efforts to apply such a framework to non-radioactive organic
and inorganic toxic chemicals have been more recent and have not as yet achieved the level of
sophistication that exists in the radioecology field.
The CalTOX program was developed for the California EPA as a set of spreadsheet
models and spreadsheet data sets to assist in assessing human exposures to toxic substances
released in multiple media (McKone 1993a,b,c). CalTOX consists of two component models: a
multimedia transport and transformations model that is based on both conservation of mass and
chemical equilibrium, and a multipathway human exposure model that includes ingestion,
inhalation, and dermal uptake exposure routes. It is a mass balancing model that also includes
the ability to quantify uncertainty and variability. The exposure assessment process consists of
relating pollutant concentrations in the multimedia model compartments to pollutant
concentrations in the media with which a human population has contact (e.g., personal air, tap
water, foods, household dusts/soils).
The Integrated Spatial Multimedia Compartmental Model (ISMCM) has been under
development for the past 15 years. The ISMCM considers all media, biological and non-
biological, in one integrated system. The model includes both spatial and compartmental
modules to account for complex transport of pollutants through the ecosystem. Assuming
NOVEMBER 1999 3-7 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 3
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
conservation of mass, ISMCM predicts transport by using estimates of intermedia transfer
factors. A newer version of ISMCM, called MEND-TOX, is currently under evaluation by EPA.
The Indirect Exposure Methodology (IEM) is a significant current EPA methodology for
multimedia, multipathway transport, fate, and exposure modeling. This methodology identifies
procedures for estimating the indirect (i.e., non-inhalation) human exposures that can result from
the transfer of emitted air pollutants to soil, vegetation, and water bodies. The IEM addresses
exposures for a variety of receptor scenarios (e.g., subsistence fisher) via inhalation, ingestion of
food, water, and soil, and dermal contact. The most up-to-date version of the IEM methodology
is scheduled to be published in late 1999 (U.S. 1999h). The updated documentation no longer
refers to the methodology as IEM; it is now referred to as the Multiple Pathways of Exppsure
(MPE) methodology. Appendix B provides a more detailed discussion of the IEM model.
3.4 STRENGTHS AND LIMITATIONS OF EXISTING MODELS
TRIM development is designed to focus on the processes that have the greatest impact on
pollutant fate and transport and on human exposure. In order to have the same scientific basis as
the rest of the TRIM system, TRIM.Expo needs to incorporate the same attributes, including: (1)
mass conservation to the extent feasible and appropriate; (2) ability to characterize uncertainty
and variability; (3) capability to assess multiple pollutants, multiple media, and multiple
exposure pathways; and (4) ability to perform iterative analyses. Hence, these four design
attributes serve as the basis for critically comparing the strengths and limitations of existing
exposure models.
By assessing the strengths and limitations of publicly available exposure models and
modeling systems in regard to the needs defined for TRIM development, a determination can be
made regarding the features of the various models that may be incorporated into TRIM. Table
B-2 in Appendix B compares the strengths and weaknesses of some of the most commonly used
EPA and non-EPA exposure models. The models in this table are included because they each
have one or more of the desirable attributes identified above needed for TRIM.Expo.
The pNEM/CO and pNEM/O3 (for ozone) models have been used extensively by OAQPS
in its reviews of the CO and ozone NAAQS, respectively (see Table B-3 in Appendix B for a
descriptive overview of many of the features of pNEM/CO). The pNEM/CO uses a stochastic
approach for selecting input variables. This stochastic approach allows both sensitivity and
uncertainty to be incorporated into the model operation. Many of the model's input variables
come from measured data, thereby decreasing the uncertainty associated with the model's
estimates. The pNEM/CO treats human exposure as a time series of the convergence of (1)
human activities occurring in a particular microenvironment and (2) air quality in those
microenvironments. The model is also designed to provide estimates of the intake dose
associated with exposures. The focus on time series modeling and intake dose allows analysts to
produce estimates of the "dose profile" of exposed people (McCurdy 1995). A disadvantage of
the pNEM/CO model in its current form is that it is difficult to execute. The pNEM/CO model,
as with all of the pNEM models, is a single pollutant, single media model.
NOVEMBER 1999 3-8 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 3
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
The CalTOX model (see Table B-4, Appendix B) consists of two main components: a
multimedia transport and transformation model and a multipathway human exposure model. The
multimedia transport and transformation model is based on both the conservation of mass and
chemical equilibrium. The multimedia transport model is a dynamic model that can be used to
assess time-varying concentrations of pollutants introduced initially into the soil or released to
the air, soil, or water. The exposure model has 23 exposure pathways encompassing all three
environmental routes of exposure, which are used to estimate average daily doses within a
human population in the vicinity of a hazardous air pollutant release site. The exposure
assessment process consists of relating pollutant concentrations in the multimedia model
compartments to pollutant concentrations in the media with which a human population has
contact (e.g., personal air, tap water, foods, house dust). This explicitly differentiates the
environmental media pollutant concentrations from the pollutant concentrations in the exposure
media to which humans are exposed. In addition, all input variables are taken from distributions.
The CalTOX model is limited in the extent of the environmental settings for which it can
be applied. For example, it has limited effectiveness for settings where there is a large ratio of
land area to surface water area. In addition, it was developed for a limited range of pollutants
(i.e., non-ionic organic chemicals in a liquid or gaseous state). As a result, CalTOX does not
provide adequate flexibility in the environmental settings or the chemical classes it models.
Also, CalTOX does not allow spatial tracking of a pollutant, hence it is not directly applicable to
the TRIM approach.
HAPEM has undergone many enhancements in recent years. The most recent of these is
the ability of the model to use air quality concentration estimates from the ASPEN. This latest
version of HAPEM is designated HAPEM4 (see Table B-5, Appendix B).. It allows exposure to
population cohorts to be simulated at the census tract level. This is a much finer spatial
resolution than was previously possible. It also means that calculation of population exposures
no longer needs to rely solely on data from fixed-site monitors. This is important for estimating
exposures to HAPs because widespread monitoring networks for these pollutants are not
available.
The HAPEM4 calculates long-term average exposure concentrations in order to address
exposures to pollutants with carcinogenic and other long-term effects. Thus, HAPEM4 does not
preserve the time-sequence of exposure events when sampling from the time/activity database.
This means that information to evaluate possible correlations in exposures to different pollutants
due to activities that are related in time is not preserved. Also, the model does not include any
measures of the ventilation rate associated with an activity, so that there is no ability to calculate
the potential dose received when engaging in various activities.
The IEM has been used by EPA in a variety of applications and has undergone extensive
scientific peer review. The methodology includes fate and transport algorithsms, exposure
pathways, receptor scenarios, and dose algorithms. It also includes procedures for estimating the
indirect (i.e., non-inhalation) human exposures and health risks that can result from the transfer
of pollutants to soil, vegetation, and water bodies.
NOVEMBER 1999 3-9 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 3
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
The IEM is limited, relative to OAQPS's needs, because the methodology, as currently
implemented, can be applied only to pollutants that are emitted to air. While IEM is a significant
current EPA methodology for multimedia, multipathway exposure modeling, it does not fully
satisfy the needs of OAQPS. An important limitation of IEM, relative to the needs of OAQPS, is
that it consists of a one-way process through a series of linked models, using as inputs the annual
average air concentrations and wet and dry deposition values from external air dispersion
modeling. As a result, it is not a truly coupled multimedia model and does not have the ability to
maintain a full mass balance or model "feedback" loops between media or secondary emissions,
nor can it provide a detailed time series estimation of media concentrations and resultant
exposures. The methodology does not provide for the flexibility OAQPS needs in site-specific
applications or in estimating population exposures. Significant site-specific adjustments must be
made to allow for spatially tracking the relationship between concentrations and exposures.
Much of the focus of the methodology is on evaluating specific receptor scenarios (e.g.,
recreational or subsistence fisher) that may be indicative of high-end or average exposures rather
than on modeling the range of exposures within a population (i.e., IEM cannot estimate
population exposure distributions). Appendix B provides more detailed discussion of the IEM
model.
The Integrated Spatial Multimedia Compartmental Model (ISMCM) has been undergoing
development at the University of California's (Los Angeles) School of Engineering and Applied
Science for the last 15 years. The latest version of ISMCM, called MEND-TOX, is currently
under evaluation at EPA's Office of Research and Development. The ISMCM considers all
media in a single integrated system. It includes both spatial and compartmental modules to
account for complex transport of pollutants through the ecosystem. The model is mass
conserving and is able to estimate intermedia transfer factors desirable for TRIM.
An important limitation of ISMCM for use in TRIM.Expo development is the lack of
flexibility in the spatial configuration of the model. The links and compartments in ISMCM are
predetermined, thus limiting its ability to be fully integrated into a system like TRIM. Another
drawback to ISMCM is that it is not structured to incorporate uncertainty and variability directly
into the model outputs.
The South Coast Risk and Exposure Assessment Model (SCREAM2) provides the ability
to model both inhalation and multipathway non-inhalation exposures (see Appendix B, Table B-
6) (Rosenbaum et al. 1994). The model can use both measured and modeled air quality data,
thus increasing the spatial resolution and number of the pollutants being studied. The
SCREAM2 also includes an indoor air model for calculating indoor air concentrations. An
internal submodel, called MULTPATH, calculates population exposures from several non-
inhalation pathways, including food ingestion, water ingestion, and dermal adsorption. The
inhalation exposure module accounts for mobility patterns of the population, indoor-outdoor
exposure concentration differences, and physical exercise levels.
The use of SCREAM2 in the TRIM.Expo framework is limited because it is a
deterministic model, so that input and output data are represented as point estimates rather than
ranges or distributions. This limitation also restricts the model's ability to explicitly characterize
uncertainty. The SCREAM2 framework consists of a one-way process through a series of linked
NOVEMBER 1999 vio TRIM.Expo TSD (DRAFT)
-------
CHAPTERS
SUMMARY REVIEW OF EXISTING EXPOSURE MODELS AND RATIONALE FOR DEVELOPING TRIM.Expo
models, based on annual average air concentrations and wet and dry deposition values from air
dispersion modeling. In contrast, TRIM.Expo will have the capability to report exposure results
for both short-term and long-term averages and will allow for "feedback" loops and secondary
emissions.
The California Population Indoor Exposure Model (CPIEM) was developed to evaluate
indoor exposures for the general California population as well as certain subgroups such as
individuals who may be highly sensitive to indoor pollutants (see Appendix B, Table B-7)
(CARB 1998a). The CPIEM combines indoor-air concentration distributions with Californians'
location and activity information to produce exposure and dose distributions for different types
of indoor environments. This is achieved through a Monte Carlo simulation whereby a number
of location/activity profiles that were collected in Air Resources Board (ARB) studies are
combined with airborne pollutant concentrations for specific types of microenvironments (e.g.,
residences, office buildings).
Concentration distributions for many pollutants and microenvironments are included in
the CPIEM database. However, for pollutants and microenvironments not included in the
database, CPIEM presents two alternatives. The first option is to estimate indoor air
concentration distributions based on distributional information for mass balance parameters such
as indoor source emission rates, building volumes, and air exchange rates. The second option is
for the user to directly specify concentration distributions. The concentration values for a
particular environment are then sampled from the distributions and multiplied by time durations
of the population groups in the environment, based on results from Californians' location/activity
profiles, to calculate time-integrated exposure.
Model limitations for CPIEM include the assumption that concentrations in different
environments on the same day are independent. For example, if outdoor concentrations have a
significant impact on indoor concentrations they may be correlated, so that the assumption of
independence may misrepresent the shape of the exposure/dose distribution. The impact could
be significant, particularly at the upper end of the distribution where the calculated exposures
could be underestimated. Also, since CPIEM uses activity profiles from a limited number of
studies (i.e., only those conducted in California), the results may not represent small subgroups
of the population or population groups in other regions of the U.S.
NOVEMBER 1999 3-11 TRIM.EXPO TSD (DRAFT)
-------
[This page intentionally left blank.]
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
4. DESIGN FRAMEWORK AND CONCEPTUALIZATION OF
TRIM.Expo
The TRIM.Expo module is expected to estimate past and future human exposure patterns
by combining pollutant concentration data with population activity tracking information. The
estimation procedure requires (1) the use of algorithms to organize and manipulate pollutant
concentration data and activity pattern information and (2) a process to define the link between
the multimedia "ambient" environment and the microenvironmental exposure media that
individuals occupy (e.g., air compartments) or contact (e.g., water, food, soil). Pollution
concentration data are required as input into TRIM.Expo and can be derived from monitoring
data, a single media transport model, or a multimedia transport model (e.g., TRIM.FaTE).
The TRIM.Expo module provides a model framework addressing features unique to each
problem and providing a characterization of the uncertainty associated with exposure estimates
for single and multiple media pollutants. This includes a process for predicting concentrations in
exposure media based on both the transfer from the ambient environment and on sources internal
to those microenvironments containing the exposure media. This chapter describes the
TRIM.Expo modeling system and how it was conceptualized to meet the above goals.
In order to characterize aggregate human exposure to a pollutant, each exposure route and
pathway must be considered. For example, a semi-volatile hazardous air pollutant (e.g., aromatic
hydrocarbon) that is released to ambient air can be transported to multiple locations where
exposure may occur. The pollutant can be (1) transported to the indoor or outdoor air
surrounding a human receptor who would then inhale the pollutant; (2) transferred by deposition
and run-off to surface water that supports fish consumed by the human receptor or provides
drinking water to the human receptor; or (3) transferred by deposition to vegetation that is
consumed by the human receptor or to vegetation that is consumed by the animals that supply
meat and milk that is consumed by the human receptor. Each scenario defines a pathway from
the pollutant's'air emission to a receptor's contact with it via an associated route of contact.
Total exposure cannot be estimated until the pathways and routes that account for a substantial
amount of the intake and uptake for a receptor population have been identified.
The TRIM.Expo module is designed to be used by analysts (e.g., modelers) and decision-
makers (e.g., regulators). As part of the initial development of TRIM. Expo, questions and issues
regarding pollutant exposure of concern to analysts and decision-makers were identified.
Some of the more significant questions pertinent to analysts include the following.
Which input properties are the most critical for modeling the movement and persistence
of chemicals in indoor and outdoor environments, and which are of lesser importance, for
estimating human exposures?
How reliable are the ambient and/or microenvironmental concentration data used as input
to the model, and how does the reliability of these concentrations limit the reliability of
the exposure estimate?
NOVEMBER 1999 ~ 4-1 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
Some of the more significant questions pertinent to decision-makers include the following.
For a given pollutant emission control measure or magnitude of reduction in ambient
pollutant levels, how much reduction of exposure (and related health effects) can be
expected?
What is an indicator of exposure that can be estimated? How much of a change can be
estimated in this indicator to provide evidence that a control measure might be effective?
How long is the lag time between a change in pollutant emissions and the estimated
change in the environment or exposure indicator?
How likely are these estimates to be wrong? How uncertain are the quantitative estimates
of exposure reduction and changes in environmental indicators of exposure?
To answer the questions identified above, a taxonomy of exposure questions was
formulated (see Section 2.3). From this, a prioritized set of exposure-model attributes was
selected, resulting in three primary model dimensions.
4.1 EXPOSURE-EVENT MODULE STRUCTURE
Exposure events are activities that bring people in contact with a contaminated exposure
medium in a specified microenvironment within a given exposure district. To construct exposure
events, an individual or a cohort must be linked with a series of time-specific activities and with
the exposure districts and microenvironments associated with those activities. In addition,
pollutant concentrations in each district-microenvironment combination must be defined through
a combination of databases and stochastic process models. This process of constructing exposure
events is illustrated in Figure 4-1. Exposure-event simulations must be able to provide a broad
range of information, including (1) detailed information on the input distributions selected (e.g.,
relative probability of values, fractiles and central tendency, confidence intervals, shape of the
distribution); (2) detailed information on the output distribution (the information here would
follow in a similar fashion to item 1, above); (3) information about the goodness-of-fit of the
data; (4) information about interdependencies and correlations between variables; (5) the number
of times the concentration exceeded specified concentration levels; (6) the average exposure
concentration exceeding some specified level; and (7) the cumulative intake or uptake during a
series of exposure events. The TRIM.Expo module will contain algorithms that determine all of
these parameters. These algorithms will include the basic exposure-event function, the average
exposure concentration, the intake-adjusted average exposure concentration, the intermedia
transfer factor, and the average daily potential dose.
NOVEMBER 1999 4-2 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Figure 4-1
Illustration of an Exposure-Event Simulation
Ambient media concentrations
Ci,alr m.k.t)
Exposure-
Event Function
,M.t) ET
Uncertainty
Variability
Ez,m(t)
Contact medium
Air
Water
Food
Soil
Ambient
zone
Time
Activity
Microenvironment
4.1.1 BASIC EXPOSURE-EVENT FUNCTION
The basic exposure-event function determines the microenvironmental exposure to an
individual or cohort from an exposure medium during time step, /. It defines exposure as the
product of concentration and exposure duration, as illustrated in Figure 4-1 and shown in
Equation 4-1. This equation can also be used to define an exposure concentration in other media,
such as water and food, although it is less intuitive than an exposure of mg/kg-body-weight for
food. For exposure media such as food and water, the potential dose rate is preferred as a basis
for intake estimates (see Section 4.1.5).
Ez,m(t) =
(4-1)
where
E.., = Exposure experienced by person z from exposure medium m during time
step t, given that person z is in exposure district i in microenvironment k
conducting activity / during that time step /. For example, the exposure in
air might be measured in units of mg-hr/m3. Note that the exposure time
need not be a whole time step.
NOVEMBER 1999
4-3
TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Cm = Concentration in exposure medium m (e.g., air, water, soil) in exposure
district /' in microenvironment k associated with activity / during time step
t. Units of measurement for air might be mg/m3, while units of
measurement of food might be mg/kg.
ET.m = Exposure duration of individual or cohort z to exposure medium m in
exposure district / in microenvironment k conducting activity / during time
step /.
z = Individual or cohort.
m = Exposure medium contacted (i.e., air, water, food).
i = Exposure district.
k = Microenvironment in which the exposure occurs (e.g., indoors at home, in
a vehicle, indoors at work).
/ = Activity code that describes what the individual is doing at the time of
exposure (e.g., resting, working, preparing food, cleaning, eating).
4.1.2 EXPOSURE OR POTENTIAL DOSE PROFILES
The time series of concentrations that could potentially result in an exposure comprises a
profile, as illustrated in Figure 4-2. As shown in Equation 4-1, exposure to a pollutant is
calculated as the product of the concentration that a person contacts and the duration of the
contact. This is shown graphically in Figure 4-2. The x-axis shows the time sequence, while the
y-axis shows the concentration of a pollutant. Therefore, the concentration for each instant in
time is given by the curve, C(t). To simplify the discussion, suppose that a person was in a
particular microenvironment from time t,., to t, and that the concentration of the pollutant (given
by c,) remained fairly constant during this time period (to make this example as simple as
possible /, -1,., = Af, is defined as one hour). The shaded rectangle in Figure 4-2 has height c, and
width A/,. The area given by this rectangle is, therefore, c, (A/,) and as stated above and shown in
Equation 4-1 the product of a concentration and a time interval is equal to exposure. Thus, the
area given by the shaded rectangle approximates the person's exposure over the time interval
from t,., to /,. The sum of the areas of many such rectangles from time a to time b approximates
the person's exposure for this interval of time and applies whether the person remains in a single
microenvironment or visits many microenvironments. Therefore, the hatched area given by B in
Figure 4-2 is the person's integrated exposure from time a to time b.
If a dose metric, such as breathing rate or ingestion rate, had been included with the
information on concentration, then the profile in Figure 4-2 would approximate the potential
dose. The dose profile combines information on pollutant concentrations, activity patterns, and
intake rates. For many pollutants (e.g., ozone, carbon monoxide), the time series of exposure or
potential dose may be more important for estimating the health impact than the overall average
NOVEMBER 1999 4^4TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
exposure or cumulative dose (McCurdy 1997). Figure 4-2 also shows some of the measures of
exposure or potential dose that may be derived from the profile. The output options for
TRIM.Expo will include exposure and potential dose profiles.
Figure 4-2
Example of an Exposure or a Potential Dose Profile and Associated Measures
(adapted from McCurdy 1997)
Ci
X
\
5
R
KEY
Time
B
p
R
t=a
t-i t
At,
i=b
integrated exposure from time t-a to t=b
time between peaks over x
respites between exceedances ofx
The method of combining the vital information that is used in a dose profile makes it
possible to develop estimates for alternative exposure and dose metrics for different averaging
times from one hour to a year or more. Examples of alternative metrics that OAQPS has needed
to investigate recently include children exposed to 8-hour daily ozone values greater than 0.08
ppm while exercising at a breathing rate of 15 L/min/m2 or higher, and cardiovascular-impaired
persons with a daily carboxyhemoglobin level of 2 percent or higher due to CO exposures
(McCurdy 1995). Indeed, by using this disaggregated approach, almost any combination of
exposure and dose metrics is possible.
With respect to health effects, two important considerations are (1) the time-scale of the
health effects that result from an exposure or repeated exposures and (2) whether there is
NOVEMBER 1999
4-5
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRJM.Expo
assumed to be a threshold concentration below which no health impacts are expected.
Depending upon the health effects associated with the pollutant of interest, the exposure and
potential dose profile may be used to derive several metrics. For example, if the time steps are
one-hour, each concentration estimate represents a one-hour average. If the pollutant's health
effect is associated with one-hour average exposures and has a lowest observed or a no observed
adverse effect level (e.g., x), important metrics might include the following (refer to Figure 4-2).
1. The number of person-hours of exposure to concentrations above x:
= P(cohort) x ฃ [ x, and
C(t) = exposure concentration for time step /.
2. The sum of the concentrations that exceed x
_ [d(t) x C(t)]
where:
/ = a, t,, /,,. .., /./, tn = b (refer to Figure 4-2).
3. The average of the concentrations that exceed x
' C(r)]
(4-4)
4. The sum of exceedances of x, when it is exceeded
(4-5)
where:
t = a, th t:,. .., tn.,, tn = b (refer to Figure 4-2).
NOVEMBER 1999 4-6 TRJM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
5. The average exceedance of x, when it is exceeded
(4-6)
Other important metrics might include the average number of sequential hours exceeding x, or
the average number of time steps between local concentration peaks (i.e., average length of
respites).
If the pollutant's health effect does not have a specific benchmark concentration (i.e.,
lowest observed or no observed adverse effect level), but is associated with a certain averaging
time (e.g., 1-hour, 8-hour, 24-hour, annual), important metrics might include:
The distribution of the maximum concentration corresponding to the averaging time of
interest for the exposure period, typically one year (e.g., distribution of the maximum 8-
hour average concentration for any time during the year), or
The average daily maximum concentration corresponding to the averaging time of
interest for the exposure period (e.g., the average of the maximum 8-hour average
concentration for each day of the year).
The TRIM.Expo module was designed to address OAQPS' need for an integrated
exposure model system that can evaluate the distribution of the population exposed to specified
levels of air pollution and the number of times they are exposed for one-hour, eight-hour,
monthly, quarterly, seasonal, and annual averaging periods (McCurdy 1995). As described
above, TRIM.Expo will fulfill this need by producing population distributions of multiple
exposure, dose, and dose-rate indicators.
4.1.3 AVERAGE EXPOSURE CONCENTRATION
During a relatively long time period (e.g., day, week, year), individuals have different
time/activity budgets and occupy different microenvironments with different pollutant
concentrations. In such cases, exposure cannot be addressed by using the basic exposure-event
function. Duan (1982) proposed a method to determine the average exposure concentration,
EC:m, that has been used by others (Ott 1984, Ott et al. 1988, Ott et al. 1992a, Klepeis 1994,
Lurmann and Korc 1994, Macintosh et al. 1995):
ECz, m - ^ Cm(l, k, /, t)ETz, m(t, kj,t) (4-7)
J. /
NOVEMBER 1999 4-7 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
where:
EC2 m = average exposure concentration in exposure medium m of individual or
cohort z over time period T, where T is used as the averaging time and is
the sum of all time steps, /
T = sum of all time steps, t.
The algorithm would determine an exposure concentration averaged over the total
potential exposure duration considered in a specific TRIM.Expo application. However, the
averaging time may not correspond to the averaging time specified for a health benchmark. For
example, the exposure averaging time for a particular application of TRIM.Expo may be a year,
while the averaging time for the health benchmark for the pollutant may be specified for a
lifetime. Such issues will need to be addressed in TRIM'S Risk Characterization module,
TRIM.Risk.
4.1.4 INTAKE-ADJUSTED AVERAGE EXPOSURE CONCENTRATION
Cumulative exposure can also be determined using overall average concentration of the
exposure medium that enters the body over time period T. The intake-adjusted average exposure
concentration, IEC., is calculated using a weighting factor based on the intake rate, such as
breathing or ingestion rate:
IEC-. = ^ (4-8)
where
IUm .(t) = rate of intake/uptake in exposure medium m by individual or cohort z
during an exposure time step /, and under other factors that are defined
above. For food ingestion, the units of measurement for /ฃ/. might be
kg-food/kg-body-weight during an exposure duration ET.
4.1.5 INTERMEDIA TRANSFER FACTOR
In many situations, the concentration in the exposure medium, Cm, is not known directly
and must be estimated from an intermedia transfer. For example, the outdoor air concentration
of a particular pollutant may be known, but the pollutant concentration indoors and the exposure
medium may be unknown. To differentiate the indoor air concentration contribution from that of
the outdoor air requires an intermedia transfer factor, which converts the time history of
concentrations that is known (i.e., outdoor air) to an estimate of the time history of the unknown
concentration (i.e., indoor air). Other examples of intermedia transfers include soil to house dust,
water to indoor air, water to fish, and soil to home-grown vegetables.
NOVEMBER 1999 4^8 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
The intermedia transfer factor, ITF, is used as follows:
CXU,/,0 = Cn(j9s)ITF(j,s=> m,k,t) (4-9)
In this expression, for exposure medium m (e.g., air, water, soil) at location / in
microenvironment k associated with activity / during time step /, Cm(i,k,l,t) is the concentration
contributed by concentration Cn(j,s) from environmental medium n at location/ and time s. ITF
(j,s -*m,k,t) is the intermedia transfer algorithm that maps concentration Cn(j,s) to concentration
Cm(i,k,l,t). Summation of the product Cn x ITF lime occurs over previous time steps that impact
current time step t and over all locations/ that impact current microenvironment k.
4.1.6 AVERAGE DAILY POTENTIAL DOSE
The exposure-event function is frequently used to assess the potential dose to an exposed
individual from his or her cumulative intake over some time period relative to an averaging time.
The average daily potential dose, ADDpol, is the potential dose per day (d). It is similar to the
intake-adjusted average exposure concentration, but defines an average rate of intake in mg/kg/d
or mg/d over the averaging time instead of a concentration.
i, k, /, t)ETz, ซ(/, k, /, t)IUz,
ADD pot = - - -- - -- (4-io)
where:
T - averaging time used to assess the health effects of the intake,
IU.mft)= rate of intake/uptake.
4.2 DEFINING THE MODEL COMPONENTS FOR A TRIM.Expo
APPLICATION
To set up a TRIM.Expo application, several steps must be performed.
1 . Identify the pollutant(s) of interest, the study area, the exposure districts, and the
population(s) of interest.
2. Determine environmental media within each exposure district. Estimate ambient media
and time-varying pollutant concentrations in the environmental media.
3. Identify exposure media for the population(s) of interest.
4. Construct and evaluate intermedia transfer factors relating time- varying exposure media
concentrations to time-varying environmental media concentrations.
NOVEMBER 1999 4^9 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
5 Divide the population(s) of interest into an appropriate set of cohorts.
6 Develop a sequence of exposure events linking the cohorts to exposure media
7. Create a dose profile for each exposure route based on the intake of pollutant(s) in each
exposure event for each cohort
Figure 4-3 illustrates a typical TRIM Expo application. Pollutants from environmental
media contained in environmental districts are linked to microenvironments through the
intermedia transfer factors Cohort activities are linked to the microenvironments as well This
information is then used to assess an exposure profile and/or cumulative intake during a defined
time period These steps are described in more detail below
Figure 4-3
Sequence of Exposure Events for a Set of Population Cohorts
Illustration of the Process by which Environmental Media Concentrations for Multiple
Exposure Districts are linked using Intermedia Transfers to Exposure Media in a Set of
Microenvironments
Ambient
environment
districts
Microenvironments
Examples
Residential zone
Transit corridor
Industrial area
Business area
transfers, e.g
penetration factors
Indoors at home
Outdoors at home
Automobile
Indoors at work
Outdoors at work
Indoors at school
Outdoors at school
Restaurant
Activity-based
sources
District 1
^District 2
District ^
District N
Int
ME A
ME B
ME C
ME X
rtrrriftrlifk
^~- -^
^ C ~")
^ -^
^ C ^)
Vv- -^
S ou rce/d ose
factors
Smoking
Cooking
Consumer
products
NOVEMBER 1999
4-10
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
4.2.1 DEFINE STUDY AREA, EXPOSURE DISTRICTS, AND ENVIRONMENTAL
MEDIA
A study area is an urban and/or local area for which environmental concentrations have
been assembled for an exposure assessment. It is divided into one or more exposure districts. In
order to perform multimedia exposure assessments, concentration data from multiple
environmental media (i.e., ambient air, vegetation, surface soil, root zone soil, deeper or vadose
zone soil, ground water, surface water) are needed for each exposure district. The concentration
information is gathered either from analysis of ambient monitoring data or from a simulation
model such as TRIM.FaTE. Each ambient environmental medium is characterized in terms of
one or more component phases - gas, water, liquid, and/or solids. Each environmental medium
is described below.
4.2.1.1 Ambient Air
The ambient (i.e., outdoor) air in an exposure district can be characterized in terms of its
gas, particulate matter, and water composition. Its volume and mass are defined by the area of
the exposure district and the depth of the lower troposphere. Pollutants in ambient air are
dispersed by atmospheric advection and diffusion and are influenced greatly by meteorological
parameters. Because particle size influences the particle's behavior in the atmosphere with
respect to deposition and settling and its ability to penetrate the lung cavities and affect human
health, it is important to consider particulate matter of various cut sizes, such as PM10 and PM2 5.
The TRIM.Expo module will consider other particle sizes, as the need arises, in the future. In
addition, during precipitation (e.g., rain, snow) and fog events, it is important to characterize the
volume fraction of air that is water.
4.2.1.2 Vegetation
Vegetation is the dominant component of the terrestrial plants exposure district.
Vegetation generally has contact with two other environmental media, air and soil. However,
plant interactions with these media are not understood well enough to define an accurate method
for predicting pollutant uptake. In order to assess potential vegetation-pathway exposures,
vegetation can be further delineated into above-ground vegetation and root crops. Above-ground
vegetation includes leafy vegetables (e.g., cabbage, cauliflower, broccoli), exposed produce (e.g.,
apples, berries, cucumber, squash), protected produce (e.g., citrus fruits), and grains (e.g., wheat,
com, rice). Root crops include, for example, carrots, beets, legumes, and melons.
4.2.1.3 Surface Soil
Surface soil consists of solid, liquid, and gas phases. Studies of radioactive fallout in
agricultural land management units reveal that, in the absence of tilling, particles deposited from
the atmosphere accumulate in and are resuspended from a thin ground or surface soil layer with a
thickness in the range 0.1 to 1 cm (Whicker and Kirchner 1987). The ground-surface-soil layer,
has a lower water content and higher gas content than underlying layers.
NOVEMBER 1999 4-11 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRJM Expo
4.2.1.4 Root Zone Soil
Root zone soil consists of solid, liquid, and gas phases. Soil-water content in the root-
zone is somewhat higher than that in surface soils because the presence of clay serves to retain
water. The roots of most plants are confined within the first meter of soil depth. In agricultural
lands, the depth of plowing is 15 to 25 cm. In addition, the diffusion depth, which is the depth
below which a pollutant is unlikely to escape by diffusion, is on the order of 1 m or less for all
but the most volatile pollutants.
4.2.1.5 Vadose Zone Soil
Vadose zone soil has solid, liquid, and gas phases. The soil in this layer typically has a
lower organic carbon content and a lower porosity than the root zone soil. It is assumed that
pollutants in this layer will move downward to the ground water zone primarily by capillary
motion of water and leaching.
4.2.1.6 Ground Water
Ground water is defined as water that can be withdrawn from the saturated zone below
the vadose zone of the soil layer. Ground water consists of both a liquid and a suspended particle
phase.
4.2.1.7 Surface Water
Surface water is composed of two phases: pure water and suspended sediment material.
Tht suspended sediment material phase contains sorbed pollutants. Surface water compartments
are assumed to be well-mixed systems and include ponds, lakes, creeks, rivers, estuaries, seas,
and oceans.
4.2.2 DEFINE EXPOSURE MEDIA AND MICROENVIRONMENTS
The exposure assessment process consists of relating pollutant concentrations in the
environmental media (e.g., ambient air, surface soil, ground water) to pollutant concentrations in
the exposure media with which a human population has contact (e.g., personal air, tap water,
foods, household dusts). The exposure media that have been identified for inclusion in
TRIM.Expo include the following.
Outdoor air;
Indoor air;
Tap water;
Home-grown food (i.e., produced by and consumed by a household);
NOVEMBER 1999 4-12 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Locally-produced food (i.e., produced by home gardens and commercial farms in contact
with air, soil, and/or water in the study area);
Prepared food;
Breast milk;
House dust;
Soil;
Swimming pools; and
Recreational surface water.
Microenvironments are locations in which a population comes into contact with
pollutants. They are well-characterized, relatively homogenous locations with respect to
pollutant concentrations for a specified time period. Table 4-1 is an initial list of the
microenvironments that will be included in TRIM.Expo., other microenvironments will be added
during the course of module development.
Table 4-1
Microenvironments to be Included in TRIM.Expo
Microenvironment #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Microenvironment
General
In vehicle
In vehicle
In vehicle
In vehicle
Indoors
Outdoors
Outdoors
Outdoors
Indoors
Outdoors
Indoors
Indoors
Indoors
Indoors
Specific
Car
Bus
Truck
Other
Public garage
Parking lot/garage
Near road
Motorcycle
Service station
Service station
Residential garage
Other repair shop
Residence - no CO source
Residence - gas stove
NOVEMBER 1999
4-13
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Microenvironment #
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Microenvironment
General
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Outdoors
Outdoors
Outdoors
Outdoors
Outdoors
Outdoors
Outdoors
In vehicle
In vehicle
Specific
Residence - attached garage
Residential - stove and
garage
Office
Store
Restaurant
Manufacturing facility
School
Church
Shopping mall
Auditorium
Health care facility
Other public building
Other location
Not specified
Construction site
Residential grounds
School grounds
Sports arena
Park/golf course
Other location
Not specified
Train/subway
Airplane
4.2.3 DEFINE RELEVANT INTERMEDIA TRANSFERS
An intermedia transfer factor is an algorithm that expresses the transfer of a pollutant
from an environmental medium to an exposure medium. The following tables summarize
intermedia transfers which need to be considered in exposure modeling. Some of these
intermedia transfers, where feasible and appropriate, will be included in TRIM.Expo, while
others may be addressed within the TRIM.FaTE module. Each cell representing an interaction is
shaded, and the exposure pathways (/. e., inhalation, ingestion, dermal contact) that results from
the intermedia transfer is indicated. The tables below show the matrix of links between the
various exposure media noted above and ambient-air phases (Table 4-2), ambient soil media
(Table 4-3), ambient water media (Table 4-4), and ambient vegetation (Table 4-5).
NOVEMBER 1999
4-14
TRIM.EXPO ISO (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Table 4-2
Matrix of Links Between Ambient-Air Phases and Various Exposure Media
EXPOSURE MEDIA
ENVIRONMENTAL MEDIUM = Ambient air
GdS
Phase PM25 PMlฐ Precipitation
outdoor , ' -^^MHBH HHIHHHHHpH^^HHHi^B^IK
lnd<
Hou
Soil
Wat
gases Inhalation
PM26 Inhalation
PM10
precipitation
>orAir
gases Inhalation
PM2 5 Inhalation
PM10
water vapor
seDust
on floors j j Ingestion/dermal
on surfaces I
_.^_.__ฃ
er ,:;i-^ j
Food
home fruits
protected
unprotected
home vegetables
above ground
root crops
home grains
meat
milk
dairy products
eggs
breast milk
prepared foods
^!8N"raBSBl^i*
t^^^^^HtMBM^CT^^-tt
Pffi^Pj^E^
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion/dermal
Inhalation
Inhalation
Ingestion/dermal
Ingestion/dermal
.. ^^^^^^^mBBlBBBi^^BKlM^^^^^^B^pBIBBMB
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
w
-WwssffSK^i^'.f.i*-. ......
Ingestion
Ingestion
Ingestion
NOVEMBER 1999
4-15
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Table 4-3
Matrix of Links Between Soil Media and Various Exposure Media
EXPOSURE MEDIA
Outdoor Air
gases
ENVIRONMENTAL MEDIUM = Soil
Surface Soil i Root Zone Soil i Vadose Zone Soil
Inhalation
Inhalation
PM,
Inhalation
PM,
Inhalation
precipitation
Indoor Air
gases
Inhalation
Inhalation
Inhalation
..................asi.::
Inhalation
Inhalation
PM,
water vapor
House Dust
on floors
Ingestion/dermal
on surfaces
Ingestion/dermal
Soil
residential
Ingestion/dermal
Ingestion/dermal
construction site
Ingestion/dermal
Ingestion/dermal
industrial site
Ingestion/dermal
agricultural site
Ingestion/dermal
Ingestion/dermal
recreational site
Ingestion/dermal
school
Ingestion/dermal
home fruits
protected
unprotected
Ingestion
Ingestion
Ingestion
home vegetables
above ground
Ingestion
Ingestion
NOVEMBER 1999
4-16
TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
EXPOSURE MEDIA
root crops
home grains
meat
milk
dairy products
eggs
breast milk
prepared foods
ENVIRONMENTAL MEDIUM = Soil
Surface Soil
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Root Zone Soil
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Vadose Zone Soil
Table 4-4
Matrix of Links Between Water Media and Various Exposure Media
ENVIRONMENTAL MEDIUM = Water
EXPOSURE MEDIA
residential
Ingestion/dermal
Ingestion/dermal
construction site
industrial site
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
agricultural site
recreational site
school
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion
Ingestion/dermal
Ingestion/dermal
Ingestion/dermal
Ingestion
NOVEMBER 1999
4-17
TRIM.Expo ISO (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
EXPOSURE MEDIA
ENVIRONMENTAL MEDIUM = Water
Surface Water
Liquid Phase
Particle
Phase
Ground Water
Liquid Phase
Particle Phase
Surface Water used for Recr
home fruits
protected
Ingestion/dermal
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
unprotected
home vegetables
above ground
root crops
Ingestion
Ingestion
Ingestion
Ingestion
home grains
meat
Ingestion
Ingestion
Ingestion
Ingestion
milk
dairy products
Ingestion
Ingestion
Ingestion
Ingestion
eggs
breast milk
prepared foods
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
NOVEMBER 1999
4-18
TRIM.Expo ISO (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Table 4-5
Matrix of Links Between Vegetation and Various Exposure Media
EXPOSURE MEDIA
ENVIRONMENTAL MEDIUM = Vegetation
Above-Ground Plants
Fruits
I Vegetables i Grains
Root Crops
Crops
home fruits
protected
unprotected
Ingestion
Ingestion
home vegetables
above-ground
Ingestion
root crops
home grains
Ingestion
Ingestion
Ingestion
meat
milk
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
jngestipn
Ingestion
dairy products
eggs
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
breast milk
prepared foods
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
Ingestion
4.2.4 DIVIDE POPULATION INTO APPROPRIATE SETS OF COHORTS
The first step in selecting population(s) of interest is to define the total population
associated with all of the exposure districts. This may be expanded to incorporate more than
residents of the defined exposure districts. For example, people who do not reside within the
exposure district but consume its agricultural products or fish may be included. Once the
population(s) are identified, the second step is to decide whether to include all of the population
members in a region or a subset of the population of concern due to a demographic factor,
activity, or health characteristic (e.g., asthmatics, outdoor workers, pregnant women).
The population of interest is then divided into cohorts. In TRIM.Expo, each individual is
exclusively assigned to a single cohort. Cohorts are distinguished from the overall population
NOVEMBER 1999
4-19
TRIM.Expo ISO (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
based on attributes including age, gender, health status (e.g., asthmatics), exposure district, and
housing types (e.g., gas stoves, HVAC systems). The number of cohorts selected depends largely
on the level of resolution of the exposure factor data used to characterize the population and on
the level of exposure resolution required. The type of pollutants and sources under consideration
can affect the number and type of cohorts as well
As previously stated, cohorts can be defined for a particular application or situation For
example, cohort exposure can be a function of demographic group, location of residence, location
of work place, and type of home ventilation system Cohort exposure can be linked to ambient
pollutant concentrations in multiple districts (e.g., home and work district). Specifying the
demographic group allows cohort exposure to be linked to activity patterns that vary with age,
work status, and other demographic variables. In some analyses, cohorts are further distinguished
according to time spent in particular microenvironments. For example, in studies of ozone
exposures, additional cohorts were created to account for children who spent significant amounts
of time outdoors and also for adults who worked outdoors These cohort designations helped
analysts estimate exposures for subgroups in the population with the greatest potential for higher
exposures
4.2.5 DEVELOP AN EXPOSURE-EVENT SEQUENCE FOR EACH COHORT
Once a set of cohorts has been selected, a sequence of exposure events is defined for each
cohort An exposure-event sequence is a chronological set of events that define the activity/time
allocation of each cohort An exposure-event sequence defines a cohort by (1) exposure district,
(2) microenvironment, and (3) activity at each time step of a calculation Implicit in the definition
of activities for the exposure-event sequence is information about the time of year or season,
through the selection of activity patterns based on the outdoor air temperature that coincided with
the collection of the activity pattern data. The following example in Table 4-6 shows simple
eight-hour exposure-event sequence. Because an exposure-event sequence is developed from an
individual's activity pattern data used to represent the cohort, concentrations in each of the
microenvironments are the same for each member of the cohort Multiple runs of the model
results in variation of the activity patterns for a given cohort.
NON-EMBER 1999 4-20 TRIM EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Table 4-6
Sample Eight-Hour Exposure-Event Sequence
Time: 06:001
Exposure District
Zone 1
Zone 2
Microenvironment
Indoors at home
Inside a
vehicle
Indoors at work
Activity
Sleeping
Dressing,
eating
Comm-
uting
Working at desk
Eating
lunch
4.2.6 DETERMINE EXPOSURE MEDIA CONCENTRATIONS AND CONTACT IN
EACH MICROENVIRONMENT
Once a sequence of exposure events is established, the exposure medium concentration
and the rate of contact between the cohort and the exposure medium in the various
microenvironments must be estimated. The route of contact/intake can greatly influence the
affected populations or the populations of interest that are studied in an exposure assessment.
For example, hand-to-mouth contacts are more important for a population of children than for a
population of adults. This information is then used to establish an exposure concentration
profile, a cumulative intake, or an average exposure medium concentration over a defined
averaging time.
4.2.7 ESTIMATE AN INTAKE RATE FOR EACH DOSE EVENT
For each exposure event in the exposure-event sequence, TRIM.Expo estimates the intake
rate of the pollutant for all relevant routes of exposure. Initially, TRIM.Expo will provide
estimates for inhalation and ingestion only. To estimate the intake rate of a pollutant via
inhalation, TRIM.Expo defines each exposure with a pollutant concentration and a ventilation
rate indicator. The applied dose is a function of the pollutant concentration in contact with the
individual or cohort, the activities that the individual or cohort are engaged in that affects
breathing rate, and the ventilation rate value that is assigned to the exposure event. Section 5.1.5
describes in detail how pollutant concentrations and the ventilation rate associated with each
exposure event will be estimated.
For ingestion, TRIM.Expo estimates the average daily potential dose (ADD) for many
different ingestion pathways and media of exposure. The ADD represents the amount of
NOVEMBER 1999
4-21
TRIM.Expo ISO (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK .VXD CONCEPTSALIZAT1ON OF TRIM Expo
pollutant that enters the mouth of the exposed individual or cohort over a defined exposure event
for a defined exposure duration Data are provided on the rate of ingestion of the exposure
medium (e.g., water, soil, food) during an exposure event and are used to calculate the ADD
Section 6 1 provides a summary of the general approach used to characterize the ADD for an
exposure medium. Subsequent sections in Chapter 6 provide details on how the algorithm applies
to specific exposure media
4.2.8 EXTRAPOLATE THE COHORT EXPOSURES TO THE POPULATIONS OF
INTEREST
After the exposures to the cohorts have been calculated, they are extrapolated to the
population at risk by estimating the population size for each cohort Population estimates
assigned to each cohort are calculated for both commuting (i.e. to work, to school) and non-
commuting cohorts The first step is to calculate the population of each demographic group that
resides in a particular home district through data available from the Bureau of the Census This
information provides the population of all non-commuting cohorts by home district
The population of non-commuting cohorts can be further divided to account for a
particular attribute shared by the cohorts The attribute should be one for which data about the
size of the population can be obtained The population size of each demographic group listed as
having attribute b is calculated through the following equation
pop (dg, ca, b) = f (ca, b) x pop (dg, ca) (4-11)
where
pop(dg,ca,b) = The population of demographic group d in census area1 ca having
The attribute of interest b.
f(ca,b) = Fraction of people in census area ca that have attribute b.
pop(dg,ca) = Total number of people in demographic group d that reside in
census area ca
The values ofpop(dg,ca,b) are summed over each home district to yield estimates ofpop(dg,h,b),
the number of people in demographic group dg within home district h that share attribute b. Any
number of attributes can be used to calculate pop(dg,h,b) if census area data are available.
The value ofpop(dg,h,b) provides an estimate of the population in each non-commuting
cohort associated with demographic group dg residing in home district h that shares the attribute
of interest b Next, the populations of the commuting cohorts in demographic group dg who
share attribute b are calculated using the fraction of all commuters residing in home district h who
travel to commute district w (including children who commute to school) using Equation 4-12
1 "Census area" can be any spatial area designated by the Bureau of the Census. For most exposure
applications, the census tract designation is used.
NOVEMBER 1999 4^22 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
com(dg, h, w, b) = pop(dg, h, b) * com(h, w) / com(h) (4-12)
where:
com(dg,h,w,b) = number of people in a commuting cohort associated with demographic
group dg in home district h and commute district \v having attribute b.
com(h,w) = number of commuters in all demographic groups that travel from home
district h to commute district w.
com(h) = total number of commuters in home district h.
Estimates ofcom(h) can be obtained from census data specific to each district.
By defining cohorts according to attributes that are important to specific applications or
situations, TRIM.Expo is able to extrapolate the cohort exposures to the general population while
still retaining information about the incidence of exposure to the specific subset of the population
under study. For example, suppose that the population of interest was individuals, ages 18 to 44,
who experience asthma. For a particular study area, information for both of these attributes is
available. Census data contains the breakdown of age information by location, while information
on the incidence of asthma is available from various sources, including the National Center for
Health Statistics and state and local health agencies. In this case, the TRIM.Expo analysis would
be arranged to capture the demographic characterization of this subset of the population by
selecting a demographic group that includes individuals from 18 to 44 years of age. The fraction
of the demographic group with asthma can be determined from health data. When combined,
this information provides an estimate of the exposures to this specific segment of the population
for each exposure district, assuming that activity patterns for asthmatics do not differ
significantly from those of the other members of the demographic group. Once the exposures of
this sensitive subpopulation are known for each exposure district, the exposures can be readily
extended to encompass the entire study area under investigation.
In the simple example given above, the fraction of the total population with the specified
attribute in a particular exposure district can be quite different from the fraction of all people in
the study area contained within that exposure district. Consider, for example, Figure 4-4 which
presents a hypothetical illustration of the example above. In this figure, each circle represents an
exposure district. The study area for this example is the sum of the three exposure districts. The
total population of this hypothetical study area is 100,000 people. The fraction and number of
the total study area population residing in each exposure district is indicated in the top half of
each circle. The fraction and number of the 18 to 44 year olds in the study area population with
asthma in each exposure district is indicated in the bottom half of each circle. As can be seen in
this example, exposure district #3, which includes 50 percent of the total population of the study
area, contains 2,500 people age 18 to 44 with asthma, only 2.5 percent of the total study area
population (i.e., all three exposure districts combined). However, exposure district #2, which
includes only 20 percent of the total population of the study area, accounts for 7 percent (7,000)
of the cases of asthma in the 18 to 44 year-olds in the study area. Finally, exposure district #1,
which has one-third of the study area's total population, has 18,000 cases of asthma in people of
age 18 to 44, 18 percent of the total study area population.
NOVEMBER 1999 4-23 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
Figure 4-4
Hypothetical Spatial Distribution of the Incidence of Asthma
Percent and number of total study
area population (all ages)
that reside In this exposure district
Percent and number of the population
(18-44 years) that reside in this
exposure district and have asthma
Percentage of the total study
area population (18-44 years)
that has asthma
SCENARIO:
Study area population (all ages) = 100,000
Study area population (18-44 years) that has asthma = 27,500 (27 5 percent)
Exposure District 1
18%
Exposure District 2
7%
Exposure District 3
2.5%
NOVEMBER 1999
4-24
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
The example above illustrates that the spatial distribution of a demographic variable, such
as the incidence of a particular disease, is not necessarily the same as the spatial distribution of
the general population. The illustration also points out that conducting exposure analyses on a
smaller spatial scale is likely to lead to more accurate results. This is particularly true when
investigating exposures related to factors that have a high degree of variability over small
distances. Different exposure estimates for the sensitive subpopulation of individuals suffering
from asthma would have been obtained if the exposures of the entire population were calculated
and then adjusted for the incidence of asthma based on the percentage of reported cases for the
study area as a whole without regard for the spatial variation of this disease.
4.2.9 FUNCTIONAL ATTRIBUTES
The TRIM.Expo module will include important attributes that are needed by analysts and
decision-makers to investigate the complex nature of multipollutant and multipathway human
exposures to hazardous and criteria air pollutants in the environment. This section briefly
describes many of the functional attributes that will be included in TRIM.Expo. The attributes
fall into two major categories: (1) important considerations related to the scientific defensibility
of the exposure estimates, and (2) design features of TRIM.Expo that help ensure the needed
flexibility and ease of use of the modeling system.
4.2.9.1 Inclusion of Indoor and Outdoor Environments and Their Emission Sources
Recent studies have shown the importance of including both indoor and outdoor
environments when assessing human exposures to toxic pollutants (Ozkaynak et al. 1996, U.S.
EPA 1987, Wallace et al. 1985, Wallace et al. 1991). These studies called the Total Exposure
Assessment Methodology (TEAM) and Particle Total Exposure Assessment Methodology
(PTEAM) studies, and other indoor/outdoor studies that have succeeded them, reported that
indoor concentrations of many toxic air contaminants were often greater than concurrent outdoor
concentrations. Other studies have reported that most people in the U.S. can spend as much as
90 percent of their time indoors or inside motor vehicles (Jenkins et al. 1992, U.S. EPA 1996a).
Conversely, exposures to the NAAQS pollutants (e.g., ozone, sulfur dioxide) are dominated by
outdoor levels or by ambient air that infiltrates to indoor environments. Therefore, it is vital that
a framework for assessing the total human exposure to pollutants include the indoor, in vehicle,
and outdoor environments.
An important microenvironment for exposure to multiple airborne pollutants is inside a
passenger vehicle. This microenvironment is treated similarly as the indoor environment,
although a significant fraction of the concentration found indoors may come from penetration of
ambient air. Nevertheless, exposures to many pollutants within a moving vehicle have been
reported to be higher than either roadside or ambient (i.e., away from a road) exposures (CARB
1998b). Lawryk et al. (1995) reported that commuting can account for a substantial amount of a
person's daily exposure to select toxic pollutants. Furthermore, both studies reported that many
factors can affect the levels of pollutants inside a vehicle, including the general condition and
maintenance of the vehicle, the vehicle's age, the density of the surrounding traffic, and the
resulting effect these factors have on the speed of the vehicle.
NOVEMBER 1999 4-25 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
Besides traveling in a vehicle, other activities that can have a significant impact on
exposures are cooking and smoking Increased concentrations in close proximity to gas stoves
have been reported for NO2 (Sega and Fugas 1991), CO (U.S. EPA 1984), and PM25 and PM10
(Ozkaynak et al 1996) The concentrations of other pollutants produced through incomplete
combustion have also been found to be higher close to a cooking appliance. Ozkaynak et al
(1996) also reported that smoking was a significant source of particles in homes in the PTEAM
study. They found that homes where smoking was reported averaged approximately 30 //g/m3
higher levels of PM10 than homes without smoking. Numerous other exposure studies have
reported higher concentrations of CO, particles, and toxic pollutants in indoor environments
(particularly the home) with environmental tobacco smoke (ETS) (Daisey et ai. 1998, Jenkins et
al 1996, Miller et al. 1998, Phillips et al. 1998, Quackenboss et al 1991, Thomas et al. 1993,
Waldmanet al 1991).
The TRIM.Expo module has several indoor and in-vehicle microenvironments, including
inside a home (several different locations), in transit (e.g , automobiles, buses, trains), indoors at
work or school, indoor recreational facilities (e.g., movie theaters), shopping malls, and
restaurants In addition, TRIM.Expo has the flexibility to include other indoor
microenvironments of importance as information becomes available.
While evidence continues to accumulate regarding the importance of indoor sources and
environments on exposures, there remains important sources of exposure outdoors, too In fact,
studies show that in the absence of indoor sources, penetration of outdoor pollutants were
responsible for elevated indoor concentrations (see, for example, Lioy et al. 1991) However,
investigations into the sources of outdoor pollutants have found that, although human activities,
such as motor vehicle use have a major impact on ambient levels of particles and toxic pollutants,
relatively small, local sources of specific toxic pollutants such as repair shops can also have a
significant impact on the exposures to a nearby population Moreover, exposures to pollutants
from certain outdoor sources can be affected by local climatology, thereby exhibiting seasonal
variations in the exposure profile An example are pollutants that are emitted by the burning of
heating fuel in the winter months (Lioy and Daisey 1990)
The TRIM Expo module currently has several outdoor environments included in its
modeling system There are outdoor microenvironments specified for at home, in transit (e.g.,
walking, bicycling), at work, and recreational locations The location where eating takes place
may also be important for estimating ingestion exposures.
4.2.9.2 Flexible, Modular, and Portable Algorithms
The TRIM Expo module is designed with features that will facilitate future expansion and
integration of the framework in anticipation of new modeling techniques and data. Additionally,
there is a need for TRIM.Expo to be compatible with other exposure modeling platforms
developed both within and outside of EPA. The design features that will allow for these
capabilities are portability, modularity, and flexibility. These three features are all interrelated in
function and purpose. They follow the recommended guidance set forth by the National
Research Council's Committee on Advances in Assessing Human Exposure to Airborne
Pollutants (NRC 1991b) Flexibility means that the model algorithms are written using precise,
NOVEMBER 1999 4-26 TRIM.Expo TSD (DRAFT)
-------
CHAPTER4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
standard encoding practices. This approach will allow the model to be updated relatively easily
as new data and new modeling techniques become available. The TRIM.Expo module
algorithms are being designed to readily incorporate new information regarding both data and
methods. As an example, comments will be included within the model's source code which will
help future developers understand how the model was constructed and explain each of the
functional components of the model. This seemingly minor point has important implications for
the future development of TRIM.Expo because (1) model development and revisions may
continue indefinitely, and (2) responsibility for future updates of the model may shift to different
groups.
Another aspect of flexibility is the use of modular design in model development.
Modular design, as its name implies, means that the components of the model involved in the
computation of exposures are separated into discreet units, where each unit is responsible for a
particular function. These discreet units, or modules, are called upon during the execution of a
model run to perform a specific function or set of functions. If a particular application does not
require the function(s) performed by a certain module, then the calculations or routines
performed by the module are not called upon. In most of the early exposure models, all of the
functions of a model were executed regardless of whether they were required for the particular
application. Using only the modules that are needed for a specific application reduces the
amount of computer resources required. Using a modular approach also aids in the development
of the model because when specific model functions are in modules, they can be revised and
tested without running the entire model. This makes isolating and correcting mistakes in the
model's algorithms easier, as well.
The functional attribute of portability is important to the development of the TRIM.Expo
modeling system. Portability refers to how easily the model can be integrated with other
modeling systems such as Models-3 or MENTOR, in the case of TRIM. Expo. The overall goal
of systems such as Models-3 and MENTOR is to simplify and integrate the development and use
of complex environmental models. The developers of these systems seek to provide a set of
ready-to-use methodological tools and linkages to relevant databases for performing assessments
of exposure. Hence, it is important that TRIM.Expo can take full advantage of the designs and
numerical analysis tools provided by these systems. Therefore, TRIM.Expo is based on an open,
user-oriented implementation that is compatible with the components of other modeling systems.
An open system design means that TRIM.Expo's development is not specific to a single
computing platform. Finally, TRIM.Expo can be run without the use of proprietary software or
model components, enhancing both the flexibility and portability of TRIM. Expo.
4.2.9.3 Explicit Treatment of Uncertainty and Variability
While the importance of characterizing uncertainty and variability explicitly and
separately is well recognized (NAS 1994, CRARM 1997, U.S. EPA 1997c), OAQPS intends to
selectively apply uncertainty and variability analyses on a case-specific basis, (i.e., for critical
parameters and, where appropriate, based on the underlying science and data).
The TRIM.Expo module explicitly treats the uncertainty in the model estimates of
exposure. Uncertainty is the lack of knowledge of the actual values of physical variables
NOVEMBER 1999 4^27 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIMExpo
(parameter uncertainty) and of physical systems (model uncertainty). For example, parameter
uncertainty results when non-representative sampling (to measure the distribution of parameter
values) gives sampling errors. Model uncertainty also results from simplification of complex
physical systems. Uncertainty can be reduced through improved measurements and improved
model formulation.
As with uncertainty, the variability in model inputs will be explicitly characterized in
TRIM.Expo, wherever feasible. Variability is not to be confused with uncertainty; they are two
separate aspects inherent in data sampled from a population. Variability represents the diversity
or heterogeneity in a population or parameter and is sometimes referred to as natural variability.
An example is the variation in the heights of people. Variability cannot be reduced by using
more measurements or measurements with increased precision (i.e., more precise measurements
of people's heights does not reduce the natural variation in their heights). However, it can often
be reduced by a more detailed model formulation. For example, modeling people's heights in
terms of their age will reduce some of the unexplained variability in the distribution of data on
heights. Variability among the members of a population in factors such as food ingestion rates,
exposure duration, and expected lifetime can be described by purely stochastic (i.e., involving
chance or probability) processes. These processes are random or variable and are not
characterized by a single value but can be described by a distribution. Thus, estimation of the
moments of the distribution (e.g., mean, variance, skewness) is possible.
The TRIM framework includes an approach to estimate uncertainty and variability in a
aiunner that allows for integration between the TRIM modules while tracking the uncertainty and
variability through the modules. The remainder of this section briefly describes the approach for
estimating uncertainty and variability. For a more rigorous explanation on uncertainty and
variability and implemention in the TRIM framework, please refer to the 1999 TRIM Status
Report (U.S. EPA 1999c).
The EPA chose a staged approach for analyzing uncertainty and variability since it has
several advantages for models as complex as TRIM. The first stage, which is comparatively easy
to implement, is known as a sensitivity and screening analysis. This stage involves identifying
influential parameters and developing an importance-ranking of parameters to focus and reduce
the number of parameters analyzed in the uncertainty and variability analysis. The sensitivity
and screening analysis computes the importance of parameters by calculating the extent to which
model results change when parameters are varied singly or in pairs. This process provides for a
first-order determination of the most influential parameters and allows further analysis to focus
on the key parameters. Furthermore, this screening approach narrows down the scope of the
detailed analysis in the second stage and reduces the number of parameters by identifying
influential parameters that should be retained for further analyses. This is a critical step toward
the goal of producing an economical representation of uncertainty and variability by excluding
unnecessary terms and parameters and still capturing all of the significant features of model
uncertainty and variability.
The second stage of TRIM.Expo's uncertainty and variability analysis is more complex
and involves a Monte Carlo approach. Monte Carlo methods analyze model uncertainty by using
statistical sampling techniques to estimate statistics that characterize uncertainty. Essentially, a
NOVEMBER 1999 4^28 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
Monte Carlo approach entails performing many model runs with model inputs that are randomly
sampled from specified distributions for the model inputs. These can be set up to characterize
the propagation of uncertainty and variability of the model input parameters, taking into account
distributions of parameter uncertainty and variability and parameter dependencies. These
simulations provide uncertainties of model outputs in terms of distributions of model outputs,
joint distributions of model inputs and outputs, and summary scalar measures (i.e., the core data
from which information about uncertainty and variability can be extracted).
This two-stage analysis of uncertainty and variability will be performed for each of the
TRIM modules to study the propagation of uncertainty. The Monte Carlo simulations for each
TRIM module will be performed sequentially in the order that the TRIM modules are run.
Sufficient information must be transmitted from one module to the next to be able to propagate
distributional information to succeeding Monte Carlo simulations. Since the amount of data
produced from Monte Carlo simulations is voluminous, the full results will be archived and a
reduced set will be retained as input for the next module. The output values from each of the
TRIM modules as well as the model inputs (i.e., parameter values) for each Monte Carlo
simulation will be saved and can be passed along to the next module for subsequent uncertainty
analysis. However, the amount of data would increase drastically from one module to the next.
Therefore, it is important to track the input parameters to each TRIM module.
To reduce the size and complexity of the transfer of uncertainty and variability
information between TRIM modules, the results can be summarized in the form of non-
parametric probability distributions that can be passed to the next module, where each
distribution to be passed is characterized non-parametrically by its percentiles.
The screening stage described above involves a sensitivity ranking analysis to select the
critical parameters to be tracked for the more detailed uncertainty analysis: all other parameters
would be set at their central tendency value. This focus on critical parameters also decreases the
amount of information to be tracked. To further reduce the volume of information, after
summarizing the results from one module as probability distributions, the transmission of
information to the next module is filtered to select the most critical parameters (e.g., those that
account for 95 percent of the variance of the uncertainty and variability).
It is important to note that since data for many of the input variables in TRIM.Expo (e.g.,
time-activity patterns, ventilation rates, air exchange rates) are not available to cover the entire
range of possible values, Monte Carlo simulations are used to better represent the distribution of
exposures that occur among the people in an exposure assessment. While this method better
captures the variability of the population's exposures, it underestimates the repetitive nature of
certain high exposure situations (e.g., long commutes in slow-moving traffic or preparing foods
in a manner that increases the release of toxic pollutants) for certain segments of the population.
Furthemore, it should be noted that much of the data used to characterize variability
related to exposure, such as time-activity patterns or other exposure factors are sometimes based
on short-term (multiday) measurements (e.g., diary studies, recall, monitoring). Despite the
limitations and assumptions built into using such data to generate long-term factors, OAQPS
believes that these data are the best available and extrapolating to generate long-term factors or
NOVEMBER 1999 ฃ29 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK .-VXD CONCEPTUALIZATION OF TRIM Expo
creating a longer term sequence of such data is the best approach. It should be reiterated that
TRIM, due to its architecture, is a dynamic modeling system, not a static model, and will be able
to undergo frequent updates that reflect new science and information. Therefore, once alternative
and improved approaches (e.g., correlated diaries for activity and consumption information at the
individual level developed through statistical analysis of multiday diaries) become available, they
will be evaluated and incorporated into TRIM where appropriate
4.3 DATA INPUT REQUIREMENTS
Inputs to TRIM Expo include environmental media concentrations, intermedia transfer
factors, and/or the input needed to calculate intermedia transfers from algorithms, cohort activity
data, and demographic and at-risk population data. The sections below summarize the principle
sources of input data in these areas
Although TRIM.Expo is primarily a stochastic model, distributions for many of the
model's inputs are either not available or are incomplete Through the use of a systematic
sensitivity analysis, exposure pathways and parameters can be identified by the model that do not
make a major contribution to the assessment endpoint nor to the overall uncertainty and
variability The use of point estimates can then be considered for these parameters (U S EPA
1997c). Conversely, TRIM Expo will use Monte Carlo analysis for those parameters that are
found to have a significant impact on exposure and that have data available. Hence, assessments
in TRIM.Expo will be able to use a combination of distributional data for input parameters that
are judged to be important by the analyst, and point estimates for those parameters that are either
determined to contribute little to the exposure outcome or that have insufficient data available
Whichever distribution type is selected for a particular parameter, the analyst will be informed by
the model of the choices made and the implications that may result
A probability density function (PDF) will be developed for many of the input variables
used in TRIM Expo For continuous random variables, the PDF expresses the probability that the
random variable falls within some very small increment (U.S. EPA 1997c). It must be emphasized
that distributions for many of the exposure factors in TRIM.Expo have not been developed to
date, hence, a major effort is required to develop the PDFs for input to the model
Probability density functions can be characterized by numerous distributional forms
(Cullen and Frey 1999). For each TRIM.Expo application, the analyst needs to address several
critical issues regarding data sorting and manipulation These issues include, but are not limited
to the following (1) how to develop a PDF based on limited data; (2) how to set the confidence
or uncertainty limit for a PDF, (3) how to truncate PDFs to eliminate unrealistic scenarios, and (4)
how to use data that is below its detection limit. The EPA has convened workshops to address
important issues regarding data distributions and their variability and uncertainty (U.S EPA
1996b, U.S. EPA 1999g) The reader is urged to refer to these workshop reports for more in-
depth information regarding these issues.
Another issue concerning the development and use of data inputs is that as studies provide
new information on exposure factors, discrepancies may appear between the factors in one study
NOVEMBER 1999 4^30 TRIM.ExPo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM.Expo
as compared to those from a different study This complicates the selection of exposure factors
for use in TRIM.Expo Therefore, whenever possible, the developers of TRIM.Expo will use
exposure factors that are consistent across EPA program offices Furthermore, they will draw on
the efforts of, and work cooperatively with, ORD and the other program offices to determine the
most appropriate exposure factors and will update these periodically to reflect the most current
knowledge in this field
4.3.1 ENVIRONMENTAL MEDIA CONCENTRATIONS
The required environmental media concentration data input for TRIM.Expo includes
ambient pollution concentration estimates For the exposure duration of interest in TRIM Expo, a
concentration profile for each of the principal environmental media (i.e., air, soil, water,
vegetation) included in the analysis for each exposure district is needed. Pollutant concentration
data for TRIM Expo is assumed to be available from analysis of monitoring data, from a
dispersion model, or from a multimedia transport model such as TRIM FaTE The two
environmental media for which monitoring data are most abundant are air and water
The OAQPS is responsible for managing a monitoring network of ambient fixed-site
monitors for air pollutants The OAQPS uses this network of monitors to help ensure that the
levels of the criteria pollutants (i.e., carbon monoxide, nitrogen dioxide, sulfur dioxide, ozone,
paniculate matter, lead) do not exceed the NAAQS established to comply with the CAA The air
quality measurements are collected by monitors at thousands of sites across the nation operated
by state and local environmental agencies Each monitor measures the concentration of a
particular pollutant in the air Monitoring data indicate the average pollutant concentration during
a time interval, usually one hour or 24 hours The monitoring stations in this network are called
the State and Local Air Monitoring Stations (SLAMS). To obtain more timely and detailed
information about air quality in strategic locations across the nation, OAQPS established an
additional network of monitors, called the National Air Monitoring Stations (NAMS). The
NAMS sites, which are part of the SLAMS network, must meet more stringent monitor siting,
equipment type, and quality assurance criteria
The CAA also stipulates that each state include a comprehensive inventory of existing
sources of air pollution and an accurate estimate of the total amount of pollutants (e.g., VOCs)
emitted to the air by each source during a calendar year These pollutant emissions estimates are
compiled by operators of industrial and commercial enterprises, state and local environmental
agencies, and EPA. The amount of pollutant is calculated using EPA-approved methods and
measurable factors such as the quantity of fuel used.
All of these emissions data are compiled in a single database called the Aerometric
Information Retrieval System (AIRS) The AIRS contains all of the air quality, emissions,
compliance, and enforcement information that OAQPS and state agencies collect to carry out
their respective programs for improving and maintaining air quality
The EPA's Office of Water (OW) manages information collected on pollutant
concentrations measured in waterbodies. The data are collected from monitoring conducted at
regular sites on a continuous basis, at selected sites on an as needed basis or to answer specific
NOVEMBER 1999 4-31 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
questions, on a temporary or seasonal basis, or on an emergency basis The responsibility to
monitor water quality rests with many different agencies. State pollution control agencies have
key monitoring responsibilities and conduct vigorous monitoring programs. Many local
governments, such as city and county environmental offices, also conduct water quality
monitoring In addition, other Federal agencies are also involved in water quality monitoring.
For example, the U.S. Geological Survey (USGS) conducts extensive chemical monitoring
through its National Stream Quality Accounting Network (NASQAN) at fixed locations on large
rivers around the country
Many agencies and organizations maintain computerized data systems to store and manage
the water quality data they or others collect. Perhaps the single largest such ambient water quality
data system is EPA's STORET (STOrage and RETrieval) system. Much of the data collected by
state, local, and federal agencies and by some private entities such as universities and volunteer
monitors are entered into STORET Raw data in STORET can be accessed, analyzed, and
summarized by many users and for many purposes
The TRIM Expo module will be capable of using data from multimedia transport models
as input data for ambient pollutant concentrations, particularly for multimedia, multipathway
pollutants One such model is the TRIM FaTE module. The TRJM.FaTE module can estimate
pollutant concentrations in multiple environmental media and biota, and accounts for transfer of
mass throughout an environmental system It models the movement of pollutant mass over time
through a bounded system, which includes both biotic and abiotic components. The boundaries of
the system are defined by the user, and can easily conform to the exposure districts in a
TRIM.Expo application
4.3.2 CONCENTRATIONS OF POLLUTANTS IN MICROENVIRONMENTS
Pollutant concentrations will need to be supplied to TRIM.Expo for all microenvironments
in which significant contributions to exposures occur. When measured data are unavailable, there
are various techniques for estimating these microenvironmental concentrations These techniques
are discussed below, with a focus on ways of estimating microenvironmental concentrations of
pollutants in air
4.3.2.1 Indoor Versus Outdoor Concentrations
Studies that simultaneously measure indoor and outdoor concentrations of various
pollutants show that numerous factors affect a pollutant's ability to penetrate to the inside of a
building from outdoors (Johnson et al. 1996c). One of the most comprehensive measurement
studies of indoor/outdoor concentrations of air pollutants conducted to date is the Total Exposure
Assessment Methodology (TEAM) (U.S. EPA 1987). This multi-year study was conducted in the
early to mid 1980's During the study, measurements of the exposures of 600 individuals to 20
target chemicals in both air and water were made. The 600 study participants were a
representative sample of a total population of 700,000 residents of cities in New Jersey, North
Carolina, North Dakota, and California One of the most important findings of the TEAM study
was that personal (air) and indoor exposures to VOCs are nearly always greater than outdoor
levels (Wallace et al 1991, U.S. EPA 1987). The TEAM studies answered a question of equal
NOVEMBER 1999 4-32 TRIM.ExPO TSD (DRAFT)
-------
CHAPTER 4
DESIGN' FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
importance for modeling exposures to air pollutants, that is. "Is there any indoor and outdoor
relationship associated with the variation of one or more air pollutants measured simultaneously at
a residence9" In the New Jersey TEAM study, results showed that (1) during times conducive to
accumulation of high outdoor VOC concentrations, substantial contributions to indoor levels can
be made, and (2) in homes where there are no indoor sources of a VOC compound, the indoor
concentration variation can be driven by outdoor VOC levels (Lioy et al. 1991).
4.3.2.2 Mass Balance Model Approach
For estimating air pollutant concentrations for indoor microenvironments, TRIM.Expo
will include the option of using a mass balance modeling approach where sufficient data exist In
general terms, the mass balance model can be described as
The change in indoor pollution concentration =
the pollutant entering from outside
- the indoor generation of pollutant
pollutant leaving the indoor microenvironment
removal of the pollutant by an air cleaning device
decay of the pollutant indoors.
Each term in the above conceptual model requires data inputs. Implicit in the above terms are
data needs for other variables such as infiltration, air exchange rate, surface reactivity, building
volume, and recirculation Additionally, air exchange removal can be separated into fractions
removed for unfiltered and filtered air, and similarly penetration of outdoor air can be separated
into fractions for penetration of filtered and unfiltered air. Another factor which has a great
influence on several of these variables is the type and use of air conditioning One of the most
ubiquitous sources of indoor air pollutants is smoking. This source type (when present) is
responsible for elevated indoor concentrations of numerous air pollutants. Section 521 of this
report describes in more detail the mass balance model that will be included in the initial inhalation
prototype of TRIM Expo
McCurdy (1994) discusses the "indoor data factors and parameters" that were required
inputs to the mass balance model used in pNEM/O3 These factors, along with seasonal
considerations, affect how much outdoor air pollution is estimated to come inside and how much
indoor sources affect the indoor exposures For the pNEM/O3, values for most of the variables in
its mass balance model are obtained by Monte Carlo sampling from empirical distributions of
measured data The variables include (McCurdy 1994):
Probability of having open windows, given the type of air conditioning system found in
the home and the outdoor temperature,
Air exchange rate, which is also affected by the "window status",
Residential surface-to-volume ratio, which is necessary to determine surface reactivity
losses; and
NOVEMBER 1999 4-33 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION OF TRIM Expo
Emission rates from any indoor sources and the probability that the sources will be
operating or not.
There is an ever increasing database for all of the factors mentioned above For example,
the Source Ranking Database provides data on indoor source emission factors (Johnston et al
1996). In addition, the U.S. Census Bureau, as part of its American Housing Survey, routinely
collects information on housing and building characteristics, including the use of air conditioning,
methods of heating, square footage, and number of rooms. Researchers have become much more
aware of the importance of seemingly minor factors such as whether a house's windows were
open or shut when conducting studies of indoor and outdoor air pollutants. As a result, new
studies are providing distributions for these important factors Johnson et al. (1996a) provides a
literature survey of recent studies for many of these factors EPA will examine available
databases such as the Source Ranking Database and American Housing Survey for relevant
information on indoor sources and housing and building characteristics that can be used in
TRIM Expo
4.3.2.3 Empirical Indoor/Outdoor Ratios Approach
The use of empirical indoor/outdoor relationships for estimating pollutant concentrations
for indoor microenvironments is an approach that is often used when measured pollutant data or
data needed for a mass balance model are lacking This approach relies on observed relationships
between outdoor concentrations and concurrent indoor or in-vehicle microenvironmental
concentrations for selected pollutants The Hazardous Air Pollutant Exposure Model
(HAPEM4), a model similar to pNEM, uses this approach for estimating indoor and in-vehicle
microenvironmental concentrations The HAPEM4 uses a method of indoor/outdoor ratios called
"microenvironmental (ME) factors" for determining the pollutant concentration in each indoor or
in-vehicle microenvironment.
The EPA's OAQPS is currently collecting and analyzing data relevant to the development
of ME factors for HAPs corresponding to those addressed in Section 112(k) of the Clean Air Act
Amendments of 1990. Eventually, ME factors will be developed for all HAPs listed under Title
III of the CAA. In addition to the traditional ME factor (also referred to as the penetration
factor), data will be collected whenever possible for a separate factor relating the closeness of key
sources to a microenvironment Referred to as the proximity factor, this factor accounts for the
higher concentrations usually expected when a receptor moves closer to a source The resulting
ME factors used in HAPEM4 are given by the expression:
C,n(i,k,t,cq) = \\r(k,cq)lCout(i,t,cq)\\a(k,cq)\\ (4-13)
where.
CJi,k,t,cq) = Average pollutant concentration in microenvironment k at exposure
district;' during time step t for calendar quarter cq.
y(k,cq) = Penetration factor (relates outdoor concentration to indoor or in-
vehicle microenvironmental concentration)
NOVEMBER 1999 4-34 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK AND CONCEPTUALIZATION" OF TRIM.Expo
Cm,(i,t,cq) = Ambient pollutant concentration taken at a fixed-site monitor or
from a modeled value.
a(k,cq) = Proximity factor (relates the closeness of key sources to a
microenvironment).
Note that the ME factors typically do not vary by exposure district or from one time step to the
next However, they may be affected by calendar quarter (i.e., by season).
4.3.3 ACTIVITY PATTERN DATA
Activity pattern data are used to determine the frequency and duration of exposure for
specific groups within various microenvironments As mentioned in Chapter 2, information on
activity patterns are taken from measured data collected during demographic surveys of
individuals' daily activities, the amount of time spent engaged in those activities, and the locations
where the activities occur Two common methods for collecting these data are through diary
studies and activity recall studies Diary studies involve a volunteer carrying a specially designed
time-activity diary with them during their daily routine. They use the diary to record the start time
of the activity, their location at the time, a description of the activity, the time the activity ended
or changed to another one, and, sometimes, an estimate of their breathing rate Diaries can be
designed to obtain additional information for specific purposes or study requirements In an
activity recall study, respondents are asked to complete a questionnaire that details their activities
from memory. The respondent is typically asked to recall his or her activities at the end of each
day for the preceding 24-hour period Data from this type of study are usually not as detailed as
However, activity recall studies can generally be conducted on larger populations than diary
studies, since large numbers of respondents can be contacted by telephone
In addition to recording the duration and location of a person's activities, important
demographic information about the person is also then collected The demographic information
usually includes the person's age, gender, and ethnic group. Most activity pattern studies also try
to collect information on other attributes of a respondent, such as highest level of education
completed, number of people in their household, whether the person or anyone in their household
is a smoker, employment status, and the number of hours spent outdoors. These are a sample of
the possible items that might be requested on a questionnaire
One of the largest databases for human activity pattern data is EPA's Comprehensive
Human Activity Database (CHAD) (Glen et al 1997). CHAD is comprised of nearly 17,000
person-days of activity pattern data At present, 140 activities and 114 locations are included in
CHAD The data have been collected and organized from eight human activity pattern surveys.
CHAD contains the sequential patterns of activities for each individual. Each activity has a
corresponding location code so that the microenvironment of each activity is known The
activities in CHAD range from one minute to multiple hours in duration (activities longer than one
hour are broken into segments and do not cross over from one clock hour to the next). The
CHAD also provides an indicator of the rate of energy expenditure during a particular activity for
each exposure event. This indicator is the metabolic equivalents or "METs" (see Section 6.1.5
for a description). In addition, CHAD includes an estimate of the body mass for each individual.
This parameter is important for calculating uptake and dose (McCurdy 1999).
NOVEMBER 1999 4-35 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 4
DESIGN FRAMEWORK .AND CONCEPTUALIZATION OF TRIM Expo
The TRIM.Expo module will have the ability to use the data in CHAD or, alternatively,
use the data from a particular survey directly, providing the data are formatted properly The
CHAD will be updated periodically as additional data from new time/activity pattern surveys
become available In this way, a user has the two option mentioned above for using time/activity
survey data in TRIM Expo.
4.3.4 DEMOGRAPHIC AND AT-RISK POPULATION DATA
One of the purposes of TRIM.Expo is to perform analyses for subsets of the population
that are particularly at risk to exposure to pollutants because of age or preexisting medical
conditions There are many sources of information detailing the demographic character of the
population for the U S. The originating source for most of this information is the Bureau of the
Census (BOC), which compiles detailed demographic information about the U.S population
every ten years The information is collected for the entire country at the census "tract" level or
for the smaller census "block " Census tracts are small, relatively permanent statistical
subdivisions of a county Census tracts usually include between 2,500 and 8,000 persons Census
tracts do not cross county boundaries The spatial size of census tracts varies widely depending
on the population density of the area. Census blocks are smaller than census tracts in areal extent
The BOC collects data on numerous aspects of the demographic character of U.S.
citizens, including national and state population trends and projections; geographical mobility;
school enrollment, educational attainment, households and families, marital status and living
arrangements; fertility; child care arrangements, child support, disability, health insurance; labor
force and occupation; income and poverty, and characteristics of various ethnic and elderly
populations Much of this information can be useful for understanding the demographic patterns
that put segments of the population at risk from exposures to environmental pollutants
Although the census is a good source of information about the geographic, housing,
business related, and demographic characteristics of the U S population, it is limited in the
amount of information available about the general health status of the population. However, there
are other sources of information available about the incidence of diseases and illnesses in the U.S.
For example, the National Center for Health Statistics (NCHS) publishes data from its National
Health Interview Survey (NHIS) The households selected to be interviewed each week in the
NHIS are a probability sample representative of the target population. Data are collected annually
from approximately 43,000 households including about 106,000 persons. Data collected includes
household composition, sociodemographic characteristics, basic indicators of health status, and
utilization of health care services Other information sources published by NCHS include the
National Health and Nutrition Examination Survey, Ambulatory Health Care Data, and the
National Maternal and Infant Health Survey. These databases can be used to estimate the size of
various at-risk population groups.
NOVEMBER 1999 4-36 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 5
INHALATION
5. INHALATION
This chapter provides details of the inhalation component of TRIM. Expo The structure
of the inhalation component will be consistent with the conceptualized framework for
TRIM Expo described in Chapter 4. In addition, the initial development of the inhalation
component will be based on the logic of pNEM This will provide a firm scientific foundation for
TRIM.Expo's inhalation exposure algorithms, so that they are responsive to OAQPS' need for a
scientifically-sound, human exposure model for inhalation.
5.1 OVERVIEW OF THE APPROACH
Because of the flexible structure of TRIM Expo, when performing an inhalation exposure
assessment, the user must make a number of selections regarding the input parameters The
following six subsections provide the generalized approach for modeling inhalation exposures in
TRIM.Expo This approach is adopted from pNEM/CO (Johnson et al 1999) Figure 5-lisa
schematic representation of the various input parameters and the resulting output of pNEM/CO
Figure 5-1
Schematic Representation of the Input Parameters
and Resulting Output of pNEM/CO
Air Exchange Rates I
Building Volumes
Emission Rates and
Use Patterns for
Indoor Sources
(e g gas appliances
passive smoking)
Ambient Fixed-Site
Concentrations
Distribution of People and
Occurrences of Exposures
Linked wtth Breathing Rale
NOVEMBER 1999
5-1
TRIM EXPO TSD (DRAFT)
-------
CHAPTER 5
IN HALATION
5.1.1 SELECTION OF STUDY AREA
The study area for inhalation exposures can be user-defined so long as estimates of
outdoor air concentrations for the pollutant(s) of interest are available. The data on outdoor
concentrations can be taken from ambient, fixed-site monitors or alternatively can be calculated
using either an air quality dispersion model (e.g., the Industrial Source Complex, or ISC, model),
TRIM.FaTE, or other air quality models. In TRIM.FaTE, the ambient air compartment is
characterized in terms of its gas phase, paniculate matter, and water composition.
The study area is divided into exposure districts. The exposure districts are spatial areas
with defined boundaries (either physical or political). If ambient monitoring data are used, then
the exposure district may be defined as the area within a given distance of the monitor. If
modeled air quality data are used, the exposure district can be defined according to the resolution
of the modeled data. Modeled air quality data that are organized in a gridded pattern can have
the exposure districts defined for each modeled grid-square or for aggregations of grid squares.
Alternatively, a modeled grid-square value can be associated with a geopolitical area for which
demographic information is available. For example, a modeled value could be assigned to the
census-tract to which it belongs, thus relating calculated outdoor concentrations to the population
information specified in census data.
5.1.2 SELECTION OF POPULATIONS OF INTEREST
When conducting an exposure assessment using TRIM.Expo, the user may select a
specific population group, or a set of groups, for which exposures will be estimated. Groups can
be defined by any number of attributes including age, gender, family income, work status, health
status (e.g., heart disease patients), or proximity to particular emission sources (e.g., natural gas
cooking fuel). Since the movements and daily activities of the population of interest will
determine which exposure media the individuals contact and in which exposure districts the
contacts occur, data about the behavior of the population are required. Typically, the entire
population for all of the exposure districts in the study area is included. Then, through the
selection of cohorts, discussed below, exposure for subsets of the population are estimated.
Characteristics for defining cohorts can be geographic factors, demographic factors, or both.
Alternatively, the user may specify a set of individuals for an exposure assessment. In this case,
the user supplies additional information about the individuals, as discussed below.
5.1.3 DEFINITION OF POPULATION COHORTS
For an exposure analysis using groups, the population of interest, once chosen, is divided
into groups with similar attributes; these groups are called cohorts. Each cohort is assumed to
contain persons with similar exposures that are taken from the same probability distribution.
Each person is associated with only one cohort. The use of cohorts is a useful technique when
estimating the exposures of a large population with inadequate information about each
individual's activity profile. Aggregating information about people who are expected to have
similar exposures makes better use of the limited data that are available.
NOVEMBER 1999 5-2 TRIM.EXPO TSD (DRAFT)
-------
CHAPTERS
INHALATION
Cohort exposure is typically assumed to be a function of demographic group, location of
residence, and location of work or school. A demographic group is comprised of all individuals
that share one or more demographic features, such as a particular age (or age range), gender,
ethnic background, or occupation. Specifying the home and commuting district of each cohort
provides a means of linking cohort exposure to ambient concentrations. Specifying the
demographic group provides a means of linking cohort exposure to activity patterns that vary
with age, work or school status, and other demographic variables. In some analyses, cohorts are
further distinguished according to factors relating to proximity to emission sources (e.g., an
indoor source such as a gas stove) or time spent in particular microenvironments.
5.1.4 DEVELOP AN INHALATION EXPOSURE-EVENT SEQUENCE FOR EACH
COHORT
When performing an inhalation exposure analysis using TRIM.Expo, information is
needed about each "location" that each individual or cohort visits during their daily activities
(e.g., in the kitchen at home, outdoors at school, indoors at work). These locations are called
microenvironments. An important feature of TRIM.Expo is the ability of the user to vary the
scale of the microenvironments to individual model applications. This feature is important
because it allows the user to relate the size of a microenvironment to the potential for exposure
from different pollutants.
In TRIM.Expo, the inhalation exposure of each individual or cohort is determined by an
exposure-event sequence specific to the individual or cohort. Furthermore, the exposure-event
sequence for a particular cohort applies to all individuals in that cohort. The exposure-event
sequence is a chronologically-ordered series of events which identifies locations and activities
and the amount of time spent performing each activity in each location. In addition, the
exposure-event sequence is specific to the day of the week. The information about the day of the
week is obtained from the activity pattern database. Each exposure-event sequence consists of a
series of events with durations ranging from 1 to 60 minutes. Once an exposure-event sequence
is selected to represent the daily exposure of a cohort, it is followed through for an entire 24-hour
period. A different exposure-event sequence is selected for each day in the study period. The
TRIM.Expo module retains the information about the day of the week and season throughout the
entire analysis because these variables can affect exposure results. Each exposure event assigns
the cohort to a particular combination of exposure district, microenvironment, and activity (e.g.,
cooking, playing, resting). Although no two individuals' exposure will be exactly the same due
to the myriad of factors that affect a person's exposure; for the purposes of estimating a
population's exposure, especially considering the dearth of long-term time/activity information
for a large enough cross section of the population, the model uses the simplifying assumption
that all individuals within a particular cohort have the same exposure.
Information about the exposure-event sequences can be obtained by sampling from a
human activity database. The human activity database is made up of diary and telephone survey
records which identify a study participant's daily activities and locations during a 24-hour period.
Because each participant of most activity diary studies provides data for only a few days, the
construction of a longer exposure-event sequence requires either the repetition of data from one
participant or the use of data from multiple participants. The latter approach is being used in the
NOVEMBER 1999 5^3 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
initial development of TRIM.Expo to better represent the variability of exposure that is expected
to occur among the persons included in the cohort. The need to extrapolate short-term activity
diary information to chronic exposure assessments is a recognized shortcoming in long-term
exposure studies. There is a critical need for data on long-term activity patterns that can be used
for constructing year-long exposure-event sequences. This issue, and discussion of an alternative
statistical approach for augmenting short-term activity pattern diary data, are expanded upon in
Section 5.4.2.
For the initial development of a TRIM.Expo Prototype, a compilation of time-activity
surveys will be used for a cohort analysis. These surveys have been organized into a single
database called the Consolidated Human Activity Database (CHAD) which was described
previously in Section 4.3.3. The developers of CHAD have supplemented the activity pattern
survey information with data showing the day of the week that a diary entry was made and also
the maximum outdoor air temperature for that day. Knowledge of the day of the week is
important when constructing exposure-event sequences since human activities are usually quite
different for weekdays than they are on weekends (U.S. EPA 1996a). Providing information
about the maximum outdoor air temperature that occurred on the day that a diary entry was made
is a useful method for selecting activity data that account for seasonal variation when
constructing year-long exposure-event sequences. For an individual analysis, the user must
either provide demographic information about the individuals so that appropriate activity pattern
data can be extracted from CHAD, or directly provide the time sequence of exposure
district/microenvironment/activity pattern combinations, as well as the demographic data related
to breathing rate (i.e., age, gender, body weight).
5.1.5 ESTIMATE POLLUTANT CONCENTRATION AND VENTILATION RATE
ASSOCIATED WITH EACH EXPOSURE EVENT
The exposure-event sequence defined for each individual or cohort is used to determine a
corresponding sequence of exposures, event-by-event. Each inhalation exposure is defined by a
pollutant concentration and a ventilation rate indicator. The applied dose is a function of the
pollutant concentration, the demographic characteristics of the individual or cohort affecting
breathing rate, and the ventilation rate values assigned to the activity.
The first step in estimating the microenvironmental pollutant concentrations is to estimate
the ambient pollutant concentrations. As discussed in Section 5.1.1, these are estimated from
either fixed-site monitoring data, through the use of an air dispersion model, or from
TRIM.FaTE. Next, microenvironmental concentrations are calculated from ambient
concentrations and data on microenvironmental emission sources for indoor microenvironments
(1) through the use of mass balance algorithms (described in Section 5.2.1), (2) with intermedia
transfer factors, and/or (3) with measurements of concentration increments associated with
indoor sources. Intermedia transfer factors are empirically derived and relate outdoor
concentrations to the concentration contributions in the various indoor microenvironments used
in TRIM.Expo. In addition, measurements of concentration increments associated with certain
outdoor microenvironments (e.g., gas stations, parking garages) may be used if they are not
modeled or monitored explicitly. For other special outdoor microenvironments (e.g., near
NOVEMBER 1999 5-4 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
roadways), statistical analysis of ambient-to-microenvironment concentration relationships may
be used to estimate microenvironmental concentrations based on fixed-site monitoring data.
Concentrations are determined for each microenvironment in each exposure district for
each time step in the exposure-event sequence. These concentrations constitute the values for
Cm(i,k,l,t) in Equation 4-1. For inhalation, the concentrations are calculated for air only;
therefore, the only exposure medium m in Equation 4-1 is air.
In TRIM.Expo, an array of microenvironmental pollutant concentration values Cm are
created for each individual or cohort. Each array consists of a set of year-long sequences of
hourly-averaged Cm values; one for each combination of exposure district, microenvironment,
and activity. In the initial development of TRIM.Expo, the district will be either the home, work,
or school district specified for the individual or cohort. For an exposure event during time step t,
an individual or cohort is assigned the value of Cm specified for that time step in the designated
exposure-district/microenvironment/activity sequence.
In addition to the pollutant concentration, a ventilation rate (VE) value is estimated for
each exposure event. VE is expressed as liters of air respired per minute (liters min"1). The
procedure for calculating VE is summarized below. An approach to estimate the various values
and relationships needed to model the ventilation rate using Equations 5-3, 5-4, and 5-5 (below)
is described in Appendices C and D of Johnston et al. (1999).
The CHAD database provides an activity indicator for each exposure event. Each activity
type is assigned a distribution of values for the metabolic equivalent of work (MET}. The MET
is a dimensionless quantity defined by the ratio:
MET = EE/RMR (5-1)
where EE is the rate of energy expenditure during a particular activity (expressed in kcal/min),
and RMR is a person's typical resting metabolic rate (also expressed in kcal/min).
A probabilistic procedure is used to assign a RMR value to cohort for a typical 365-day
exposure period. An EE value is calculated for each exposure event by the equation:
EE0(r,p,z) = [MET(r,p,z)][RMR(z)] (5-2)
in which EE0(r,p,z) is the average energy expenditure rate (kcal min'1) for cohort z during
exposure event r on day/?; MET(r,p,z) is a value randomly selected from the distribution of MET
values associated with each activity type in CHAD; and RMR(z) is the RMR value randomly
generated for cohort z.
Energy expenditure requires oxygen, which is supplied through ventilation (respiration).
Let ECF(y) indicate an energy conversion factor defined as the volume of oxygen required to
produce one kilocalorie of energy in person >>. The oxygen uptake rate (VO2} associated with a
particular activity can be expressed as:
NOVEMBER 1999 5-5 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
V02(r,p,z) = [ECF(y)][EEa(r,p,z)] (5-3)
in which VO2(r,p,z) has units of liters oxygen min"1, ECF(y) has units of liters oxygen kcal"1, and
EEa(r,p,z) has units of kcal min"1. The value of VO2(r,p,z) is determined from MET(r,p,z) by
substituting Equation 5-2 into Equation 5-3 to produce the relationship:
V02(r,p,z) = [ECF(y)][MET(r,p,z)][RMR(y)] (5-4)
Ventilation rate (Fฃ) tends to increase as VO2 increases up to the point of maximum
oxygen uptake (VO2mca). The relationship is known to be non-linear, with the slope of the
relationship usually increasing at higher values of VO2. The relationship between VE(r,p,z)) and
VO2(r,p,z) is modeled by the generic equation:
ln[VE(r,p,z)/BM(y)J = a + (\>){ln[VO2(r,p,z)/BM(y)]} + d(y) + e(r,p,z) (5-5)
in which VE(r,p,z) is the VE value associated with the rth event of day p for person y, BM(y) is the
body mass assigned to person y, and a and b are constants determined by the age and gender of
person y. The term d(y) is a random variable selected for each person from a normal distribution
with mean equal to zero and standard deviation equal to ad. The term e(r,p,z) is a random
variable selected for each individual event from a normal distribution with mean equal to zero
and standard deviation equal to oe.
5.1.6 EXTRAPOLATE THE COHORT INHALATION EXPOSURES TO THE
POPULATIONS OF INTEREST
For a population analysis, the inhalation exposures calculated for the cohorts can be
extrapolated to the larger general population by estimating the number of individuals in each
cohort. First, the population of each demographic group that resides within a particular exposure
district is extracted from census data specific to that district. This gives an estimate of the
population of each non-commuting cohort residing within each exposure district. Then, as
described in Equation 4-12 (shown below), the populations of the commuting cohorts (assumed
to include all cohorts of working adults and school children) are determined by the expression:
com(dg,h,\v,b) = pop(dg,h,b) x com(h,\v) /com(h)
where com(dg,h,\v,b) is the number of persons in the commuting cohort associated with
demographic group dg, residing in exposure district h (i.e., the home district), commuting to
exposure district w (i.e., the commute district) and having attribute b (e.g., the incidence of a
particular disease or ailment). The pop(dg,h,b) is the population of demographic group dg
residing in exposure district h that has attribute b. The com(h,\v) is the number of commuters in
all demographic groups that commute from their residence in exposure district h to work or
school in exposure district w, and com(h) is the total number of commuters that reside in
exposure district h.
NOVEMBER 1999 5-6 TRI M. EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
5.2 PRESENTATION OF THE MODEL ALGORITHMS BY
MICROENVIRONMENTAL LOCATION
The TRIM.Expo module will be able to model inhalation exposures for several indoor, in-
vehicle, and outdoor microenvironments As mentioned earlier in this chapter, a user will have
the ability to specify additional microenvironments of various scales to fit individual modeling
requirements In this section, the general algorithms are presented for each of these locations
The methodology presented here for calculating indoor, in-vehicle, and outdoor
microenvironmental concentrations conforms to the requirements specified earlier for TRIM
framework development. One of the important goals for modeling microenvironmental
concentrations is that the methodology conserve mass, where appropriate and feasible
Alternative methodologies will also be included as options when the information required for mass
balance is not available
5.2.1 MICROENVIRONMENTAL LOCATIONS SPECIFIC TO INDOOR AIR AND
INSIDE VEHICLES
The TRIM Expo module will include algorithms that can be used to estimate pollutant
concentration for several indoor microenvironments, such as residences, residential garages, the
work place, school, and other indoor locations such as restaurants and stores In addition, the
inside of passenger vehicles, such as automobiles and buses, will be treated similarly to the indoor
microenvironment In general, there are two types of contributions to the pollutant
concentrations in these microenvironments infiltration of air from outside the microenvironment's
boundaries, and direct emission of a pollutant of concern from a source within the
microenvironment Infiltration will be modeled by TRIM.Expo using either a mass balance
approach, described below, or an assumed transfer factors (ME factors)
For indoor emission sources, TRIM.Expo will provide two options. For the first option,
information on the emission rate (in units of mass/time) for a source is used as an input to the
mass balance model For the second option, sample values are drawn stochastically from a
distribution that relates the presence of an indoor source in a particular microenvironment to
incremental increases in pollutant levels For example, to estimate the contribution to pollutant
concentrations in the home from tobacco smoking using the first option, the user may specify the
frequency of smoking (e.g., the number of cigarettes per hour), which TRIM.Expo will use to
derive a pollutant emission rate for input to the mass balance equation. If smoking frequency
information is unavailable, the user may simply indicate that smoking occurs in the home. In that
case, TRIM Expo will estimate the contribution to pollutant concentrations from smoking by
sampling from a distribution of the measured increase in pollutant concentrations in homes of
smokers Alternatively, the user may supply his or her own distributional data.
The TRIM.Expo mass balance model and description are adapted from Johnson et al.
(1996b) The mass balance model is based on the generalized mass balance model presented by
Nagda et al. (1987) for a single indoor compartment. As originally proposed, this model uses the
assumption that pollutant concentration decays indoors at a constant rate. However, Johnson et
al (1996b) reports that pollutant decay rate is a function of the indoor pollutant concentration.
NOVEMBER 1999 5-7 TRIM. EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
Therefore, in TRIM.Expo, the model of Nagda and co-workers was revised to incorporate an
alternative assumption that the indoor decay rate is proportional to the indoor concentration. The
resulting model is expressed by the differential equation:
C = (1 - F_) vC + - M\C. - FC - - - (5-6)
in ^ B' out m d m '
cy
where:
COT = indoor concentration (mass/volume)
FB = fraction of the outdoor pollutant concentration intercepted by the building
or structure (dimensionless fraction)
v = air exchange rate (1/time)
Fj = pollutant decay coefficient (1/time)
Cml = outdoor concentration (mass/volume)
S = indoor generation rate (mass/time)
cV = effective indoor volume where c is a dimensionless fraction (volume)
M = mixing factor (i.e., the portion of the ventilation air flow that is completely
mixed with room air) (dimensionless fraction)
q = flow rate through air-cleaning device (volume/time)
F = efficiency of the recirculation air-cleaning device (dimensionless fraction)
The model is further generalized to include a mixing factor for outdoor air infiltration and
the possibility of infiltrated air from outside being filtered as follows:
. rr qF C
4c = (1 - FD) Mv C + - Mv C - FC - + (1 - F,) MvC - M\C.
fa in ^ B' u out f,y u in dm f.y x 2' ) out J m
(5-7)
where:
vu = air exchange rate, unfiltered (I/time)
vf = air exchange rate, filtered (I/time)
F, = efficiency of the recirculation air-cleaning device (dimensionless fraction)
F2 = efficiency of the outdoor makeup-air cleaning device (dimensionless
fraction)
Equation 5-7 can be simplified by substituting a "penetration factor," Fp, for the fraction
of the outdoor concentration intercepted by the enclosure and an "effective volume," Ve, for cV.
Fp and Ve are given by Equations 5-8 and 5-9, respectively:
Fp=\-FB (5-8)
Ve = cV (5-9)
NOVEMBER 1999 5^8 TRIM.EXPOTSD (DRAFT)
-------
CHAPTER 5
INHALATION
Substituting Equations 5-8 and 5-9 into Equation 5-7 results in:
C
C=FMvC + - Mv C . - F.C - i-^ + (1 - FJ M\ C , - M\C
in p u out y u in dm y v 2y y out J a
e e
(5-10)
Combining and rearranging terms yields:
C = M(F v + [1 - FJv)C + - M(v + \)C - F.C -
v J out v u m dm
. -
p u 2J f out y v u f m dm y
e e
(5-11)
Equation 5-1 1 can be simplified by combining terms proportional to Cm:
C = M(F v + [1 - FJv^C + - v'C
ฃJj in ^ p u L 2J f out y in
e
where:
e
v' = M(vu + vp + Fd + (5-13)
It can be shown that Equation 5-12 has the following approximate solution:
Cm (t) = kjCln (t-At) + k2C'out + k3 (5-14)
where:
k,=e-** (5-15)
M(F v + [1 - FJv)
'* '
k3 = (S/ v'VJ (1-kJ. (5-17)
NOVEMBER 1999 5^9 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
C'olll is the average value of the outdoor concentration over the interval / to t + At.
The average indoor concentration for hour h is given by C'm in the expression:
C ; (hj = a,Cm (h-1) + a2C 'oul (ho) + a3 (5- 1 8)
where Cm(h-l) is the instantaneous indoor concentration at the end of the preceding hour and C'm,
(h) is the average outdoor concentration for hour h. Also, a,, a2, and a3 are given by:
(5-19)
M(F v + [1 - F.]v)
- ฃJฑ - . - LX (1 - Z(h0)) (5-20)
"a = (1 - WW (5-21)
z(h(t) = (1 - ?V') / v' (5-22)
A steady-state version of the mass balance model (Equation 5-12) can be developed if it is
assumed that the change in indoor concentration with time is zero.
When information about the indoor emission source strength is not available, a3 may be
sampled from a distribution of measured incremental concentrations associated with the presence
of the indoor source. Some of these distributions will be included in TRIM.Expo. The user will
also have the option of supplying his or her own indoor source distribution.
5.2.2 MICROENVIRONMENTAL LOCATIONS SPECIFIC TO AMBIENT AIR
One of the options for obtaining pollutant concentration data for outdoor locations in
TRIM.Expo is through the use of air dispersion modeling. The output files from TRJM.FaTE or
another air model may be used if data are properly formatted. Regardless of the method for
modeling the dispersion of pollutants, the ambient pollutant concentrations at receptors must be
related to geopolitically defined exposure districts where people live, work, or attend school. For
example, suppose the user wants to design an exposure analysis to cover a large city and the
surrounding suburbs with an areal extent of 100 km within an urban area. For this example, a
typical TRIM.Expo exposure analysis could be conducted using census tracts as the exposure
districts. In that case, time sequences of hourly-averaged estimates of outdoor pollutant
concentrations for each census tract in the study area are required. These concentration estimates
could come directly from the air dispersion model, if census tract centroids were used as the
model receptors, or they may be derived from concentration estimates made at other receptor
points in the study area. There are several ways to do this.
NOVEMBER 1999 5^10 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
If the spatial resolution of model receptors is finer than the spatial resolution of exposure
districts, concentrations can be assigned to exposure districts from modeling receptors that fall
within each exposure district according to the formula:
c,i,0 (5-23)
where C, (v,t) is the average ambient pollutant concentration in exposure district / from the v
modeled receptor points for time step t, v, is the number of receptor points in exposure district /,
and C(c,i,t) is the ambient concentration at receptor point c within exposure district / during time
step t. The values ofC(c,i,t) are summed over the total number of receptor points in each
exposure district (i.e., v,).
If the spatial resolution of modeled receptor points is more coarse than the resolution of
exposure districts, concentrations can be spatially interpolated to the exposure district centroids
using the formula:
'^ (5-24)
where: C(c,t) = estimated concentration at receptor point c for time step t
d(c,i) = distance from a receptor point c and the centroid of exposure district /'.
Alternatively, exposure districts could be redefined as aggregations of contiguous census tracts,
with each tract assigned the concentration estimate at the nearest modeled receptor.
The second method for obtaining a time sequence of outdoor concentrations for each
exposure district is through the use of monitored ambient data. There are several limitations to
the use of monitoring data for air toxics. At present, the number of routine monitoring sites for
air toxics is much smaller than for criteria air pollutants. Also, air toxics are often measured as
24-hour integrated samples taken every sixth or twelfth day. However, with increased concern
about health effects from toxic air pollutants, EPA plans to increase the extent of its monitoring
efforts. In addition, future development of TRIM.Expo will make it easier to use a variety of
mathematical tools and spatial interpolation techniques such as krieging for estimating outdoor
pollutant concentrations.
Because the spatial resolution of monitors, even for criteria air pollutants, is typically
rather coarse, it is customary to specify exposure districts by assigning concentrations to census
tracts according to the values measured at the nearest monitor. Using this method, an air
pollutant's concentration is assumed to be the same for all census tracts within a particular
exposure district. In making these assignments, attention should be paid to the spatial area of
representation for the monitors. The EPA has four different monitor classifications as follows:
NOVEMBER 1999 5-11 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
1. Micro-scale: representative of from a few to 100 m;
2. Middle-scale: representative from 100 to 500 m;
3. Neighborhood-scale: representative from 0.5 to 4 km; and
4. Urban-scale: representative from 4 to 50 km.
There are a number of outdoor locations that may have enhanced pollutant
concentrations, such as gasoline stations, parking garages, and near roadways. As noted above, if
these microenvironments are not modeled or monitored explicitly for the particular study area, it
may be necessary to derive concentration estimates from measurements in those
microenvironments from other locations. Alternatively, it may be necessary to derive the
concentrations from the more generalized outdoor estimates and information about the
relationship between the generalized outdoor concentration and the outdoor microenvironment
(e.g., the distribution of ratios of CO concentrations near roadways to concentrations at other
outdoor locations). Some of this information will be provided in TRIM.Expo. The user will also
have the option of providing his or her own distributions.
5.3 INTEGRATION OF EXPOSURE ACROSS MULTIPLE LOCATIONS
AND TIMES
Figure 5-2
Hypothetical Exposure Profile
Covering 24 Time Steps
The TRIM.Expo module will have
the ability to integrate exposures of
varying durations across numerous
microenvironments. This is accomplished
by TRIM.Expo's exposure
characterization process. The purpose of
the exposure characterization process is to
combine and simultaneously track all of
the relevant information needed to assess
exposures across several exposure media
occurring with varying time-durations. In
this chapter, the exposure media of
concern are outdoor and indoor air. As
noted above, for each individual or
cohort, a sequence of exposure events is
defined. Exposure-event sequences are
chronological sets of events that define
the time-activity allocation of the individual or cohort. A simple example of an exposure-event
sequence is presented in Table 4-6 (Section 4.2.5). The exposure-event sequence tracks the
individual or cohort by (1) exposure district, (2) microenvironment, and (3) activity at each time
step. Each exposure event will be associated with an exposure concentration and a breathing
rate. The TRIM.Expo algorithms will use the information on the exposure concentration at each
time step to create an exposure time series or profile. Figure 5-2 shows an example of an
Time
NOVEMBER 1999
5-12
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
exposure profile covering 24 time steps. By combining the exposure concentration and the
breathing rate at each time step, TRIM.Expo will also create a potential dose profile (see Section
4.1.2). McCurdy (1997) summarizes the definition of dose profile as the collection of the
instantaneous intake doses over a time interval (to, t,), where the instantaneous intake dose is the
rate at which the pollutant penetrates into the target at a given instant of time.
5.4 SUMMARY OF INPUTS AND VALUES
Because the types of data used in a TRIM.Expo exposure analysis are quite diverse, each
one is described in a separate section below. These sections summarize these data inputs and
provide values for them wherever possible.
5.4.1 DATA INPUTS FOR THE MASS BALANCE MODEL
The mass balance model (Equation 5-12) requires information on the air exchange rate,
the building volume, the indoor generation of the pollutant, the fraction of the pollutant
penetrating the building from outdoors, and the pollutant decay rate. Many of these parameters
depend on several factors. For example, the fraction of pollutant mass in infiltration air that
actually enters the building (i.e., the penetration rate) depends upon the state of the pollutant (i.e.7
whether it is a gas, a fine particle, or a coarse particle), the type of building construction, whether
the building's windows are open, if windows are open, by how much, and the type of air
conditioning and/or air handling system. Therefore, data are not available for every pollutant and
every scenario.
Two factors that are important to the calculation of indoor concentrations which are not
pollutant-specific are the air exchange rate and building volume. Data on these factors have been
collected by numerous studies in different parts of the U.S. Some of these data were summarized
in Johnson et al. (1999) for the pNEM/CO model and are shown in Tables 5-1 and 5-2. Table 5-
1 shows the sources of information on the distributions for air exchange and building volumes.
Table 5-2 shows the references for specific microenvironmental air exchange rate data. These
microenvironments correspond to those currently used in pNEM/CO. Information on building
volume and air exchange rate will need to be developed for additional microenvironments for
TRIM.Expo system applications to other pollutants based on these and other databases.
NOVEMBER 1999 5-13 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
Table 5-1
Distributions and References for Air Exchange and Building Volume Data
Parameter
Air exchange rate, exchanges/h:
residence - windows closed
Air exchange rate, exchanges/h:
residence - windows open
Air exchange rate, exchanges/h:
nonresidential, enclosed
microenvironments, including
motor vehicles
Residential volume, cubic meters
Distribution of Parameter
Lognormal distributions by season
Lognormal distribution
See Table 5-2
Lognormal distribution
Reference
Murray and Burmaster 1995
Johnson, Weaver, Mozier et al.,
1998
See Table 5-2
Bureau of Census 1 995
Table 5-2
Distributions and References for Specific Microenvironmental
Air Exchange Rate Data
Microenvironment
General
Location
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Vehicle
Vehicle
Vehicle
Specific
Location
Nonresidence A
Nonresidence B
Nonresidence C
Nonresidence D
Nonresidence E
Residential
garage
Automobile
Other
Airplane
Activity Diary Locations Included in
Microenvironment
Service station or auto repair
Other repair shop
Shopping mall
Restaurant
Other indoor location
Auditorium
Store
Office
Other public building
Health care facility, School,
Church, Manufacturing facility
Residential garage
Automobile
Bus, Truck, Bicycle, Motorcycle,
Train/subway,
Other vehicle
Airplane
Distribution of Air
Exchange Rate
Distribution
Type
Lognormal
Lognormal
Lognormal
Lognormal
Lognormal
Lognormal
Lognormal
Lognormal
NA
Source of
Data
a
a
a
a
b
a
c
c
-
. -~.. _ ,) all non-school AER values provided by Turk et al (1989) and CEC (Lagus Applied Technology, Inc 1995)
"Data set containing all AER values provided by Turk et al (1989) and CEC (Lagus Applied Technology, Inc. 1995).
cOtt, Switzer, and Willis (1994)
NOVEMBER 1999
5-14
TRJM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
5.4.2 DATA INPUTS FOR TIME/ACTIVITY PATTERNS
The time/activity data for use in TRIM.Expo were obtained from CHAD. The CHAD is
comprised of approximately 17,000 person-days of 24-hour time/activity data developed from
eight surveys (Glen et al. 1997). The surveys include probability-based recall studies conducted
by EPA and the California Air Resources Board, as well as real-time diary studies conducted in
individual U.S. metropolitan studies using both probability-based and volunteer subject panels.
All ages of both genders are represented in CHAD. The data for each subject consist of one or
more days of sequential activities in which each activity is defined by start time, duration,
activity type (140 categories), and microenvironment classification (110 categories). Activities
vary from one minute to one hour in duration, with longer activities being subdivided into clock-
hour durations to facilitate exposure modeling. Refer to Section 4.3.3 for a more detailed
discussion of CHAD.
Extrapolating the information from short-term recall surveys to longer-term chronic
exposure assessments is currently a potential source of uncertainty in exposure modeling.
Additional research into longer term activity pattern data is needed to address this shortcoming.
The EPA's Office of Research and Development is embarking on research to develop statistical
methods to develop long-term exposure profiles from the activity pattern survey data that is
currently available (Ozkaynak 1999). As part of this effort, careful analysis of multiday diaries
from currently available surveys will be used to develop alternative statistical approaches for
generating correlated diaries for activity and consumption information at the individual level.
Once these statistical approaches are developed, they will be incorporated into TRIM.Expo as
appropriate. Ultimately, year-long or greater measured exposure data will need to be collected to
verify the validity of these statistical techniques.
A statistical technique developed to augment the activity pattern data for pNEM/CO will
be used during the initial development of TRIM.Expo. Earlier versions of pNEM/CO defined
cohorts solely according to home district, demographic group, work district (if applicable), and
residential cooking fuel. The new feature installed in pNEM/CO (Version 2.0) permits the user
to specify a "replication" value (n) such that the model will produce n cohorts for each
combination of the above four indices. Because pNEM/CO uses a Monte Carlo process to
construct an activity pattern for each cohort, each of the n cohorts associated with a particular
combination of home district, demographic group, work district, and residential cooking fuel is
associated with a distinct exposure sequence. The replication feature permits the analyst to
divide the population of interest into a larger number of smaller cohorts; a process which
pNEM/CO's developers report decreases the "lumpiness" of the exposure simulation. For
example, if a replication value of five (n = 5) is specified, the pNEM/CO model analyzes five
times the number of cohorts it would have considered if the cohorts had been defined solely by
home district, demographic group, work district, and residential cooking fuel. The use of
replication values is a technique that is intended to enhance the utility of existing data while
more robust statistical techniques are developed and additional data are collected on chronic or
longitudinal exposures of individuals within the population.
NOVEMBER 1999 5-15 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 5
INHALATION
5.4.3 DATA INPUTS FOR VENTILATION RATE
As described in Section 5.1.5, CHAD provides an activity indicator for each exposure
event. In turn, a distribution of values for the ratio of oxygen uptake rate to body mass (referred
to as metabolic equivalents or "METs") is provided for each activity type listed in CHAD. The
forms and parameters of these distributions were determined through an extensive review of the
exercise and nutrition literature. The primary source of distributional data was a compendium
developed specifically to facilitate the coding of physical activities and to promote comparability
across studies by Ainsworth et al. (1993). Table 5-3 contains a list of the parameters used in
pNEM/CO (Version 2.0) for estimating ventilation rates.
Table 5-3
Parameters Used to Estimate Ventilation Rates
Parameter
Body mass
Metabolic equivalence
Resting metabolic rate
Normalized oxygen
uptake rate
Abbreviation
BM
MET
RMR
NV02max
Functional Form
Lognormal distribution
Distribution specified in
CHAD Database
Regression equations
specific to age and gender
Normal distribution
Source of Data
Brainard and Burmaster 1992
Johnson et al. 1999
Schofield 1985, as compiled by
Johnson et al. 1999
Astrand 1960, Mercieret. al. 1991,
Katch and Park 1975, Heil et. al.
1995, Mermier et al. 1993, Rowland
etal. 1987
NOVEMBER 1999
5-16
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGEST1ON
6. INGESTION
For many pollutants, an understanding of the ingestion exposure pathway is critical to
accurately assessing potential health risks. This chapter provides a detailed presentation of the
algorithms used in TRIM.Expo to assess, compare, and combine different ingestion pathways of
exposure. Ingestion exposures are characterized in TRIM.Expo using the potential average daily
dose (ADD). The ingestion exposure ADD represents the amount of pollutant that enters the
mouth of the exposed individual over a defined exposure event or during a defined exposure
duration. This chapter begins by providing an overview of the general approach used in
TRIM.Expo to characterize the equivalent daily intake from an exposure medium. Of particular
importance here is how the attributes of ingestion exposure define the spatial, temporal, and
cohort resolution of the ingestion exposure algorithms. This is followed by a description of the
media-specific algorithms used to estimate these exposures in TRIM.Expo. The chapter
concludes with a presentation of the inputs and default values necessary to assess the ingestion
exposure pathway.
6.1 OVERVIEW OF THE APPROACH
In Section 2.3.2, all exposure attributes relevant to exposure routes and pathways were
organized into a set of three key exposure dimensions that has a significant impact on the
structure of the exposure model (e.g., on the exposure media included, the degree of spatial
resolution, and the level of temporal and spatial aggregation). These three important key
exposure dimensions of the exposure assessment problem were determined to be the (1) route of
exposure, (2) time scale of an exposure event relevant to the pollutant's associated effects, and
(3) degree of location dependence (i.e., dependence of exposure on the location of the exposed
subject).
In addition to incorporating the above exposure dimensions, ingestion exposure-event
functions must also account for some unique modeling constraints. For example, slower rates of
temporal variation in pollutant concentrations are exhibited in soil, water, vegetation, and
animals because of their larger mass and tendency to retain pollutants. Furthermore, the food
and water consumption data available from sources, such as EPA's Exposure Factors Handbook
(U.S. EPA 1997b), provide only seasonal resolution which limits the use of this data for shorter-
term assessments. Thus, ingestion exposure algorithms for media such as water, soil, and home-
grown or locally-produced foods cannot support as high a degree of time and spatial resolution as
inhalation exposures.
Based on the above constraints, the exposure-event functions for ingestion focus on
pollutant concentration variations among exposure districts. Activities (i.e., ingestion patterns)
are characterized as daily equivalent intakes representative of each month. Concentrations are
assessed at a time resolution of daily or greater and are then averaged for a representative
exposure duration (i.e., monthly).
To characterize an ingestion exposure, the ingestion intake algorithms require
information on the following factors:
NOVEMBER 1999 6^jTRIM.Expo TSD (DRAFT)
-------
CHAPTER 6
INGESTION
The pollutant concentration in the exposure medium during an exposure event;
The ingestion rate of the exposure medium during an exposure event;
The body weight of the individual exposed or the body weight distribution of the exposed
cohort;
The fraction of the cohort's ingested water, soil, and food products that come from an
exposure district with a specified pollutant concentration in the exposure medium; and
The frequency of exposure events over the exposure duration.
Figure 6-1 below illustrates how this information is linked to characterize exposure and
intake over an exposure duration.
figure o-l
An Illustration of How Information is Combined to Assess
Ingestion Exposure over a Defined Exposure Duration
Body
Time step size weight
^
C
i
t
V
I
Intake rate during ^
the time step "^ "
1 Expos
freque
^^^
""N. ^ff^
Cumulative
ime step w^ a"
J
t
Exposure media
concentration
- water
-food . ,
-soil ^ lnterr
^^p^ trans
Consumption data Exposure
- water duration, ED
- food
I fr-til
1
ure [^
ncy ^
^ Average intake N,
/ over the
mmation over ^ ^^^/ exposure
time steps ^ ^ duration, ED, in
\. mg/kg-d ,
nedia ^ Environmental
-air
-sou
- water
NOVEMBER 1999
6-2
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
In order to organize this information into a form that expresses population exposure, an
algorithm of the following form is used for each exposure medium considered in TRIM.Expo:
ADD:,m(T) =
Cm(i,kJ,t)EFz,m(i,t)ET(t)
T
(6-1)
where:
[Lm(k,l)/BW] =
Cm(iMt)
EF:Ji,t)
ET(t)
m
for individual or cohort z the average daily dose or intake of a
pollutant from exposure medium m averaged over the time period
T, which is one day or greater
the equivalent rate of intake of exposure medium m (expressed
kg/d, L/d) over the time step / by individual or cohort z divided by
that individual's body weight (BW). The microenvironment and
activity codes k and / are not used in the ingestion calculations, but
are included here to provide consistency with the general exposure
model presented in Chapter 2. This parameter can be obtained from
EPA's Exposure Factors Handbook (U.S. EPA 1997b) for most
exposure media
the pollutant concentration in mg/kg in exposure medium m (e.g.,
tap water, dairy products, meat, fish, vegetables) in exposure
district / or location of the exposure media itself, averaged over the
time step t. These concentrations are obtained from environmental
samples or from a fate and transport model such as TRIM.FaTE
the frequency (fraction of days in the time step) that individual or
cohort z from district / has contact with exposure medium m during
time step t. This term is used to make adjustments for factors such
as drinking water consumption that can be dispersed among several
exposure districts. Typically, for food, the exposure frequency is
set equal to the number of days in the time step because the
exposure frequency is already implicitly incorporated into the daily
intake rate.
the duration of time step / in months or days depending on the time
resolution required
the exposure medium contacted (e.g., air, water, food)
NOVEMBER 1999
6-3
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
/ = the geographical location in which the exposure takes place (i.e.,
the exposure district) or the location of origin of the exposure
media of concern (e.g., drinking water).
k = microenvironment in which the exposure occurs, [e.g., indoors at
home; in a vehicle; indoors at work (not an important attribute for
ingestion exposures)]
/ = activity code that describes what the individual is doing at the time
of exposure (e.g., resting, working, preparing food, cleaning,
eating)
The summation of pollutant exposures from all exposure media is over all time steps, /, that
comprise the time period, T.
6.1.1 SELECTION OF POPULATION COHORTS
For ingestion exposures, important attributes that define cohorts include age, gender,
exposure district, water supply, consumption of home-grown foods, and consumption of locally-
produced foods. Table 6-1 summarizes the initial set of cohort attributes used in TRIM.Expo for
ingestion exposures. These attributes were selected based on information available in key
references such as EPA's Exposure Factors Handbook. There are fifteen primary attributes listed
in this table that can be used to construct cohorts for ingestion exposure assessments, thus
potentially leading to a large number of cohorts. However, many of the primary attributes can be
combined (e.g., home-grown foods can be aggregated into a single primary attribute instead of
four separate categories as shown in Table 6-1). By combining attributes, the number of cohorts
in the exposure analysis is significantly reduced.
Table 6-1
Primary Attributes of Exposure Cohorts
Primary Cohort Attributes
(1) Age
Child
Adult
(2) Gender
Male
Female
Exposure district
(3) Residential
(4) Work-school
NOVEMBER 1999
6-4
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
Primary Cohort Attributes
(5) Water supply
Surface water
Ground water (public well)
Ground water (private well)
Home-grown food
(6) Fruits, vegetables, grains
(7) Dairy products
(8) Eggs
(9) Meat
Locally-produced food supplies
(10) Fruits, vegetables, grains
(11) Dairy products
(12) Eggs
(13) Meat
(14) Local fishing
(15) Local hunting
6.1.2 TIME RESOLUTION OF EXPOSURE EVENTS
The TRIM.Expo module is designed to allow the user to specify time steps ranging from
hours to years. However, because of limited data on food consumption and biotransfer factors,
the time resolution of ingestion exposure events can be set to either daily or monthly time steps.
The time step can be increased or decreased, depending on the pollutants being studied and the
time-scale of the health effects.
6.1.3 EXPOSURE MEDIA CONSIDERED
The ingestion exposure media that are included in the current conceptual design of
TRIM.Expo are
Surface water;
Ground water;
Soil and house dust;
Home-grown fruits, vegetables, and grains;
NOVEMBER 1999
6-5
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
Home-produced eggs, dairy products, meat and fish; and
Locally-produced fruits, vegetables, grains, dairy products, meat, and fish.
Surface water and ground water are both considered to be sources of drinking water, but
direct ingestion of surface water and ground water can also occur during swimming and other
recreational activities. Ingestion of soil is modeled to occur outdoors in a residential
environment, whereas ingestion of dust is assumed to occur indoors at a residence or in a work
environment.
Food supplies are categorized to be home-grown, locally-grown, or some combination of
the two categories. TRIM.Expo defines home-grown foods as those foods produced on the land
associated with a household and, for the most part, consumed within that household; whereas
locally-produced foods are those that are produced in home gardens and commercial farms in
contact with air, soil, and/or water that are in the study area. For the purposes of TRIM.Expo, it
is assumed that the attributes associated with the consumption of locally-grown foods is the same
for all districts within a study area (e.g., urban air shed) and that the quantity of food that is
designated as home-grown is applied separately based on whether the exposure district is urban,
suburban, or rural.
For lipophilic pollutants (e.g., dioxins, furans, polychlorinated biphenyls, pesticides) and
for metals (e.g., lead, mercury), exposures through food have been demonstrated to be the
dominant contributors to total dose for non-occupationally exposed populations (Travis and
Hester 1991). However, overall uncertainties in estimating potential doses through food chains
are much larger than uncertainties associated with other exposure pathways (Jones et al. 1991,
McKone and Daniels 1991, McKone and Ryan 1989).
6.1.3.1 Ingested Water
Ingested water is defined as water that is (1) consumed directly from the tap; (2)
consumed in food and beverages; and/or (3) ingested during water recreation. The primary
source of ingested water is tap water that is drawn from ground water, surface water, or some
combination of the two sources. Water ingested during recreation could be either surface water
or ground water.
Surface water includes water obtained from estuaries, lakes, rivers, and wetlands.
Therefore, exposure to surface water can occur from the use of tap water and from recreational
activities, such as swimming and other water sports. Ground water is found in the saturated zone
of the subsurface environment. Human exposure to ground water can occur once it is withdrawn
for tap/drinking water; used in cooking and processing foods; used for irrigation; used as water
supply in recreational activities (e.g., to fill a swimming pool); and/or supplied to animals
reared/bred for human consumption.
NOVEMBER 1999 6-6 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
6.1.3.2 Food
Currently TRIM.Expo includes several food exposure media, such as fruits, vegetables,
grains, milk, dairy products, eggs, meat, and fish. TRIM.Expo differentiates between home-
grown foods and locally-produced foods, and then focuses on which of these raw foods are
produced and consumed within the set of exposure districts being considered in the risk and
exposure analysis.
Fruits and vegetables are further divided into categories of unprotected and protected
produce. Protected produce crops have skins or shells that are not usually consumed (e.g., citrus
fruits, peanuts, beans). Unprotected produce typically have no outer covering or have skins or
shells that are commonly consumed (e.g., leafy vegetables, grapes, most grains). In addition,
vegetables are distinguished between those that have edible parts that grow above the ground
(i.e., above-ground crops) versus those that are root crops (i.e., below-ground crops). In the case
of root (below-ground) crops, EPA's Exposures Factors Handbook (U.S. EPA 1997b) does not
differentiate between protected or unprotected produce. Table 6-2 provides examples of fruits,
vegetables, and grains defined by the above categories.
Table 6-2
Taxonomy of Food Types Categorized as Fruits, Vegetables, and Grains
Type of Crop
Above-ground crops
Below-ground crops
(root crops)
Protected/
Unprotected
unprotected
protected
Fruits
grapes, berries, apples
citrus
none considered
Vegetables
lettuce, broccoli
beans, peas
carrots, radishes,
peanuts, potatoes
Grains
wheat, barely oats
none considered
none considered
Cooked and Processed Food
Exposures to pollutants in cooked and processed foods depend on the preparation
techniques used to combine and convert raw foods into edible, consumption food products.
Meats, eggs, dairy products, and grains are almost always processed and cooked prior to human
consumption. Since cooking and food processing can result in the transformation of many
pollutants, intermedia transfer factors are needed to characterize how preparation and cooking
alter raw food products. However, for the most part, such factors are unavailable. EPA's
Exposure Factors Handbook (U.S. EPA 1997b) contains some of this information.
Uncooked Food
Most fruits and some vegetables are served uncooked, reducing the need for intermedia
transfer factors to distinguish differences between ambient pollutant concentrations in the
vegetation and the pollutant concentration at the time of consumption. Even though fruits and
vegetables can be washed before they are eaten, current research suggests that for fine particles
and pollutants dissolved in the lipid phase of the vegetation, washing does little to reduce
NOVEMBER 1999
6-7
TRIM.Expo ISO (DRAFT)
-------
CHAPTER 6
INGESTION
pollutant concentration (Jones et al. 1991) and thus, it may not be necessary to incorporate a
factor that accounts for how washing alters pollutant concentrations.
6.1.3.3 Soil and Dust
Three soil categories, classified by the depth of soil-surface soil, root zone soil, and
vadose zone soil-are used to assess soil contamination in the outdoor environment (see the
TRIM.FaTE TSD, Volumes I and II for more information; U.S. EPA 1999a, b). Both surface
soil and root zone soil are considered to be sources of pollutant transfers from soil to vegetation.
Surface soil is also assumed to be the primary source for direct ingestion of soil outdoors.
In the indoor environment, the hand-to-mouth activities of children and adults are
assumed to give rise to contact with house dust as opposed to actual soil ingestion. House dust
suspended in the indoor air environment originates from three sources: (1) airborne particles that
penetrate from outdoor air to indoor air; (2) surface soil and dust tracked into buildings on shoes
or clothes, by pets, or other vectors; and (3) a variety of sources related to occupant activities,
material degradation, and household products. In the current version of TRIM.Expo, the
pollutant concentration in house dust is assumed to be equal to that in surface soil.
6.1.4 EXPOSURE LOCATIONS
The pollutant concentrations in the exposure media of drinking water, locally-grown
foods, and recreational food products are likely to be either highly dependent on the original
location of the media or aggregated among several exposure districts. The magnitude of
exposure to an individual or cohort is not strongly dependent on the exposure district in which
the receptor resides; instead, the magnitude of exposure will depend on the fraction of food and
water supply that comes from local sources and on their proximity to sources of pollution. The
pollutant concentrations in these exposure media are considered to be weakly dependent on the
resident location of the receptor since the driving factor for exposure is the pollutant
concentrations found in the locations from where water and food are obtained, rather than the
pollutant concentrations found in the area where the receptor resides. In cases where food and
water distribution systems are not well characterized, pollutant concentrations in all exposure
media are aggregated among several exposure districts (i.e., those that supply drinking water
and/or food) and then delivered to the various population cohorts.
In contrast to exposure media that are weakly dependent on the receptor location,
exposure media such as home-grown foods, soil, and house dust have pollutant concentration
levels that depend almost completely on pollutant levels in the air and soil of the exposure
district in which the exposed receptor resides.
6.1.4.1 Residential Exposure Locations
For the modeling purposes of TRIM.Expo, it is assumed that ingestion exposure events
occur mostly in the residential microenvironment located in the residential exposure district of
the exposed receptor(s). Ingestion of home-grown foods, soil, and/or dust are assumed to occur
NOVEMBER 1999 6^8 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 6
INGESTION
exclusively at the primary residential location of the exposed individual or cohort. Ingestion of
water and locally-grown foods are assumed to take place primarily in the residential exposure
districts. However, ingestion of the above products can also take place at work and/or at
restaurants that are not necessarily in the residential exposure district of the individual or cohort.
However, since these latter pollutant concentrations are aggregated among several exposure
districts, the exact location where water or locally-grown foods are consumed is not important.
In such cases, the information that is needed is (1) the quantity of locally-grown foods and
locally-supplied water consumed by an individual or cohort and (2) the exposure districts from
which the food and water are obtained.
6.1.4.2 Other Exposure Locations
As noted above, the exposure location for pollutants found in water and locally-grown
foods is primarily residential, but the estimation of the pollutant levels in food and water are
based on a combination of several exposure districts in which these pollutants are found. A
similar approach is used for exposure to pollutants found in meat and fish derived from local
hunting and fishing. The exposure location is assumed to be the residence of the cohort, but the
pollutant concentrations in the watershed or habitat in which hunting and fishing take place.
6.2 PRESENTATION OF THE MODEL ALGORITHMS BY EXPOSURE
MEDIA
This section presents the ingestion exposure algorithms, organized by exposure media
(i.e., water, soil, food) that are currently used in TRIM.Expo to express the intake of pollutant
concentrations found in the environmental media.
6.2.1 INGESTED WATER
Figure 6-2 illustrates the ingestion pathways of surface water and ground water for
humans. TRIM.Expo currently incorporates only the direct ingestion of tap water that is derived
from surface water or ground water.
NOVEMBER 1999 6-9 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 6
INGESTION
Figure 6-2
Exposure Pathways Considered for Surface and Ground Water
Surface
water
supply
Ground
water
supply
Dilution/
Processing
Intrinsic water in food
0
Indirect intake pathways
Swimming water (rivers, lakes, & pools)
Direct intake pathways
Inake of tap water,
beverages, and
cooking water
Ingestion while
swimming
Tap water intake includes household drinking water that is consumed directly or
consumed indirectly in a beverage (e.g., orange juice, soft drinks, coffee, tea) or in adding
intrinsic water to food. For the direct ingestion of tap water, the applicable exposure algorithm
is:
BW:
,l, t)EFz, w(z, t)ET(t)
(6-2)
where:
CJiMt)
[LMBW.J =
EF:JW(i,t)
pollutant concentration in tap water, mg/L.
the rate of intake of tap water (fw), L/kg/d, by individual or cohort
z divided by that individual's body weight (BW). The
microenvironrnent and activity codes k and / are not used in this
calculation.
the exposure frequency, number of days per month equivalent, that
individual or cohort z obtains tap water from exposure district i
(this is likely to be location independent).
NOVEMBER 1999
6-10
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
Pollutant concentration in tap water may not be directly measurable or derivable from the
output of models, such as TRIM.FaTE. In such cases, intermedia transfer factors can be used to
calculate the pollutant concentrations in tap water based on pollutant concentrations found in
surface water and ground water:
CJi,k,l,t) = PF(w)CMl,t) (6-3)
PF(w) Cgw(i,k,l,t) (6-4)
PF(\v) - the intermedia transfer that expresses the concentration in tap water
relative to ground or surface water used in exposure district i.
PF(w) is the processing dilution factor that accounts for the
removal of pollutants by processing. In the current version of
TRIM.Expo, PF(w) is set to 1 but can be set to a value specific to
each exposure district.
The Environmental Fate and Effects Division of EPA's Office of Pesticide Programs (OPP) is in
the process of producing a document that discusses the concept of the intermedia transfer
(PF(w)) that relates the pollutant concentration in tap water to that found in ground water or
surface water.
6.2.2 INGESTION OF SOIL AND HOUSE DUST
Both adults and children inadvertently ingest small amounts of soil through
hand-to-mouth activities. Children who spend a great deal of time outdoors contact and ingest
soil, and adults ingest soil through activities such as gardening, outdoor labor, and cleaning. In
some cases, individuals suffer from a condition called pica behavior, and are known to
intentionally consume large quantities of soil.
Several studies have been conducted to characterize soil ingestion by children (e.g., see
U.S. EPA 1997b, Stanek et al. 1998, Calabrese and Stanek 1995, Sedman and Mahmood 1994,
Thompson and Burmaster 1991, Davis et al. 1990). Some of these studies make use of soil
loading on children's hands in combination with observations of hand-to-mouth activity to
estimate soil uptake. Another approach to estimating soil ingestion uses tracer elements in feces.
The process involves analyzing both the feces of children and the soil in their playgrounds for
elements that are thought to be poorly absorbed in the gut, such as aluminum, silicon, and
titanium. Then, assuming that there are no non-soil sources of these elements and using a fecal
excretion rate, the soil ingestion for each child is estimated based on the mass of each tracer
element detected in the feces relative to that found in soil. In such studies, hospitalized children
who have little contact with soil are often used as control groups.
Hand-to-mouth activities lead to the ingestion of pollutants in soil (outdoors) and in
house dust (indoors). Pollutants in house dust are attributable in part to pollutants found in the
surface soil surrounding the residence because of the resuspension of outdoor surface soil, and its
infiltration through windows and openings, and its subsequent deposition onto indoor surfaces.
The pollutants can also be brought indoors from the outdoor soil by soil tracking (i.e., soil carried
NOVEMBER 1999 6^1 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 6
INGESTION
into the house by shoes, clothing, and pets). Figure 6-3 illustrates soil ingestion pathways that
are considered in TRIM.Expo.
Figure 6-3
Exposure Pathways Considered for Contact with Soil and House Dust
Outdoor
environment
Soil
Particle
infiltration
Tracking &
Dilution
Indoor
environment
Dust
Ingestion of
soil during
outdoor
activities
Ingestion of
dust during
indoor
activities
6.2.2.1 Soil Ingestion (Outdoors)
For direct ingestion of surface soil in the residential outdoor environment, the applicable
exposure algorithm is:
ADD-. SS.,(T) =
I,(*
BW-.
Cซ(i,k,l,t)EF:,ss(i,t)ET(t)
(6-5)
where:
[LJk,l)/BW] =
EF:,Ji,t)
the annually averaged daily rate of intake of surface soil (ss\
kg/kg/d, by individual or cohort z divided by a representative
individual's body weight (BW). The microenvironment and
activity codes k and / are not used in this calculation.
the exposure frequency, which is the fraction of days in a month
that individual or cohort z has contact with outdoor soil in
exposure district;'. This factor is used to make adjustments for
time within or outside of the exposure district - the number of days
NOVEMBER 1999
6-12
TRIM.Expo ISO (DRAFT)
-------
CHAPTER 6
__ INGESTION
per month outside with soil contact is already accounted for in the
parameter above.
Ci5(i,k,l,tj - the pollutant concentration, mg/kg, in the surface soil or outside
surface dust (of urban areas) for exposure district / during the time
step /.
6.2.2.2 Dust Ingestion (Indoors)
For the direct inge^tion of pollutants from house dust in the indoor environment, it is
assumed that residential surface soil is the source of the dust pollutants. The applicable exposure
algorithm is:
ADD-..hd.i(T}= - - - - - (6-6)
where:
Q/'. V = the pollutant concentration, mg/kg, in the house dust (hd) of
exposure district / during the time-step /.
[LhJ(k,l)/BW] = the annually averaged daily rate of intake of soil, kg/kg/d, by
individual or cohort z divided by a representative individual's body
weight (BW). The microenvironment and activity codes k and / are
not used in this calculation,
EF.,Ji,t) = the exposure frequency, fraction of days per month equivalent, that
individual or cohort z has contact with house dust in exposure
district /.
The concentration of a pollutant in house dust is calculated from two intermedia transfer
factors:
Chd(i,k,l,t) = fhdsINss(i,t)Css(i,t) + (1 - fhds)\ ^Cap(i,t) (6-7)
pap
C^(i,t) = the pollutant concentration in soil of exposure district / during time
step /, mg-kg.
Cap(i, t) = the pollutant concentration in the particulate phase of exposure
district / during time step /, rng-m3.
NOVEMBER 1999
-------
CHAPTER 6
INGESTION
_/f,ds = the fraction of indoor dust that originates from outdoor soil. The
remaining fraction is assumed to originate from paniculate matter
in air. Currently, this fraction is set to 0.5.
INss(i,t) = the combined soil infiltration and soil tracking factor that
expresses the likely increase or decrease of pollutant concentration
in indoor soil relative to outdoor surface soil in exposure district /'.
TRIM.Expo currently sets INss(i,t) to 1 but it can be set to a value
specific to each exposure district and each time step.
fINap(i,t)/pap] = the particle infiltration factor that expresses the likely increase or
decrease of pollutant concentration in indoor dust relative to
pollutant concentration in outdoor particulate matter in exposure
district /. pap is the density of airborne particles of-2,400 kg-m3.
In the current version of TRIM.Expo, INap(i,t) is set to 1 but can be
set to a value specific to each exposure district and each time step.
6.2.3 INGESTION OF POLLUTANTS IN HOME-GROWN PRODUCE OR
HOME-BRED ANIMALS
Soil pollutants can be transferred to the edible parts of vegetation from surface soil by
resuspension and deposition, rain splash, volatilization followed by partitioning, and by root
uptake for below-ground vegetation (Jones et al. 1991). The level of exposure to the soil
pollutants found in vegetation food products often depends on translocation (i.e., the process by
which a pollutant is transferred from one part of a plant to another). Translocation can cause
significant differences in pollutant concentrations in the total plant and the edible portion of the
plant (i.e., fruit, seeds). In addition, ingestion of contaminated soil or grains by animals
reared/bred for human consumption can lead to contaminated, animal-based food products, such
as meat, milk, dairy products, and eggs.
For the purposes of TRIM.Expo, home-grown foods are defined as foods grown on the
land associated with a household and, for the most part, consumed within that household. Figure
6-4 illustrates the exposure pathways considered in constructing food-based exposures from
home-grown foods and the algorithms used for modeling pollutant exposures from home-grown
food products are discussed in the sections below. In the TRIM.Expo framework, home-grown
foods are studied in a single exposure district. To calculate individual or cohort consumption of
home-grown food products, the annual average consumption of the food product for individual or
cohort z is derived from the national food consumption survey data (CSFII1996) and then used
to calculate the fraction that can be allocated to home-grown foods.
NOVEMBER 1999 6-14 TRIM.ExPoTSD (DRAFT)
-------
CHAPTER 6
INGESTION
Figure 6-4
Exposure Pathways Considered for Contact with Food Products
Outdoor environment
at a residence within a specified exposure district
Air
Home-grown
fruits,
vegetables,
& grains
6.2.3.1 Vegetables, Fruits, and Grains
Vegetables, fruits, and grains include leafy vegetables (e.g., cabbage, cauliflower,
broccoli, celery, lettuce, spinach), exposed produce (e.g., apples, pears, berries, cucumber,
squash, grapes, peaches, tomatoes, string beans), protected produce or root crops (e.g., carrots,
beets, turnips, potatoes, legumes, melons, citrus fruits) and grains (e.g., wheat, corn, rice, barley,
millet).
Pollutants in the root zone soil gas and liquid can be taken up by plant roots and
potentially transferred to above-ground plant parts in the transpiration stream. The ease with
which non-ionized pollutants are taken up from the soil into root material is influenced by the
pollutant's octanol/water partition coefficient, K^, which is commonly used as a measure of
lipophilicity and water solubility. Pollutants with high K^ values tend to either be strongly
sorbed onto organic material in the root, making them less available for movement in the
transpiration stream, or sorbed onto organic material in the soil, reducing their availability to root
uptake. Increases in the water solubility of a pollutant tend to increase the amount of pollutant
available for uptake from the soil water and increase the likelihood of movement with the
transpiration stream; however, the root membrane of most plants restricts uptake of highly
soluble or ionized species.
NOVEMBER 1999
6-15
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGEST1ON
The algorithm for ingestion of pollutants in above-ground fruits, vegetables, and grains
has the general form:
l2's(kJAcg(i,t)EF:,g(i,t)ET(t)
ADDs,jvg,i(T) =
1,1
BWz
'-+ \C^i,t)EF,.^i,t)ET(t)
(6-8)
Zl 7r,/7/v(ft,
'( RW-
where:
Cg(i,t) = the pollutant concentration, mg/kg, in the grains (g) of exposure
district / during the time step t.
Cefi,(i,t) = the pollutant concentration, mg/kg in the exposed fruits and
vegetables (efv) of exposure district / during the time step t.
Cpfi(i,t) = the pollutant concentration, mg/kg in the protected fruits and
vegetables (pfv) of exposure district / during the time step t.
[Note that these three concentrations must be obtained from measurements or from a model
suchasTRIM.FaTE].
[LM/BW.J
[Lefv(k,l)/BW.J
[Lpfv(k,l)/BW.J
the monthly or seasonally averaged daily rate of intake of
grains, kg/kg/d, by individual or cohort z divided by a
representative individual's body weight (BW). The
microenvironment and activity codes A: and / are not used in
this calculation.
the monthly or seasonally averaged daily rate of intake of
exposed fruits and vegetables, kg/kg/d, by individual or
cohort z divided by a representative individual's body
weight (BW). The microenvironment and activity codes k
and / are not used in this calculation,
the monthly or seasonally averaged daily rate of intake of
protected fruits and vegetables, kg/kg/d, by individual or
cohort z divided by a representative individual's body
NOVEMBER 1999
6-16
TR1M.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGEST10N
weight (BW). The microenvironment and activity codes k
and / are not used in this calculation,
EF.g(i,t) = the exposure frequency, fraction of days per month
equivalent, that individual or cohort z consumes home-
grown grains in exposure district /. If the daily intake rate
implicitly includes the exposure frequency, then this term is
set equal to 1.
EF. efv(i, t) = the exposure frequency, fraction of days per month
equivalent, that individual or cohort z consumes home-
grown exposed fruits and vegetables in exposure district /.
If the daily intake rate implicitly includes the exposure
frequency, then this term is set equal to 1.
EF:pfi.(i,t) = the exposure frequency, fraction of days per month
equivalent, that individual or cohort z consumes home-
grown protected fruits and vegetables in exposure district /.
If the daily intake rate implicitly includes the exposure
frequency, then this term is set equal to 1.
According to Yang and Nelson (1986), about half of all produce (i.e., fruits, vegetables)
consumed by humans consists of leafy vegetables and exposed produce that intercept pollutants
from the atmosphere. The remaining half consists of protected produce or root crops, where
pollutant transfer to the edible portion is primarily by root uptake. All grain crops are assumed
to be exposed primarily to air pollutants.
6.2.3.2 Dairy Products
To calculate human exposures to pollutants found in dairy products, the biotransfer
factors for milk versus the pollutant intake by dairy cattle must be determined, along with the
parameters that describe the pasture, water, and soil intake rates of dairy cattle. For the purposes
of TRIM.Expo, pasture is defined to be all foodstuffs that are grown on the farm to feed the
animals (e.g., open pasture grass, grains, corn).
The algorithm for ingestion of pollutants in dairy products has the form:
L,dp(k,l)
IJ
/, i)EFz, dP(i, t}ET(t}
ADDz, dP, t(T) = ' (6-9)
where:
NOVEMBER 1999 6-17 TRJM.Expo TSD (DRAFT)
-------
CHAPTER 6
INGESTION
Cdp(i,t)
[Ldp(k,l)/BW.J =
EF:,dp(U)
6.2.3.3 Eggs
the pollutant concentration, mg/kg, in the dairy products (dp) of
exposure district z during the time step / obtained from
measurements or from a model such as TRIM.FaTE.
the monthly or seasonally averaged daily rate of intake of dairy
products, kg/kg/d, by individual or cohort z divided by a
representative individual's body weight (BW). The
microenvironment and activity codes k and / are not used in this
calculation.
the exposure frequency, expressed as the fraction of days per
month equivalent, that individual or cohort z consumes home-
grown dairy products in exposure district /'. If the daily intake rate
implicitly includes the exposure frequency, then this term is set
equal to 1.
To calculate human exposures to pollutants found in eggs laid by home-grown hens, the
biotransfer factors for eggs versus the pollutant intake by chickens, and the parameters that
describe the feed, water, and soil intake rates of chickens must be characterized.
The algorithm for ingestion of pollutants in eggs has the form:
ADD;,egg.1(T) =
Cegg(i,t)EFz,egg(i,t)ET(t)
(6-10)
where:
Ctgg(i,t)
[Legg(k,l)/BW:]
EF:.egg(i,t)
the pollutant concentration, mg/kg, in eggs in the exposure
district / during the time step t obtained from measurements
or from a model such as TRIM.FaTE.
the monthly or seasonally averaged daily rate of intake of
eggs, kg/kg/d, by individual or cohort 2 divided by a
representative individual's body weight (BW). The
microenvironment and activity codes k and / are not used in
this calculation,
the exposure frequency, expressed as the fraction of days
per month equivalent, that individual or cohort z consumes
home-grown eggs in exposure district i. If the daily intake
rate implicitly includes the exposure frequency, then this
term is set equal to 1.
NOVEMBER 1999
6-18
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGEST1ON
6.2.3.4 Meat and Poultry
To calculate human exposures to pollutants found in meat, the biotransfer factors for meat
versus pollutant intake by animals reared/bred for human consumption and the parameters that
describe the pasture, water, and soil intake rates of these animals must be quantified. For the
purposes of TRIM.Expo, cattle is used to represent all animals that are reared/bred for human
consumption.
The algorithm for ingestion of pollutants found in meat has the form:
y (L.mp(kJ}\
^'l BW-. ) mp(l
,t)EFz.mp(i,t)ET(t)
ADD.-.mp.,(T)= ,..,
where:
Cmp(i,t) = the pollutant concentration, mg/kg in the home-grown meat
and poultry (mp) of exposure district / during the time step /
obtained from measurements or from a model such as
TRIM.FaTE.
[I.mp(k,l)/BW^\ = the average daily rate of intake of meat and poultry,
kg/kg/d, by individual or cohort z divided by a
representative individual's body weight (5W7)- The
microenvironment and activity codes k and / are not used in
this calculation.
EF.mp(i,t) = the exposure frequency of meat and poultry consumption of
individual or cohort z in exposure district / (expressed as
the fraction of days per month equivalent). If the daily
intake rate implicitly includes the exposure frequency, then
this term is set equal to 1.
6.2.4 LOCALLY-GROWN COMMERCIAL FOODS
The following subsections describe algorithms for estimating intake of pollutants found
in locally-grown foods. These types of algorithms have been previously developed by McKone
and Ryan (1989), McKone and Daniels (1991), Travis and Blaylock (1992), and McKone
(1993a, b, c). The form and ranges of values used in these models have been validated for a
limited number of compounds by Bennett (1981,1982) and by Travis and Blaylock (1992).
Locally-grown foods (i.e., grown in home gardens and commercial, local farms) are
defined as foods that are not only produced, but also consumed within the same urban air shed
being modeled by TRIM.Expo. The pollutant concentrations in air, soil, and/or water that are
used to assess concentrations in locally-grown foods make use of the average pollutant
NOVEMBER 1999
-------
CHAPTER 6
INGESTION
concentration in the locations of farms producing such foods. For modeling purposes, it is ideal
if distributions for describing the local concentrations of pollutants in produce, grain, milk and
dairy products, meat, eggs, and fish are available; but if these values are not available, they must
be developed.
6.2.4.1 Vegetables, Fruits, and Grains
The algorithms used to calculate intake of pollutants in locally-grown vegetables, fruits,
and grains are the same as those provided in Section 6.2.3.1 with the following replacements:
EF:f(\,t) = the exposure frequency, fraction of days per month equivalent, that
individual or cohort z consumes locally-produced grains in
exposure district /'. As with home-grown produce and home-bred
animals, if the daily intake rate of locally-grown produce implicitly
includes the locally-grown produce exposure frequency, then this
term is set equal to 1.
EF.^.(i,t) = the exposure frequency, fraction of days per month equivalent, that
individual or cohort z consumes locally-produced exposed fruits
and vegetables in exposure district /'. As with home-grown
produce and home-bred animals, if the daily intake rate of locally-
grown produce implicitly includes the locally-grown produce
exposure frequency, then this term is set equal to 1.
EF. Pfi.(i, t) - the exposure frequency, fraction of days per month equivalent, that
individual or cohort z consumes locally-produced protected fruits
and vegetables in exposure district /'. As with home-grown
produce and home-bred animals, if the daily intake rate of locally-
grown produce implicitly includes the locally-grown produce
exposure frequency, then this term is set equal to 1.
Cg(avg,i) replaces Cg(i,t), where Cg(avg,t) is the averaged pollutant concentration, mg/kg,
in grains based on the average concentration in the locations where local foods are
produced during the time step /.
Cefv(avg,t) replaces Cefv(i,t), where Cefi(avg,t) is the averaged pollutant concentration,
mg/kg, in exposed fruits and vegetables based on the average concentration in the
locations where local foods are produced during the time step /.
Cpfv(avg,t) replaces Cpfv(i,t), where Cpfv(avg,t) is the averaged pollutant concentration,
mg/kg, in protected fruits and vegetables based on the average concentration in the
locations where local foods are produced during the time step /.
NOVEMBER 1999 6-20 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGEST1ON
6.2.4.2 Dairy Products
The algorithms used to calculate intake of pollutants from locally-produced dairy
products are the same as those provided in Section 6.2.3.2 with the following replacements:
EF.dp(i,t) = the exposure frequency, expressed as the fraction of days per
month equivalent, that individual or cohort z consumes
locally-produced dairy products in exposure district /. As with
home-grown produce and home-bred animals, if the daily intake
rate of locally-produced dairy products implicitly includes the
locally-produced dairy products exposure frequency, then this term
is set equal to 1.
Cdp(avg,t) replaces Cdp(i,t), where Cdp(avg,t) is the spatially averaged pollutant
concentration, mg/kg, in the milk derived from all suburban and rural exposure districts
where local dairy products are produced during the time step t.
6.2.4.3 Eggs
The algorithms used to calculate intake of pollutants found in locally-produced eggs are
the same as those provided in Section 6.2.3.3 with the following replacements:
EF. egg(i, t) = the exposure frequency, fraction of days per month equivalent, that
individual or cohort z consumes locally-produced eggs in exposure
district /. As with home-grown produce and home-bred animals, if
the daily intake rate of locally-produced eggs implicitly includes
the locally-produced eggs exposure frequency, then this term is set
equal to 1.
Ce(avg,t) replaces Cegg(i,t), where Cegg(avg,t) is the averaged pollutant concentration,
mg/kg, in eggs based on the average concentration in the locations where eggs are
produced locally during the time step /.
6.2.4.4 Meat and Poultry
The algorithms used to calculate intake of pollutants found in locally-produced meat and
poultry are the same as those provided in Section 6.2.3.4 with the following replacements:
EF.mp(i, t) = the exposure frequency, number of days per month equivalent, thai
individual or cohort z consumes locally-produced meat products in
exposure district /. As with home-grown produce and home-bred
animals, if the daily intake rate of locally-produced meat products
implicitly includes the locally-produced meat products exposure
frequency, then this term is set equal to 1.
NOVEMBER 1999 6-21 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
Cmp(avg,t) replaces Cmp(i,t), where Cmp(avg,t) is the averaged pollutant concentration,
mg/kg, in meat based on the average concentration in the locations where local meat
products are produced during the time step /.
6.2.4.5 Fish (Commercial, Subsistence, and Recreational)
Exposures to pollutants that are found in fish are assumed to occur only on a local scale,
with no residential (home) scale exposures. The algorithm for ingestion of pollutants found in
fish has the form:
ADD=.f.,(T) =
where:
C/avg,t)
EF.j(i,t)
the average daily rate of intake offish (/), kg/kg/d, by individual or
cohort 2 divided by a representative individual's body weight (BW).
The microenvironment and activity codes k and / are not used in
this calculation.
the spatially or market averaged pollutant concentration, mg/kg. in
the fish of all exposure districts in the air shed being considered
during the time step /.
the exposure frequency, expressed as the fraction of days per
month or its equivalent), that individual or cohort z in exposure
district / consumes locally-caught fish. When the daily intake rate
implicitly includes exposure frequency this term can be set equal to
1.
The TRIM. Expo module utilizes three types of cohorts to reflect differences in the
exposure frequency to fish - those who buy and consume locally-raised fish, but do not catch
their own fish; those who consume locally-raised fish that they also catch on their own (e.g.,
recreational fishermen); and those who are subsistence fishermen who catch fish for a living and
also eat the fish that they catch.
6.2.5 RECREATIONAL SPORT MEAT (HUNTING)
Exposures to pollutants found in game animals (e.g., deer, water fowl) are assumed to
occur only on a local scale, with no residential (home) scale exposures. The algorithm for
ingestion of pollutants in meat from game animals is as follows:
NOVEMBER 1999
6-22
TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 6
INGESTION
ADD;, sm, i(T) = (6-13)
where
[L,sm(k,l)/BW:] = the time-step averaged daily rate of intake of sport meat
(sm), kg/kg/d, by individual or cohort z divided by a
representative individual's body weight (BW). The
microenvironment and activity codes k and / are not used in
this calculation.
Csm(avg, t) = the averaged pollutant concentration, mg/kg, in the meat of
game animals residing in the air shed being considered
during the time step /.
There is currently no separate algorithm available in TRIM.Expo to determine
concentrations in game animal tissues. This information must be obtained from available data or
can be obtained from the output of TRIM.FaTE.
6.3 INTEGRATION OF EXPOSURES ACROSS MULTIPLE
INGESTION MEDIA
The integration of ingestion exposures across multiple media for an individual or cohort
is based on matching time scales. All ingestion intakes within a given time step are summed, and
currently, relatively large time steps (monthly) are used in TRIM.Expo, such that the time
aggregation of ingestion exposures is straight forward and does not introduce a significant source
of uncertainty or confusion. Even daily aggregation of ingestion exposures should not be
difficult since TRIM.Expo is capable of working in hourly time steps. However, more
importantly, if TRIM.Expo is used to integrate ingestion exposures in hourly time steps, then a
comprehensive inventory of micro-activity data on the daily water and food intake and
hand-to-mouth activities of cohorts and individuals are needed. Such data are not yet available.
With regard to ingestion, some products are bolous exposures which are discrete while
others are aggregated (e.g., a beef meal would be associated with a specific farm and have a
spatially dependent concentration, while milk would likely be a diluted exposure composed of a
mixture of milk from all neighboring dairy farms). Therefore, where appropriate these exposures
should be modeled on a meal-by-meal basis (i.e., for each meal, the source of the food item is
randomly selected based on the proportion of its contribution to the total commodity load and,
thus, the pollutant concentration in the food item would be specific to that farm). Probabilistic
approaches would be repeated for each meal to get a more representative estimate of dietary
exposure. This process is described in more detail in EPA's guidance on health risks associated
with multiple exposure pathways to combustor emissions (U.S. EPA 1997d).
NOVEMBER 1999 6-23 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 6
INGESTION
The population risk (i.e., incidence of effects) from pollutants that are believed to exhibit
non-threshold mechanisms (i.e., linear carcinogens) can be estimated based on the amount of
pollutants entering the food supply each year (U.S. EPA 1997d). This method, as specified in
EPA guidance (U.S. EPA 1997d), uses the annual amount of food produced in various food-
producing regions of the study area as the metric of concern and then estimates the subsequent
exposures to pollutants in the food.
6.4 DISCUSSION OF ALGORITHM INPUTS AND VALUES
An important attribute of exposure models is the ability to account for factors that control
variation in human contact (i.e., age, gender, location, activity patterns). Exposure assessments
for ingestion pathways use of a number of factors that are both variable and uncertain. For each
relevant population cohort, a number of exposure factors described in the previous sections can
be used to characterize contact and intake. These factors are used to describe specific ingestion
behaviors (e.g.. rates of ingestion for specific media such as fish and water for specific cohorts)
or to describe the characteristics of the populations themselves (e.g., body weight). For each of
these parameters, it is necessary to develop a range of values that represent the population
cohorts. Currently, OAQPS is compiling and evaluating data for each parameter for
TRIM.Expo. The EPA's Exposure Factors Handbook (U.S. EPA 1997b) is one source being
used extensively to derive such factors. EPA's National Center for Environmental Assessment
(NCEA) is currently conducting research on how to derive distributions for many of the ingestion
exposure factors identified in this chapter. When these distributions become available, they will
be adopted as appropriate for use in TRIM.Expo. In the meantime, efforts are underway by
OAQPS and other EPA program offices to develop exposure factors and associated distributions
for specific parameters for use in risk assessments. When these are available, they will be
published and be the subject of subsequent reviews.
NOVEMBER 1999 6-24 TRJM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
7. REFERENCES
Ainsworth, B.E., W.L. Haskell, A.S. Leon and D.R. Jacobs, Jr. 1993. Compendium of Physical
Activities: Classification of Energy Costs of Human Physical Activities. Med. Sci. Sports Exer.
25:71-80.
Akland, G., T. Hartwell, T. Johnson and W.R. Whitmore. 1985. Measuring Human Exposure to
Carbon Monoxide in Washington, D.C., and Denver, Colorado, During the Winter of 1982-1983.
Environ. Sci. Technol. 19:911-918.
Allott, R.W., M. Kelly and C.N. Hewitt. 1992. Behavior of Urban Dust Contaminated by
Chernobyl Fallout: Environmental Half-lives and Transfer Coefficients. Environ. Sci. Technol.
26:2142-2147.
Astrand, I. 1960. Aerobic Work Capacity in Men and Women with Special Reference to Age.
Acta. Physiol. Scand. 49(Supp. 169).
Bennett, B.G. 1981. The Exposure Commitment Method in Environmental Pollutant
Assessment. Environ. Monit. Assess. 1:21-36.
Bennett, B.G. 1982. Exposure of Man to Environmental Nickel - An Exposure Commitment
Assessment. Sci. Total Environ. 22:203-212.
Brainard, J. and D. Burmaster. 1992. Bivariate Distributions for Height and Weight of Men and
Women in the United States. Risk Analysis. 12:267-275.
Buck, R.J., K.A. Hammerstrom, and P.B. Ryan. 1995. Estimating Long-term Exposures from
Short-term Measurements. J. Expos. Anal, and Environ. Epidem. 5, 359-373.
Bureau of Census. 1994. STP154: Census Tract of Work by Census Tract of Residence. U.S.
Department of Commmerce, Population Division, Journey-to-Work and Migration Statistics
Branch.
Bureau of Census. 1995. American Housing Survey for the Denver Metropolitan Area in 1995.
Current Housing Reports. Report No. HI 70/95-46. U. S. Department of Commerce.
Burmaster, D.E. and K. von Stackelberg. 1991. Using Monte Carlo Simulations in Public
Health Risk Assessments: Estimating and Presenting Full Distributions of Risk. J. Expos. Anal.
Environ. Epidem. 1:491-512.
Burmaster, D.E. and E.A.C Crouch. 1997. Lognormal Distributions of Body Weight as a
Function of Age for Males and Females in the United States, 1976-1980. Risk Analysis.
17:499-505.
Calabrese, E.J. and E.J. Stanek. 1995. Resolving Intertracer Inconsistencies in Soil Ingestion
NOVEMBER 1999 ~\ TRIM.Expo TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Estimation. Environ. Health Persp. 103:454-457.
CARB. 1998a. California Air Resources Board. Development of a Model for Assessing Indoor
Exposure to Air Pollutants. Report No. IE-2631. Sacramento, CA: Research Division. January.
CARB. 1998b. California Air Resources Board. Measuring Concentrations of Selected Air
Pollutants Inside California Vehicles. Contract No. 95-339. Sacramento, CA: Research
Division. December.
Commission of Risk Assessment and Risk Management (CRARM). 1997. Risk Assessment and
Risk Management in Regulatory Decision-Making. Final Report, Volume 2.
Continuing Survey of Food Intakes by Individuals (CSFII) 1996 and the Diet and Health
Knowledge Survey 1994-96. Food and Nutrient Intake by Individuals in the United States by
and Age, 1994-96. NFS Report No 96-2. U.S. Department of Agriculture.
Cullen, A.C. and H.C. Frey. 1999. Probabilistic Techniques in Exposure Assessment: A
Handbook for Dealing with Variability and Uncertainty in Models and Inputs. Plenum Press,
New York.
Daisey, J.M., K.R.R. Mahanama and A.T. Hodgson. 1998. Toxic Volatile Organic Compounds
in Simulated Environmental Tobacco Smoke: Emission Factors for Exposure Assessment. J.
Expos. Anal. Environ. Epidem. 8:313-334.
Davis, S., P. Waller, R. Buschbom, J. Ballou and P. White. 1990. Quantitative Estimates of Soil
Ingestion in Normal Children Between the Ages of 2-years and 7-years - Population-based
Estimates Using Aluminum, Silicon, and Titanium as Soil Tracer Elements. Arch. Environ.
Health. 45:112-122.
Droppo, J.G., Jr., J.W. Buck, D.L. Strenge and B.L. Hoopes. 1992. Risk Computation for
Environmental Restoration Activities. J. ofHaz. Mat. 35:341-352.
Duan, N. 1982. Models for Human Exposure to Air Pollution. Environment International 8:
305.
Ershow, A.G. and F.P. Cantor. 1989. Total Water Intake and Tapwater Intake in the United
States: Population-based Estimates of Quantiles and Sources. Report prepared under National
Cancer Institute Order #263-MD-810264 with the Life Sciences Research Office, Federation of
American Societies for Experimental Biology, Bethesda, MD.
Freijer, J.I., H.J. Th. Bloemen, S. de Loos, M. Marra, P.J.A. Rombout, G.M. Steentjes and M.P.
van Veen. 1997. Modelling Exposure of the Dutch Population to Air Pollution. In: B.J.M. Ale,
M.P.M. Janssen, and M.J.M. Pruppers, eds. Proceedings of the Risk 97: International
Conference Mapping Environmental Risks and Risk Comparison, RAI International Congress
Centre, Amsterdam, The Netherlands, pp. 105-111.
NOVEMBER 1999 7^2 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
General Sciences Corporation (GSC). 1988. PCGEMS User's Guide, Release 1.0. Contract No.
68-02-4281. Prepared for U.S. Environmental Protection Agency, Office of Toxic Substances,
Exposure Evaluation Division.
Georgopoulos, P.O., A. Walia, A. Roy and P.J. Lioy. 1997. Integrated Exposure and Dose
Modeling and Analysis System. 1. Formulation and Testing of Microenvironmental and
Pharmacokinetic Components. Environ. Sci. Technol. 31:17-27.
Glen, G., Y. Lakkadi, J.A. Tippett and M. del Valle-Torres (Prepared by ManTech
Environmental Technology, Inc.). 1997. Development of NERL/CHAD: The National
Exposure Research Laboratory Consolidated Human Activity Database. EPA Contract No. 68-
D5-0049.
Hayes. S.R., C. Seigneur and G.W. Lundberg. 1984. Numerical Modeling of Ozone Population
Exposure: Application to a Comparison of Alternative Ozone Standards. San Rafael, CA:
Systems Applications, Inc.
Hayes, S.R. and G.W. Lundberg. 1985. Further Improvement and Sensitivity Analysis of an
Ozone Population Exposure Model. San Rafael, CA: Systems Applications, Inc.
Heil, D.P., P.S. Freedson, L.E. Ahlquist, J. Price and J.M. Rippe. 1995. Nonexercise Regression
Models to Estimate Peak Oxygen Consumption. Med. Sci. Sports Exer. 27:599-606.
Jenkins, P.L., T.J. Phillips, E.J. Mulberg, and S.P. Hui. 1992. Activity Patterns of Californians:
Use of and Proximity to Indoor Pollutant Sources. Atmos. Environ. 26A:2141-2148.
Jenkins, R.A., A. Palausky, R.W. Counts, C.K. Bayne, A.B. Dindal and M.R. Guerin. 1996.
Exposure to Environmental Tobacco Smoke in Sixteen Cities in the United States as Determined
by Personal Breathing Zone Air Sampling. J. Expos. Anal. Environ. Epidem. 6:473-502.
Johnson, T.R., L. Wijnberg, J. Capel and Vostal. 1992a. The Use of Activity Diary Data to
Estimate the Probability of Exposure to Air Pollution. In: R.L. Burglund, ed. Tropospheric
Ozone and the Environment II. Pittsburgh, PA: Air and Waste Management Association, pp.
713-724.
Johnson, T., J. Capel, R. Paul and L. Wijnberg. 1992b. Estimation of Carbon Monoxide
Exposures and Associated Carboxyhemoglobin Levels in Denver Residents Using a Probabilistic
Version of NEM. Contract No. 68-DO-0062. Prepared by International Technology Corporation
for U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards. July.
Johnson, T., M. McCoy and J. Capel. 1993a. Enhancements to the Hazardous Air Pollutant
Exposure Model (HAPEM) as Applied to Mobile Source Pollutants. Contract No. 68-DO-0062.
Prepared by International Technology Corporation for U. S Environmental Protection Agency,
Office of Air Quality Planning and Standards. September.
NOVEMBER 1999 7-3 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Johnson, T., M. McCoy, J. Capel, M. Alberts and B. Morrison. 1993b. Estimation of
Incremental Benzene Exposures and Associated Cancer Risks Attributable to a Petroleum
Refinery Waste Stream Using the Hazardous Air Pollutant Exposure Model (HAPEM). Paper
presented at the 86th Annual Meeting of the Air and Waste Management Association, Paper
#A1389, Denver, Colorado.
Johnson, T.R. 1995. Recent Advances in the Estimation of Population Exposure to Mobile
Source Pollutants. J. Expos. Anal. Environ. Epidem. 5:551-571.
Johnson, T., J. Warnasch, M. McCoy, J. Capel and M. Riley. 1996a. Developmental Research
for the Hazardous Air Pollutant Exposure Model (HAPEM) as Applied to Mobile Source
Pollutants. Contract No. 63-D-30094. Prepared by International Technology Corporation for
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards. February.
Johnson, T., J. Capel and M. McCoy. 1996b. Estimation of Ozone Exposures Experienced by
Urban Residents Using a Probabilistic Version of NEM and 1990 Population Data. Contract No.
68-DO-0062. Prepared by International Technology Corporation for U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards. April.
Johnson, T., J. Capel, M. McCoy and J. Mozier. 1996c. Estimation of Ozone Exposures
Experienced by Outdoor Workers in Nine Urban Areas Using a Probabilistic Version of NEM.
Contract No. 63-D-30094. Prepared by International Technology Corporation for U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards. April.
Johnson, T., M. Weaver and J. Mozier. 1998. Variation of Residential Air Exchange Rates
under Scripted Ventilation Conditions. Paper presented at the International Symposium on the
Measurements of Toxic and Related Pollutants, Air and Waste Management Association, Gary,
NC. September 1-3, 1998.
Johnson, T., G. Mihlan, J. LaPointe, K. Fletcher and J. Capel. 1999. Estimation of Carbon
Monoxide Exposures and Associated Carboxyhemoglobin Levels in Denver Residents Using
pNEM/CO (Version 2.0 ). Contract No. 68-D6-0064. Prepared by TRJ Environmental, Inc. for
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards. March.
Johnston, P.K., C.A. Cinalli, J.R. Girman, and P.W. Kennedy. 1996. Priority Ranking and
Characterization of Indoor Air Sources. In Characterizing Sources of Indoor Air Pollution and
Related Sink Effects. B. A. Tichenor, Ed. ASTM STP 1287. American Society for Testing and
Materials. 392-400.
fA
Jones K.C., T. Keating, P. Diage and A.C. Chang. 1991. Transport and Food Chain Modeling
and its Role in Assessing Human Exposures to Organic Chemicals. J. Environ. Qual. 20:317-
329.
Katch, V. L. and M. W. Park. 1975. Minute-by-Minute Oxygen Requirement and Work
Efficiency for Constant-load Exercise of Increasing Duration. Res. Q. 46:38-47.
NOVEMBER 1999 7^4 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Klepeis, N., P. Switzer, et al. 1994. A Total Human Exposure Model (THEM) for Respirable
Suspended Particles. Air and Waste Management Association Meeting.
Lagus Applied Technology. 1995. Air Exchange Rates in Non-residential Buildings in
California. Document No. P400-91-034. Prepared for the California Energy Commission.
Law, P.L., P.J. Lioy, M.P. Zelenka, A.H. Huber and T.R. McCurdy. 1997. Evaluation of a
Probabilistic Exposure Model applied to Carbon Monoxide (pNEM/CO) using Denver Personal
Exposure Monitoring Data. J. Air & Waste Manage. Assoc. 47:491-500.
Lawryk, N., P.J. Lioy, and C.P. Weisel. 1995. Exposure to Volatile Organic Compounds in the
Passenger Compartment of Automobiles during Periods of Normal and Malfunctioning
Operation. J. Expos. Anal. Environ. Epidem. 5:511-531.
Layton, D.W., L.R. Anspaugh, K.T. Bogen and T. Straume. 1992. Risk Assessment of Soil-
based Exposures to Plutonium at Safety-Shot Sites Located on the Nevada Test Site and
Adjoining Areas. UCRL-ID-112605DR. Livermore, CA: Lawrence Livermore National
Laboratory.
Lioy, P.J. and J.M. Daisey. 1990. Toxic Air Pollution. Chelsea, MI: Lewis Publishers.
Lioy, P.J., L. Wallace and E. Pellizzari. 1991. Indoor/Outdoor, and Personal Monitor and
Breath Analysis Relationships for Selected Volatile Organic Compounds Measured at Three
Homes During New Jersey TEAM - 1987. J. Expos. Anal. Environ. Epidem. 1:45-61.
Lurmann, F.W.. A.M. Winer and S.D.Colome. 1990. Development and Application of a New
Regional Human Exposure (REHEX) Model. In: Total Exposure Assessment Methodology.
Pittsburgh, PA: Air and Waste Management Association, pp. 478-498.
Lurmann, F.W., S.D. Colome and H. Hogo. 1992. Modeling Current and Future Human
Exposure to Ozone in Southern California. In: R.L. Burglund, ed. Tropospheric Ozone and the
Environment II. Pittsburgh, PA: Air and Waste Management Association, pp. 725-744.
Lurmann, F. and E. Korc. 1994. Characterization of Human Exposure to Ozone and PM-10 in the
San Francisco Bay Area, Report No. STI-931501416-FR. Sonoma Technology, Inc. San Rafael,
CA.
Macintosh, D.L., J. Xue, H. Ozkaynak, J.D. Spengler and P.B. Ryan. 1995. A Population-based
Exposure Model for Benzene. J. Expos. Anal. Environ. Epidem. 5:375-403.
McCurdy, T. and T. Johnson. 1989. Development of a Human Exposure Model for Ozone.
Paper presented at the Total Exposure Assessment Methodology-A New Horizon, EPA/AWMA
Specialty Conference, Las Vegas, NV.
McCurdy, T.R. 1994. Human Exposure to Ambient Ozone. In: D.J. McKee, ed. Tropospheric
NOVEMBER 1999 ~5 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Ozone: Human Health and Agricultural Impacts. Ann Arbor, MI: Lewis Publishers, pp. 85-127.
McCurdy, T. 1995. Estimating Human Exposure to Selected Motor Vehicle Pollutants Using
the NEM Series of Models: Lessons to be Learned. J. Expos. Anal. Environ. Epidem.5:533-550.
McCurdy, T. 1997. Modeling the Dose Profile in Human Exposure Assessments: Ozone as an
Example. Reviews in Toxicology. 1:3-23.
McCurdy, T. 1999. Conceptual Basis for Multi-route Intake Dose Modeling Using an Energy
Expenditure Approach. Journal of Exposure Analysis and Environmental Epidemiology
accepted for publication.
McKone, T.E. and P.B. Ryan. 1989. Human Exposures to Chemicals Through Food Chains:
An uncertainty analysis. Environ. Sci. Technol. 23:1154-1163.
McKone, T.E. and J.I. Daniels. 1991. Estimating Human Exposure Through Multiple Pathways
from Air, Water, and Soil. Regul. Toxicol. Pharmacol. 13:36-61.
McKone, T.E. 1993a. CalTOX, a Multimedia Total-Exposure Model for Hazardouswastes
Sites Part I: Executive Summary. UCRL-CR-11456, Pt. I. Livermore, CA: Lawrence
Livermore National Laboratory.
McKone, T.E. t993b. CalTOX, a Multimedia Total-Exposure Model for Hazardous Wastes
Sites Part II: the Dynamic Multimedia Transport and Transformation Model.
UCRL-CR-111456, Pt. II. Livermore, CA: Lawrence Livermore National Laboratory,
McKone, T.E. 1993c. CalTOX, a Multimedia Total-Exposure Model for Hazardous Wastes
Sites Part III: The Multiple-Pathway Exposure Model. UCRL-CR-111456, Pt. III. Livermore,
CA: Lawrence Livermore National Laboratory.
Mercier, J., A. Varray, M. Ramonatxo, B. Mercier and C. Prefaut. 1991. Influence of
Anthropomorphic Characteristics on Changes in Maximal Exercise Ventilation and Breathing
Pattern During Growth in Boys. Eur. J. Appl. Physiol. 63:235-241.
Mermier, J., J.M. Samet, W.E. Lambert and T.W. Chick. 1993. Evaluation of the Relationship
Between Heart Rate and Ventilation for Epidemiologic Studies. Arch. Environ. Health. 48:263-
269.
Miller, S.L., S. Branoff and W.W. Nazaroff. 1998. Exposure to Toxic Air Contaminants in
Environmental Tobacco Smoke: an Assessment for California Based on Personal Monitoring
Data. J. Expos. Anal. Environ. Epidem. 8:287-311.
Morgan, G. M. and M. Henrion. 1990. Uncertainty: A Guide to Dealing with Uncertainty on
Quantitative Risk and Policy Analysis. Cambridge, Cambridge University Press.
NOVEMBER 1999 7-6 TRIM. EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Murray, D. M. and D. E. Burmaster. 1995. Residential Air Exchange Rates in the United States:
Empirical and Estimated Parametric Distributions by Season and Climatic Region. Risk
Analysis. 15:459-465.
Nagda, N.L., H.E. Rector and M.D. Koontz. 1987. Guidelines for Monitoring Indoor Air
Quality. Washington, D.C.: Hemisphere Publishing Corporation.
Nazaroff, W.W. and G.R. Cass. 1986. Mathematical Modeling of Chemically Reactive
Pollutants in Indoor Air. Environ. Sci. Technol. 20:924-934.
Nazaroff, W.W. and G.R. Cass. 1989. Mathematical Modeling of Indoor Aerosol Dynamics.
Environ. Sci. Technol. 23:157-166.
National Research Council (NRC). 1991. Human Exposure Assessment for Airborne Pollutants:
Advances and Opportunities. Washington, D.C.: National Academy Press.
NRC/National Academy of Sciences (NAS). 1994. Science and Judgment in Risk Assessment.
Washington, D.C.: National Academy Press.
Ott, W. 1982. Concepts of Human Exposure to Air Pollution. Environ. Int. 7:179-186.
Ott, W. 1984. Exposure Estimates Based on Computer Generated Activity Patterns. J. Clin.
Toxicol. 21:97-128.
Ott, W.R., D. Mage and L. Wallace. 1988. Validation of the Simulation of Human Activity and
Pollutant Exposure (SHAPE) Model using Paired Days from the Denver, Colorado, Carbon
Monoxide Field Study. Atmos. Environ. 22:2101-2113.
Ott, W., L. Langan, et al. 1992. A Time Series Model for Cigarette Smoking Activity Patterns:
Model Validation for Carbon Monoxide and Respirable Particles in a Chamber and an
Automobile. Journal of Exposure Analysis and Environmental Epidemiology. 2(Suppl. 2):
175-200.
Ott W ; Switzer P ; Willits N. 1994. Carbon monoxide exposures inside an automobile
traveling on an urban arterial highway. J. Air & Waste Manage. Assoc. 44(8): 1010-8
Ozkaynak, H., P.B. Ryan, G.A. Allen and W.A. Turner. 1982. Indoor Air Quality Modeling:
Compartmental Approach with Reactive Chemistry. Environ. Int. 8:461-471.
Ozkaynak, H., J. Xue, J. Spengler, L. Wallace, E. Pellizzari and P. Jenkins. 1996. Personal
Exposure to Airborne Particles and Metals: Results from the Particle TEAM Study in Riverside,
California. J. Expos. Anal. Environ. Epidem. 6:57-78.
Ozkaynak, H. 1999. Personal communication (electronic mail correspondence).
NOVEMBER 1999 7-7 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Ozkaynak H., Zufall, M., Burke, J., Xue, J., Zidek, J. 1999a. Predicting Population Exposures
to PM10 and PM2.5. Presented at PM Colloquium, Durham, NC. June, 1999.
Ozkaynak H., Zufall, M., Burke, J., Xue, J., Zidek, J. 1999b. A Probabilistic Population
Exposure Model for PM10 and PM2.5. Presented at 9th Conference of the International Society
of Exposure Analysis, Athens, Greece. September 5-8, 1999.
Palma, T., M. Riley and J.E. Capel. 1996. Development and Evaluation of Enhancements to the
Hazardous Air Pollutant Exposure Model (HAPEM-MS3). Contract No. 63-DO-30094.
Prepared by International Technology Corporation for U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards. September.
Panguluri, S., T. Johnson and J. Capel. 1998. Development of pHAP-PC Model: A Summary
Report. Contract No. 63-D-30094. Prepared by International Technology Corporation for U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, June 1998.
Phillips, K., M.C. Bentley, D.A. Howard and G. Alvan. 1998. Assessment of Environmental
Tobacco Smoke and Respirable Suspended Particle Exposures for Nonsmokers in Kuala Lumpur
Using Personal Monitoring. J. Expos. Anal. Environ. Epidem. 8:519-542.
Quackenboss, J.J., M. Krzyzanowski and M.D. Lebowitz. 1991. Exposure Assessment
Appioaches to Evaluate Respiratory Health Effects of Partioulate Matter and Nitrogen Dioxide.
J. Expos. Anal. Environ. Epidem. 1:83-107.
Raunemaa, T., M. Kulmala, H. Saari, M. Olin and M.H. Kulmala. 1989. Indoor Air Aerosol
Model - Transport Indcors and Deposition of Fine and Coarse Particles. Aerosol Sci. Technol.
11:11-25.
RIVM, VROM, WVC. 1994. Uniform System for the Evaluation of Substances (USES),
Version 1.0. VROM Distribution No. 11144/150. National Institute of Public Health and
Environmental Protection (RIVM) Ministry of Housing, Spatial Planning and Environment
(VROM), Ministry of Welfare, Health, and Cultural Affairs (WVC), The Hague, The
Netherlands.
Roddin, Ellis, and Siddiqee. 1979. Background Data for Human Activity Patterns, Vols. I and
II." SRI International
Rosenbaum, A.S., G.E. Anderson and G.W. Lundberg. 1994. User's Guide to the South Coast
Risk and Exposure Assessment Model - Version 2 (SCREAM-II) for the Personal Computer.
San Rafael, CA: Systems Applications International.
Rowland, T.W., J.A. Auchinachie, T.J. Keenan and G.M. Green. 1987. Physiologic Responses
to Treadmill Running in Adult and Prepubertal Males. Inter. J. Sports Med. 8:292-297.
Ryan, P.B., J.D. Spengler, and R. Letz. 1983. The Effects of Kerosene Heaters on Indoor
NOVEMBER 1999 ?1 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Pollutant Concentrations: a Monitoring and Modeling Study. Atmos. Environ. 17:1339.
SAL 1999. Systems Applications International. User's Guide: Assessment System for
Population Exposure Nationwide (ASPEN, Version 1.1)- Volumes I and II. SYSAPP-98/25r2.
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards.
Schofield, W.N. 1985. Predicting Basal Metabolic Rate, New Standards, and Review of
Previous Work. Hum. Nutr. Clin. Nutr. 39C(Supp.l):5-41.
Sedman, R.M. and R.J. Mahmood. 1994. Soil Ingestion by Children and Adults Reconsidered
Using the Results of Recent Tracer Studies. J. Air & Waste Manage. Assoc. 44:41-144.
Sega, K. and M. Fugas. 1991. Different Approaches to the Assessment of Human Exposure to
Nitrogen Dioxide. J. Expos. Anal. Environ. Epidem. 1:227-234.
Stanek, E.J., E.J. Calabrese, K. Mundt, P. Pekow and K.B. Yeatts. 1998. Prevalence of Soil
Mouthing/ingestion among Healthy Children Aged 1 to 6. J. of Soil Contam. 7:227-242.
Stock, T.H., D.J. Kotchman and C.F. Contant. 1985. The Estimation of Personal Exposures to
Air Pollutants for a Community-based Study of Health Effects in Asthmatics - Design and
Results of Air Monitoring. J. Air Pollut. Control Assoc. 35:1266-1273.
Strenge, D.L. and P.J. Chamberlain. 1995. Multimedia Environmental Pollutant Assessment
System (MEPAS): Exposure Pathway and Human Health Impact Assessment Models. PNL-
10523 3 ichland, WA: Pacific Northwest Laboratory.
Thibodeaux, L.J. 1979. Chemodynamics, Environmental Movement of Chemicals in Air,
Water, and Soil. New York, NY: John Wiley and Sons.
Thomas, K.W., E.D. Pellizzari, C.A. Clayton, R.L. Perritt, R.N. Dietz, R.W. Goodrich, W.C.
Nelson and L.A. Wallace. 1993. Temporal Variability of Benzene Exposures for Residents in
Several New Jersey Homes with Attached Garages or Tobacco Smoke. J. Expos. Anal. Environ.
Epidem. 3:49-73.
Thompson, K.M. and D.E. Burmaster. 1991. Parametric Distributions for Soil Ingestion by
Children. Risk Analysis. 11:339-342.
Travis, C.C. and S.T. Hester. 1991. Global Chemical Pollution. Environ Sci Technol.
25:814-819.
Travis, C.C. and B.P. Blaylock. 1992. Validation of a Terrestrial Food Chain Model. J. Expos.
Anal. Environ. Epidem. 2:221-239.
Turk, B.H., D.T. Grimsrud, J.T. Brown, K.L. Geisling-Sobotke, J. Harrison, and R.J. Prill. 1989.
NOVEMBER 1999 7^9 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Commercial Building Ventilation Rates and Particle Concentrations. ASHRAE Transactions.
95(Part 1).
U.S. EPA. 1983. U.S. Environmental Protection Agency. The NAAQS Exposure Model
Applied to Carbon Monoxide. EPA-450/5-83-004. Research Triangle Park, NC: Office of Air
Quality Planning and Standards.
U.S. EPA. 1984. U.S. Environmental Protection Agency. A Study of Personal Exposure to
Carbon Monoxide in Denver, Colorado. EPA-600/54-84-014. Research Triangle Park, NC:
Office of Research and Development.
U.S. EPA. 1986. U.S. Environmental Protection Agency. Users Manual for the Human
Exposure Model (HEM). EPA-540/5-86-001. Research Triangle Park, NC: Office of Air
Quality Planning and Standards. June.
U.S. EPA. 1987. U.S. Environmental Protection Agency. The Total Exposure Assessment
Methodology (TEAM) Study: Summary and Analysis (Volume 1). EPA-600/6-87-002a, NTIS
#PB 88-100060. Washington, D.C.: Office of Acid Deposition. June.
U.S. EPA. 1990a. U.S. Environmental Protection Agency. Benzene Exposure Assessment
Model (BEAM): 2nd Interim Report. EPA/600/X-90/027. Las Vegas, NV: Environmental
Monitoring Systems Laboratory.
U.S. EPA. 1990b. U.S. Environmental Protection Agency. Methcdoiogy for Assessing Health
Risks Associated with Indirect Exposure to Combustor Emissions. Interim Final.
EPA/600/6-90/003. Washington DC, Office of Health and Environmental Assessment.
U.S. EPA. 1991. U.S. Environmental Protection Agency. HEM-II User's Guide. EPA/450/3-
91-0010. Research Triangle Park, NC: Office of Air Quality Planning and Standards.
U.S. EPA. 1992a. U.S. Environmental Protection Agency. Guidelines for Exposure
Assessment: Notice. Fed. Reg. 57(104):22888-22938. May 29.
U.S. EPA. 1992b. U.S. Environmental Protection Agency. Toxic Modeling System Long-Term
(TOXLT): User's Guide. EPA-450/4-92-003. Research Triangle Park, NC: Office of Air
Quality Planning and Standards. November.
U.S. EPA. 1992c. U.S. Environmental Protection Agency. A Tiered Modeling Approach for
Assessing the Risks Due to Sources of Hazardous Air Pollutants. EPA-450/4-92-001.
Research Triangle Park, NC: Office of Air Quality Planning and Standards.
U.S. EPA. 1992d. U.S. Environmental Protection Agency. Multimedia Contaminant Fate,
Transport, and Exposure Model: Documentation and User's Manual. Washington, D.C.:
Office of Health and Environmental Assessment and Office of Environmental Processes and
Effects Research.
NOVEMBER 1999 7-10 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
U.S. EPA. 1993a. U.S. Environmental Protection Agency. Estimation of the Exposure to
Benzene of Selected Populations in the State of Texas Using the Benzene Exposure Assessment
Model (BEAM). Las Vegas, NV: Environmental Monitoring Systems Laboratory.
U.S. EPA. 1993b. U.S. Environmental Protection Agency. Addendum to Methodology for
Assessing Health Risks Associated with Indirect Exposure to Combustor Emissions. External
Review Draft. EP A/600/AP-93/003. Washington DC, Office of Health and Environmental
Assessment.
U.S. EPA. 1994a. U.S. Environmental Protection Agency. Report of the Agency Task Force on
environmental regulatory modeling. Guidance, support needs, draft criteria and charter. EPA
500-R-94-001. Washington, DC: Office of Solid Waste and Emergency Response.
U.S. EPA. 1994b. U.S. Environmental Protection Agency. Toxic Modeling System Short-Term
(TOXST), User's Guide: Volume I. EPA-453/R-94-058A. Research Triangle Park, NC: Office
of Air Quality Planning and Standards. July.
U.S. EPA. 1996a. U.S. Environmental Protection Agency. Analysis of the National Human
Activity Pattern Survey (NHAPS) Respondents from a Standpoint of Exposure Assessment.
EPA/600/R-96/074. Washington, D.C.: Office of Research and Development. July.
U.S. EPA. 1996b. U.S. Environmental Protection Agency. Summary Report for the Workshop
on Monte Carlo Analysis. EPA/630/R-96/010. Washington, D.C.: Office of Research and
Development. September.
U.S. EPA. 1996c. U.S. Environmental Protection Agency. Exposure Models Library and
Integrated Model Evaluation System. Revised. EPA/600/C-92/002. Washington, D.C.: Office
of Research and Development. March.
U.S. EPA. 1997a. U.S. Environmental Protection Agency. Mercury Study Report to Congress
(Volumes I-VIII). EPA-452/R-97-005. Office of Air Quality Planning and Standards.
U.S. EPA. 1997b. U.S. Environmental Protection Agency. Exposure Factors Handbook; Vol I:
General Factors, Vol. II: Food Ingestion Factors, Vol. Ill: Activity Factors. EPA/600/P-
95/002Fa (Vol. I), EPA/600/P-95/002Fb (Vol. II), EPA/600/P-95/002Fc (Vol. III). Washington,
D.C.: Office of Research and Development. August.
U.S. EPA. 1997c. U.S. Environmental Protection Agency. Guiding Principles for Monte Carlo
Analysis. EPA/630/R-97/001. Washington, D.C.: Office of Research and Development. March.
U.S. EPA. 1997d. U.S. Environmental Protection Agency. Methodology for Assessing Health
Risks Associated with Multiple Exposure Pathways to Combustor Emissions. Update to
EPA/600/6-90/003. National Center for Environmental Assessment. NCEA 0238. Cincinnati,
OH.
NOVEMBER 1999 7-11 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
U.S. EPA. 1998a. U.S. Environmental Protection Agency. The Total Risk Integrated
Methodology. Implementation of the TRIM Conceptual Design through the TRIM.FaTE
Module. A Status Report. EPA-452/R-98-001. Research Triangle Park, NC: Office of Air
Quality Planning and Standards. March.
U.S. EPA. 1998b. U.S. Environmental Protection Agency. Advisory on the Total Risk
Integrated Methodology (TRIM). EPA-SAB-EC-ADV-99-003. Science Advisory Board.
U.S. EPA. 1998c. U.S. Environmental Protection Agency. Peer Review Handbook. EPA/100-
B-98-001. January. Washington, DC: Office of Science Policy.
U.S. EPA. 1998d. U.S. Environmental Protection Agency. Study of Hazardous Air Pollutants
from Electric Utility Steam Generating Units - Final Report to Congress. EPA 453/R-989-004a.
February.
U.S. EPA. 1998e. U.S. Environmental Protection Agency. Human Health Risk Assessment
Protocol for Hazardous Waste Combustion Facilities. Peer Review Draft. EPA530-D-98-001 A.
Office of Solid Waste and U.S. EPA Region 6, Multimedia Planning and Permitting Division.
U.S. EPA. 1998f. U.S. Environmental Protection Agency. EPA Third-generation Air Quality
Modeling System. EPA-600/R-98/069(b). Washington, D.C.: Office of Research and
Development. June.
U.S. EPA. 1998g. U.S. Environmental Protection Agency. THERdbASE, Version 1.2, User's
Manual. Las Vegas, NV: Office of Research and Development. August.
U.S. EPA 1999a. U.S. Environmental Protection Agency. Total Risk Integrated Methodology
(TRIM). TRIM.FaTE Technical Support Document. Volume I: Description of Module. EPA-
453/D-99-002A. Research Triangle Park, NC: Office of Air Quality Planning and Standards.
November.
U.S. EPA 1999b. U.S. Environmental Protection Agency. Total Risk Integrated Methodology
(TRIM). TRIM.FaTE Technical Support Document. Volume II: Description of Chemical
Transport and Transformation Algorithms. EPA-453/D-99-002B. Research Triangle Park, NC:
Office of Air Quality Planning and Standards. November.
U.S. EPA 1999c. U.S. Environmental Protection Agency. Total Risk Integrated Methodology
(TRIM). Status Report. EPA-453/R-99-010. Research Triangle Park, NC: Office of Air Quality
Planning and Standards. November.
U.S. EPA. 1999d. U.S. Environmental Protection Agency. Systems Installation and Operation
Manual for the EPA Third-Generation Air Quality Modeling System (Models-3 Version 3.0).
Atmospheric Modeling Division, National Exposure Research Laboratory, Research Triangle
Park, NC and EPA Systems Development Center (A contractor operated facility), Science
Applications International Corporation, Arlington, VA. June.
NOVEMBER 1999 7^12 TRIM.Expo TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
U.S. EPA. 1999e. U.S. Environmental Protection Agency. National Air Toxics Program: The
Integrated Urban Strategy. Federal Register 64: 38705-38740. July 19.
U.S. EPA. 1999f. U.S. Environmental Protection Agency. Residual Risk Report to Congress.
EPA - 453/R-99-001. Office of Air Quality Planning and Standards, Research Triangle Park,
NC. March.
U.S. EPA. 1999g. U.S. Environmental Protection Agency. Report of the Workshop on
Selecting Input Distributions for Probabilistic Assessments. EPA/630/R-98/004. Washington,
D.C.: Office of Research and Development. January.
U.S. EPA. 1999h. Methodology for Assessing Health Risks Associated with Multiple Pathways
of Exposure to Combustor Emissions. National Center for Environmental Assessment. Office of
Research and Development. EPA 600/R-98/137.
Van Veen, M.P. 1995. CONSEXPO - A program to estimate consumer product exposure and
uptake. Report No. 612810.002. Bilthoven, The Netherlands: RIVM.
\Vadden, R. A. and P. A. Scheff. 1983. Indoor Air Pollution Characterization, Prediction and
Control. New York, NY: Wiley Interscience.
Waldman, J.M., P.J. Lioy, A. Greenberg and J.P. Butler. 1991. .Analysis of Human Exposure to
3enzo(a)pyrene via Inhalation and Food Ingestion in the Total Human Environmental Exposure
Study (THEES). J. Expos. Anal. Environ. Epidem. 1:193-225.
Wallace, L.A., E.D. Pellizzari, T. Hartwell, C.M. Sparacino, L.S. Sheldon and H.S. Zelon. 1985.
Personal Exposures, Indoor/outdoor Relationships, and Breath Levels of Toxic Pollutants
Measured for 355 Persons in New Jersey. Atmos. Environ. 19:1651-1669.
Wallace, L., W. Nelson, R. Ziegenfus, E. Pellizzari, L. Michael, R. Whitmore, H. Zelon, T.
Hartwell, R. Perritt and D. Westerdahl. 1991. The Los Angeles TEAM Study: Personal
Exposures, Indoor-outdoor Air Concentrations, and Breath Concentrations of 25 Volatile
Organic Compounds. J. Expos. Anal. Environ. Epidem. 1:157-192.
Whicker, F.W. and T.B. Kirchner. 1987. PATHWAY: A dynamic food-chain model to predict
radionuclide ingestion after fallout deposition. Health Phys. 52:717-737.
Wilkes, C.R., M.J. Small, J.B. Andelman, N.J. Giardino and J. Marshall. 1992. Inhalation
Exposure Model for Volatile Chemicals from Indoor Uses of Water. Atmos. Environ.
26A:2227-2236.
Winer, A.M., et al. 1989. Characterization of Air Pollution Exposures in the California South
Coast Air Basin: Application of a new Regional Human Exposure (REHEX) model. Riverside,
CA: University of California.
NOVEMBER 1999 7-13 TRIM.EXPO TSD (DRAFT)
-------
CHAPTER 7
REFERENCES
Yang, Y.Y. and C.B. Nelson. 1986. An Estimation of Daily Food Usage Factors for Assessing
Radionuclide Intakes in the U.S. Population. Health Phys. 50:245-257.
Zartarian, V.G. 1996. DERM (Dermal Exposure Reduction Model): A physical-stochastic
model for understanding dermal exposure to chemicals. Doctoral Dissertation. Stanford
University, Stanford, CA.
Zartarian V.G., Ozkaynak H., Burke J.M., Zufall M.J., Rigas M.L., and Furtaw Jr. E.J. A
Modeling Framework For Estimating Children's Residential Exposure and Dose to Chlorpyrifos
Via Dermal Residue Contact and Non-Dietary Ingestion. Submitted to Environmental Health
Perspectives. September 1999.
NOVEMBER 1999 7-14 TRIM.EXPO TSD (DRAFT)
-------
Absorbed dose
APPENDIX A
GLOSSARY
APPENDIX A
Glossary
The amount of pollutant that crosses a specific absorption barrier
(e.g., the exchange boundaries of skin, lung, or digestive tract)
through uptake processes. Absorbed dose is calculated from the
intake and absorption efficiency and is usually expressed as the
mass of pollutant absorbed into the body per unit time (mg/kg/d).
For inhalation exposure, absorbed dose is the amount of material
that passes from the lung volume into the blood. For ingestion
exposure, absorbed dose is the quantity of pollutant that passes
from the volume of the gastrointestinal tract across the gut wall
and into the blood stream. For dermal exposure, absorbed dose is
the quantity of material that passes through the stratum corneum
into the living cells of the epidermis and dermis and then into the
blood stream. Sometimes referred to as internal dose.
Any of the exchange boundaries of the body (e.g., the skin, lung
tissue, digestive tract, gastrointestinal tract wall) that allow
differential diffusion of various pollutants across the boundary.
A series of discrete events of varying time intervals describing
information about an individual's lifestyle and routine. The
information contained in an activity pattern typically includes the
locations that the individual visited (usually described in terms of
microenvironments), the amount of time spent in those locations,
and a description of what the individual was doing in each location
(e.g., sleeping, eating, exercising). All of the information for an
activity pattern is gathered during an "activity pattern survey,"
usually through the use of questionnaires or diaries. Each activity
pattern survey is designed to collect information on activities
needed for a particular study or purpose. Activity patterns are also
referred to as "time/activity patterns."
The amount of a pollutant given in mg/kg/d that comes in contact
with the living tissue of an organism by entering into the lungs, by
entering the gastrointestinal tract, and/or by crossing the stratum
comeum into the living cells of the epidermis. In some
experimental designs, the applied dose is referred to as the
administered dose.
Average daily dose (ADD) Dose rate within a population averaged over body weight and an
averaging time and typically expressed in terms of mg/kg/d.
Absorption barrier
Activity pattern
Applied dose
NOVEMBER 1999
A-l
TRJM.EXPO TSD (DRAFT)
-------
APPENDIX A
GLOSSARY
Biologically effective dose
Breathing zone
Cohort
Demographic group
Dermal
The amount of a deposited or absorbed pollutant that reaches the
cells or target site where an adverse effect occurs or where that
pollutant interacts with a membrane surface
A zone of air in the vicinity of an organism from which respired air
is drawn Personal monitors are often used to measure pollutants in
the vicinity of the breathing zone.
A group of people within a population who are assumed to have
similar exposures are taken from the same probability distribution
during a specified exposure period.
The use of cohorts is useful when modeling the exposures of a large
population Since adequate data on the exposures of each
individual in a population does not exist, information about people
who are expected to have similar exposures are aggregated
together in order to make better use of the limited data that is
available
Cohorts can be defined for each application or situation In the
latest pNEM/CO model, for example, cohort exposure was taken to
be a function of demographic group, location of residence, location
of work place, and type of cooking fuel (natural gas or other)
Specifying the home and work district of each cohort provided a
means of linking cohort exposure to ambient CO concentrations
Specifying the demographic group provided a means of linking
cohort exposure to activity patterns which vary with age, work
status, and other demographic variables. Specifying the type of
cooking fuel provided a means of linking cohort exposure to
proximity to a particular emission source In some analyses,
cohorts are further distinguished according to factors relating to
time spent in particular microenvironments In the pNEM analyses,
the population-of-interest is divided into a set of cohorts such that
each person is assigned to one and only one cohort
A group of people within a population sharing common
demographic characteristics such as gender, race, household
income, working status, or incidence of a particular disease or
ailment. These groups can be defined differently depending on the
study or application and much of this information can be gathered
from the census
The external skin surface of an organism.
NOVEMBER 1999
A-2
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX A
GLOSSARY
Dermal exposure
Dose
Dose rate
Environmental media
Exposure
Exposure assessment
Exposure district
Exposure event
Exposure factor
The contact between a pollutant and the external skin surface of a
biological organism.
The quantity of energy or pollutant available for interaction with
metabolic processes or biological receptors after crossing the outer
boundary of an organism. See related terms: absorbed dose,
applied dose, biologically effective dose, delivered dose, internal
dose, and potential dose.
Dose per unit time (mg/d). Dose rate is often expressed on a
per-unit-bodyweight-basis (e.g., mg/kg/d) and may be expressed as
an average over a long time period (e.g., a lifetime). Also referred
to as dosage.
The components of the physical environment that carry a pollutant,
and through which pollutants can move and reach the organisms.
The environmental media in TRIM.Expo include ambient air,
ground water, surface water, surface soil, root zone soil, vadose
zone soil, and several classes of vegetation.
The contact between a target organism and a pollutant at the outer
boundary of the organism. Exposure may be quantified as the
amount of pollutant available at the boundary of the receptor
organism per specified time period. As an example, inhalation
exposure over a period of time may be represented by a time-
dependent profile of the exposure concentrations.
Measurement or estimation of the magnitude, frequency, duration,
and route of exposure of biological organisms to pollutants in the
environment for a specified time period. An exposure assessment
also describes the nature of exposure and the size and nature of the
exposed populations.
A geographic location within a defined physical or political region
where there is potential contact between an organism and a
pollutant, and for which environmental media concentrations have
been estimated either through modeling or measurement.
A human activity that results in contact with a contaminated
medium within a specified microenvironment at a given
geographic location.
A normalizing or standardizing factor used in an exposure
assessment as a surrogate for specific information that is not
available for a particular subject, cohort, or demographic group.
NOVEMBER 1999
A-3
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX A
GLOSSARY
Exposure media
Exposure pathway
Exposure route
Fixed-site monitoring
Ingestion
Ingestion exposure
Inhalation
Inhalation exposure
Intake
These factors are often drawn from a distribution or a range of data
[see for example, EPA's Exposure Factors Handbook (U.S. EPA
1997b)].
The part of the physical environment that surrounds or contacts
organisms at the time of an exposure. The exposure media in
TRIM.Expo include outdoor air, indoor air (multiple
microenvironments), tap water, home-grown food, locally-
produced food, prepared food, breast milk, house dust, soil,
swimming pools, and other recreational surface water.
The physical course of a pollutant from the source to the exposed
organism. An exposure pathway describes a unique mechanism by
which an individual or population is exposed to pollutants or
physical agents at, or originating from, a site. Each exposure
pathway includes a source, or release from a source, an exposure
point, and an exposure route. If the exposure point differs from the
source, a transport/exposure medium (such as air) or media (in
cases of intermedia transport, such as water to air) is also included.
The way a pollutant enters an organism after contact, including
inhalation, ingestion, or dermal absorption.
Sampling of an environmental or ambient medium for a pollutant's
concentration at the same location continuously or repeatedly over
some length of time.
An exposure route whereby pollutants enter the body by the mouth
for digestion or absorption.
The contact between a pollutant and the boundary in the area
surrounding the mouth at the time of ingestion. The contact
boundary is often defined for each particular situation or study.
An exposure route whereby air and pollutants are drawn into the
lungs via the nasal or oral respiratory passages.
The contact between an airborne pollutant and a human, or other
animal, at the time of inhalation.
The process by which a pollutant crosses the outer boundary of an
organism prior to passing an absorption barrier (e.g., through
ingestion or inhalation).
NOVEMBER 1999
A-4
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX A
GLOSSARY
Intermedia transfer
Internal dose
An algorithm for "linking" the environmental media with the
microenvironmental media that exposed individuals occupy (e.g.,
air compartments) or the exposure media with which they come in
contact (e.g., air, water, food, soil). An intermedia transfer
algorithm relates the pollutant concentration in a
microenvironmental medium to the concentration in an ambient
environmental medium that provides an input to that
microenvironment.
See absorbed dose.
Lifetime average daily dose LADD is the average daily dose within a population when the
averaging time is the expected individual lifetime and is usually
expressed in terms of mg/kg/d for compounds with carcinogenic or
chronic effects.
Microenvironment
Monte Carlo technique
Multipathway exposure
Potential dose
Risk
Risk assessment
A defined space in which human contact with an environmental
pollutant takes place and which can be treated as a
well-characterized, relatively homogeneous location with respect
to pollutant concentrations for a specified time period.
A statistical method that uses repeated random sampling from the
distribution of values for each of the parameters in a generic
(exposure or dose) equation to derive an estimate of the
distribution of (exposures or doses in) the population.
Exposure in which the pollutant travels via more than one
environmental medium from its source to the point of contact with
the exposed organism. For example, a pollutant which is released
into the air, but is then deposited into a water body before coming
in contact with a person.
An approximation to the applied dose that is simply the amount of
a pollutant in the material ingested, air inhaled, or material applied
to the skin. The potential dose for inhalation and ingestion is
analogous to the administered dose in a dose-response experiment.
For the dermal route, the potential dose is the amount of pollutant
applied or the amount of pollutant in the exposure medium applied
to the skin.
The probability of deleterious effects that may result from an
action or inaction.
The process of evaluating the toxic properties of a pollutant and the
conditions of human exposure to the pollutant in order to ascertain
NOVEMBER 1999
A-5
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX A
GLOSSARY
the likelihood that exposed humans will be adversely affected and
to characterize the nature of the effects they may experience. Risk
assessments historically have included the following four steps: (1)
hazard identification (determination of whether or not a particular
pollutant is causally linked to a particular adverse health effect);
(2) exposure assessment (determination of the amount of the
pollutant that humans are exposed to and the conditions of that
exposure); (3) dose-response assessment (quantification of how
adverse effects change with dose); and (4) risk characterization (the
final analysis that integrates the scientific findings of the previous
three components to assess the overall conclusions about potential
human risk including a description of the expected nature and
severity of harm associated with the risk).
Uptake
The process in which a pollutant crosses an absorption barrier and
is absorbed into the body.
NOVEMBER 1999
A-6
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
APPENDIX B
Comparison/Critique of Exposure Models
This appendix provides a review of several existing and emerging concentration and
human exposure assessment models The review in this chapter is organized according to the
general characteristics of the concentration and human exposure models. To begin the review,
models were generally classified according to whether they were (1) concentration models or (2)
exposure models Models were then further characterized into more specific categories. For
example, the concentration models were further subdivided into outdoor air (Section B.I) and
indoor air (Section B 2) concentration models. The exposure models were subdivided according
to the exposure media represented, air (Section B.3), consumer products (Section B 4), dietary
(Section B 5), and multimedia (Section B 6) Exposure simulation modeling systems were also
reviewed (Section B.7) These are not individual models per se, rather they are a compilation of
various components that are integrated through a common computer system The parts of this
system can include such varied components as air quality models (e.g., atmospheric dispersion
models or other types of fate/transport models), Geographic Information System (GIS)
capabilities, environmental and various other databases, as well as exposure models and
physiologically-based pharmacokinetic (PBPK) models
This appendix is organized by model categories - air concentration models (i.e., indoor
and outdoor air quality models) and human exposure models, therefore facilitating comparison of
models with similar characteristics For the air concentration models (Section B 1 & B.2),
"Summary Features" are provided that display the key attributes of each model
B.I OUTDOOR AIR CONCENTRATION MODELS
Three models were identified that primarily assess outdoor air concentrations TOXLT,
TOXST, and ASPEN Both TOXLT and TOXST can be accessed via the EPA's Exposure
Models Library (EML) (U S EPA 1996c).
B.I.I TOXLT (Toxic Modeling System Long-Term)
The Toxic Modeling System Long-Term (TOXLT) is a PC-based model that was
developed in conjunction with the release of the EPA's Industrial Source Complex (ISC2)
Dispersion Models (U.S EPA 1992b) Both the TOXLT and ISC2 models coincided with the
promulgation of the EPA's guidance entitled, "A Tiered Modeling Approach for Assessing the
Risks due to Sources of Hazardous Air Pollutants" (U.S. EPA 1992c) The TOXLT computer
system was established by OAQPS to examine both the lifetime cancer risks and the chronic
noncancer hazard indexes associated with toxic pollutants. The purpose of TOXLT is to assist in
the evaluation of the lifetime cancer risks and chronic noncancer hazards that may result from
long-term exposure to toxic air pollutants. The ISCLT2 model is used to simulate annual average
pollutant concentrations which are then used to estimate cancer risk levels or hazard index values
at each user-specified receptor. These outputs presume (1) a hypothetical individual exists at
each receptor, (2) no contribution from "background" sources (i.e., sources not specifically
included in the simulation); and (3) pollutant contributions in a mixture are additive (i.e., there are
NOVEMBER 1999 B-l TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
no synergistic or antagonistic interactions between pollutants).
Summary Features - TOXLT
Environmental media: Ambient air
Pollutants. Multiple gas- and particle-phase agents
Time scale Long-term (annual average)
Stochastic No
Variability No
Uncertainty No
B.1.2 TOXST (Toxic Modeling System Short-Term)
The former Integrated Toxic Expected Exceedance Model (INTOXX), which was based
on the superseded version of the Industrial Source Complex Short-Term (ISCST) Model, was
revised to become the Toxic Modeling System Short-Term (U S. EPA 1994b) TOXST
addresses the problem of estimating expected exceedances of specified short-term health effects
thresholds in the vicinity of continuous and intermittent toxic pollutant releases Certain industrial
facilities emit airborne toxic chemicals known to be harmful when their concentrations exceed a
specified health effect threshold value for a specified length of time However, releases of such
chemicals often occur intermittently This random emission pattern makes it difficult to predict
the frequency with which ambient concentrations will exceed the health effect threshold TOXST
attempts to avoid the problems of underestimation and overestimation of exceedance rates
resulting from random emission patterns by using the Monte Carlo simulation of source emissions
of user-specified durations and rates at randomly selected points in time over a simulated long
period of time In addition, TOXST maintains the capability of simulating continuous emission
sources along with intermittent emission, thereby providing a more realistic simulation of actual
industrial operations
Summary Features - TOXST
Environmental media Ambient air
Pollutants Multiple gas- and particle-phase agents
Time scale Short term to long-term (~ 1 hour to continuous)
Stochastic Yes
Variability Yes
Uncertainty No
B.I.3 ASPEN (Assessment System for Population Exposure Nationwide)
The Assessment System for Population Exposure Nationwide (ASPEN) was originally
developed with support from EPA's Office of Policy (OP) [formally the Office of Policy,
Planning, and Evaluation (OPPE)] The model is being applied by OAQPS as part of their
National Air Toxics Assessment (NATA) activities ASPEN, which is used for hazardous air
pollutants, consists of three separate modules (SAI 1999).
1 A dispersion module estimates ambient concentration increments at a set of fixed receptor
locations in the vicinity of an emission source;
NOVEMBER 1999 1^2 TRIM.Expd TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
2 A mapping module interpolates ambient concentration increment estimates from the grid
receptors to census tract centroids and sums contributions from all modeled sources; and
3. An exposure module, currently under development, will estimate the average
concentration increment to which the population of a census tract is exposed, accounting
for time spent in both indoor and outdoor microenvironments and time spent in other
census tracts
The ASPEN dispersion module, like its predecessors HEM and SCREAM2, uses a
Gaussian model formulation and climatological data to estimate long-term average
concentrations For each source, the model calculates ground-level concentrations as a function
of radial distance and direction from the source for a set of receptors laid out in a radial grid
pattern The concentrations represent the steady-state concentrations that would occur with
constant emissions and meteorological parameters. For each grid receptor, concentrations are
calculated for each combination of stability class, wind speed, and wind direction These
concentrations are then averaged together. The resulting output of ASPEN's dispersion module
is a grid of annual average outdoor concentration estimates for each source/pollutant
combination Improvements to HEM and SCREAM2 that have been incorporated into ASPEN
include
Expansion of reactive decay options;
Inclusion of simple treatment of secondary formation,
Improvement of the deposition algorithm,
Improved treatment of locations near major point sources, and
Improved treatment of area and mobile source emissions
The annual average concentration estimates from ASPEN's dispersion module are then
interpolated from the grid receptors to census tract centroids with ASPEN's mapping module.
The contributions from all modeled sources are summed to give estimates of cumulative ambient
concentration increments in each census tract. The concentration estimates are designed to
represent population-weighted concentration averages for each census tract
The number of emission sources, receptors, and pollutants for an ASPEN application are
virtually unlimited It has been applied to more than 200,000 point, area, and mobile emission
sources of 148 hazardous air pollutants to estimate outdoor concentrations in the more than
60,000 census tracts in the contiguous U.S. Work is underway to link an appropriate exposure
module to ASPEN.
Summary Features - ASPEN
Environmental media: Ambient air
Pollutants Multiple gas- and particle-phase agents
Time scale Long-term (annual average)
Stochastic1 No
Variability. No
Uncertainty: No
NOVEMBER 1999 B^3 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.2 INDOOR AIR CONCENTRATION MODELS
Several approaches have been used to estimate expected indoor air pollutant
concentrations (for a review, see Wadden and Scheff 1983) These approaches include
deterministic models based on a pollutant mass balance around a particular indoor air volume; a
variety of empirical approaches based on statistical evaluation of test data and (usually) a least-
squares regression analysis; or a combination of both approaches - empirically fitting the
parameters of a mass balance model with values statistically derived from experimental
measurements All three approaches have advantages and disadvantages The mass balance
models provide more generality in their application, but often the information on various input
parameters is unavailable to carry out a mass balance approach The empirical models, when
applied within the range of measured conditions for which they were fitted, provide more accurate
information Mass balance models include single and multiple compartment models Often the
component of the indoor air mass balance models that is most difficult to represent is the role of
indoor surfaces as sources or sinks for pollutants
B.2.1 INDOOR, EXPOSURE, and RISK (EPA ORD Indoor Air Quality Models)
INDOOR, EXPOSURE, and RISK are a series of three indoor air quality models
developed by the Indoor Air/Radon Mitigation Branches of EPA's National Risk Management
Research Laboratory within the Office of Research and Development (ORD) The first model,
INDOOR, was designed to calculate the indoor pollutant concentrations from indoor sources
The second model, EXPOSURE, extended INDOOR to allow calculation of individual exposure
The RISK model extends EXPOSURE to allow analysis of individual risk to indoor pollutant
sources Risk estimates generated by models such as RISK are useful mainly for the purpose of
comparing scenarios, rather than for estimating risks to individuals or populations
The RISK model uses data on source emissions, room-to-room air flows, air exchange
with the outdoors, and indoor sinks to predict concentration-time profiles for all the rooms The
concentration-time profiles are then combined with individual activity patterns to estimate
exposure Risk is calculated using a risk calculation framework The model allows analysis of the
effects of air cleaners located in the central air circulating system and/or individual rooms on IAQ
and exposure The model allows simulation of a wide range of sources including long-term steady
state sources, intermittent sources, and decaying sources. Several sources can be modeled in each
room The model allows the.analysis of the effects of sinks and sink re-emissions on IAQ The
results of test house experiments were compared with model predictions. The agreement between
predicted concentration-time profiles and the test house data was good The model is designed to
run in the Windows operating environment
Summary Features - INDOOR. EXPOSURE. RISK
Environmental media. User-specified indoor sources
Pollutants. Multiple chemicals and radon
Time scale Annual average
Stochastic No
Variability. Yes
Uncertainty. No
NOVEMBER 1999 B-4 TRIM. EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.2.2 MA VRIQ (Model for Analysis of Volatiles and Residential Indoor Air Quality)
The Model for Analysis of Volatiles and Residential Indoor Air Quality (MAVRIQ )
(Wilkes et a) 1992) was developed jointly at Carnegie Mellon University and the University of
Pittsburgh It is a compartmental mass balance model that was developed to address human
exposure to volatile organic compounds released from showers and other household water uses
In MAVRIQ, the indoor environment is divided into multiple compartments with constant or
varying air flows Based on water supply concentrations as an input, MAVRIQ accounts for
pollutant generation, chemistry, and transport kinetics and characteristics of the exposed
individual (/'. e., water use activity, location, breathing rates)
Summary Features - MA VRIQ
Environmental media Ground or surface water
Pollutants Volatile organic compounds
Time scale Short-term to long-term (~ 1 hour to continuous)
Stochastic Yes (but, only for parts of the model)
Variability Yes (must be entered repetitively)
Uncertainty No
B.2.3 CONTAM (various versions)
The National Institute of Standards and Technology (NIST) has over the past several
years developed a series of public domain computer programs for calculating air flow and
pollutant dispersal in multi-zone buildings, including CONTAM86, CONTAM87, and
CONTAM94 These programs take a multi-zone network approach to airflow analysis Airflow
paths include doorways, small cracks in the building envelope, and a simple model of the air
handling system CONTAM94, the most recent version of CONTAM, works on an Intels-based
PC in the DOS environment A graphical interface is used to create and edit building
descriptions Future versions of this program are expected to include the capability for carrying
out exposure assessments
Summary Features - CONTAM
Environmental media User-specified indoor sources
Pollutants Generic
Time scale Short term to long-term (~ 1 hour to continuous)
Stochastic No
Variability: No
Uncertainty. No
B.2.4 AMEM (ADL Migration Exposure Model)
The ADL Migration Exposure Model was developed by EPA's Office of Pollution
Prevention and Toxics (OPPT) to estimate the migration of chemicals from polymeric materials
such as television cabinets, water pipes, curtain backings, plastic toys, or other products
containing polymers in home environments where these chemicals could become sources of
indoor air pollution or contaminate potable water. Once the fraction of chemical that can migrate
NOVEMBER 1999 frl TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
from a product is estimated, external models can be used to estimate the exposures and risk to
people from contaminated indoor air or water The AMEM provides estimates for screening-level
assessments when data are not available The goal of the model is to identify concentrations that
result in possible health concerns to justify further emission testing of the product for polymeric
materials
Summary Features - AMEM
Environmental media Indoor environment
Pollutants Chemicals emitted by polmeric materials
Time scale Short term
Stochastic No
Variability No
Uncertainty No
NOVEMBER 1999 B-6 TRTM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.3 HUMAN INHALATION EXPOSURE MODELS
Most human exposure models have the capability to track humans, either as individuals or
in groups (i.e., cohorts1) through their daily routines. The tracking process includes knowing
where the person is, what they are doing (i.e., their activities, including knowledge of the physical
demand that the person is exerting during an activity), and the concentration of the pollutants that
they come into contact with as they move about.
With the knowledge that outdoor air pollutants are able to penetrate into the interior of
buildings and that many air pollutants are emitted by both outdoor and indoor sources, a great
deal of work has focused on combining the features of both indoor and outdoor exposure models.
These models differ in their approach, but all of them estimate exposures to outdoor pollutants
that penetrate into buildings The building type of greatest concern for estimating indoor
exposures is probably residential buildings A national study on human activities found that on
average, U S citizens spend 69 percent of their time indoors at home (U.S. EPA 1996a) This
percentage can be higher for some subsets of the population, for example, the very young or the
elderly Therefore, it is important to be able to model the pollutants of concern to human health
that can penetrate a building's structure. The models in this section have, through various
techniques, attempted to develop an integrated assessment of the exposure mechanisms associated
with airborne pollutants from both indoor and outdoor sources Descriptions of the models are
provided below
B.3.1 NEM and pNEM [The (probabilistic) National Ambient Air Quality Standards
Exposure Models]
The EPA has used the NEM modeling methodology since the late 1970s when three
pollutant-specific versions of the NAAQS Exposure Model (NEM) were developed for ozone
(O3), carbon monoxide (CO), and particulate matter The early versions of the NEM were
referred to as "deterministic" as they did not attempt to model the random processes of people's
activities as part of the exposure simulation. The models simulate the movements of specific
subgroups or cohorts within a population through zones of varying air quality Each zone is
typically defined by a geographic location and a microenvironment The movements of each
cohort are determined by the use of activity diary data specific to the demographic characteristics
of the cohort The activity data are also specific to day of the week, season, and temperature
Depending on the application, cohort movements may account for trips to work places or to
schools
From its inception,NEM was designed to treat human exposure to airborne pollutants as a
time series of a joint set of human activities occurring in a particular microenvironment and air
quality (as measured by the concentration of a pollutant) in that same microenvironment.
Maintaining the time series allows estimates for alternative exposure and dose metrics to be
developed, a capability that has proven invaluable (McCurdy 1997).
1 A cohort is comprised of persons with similar demographic characteristics. A cohort is defined by a
specific combination of home district (where they reside), work district (their place of employment), and
demographic variables (e.g., gender, age)
NOVEMBER 1999 B-7 TRJM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
In 1988, the Monte Carlo technique for randomly selecting important variables was
incorporated into the model simulations. These models were referred to as "probabilistic" and
hence are known as probabilistic NEM, or pNEM The models are pollutant specific. To identify
one version of the model from another, the chemical symbol is appended to the pNEM-acronym;
hence, the model for ozone is pNEM/O3 and for carbon monoxide is pNEM/CO These two
models are the most commonly used pNEM models today Developing the probabilistic versions
of the model was an important step because it meant that the entire distribution of available data
for each variable in a model could now be used. This makes it easier to address variability and
uncertainty regarding the variables that were in the model (McCurdy 1997).
The first pNEM developed was that for ozone, or pNEM/O3. The early pNEM/O3 used a
regression-based relationship to estimate indoor concentrations of ozone using concentrations
measured outdoors. Then, in 1991, a new version of pNEM applicable to carbon monoxide
(pNEM/CO) was developed (Johnson et al 1992b). This model was the first to use a mass
balance model to estimate indoor pollutant concentrations The mass balance model is based on
the generalized mass balance model presented by Nagda et al (1987). In general terms, the mass
balance model can be described as
The change in indoor pollution concentration =
the pollutant entering from outside
- the indoor generation of pollutant
pollutant leaving the indoor microenvironment
removal of the pollutant by an air cleaning device
decay of the pollutant indoors.
Another version of the pNEM/O3 soon followed which also used a mass balance model to
estimate indoor ozone concentrations Several other refinements were included in this new
version of the pNEM/O3 Some of these include the use of more recent census data for
determining demographic information, a new commuting algorithm, and an increase in the number
of fixed-site monitors able to represent each urban area.
Early in 1994, a special version of pNEM/O3 applicable to outdoor workers was
developed and used to estimate ozone exposures for outdoor workers in several cities in the U S
(Johnson et al 1996c) In a follow-up effort, another version of pNEM/O3 specific to children
who were active in outdoor activities was developed (Johnson et al. 1996c).
More recently, enhancements have been completed on the pNEM/CO The latest version
of the model is pNEM/CO (Version 2 0) Improvements to this model include the use of an
expanded human activity database. This database is called the Comprehensive Human Activity
Database (CHAD) (McCurdy 1999) The CHAD is comprised of over 17,000 person-days of
activity pattern data. The data have been collected and organized from eight human activity
pattern surveys The CHAD contains the sequential patterns of activities for each individual,
which is particularly important to estimating the dose profile for CO by the model. Another
enhancement to pNEM/CO (Version 2.0) is the inclusion of a special commuting database
developed by the Bureau the model for creating an "origin-destination" table to indicate the
patterns of commuting trips made by working cohorts among the defined exposure districts
NOVEMBER 1999 B^8 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
(Johnson et al 1999)
Improvements have been made to both the algorithms and inputs to the mass balance
model in pNEM/CO (ver. 2 0). The pNEM/CO methodology includes a mass balance model,
which is used to estimate CO concentrations when a cohort is assigned to an indoor or motor
vehicle microenvironment The mass balance model is based on the generalized mass balance
model presented by Nagda et al. (1987) As originally proposed, this model assumed that
pollutant concentration decays indoors at a constant rate. For use in pNEM/CO, the Nagda
model was revised to incorporate an alternative assumption that the indoor decay rate is
proportional to the indoor concentration (Johnson et al 1999). This alternative assumption is
believed to more closely model the actual decay rate that takes place indoors. In addition, new
databases and improved algorithms have been included for determining air exchange rates, the
probability of gas stove use, gas stove-burner emission rates, pilot light emission rates, and
residential volumes used in the mass balance model (Johnson et al. 1999)
The pNEM/O3 and pNEM/CO are part of a small group of exposure models in which
attempts have been made to evaluate their results using personal exposure monitoring data.
Johnson et al (1996c) describes initial efforts to evaluate the pNEM/O3 In this effort, pNEM/O3
exposure estimates for Houston, Texas were compared with personal exposure monitoring data
collected in 1981 during the Houston Asthmatic Study (HAS) (Stock et al 1985). A special
version of the pNEM/O3 was created, which corresponded to the data collection criteria for the
HAS Results were compared for distributions of both one-hour ozone exposure estimates and
one-hour daily maximum ozone exposure estimates. In general, the results suggested that the
model overpredicted the HAS exposures in the range below 70 ppb and underpredicted exposures
above 70 ppb Developers of the pNEM/O3 believed that the exposure estimates of the model
were particularly sensitive to the distribution of ozone decay rates used in the model's mass
balance algorithm (Johnson et al 1996c)
During early model development, the execution of the pNEM series of models was
conducted only on an EPA mainframe computer because of the large input and output data files
required to run the model However, in the summer of 1999, pNEM/CO (Version 2.0) was
migrated (that is transferred) to run on a PC. OAQPS and ORD's NERL are continuing to
support efforts to improve the efficiency of pNEM on a PC and to provide documentation and
user's guides for the PC version The documentation and code for the PC version of pNEM/CO
should be available for public release in 2000.
The pNEM/CO (Version 1) was evaluated using CO exposure data collected during the
Denver Personal Exposure Monitoring Study conducted during the winter of 1982/83 (Akland et
al 1985). Researchers analyzed the Denver data to determine the one-hour daily maximum and
the 8-hour daily maximum CO exposures associated with each person-day of data. Then, the
pNEM/CO was run to simulate the conditions of the Denver Personal Exposure Monitoring
Study The exposure estimates from this application were tabulated according to the classification
of each cohort with respect to the type of cooking fuel used (i.e., natural gas or other). The
researchers found relatively good agreement between the observed and estimated distributions for
the one-hour daily maximum analyses, except for the values above the 99th percentile. They did
not find as good agreement between the observed and estimated distributions for the eight-hour
NOVEMBER 1999 ~B^9 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
daily maximum exposures The researchers reported that in each case, the distribution obtained
from pNEM/CO overestimated the exposure values at low exposures and underestimated the
exposure values at high exposures The researchers point out that the estimated and observed
distributions for the eight-hour daily maximum exposures agreed most closely in the range of CO
concentrations between 5 and 12 ppm (Johnson et al. 1992b)
The pNEM/CO (Version 1) was evaluated similarly to the Johnson et al (1992b)
evaluation effort but with the additional use of the Kolmogorov-Smirnov test statistic to compare
the observed and simulated cumulative frequency distributions for the one-hour daily maximum
exposure (1DME) and the eight-hour daily maximum moving average exposure (8DME) (Law et
al 1997) A similar effect to that seen in the evaluation of the pNEM/O3 occurred for pNEM/CO
For 1DME, the pNEM/CO exposure estimates agreed most closely with observed exposures
within the middle of the distribution, that is. in the range of approximately 6 to 13 ppm
However, the model overestimated values at low exposures (i.e., less than 6 ppm) and
underestimated values at high exposures (i.e., greater than 13 ppm) For 8DME, the estimated
exposures agreed well with observed exposures in the range of CO concentrations between about
5 5 and 7 ppm However, the model overestimated values below 5.5 ppm and underestimated
values above 7 ppm (Law et al 1997)
B.3.2 HAPEM (The Hazardous Air Pollutant Exposure Model)
In 1985, the EPA's Office of Mobile Sources (OMS) developed a model for estimating
human exposure to nonreactive pollutants emitted by mobile sources This model is similar to the
pNEM in that it simulated the movements of population groups between home and work locations
and through various microenvironments However, they differed in the temporal resolution used
for expressing the exposure estimates The pNEM provided hourly exposure estimates which
could be averaged over longer time periods, whereas the HAPEM provided annual average
exposure estimates The HAPEM included a facility for estimating cancer incidence through the
use of risk factors developed by the EPA, but the pNEM does not include this capability
Then, in 1991, OMS extended this modeling methodology to estimate annual average
carbon monoxide (CO) exposures in urban and rural areas under specified control scenarios The
model was now called the Hazardous Air Pollutant Exposure Model for Mobile Sources
(HAPEM-MS) The annual average CO exposures could be used to estimate annual average
exposures to various hazardous air pollutants associated with mobile sources In each case, it was
necessary to assume that the annual average exposure to a particular hazardous air pollutant was
linearly proportional to the annual average CO exposure. The model was executed for specified
urban areas that had ambient fixed- te CO monitors.
Shortly after, under the direction of EPA's Office of Research and Development (ORD),
an enhanced version of the HAPEM-MS was developed. This model was labeled the HAPEM-
MS2. It sub-divided the annual exposures by calendar quarter (i.e., 3-month periods) to better
estimate exposures to mobile sources as a consequence of outdoor air temperature. The
HAPEM-MS2 also increased the number of microenvironments to 37, increased the number of
NOVEMBER 1999 B-10 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
demographic groups2 to 23, and increased the size of the activity pattern database (Johnson et al.
1993a)
In 1996, the EPA's ORD further enhanced the HAPEM by creating another generation of
the model called the HAPEM-MS3. The enhancements included adding the ability to customize
the demographic groups, updating the census data by using the 1990 census, and developing an
algorithm for estimating ambient impacts in residences with attached garages (Palma et al. 1996).
Until the spring of 1998, execution of the HAPEM-MS3 operated only on an EPA
mainframe computer. During early model development, this limitation was necessary as the
model requires large data files for storage and large internal arrays for calculation Then, by
1998, with advances in computing technology, it became possible to have the HAPEM-MS3
executed on a "workstation " To this end, in the spring of 1998, the HAPEM-MS3 was migrated
(that is transferred) to the UNIX operating system on a workstation. During the migration,
further enhancements to the model were made, including a new time-activity database derived
from the CHAD, a new air quality program that automatically selects sites, and a more efficient
implementation of the commuting algorithm
Immediately after the release of the UNIX-version of the HAPEM-MS3, the ORD again
made substantial improvements to the model The newest model had two distinct improvements
over the 1998 UNIX-version. First, the areal extent of the model was expanded to include the
entire contiguous United States at the census tract-level In order to make this possible, the
second innovation to the model was the facility to use modeled air quality data as well as AIRS
data With this improvement, the model for the first time was able to directly estimate exposures
to hazardous air pollutants, hence, the model was renamed again by dropping the mobile source (-
MS) acronym This latest version of the model, called the HAPEM4, has other enhancements as
well These include broader flexibility in defining the study area (this can range from a census
tract up to the entire contiguous U S ), an updated database of temperatures, an updated
commuting algorithm, population data for all census tracts in the country, and the ability to
change internal modeling parameters such as the number of microenvironments and the
demographic group designations
B.3.3 HAPEM-PS (The Hazardous Air Pollutant Exposure Model for Point Sources)
The Hazardous Air Pollutant Exposure Model for Point Sources (HAPEM-PS) was
initially developed for OAQPS. In its original form, HAPEM-PS was intended to be applied to
factories, refineries, and other stationary point emission sources. The HAPEM-PS requires an air
quality indicator (e.g., annual mean concentration) for each point in a receptor grid surrounding
the point source under evaluation. Receptor air quality values are typically determined through
the use of emissions data and a dispersion model. The HAPEM-PS has not had the same
extensive enhancements that the HAPEM-MS has had since the early 1990s
2 A demographic group is defined by specific demographic characteristics taken from the census. For
example, in HAPEM demographic groups are typically defined by gender, age, race, and working status (i.e.,
either working or non-working).
NOVEMBER 1999 &Ti TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Like the HAPEM-MS, HAPEM-PS defines exposures for sets of cohorts However, the
population of concern in HAPEM-PS is usually defined as all persons residing within a specified
distance from a particular emission source The pollutant concentrations at the receptors are
typically estimated by the ISCLT dispersion model. Input data for the ISCLT model include local
meteorological data and an estimate of the pollutant emissions from the source. In a typical
HAPEM-PS application, the ISCLT model is used to estimate the annual average pollutant
concentrations at the centroid of each of the census units used to define the home and work
districts and at regularly-spaced receptors along the emission source property line. The HAPEM-
PS output provides a histogram of the total number of people exposed at pollutant concentration
level intervals. The HAPEM-PS output also provides the annual cancer incidence by home
district, the home district population, and the cancer incidence per million individuals for each
home district Finally, the output includes the value and cohort of the maximum exposure and the
values and home districts of the maximum lifetime cancer incidence and incidence rate (Johnson et
al 1993b)
B.3.4 AirPEx (Air Pollution Exposure Model)
The AirPEx was developed at the National Institute of Public Health and the Environment
in the Netherlands as a tool for analyzing the inhalation exposures of the Dutch population to air
pollutants The model was designed to assess and evaluate the time- and space-dependency of
inhalation exposures of humans It can be used to evaluate individuals, as well as populations and
subpopulations
The AirPEx estimates personal exposure for one-hour time intervals The exposure
parameters calculated include (1) the potential exposure concentrations (the air concentration as a
function of time and space (i.e., microenvironments)), (2) the actual exposure concentration (the
concentration that a person moving through the microenvironment experiences as a function of
time, (3) the intake rate (the rate at which a pollutant enters the respiratory tract per unit time, (4)
the standardized intake rate (the intake rate standardized to the target organ (e.g., lung, body
mass), (5) the frequency and time fraction that a person is in contact with concentrations above a
certain threshold value, (6) the critical intake (the excess intake at exposure concentrations above
critical concentrations) (Freijer et al 1997) Averages for each of these variables can be obtained
for an exposure period by integrating them over the whole period and dividing by the time span of
the period
Population exposures are approximated by repeating individual calculations for a large
sample of individuals taken randomly from the whole population. The distribution of the
calculated individual average exposures are approximated for the whole population by the
probability density function. Analysis of the distribution in terms of percentiles yields information
on the median, and extremes in the exposure levels are quantified by the 10th and 90th percentiles
(Freijer et al 1997).
The model itself consists of three modules assembled in the Windows* environment. The
program retrieves data from two databases and uses numerous compound specific parameters.
Default values for benzene, B(a)P, ozone, and PM are included in the program. Users can
NOVEMBER 1999 B-12 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
override the default values and supply their own values for these four compounds, as well as for
other compounds The main module calculates individual exposure measures from time-series of
air quality data and human activity patterns. Exposures are estimated in 15-minute discrete time-
steps for various microenvironments. The AirPEx currently uses a database containing 4,985
daily activity patterns with 15-minute time resolution for the population in the Netherlands. Time
series of air quality are supplied with one-hr resolution. A second module selects records from
the activity pattern database. It estimates population exposures by repeating individual exposure
calculations for all selected activity patterns and then combining this information to construct
normalized frequency distributions A third module displays the results of the exposure
calculations and analyzes the distributions by percentiles. An important feature is the ability to
analyze the socioeconomic characteristics of the individuals having the highest exposures to
enable identification of high risk groups
B.3.5 HEM (Human Exposure Model)
In 1980, the EPA's OAQPS developed the Human Exposure Model (HEM). The model
was designed to screen point sources of air pollutant emissions efficiently, ranking the sources
according to their potential cancer risks Then, in 1990, an updated version (HEM-II) that had
additional modeling capabilities needed to address issues related to the analysis of toxic air
pollutants was released The HEM-II was intended for use in evaluating potential human
exposure and risks from sources of toxic air pollutants (U S EPA 1991) HEM-II retained the
capability of screening point sources for a single pollutant in order to rank sources according to
cancer risks The HEM-II also allows more refined analyses of individual point sources and study
of entire urban areas that include multiple point sources, multiple pollutants, area sources, and
dense population distributions.
The HEM-II uses the Industrial Source Complex Long-Term (ISCLT) Model for
estimating dispersion The HEM-II also provides the ability of moving the exposed population
into up to ten microenvironments These may include indoors at home, indoors at work, in
transit, and movement out of the study area For each application, parameters can be defined for
indoor-outdoor concentration ratios for each microenvironment, the percentage of the exposed
population to be assigned to the microenvironment, and the amount of time, on an annual basis,
estimated to be spent in each microenvironment. New to this revised version of the model is the
quantification of several key uncertainties Using the Monte Carlo technique, six input variables
can be described by distributions They are the unit risk factor, the emission rate,
microenvironmental concentrations, the time spent in a microenvironment, years spent in current
residence, and the variability in concentrations predicted at the receptors A choice of several
statistical distributions can be selected for each input variable.
The HEM-II contains a limited STability ARray (STAR) database within the model.
Complex emissions inventories can also be modeled This includes modeling area sources (e.g.,
mobile sources, residential heating) simultaneously with point sources. A choice of grid systems,
including a Cartesian grid that will accommodate areas with high population density and
numerous air pollution sources, is offered for calculating exposures. Population growth can be
simulated, either from the base year of the population database to the current year or to a future
NOVEMBER 1999 1^13 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
year The model allows the user to account for differences between microenvironments (i.e.,
indoor and outdoor concentrations). Census coverages are for the entire U.S at the block group
level The results of the model's output can be shown graphically.
B.3.6 SHAPE (Simulation of Human Activities and Pollutant Exposure)
EPA's ORD developed the Simulation of Human Activities and Pollutant Exposure
(SHAPE) (Ott et al 1988) SHAPE generates carbon monoxide inhalation exposure profiles for
different human subgroups. It considers exposure to carbon monoxide in air through the
inhalation pathway only The model has two major input components. (1) human location
patterns, and (2) microenvironmental carbon monoxide concentrations. It matches available
location/activity data with environmental concentration data to obtain exposure profiles for
24-hour periods Exposure concentrations are obtained by applying a superposition principle to
contributions from the ambient and different microenvironments In addition to the limited
evaluations conducted on pNEM, SHAPE is one of the only other exposure models where
attempts were made to evaluate the model's estimates using personal exposure data The EPA's
Denver/Washington, D C personal exposure database was used to test the model's predictions
against 24-hour exposure profiles for more than 1,200 persons (Ott et al. 1988).
B.3.7 BEAM (Benzene Exposure Assessment Model)
The Benzene Exposure Assessment Model's (BEAM) initial development by EPA's ORD
in the late 1980s was spurred, at least in part, as a result of benzene being listed as a hazardous air
pollutant by the Clean Air Act and because benzene is regarded as a human carcinogen (U S EPA
1990a) The model utilizes microenvironmental benzene concentration data coupled with human
activity pattern data to estimate exposure to benzene The BEAM estimates benzene inhalation
exposure profiles for different human subgroups. It considers exposure to benzene in air through
the inhalation pathway only The model has three major input components: 1) human location
patterns, 2) ambient (background) benzene concentrations, and 3) microenvironmental benzene
concentrations It matches available location/activity data with environmental concentration data
to obtain exposure profiles for 24-hour periods. Exposure concentrations are obtained by
applying a superposition principle to contributions from the ambient and different
microenvironments Inhalation dose is then obtained by applying inhalation rate to exposure
concentrations. The BEAM was patterned after the Simulation of Human Air Pollution Exposure
(SHAPE) model.
The Total Exposure Assessment Methodology (TEAM) studies conducted between 1979
and 1988 by the U.S. EPA have been cited as providing the impetus for developing a human
exposure assessment model for benzene The TEAM studies strongly indicated that human
exposure to certain classes of volatile organic compounds (VOC), including benzene, occurs
primarily within the confines of very restrictive microenvironments, primarily indoors, but
outdoors as well The TEAM studies also suggested that the traditional method of estimating
exposures to benzene (as well as most VOCs) did not adequately account for the contribution of
benzene from small, nearby sources. Therefore, the developers of the BEAM endeavored to
NOVEMBER 1999 B-14 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
develop an exposure simulation model for benzene that utilized microenvironmental benzene
concentrations data coupled with human activity pattern data to estimate human exposure to
benzene (U S EPA 1990a, U.S EPA 1993 a) The BEAM is one of the models included in
EPA's THERdbASE (see Section B 7.2).
B.3.8 pHAP (probabilistic Hazardous Air Pollutant exposure model)
The probabilistic Hazardous Air Pollutant (pHAP) exposure model was developed by the
EPA to estimate exposures to HAPs for the population residing in a specified study area The
mode) uses census data, ambient air quality data, meteorological data, and human activity pattern
data to simulate exposures to the population from several different HAPs The original version of
the pHAP model was developed for a mainframe computing system. The mainframe version was
developed and used to estimate benzene exposures to residents of a study area in Phoenix,
Arizona for the year 1990 Subsequent to that development, a PC-version of the model was
developed and tested using the same study area The PC-version of the model is called, pHAP-
PC The pHAP-PC model utilizes a Graphical User Interface (GUI) to provide a Windows^-like
environment (Panguluri et al 1998)
B.3.9 CPIEM (California Population Indoor Exposure Model)
The California Population Indoor Exposure Model was developed for the California Air
Resources Board's (ARB) Indoor Program to evaluate indoor exposures for the general
California population as well as certain subgroups such as individuals who may be highly sensitive
to indoor air pollutants The ARB required a model which could estimate the average and peak
indoor exposures for the population and sensitive subgroups The CPIEM combines indoor air
concentration distributions with Californians' location and activity information to produce
exposure and dose distributions for different types of indoor environments This task is achieved
through a Monte Carlo simulation whereby a number of location/activity profiles that were
collected in ARB studies are combined with airborne pollutant concentrations for specific types of
microenvironments (e.g., residences, office buildings)
Concentration distributions for many pollutants and microenvironments are included in the
CPIEM database. However, for pollutants and microenvironments not included in the database,
the CPIEM presents two alternatives The first is to estimate indoor air concentration
distributions based on distributional information for mass balance parameters described below.
The second is for the user to directly specify concentration distributions. The concentration
values for a particular environment are then sampled from the distributions.
The simulation of indoor concentrations accounts for various types of indoor sources as
well as outdoor concentrations, air exchange rates, and losses to indoor sinks The concentration
component (called a module) of the model uses a mass balance equation, based on the principle of
conservation of mass, to estimate concentration distributions for specific types of indoor
environments such as residences, offices, and schools. This module samples values from user-
specified distributions for parameters such as emission rates for indoor sources, building volumes,
NOVEMBER 1999 B-15 TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
outdoor air concentrations, and indoor-outdoor air exchange rates, which are used as inputs to
the mass balance equation
Multiplication of the concentration values by breathing rates determined from the
location/activity profiles and pulmonary ventilation data yields an estimate of the potential inhaled
dose distribution for each modeled environment. The model then aggregates the environment-
specific exposure and dose estimates to develop distributions of "total indoor air" exposures and
doses That is, the portion of the total (24 hour) exposure/dose associated with time spent
indoors Because outdoors are included as one of the environments in the model, it is also
possible to simulate the total (both indoor and outdoor) exposure and dose distributions (CARB
1998)
NOVEMBER 1999 B-16 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.4 CONSUMER PRODUCT EXPOSURE MODELS
A few exposure models have been developed to assess potential exposures associated with
the use of household consumer products This type of exposure model is not as numerous as
those found in the previous three sections because these models deal solely with pollutants
emitted by the consumer products The emissions from these products occurs primarily indoors
where relatively few regulations exist for controlling toxic emissions However, unlike the
exposure models in the previous section that estimated exposures exclusively by the inhalation
route, two of the models in this section can also estimate exposures by ingestion and/or dermal
contact
B.4.1 CONSEXPO (CONSumer EXPOsure Model)
The CONSumer product EXPOsure model (CONSEXPO) (Van Veen 1995) developed by
the National Institute of Public Health and the Environment (RIVM) [Netherlands] uses simple
exposure and uptake models to assess the potential health impacts of consumer products In
order to cope with the diversity in consumer products, it is based on a general model framework
that provides a general setting for widely differing exposure situations, and, secondly, it offers a
number of predefined exposure/uptake models, which the user can link to build a complete
exposure/uptake model The starting points are the inhalation, dermal, and ingestion exposure
pathways For each of these pathways, a limited number of models is available to model
exposure/uptake The program reports several important exposure variables, namely, the per
event concentration, the yearly averaged concentration, the fraction taken up, the amount taken
up during a year (per year and summed), and the uptake per kilogram body weight per day The
program also allows for stochastic parameters, in order to propagate the effects of variable and/or
uncertain parameters to the final exposure/uptake estimates If one uses the stochastic
parameters, the resultant distributions can be displayed and studied
B.4.2 SCIES (Screening Consumer Inhalation Exposure Software)
The Screening Consumer Inhalation Exposure Software (SCIES) was developed to assist
the Economics, Exposure, and Technology Division of U.S EPA's Office of Pollution Prevention
and Toxics in performing screening-level assessments of the potential dose rates resulting from
inhalation of new and existing chemicals in consumer products. The model calculates
screening-level estimates of average individual inhalation potential dose rates to components of
consumer products that can be classified into 11 different product categories. The model
estimates potential dose rates for both actively-exposed users of the product and
passively-exposed non-users. Default values are suggested for each parameter required to run the
model for each product category. These values are based on exposure scenarios, volatility
classifications, and residence occupancy patterns. The model combines results of an effort to
measure ventilation flows within residences with a 2-zone mass balance model to allow estimation
of potential dose rates to both consumer product users and non-users.
B.4.3 DERMAL
NOVEMBER 1999 B-17 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
DERMAL was developed to assist the Economics, Exposure, and Technology Division of
U S EPA's Office of Pollution Prevention and Toxics in performing screening-level assessments
of the potential dose rates resulting from dermal contact with consumer products containing new
and existing chemicals in consumer products. The model calculates screening-level estimates of
annual individual dermal potential dose rates to components of 16 consumer product categories.
Exposures are calculated based on the weight fraction of the chemical of interest in the consumer
product and assuming deposition of a film of liquid on to the dermal surface from contact with the
product Conservative default values are provided for most of the input parameters required to
run the model for each of the 16 consumer product categories
B.4.4 MCCEM (Multi-Chamber Concentration and Exposure Model)
The MCCEM is an interactive model developed for the EPA's OPPT and updated for the
EPA's ORD It allows users to model indoor air contamination for use in assessing potential
inhalation exposures caused by consumer products The objective of MCCEM is to allow users
to be able to assess the risk from exposure to pollutants emitted by consumer products The
model uses a spreadsheet format to estimate indoor concentrations for, and individual exposures
to, chemicals released from products in residences Concentrations can be modeled in as many as
four zones within a house. The model can supply air exchange rates and interzonal airflows for
different types of residences. Time-varying emission rates can be input for a pollutant in each
zone of the residence, for outdoor concentrations, and for the zone where an individual is located
In this way. the model develops a time-series exposure profile for the individual
The MCCEM allows the user to explore the sensitivity of the model results to changes in
one or more of the input parameters The parameters that can be modified in the sensitivity
analysis include the infiltration rate, the source rate, the decay rate, and the outdoor
concentration
NOVEMBER 1999 B-18 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.5 DIETARY EXPOSURE MODELS
B.5.1 DEPM (Dietary Exposure Potential Model)
The Dietary Exposure Model (DEPM) is a model and database system developed by U.S.
EPA's ORD to correlate food information in a format for dietary exposure modeling. Currently,
the database system includes information from several national, government-sponsored surveys
and monitoring programs In the model, food consumption is based on 11 food groups,
containing approximately 800 core food types, established from over 6500 common food items.
A unique feature of the DEPM is the use of recipes, developed for exposure analysis, that link
consumption survey data to the pollutant residue information. The summary databases are
aggregated in a fashion to allow the analyst selection of demographic factors, such as age/sex
groups, geographical regions, ethnic groups and economic status The model was developed for
personal computers with the data files designed in dBASE IV with FoxPro for Windows
applications programs for queries and reporting
NOVEMBER 1999
B-19
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.6 MULTIMEDIA EXPOSURE MODELS
This section presents a review of a number of multimedia models that address exposure
links among multiple ambient environmental media and multiple exposure media
B.6.1 The Exposure Commitment Method
One of the earliest approaches for systematically assessing multipathway exposures to
environmental pollutants is termed the Exposure Commitment Method, developed by Bennett
(1981) The basic objective of this approach is to calculate exposure commitments (i.e., pollutant
concentration in human tissue), which are calculated from transfer factors that are estimated as
the ratios of the steady-state concentrations of a pollutant in adjoining compartments of an
exposure pathway An exposure commitment is determined by multiplying the transfer factors
associated with the adjoining compartments of a given pathway of exposure, for example, air to
plants to livestock to diet This method has been applied to organic chemicals and metals The
published applications of the exposure commitment methodology depend on measured
concentrations of the substances in different compartments to estimate transfer factors The
retrospective nature of this approach limits its usefulness for predicting exposures to chemicals for
which there is little or no monitoring data available
B.6.2 Layton et al. (1992) Indoor/Outdoor Air/Soil Transport Model
In recent years various researchers have begun to model algorithms that address the
movement of fine and coarse particles in the indoor environment by processes such as
resuspension, deposition, and soil tracking (see Raunemaa et al 1989, Nazaroffand Cass 1989,
Allott et al 1992) Nevertheless, none of these algorithms provide an integrated simulation of
major transport processes and indoor/outdoor relationships for toxic substances in air, water and
soil In order to estimate concentrations of pollutants in the media identified here, one needs an
indoor transport model that simulates (1) the movement of pollutants from the outdoor
environment (air and soils) to the indoor environment and (2) the resulting concentrations in
indoor media (air and floor dusts) resulting from both outdoor and indoor sources of the target
pollutants
Layton et al (1992) have developed for particle bound radio nuclides a pollutant transport
model that accounts for (1) the movement of pollutants from the outdoor environment (air and
soils) to the indoor environment and (2) the resulting levels in human contact media G/g/m2 of
floor dust and ,ug/m3 in air) derived from both outdoor and indoor sources This model can be
linked directly to the multimedia model to evaluate indoor/outdoor relationships. This model was
used to evaluate the relative importance of various kinds of human factors in mediating human
contacts with substances in air and dust The loading of soil/dust on floor surfaces and the
resuspension of particles from floors, for example, should increase as the number of household
occupants increases. Increased loading of soil/dust on floors should in turn lead to more
cleaning/vacuuming that redistributes pollutants onto contact surfaces throughout a house. For
example, the tracking-in process, vacuuming, and particle penetration of a building shell tend to
produce smaller sized particles in the indoor environment, making them more bioavailable.
NOVEMBER 1999 B-20 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.6.3 CalTOX (California Total Exposure Model for Hazardous Waste Sites)
The Department of Toxic Substances "Control (DTSC) within the California
Environmental Protection Agency, has the responsibility for managing the State's
hazardous-waste program As part of this program, the DTSC funded the development of the
CalTOX program CalTOX has been developed as a set of spreadsheet models and spreadsheet
data sets to assist assessing human exposures and defining soil clean-up levels at uncontrolled
hazardous wastes sites (McKone 1993a, b, c) More recently, CalTOX has been modified for use
in establishing waste classification for landfills and hazardous waste facilities in California.
CalTOX addresses contaminated soils and the contamination of adjacent air, surface water,
sediments, and ground water The modeling components of CalTOX include a multimedia
transport and transformation model, exposure scenario models, and add-ins to quantify
uncertainty and variability The multimedia transport and transformation model is a dynamic
model that can be used to assess time-varying concentrations of pollutants introduced initially to
soil layers or for pollutants released continuously to air, soil, or water This model assists the user
in examining how chemical and landscape properties impact both the ultimate route and quantity
of human contact Multimedia, multiple pathway exposure models are used in CalTOX to
estimate average daily doses within a human population. The exposure models encompass
twenty-three exposure pathways. The exposure assessment process consists of relating pollutant
concentrations in the multimedia model compartments to pollutant concentrations in the media
with which a human population has contact (personal air, tap water, foods, household dusts soils)
B.6.4 MMSOILS: Multimedia Contaminant Fate, Transport, and Exposure Model
MMSOILS was developed by EPA's ORD to estimate the human exposure and health risk
associated with releases of contamination from hazardous waste sites (U S. EPA 1992d) It is a
multimedia model addressing the transport of a chemical in ground water, surface water, soil
erosion, the atmosphere, and accumulation in food The human exposure pathways considered in
the methodology include soil ingestion, air inhalation of volatiles and particulates, dermal contact,
ingestion of drinking water, consumption offish, consumption of plants grown on contaminated
soil, and consumption of animals grazing on contaminated pasture For multimedia exposures, the
methodology provides estimates of human exposure through individual pathways and combined
exposure through all pathways considered. The risk associated with the total exposure dose is
calculated based on chemical-specific toxicity data The intended use of MMSOILS is for
screening and relative comparison of different waste sites, remediation activities, and hazard
evaluation The methodology can be used to provide an estimate of health risks for a specific
site, but the uncertainty of the estimated risk may be quite large (depending on the site
characteristics and available data) and this uncertainty must be considered in any decision-making
process.
B.6.5 RESRAD (RESidual RADiation)
The Residential Radiation (RESRAD) model was developed by Argonne National
Laboratory to evaluate residual concentrations of radio nuclides in soil, concentrations of airborne
NOVEMBER 1999 B-21 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
radon decay products, external gamma radiation levels, surface contamination levels, and radio
nuclide concentrations in air and water and to determine radiation dose and excess lifetime cancer
risks to an on-site resident (a maximally exposed individual or a member of a critical population
group).
RESRAD was developed for the U S DOE and is accepted for use in remedial action
activities RESRAD determines site-specific residual radioactive material cleanup guidelines
based on calculations of the radiation dose to hypothetical residents or workers on the site The
nine environmental pathways considered in RESRAD are direct exposure, dust inhalation, radon,
and ingestion of plant foods, meat, milk, aquatic foods, water, and soil.
RESRAD code has been adapted to include both chemical and radiological health risks
Other recent RESRAD developments include the incorporation of uncertainty analysis and
decontamination and decommissioning analysis capabilities The development of the code is
funded by the DOE
B.6.6 USES (Unified System for the Evaluation of Substances)
The Uniform System for the Evaluation of Substances (USES) (RIVM 1994), was
developed in the Netherlands by the National Institute of Public Health and the Environment
(RIVM), Ministry of Housing, Spatial Planning and Environment (VROM), and the Ministry of
Welfare, Health, and Cultural Affairs (WVC) USES provides a single framework for comparing
the potential risks of difference chemical substances released to multiple media of the
environment It is an integrated modeling system that includes multiple environmental media and
multiple human exposure pathways. The exposure assessment in USES starts with an estimate of
substance emissions to water, soil, and air during the various life-cycle stages of a substance and
follows its subsequent distribution in the total environment The result of this type of multi-media
assessment are the Predicted Environmental Concentrations (PECs) and an estimate of the daily
intake by human receptors In general, PECs are compared to "no-effects" levels for organisms in
the environment, which are derived by extrapolating single-species toxicity tests to field
situations The estimated daily intake by humans is compared to the "no-observed-adverse-effect"
level for mammals or to the "no-effect" level for humans
B.6.7 BEADS (The Benzene Exposure and Absorbed Dose Simulation)
BEADS (Macintosh et al 1995) was developed at the Harvard School of Public Health
through support from EPA's ORD. It is a population-based, multiple exposure pathway
microenvironmental model of 24-hour average inhalation exposures and total absorbed doses of
benzene The model was developed. (1) to provide a tool for estimating the distribution of
benzene personal air concentrations and total absorbed doses for a large population, (2) to
examine the determinants of inter-individual variability of exposures and absorbed doses of
benzene, and (3) to explore the accuracy and precision of predictions of population exposures and
absorbed doses of benzene made with monitoring results from past field studies. A
two-dimensional Monte Carlo simulation approach is used in the model to estimate the
NOVEMBER 1999 " B-22 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
uncertainty about the predicted population exposure and absorbed dose distributions. A principal
advantage of this approach to uncertainty analysis is that the relative contribution of the input
variables to prediction uncertainty can be easily identified. Decisions can then be made regarding
the appropriate measures to be taken to reduce the parameter uncertainty, where the overall goal
is to minimize prediction uncertainty.
The BEADS model includes a probabilistic non-sequential (non-temporal) simulation of
time activity patterns (TAP) and an anthropometric module used to correlate exposure factors in
order to estimate absorbed dose (from inhalation, ingestion, and dermal absorption). Short term,
high concentration exposucgs are not accounted for. The multimedia exposure and absorbed dose
model underwent preliminary evaluation to estimate the benzene distribution of personal air
concentration that would be expected for a large population depending on a microenvironment
and exposure scenario in air (all via inhalation) or water (via ingestion, dermal uptake, or
inhalation) Estimated distributions of personal and microenvironmental benzene exposures
compared well with previous monitoring results (TEAM-Total Exposure Assessment
Methodology) except at the upper ends
B.6,8 DERM (Dermal Exposure Reduction Model)
The physical-stochastic model DERM, was developed by Stanford's Environmental
Engineering and Science Group to' estimate personal dermal exposure incurred via multiple
contact mechanisms as a function of time (Zartarian 1996) This is the first exposure model to
calculate dermal exposure as a function of actual human activity data An important output of
DERM is the dermal exposure profile, which plots mass of pollutant loading on the skin as a
function of time Such profiles are the basis of understanding the pathways by which dermal and
non-dietary ingestion exposure are incurred (e.g., liquid immersion, surface contact with liquids,
soil, or dust, aerosol deposition, hand-to-mouth contact) DERM is a personal, physical-
stochastic model designed to evaluate the sources of uncertainty in the calculations, to understand
the important dermal exposure contact pathways, and to help determine the best ways to control
those factors that contribute most significantly to exposure DERM has also been designed to
assess ingestion exposures for hand-to-mouth activities
B.6.9 SCREAM2 (South Coast Risk and Exposure Assessment Model, ver. 2)
The SCREAM2 (Rosenbaum et al 1994) was developed with support from California's
South Coast Air Quality Management District and provides the ability to model both inhalation
and multipathway non-inhalation exposures. A submodel, called MULTPATH, calculates
population exposures to air toxics through non-inhalation pathways. The MULTPATH submodel
includes the following pathways soil ingestion, soil dermal contact, home-grown produce
ingestion, commercial (locally-grown) produce ingestion, commercial (locally-raised) animal
product ingestion, surface drinking water (local source) ingestion, fish (local source) ingestion,
and breast milk ingestion Average daily doses of a pollutant to the population are estimated from
concentrations in each medium on the basis of age-specific ingestion rates and body weights and
local population age profiles Exposure estimates are made with respect to contamination of
NOVEMBER 1999 B^23 TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
commercial foodstuffs (produce and animal products) and surface water used for drinking and/or
fishing by assuming that they are consumed entirely and uniformly by the population of the
modeled area in proportion to the average ingestion rates for each age group. The individual
carcinogenic health risk associated with ingestion exposure over a 70-year lifetime is estimated as
the product of the dose and a cancer potency slope. MULTPATH requires site-specific
information on the locations and yields of all commercial produce and commercial animal-raising
operations. In addition, the locations and information on certain physical parameters of all water
bodies used as sources of drinking water or fishing are required
For inhalation exposures, the 24-hourly concentrations for each census block group are
estimated using a Gaussian air dispersion- and mapping modules. For each source, the model
calculates ground-level concentrations as a function of radial distance and direction from the
source for a set of receptors laid out in a radial grid pattern. The concentrations represent the
steady-state concentrations that would occur with constant emissions and meteorological
parameters
The inhalation exposure module accounts for mobility patterns of the population, indoor-
outdoor exposure concentration differences, and physical exercise levels. The module estimates
exposures for each individual census block group, aggregating exposure throughout the modeled
area for a number of subregions, called exposure districts Exposure is estimated for 12 basic
population age/occupation groups which are further subdivided into 56 subgroups, distinguished
by their hourly activity patterns For each hour, a subgroup is assigned to a geographic location
(i.e., either a home or work district), one of several different indoor or outdoor
microenvironments, and one of three physical exercise levels (low, moderate, or heavy) There
are population activity patterns defined for weekdays, Saturdays, and Sundays
In order to track the geographic locations of the population from hour to hour, the
inhalation exposure/risk module again divides the 56 subgroups into cohorts, defined on the basis
of common age-occupation groupings, and again by home and work exposure districts The
population composition and mobility data are compiled at the exposure-district level, while
concentrations are for the smaller census block-group level
SCREAM2 can use the Indoor Air Quality Model (IAQM) submodel to calculate indoor
pollutant concentrations IAQM simulates indoor air quality by means of a dynamic mass balance
equation, with a building being represented by a single compartment. Outdoor air is permitted to
leak into and out of the building, and indoor recirculation and makeup air can be supplied as
appropriate through simulation of a heating, ventilation, and air conditioning (HVAC) system In
addition, indoor sources, which are specified in terms of source strength and time profile, can be
modeled with the IAQM. Pollutant losses indoors are simulated in terms of adsorption onto
surfaces or deposition due to settling, with surface reactivity or deposition rates dependent on the
pollutant.
NOVEMBER 1999 B-24 TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.6.10 Integrated Spatial Multimedia Compartmental Model (ISMCM)
The Integrated Spatial Multimedia Compartmental Model has been under development
with the School of Engineering and Applied Science at the University of California Los Angeles
for approximately the last 15 years A newer version of the ISMCM, called MEND-TOX, is
currently undergoing evaluation at the EPA ORD's National Exposure Research Laboratory
(NERL)
The ISMCM considers all media in one integrated system It includes both spatial and
Compartmental modules to account for complex transport of pollutants through the ecosystem.
Assuming mass conservation, ISMCM is able to predict transport based on a mechanistic
description of environmental processes, including estimation of intermedia transfer factors
The ISMCM is not structured to incorporate uncertainty/variability analyses directly into
the model operation Furthermore, the links and compartments (spatial configuration) of the
ISMCM are predetermined, thereby making it less useful in a system that is to be fully integrated
B.6.11 Indirect Exposure Methodology (IEM) Model
The U.S. EPA began developing the Indirect Exposure Methodology (IEM) in the 1980s
The IEM consists of fate and transport algorithms that estimate the media concentrations resulting
from the multipathway transfer of air pollutants to soil, vegetation, and water bodies. The
algorithms in IEM are designed to predict exposures for pollutants for which indirect impacts may
be important (i.e., organic and inorganic pollutants that tend to be long-lived, bioaccumulating,
non- [or at most semi-] volatile, and more associated with soil and sediment than with water) An
interim document summarizing the IEM methodology was published in 1990 (U S EPA 1990b),
and a major addendum was issued in 1993 (U S EPA 1993b) The Agency's Office of Solid
Waste and Emergency Response (OSWER) has adapted IEM and compiled detailed information
on many of IEM's input parameters and algorithms in the Human Health Risk Assessment
Protocol for Hazardous Waste Combustion Facilities (U.S. EPA 1998e) The algorithms in IEM
are under continuous refinement, and revised documentation addressing SAB and public
comments on the 1993 Addendum is pending (U.S. EPA 1999h) This revised document no
longer uses the IEM terminology; instead, the document refers to MPE (multiple pathways of
exposure) assessment.
The IEM estimates human exposure to pollutants via several routes, including inhalation,
dermal contact, and food, water, and soil ingestion. Exposures are estimated using environmental
media concentrations, transfer factors (e.g., bioaccumulation factors) where appropriate, and
measures of human activity and exposure characteristics (e.g., consumption rates for food types)
such as those available in EPA's Exposure Factors Handbook (U.S. EPA 1997b). The IEM is
designed to estimate intakes for specific, predetermined receptor scenarios (e.g., subsistence
gardener, recreational fisher, average urban resident) that may be indicative of high-end or
average exposures to a pollutant.
NOVEMBER 1999 B-25 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Because it is designed to estimate exposure for individuals classified into specific
scenarios, IEM does not readily allow for the modeling of a distribution of exposures within a
population It is not designed to provide estimates of population exposures. In addition, IEM is
set up to model a long-term emission source, and the fate and transport component of IEM
consists of a set of linked, one-way algorithms that do not allow for tracking transformations
between different chemical species or feedback between different media. Pollutants are input to
the model as annual average air concentrations and wet and dry deposition rates for a specific
location or as areal averages for a given space (e.g., a watershed). Thus, the model cannot
provide a detailed time series estimation of media concentrations and the resulting human
exposures, and spatial variations in exposure can be approximated only through substantial site-
specific model adjustment and repeated model runs In addition, IEM is a deterministic model and
is not designed to estimate the uncertainty or variability associated with exposure estimates.
These characteristics make IEM suitable for scenario-specific, screening-level exposure
assessments and the determination of exposure routes of potential concern for a long-term
emission source but less appropriate for estimating distributions of population exposures over
time and space and across various pathways
B.7 EXPOSURE SIMULATION MODEL SYSTEMS
In addition to exposure models, there are a number of exposure modeling systems. These
are systems or libraries that can contain transport models, exposure models, data files, and the
associated software for linking these models with the various input and output files.
B.7.1 GEMS (Graphical Exposure Modeling System)
The EPA's OPPT developed the Graphical Exposure Modeling System (GEMS) to
support exposure and risk assessments by providing access to single medium and multimedia fate
and exposure models, physical and chemical properties estimation techniques, statistical analysis,
and graphics and mapping programs with related data on environments, sources, receptors, and
populations Under development since 1981, GEMS provides analysts with an interactive, easily
learned interface to various models, programs, and data needed for exposure and risk
assessments. PC-GEMS (GSC 1988) is a stand-alone version of GEMS that can be run on a
personal computer
The environmental models in GEMS are atmospheric, surface water, land unsaturated
(soil) and saturated (ground water) zones, and multimedia in nature. Methods for estimating
octanol-water partition and adsorption coefficients, bioconcentration factor, water solubility,
melting and boiling point, vapor pressure, Henry's constant, acid dissociation constant,
lake/stream volatilization rate, and atmospheric half-life are available. Data sets are related to
environmental characteristics (climate, soil, rivers, ground water, vegetation), source releases
(POTWs and industrial water discharges, Census business patterns, RCRA permit sites), and
receptors (population and household estimates for 1970, '80, '90, and '95 by small area census
district; and drinking water facility information)
NOVEMBER 1999 B-26 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
B.7.2 THERdbASE (Total Human Exposure Risk database and Advanced Simulation
Environment)
THERdbASE has been developed by EPA's ORD as a PC-based computer modeling and
database system that contains exposure and risk related information. The system provides a
framework for the construction of a suite of exposure and risk related models within the Modeling
Engine by using information available in data files within the Database Engine Data can be
viewed as a table, coded fields can be viewed as decoded fields, fields can be set to "show" or
"hide" mode, and multiple data files can be viewed at the same time. In the "advanced" mode,
user files can be edited. Data records can be queried and simple statistics (summary statistics
mean, standard deviation, minimum and maximum; percentile values at desired intervals, and
linear regression on two numerical data fields) can be performed Data can be printed, saved, or
exported New user files can be created and data can be imported Input to models is achieved
through a standardized procedure. Inputs can be provided as single values, custom distributions
(normal, lognormal), distributions based on data files present in THERdbASE, or as specific
percentile values Efficient algorithms are provided to optimally access input data, to perform the
numerical simulations, and to generate appropriate output data. Multiple model runs can be done
through a batch process Results can be output as either THERdbASE data files or as pre-set
graphs (US EPA 1998g).
Information about THERdbASE is available on the EPA's Internet website
(http://www epa gov/nerlpage/heasd/therdbase.htm) The Internet version of THERdbASE
includes the following models.
Location Patterns
Chemical Source Release - Instantaneous Emission
Chemical Source Release - Timed Application
Indoor Air (2-Zone)
Indoor Air (N-Zone)
Exposure Patterns For Chemical Agents
Benzene Exposure Assessment Model (Beam)
Source Based Exposure Scenario (Inhalation + Dermal)
Film Thickness Based Dermal Dose
PBPK Based Dermal Dose
The Internet version of THERdbASE also includes the following databases
1990 Bureau of Census Population Information
California Adult Activity Pattern Study (1987 - 88)
AT&T-sponsored National Activity Pattern Study (1985)
1992-94 National Human Activity Patterns Study (NHAPS)
Chemical Agents from Sources
Chemical Agent Properties
Air Exchange Rates
Information from EPA's TEAM (Total Exposure Assessment Methodology)
Studies
Information from EPA's NOPES (Non-Occupational Pesticides Exposure Study)
Studies
NOVEMBER 1999 B-27 TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Human Physiological Parameters
B.7.3 MEPAS (Multimedia Environmental Pollutant Assessment System)
The Multimedia Environmental Pollutant Assessment System (MEPAS) (Strenge and
Chamberlain 1995, Droppo et al. 1992) was developed by Battelle Pacific Northwest Laboratory
(PNL) for the U S Department of Energy The system was developed to rank DOE sites having
potential hazardous chemical and radioactive releases The key objective of MEPAS is to rank
sites by calculating human health risk to the population surrounding the site. MEPAS calculates
"hazard potential index" (HPI) values for a site by summing up risk factors associated with
various exposure scenarios This system has wide applicability to a range of environmental
problems using air, ground water, surface water, overland, and exposure models. MEPAS
integrates source, transport, and exposure models into a single system The algorithms in
MEPAS accommodate the following ten components 1) source terrain, 2) overland pathway, 3)
ground water (vadose and saturated zones) pathway, 4) surface water pathway, 5) atmospheric
pathway, 6) exposure routes, 7) hazard assessment (chemical carcinogens / non-carcinogens,
radio nuclides), 8) pollutant/transport and exposure scenarios, 9) a user-friendly PC shell, and 10)
a chemical database
Pollutant transport media are ground water, overland flow, surface water, and
atmosphere Human uptake occurs through ingestion (of contaminated water, soil, crops, animal
products, and aquatic foods), inhalation (of airborne pollutants), and dermal contact (with
chemicals and radio nuclides) The hydrologic media consists of the hydrologic source term,
unsaturated and saturated ground water zones, and surface water/runoff The source term can be
computed internally or specified at receptor locations or by specified flow. The source term
geometries include point, line, and area sources Limitations of the hydrologic pathway include
negligible leaching of the source by ground water, and flow in the virtual direction only
The atmospheric pathway consists of the atmospheric source term and atmospheric
transport processor Source terms consist of point sources and area sources Atmospheric
transport of pollutants utilizes an enhanced Gaussian plume model, and computes long term
exposure for a sixteen sector grid using average stack parameters Enhancements to the plume
model include deflection of wind speed to account for variability in local surface roughness, and
can consider radioactive decay depletion and first order chemical reactions. Only simple sources
can be modeled, and paniculate pollutants originate from area sources only.
MEPAS calculates an average dose over 70 years time increments for a number of user
specified receptor locations. Dose is calculated for each transported pollutant. For radioactive
pollutants the dose is expressed as the effective dose equivalent from each pollutant. MEPAS
uses the ICRP dose conversion factors to convert the rate of exposure to dose.
B.7.4 EML/IMES (Exposure Models Library / Integrated Model Evaluation System)
The Exposure Models Library (EML) was developed by the U.S. EPA's ORD and is a
collection of exposure models distributed in a CD-ROM (U.S. EPA 1996c). The purpose of this
disk is to provide a compact and efficient means to distribute exposure models, documentation,
NOVEMBER 1999 B-28 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
and the Integrated Model Evaluation System (IMES) The EML disk contains over 120 models
which may be used for exposure assessments and transport modeling The model files may
contain source and/or executable code, sample input files, and other data files, sample output files,
and in many cases, model documentation in WordPerfect8, ASCII text, or other similar formats
IMES assists in selecting appropriate models, provides literature citations on model validations,
and demonstrates model uncertainty protocols The IMES software is an MS-DOS application,
can be used on an Intel-based PC, and is capable of running on a network. Model codes and
documentation can be downloaded from the CD-ROM to a hard drive The most recent version,
which is the third edition, has an HTML interface to view model directories and Internet source
for some models
B.7.5 MENTOR (Modeling ENvironment for TOtal Risk)
The Modeling ENvironment for TOtal Risk (MENTOR) project, is being developed
through funding from EPA's National Exposure Research Laboratory (NERL) The objective of
the on-going MENTOR project is to develop, apply through case studies, and evaluate state-of-
the-art computational tools, that will support multipathway, multiscale source-to-dose studies and
exposure assessments for a wide range of environmental pollutants Particular emphasis in
MENTOR is placed on integrating methods for prognostic and diagnostic exposure/dose analyses,
by utilizing, in combination, environmental, microenvironmental, and biomonitoring information
to evaluate assumptions regarding routes and pathways of exposure
MENTOR merges the methods and tools of the comprehensive Exposure and Dose
Modeling and Analysis System (EDMAS) with those currently available in pNEM, and extends
them for application to situations that are relevant to multimedia and multipathway exposures
EDMAS is an expandable library of interlinked computation modules (Georgopoulos et al 1997).
MENTOR incorporates models, databases, and analytic tools which can
probabilistically estimate exposures (and doses) to individuals, populations, and susceptible
subpopulations as well as predict and diagnose the complex relationships between source and
dose MENTOR is designed as a multiscale modeling system, that allows following in a
mechanistically consistent manner the evolution of physicochemical phenomena over spatial scales
ranging from geographic regions to personal and residential microenvironments It provides a
consistent link with biological uptake and disposition models MENTOR is also multiscale in
time, designed to support modeling of processes in ranges from minutes to decades
MENTOR has a modular structure with an interface that offers linkage to both a
Geographic Information System (ArcView and Arclnfo) and a relational database management
system (Oracle) for "defining" an application or case study. MENTOR incorporates libraries of
environmental and biological process models, including macroenvironmental,
ecological/food-web, local multimedia, microenvironmental, activity pattern/exposure event,
biological fate and transport, and dose response modules. MENTOR will eventually provide an
extensible set of ready-to-use methodological tools, as well as linkages to relevant databases, for
performing assessments of exposure/dose for populations or specific individuals, and for a variety
of user-defined scenarios
NOVEMBER 1999 B-29 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
The MENTOR development and application effort is being pursued at the Computational
Chemodynamics Laboratory (CCL) of the Environmental and Occupational Health Sciences
Institute (EOHSI), which is a joint project of the University of Medicine and Dentistry of New
Jersey (UMDNJ) and Rutgers University
B.7.6 MODELS-3/Multimedia Integrated Modeling System (MIMS)
The U.S EPA ORD's NERL is developing Models-3 Community Multi-scale Air Quality
(CMAQ) modeling system. It is a flexible software system designed to simplify the development
and use of environmental assessment and decision support tools for a wide range of applications
from regulatory and policy analysis to understanding the interactions of atmospheric chemistry
and physics This newest generation of environmental modeling software has been under
development for the past seven years
Models-3, in combination with CMAQ, form a third generation air quality modeling and
assessment system. First generation air quality models dealt with tropospheric air quality with
simple chemistry at local scales using Gaussian plume formulation as the basis for prediction
Second generation models covered a broader range of scales (i.e., local, urban, and regional) and
pollutants, addressing each scale with a separate model and often focusing on a single pollutant
Third generation models treat multiple pollutants simultaneously up to continental scales and
incorporate feedback between chemical and meteorological components Future development is
planned for a fourth generation system which would extend linkages and process feedback to
include air, water, land, and biota to provide a more holistic approach to simulation of transport
and fate of chemical and nutrients throughout an ecosystem (U S EPA 1998f) This system,
called the Multimedia Integrated Modeling System (MIMS), is described below
Models-3 has a unique framework and science design that enables scientists and regulators
to build their own modeling systems to suit their needs The CMAQ system has capabilities for
urban to regional-scale air quality simulation of tropospheric ozone, acid deposition, visibility, and
fine particles The Models-3 framework contains components that assist the model developer
with creating, testing, and performing comparative analysis of new versions of air quality models
and enables the user to execute air quality simulation models and visualize the results The overall
goal of Models-3 is to simplify and integrate the development and use of complex environmental
models, beginning with air quality and deposition models (U.S. EPA 1998f)
MIMS will have capabilities to represent the transport and fate of nutrients and chemical
stressors over multiple scales. It will be designed to improve the environmental management
community's ability to evaluate the impact of air and water quality and watershed management
practices on stream and estuarine conditions. The system will provide a computer-based problem
solving environment for testing understanding of multimedia (atmosphere, land, water)
environmental problems, such as the movement of chemicals through the hydrologic cycle, or the
response of aquatic ecological systems to land-use change, with initial emphasis on the fish health
endpoint The design will attempt to combine the state-of-the-art in computer science, system
design, and numerical analysis (i.e. object oriented analysis and design, parallel processing,
advanced numerical libraries including analytic elements) with the latest advancements in process
NOVEMBER 1999 B-30 TRIM.ExPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
level science (process chemistry, hydrology, atmospheric and ecological science) The problem
solving environment will embrace the watershed/airshed approach to environmental management,
and build upon the latest technologies for environmental monitoring and geographic
representation
B.7.7 SHEDS (Stochastic Human Exposure and Dose Simulation) Model
The Stochastic Human Exposure and Dose Simulation (SHEDS) Model is a probabilistic,
physically-based model that simulates aggregate exposure and dose for population cohorts and
multimedia pollutants of interest It is being developed by the U S EPA ORD's NERL. At
present the model is applied to assess children's exposures to pesticides (SHEDS-Pesticides) and
population exposures to PM (SHEDS-PM) The key objectives of SHEDS are1 (1) to improve
the risk assessment process by predicting both inter-individual variability and uncertainties
associated with the upper percentiles (e.g., >90th percentile) of population exposure and dose
distributions, (2) to improve the risk management process by identifying critical exposure routes
and pathways; and (3) to provide a framework for identifying and prioritizing measurement needs
and to formulate the most appropriate hypotheses and designs for exposure studies
SHEDS-PM estimates the population distribution of PM exposure by sampling from
distributions of ambient PM concentrations and from distributions of emission strengths for
indoor sources of PM, such as cigarette smoking and cooking A steady-state mass balance
equation is used to calculate PM concentrations for the home microenvironment The physical
factors data used in the equation (e.g., air exchange rate, penetration rate, deposition rate) are
also sampled from distributions Non-residential microenvironmental concentrations are
calculated based on penetration of outdoor PM and indoor sources Additional model inputs
include demographic data for the population being modeled and human activity pattern data from
the National Human Activity Pattern Survey (NHAPS) Output from the SHEDS-PM model
includes distributions of PM exposures in various microenvironments (e.g., indoors at home, in
vehicles, outdoors) and the relative contributions of these various microenvironments to the total
exposure
The first generation of SHEDS-PM has been applied to the population of Vancouver,
Canada using spatially interpolated ambient PM10 measurements (Ozkaynak et al. 1999a, b)
Subsequent generations will focus on modeling both PM10 and PM2 5 exposure and dose in a
selected U S. city
SHEDS-Pesticides predicts children's aggregate population exposure and dose to
pesticides. It simulates individuals from the user-specified population cohort by selecting daily
sequential time/location/activity diaries from surveys contained in EPA's CHAD (e.g., the
National Human Activity Pattern Survey) For each individual, SHEDS-Pesticides constructs
daily exposure and dose time profiles for the inhalation, dietary and non-dietary ingestion, and
dermal contact exposure routes, then aggregates the dose profiles across routes A single-
compartment pharmacokinetic component has been incorporated into the first generation
SHEDS-Pesticides model to predict real-time pollutant or metabolite concentrations in the blood
compartment or eliminated urine Exposure and dose metrics of interest (e.g., peak, time-
averaged, time-integrated) are extracted from the individual's profiles, and the process is repeated
NOVEMBER 1999 I3l TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
thousands of times to obtain population distributions. This approach allows identification of the
relative importance of routes, pathways, and model inputs Two-stage Monte-Carlo sampling is
applied to predict the range and distribution of aggregate doses within the specified population
and the uncertainties associated with percentiles of interest
SHEDS-Pesticides samples, for each individual, location-specific air, dust, soil, and
surface residue levels, meal-specific food and beverage residues; exposure factors (e.g., residue-
to-skin transfer efficiency, saliva and washing removal efficiency, soil adherence, surface area
contacted), uptake factors (e.g., inhalation and dietary absorption fractions), and pharmacokinetic
rate constants (e.g., dermal absorption, gastrointestinal absorption, elimination) from user-
specified probability distributions. For each location/activity combination in the individual's
diary, SHEDS-Pesticides combines the air concentration, activity-specific inhalation rate (derived
from distributions of MET energy expenditures), and inhalation absorption fraction to estimate
real-time inhalation absorbed dose. For each eating or drinking event, SHEDS-Pesticides
combines the mass residue ingested by the ingestion absorption fraction to obtain mass in the
gastrointestinal tract, then applies a gastrointestinal absorption rate constant to estimate real-time
mass in the blood compartment If the eating event is non-dietary, the mass of residue ingested is
that on the object mouthed or the skin at that instant in time The dermal loading over time is
obtained by simulating exposures from discrete dermal contact events (i.e., contacts between the
skin surface and different objects such as smooth surfaces, textured surfaces, mouth, turf) within
each macro-activity Probabilities of skin contacts with different surfaces for a given contact
event are obtained from the contact frequency and duration information collected via videography
studies For each dermal contact event, the model combines available mass on the skin by a
dermal absorption rate constant to estimate real-time dermal absorbed dose.
To help meet the requirements of the Food quality Protection Act of 1996 (FQPA), the
initial focus of the SHEDS-Pesticides model has been residential exposures of children to
pesticides Model estimates for chlorpyrifos have been obtained for several application methods
(i.e., broadcast, crack, and crevice) and age groups (0-4 years, 5-9 years) for acute, short-term,
and chronic post-application time periods, and then weighted with available pesticide use and
frequency information to develop aggregate population estimates A paper describing the initial
SHEDS-Pesticides modeling framework, presenting the chlorpyrifos case study, and
demonstrating that the modeled estimates compare well against available published measurement
data has been submitted for publication (Zartarian et al 1999).
While the first generation SHEDS-Pesticides model was developed with a special
emphasis on characterizing critical exposure pathways and factors for residential exposures of
children to pesticides, the next generation will characterize both aggregate and cumulative dose
associated with human exposure (i.e., for both adults and children) to a variety of environmental
pollutants in addition to pesticides, including other persistent organic pollutants, metals, and air
toxics SHEDS-Pesticides will eventually be expanded to include source-to-concentration (i.e.,
fate and transport models) and more complete exposure-to-dose models (pharmacokinetic or
dosimetric models)
Each iteration of SHEDS will use the best available data to identify critical pathways of
human exposure and dose and the major uncertainties in those pathways. Model inputs and
NOVEMBER 1999 ~ B^32 TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
assumptions will continue to be reined as new measurement data become available (e.g., pesticide
usage survey data, residue and concentration distributions in space and time, residue-to-skin
transfer efficiencies, uptake data, microlevel activity data, emission source strengths, air exchange
rates, penetration rates) The model will be tested against field measurement programs for
refinement and subsequent evaluation.
Table B-l
Model Features and the Exposure Models Associated with Each Feature
Feature
Model
Remarks
Short-term
(; 1 h)
exposure events
CPIEM
Other built-in time scales include 24 hours, 12 hours (daytime), 12
hours (nighttime), and 8 hours
pNEM
Inhalation exposures can be calculated for i minute to 24 hours
depending on the pollutant
Exposures are generally for 1-hour (5-mmute duration for SO2)
However, activities can be as short as 1-mmute in duration
HAPEM-PS
Annually averaged 1-hour increments
HAPEM
Seasonally (3-month) averaged 1-hour increments
AirPEx
15-mmutes
SHAPE
Exposures are generally for 1-hour However, activities can be as
short as 1-mmute in duration
DERM
Dermal exposures are estimated over the day based on dermal
contact events ranging from seconds to minutes
SHEDS
(under development)
Inhalation exposures can be calculated for 1 minute to 12 hours,
depending on the pollutant and diary selected. Dermal and non-
dietary mgestion exposures can be as short as a 5-second duration
Diaries are used to determine mgestion events Modeled dietary
ingestion exposures for each eating and drinking event are assumed
instantaneous, but absorption is calculated on a 30-mmute time scale
BEAM
1 hour
pHAP
1 hour
Long-term
exposure events
CalTOX
Annual
SHEDS
(under development)
Daily exposure and dose profiles could be repeated over longer
durations based on multi-day activity and contact frequency
information
HEM
Annual
SCREAM2
Annual.
Exposure media
CPIEM
Indoor air (multiple microenvironments)
pNEM
Indoor air and outdoor air
HAPEM
Indoor air and outdoor air
NOVEMBER 1999
B-33
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Feature
Model
Remarks
AirPEx
Indoor air and outdoor air
HEM
Indoor air and outdoor air
SHAPE
Indoor air and outdoor air
BEAM
Indoor air and outdoor air
pHAP
Indoor air and outdoor air
CONSEXPO
Indoor air, personal air, and contact with surfaces (both oral and
dermal)
CalTOX
Indoor air, outdoor air, soil, house dust, tap water, food from home
gardens and locally-produced fruits, vegetables, grams, fish, meat,
milk, eggs, and dairy products
SHEDS
(under development)
Indoor air, outdoor air, soil, house dust, surface residues (indoor and
lawn), hand and object residues, tap water, food and beverage
residues
MMSOILS
Indoor air, outdoor air, soil, house dust, tap water, food from home
gardens and locally-produced fruits, vegetables, grains, fish, meat,
milk, eggs, and dairy products
USES
Indoor air, outdoor air, soil, house dust, tap water, food from home
gardens and locally-produced fruits, vegetables, grains, fish, meat,
milk, eggs, and dairy products
BEADS
Indoor air, outdoor air, and tap water
SCREAM2
Indoor air, outdoor air, soil, tap water, food from home gardens,
locally-grown produce, locally-raised animal products, including fish,
and breast milk
DERM
Liquid, air, soil on surfaces, dust on surfaces, residues on surfaces
DERMAL
Contact with surfaces (dermal)
Models
inhalation only
CPIEM
Designed for indoor exposures to numerous pollutants
pNEM
Models are pollutant-specific (e g , ozone, carbon monoxide, sulfur
dioxide) Current development is on CO-version
HAPEM
For criteria and modeled air toxic pollutants
AirPEx
Default values available for benzene, B(a)P, ozone, and PM
HEM
Multiple gas- and particle-phase agents from outdoor sources
SHAPE
Model designed for inhalation exposures to CO
BEAM
Model for benzene only
pHAP
Multiple gas- and particle-phase agents from outdoor sources Has
been tested using benzene
Models non-
mhalation routes
of exposure
CONSEXPO
Inhalation, ingestion, and dermal contact of chemicals from consumer
products Indoor sources only
NOVEMBER 1999
B-34
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Feature
Model
Remarks
CalTOX
Inhalation, ingestion, and dermal contact of organic chemicals and
some metals.
MMSOILS
Inhalation, ingestion, and dermal contact of organic chemicals and
metal species from hazardous waste sites
USES
Inhalation, ingestion, and dermal contact of organic chemicals and
some metal species
BEADS
Inhalation, ingestion, and dermal contact to benzene only
SHEDS
(under development)
Inhalation, dietary and non-dietary ingestion, dermal contact for
multimedia pollutants
SCREAM2
Inhalation, ingestion, and dermal contact to numerous modeled air
toxics
DERM
Dermal exposure to pesticides
DERMAL
Calculates screening-level estimates of annual individual dermal
potential dose rates to components of 16 consumer product
categories
Explicit
treatment of
variability
CPIEM
Several parameters are sampled from distributions, including the
concentration data A variety of formats for describing the
concentration distribution is allowed by the model, including provision
of a data file containing the concentration values A random number
seed can be selected by the user or the model can use the system's
clock to determine the seed
CalTOX
Variability in exposure factors and in landscape factors are explicitly
represented by probability distributions
pNEM
Environmental, demographic (e g , activity pattern and ventilation
data), and mass balance inputs are characterized by distributions
HAPEM
Most input parameters are characterized by distributions
AirPEx
Exposure distributions for the population are characterized by the
normalized cumulative frequency distribution
SHEDS
(under development)
Model samples from user-specified input probability distributions for
residues and exposure factors Monte Carlo sampling allows analyses
of input and output variability
HEM
SHAPE
Model generates a hypothetical population, sampling each descriptive
parameter of an individual (e g., age, gender, body mass) from a
user-specified distribution function
DERM
Variability quantified by applying bootstrap method to dermal
exposure estimates for individuals
BEAM
pHAP
Explicit
treatment of
uncertainty
CPIEM
User specifies all inputs. Then, the model is run several times with all
inputs kept constant except the random number seed. Variability for
each output parameter across repeated model runs is characterized
through a measure such as the coefficient of variation.
NOVEMBER 1999
B-35
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Feature
Model
Remarks
SHEDS
(under development)
Two-stage Monte Carlo sampling allows for explicit characterization of
both uncertainty and variability Model samples from user-specified
input probability distributions and their associated uncertainty
distributions
CalTOX
All chemical-specific property data are represented by probability
distributions
HEM
pNEM
The model is generally run 10 times using a Monte Carlo simulation
approach for each regulatory scenario analyzed
Mass balance
approach for
indoor
concentrations
pNEM
Mass balance model accounts for outdoor concentration, air exchange
rate, building volume, building penetration rate, deposition rate, and
indoor emission rates and usage patterns for some indoor sources
SHEDS
(under development)
Mass balance model accounts for outdoor concentration, air exchange
rate, building volume, and indoor emission rates, and usage patterns
CPIEM
Uses a mass balance equation, based on the principle of conservation
of mass, to estimate concentration distributions for specific types of
indoor environments such as residences, offices, and schools
SCREAM2
Optional use of mass balance model that accounts for outdoor
concentration, air exchange rate, and indoor emission rate
Regression or
I/O ratios for
indoor
concentrations
HAPEM
Uses microenvironmental factors
AirPEx
Parameters relating concentration measured at monitoring site to four
macroenvironments (rural, urban, city, and transit) which are in turn
related to three indoor microenvironments (home, vehicle, elsewhere)
by I/O ratios Also, three additive terms account for indoor sources in
the same three microenvironments
SHEDS
(under development)
For certain microenvironments, empirical mass balance models in the
form of regressions are used
HEM
Uses I/O ratios for microenvironments
SHAPE
Exposure concentrations are obtained by applying a superposition
principle to contributions from the ambient and different
microenvironments
BEAM
Same as SHAPE
SCREAM2
Optional use of I/O ratios for microenvironments
Includes
ventilation rate
pNEM
Ventilation rate (VE) value estimated based on estimated body mass,
gender, and other information from energy expenditure literature for
each exposure event
SHEDS
(under development)
Ventilation rate (VE) value estimated based on estimated body mass,
gender, and other information from energy expenditure literature for
each exposure-event
NOVEMBER 1999
B-36
TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Feature
Model
Remarks
CPIEM
Breathing rates supplied by the model are specific to three age/gender
groups (adult males, adult females, and children under age 12) and
four activity levels (resting, light, moderate, and heavy).
SCREAM2
For each hour, the activity pattern of each subgroup is assigned to
one of three activity levels (low, moderate, or heavy) The user may
designate enhancement of inhaled dosage of the pollutant for the
various activity levels with scaling factors
AirPEx
Ventilation rate is a function of a person's body mass and level of
activity Uses five levels of activity, ranging from "sleeping" to "heavy
exercise"
Can estimate
dose
pNEM
The CHAD database provided an activity indicator for each exposure
event Each activity type was assigned a distribution of values for the
metabolic equivalent of work (MET) The MET is dimensionless, given
by the ratio of the rate of energy expenditure during a particular
activity (expressed in kcal/mm) and a person's typical resting
metabolic rate (also expressed in kcal/mm)
SHEDS
(under development)
Inhalation For each modeled individual's sequential location/activity
combination in daily diaries, model combines air concentration,
activity-specific inhalation rate (derived via METs), and absorption
fraction to estimate inhalation absorbed dose Next generation will
include more PK models to calculate dose
Ingestion For each eating and drinking event, model combines mass
residue ingested and absorption fraction to obtain mass in
gastrointestinal (Gl) tract, then applies a Gl absorption rate constant
to estimate mass in blood compartment
Dermal For each dermal contact event, model combines available
mass on skin and dermal absorption rate constant to estimate dermal
absorbed dose
To obtain aggregate absorbed and eliminated dose, model sums time
profiles across all routes
CPIEM
Breathing rates and activity levels are used by the model to calculate
the potential inhaled dose received by each individual in each
microenvironment
SCREAM2
Average daily doses of a pollutant to the population are estimated
from concentrations in each medium on the basis of age-specific
ingestion rates and body weights and local population age profiles
Includes indoor
sources
CPIEM
The model samples values from user-specified distributions for
emission rates for indoor sources
SHEDS
(under development)
The model samples values from user-specified distributions for indoor
source concentrations
pNEM
Includes CO emitted by gas stove operation and passive smoking.
HAPEM
Includes an additive term for indoor source contributions (user
specified)
SCREAM2
Includes an additive term for indoor source contributions (user
specified)
Includes
smoking as a
source
pNEM
Contribution from smoking is modeled.
NOVEMBER 1999
B-37
TRIM.EXPO TSD (DRAFT)
-------
APPENDLX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Feature
Calculates
exposures of
commuters
Includes
dispersion
algorithms to
calculate
outdoor air
concentrations
Model
SHEDS
(under development)
pNEM
SHEDS
(under development)
HAPEM
SCREAM2
SCREAM2
SHEDS
(under development)
HEM
Remarks
Contribution from smoking is modeled
The populations of the commuting cohorts (assumed to include all
working cohorts) were determined by the expression
Com(d,h,f,w) = Pop(d,h,f) x Com(h,w)AVork(h)
where Com(d,h,f,w) is the number of persons in the commuting cohort
associated with demographic group d, home district h, cooking fuel f,
and work district w, Pop(d,h,f) values provided an estimate of the
population of each non-commuting cohort residing within home district
/?, Com(h.w) is the number of workers in all demographic groups that
commute from home district h to work district w, and Work(h) is the
total number of workers in home district h Estimates of Work(h) were
developed from census data specific to each district The m-vehicle
concentration is calculated using a mass balance model
Exposures of commuters are modeled. A similar approach to pNEM
is being considered
Commuting patterns of workers between exposure districts are
modeled The "travel time to work" data from the 1990 census are
used to develop the commuting patterns A program uses these data
to build an array of probabilities of the movement of working
commuters for each census tract-to-tract combination
Commuting patterns of workers between exposure districts are
modeled The patterns were estimated from travel survey data
collected by the Southern California Association of Governments in a
manner similar to pNEM/CO
Uses a Gaussian model formulation and climatological data to
estimate long-term average concentrations as a function of radial
distance and direction from a source for a set of receptors laid out in a
radial grid pattern
Uses Bayesian spatial and temporal interpolation methods to estimate
outdoor air concentrations in different census tracts
Uses the Industrial Source Complex Long-Term (ISCLT2) Model for
estimating dispersion
NOVEMBER 1999
B-38
TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Table B-23
Strengths and Weaknesses of Different Models and Modeling Systems
Model
Strengths
Weaknesses
pNEM/CO
(inhalation)
1. Most input parameters are probabilistic
2 Can calculate delivered dose as a
function of the pollutant concentration
and ventilation rate values assigned to
the event and the demographic
characteristics of the cohort
3 Mass balance model for indoor
microenvironments
4 Calculates the exposures to the portion of
the population that commute to work
5 Estimates exposures to those exposed to
passive smoking
1. Single exposure route (inhalation) only
2. Each version of pNEM is specific to a single
pollutant
CalTOX
(multimedia)
1 Multipathway and multimedia
2 All input parameter values are
distributions
3 Explicit treatment of pollutant
concentrations in various environmental
media
4 Mass balance model
1 Does not allow spatial tracking of a pollutant
2 Limited number of chemical species for
which the model is applicable
3 Limited in the extent of the environmental
settings for which it can be applied
HAPEM4
(inhalation)
1 Most input parameters are probabilistic
2 Model can use both measured air quality
data and modeled data from the ASPEN
model or from air dispersion models
3 Can model population exposures down to
the census tract-level
4 User can easily specify different
demographic groups (providing they have
data for the groups)
5 User can specify a "lag" factor for
calculating indoor concentrations from
pollutants penetrating from outside
1 Single exposure route (inhalation) only
2 The sequence of exposure events for
activities is not preserved
3 Does not provide any estimate of ventilation
rate or delivered dose
4 Currently, does not account for exposures to
passive smoking
3 This table only includes models and modeling systems that are currently publicly available. OAQPS will
continue to monitor and incorporate into TRIM.Expo, where appropriate, features, algorithms, and/or models that
are under development. This includes ongoing work by the U.S. EPA on models and modeling systems, such as
MENTOR and SHEDS
NOVEMBER 1999
B-39
TRIM EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Model
Strengths
Weaknesses
IEM
(multimedia)
Establishes procedures for estimating the
indirect human exposures and health
risks that can result from the transfer of
air pollutants to soil, vegetation, and
water bodies
Addresses exposures via multiple routes
(inhalation, food, water, and soil
mgestion, and dermal contact)
Has undergone extensive scientific
review
Has been widely used in EPA screening-
level risk assessments
Relatively simple spreadsheet model
Based on annual average air concentrations
and deposition rates
Structured as a one-way process through a
series of linked models Not a truly coupled
multimedia model Does not have the ability
to model feedback loops between media or
secondary emissions
Not designed to readily address spatial
variability in exposures
Not designed for probabilistic variability and
uncertainty analysis
Not suitable for estimating population
exposures.
Does not provide a detailed time series of
media concentrations or the resulting
exposures
Can only be applied to chemicals that are
emitted into the air
ISMCM
(multimedia)
Considers all media, biological and non-
biological, in one integrated system
Includes both spatial and compartmental
modules to account for complex transport
of pollutants through the ecosystem
Mass conserving model
Includes estimation of intermedia transfer
factors
Links and spatial compartments are
predetermined
Not structured to incorporate
uncertainty/variability directly into the model
operation
SCREAM2
(multimedia)
Can calculate indoor pollutant
concentrations using the Indoor Air
Quality Model (IAQM) or by using
indoor/outdoor ratios
Multipathway inhalation, mgestion, and
dermal
Results reported in terms of both
concentrations/dosages and risks
Includes air dispersion algorithms to
calculate air concentrations from
emissions
1 Deterministic
2 Results reported for annual average
exposures only
NOVEMBER 1999
B-40
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON'/CRITIQUE OF EXPOSURE MODELS
Table B-3
Model Features for pNEM/CO
Attribute
General
Modeled area, study
population, and
modeling period
Component
Model name
Pollutants of concern
Reference
Model status
Contact/Affiliation
Stochastic''
Variability?
Uncertainty''
Study areas where model
has been applied
Spatial designation of
study area
Sub-area designations
Exposure duration for
modeling
General population of
interest
Special subgroups or
designations
Remarks
pNEM/CO
Carbon monoxide
Johnson et al 1999 for U S EPA, OAQPS Estimation of
Carbon Monoxide Exposures and Associated
Carboxyhemoglobm Levels in Denver Residents Using
pNEM/CO (Version 2 0).
Operates on either mainframe or PC Further
enhancements are currently ongoing
Harvey Richmond (U S EPA, OAQPS)
(919)541-5271
Yes - most variables chosen stochastically
Yes - year-long exposure-event sequences (EES) use data
from multiple subjects to better represent the variability of
exposure that is expected to occur among the persons
included in the cohort
Yes
Most recently to Denver Application to Los Angeles is
planned
50 km radius surrounding the city center of Denver
Six exposure districts, each 10 km in radius, surrounding
fixed site CO monitors
Typically one year
Typically defined as people with specific demographic or
health status characteristics (e g , adults with ischemic
heart disease)
Demographic groups (DG)
1 Children 0 to 17 years,
2 Males, 18 to 44, working,
3 Males, 18 to 44, non-working,
4 Males 45 to 64, working,
5 Males 45 to 64, non-working,
6. Males 65+,
7 Females, 18 to 44, working,
8 Females, 18 to 44, non-working,
9 Females 45 to 64, working,
10 Females 45 to 64, non-working,
11. Females 65+
NOVEMBER 1999
B-41
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Exposure events
Component
Special attributes for
subgroups
Source of demographic
data for study population
Environmental media
Exposure media
Pathways
Routes
Time resolution of
exposure events
Integration of exposures
across multiple media
Method for determining
pollutant contact rate
Activity pattern
methodology
Source of activity pattern
data
Time resolution of activity
patterns
Microenvironments
(Inhalation)
Exposure locations
(Ingestion)
Model calculates exposure
of commuters
Remarks
Each DG further subdivided into cohorts identified as a
distinct combination of (1) home district, (2) demographic
group, (3) work district (if applicable), (4) residential
cooking fuel, and (5) replicate number
1990 Bureau of the Census
Ambient air.
Indoor air, outdoor air
Ambient air and indoor air to personal air
Inhalation
Inhalation'
One minute
Inqestion N/A
N/A
Inhalation
Ventilation rate (VE) value estimated for each exposure-
event VE expressed as liters of air respired per minute
(liters mm"1)
Inqestion N/A
In typical pNEM applications, the EESs are determined by
assembling activity diary records relating to individual 24-
hour periods into a year-long series of records
Each exposure event within an EES was defined by (1 )
district, (2) CHAD location descriptor, (3)
microenvironment, (4) CHAD activity descriptor, and (5)
passive smoking status
Comprehensive Human Activity Database (CHAD)
One minute
1 Indoors - residence
2-6 Indoors - nonresidence A- E
7. Indoors - residential garage
8. Outdoors - near roadway
9 Outdoors - other locations
10. Vehicle - automobile
11 Vehicle- other
12 Outdoors - public parking or fueling
N/A
Yes -the number of commuters in each working cohort is
calculated based on census data
In vehicle concentrations are estimated using a mass
balance model (see Concentrations and Sources).
NOVEMBER 1999
B-42
TRIM.EXPO TSD (DRAFT)
-------
APPENTDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Concentrations and
sources
Extrapolation to
study population
Component
Outdoor concentration
determination method
Indoor concentration
determination method
In-vehicle concentration
estimation
Passive smoking
Other indoor sources
Method of allocating
estimated exposures to
study population
Remarks
Hourly-average CO concentrations for outdoor
microenvironments based on data from fixed-site monitor
and statistical relationship between fixed-site data and
personal monitoring for outdoor microenvironments from a
previous personal exposure study in Denver Exposure
districts are defined by monitor locations
Mass balance model used to estimate CO concentrations
when a cohort is assigned to an indoor or motor vehicle
microenvironment
Mass balance model which accounts for outdoor
concentration, air exchange rate, and passive smoking
status of occupants
CO contribution from indoor and m-vehicle passive
smoking is modeled using a mass balance model
Gas stoves
Entire population is simulated through the use of cohorts
and census data relating cohorts to study area population
NOVEMBER 1999
B-43
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Table B-4
Model Features for CalTOX
Attribute
General
Modeled area, study
population, and
modeling period
Component
Model name
Pollutants of concern
References
Model status
Contact/Affiliation
Stochastic"?
Variability?
Uncertainty7
Study areas where model
has been applied
Spatial designation of
study area
Sub-area designations
Exposure duration for
modeling
Remarks
CalTOX (California Total Exposure Model for Hazardous
Waste Sites)
Potential toxic chemicals placed in landfills and in
controlled and formerly uncontrolled hazardous waste
sites
Chemicals on the Toxic Release Inventory (TRI) list
emitted to air or water
McKone, TE 1993 CalTOX, A Multimedia Total-
Exposure Model for Hazardous Wastes Sites Lawrence
Livermore National Laboratory, Livermore, CA,
UCRL-CR-1 1 1456 Part I Executive Summary, Part II
The Dynamic Multimedia Transport and Transformation
Model, Part III The Multiple-Pathway Exposure Model
Two versions of CalTOX are currently available from Cal-
EPA the original version developed for soil clean-up goals
and a second version used for waste classification
A third version has been developed for use with the
Environmental Defense Fund (EDF) Scorecard project
Tom McKone, Lawrence Berkeley National Laboratory
(510-642-8771)
Cal-EPA version Ned Butler (916-323-3751)
Yes all model inputs are represented by a mean value,
coefficient of variability, and default distribution
Yes variability in exposure factors and in landscape
factors are explicitly represented by probability
distributions
Yes all chemical-specific property data are represented
by probability distributions
For setting soil clean-up goals and for assessing residual
risk at municipal landfills, CalTOX was used to represent
the variability among all California land areas
For the EDF Scorecard project, CalTOX was used to
represent the fate of air and water emissions in all 48
conterminous U.S states
The designated study area is the environment impacted by
either a waste site, air emissions, or water releases 5,000
GIS land units were used to establish variability among
California locations For the EDF version, county-level
climate and land data are used to establish state-level
variability
The exposure location can be modeled with a residential,
agricultural, commercial, or industrial scenario
Exposure duration is variable depending on how long the
exposure individual remains at the exposure location.
Values as long as 70 years can be used
NOVEMBER 1999
B-44
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPAR1SONT/CRIT]QUE OF EXPOSURE MODELS
Attribute
Exposure events
Component
General population of
interest
Special subgroups or
designations
Special attributes for
subgroups
Source of demographic
data for study population
Environmental media
Exposure media
Pathways
Routes
Time resolution of
exposure events
Integration of exposures
across multiple media
Method for determining
pollutant contact rate
Remarks
Populations in the landscape impacted by a waste site, air
release, or water release
Residential populations
Children (0 to 12 years)
Adults (12 to 70 years)
Agricultural populations
Those working or shopping at a commercial site
Those working at an industrial site
Those with home gardens have been singled out for
special attention
EPA Exposure Factors Handbook.
Ambient air, surface soil, root zone soil, surface water and
ground water
Indoor air, outdoor air, soil, house dust, tap water, food
from home gardens and locally-produced fruits,
vegetables, grains, fish, meat, milk, eggs, and dairy
products
Twenty-three pathways linking ambient media to exposure
media
Inhalation, mgestion, dermal contact
Inhalation 1 day resolution to build the exposure scenario
that is repeated over longer duration based on exposure
frequency
Inqestion 1 year
Dermal contact 1 day resolution to build the exposure
scenario that is repeated over a longer duration based on
exposure frequency data
Exposures are aggregated across all media and pathways
to construct a total intake by route-inhalation, mgestion,
and dermal contact
Inhalation' Activity adjusted breathing rate per unit body
weight
Ingestion Food consumption per unit body weight for each
major food category is adjusted by fraction of food
consumed within the contaminated landscape
Dermal contact Activity data are used to establish lonq-
term average rate and duration of contact with tap water
and soil
NOVEMBER 1999
B-45
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Concentrations and
sources
Extrapolation to
study population
Component
Activity pattern
methodology
Source of activity pattern
data
Time resolution of activity
patterns
Microenvironments
(Inhalation)
Exposure locations
(Ingestion)
Model calculates exposure
of commuters
Outdoor concentration
determination method
Indoor concentration
determination method
In-vehicle concentration
estimation
Passive smoking
Other indoor sources
Method of allocating
estimated exposures to
study population
Remarks
For inhalation, the day is divided into the amount in each
microenvironment and a breathing rate is assigned to each
microenvironment
For ingestion, consumption of each major food category is
divided into local and non-local sources
For dermal contact, the number and duration of water or
soil contact events in a day are used
EPA Exposure Factors Handbook
1 to 24 hours for inhalation
Annual consumption patterns for ingestion
Minutes to hours for dermal contact
Outdoors at home,
Outdoors away from home,
Indoors at home,
Indoors in the bathroom
Residential environment
No
Multimedia mass balance using a dynamic regional
fugacity model Sources to air, soil, and water are allowed
Based on a simple penetration model for outdoor air, a
dust tracking model for soil pollutants, a transfer model for
soil gas drawn into home, and a transfer model from water
to indoor air
Not explicitly represented, assumed to be equal to outdoor
concentration
Not included
Stripping of chemicals Household water uses
Tracking of soil to house
Transfer of volatile chemicals from soil gas below homes
Entire population is simulated through the use of
probability distributions to represent variability and
uncertainty in source-to-dose relationships
NOVEMBER 1999
B-46
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Table B-5
Model Features for HAPEM4
Attribute
General
Modeled area, study
population, and
modeling period
Component
Model name
Pollutants of concern
Reference
Model status
Contact/Affiliation
Stochastic''
Variability''
Uncertainty9
Study areas where model
has been applied
Spatial designation of
study area
Sub-area designations
Exposure duration for
modeling
General population of
interest
Remarks
HAPEM4 (Hazardous Air Pollutant Exposure Model)
Criteria pollutants for which measured data, or HAPs for
which modeled data can be obtained Most recent
application was for benzene
Operates on UNIX workstations
Ted Palma (U S EPA, OAQPS)
(919)541-5470
Yes - most variables chosen stochastically
Yes - distributions available for many input variables
Not expl/citely handled by the current model
Most recently Houston and Phoenix
*Note - HAPEM4 may also be run in a measured data
mode using data from fixed-site monitors In this mode,
the spatial designation and study areas are similiar to
those in pNEM
Houston and Phoenix Metropolitan Statistical Areas
(MSAs)
*ln measured data mode study areas are typically a circle
with a predetermined radius surrounding the city center
Census tracts
*ln measured data mode up to 18 exposure districts
surrounding fixed-site monitors can be chosen for each
MSA
Typically one year (results are currently aggregated and
reported as 3 month "seasons")
Typically defined as all persons described by the
demographic groups for each sub-area designation
NOVEMBER 1999
B-47
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Exposure events
Component
Special subgroups or
designations
Special attributes for
subgroups
Source of demographic
data for study population
Environmental media
Exposure media
Pathways
Routes
Time resolution of
exposure events
Integration of exposures
across multiple media
Method for determining
pollutant contact rate
Remarks
Demographic groups (DG) in latest version (other groups
for which census data are available may be input by the
user)
1. Caucasians,
2 African-Americans,
3 All people not in DG 1 or 2,
4 Children 0 to 5 years,
5 Children 6 to 11 years,
6 Children 12 to 17 years,
7 Males 18 to 64,
8 Females 18 to 64,
9 Males 65+,
10 Females 65+;
1 1 People with gas stoves at home,
12 People without gas stoves at home,
13 People with attached garages at home,
14 People without attached garages at home,
15 People age 18 and older who work and were outdoors
for at least 240 minutes (4 hrs) on the day recorded,
16 All persons
No further sub-designations of population for this
application However, in future development, cohorts may
be identified for commuting by specifying a home district
and work district Other cohort attributes may also be
specified in future versions
1990 Bureau of the Census
Ambient air
Indoor air and outdoor air
'Ambient air and indoor air to personal air
Inhalation
Inhalation
1 to 60 minutes, obtained from CHAD
Inqestion N/A
N/A
Inhalation. N/A
Inqestion N/A
NOVEMBER 1999
B-48
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Remarks
Activity pattern
methodology
Information on the time spent in various
microenvironments (ue) for each individual are used.
* Note, this is not an activity sequence, rather it is the total
time spent in each ue during each 1-hour block of time
throughout the day
A daily record (comprised of 24 separate hours) is chosen
using the Monte Carlo technique The record chosen is
matched to the day being modeled by using the maximum
outdoor temperature for each hour The number of
minutes spent in a particular ue for each individual
selected and for each hour of the day are calculated. The
exposures reported for each DG are the averages of all
individuals sampled from each group
Source of activity pattern
data
Comphrensive Human Activity Database (CHAD)
Time resolution of activity
patterns
1 to 60 minutes
Microenvironments
(Inhalation)
1 In vehicle - car
2 In vehicle - bus
3 In vehicle -truck
4 In vehicle - other
5 Indoors - public garage
6 Outdoors - parking lot/garage
7 Outdoors - near a road
8 Outdoors - motorcycle
9, Indoors - service station
10 Outdoors - service station
11 Indoors - residential garage
12 Indoors - other repair shop
13 Indoors - residence, no inside sources of CO
14 Indoors - residence, gas stove present
15 Indoors - residence with attached garage
16 Indoors - residence with stove and attached garage
17 Indoors - office
18 Indoors - store
19 Indoors - restaurant
20 Indoors - manufacturing facility
21 Indoors - school
22 Indoors - church
23 Indoors - shopping mall
24 Indoors - auditorium
25 Indoors - health care facility
26 Indoors-other public building
27. Indoors other location
28 Indoors - not specified
29. Outdoors - construction site
30 Outdoors - residential grounds
31. Outdoors - school grounds
32 Outdoors - sports arena
33 Outdoors - park/golf course
34. Outdoors - other location
35 Outdoors - not specified
36. In vehicle - train/subway
37 In vehicle-airplane
NOVEMBER 1999
B-49
TRIM. EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Concentrations and
sources
Extrapolation to
study population
Component
Exposure locations
(Ingestion)
Model calculates exposure
of commuters
Outdoor concentration
determination method
Indoor concentration
determination method
In-vehicle concentration
estimation
Passive smoking
Other indoor sources
Method of allocating
estimated exposures to
study population
Remarks
N/A
Yes - commuting patterns of workers between exposure
districts are modeled. The "travel time to work" data from
the 1990 census are used to develop the commuting
patterns. A program uses this to build an array of
probabilities of the movement of working commuters for
each census tract-to-tract combination
Modeled air quality mode. The model is currently
configured to read output at the census tract-level from the
ASPEN (Assessment System for Population Exposure
Nationwide) model.
Measured data mode Hourly-average CO concentrations
taken from fixed-site monitors Exposure districts defined
by monitor locations
Uses microenvironmental factors These factors are
obtained from field studies through a linear regression of
microenvironmental concentrations against fixed-site
monitoring concentrations
Microenvironmental factor (same as indoor concentration
determination method above)
N/A
Gas stoves and residences with attached garages through
the use of additive factors (user supplied)
Entire population is simulated through the use of cohorts
and census data relating cohorts to study area population
NOVEMBER 1999
B-50
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Table R-6
Model Features for SCREAM2
Attribute
Comoonent
Remarks
General
Model name
SCREAM2
(South Coast Risk and Exposure Model, Version 2 0)
Pollutants of concern
HAPs for which emissions data can be obtained
Reference
Rosenbaum, A S. User's Guide for an Enhanced Version
of the South Coast Air Quality Management District's Air
Toxics Risk and Exposure Assessment Model (SCREAM2
- PC version)
Model status
Operates on UNIX workstations and PCs
Contact/Affiliation
Unix Henry Hogo, South Coast Air Quality Management
District, (909)396-3100
PC Arlene Rosenbaum, ICF Consulting Group,
(415)507-7192
Stochastic?
Variability?
No
Uncertainty?
Modeled area,
study population,
and modeling
period
Study areas where model
has been applied
California's South Coast Air Basin
Spatial designation of
study area
All block groups in the South Coast Air Basin as defined
by the Bureau of the Census
Sub-area designations
Block group centroids
Exposure duration for
modeling
One year
General population of
interest
Typically defined as all persons described by the
demographic groups for each sub-area designation.
Special subgroups or
designations
Default demographic groups (other groups for which
census data are available may be input by the user)
1 Students 18 and over,
2 Managers and professionals,
3 Sales workers,
4. Clerical and kindred workers,
5 Craftsmen and kindred workers,
6 Farmers,
7. Operatives and laborers,
8. Service, military, and private household workers;
9 Housepersons,
10 Unemployed and retired persons,
11. Children under 5,
12 Children 5 to 17
Special attributes for
subgroups
Each demographic group is further sub-divided into
cohorts identified as a distinct combination of (1) home
district, (2) demographic group, and (3) work district (if
applicable),
NOVEMBER 1999
B-51
TRIM EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Remarks
Source of demographic
data for study population
1990 Bureau of the Census and Southern California
Association of Governments (commuting)
Exposure events
Environmental media
Ambient air, soil, surface water, food crops
Exposure media
Indoor air, outdoor air, soil ingested or contacted, drinking
water, food ingested, and breast milk ^^
Pathways
Ambient air, indoor air to personal air, soil to ingested
soil, contacted soil, soil to root crops and leafy crops to
ingested food, soil and surface water to animal food
sources to ingested food, ambient air and soil to surface
water to drinking water, ingested food and drinking water
to ingested breast milk
Routes
Inhalation, ingestion, dermal contact
Time resolution of
exposure events
Inhalation 1 hour
Inqestion daily
Integration of exposures
across multiple media
Yes
Method for determining
pollutant contact rate
Inhalation N/A
Inqestion N/A
Activity pattern
methodology
Information on the fraction of time spent in various
microenvironments by each demographic group for each
of 24 hours is used
Source of activity pattern
data
Roddm, Ellis, and Siddiqee (1979) "Background Data for
Human Activity Patterns, Vols I and II " SRI
International
Time resolution of activity
patterns
1 hour
Microenvironments
(Inhalation)
1 Indoors - residence
2a Indoors - office
2b Indoors - school
3 In vehicle
4 Outdoors - near a road
5 Outdoors - other location
Exposure locations
(Ingestion)
Home
Model calculates exposure
of commuters
Yes - commuting patterns of workers between exposure
districts are modeled Commuting data provided by the
Southern California Association of Governments
Concentrations and
sources
Outdoor concentration
determination method
Estimated from data on emission sources with the air
dispersion module
Indoor concentration
determination method
Alternative 1 the Indoor Air Quality Model (IAQM), a
mass balance model using outdoor concentration, air
exchange rates, filtration rates, and mixing ratios
Alternative 2 indoor/outdoor concentration ratios obtained
from field studies
In-vehicle concentration
estimation
Same as for indoor concentration determination method
NOVEMBER 1999
B-52
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Extrapolation to
study population
Comoonent
Passive smoking
Other indoor sources
Method of allocating
estimated exposures to
study population
Remarks
N/A
Additive factors (user supplied)
Entire population is simulated through the use
and census data relating cohorts to study area
of cohorts
population
NOVEMBER 1999
B-53
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMP PRISON/CRITIQUE OF EXPOSURE MODELS
Table B-7
Model Features for CPIEM
Attribute
Component
Remarks
General
Model name
Pollutants of concern
Reference
Model status
Contact/Affiliation
Stochastic?
Variability?
Uncertainty"
CPIEM (California Population Indoor Exposure Model)
Benzene, berizo[a]pyrene, CO, chloroform, formaldehyde,
NO-, PM10, perchloroethylene, trichloroethylene, and PAHs
User has the ability to input data for other chemicals
California Air Resources Board 1998 Development of a
model for assessing indoor exposure to air pollutants
Sacramento, CA Report No A933-157 January 1998
Operates on PCs
Susan Lum (California Air Resources Board)
(916)323-5043
Yes
Yes
The model needs to be run several times with all inputs the
same except the random number seed Variance in each
output parameter across the repeated model runs is then
used to characterize the uncertainty through a measure
such as the coefficient of variation
Modeled area, study
population, and
modeling period
Study areas where model
has been applied
Most recently applied to the South Coast region
(encompasses Los Angeles and surrounding areas)
Spatial designation of
study area
South Coast region, San Francisco Bay area, or the
remainder of the State of California
Sub-area designations
County
Exposure duration for
modeling
Varies (user-defined)
General population of
interest
California population
Special subgroups or
designations
Certain identified subgroups of the population such as
individuals who may be sensitive to indoor pollutants
Special attributes for
subgroups
For purposes of calculating potential dose, the population
was further segregated into the following groups' adult
males, adult females, and children (age 12 and younger)
Source of demographic
data for study population
Exposure events
Environmental media
Indoor air and outdoor air.
Exposure media
Indoor air and outdoor air
Pathways
Ambient air and indoor air to personal air.
Routes
Inhalation
NOVEMBER 1999
B-54
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Time resolution of
exposure events
Integration of exposures
across multiple media
Method for determining
pollutant contact rate
Activity pattern
methodology
Source of activity pattern
data
Time resolution of activity
patterns
Microenvironments
(Inhalation)
Exposure locations
(Ingestion)
Model calculates exposure
of commuters
Remarks
Inhalation 1-hour periods, but these may be aggregated
into an 8-hour exposure event Also, a 12- or 24-hour
exposure event may be specified (the basis for these are a
single activity per individual)
Ingestion N/A
N/A
Inhalation Breathing rates are supplied by the model for
three age/sex groups (adult males, adult females, and
children under age 12) and four activity levels (resting,
light, moderate, heavy) Using information sampled on the
quantity of time spent in an environment and the
concentration in each environment, combined with the
breathing rates, the model calculates the potential inhaled
dose received by each individual in each environment
Ingestion N/A
Location codes were grouped into nine types of
environments Then, the time spent was summed across
locations within each environment type This was done
separately for each individual for a 24-hour period, 12-hour
daytime and nighttime periods, 24 sequential 1-hour
periods, and 24 running 8-hour periods Within each of the
nine environment types, each individual's time was further
disaggregated according to four activity levels (resting,
light, moderate, heavy) In addition, demographic
information (e g , age, gender, location of residence, month
and day of week when the activity was recorded, work
status, income) on each person was matched to their
location/activity information An index number was
assigned sequentially to each record of each file in order to
enable linking the information
Two ARB-sponsored studies The first for a target
population of adults (18 years and older) and adolescents
(12 to 17 years), provided 1,762 profiles, the second, for a
target population of children (aged 11 years and younger),
provided 1,200 profiles
One minute
Environment types residences (numerous
microenvironments), offices (office, bank, or post office),
industrial plants; school, travel in enclosed vehicles
(several microenvironments); stores and other public
buildings (several microenvironments), restaurants and
lounges, other indoor locations (several
microenvironments), outdoors (several
microenvironments)
N/A
Distributional data on concentrations for travel in enclosed
vehicles were taken from studies identified in the literature
NOVEMBER 1999
B-55
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Concentrations and
sources
Extrapolation to
study population
Component
Outdoor concentration
determination method
Indoor concentration
determination method
In-vehicle concentration
estimation
Passive smoking
Other indoor sources
Method of allocating
estimated exposures to
study population
Remarks
Two options are provided for inputting outdoor
concentrations daily-average values and hourly-average
values Sources of data identified for input (for
formaldehyde, several VOCs, B(a)P, PM10, NO2, and CO)
include state/local ambient air monitoring networks and
some indoor air monitoring studies where outdoor
concentrations were measured in conjunction with indoor
measurements
Mass balance model
Each source is associated with both a pollutant (a source
may be associated with more than one pollutant) and a
source category ( long-term, episodic, or frequent)
For each source category, different input parameters are
required by the model which are sampled from
distributions
The model calculates the initial indoor concentration and
the hourly emission rates These contributions are then
used by the mass balance routine to calculate the indoor
concentrations
Distributional data on concentrations for travel in enclosed
vehicles were taken from studies identified in the literature
Examples of the indoor sources considered for pollutants
provided with the model are tobacco smoking, unburned
natural gas (leaks), various consumer products, wood
burning, range cooking, range pilot light, chlorinated water
supply, pressed wood products, vacuuming/sweeping, and
dry-cleaned clothes
*Note Data are not available for each pollutant being
emitted from each of these identified sources
NOVEMBER 1999
B-56
TRIM.Expo TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Table B-8
Model Features for SHEDS
Attribute
Component
Remarks
General
Model name
Pollutants of concern
References
Model status
Contact/Affiliation
SHEDS
(Stochastic Human Exposure and Dose Simulation) Model
pesticides, particulate matter, and other POPs, metals, and
air toxics
Zartanan V G , Ozkaynak H., Burke J M., Zufall M J ,
Rigas M L , and Furtaw Jr. E.J A Modeling Framework
For Estimating Children's Residential Exposure and Dose
to Chlorpynfos Via Dermal Residue Contact and Non-
Dietary Ingestion Submitted to Environmental Health
Perspectives September 1999
Ozkaynak H , Zufall, M , Burke, J , Xue, J , Zidek, J.
1999a Predicting population exposures to PM10 and
PM2 5 Presented at PM Colloquium, Durham, NC. June,
1999
Ozkaynak H , Zufall, M , Burke, J , Xue, J , Zidek, J.
1999b A Probabilistic Population Exposure Model for
PM10 and PM2 5 Presented at 9th Conference of the
International Society of Exposure Analysis, Athens,
Greece September 5-8, 1999
Prototype first-generation SHEDS-Pesticides model has
been developed A case study for children and chlorpynfos
has been conducted using 1-stage Monte Carlo sampling
Currently case study is being completed for 2-stage Monte
Carlo sampling to conduct variability and uncertainty
analyses Next generation will focus on pesticides and
other multimedia, multipathway pollutant for different
population cohorts, and will assess both cumulative and
aggregate dose
Prototype first-generation SHEDS-PM model has been
developed A PM10 case study for Vancouver, Canada
has been conducted using 1-stage Monte Carlo sampling
Subsequent generations will focus on modeling both PM10
and PM2 5 exposure and dose in a selected U S city,
using 2-stage Monte Carlo sampling
Haluk Ozkaynak, U S. EPA National Exposure Research
Laboratory (919-541-5172)
Valerie Zartarian (SHEDS-Pesticides), U S. EPA National
Exposure Research Laboratory (703-648-5538)
Janet Burke (SHEDS-PM), U S. EPA National Exposure
Research Laboratory (919-541-0820)
Maria Zufall (SHEDS-PM), U.S EPA National Exposure
Research Laboratory (919-541-5461)
NOVEMBER 1999
B-57
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Remarks
Stochastic?
Yes - all model inputs are represented by a probability
distribution
Variability7
Yes - variability of inputs is explicitly represented by user-
specified probability distributions and Monte Carlo
sampling is applied to quantify variability in model outputs
Uncertainty''
Yes - for each input parameter probability distribution,
associated uncertainty distributions can also be specified
2-stage Monte Carlo sampling is applied to analyze
uncertainty in model outputs.
Modeled area, study
population, and
modeling period
Study areas where model
has been applied
Spatial designation of
study area
Sub-area designations
Exposure duration for
modeling
General population of
interest
Special subgroups or
designations
First-generation SHEDS-Pesticides model has been
applied to estimate children's indoor and outdoor (home
lawn) exposures and doses to chlorpynfos Children were
sampled from the National Human Activity Pattern Survey,
which includes all 48 conterminous U S states
First-generation SHEDS-PM model has been applied to
estimate PM10 exposures to individuals m Vancouver,
Canada
The designated study area is the geographic location
associated with the user-specified cohort from national
time/location/activity surveys in EPA's Consolidated
Human Activity Database (CHAD)
Sub-area designations are the set of locations
(microenvironments) occupied by individuals sampled from
the time/location/activity pattern surveys m EPA's
Consolidated Human Activity Database (CHAD)
Daily exposure and dose profiles are modeled for each
individual The time scales used to generate these profiles
differ by route For inhalation, the time scale ranges from
1 minute to 12 hours, depending on the pollutant and diary
selected For dermal contact and non-dietary ingestion,
the time scale is 5 seconds. For dietary ingestion, the time
scale for ingestion is instantaneous for each eating/drinking
event, but absorption is calculated on a 30-mmute time
scale after each ingestion event The daily profiles
correspond to time periods associated with the user-
specified environmental concentrations (e g , <1 day, 1-7
days, 8-30 days, 30-365 days post-pesticide application)
The U S population as represented by individuals in
time/location/activity surveys contained m EPA's
Consolidated Human Activity Database (CHAD)
The first generation of SHEDS-Pesticides focuses on
residential exposures to children (0-4 years and 5-9 years)
The next generation will be able to address residential,
non-residential, and occupational exposures for other
cohorts of interest
The first generation of SHEDS-PM estimates exposures by
gender and age category
NOVEMBER 1999
B-58
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Remarks
Special attributes for
subgroups
The first generation of SHEDS-Pesticides focuses on
children living in residences with lawns
Source of demographic
data for study population
Time/location/activity surveys in EPA's CHAD
Exposure events
Environmental media
SHEDS is a concentration-to-dose model in which the user
enters distributions for residues or concentrations in
exposure media rather than environmental media (except
for ambient air and surface soil)
Exposure media
indoor air, outdoor air, commuting/m-vehicle air, soil,
house dust, surface residues (indoor and lawn), hand and
object residues, tap water; food and beverages
Pathways
Ingestion of residues m food and beverages by
eating/drinking event,
Dermal contact with soil, dust, or residues on surfaces,
Non-dietary mgestion of residues on skin and objects
mouthed,
Inhalation of pollutants in indoor, outdoor, and
commutmg/in-vehicle locations
Routes
Inhalation, Dietary Ingestion (food, drinking water, other
beverages), Non-Dietary Ingestion (hand-to-mouth and
object-to-mouth), Dermal Contact
Time resolution of
exposure events
Inhalation Daily profiles with down to 1 minute resolution
Dermal and Non-Dietary Inqestion Daily profiles with 5-
second resolution
Dietary Ingestion Daily profiles with 30-mmute resolution
for absorption, residues ingested by eating events
assumed to be instantaneous
Integration of exposures
across multiple media
Exposures are estimated for each route and pathway via
sequential exposure profiles Corresponding dose profiles
for each route and pathway are calculated then summed
across routes.
Inhalation For sampled individual's daily sequential
time/location/activity diary events, combine location-
specific air concentrations (drawn from input probability
distributions), activity-specific inhalation rates (in CHAD,
derived using METS), and inhalation absorption fraction
Method for determining
pollutant contact rate
Dietary Ingestion For sampled individual's daily sequential
time/location/activity diary events, combine total residue
mass ingested during each eating/drinking event (sampled
from measured or modeled distributions) with dietary
absorption fraction.
Dermal Contact: For sampled individual's daily sequential
time/location/activity diary events, simulate sequences of
microlevel object contact events using probabilities
developed from videography study data For each discrete
microlevel contact event, combine surface residue, transfer
or removal efficiency, and dermal or Gl absorption rate
constant
NOVEMBER 1999
B-59
TRIM.EXPO ISO (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Remarks
Activity pattern
methodology
Source of activity pattern
data
Time resolution of activity
patterns
Microenvironments
(Inhalation)
Exposure locations
(Ingestion)
Daily time/location/activity profiles are obtained from
surveys contained in EPA's CHAD (e.g , NHAPS, CARB, U
Michigan) For inhalation, the day is divided into 1-mmute
to 12 hour sequential activities (depending on the pollutant
and diary selected). For ingestion, the day is divided into
30-mmute sequential macro-activities and eating/drinking
events For dermal and non-dietary ingestion, the day is
divided first into 30-mmute sequential macro-activities,
then each macro-activity is divided into 5-second contact
events Daily activity pattern time profiles for each route
are combined with concentrations, exposure factors, and
dose factors to yield daily exposure and dose profiles.
For macrolevel activity patterns, time/location/activity
surveys contained in EPA's CHAD (e g , NHAPS, CARB, U
Michigan) are used
For microlevel activity patterns, data on contact frequency
and duration for different body parts and surfaces are
obtained from available videography studies
Inhalation 1 minute to 12 hours, depending on the
pollutant and diary selected
Dietary ingestion. Daily consumption patterns, with
instantaneous ingestion assumed for each eating/drinking
event, and 30-mmute time steps for Gl absorption
Dermal and non-dietary ingestion 5-second time steps for
contact events occurring within each macro-activity
In the first-generation of SHEDS indoors at home and
outdoors at home, non-residential locations (SHEDS-PM),
and in vehicles (SHEDS-PM)
Microenvironments in time/location/activity surveys in
which sampled individuals ingest food or beverages
Model calculates exposure
of commuters
First generation of SHEDS calculates m-vehicle exposures,
but does not explicitly incorporate pollution gradients
during commuting Next generation will include air
concentrations in specified cohort's commuting area
Concentrations and
sources
Outdoor concentration
determination method
SHEDS requires user to enter distributions for
concentrations in exposure media Distributions for
outdoor concentrations can be derived from measurements
or from fate and transport models
Indoor concentration
determination method
SHEDS-Pesticides requires user to enter distributions for
concentrations in exposure media Distributions for indoor
concentrations can be derived from measurements or from
fate and transport models
SHEDS-PM uses a physical or empirical mass balance
model to estimate indoor concentrations
In-vehicle concentration
estimation
SHEDS requires user to enter distributions for
concentrations in exposure media In-vehicle air
concentrations are modeled using available
measurements
NOVEMBER 1999
B-60
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
Attribute
Component
Remarks
Passive smoking
Indoor PM passive smoking concentrations are modeled
using a mass balance model
Other indoor sources
Other indoor sources (e g , tracked-m soil, stripping of
chemicals via household water use, pesticide application
rates) are not explicitly included in SHEDS-Pesticides, but
are implicitly included through user-specified
concentrations for indoor air and surfaces
SHEDS-PM includes cooking and resuspension as other
indoor sources of particles
Extrapolation to
study population
Method of allocating
estimated exposures to
study population
Daily exposures and doses are simulated for individuals in
the specified cohort by combining actual macro-level and
micro-level activity data with residues or concentrations in
exposure media and exposure and dose factors 2-stage
Monte Carlo sampling is applied to simulate population
distributions
NOVEMBER 1999
B-61
TRIM. EXPO TSD (DRAFT)
-------
APPENDIX B
COMPARISON/CRITIQUE OF EXPOSURE MODELS
(This page intentionally left blank.]
NOVEMBER 1999 B-62 TRIM.Expo TSD (DRAFT)
-------
APPENDIX C
LIST OF TRIM.EXPO INPUT PARAMETERS
APPENDIX C
List of TRIM.Expo Input Parameters
Table C-l1
Example Input Parameters for Calculating Inhalation Exposures
Parameter
by Category
Units
Distribution
Type
Reference/Source
Building Parameters (for use in mass balance model)
Air exchange rates (ACH),
residences - windows closed
ACH, residences - windows
open
ACH, non-residential,
enclosed microenvironments,
filtered air
ACH, non-residential,
enclosed microenvironments,
unfiltered air
Efficiency of air cleaning
device (F)
Flow rate through air cleaning
device (q)
Fraction of outdoor pollutant
intercepted by enclosure (FB)
Indoor generation rate (S)
Indoor building volume (V)
Mixing factor (m)
Pollutant decay coefficient (Fd)
1/time
1/time
1/time
1/time
dimensionless
fraction
volume/time
dimensionless
fraction
mass/time
volume
dimensionless
1/time
Lognormal
Lognormal
Murray and Burmaster 1995
Johnson etal. 1998
Demographic Parameters
Age
Gender
Census data, Comprehensive Human
Activity Database (CHAD)
Census data, Comprehensive Human
Activity Database (CHAD)
' The input parameters and distribution types reflect the current initial choices that EPA is planning on
using in developing the initial TRIM.Expo inhalation Prototype and are subject to change.
NOVEMBER 1999
C-l
TRIM.Expo TSD (DRAFT)
-------
APPENDIX C
LIST OF TRIM.EXPO INPUT PARAMETERS
Parameter
by Category
Units
Distribution
Type
Reference/Source
Environmental Parameters
Temperature
degrees
National Climatic Data Center
Physiological Parameters (for calculating ventilation rate)
Body mass (BM)
Metabolic equivalence (MET)
Oxygen uptake rate (VO2)
Normalized oxygen uptake
rate (NVO2)
Resting metabolic rate (RMR)
kilograms (kg)
dimensionless
liters/min
ml/min/kg
kcal/mm
From
regression fit
specific to age
and gender
Brainard and Burmaster 1992, Burmaster
and Crouch 1997
Astrand 1960, Mercieret al. 1991, Katch
and Park 1975, Heil et al 1995, Mermier et
al. 1993, Rowland et al. 1987
Astrand 1960, Mercieret al. 1991, Katch
and Park 1975, Heil et al. 1995, Mermier et
al. 1993, Rowland etal. 1987
Schofield 1985
Pollutant Parameters
Ambient pollutant
concentrations
Microenvironmental
concentrations
mass/volume
mass/volume
Aerometric Information Retrieval System
(AIRS), dispersion model, TRIM.FaTE
Mass balance model, direct measurement,
intermedia transfer factors
NOVEMBER 1999
C-2
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX C
LIST OF TRJM.Expo INPUT PARAMETERS
Table C-22
Example Input Parameters for Calculating Ingestion Exposures
Parameter
by Category
Units
Distribution
Type
Reference/Source
Demographic Parameters
Age
Gender
ซ*
Body mass (BM)
kg
Census data, Comprehensive Human
Activity Database (CHAD)
Census data, Comprehensive Human
Activity Database (CHAD)
Brainard and Burmaster 1992, Burmaster
and Crouch 1997
Parameters Specific to Ground Water and Surface Water Intake
Concentration in tap water (C^)
Concentration in ground water
or surface water (Cg^jJ
Exposure duration (ED)
Exposure frequency (EF2tw)
Rate of intake of tap water (I2tw)
mg/L
mg/L
time
d/month (or
equivalent)
L/kg/d
EPA Exposure Factors Handbook 1997b,
Ershow and Cantor 1989, Canadian Ministry
of Health and Welfare 1981
Parameters Specific to Soil (Outdoors) Intake
Pollutant concentration in
surface soil (Css)
Exposure frequency (EFZ ss)
Annually averaged daily rate of
intake of soil (I2SS)
mg/kg
d/month (or
equivalent)
kg/kg/d
Parameters Specific to Dust (Indoors) Intake
Pollutant concentration in
house dust [Cnd(i,t)] (for
exposure district, i, during time
step t)
Pollutant concentration in
surface soil [C5S(i,t)]
Pollutant concentration of air
particles [Cap(i,t)]
Exposure frequency (EF2ihd)
Annually averaged daily rate of
intake of soil (lz S5)
mg/kg
mg/kg
mg/m3
d/month (or
equivalent)
kg/kg/d
2 The input parameters and distribution types reflect the current initial choices that EPA is planning on
using in developing the initial TRIM.Expo ingestion Prototype and are subject to change.
NOVEMBER 1999
C-3
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX C
LIST OF TRJM.Expo INPUT PARAMETERS
Parameter
by Category
Fraction of indoor dust that
originates from outdoor soil
('/*)
Units
dimensionless
Distribution
Type
Reference/Source
Parameters Specific to Home-grown Vegetables, Fruits, and Grains Intake
Pollutant concentration - air
IC.(i.t)]
Pollutant concentration in root
zone soil (Cra)
Pollutant concentration in:
1) grains [C8(i,t)],
2) exposed fruits and
vegetables [Cetv(i,t)],
3) protected fruits and
vegetables [Cpfv(i,t)]
Exposure frequency (number of
days per month equivalent, that
individual z consumes
homegrown foods in exposure
district in):
1) grains [EF,,(i.t)].
2) exposed fruits and
vegetables [EFzefv(i,t)J,
3) protected fruits and
vegetables [EFzpfv(i,t)]
Annually averaged daily rate of
intake of grains (lzg)
Annually averaged daily rate of
intake of exposed fruits and
vegetables (lzefg)
Annually averaged daily rate of
intake of protected fruits and
vegetables (lzpfg)
mg/m3
mg/kg
mg/kg
d/month (or
equivalent)
kg/kg/d
kg/kg/d
kg/kg/d
Parameters Specific to Home-grown Dairy Product Intake
Pollutant concentration in dairy
products of exposure district, i
at time step, t [Ck(i,t)]
Pollutant concentration in
pasture [Cp(i,t)]
Pollutant concentration in
surface soil [Cs(i,t)]
Pollutant concentration in the
water (Cw(i,t)]
Biotransfer factor from cattle
diet to dairy products
[Ck(i,t)/lndJ
Exposure frequency [EFZ k(i,t)]
mg/kg
mg/kg
mg/kg
mg/kg
d/kg
d/month (or
equivalent)
NOVEMBER 1999
C-4
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX C
LIST OF TRIM.EXPO INPUT PARAMETERS
Parameter
by Category
Annually averaged daily rate of
intake of dairy products (I2,J
Ingestion rate of pasture by
dairy cattle (lpdc)
Ingestion rate of soil by dairy
cattle (lsde)
Ingestion rate of water by dairy
cattle (lwdc)
Units
kg/kg/d
kg/d
kg/d
kg/d
Distribution
Type
Reference/Source
Parameters Specific to Home-grown Egg Intake
Pollutant concentration of
exposure district i, at time step,
t [C.(i,t)]
Pollutant concentration in
pasture [Cp(i,t)]
Pollutant concentration in
surface soil [Cs(i,t)]
Pollutant concentration in the
water [Cw(i.t>]
Biotransfer factor from hen diet
to eggs [C.(i,t)/lnhn]
Exposure frequency [EFze(i,t)]
Annually averaged daily rate of
intake of eggs (I2e)
Ingestion rate of pasture by
chickens (lphn)
Ingestion rate of soil by
chickens (lshn)
Ingestion rate of water by
chickens (lwac)
mg/kg
mg/kg
mg/kg
mg/kg
d/kg
d/month (or
equivalent)
kg/kg/d
kg/d
kg/d
kg/d
Parameters Specific to Home-grown Meat and Poultry Intake
Pollutant concentration of
exposure district, i at time step,
t [Ck(i.t)]
Pollutant concentration in
pasture [Cp(i.t)J
Pollutant concentration in
surface soil [Cs(i,t)]
Pollutant concentration in the
water [Cw(i,t)]
Biotransfer factor from cattle
diet to meat [C^.tXInJ
Exposure frequency [EFZ k(i,t)]
mg/kg
mg/kg
mg/kg
mg/kg
d/kg
d/month (or
equivalent)
NOVEMBER 1999
C-5
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX C
LIST OF TRIM.EXPO INPUT PARAMETERS
Parameter
by Category
Annually averaged daily rate of
intake of meat (lzt)
Ingestion rate of pasture by
beef cattle (IpJ
Ingestion rate of soil by beef
cattle (lSbc)
Ingestion rate of water by beef
cattle (IwJ
Units
kg/kg/d
kg/d
kg/d
kg/d
Distribution
Type
Reference/Source
Parameters Specific to Locally -grown Vegetables, Fruits, and Grains Intake3
Spatially-averaged pollutant
concentration - air [Ca(avg,t)]
Spatially-averaged pollutant
concentration in.
1) grains [Cg(avg,t)],
2) exposed fruits and
vegetables [Cefv(avg,t)],
3) protected fruits and
vegetables [Cptv(avg,t)]
Exposure frequency (number of
days per month equivalent, that
individual 2. consumes locally-
produced foods in exposure
district in)-
1) grains [EFzg(!,t)],
2) exposed fruits and
vegetables [EFze(v(i,t)],
3) protected fruits and
vegetables [EFzpfv(i,t)]
mg/m3
mg/kg
d/month (or
equivalent)
Parameters Specific to Locally-grown Dairy Product Intake4
Spatially-averaged pollutant
concentration - in the pastures
of all suburban and rural
exposure districts where local
dairy products are produced
during the time step, t
[Cp(avg,t)]
Exposure frequency (number of
days per month equivalent, that
individual z consumes locally-
produced dairy products in
exposure district in): [EFzk(i,t)]
mg/kg
d/month (or
equivalent)
3 The parameters used to calculate the intake of pollutants for locally-grown vegetables, fruits, and grains
are the same as those for home-grown with the following replacements.
4 The parameters used to calculate the intake of pollutants for locally-grown dairy products are the same as
those for home-grown with the following replacements.
NOVEMBER 1999
C-6
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX C
LIST OF TRJM.Expo INPUT PARAMETERS
Parameter
by Category
Units
Distribution
Type
Reference/Source
Parameters Specific to Locally-grown Egg Intake5
Spatially-averaged pollutant
concentration - in the pastures
of all suburban and rural
exposure districts where local
eggs are produced during the
time step, t [Cp(avg,t)].
Exposure frequency (number of
days per month equivalent, that
individual z consumes locally-
produced eggs in exposure
district i): [EFIie(i,t)J
mg/kg
d/month (or
equivalent)
Parameters Specific to Locally-grown Meat and Poultry Intake6
Spatially-averaged pollutant
concentration - in the pastures
of all suburban and rural
exposure districts where local
meat products are produced
during the time step, t
[Cp(avg,t)]
Exposure frequency (number of
days per month equivalent, that
individual z consumes locally-
produced meat products in
exposure district i): [EFzt(i,t)J
mg/kg
d/month (or
equivalent)
Parameters Specific to Local Fish Intake
Spatially-averaged pollutant
concentration - in the water of
all exposure districts in the air
shed being considered during
the time step, t [C,(avg,t)]
Biotransfer factor from water to
fish (BCF)
Exposure frequency (number of
days per month equivalent, that
individual z, in exposure district
i, consumes locally caught
fish): [EFIik(i,t)]
Annually-averaged daily rate of
intake of fish (lz ,)
mg/L
L/kg
d/month (or
equivalent)
kg/kg/d
5 The parameters used to calculate the intake of pollutants for locally-grown egg products are the same as
those for home-grown with the following replacements.
6 The parameters used to calculate the intake of pollutants for locally-grown meat and poultry products are
the same as those for home-grown with the following replacements.
NOVEMBER 1999
C-7
TRIM.EXPO TSD (DRAFT)
-------
APPENDIX C
LIST OF TRIM.EXPO INPUT PARAMETERS
Parameter
by Category
Units
Distribution
Type
Reference/Source
Parameters Specific to Recreational Sport Meat (Hunting)
Spatially-averaged pollutant
concentration - in the meat of
game animals residing in the
air shed being considered
during the time step, t
[Csm(avg,t)]
Time step averaged daily rate
of intake of sport meat (I2 sm)
mg/kg
kg/kg/d
NOVEMBER 1999
C-8
TRIM.EXPO TSD (DRAFT)
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
I REPORT NO.
EPA-453/D-99-001
3. RECIPIENT'S ACCESSION NO.
4 TITLE AND SUBTITLE
Total Risk Integrated Methodology. TRIM.Expo Technical
Support Document.
5 REPORT DATE
November, 1999
6 PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8 PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
10 PROGRAM ELEMENT NO
11. CONTRACT/GRANT NO
12 SPONSORING AGENCY NAME AND ADDRESS
13 TYPE OF REPORT AND PERIOD COVERED
External Review Draft
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report is part of a series of documentation for the Total Risk Integrated Methodology (TRIM).
TRIM is a time series modeling system, with multimedia capabilities, designed for assessing human health
and ecological risks from hazardous and criteria air pollutants. The detailed documentation of TRIM'S logic,
assumptions, equations, and input parameters is provided in comprehensive technical support documents for
each of the three TRIM modules, as they are developed. This report documents the Exposure Event module
ofTRIM(TRIM.Expo).
17
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
c. COSAT1 Field/Group
Risk assessment
Multimedia modeling
Exposure assessment
Air Pollutants
Air Pollution
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
218
20. SECURITY CLASS (Page)
Unclassified
22 PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
-------
------- |